Functional endothelin ETB receptors are selectively expressed in human oligodendrogliomas

Molecular Brain Research 137 (2005) 77 – 88 www.elsevier.com/locate/molbrainres

Research report

Functional endothelin ETB receptors are selectively expressed in human oligodendrogliomas E. Anguelovaa,1, F. Beuvonb,1, N. Leonardb, N. Chaverota, P. Varletb, P.-O. Courauda, C. Daumas-Duportb, S. Cazaubona,T a

Department of Cell Biology, Institut Cochin, INSERM U567, CNRS UMR 8104, IFR116, 22 rue Me´chain, 75014 Paris, France b Hoˆpital Ste Anne, Service d’Anatomie Pathologique, 1 rue Cabanis, 75674 Paris, France Accepted 13 February 2005 Available online 1 April 2005

Abstract Endothelin-1 (ET-1), a vasoactive and mitogenic peptide mainly produced by vascular endothelial cells, may be involved in the progression of several human tumors. Here, we present an immunohistochemical analysis of the expression pattern of ET-1 receptor subtypes (ETA-R and ETB-R) and a functional study of their potential role in human oligodendrogliomas and oligoastrocytomas. By comparison, we assessed the corresponding expression patterns of glioblastomas. Interestingly, a nuclear localization of ET-1 receptor subtypes (associated or not with a cytoplasmic labeling) was constantly observed in tumor cells from all three glioma types. Moreover, we noted a distinct receptor distribution in the different gliomas: a nuclear expression of ETB-R by tumor cells was found to be restricted to oligodendrogliomas and oligoastrocytomas, while a nuclear expression of ETA-R was only detected in tumor cells from some glioblastomas. Using primary cultures of oligodendroglial tumor cells, we confirmed the selective expression of nuclear ETB-R, together with a plasma membrane expression, and further demonstrated that this receptor was functionally coupled to intracellular signaling pathways known to be involved in cell survival and/ or proliferation: extracellular signal-regulated kinase and focal adhesion kinase activation, actin cytoskeleton reorganization. In addition, impairment of ETB-R activation in these cells by in vitro treatment with an ETB-R-specific antagonist induced cell death. These data point to ET-1 as a possible survival factor for oligodendrogliomas via ETB-R activation and suggest that ETB-R-specific antagonists might constitute a potential therapeutic alternative for oligodendrogliomas. D 2005 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Neuro-oncology Keywords: Oligodendroglioma; Oligoastrocytoma; Endothelin-1; ETB receptor; Signaling pathways; Survival factor

1. Introduction Endothelin-1 (ET-1), a 21-amino acid peptide initially isolated as one of the most potent vasoconstrictors secreted by endothelial cells [52], is a member of a small family of three isopeptides (ET-1, ET-2, and ET-3) with pleiotropic activity, including cell proliferation and hormone secretion. T Corresponding author. Fax: +33 1 40 51 64 30. E-mail address: [email protected] (S. Cazaubon). 1 These authors contributed equally to this work. 0169-328X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2005.02.015

In normal brain, the presence of ET-1 and ET-3 has been detected in endothelial cells and neurons [47]. ET-1 may also be synthesized by astrocytes during reactive gliosis, inducing astrocyte proliferation [21], and under pathological conditions such as stroke and Alzheimer’s disease [53]. Hypoxia, which induces the secretion of angiogenic factors, such as vascular endothelial growth factor (VEGF), through transcriptional activation, also activates ET-1 gene transcription and results in increased ET-1 expression [46]. The biological effects of ET-1 in brain include glucose uptake [44], glutamate efflux [40], nerve growth factor

78

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

expression [24], and stimulation of astrocyte proliferation [10]. Indeed, ET-1 is known to be mitogenic and antiapoptotic for many cell types and numerous studies suggest its involvement in a variety of neoplasms [4,32]. This hypothesis is supported by the observation of ET-1 synthesis and secretion by a variety of human cancer cells: ovarian carcinoma [3], colon cancer [1], prostate cancer [31], and melanoma [22]. Moreover, preclinical models are in favor of a contribution of ET-1 in the process of angiogenesis associated with tumor progression. Indeed, ET-1 may modulate, either directly or via VEGF production, various stages of neovascularization including proliferation, migration, and invasion of endothelial cells [32]. ET-1 pleiotropic activity is mediated by binding to specific cell surface receptors: ETA receptor (ETA-R) [2] and ETB receptor (ETB-R) [39], which belong to the Gprotein-coupled, seven transmembrane domain receptor family. Pharmacological studies have shown that ETA-R preferentially binds ET-1, while ETB-R binds all three ETs with similar affinity [5]. In situ hybridization and Northern blot analyses reveal that mRNAs for rat ETA-R and ETB-R exhibit distinct cellular expression patterns both in brain and peripheral tissues. The ETA-R mRNA appears to have a restricted distribution, being predominantly expressed in vascular smooth muscle cells of peripheral tissues, bronchial smooth muscle cells, myocardium, and the pituitary gland. In contrast, the ETB-R mRNA is more widely distributed, with a prominent expression in brain, mostly in glial cells [20]. Whereas the presence of mRNA coding for components of the ET-1 system was detected in some brain tumors including meningiomas and glioblastomas [17,33], the putative expression of ET-1 and/or its receptors in oligodendrogliomas has not yet been reported. According to the World Health Organization (WHO) classification, oligodendrogliomas or anaplastic oligodendrogliomas are primary neoplasms of the central nervous system, which are predominantly composed of cells morphologically resembling oligodendrocytes [23]. These tumors are often estimated to represent less than 10% of all glial tumors [19]; however, the actual frequency of oligodendrogliomas may have been largely underestimated because of lack of specific markers of tumoral oligodendrocytes. Oligoastrocytomas or anaplastic oligoastrocytomas are mixed gliomas which show a substantial proportion of neoplastic oligodendrocytes and astrocytes [23]. The histological features of anaplasia used for the distinction of grade III (anaplastic) from grade II oligogendrogliomas or oligoastrocytomas include nuclear atypia, cellular pleomorphism, high cellularity, and high mitotic activity. In addition, microvascular proliferation and necrosis may be present. We recently observed that oligodendrogliomas and oligoastrocytomas may grow slowly for many years as isolated tumor cells that permeate the brain parenchyma, without microangiogenesis. The emergence of microangiogenesis was shown to be associated with the formation of foci of solid tumor tissue and to represent a crucial event toward more

aggressive behavior [14,15]. In line with this observation, we have proposed a new grading system of oligodendrogliomas and oligoastrocytomas in which endothelial hyperplasia and contrast enhancement, both reflecting microangiogenesis, discriminate two grades of malignancy. The grade A is identified by the absence of endothelial hyperplasia and of contrast enhancement, whereas the grade B corresponds to the presence of at least one of these two characters [14,15]. Accordingly, we have found that expression of the angiogenic factor VEGF is rarely detected in grade A tumors but is constantly observed in grade B tumors [49]. In order to assess the potential involvement of the ET-1 system in the progression of oligodendrogliomas and oligoastrocytomas, we have performed an immunohistochemical analysis of ET-1 receptor subtype expression in these tumors. In parallel, we assessed the corresponding expression patterns of glioblastomas, that are known as highly angiogenic brain tumors. Based on our observations, we then investigated, in primary cultures of oligodendroglial tumor cells, the functional relevance of ET-1 receptor subtype expression. We report here the selective expression in oligodendrogliomas of ETB-R functionally coupled to several intracellular pathways and its possible involvement in cell survival.

