By Mercier, Marie Le Mathieu, Veronique; Haibe-Kains, Benjamin; Bontempi, Gianluca; Mijatovic, Tatjana; Decaestecker, Christine; Kiss, Robert; Lefranc, Florence
Abstract Galectin (Gal) 1 is a hypoxia-regulated proangiogenic factor that also directly participates in glioblastoma cell migration. To determine how Gal-1 exerts its proangiogenic effects, we investigated Gal-1 signaling in the human Hs683 glioblastoma cell line. Galectin 1 signals through the endoplasmic reticulum transmembrane kinase/ribonuclease inositol-requiring 1alpha, which regulates the expression of oxygen-regulated protein 150. Oxygen- regulated protein 150 controls vascular endothelial growth factor maturation. Galectin 1 also modulates the expression of 7 other hypoxia-related genes (i.e. CTGF, ATF3, PPP1R15A, HSPA5, TRA1, and CYR61) that are implicated in angiogenesis. Decreasing Gal-1 expression in Hs683 orthotopic xenografts in mouse brains by siRNA administration impaired endoplasmic reticulum stress and enhanced the therapeutic benefits of the proautophagic drug temozolomide. These results suggest that decreasing Gal-1 expression (e.g. through brain delivery of nonviral infusions of anti-Gal-1 siRNA in patients) can represent an additional therapeutic strategy for glioblastoma.
Key Words: Angiogenesis, Endoplasmic reticulum stress, Galectin 1, Glioblastoma, Hypoxia.
INTRODUCTION
Malignant gliomas, particularly glioblastoma multiforme (GBM), diffusely invade brain tissue as single migrating cells with reduced levels of apoptosis and consequent resistance to the cytotoxic effects of proapoptotic drugs (1). Current treatments of GBM patients include maximum surgical resection, followed by radiation and chemotherapy with the proautophagic drug temozolomide (2). Despite aggressive treatment regimens, the prognosis for GBM remains poor. Glioblastoma multiformes are distinguished pathologically from lower-grade tumors by the presence of necrosis and endothelial cell proliferation (3). Necrotic foci in malignant gliomas are typically surrounded by the distinct configuration of “pseudopalisading” tumor cells. These cells are severely hypoxic, overexpress hypoxia- inducible factor, and secrete proangiogenic factors such as vascular endothelial growth factor (VEGF) and interleukin 8 (3). An emerging model suggests that pseudopalisades represent a wave of tumor cells that are actively migrating away from the central hypoxic foci that arise as a consequence of a vascular insult (3).
Galectin (Gal) 1 belongs to a family of mammalian lectins with specificity for beta-galactosides (4, 5). It influences glioma cell migration both in vitro and in vivo (6, 7). Galectin 1 is expressed by tumor endothelial cells (8-10), and it contributes to different stages in tumor progression such as immune escape and metastasis (11, 12). Its expression is increased under hypoxic conditions (13), and it exerts potent proangiogenic effects (9). In addition to increasing Gal-1 expression, hypoxia itself influences angiogenesis (14, 15), and deficiencies in the oxygenation of solid tumors, including GBMs (3), are associated with a poor patient prognosis due to changes in cell metabolism, angiogenesis, invasiveness, and resistance to therapy (16). Endothelial cell Gal-1 is therefore a potential target for anti-cancer therapeutics (9, 10).
Recently, hypoxia has been shown to suppress protein synthesis through the regulation of the initiation step of mRNA translation (17). This seems to be a common feature of cell responses to hypoxia and is mediated by 2 distinct pathways. The first occurs rapidly, is transient, and is associated with activation of the unfolded protein response (UPR) that occurs in response to endoplasmic reticulum (ER) stress (17). Translation inhibition during this initial phase is due to phosphorylation of eukaryotic initiation factor 2alpha in a protein kinase-like ER kinase (PERK)-dependent manner (17-19). Although this effect is transient, the overall levels of translation remain low during hypoxia due to the inhibition of a second eukaryotic initiation complex (eIF4F) (17, 19). This second mechanism is multifactorial but in part results from inhibition of mammalian target of rapamycin kinase (17, 19). Inactivation of PERK or eukaryotic initiation factor 2alpha phosphorylation impairs cell survival in hypoxia (18, 19).
In this study, we investigated whether transiently impairing Gal- 1 expression in human Hs683 GBM cells can modify the pattern of ER stress and/or the UPR profile and how such modifications impact tumor angiogenesis. Our data provide insight into glioblastoma pathogenesis and suggest new approaches for treating GBMs.
MATERIALS AND METHODS
Cell Cultures and Compounds
The Hs683 human glioblastoma cell line (American Type Culture Collection code HTB-138) was obtained from the American Type Culture Collection (Manassas, VA) and maintained in our laboratory as detailed previously (20). Primary human umbilical vein endothelial cells (HUVECs) were obtained from umbilical cords at the Erasmus Academic Hospital. The cells were isolated by collagenase treatment and were grown in endothelial cell growth media EGM-2 MV bulletkit from Lonza (Verviers, Belgium). Human umbilical vein endothelial cell cultures were used between Passages 5 and 10. Temozolomide was purchased from Schering Plough (Brussels, Belgium).
Hs683 Human GBM Orthotopic Xenografts in Nude Mice
In vivo orthotopic xenografts of human Hs683 glioblastoma cells to nude mice were performed as previously described (20). In each experiment, 8-week-old female nu/nu mice (21-23 g; Iffa Credo, Charles River Laboratories, L’Arbresle, France) had Hs683 tumor cells stereotactically implanted on the same day. Each experimental group contained 11 mice. All of the in vivo experiments described in the present study were performed on the basis of Authorization No. LA1230509 of the Animal Ethics Committee of the Federal Department of Health, Nutritional Safety and the Environment (Belgium).
Anti-Gal-1 siRNA
The sense sequence of the anti-Gal-1 siRNA (Eurogentec; Seraing, Belgium) used in the current work was 5′-GCUGCCAGAUGGAUACGAADTDT- 3′, and the anti-sense sequence was 5′-UUCGUAUCCAUCUGGCAGCDTDT-3′. A corresponding scrambled siRNA was used as a control (sense, 5′- CUACGAUGCUGCUUAGCUCDTDT-3′; anti-sense, 5′-GAGCUAAGCAGCAUCGUAGDTDT- 3′). The same short strands of siRNA coupled with fluorescein were used to detect the in vivo distribution of the anti-Gal-1 siRNA in the brains of nude mice. The anti-sense and sense strands of the siRNA were annealed by the manufacturer in 50 mmol/L of Tris, pH 7.5 to 8.0, 100 mmol/L of NaCl in diethylpyrocarbonate-treated water. The final concentration of siRNA duplex was 100 [mu]mol/L. The anti- sense and sense strands of the scrambled control were annealed in the same way.
