Brain Advance Access originally published online on July 10, 2006
Brain 2006 129(9):2426-2435; doi:10.1093/brain/awl173
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Irradiation and hypoxia promote homing of haematopoietic progenitor cells towards gliomas by TGF-ß-dependent HIF-1
-mediated induction of CXCL12
1 Laboratory of Molecular Neuro-Oncology, Department of General Neurology and Hertie Institute for Clinical Brain Research University of Tübingen, Tübingen, Germany 2 Department of Internal Medicine II (Hematology) University of Tübingen, Tübingen, Germany
Correspondence to: Wolfgang Wick, MD, Laboratory of Molecular Neuro-Oncology, 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 IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Previously we defined a pathway of transforming growth factor beta (TGF-ß) and stromal cell-derived factor-1/CXC chemokine ligand 12 (SDF-1
/CXCL12) dependent migration of adult haematopoietic stem and progenitor cells (HPC) towards glioma cells in vitro and their homing to experimental gliomas in vivo. Hypoxia is a critical aspect of the microenvironment of gliomas and irradiation is an essential part of the standard therapy. To evaluate the therapeutic potential of HPC as vectors for a cell-based therapy of gliomas, we investigated the impact of hypoxia and irradiation on the attraction of HPC by glioma cells. Temozolomide (TMZ) treatment and hyperthermia served as controls. Supernatants of irradiated or hypoxic LNT-229 glioma cells promote HPC migration in vitro. Reporter assays reveal that the CXCL12 promoter activity is enhanced in LNT-229 cells at 24 h after irradiation at 8 Gy or after exposure to 1% oxygen for 12 h. The irradiation- and hypoxia-induced release of CXCL12 depends on hypoxia inducible factor-1 alpha (HIF-1), but not on p53. Induction of transcriptional activity of HIF-1 by hypoxia or irradiation requires an intact TGF-ß signalling cascade. This delineates a novel stress signalling cascade in glioma cells involving TGF-ß, HIF-1 and CXCL12. Stress stimuli can be irradiation, hypoxia or TMZ, but not hyperthermia. Cerebral irradiation of nude mice at 21 days after intracerebral implantation of LNT-229 glioma induces tumour satellite formation and enhances the glioma tropism of HPC to the tumour bulk and even to these satellites in vivo. These data suggest that the use of HPC as cellular vectors in the treatment of glioblastoma may well be combined with irradiation or other anti-angiogenic therapies that induce tumour hypoxia.
Key Words:
Haematopoietic progenitor cells; glioma; irradiation; non-lethal hypoxia; CXCL12; HIF-1; TGF-ß
Abbreviations: CXCL12, CXC chemokine ligand; DAPI, 4',6'-diamidino-2-phenylindole dihydrochloride; ELISA, enzyme-linked immunosorbent assay; HIF, hypoxia-inducible factor; HPC, haematopoietic progenitor and stem cells; MMP, matrix metalloproteinase; sKitL, soluble Kit ligand; SCF, stem cell factor; SDF, stromal cell-derived factor; TGF, transforming growth factor; VEGF, vascular endothelial growth factor
Received January 25, 2006. Revised April 22, 2006. Accepted June 2, 2006.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Stem cells emerge as an attractive option for a cell-based therapy of invasive neoplasms. Granulocyte-colony-stimulating factor (G-CSF) mobilized peripheral blood CD34+ haematopoietic progenitor and stem cells (HPC) localize to human gliomas after systemic intravenous delivery (Tabatabai et al., 2005
).
Almost all patients with malignant gliomas will be treated with radiotherapy at some stage of their disease (DeAngelis, 2001). Tumours respond to radiation by secreting cytokines that are determinants of tumour radiosensitivity, and intratumoural hypoxia plays a detrimental role in this process (Moeller et al., 2004). The tropism of HPC for gliomas is at least in part mediated by specific growth factors and chemokines. The main downstream target in a transforming growth factor-ß (TGF-ß) dependent signalling cascade is stromal cell-derived factor-1 (SDF-1)/CXC chemokine ligand 12 (CXCL12) (Tabatabai et al., 2005). This chemokine has also been identified as a central mediator of mesenchymal stem cell attraction by human gliomas (Nakamizo et al., 2005). It interacts with chemokine receptor CXCR4 on HPC. This tropism is facilitated by matrix metalloproteinase-9 (MMP-9) dependent cleavage of soluble Kit ligand (sKitL) released by glioma cells that interact with its receptor cluster of differentiation CD117 on HPC (Tabatabai et al., 2005). The expression of MMP-9 in glioma cells is induced by irradiation and hypoxia (Wild-Bode et al., 2001; Brat et al., 2004). Sublethal irradiation even enhances the migration and invasiveness of glioma cells (Wild-Bode et al., 2001). Apart from invasiveness, hypoxia is a characteristic aspect of malignant gliomas. It drives endothelial cell motility in vitro (Brat et al., 2004) and results in enhanced sensitivity to apoptotic stimuli as well as resistance towards epidermal growth factor receptor (EGFR) inhibitory strategies (Steinbach et al., 2004). Mast cell progenitors are attracted by ischaemia- and irradiation-induced release of vascular endothelial growth factor (VEGF) from resident mast cells at an ischaemic site; irradiation induces VEGF by an MMP-9- and sKitL-dependent process (Heissig et al., 2005).