2. Materials and methods 2.1. Brain tumors The brain tumor series analyzed in the present study included 40 primary untreated gliomas from patients who had undergone surgery at Ste-Anne Hospital (Paris, France). According to the WHO classification, the series included: 20 oligodendrogliomas (15 grade II and 5 grade III), 10 oligoastrocytomas (7 grade II and 3 grade III), and 10 glioblastomas. According to the Ste-Anne grading system [43], oligodendrogliomas comprised 10 grade A (corresponding to WHO grade II) and 10 grade B (corresponding to WHO grade II or grade III), and oligoastrocytomas comprised 5 grade A (corresponding to WHO grade II) and 5 grade B (corresponding to WHO grade II or grade III). The histological assessment of the selected tumors is presented in Fig. 1. Grade A oligodendrogliomas: in all 10 cases, the isolated tumor cells exhibited typical round nuclei with a well-visible nuclear membrane and conspicuous dots of chromatin. These tumor cells often showed scant or barely discernible cytoplasm, but all samples also contained tumor cells with a clear cytoplasm giving them a typical ‘‘fried egg’’ appearance (Fig. 1A, inset). In all cases, GFAP immunostaining demonstrated the presence of typical reactive astrocytes. GFAP-positive oligodendrocytes (so-called minigemistocytes) were observed in 3 cases (not shown). Grade B oligodendrogliomas: the solid tumor tissue was predominantly made of tumor cells showing a typical ‘‘fried

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

79

Fig. 1. Histology of primary gliomas. Representative hemalun-phloxin staining of selected tumors: (A) Grade A oligodendroglioma composed of isolated tumor cells permeating the white matter, many of them showing a clear perinuclear halo (‘‘fried egg appearance’’). (B) Grade B oligodendroglioma (solid tumor tissue) showing a ‘‘honeycomb’’ pattern and branching capillaries with obvious endothelial hyperplasia; (C and D) Grade A (C) and grade B (D) oligoastrocytomas show a high polymorphism and are composed of GFAP-negative oligodendroglial tumor cells intermigled with GFAP-positive cells (insets), that may exhibit a variable cytological appearance, including ‘‘minigemistocytes’’ with a small eccentric cytoplasm (inset C, arrow), and gemistocytes with a large cytoplasm (inset D). (E1 and E2) Glioblastoma (solid tumor tissue) showing microvascular proliferation (E1 arrows) and foci of necrosis (E2). (Original magnification: A to E: 20; inset A: 60, insets B, C, D: 40.)

egg’’ appearance, together with branched capillary blood vessels showing endothelial hyperplasia (Fig. 1B, inset). Six out of 10 of these tumors contained numerous GFAPpositive minigemistocytes (not shown). Grade A and grade B oligoastrocytomas : these tumors were composed of a variable proportion of isolated tumor cells showing the typical cytological features of tumoral oligodendrocytes and GFAP-positive cells (Figs. 1C and D, inset). The GFAP-positive cells showed variable cytological appearance and included minigemistocytes (Fig. 1C, inset), gemistocytes (Fig. 1D, inset), and reactive astrocytes with typical stellate processes. All of 5 grade A oligoastrocytomas included in the present study exhibited minigemistocytes (Fig. 1C). The microvasculature of the solid tumor tissue of grade B oligoastrocytoma always exhibited endothelial hyperplasia (Fig. 1D). Glioblastomas: in all the 10 cases examined, the solid tumor tissue typically exhibited microvascular proliferation (Fig. 1E1) and foci of necrosis often surrounded by pseudo-

palissading tumor cells (Fig. 1E2). The solid tumor tissue was mostly constituted by poorly differentiated, weakly GFAP-positive or negative cells (not shown). The isolated tumor cells were always GFAP-negative. Although the WHO classification admits the presence of oligodendroglial components in glioblastomas, tumors with an obvious presence of oligodendroglial tumor cells were excluded from our study. As controls, non-tumoral brain samples were obtained from surgical resection for chronic pharmaco-resistant epilepsy (n = 5) and from reactive astrocytic gliosis around hematomas (n = 5) and metastasis (n = 5). 2.2. Reagents Rabbit polyclonal anti-ET-1 antibodies were purchased from Chemicon; rabbit polyclonal anti-ETA-R and anti-ETBR were obtained from Euromedex. Mouse monoclonal antibodies specific to CD11b, N-CAM, and PECAM-1

80

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

(CD31) were from BD Biosciences Pharmingen; mouse monoclonal anti-GFAP antibodies were from Sigma. Mouse monoclonal anti-phospho ERK1/ERK2 (Thr202/Tyr204) antibodies were from New England Biolabs, Inc.; rabbit polyclonal anti-phospho FAK (Y397) antibodies were from Biosource International; rabbit polyclonal anti-ezrin antibodies were kindly provided by Dr. P. Mangeat (University Montpellier II, France). ET-1 antagonists selective for ETBR (BQ 788) or ETA-R (BQ 123) were purchased from Neosystem. 2.3. Histology and immunohistochemistry In each case, 1 to 3 representative samples were selected in order to analyze both the solid tumor tissue and isolated tumor cells in grade B oligodendrogliomas, grade B oligoastrocytomas, and glioblastomas. These samples were initially fixed in formalin zinc (formol 4%, zinc 3 g/l, NaCl 9 g/l) and paraffin-embedded. From the corresponding blocks, 4-Am serial sections were performed and proceeded for hemalun phloxin staining and immunostaining. The 4-Am sections were deparaffinized in toluene and endogenous peroxidase was inactivated in 3% hydrogen peroxide for 5 min. Sections were incubated for 1 h at room temperature with primary antibodies directed to GFAP and ETB-R; the incubation was performed overnight for ETA-R immunostaining (used at 1:200 dilution). Sections were then incubated with biotin-conjugated secondary antibodies and streptavidin – peroxidase complex (Immunotech, Marseille, France) and visualized with diaminobenzidine (Dako, Denmark), counterstained with hematoxylin, dehydrated and mounted. Omission of primary antibodies or neutralization of primary antibodies with the specific peptides used for the immunization, was used as control for the evaluation of non-specific staining. 2.4. Primary cultures of oligodendroglioma cells Surgically removed grade B oligodendrogliomas were collected in serum-free Dulbecco’s modified Eagle’s medium (DMEM) and processed for cell culture within 3 h post-surgery. Tissues were minced and mechanically dissociated through a Pasteur pipette. Dissociated cells were seeded onto poly-l-ornithine (1.5 Ag/ml)-precoated dishes of 100 mm diameter (107 cells/100-mm dish) and grown in DMEM containing 1 g/l glucose, supplemented with 10% fetal calf serum, 1% human serum, and 10 mM HEPES pH 7.4. After 1 week in culture, the medium was changed and the cells were allowed to grow to confluence. Cells were used between passages 3 and 6. 2.5. Immunofluorescence For immunofluorescence analysis, oligodendroglial tumor cells were re-plated on poly-l-ornithine (1.5 Ag/ ml)-precoated thermanox plastic coverslips (Dutscher SA).

