Genetic Signaling in GBM 805
in any way coordinated, the individual or coordinated contribution of BTSCs to
tumor cell proliferation, and how BTSC-specific proteins interact with each other
and with chemoactive, antineoplastic agents should be of use not only in expanding
our understanding of how brain cancers develop, but also in the design of future
neurotherapeutic approaches and multimodal treatment strategies.
4 Gene Expression in the Human Brain
Mammalian brain cells have an intrinsically higher index of genetic output and
complexity compared to gene transcription profiles in other cells and tissue cell
systems (Lukiw et al., 2000; Sutcliffe, 2001; Colangelo et al., 2002; Mattick
and Makunin, 2005). Furthermore, complex biological behaviors and functional
neurochemical mechanisms that accompany aging and neuropathology, including
neuro-oncological change, are probably not controlled by single genes but rather
by groups of functionally related genes sometimes referred to as gene families. The
use of DNA array technologies is currently capable of interrogating 33,000 genes
on a single DNA array, or the levels of all expressed genes in a single brain biopsy
sample (Affymetrix Corporation, Santa Clara, CA; Lukiw, 2004; Lukiw et al., 2005;
Macdonald et al., 2007). Although some pathogenically related upregulated onco-
genic genes thus far identified may share common overlapping functions such as
stress response and adaptive processes that support several related aspects of brain
inflammation, apoptosis and/or angiogenesis, glioma and GBM may represent some
of the most heterogeneous gene expression patterns of any neurological disease
known (Kavsan et al., 2006; Idbaih et al., 2007; Juric et al., 2007; Culicchia et al.,
2008).
5 Gene Expression in Brain Cancers
Molecular–genetic and population-based studies have identified several gene muta-
tions that associate with brain tumor development (Tatter, 2005; Nicholas et al.,
2006; Van den Bent and Kros, 2007; Juric et al., 2007; Kavsan et al., 2007;
Johansson Swartling, 2008). These may be linked to an ordered accumulation of
multiple genetic mutations on multiple chromosomes and sequential, temporally
mediated pathogenic interactions. For example, in about one third of all cases, the

transition from healthy astroglial cells to astrocytoma has been associated with p53
gene mutations at chromosome 17p, and mutations in p53, a tumor suppressor gene,
were among the first genetic alterations ever to be identified in astrocytic brain
tumors (Tatter, 2005; Nicholas et al., 2006; Van den Bent and Kros, 2007). Deletion
or alteration of the p53 gene appears to be present in approximately 25–40% of all
GBMs, and p53 immunoreactivity also appears to be associated with tumors that
arise in younger patients. Brain cancers also often exhibit loss of heterozygosity
(LOH), and LOH at chromosome 10q is the most frequent gene alteration for both
806 W.J. Lukiw and F. Culicchia
primary and secondary glioblastomas, ranging from 60 to 90% of all cases. This
mutation appears to be specific for GBM, is found rarely in other tumor grades,
and is associated with poor survival. LOH at 10q plus 1 or 2 of the additional gene
mutations appear to be frequent alterations and are most likely major players in
the development of GBM. LOH leads in almost one half of all subsequent cases to
anaplastic astrocytoma associated with the retinoblastoma (Rb) gene at chromosome
13q. Involvement of additional mutations at chromosomes 9p and 19q, followed by
GBM development associated with chromosome 10 mutations and amplification of
the epidermal growth factor receptor gene is further postulated in this highly com-
plex pathway of oncogenic development (see, e.g., Tatter, 2005; Nicholas et al.,
2006; Van den Bent and Kros, 2007; Johansson Swartling, 2008). The involvement
of multiple gene loci and reduced or incomplete penetrance of these gene muta-
tions indicate that the resulting altered developmental or oncogenic pathways induce
tumors possessing a highly variable phenotype and heterogeneous morphology.
Brain cancers are also genetically associated with homeostatic disturbances in the
epidermal growth factor receptor (EGFR), MDM2, platelet-derived growth factor-
alpha (PDGFα) and PTEN genes. The EGFR gene is involved in the control of
cell proliferation and EGFR overexpression and mutant EGFR expression occurs
in approximately 50% of patients with GBM (Nicholas et al., 2006; Voelzke et al.,
2008). In fact multiple genetic mutations in EGFR are apparent, including overex-
pression of the receptor as well as rearrangements that result in truncated isoforms

