J Neurosurg 114:663–669, 2011

Cadherin-dependent adhesion of human U373MG

glioblastoma cells promotes neurite outgrowth
and increases migratory capacity

Laboratory investigation
Christopher P. Cifarelli, M.D., Ph.D., Brian Titus, M.D., Ph.D.,
and Hian Kwang Yeoh, M.B.Ch.B.

Department of Neurological Surgery, University of Virginia, Charlottesville, Virginia

Object. The current management of primary CNS tumors involves a multimodal approach, incorporating cytore-
ductive techniques including resection, radiotherapy, and antiproliferative chemotherapeutic agents. Despite these
attempts, the majority of patients with a diagnosis of a high-grade glioma have a dismal prognosis, with the leading
cause of treatment failure and tumor recurrence attributable to local invasion of adjacent brain parenchyma. The cur-
rent study examines the capacity of glioma tumor cells to undergo neurite outgrowth and local migration, specifically
focusing on the role of the cadherin cell adhesion system.
Methods. Using a recombinant cadherin ectodomain protein, U373MG human glioblastoma cells were assessed
for their ability to adhere and migrate in a cadherin-dependent manner in culture. Adhesion was evaluated via growth
assessment and neurite length at 72 hours growth on an immobilized cadherin substrate and compared with other
matrix adhesion proteins, such as Type IV collagen and vitronectin. Migratory capacity was measured via modified
transwell assays, also using recombinant cadherin ectodomain in comparison with collagen and vitronectin.
Results. Cadherin adherent cells adopt a fasciculated morphology, with a significant increase in neurite exten-
sion, measuring 104 ± 13.3 μm in length, compared with background adhesion on bovine serum albumin and non-
functional cadherin ectodomain controls measuring 55 ± 4.4 and 47 ± 3.84 μm, respectively (p = 0.029). Significant
increases in neurite length compared with controls were also observed in the vitronectin (81 ± 4.69 μm) and Type IV
collagen (91 ± 7.7 μm) groups (p = 0.017 and 0.025, respectively). With respect to migration, U373 cells demonstrate
increased invasion in response to cadherin ectodomain exposure, whereas vitronectin and Type IV collagen were not
potent initiators of migration through the transwell barrier. Both adhesion and migration outcomes were noted in the
absence of any relative changes in cell proliferation, indicating a primary role for the cadherin-based adhesion system
in tumor invasion.
Conclusions. Cadherin-based adhesion promotes increased adhesion, neurite outgrowth, and migration in hu-
man U373MG glioblastoma cells, providing a novel area of research for the development of therapeutic targets ad-
dressing local tumor invasion. (DOI: 10.3171/2010.3.JNS091451)

Key Words      •      cadherin      •      migration      •      adhesion      •      glioblastoma      •

neurite      •      U373MG cell

P
rimary tumors of the CNS account for approximate-
ly 2% of all adult cancer diagnoses, with approxi-
mately 20,500 new cases of malignant CNS cancer
diagnosed annually.2 Of these, approximately 30% are
found to be invasive high-grade tumors (WHO Grade III/
IV) at the time of diagnosis.7 The current treatment para-
digm for a newly diagnosed high-grade glioma involves
a patient-specific combination of biopsy and/or resection
followed by postoperative chemoradiotherapy.31 Despite
this intervention, the prognosis remains dismal, with the
Abbreviations used in this paper: BSA = bovine serum albumin;
DMEM = Dulbecco modified eagle medium; PBS = phosphate-
buffered saline
leading cause of treatment failure and tumor recurrence
attributable to local invasion of adjacent brain parenchy-
ma.12 Currently, adjuvant chemobiological therapy for
gliomas does not directly target the ability of tumor cells
to migrate and seed surrounding tissues, despite mount-
ing evidence that increased expression of neural migra-
tory markers in subsets of patients with glioblastoma pre-
dict more rapid disease progression.22,36
A substantial body of evidence exists demonstrating
the importance of several cell adhesion molecules as well
as extracellular matrix proteins in the mediation of cell
migration during neural development, as well as in onco-
genic transformation and CNS tumor invasion.9,34 Among
the most widely studied cell adhesion molecules, cad-
herins exist as single-pass transmembrane glycoproteins