Given the pivotal roles of hypoxia in glioma biology and of radiotherapy for the treatment of these tumours, we have been interested in the influence of both on the tropism of HPC to glioma and aimed at answering the following questions: (i) Is the tropism of HPC to glioma cells in vitro and in vivo influenced by hypoxia or irradiation? (ii) Is there a stimulus-provoked release of CXCL12 in glioma cells? (iii) What are the regulatory mechanisms underlying the stimulus-provoked release of CXCL12? We examined three clinically relevant stimuli, irradiation, hypoxia and alkylating chemotherapy with TMZ, as well as experimental hyperthermia on glioma cells and analysed their impact on glioma-directed HPC migration.
Using an orthotopic brain tumour model and focal cerebral irradiation, we show that irradiated brain tumours are more invasive than controls and that irradiation promotes HPC homing to the tumour bulk as well as to irradiation-induced tumour satellites in vivo. Irradiation, hypoxia and TMZ, but not hyperthermia, induce secretion of CXCL12 by glioma cells, resulting in an enhanced glioma tropism of HPC. In summary, we provide evidence that irradiation- and also hypoxia-induced attraction of HPC by glioma cells involves a TGF-ß-dependent, hypoxia-inducible factor-1 alpha (HIF-1) mediated release of CXCL12.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
HPC
Isolation of HPC was performed using anti-CD34 immunomagnetic microbeads as described previously (Bautz et al., 2001
).
Glioma 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 (Friese et al., 2004) and the generation of LNT-229 sublines depleted from p53 by RNA interference have been described previously (Wischhusen et al., 2003). LNT-229 puro TGF-ß and LNT-229 puro p53 have been generated as controls with the respective experimental sublines.
Cell culture
Lysates and supernatants were prepared as described previously. Incubation of cells in hypoxia was performed at 37°C and 1% O2 (Wick et al., 2002) for 3, 6, 9, 12 and 24 h. Cells were irradiated at 8 Gy (
-Cell, Nordion, Kanata, CA, USA) and analysed 12, 24, 36 and 48 h post-irradiation, or treated with 250 µM TMZ for 24 h (Schering-Plough, Kenilworth, NJ, USA) (Wick et al., 2002). Incubation under hyperthermic conditions was performed according to Hermisson which et al. (2000). In brief, cells were seeded and cultured for 48 h at 42°C, 21% O2, 5% CO2.
Chemotaxis assays and quantification of migrated cells
All assays and quantification of migrated cells were performed as described previously (Tabatabai et al., 2005). Supernatants applied in the migration assays are normalized to the number of glioma cells and protein content. Quantities containing 100 µg protein have been used in the lower chamber.
Reporter assays
Dual luciferase/renilla assays (Dyer et al., 2000) were performed with co-transfection of 200 ng of the respective reporter constructs and 20 ng of pRL-CMV. The CXCL12 reporter constructs were generated by Ceradini (2004): pGL3b.SDF1.full with the full-length promoter of CXCL12, pGL3b.SDF1.–544 with deletion of both HIF-binding sites in the promoter and pGL3b.SDF1.Mut.HBS1 with mutation of one HIF-binding site. For the HIF luciferase activity assay pT81/3xHRE was used. Luciferase activity was normalized to constitutive renilla activity (pRL-CMV). Recombinant TGF-ß2 was purchased from R&D Systems (Wiesbaden-Nordenstadt, Germany). SD-208 was used according to Uhl et al. (2004).
Cloning of pSUPER siHIF-1
The following oligonucleotide sequences for targeting HIF-1 were cloned into pSUPER: 5'-CTCAAGCAACTGTCATATA-3' (sequence no. 1) and TGCCACCACTGATGAATTA (sequence no. 2). The efficacy of gene suppression by both sequences was assessed by immunoblot after co-transfection of the respective pSUPER siHIF-1 plasmid with pCMX-SAH/Y145F (GFP)-HIF-1 in HEK293 cells and by reporter assays after co-transfection of pSUPER siHIF-1 with pT81/3xHRE. pSUPER siHIF-1 no. 1 was used for the delineated experiments.
Immunoblot
Immunoblots were performed as described previously (Weiler et al., 2005). Antibodies were HIF-1 (R&D Systems, Wiesbaden-Nordenstadt, Germany) and ß-actin (Santa Cruz, CA, USA). Quantification was done by multiplying the respective signal area of each protein band with its mean intensity using Corel Photo Paint 11 software (Corel Cooperation, Ottawa, Canada) and normalizing to the values of untreated and control samples.
Enzyme-linked immunosorbent assay (ELISA)
The concentrations of the following chemokines were detected in cell culture supernatants by Quantikine Immunoassays from R&D Systems: CXCL12 (Cat. No. DSA00), stem cell factor (SCF) (Cat. no. DCK00).