(Voelzke et al., 2008). However, all the clinically relevant EGFR mutations appear
to generate a similar phenotype resulting in increased EGFR activity. Amplification
or overexpression of MDM2 constitutes an alternative mechanism to escape from
p53-regulated control of cell growth by binding to p53 and blunting its activity.
Overexpression of MDM2 is the second most common gene mutation in GBMs
and is observed in 10–15% of patients. Some studies show that this mutation asso-
ciates with a poor prognosis. The PDGF gene acts as a major mitogen for glial
cells by binding to the PDGF receptor (PDGFR) and amplification or overexpres-
sion of PDGFR is typical (60%) in the pathway leading to secondary glioblastomas.
PTEN (also known as MMAC and TEP1) encodes a tyrosine phosphatase located at
chromosome 10q23.3 that functions as a cellular phosphatase, turning off signaling
pathways, and is consistent with possible tumor-suppression activities. When phos-
phatase activity is lost because of genetic mutation, signaling pathways can become
activated constitutively, resulting in aberrant proliferation. PTEN mutations have
been found in as many as 30% of all glioblastoma cases studied (Koul, 2008; Cheng
et al., 2009).
Current DNA array technologies and statistical and comparative bioinformat-
ics analysis enable a comprehensive examination of the expression of all genes
associated with brain health and disease. Genomewide gene expression patterns
of neoplastic brain cells in their various developmental stages provide a fascinat-
ing reflection of the physiological and pathological status of those pathogenic brain
cells. Robust gene expression analyses have been applied to whole tumor cells and
some recurrent themes, besides such variables as patient age, sex, affected lobe and
disease onset, duration, and other clinical parameters, are emerging for specific
pathology-related genes. Interestingly, the progression from low-grade glioma to
Genetic Signaling in GBM 807
high-grade GBM may be associated with distinct molecular–genetic changes that
vary according to WHO grade (MacDonald et al., 2007; Juric et al., 2007). Several
excellent reviews of DNA array analysis of glioma and GBM have recently appeared
and the material in t hem is not reiterated here (Boudreau et al., 2005; Belda-Iniesta

et al., 2006; Kavsan et al., 2005; Tso et al., 2006; Faury et al., 2007; Idbaih et al.,
2007; Juric et al., 2007; MacDonald et al., 2007; Johansson Swartling, 2008). Rather
we focus on some current observations on increases in the expression of glioma
and GBM of several altered markers involved in brain cell contact, cell cycle, cell
death, and vascular proliferation markers at the level of gene expression in virtually
all brain tumors examined: beta-amyloid precursor protein (βAPP), the apoptosis
effector caspase-3, a cell–cell contact neuronal-enriched protein pentraxin-2, and
the angiogenesis promoting vascular endothelial growth factor (VEGF).
5.1 Beta-Amyloid Precursor Protein (βAPP)
Beta amyloid precursor protein (βAPP), a brain-abundant trans-membrane glyco-
protein “sensor” implicated in neuronal–glial intercellular contact and progressive
apoptotic and necrotic brain cell death appears to be part of a pathogenic gene family
that associates with glial cell proliferation in glioma, GBM, and neurodegeneration
(Colangelo et al., 2002; Lukiw, 2004; Fuso et al., 2007; Culicchia et al., 2008).
In fact most of the original studies on the role of βAPP in neurobiology and neu-
rodegeneration were first performed in transformed human glioblastoma cell lines
(Lahiri et al., 1997; Paris et al., 2005; Fuso et al., 2007; Culicchia et al., 2008).
Cancer-affected glial cells are characterized by highly unusual and diverse mor-
phology that often correlates to the grade of the neoplasm and display noncontact
inhibited cells, lack of cell–cell adhesion and connectivity, and a highly varied range
of cellular morphology (Fig. 1) (Hyun Huang et al., 2007; Caltagarone et al., 2007;
Culicchia et al., 2008). Bizarre glial cell morphology in malignant gliomas and
GBM have been correlated with the depletion of cytoskeletal-matrix actin-bundling
proteins and alterations in integrin-mediated communication between the extracellu-
lar matrix and the actin cytoskeleton (Venezia et al., 2007; Caltagarone et al., 2007;
Young-Pearse et al., 2007). Although the abundant cytoskeletal protein β-actin itself
is not upregulated, β-actin-associated cytoskeletal proteins and integrins that further
drive glial cell division and proliferation processes during the cell cycle have been
implicated in brain cancer development. Interestingly βAPP structural orientation
within the membrane, βAPP trafficking, and intracellular signaling functions are

