Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Analysis of human brain endothelial cells versus human tumor associated brain endothelial cell tight junction and adherens junction expression differences in glioblastoma multiforme
(USC Thesis Other)
Analysis of human brain endothelial cells versus human tumor associated brain endothelial cell tight junction and adherens junction expression differences in glioblastoma multiforme
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Analysis of Human Brain Endothelial Cells Versus Human Tumor
Associated Brain Endothelial Cell Tight Junction and Adherens
Junction Expression Differences in Glioblastoma Multiforme
by
Aida Martinez
A Thesis Presented to the
FACULTY OF THE USC Keck School of Medicine
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
Experimental and Molecular Pathology
May 2020
Copyright 2020 Aida Martinez
ii
Acknowledgements
To my parents for being the pillar to all my aspirations, specially my father for helping me
discover the force in me
To Dennis, for being my diary through this journey.
Special thank you for Dr. Hofman's lab group for guiding me through this journey
iii
Table of Contents
Acknowledgements……………………………………………………………………………………ii
List of Tables……………………………………………………………………………....................iv
List of Figures…..……..……………………………………………………………………….. ...…...v
Abstract………………………………………………………………………………………………...vi
Introduction to the Field……………………………………………………………………………….1
Glioblastoma Multiforme and the Blood Brain Barrier………………………………………..1
Integrin and Their Role in Glioblastoma Multiforme………………………...………………..4
Brain Endothelial Cells and Tumor Associated Endothelial Cells in Glioblastoma
Multiforme...…………………………………………………………………...………………….5
Treatment of Glioblastoma Multiforme…………………………………………………………6
Materials and Methods………………………………………………………………………………..8
Cell Culture Reagents……………………………………………………………………………8
Endothelial Cell Isolation and Culture …………….……………………………………………8
Cytotoxicity Assay………………………………………………………………………………..9
Cell Seeding and Immunocytochemical (ICC) Analysis……………………………………10
Immunohistochemistry (IHC) Analysis……………………………………………………..…11
Treatment Specifications (in vitro)…………………………………………………………….11
In vivo Analysis………………………………………………………………………………….11
Results……………………………………………………...…………………………………………13
Characterization of Brain Endothelial Cells and Tumor Associated Endothelial Cells
Using Endothelial and Mesenchymal Markers………………………………...……………..13
BEC and TuBEC Viability When Treated with NEO212……………………………...……..15
Effects of NEO100 and NEO212 on the Cell-Cell Adhesion Marker Expression in
Brain Endothelial Cells and Tumor Associated Endothelial Cells (in vitro)………….........16
Effects of NEO100 and NEO212 on the Integrin Marker Expression in Brain Endothelial
Cells Versus Tumor Associated Endothelial Cells (in vitro)………………………………...21
Effects of NEO100 and NEO212 on the Cell-Cell Adhesion Marker Expression in Brain
Endothelial Cells versus Tumor Associated Endothelial Cells (in vivo)…...………………21
Summary Table of Results…………………...………………………………………………...25
Discussion………………………………………………………………………………………..…..25
Future Directions…………...…………………………………………………………………….....29
Supplementary Data………………………………………………………………………………...31
Materials and Methods: Antibody Reagents (Figure A)………………………..………….31
Immunohistochemistry (IHC) other markers (Figure B)………………………………..….32
Immunocytochemistry (ICC) other markers (Figure C)………………..…………………..35
References………………………………………………………………………………..………….38
iv
List of Tables
Table 1: Summary of Markers Expression on Brain Endothelial Cells and Tumor Associated
Brain Endothelial Cells……………………………………………………………………….………..25
Table 2 (Supplementary Figure A): Antibody Reagents………….………………………………..31
v
List of Figures
Figure 1: Schematic Representation of the Blood-Brain Barrier……………………….…………..5
Figure 2: Chemical Structures of Temozolomide,, NEO100 and NEO212……….………………7
Figure 3: Simplified Diagram of Cell Isolation Process…………………….……………………….9
Figure 4: Timeline of in vivo Treatment…………………...…………….…………………………..12
Figure 5: Immunohistochemical and Immunocytochemical Analysis of PECAM (CD31) and
Vimentin In Human Brain Endothelial Cells versus Tumor Endothelial Cells and
Mouse Normal Brain Tissue versus Mouse Brain Tissue With Tumor…………...….13
Figure 6: Cell Viability Under Different NEO212 Concentrations for Brain Endothelial Cells
and Tumor Brain Endothelial Cells………………..……………………………………...16
Figure 7: Immunocytochemical Analysis of Human Brain Endothelial and Tumor Associated
Brain Endothelial Cell Tight Junction and Adherens Junction Expression Using
Markers Claudin 6, Claudin 5, β-Catenin and, JAM-1….…………..…....……...……..19
Figure 8: Immunohistochemical Analysis of Mouse Normal Brain and Mouse Brain With
Tumor Using Markers Claudin 6, Claudin 5, β -Catenin, and JAM-1…....……….…..22
Figure A (Supplementary): Antibody Reagents………………………………………………..…..31
Figure B (Supplementary): Immunohistochemical Analysis of Additional Markers
(Supplementary Table A) on Mouse Normal Brain and Mouse Brain with Tumor….32
Figure C (Supplementary): Immunocytochemical Analysis of Additional Markers
(Supplementary Table A) on Human brain Endothelial and Tumor Associated Brain
Endothelial Cell…………………………………..…………………...……………………35
vi
Abstract
Glioblastoma Multiforme (GBM) is an astrocytoma with high aggressiveness and
resistance known to have extensive number of vessels. The increase in vessel formation during
GBM happens via the dissociation of cell-cell junction proteins which maintain the blood-brain-
barrier (BBB) and maintain vessel integrity from incoming macromolecules, toxins, and
xenobiotics. The loss of tight junctions and adherens junctions contributes to this dissociation
while integrins allow for endothelial cell migration and angiogenesis. Part of the transition from
normal to leaky vessels is due to an increase in mesenchymal markers in endothelial cells, thus
rendering this transition as endothelial to mesenchymal transition (EndMT) from brain
endothelial cells (BECs) to tumor- associated endothelial cells (TuBECs).This study uses
immunocytochemistry and immunohistochemistry as methods to study a panel of markers
comparing BEC versus TuBEC. Additionally, the study uses NEO212 as a tool to understand if
this novel conjugate drug can normalize phenotypic differences from TuBEC to BEC. Hence, the
study finds that claudin 6, a marker not yet associated with the BBB increases in tumor
endothelial cells, while Claudin 5 and 3 increase rather than decrease as expected from other
studies. Further β-Catenin, a cytoplasmic protein associated with cadherin adherens junctions,
increases cytoplasmically. However, when treated with NEO212, this expression as well as the
expression of claudin 6 decreases back to similar yet not same intensity to BECs. This shows
promising results to employ NEO212 in further studies for GBM treatment, as well as focus on
novel markers to better understand EndMT transition from BEC to TuBEC.
1
Introduction to the Field
Glioblastoma multiforme (GBM) is a type of brain tumor characterized by its high
aggressiveness, recurrence and resistance. GBM composes the majority of astrocytomas in the
glioma tumor group accounting for 80% of all brain tumors (Schwartzbaum, Fisher, Aldape, &
Wrensch, 2006) the rest of them being oligodendrogliomas, mixed gliomas and ependymomas
(Hanif, Muzaffar, Perveen, Malhi, & Simjee, 2017). The most common age of incidence is between
the ages of 55 to 60 years, being more common in males than in females (Thakkar et al., 2014).
Despite the efforts to treat the tumor, the median survival approximately is from 15 to 18 months
(Cho et al., 2014). Even though vast studies have focused on understanding the neuroglial stem
or progenitor cells that originate the intrinsic brain tumor (Le Rhun et al., 2019), not much is known
about the vasculature that surrounds these tumors, particularly about the so-called tumor-
associated brain endothelial cells (TuBECs).
