A novel splicing isoform of protein arginine methyltransferase 1 (PRMT1) that lacks the dimerization arm and correlates with cellular malignancy.

Methylation of arginine residues is an important modulator of protein function that is involved in epigenetic gene regulation, DNA damage response and RNA maturation, as well as in cellular signaling. The enzymes that catalyze this post-translational modification are called protein arginine methyltransferases (PRMTs), of which PRMT1 is the predominant enzyme. Human PRMT1 has previously been shown to occur in seven splicing isoforms, which are differentially abundant in different tissues, and have distinct substrate specificity and intracellular localization. Here we characterize a novel splicing isoform which does not affect the amino-terminus of the protein like the seven known isoforms, but rather lacks exons 8 and 9 which encode the dimerization arm of the enzyme that is essential for enzymatic activity. Consequently, the isoform does not form catalytically active oligomers with the other endogenous PRMT1 isoforms. Photobleaching experiments reveal an immobile fraction of the enzyme in the nucleus, in accordance with earlier results from our laboratory that had shown a tight association of inhibited or inactivated PRMT1 with chromatin and the nuclear scaffold. Thus, it apparently is able to bind to the same substrates as catalytically active PRMT1. This isoform is found in a variety of cell lines, but is increased in those of cancer origin or after expression of the EMT-inducing transcriptional repressor Snail1. We discuss that the novel isoform could act as a modulator of PRMT1 activity in cancer cells by acting as a competitive inhibitor that shields substrates from access to active PRMT1 oligomers.

RON receptor tyrosine kinase in human gliomas: expression, function, and identification of a novel soluble splice variant.

Malignant gliomas are incurable because of their diffuse infiltration of the surrounding brain. The recepteur d'origine nantais (RON) receptor tyrosine kinase is highly expressed in several epithelial cancer types and mediates tumorigenic, pro-invasive as well as metastatic effects. Analyzing RON expression in human gliomas, we found that different splice variants with known oncogenic activity are expressed in glioblastomas (GBM). In addition, the RON ligand macrophage-stimulating protein (MSP) is secreted by cultured GBM cells. MSP showed no mitogenic effect on GBM cells but displayed significant chemotactic activity for several GBM cell lines. We identified a novel splice variant, RONDelta90, which is generated by a transcript missing exon 6. As a result of a frameshift, translation is terminated in exon 7, resulting in a truncated soluble protein. RONDelta90 transcripts are expressed in normal human brain as well as in low grade astrocytomas but only in approximately 50% of highly malignant astrocytomas. In addition, RONDelta90 is detectable in supernatants of GBM cell lines. We cloned the RONDelta90 cDNA, and purified the recombinant protein from transfected cells. RONDelta90 inhibited MSP-induced phosphorylation of cellular RON and also attenuated basal activation levels. In addition, RONDelta90 inhibited MSP-induced glioma cell migration as well as random motility. To conclude, RONDelta90 is a novel soluble receptor variant with antagonistic activity that may act as a physiological modulator of RON signaling. The expression of several oncogenic RON splice variants in malignant gliomas suggests that these could represent candidate targets for treatment with agents inhibiting RON activity.

Hypoxia can induce c-Met expression in glioma cells and enhance SF/HGF-induced cell migration.

The c-Met receptor and its ligand scatter factor/hepatocyte growth factor (SF/HGF) are strongly overexpressed in malignant gliomas. Signaling through c-Met as well as exposure to hypoxia can stimulate glioma cell migration and invasion. In several cancer cell types, hypoxia was shown to activate the c-met promoter, which contains hypoxia inducible factor-1 (HIF-1) binding sites. We hypothesized that hypoxia might upregulate c-Met also in glioma cells. Analyzing 18 different glioblastoma cell lines and 10 glioblastoma primary cultures, we found that in 50% of both the cell lines and the primary cultures c-Met protein levels were increased following exposure to hypoxia. Upregulation of c-met in response to hypoxia was also detected at the transcriptional level. In all primary cultures and in 16 of the 18 cell lines (89%), HIF-1 alpha levels were increased by hypoxia. Transfection of siRNA against HIF-1 alpha abgrogated the hypoxic induction of c-Met, suggesting that c-Met expression is upregulated by a HIF-1 alpha-dependent mechanism. Hypoxia sensitized glioblastoma cell lines which showed hypoxic induction of c-Met to the motogenic effects of SF/HGF. These findings suggest that approximately half of all human glioblastomas respond to hypoxia with an induction of c-Met, which can enhance the stimulating effect of SF/HGF on tumor cell migration.

A novel one-armed anti-c-Met antibody inhibits glioblastoma growth in vivo.