Human umbilical vein endothelial cells were transfected with INTERFERin(R) siRNA transfection reagent (Polyplus Transfection, Willemdorp, Belgium) according to the instructions provided by the manufacturer. The expression levels of Gal-1 in the transfected cells were evaluated at Day 5 posttransfection using computer- assisted fluorescence microscopy (Olympus AX70; Omnilabo, Antwerp, Belgium) equipped with a MegaView 2 digital camera and Analysis software (Soft Imaging System, Munster, Germany) as detailed previously (21).
Anti-Gal-1 siRNA Administration into Mouse Brains
Osmotic minipumps (Alzet, Charles River Laboratories) were used to infuse at a rate of 0.25 [mu]l/hour vehicle (0.9% NaCl) or nonviral siRNA. Either 0.3 mg of scrambled siRNA (control) or 0.3 mg of anti-Gal-1 siRNA were infused during a period of 2 weeks, as detailed elsewhere (22). Infusions were effected via the ventricular system of the brains of adult mice as described by Thakker et al (23). In addition to the continuous delivery of vehicle, control siRNA, or anti-Gal-1, siRNA were delivered by minipumps; vehicle, control siRNA, and siRNA were also directly injected into the same groups of Hs683 orthotopic xenografts on 3 occasions according to the schedule described in Figure 1A.
Each mouse bearing an Hs683 GBM in its brain underwent euthanasia in a carbon dioxide chamber for ethical reasons when it lost 15% of its weight compared with the first day of tumor grafting. The brain was removed from the skull, fixed in buffered formalin for 5 days, embedded in paraffin, and cut into 5-[mu]m-thick sections; histologic slides were stained with hematoxylin and eosin for vessel counts. To quantify angiogenesis, a grid was used to count numbers of blood vessels as illustrated previously (21). The types of vessels counted are illustrated in Figure 2A. Five fields at a Gx200 magnification were analyzed in each of 5 slides of each tumor; thus, counts were from 25 fields per tumor.
Genomic and Proteomic Analyses
Hs683 cells were transfected twice with control or siRNA directed against Gal-1, as we recently detailed with respect to mouse B16F10 melanoma cells (21). Briefly, Hs683 cells were either left untreated or were transfected during 16 hours with anti-Gal-1 siRNA or control siRNA (Day 0). On Day 1, the transfection procedure was repeated. On Day 2, the cells were washed with phosphate-buffered saline (PBS) and incubated under normal cell culture conditions. On Day 3, each group of cells was pooled and replated for subsequent experiments. On Days 5, 7, and 9, cells were either scraped into cold PBS buffer (for RNA extraction) or lysed directly in boiling lysis buffer (10 mmol/L of Tris, pH 7.4; 1 mmol/L of Na^sub 3^O^sub 4^V, 1% sodium dodecyl sulfate, pH: 7.4) for Western blot (WB) analyses with an anti-Gal-1 antibody (Preprotech TebuBio, Boechout, Belgium); this was performed for each experiment to confirm the effectiveness of the anti-Gal-1 siRNA treatment. RNA extraction and the determination of the quality and the integrity of the extracted RNA were assessed as detailed elsewhere (21). Full genome analyses were performed on Day 5 of transfection at the VIB MicroArray Facility (UZ Gasthuisberg, Catholic University of Leuven, Leuven, Belgium) using the Affymetrix Human Genome U133 set Plus 2.0 (High Wycome, UK).
Microarray Data Analyses
In addition to R, an open-source software environment for statistical computing (24), a set of functions called BioConductor (25) was used for the analysis and interpretation of the genomic data. The quality controls in the Affymetrix microarray experiments were performed with the Simpleaffy package (26) and agreed with the Affymetrix guidelines (Affymetrix, GeneChip Expression Analysis, 2002). The background correction, expression quantification, and normalization were performed using Robust Multichip Analysis (27). To select differentially expressed genes between 2 experimental conditions, probes were first identified for which no overlap occurred between intervals in the expression values obtained for each condition. The fold change between 2 experimental conditions was computed for each of these probes (without any value overlap) as the ratio between the 2 nearest unlog expression values observed for the 2 different conditions (i.e. the ratio closest to 1 between any 2 values from the 2 different conditions). Probes for which these ratios were more than 2.0 or less than 0.5 were then selected. The annotations of candidate genes were retrieved from the Affymetrix website through the BioConductor package “hgu133plus2”. The Expression Analysis Systematic Explorer software package (version 2.0; 28) was used to gather biologic information on the genes detected as overexpressed or downregulated according to the microarray analyses. This software package was then used to rank functional gene clusters by means of the statistical overrepresentation of individual genes in specific categories relative to all the genes in the same category on the microarray.
FIGURE 1. (A, B) Experimental protocol. Wild-type Hs683 cells were grafted orthotopically on Day 0 (DO) (A). Osmotic minipumps delivered vehicle (B; solid blue line), control siRNA (0.3 mg; B; solid pink line), or anti-Gal-1 siRNA (0.3 mg; B; solid green line) for 2 weeks into the third ventricle. Vehicle, control siRNA, and anti-Gal-1 siRNA were also directly injected into the same groups on Day 12 (D12), Day 19 (D19), and Day 26 (D26) posttumor grafting (A; orange arrows). Half of the mice in each WT, control siRNA- and SI- transfected group received 12 intravenous (tail vein) injections of either 50 [mu]l saline (solid color lines; 11 mice per group) or 50 [mu]l of a solution containing temozolomide to result in dosing at 40 mg/kg (B; hatched lines in; 11 mice per group), with the first temozolomide injection administered on D5 postgrafting (A; blue arrows). (C) Typical Hs683 orthotopic xenograft in the brain. (Ca) Conventional immunohistochemistry for Gal-1. (Cb-Cd) Immunofluorescence immunohistochemistry for Gal-1 (Cb, 5 mm from the field shown in Ca) and SI (Cc) in the same Hs683 orthotopic xenograft. (Cc) There is penetration of the SI-fluorescein isothiocyanate into the tumor (green fluorescence). The hatched white area delineates high siRNA concentrations; the 2 white arrows indicate areas of low SI concentration. (Cd) Galectin 1 expression (red fluorescence) in the same slide as in (Cc) indicates specific decreased expression in the area of siRNA penetration but higher levels in the areas of low SI concentration. (D) Western blotting analyses for Gal-1 expression in WT, SCR-, and SI-transfected Hs683 cells. D, day; SCR, scrambled siRNA; SI, anti-Gal-1 siRNA; WT, wild type.
