MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

NEURO-ONCOLOGY

0889-8588/01 $15.00

+ .OO

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS Sandra A. Rempel, PhD

A disruption in the balance of expression of the positive and negative effectors that control the cell cycle, neovessel formation, and cell motility can give rise to a tumor and promote malignant progression. The disruption may be simple, resulting from changes in the level of expression of a normal gene, or may be more complex, involving the loss of the gene or mutation of the gene product. Great strides have been made in characterizing altered gene expression and acquired genetic mutations in central nervous system (CNS) tumors; however, the ability of a tumor cell to divide, initiate new blood vessel formation, and migrate into adjacent brain tissue may depend not only on the intrinsic changes within the tumor itself, but also may rely on the extrinsic environment in which the tumor arises. This article focuses on the known or implicated genetic alterations, in light of the tumorextracellular matrix (ECM) interactions, that promote tumor proliferation, angiogenesis, and invasion. PROLIFERATION The signal for a normal cell to divide often is initiated at the cell surface, where growth factors, released from the surrounding ECM or secreted by the tumor itself, interact with their cognate plasma membrane-associated receptors (Fig. 1). This interaction triggers a series of cytoplasmic signaling events that ultimately affects gene expression in

From the Barbara Jane Levy Laboratory of Molecular Neuro-Oncology, Hermelin Brain Tumor Center, Department of Neurosurgery, Henry Ford Hospital, Detroit, Michigan

HEMATOLOGY/ONCOLOGY CLINICS OF NORTH AMERICA

-

VOLUME 15 * NUMBER 6 DECEMBER 2001

979

980

REMPEL

Figure 1. Biological interactions involved in the acquisition of tumor phenotypes. The ability of a tumor cell to divide, initiate new blood vessel formation, and migrate into adjacent brain tissue may depend not only on the intrinsic changes within the tumor itself, but may also rely on the extrinsic environment in which the tumor arises (1) Cell-extracellular matrix (ECM) interactions. Receptors at the cell surface act as sensors to relay information concerning the external environment by the binding of growth factors, extracellular matrix proteins, or secreted proteins. (2) Cytoskeletal changes. Proteolysis of ECM proteins or the binding of secreted proteins alter integrin-mediated anchorage, focal adhesions, and cytoskeletal architecture to alter adhesion, migration, and proliferation. (3) Signaling. Changes in cytoskeletal structure and growth factor-induced receptor dimerization induce cytoplasmic signaling pathways that transduce the information received at the cell surface into the nucleus. (4) Changes in gene expression. The transduced signals induce expression of genes necessaty to promote tumor cell proliferation, angiogenesis, and invasion. Genetic mutations or alterations in gene expression at any level can promote tumorigenesis. (Data from Rempel SA: Molecular biology of central nervous system tumors. Curr Opin Oncol 10:179-185, 1998; Shibuya M, Ito N, Claesson-Welch: Structure and function of vascular endothelial growth factor receptor-1 and -2. Curr Top Microbiol lmmunol 2375983, 1999.)

the nucleus to promote cell division. Alterations and mutations in gene expression have been described for molecular participants in each of these biologic pathways that together or independently give rise to tumor cell proliferation. Growth Factors and Their Receptors

One of the earliest events in CNS tumorigenesis is an increased expression of growth factors or their receptors or both (Table 1).While

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

981

Table 1. POSITIVE AND NEGATIVE EFFECTOR GENES IMPLICATED IN CENTRAL NERVOUS SYSTEM TUMOR PROGRESSION, CELL PROLIFERATION, ANGIOGENESIS, AND TUMOR CELL INVASION Human Nervous System Neoplasia

Negative Effectors/TumorSuppressor Genes

Gliomas Diffuse fibrillary astrocytoma Progression (secondary) pathway Astrocytoma 17p (p53), 22q (not NF2) 3p (VHL), 16p (TSC2) P27 SPARC

Positive Effectors/ Oncogenes

PDGFs/PDGF-Rs, IGFs/IGFR, IGFBP2, EGF, a/bFGF/ a/bFGFRs, TGF-a/P, TNFa,Scatter factor/c-met, ROSl MMP-2, MMP-9, stromelysin, matrilysin, cathepsin B, uPA/uPAR SPARC, Tenascin VEGF/R, IL13-R JAKI and 2, Stats 1, 3, 5, 6 Anaplastic 9p (pZ6 and pZ5), 13q (Rb), CDK4, cyclin D1 and D3 astrocytoma lp, 19q, 17~13.1(not p53), PCNA, granulin ~ 1 4 ~ ~ Glioblastoma lop, 10q23 (MACCZPTEN) lq32 multiforme 1Oq24-25 (MXIZ?, LGII?) CDK2, CDK6, (MDM4?) Ras 1Oq25.3-26.1(DMBTZ) 18q p57, E2F thrombospondin De novo (primary) pathway 10 7 (EGFR), 9p (PAX5), 12 Glioblastoma (MDM2) Scatter factor/c-met Oligodendroglioma 7 (EGFR) Low grade lp, 19q (DZ9S412-STD), 4, 14, 15, 18 Anaplastic 9p, 10(10q25-26), Y 8, 11, X, Y, 22, 17 p16, PIS Epend ymoma 22q (NF2), 16p (TSC2), 6q lq, 9 Low grade lOq, 13q, 17, X PDGFs/PDGFRs, SV40, IGF I1 MDM2 ShcA VEGF, Scatter factor/c-met JAK1 and 2, Stats 1, 3, 5, 6 Anaplastic P53 Medullobalstoma (PNET) 17p (not p53), 6q, 16q, 9q22 7, 17q, 1, 4, 6, IGF I & 11, IGFRI, MYC-N & MYC-C (PTCH), lOq25.3-26.1 (DMBTZ) VEGF JAK1 and 2, Stats 1, 3, 5, 6 APC Table continued on following page

982

REMPEL

Table 1. POSITIVE AND NEGATIVE EFFECTOR GENES IMPLICATED IN CENTRAL NERVOUS SYSTEM TUMOR PROGRESSION, CELL PROLIFERATION, ANGIOGENESIS, AND TUMOR CELL INVASION (Continued) Human Nervous System Neoplasia

Meningioma Benign

Atypical

Malignant Neurofibroma Benign Malignant (MPNST) Schwannoma Sporadic NF2-associated Neuroblastoma

Negative EffectorwTumorSuppressor Genes

Positive Effectors1 Oncogenes

22q (NF2; subtype specificity) 18 (DAL-I)

EGF/c-erbB2, bFGF/bFGFR, IGFs/IGFRs, IGFBP1-3, PDGF-BB/PDGFR, TGF-cx/ p, ILl3-R, ROS1, c-myc, cfos, RAs* MMP-9 SPARC VEGF/VEGFR JAK2, Stats 1-6 lp36 (ALPL), 1 ~ 3 4 - 1 ~ 3 2 , iq, 9q, 12q, 15q,17q, 20 1 ~ 2 21~21.1-lp13 , 2p, 6q, 10, 14q24.3-31, 14q32.132.2, 18q P53 Amp 17q 9P9 ( P W

17q11.2 (NFI), 17p (p53) 13q14-qZ1, 9p, 3, 18p p16, p27

17q24-qter (not NFI) Ras, cyclin E

22q (NF2), lp? 22q (NF2)

bFGF, IGFs, CD44

llq, 14q, 17p and q, lp36 (p73), 3p, activin A

N-myc, NM23H1, VEGF, bFGF, IGFs, PDGFA, TGF-a

Note. When known or suspected (?), the tumor-suppressor loci or genes corresponding to the chromosomal loci are indicated in parentheses. Genes are indicated when they first appear mutated or altered in expression in tumor progression. *Neither mutations nor altered expression has been reported; however, inhibition of expression indicates that it plays a crucial role in signal transduction. MPNST = Malignant peripheral nerve sheath tumors. See text for other abbreviations.

growth factors are released from the surrounding ECM by protease digestion of the ECM (see Fig. l), many well-characterized growth factors and their receptors have also been implicated in autocrine growth stimulation of CNS tumors (see Fig. 2; Table 1). These factors include platelet-derived growth factors (PDGFs) A and B, epidermal growth factor (EGF), transforming growth factor (TGF)-a, acidic and basic fibroblast growth factor (FGFs), insulin-like growth factors (1GF)-1 and IGF-2, and their cognate receptors.51, PDGFA and PDGFB form three dimers, PDGF-AA, PDGF-BB, and PDGF-AB. Binding to their receptors PDGFR-a and PDGFR-P induces the PDGFR dimers a-a,p-p, and a-p. The factors and the receptors are expressed widely in gliomas, meningiomas, ependym0mas,7~and neuroblastomaszZ(see Table 1). Reactivation of expression of the normal, developmentally regulated genes appears to be the cause of overexpres-

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

983

sion in most cases since amplification and rearrangements occur in only 10% to 20% of gliomas51;however, there may be tissue-type specificity because amplification of PDGFR-a appears to be restricted to oligodendrogliomas with anaplastic features.13yThe induction of PDGF expression appears to be a sufficient and early event in tumorigenesis; the injection of a retrovirus coding the PDGFB chain into mouse brains induced tumors with features similar to gliobastoma multiforme (GBM) or primitive neuroectodermal tumor (PNET).154Similar tumor induction in p16 null mice provided permanent cell lines for further analyses. These tumor cells expressed not only the PDGFB chain, but also increased PDGFR-a. The resultant autocrine stimulation could be specifically inhibited in ~ i t r 0 .Such l ~ ~ experiments show the significant contributions to tumorigenesis that arise from the change in the expression of a single growth factor. The acidic and basic FGFs (aFGF, bFGF) belong to a class of heparinbinding factors that bind to isoforms of their cognate receptors FGFRl and FGFR2.5I. 8o Increases in aFGF and bFGF or FGFRs have been reported in astrocytoma, meningioma, schwannoma, and neuroblastoma (see Table 1).Coexpression of the factors and their receptors contributes to autocrine stimulation since glioma growth was inhibited in the presence of antisense bFGF oligonucleotides'00,lo3 or transfected antisense vectors.120The mitogenic effects may be the result of overexpression of normal proteins; however, studies on glioblastomas have shown that the a-exon of FGFRl was alternatively spliced out of the transcript resulting in a high-affinity receptor believed to contribute to glial malignan~y.~~ Thus, changes in either the growth factor levels (as described previously for PDGF) or changes to the receptor can have significant impact on tumor growth. IGF-1 and IGF-2 are endocrine factors that circulate in the bloodstream and bind to their cognate receptors, (IGFlR), a tyrosine kinase receptor, and IGF2R, the mannose 6-phosphate receptor.51 The IGFs also bind to IGF binding proteins (IGFBP), thought to modulate IGF function. IGF-1, IGF-2, or IGFlR are increased in glioma, meningioma, schwannoma, neuroblastoma, ependymoma, and medulloblastoma (see Io6, lo8,117 IGFBPl is increased in gliomas and IGFBPl through Table l).38, IGFBP3 are increased in meningiomas,38 suggesting a role for these proteins in brain tumors. For most CNS tumors, the autocrine stimulation by IGFs is proposed (because the factors or receptors or both are up-regulated in expression) but has not yet been established; however, in medulloblastomas, studies suggest that the lack of PTCHIoypromotes enhanced IGF2 expression and that this overexpression is indispensable for tumor formation.4yAs a result, the IGF system is a promising therapeutic target. Ligand binding of a member of the EGF family of growth factors (EGF, TGF-a, epiregulin, f3-cellulin, HB-EGF, amphiregulin, or NRGs [ - 1, - 2, - 3, - 41) to an EGF receptor (EGFR/c-erbB1, B2, B3, and B4) leads to receptor dimeri~ation.~~ The pleiotropic effects of EGFR activation are due to the specificity and affinities of the various ligands for