MMP-9 zymography
MMP-9 activity was assessed by zymography as described by Friese et al. (2004).
Animal studies
We used the orthotopic xenograft nude mice paradigm with LNT-229 glioma cells. Mice were divided into four experimental groups: (i) sham surgery at Day 0, which consisted of an incision with a Hamilton syringe (Martinsried, Germany) without implantation of glioma cells; (ii) sham surgery at Day 0 plus cerebral irradiation (8 Gy) at Day 21; (iii) implantation of 7.5 x 104 LNT-229 cells into the right striatum at Day 0; and (iv) implantation of 7.5 x 104 LNT-229 cells at Day 0, followed by cerebral irradiation at 8 Gy at Day 21. Two days after irradiation or 23 days after surgery, all mice received intravenous (i.v.) injection of 106 peripheral G-CSF-mobilized PKH26-stained HPC. Forty-eight hours after i.v. injection of HPC, mice were killed to obtain histologies. Sections were counterstained with 4',6'-diamidino-2-phenylindole dihydrochloride (DAPI) (Vectashield Mounting Medium with DAPI, H1200, Vector Laboratories, Burlingame, CA, USA). All animal studies were performed according to the German animal protection law under the permission of the local authorities in Tübingen (N3/03). Quantification of PKH26-positive HPC in the histologies was done as follows: for each animal the rostrocaudal section harbouring the maximal tumour area, designated central tumour section, and three additional sections at multiples of 160 µm cranial and caudal of the central tumour sections were determined and PKH26-positive HPC counted on these sections. The number of PKH26-positive HPC was determined in cells/106 µm2 tumour area. Mean values were calculated from three animals in the control and the irradiation group.
Cerebral irradiation
For local irradiation, brains of nude mice were irradiated at 8 Gy using electrons from a standard Linac radiation source. Positioning and shielding of the animals were achieved by a lead/plastic device that allows the exact application of the radiation with a 90% isodose to the targeted 7 x 7 mm brain section, sparing the throat of the mice.
Immunohistochemistry
For MMP-9 staining, sections were fixed in acetone and blocked in 3% H2O2, Dako A+B (Biotin blocking X0590), 3% skim milk. Polyclonal rabbit anti-human MMP-9 (Zytomed, San Francisco, CA, USA) was added for 30 min at room temperature at 1 : 75 dilution, and then stained with secondary anti-rabbit antibody (Vectastain, PK 4001, Vector Laboratories) and developed with DAB Vectastain (SK-4100, Vector Laboratories). For HIF-1 staining, a polyclonal rabbit anti-human antibody was used (R&D Systems).
Statistical analysis
Quantitative data were obtained for chemotaxis, ELISA and luciferase activity as indicated. Data are expressed as mean and SEM. Statistical significance was assessed by one-way ANOVA (analysis of variance) followed by Tukey's post hoc test (Excel, Microsoft, Seattle, WA, USA). All experiments reported here were performed at least three times in triplicate with similar results.
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Hypoxia, irradiation and TMZ promote release of CXCL12 by LNT-229 glioma cells in vitro
Previously, we delineated a cascade for the glioma tropism of HPC that revealed CXCL12, released in a TGF-ß-dependent manner, as a central chemokine. We evaluate here whether the secretion of CXCL12 by LNT-229 cells is influenced by cellular stress, namely hypoxia, therapeutic irradiation, treatment with TMZ or hyperthermia.
CXCL12 levels in the supernatant of LNT-229 puro (control) cells are significantly enhanced by hypoxia at 1% O2, single fraction of irradiation at 8 Gy or treatment with 250 µM TMZ, but not by hyperthermia at 42°C (Table 1). The time course for hypoxia shows peak levels of CXCL12 at 24 h after the onset of hypoxia, whereas CXCL12 peaks with a gradual increase over 48 h after challenge with irradiation and TMZ (data not shown). Given the central role of p53 in stress-induced cellular responses, we next examined the role of p53 in baseline and stress-induced CXCL12 release. Baseline secretion in LNT-229 sip53 cells is unaltered compared with controls. p53-deficient LN-308 cells (data not shown) as well as LNT-229 sip53 cells, similar to controls, exhibited a stimulus-evoked release of CXCL12 by hypoxia, irradiation or treatment with TMZ, but not by hyperthermia. Since we have previously identified TGF-ß as a central player in the glioma-mediated attraction of HPC, we analysed the baseline and stress-induced CXCL12 release in LNT-229 siTGF-ß1/2 cells, too. Interestingly, LNT-229 siTGF-ß1/2 cells release very low quantities of CXCL12 into the supernatant per se, and neither hypoxia, irradiation nor TMZ treatment induce CXCL12 (Table 1). However, the treatment of LNT-229 siTGF-ß1/2 cells with TGF-ß2 for 7 days reconstitutes both baseline CXCL12 expression (Tabatabai et al., 2005) and stimulus-induced secretion of CXCL12 after hypoxia and irradiation (Table 1). CXCL12-mediated glioma tropism of HPC is enhanced by MMP-9-mediated cleavage of membrane-bound KitL (mKitL) to sKitL, which interacts with CD117 on HPC (Tabatabai et al., 2005). Both hypoxia and irradiation increase MMP-9 activity (Fig. 1) and enhance sKitL release by glioma cells (Table 1). LNT-229 siTGF-ß1/2 cells show diminished MMP-9 activity (Friese et al., 2004). These cells release less SCF under basal conditions, and there is no elevation of SCF concentration in the supernatants of hypoxic or irradiated LNT-229 siTGFß1/2cells (Table 1).