associated with β-actin-associated proteins and β-actin-mediated cellular shape. It
is difficult to rationalize whether βAPP upregulation in glioma and GBM is either
a consequence of, or contributory factor to altered microfilament of microtubule
cytoarchitecture, or increased cellular proliferation, or both (Venezia et al., 2007;
Caltagarone et al., 2007; Young-Pearse et al., 2007; Culicchia et al., 2008).
Although βAPP is known to be variably upregulated in chronic neurodegenera-
tive disease, depending on the stage of the disease, the observation of upregulated
proinflammatory and amyloidogenic neural degenerative markers in glioma and
808 W.J. Lukiw and F. Culicchia
GBM is a relatively recent one (Lukiw, 2004; Lukiw et al., 2005; Fuso et al.,
2007; Sin et al., 2008; Culicchia et al., 2008). Excessive βAPP-mediated signaling is
thought to be responsible, in part, for driving neural inflammation, glial cell growth
and expansion, and brain cell degeneration events such as apoptosis (Melhorn et al.,
2000; Radde et al., 2007; Venezia et al., 2007; Herber et al., 2007). Ischemic brain
damage is also known to induce inflammatory cytokine and βAPP over-expression
that further induce widespread brain cell death via apoptosis or necrosis (Pluta,
2002; Bates et al., 2002). Chronic gliosis is, in addition, associated with altered
processing of βAPP in vivo, and thus may trigger pathological changes associated
with aberrant interneural communication between brain cells, thus contributing to
progressive alterations in glial cell morphology (Pluta, 2002; Bates et al., 2002;
Young-Pearse et al., 2007).
Increased upregulation of βAPP expression in glioma and GBM further suggests
that unscheduled proliferative events of brain cells are accompanied by the signifi-
cant elevation of integral transmembrane receptors that are pathogenic markers for
neurodegeneration. βAPP appears to be part of a poorly understood cell-contact
signaling pathway whose disruption induces cell-cycle s ignaling, mitotic abnormal-
ities, and glial cell expansion (Paris et al., 2005; Venezia et al., 2007; Young-Pearse
et al., 2007; Fuso et al., 2007; Lukiw et al., 2008). Increased βAPP expression has
long been associated with gliosis, the localized expansion of astrocyte populations,
and the production of dense fibrous networks of neuroglia in the area of a pathogenic