J Neurosurg / Volume 114 / March 2011 663


that mediate Ca2+-dependent cell adhesion as well as cell
signaling events through the Wnt pathway.29,33 Of specific
interest with respect to the CNS is neural (N-) cadherin,
which has been shown to be integral in the process of
neurite outgrowth and neural development.6,27,37,38 With
respect to neurooncology, the role of cadherins has been
enigmatic. Several investigative groups have attempted to
correlate a direct relationship between cadherin expres-
sion and degree of tumor invasion with limited success.3,30
This finding is further supported by Perego et al.24 who
demonstrated that qualitative disruption of the cadherin-
catenin complex in glioblastoma cell lines is responsible
for increased tumor invasion rather than quantitative dif-
ferences in expression levels.
In the present study, the role of cadherin-mediated
adhesion is explored using the U373 glioma cell line to
examine neurite extension and cell invasion. Using a re-
combinant cadherin ectodomain in an in vitro system, the
data indicate that exposure of U373 cells to the cadherin
binding ectodomain is sufficient to promote adhesion and
a relative increase in cell invasion in comparison with
other extracellular matrix adhesive substrates. Moreover,
this occurs independent of a change in the proliferation
index of the cells with respect to the various adhesive
substrates.
Methods
Cell Culture and Cadherin Expression
The U373MG and U87MG cells were acquired from
the American Type Culture Collection (ATCC) and
maintained in culture using DMEM (Sigma) containing
10% fetal bovine serum, and 1% penicillin, streptomycin,
and amphotericin B (GibcoBRL) at 37°C in 5% CO2. Ex-
pression of cadherins in U373MG and U87MG cells was
determined by sodium dodecyl sulfate–polyacrylamide
gel electrophoresis and immunoblot. Briefly, confluent
U373MG and U87MG cells were treated with radioim-
munoprecipitation assay extraction buffer at 4°C for 1
hour. Soluble protein was prepared by centrifugation at
13,000 G with retention of the supernatant. Protein con-
centration was determined according to the bicinchoninic
acid protein determination assay (Pierce Chemical Co.).
The 50-µg samples were resolved using sodium dodecyl
sulfate–polyacrylamide gel electrophoresis on a 7.5%
acrylamide gel and transferred to nitrocellulose for im-
munoblotting. Western blot analysis was performed using
the pancadherin rabbit polyclonal antibody (1:5000; Neo-
markers) in 5% dry milk in Tris-buffered saline as a pri-
mary antibody and a horseradish peroxidase-conjugated
goat antirabbit immunoglobulin G (Sigma-Aldrich) sec-
ondary antibody (1:10,000) in PBS. Blots were developed
with the WestDura SuperSignal reagent (Pierce Chemical
Co.) on Biomax MR film (Kodak).
Cadherin-Fusion Protein Preparation
Recombinant C-cadherin-Fc fusion protein consist-
ing of the complete extracellular adhesive EC1 to EC5
domains was prepared according to Chappuis-Flament et
al.8 Briefly, Chinese hamster ovary cells stably transfected
664
C. P. Cifarelli, B. Titus, and H. K. Yeoh
with the XC-cadherin-Fc/pEE14 vector were cultured in
complete Glasgow minimum essential medium (Sigma-
Aldrich) and selected for the presence of glutamine syn-
thetase on pEE14 in 25-µM methionine sulfoximine (Sig-
ma-Aldrich) until confluent. The conditioned media was
then centrifuged and purified on an affinity column using
Protein A Sepharose (Repligen Corp.). The EC1–5-Fc fu-
sion protein was eluted with 150-mM glycine (pH 2.0)
in 1-ml fractions. Each tube was immediately neutralized
with 1.5 M Tris-HCl (pH 8.8). Absorbance at 280 nm was
determined in each elution fraction and aliquots contain-
ing protein were pooled and desalted on a G-25 Sephadex
column (J.T. Baker, Inc.). Final protein concentration was
determined by A280 on a NanoDrop 1000 spectrophotom-
eter (Thermo Scientific). The EC1–5-Fc Trp mutant con-
struct was prepared according to the same protocol.
Cadherin-Dependent Adhesion Assay
A solution of nitrocellulose (0.5 cm2/ml methanol)
was pipetted onto a sterile 100-mm untreated culture dish
and allowed to air dry as modified from the technique de-
scribed by Lagenaur and Lemmon.18 The 10-µg aliquots
of EC1–5-Fc, EC1–5-Fc Trp mutant, Type IV collagen
(Cohesion), and vitronectin (Sigma-Aldrich) were applied
in spots to the nitrocellulose-coated dish. After 3 minutes
at room temperature, the dishes were rinsed 2 times with
5 ml of Hank’s buffered saline solution (Sigma-Aldrich).
To block nonspecific binding, 5 ml of a solution of 1%
BSA in PBS (pH 7.4) was applied to each plate and incu-
bated for 1 hour at 37°C in 5% CO2. The dishes were then
rinsed 2 times with Hank’s buffered saline solution and
once with DMEM. One hundred million (108) cells were
applied to each dish in 5-ml DMEM containing 1% peni-
cillin/streptomycin/amphotericin B and incubated at 37°C
in 5% CO2. At 24 hours, media was exchanged and non-
adherent cells were removed from culture. Adhesion was
documented via light microscopy using a Zeiss Axiovert
40 light microscope (Carl Zeiss). Morphological analysis
was performed using Image J Image Processing (NIH).1
Statistical analyses, including the 2-tailed Student t-test,
were performed using S-Plus 8.0 for Windows (Insightful
Corporation). A probability value < 0.05 was considered
statistically significant.
Cadherin-Mediated Invasion Assay
The 10-µg aliquots of each adhesion substrate (EC1–
5-Fc, EC1–5-Fc-Trp mutant, Type IV collagen, vitronec-
tin, and BSA) were sterile pipetted onto the undersurface
of 25-mm diameter (8.0-µm pore size) Transwell inserts
(NalgeNunc) and allowed to air dry over 2 hours. The
inserts were then placed in 6-well culture dishes (Costar,
Corning). Two milliliters of DMEM with 10% fetal bo-
vine serum and 1% penicillin/streptomycin/amphotericin
B was placed in the lower chamber and 2 ml of serum-
free DMEM with 1% penicillin/streptomycin/amphoteri-
cin B was pipetted into the upper chamber. Aliquots of
3 × 104 U373MG astrocytoma cells in DMEM with 1%
penicillin/streptomycin/amphotericin B were placed into
the each upper chamber and incubated at 37°C in 5% CO2
and maintained for 96 hours. Each invasion assay was
J Neurosurg / Volume 114 / March 2011
Cadherin-dependent migration of glioblastoma cells