|
|
Supernatants of hypoxic and irradiated LNT-229 cells enhance migration of HPC in vitro
We examined next whether stimulus-induced CXCL12 release into the supernatant of LNT-229 cells leads to an increased migration of HPC. The migration of HPC was enhanced by the supernatants of hypoxic, irradiated as well as TMZ-treated but not by the supernatant of heated glioma cells (Fig. 2A). The increased attraction of HPC by the supernatants of hypoxic or irradiated LNT-229 sublines and LN-308 glioma cells is diminished after addition of neutralizing CXCL12 antibodies as in wild-type LNT-229 cells (Fig. 2B and C). Supernatants of LNT-229 sip53 with reduction of p53 expression by RNA interference (Wischhusen et al., 2003
) as well as supernatants of p53-deficient LN-308 glioma cells lead to enhanced HPC attraction after exposure of the respective cells to hypoxia or irradiation (Fig. 2B). Thus, enhanced migration of HPC towards hypoxic and irradiated glioma supernatant does not depend on the p53 status of glioma cells. In contrast, LNT-229 siTGF-ß1/2 cell-derived supernatants show decreased baseline attraction of HPC, and preconditioning by hypoxia or irradiation does not enhance migration of HPC (Fig. 2C).
|
From these experiments, we came to the following conclusions: First, CXCL12 release by glioma cells was enhanced by exposure to hypoxia, irradiation or TMZ, but not by hyperthermia. Secondly, this process depends on TGF-ß but not on p53. Thirdly, elevated concentrations of CXCL12 in the supernatants of pre-treated glioma cells lead to enhanced attraction of HPC.
Increased activity of cxcl12 promoter after hypoxia and irradiation depends on HIF-1
Reporter assays were used to analyse the mechanisms underlying the hypoxia- and irradiation-induced increase in CXCL12 secretion by glioma cells. The pGL3b.SDF1.full construct (Ceradini et al., 2004) contains the full cxcl12 promoter region. Luciferase assays showed elevated relative luciferase activity after treatment of transfected LNT-229 cells with hypoxia or TMZ, but not after culturing the cells at 42°C for 48 h. These data indicate that the activity of cxcl12 promoter was enhanced by hypoxia or chemotherapy, but not by hyperthermia. TGF-ß gene silencing in LNT-229 siTGFß1/2 cells resulted in reduced baseline cxcl12 promoter activity. In addition, the induction of relative luciferase activity after cellular stress stimuli did not occur in LNT-229 siTGFß1/2 cells. In contrast, the silencing of p53 did not lead to changes in the cxcl12 promoter activity: As in wild-type LNT-229 cells, cellular stress stimuli resulted in significantly enhanced cxcl12 promoter activity in LNT-229 sip53 cells (Fig. 3A). The promoter of cxcl12 contains two binding sites for HIF-1. HIF-1 induces CXCL12 in hypoxic endothelial cells (Ceradini et al., 2004). Reporter assays with the construct pT81/3xHRE-Luc showed that the transcriptional activity of HIF-1 in glioma cells is enhanced by hypoxia or irradiation (Fig. 3B), as well as by treatment of LNT-229 glioma cells with TMZ (data not shown). Hyperthermia, however, did not lead to the induction of HIF-1 transcriptional activity (data not shown). In the next steps, we analysed the role of HIF-1 in the induction of CXCL12 by cellular stress stimuli. We used a reporter construct, in which both HIF-1 binding sites of the cxcl12 promoter had been deleted (pGL3b.SDF1.–544; Ceradini et al., 2004). Deletion of both HIF-1 binding sites in the cxcl12 promoter resulted in the absence of changes in the cxcl12 promoter activity after exposure to hypoxia or irradiation (Fig. 3B), while the baseline promoter activity remained unchanged. We next used the construct pGL3b.SDF1.MUT.HBS, which contained a mutation in one HIF-1 binding site of the cxcl12 promoter. Reporter assays with this construct showed that treatment of LNT-229 cells with hypoxia or irradiation leads to marginal induction of cxcl12 promoter activity. Taken together, these experiments indicate that both HIF-1 binding sites in the promoter are necessary to fully generate hypoxia or irradiation-induced cxcl12 promoter activity. However, the absence of intact HIF-1 binding sites did not modulate the baseline cxcl12 promoter activity (Fig. 3B), suggesting that HIF-1 does not contribute to the baseline cxcl12 promoter activity in LNT-229 cells. Treatment of LNT-229 cells with TMZ led to enhancement of cxcl12 promoter activity, whereas hyperthermia did not (Fig. 3A).