984

REMPEL

the various receptors.158In CNS tumors, the ligands TGF-a and EGF contribute significantly to the autocrine stimulation of EGFR.146TGF-a is increased in g l i ~ m a s meningiomas,5O ,~~~ and neuroblastomas,22and when expression was inhibited in U251 glioma cells using antisense vectors, reduced expression was accompanied by approximately 50% growth inhibition.147EGF and EGFR are increased in meningiomas and gliomas; however, the EGF receptor is amplified with rearrangements or mutations in approximately 40% to 50% of all GBMs (presumably de novo astrocytic GBMS,~~ see Table 1) and in low-grade and high-grade oligodendrogliomas.121 In a study of 44 GBMs, 17 tumors had amplified EGFR with alterationsz5 Thus, the commonest mutation resulted in a transcript lacking amino acids 6 through 273 (67%). Other mutations included transcripts encoding a truncated protein (amino acids from 958 [15%]) or a protein lacking amino acids 521 through 603 (15%). Of tumors with amplification, 33% showed multiple mutations. The transforming effects of EGFR in gliomas appear to be complex and conferred through ligand-independent, constitutively activated isoforms. The ability of a tumor to generate multiple, functional EGFR mutants may contribute to the ability of those gliomas to evade multimodal therapies and greatly increases the complexity in designing therapeutic targeting based on EGFR stirn~lation.~~ Signaling

When the growth factors bind to their receptors, they induce receptor dimerization (see Fig. 2). This pairing activates the kinase domain of both receptors, inducing the transautophosphorylation of the receptors in several locations.80This phosphorylation creates docking sites to recruit the signaling proteins SHc adaptor protein (Shc) or growth factor receptor binding protein-2 (Grb2). Grb2 binds to the son of sevenless (SOS) protein that engages and phosphorylates the Ras protein, which, in turn, induces two major signaling pathways (Fig. 2). In pathway 1,115 Ras activates Raf, followed by mitogen-activated protein kinase (MAPK) kinase kinase (MEK) 1/2, followed by MAPK/extracellular signal-regulated kinase (ERK) activation. Phosphorylated MAPK translocates into the nucleus, where it up-regulates the transcription of the protooncogenes c-myc and c-fos. c-Myc and c-Fos regulate the transcription of proliferation-associated genes, which, in turn, promote cell division. In pathway 2, Ras activates phosphatidylinositol-3kinase (PI3K) which, in turn, induces a cascade of events. v-akt mucine thyoma viral homolog (Akt)/PKB becomes phosphorylated and induces phosphorylation of the IKB kinase (IKK) complex. The complex induces the release of nuclear factor KB (NF-KB)by phosphorylation of its inhibitor IKB. Free NF-KB translocates to the nucleus, where it regulates expression of antiapoptotic genes, also promoting cell division.88 Mutations in signaling molecules have not been documented for brain tumors. It appears that the increase in enzyme activation in response to increased growth factor activation is sufficient to induce prolif-

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

985

Figure 2. Growth factor and cytokine signaling. Growth factors: Growth factors bind to their cognate tyrosine kinase receptors to induce dimerization, transautophosphorylation, and the creation of docking sites for signaling proteins, such as Shc and Grb2. The Grb2/SOS complex activates Ras that, in turn, activates the downstream Raf/MEK/MAPK cascade or the PI-3WAkt nuclear factor KB (NF-KB) cascade. The activated MAPK and free NF-KB translocate into the nucleus where they induce the expression of pro-proliferation and antiapoptotic genes, respectively. Cytokines: Cytokines bind to their receptor dimers that are devoid of kinase activity. When activated by specific ligands, the receptors recruit one of the four members of the cytoplasmic protein tyrosine kinases, the Janus kinases (JAKs), including Tyk2, JAK1, JAK2, and JAKS. The downstream effectors are a family of proteins called signal transducer and activator of transcription (STATs). Following JAK receptor tyrosine phosphorylation, the STAT proteins homo- or heterodimerize and translocate to the nucleus where they bind to specific DNA elements that regulate cytokine-inducible genes. The net effect of these interactions is increased cell proliferation. (Data from Croteau D, Mikkelsen T, Rempel SA, et al: New innovations and developments for glioma treatment. Clin Neurosurg 48:60-81, 2001; Shibuya M, Ito N, Claesson-Welch: Structure and function of vascular endothelial growth factor receptor-1 and -2. Curr Top Microbiol lmmunol 237:5983, 1999.)

eration in astrocytomas and meningiomas. Levels of activated Ras are increased in glioblastomas in comparison with levels in nonneoplastic brain.46The importance of Ras in this signaling cascade has been shown by inhibition of proliferation by blocking Ras farnesylation in glioblastoma cells by the drug lovastatin7 or by the lnhibition of Ras function using adenovirus-mediated gene transfer of dominant negative Ha-ras into meningioma cells.137These antitumoral effects and the pivotal role that Ras plays in controlling several signal transduction pathways suggest that it may be an important therapeutic target; however, the activation of Ras also may be mediated by growth factor-independent mechanisms because it is postulated that the loss of the guanosine triphosphatase (GTPase) activating protein, neurofibromin, in neurofibromatosis type (NF1) tumors leads to Ras activation (see Fig. 2).46

986

REMPEL

Other mitogen-stimulated signaling molecules elevated in brain tumors include the phospholipase C (PLC) family of protein^"^ and the other members of the Ras signaling pathways, including Shc2, Grb2/ SOS, Raf-MAPK, and PI-3K/Akt.85,115 These observations further emphasize the importance of these signaling pathways for CNS tumor proliferation and suggest that targeted inhibition of selected signaling pathway components could be exploited to inhibit tumor growth.115 Other Ligand and Receptor Signaling

Cytokine receptors are devoid of intrinsic tyrosine kinase activity. When activated by specific ligands, they recruit one of the four members of the cytoplasmic protein tyrosine kinases, the Janus kinases (JAKs), including Tyk2, JAK1, JAK2, and JAK3 (see Fig. 2). The downstream effectors are a family of proteins called STATs (signal transducer and activator of transcription). After JAK receptor tyrosine phosphorylation, the STAT proteins homodimerize or heterodimerize and translocate to the nucleus, where they bind to specific DNA elements that regulate cytokine-inducible genes. Increased levels of JAKl and JAK2 have been found in gliomas, medulloblastomas, and ependymomas.12Increases in the STATs 1,3,5, and 6 were observed in these tumors.12In meningioma, JAK2 and all STATs were highly elevated in contrast with levels observed in dura matter.89 Interleukin-13 receptor (IL-13R) is highly expressed in all grades of gliomas but especially high-grade gliomas18,19, ”; is present in a small percentage of meningiomas (2 of 20); and is absent in medulloblastomas.18Although it is unknown whether the receptor functions to contribute to tumor pathology, its selective expression to tumor cells suggests it may be a therapeutic target.19,72 Increased proliferation may be due not only to increased cell division, but also may be a result of inhibition of apoptosis.22The inhibition of apoptosis is mediated by a diverse collection of survival molecules that interfere with the ligand-induced or p53-induced apoptotic pathways. Proteins considered to be members of this group include Bc12 and Akt22(see Fig. 2). There are a growing number of proteins showing survival attributes, and it is anticipated that the inhibition of function of such proteins may promote apoptosis, providing possible targets for therapeutic intervention. SETA is one such candidate (see Fig. 2). It is increased in brain tumors when compared with that in normal brain? binds to the signal transduction protein cb1,6 and decreases apoptosis when transfected into cells,14 suggesting that it may play a role in the signal transduction of a survival pathway. Cell Cycle Regulation

The mitogenic stimuli that are transduced by the cytoplasmic signaling intermediates affect gene expression in the nucleus (Fig. 1)to initiate

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

987

cell cycle progression (Fig. 3). Here the cell possesses an intricate series of checks and balances regulated by positive effectors and negative regulators that control proliferation. The positive effectors include the cyclin-dependent kinase (CDK)/cyclin complexes (cdk 4,6/cyclin 1,2,3) and the protooncogenes MDM2, possibly its relative MDM4, and PAX5. The negative effectors include the cyclin-dependent kinase inhibitors (p15, p16, p18, p19, p21, and p27), and the pivotal tumor-suppressor genes Rb and p53. A positive growth signal transmitted through these cell cycle molecules results in the phosphorylation of Rb, dissociation of the Rb/E2F complex, and the release of E2F, which is essential for the cell to cross the restriction point (see Fig. 3).60, 136 Many of the mutations or genetic alterations observed in CNS tumors to date occur in the cell cycle control genes (see Table 1).These mutations or alterations promote unregulated cell proliferation, an essential biologic function for tumors

Figure 3. Cell cycle regulation. The mitogenic stimuli that are transduced by the cytoplasmic signaling pathway intermediates impact on gene expression in the nucleus to initiate cell cycle progression. Here, the cell possesses an intricate series of checks and balances regulated by positive effectors and negative regulators. The positive effectors include the cyclin-dependent kinase (CDK)/cyclin complexes (cdk 4, 6/cyclin 1, 2, 3) and the protooncogene proteins PAX5, MDM2, and MDM4. The negative effectors include the CDK inhibitors (p15, p16, p18, p19, p21, and p27), and the pivotal tumor suppressors Rb and p53. A positive growth signal transmitted through these cell cycle molecules results in the phosphotylation of Rb, dissociation of the Rb/E2F complex, and the release of E2F which is essential for the cell to cross the restriction point. Genetic mutations or alterations in gene expression reported for CNS tumors are indicated. (Data from Hunter T, Pines J: Cyclins and cancer II: Cyclin D and CDK inhibitors come of age. Cell 79573482, 1994.)