lesion. Similar gliosis-related increases in the expression of βAPP in glioma and
GBM and in the neurodegenerating brain tissues support some underlying com-
monality in disorganized interneural signaling, aberrant cytoarchitecture and neural
cell shape characteristic of both neurological conditions. Interestingly, Alzheimer’s
disease and GBM have similar age-specific incidence rates and accumulation of
senile plaque deposits consisting of amyloid beta (Aβ) peptides derived from the
secretase cleavage of βAPP holoprotein. About one third of all GBM cases exhibit
age-related plaque scores indicative or suggestive of AD; and progressive neurode-
generative pathology is present in about half of all cases of GBM (Nelson, 2002;
Lukiw et al., 2008).
5.2 Caspase-3
Whether brain cell death in the malignant neoplasms, triggered by hypoxia, lack
of nutritive support or other pathogenic factors, is driven by apoptosis or necrosis
is not well understood. In fact both neural-destructive processes may be operating
simultaneously. Apoptosis and necrosis appear to lie at either end of a spectrum
of functional brain cell dysfunction and progressive cell loss spanning from pro-
grammed cell death (apoptosis; internucleosomal DNA fragmentation) at one end
to induced and premature cell death (necrosis; randomized DNA fragmentation) of
brain cells at the other. Cysteine–aspartic acid protease-3 (caspase-3), a key mem-
ber of a family of 11 human cysteine proteases, plays key essential effector roles in
Genetic Signaling in GBM 809
both apoptosis and necrosis and in neuroinflammatory aspects of neurodegeneration
and brain tumor growth. There is evidence for both caspase-3 upregulation (Ray
et al., 2002; Lukiw et al., 2009) and caspase-3 downregulation in human brain
tumors (Stegh et al., 2008). Nonhomeostatic levels of caspase-3 indicate alterations
in the cellular balance of both pro-apoptotic and anti-apoptotic signals (Takuma
et al., 2004; Lefranc et al., 2007). The increased expression of the pro-apoptotic
Bax protein, upregulation of calpain and caspase-3, and occurrence of internucleo-
somal DNA fragmentation indicate that one mechanism of cell death in malignant
brain tumors is apoptosis (Ray et al., 2002; Lukiw et al., 2009). These results may

be explained by the fact that the apoptotic process only approaches the stage of
caspase-3 activation, followed by a subsequent variable activation of the apoptotic
cascade and “programmed” cell death mechanism, resulting in apoptotic blockage
and an accumulation of brain cell mass.
5.3 Pentraxin-2 (NP2; NPTX2)
Pentraxins are a family of pentameric calcium-dependent ligand-binding proteins
bearing a highly distinctive structure similar to that of the ring-shaped lectins
(Emsley et al., 1994). Pentraxins represent a novel neuronal uptake pathway that
functions during intercellular and extracellular signaling, synapse formation and
clustering, remodeling, and cell–cell contact (Gerrow and El-Husseini, 2007).
“Short” pentraxins include the inflammation-related serum amyloid P component
(SAP) and C reactive protein (CRP) and the “long” pentraxins include PTX3
(a cytokine-modulated molecule) and several prominent brain-enriched pentrax-
ins such as neuronal pentraxin-2 (NP2; NPTX2). Interestingly, NPTX2, normally
expressed in the CNS, is a member of a family of proteins related to CRP and
other acute-phase inflammatory mediators, and has been found to be correlated
with glioma and GBM edema, the swelling of soft tissues as the result of loss
of brain water balance and excessive water accumulation. Increased NPTX2 are
in turn strongly associated with poorer survival rates in tumors with the high-
est levels of edema (Hsu and Perin, 1995; Goodman et al., 1996). Several gene
expression studies have shown NPTX2 to be consistently and significantly upreg-
ulated in glioma and GBM (Carlson et al., 2007; Pope et al., 2008; unpublished
observations). It is important to note that the NPTX2 upregulation associated with
angiogenic- and edema-related signaling is often coregulated with the simultaneous
upregulation of vascular endothelial growth factor (VEGF) and the proliferation of
neovascularization.
5.4 Vascular Endothelial Growth Factor (VEGF)
Angiogenesis, the proliferation of vascular growth that provides nutritive sup-
port to the expanding tumor cell mass, is one of the hallmarks of all cancers.
810 W.J. Lukiw and F. Culicchia