|
To further analyse the role of HIF-1
in the induction of CXCL12, we targeted hif-1 in glioma cells by RNA interference. Therefore, we generated the plasmid pSUPER siHIF-1. Successful targeting of hif-1 by pSUPER siHIF-1 was confirmed by reporter assays and immunoblot (Fig. 3C). Co-transfection of pGL3b.SDF.full and pSUPER siHIF-1 led to loss of relative luciferative activity in LNT-229 cells. These data indicate that targeting of hif-1 by RNA interference inhibited the induction of the promoter activity of cxcl12 by hypoxia or irradiation in glioma cells (Fig. 3C). Similarly, TMZ-induced cxcl12 promoter activity was inhibited by targeting of hif-1 (data not shown).
On the basis of these experiments, we conclude that elevated cxcl12 promoter activity in glioma cells after exposure to hypoxia or irradiation critically depends on HIF-1. These experiments place HIF-1 upstream of CXCL12 in our paradigm.
Transcriptional activity of HIF-1 in LNT-229 cells requires intact signalling of TGF-ß
HIF-1 regulation by hypoxia is evident in our paradigm. Since there is increasing evidence that HIF-1 interrelates with TGF-ß, for example, by hypoxia-enhanced expression of the proprotein convertase furin (McMahon et al., 2005), we next analysed the connection between HIF-1 and TGF-ß in glioma cells. Immunoblots for HIF-1 showed a lack of HIF-1 induction in LNT-229 siTGFß1/2 cells in response to hypoxia. Not only hypoxia- but also irradiation-inducible HIF-1 protein expression was observed in LNT-229 puro (control) cells. In LNT-229 siTGF-ß1/2 cells, however, there was a lack of HIF-1 induction in response to irradiation. Importantly, there was no induction of TGF-ß protein level by hypoxia (data not shown) or irradiation in LNT-229 cells (Wild-Bode et al., 2001). With these data we postulate an upstream role of TGF-ß for HIF-1 protein expression and transcriptional activity (Fig. 4). This was confirmed by similar experiments using SD-208, an inhibitor of the TGF-ß receptor I kinase (Uhl et al., 2004). SD-208 inhibited baseline and stimulus-induced transcriptional activity of hif-1 compared with controls, similar to RNA interference for TGF-ß. Reconstitution of LNT-229 siTGF-ß cells with recombinant TGF-ß2 led to the phosphorylation of Smad proteins (data not shown) and restored the transcriptional activity of hif-1 in LNT-229 siTGF-ß cells in response to hypoxia or irradiation (Fig. 4B), but not in cells co-treated with SD-208 (data not shown).
|
P < 0.05 for the effect of recombinant TGF-ß).Irradiation enhances migration of HPC towards LNT-229 glioma in vivo
Pre-irradiated 9L glioma cells implanted into rat brains display invasive growth compared with non-irradiated cells (Wild-Bode et al., 2001
). But there is so far no evidence that irradiation induces invasiveness of preformed gliomas nor that HPC are targeting these invasive glioma satellites.
Interestingly, focal irradiation induced the formation of glioma cell satellites distant from the bulk tumour (Fig. 5A). We used the paradigm of specific glioma tropism of HPC after i.v. injection in an LNT-229 orthotopic xenograft model (Tabatabai et al., 2005). There were no HPC in the perilesional area of a sham surgery. Further, irradiation of a lesion induced by sham surgery did not attract HPC to the site of the lesion or to the contralateral brain (Fig. 5B). Of note, the tropism of HPC to a pre-implanted intracerebral LNT-229 glioma was further enhanced 2-fold after focal irradiation. HPC were found within the tumour but neither in the adjacent brain tissue nor elsewhere distant from the visible tumour mass (Fig. 6A). Immunohistochemistry showed upregulation of HIF-1 and MMP-9 in glioma cells after irradiation in vivo (Fig. 6B). Importantly, HPC were found not only in the tumour bulk but also in the invasive satellites (Fig. 6C).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The standard of care of newly diagnosed malignant glioma involves radiotherapy of the brain tumour region and chemotherapy with TMZ (DeAngelis, 2001
; Stupp et al., 2005). The irradiation of cultured tumour cells leads to oxygen level-dependent cytokine release, which may determine the tumour radiosensitivity (Moeller et al., 2004). The tropism of HPC for gliomas is at least in part mediated by specific growth factors and chemokines. The main downstream mediator of HPC attraction in a TGF-ß-dependent signalling cascade in our glioma model is CXCL12 (Tabatabai et al., 2005). The central hypothesis investigated here proposes that hypoxia and irradiation of glioma cells lead to the release of cytokines that modulate the attraction of HPC by human gliomas. The stimulus exerted by a brain lesion, for example, by a tumour or infarcted brain, to susceptible cells resembles an inflammatory response (Imitola et al., 2004) that could be exogenously enhanced by other stimuli.