988

REMPEL

of all grades (see Table 1).There are two major axes that are implicated in the deregulation of the cell cycle, and these are centered on the tumorsuppressor genes p53 and Rb (Fig. 3). p 16-CDK4, G/CycIin D-Rb 1 Axis

Loss of the wild-type allele with mutation of the remaining allele or loss by homozygous deletion has been well documented for all components of the plb-CDK4,6/cyclin D-Rbl axis.22Deletion or mutation of each of these genes is prevalent in astro~ytomas~~, lo7,133, 134 and oligodendrogliomas,3.94 and they are associated with specific stages of glioma progression (see Table 1). Although it was thought that loss of p16 was restricted to tumors of glial origin: p26 mutations have been reported in malignant meningi~ m a s and ' ~ ~malignant peripheral nerve sheath tumors (MPNSTs).lo4 C0K4134and CDK617 amplification have been reported in gliomas. Although uncommon, amplification and overexpression of cyclin 0 1 and 0 3 also have been reported (4 of 102 GBMs and 2 of 8 anaplastic astrocytomas).8The loss of Rb in gliomas owing to mutations, deletions, and rearrangements is well documented.22, p21 -p53-MDMUMDM4/PAX5-p 14 Axis

The p21-~53-MDM2/MDM4/PAX5-p14 axis involves the p53 gene, which is mutated in approximately 60% to 65% of astrocytomas. Mutation in this gene is one of the earliest observed in astrocytomas, and the resultant lack of DNA damage monitoring most likely contributes to the short time to tumor recurrence and increased malignant progression p53 mutations also occur, although less observed in these frequently, in atypical meningiomas,16high-grade ependymoma~,'~~ and neurofibromasZ2(see Table 1).Mutations to p53 affect p21, which regulates the same cyclin D/CDK4 complex of the p16 pathway, indirectly regulating Rb phosphorylation (see Fig. 3). Although mutations or deletions of the p21 gene are rare in brain tumors, expression of p21 was decreased in seven malignant glioma cell lines; however, the decreased expression did not correlate with the mutational status of p53,2I suggesting that the regulation of p21 is complex and there are other unknown regulators of p21. Inhibitors of p53 include PAX5,141-143MDM2,63 and possibly MDM4.lZ7PAX5 represses the transcriptional activation of the p53 gene by binding to a 5' regulatory region that is necessary for p53 promoter activity.141 Increased expression of PAX5 has been reported only for astrocytomas. Overexpression was observed in areas exhibiting an increased expression of the EGFR, suggesting that these two genes may be connected intricately to the de novo glioblastoma pathway.142MDM2 functions to inhibit p21 transcriptional activation by p53 by binding to and concealing the p53 DNA binding domain and inducing the rapid degradation of the p53 protein.%,78 MDM2 amplification and overexpres-

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

989

sion occur in a greater percentage of de novo glioblastomas as compared with progression-related glioblastomas, suggesting that MDM2 overexpression also may constitute an alternate molecular mechanism of escape from p53-regulated growth contr01.~MDM2 is amplified and overexpressed in ependymomas.'44MDM4 was amplified and overexpressed in 5 of 208 malignant gliomas that did not have p53 mutations or MDM2 mutations, suggesting that it plays a role in suppressing p53 function in a few tumors.*27 Another important regulator of this pathway is ~ 1 4This ~ protein ~ ~ . arises from the alternative splicing of the CDKN2A locus that also codes for the p16 protein. p14 binds to the p53/MDM2 complex and inhibits MDM2-mediated degradation of p53. In a study of 190 astrocytic gliomas, 40% of GBMs and 5% of anaplastic astrocytomas were found to have mutant or deleted ~ 1 4 . ~ ~ There is evidence that mutation to any one of the members of these two axes is sufficient to promote cell division? 55, supporting the existence of redundant pathways to malignant progression. That mutation of one of these genes is sufficient to subvert cell cycle control is supported by the reports showing that the reexpression of ~ 1 6 ,~' 1~5~: ~~ Rb,Z8and ~ 5 in 3glioma ~ ~ tumor cells inhibits their growth. Although most mutated or altered genes function within these two major axes, almost every gene known to date that contributes to regulation of the G,/S phase of the cell cycle has been found to be mutated, even if only in a small percentage of tumors. As illustrated in Figure 3 and Table 1, mutations or expression changes have been observed for p57,I5O PCNA,'02 TGFs and other growth factors as described earlier, p27,77,97 and p1861,114 in at least one CNS tumor type. Although it appears that mutation in only one gene is sufficient to disrupt cell cycle control, tumors have been identified that have mutations in more than one of the cell cycle genes.66 Cell-Cell Contact and Cell Proliferation

Cell-cell contact is important in the negative regulation of cell proliferation. Upon cell-cell contact of cell adhesion molecules, a signal to inhibit proliferation is transduced intracellularly through members of the 4.1 family of membrane-associated protein^'^' that include the ezrinradixin-moesin proteins, merlin and DAL-1. These proteins function to link the actin cytoskeleton to cell membrane glycoproteins (Fig. 4). Disruption to this signaling pathway either by loss of receptor expression and function or loss of downstream 4.1 family proteins results in loss of the inhibitory signal and proliferation proceeds. Merlin is encoded by the NF2 gene. In normal Schwann cells, the protein is localized to the intracellular side of the cell membrane and is concentrated at cell-cell adhesion Loss of heterozygosity and mutations of the NF2 gene have been reported for meningiomas, schwannomas, and ependymomas. Analysis of the merlin protein corre-

Figure 4. Receptor interactions in proliferation, adhesion, and migration. Cell adhesion molecules: These transmembrane proteins modulate cell-cell interactions. The cytoplasmic domains interact with 4.1 superfamily members, such as ezrin-radixin-moesin (ERM), merlin, and DAL-1. These proteins link the receptors to the F-actin cytoskeleton. Either loss of cell-cell contact in the ECM or loss of a 4.1 family member disrupts the F-actin cytoskeleton. By mechanisms as yet unknown, this disruption induces a signaling response that promotes cell proliferation. Integrins: lntegrins are transmembrane proteins that modulate cell-ECM interactions. The extracellular interactions can be modulated via (1) the loss of ECM contact by protease degradation, (2) by activation of plasmin from plasminogen and additional subsequent protease activation, (3) UPNUPARbinding to integrins, or (4) by focal adhesion disassembly by secreted proteins, including SPARC, thrombospondin, and tenascin. The integrin-mediated intracellular signaling cascades are initiated directly, as described previously, by the clustering and adhesion of integrins to ECM proteins or through other receptor types, including the receptor tyrosine kinases involved in growth factor signaling, including the VEGFRs and the scatter factor receptor c-met. The subsequent integrin-mediated activation of focal adhesion kinase (FAK) can promote (1) cell proliferation by activation of the Ras/MAPK pathway, (2) the inhibition of apoptosis, (3) cell spreading, and (4) migration. Although the mechanisms that determine which of the downstream pathways is induced remain unclear, there appears to be cross-talk between the pathways that may influence the signaling outcome. FAK activation is countered negatively by the tumor suppressor, MACCWTEN, that reduces the tyrosine phosphorylation of FAK and pl30Cas, and inhibits the phosphorylation of Shc. (Data from Rempel SA: Molecular biology of central nervous system tumors. Curr Opin Oncol 10:179-185, 1998; Shibuya M, It0 N, Claesson-Welch: Structure and function of vascular endothelial growth factor receptor-1 and -2. Curr Top Microbiol lmmunol 23759-83, 1999; Turunen 0, Sainio M, Jaaleainen J, et al: Structure-function relationships in the ezrin family and the effect of tumor-associated point mutations in neurofibromatosis 2 protein. Biochim Biophys Acta 1387:l-16, 1998; and Cary LA, Guan JL: Focal adhesion kinase in integrin-mediated signaling. Front Biosci 41102-1 13, 1999.)

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

991

lated with reduced expression of protein in these tumors as well.47,86 In meningioma, loss was observed particularly in the fibrous and transitional 86 In contrast, merlin is up-regulated in astrocytomas compared with normal astrocytes. The loss of merlin appears integral to the pathogenesis of schwannoma, meningioma, and ependymoma but not astrocytic tumors.56 DAL-1 is also a member of the 4.1 family of membrane-associated p~0teins.I~~ It is localized to cell-cell contact points in areas rich in catenin and cadherins, suggesting it is a critical sensor of cell-cell contact that mediates growth arrest by signaling through actin cytoskeletalassociated proteins.149DAL-1 normally is expressed in high levels of the brain but is lost in approximately 60% of sporadic meningiomas and in all grades of The loss of DAL-1 in meningiomas may contribute to meningioma cell proliferation by pathways yet unknown (see Fig. 4). The regulation of tumor cell proliferation is complex and may be initiated by events occurring in the ECM, in the cytoplasm, or in the nucleus. When it comes to treating brain tumor patients with antiproliferative gene therapy approaches, future considerations should include the screening of each tumor for patient-specific mutations to design individual tumor-specific treatment approaches. Depending on the number of mutations identified, it may be prudent not to target the mutated gene or genes. Depending on where the mutation occurs in the cascade of events, it may be more efficacious to target a downstream gene that is affected by the mutation. Because the ultimate outcome of cell cycle signaling involved in Go-G, phase of the cell cycle is to release E2F from Rb, inhibition of E2F may alleviate the effects of mutations elsewhere in the cell cycle. Targeting E2F has been used successfully to inhibit cell division.29 ANGIOGENESIS

Tumor growth cannot be sustained without angiogenesis (the induction of new blood vessels from preexisting vessels). This is a requirement of benign and more aggressive tumors. The formation of new vessels requires the coordination of angiogenic and antiangiogenic factors that regulate activation, proteolysis of ECM, disruption of cell adhesion, migration and chemotaxis of endothelial cells, proliferation, and inhibition of endothelial cell proliferation.lOlAngiogenesis shares biologic processes in common with tumor cell proliferation (as previously discussed) and tumor cell invasion (see under Invasion). The establishment and remodeling of the blood vessels is initiated by growth signals interacting with their ligands at the cell surface (see Fig. 1). These surface signals are transduced by signaling molecules through the cytoplasm into the nucleus (see Fig. 2). The proliferative phase of angiogenesis shares similar biologic pathways used for tumor cell proliferation. In an analogous fashion, a disruption in the balance of either the angiogenesis-promoting genes or the angiogenesis-inhibiting