Protein Expression Measurements
Western blot and immunofluorescence (IF) analyses were performed as detailed previously (6, 21, 22). Control experiments, including the omission of the incubation step with the primary antibodies (negative control), were performed. Equal loading was verified by the bright Ponceau red coloration of the membranes. The integrity and quantity of the extracts were assessed by means of alpha-actin or tubulin immunoblotting. The proteins submitted as blots to WB and/ or entire cells to IF analyses were detected using the following primary antibodies: anti-microvascular differentiation gene (MDG1; dilution, 1/1000; AbCam, Cambridge, UK), anti-oxygen-regulated protein 150 (ORP150; dilution, 1/100; IBL, Minneapolis, MN), anti- VEGF (dilution, 1/50; SantaCruz Tebu-Bio, Boechout, Belgium), anti- activating transcription factor (ATF) 6 (dilution, 1/500 WB; 1/100 IF; Imgenex; Bio-Connect BV, Huissen, The Netherlands), anti-ER transmembrane kinase/ribonuclease inositol-requiring (IRE) 1alpha (dilution, 1/500 WB; 1/50 IF; Cell Signaling; Bioke, Leiden, The Netherlands), anti-PERK (dilution, 1/500 WB; 1/50 IF; Cell Signaling; Bioke), anti-X-box-binding protein 1 (XBP1; dilution, 1/ 500 WB; 1/100 IF, Biolegend; Sanbio BV, Uden, The Netherlands), upstream binding factor (dilution, 1/200; SantaCruz TebuBio), anti- Gal-3 (dilution, 1/100; Novocastra; Newcastle, UK); anti-Gal-2 (dilution, 1/100), and anti-Gal-9 (dilution, 1/100) were purchased from R&D Systems (Abingdon, UK). Secondary antibodies were purchased from Pierce (PerbioScience, Erembodegem, Belgium) for the WBs and from Molecular Probes (Invitrogen, Merelbeke, Belgium) for fluorescent detection (Alexafluor conjugated antibodies). Western blots were developed using the Pierce Supersignal Chemiluminescence system.
FIGURE 2. (A) Blood vessels in Hs683 orthotopic xenographs. (Aa) Typical morphology of the blood vessels (white arrows; hematoxylin and eosin staining: x 200). (Ab) Vessel counts in SI-transfected, SCR-transfected, and WT Hs683 orthotopic xenografts (Fig. 1B). (B) Wild-type (Ba) and SCR-transfected (Bb) Hs683 cells developed vasculogenic capillary networks when cultured on Matrigel, whereas SI-transfected cells did not (Bc). The SI that decreased Gal-1 expression in Hs683 glioblastoma multiforme cells (Fig. 1D) also decreased Gal-1 expression in human umbilical vein endothelial cells (HUVECs) (Cd, with Cc as bright field control) when compared with SCR-transfected HUVECs (Cb, with Ca as bright field control). (D) Although the SI decreased Gal-1 expression in HUVEC cells (Cd) compared with WT (not shown) and SCR-transfected cells (Cb), it did not prevent HUVECs from forming capillary networks (Dc) when compared with WT (Da) and SCR-transfected HUVECs (Db). SI, anti-Gal- 1 siRNA; SCR, control siRNA; WT, wild type.
TABLE 1. Gene Categories Overrepresented in the Set of Genes Detected as Differentially Expressed After Decreasing Gal-1 Expression in Human Hs683 GBM Cells
The staining patterns of PERX and IRE-1alpha in Hs683 cells cultured in vitro were analyzed by means of a computer-assisted fluorescent Olympus AX70 microscope (Omnilabo) equipped with a MegaView2 digital camera and analysis software (Soft Imaging System), as detailed previously (21). For immunohistochemical detection of ORP150 in Hs683 orthotopic xenografts, the tumors were fixed for 24 hours in 4% formaldehyde, dehydrated, and routinely embedded in paraffin. Immunohistochemistry was performed on 5-[mu]m- thick sections on silane-coated glass slides. The sections were incubated with 0.4% hydrogen peroxide for 5 minutes to block endogenous peroxidase activity, rinsed in PBS (0.04 mol/L Na^sub 2^HPO^sub 4^, 0.1 mol/L KH^sub 2^PO, and 0.12 mol/L NaCl; pH 7.4), and exposed successively for 20 minutes to avidin (0.1 mg/ml in PBS) and biotin (0.1 mg/ml in PBS) to inactivate endogenous biotin. After rinsing in PBS, the sections were incubated for 20 minutes with 0.5% casein in PBS and exposed sequentially at room temperature to 1) the primary specific anti-ORP150 antibody (see previous discussion), 2) the biotinylated secondary antibody (DakoCytomation, Glostrup, Denmark), and 3) the avidin-biotin-peroxidase complex (ABC kit; DakoCytomation). The presence of labeled peroxidase on the sections was visualized by incubation with a chromogen substrate containing diaminobenzidine and H^sub 2^O^sub 2^. For controls, the primary specific antibody was omitted or replaced by nonimmune antisera. In all cases, these controls were negative.
The presence of labeled peroxidase was quantitatively determined using a KS400 imaging system (Carl Zeiss Vision, Hallbergmoos, Germany). For each Hs683 tumor, 15 fields corresponding to a total surface ranging from 60,000 to 120,000 [mu]m^sup 2^ were scanned. The analysis of the immunohistochemical expression of ORP150 by computer-assisted morphometry was quantitatively expressed by the labeling index, that is, the percentage of cells positively stained with the ORP150 antibody.
Transcription Factor Activity Analyses
Hs683 cells were transfected twice with anti-Gal-1 siRNA or control siRNA, and 1 group of cells was incubated with recombinant Gal-1 (0.1 ng/ml) for 24 hours. On Day 4, cells were scraped into cold PBS buffer, and nuclear extracts were isolated from these cells using the nuclear extraction kit from Active Motif (Rixensart, Belgium; catalog no. 40010).
Arrays were performed using TranSignal protein/DNA arrays according to the manufacturer’s instructions. In brief, biotin- labeled DNA-binding oligonucleotides (TranSignal probe mix; Panomics) were incubated with 15 [mu]g of nuclear extract to allow the formation of protein/DNA (or TF/DNA) complexes. The protein/DNA complexes were then separated from the free probes using the provided spin columns. The probes were hybridized to the TranSignal array membrane overnight at 42[degrees]C. Each membrane was then incubated with streptavidin-horseradish peroxidase conjugate, developed with a substrate solution containing luminol enhancer and peroxide solution, and then exposed using Hyperfilm enhanced chemiluminescence (Amersham). FIGURE 3. (A, B) Immunofluorescence analyses (with bright field controls) of ORP150 in scrambled siRNA treated (SCR) (A) and anti-galactin 1 siRNA treated (SI) (B) transfected Hs683 cells 5 days after transaction. (C) Western blot analyses of ORP150 expression in wild-type (WT), SCR-, or SI- transfected Hs683 cells (after 5, 7, or 9 days). (D) Immunohistochemical expression of ORP150 in an SCR (Da)- and SI (Db)- transfected Hs683 GBM (black arrows). (Dc) Quantitative determination by computer-assisted microscopy of ORP150 expression. (E, F) Immunofluorescence staining with bright field controls of vascular endothelial growth factor in SCR (E)- and SI (F)- transfected Hs683 cells after 5 days. Greater staining in (F) indicates vascular endothelial growth factor retention within the cells.