Expression of cell markers and membrane receptors was assessed on cells grown in complete medium, while ET-1induced cytoskeletal reorganization and intracellular signaling were performed on cells starved in serum-free medium for 15 h. After washes with PBS, the cells were fixed with paraformaldehyde (4%) in PBS for 10 min, protected with glycine 0.1 M for 10 min and blocked with BSA (2%)/ saponin (0.05%) PBS for 1 h. The cells were incubated for 1 h with the corresponding primary antibody (1/200), or with FITC-conjugated phalloidin for actin labeling (1/200). Antimouse or anti-rabbit antibodies conjugated to Cy3 (1/150) were used as secondary antibodies. Immunofluorescence images were collected in a scanner confocal microscope (MCR.1000, Bio-Rad). 2.6. SDS-PAGE and Western blot analysis Primary cultures of oligodendrogliomas (2.5  106) were maintained in serum-free medium for 48 h before incubation with ET-1 (50 nM) for increasing periods of time. After treatment, cells were rapidly rinsed with ice-cold PBS containing 0.1 mM orthovanadate, scraped in 40 Al SDS sample buffer (125 mM Tris – HCl pH 6.8, 4% sodium dodecylsulfate, 5% glycerol, 50 mM dithiothreitol, 1 mM orthovanadate, 0.05 Ag/ml bromophenol blue) and heated at 95 -C for 10 min. Cell lysates were then analyzed by Western blot as previously described [30, 31] using indicated antibodies at 0.5 to 2 Ag/ml. When indicated, preincubation of the antibody with an excess of the peptide antigen (10 Ag/ml) used for rabbit immunization was performed as a control of antibody specificity. 2.7. Light microscopy of primary cultures of oligodendrogliomas For light microscopy, cells were grown in 6-well dishes for 96 h in DMEM containing 1 g/l glucose, supplemented with 10% fetal calf serum, 1% human serum and 10 mM HEPES pH 7.4. After three washes with serum-free medium, cells were treated with ET-1 antagonists for the indicated periods of time. Cells were then fixed with paraformaldehyde (4%) in PBS for 10 min. Photographs were taken using a LEICA microscope (DM IRB) at 10 magnification. 2.8. Cytotoxicity assay A cytotoxicity detection kit (Roche, Basel, Switzerland), based on the quantification of lactate dehydrogenase (LDH) activity released from the cytosol of damaged cells into the supernatant, was used for the quantification of cellular viability. The culture supernatants from cells treated or not with ET-1 antagonists selective for ETB-R (BQ 788) or ETAR (BQ 123) were collected and incubated with the substrate mixture, according to the manual’s instructions. LDH activity was determined by enzymatic reaction whereby

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

the tetrazolium salt was reduced to formazan, and absorption was measured at 490 nm.

3. Results 3.1. Immunohistological analysis of ET-1 receptor subtype expression in primary gliomas By immunohistological analysis, expression of both ET-1 and its synthesis enzyme endothelin converting enzyme-1 (ECE-1) was detected, as expected, in the tumor cells from all primary gliomas examined in this study (not shown). The expression pattern of the ET-1 receptor subtypes ETA-R and ETB-R was then investigated in oligodendrogliomas (10 grade A and 10 grade B), oligoastrocytomas (5 grade A and 5 grade B), and glioblastomas (10 cases), as well as in nontumoral brain samples (n = 10), as controls.

81

3.2. ETB-R immunoreactivity In control brain parenchyma, only some endothelial cells of the capillaries and few neuronal bodies were weakly ETB-R-positive (not shown). In cases of reactive gliosis associated with either metastasis or abscess, reactive astrocytes showing a strong ETB-R cytoplamic expression were constantly observed (Fig. 2A). The proportion of ETBR-positive reactive astrocytes was however highly variable from one case to another. It should be noted that most of the positive reactive astrocytes seen in the periphery of metastasis exhibited a gemistocytic appearance (Fig. 2A, inset). In oligodendrogliomas and oligoastrocytomas, ETB-R immunoreactivity was detected in tumor cells (Table 1). Numerous oligodendroglial tumor cells with a nuclear pattern of ETB-R expression were observed in most grade A (7/10) and grade B (8/10) oligodendrogliomas (Figs. 2B and D) as well as in grade A (4/5) and grade B (4/5)

Fig. 2. ETB-R immunostaining of primary gliomas. (A) Brain parenchyma around a metastasis composed of reactive astrocytes with gemistocytic appearance that show cytoplasmic ETB-R staining. (B) Grade A oligodendroglioma (cortical infiltration) composed of several tumor cells with nuclear ETB-R staining. (C) Grade A oligoastrocytoma (white matter infiltration) composed of many ETB-R-positive cells: the tumoral oligodendrocytes typically show nuclear ETB-R staining, whereas astrocytic population shows exclusively cytoplasmic ETB-R staining (inset) that is similar to that of reactive astrocytes seen in A. (D) Grade B oligodendroglioma (solid tumor tissue) composed of tumor cells that mostly show nuclear ETB-R staining and ETB-R-positive micro blood vessels. (E) Grade B oligoastrocytoma (solid tumor tissue) composed of ETB-R-positive cells including minigemistocytes that typically exhibit small positive eccentric cytoplasm and negative nucleus (inset). (F) Glioblastoma (solid tumor tissue) showing only rare positive cells that typically exhibit cytoplasmic ETB-R labeling; microvascular proliferation is observed but without ETB-R staining (arrows head). (Original magnification: A and C 10; B 40; D, E, and F 20; inset 60.)