Glioblastoma Multiforme and Blood Brain Barrier
During GBM, the endothelial cells that are part of the blood brain barrier (BBB) become
dissociated from one another and as a result become tumor-associated endothelium. Part of the
reason of the complexity of GBM, besides the heterogeneity of the tumor cells (primary GBMs
arise de novo, while secondary GBMs progress slowly from lower grade astrocytoma) (Smith and
Ironside, 2007), is the surrounding highly dense vasculature that contributes to the disease
progression. The BBB is essential for the normal brain physiology (Figure 1A-B). It is part of the
neurovascular unit (NVU), which is composed of pericytes, neurons, astrocytes, and brain
endothelial cells (BECs) (Muoio, Persson, & Sendeski, 2014). The BBB component regulates
nutrients, oxygen, temperature, electrical resistance, blood flow to the surrounding tissue and
conserves the cerebral homeostasis. It has low permeability compared to peripheral endothelial
cells (Tajes et al., 2014) in order to regulate the passage of macromolecules like xenobiotics and
2
toxins (Abbott, 2002), as well as the inflammatory responses and other fluctuating plasma
compositions that may disturb neural function and synaptic dysfunction (Abbott, 2002; Zlokovic,
2008). Thus, the endothelial lining of the cerebral microvasculature is non fenestrated unlike any
other surrounding systemic endothelial monolayer. Molecules can only cross via transmembrane
diffusion, adsorptive endocytosis, vesicular transporters, or efflux pumps (Banks, 2009). When in
the circumstance of GBM, tumor-associated blood vessels undergo a phenotypic and functional
transition that allows for tumor progression by allowing tumor growth and facilitated metastasis.
Studies have shown that tumor endothelium is of irregular size, sharp, fragile and with abnormal
sprouting in hypoxic regions of the tumor microenvironment. Hence, there is a dissociation of the
extracellular matrix (ECM) (Gerhardt et al., 2003) guided by pro-angiogenic factors like matrix
metalloproteinases (MMPs). This favors the dissociation of cell-cell unity so that cells migrate for
novel vessel sprouting and abnormal vessel formation. This process leads to fenestrated and
irregularly leaky blood vessels, which contribute to microvascular permeability and compromises
vascular integrity (Dudley, 2012).
Tight junctions, adherens junctions, and their respective accessory proteins on the
cytoplasmic side contribute to the dissociation of blood vessels in GBM and hence the integrity of
the endothelial monolayer. Tight junctions (or occluding junctions) are found in the most apical
part of the polarized brain endothelium, forming a complex network of strands that constitute a
series of punctate or extended contacts (Staehelin, 1974) (Figure 1C). Their main known function
is to bring cells in close apposition in order to create a selective para-cellular barrier (Gumbiner,
1987). They also maintain the integrity of the endothelial vessel wall by associating with
cytoplasmic adaptor proteins that form complexes and connect them to the cytoskeleton (Bauer,
Krizbai, Bauer, & Traweger, 2014). The most common members of the tight junction family are
claudins, occludins and junctional adhesion molecules (JAMs). They are associated with
cytoplasmic components that connect them to the actin cytoskeleton. Claudins and occludins are
3
similar proteins both with four transmembrane domains and both their N-and C-terminal domains
are located in the cytoplasm. In contrast, JAMS only have one transmembrane domain and it is
believed to be not necessary for the establishment of tight junctions, but JAM-1 may facilitate the
assembly of other junction proteins for tight junction formation and establishment of cell polarity
(Luissint, Artus, Glacial, Ganeshamoorthy, & Couraud, 2012). The role of claudins in CNS
tumorigenesis has been contradictory. They have been demonstrated to be tumor suppressive or
to play a role in tumor epithelial to mesenchymal transition (EMT) by either being affected by the
transition or influence the transition (Salvador, Burek, & Förster, 2016). For example, the levels
of claudin 1 and claudin 5 decrease in accordance to increased GBM tumor grade, which could
help explain tumor invasiveness (Karnati et al., 2014; Swisshelm, Macek, & Kubbies, 2005).
On the contrary, other studies have found that claudin 1 expression during post stroke recovery
is an impediment for BBB recovery because its expression coincides with a decreased expression
of claudin 5 (Sladojevic et al., 2019). Although tight junctions are understood to be important
elements of the BBB, their role in GBM endothelial cells is yet to be extensively studied.
Aside from tight junctions, adherens junctions are also important components of the BBB
and affect vascular integrity when dissociated from cell-cell contacts during GBM. Their normal
physiological function is to modulate cell-cell contacts, promote the maturity of these contacts,
and regulate tensile forces (Tietz & Engelhardt, 2015) between cells. Like tight junctions and
JAMs, adherens junctions are also associated with the Zonula Occludens (ZO) family of
cytoplasmic proteins. These are scaffold proteins that bind the junction proteins to the actin
cytoskeleton along with other associated proteins. Across the plasma membrane, the main
component of the adherens junction family in the BBB is the calcium dependent transmembrane
glycoprotein VE-Cadherin, which is important for BBB integrity. In GBM, VE-Cadherin is
downregulated When bound to cytoplasmic protein β-catenin (α, γ, p120 are also
associated),VE-Cadherin allows for endothelial cells to recognize each other and to bind to the
4
actin cytoskeleton (Stamatovic, Keep, & Andjelkovic, 2008).Besides providing a binding bridge
between VE-Cadherin and the cytoskeleton, β-catenin is also a signal transduction protein in
the canonical Wnt pathway involved with proliferation, migration, differentiation, and stem cell
renewal. One study demonstrated that the canonical Wnt/β-Catenin pathway is involved in
reduced blood vessel density and normalized blood vessels in GBM (Reis et al., 2012).
However, during GBM, loss of adherens junctions like VE-Cadherin is associated with increased
epithelial to mesenchymal transition of normal to tumor cells and aids in their cell migration
(Zeng, Fee, Rivas, Lin, & Adamson, 2014) which could be explained by the lack of cell-cell
contact.
Integrins and Their Role in GBM
Much is yet to be understood of the role that integrins playa in the normal bran endotelial
cell transition to brain tumor associated endothelial cell in GBM. Integrins are transmembrane
heterodimeric proteins that consist α/ β subunit of which combinations provide unique integrin
function and phenotype (Malric et al., 2017). They enable the cell to adhere to extracellular
proteins or proteins on other cell surfaces like immune cells (Takada, Ye, & Simon, 2007).
Integrins also have an important role enabling the EndMT transition. In cancer they are involved
in invasion, metastasis, and ECM remodeling via regulation of protease release to provide tract
for movement in order for these mesenchymal-like endothelial cells to form new vessels.
Additionally, integrins can regulate how a tumor proliferates and survives by the way the tumor
binds to its surrounding matrix and influences intracellular signaling. Integrins like αv β3 have
been linked to increase in both GBM tumor cells (Schnell et al., 2008). Furthermore, several
integrins like the β1 family have been shown to be implicated in GBM angiogenesis (Malric et
al., 2017). However, whether this integrin increase occurs in glioma tumor cells or TuBECs is
not clear. Therefore, further analysis of the expression pattern of integrins on control BECs
5
versus TuBECs could provide clarity on the functional and phenotypical differences seen on
GBM blood vessels.