Statistical Analyses
Survival analyses were carried out by means of Kaplan-Meier curves and Gehan generalized Wilcoxon test. Statistical comparisons between control and treated groups were established by carrying out the Kruskal-Wallis test (a nonparametric 1-way analysis of variance) or the Mann-Whitney test (in the case of 2 groups). All of the statistical analyses were undertaken using Statistica (Statsoft, Tulsa, OK).
RESULTS
Delivering Anti-Gal-1 siRNA into Hs683 CBM Orthotopic Xenografts Increases the Beneficial Effects of Temozolomide
Wild-type Hs683 glioblastoma cells (not transfected with either scrambled or anti-Gal-1 siRNA) were grafted into the brains of nude mice. From Day 5 postgrafting of the cells, osmotic minipumps infused either vehicle alone, control siRNA, or anti-Gal-1 siRNA into the mice as indicated in Figure 1A. Vehicle, control siRNA, and anti-Gal-1 siRNA were also directly injected at 3 time points into Hs683 orthotopic xenografts (Fig. 1A). The mice also received 12 injections (3 per week for 4 weeks, i.v.) of either 50 [mu]l saline (control) or 50 [mu]l of a solution containing temozolomide equivalent to a dose of 40 mg/kg; the first injection started on Day 5 postgrafting (Fig. 1A). Temozolomide significantly increased the survival of Hs683 GBM-bearing mice in all experimental groups (i.e. wild type + vehicle vs wild type + temozolomide [p
FIGURE 4. Immunofluorescence staining with bright field controls of microvascular differentiation gene (MDG1) in control siRNA- and anti-galectin 1 siRNA-transfected Hs683 cells after 5 days. The UBF is a specific nucleolar marker. MDG1 colocalizes with UBF in the nucleoli along with decreased Gal-1 expression. UBF, upstream binding factor.
Well-established Hs683 orthotopic xenografts were present in the brains 30 days postimplantation, and Gal-1 expression was detected in the xenografts (Fig. 1Ca, Cb). Anti-Gal-1 siRNA-fluorescein isothiocyanate (delivered by micro-pump and indicated by the green fluorescence) penetrated into the tumors (Fig. 1Cc). Double immunostaining demonstrated that Gal-1 expression was decreased in areas with high anti-Gal-1 siRNA concentrations, whereas Gal-1 expression remained high in areas where the concentration of the anti-Gal-1 siRNA was lower (Fig. 1Cd). The decrease in Gal-1 expression in Hs683 GBM cells lasted for at least 6 days after the cells were transfected in vitro with the anti-Gal-1 siRNA (Fig. 1D). The anti-Gal-1 siRNA did not modify levels of expression of Gal-2, Gal-3, or Gal-9 (data not shown).
Decreasing Gal-1 Expression in Hs683 GBM Cells Impairs In Vivo Angiogenesis and In Vitro Vasculogenic Mimicry
Tumors were collected from the brains of xenograft-bearing mice from the experiment detailed in Figure 1. Figure 2Aa illustrates the typical blood vessels seen in these tumors. In vivo delivery of the anti-Gal-1 siRNA significantly (p
FIGURE 5. Immunofluorescence analyses (with bright field controls) of XBP-1, ATF6, P-PERK, and of inositol-requiring (IRE) 1alpha in SCR (A, C, E, G)- or SI (B, D, F, H)-transfected (after 5 days) Hs683 cells. Phospho-PERK (I) and IRE-1alpha (I) immunofluorescence staining expression was quantitatively determined by means of computer-assisted microscopy. (J) Western blotting analyses of IRE-1alpha expression in wild-type (Wt), SCR, or SI transfected after 4 or 5 days in Hs683 cells (loading control; tubulin). ATF, activating transcription factor; P-PERK, phospho- protein kinase-like endoplasmic reticulum kinase; SCR, control siRNA; SI, anti-galectin 1 siRNA XBP-1, X-box-binding protein.
Aggressive tumors, including malignant melanomas and gliomas, may generate tumor cells that form vascular channels that facilitate tumor perfusion independent of tumor angiogenesis (29, 30). Because hypoxia influences angiogenesis, vasculogenic mimicry channel formation (i.e. generation of capillary-like networks in vitro), and tumor-invasion-related protein expression in melanoma (31), we investigated whether Hs683 GBM cells are able to perform vasculogenic mimicry-like processes in vitro. When wild-type (Fig. 2Ba) and control-transfected (Fig. 2Bb) Hs683 GBM cells were cultured on Matrigel, they generated networking processes of vasculogenic mimicry, whereas anti-Gal-1 siRNA-transfected Hs683 cells did not (Fig. 2Bc). In contrast, normal HUVECs express high levels of Gal-1 (Fig. 2Ca, Cb), significantly decreasing the Gal-1 expression in approximately 70% of these cells (Figs. 2Cc, Cd), did not impair their ability to form capillary networks (Fig. 2Dc) when compared with wild-type (Fig. 2Da) or control-transfected HUVECs (Fig. 2Db). These results were reproduced in 3 distinct batches of HUVECs (data not shown). Thus, the anti-angiogenic effects obtained on reducing Gal-1 expression in Hs683 tumors seem to be relatively specific to tumor angiogenesis.
FIGURE 6. Endoplasmic reticulum (ER) stress-related control of UPRE- and/or ER stress element (ERSE) sequence-dependent gene expression depicted according to the references cited (Table 2). ATF, activating transcription factor; Bip/CRP78: immunoglobulin heavy chain-binding protein/glucose-regulated protein, 78 kd; elF2, eukaryotic initiation factor 2; GADD, growth arrest and DNA damage; HERP, homocysteine-inducible ER stress-inducible protein; Hsp70, 70- kDa heat shock protein; IRE, inositol requiring; NF-Y, nuclear transcription factor Y; ORP150, 150-kDa oxygen-related protein; PDIA4, protein disuifide isomerase associated 4; PERK, protein kinase-like ER kinase; U-Pr, unfolded protein; UPRE, unfolded protein response element; XBP-1, X-box-binding protein.