82

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

Table 1 Cellular ETB-R expression in primary gliomas ETB-R expression

Cells in infiltrated tissue or in solid tumor tissue Nuclear Cytoplasmic (nuclear and exclusively cytoplasmic)

Grade A oligodendrogliomas 7/10 (5) Grade B oligodendrogliomas 8/10 (4) Grade A oligoastrocytomas 4/5 (1) Grade B oligoastrocytomas 4/5 (3) Glioblastomas 0/10 (0)

3/10 6/10 5/5 5/5 6/10

Endothelial cells in newly formed micro blood vessels

– 9/10 – 4/5 2/10

For each type of primary tumors, the number of cases with positive cells in infiltrated tissue or in solid tumor tissue is indicated: cases with cells showing a nuclear localization (including those with both nuclear and cytoplasmic staining, number in brackets) and cases with cells showing an exclusive cytoplasmic localization. Reactive astrocytes were not considered in this table. The number of cases with positive endothelial cells in newly formed micro blood vessels is indicated for angiogenic tumors only: glioblastomas, grade B oligodendrogliomas, and grade B oligoastrocytomas.

oligoastrocytomas (Figs. 2C and E). The distribution of positive cells and the intensity of immunostaining, however, were highly variable, both in the infiltrated tissue and the solid tumor tissue. Among the oligodendroglial tumor cells with a positively stained nucleus, few of them also exhibited a cytoplasmic labeling (Table 1: numbers in brackets). The same observation was made in oligoastrocyomas, independently of tumor grade. In contrast, minigemistocytes and astrocytic components present in oligoastrocytomas only exhibited a cytoplasmic ETB-R expression: strongly positive minigemistocytes were detected in some grade A (3 /10), in a majority of grade B (6/10) oligodendrogliomas, and in all cases of oligoastrocytomas (10/10) (Fig. 2E, inset). Oligoastrocytomas (grades A and B) also exhibited numerous strongly ETBR-positive astrocytic cells; although these cells showed a high polymorphism, they often exhibited a gemistocytic appearance (Fig. 2C, inset). In addition, in all cases of oligodendrogliomas and oligoastrocytomas (grades A and B), reactive astrocytes with a typical stellate appearance were found to strongly express ETB-R in their cytoplasm, with no nuclear staining (not shown and excluded from analysis in Table 1). ETB-R-positive reactive astrocytes were easily identifiable within the infiltrated tissue, where they were more abundant in grade B than grade A tumors. Finally, endothelial cells in newly formed micro blood vessels exhibited a positive ETB-R staining (with a nuclear and/or cytoplasmic pattern) in the solid tumor tissue of most grade B oligodendrogliomas (9/10) and oligoastrocytomas (4/5) (Figs. 2D and E). Few endothelial cells from capillaries in the infiltrated parenchyma could also be positive (not shown). In contrast with oligodendrogliomas and oligoastrocytomas, no nuclear ETB-R staining was detected in glioblastomas (Table 1, Fig. 2F). Nevertheless, few ETB-R-positive

cells, with an exclusive cytoplasmic staining, were observed in the solid tumor tissue of 6/10 glioblatomas, with a scattered or perivascular location (Fig. 2F). Their cytoplasmic labeling pattern reminiscent of that of reactive astrocytes (see above), together with a lack of obvious nuclear atypia in most cases, suggest that these ETB-Rpositive cells might be residual astrocytes rather than tumor cells. In all glioblastomas, typical reactive stellate astrocytes with strong cytoplasmic ETB-R expression could be easily identified within the infiltrated tumor tissue (not shown). Also in contrast with oligodendrogliomas and oligoastrocytomas, endothelial cells from newly formed micro blood vessels in glioblastomas were ETB-R negative (Fig. 2E, arrows), a weak labeling being observed very focally in only 2/10 cases. 3.3. ETA-R immunoreactivity In control brain parenchyma, ETA-R immunoreactivity was exclusively localized to neuronal bodies (not shown). In all cases of reactive gliosis associated with either metastasis (5 cases) or abscess, reactive astrocytes were always found negative for ETA-R expression (not shown). Remarkably, no ETA-R immunoreactivity was detected in oligodendrogliomas and oligoastrocytomas (Table 2, Fig. 3B): tumor cells were all negative, as well as endothelial cells from newly formed micro blood vessels and as expected, reactive astrocytes. In contrast, a majority of glioblastomas showed positive ETA-R immunoreactivity (Table 2, Fig. 3A). In 6/10 cases, tumor cells expressing ETA-R were detected in both infiltrated tissue and solid tumor tissue, with a restricted nuclear pattern of expression (Fig. 3A, inset). In glioblastomas, endothelial cells from newly formed micro blood vessels were uniformly ETA-R-negative, as well as reactive astrocytes. These results demonstrate that ETA-R and ETB-R display a differential distribution in primary gliomas, both in the Table 2 Cellular ETA-R expression in primary gliomas ETA-R expression

Grade A oligodendrogliomas Grade B oligodendrogliomas Grade A oligoastrocytomas Grade B oligoastrocytomas Glioblastomas

Cells in infiltrated tissue or in solid tumor tissue Nuclear

Cytoplasmic exclusively

0/10 0/10 0/5 0/5 6/10

0/10 0/10 0/5 0/5 0/10

Endothelial cells in newly formed micro blood vessels

– 0/10 – 0/5 0/10

For each type of primary tumors, the number of cases with positive cells in infiltrated tissue or in solid tumor tissue is indicated: cells showing a nuclear localization and cells showing an exclusive cytoplasmic localization. Reactive astrocytes are not included in this table. Positive endothelial cells in newly formed micro blood vessels are indicated for glioblastomas, grade B oligodendrogliomas, and grade B oligoastrocytomas.

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

83

Fig. 3. ETA-R immunostaining of primary gliomas. (A) Glioblastoma (solid tumor tissue) composed of many ETA-R positive cells that typically show a nuclear staining (inset). (B) Oligodendroglioma grade B (solid tumor tissue) showing no ETA-R positive cells. (Original magnification: A and B 20, inset 40.)

infiltrated tissue and the solid tumor tissue. Whereas a nuclear localization of these receptors is constantly observed in all tumor cell types, ETB-R is exclusively found in tumor cells from oligodendrogliomas and oligoastrocytomas. By contrast, glioblastoma tumor cells either express no detectable ET-1 receptors or only express ETA-R. Together with our observation of ET-1 expression in all gliomas included in the present study (not shown), these data suggest a potential autocrine and/or paracrine action of ET-1 on tumor cells. 3.4. Intracellular signaling coupled to ETB-R in primary cultures of grade B oligodendrogliomas In order to assess the potential functional relevance of the highly selective expression of ETB-R in oligodendrogliomas, primary cultures of grade B oligodendrogliomas were established. They were found free of putative contaminant cells by immunofluorescence analysis for the microglial cell marker CD11b, the neuronal marker NCAM, the astrocytic marker GFAP, and the endothelial marker PECAM-1 (CD31) (data not shown). As expected from our in situ immunohistochemical observations, primary cultures of oligodendroglial tumor cells clearly expressed ETB-R and remained ETA-R-negative. A nuclear expression as well as a diffuse staining of ETB-R characteristic of cell surface expression were observed by indirect immunofluorescence (Fig. 4A). Western blot analysis of whole cell lysates further confirmed the expression of ETB-R by oligodendroglial tumor cells (Fig. 4B) and the specificity of the antibody used in the immunohistological analysis reported above. Cellular response to growth factors is generally associated with the activation of multiple phosphorylation cascades, including extracellular signal-regulated kinase (ERK) and focal adhesion kinase (FAK) pathways, which lead to cytoskeleton rearrangement and ultimately to cell survival and/or proliferation. We established in a previous study that ET-1 induces the proliferation of cultured astrocytes through the activation of both ERK- and FAKdependent pathways, FAK activation being associated with actin-based cytoskeleton rearrangement [10]. To test whether ET-1 might activate these pathways in oligoden-