992

REMPEL

genes can lead to the stimulation of endothelial cell proliferation required for neovessel formation. Growth Factors and Their Receptors

The tumor is the major source of these factors that, when secreted into the ECM, act on their cognate receptors on endothelial cells in a paracrine manner. Of these factors, vascular endothelial growth factor (VEGF), plays a major role because its receptors are present only on endothelial cells. VEGF is a dimeric glycoprotein with four isoforms .that are highly specific for their cognate receptors Flt-I and KDR.160 VEGF is considered to be the most important mediator of neovascularization in gliomas, and its expression increases during glioma progression.112 Overexpression of VEGF has been shown for meningiomas,Q 118 ependymomas, hemangioblastomas, and neuroblast~mas.~~,It appears that the p16 gene, a gene that plays a role in cell cycle regulation (see earlier), also plays an important role in the regulation of VEGF as the restoration of wild-type p16 into U87 glioma cells significantly reduced the level of VEGF and the number of neovessels formed in V ~ V O . ~ ~ The VEGF receptors belong to the family of tyrosine kinase receptors that initiate signaling events on dimerization (see Fig. 2). Both receptors are increased on endothelial cells in malignant astrocytomas, anaplastic oligodendrogliomas, and ependym~mas.'~ The expression of flt-1 receptor occurs early in glioma tumorigenesis, whereas the KDR gene is upregulated during glioma progression.ll, 111 The dependency of tumor growth on the acquisition of blood vessels was shown by treating C6 glioblastoma tumors with a FLK-l(KDR) dominant-negative mutantg6 Although the mechanism of action of VEGF in glioma endothelial cells is less well characterized, studies of VEGF stimulation of aortic endothelial cells indicate that the ligand-induced activation of KDR results in the phosphorylation of PLC and downstream activation of PI3K kinase (see Fig. 2), providing insight into downstream effectors.160 Scatter factor ([SF]/hepatocyte growth factor [HGF]) is a disulfidelinked heterodimer composed of an a and a p chain.51It exerts pleiotropic effects on tumors, including the stimulation of angiogenesis through interactions with its receptor c-met, a tyrosine kinase receptor, present on endothelial cells.51,81 SF and c-met are expressed in low-grade and high-grade astrocytomas, oligodendrogliomas, ependymomas, and glioblastomas128;however, the receptor is expressed not only on the endothelial cells but also on the tumor cells, suggesting that SF affects not only angiogenesis but also tumor cell growth.128When SF was transfected into 9L glioma cells that do not express the factor but do express c-met, glioma growth and angiogenesis were enhanced.84 Reciprocally, U87 glioma growth and angiogenesis were inhibited by transfection with HGF/NK2, an antagonist of SF.45These data suggest that the SFlc-met autocrine and paracrine mechanisms are potential therapeutic targets.

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

993

Several of the previously described growth factors and their tyrosine kinase receptors play a role in angiogenesis (see Fig. 2). Basic FGF (bFGF) expression is up-regulated early and consistently throughout glioma progression135and has been implicated in neovascularization of schwannomas (see Table 1). It was thought that bFGF has little effect on angiogenesis alone but has an effect when combined with VEGF.135The delayed effect of FGF on angiogenesis also may be explained by delayed expression of the cognate receptors, FGFRl and FGFR2, that are observed on endothelial cells in malignant gliomas? Similarly, PDGF expression is up-regulated early in low-grade gliomas, but PDGFR-P is expressed in the endothelial cells of high-grade gliomas.llo Another set of growth factors that contribute to angiogenesis are TGF-P1, TGF-P2, and TGF-P3 that bind to serine/threonine kinase receptor types I and 113 70 TGF-P1, TGF-P2, and TGF-P3 are secreted by malignant glioma cells.67,75 These factors presumably bind to their receptors that are present in endothelial cells and stimulate proliferation in a paracrine fashion. Other cytokines and their receptors may play a role in neovessel formation. The up-regulated expression of the cytokine SDFl and that of its cognate receptor CXCR4 was observed in gliomas in regions of angiogenesis.lZ5Experiments using endothelial cells from other human tissue sources indicate that CXCR4 expression is induced by bFGF and VEGF and that SDFl is able to induce chemotaxis of the endothelial cells by CXCR4 in vitro and capable of inducing neovessel formation in vivo.130 These data suggest that SDFl expression in brain tumors may serve as a potent chemoattractant for CXCR4-expressing endothelial cells and contribute to neovessel formation. Secreted Proteins

Matricellular proteins are members of a nonhomologous group of proteins that are secreted into the ECM (see Fig. l),where they interact with growth factors, matrix proteins, and cell-surface receptors, including integrins (see Fig. 2) but do not function as structural matrix components.lol They function to disrupt cell-matrix interactions that affect endothelial cell migration and proliferation. Members of this group of proteins include SPARC (secreted protein acidic and rich in cysteine), thrombospondin, and tenascin (see Table 1 and Fig. 4). SPARC is a proangiogenic and antiangiogenic factor. It influences (directly or indirectly) several biologic functions that promote endothelial cell migration.lOl,131 It promotes deadhesion of cells to substrate by disrupting focal adhesions and redistributing actin filaments. It is proposed that this reduced adherence promotes endothelial cell motility." SPARC indirectly promotes ECM degradation through increased levels of matrix metalloproteinases and plasminogen activator inhibitor-1 and influences the ECM environment through the reduction of ECM production.131 Inhibition of endothelial cell proliferation is accomplished, in

994

REMPEL

part, indirectly by binding to VEGF and PDGF in the ECM and inhibiting the binding of these growth factors to their receptors, preventing the growth signaling mechanism (as described previously). SPARC also may function intracellularly to inhibit proliferation as it is taken up by endothelial cells in culture and transported into the nucleus, where it may function to delay cell-cycle progression.41 SPARC is up-regulated in neovessel endothelial cells in g l i o m a ~and ’ ~ ~meningioma~,’~~ suggesting that it may play a role in brain tumor angiogenesis. Thrombospondins are a family of multimodular proteins that bind to the ECM with strong affinity and, similar to SPARC, promote deadhesion of cells from substrate43Of the five members of the family, TSPl and TSP2 are antiangiogenic, inhibiting endothelial cell migration and proliferation. Both proteins have been implicated in the inhibition of glioma angiogenesis. TSPl is positively regulated by p53 and r n r ~ 2 3in~ ~ addition to an unknown genes on chromosome 10 implicated by the reintroduction of chromosome 10 into glioma cells.58The wild-type chromosome 10 reintroduction was accompanied by increased thrombospondin expression and inhibition of angiogenesis that was reversed by the presence of antithrombospondin a n t i b ~ d i e sThe . ~ ~ loss of expression of thrombospondin can occur through disruption of genes that frequently are targeted in gliomas (see earlier), in particular, the loss of p53 or loss of chromosome In another study, the loss of TSP2 in gliomas was correlated with increased vessel counts and density.” Together, these studies suggest that the loss of either TSPl or TSP2 may contribute to glioma angiogenesis and may have therapeutic potential as angiogenesis-inhibiting factors. Tenascin C, another matricellular protein with deadhesive properties, has been associated with angiogenesis-promoting activitie~.~~ Tenascin C is expressed in human glioma in vitro and in vivo, where its expression is associated strongly with neovessels.68, The inappropriate expression of several growth factors or secreted proteins tips the balance in favor of overall tumor neovascularization, suggesting that inhibition of angiogenesis may be an important therapeutic strategy. Targeting angiogenesis may be difficult, however, because many CNS tumors, including gliomas, meningiomas, and neurobla~tomas?~ express several angiogenic factors. Their redundancy and synergy in affecting angiogenesis may be prohibitive until the regulation of their expression is characterized more fully.23It was shown that the inhibition of FGF expression by dominant negative FGFRl and FGFR2 receptors impacted tumor growth through angiogenesis-dependent and angiogenesis-independent pathways.2 Its impact on angiogenesis suggests that FGF expression is important to the later expression of VEGF and that targeting FGF would have a significant therapeutic impact by inhibiting tumor growth, in part by preventing the downstream activation of VEGF-induced angiogenesis. By implication, CXCR4 expression (see earlier) would be prevented by the inhibition of bFGF and VEGF. Deciphering the cascade of events will have important consequences for treatment designs.

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

995

INVASION

The invasive or infiltrative phenotype renders brain tumors most difficult to cure. When tumor cells have disseminated into the adjacent brain tissue, recurrence and often tumor progression are inevitable. Understanding the biologic processes important in the modulation of tumor cell adhesion and migration may lead to the identification of novel therapeutic targets capable of inhibiting invasion specifically. Invasion as a biologic phenotype is complex, requiring the coordination of several processes, including the modulation of matrix degradation and cellmatrix interactions that in turn are mediated through extracellular and intracellular interactions (see Fig. 4).