Vascular endothelial growth factor stands out as a key mediator of tumor-associated
angiogenesis among a complex signaling system involving pro- and antiangiogenic
factors (Chamberlain, 2008; Grothey et al., 2008; Pope et al., 2008). The upregula-
tion of VEGF, originally described as a vascular permeability factor in brain tumors,
has often been proposed to be the major cause of both vasogenic edema in gliomas
and neovascularization (NV) (Bruce et al., 1987; Buie and Valgus, 2008; Norden
et al., 2008). A consistent observation in brain cancer biology is that malignant
gliomas invariably express vast amounts of VEGF, now regarded as an important
pathogenic marker of angiogenesis and NV, essential for the proliferation and the
survival at any cost for malignant glioma cells. NV is orchestrated by the coordinate
induction of a family of growth-factor genes and most prominently by VEGF which
also possesses endothelial cell-specific mitogenic effects that closely correlate with
NV during embryonic development and normal systemic physiology, fetal anemia,
in retinal NV, in models of hypoxic ischemia, and in malignant tumors. These com-
bined observations are suggestive of VEGF’s key role in vascular proliferation in
growth, health, and disease. Hypoxia is thought to be one crucial physiological
stimulus for VEGF upregulation that precedes NV, and low cellular oxygen tension
rapidly induces a number of transient genetic signals through which this is accom-
plished (Larrivee and Karsan, 2000; Hasan and Jayson, 2001; Giles, 2001; L. Lukiw
et al., 2003; Norden et al., 2008). The multiple roles of VEGF in brain tumor
development and proliferation and anti-VEGF-based therapies have been recently
examined in the last year in several excellent reviews and interested readers are
encouraged to refer to these thoughtful works and the published papers referenced
within (Brandsma et al., 2008; Chamberlain, 2008; Grothey and Ellis, 2008; Pope
et al., 2008; Reardon et al., 2008; Wong and Brem, 2008).
6 Specific Alterations in the Expression
of Brain-Enriched Genes
Neurological disorders including glioblastoma involve a highly complex patho-
genesis with multiple etiological factors, and this is reflected in the expression of
brain genes in this disease. Several glioma and GBM tumor cell lines have been

immortalized and “standardized brain tumor cell cultures” are available to oncol-
ogy researchers through government-funded sources such as the American Type
Tissue Collection (ATCC, Bethesda MD). Commonly used human brain cell cul-
tures include glioma cell line HTB-138 (Hs683) and glioblastoma tumor cell lines
CRL-2020 (DBTRG-05MG), CRL-1690 (T98G), CRL-2365 (M059K), and CRL-
2366 (M059J). The majority of these standardized neoplastic, immortalized cell
lines develop as a mixture of floating and adherent cells growing as heterogeneous
clusters of neuroblastic cells with multiple, short, fine cell processes (neurites) that
often aggregate, forming clumps, detach from solid surface, and float within the cell
culture medium (Fig. 1). Total DNA, RNA, and protein can be effectively and effi-
ciently isolated from these archived cell lines and are subsequently used for gene
Genetic Signaling in GBM 811
expression analysis and downstream molecular–genetic investigations. Recent stud-
ies in these cell lines have indicated increases in the integral membrane β-amyloid
precursor protein (βAPP) as a proinflammatory, neurodegenerative, and proliferative
pathogenic marker (Culicchia et al., 2008). Indeed, from the perspective of dys-
regulated pathogenic gene expression, glioma and glioblastoma multiforme display
significant upregulation of disease markers such as βAPP and caspase-3 with fea-
tures of rapid-onset, progressive, glial cell proliferating, degenerative brain disease.
The known disease-related functions of these inflammatory and neurodegenera-
tive markers may further contribute to the pathogenic phenotype and unscheduled
misregulated propagation of glial cells in the brain. The important point is that
brain tumors consist of a spectrum of tumors of varying differentiation, malignancy
grades, and gene expression profiles. Despite the fact that all tumors have an initially
invasive phenotype, early genetic events appear to differ between astrocytic and
oligodendroglial tumors, and this may form, in part, the molecular genetic basis for
variation in brain cancer cell composition that complicates more effective therapies.
Knowledge of malignant glioma genetics has already affected clinical management
of these tumors, and researchers and clinicians can only hope that further knowl-
edge of the evolution of the molecular pathology of malignant gliomas will result in