Figure 1: A. Schematic representation of the blood-brain barrier (BBB).The BBB is mainly composed of
neurons, astrocytes, BECs, extracellular matrix and pericytes. B. In the presence of tumor, the vasculature undergoes
morphological and functional differences including dissociation of tight junctions, which enables the paracellular
passage of molecules such as blood borne pathogens and or toxins C. Conceptual diagram of the main cell-cell junction
proteins that are responsible for the brain vascular integrity. Claudins, occludins and Junctional Adhesion Molecules
(JAM) are the main types of tight junctions. Adherens junctions are composed mainly of cadherins ( shown is vascular
endothelial (VE)) in the brain and their accessory proteins on the cytoplasmic side that bridge them to the actin
cytoskeleton. In picture, these proteins are shown as P120, catenins, vinculin, AF6, α actinin. Other important cell-cell
proteins are integrins,which are mainly responsible for cell migration and ECM adhesion. While ZO-1-3 are mainly
associated with tight junctions as scaffold proteins, they may be also be associated with the catenins joining these two
(Stamatovic, Johnson, Keep, & Andjelkovic, 2016).
BEC and TuBEC in Glioblastoma Multiforme
Studies have documented some phenotypic and functional differences comparing control
healthy brain endothelial cells (BECs) and tumor associated brain endothelial cells (TuBECs)
derived from GBM. Functionally, BECs are angular, plump, and with vast cytoplasmic space.
When sub-confluent, these show a “cobble-stone appearance.” In contrast, TuBECs are
elongated with flat appearance and multiple processes that seem to branch into one another
A B
C
6
(Charalambous, Hofman, & Chen, 2005). At the subcellular level, the phenotype of the BECs
undergoes a progressive loss of markers like von Willbrand factor (VWF), PECAM (CD31) (Marín-
Ramos et al., 2019) and E-cadherin to become TuBECs. In turn, they acquire α-smooth muscle
actin (α-SMA), N-cadherin, vimentin and increased cell migration and invasion like mesenchymal
cells (Marín-Ramos et al., 2019). This transition was termed Endothelial to Mesenchymal
transition (EndMT) (Dejana, Hirschi, & Simons, 2017; Kalluri & Weinberg, 2009), and it is similar
to epithelial to the mesenchymal transition (EMT) that occurs in normal epithelial cells. This
transition increases cell migration and invasion for tumorigenesis and metastasis (Kalluri &
Weinberg, 2009; Potenta et al., 2008). Thus, the EndMT transition is critical for the progression
of tumor-related angiogenesis because it enables endothelial cells to acquire motility and
sprouting to benefit tumor growth majorly for nutrient and oxygen acquisition.
Functionally, BECs proliferate at a faster rate yet they migrate slower while TuBECs
proliferate at a slower rate and migrate faster. Thus, this could be a dichotomy whereby TuBECs
mainly want to promote the phenotype for the plasticity of new vessel sprouts rather than fully
developing into a normally formed “plump” BEC.
Treatment of Glioblastoma Multiforme
The current standard of care treatment for GBM is the chemotherapy alkylating agent
Temozolomide (TMZ), which is also given in conjunction with surgery and radiotherapy.
However, despite the available therapy, the median survival is still 15-18 months (Cho et al.,
2014), with 90% of the patients developing resistance to the treatment (Stupp et al., 2005).
Once the tumors recur, they are usually resistant to TMZ and no alternative treatments are
available. A novel drug, NEO212 (Figure 2), has emerged as a better alternative of treatment
that is cytotoxic to both TMZ resistant and sensitive glioma cells in vivo and in vitro. This drug is
a conjugate of NEO100 (perillyl alcohol; POH) and TMZ. NEO100 is a monoterpene that has
7
been given as an oral medication to treat a variety of cancers. It is a cytotoxic agent that inhibits
the cell cycle, is a Ras inhibitor, and upregulates the pro apoptotic protein Bax (Yeruva, Pierre,
Elegbede, Wang, & Carper, 2007). NEO100 has also been shown to be cytotixic to TMZ
resistant and sensitive glioma cells and reduce their invasiveness (Cho et al., 2012). NEO212
favors over TMZ alone because NEO100 provides lipophilicity for the drug to cross the blood
brain barrier and so lower doses of both NEO100 and TMZ could be administered to minimize
toxicity (Lanevskij, Japertas, & Didziapetris, 2013) (Figure 2). NEO212 has been shown to block
EndMT in BECs and it has the potential to revert this transition back from TuBEC to BEC
(Marín-Ramos et al., 2019). It has also been shown to impair vessel tubule formation of these
mesenchymal-like endothelial cells. Because EndMT is involved in the development of tumor
vasculature, tumor progression, vessel permeability, and angiogenesis (Potenta, Zeisberg, &
Kalluri, 2008), NEO212 and its component NEO100 are potential drugs to study if they have an
effect on the expression levels of cell-cell junction proteins, cytoskeleton, and cellular adhesion
molecules that may play a role in the EndMT transition from BEC to TuBEC.
Figure 2: NEO212 is a conjugate of NEO100 (perillyl alcohol; POH) and temozolomide (TMZ), the
current standard care of treatment chemotherapy for GBM. Chemical structures depicted using ChemDraw
software
Hence, the focus of this project is 1) to understand the differences in the expression of
some tight junction proteins, adherens junctions, and integrins between BECs and TuBECs; and
2) to understand whether NEO212 and/or NEO100 have the potential to normalize the phenotype
8
of TuBECs to make them similar to BECs. Understanding the expression differences between
BECs and TuBECs would provide insight into the disintegration of the blood vessel wall and
EndMT transition, which contributes to promote angiogenesis and tumor invasion (Potenta et al.,
2008) important for the progression and complexity of GBM.
Materials and Methods
Cell Culture Reagents
Both TuBEC and BEC were cultured in Endothelial cell Growth Basal Medium-2 (EBM-2)
medium (Lonza,19860) supplemented with 5% FBS, 1% pen-strep,1.4 µM hydrocortisone, 5
µg/mL ascorbic acid, 1% lipid concentrate, 10 mM HEPES, and 1 ng/mL of bFGF. They were
cultured on gelatin coated flask. Cells were grown at 37 °C and 5% CO2.
A list of the antibody reagents used is available in Supplementary Data (Figure A)
Endothelial Cell Isolation and Culture
Human normal and human GBM brain tissues were obtained from available GBM
specimens or epileptic patient surgeries. These tissues were the source of the brain endothelial
cells and tumor associated endothelial cells used for immunocytochemistry. First, the tissue was
washed with the appropriate medium that contains 1% penicillin and streptomycin (pen-strep).
The tissues were then cut into miniscule pieces while inside medium with 2% FBS. The mixture
was then transferred to a centrifuge tube and new medium was added. The mixture concentration
was reduced to 15% after adding 30% dextran solution. This mixture was then centrifuged at
10,000 rpm during 20 minutes. In order to isolate the micro vessels, the pellet was resuspended
in 1 mg/mL collagenase dispase in EBM-2 media with 1% FBS and shook for one hour inside a
37°C incubator. Then 10 mL of EBM-2-2% FBS was added to the pellet, centrifuged at 1200 rpm
for 5 minutes. Then the pellet was resuspended again in 20 ml of EBM-2-2% FBS and centrifuged
one final time. Then the final pellet was suspended in EBM-2 with 5% FBS, 1% penicillin.
9
Endothelial cells were then purified and sorted using FACS dil acetylated low density lipoprotein
used to detect vertebrate endothelial and microglial cells.
Cells were seeded on 75 mL culture flasks and 50% of the media (approx. 5 ml) was
changed every 3-4 days until they were 80- 90% confluent. Cells
were not passaged unless they had at least 24 hours following media change to allow them to
adapt to the new media, temperature and pH changes that the previous media change may have
caused.
Figure 3. Simplified diagram demonstrating flow diagram of the cell isolation process. Both TuBEC and BEC
are cultured in EBM media with 2% FBS, allowed to proliferate and further dissociated in new media until the
establishment of cell lines. Pictured above is a overview of different cells that could be established using this process
not exclusive to endothelial cells. Image adapted from Vescovi et. al, 2006.