performed in triplicate. At the end of 96 hours, the inferi-
or surface of the transwell membrane was rinsed with the
lower chamber media to form the pooled lower chamber
as described by Kleinman and Jacob.17 Cell counts were
performed via 0.4% trypan blue exclusion assay using the
lens with a magnification of 10 on a Nikon light micro-
scope. Statistical analyses, including the 2-tailed Student
t-test, were performed using S-Plus 8.0 for Windows.
Results
U373MG Cells Express Cadherins in Culture
As previously described,30 our U373MG cells consis-
tently expressed cadherins based on immunoblotting for
the cadherin cytoplasmic domain (Fig. 1A). Moreover, the
expression of cadherin-reactive bands was substantially
greater in the U373MG cells when compared with equal-
ly confluent U87MG positive control cultures at Day 14,
suggesting either an increased expression level in U373
cells or relatively increased proteolysis of cadherins in
the U87 cell line. Regardless, the constitutive expression
of cadherin subtypes on the cell surface of the U373 cells
throughout the culture period made the cell line suitable
for examining cadherin-mediated neurite growth and cell
migration.
Cadherin Adherent Cells Adopt Fasciculated Morphology
The U373 adhesion cultures were continued for 14
days, with documentation of neurite outgrowth and cell
morphology on Day 3. The U373 cells bind to an EC1–
5-coated culture dish, with interface between cadherin-
mediated and nonspecific BSA adhesion demarcated by
a concentric border (Fig. 1B). Here, the EC1–5 recom-
binant protein serves to mimic the extracellular binding
domain of the classic cadherin subtype, providing a sys-
tem of purely cadherin-dependent adhesion. Morphologi-

Fig. 1.  Cadherin expression, neurite extension, and morphological analysis of U373 cells.  A: Western blot showing cadherin
expression in U87 and U373 cell lines. A 120–130 kDa pancadherin reactive band (arrow) is present in both U87 and U373 ra-
dioimmunoprecipitation assay extracted fractions, with significantly higher reactivity for the full-length protein in the U373 cell ex-
tract. The relative increased proportion of the 60–80 kDa fragment in the U87 cells may represent increased baseline proteolysis
compared with the U373 cells.  B: Phase-contrast microscopic image showing immobilized cadherin-dependent adhesion and
neurite extension. The U373 cells cultured for 3 days on nitrocellulose bound recombinant EC1–5 cadherin fusion protein (right)
demonstrate qualitative increases in cell spreading and neurite length compared with cells grown on BSA-coated surface (left)
as demarcated by the dashed line.  C: Cell morphology analysis of cadherin-based adhesion compared with controls. The U373
cells grown in adhesion culture on either recombinant EC1–5 or BSA were categorized based on growth patterns as cell clusters
(≥ 2 cells) versus single cells, with a significant (*p = 0.0005) increase in single cell growth in the cadherin-adherent cells.  D:
Proliferation of cadherin-adherent cells. The U373 cells grown on immobilized cadherin ectodomain do not demonstrate a signifi-
cant increase in cell proliferation compared with controls grown on BSA-coated culture dishes (p = 0.118).