novel therapies that employ multiple, multimodal treatment strategies.
7 Micro RNAs (miRNAs): Specific miRNA and mRNA
Alterations in Human Brain Cancer
Micro RNAs (miRNAs) are small RNA polymerase II and III transcribed, noncoding
RNA molecules that play important posttranscriptional regulatory roles by recogniz-
ing and binding to the 3

untranslated region (3

UTR) of mature messenger RNAs
(mRNAs). By doing so, miRNAs repress translation and expression of their partic-
ular mRNA targets (Mattick and Makunin, 2005; Cao et al., 2006; Lukiw, 2007;
Lukiw and Pogue, 2007; Cho, 2007; Amaral et al., 2008; Dogini et al., 2008; Zeng
2009). Transcription of protein-encoding genes and miRNAs by RNA polymerase
II and III and their interrelated functions in the modulation of gene expression sug-
gests the possibility of some coordinated mode of interaction, possibly through
miRNA interaction with specific transcription factors (Mattick and Makunin, 2005;
Hobert, 2008; Lukiw et al., 2008; Makeyev and Maniatis, 2008; Williams et al.,
2008; Amaral et al., 2008). Interestingly, small signaling molecules such as miRNA
may transfer epigenetic information not only within cells but also between cells and
organ systems as part of a dynamic RNA-mediated interplay between the environ-
ment and the genome (Zhao et al., 2006; Hill et al., 2009; Mattick et al., 2009). Such
novel genetic mechanisms may explain, in part, cancer invasiveness and metasta-
sis throughout cells, organs, and tissue systems (Louis, 2006; Hyun Hwang et al.,
2008).
To date about 911 miRNAs in the human brain have been identified (Lukiw,
2007; Lukiw and Pogue, 2007). The miRNA-mediated regulation of messenger
RNA (mRNA) complexity in the human central nervous system is evolving as a
812 W.J. Lukiw and F. Culicchia
critical and determining factor in regulating CNS-specific gene expression during

development, plasticity, aging, and disease (Hobert, 2008; Makeyev and Maniatis,
2008; Williams et al., 2008). Several excellent recent reviews on miRNA specia-
tion in the CNS and specific examples in brain tumors have recently appeared in
the literature and the authors would encourage interested researchers, clinicians,
and medical and graduate students to read them over (Mattick and Makunin, 2005;
Ciafrè et al., 2005; Zhang et al., 2007; Mourelatos, 2008; Nicoloso and Calin, 2008;
Papagiannakopoulos and Kosik, 2008; Silber et al., 2008; Hobert, 2008;Makeyev
and Maniatis, 2008; Williams et al., 2008; Zeng, 2009, Lukiw et al., 2009).
Current studies indicate that specific miRNAs may function at multiple hierar-
chical levels in gene regulatory networks, from targeting hundreds of effector genes
to controlling the levels of global regulators of transcription and alternative pre-
mRNA splicing (Cao et al., 2006; Makeyev and Maniatis, 2008; Silber et al., 2008).
An expanding number of miRNAs have been reported to be altered in abundance
in glioma and GBM, and largely because of their disease-related expression and
selection of specific mRNA targets in the brain, these miRNAs are strongly impli-
cated as important regulatory controls in neoplastic onset and evolution. In general,
abrogation of global miRNA-mediated mRNA processing and homeostatic control
is associated with accelerated cellular transformation and tumorigenesis, and some
specific examples are given below (Lukiw, 2004; Pogue and Lukiw, 2004; Lukiw
and Bazan, 2006; Kumar et al., 2007; Lukiw and Bazan, 2008; Lukiw, 2009; Zeng,
2009).
Several decreased or increased miRNA species implicated in miRNA-mediated
brain cell tumor growth, oncogenesis, apoptosis, and survival (sometimes referred
to as oncomirs) are miRNA-124 and miRNA-137 (Ciafrè et al., 2005; Cho, 2007;
Silber et al., 2008; Papagiannakopoulos and Kosik, 2008; unpublished observa-
tions). In one recent study the expression levels of miRNA-124 and miRNA-137
were found to be significantly decreased in anaplastic astrocytoma (WHO grade
III) and GBM (WHO grade IV) relative to nonneoplastic control tissue (Silber
et al., 2008; Papagiannakopoulos and Kosik, 2008). Interestingly, when miRNA-
124 was introduced into nonneuronal mammalian cells a preferential reduction in