droglial tumor cells, we performed a Western blot analysis using antibodies specific for the phosphorylated and activated form of these kinases. Treatment of tumor cells with ET-1 (50 nM) increased the phosphorylation level of both ERK isoforms (ERK1 and ERK2), and FAK (Fig. 4C), known to be associated with their activation. These responses were transient, with a maximum at 15 min for ERK phosphorylation and 30 min for FAK phosphorylation. Actin-based cytoskeleton rearrangement was evaluated by confocal microscope analysis using FITCconjugated phalloidin as marker of polymerized actin. Treatment of tumor cells with ET-1 for 30 min induced a marked increase in the formation of stress fibers, including actin cables at the cell periphery, associated with clear modifications of cell morphology (Fig. 4D). Similar results were observed with three distinct primary cultures of grade B oligodendrogliomas (not shown). Altogether, these observations indicate that ETB-R expressed by oligodendroglial tumor cells is functionally coupled to intracellular signaling pathways, generally known to be involved in cell survival and/or proliferation and point to ET-1 as a growth or survival factor for oligodendroglial tumor cells. 3.5. Loss of oligodendroglioma viability in the presence of the selective ETB-R antagonist BQ788 To determine whether ETB-R might indeed regulate cell survival and/or proliferation of oligodendroglial tumor cells, primary cultures (Fig. 5A) were treated with specific ETB-R or ETA-R antagonists, respectively, BQ788 or BQ123. A 3day treatment with 200 AM of BQ788 induced drastic morphological changes associated with the formation of large cytoplasmic vacuoles (Fig. 5B). In contrast, treatment with the specific ETA-R antagonist BQ123 (Fig. 5C) did not induce any detectable change, confirming that the cytotoxic effect of BQ788 was specifically mediated by ETB-R. More pronounced morphological changes, together with a drastic reduction in the number of viable cells, were observed in the presence of a higher concentration of BQ 788 (400 AM) (Fig. 5D) or following a longer treatment (up to 5 days, not shown); still no effect was observed with the same concentration of BQ123 (Fig. 5E). Cytotoxicity was quantified by assessing LDH activity released in cell

84

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

Fig. 4. Intracellular signaling coupled to ETB-R in primary cultures of grade B oligodendrogliomas. (A) Primary cultures of grade B oligodendrogliomas were labeled with anti-ETB-R antibodies (1/200) and visualized with confocal microscope. (B) Whole-cell lysates (20 Ag of protein/lane) from primary cultures of grade B oligodendroglioma were submitted to Western blot analysis using anti-ETB-R antibodies (1/200) without (lane 1) or with preincubation (lane 2) with the peptide antigen (10 Ag/ml), as a control of antibody specificity. (C) Primary cultures of grade B oligodendrogliomas were starved in serum-free medium for 48 h and then treated with ET-1 (50 nM) for increasing periods of time (0, 15, 30, 60 min). Whole-cell lysates were analyzed by Western blot using antibodies specific to the phosphorylated form of ERK (ERK1 and ERK2) (1 Ag/ml). The same blot was then incubated with antibodies specific to the phosphorylated form of FAK (1 Ag/ml), and with antibodies specific to ezrin, an actin-associated protein (1/2000) for internal control of the quantity of protein loaded in each lane. (D) Primary cultures of grade B oligodendrogliomas were starved in serum-free medium for 48 h and then treated with ET-1 (50 nM) for 30 min. Fixed cells were labeled with FITC-conjugated phalloidin and cytoskeletal reorganization was submitted to confocal microscope analysis. The results are representative of three independent experiments.

supernatants in the same treatment conditions. Consistent with the morphological changes, increased LDH activity was exclusively observed in the presence of BQ 788 (Fig. 6). The cytotoxic effect of BQ788 was dose- and timedependent: significant cell death was detected after a 3-day treatment with 200 AM BQ788, whereas 400 AM induced 90 –100% cell death (Fig. 6). In the same conditions, BQ123 had no effect on cell viability. Similar results were observed with three distinct primary cultures of grade B oligodendrogliomas (not shown). These results demonstrate that a selective ETB-R antagonist can impair survival of oligodendroglial tumor cells in vitro and suggest that ET-1 may behave in situ as an autocrine or paracrine survival factor for these tumor cells.

4. Discussion The most important observations of the present study are the differential expression of ET-1 receptor subtypes (ETA-R and ETB-R) in primary human gliomas together with their nuclear localization in tumor cells. Indeed, ETB-R was expressed, independently of tumor grade, in the nucleus of tumor cells of most oligodendrogliomas and oligoastrocytomas, but not of glioblastomas, whereas ETA-R, when detected, was found exclusively in the nucleus of glioblastoma tumor cells. Using primary cultures of grade B oligodendrogliomas, we have evidenced the functional coupling of ETB-R to intracellular signaling pathways (ERK and FAK activation, stress fiber formation) in these

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

85

Fig. 5. Effect of ETB-R antagonist to the morphology of primary cultures of grade B oligodendrogliomas. Oligodendroglial tumor cells were cultured for 3 days without antagonist (A), with antagonist selective for ETB-R: BQ788 at 200 AM (B) or at 400 AM (D), with antagonist selective for ETA-R: BQ123 at 200 AM (C) or 400 AM (E). Pictures of cells were taken with bright-field optics at 10 magnification. The results are representative of five independent experiments.

Fig. 6. Effect of ETB-R antagonist to the viability of primary cultures of grade B oligodendrogliomas. Cell death was quantified by measuring LDH release from damaged cells after 3 days (open bars) and 5 days (black bars) in culture, in the absence (Control) or in the presence of ETB-R antagonist (BQ788 at 200 AM, 400 AM) or ETA-R antagonist (BQ123 at 200 AM, 400 AM). The results are representative of three independent experiments.

tumor cells. Interestingly, our in vitro study with ET-1 receptor subtype specific antagonists demonstrates that ET-1 binding to ETB-R mediates the survival of cultured oligodendroglial tumor cells. The expression of the endothelins ET-1 and ET-3 and their receptors was extensively described in human and rat brains. Endothelin genes and proteins have been identified in various cerebral areas: cortex, hypothalamus, hippocampus, substantia nigra, and amygdala. At the cellular level, ETs are present in neurons, astrocytes, and vascular endothelial cells, whereas ETA-R is expressed by neurons and ETB-R by astrocytes and endothelial cells [47]. Our observations with non-neoplastic brain tissue confirmed this pattern of expression of ET-1 system components. Synthesis of ET-1 is also known to occur in several human tumors including glioblastomas [17,32]. Our observations extend these data to other types of gliomas (oligoastrocytomas and oligodendrogliomas) and document the expression and function of ET-1 receptor subtypes in these tumors.