J Neurosurg / Volume 114 / March 2011 665


cal comparison between background and EC1–5-bound
cells demonstrates a significant (p = 0.0005) increase in
single cell growth in the cadherin fields, with 81 ± 1.8%
(mean ± SEM) of cadherin adherent cells growing as sin-
gle cells (Fig. 1C). Conversely, 82 ± 2.8% of BSA-bound
U373 cells remain in clusters (Fig. 1D). This increased
proportion of single cells spreading on the cadherin ecto­
domain does not appear to be due to any relative change
in proliferation between control (116 ± 6.92 cells/hpf) and
cadherin adherent cells (136 ± 14.3 cells/hpf; p = 0.118;
Fig. 1D.) After 14 days in culture, the U373 cells remain
adherent to the recombinant cadherin ectodomain and de-
velop a tightly packed fasciculated appearance with neu-
ritic processes aligned in parallel, in distinct comparison
with BSA-bound cells with retained clusters and radially
oriented neurite projections (Fig. 2).
Immobilized Recombinant Cadherin Ectodomain Promotes
Neurite Extension in Vitro
At Day 3 of adhesion culture using nitrocellulose-
bound 10-µg aliquots of either recombinant cadherin
ectodomain, vitronectin, Type IV collagen, BSA, or
the nonadhesive recombinant EC1–5 Trp mutant (nega-
tive control for the Fc-fusion protein fragment), neurite
extensions were measured by stereological analysis, as
described by Rønn et al.28 The EC1–5 Trp mutant con-
trol, previously described by Boggon et al., 5 serves as the
most precise means of delineating the role of cadherin
adhesion in our model system, as the protein fragment
is identical to the EC1–5 cadherin ectodomain with the
exception of a single mutation, rendering the adhesive
properties nonfunctional. Background adhesion, as noted
in BSA and nonfunctional cadherin ectodomain (EC1–5
mut), demonstrated neurites measuring 55 ± 4.4 µm and
47 ± 3.84 µm, respectively (Fig. 3). Significant increas-
es in neurite length over controls were observed in all
groups, including vitronectin (81 ± 4.69 µm; p = 0.017),
Type IV collagen (91 ± 7.7 µm; p = 0.025), and recombi-
nant cadherin EC1–5 (104 ± 13.3 µm; p = 0.029), with the
greatest increase in EC1–5 and Type IV collagen adher-
ent cells (Fig. 3).
U373 Cells Demonstrate Increased Invasion in Response to
Cadherin Ectodomain Exposure
The ability of U373 cells to traverse an artificial
membrane coated with recombinant cadherin ectodo-
main, vitronectin, Type IV collagen, BSA, or the non-
adhesive recombinant EC1–5 Trp mutant was assessed.
Noncoated control membranes were treated with PBS
only. Cell counts of the lower chamber, including those
cells able to be gently rinsed from the inferior surface of
the membrane, demonstrated a significant increase in the
relative invasion in cadherin ectodomain exposure cells
(Fig. 4). At 96 hours incubation, 528 ± 24.7 cells (p =
0.00023) in the EC1–5 wells had migrated through the
coated membrane, compared with 155 ± 30 cells in the
BSA controls and 100 ± 17.4 in the EC1–5 mutant nega-
tive controls. Surprisingly, Type IV collagen served as an
impediment to invasion of the U373 cells, with only 67 ±
7.5 lower chamber cells noted at the end point of the assay
666
C. P. Cifarelli, B. Titus, and H. K. Yeoh
Fig. 2.  Microscopic image showing U373 cells grown on cadherin
ectodomain substrate adopt a fasciculated appearance. At 14 days in
culture, the U373 cells grown on the recombinant ectodomain fragment
appear densely packed with parallel alignment, compared with the BSA
adherent controls than maintain a loosely defasciculated growth pat-
tern. Original magnification × 40.
(Fig. 4B), but served as a substrate with significant cell
proliferation (Fig. 4C). Conversely, the EC1–5 substrate
did not produce an increase in cell number in comparison
with BSA controls (Fig. 4C).
Discussion
Significant improvements are still needed in the
multimodal treatment of glioblastoma. With respect to
using tumor cell migration as a potential therapeutic tar-
get, several studies have demonstrated that the majority
of disease progression in glioblastoma treated with the
standard doses of concurrent radiotherapy and temozolo-
mide tends to be local, usually within 2 cm of the initial
tumor rather than at distant sites.13,32 Local therapies, such
as interstitial carmustine, have attempted to use this find-
ing by providing increased antiproliferative effects at the
site of the resection margin with only a subtle increase
in survival and no change in local recurrence pattern.13
Specific adjuvant biological agents aimed at inhibiting
migration of tumor cells have yet to be tested within the
current treatment paradigm, but provide promising tar-
gets based on expression profiles and clinical progression
analysis.10,19,36
Cadherin expression, including classic members of
the cadherin superfamily such as E-, N-, and P-cadherin
has been shown in various glioma cell lines.3,24,30 In the
current study, we employ the technique of providing a
cadherin-specific extracellular adhesive domain to ana-
lyze both adhesion and migration via incorporation of
the recombinant EC1–5 protein as a binding substrate.
Our negative control, the recombinant EC1–5 Trp mutant
protein, is identical to the EC1–5 protein, with the ex-
ception of a single mutation within the primary cadherin
adhesion domain, thereby creating a specific negative
control for our system. As the cadherin adhesion com-
plex includes several distinct adapter proteins linking the
cadherin molecule to the actin cytoskeleton, controlled
expression of various cadherin molecules into a cadher-
J Neurosurg / Volume 114 / March 2011
Cadherin-dependent migration of glioblastoma cells