Decreasing Gal-1 Expression in Hs683 GBM Cells Impairs the ER Stress Response
Hs683 cells in which Gal-1 expression was reduced by anti-Gal-1 siRNA transfection were subjected to a microarray analysis. One hundred thirty-three genes were found to be overexpressed by greater than 200% (“expression level”>2.0 in Table 1) or decreased by greater than 50% (“expression level”
Transient Cal-1 Reduction in Hs683 GBM Cells Reduces MDG1 and ORP150 Expression and Impairs Angiogenesis Via Retention of VEGF
Among the list of genes in Table 1, ORP150 and the MDG1 are known to have major roles in angiogenesis. Oxygen-regulated protein 150 is an inducible ER chaperone that plays a major role in tumor-mediated angiogenesis via the processing of VEGF (32). Oxygen-regulated protein 150 was first identified and cloned from cultured astrocytes on the basis of their ability to withstand and even produce neurotrophic factors in response to severe hypoxia (33). We first validated the gene expression profiling results for ORP150 by IF (Figs. 3A, B) and WB (Fig. 3C). We further validated our in vitro (Figs. 3A-C) data by immunohistochemical staining of the tumor samples. The in vivo delivery of anti-Gal-1 siRNA significantly (p
TABLE 2. Identification of the Transcription Factors Targeting the Genes*
The expression of angiogenic factors such as VEGF under conditions of cell stress involves both transcriptional and translational events and an important role for inducible ER chaperones (32, 33). The in vivo reduction of ORP150 expression in C6 rat gliomas induces a decrease in angiogenesis in C6 gliomas, whereas the in vitro inhibition of ORP150 expression decreases the release of VEGF into the supernatant of C6 cultured tumor cells (32). In ORP150 anti-sense transfected C6 rat glioma cells, VEGF accumulates within the ER (32). Increased levels of ORP150 promote VEGF processing with subsequent transport from the ER to the Golgi, followed by secretion from the cell (34). The cell death mechanism in tumors arising from ORP150 anti-sense C6 transfectants results largely from the inhibition of angiogenesis, which is caused by the retention of VEGF in the ER (32). In the present study, the decrease in Gal-1 expression in Hs683 GBM cells resulted in a reduction in VEGF secretion into the culture media (data not shown) and marked accumulation in anti-Gal-1 siRNA-transfected (Fig. 3F) compared with control siRNA-transfected (Fig. 3E) Hs683 cells. This likely resulted in impairment of angiogenesis in vivo (Fig. 2Ab). Both the mRNA (Table 1) and cellular expression (Fig. 4) of MDG1 (DNAJB9; Erdj4) were markedly decreased in Hs683 GBM cells with reduced Gal- 1 expression. High levels of MDG1 gene expression reflect the activation state of endothelial cells (35), and its upregulation in endothelial cells is induced by ER stress (36). Use being made of the upstream binding factor (UBF) as a marker of nucleoli (37). Microvascular differentiation gene may also play a number of roles in stabilizing 78-kd glucose-regulated protein (GRP78) binding to unfolded substrate proteins in a J domain-dependent manner, thus preventing the accumulation of unfolded proteins in the ER, thereby protecting cells from ER stress (36). The mRNA levels of GRP78 were decreased more than 4-fold when Gal-1 expression was reduced in Hs683 cells (Table 1). Microvascular differentiation gene was originally isolated from differentiating microvascular endothelial cells cultured in collagen Type I gels (3D culture) and is known to be upregulated in primary endothelial and mesangial cells when subjected to various stress stimuli (36). Microvascular differentiation gene proteins are located in the cytoplasm under control conditions, but stress induces their translocation into the nucleus where they accumulate in the nucleoli (36). Similarly, the transient reduction of Gal-1 in Hs683 GBM cells caused decreases in cytoplasmic MDG1 expression with concomitant accumulation in nucleoli (Fig. 4).
Gal-1 May Modulate ORP150 Expression Through IRE-1alpha
Galectin 1 expression in Hs683 cells did not modify the expression levels of XBP-1 (Figs. 5A, B), ATF6 (Figs. 5C, D) or RNA- dependent PERK (Figs. 5E, F). In contrast, decreasing Gal-1 expression was accompanied by a marked decrease in IRE-1alpha expression (Figs. 5G, I; p
Clear relationships have been identified for the targets analyzed in Figure 5 and the signaling pathways in which they are involved (Fig. 6). Inositol-requiring 1alpha has a major role in mRNA splicing processes, that is, cellular processes in which Gal-1 participates (see Discussion section).
TABLE 3. Hypoxia-Related Genes (in Addition to ORP150 and MDG1) Involved in Angiogenesis, the mRNA Expression Levels of Which Were Modified in Anti-Galactin 1 siRNA-Treated Hs683 GBM Cells
The ER stress response element (ERSE) sequence evidenced in the ORP150 gene is also present in the promoters of other genes, including, for example, GRP78 and homocysteine-inducible ER stress- inducible protein (Fig. 6). The expression of each of the genes for these proteins is also downregulated in anti-Gal-1 siRNA- transfected Hs683 cells (Table 1). Table 2 gives an overview of potential Gal-1-targeted genes (Table 1), the promoters of which contain sequences that are targeted by transcription factors involved in ER stress and/or the UPR (Fig. 6). In brief, the data in Table 2 suggest that decreasing Gal-1 expression in Hs683 GBM cells induces a decrease in the mRNA levels of a cluster of genes that are directly implicated in ER stress and/or the UPR and the promoters of which are activated by ERSE or ERSE-related targeting transcription factors.
Decreasing Gal-1 Expression Decreases mRNA of Hypoxia-Related Genes that are Implicated in Angiogenesis
Knowing that hypoxia stimulates Gal-1 expression (13) and that Gal-1 is a potent proangiogenic factor (9) that exerts a major influence on MDG1 and ORP150 at both gene expression (Table 1) and protein (Figs. 3, 4) levels, the list of 133 genes whose expression was modified on Gal-1 knockdown was investigated for those known to be hypoxia regulated and implicated in angiogenesis. Table 3 reveals that 6 hypoxia-related genes (CTGF, ATF3, PPP1R15A, HSPA5, TRA1, and CYR61) implicated in angiogenesis (in addition to ORP150 and MDG1) had their mRNA levels decreased in Hs683 cells when Gal-1 expression was reduced. Galectin 1 thus seems to be a key effector of hypoxia- mediated modification in the expression of several genes that are implicated in angiogenesis.
DISCUSSION
Hypoxia stimulates Gal-1 expression (13) and activates the UPR (54, 55). Hypoxia also suppresses protein synthesis through the regulation of the initiation step of mRNA translation (17). The present study suggests that Gal-1 can be a key mediator in these processes by modulating a general mechanism of tumor cell defense. Because Gal-1 is itself stimulated by hypoxia, this would seem to be the case at least in the human Hs683 GBM and the mouse B16F10 melanoma (23) models. The protective effects of Gal-1 in cancer cells seem to relate at least in part to the control of the expression of a cluster of hypoxia-related genes that are known to be involved in angiogenesis and include ORP150. Galectin 1 may thus control ORP150 expression, and therefore VEGF maturation, through activation of IRE-1alpha (Figs. 5, 6). Both Gal-1 and IRE-1alpha are implicated in mRNA splicing.