86

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

Our observation that glioblastoma tumor cells express ETA-R in 6/10 cases contrasts with a recent publication reporting ETB-R expression in these tumors [17]. The reason for the discrepancy between that report and the present study might be due to the fact that glioblastomas with oligodendroglial components were excluded from our study, as well as to the poor reproducibility of the WHO histological classification of gliomas [9,12,30]. Also, it has to be considered that the cellular resolution of in situ hybridization as used in that study [17] is much lower than that of the immunohistological analysis presented herein. Although, in our study, we could detect few ETB-R-positive cells within the solid tumor tissue of glioblastomas, the lack of obvious nuclear atypia and of nuclear labeling was suggestive of residual or reactive astrocytes rather than tumor cells. It should be noted that we also constantly observed typical reactive astrocytes with a strong cytoplasmic labeling for ETB-R, in all types of gliomas, in line with the well-documented expression of ETB-R by astrocytes in normal and injured brain [26,38]. Finally, we also detected ETB-R expression in endothelial cells from newly formed micro blood vessels present in grade B oligodendrogliomas (9/10) and grade B oligoastrocytomas (4/5). In contrast, ETB-R expression was rarely detected in glioblastoma newly formed micro blood vessels (2/10), possibly due to the aberrant neoangiogenesis and the abnormal phenotype of vascular endothelium reported in these tumors [37,51]. These observations suggest that ET-1 system may not be directly involved in neoangiogenesis associated with glioblastoma tumor progression, in contrast to VEGF, whose expression correlates with tumor progression [49]. However, such results do not rule out a putative participation of the ET-1 system in neoangiogenesis associated with grade B oligodendrogliomas and oligoastrocytomas in line with other observations [32]. We report in the present study the expression of ETB-R (but not ETA-R) by tumor cells of most oligodendrogliomas and oligoastrocytomas, independently of WHO or Ste-Anne tumor grade (Table 1). Indeed, ETB-R expression was observed in oligodendroglial tumor cells within the infiltrated tissue (grades A and B) and the solid tumor tissue (grade B), always showing a nuclear localization (associated or not with a cytoplasmic labeling). Astrocytic components present in oligodendrogliomas and oligoastrocytomas showed a cytoplasmic ETB-R labeling, this pattern of expression together with the cell morphology being characteristic of reactive astrocytes. Similar to astrocytes, minigemistocytic cells were found to express ETB-R exclusively in the cytoplasm, clearly differing from oligodendroglial tumor cells. In contrast to oligodendrogliomas and oligoastrocytomas, no cells with a nuclear ETB-R staining were detected in glioblastomas that instead expressed ETA-R (6/10 cases), always with a nuclear localization. Several studies have reported a nuclear localization for other transmembrane receptors, including growth factor receptors: EGF, NGF, and FGF receptors

[28,29,35,36], and G-protein-coupled, seven transmembrane domain receptors: the chemokine receptor CXCR4, the angiotensin II AT1, bradykinin B2, and apelin receptors [8,27,42]. On the basis of these observations, novel roles for these receptors in cell signaling and function have been suggested. In line with this hypothesis, heterotrimeric G proteins [50] and various components of signaling pathways such as phosphoinisitols, 1,2-diacylglycerol, phospholipase C, and ERK were also detected in the cell nucleus [13,18,34,48]. The presence of protein kinase activity in the nucleus was further illustrated by studies showing the phosphorylation of nuclear proteins in response to growth factors including EGF and ET-1 [6,7]. In addition, an unanticipated role for the EGF receptor in the nucleus as a transcription factor was proposed, based on the observation that this receptor may interact with the promoter of cyclin D1, an actor of cell cycle progression [28]. In conclusion, in line with the observations reported here, the concept of the nuclear expression of functional plasma membrane receptors is now emerging, with their possible involvement in cell proliferation [28,35]. To our knowledge, the putative role of ET-1 and ETB-R in oligodendroglioma progression was not documented so far. Using primary cultures of grade B oligodendroglial tumor cells, we have here assessed the functionality of ETBR expression in these tumors. Confocal microscopy analysis confirmed the strong nuclear expression of ETB-R in these cells, together with a diffuse staining characteristic of plasma membrane expression. In vitro treatment with ET-1 indicated that oligodendroglial tumor cells express functional ETB-R coupled to multiple intracellular signaling pathways known to be involved in cell survival and/or proliferation: ERK and FAK activation, and stress fiber formation. Indeed, in various cell types, activation of the ERK pathway plays a central role in these responses [41], together in adherent cells, with activation of FAK, a protein kinase localized at the contacts between the cell and the extracellular matrix [11]. Activation of FAK is associated with actin cytoskeleton rearrangement, leading to stress fiber formation. We have previously demonstrated in astrocytes that ERK and FAK pathways are converging to the induction of cyclin D1 and D3 expression, both pathways therefore contributing to cell cycle progression [10,45]. We now show that ETB-R in oligodendrogliomas is involved in cell survival, its inhibition by the specific ETB-R antagonist BQ788 dramatically inducing cell death. Moreover, this cytotoxic effect of BQ788 was also observed with primary cultures of two distinct grade B oligoastrocytomas (not shown), further supporting the similar clinical behavior of oligodendrogliomas and oligoastrocytomas [16]. Altogether, these in vitro data strongly suggest that ET-1, which is locally produced in vivo by tumor cells, brain vascular endothelial cells, and reactive astrocytes, may be a survival factor for oligodendrogliomas and oligoastrocytomas. This conclusion is supported by a recent study reporting both in vitro and in vivo cytotoxic activity of an ETB-R antagonist

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

on melanomas [25], as well as by the initiation of clinical trials in various cancers based on treatment with ET-1 receptor antagonists [32]. In conclusion, the present study demonstrates that ET-1 receptor subtypes are differentially expressed in human primary gliomas: we observed a nuclear expression of ETB-R selectively in tumor cells from oligodendrogliomas or oligoastrocytomas, while ETA-R was only detected in some glioblastomas. Moreover, it is tempting to speculate from our in vitro observations, that ETB-R antagonists might be considered as potential therapeutic agents in oligodendrogliomas.

Acknowledgments This work was supported by the Centre National de la Recherche Scientifique, the Association pour la Recherche sur le Cancer (ARC), the Ligue Nationale Franc¸aise contre le Cancer, the Institut National de la Sante´ et de la Recherche Me´dicale, the Universite´ Rene´ Descartes-Paris5, and the Ministe`re de la Jeunesse, de l’Education nationale et de la Recherche.