Fig. 3.  Graph showing immobilized recombinant cadherin ectodo-

main promotes neurite extension. The U373 cells grown for 3 days on
immobilized EC1–5, as well as the nonadhesive EC1–5 mutant control,
were analyzed for neurite growth in comparison with Type IV collagen,
vitronectin, and BSA controls. Significant increases in neurite length
were observed in Type IV collagen, vitronectin, and EC1–5 (cadherin)
based adhesion (p < 0.05 in all groups) in comparison with BSA and the
EC1–5 negative control. *p = 0.017, **p = 0.029, ***p = 0.025.

in-free cell system creates the obstacle of controlling for

the appropriate coexpression of these other proteins, in-
cluding β-catenin, whose role as a potent mediator of the
Wnt signaling cascade would severely hinder the techni-
cal feasibility of the study. Moreover, to study cadherin
function in a model system devoid of cadherins, or where
cadherin expression levels were natively low, would limit
any conclusions that could be drawn with respect to the
identity of therapeutic targets.
Unlike other prominent biomarkers of disease, such
as p53 mutations or mitotic index as determined by
MIB-1 staining, cadherin expression is not as easily cor-
related with disease progression. Currently, only breast
carcinoma and bladder cancer employ cadherin expres-
sion, or the presence of cadherin proteolysis fragments,
in prognostication, and even in these situations no causal
relationships exist.4,11,16,21 The expression of cadherin cell
adhesion molecules in tumors of the CNS is not a novel
concept, but the precise role that cadherins play in the
pathogenesis of CNS lesions is still the subject of consid-
erable debate. The examination of expression profiles of
cadherins in tissue sections from invasive tumors has led Fig. 4.  Analysis of the ability of U373 cells to traverse an artificial
several investigators to conclude that N-cadherin expres- membrane.  A: Transwell invasion assay. The U373 cells appearance
sion is correlated with increased invasion, while others as grown on cadherin-coated transwell membranes with an 8.0-µm
have noted the converse to be true.3,14 Similarly, loss of E- pore size display similar growth patterns to those grown on cadherin-
coated culture dishes at 3 days of culture.  Crystal violet, original mag-
cadherin cell surface expression is believed to be a central nification × 40.  B: Graph of U373 cell migration. The cadherin ect-
event in the epithelial-mesenchyme transformations asso- odomain coated transwell membranes promote a significant increase
ciated with tumor invasion in several cancers.15,25 Aside in migratory capacity as measured by the number of cells passing
from tissue- and organ-specific differences accounting through into the lower chamber. Cadherin-adherent cells demonstrate
for such incongruities, the conceptualization of cadherins nearly a 3-fold increase in migration through the transwell membrane
as static adhesion proteins in these studies is likely to be compared with both the controls and the Type IV collagen substrate (*p
the greatest obstacle in delineating the role of cadherins = 0.00023).  C: Transwell migration is not a function of cell prolifera-
in tumor pathogenesis. tion in U373 cells. Analysis of the number of cells remaining bound to
the transwell membrane following 3 days of culture does demonstrate
Our data support the concept that cadherin binding a significant relationship between migratory capacity and proliferation
promotes both adhesion and migration as defined by the with respect to cadherin-based adhesion (p = 0.088), but did reveal a
ability of the U373 cells to spread as large fascicles and significant increase in Type IV collagen-mediated proliferation (**p =
traverse a cadherin-coated membrane. Earlier work uti- 0.002). *p = 0.02.