Extracellular Events in Tumor Invasion Regulation of Matrix Degradation For a tumor cell to invade brain tissue, the adjacent matrix must be degraded partially to make room for the tumor cell. The matrix of CNS gray and white matter is composed of hyaluron, glycosaminoglycans, versican, pgTl with regional differences in aggrecan, and GHAP.34The matrix proteins collagens IV and V, laminin, fibronectin, vitronectin, entactin, and heparin sulfate are found in the brain vascular basement membranes.= Many matrix-degrading proteinases are implicated in brain tumor invasion.34,153 Matrix metalloproteinases (MMPs) can degrade almost all known ECM component^.'^ There are at least 20 members of this family that are grouped according to their substrate specificity, including collagenases, gelatinases, stromelysins, membrane-type MMPs, and others. MMP-2 and MMP-9 are expressed in gliomas, and further increased expression of MMP-9 is observed in higher grades. Although the role of MMPs in glioma invasion is not clearly understood, it has been shown that the inhibition of MMP-9 using antisense gene transfer impairs glioblastoma invasion in vivo,76 and MMP inhibitors reduced glioma invasion in vitro,lMsuggesting that these molecules may serve as therapeutic targets. The cysteine proteinase cathepsin B functions as an endopeptidase and exopeptidase and has a broad range of substrates.20Its expression is increased in human gli0mas,9~is increased further with increasing tumor grade, and is expressed in invading tumor cells, suggesting it may contribute in glioma invasion.122 To address this question, subclones of U251 tumor cells that express varying levels of cathepsin B were used in an in vitro spheroid confrontation assay. Cathepsin B expression level correlated with the extent of tumor invasion into the normal brain spheroid. This invasion could be inhibited by the addition of specific cysteine protease inhibitors.*O Active cathepsin B was localized to the cell surface of these cells in culture, and it was found to be coincidentally

996

REMPEL

active in regions of laminin degradation.20It appears that this enzyme also contributes to glioma-induced matrix degradation. The plasminogen-plasmin system consists of the inactive plasminogen, the activated plasmin, urokinase-type plasminogen activator (uPA) and tissue-type plasminogen activator (PA), and the plasminogen activator inhibitors type 1 (PAI-1) and type 2 (PAI-2). On the cell surface, uPA binds to its receptor and converts plasminogen to plasmin (see Fig. 4). Plasmin is a broad-spectrum protease that degrades ECM and activates latent forms of collagenases that in turn hydrolyze collagenous components that are not substrates of plasmin. The up-regulation of uPA can initiate a cascade of events to promote pericellular matrix degradation and promote cell migration. uPA and uPAR are coexpressed in malignant astrocytomas, and uPA was found to be expressed predominantly at the leading edge of the tumors, suggesting that they may contribute to brain tumor invasion.37When glioma. cells were infected with adenovirus-mediated antisense uPAR, glioma invasion and tumor growth were A disruption in the balance of these proteinases and their inhibitors may contribute to tumor invasion simply by promoting increased degradation of the ECM, thereby removing the physical barrier and allowing tumor cells to move away. Their effects are pleiotropic, however, likely owing to their modulation of the release of growth and angiogenic factors from the ECM (see Figs 1 and 4) and the interruption of integrinECM interactions (see Fig. 4). Integrin-Extracellular Matrix Interactions Integrins are transmembrane receptors composed of heterodimers of a and p subunits that mediate adhesion of cells to the surrounding ECM.l13 Changes to integrin and ECM components and their resultant interactions have profound effects on tumor cell adhesion and migration. It seems that adhesion and migration are different for different glioma 90 and preference for a matrix substrate depends at least cell 014, a5, partially on the integrins expressed. To date, the integrins ~~2,013, a6, pl, p4, av, avp3, and avp5 are found to be significantly expressed in gliomas and modulate adhesion and migration.26,35,39, 53, 152 Although astrocytic tumors migrate on a broad spectrum of ECM constituents to varying degrees, PNETs appear to be more restricted, with migration limited to laminin, fibronectin, and type IV It seems that the tumor cells themselves may modulate the ECM environment by secreting matrix molecules and in so doing provide a suitable matrix for adhesion and migration. The synthesis of vitronectin by GBM cells invading into adjacent brain has been and glioma cell lines, U87MG, U251MG, ANl/lacZ, and HF66 all expressed fibronectin, laminin, tenascin, collagen type IV, and chondroitin sulfate to varying It is speculated that the diversity and amount of ECM production by the tumor cells, combined with their integrin expres-

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

997

sion profiles, determine the extent of adhesion, which in turn affects the ability of the tumor cell to migrate. The uPA/uPAR complex also plays a role in tumor cell migration (see Fig. 4). uPAR is an adhesion receptor for vitronectin, and the uPA/ . ~ ~ ~ human glioblasuPAR complex binds to p l and p2 i n t e g r i n ~When toma cells were transfected with an antisense uPAR construct, transfectants showed lower invasion through matrigel; decreased migration on collagens; and increased adhesion to collagen, fibronectin, and laminin the invasiveness of these cells with the antisense in ~ i t r oBy . ~reversing ~ construct, the data suggest that the up-regulation of uPAR that is observed in gliomas could enhance invasion. Because uPAR binds to integrins, one possible mechanism is through the sequestration of integrins by enhanced uPAR so that cell-matrix interactions decrease, and invasion is enhanced.99 Secreted Proteins The matricellular protein SPARC is implicated in tumor cell invasion because of its ability to modulate cellular adhesion and deadhesion from matrix Although best characterized for its role in angiogenesis (see earlier), this protein also is expressed by tumor cells. SPARC is upregulated in g l i o m a ~ , ’invasive ~~ meningi~mas,’~~ and invading C6 glioma cells in ~ i v 0 . When l ~ ~ transfected into U87 cells, SPARC promoted invasion as assessed by an in vitro spheroid confrontation assay?’ SPARC may promote deadhesion through the interruption of vitronectinintegrin interactions because decreased tumor cell migration on vitronectin was associated with increased SPARC secretion in vitro, regardless of tumor SPARC also binds to collagen IV, and migration of SPARC-transfected clones on this matrix was found to be complex and dependent on the amount of SPARC secreted.lz6SPARC plays a role in angiogenesis (see earlier) and invasion, and the targeting of this protein may affect both tumor phenotypes. Brevican (BEHAB) is a brain-specific hyaluronan-binding protein. Its expression is increased in gliomas, and it is secreted into the ECM.69* 169 Although the full-length protein does not contribute to invasion, the 5’ peptide released on cleavage by ADAMTS4, a disintegrin and metalloproteinase with thrombospondin motifs, contributes to invasion in v ~ v oTargeting .~~ this metalloproteinasemay provide a novel therapeutic strategy to target the invasive phenotype specifically. Tumor cell-ECM interactions are complex, involving many proteinprotein interactions that regulate both the extent of matrix degradation and substrate production to provide the appropriate level of attachment to permit tumor cell migration into adjacent brain tissue. lntracellular Signaling Events in Tumor Invasion Integrins not only modulate extracellular events but also intracellular signaling because the integrins essentially serve as transmembrane

998

REMPEL

bridges between the ECM and the actin cytoskeleton. Integrins are involved in either direct or indirect signaling mechanism^.^^ Direct signaling results with the clustering and adhesion of integrins to ECM proteins; activation of cytoplasmic tyrosine kinases; and the induction of downstream effector pathways that modulate apoptosis, migration, spreading, and proliferation (see Fig. 4). Indirect signaling results when integrin signaling events are initiated through other receptor types, including the receptor tyrosine kinases involved in growth factor signaling, including V E G F R S ~and ~ the scatter factor receptor c-met*' (see Fig. 4 and earlier). Although the mechanisms are not understood completely, several key players involved in positively and negatively regulating integrin-initiated signaling pathways have been identified (see Fig. 4). Of these, focal adhesion kinase (FAK) promotes cell proliferation by activation of the RasIMAPK pathway and the mhibition of apoptosis.10 FAK also promotes cell spreading and migration. Although the mechanism that determines which of,the downstream pathways is induced by FAK activation is unclear, there appears to be cross-talk between the pathways that may influence the signaling outcome (see Fig. 4).'05 The influence of FAK activation by integrin signaling is countered negatively by the tumor suppressor MACC1/PTEN.87,140 MACCIIPTEN functions to inhibit FAK-induced cell migration and cell proliferation.44 MACCl/PTEN directly associates with FAK and can reduce its tyrosine phosphorylation and that of its downstream effector, ~130caS.~ When overexpressed, either FAK or ~1300"can antagonize MACCIIPTEN efMACCZIPTEN also inhibits the direct and indirect fects on mig1-afi0n.l~~ integrin-mediated activation of the Ras/MAPK pathway by inhibiting phosphorylation of Shc (see Fig. 4). In CNS tumors, alterations of MACCZIPTEN are associated with gliomas. The gene is mutated in 17% to 24% of xenografted and primary glioblastomas140and occurs occasionally in lower grade gliomas. The loss of MACCIIPTEN lifts the negative regulation of FAK, promoting cell proliferation or migration or both. The ability to affect growth and migration results from different domain activities. Growth suppression by MACCI/PTEN is through the phosphatase catalytic domain?* whereas the inhibition of invasion is independent of the phosphatase a~tivity.~' Cell surface receptor interactions play pivotal roles in signaling information regarding conditions in the immediate environment outside the cell into the cell. The transduced signal in turn induces a response within the cell that provokes the cell to undertake a specific behavior. A thorough understanding of the mechanisms of action of all the molecular participants will permit the design of future therapeutic strategies. SUMMARY

Many genetic alterations that contribute to CNS tumorigenesis and progression have been identified. One goal of such studies is to identify loci that would serve as diagnostic prognostic markers or both. A sig-

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

999

nificant advance is the observation that chromosome l p loss identified anaplastic oligodendroglioma and a subset of high-grade glioma patients who responded to chemotherapy and had longer survival times.9,64 Combined l p and 19q loss was a predictor of prolonged survival of patients having pure oligodendr~gliomas.~~~ Such markers eventually may be used to identify patients upfront who would benefit from treatment, while sparing patients who would not benefit. Although many molecular participants involved in the biologic pathways that promote proliferation, angiogenesis, and invasion have been elucidated, there are still many gaps in clinicians’ knowledge. It is ~~ expected that the use of the human genome project i n f ~ r m a t i o nand databases such as sAGEma~,7~ in combination with techniques such as cDNA arrays and proteomics, will facilitate greatly the identification of 59 cDNA arrays and tissue novel genes that contribute to CNS arrays13*will permit the construction of CNS-specific screening tools that will permit the identification of tumor-specific mutations and alterations so that patient-specific therapies can be designed. ACKNOWLEDGMENTS The author is grateful to Dr Oliver Bogler, for critical review of this article, and acknowledges Devina Khan, for assistance with the illustrations. The author apologizes to all colleagues whose work could not be cited because of space considerations.