References [1] H. Ali, M. Loizidou, M. Dashwood, F. Savage, C. Sheard, I. Taylor, Stimulation of colorectal cancer cell line growth by ET-1 and its inhibition by ET(A) antagonists, Gut 47 (2000) 685 – 688. [2] H. Arai, S. Hori, I. Aramori, H. Ohkubo, S. Nakanishi, Cloning and expression of a cDNA encoding an endothelin receptor, Nature 348 (1990) 730 – 735. [3] A. Bagnato, D. Salani, V. Di Castro, J.R. Wu-Wong, R. Tecce, M.R. Nicotra, A. Venuti, P.G. Natali, Expression of endothelin 1 and endothelin A receptor in ovarian carcinoma: evidence for an autocrine role in tumor growth, Cancer Res. 59 (1999) 720 – 727. [4] B. Battistini, P. Chailler, P. D’Orleans-Juste, N. Briere, P. Sirois, Growth regulatory properties of endothelins, Peptides 14 (1993) 385 – 399. [5] W.A. Bax, P.R. Saxena, The current endothelin receptor classification: time for reconsideration? Trends Pharmacol. Sci. 15 (1994) 379 – 386. [6] L. Bian, Z. Lei, C.V. Rao, Mitogen-activated protein kinase is involved in epidermal-growth-factor-regulated protein phosphorylation in nuclear membranes isolated from JEG-3 human choriocarcinoma cells, Eur. J. Biochem. 253 (1998) 545 – 551. [7] B. Boivin, D. Chevalier, L.R. Villeneuve, E. Rousseau, B.G. Allen, Functional endothelin receptors are present on nuclei in cardiac ventricular myocytes, J. Biol. Chem. 278 (2003) 29153 – 29163 (electronic publication 2003 May 19). [8] G.W. Booz, K.M. Conrad, A.L. Hess, H.A. Singer, K.M. Baker, Angiotensin-II-binding sites on hepatocyte nuclei, Endocrinology 130 (1992) 3641 – 3649. [9] J.M. Bruner, L. Inouye, G.N. Fuller, L.A. Langford, Diagnostic discrepancies and their clinical impact in a neuropathology referral practice, Cancer 79 (1997) 796 – 803. [10] S. Cazaubon, N. Chaverot, I.A. Romero, J.A. Girault, P. Adamson, A.D. Strosberg, P.O. Couraud, Growth factor activity of endothelin-1 in primary astrocytes mediated by adhesion-dependent and -independent pathways, J. Neurosci. 17 (1997) 6203 – 6212. [11] E.A. Clark, J.S. Brugge, Integrins and signal transduction pathways: the road taken, Science 268 (1995) 233 – 239.

87

[12] S.W. Coons, P.C. Johnson, B.W. Scheithauer, A.J. Yates, D.K. Pearl, Improving diagnostic accuracy and interobserver concordance in the classification and grading of primary gliomas, Cancer 79 (1997) 1381 – 1393. [13] C.S. D’Santos, J.H. Clarke, R.F. Irvine, N. Divecha, Nuclei contain two differentially regulated pools of diacylglycerol, Curr. Biol. 9 (1999) 437 – 440. [14] C. Daumas-Duport, M.L. Tucker, H. Kolles, P. Cervera, F. Beuvon, P. Varlet, N. Udo, M. Koziak, J.P. Chodkiewicz, Oligodendrogliomas. Part II: a new grading system based on morphological and imaging criteria, J. Neuro-Oncol. 34 (1997) 61 – 78. [15] C. Daumas-Duport, P. Varlet, M.L. Tucker, F. Beuvon, P. Cervera, J.P. Chodkiewicz, Oligodendrogliomas. Part I: patterns of growth, histological diagnosis, clinical and imaging correlations: a study of 153 cases, J. Neuro-Oncol. 34 (1997) 37 – 59. [16] C. Daumas-Duport, F. Beuvon, P. Varlet, C. Fallet-Bianco, [Gliomas: WHO and Sainte-Anne Hospital classifications], Ann. Pathol. 20 (2000) 413 – 428. [17] G. Egidy, L.P. Eberl, O. Valdenaire, M. Irmler, R. Majdi, A.C. Diserens, A. Fontana, R.C. Janzer, F. Pinet, L. Juillerat-Jeanneret, The endothelin system in human glioblastoma, Lab. Invest. 80 (2000) 1681 – 1689. [18] I. Faenza, A. Matteucci, L. Manzoli, A.M. Billi, M. Aluigi, D. Peruzzi, M. Vitale, S. Castorina, P.G. Suh, L. Cocco, A role for nuclear phospholipase Cbeta 1 in cell cycle control, J. Biol. Chem. 275 (2000) 30520 – 30524. [19] A. Helseth, S.J. Mork, Neoplasms of the central nervous system in Norway: III. Epidemiological characteristics of intracranial gliomas according to histology, APMIS 97 (1989) 547 – 555. [20] S. Hori, Y. Komatsu, R. Shigemoto, N. Mizuno, S. Nakanishi, Distinct tissue distribution and cellular localization of two messenger ribonucleic acids encoding different subtypes of rat endothelin receptors, Endocrinology 130 (1992) 1885 – 1895. [21] M.H. Jiang, A. Hoog, K.C. Ma, X.J. Nie, Y. Olsson, W.W. Zhang, Endothelin-1-like immunoreactivity is expressed in human reactive astrocytes, NeuroReport 4 (1993) 935 – 937. [22] K. Kikuchi, H. Nakagawa, T. Kadono, T. Etoh, H.R. Byers, M.C. Mihm, K. Tamaki, Decreased ET(B) receptor expression in human metastatic melanoma cells, Biochem. Biophys. Res. Commun. 219 (1996) 734 – 739. [23] P. Kleihues, D.N. Louis, B.W. Scheithauer, L.B. Rorke, G. Reifenberger, P.C. Burger, W.K. Cavenee, The WHO classification of tumors of the nervous system, J. Neuropathol. Exp. Neurol. 61 (2002) 215 – 225 (discussion 226 – 9). [24] R.G. Ladenheim, I. Lacroix, N. Foignant-Chaverot, A.D. Strosberg, P.O. Couraud, Endothelins stimulate c-fos and nerve growth factor expression in astrocytes and astrocytoma, J. Neurochem. 60 (1993) 260 – 266. [25] R. Lahav, G. Heffner, P.H. Patterson, An endothelin receptor B antagonist inhibits growth and induces cell death in human melanoma cells in vitro and in vivo, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 11496 – 11500. [26] F. Lazarini, A.D. Strosberg, P.O. Couraud, S.M. Cazaubon, Coupling of ETB endothelin receptor to mitogen-activated protein kinase stimulation and DNA synthesis in primary cultures of rat astrocytes, J. Neurochem. 66 (1996) 459 – 465. [27] D.K. Lee, A.J. Lanca, R. Cheng, T. Nguyen, X.D. Ji, F. Gobeil Jr., S. Chemtob, S.R. George, B.F. O’Dowd, Agonist-independent nuclear localization of the Apelin, angiotensin AT1, and bradykinin B2 receptors, J. Biol. Chem. 279 (2004) 7901 – 7908 (electronic publication 2003 Nov 26). [28] S.Y. Lin, K. Makino, W. Xia, A. Matin, Y. Wen, K.Y. Kwong, L. Bourguignon, M.C. Hung, Nuclear localization of EGF receptor and its potential new role as a transcription factor, Nat. Cell Biol. 3 (2001) 802 – 808. [29] P.A. Maher, Nuclear translocation of fibroblast growth factor (FGF) receptors in response to FGF-2, J. Cell Biol. 134 (1996) 529 – 536.