J Neurosurg / Volume 114 / March 2011 667


lizing immobilized cadherin ectodomain has also shown
similar neurite outgrowth patterns in developing nervous
tissue, including embryonic avian retina,23 but this is the
first study to demonstrate these findings in glioma cell
populations. Moreover, the ability of the recombinant
cadherin ectodomain to promote migratory capacity
without significant changes in cell proliferation implies
that a phenotypic change is responsible for the increased
mobility. The precise mechanism underlying this find-
ing remains unknown. The varied nature of cadherin
function within the cell during development, in adult tis-
sue, and, by extension, in pathological states such as tu-
mor cell migration, make it difficult to identify a single
mechanism whereby cadherin-mediated adhesion could
significantly affect migratory capacity in an oncogenic
population. As a mediator of homophilic cell-cell bind-
ing, its role as an adhesion molecule can certainly explain
the substrate binding via the immobilized recombinant
ectodomain, but the ability of the U373 cells exposed to
cadherin-binding sites to migrate through the transwell
membrane and, thus, past the binding substrate implies
that cadherin exposure in this glioma cell model mediates
a more significant change in cell behavior. In comparison
with known adhesive substrates, such as vitronectin and
collagen, where integrin heterodimers effectively bind
the cell to the extracellular matrix protein and promote
significant neurite growth, cadherin adhesion binds and
promotes neurite growth as well as invasion.
The dynamic nature of cadherin adhesion at the de-
veloping growth cone and, in the current study, at the site
of neurite extension in migrating CNS tumor cell popula-
tions is underscored by recent studies that have focused
on mediators of cadherin proteolysis. Membrane-bound
members of the family of matrix metalloproteases, known
as adamalysins, are believed to function by cleaving the
cadherin ectodomain, specifically ADAM10 in the pro-
teolysis of E-cadherin and N-cadherin.20,26,35 By cleaving
the extracellular cadherin domain, the adhesive junction is
disrupted and the migrating cell border is free to advance,
adding a needed element of plasticity to a seemingly stat-
ic system of cell-cell adhesion. As this field continues to
develop, specific inhibitors directed at this process may
serve as potential therapeutic agents for blocking local
tumor invasion following surgical intervention.
The heterogenous nature of the extracellular milieu
creates a substantial obstacle for the development of novel
therapeutics as no single target is likely to be ideal for
control of local tumor invasion. In the current study, we
attempted to define the specific role of cadherins in migra-
tion and invasion, and have shown that human U373MG
glioblastoma cells demonstrate increased adhesion, neu-
rite outgrowth, and chemotaxis in a cadherin-dependent
in vitro assay. What remains to be further elucidated are
the numerous signaling pathways that are activated in the
in vivo response to the local tumor microenvironment,
including nontumor cell-cell interactions, extracellular
matrix proteins, additional cell adhesion molecules, as
well as proteolytic elements. Certainly, future studies will
need to incorporate animal models in an effort to address
these critical questions prior to exploring any therapeutic
options. Yet given the current findings, we believe that
668
C. P. Cifarelli, B. Titus, and H. K. Yeoh
the disruption of tumor cell invasion via manipulation of
cadherin-dependent cell adhesion and migration will be a
viable avenue for future preclinical trials.
Conclusions
Based on our findings, we conclude that cadherin-
mediated cell adhesion promotes increased cell-surface
binding, neurite outgrowth, and migration in human
U373MG glioblastoma cells. Moreover, we believe that
these data provide a novel area of research for the de-
velopment of therapeutic targets to address mechanisms
of local tumor invasion in the treatment of high-grade
gliomas.
Disclosure
The authors report no conflict of interest concerning the mate-