PURPOSE: Expression of the receptor tyrosine kinase c-Met and its ligand scatter factor/hepatocyte growth factor (SF/HGF) are strongly increased in glioblastomas, where they promote tumor proliferation, migration, invasion, and angiogenesis. We used a novel one-armed anti-c-Met antibody to inhibit glioblastoma growth in vivo. EXPERIMENTAL DESIGN: U87 glioblastoma cells (c-Met and SF/HGF positive) or G55 glioblastoma cells (c-Met positive and SF/HGF negative) were used to generate intracranial orthotopic xenografts in nude mice. The one-armed 5D5 (OA-5D5) anti-c-Met antibody was infused intratumorally using osmotic minipumps. Following treatment, tumor volumes were measured and tumors were analyzed histologically for extracellular matrix (ECM) components and proteases relevant to tumor invasion. Microarray analyses were done to determine the effect of the antibody on invasion-related genes. RESULTS: U87 tumor growth, strongly driven by SF/HGF, was inhibited > 95% with OA-5D5 treatment. In contrast, G55 tumors, which are not SF/HGF driven, did not respond to OA-5D5, suggesting that the antibody can have efficacy in SF/HGF-activated tumors. In OA-5D5-treated U87 tumors, cell proliferation was reduced > 75%, microvessel density was reduced > 90%, and apoptosis was increased > 60%. Furthermore, OA-5D5 treatment decreased tumor cell density > 2-fold, with a consequent increase in ECM deposition and increased immunoreactivity for laminin, fibronectin, and tenascin. Microarray studies showed no increase in these ECM factors, rather down-regulation of urokinase-type plasminogen activator and matrix metalloproteinase 16 in glioblastoma cells treated with OA-5D5. CONCLUSIONS: Local treatment with OA-5D5 can almost completely inhibit intracerebral glioblastoma growth when SF/HGF is driving tumor growth. The mechanisms of tumor inhibition include antiproliferative, antiangiogenic, and proapoptotic effects.

RNA interference targeting protein tyrosine phosphatase zeta/receptor-type protein tyrosine phosphatase beta suppresses glioblastoma growth in vitro and in vivo.

The protein tyrosine phosphatase zeta/receptor-type protein tyrosine phosphatase beta (PTPzeta/RPTPbeta) and its ligand pleiotrophin (PTN) are overexpressed in human glioblastomas. Both molecules are involved in neuronal cell migration during CNS development. In addition, PTN can induce glioma cell migration which is at least in part mediated through binding to PTPzeta/RPTPbeta. To study the relevance of this ligand-receptor pair for glioma growth in vitro and in vivo, we transfected the human glioblastoma cell line U251-MG with small interfering RNA (siRNA) directed against PTPzeta/RPTPbeta. Stable siRNA transfection resulted in strong down-regulation of PTPzeta/RPTPbeta expression. When injected subcutaneously into nude mice, clones that expressed normal levels of PTPzeta/RPTPbeta (PTPzeta + clones) formed exponentially growing tumours, whereas tumour growth was almost completely abrogated for clones that expressed reduced PTPzeta/RPTPbeta levels (PTPzeta - clones). Similar results were obtained using an orthotopic intracerebral model. Proliferation of PTPzeta - cells in vitro was significantly reduced compared with that of control clones. Matrix-immobilized PTN stimulated the proliferation of PTPzeta + cells but not of PTPzeta - cells. Haptotactic migration induced by PTN was reduced for PTPzeta - clones compared with control clones. Our findings suggest that antagonization of PTPzeta/RPTPbeta expression can inhibit glioma growth in vivo and may thus represent a potentially promising treatment strategy.

Contactin is expressed in human astrocytic gliomas and mediates repulsive effects.

Contactin is a cell surface adhesion molecule that is normally expressed by neurons and oligodendrocytes. Particularly high levels of contactin are present during brain development. Using subtractive cloning, we identified contactin transcripts as overexpressed in glioblastomas compared with normal brain. We confirmed contactin overexpression in glioblastomas at the protein level, and localized contactin to the surface of glial fibrillary acidic protein (GFAP)-expressing glioblastoma cells. In contrast, normal astrocytes did not express contactin. Analyzing different types of astrocytic tumors, we detected an association between increasing malignancy grade and contactin expression. Functionally, contactin had repellent effects on glioma cells in vitro, as demonstrated by adhesion and migration assays. Overexpression of contactin by transfection into glioblastoma cells did not alter the proliferation rate or adhesion to various extracellular matrix proteins as well as adhesion to cells expressing the specific contactin ligand the protein tyrosine phosphatase zeta (PTPzeta). Our findings suggest that contactin has repellent effects on glioma cells to which it is presented as a ligand, but it does not alter the proliferative or adhesive capacities of cells that overexpress the molecule. The repulsive properties of contactin may be a key factor in glioma disaggregation, and may contribute to the diffuse infiltration pattern characteristic of glioma cells in human brain.