88

E. Anguelova et al. / Molecular Brain Research 137 (2005) 77 – 88

[30] M.A. Mittler, B.C. Walters, E.G. Stopa, Observer reliability in histological grading of astrocytoma stereotactic biopsies, J. Neurosurg. 85 (1996) 1091 – 1094. [31] J.B. Nelson, K. Chan-Tack, S.P. Hedican, S.R. Magnuson, T.J. Opgenorth, G.S. Bova, J.W. Simons, Endothelin-1 production and decreased endothelin B receptor expression in advanced prostate cancer, Cancer Res. 56 (1996) 663 – 668. [32] J. Nelson, A. Bagnato, B. Battistini, P. Nisen, The endothelin axis: emerging role in cancer, Nat. Rev., Cancer 3 (2003) 110 – 116. [33] U. Pagotto, T. Arzberger, U. Hopfner, A. Weindl, G.K. Stalla, Cellular localization of endothelin receptor mRNAs (ETA and ETB) in brain tumors and normal human brain, J. Cardiovasc. Pharmacol. 26 (Suppl. 3) (1995) S104 – S106. [34] J. Pouyssegur, P. Lenormand, Fidelity and spatio-temporal control in MAP kinase (ERKs) signalling, Eur. J. Biochem. 270 (2003) 3291 – 3299. [35] E.M. Rakowicz-Szulczynska, U. Rodeck, M. Herlyn, H. Koprowski, Chromatin binding of epidermal growth factor, nerve growth factor, and platelet-derived growth factor in cells bearing the appropriate surface receptors, Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 3728 – 3732. [36] E.M. Rakowicz-Szulczynska, M. Herlyn, H. Koprowski, Nerve growth factor receptors in chromatin of melanoma cells, proliferating melanocytes, and colorectal carcinoma cells in vitro, Cancer Res. 48 (1988) 7200 – 7206. [37] G. Rascher, A. Fischmann, S. Kroger, F. Duffner, E.H. Grote, H. Wolburg, Extracellular matrix and the blood – brain barrier in glioblastoma multiforme: spatial segregation of tenascin and agrin, Acta Neuropathol. (Berl.) 104 (2002) 85 – 91. [38] S.D. Rogers, C.M. Peters, J.D. Pomonis, H. Hagiwara, J.R. Ghilardi, P.W. Mantyh, Endothelin B receptors are expressed by astrocytes and regulate astrocyte hypertrophy in the normal and injured CNS, Glia 41 (2003) 180 – 190. [39] T. Sakurai, M. Yanagisawa, Y. Takuwa, H. Miyazaki, S. Kimura, K. Goto, T. Masaki, Cloning of a cDNA encoding a non-isopeptideselective subtype of the endothelin receptor, Nature 348 (1990) 732 – 735. [40] Y. Sasaki, M. Takimoto, K. Oda, T. Fruh, M. Takai, T. Okada, S. Hori, Endothelin evokes efflux of glutamate in cultures of rat astrocytes, J. Neurochem. 68 (1997) 2194 – 2200. [41] R. Seger, E.G. Krebs, The MAPK signaling cascade, FASEB J. 9 (1995) 726 – 735. [42] K. Shibuta, M. Mori, K. Shimoda, H. Inoue, P. Mitra, G.F. Barnard,

[43]

[44]

[45]

[46]

[47] [48]

[49]

[50]

[51]

[52]

[53]

Regional expression of CXCL12/CXCR4 in liver and hepatocellular carcinoma and cell-cycle variation during in vitro differentiation, Jpn. J. Cancer Res. 93 (2002) 789 – 797. D.B. Stanimirovic, T. Yamamoto, S. Uematsu, M. Spatz, Endothelin-1 receptor binding and cellular signal transduction in cultured human brain endothelial cells, J. Neurochem. 62 (1994) 592 – 601. A. Tabernero, C. Giaume, J.M. Medina, Endothelin-1 regulates glucose utilization in cultured astrocytes by controlling intercellular communication through gap junctions, Glia 16 (1996) 187 – 195. A. Teixeira, N. Chaverot, A.D. Strosberg, S. Cazaubon, Differential regulation of cyclin D1 and D3 expression in the control of astrocyte proliferation induced by endothelin-1, J. Neurochem. 74 (2000) 1034 – 1040. M.C. Tsang, A.C. Lo, P.T. Cheung, S.S. Chung, S.K. Chung, Perinatal hypoxia-/ischemia-induced endothelin-1 mRNA in astrocyte-like and endothelial cells, NeuroReport 12 (2001) 2265 – 2270. M. van den Buuse, K.M. Webber, Endothelin and dopamine release, Prog. Neurobiol. 60 (2000) 385 – 405. L.R. Vann, F.B. Wooding, R.F. Irvine, N. Divecha, Metabolism and possible compartmentalization of inositol lipids in isolated rat-liver nuclei, Biochem. J. 327 (1997) 569 – 576. P. Varlet, J.S. Guillamo, F. Nataf, M. Koziak, F. Beuvon, C. DaumasDuport, Vascular endothelial growth factor expression in oligodendrogliomas: a correlative study with Sainte-Anne malignancy grade, growth fraction and patient survival, Neuropathol. Appl. Neurobiol. 26 (2000) 379 – 389. F.S. Willard, M.F. Crouch, Nuclear and cytoskeletal translocation and localization of heterotrimeric G-proteins, Immunol. Cell Biol. 78 (2000) 387 – 394. H. Wolburg, K. Wolburg-Buchholz, J. Kraus, G. Rascher-Eggstein, S. Liebner, S. Hamm, F. Duffner, E.H. Grote, W. Risau, B. Engelhardt, Localization of claudin-3 in tight junctions of the blood – brain barrier is selectively lost during experimental autoimmune encephalomyelitis and human glioblastoma multiforme, Acta Neuropathol. (Berl.) 105 (2003) 586 – 592. M. Yanagisawa, H. Kurihara, S. Kimura, Y. Tomobe, M. Kobayashi, Y. Mitsui, Y. Yazaki, K. Goto, T. Masaki, A novel potent vasoconstrictor peptide produced by endothelial cells, Nature 332 (1988) 411 – 415. M. Zhang, Y. Olsson, Reactions of astrocytes and microglial cells around hematogenous metastases of the human brain. Expression of endothelin-like immunoreactivity in reactive astrocytes and activation of microglial cells, J. Neurol. Sci. 134 (1995) 26 – 32.