Cytotoxicity Assay
On a 96 well plate, 2.5x10³ human normal brain endothelial cells or human tumor
associated brain endothelial cells (obtained from primary patient brain tissue per well) were
seeded. After 24 h to allow for cell attachment, the EBM-2 with 10% FBS medium was replaced
with increasing doses of NEO212 or the corresponding equivalent volume of dimethyl sulfoxide
(DMSO) for vehicle. After 72 hours, the medium was changed with EBM-2 medium containing 2
mg/mL of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) reagent for four
hours at 37°C in the dark. Then the medium was removed and the formazan crystals were
dissolved in DMSO (100 µL/well). The absorbance at 570 nm (using 630 nm as a reference
wavelength) was then measured using an Asys UVM 340 microplate reader. NEO100 dose was
obtained from a previous study indicating that 2mM/mL NEO100 is subcytotoxic for confluent
EBM+ 2%FBS Media
EBM+ 2%FBS Media
EBM+
2%FBS Media
Optional : co-culture cells
10
BECs (Cho et al., 2012). This dose was adjusted to 1mM/mL of NEO100 for TuBECs taking into
account the lower cell viability of TuBECs when treated with NEO212. However, a proper dose of
POH for TuBECs is yet to be experimentally tested. One experimental triplicate was performed
on human TuBEC (n=1) or human BEC (n=1) for NEO212 treated.
Cell Seeding and Immunocytochemical (ICC) Analysis
Prior to BEC and TuBEC seeding, a round 15 mm glass plate was added to each well of
a 24 well plate and coated with 1% collagen (for integrin antibodies) or 1% gelatin ( for all other
antibodies)(supplementary table Figure A). Once they reached 85% confluency, cells were
seeded onto each plate at a density of 30,000 cells per well using conventional hemocytometry.
After 24 hours, cells were treated for 72 hours. Then, cells were fixed with 4% paraformaldehyde
(PFA) in PBS and washed gently with PBS once. Slides were permeabilized with PBS-T for 5
minutes, and washed again with PBS to remove any excess PBS-T. The cells were blocked using
200 µL of 1% BSA in TBS-T for 20 minutes and then incubated with 200 µL of the corresponding
primary antibody (Figure 1A in Supplementary Data) overnight. The next day cell slides were
washed with PBS 3 times before being incubated with 200 µl secondary antibody in 1%BSA in
TBST-T (either biotinylated goat anti-mouse 1:200, biotinylated anti-rabbit 1:300, biotinylated
rabbit anti-rat 1:200 for 45 minutes. Subsequently, 150 µL of ABC Elite (avidin-biotin complex) in
PBS was added to the cell slides for 30 minutes at room temperature at shaker speed 2. ABC
had been prepared at least 30 minutes prior to be added to the slides. Following, the slides were
washed 3 times with PBS and then added 150 µL of AEC (2.12 ml dH20, 25 µL acetic acid 1M,
100 µL sodium acetate 1M, 12.5 3% H202, 125 µL AEC substrate). The presence of a light brown
or brick red precipitate meant positive staining. Lastly, slides were counterstained with
hematoxylin for 30 seconds each to stain and sealed with VectaMount AQ on glass slide. Slides
were imaged using a microscope provided by the Cell and Tissue Imaging core of USC Research
11
Center for Liver Diseases. Each complete ICC analysis was done in experimental triplicate using
one patient sample (n=1) for BEC or TuBEC. Staining was subjectively analyzed based on brown
or brick red color being positive staining, versus blue or blue gray being negative staining.
Immunohistochemical Analysis (IHC)
Frozen tissues were cryostat sectioned with a thickness of 10 microns. Each tissue slide
was fixed in acetone for 10 minutes. Then, tissues were washed in PBS for 5 minutes 3 times and
incubated in Sea Block solution (Sea Block:PBS 1:1) for 20 minutes. Tissues were then washed
with PBS 3 times each for 5 minutes and then incubated with primary antibodies (Figure 1A in
Supplementary Data) overnight. The next day, tissue was washed with PBS 3 times for 5 minutes
each. Immunostaining was performed as described above.Images presented in the study were
based on best image of 3 individual experiments (n=3).
Treatment Specifications (In vitro)
NEO212 and NEO100 were provided by NeOnc Technologies Inc.The stocks were
prepared in DMSO. Final concentration of DMSO in the culture medium was 0.158%.
In vivo
All animal protocols were approved by the USC Institutional Animal Care and Use
Committee (IACUC). A total of 1x10
5
GL261 cells were implanted into the subcortical
parenchyma of each C57BL/6 female mice. The total implantation volume was 1 µL.NEO212
was diluted individually in ethanol: glycerol 1:1 with a final concentration of 10% DMSO. Vehicle
(10% DMSO in ethanol: glycerol 1:1) were administered individually to each mouse,NEO212
given at a dose of 100 µL 50 mg/kg (Marín-Ramos et al., 2019) subcutaneously. After seven
days post implantation of tumor cells, treatment was given daily every 5 days and off for 2 days
(Figure 4). A total of 9 mice were used, 3 mice were euthanized for normal brain, 3 were treated
with vehicle , and 3 were treated with NEO212. Mice were euthanized at a humane endpoint
except normal mice.
12
Figure 4: Timeline of in vivo treatment. After 7 days post implantation, mice were treated for 5 days (blue) and off
for 2 days (red) for vehicle or NEO212 treated. Each mouse was euthanized (black circle) at different times between
10-20 days after treatment depending on when they reached a humane endpoint.
&
13
Results
Characterization of BECs and TuBECs Using Endothelial and Mesenchymal Markers
Data in this study confirms that TuBECs closely resemble mesenchymal cells due to
higher mesenchymal associated marker expression compared to control BECs which, as
expected, express endothelial markers. TuBECs resemble mesenchymal cells because of their
increased expression of markers like α-smooth muscle actin and vimentin (Kalluri & Weinberg,
2009), while the expression of vascular endothelial cell markers like cell adhesion molecule
PECAM (CD31) is decreased (Marín-Ramos et al., 2019). Taking these findings into
consideration, PECAM and vimentin were selected as staining markers in order to confirm that
BECs used for this study were different from TuBECs. Both immunocytochemistry (ICC) of
patient-derived endothelial cells and immunohistochemistry (IHC) of mice brains bearing tumors
were performed. The results in Figure 5A show a decreased staining of vascular endothelial
marker PECAM (CD31) from BEC to TuBEC in the cytoplasm. Both cell types show PECAM
(CD31) marker distribution localized to the cytoplasm rather than the cell membrane. Vimentin
appears absent in BECs while there is an increased staining in TuBECs. Similarly, IHC analysis
of ex vivo mouse brain tissue (Figure 5B) and a closer view of the blood vessels (Figure 5C) show
a decreased expression of PECAM (CD31) from normal brain to vehicle treated tumor brain while
vimentin expression increases for the same conditions. This confirms the identity of BEC and
TuBEC onto which ICC and IHC analysis were done.
Figure 5 (below). Photomicrographs showing marker expression differences of PECAM (CD31) and vimentin
between Brain Endothelial Cells (BECs) and Tumor Associated Brain Endothelial Cells (TuBECs). A BECs show
a slight greater staining intensity of PECAM (CD31) but greatly lower vimentin stain compared to TuBECs. Arrow
depicts cytoplasmic area of stain: both cells show PECAM in the cytoplasm not concentrated at cell-cell junction.