References 1. Arap W, Nishikawa R, Fumari FB, et al: Replacement of the pl6/CDKN2 gene suppresses human glioma cell growth. Cancer Res 55:1351-1354, 1995 2. Auguste P, Gursel DB, Lemiere S, et a1 Inhibition of fibroblast growth factor/ fibroblast growth factor receptor activity in glioma cells impedes tumor growth by both angiogenesis-dependent and -independent mechanisms. Cancer Res 61:17171726, 2001 3. Barker FG, Chen P, Furman F, et al: p16 deletion and mutation analysis in human brain tumors. J Neurooncol31:17-23, 1997 4. Biernat W, Kleihues P, Yonekawa Y, et al: Amplification and overexpression of MDM2 in primary (de novo) glioblastomas. J Neuropathol Exp Neurol56:180-185, 1997 5. Bogler 0, Funari FB, Kindler-Roehrborn A, et al: SETA: A novel SH3 domaincontaining adapter molecule associated with malignancy in astrocytes. Neurooncology 2:6-15,2000 6 . Borinstein SC, Hyatt MA, Sykes VW, et al: SETA is a multifunctional adapter protein with three SH3 domains that binds Grb2, Cbl, and the novel SB1 proteins. Cell Signal 12:769-779, 2000 7. Bouterfa HL, Sattelmeyer V, Czub S, et al: Inhibition of ras farnesylation by lovastatin leads to downregulation of proliferation and migration in primary cultured human glioblastomas cells. Anticancer Res 20:2761-2772, 2000 8. Buschges R, Weber RG, Actor B, et al: Amplification and expression of cyclin D genes (CCND1, CCND2, and CCND3) in human malignant gliomas. Brain Pathol 9:435-443, 1999 9. Cairncross LG, Ueki K, Zlatescu MC, et al: Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst 903473-1479, 1998

10. Cary LA, Guan J-LFocal adhesion kinase in integrin-mediated signaling. Front Biosci 4~102-113,1999 11. Carroll RS, Zhang J, Bello L, et al: KDR activation in astrocytic neoplasms. Cancer 861335-1341, 1999 12. Cattaneo E, Magrassi L, De-Fraja C, et al: Variations in the levels of the JAK/STAT and ShcA protein in human brain tumors. Anticancer Res 18:2381-2388, 1998 13. Chan AS, Leung SY, Wong MP, et al: Expression of vascular endothelial growth factor and its receptors in the anaplastic progression of astrocytoma, oligodendroglioma, and ependymoma. Am J Surg Pathol22816-826,1998 14. Chen 8, Borinstein SC, Gillis J, et a1 The glioma-associated protein SETA interacts with AIPl/Alix and ALG-2 and modulates apoptosis in astrocytes. J Biol Chem 27519275-19281, 2000 15. Chintala SK, TOM JC, Rao J S Matrix metalloproteinases and their biological function in human gliomas. Int J Dev Neurosci 17495-502, 1999 16. Cho H, Ha SY, Park SH, et al: Role of p53 gene mutations in tumor aggressiveness of intracranial meningiomas. J Korean Med Sci 14:199-205, 1999 17. Costello JF, Plass C, Arap W, et al: Cyclin-dependent kinase 6 (CDK6) amplification in human gliomas identified using two-dimensional separation of genomic DNA. Cancer Res 571250-1254, 1997 18. Debinski W, Gibo DM, Slagle B, Ot al: Receptor for interleukin 13 is abundantly and specifically overexpressed in patients with glioblastoma multiforme. Int J Oncol 13:481486, 1999 19. Debinski W, Slagle B, Gibo DM, et al: Expression of a restrictive receptor for interleukin 13 is associated with glial transformation. J Neurooncol48:103-111, 2000 20. Demchik LL, Sameni M, Nelson K, et a1 Cathepsin B and glioma invasion. Int J Dev Neurosci 17483-494, 1999 21. Dirks PB, Hubbard SL, Murakami M, et al: Cyclin and cyclin-dependent kinase expression in human astrocytoma cell lines. J Neuropathol Exp Neurol 56:291-300, 1997 22. Dudas SP, Rempel SA: Development, molecular genetics, and gene therapy of glial tumors. In Rock JP, Rosenblum ML, Shaw E, et a1 (eds): The Practical Management of Low Grade Primary Brain Tumors. Philadelphia, Lippincott-Raven, 1999, pp 193-229 23. Eggert A, Ikegaki N, Kwiatkowski J, et a1 High-level expression of angiogenic factors is associated with advanced stage in human neuroblastomas. Clin Cancer Res 61900-1908,2000 24. Enam SA, Rosenblum ML, Edvardsen K Role of extracellular matrix in tumor invasion: Migration of glioma cells along fibronectin-positive mesenchymal cell processes. Neurosurgery 42:599408,1998 25. Frederick L, Wang X-Y, Eley G, et al: Diversity and frequency of epidermal growth factor receptor alterations in human glioblastomas. Cancer Res 60:1383-1387,2000 26. Friedlander DR, Zagzag D, Shiff B, et al: Migration of brain tumor cells on extracellular matrix proteins in vitro correlates with tumor type and grade and involves a v and p l integrins. Cancer Res 56:1939-1947, 1996 27. Fueyo J, Gomez-Manzano C, Yung WKA, et al: Adenovirus-mediated pl6/CDKN2A gene transfer induces growth arrest and modifies the transformed phenotype of glioma cells. Oncogene 12102-110, 1996 28. Fueyo J, Gomez-Manzano C, Yung WKA, et al: Suppression of human glioma growth by adenovirus-mediated Rb gene transfer. Neurology 50:1307-1315, 1998 29. Fueyo J, Gomez-Manzano C, Yung WKA, et al: Overexpression of E2F-1 in glioma triggers apoptosis and suppresses tumor growth in vitro and in vivo. Nat Med 4:685490, 1998 30. Fulci G, Ishii N, Van Meir EG: p53 and brain tumors: From gene mutations to gene therapy. Brain Pathol8:599413, 1998 31. Fuller GN, Rhee CH, Hess KR, et a1 Reactivation of insulin-like growth factor binding protein 2 expression in glioblastoma multiforme: A revelation by parallel gene expression profiling. Cancer Res 59:4228-4232, 1999 32. Funari FB, Lin H, Huang H-JS, et al: Growth suppression of glioma cells by PTEN

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

requires a functional phosphatase catalytic domain. Proc Natl Acad Sci U S A 9412479-12484, 1997 Fuxe J, Akusjarvi G, Goike HM, et al: Adenovirus-mediated overexpression of inhibits human glioma cell growth, induces replicative senescence, and inhibits telomerase activity similarly to ~ 1 6 " ~ ~Cell ' . Growth Differ 11:373-384, 2000 Giese A, Westphal M: Gliomainvasion in the central nervous system. Neurosurgery 391235-252. 1996 Gladson CL, Cheresh DA: Glioblastoma expression of vitronectin and the avp3 integrin. J Clin Invest 88:1924-1932, 1991 Gladson CL, Wilcox JN, Sanders L, et al: Cerebral microenvironment influences expression of the vitronectin gene in astrocytic tumors. J Cell Sci 108:947-956, 1995 Gladson CL, Pijuan-Thompson, Olman MA, et al: Up-regulation of urokinase and urokinase receptor genes in malignant astrocytoma. Am J Pathol 146:1150-1160, 1995 Glick RF, Lichtor T, Unterman TG: Insulin-like growth factors in central nervous system tumors. J Neurooncol35:315-325, 1997 Goldbrunner RH, Haugland HK, Klein CE, et al: ECM dependent and integrin mediated tumor cell migration of human glioma and melanoma cell lines under serum-free conditions. Anticancer Res 16:3697-3688, 1996 Golembieski WA, Ge S, Nelson K, et al: Increased SPARC expression promotes U87 glioblastoma invasion in vitro. Int J Dev Neurosci 17463472, 1999 Gooden MD, Vernon RB, Bassuk JA, et a1 Cell-cycle dependent nuclear localization with the matricellular protein SPARC: Association with the nuclear matrix. J Cell Biochem 74152-167, 1999 Greenwood JA, Murphy-Ullrich JE: Signaling of de-adhesion in cellular regulation of motility. Microsc Res Tech 43:420432, 1998 Greenwood JA, Pallero MA, Theibert AB, et al: Thrombospondin signaling of focal adhesion disassembly requires activation of phosphoinositide 3-kinase. J Biol Chem 273:1755-1763, 1998 Gu J, Tamura M, Pankov R, et al: Shc and FAK differentially regulate cell motility and directionality modulated by PTEN. J Cell Biol 146389403, 1999 Guerin C, Luddy C, Abounder R, et a1 Glioma inhibition by HGF/NK2, an antagonist of scatter factor/hepatocyte growth factor. Biochem Biophys Res Commun 273:287293, 2000 Guha A: Ras activation in astrocytomas and neurofibromas. Can J Neurol Sci 25:267281, 1998 Gutmann DH, Giordano MJ, Fishback AS, et al: Loss of merlin expression in sporadic meningiomas, ependymomas and schwannomas. Neurology 49:267-270, 1997 Gutmann DH, Donahoe J, Perry A, et al: Loss of DAL-1, a second protein 4.1-related tumor suppressor, is an important early event in the pathogenesis of meningiomas. Hum Mol Genet 9:1495-1500, 2000 Hahn H, Wojnowski L, Specht K, et a1 Patched target Igf2 is indispensable for the formation of medulloblastoma and rhabdomyosarcoma. J Biol Chem 275:2834128344.2000 ,~ - - Halper J, Jung C, Perry A, et al: Expression of TGFa in meningioma. J Neurooncology 45~127-134,1999 Hamel W, Westphal M: Growth factors in gliomas revisited. Acta Neurol 142:11> 138,2000 Harada H, Nakagawa K, Iwata S, et al: Restoration of wild-type p16 down-regulates vascular endothelial growth factor expression and inhibits angiogenesis in human gliomas. Cancer Res 593783-3789, 1999 Haugland HK, Tysnes BB, Tysnes OB: Adhesion and migration of human glioma cells are differently dependent on extracellular matrix molecules. Anticancer Res 171035-1043, 1997 Haupt Y, Maya R, Kazaz A, et al: MDM2 promotes the rapid degradation of p53. Nature 387296-299, 1997 He J, Olson JJ, James CD: Lack of p16INK4 or retinoblastoma protein (pRb), or amplification-associated overexpression of cdk4 is observed in distinct subsets of malignant glial tumors and cell lines. Cancer Res 55:4833-4836, 1995 ~~~

50.

51. 52. 53. 54. 55.