Inhibition of glioblastoma angiogenesis and invasion by combined treatments directed against vascular endothelial growth factor receptor-2, epidermal growth factor receptor, and vascular endothelial-cadherin.

PURPOSE: Inhibition of angiogenesis can influence tumor cell invasion and metastasis. We previously showed that blockade of vascular endothelial growth factor receptor-2 (VEGFR-2) with the monoclonal antibody DC101 inhibited intracerebral glioblastoma growth but caused increased tumor cell invasion along the preexistent vasculature. In the present study, we attempted to inhibit glioma cell invasion using a monoclonal antibody against the epidermal growth factor receptor (EGFR), which in the context of human glioblastomas, has been implicated in tumor cell invasion. In addition, we analyzed whether blockade of vascular endothelial (VE)-cadherin as a different antiangiogenic target could also inhibit glioblastoma angiogenesis and growth. EXPERIMENTAL DESIGNS: Nude mice who received intracerebral glioblastoma xenografts were treated using monoclonal antibodies against VEGFR-2 (DC101), EGFR (C225), and VE-cadherin (E4G10) either alone or in different combinations. RESULTS: Increased tumor cell invasion provoked by DC101 monotherapy was inhibited by 50% to 66% by combined treatment with C225 and DC101. C225 inhibited glioblastoma cell migration in vitro, but had no effect on the volume of the main tumor mass or on tumor cell proliferation or apoptosis in vivo, either alone or in combination with DC101. The anti-VE-cadherin monoclonal antibody E4G10 was a weaker inhibitor of tumor angiogenesis and growth than DC101, and also caused a weaker increase in tumor cell invasion. CONCLUSIONS: Inhibition of angiogenesis achieved by blocking either VEGFR-2 or VE-cadherin can cause increased glioma cell invasion in an orthotopic model. Increased tumor cell invasion induced by potent inhibition of angiogenesis with DC101 could be inhibited by simultaneous blockade of EGFR.

Expression and function of the receptor protein tyrosine phosphatase zeta and its ligand pleiotrophin in human astrocytomas.

Using subtractive cloning combined with cDNA array analysis, we previously identified the genes encoding for the protein tyrosine phosphatase zeta/receptor-type protein tyrosine phosphatase beta (PTPzeta/RPTPbeta) and its ligand pleiotrophin (PTN) as overexpressed in human glioblastomas compared to normal brain. Both molecules have been implicated in neuronal migration during central nervous system development, and PTN is known to be involved in tumor growth and angiogenesis. We confirm overexpression of both molecules at the protein level in astrocytic gliomas of different malignancy grades. PTPzeta/RPTPbeta immunoreactivity was associated with increasing malignancy grade and localized predominantly to the tumor cells. PTN immunoreactivity as determined by ELISA and immunohistochemistry analysis was increased in low-grade astrocytomas compared to normal brain. Further increase in malignant gliomas was marginal, and thus no correlation with malignancy grade or microvessel density was present. However, PTN levels were significantly associated with those of fibroblast growth factor-2, suggesting co-regulation of both factors. Functionally, PTN induced weak chemotactic and strong haptotactic migration of glioblastoma and cerebral microvascular endothelial cells. Haptotaxis of glioblastoma cells towards PTN was specifically inhibited by an anti-PTPzeta/RPTPbeta antibody. Our findings suggest that upregulated expression of PTN and PTPzeta/RPTPbeta in human astrocytic tumor cells can create an autocrine loop that is important for glioma cell migration. Although PTN is a secreted growth factor, it appears to exert its mitogenic effects mostly in a matrix-immobilized form, serving as a substrate for migrating tumor cells.

The ubiquitous chromatin protein DEK alters the structure of DNA by introducing positive supercoils.

We have investigated the molecular mechanism by which the proto-oncogene protein DEK, an abundant chromatin-associated protein, changes the topology of DNA in chromatin in vitro. Band-shift assays and electron microscopy revealed that DEK induces both intra- and intermolecular interactions between DNA molecules. Binding of the DEK protein introduces constrained positive supercoils both into protein-free DNA and into DNA in chromatin. The induced change in topology is reversible after removal of the DEK protein. As shown by sedimentation analysis and electron microscopy, the DEK-induced positive supercoiling causes distinct structural changes of DNA and chromatin. The observed direct effects of DEK on chromatin folding help to understand the function that this major chromatin protein performs in the nucleus.