Vimentin does not stain in BECs, but arrow shows area of stain in TuBEC the cytoplasm, not cell-cell junction. B Normal
brain tissue shows more staining of PECAM in normal brain tissue ( arrow) while this decreases in vehicle treated tumor
tissue (tumor area pointed by arrow). Vimentin in normal brain does not stain compared to the increase both at the
tumor site (arrow) and the normal peritumoral (star) area. C A closer view (54x) of blood vessels (arrows) from normal
brain tissue versus tumor vehicle treated tissue. PECAM (CD31) stain is present at endothelial junctions while absent
in tumor vessels. On the contrary Vimentin is not seen expressed in BEC but is seen in TuBEC vessels that are
surrounded by tumor cells. All negative control experiments (secondary antibody only) show no staining, meaning there
14
is no non-specific staining of antibodies. Results reported represent one of three experimental experiments for BEC or
TuBEC (n=1) for in vitro (A). B-C represent best images of three individual experiments (n=3). NB is normal brain while
T is tumor.
PECAM (CD31)
40x
40x
40x
PECAM (CD31)
40x
A
Vimentin
40x
Vimentin
40x
Normal Brain BEC TuBEC Secondary Only
B
40x
40x
40x
40x
N
B
T
T
40x
NB
40x
15
BEC and TuBEC Viability When Treated with NEO212
MTT assays showed that the viability of BEC and TuBEC was not affected by NEO212 at
the concentration selected for this study (50 uM) and this dose is sub-cytotoxic to the cells. To
determine whether NEO212 affects the viability of BECs and TuBECs at the dose used for the
study after 72 hours, an MTT assay was performed (Figure 6). At first concentrations in culture
(0-20 μM) on average TuBECs were more resistant than BECs by 11% cell viability difference.
This is expected as TuBECs develop resistance to standard TMZ chemotherapy treatment under
several DNA repair mechanisms but the resistance is not as great in TMZ as it is for NEO212 with
same dose levels (Cho et al., 2014). At 50 μM, cell viability is very similar with about 1% difference
meaning that at this same dose NEO212 is subcytotoxic to both cell types with small percentage
difference in cell viability between both. However, at increasing μM concentrations of NEO212,
BECs remained more resistant compared to TuBECs with about a 14% cell viability difference.
Dose of 1mM for NEO100 were taken from previous study finding the IC50 of endothelial cells
treated with NEO100 is greater 2mM/L. Cho et al., 2012). However, because this study was done
54x
PECAM (CD31)
Vimentin
54x
54x
54x
C
Normal Brain BEC TuBEC Secondary Only
40x
40x
16
for confluent BECs, the dose was adapted to 1mM/L to be below subcytotoxic levels and be
treated on TuBEC too.
Figure 6. Linear plot demonstrating the cell viability under different NEO212 concentrations for BEC and
TUBEC Based on results, NEO212 is not cytotoxic for BEC and TuBEC at 50 uM after 72 hours. Data represents
average of experimental triplicates for BEC or TuBEC (n=1) for each. Error bars based on standard deviation average
of three experimental triplicates.
Effects of NEO100 and NEO212 on the Cell-Cell Adhesion Marker Expression in
BECs versus TuBECs (In vitro)
Tight junctions: Claudin 6 and JAM-1 decreases from BEC to TuBEC while JAM-1
decreases from BEC to TuBEC
In order to understand the tight junction expression differences between endothelial cells
in normal brain (BECs) versus GBM tumor brain vessels (TuBECs), cells were first analyzed in
0
20
40
60
80
100
120
0 20 40 60 80 100 120
% Cell Viability (relative to
vehicle)
NEO 212 ( micro M)
Cell Viability When Treated With NEO212
17
vitro using immunocytochemical analysis of claudins, JAM-1, and their accessory protein ZO-1.
An in vitro analysis was done first in order to specifically see endothelial cell marker expression
without tissue micro-environment ( such as presence of glial tumor cells, pericytes, or glial stem
cells) that could influence the expression of the proteins on endothelial cells. Patient-derived
endothelial cells were cultured and seeded into round 15 mm clear glass slides and treated with
vehicle, NEO100 (1 mM), NEO212 (50 μM) all at 0.158% DMSO. Seeded cells were then
analyzed by immunohistochemistry after 72 hours in incubation with treatment. Analysis was done
subjectively comparing TuBEC staining under the three fields (vehicle,NEO100, and NEO212) to
normal staining in BECs. Staining was determined to be positive if there is a brown or red satin
versus only blue signifying absence of stain. Because claudin 1,3,5, and 12 are highly expressed
in endothelial cells in neural tissue and considered to be responsible for the function of the BBB
(Berndt et al., 2019), the expression of claudin 5,3, and 6 was studied. There was a decrease of
claudin 6 from vehicle treated BECs to vehicle TuBECs. As observed in vehicle treated TuBECs,
NEO100 treated BECs show similar staining, hence NEO100 does not have an effect on the
expression of this marker compared to vehicle. However, the staining of claudin 6 increases in
TuBECs treated with NEO212 compared to vehicle treated TuBEC. Although the staining intensity
is not equivalent to vehicle treated BEC (slightly less intensity compared to BECs), NEO212
maintains more expression of this marker on tumor endothelial cells (Figure 7A). Staining for
Claudin 6 remains the same throughout vehicle and NEO100 treatment, and staining intensity
seems to increase when treated with NEO212.Throughout the three treatments, the morphology
of TuBECs remains elongated, not plump and “ cobble stone” in appearance compared to BECs
signifying that treatments did not have an effect on the morphology of the cells stained for claudin
6 but yes on marker expression. Claudin 5,on the contrary, shows an appearance in marker
expression from not staining positive (absence of brown/red staining) in BEC vehicle treated to
TuBEC vehicle treated. Vehicle treated TuBEC to NEO100 treated TuBEC show similar staining
18
intensities however, the morphology changes significantly from more plump to thin elongated
TuBECs. This morphology is restored in TuBECs treated with NEO212, however the staining
intensity for in NEO212 treated TuBECs remains similar to vehicle and NEO100 treated TuBEC.
BEC across the three treatment types remained absent of claudin 5 expression (Figure 7B).
Additional tight junction markers tested like JAM-1 showed an increase in expression from BEC
to TuBEC vehicle treated, and absent when treated with NEO100, and NEO212 (Figure 7D).
Hence, NEO100 and NEO212 could have an effect on the expression on JAM-1 IN TuBECs. To
further test if there was cytoplasmic accessory protein supporting the expression of the tight
junctions, ZO-1 was tested using the same ICC technique (in vitro). However, there was no
evident staining (absence of brown or red stain) in vitro for this marker for both cell types across
all three treatment types (Supplementary Figure C)
Adherens Junction: β-Catenin increases from BEC to TuBEC
β-Catenin expression increases from BEC to TuBEC. β-Catenin is an important cell-cell
junction protein because it links vascular endothelial (VE)-Cadherin in the BBB to the cytoskeleton
of cells (Guo, Breslin, Wu, Gottardi, & Yuan, 2008). It is also involved in the Wnt/β-catenin
canonical pathway inhibiting angiogenesis and normalizing blood vessels when localized in the
nucleus of endothelial cells (Reis et al., 2012). To verify the presence and to what extent is it
expressed in BECs and TuBECs, β-catenin expression was also tested. Both BEC and TuBEC
show expression of β-catenin, however, TuBECs show an increased staining intensity when
treated with vehicle compared to vehicle treated BEC. This expression remained high in NEO100
treated TuBEC, however the cytoplasmic space of NEO100 treated TuBECs is significantly
reduced, hence the expression of β-catenin is less evident compared to less elongated vehicle
treated TuBECs. The expression of β-catenin in NEO212 TuBEC remains the same staining
intensity compared to vehicle and NEO100 treated TuBEC, however it is still greater compared
to vehicle treated BECs. The morphology once again is retained when treating TuBEC with
19
NEO212 and closer to control BECs, less elongated. Whether the expression is also expressed
in the nucleus of TuBECs remains yet to be determined.