We demonstrate here a pathway indicating increased secretion of CXCL12 by glioma cells after hypoxia, irradiation and treatment with TMZ, but not by hyperthermia (Table 1), leading to CXCL12-dependently enhanced attraction of HPC by glioma cells (Fig. 2A). Hypoxia and irradiation also lead to increased activity of MMP-9 and enhanced secretion of sKitL by glioma cells. In a limb ischaemia model, ionizing radiation induces MMP-9 and sKitL, leading to tissue revascularization through VEGF release from resident mast cells and MMP-9-mediated progenitor cell mobilization (Heissig et al., 2005). The importance of this response has been illustrated in a hypoxia paradigm where secretion of CXCL12 by endothelial cells was induced after hypoxia (Ceradini et al., 2004).
TGF-ß plays a central role in the attraction of HPC by gliomas. Furin is expressed in many tumours, proteolytically activates TGF-ß (Bassi et al., 2005) and thus mediates motility of glioma cells (Wick et al., 2004). In addition, the proprotein convertase furin is regulated by hypoxia. In endothelial cells, there is an increase in TGF-ß2 mRNA and protein levels in response to hypoxic treatment at 1% O2 (Akman et al., 2001). To our surprise, TGF-ß protein expression was not regulated by hypoxia (data not shown) nor by irradiation in glioma cells (Wild-Bode et al., 2001). However, the importance of TGF-ß is evident by the absence of hypoxia-, irradiation- or TMZ-induced CXCL12 secretion in LNT-229 siTGF-ß1/2 cells. Another pivotal mediator of cellular stress signals, p53, in contrast, was not necessary for hypoxia- and irradiation-induced CXCL12 secretion or attraction of HPC by glioma cells (Table 1 and Fig. 2B and C). This is in line with our previous observation that the induction of migration and invasiveness of glioma cells in vitro by sublethal irradiation was also p53-independent (Wild-Bode et al., 2001). In contrast, the radioresistance of gliomas at least in part might be due to an impairment of p53 responses (Avenia et al., 2006). Recently, a successful attempt shows reduced glioma cell survival at doses <2 Gy in combination with adenoviral chimeric tumour suppressor 1 (CTS1) transfer. CTS1 is an artificial p53-based gene designed to resist various pathways of p53 inactivation (Naumann et al., 2001).
Oxygen availability is a determinant in the setting of chemotactic responsiveness to CXCL12 by upregulation of CXCR4 (Schioppa et al., 2003). In endothelial cells, CXCL12 gene expression is regulated by HIF-1 in response to hypoxia. The trafficking of circulating stem and progenitor cells to areas of tissue damage by infarction was sensitive to blocking of CXCL12 (Ceradini et al., 2004). We find that CXCL12 promoter induction by hypoxia depends on HIF-1 and that this induction is absent in TGF-ß-depleted cells (Fig. 3). These data place HIF-1 between TGF-ß and CXCL12 (Figs 3 and 7). So far, a transcriptional cooperation between hypoxia and TGF-ß has been suggested because in endothelial cells the stimulation of the endoglin expression by hypoxia is profoundly enhanced by TGF-ß. A multiprotein complex involving Sp1, Smad3 and HIF-1 is proposed to be responsible for this transcriptional effect (Sanchez-Elsner et al., 2002). From the results in Figs 3 and 4, we conclude that HIF-1 promotor transcriptional activity and protein expression in LNT-229 glioma cells require TGF-ß signalling. Consequently, the lack of induction of CXCL12 mediated by hypoxia and irradiation in LNT-229 siTGF-ß1/2 cells is probably due to reduced hif-1 expression in these cells. The profound and similar effect of irradiation and hypoxia on HIF-1a-dependent CXCL12 expression is surprising. Enhanced CXCL12 release from immature osteoblasts and endothelial cells has been reported following radiation, and it dramatically increased the potential of HPC to engraft the bone marrow of transplanted NOD/SCID mice (Ponomaryov et al., 2000). Further evidence for the importance of HIF-1 for irradiation-induced chemokine expression comes from HIF-1 depletion and studies with the inhibitor 1-benzyl-3-(5'-hydroxymethyl-2'-furyl)indazole (YC-1). HIF-1-depleted cells failed to induce VEGF after radiation, and YC-1 resulted in a substantial sensitization to radiotherapy in a tumour implanted into a window chamber by destruction of the vasculature (Moeller et al., 2004).
|
We have previously demonstrated that intracerebral glioma cells attract i.v. injected human HPC (Tabatabai et al., 2005
). Here, we ascertain that the syringe lesion per se (sham surgery) without implantation of glioma cells did not attract HPC to the brain. Further, cerebral irradiation of the lesional area also failed to attract HPC in the absence of glioma cells (Fig. 5B).