the amounts of nonneuronal mRNAs, including those encoding protein required
for cell proliferation or neural stem cell function was observed, and promotion
of a neuronal-like mRNA profile (Conaco et al., 2006; Makeyev et al., 2007).
Conversely, a depletion of miRNA-124 from primary neurons accumulated a
number of nonneuronal mRNA targets, suggesting that miRNA-124 ensures that
progenitor genes are posttranscriptionally inhibited in neurons (Makeyev et al.,
2007; Cao et al., 2007). Such evidence suggests the roles of miRNAs are in con-
trolling cell fate and the proliferating capacity of brain cells. That miRNA-124 and
miRNA-137 induce differentiation of adult neural stem cells, oligodendroglioma-
derived stem cells, and human GBM-derived stem cells and induce cell cycle arrest
in GBM suggests that targeted delivery of these highly soluble and mobile small
RNAs to glioma and GBM cells may provide an efficacious and novel therapeutic
treatment strategy for containing the growth of cancerous brain cells (Silber et al.,
2008; unpublished observations).
Genetic Signaling in GBM 813
Another miRNA species implicated in cell tumor growth, oncogenesis, apoptosis,
and survival is miRNA-221 (Ciafrè et al., 2005; Gillies and Lorimer, 2007; Medina
et al., 2008; Lukiw et al., 2009). Support for the pathogenic role of miRNA-221
in tumor growth comes from the recent observations that upregulated miRNA-221
targets the cell growth suppressive cyclin-dependent kinase inhibitors p27 and p57,
thus linking a cell-cycle checkpoint at S phase initiation with growth factors that
trigger cell proliferation (Li et al., 1999; le Sage et al., 2007; Mellai et al., 2008;
Medina et al., 2008; Lukiw et al., 2009). Other recent work reported a selective
upregulation of miRNA-221 and downregulation of a miRNA-221 messenger RNA
target encoding the survivin-1 homologue BIRC1, a neuronal inhibitor of apopto-
sis protein and brain cell marker for neural degeneration (Lukiw et al., 2009). In
these later studies the expression of BIRC5 (survivin-1) and caspase-3 was found
to be significantly upregulated, particularly in the more advanced stages of GBM.
It is important to note that paracrine signaling between adjacent brain cells may
contribute to significant positive feedback regulation and the progressive intercellu-

lar proliferation of pathogenic signaling in both degenerating brain cells and brain
tumors (Zhao et al., 2006; Culicchia et al., 2008). Indeed, tumor invasion occurs not
only through dysfunction of the adhesive properties of tumor cells but also in their
pathogenic secretion of small lysosomal proteolytic enzymes such as cathepsin-L
(Levicar et al., 2002, 2003). Use of online accessible miRNA–mRNA database
searches, other miRNA-221-targeted components of apoptotic signaling in glioma
and GBM, and interactions with the Bcl-2 protein family of apoptosis include anti-
apoptotic protein Bcl-2-binding component 3 and other Bcl-2-modifying factors
(Sanger mirBase version 10.1; hence miRNA-
221 may further modulate apoptotic signaling via quenching or augmentation of the
expression of a number of alternate antiapoptotic mRNA targets, such as additional
Bcl-2-modifing factors. Again the small size and high solubility of specific brain-
enriched miRNA species suggests that they may perform ancillary intracellular and
extracellular signaling roles involved in the spreading and propagation of tumor cell
growth and associated metastatic events (Lukiw, 2007; Lukiw and Pogue, 2007;
Felicetti et al., 2008; Mattick et al., 2009). More recently, miRNA-125b has been
shown to be upregulated in interleukin-6 (IL-6)-stressed normal human astrocytes
(NHA), a treatment known to induce astrogliosis, and in vitro, anti-miRNA-125b
added exogenously to IL-6-stressed NHA cultures attenuated both glial cell prolif-
eration and increased the expression of the cyclin-dependent kinase inhibitor 2A
(CDKN2A), a miRNA-125b target, and negative regulator of cell growth (Pogue
et al., 2010).
8 Therapeutic Strategies for the Clinical Management
of Glioma and GBM
Current treatment strategies for GBM are multimodal and typically involve sur-
gical resection followed by radiation therapy and chemotherapy. Upon tumor
recurrence repeat resection or stereotactic radiosurgery followed by additional
814 W.J. Lukiw and F. Culicchia
radiotherapy and chemotherapy may improve outcome in certain cases; several
new strategies have been developed to optimize designer therapies for GBM