rials or methods used in this study or the findings specified in this
paper.
Author contributions to the study and manuscript preparation
include the following. Conception and design: Cifarelli. Acquisition
of data: Cifarelli, Titus, Yeoh. Analysis and interpretation of data:
Cifarelli. Drafting the article: Cifarelli. Critically revising the article:
Cifarelli, Titus. Reviewed final version of the manuscript and
approved it for submission: Cifarelli, Titus, Yeoh. Statistical analy-
sis: Cifarelli. Study supervision: Cifarelli.
Acknowledgments
The authors wish to thank Dr. Barry M. Gumbiner and Mar­
tha­joy Spano for the reagents and technical advice and Drs. Jason P.
Sheehan and Gregory A. Helm for their continued support.
References
  1.  Abramoff MD, Magelhaes PJ, Ram SJ: Image processing with
ImageJ. Biophotonics Int 11:36–42, 2004
  2.  American Cancer Society: Cancer Facts & Figures 2008.
(http://www.cancer.org/docroot/stt/content/stt_1x_cancer_
facts_and_figures_2008.asp) [Accessed March 25, 2010]
  3.  Asano K, Duntsch CD, Zhou Q, Weimar JD, Bordelon D, Rob-
ertson JH, et al: Correlation of N-cadherin expression in high
grade gliomas with tissue invasion. J Neurooncol 70:3–15,
2004
  4.  Baumgart E, Cohen MS, Silva Neto B, Jacobs MA, Wotko-
wicz C, Rieger-Christ KM, et al: Identification and prognos-
tic significance of an epithelial-mesenchymal transition ex-
pression profile in human bladder tumors. Clin Cancer Res
13:1685–1694, 2007
  5.  Boggon TJ, Murray J, Chappuis-Flament S, Wong E, Gumbin-
er BM, Shapiro L: C-cadherin ectodomain structure and im-
plications for cell adhesion mechanisms. Science 296:1308–
1313, 2002
  6.  Burden-Gulley SM, Payne HR, Lemmon V: Growth cones are
actively influenced by substrate-bound adhesion molecules. J
Neurosci 15:4370–4381, 1995
  7.  Central Brain Tumor Registry of the United States: CBTRUS
Statistical Report: Primary Brain Tumors in the United
States, 2000–2004. Central Brain Tumor Registry of the
United States, 2008
  8.  Chappuis-Flament S, Wong E, Hicks LD, Kay CM, Gum-
biner BM: Multiple cadherin extracellular repeats mediate
homophilic binding and adhesion. J Cell Biol 154:231–243,
2001
  9.  Couldwell WT, de Tribolet N, Antel JP, Gauthier T, Kuppner
J Neurosurg / Volume 114 / March 2011
Cadherin-dependent migration of glioblastoma cells
MC: Adhesion molecules and malignant gliomas: implica-
tions for tumorigenesis. J Neurosurg 76:782–791, 1992
10.  de Groot J, Milano V: Improving the prognosis for patients
with glioblastoma: the rationale for targeting Src. J Neuroon-
col 95:151–163, 2009
11.  Gadducci A, Ferdeghini M, Cosio S, Annicchiarico C, Ciampi
B, Bianchi R, et al: Preoperative serum E-cadherin assay in
patients with ovarian carcinoma. Anticancer Res 19 (1B):
769–772, 1999
12.  Giese A, Bjerkvig R, Berens ME, Westphal M: Cost of migra-
tion: invasion of malignant gliomas and implications for treat-
ment. J Clin Oncol 21:1624–1636, 2003
13.  Giese A, Kucinski T, Knopp U, Goldbrunner R, Hamel W, Me-
hdorn HM, et al: Pattern of recurrence following local chemo-
therapy with biodegradable carmustine (BCNU) implants in
patients with glioblastoma. J Neurooncol 66:351–360, 2004
14.  Hazan RB, Phillips GR, Qiao RF, Norton L, Aaronson SA:
Exogenous expression of N-cadherin in breast cancer cells
induces cell migration, invasion, and metastasis. J Cell Biol
148:779–790, 2000
15.  Hirohashi S: Inactivation of the E-cadherin-mediated cell ad-
hesion system in human cancers. Am J Pathol 153:333–339,
1998
16.  Kashibuchi K, Tomita K, Schalken JA, Kume H, Takeuchi T,
Kitamura T: The prognostic value of E-cadherin, alpha-, beta-
and gamma-catenin in bladder cancer patients who underwent
radical cystectomy. Int J Urol 14:789–794, 2007
17.  Kleinman HK, Jacob K: Invasion assays. Curr Protoc Cell
Biol Chapter 12:Unit 12.2, 2001
18.  Lagenaur C, Lemmon V: An L1-like molecule, the 8D9 an-
tigen, is a potent substrate for neurite extension. Proc Natl
Acad Sci U S A 84:7753–7757, 1987
19.  Lu KV, Zhu S, Cvrljevic A, Huang TT, Sarkaria S, Ahkavan
D, et al: Fyn and SRC are effectors of oncogenic epidermal