Cultured astrocytes that are exposed to hypoxic stress overexpress GRP78/Bip, GRP94, and ORP150 (56). Furthermore, the localization of these 3 proteins to the ER (57) suggests that hypoxia induces a stress response the focal point of which may be in this organelle (56). Decreasing Gal-1 expression, even transiently, markedly decreased the levels of expression of GRP78/Bip, GRP94, and ORP150 and numerous other hypoxia-related genes (Tables 1-3) in Hs683 cells. Oxygen-regulated protein 150 may be cytoprotective against ischemia/reperfusion injury in the human (58) and rodent brain (59) and in renal tubular epithelium (60) via reduction of ER stress and probably also in the inhibition of apoptosis (60). As in the present study, inhibition of ORP150 expression in cultured macrophages causes retention of VEGF within the ER, whereas overexpression of ORP150 promotes the secretion of VEGF into hypoxic culture supernatants (32, 34). Increased levels of ORP150 promote VEGF processing with subsequent transport from the ER to the Golgi, followed by export out of the cell (32, 34). Decreasing Gal-1 expression in Hs683 cells markedly decreased ORP150 expression and resulted in accumulation of VEGF in Hs683 cells (Fig. 3).
In mammalian cells, there are 4 sensors that respond to ER stress: IRE-1alpha, IRE-1beta, PERK, and ATF6 (34, 61) (Fig. 6). Protein kinase-like ER kinase plays an important role in the attenuation of translation and apoptosis during ER stress by activating a pathway controlled by ATF4, GADD 153 and the CCAAT/ enhancer-binding protein homologous protein (61). The present study shows that decreasing Gal-1 transiently in Hs683 cells did not modify the expression of PERK (either at the genomic or proteomic levels) or of ATF4 at the genomic level (data not shown). It also did not induce apoptosis in these cells (data not shown).
Inositol-requiring 1beta is reported to be involved in the ER stress-dependent degradation of rRNA (61, 62). The present study reveals that decreasing Gal-1 transiently did not modify IRE-1beta expression at either the genomic or proteomic level (data not shown) in HS683 cells. Inositol-requiring 1alpha and ATF6 play essential roles in the induction of ER chaperones. Activating transcription factor 6 is a 90-kd Type II ER transmembrane protein containing a DNA-binding domain in the N-terminal cytoplasmic region, whereas its C-terminal domain resides in the ER lumen (61). Upon stimulation by ER stress, ATF6 is cleaved by Site 1 and Site 2 proteases, the same enzymes that cleave the sterolresponse-element binding protein (44). The 50-kd N-terminal half of the protein is transported to the nucleus and recognizes the ERSE (consensus sequence CCAATN^sub 9^C- CACG) in the promoters of ER chaperone genes, including GRP78, GRP94, ORP150, homocysteine-inducible ER stress-inducible protein, and PDIA4/ERP70 (Fig. 6; Table 2). These expression levels of all of these genes were modified by transiently decreasing Gal-1 expression. Activating transcription factor 6 binds ERSE only in the presence of nuclear factor Y, which recognizes the CCAAT box (61) (Fig. 6). Endoplasmic reticulum stress-activated ATF6 induces the transcription of XBP1 mRNA through the ERSE element of the XBP1 gene (61). Decreasing Gal-1 transiently in Hs683 cells did not modify XBP1 expression either at genomic or proteomic levels, nor its activity at the transcriptomic level (data not shown).
X-box-binding protein 1 mRNA is spliced by the IRE-1alpha protein under conditions of ER stress (61). The XBP1 protein translated from the spliced mRNA is active for transcription, whereas those translated from the unspliced mRNA is inactive and rapidly degraded (61). Active XBP1 recognizes ERSE in the presence of nuclear factor Y and activates the transcription of the XBP1 gene itself and of ER chaperones (44). The current data revealed that decreasing Gal-1 transiently in Hs683 cells markedly modified the level of expression of IRE-1alpha (Fig. 5) in these cells, thus preventing XBP1 splicing.
The initial hint that Gal-1 might play a role in premRNA splicing came from the fact that nuclear extracts from HeLa cells contain both Gal-1 and Gal-3, and that the depletion of both Gals from these nuclear extracts either by lactose affinity adsorption or by double- antibody adsorption results in a loss of splicing activity (63, 64). Moreover, Gal-1 and Gal-3 were shown by double-IF experiments to be colocalized in nuclear speckles, with known splicing factors such as the core polypeptides of small nuclear ribonucleoproteins bearing the Sm epitope and the small nuclear ribonucleoprotein splicing factor SC35. Furthermore, it was shown in a 2-hybrid screen that Gal- 1 directly binds to Gemin4, a component of nuclear complexes containing survival of motor neuron proteins (63). It was also shown that antisera to Gal-1 and Gal-3 coimmunoprecipitate with protein factors, including HnRNP C1/C2 and Slu7, which are components of the spliceosomal complexes (65). The novel aspects of Gal-1-related function in the ER stress response that are highlighted in the present study may be amenable to therapeutic manipulation either by the in vivo delivery of anti-Gal-1 siRNA as demonstrated here or through chemicals that suppress Gal-1 expression and/or biologic activity (7, 8, 66, 67). The in vivo delivery of antiGal-1 siRNA can be performed directly in cases of GBMs using Ommaya reservoirs, thus minimizing/avoiding systemic release/exposure and potential hepatotoxicity. Decreasing the levels of Gal-1 expression in gliomas may thus i) contribute to decreasing the migration levels of individual GBM cells diffusely invading the brain parenchyma (6, 7; Fig. 2Bc); ii) weaken glioma cell defenses by reducing their ability to respond to ER stress and/or the UPR; and iii) impair angiogenesis. These effects were collectively seen in the present study to reinforce the therapeutic benefits of the proautophagic drug temozolomide, the current standard chemotherapeutic treatment for GBM patients, in a glioblastoma model in vivo.
The present study does not show direct cause and effect of the apparent improvement in response to temozolomide on Gal-1 knockdown. Our observations may be consistent with the Gal-1 knockdown- mediated effects on hypoxia/angiogenesis and stress genes, or with the fact that these genes are implicated in pathways that may account for the improved response to temozolomide. We are currently performing experiments to further elucidate the mechanism for the Gal-1 knockdown enhancement of the therapeutic benefits of temozolomide in experimental glioma models.
REFERENCES
1. Lefranc F, Brotchi J, Kiss R. Possible future issues in the treatment of glioblastomas: Special emphasis on cell migration and the resistance of migrating glioblastoma cells to apoptosis. J Clin Oncol 2005;23:2411-22.