(Sathornsumetee and Rich, 2008; Sathornsumetee and Reardon, 2009; Tentori and
Graziani, 2009). The most commonly used chemotherapeutic drug for GBM, temo-
zolomide (TMZ), typically administered both during and after radiotherapy, is a
potent DNA methylating agent that generates a wide s pectrum of random methyl
adducts in the genome. The antitumor activity of TMZ and related alkylating
agents has been mainly attributed to the production of O(6)-methylguanine as
a potent cytotoxic and antimitotic (Tentori and Graziani, 2009). TMZ also pro-
motes autophagic cell death, a caspase-independent process characterized by the
accumulation of cytoplasmic autophagic vacuoles and accompanied by extensive
degradation of polyribosomes, the endoplasmic reticulum, and the Golgi apparatus
that precedes the destruction of the nucleus (Lefranc et al., 2007). As brain neo-
plasms are generally associated with altered βAPP, pentraxin, caspase-3, VEGF
expression, and the kinases that modify these effector molecules, co-ordinated
inhibition of these oncogenic markers might be an effective therapeutic strat-
egy. These kinds of treatment approaches have recently been reviewed (Anderson
et al., 2008; Lakka and Rao, 2008; Hide et al., 2008; Norden et al., 2008). Anti-
inflammatory, anti-βAPP, and antiamyloid pharmacologic strategies directed at
neurodegenerative processes may also have some therapeutic value in the treat-
ment of glioma- or glioblastoma-affected brain cells (Nelson, 2002; Lukiw and
Bazan, 2006; Lukiw and Bazan, 2008; Tschape and Hartmann, 2008; Culicchia
et al., 2008).
Unfortunately, chemotherapeutic drug resistance occurs relatively often and
effective drug delivery to cancer targets remains an accessory concern affecting the
clinical response in brain cancer patients. Because malignant gliomas are highly
vascularized tumors that produce VEGF, a key mediator of angiogenesis, and given
the fact that angiogenesis is essential for the proliferation and survival of malignant
glioma cells, angiogenesis antagonists such as angiostatin, endostatin, and vaso-
statin may provide yet another specifically targeted, therapeutic strategy. Recent
studies have investigated the use of bevacizumab—a humanized monoclonal anti-
body against VEGF—for patients with recurrent malignant glioma, however, the

results have been inconsistent, and larger, randomized clinical trials are needed
to determine the magnitude of the benefit (Buie and Valgus, 2008; Norden et al.,
2008). Moreover, angiogenesis antagonists have numerous unwanted side effects
in interfering with normal wound healing, bleeding, and blood clotting, and are
associated with heart, immune, and reproductive dysfunction (Norden et al., 2008).
Interestingly, gamma- and beta-secretases that act on βAPP processing appear to
play an essential role during angiogenesis, and inhibitors of these secretases may
constitute a novel evolving class of antiangiogenic and antitumoral compounds
(Paris et al., 2005).
Just as for angiogenesis antagonists, toxic metal-based anticancer drugs
(MBADs), including cisplatin, carboplatin, and oxaliplatin, and other arsenic-,
cadmium-, copper-, gallium-, lanthanum-, platinum-, ruthenium-, or titanium-
containing antitumor drug complexes have adverse effects on physiological systems