The stimulus-induced enhanced attraction of HPC by glioma cells offers an attractive perspective for the clinical application of these cells as cellular vectors. Heissig et al. (2005) demonstrate that low-dose irradiation induces VEGF and sKitL, promoting mast cell migration from the bone marrow to the ischaemic site in a limb ischaemia model. The present study is the first to demonstrate that irradiation induces invasiveness of pre-implanted gliomas in an orthotopic model (Fig. 5A). Immunohistochemistry shows an upregulation of MMP-9 and HIF-1 in the irradiated tumours (Fig. 6B). The data in vivo and in vitro suggest a TGF-ß and HIF-1-dependent mechanism of irradiation-induced alterations in the secretion of CXCL12 by glioma cells resulting in the enhanced migration of HPC towards malignant glioma cells in vitro and in vivo (Fig. 7). Most interestingly, the in vivo data indicate the propensity of HPC to track not only large tumour bulks (Fig. 6A) but also smaller number of tumour cells surrounded in the normal brain (Fig. 6C).
Other studies have focused on the potential negative aspects of circulating stem and progenitor cells when they are recruited by the tumour microenvironment and contribute to the tumour vasculature (Semenza, 2003; Vajkoczy et al., 2003). HIF-1 expression has been a negative prognostic marker in several human cancers (Semenza, 2003). With our present work, we suggest a clinical setting that might allow the application of HPC as cellular vehicles to deliver therapeutic molecules. As the molecular mechanisms leading to glioma tropism of HPC are dependent on TGF-ß, a potential HPC-based therapy against glioma might not work in combination with TGF-ß-antagonistic strategies. Instead, HPC-based therapy may be compatible with and even positively influenced by irradiation and novel anti-angiogenic therapies leading to hypoxia.
|
Acknowledgements |
|---|
We thank Ulrike Obermüller and Gabriele von Kürthy for excellent technical assistance. This work was supported by the Landesstiftung Baden-Württemberg, State of Baden-Württemberg, Germany (P-LS-AS/HSPA7-12).
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Akman HO, Zhang H, Siddiqui MA, Solomon W, Smith EL, Batuman OA. (2001) Response to hypoxia involves transforming growth factor beta2 and Smad proteins in human endothelial cells. Blood 98:3324–31.
Avenia PD, Porrello A, Berardo M, Angelo MD, Soddu S, Arcangeli G, et al. (2006) Tp53-gene transfer induces hypersensitivity to low doses of X-rays in glioblastoma cells: a strategy to convert a radio-resistant phenotype into a radiosensitive one. Cancer Lett 231:102–12.[CrossRef][ISI][Medline]
Bassi DE, Fu J, Lopez de Cicco R, Klein-Szanto AJ. (2005) Proprotein convertases: ‘master switches’ in the regulation of tumor growth and progression. Mol Carcinog 44:151–61.[CrossRef][ISI][Medline]
Bautz F, Denzlinger C, Kanz L, Mohle R. (2001) Chemotaxis and transendothelial migration of CD34(+) hematopoietic progenitor cells induced by the inflammatory mediator leukotriene D4 are mediated by the 7-transmembrane receptor CysLT1. Blood 97:3433–40.
Brat DJ, Castellano-Sanches AA, Hunter SB, Pecot M, Cohen C, Hammond EH, et al. (2004) Pseudopalisades in glioblastoma are hypoxic, express extracellular matrix proteases, and are formed by an actively migrating cell population. Cancer Res 64:920–7.
Ceradini DJ, Kulkarni AR, Callaghan MJ, Tepper OM, Bastidas N, Kleinman ME, et al. (2004) Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med 10:858–64.[CrossRef][ISI][Medline]
DeAngelis LM. (2001) Brain tumors. N Engl J Med 344:114–23.
Dyer BW, Ferrer FA, Klinedinst DK, Rodriguez R. (2000) A noncommercial dual luciferase enzyme assay system for reporter gene analysis. Anal Biochem 282:158–61.[CrossRef][ISI][Medline]
Friese MA, Wischhusen J, Wick W, Weiler M, Eisele G, Steinle A, et al. (2004) RNA interference targeting transforming growth factor-beta enhances NKG2D-mediated antiglioma immune response, inhibits glioma cell migration and invasiveness, and abrogates tumorigenicity in vivo. Cancer Res 64:7596–603.
Hermisson M, Wagenknecht B, Wolburg H, Glaser T, Dichgans J, Weller M, et al. (2000) Sensitization to CD95 ligand-induced apoptosis in human glioma cells by hyperthermia involves enhanced cytochrome c release. Oncogene 19:2338–45.[CrossRef][ISI][Medline]
Heissig B, Rafii S, Akiyama H, Ohki Y, Sato Y, Rafael T, et al. (2005) Low-dose irradiation promotes tissue revascularization through VEGF release from mast cells and MMP-9-mediated progenitor cell mobilization. J Exp Med 202:739–50.
Imitola J, Raddassi K, Park KI, Mueller FJ, Nieto M, Teng YD, et al. (2004) Directed migration of neural stem cells to sites of CNS injury by the stromal cell-derived factor 1alpha/CXC chemokine receptor 4 pathway. Proc Natl Acad Sci USA 101:18117–22.