2. Lefranc F, Sadeghi N, Camby I, et al. Present and potential future issues in glioblastoma treatment. Expert Rev Anticancer Ther 2006;6:719-32
3. Rong Y, Durden DL, Van Meir EG, et al. ‘Pseudopalisading’ necrosis in glioblastoma: A familiar morphologic feature that links vascular pathology, hypoxia, and angiogenesis. J Neuropathol Exp Neurol 2006;65: 529-39
4. Liu FT, Rabinovich GA. Galectins as modulators of tumour progression. Nat Rev Cancer 2005;5:29-41
5. Camby I, Le Mercier M, Lefranc F, et al. Galectin-1: A small protein with major functions. Glycobiology 2006;16:137R-57R
6. Camby I, Belot N, Lefranc F, et al. Galectin-1 modulates human glioblastoma cell migration into the brain through modifications to the actin cytoskeleton and levels of expression of small GTPases. J Neuropathol Exp Neurol 2002;61:585-96
7. Camby I, Belot N, Rorive S, et al. Galectins are differentially expressed in supratentorial pilocytic astrocytomas, astrocytomas, anaplastic astrocytomas and glioblastomas, and significantly modulate tumor astrocyte migration. Brain Pathol 2001;11:12-26
8. Baum LG, Seilhamer JJ, Pang M, et al. Synthesis of an endogeneous lectin, galectin-1, by human endothelial cells is up- regulated by endothelial cell activation. Glycoconj J 1995;12:63-68
9. Thijssen VL, Postel R, Brandwijk RJ, et al. Galectin-1 is essential in tumor angiogenesis and is a target for antiangiogenesis therapy. Proc Natl Acad Sci U S A 2006;103:15975-80
10. Thijssen VL, Poirier F, Baum LG, et al. Galectins in the tumor endothelium: Opportunities for combined cancer therapy. Blood 2007; 110:2819-27
11. Perillo NL, Pace KE, Seilhamer JJ, et al. Apoptosis of T cells mediated by galectin-1. Nature 1995;378:736-39
12. Rabinovich GA, Gabrilovich D, Sotomayor EM. Immunosuppressive strategies that are mediated by tumor cells. Annu Rev Immunol 2007;25: 267-96
13. Le QT, Shi G, Cao H, et al. Galectin-1: A link between tumor hypoxia and tumor immune privilege. J Clin Oncol 2005;23:8932-41
14. Keith B, Simon MC. Hypoxia-inducible factors, stem cells, and cancer. Cell 2007;129:465-72
15. Kim JW, Gao P, Dang CV. Effects of hypoxia on tumor metabolism. Cancer Metastasis Rev 2007;26:291-98
16. Magagnin MG, Koritzinsky M, Wouters BG. Patterns of tumor oxygenation and their influence on the cellular hypoxic response and hypoxia-directed therapies. Drug Resist Updat 2006;9:185-97
17. van den Beucken T, Koritzinsky M, Wouters BG. Translational control of gene expression during hypoxia. Cancer Biol Ther 2006;5:749-55
18. Koumenis C. ER stress, hypoxia tolerance and tumor progression. Curr Mol Med 2006;6:55-69
19. Pouyssegur J, Dayan F, Mazure NM. Hypoxia signalling in cancer and approaches to enforce tumour regression. Nature 2006;441:437-43
20. Branle F, Lefranc F, Camby 1, et al. Evaluation of the efficiency of chemotherapy in in vivo orthotopic models of human glioma cells with and without 1p19q deletions and in C6 rat orthotopic allografts serving for the evaluation of surgery combined with chemotherapy. Cancer 2002;95:641-55
21. Mathieu V, Le Mercier M, De Neve N, et al. Galectin-1 knockdown increases sensitivity to temozolomide in a B16F10 mouse metastatic melanoma model. J Invest Dermatol 2007;127:2399-2410
22. Lefranc F, Sadeghi N, Metens T, Brotchi J, Salmon I, Kiss R. Characterization of gastrin-induced cytostatic effect on cell proliferation in experimental gliomas. Neurosurgery 2003;52:881-90
23. Thakker DR, Natt F, Husken D, et al. Neurochemical and behavioral consequences of widespread gene knockdown in the adult mouse brain by using nonviral RNA interference. Proc Natl Acad Sci USA 2004; 101:17270-75
24. Development Core Team. 2006. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing.
25. Gentleman RC, Carey VJ, Bates DM, et al. Bioconductor: Open software development for computational biology and bioinformatics. Genome Biol 2004;5:R80
26. Wilson CL, Miller CJ. Simpleaffy: A BioConductor package for Affymetrix Quality Control and data analysis. Bioinformatics 2005;21: 3683-85
27. Bolstad BM, Irizarry RA, Astrand M, et al. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 2003;19:185-93
28. Hosack DA, Dennis G Jr, Sherman BT, et al. Identifying biological themes within lists of genes with ease. Genome Biol 2003;4:R70
29. Hendrix MJ, Seftor EA, Hess AR, et al. Vasculogenic mimicry and tumour-cell plasticity: Lessons from melanoma. Nat Rev Cancer 2003;3: 411-21
30. Yue WY, Chen ZP. Does vasculogenic mimicry exist in astrocytoma? J Histochem Cytochem 2005;53:997-1002
31. Sun B, Zhang D, Zhang S, et al. Hypoxia influences vasculogenic mimicry channel formation and tumor invasion-related protein expression in melanoma. Cancer Lett 2007;249:188-97
32. Ozawa K, Tsukamoto Y, Hori O, et al. Regulation of tumor angiogenesis by oxygen-regulated protein 150, an inducible endoplasmic reticulum chaperone. Cancer Res 2001;61:4206-13
33. Kuwabara K, Matsumoto M, Ikeda J, et al. Purification and characterization of a novel stress protein, the 150-kDa oxygen- regulated protein (ORP150), from cultured rat astrocytes and its expression in ischemic mouse brain. J Biol Chem 1996;271:5025-32
34. Ozawa K, Kondo T, Hori O, et al. Expression of the oxygen- regulated protein ORP150 accelerates wound healing by modulating intracellular VEGF transport. J Clin Invest 2001;108:41-50
35. Berger BJ, Muller TS, Buschmann IR, et al. High levels of the molecular chaperone Mdgl/ERdj4 reflect the activation state of endothelial cells. Exp Cell Res 2003;290:82-92
36. Prols F, Mayer MP, Renner O, et al. Upregulation of the cochaperone Mdgl in endothelial cells is induced by stress and during in vitro angiogenesis. Exp Cell Res 2001;269:42-53
37. Spector DL. Nuclear domains. J Cell Sci 2001;114:2891-93
38. Yamamoto K, Yoshida H, Kokame K, Kaufman RJ et al. Differential contributions of ATF6 and XBPl to the activation of endoplasmic reticulum stress-responsive cis-acting elements ERSE, UPRE and ERSE-II. J Biochem 2004;136:343-50
39. Yoshida H, Haze K, Yanagi H, et al. Identification of the as- acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose-regulated proteins. Involvement of basic leucine zipper transcription factors. J Biol Chem 1998;273: 33741-49