1001

~~

56. Hitotsumatsu T, Iwaki T, Kitamoto T, et al: Expression of neurofibromatosis 2 protein in human brain tumors: An immunohistochemical study. Acta Neuropathol 93:225232, 1997 57. Howe A, Aplin AE, Alahari SK, et al: Integrin signaling and cell growth control. Curr Opin Cell Biol 10:220-231, 1998 58. Hsu SC, Volpert OV, Steck PA, et al: Inhibition of angiogenesis in human glioblastomas by chromosome 10 induction of thrombospondin-1. Cancer Res 56:5684-5691, 1996 59. Huang H, Cotello S, Kuuer M, et al: Gene expression profiling of low-grade diffuse astrocytomas by cDNA arrays. Cancer Res 60:6868-6874,2000 60. Hunter T, Pines J: Cyclins and cancer: 11. Cyclin D and CDK inhibitors come of age. Cell 79:573-582, 1994 61. Husemann K, Wolter M, Buschges R, et a1 Identification of two distinct deleted regions on the short arm of chromosome 1 and rare mutation of the CDKN2C gene from lp32 in oligodendroglial tumors. J Neuropathol Exp Neurol58:1041-1050, 1999 62. Ichimura K, Schmidt EE, Goike HM, et a1 Human glioblastomas with no alterations of the CDKN2A (pl6INK4A, MTS1) and CDK4 genes have frequent mutations of the retinoblastoma gene. Oncogene 13:1065-1072, 1996 63. Ichimura K, Bolin MB, Goike HM, et a1 Deregulation of the p14ARF/MDM2/p53 pathway is a prerequisite for human astrocytic gliomas with G1-S transition control gene abnormalities. Cancer Res 60:417424, 2000 64. In0 Y, Zlatescu MC, Sasaki H, et al: Long survival and therapeutic responses in patients with histologically disparate high-grade gliomas demonstrating chromosome l p loss. J Neurosurg 92983-990,2000 65. International Human Genome Sequencing Consortium: Initial sequencing and analysis of the human genome. Nature 409:860-941, 2000 66. Ishii N, Maier D, Merlo A, et al: Frequent co-alterations of TP53, pl6/CDKN2A, pl4ARF, PTEN tumor suppressor genes in human glioma cell lines. Brain Pathol 9:469479, 1999 67. Isoe S, Naganuma H, Nakano S, et al: Resistance to growth inhibition by transforming growth factor-p in malignant glioma cells with functional receptors. J Neurosurg 88~529-534,1998 68. Jallo GI, Friedlander DR, Kelly PJ, et al: Tenascin-C expression in the cyst wall and fluid of human brain tumors correlates with angiogenesis. Neurosurgery 41:10521059, 1997 69. Jaworski DM, Kelly GM, Piepmeier JM, et al: BEHAB (brain-enriched hyaluronan binding) is expressed in surgical samples of glioma and in intracranial grafts of invasive glioma cell lines. Cancer Res 56:2293-2298, 1996 70. Jensen RL: Growth factor-mediated angiogenesis in the malignant progression of glial tumors: A review. Surg Neurol49:189-196, 1998 71. Jin W, McCutcheon IE, Fuller GN, et al: Fibroblast growth factor receptor-1 aexon exclusion and polypyrimidine tract-binding protein in glioblastoma multiforme tumors. Cancer Res 60:1221-1224, 2000 72. Joshi BH, Plautz GE, Puri RK: Interleukin-13 receptor a chain: A novel tumorassociated transmembrane protein in primary explants of human malignant gliomas. Cancer Res 60:116&1172,2000 73. Kazuno M, Tokunaga T, Oshika Y, et a1 Thrombospondin-2 (TSP2) expression is inversely correlated with vascularity. Eur J Cancer 35:502-506, 1999 74. Kirsch M, Wilson JC, Black I? Platelet-derived growth factor in human brain tumors. J Neurooncol35:289-301, 1997 75. Kjellman C, Olofsson SP, Hansson 0, et al: Expression of TGF-p isoforms, TGF-p receptors, and SMAD molecules at different stages of human glioma. Int J Cancer 89:251-258, 2000 76. Kondraganti S, Mohanam S, Chintala AK, et al: Selective suppression of matrix metalloproteinase-9 in human glioblastoma cells by antisense gene transfer impairs glioblastoma cell invasion. Cancer Res 60:6851-6855, 2000 77. Kourea HP, Cordon-Cardo C, Dudas M, et al: Expression of p27(kip) and other cell cycle regulators in malignant peripheral nerve sheath tumors and neurofibromas: The

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

1003

emerging role of p27(kip) in malignant transformation of neurofibromas. Am J Pathol 155:1885-1891, 1999 78. Kubbutat MH, Jones SN, Vousden K Regulation of p53 stability by MDM2. Nature 387299-303, 1997 79. La1 A, Lash AE, Altschul SF, et al: A public database for gene expression in human cancers. Cancer Res 59:5403-5407, 1999 80. Lala P, Angelov L, Provias J, et al: Growth factors and nervous system tumors. In McL.Black P, Loeffler JS (eds): Cancer of the Nervous System. Cambridge, MA, Blackwell Science, 1997, pp 744-772 81. Lamszus K, Schmidt NO, Jin L, et al: Scatter factor promotes motility of human glioma and neuromicrovascular endothelial cells. Int J Cancer 75:19-28, 1998 82. Lamszus K, Lengler U, Schmidt NO, et a1 Vascular endothelial growth factor, hepatocyte growth factor, basic fibroblast growth factor, and placenta growth factor in human meningiomas and their relation to angiogenesis and malignancy. Neurosurgery 46:938-948, 2000 83. Lang FF, Yung WK, Sawaya R, et al: Adenovirus-mediated p53 gene therapy for human gliomas. Neurosurgery 45:1093-1104, 1999 84. Laterra J, Nam M, Rosen E, et al: Scatter factor/hepatocyte growth factor gene transfer enhances glioma growth and angiogenesis in vivo. Lab Invest 76:565-577, 1997 85. Lau N, Feldkamp MM, Roncari L, et al: Loss of fibromin is associated with activation of RAS/MAPK and PI3-K/AKT signaling in a neurofibromatosis 1 astrocytoma. J Neuropathol Exp Neurol59:759-767, 2000 86. Lee JH, Sundaram V, Stein DJ, et a1 Reduced expression of schwannomin/merlin in human sporadic meningiomas. Neurosurgery 40:578-587, 1997 87. Li J, Yen C, Liaw D, et al: PTEN, a putative tyrosine phosphatase gene mutated in human brain, breast, and prostate. Science 275:1943-1947, 1997 88. Li N, Karin M: Signaling pathways leading to nuclear factor-kappa B activation. Meth Enzymol319:273-279, 2000 89. Magrassi L, De-Fraja C, Conti L, et al: Expression of the JAK and STAT superfamilies in human meningiomas. J Neurosurg 9k440-446, 1999 90. Mahesparan R, Tysnes BB, Read TA, et al: Extracellular matrix-induced cell migration from glioblastoma biopsy specimens in vitro. Acta Neuropathol97231-239, 1999 91. Maier D, Jones G, Li X, et a1 The PTEN lipid phosphatase domain is not required to inhibit invasion of glioma cells. Cancer Res 59:5479-5482, 1999 92. Matthews RT, Gary SC, Zerillo C, et a1 Brain-enriched hyaluronan-binding (BEHAB)/ brevican cleavage in a glioma cell line is mediated by a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) family member. J Biol Chem 275:2269522703, 2000 93. Menon I'M, Gutikrrez JA, Rempel S A A study of SPARC and vitronectin localization and expression in pediatric and adult gliomas: High SPARC expression correlates with decreased migration on vitronectin. Int J Oncol 17:683-693, 2000 94. Miettinen H, Kononen J, Sallinen P, et al: CDKN2/p16 predicts survival in oligodendrogliomas: Comparison with astrocytomas. J Neurooncol41:205-211, 1999 95. Mikkelsen T, Yan PS, Ho KL, et al: Immunolocalization of cathepsin B in human glioma: Implications for tumor invasion and angiogenesis. J Neurosurg 83:285-290, 1995 96. Millauer B, Shawver LK, Plate KH, et a1 Glioblastoma growth inhibited in vivo by a dominant-negative Flk-1 mutant. Nature 367576-579, 1993 97. Mizumatsu S, Tamiya T, On0 Y, et al: Expression of cell cycle regulator p27kipl is correlated with survival of patients with astrocytoma. Clin Cancer Res 5:551-557,1999 98. Mohan PM, Chintala SK, Mohanam S, et al: Adenovirus-mediated delivery of antisense gene to urokinase-type plasminogen activator receptor suppresses glioma invasion and tumor growth. Cancer Res 59:3369-3373, 1999 99. Mohanam S, Chintala SK, Go Y, et al: In vitro inhibition of human glioblastoma cell line invasiveness by antisense uPA receptor. Oncogene 14:1351-1359, 1997 100. Morrison RS Suppression of basic fibroblast growth factor expression by antisense

1004

REMPEL

oligonucleotides inhibits the growth of transformed human astrocytes. J Biol Chem 266~728-734,1991 101. Motamed K, Sage EH: Regulation of vascular morphogenesis by the matricellular protein SPARC. Kidney Int 51:1383-1387, 1997 102. Mottolese M, Natali PG, Coli A, et al: Comparative analysis of proliferating cell nuclear antigen and epidermal growth factor receptor expression in glial tumours: Correlation with histological grading. Anticancer Res 18:1951-1956, 1998 103. Murphy PR, Sat0 Y, Knee RS: Phosphorothioate anti-sense oligonucleotides against basic fibroblast growth factor inhibit anchorage-dependent and anchorage-independent growth of a malignant glioblastoma cell line. Mol Endocrinol 6S77-884, 1992 104. Nielsen GP, Stemmer-Rachaminov AO, Ino Y, et al: Malignant transformation of neurofibromas in neurofibromatosis 1 is associated with CDKN2A/pl6 inactivation. Am J Pathol 155379-1884, 1999 105. Nguyen DHD, Catling AD, Webb DJ, et al: Myosin light chain functions downstream of RAS/ERK to promote migration of urokinase-type plasminogen activator-stimulated cells in an integrin-selective manner. J Cell Biol 146:149-164, 1999 106. Ogino S, Kubo S, Abdul-Karim FW, et al: Comparative immunohistochemical study of insulin-like growth factor I1 and insulin-like growth factor receptor type 1 in pediatric brain tumors. Pediatr Dev Pathol423-31, 1001 107. Ono Y, Tamiya T, Ichikawa T, et al: Malignant astrocytomas with homozygous CDKN2/p16 gene deletions have higher Ki-67 proliferation indices. J Neuropathol Exp Neurol55:1026-1031, 1996 108. Patti R, Reddy CD, Geoerger B, et al: Autocrine secreted insulin-like growth factor-1 stimulates MAP kinase-dependent mitogenic effects in human primitive neuroectodermal tumor/medulloblastoma. Int J Oncol 16:577-584,2000 109. Pietsch T, Waha A, Koch A, et al: Medulloblastomas of the desmoplastic variant carry mutations of the human homologue of Drosophila patched. Cancer Res 5720852088, 1997 110. Plate KH, Breier G, Farrell C, et al: Platelet-derived growth factor receptor-B is induced during tumor development and upregulated during tumor progression in endothelial cells in human gliomas. Lab Invest 67529-534, 1992 111. Plate KH, Risau W Angiogenesis in malignant gliomas. Glia 15:339-347, 1995 112. Plate KH, Wamke PC: Vascular endothelial growth factor. J Neurooncol 35:365-372, 1997 113. Plow EF, Haas TA, Zhang L, et al: Ligand binding to integrins. J Biol Chem 275:2178521788, 2000 114. Pohl U, Caimcross JG, Louis DN: Homozygous deletions of the CDKN2C/pl8ink4c gene on the short arm of chromosome 1 in anaplastic oligodendrogliomas. Brain Pathol 9:639443, 1999 115. Pollack IF, Bredel M, Erff M: Application of signal transduction inhibition as a therapeutic strategy for central nervous system tumors. Pediatr Neurosurg 29:228244, 1998 116. Previtali S, Quattrini A, Nemni R, et al: a6P4 and a6Pl integrins in astrocytomas and other CNS tumors. J Neuropathol Exp Neurol 55:456-465, 1996 117. Provias JP, Becker LE: Cellular and molecular pathology of medulloblastoma. J Neurooncol29:35-43, 1996 118. Provias J, Claffey K, delAguila L, et al: Meningiomas: Role of vascular endothelial growth factor/vascular permeability factor in angiogenesis and periturnoral edema. Neurosurgery 40:1016-1026, 1997 119. Rebecchi MJ, Pentyala SN: Structure, function, and control of phosphoinositidespecific phospholipase C. Physiol Rev 80:1291-1335, 2000 120. Redekop GJ, Naus CC: Transfection with bFGF sense and antisense cDNA resulting in modification of malignant glioma growth. J Neurosurg 82:83-90, 1995 121. Reifenberger J, Reifenberger G, Ichimura K, et al: Epidermal growth factor receptor expression in oligodendroglial tumors. Am J Pathol 149:29-35, 1996 122. Rempel SA, Rosenblum ML, Mikkelsen T, et a1 Cathepsin B expression and localization in glioma progression and invasion. Cancer Res 54:6027-6031, 1994 123. Rempel SA, Golembieski WA, Ge S, et al: SPARC: A signal of astrocytic neoplastic