McMahon S, Grondin F, McDonald PP, Richard DE, Dubois CM. (2005) Hypoxia-enhanced expression of the proprotein convertase furin is mediated by hypoxia-inducible factor-1. J Biol Chem 280:6561–9.
Moeller BJ, Cao Y, Li CY, Dewhirst MW. (2004) Radiation activates HIF-1 to regulate vascular radiosensitivity in tumors: role of reoxygenation, free radicals, and stress granules. Cancer Cell 5:429–41.[CrossRef][ISI][Medline]
Nakamizo A, Marini F, Amano T, Khan A, Studeny M, Gumin J, et al. (2005) Human bone marrow-derived mesenchymal stem cells in the treatment of gliomas. Cancer Res 65:3307–18.
Naumann U, Kügler S, Wolburg H, Wick W, Rascher G, Schulz JB, et al. (2001) Chimeric tumor suppressor 1 (CTS1), a p53-derived chimeric tumor suppressor gene, kills p53 mutant and p53 wild-type glioma cells in synergy with irradiation and CD95 ligand. Cancer Res 61:5833–42.
Ponomaryov T, Peled A, Petit I, Taichman RS, Habler L, Sandbank J, et al. (2000) Induction of chemokine stromal-derived factor-1 following DNA damage improves human stem cell function. J Clin Invest 106:1331–9.[ISI][Medline]
Sanchez-Elsner T, Botella LM, Velasco B, Langa C, Bernabeu C. (2002) Endoglin expression is regulated by transcriptional cooperation between the hypoxia and transforming growth factor-beta pathways. J Biol Chem 46:43799–808.
Schioppa T, Uranchimeg B, Saccani A, Biswas SK, Doni A, Rapisarda A, et al. (2003) Regulation of the chemokine receptor CXCR4 by hypoxia. J Exp Med 198:1391–402.
Semenza GL. (2003) Targeting HIF-1 for cancer therapy. Nat Rev Cancer 3:721–32.[CrossRef][ISI][Medline]
Steinbach JP, Klumpp A, Wolburg H, Weller M. (2004) Inhibition of epidermal growth factor receptor signaling protects human malignant glioma cells from hypoxia-induced cell death. Cancer Res 64:1575–8.
Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al. (2005) Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352:987–96.
Tabatabai G, Bähr O, Möhle R, Eyupoglu AM, Boehmler J, Wischhusen J, et al. (2005) Lessons from the bone marrow: how malignant glioma cells attract adult haematopoietic progenitor cells. Brain 128:2200–11.
Uhl M, Aulwurm S, Wischhusen J, Weiler M, Almirez R, Mangadu R, et al. (2004) SD-208, a novel transforming growth factor beta receptor I kinase inhibitor, inhibits growth and invasiveness and enhances immunogenicity of murine and human glioma cells in vitro and in vivo. Cancer Res 64:7954–61.
Vajkoczy P, Blum S, Lamparter M, Mailhammer R, Erber R, Engelhardt B, et al. (2003) Multistep nature of microvascular recruitment of ex vivo-expanded embryonic endothelial progenitor cells during tumor angiogenesis. J Exp Med 197:1755–65.
Weiler M, Bähr O, Hohlweg U, Naumann U, Rieger J, Huang H, et al. (2005) BCL-xL time-dependent dissociation between modulation of apoptosis and invasiveness in human malignant glioma cells. Cell Death Differ Oct 28: Epub ahead of print.
Wick A, Wick W, Waltenberger J, Weller M, Dichgans J, Schulz JB. (2002a) Neuroprotection by hypoxic preconditioning requires sequential activation of vascular endothelial growth factor receptor and Akt. J Neurosci 22:6401–7.
Wick W, Wick A, Schulz JB, Dichgans J, Rodemann HP, Weller M. (2002b) Prevention of irradiation-induced glioma cell invasion by temozolomide involves caspase 3 activity and cleavage of focal adhesion kinase. Cancer Res 62:1915–9.
Wick W, Wild-Bode C, Frank B, Weller M. (2004) BCL-2-induced glioma cell invasiveness depends on furin-like proteases. J Neurochem 91:1275–83.[CrossRef][ISI]
Wild-Bode C, Weller M, Rimner A, Dichgans J, Wick W. (2001) Sublethal irradiation promotes migration and invasiveness of glioma cells: implications for radiotherapy of human glioblastoma. Cancer Res 61:2744–50.
Wischhusen J, Naumann U, Ohgaki H, Rastinejad F, Weller M. (2003) CP-31398, a novel p53-stabilizing agent, induces p53-dependent and p53-independent glioma cell death. Oncogene 22:8233–45.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Moskovits, A. Kalinkovich, J. Bar, T. Lapidot, and M. Oren p53 Attenuates Cancer Cell Migration and Invasion through Repression of SDF-1/CXCL12 Expression in Stromal Fibroblasts. Cancer Res., November 15, 2006; 66(22): 10671 - 10676. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||