40. Yoshida H, Okada T, Haze K, et al. ATF6 activated by proteolysis binds in the presence of NF-Y (CBF) directly to the cis- acting element responsible for the mammalian unfolded protein response. Mol Cell Biol 2000;20:6755-67
41. Kokame K, Kato H, Miyata T. Identification of ERSE-II, a new cis-acting element responsible for the ATF6-dependent mammalian unfolded protein response. J Biol Chem 2001;276:9199-9205
42. Wang Y, Shen J, Arenzana N, et al. Activation of ATF6 and an ATF6 DNA binding site by the endoplasmic reticulum stress response. J Biol Chem 2000;275:207013-20
43. Mizobuchi N, Hoseki J, Kubota H, et al. ARMET is a soluble ER protein induced by the unfolded protein response via ERSE-II element. Cell Struct Funct 2007;23:41-50
44. Wolfgang CD, Chen BPC, Martindale JL, et al. Gadd153/Chop10, a potential target gene of the transcriptional repressor ATF3. Mol Cell Biol 1997;17:6700-6707
45. Kaneda S, Yura T, Yanagi H. Production of three distinct mRNAs of 150 kDa oxygen-related protein (ORP150) by alternative promoters: Preferential induction of one species under stress conditions. J Biochem 2000;128:529-38
46. Shimo T, Kubota S, Kondo S, et al. Connective tissue growth factor as a major angiogenic agent that is induced by hypoxia in a human breast cancer cell line. Cancer Lett 2001;174:57-64 47. Ameri K, Hammond EM, Culmsee C, et al. Induction of activating transcription factor 3 by anoxia is independent of p53 and the hypoxic HIF signaling pathway. Oncogene 2007;26:284-89
48. Blais JD, Filipenko V, Bi M, et al. Activating transcription factor 4 is translationally regulated by hypoxic stress. Mol Cell Biol 2004;24: 7469-82
49. Kaufman RJ. Stress signaling from the lumen of the endoplasmic reticulum: Coordination of gene transcriptional and translational controls. Genes Dev 1999;13:1211-33
50. Hori O, Matsumoto M, Kuwabara K, et al. Exposure of astrocytes to hypoxia/reoxygenation enhances expression of glucose- regulated protein 78 facilitating astrocyte release of the neuroprotective cytokine interleukin 6. J Neurochem 1996;66:973-79
51. Paris S, Denis H, Delaive E, et al. Up-regulation of 94-kDa glucose-regulated protein by hypoxia-inducible factor-1 in human endothelial cells in response to hypoxia. FEBS Lett 2005;579:105-14
52. Babic AM, Kireeva ML, Kolesnikova TV, et al. CYR61, a product of a growth factor-inducible immediate early gene, promotes angiogenesis and tumor growth. Proc Natl Acad Sci U S A 1998;95:6355- 60
53. Kunz M, Moeller S, Koczan D, et al. Mechanisms of hypoxic gene regulation of angiogenesis factor Cyr61 in melanoma cells. J Biol Chem 2003;278:45651-60
54. Zhang K, Kaufman RJ. The unfolded protein response: A stress signaling pathway critical for health and disease. Neurology 2006;66:S102-9
55. Zhao L, Ackerman SL. Endoplasmic reticulum stress in health and disease. Curr Opin Cell Biol 2006;18:444-52
56. Ozawa K, Kuwabara K, Tamatani M, et al. 150-kDa oxygen- regulated protein (ORP150) suppresses hypoxia-induced apoptotic cell death. J Biol Chem 1999;274:6397-6404
57. Pelham HR. Speculations on the functions of the major heat shock and glucose-regulated proteins. Cell 1986;46:959-61
58. Tamatani M, Matsuyama T, Yamaguchi A, et al. ORP150 protects against hypoxia/ischemia-induced neuronal death. Nat Med 2001;7:317- 23
59. Kitano H, Nishimura H, Tachibana H, et al. ORP150 ameliorates ischemia/reperfusion injury from middle cerebral artery occlusion in mouse brain. Brain Res 2004;1015:122-28
60. Bando Y, Tsukamoto Y, Katayama T, et al. ORP150/HSP12A protects renal tubular epithelium from ischemia-induced cell death. FASEB J 2004;18:1401-3
61. Nozaki J, Kubota H, Yoshida H, et al. The endoplasmic reticulum stress response is stimulated through the continuous activation of transcription factors ATF6 and XBP1 in Ins2+/Akita pancreatic beta cells. Genes Cells 2004;9:261-70
62. Iwawaki T, Hosoda A, Okuda T, et al. Translational control by the ER transmembrane kinase/ribonuclease IRE1 under ER stress. Nat Cell Biol 2001;3:158-64
63. Park JW, Voss PG, Grabski S, et al. Association of galectin- 1 and galectin-3 with Gemin4 in complexes containing the SMN protein. Nucleic Acids Res 2001;29:3595-3602
64. Patterson RJ, Wang W, Wang JL. Understanding the biochemical activities of galectin-1 and galectin-3 in the nucleus. Glycoconj J 2004;19:499-506
65. Wang W, Park JW, Wang JL, et al. Immunoprecipitation of spliceosomal RNAs by antisera to galectin-1 and galectin-3. Nucleic Acids Res 2006;34:5166-74
66. Ingrassia L, Camby I, Lefranc F, et al. Anti-galectin compounds as potential anti-cancer drugs. Curr Med Chem 2006;13:3513- 27
67. Rabinovich GA, Cumashi A, Bianco GA, et al. Synthetic lactulose amines: Novel class of anticancer agents that induce tumor- cell apoptosis and inhibit galectin-mediated homotypic cell aggregation and endothelial cell morphogenesis. Glycobiology 2006;16:210-20
Marie Le Mercier, MSc, Veronique Mathieu, MD, PhD, Benjamin Haibe- Kains, MSc, Gianluca Bontempi, PhD, Tatjana Mijatovic, PhD, Christine Decaestecker, PhD, Robert Kiss, PhD, and Florence Lefranc, MD, PhD
From the Laboratory of Toxicology (MLM, VM, TM, CD, RK, FL), Institute of Pharmacy, Free University of Brussels (ULB); MicroArray Unit (BHK), Jules Bordet Institute; Machine Learning Group (BHK., GB), Department of Computer Science, ULB; and Department of Neurosurgery (FL), Erasmus University Hospital, ULB, Brussels, Belgium.
Send correspondence and reprint request to: Robert Kiss, PhD, Laboratory of Toxicology, Institute of Pharmacy, Free University of Brussels, Campus de la Plaine CP205/1-Boulevard du Triomphe, 1050 Brussels, Belgium; E-mail: [email protected]
F.L. is a Clinical Research Fellow; C.D. is a Senior Research Associate; and R.K. is Director of Research with the Belgian National Fund for Scientific Research, FNRS, Belgium. V.M. and M.L.M. are holders of a “Grant Televie” from the FNRS.
This study was supported by grants awarded by the Fonds de la Recherche Scientifique Medicale (FRSM, Belgium) and by the Fonds Yvonne Boel (Brussels, Belgium).
Copyright Lippincott Williams & Wilkins May 2008
(c) 2008 Journal of Neuropathology and Experimental Neurology. Provided by ProQuest Information and Learning. All rights Reserved.