Figure 7 ( below): Immunocytochemical analysis of tight junction and adherens junction expressions: Claudin
6 (A), Claudin 5 (B), JAM-1 (C), and β-Catenin (D). Top row for each marker shows BEC under all treatment
conditions: vehicle treated, NEO100 (1 nM), and NEO212 (50 μM) treated (all treatments are in .158% DMSO). Bottom
row of each marker shows TuBECs with same testing conditions as BEC.A Arrow points at BEC with staining intensity.
This staining intensity decreases in vehicle treated TuBEC and remains similar to TuBEC treated with POH. When
treated with NEO212, this staining intensity (based on brown staining) increases in TuBECs, however the intensity
seems to be heterogenous in some areas ( as pointed by black arrow arrow) compared to less stained surrounding
areas (red arrow). BECs all have high staining intensities for claudin 6. Arrows point at stronger staining intensity areas
hence, the staining pattern is not homogenous throughout the cells. B BECs do not stain for claudin 5 throughout all
three conditions. Arrows point at areas of greater cell density where it would be expected to observe cell- cell staining
of this tight junction, however, no stain is seen. In TuBECs, however, arrows at three conditions point at cells with
staining for claudin 5 seen in the cytoplasm. TuBEC staining for claudin 5 remains positive ( brown) and similar across
the three conditions. C. β-catenin expression increases in brown staining intensity from normal BEC to TuBEC. Arrows
point at cells where the stain is seen ( at the bottom row), mainly cytoplasmically however, if β-catenin is localized to
the nucleus is yet to be determined. When treated with NEO212, TuBEC cells are not elongated and thin compared to
TuBEC that are treated with NEO100. Hence, this morphology difference could be due to NEO100 dose effects on
TuBECs, which affects TuBECs in their morphology and reduces their cytoplasmic space. The arrow at the top row
demonstrate staining is very faint in BEC compared to a stronger, bolder staining of TuBEC β-catenin across the three
conditions. D Compared to BEC which show no staining( no brown) on any of the three conditions, the expression of
JAM-1 does increase in TuBEC. JAM-1 is not seen ( absence of brown) when treating TuBEC with NEO100 or
NEO212. Hence the arrows on bottom row center and right show an absence of staining between cell-cell contact,
where JAM-1 would be expected to be on a tight junction.
A
Claudin 6
BEC TuBEC
40x
40x
40x
40x
40x
Secondary Only
40x
40x
20
C
β
-
C
a
t
e
n
i
n
C
D
β-
C
at
en
in
C
B
β Catenin
BEC TuBEC BEC TuBEC
40x
40x
BEC TuBEC
D
40x
40x
21
Effects of NEO100 and NEO212 on the Integrin Marker Expression in BECs versus
TuBECs (In vitro)
To determine whether integrins could have a role in the transition from BECs to TuBECs,
we chose three markers to analyze αvβ3, α6, and β1. Instead of 1% gelatin, cells were cultured
on rat collagen coated 75 ml flasks (50 μL/mL). Although studies have shown integrins to be
important in blood vessel growth because they are the main adhesion receptors that endothelial
cells use to bind to the extracellular environment, hence allowing for proliferation, migration and
invasion (Stupack & Cheresh, 2004), our studies did not show marker intensity differences
between TuBECs and BECs during immunohistochemical analysis nor immunocytochemistry
(Supplementary Figure B and C).
Cell-Cell Adhesion Marker Expression in BEC versus TuBEC (In vivo)
To further test whether the expressions in vitro are also applicable in vivo, the markers
were tested in mice brain implanted with tumor cells. A total of 9 mice were tested, 3 vehicle
treated, 3 NEO212 treated and 3 normal brain. However, shown here is the best representative
of three individual experiments (n=3). Each mouse was treated post 7 days and treated 5
consecutive days and off for two days for 10-20 days. All markers were tested using IHC analysis
(full table of results see Table 1). However, only results similar to ICC were reported here (for full
IHC and ICC results, see Supplementary Data Figure B-C). Normal mouse brain shows a high
expression of claudin 6 on endothelial cells lining vessel walls (bottom left of figure 8A). Vehicle
treated tumor tissue shows a decrease in claudin 6 expression lining the vessel (Figure 8A) and
a slight increase in staining on the normal peritumoral side compared to normal tumor brain
(shown at Figure 8A top middle region).Results are consistent with ICC results whereby vehicle
treated BECs show more staining compared to vehicle treated TuBECs which also show very
faint, brown positive staining in cells. NEO212 seems to normalize this intensity and shows more
staining on the blood vessel in the tumor area compared to vehicle treated tumor vessel (8A;
22
bottom right). This staining is normalized also in the peritumoral normal brain region (Figure 8A
top right) and is less intense compared to vehicle treated peritumoral region. This could be due
to the effect that glial cells have on adjacent peritumoral regions that could include microvessels.
However, the effect of tumor on adjacent normal tissue is yet to be explored. Claudin 5 expression
increases from an almost absent blood vessel staining in normal brain to a clearly outlined blood
vessel (Figure 8B bottom row) in vehicle treated tumor blood vessel (8B;bottom center). The
staining intensity is less opaque when treated with NEO212 (Figure 8B; top and bottom right)
similar to ICC, whereby NEO212 has an effect on the expression of the marker. Unlike ICC, β-
catenin intensity is not present in normal brain blood vessels. But, like ICC, the staining intensity
is high as seen in vehicle brain with tumor and it does not decrease staining intensity in NEO212
treated tumor brain blood vessels compared to vehicle treated tumor brain (Figure 8C; bottom
row). Lastly, JAM-1 did not show any marker staining in normal brain. However, staining in blood
vessels was present in tumor vehicle treated brain and remains in NEO212 treated tumor brain
blood vessels. Thus JAM-1 expression in vehicle mouse brain with tumor is not affected by
NEO212 unlike as shown in vitro for TuBECs treated with NEO100 or NEO212.
Figure 8 ( below): Representative images of immunohistochemical staining of mouse normal brain, vehicle and
NEO212 treated tumor brain using markers Claudin 6 (A), Claudin 5 (B), β Catenin (C), and JAM-1 (D). Top row
represents at a glance view of tissue (40x) showing overall stain intensity that could represent micro vessels amongst
the tissue. Bottom row for each marker represents a closer view of staining at the blood vessel located within tumor
region (80x). Arrows show blood vessels in mouse brain either with tumor ( bottom row, left) or with tumor ( bottom row
center and right). NEO100 was not tested for this in vivo analysis because the ICC analysis did not show a significant
change from vehicle treated TuBEC to vehicle treated TuBEC. Hence, to best use available testing subjects, only
NEO212 and vehicle were tested as treatments. Observed in all tissues of normal brain (top left) is tissue with high
density of myelin and distanced cell structures, a characteristic of white matter of the brain. Tumor cells in brain tissue
(center, bottom right corner of image) are more separated and with larger nuclei compared to normal glial cells in normal
tissue. Top row left shows an overview of normal brain tissue microvessels in the tissue for every marker. Tissues are
counterstained with hematoxylin to stain nucleus ( blue circles). A. Like ICC, claudin 6 expression decreases from
normal brain blood vessels to blood tumor associated vessels in brain with tumor. Arrow points at one microvessel
outlined by a positive stain (faint brown) of claudin 6.At the top center, the staining of claudin 6 is seen very faint ( faint
brown) in the tumor area blood vessels ( arrow). However, there is an increase in claudin 6 staining in the peritumoral
area (star). When treated with NEO212 the top right shows an increase in staining of blood vessels as seen by the
arrow. Bottom row three images depict the same pattern as described but at a higher magnification (54x) to clearly
depict blood vessels (arrows). B. There was no staining for claudin 5 in normal mouse brain observed in normal mouse
brain. However this staining increases in tumor blood vessel ( top center arrow) and also staining is seen in the
peritumoral area (star). When treated with NEO212, staining is more faint (left top and center bottom image; arrow)
compared to a greater outlined blood vessel (center top and center bottom image; arrow).C There is no positive stain
( no brown) for β-catenin observed in brain normal blood vessel which is seen in blue ( bottom left ; arrow) or at lower
magnification (40x depicts no clearly outlined microvessels).However this staining increases in tumor blood vessel
23
(outlined in brown) and also the peritumoral area (star). Treating the tissue with NEO212 however, does not change
this staining intensity for β-catenin in blood vessels (right top and right bottom image; arrow). D. Like in vitro analysis,
normal brain has no positive staining for JAM-1 (bottom image; arrow). No clear depiction of microvessels is seen in
normal brain at 40x magnification. However arrows point at blood vessels within the tumor which show positive staining
for JAM-1 and which this staining remains even when treated with NEO212 (bottom and top center and right images;
arrow), unlike ICC. NB is normal brain and T is tumor. Images represent best image of three independent experiments
(n=3).