growth factor receptor signaling in glioblastoma patients.
Cancer Res 69:6889–6898, 2009
20.  Maretzky T, Reiss K, Ludwig A, Buchholz J, Scholz F, Proksch
E, et al: ADAM10 mediates E-cadherin shedding and regu-
lates epithelial cell-cell adhesion, migration, and beta-catenin
translocation. Proc Natl Acad Sci U S A 102:9182–9187,
2005
21.  Mohammadizadeh F, Ghasemibasir H, Rajabi P, Naimi A,
Eftekhari A, Mesbah A: Correlation of E-cadherin expression
and routine immunohistochemistry panel in breast invasive
ductal carcinoma. Cancer Biomark 5:1–8, 2009
22.  Nakada M, Anderson EM, Demuth T, Nakada S, Reavie LB,
Drake KL, et al: The phosphorylation of ephrin-B2 ligand
promotes glioma cell migration and invasion. Int J Cancer
126:1155–1165, 2010
23.  Paradies NE, Grunwald GB: Purification and characterization
of NCAD90, a soluble endogenous form of N-cadherin, which
is generated by proteolysis during retinal development and
retains adhesive and neurite-promoting function. J Neurosci
Res 36:33–45, 1993
24.  Perego C, Vanoni C, Massari S, Raimondi A, Pola S, Catta-
neo MG, et al: Invasive behaviour of glioblastoma cell lines is
associated with altered organisation of the cadherin-catenin
adhesion system. J Cell Sci 115:3331–3340, 2002
J Neurosurg / Volume 114 / March 2011
25.  Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G: A
causal role for E-cadherin in the transition from adenoma to
carcinoma. Nature 392:190–193, 1998
26.  Reiss K, Maretzky T, Ludwig A, Tousseyn T, de Strooper B,
Hartmann D, et al: ADAM10 cleavage of N-cadherin and reg-
ulation of cell-cell adhesion and beta-catenin nuclear signal-
ling. EMBO J 24:742–752, 2005
27.  Riehl R, Johnson K, Bradley R, Grunwald GB, Cornel E, Lil-
ienbaum A, et al: Cadherin function is required for axon out-
growth in retinal ganglion cells in vivo. Neuron 17:837–848,
1996
28.  Rønn LC, Ralets I, Hartz BP, Bech M, Berezin A, Berezin
V, et al: A simple procedure for quantification of neurite out-
growth based on stereological principles. J Neurosci Meth-
ods 100:25–32, 2000
29.  Shapiro L, Love J, Colman DR: Adhesion molecules in the
nervous system: structural insights into function and diversity.
Annu Rev Neurosci 30:451–474, 2007
30.  Shinoura N, Paradies NE, Warnick RE, Chen H, Larson JJ,
Tew JJ, et al: Expression of N-cadherin and alpha-catenin in
astrocytomas and glioblastomas. Br J Cancer 72:627–633,
1995
31.  Stupp R, Hegi ME, Gilbert MR, Chakravarti A: Chemoradio-
therapy in malignant glioma: standard of care and future di-
rections. J Clin Oncol 25:4127–4136, 2007
32.  Stupp R, Hegi ME, van den Bent MJ, Mason WP, Weller M,
Mirimanoff RO, et al: Changing paradigms—an update on the
multidisciplinary management of malignant glioma. Oncolo-
gist 11:165–180, 2006
33.  Takeichi M: The cadherin superfamily in neuronal connec-
tions and interactions. Nat Rev Neurosci 8:11–20, 2007
34.  Tysnes BB, Mahesparan R: Biological mechanisms of glio-
ma invasion and potential therapeutic targets. J Neurooncol
53:129–147, 2001
35.  Uemura K, Kihara T, Kuzuya A, Okawa K, Nishimoto T, Ni-
nomiya H, et al: Characterization of sequential N-cadherin
cleavage by ADAM10 and PS1. Neurosci Lett 402:278–283,
2006
36.  Wang LF, Fokas E, Juricko J, You A, Rose F, Pagenstecher A,
et al: Increased expression of EphA7 correlates with adverse
outcome in primary and recurrent glioblastoma multiforme
patients. BMC Cancer 8:79, 2008
37.  Wanner IB, Guerra NK, Mahoney J, Kumar A, Wood PM,
Mirsky R, et al: Role of N-cadherin in Schwann cell precur-
sors of growing nerves. Glia 54:439–459, 2006
38.  Wanner IB, Wood PM: N-cadherin mediates axon-aligned
process growth and cell-cell interaction in rat Schwann cells.
J Neurosci 22:4066–4079, 2002
Manuscript submitted October 19, 2009.
Accepted March 22, 2010.
Please include this information when citing this paper: pub-
lished online April 23, 2010; DOI: 10.3171/2010.3.JNS091451.
Address correspondence to: Christopher P. Cifarelli, M.D., Ph.D.,
Department of Neurological Surgery, University of Virginia Health
System, P.O Box 800212, Charlottesville, Virginia 22908-0212.
email: cpcifarelli@virginia.edu.
669