MOLECULAR BIOLOGY OF NERVOUS SYSTEM TUMORS

1005

transformation and reactive response in human primary and xenograft gliomas. J Neuropathol Exp Neurol571112-1121, 1998 124. Rempel SA, Ge S, Gutikrrez JA: Characterization of SPARC overexpression in meningiomas: A candidate diagnostic marker of invasive meningiomas. Clin Cancer Res 5:237-241, 1999 125. Rempel SA, Dudas S, Ge S, et al: Identification and localization of the cytokine SDFl and its receptor CXCR4 to regions of necrosis and angiogenesis in human glioblastoma. Clin Cancer Res 6:102-111, 2000 126. Rempel SA, Golembieski WA, Fisher J, et a1 SPARC modulates cell growth, attachment, and migration of U87 Glioma cells on brain extracellular matrix proteins. J Neurooncol, (spec iss) (in press) 127. Riemenschneider MJ, Biischges R, Wolter M, et a1 Amplification and overexpression of MDM4 (MDMX) gene from lq32 in a subset of malignant gliomas without TP53 mutation or MDM2 amplification. Cancer Res 59:6091-6096, 1999 128. Rosen EM, Laterra J, Joseph A, et al: Scatter factor expression and regulation in human glial tumors. Int J Cancer 67248-255, 1996 129. Rushing EJ, Brown DF, Hladik CL, et al: Correlation of bcl-2, p53 and MIB-1 expression with ependymoma grade and subtype. Mod Pathol 11:467470, 1998 130. Salcedo R, Wasserman K, Young HA, et al: Vascular endothelial growth factor and basic fibroblast gro'M'th factor induce expression of CXCR4 on human endothelial cells: In vivo neovascularization induced by stromal-derived factor 1. Am J Pathol 15411125-1135, 1999 131. Sage EH: Terms of attachment: SPARC and tumorigenesis. Nat Med 3:144-146, 1997 132. Sallinen SL, Sallinen PK, Haapasalo HK, et a1 Identification of differentially expressed genes in human gliomas by cDNA microarray and tissue chip techniques. Cancer Res 60:6617-6622,2000 133. Schmidt EE, Ichimura K, Messerle KR, et al: Infrequent methylation of CDKN2A(MTSZ/pl6) and rare mutation of both CDKN2A and CDKN2B(MTS/pl5) in primary astrocytic tumours. Br J Cancer 75:2-8, 1997 134. Schmidt EE, Ichimura K, Reifenberger G, et a1 CDKN2 (pl6/MTS1) gene deletion or CDK4 amplification occurs in the majority of glioblastomas. Cancer Res 54:63216324, 1994 135. Schmidt NO, Westphal M, Hagel C, et al: Levels of vascular endothelial growth factor, hepatocyte growth factor/scatter factor and basic fibroblast growth factor in human gliomas and their relation to angiogenesis. Int J Cancer 8410-18, 1999 136. Sherr CJ: Cancer cell cycles. Science 2741672-1677, 1996 137. Shu J, Lee JH, Harwalkar JA, et al: Adenovims-mediated gene transfer of dominant negative Ha-Ras inhibits proliferation of primary meningioma cells. Neurosurgery 44~579-587, 1999 138. Smith JS, Perry A, Bore11 TJ, et a1 Alterations of the chromosome arms l p and 19q as predictors of survival in oligodendrogliomas, astrocytomas, and mixed oligoastrocytomas. J Clin Oncol 18:636-645, 2000 139. Smith JS, Wang XY, Qian J, et al: Amplification of the platelet-derived growth factor receptor-A (PDGFRA) gene occurs in oligodendrogliomas with grade IV anaplastic features. J Neuropathol Exp Neurol 59:495-503, 2000 140. Steck PA, Pershouse MA, Jasser SA, et al: Identification of a candidate tumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers. Nat Genet 15:356-362, 1997 141. Stuart ET, Haffner R, Oren M, et al: Loss of p53 function through PAX-mediated transcriptional repression. EMBO J 14:5638-5645, 1995 142. Stuart ET, Kioussi C, Aguzzi A, et al: PAX5 expression correlates with increasing malignancy in human astrocytomas. Clin Cancer Res 1:207-214, 1995 143. Stuart ET, Gmss P: PAX: Developmental control genes in cell growth and differentiation. Cell Growth Diff 7405412, 1996 144. Suzuki SO, Iwaki T Amplification and overexpression of mdm2 gene in ependymomas. Mod Pathol 13:548-553, 2000 145. Tamura M, Gu J, Takino T, et al: Tumor suppressor PTEN inhibition of cell invasion,

1006

REMPEL

migration, and growth differential involvement of focal adhesion kinase and pl3OCas. Cancer Res 59:442449, 1999 146. Tang P, Steck PA, Yung WKA: The autocrine loop of TGF-alpha/EGFR and brain tumors. J Neurooncol35:30>314, 1997 147. Tang P, Jasser SA, Sung JC, et al: Transforming growth factor-a antisense vectors can inhibit glioma cell growth. J Neurooncol43:127-135, 1999 148. Tom JC, Kerkau S, Hanke A, et al: Effect of synthetic matrix-metalloproteinase inhibitors of invasive capacity and proliferation of human malignant gliomas in vitro. Int J Cancer 80:764-772, 1999 149. Tran YK, Bogler 0, Gorse KM, et a1 A novel member of the NF2/ERM/4.1 superfamily with growth suppressing properties in lung cancer. Cancer Res 59:3543, 1999 150. Tsugu A, Sakai K, Dirks P, et al: Expression of p5Pp2 potently blocks the growth of human astrocytomas and induces cell senescence. Am J Pathol 157919-932,2000 151. Turunen 0, Sainio M, Jaaskelainen J, et al: Structure-function relationships in the ezrin family and the effect of tumor-associated point mutations in neurofibromatosis 2 protein. Biochim Biophys Acta 13871-16,1998 152. Tysnes BB, Larsen LF, Ness GO, et al: Stimulation of glioma-cell migration by laminin and inhibition of anti-a3 and anti-pl integrin antibodies. Int J Cancer 67777-784,1996 153. Uhm JH, Dooley NP, Villemure JG, et al: Mechanisms of glioma invasion: Role of matrix-metalloproteinases.Can J Neurol Sci 243-15, 1997 154. Uhrbom L, Hesselager G, Nist6r M, et al: Induction of brain tumors in mice using a recombinant platelet-derived growth factor b-chain retrovirus. Cancer Res 58:52755279, 1998 155. Uhrbom L, Hesselager G, Ostman A, et al: Dependence of autocrine growth factor stimulation in platelet-derived growth factor-B-induced mouse brain tumor cells. Int J Cancer 85:398-406,2000 156. Vajkoczy P, Menger MD, Goldbrunner R, et al: Targeting angiogenesis inhibits tumor infiltration and expression of the pro-invasive protein SPARC. Int J Cancer 87261268, 2000 157. Weber RG, Bostrom J, Wolter M, et a1 Analysis of genomic alterations in benign, atypical, and malignant meningiomas: Toward a genetic model of meningioma progression. Proc Natl Acad Sci U S A 9414719-14724, 1997 158. Yarden Y, Sliwkowski MX: Untangling the ErbB signaling network. Mol Cell Biol 2127-137, 2001 159. Yebra M, Goretzki L, Pfeifer M, et al: Urokinase-type plasminogen activator binding to its receptor stimulates tumor cell migration by enhancing integrin-mediated signal transduction. Exp Cell Res 250:231-240,1999 160. Zachary I: Vascular endothelial growth factor: How it transmits its signal. Exp Nephrol6:480487, 1998 161. Zagzag D, Friedlander DR, Miller DC, et al: Tenascin C in astrocytomas correlates with angiogenesis. Cancer Res 55:907-914, 1995 162. Zagzag D, Friedlander DR, Dosik J, et al: Tenascin-C expression by angiogenic vessels in human astrocytomas and by human brain endothelial cells in vivo. Cancer Res 56182-189, 1996 163. Zhang H, Kelly G, Zerillo C, et al: Expression of cleaved brain-specific extracellular matrix protein mediates glioma cell invasion in vivo. J Neurosci 18:2370-2376, 1998

Address reprint requests to Sandra A. Rempel, PhD Barbara Jane Levy Laboratory of Molecular Neuro-Oncology Department of Neurosurgery Education and Research Building, Room 3096 Henry Ford Hospital 2799 W Grand Boulevard Detroit, MI 48202 e-mail: nssanQneuro.hfh.edu