Claudin 5
A
B
Secondary Only
40x
40x
T
NB
T
NB
NB NB
T
NB
24
C
D
40x
40x
NB
NB
NB
NB
T
T
T
T
T
Secondary Only
25
======================================================================
Tumor
Marker Normal Brain Vehicle NEO212
α 6 Integrin - - -
αv β3 integrin - - -
β 1 Integrin - - -
β Catenin - ++ ++
CD44 + ++ ++
Claudin 3 +++ +++ +++
Claudin 5* - +++ ++
Claudin 6* ++ + + ++
E- Selectin +++ ++ ++
ICAM +++ ++++ ++
JAM-1 - + +
ZO-1 - - -
• Abbreviation: UNT= untreated, VEH= vehicle treated
• - = No staining detected
• += low intensity, ++=moderately low intensity , +++= moderately high intensity, ++++=
high intensity.
* = results consistent with ICC results
Discussion
The BBB is component of the neurovascular unit (NVU). Its function is to keep
macromolecules such as toxins or xenobiotics from crossing to the brain parenchyma. Although
much is known about the BBB in normal brain, much is yet to be understood about the changes
that occur in the endothelial cells that compose it during pathological circumstances such as
glioblastoma multiforme (GBM). The integrity of the BBB is maintained with the help of special
cell-cell proteins called tight junctions and adherens junctions, which are responsible for high
endothelial resistance and cellular polarity respectively at the BBB. Unfortunately, during GBM,
Table 1(below): Summary of markers expression on BECs versus TuBECs. Overall representation of staining
intensity as seen in vivo for all markers of the study. Only markers with asterisks demonstrated consistent results
with ICC. In vivo analysis of tissue had more staining compared to ICC analysis. This could be to the nature of the
surrounding microenvironment that surrounds endothelial cells. Hence, in turn this could have influenced in vitro
results. The analysis was done based on brown or brick red staining intensities to represent positive staining for the
marker. Negative staining is seen in blue.The analysis was done comparing images across the different field
conditions using normal brain ( and BEC for ICC) as baselines of expression.
26
many changes occur including hyperactive vessel growth, hemorrhage, and leaky vessels with
dead – end structures (Charalambous, Hofman, & Chen, 2005). The former is all due to
dissociation of the junctions that tighten the blood vessel wall. Another type of proteins that play
a role in the blood vessel phenotypic modifications that take place during GBM are integrins,
heterodimeric proteins that are associated with providing track for endothelial cell migration and
angiogenesis, a feature that ultimately benefits tumor progression. Recent studies have shown
effort to understand the phenotypic changes that brain associated endothelial cells (BEC) undergo
to become tumor-associated endothelial cells (TuBEC) such as tubule formation or migration for
example (Marín-Ramos et al., 2019).Additionally, one of these changes is the endothelial-to-
mesenchymal transition (EndMT), a process similar to epithelial to mesenchymal transition (EMT)
in which BECs undergo the change to resemble a more mesenchymal phenotype expressing
markers such as alpha smooth muscle actin, vimentin, and CD44 to name a few. This study
analyzes different markers including tight junctions, adherens junctions, and integrins, to see if
there are any expression differences between BECs and TuBECs. Additionally, we analyzed
whether NEO100 (perillyl alcohol; POH) or the novel conjugate NEO212 could normalize the
expression of these markers in TuBECs and make it back similar to BEC. Overall because of the
importance that junctional proteins and integrins could have in these phenotype changes
including EndMT, this study is done to give light for future studies to understand the molecular
changes that occur in BECs to transform into TuBECs.
Claudins were the first markers to be tested in this study due to their known presence in
BBB ( specially claudin-3, and -5) and their essential role in the formation and maintenance of the
BBB. Contrary to findings that claudin 5 is significantly downregulated in hyperplastic vessels of
GBM (Liebner et al., 2000), in vitro and in vivo analysis of this marker showed an increase in
expression from normal endothelial cells to tumor associated endothelial cells. When treated with
NEO212 both in vitro and in vivo, claudin 5 expression is absent when compared to vehicle
27
treated tumor endothelial cell, however it does not go back to normal endothelial cell expression
levels in vivo ( Figure 8B) or in vitro ( Figure 7B). The morphology of NEO212 treated TuBECs is
retained to a more normal looking endothelial cell with a vast cytoplasm when compared to
NEO100 which made TuBECs look elongated compared to normal BECs, Hence, morphologic
revertion occurs with the treatment of this novel conjugate in TuBECs. When using NEO100, the
morphology of TuBEC is affected rather than made look more similar to BECs and the staining
does not change compared to vehicle in vitro, hence the effectivity of POH as a treatment is still
questioned. Unlike other studies which express a decrease in claudin 5 in hyperplastic vessels
during GBM, this expression variability could be due to the tumor grade and the degree of
hyperplasticity of blood vessels during the time of tissue collection. Additionally, these studies are
done using human GBM blood vessels, which could present variability when studying mouse
tissue in vivo and in vitro analysis of human BECs and TuBECs do not have the without the
supporting pericytes or surrounding tumor microenvironment surrounding a GBM vessel. Claudin
3, although known to be a central component of the BBB, showed no difference in expression in
vivo and lacked expression in vitro regardless of treatment type (Supp. Figure B-C). As it happens
in the case of claudin 5, the lack of presence of pericytes (which give endothelial cells of blood
vessels support and stability) in vitro could explain the absence of positive staining contrary to
studies that confirm the presence of claudin 3 in tight junctions of the BBB(Schrade et al., 2012)
Lastly, although claudin 6 is known to play a role in other systemic neoplasms like teratoid
rhabdoid tumors (AT/RT) (Sullivan et al., 2012), little is known about its potential roles in GBM.
This study found that there is a decrease in claudin 6 both in vitro and in vivo from normal to tumor
endothelial cell. This levels of expression increased in vitro when compared to vehicle treated
TuBECs and NEO100 treated TuBECs which did not show a expression difference to vehicle
treated BECs. Similarly, NEO212 increases the expression of claudin 6 when compared to vehicle
treated tumor tissue vessels, however not to the expression levels of normal brain in vitro or
28
vehicle treated BEC in vitro. Hence, although NEO212 increases the expression from vehicle
treated tumor, levels of marker are not normalized to normal endothelial cell expression levels.
Claudin 6 could be a marker for future studies to understand the role it has in tight junction control
in the BBB as there are no studies that indicate its presence in the BBB.
Although JAMs are implicated in the early attachment of adjacent cell membranes and
actively involved in the BBB, we observed no expression of JAM-1 in the normal brain vasculature
in vitro. However, the expression of JAM-1 was increased in TuBECs compared to control BECs,
as well as in tumor bearing brains, compared to normal brains. Treating TuBECs with vehicle,
NEO100, or NEO212, decreased this expression which questions the role that JAM-1 may have
in tumorigenesis since all three treatments were able to remove the expression rather than
increase it more. In vivo, JAM-1 increases in blood vessels ( Figure 8 D) and its levels are not
affected by NEO212 treatment which varies from in vitro results. These results may vary due to
the surrounding microenvironment of the tumor which could support the sustained expression of
JAM-1.
Also in this study we observed an increase in the levels cytoplasmic staining of the
adherens junction associated cytoplasmic protein β–catenin both in vitro and in vivo. Although
this study did not test for VE-Cadherin, it is known to decrease in the Endothelial to Mesenchymal
(EndMT) transition from BECs to TuBECs (Marín-Ramos et al., 2019). Because β-catenin is
associated with cadherins, its staining in the cytoplasm could be due to its release due to
dissociation from the plasma membrane adherens junction. Also, studies have found that β-
catenin reduces vascular density, and normalizes vessels via the canonical Wnt/β-Catenin
pathway thanks to Wnt ligand expression from the tumor (Reis et al., 2012). Hence, the
cytoplasmic staining as seen in vitro could help explain an unstable β-catenin and hence
increased vascular density and abnormal vessels. However, localization of β-catenin at the
29
nucleus is yet to be determined experimentally to confirm involvement of the Wnt/β-
Cateninpathway. When treated with NEO212 and NEO100, its cytoplasmic expression remains.
Whether this increase of β-catenin affects EndMT transition via α actin cytoskeleton
rearrangement is yet to be understood and explored in future studies.
Other markers including cell adhesion molecules ( CD44,E-Selectin,ICAM), integrins, and
ZO-1 did not show comparable results both in vivo and in vitro and therefore were not described
in this study (Supplementary Figure B and C). This could be due to varying reasons including the
antibody reactivity (Supplementary Figure A) being reactive to mouse and causing varying non-
specific effects, or not being reactive to human as is the case for in vitro human BECs and
TuBECs. None of the integrins demonstrated significant results in vitro to support in vivo results.
Hence this could be due to the non-specific availability of ligand in rat collagen as is the case for
αvβ3 which binds to vitronectin or fibronectin for example (Butler, Williams, & Blystone, 2003).
Although collagen is another binding extracellular matrix for some integrins, the integrins used in
this study could have not been expressed due to the unavailability of their ligand due to monetary
reasons. Additionally, integrins like β1 Integrin or the former αvβ3 have a reactivity for human,
which could help explain the lack of appearance when stained on mouse brain tissue.
Future Direction
For future direction ,several additional assays and changes could be done in this study to
support the results. For example, it is important to asses the quantitative expression of proteins
in normal and tumor endothelial cells using western blot (WB) that can confirm protein expression,
and then qPCR to accurately depict mRNA expression of these proteins. These two studies can
validate the qualitative assay of immunostaining.
In order to re-test the markers (claudin 6, β-Catenin, claudin 5, and JAM-1) using western
blot, appropriate antibodies should be assessed. The antibody should have the appropriate
30
reactivity either it be for mouse or for human. Additionally, the reagent source could have had an
effect in the staining results of this study both in vitro and in vivo. The dilution for antibodies used
was 1:100, so increasing the concentration of antibody in the dilutions to 1:50, for example, could
increase the accuracy of the stain and resolve if there are any false negative results as seen for
markers in supplementary figure C. Likewise, decreasing the concentration of antibody in the
dilutions to 1:200 could also resolve any false positive results due to non specific binding of the
antibody due to its concentration. Lastly, the age of the antibodies used was within 5 years range,
so to preserve the integrity of the antibody and its function, an antibody not older than 1 year
could be used to increase the accuracy of results and binding to the epitope of the marker.
In order for the study to be translational to human, it is important to study human related
GBM endothelial cells surrounded by a GBM-like microenvironment. So instead of testing mouse
brain tissue from in vivo experiments, a human endothelial cell could be co-cultured in the
appropriate BBB ECM surrounded by members of the BBB like human derived pericytes,
astrocytes, and neurons, to properly analyze what effect these have on the expression of tight
junction, adherens junctions, and integrin markers in GBM associated endothelial cells. This will
ensure that the model is closely related to a human brain tissue rather than a mouse brain tissue.
Additionally, this latter model could be co-cultured with glioma stem cells and glioma cells to
understand the tumor interplay with endothelial cells
To study the molecular pathway involved with the increase of β-Catenin, this study could
gear to understand if EndMT is associated with the Wnt/β-Catenin pathway and how this pathway
could interact with the current pathway known to contribute to this transition the TGF-β/Notch
pathway (Marín-Ramos et al., 2019). Additionally, the expression of claudin 6 and 5 could
influence cellular proliferation of endothelial cells since claudin 5,7,and 18 are involved in
suppressing proliferation of lung cellular carcinoma via inhibition of the AKT pathway and
31
upregulation of cyclin -dependent kinase inhibitor (CKI), p21, suppressor of the cell cycle (Akizuki,
Shimobaba, Matsunaga, Endo, & Ikari, 2017). So studying the AKT pathway and how these
markers influence it could be involved in cell proliferation of tumor endothelial cells. Since JAM-1
is also involved in the tight junction complex of endothelial cells, JAM-1 could also be studied to
see its involvement in the AKT pathway and how this influences endothelial cell proliferation. All
these studies should be tested while cells are treated with NEO212, NEO100, and Vehicle
measured by time to understand the changes of protein expression with time rather than imaged
on one time point. Using treatment while studying the expression of these different cells would
help translate to clinical work and be employed in real life patients in an optimistic future. Overall,
NEO212 is a new conjugate drug treatment that is promising to help revert TuBECs to BECs in
hopes to maintain blood vessel integrity in GBM
Supplementary Data
Supplementary Figure A: (From Materials and Methods: Cell reagents): A list of the antibodies used including
secondary antibodies.
Antibody Name Catalog # Company Name Host Reactivity
β-Catenin SC1496R SantaCruz Biotechnology Rabbit Mouse,Rat
Human
Claudin 3 ab15102 abcam Rabbit Mouse,Rat,
Human
Claudin 5 ab15106 abcam Rabbit Human, Mouse
Claudin 6 ab107059 abcam Rabbit Human, Mouse
ZO-1 ab2272 Millipore Sigma Rabbit Human, Mouse
JAM-1 ab180821 abcam Rabbit Mouse,Rat,
Human
PECAM (CD31) SC1506 SantaCruz Biotechnology Goat Human,
Mouse,Rat
E-Selectin 553749 BD Pharmigen Rat Mouse
ICAM (CD54) 202405 BioLegend Mouse Rat
CD44 559046 BD Pharmigen Mouse Human
Vimentin SC32322 SantaCruz Biotechnology Mouse Mouse,Rat,
Human
α-6 Integrin 3750S Cell Signaling Technology Rabbit Human, Mouse,
Rabbit, Monkey
αvβ3 Integrin MAB1976 Millipore Sigma Mouse Human
β1 Integrin SC8978 SantaCruz Biotechnology Rabbit Human
32
A
Blood Vessel Overview
Blood Vessel Overview Blood Vessel Overview
Secondary Only
40x
40x
40x
NB
NB
NB
NB
NB
NB
T T
T
T
T
T
Blood Vessel Overview
33
B
40x
40x
40x
Secondary Only
Blood Vessel Overview Blood Vessel Overview
Blood Vessel Overview
NB NB
NB NB
NB
NB
T
T
T
T
T T
34
Figure B (above): Photomicrograph sections of different markers stained using immunohistochemical
analysis. On the top row of every marker is a 40x view of mouse normal brain, tumor treated with vehicle, and tumor
treated with NEO212. On bottom row of each marker is a zoomed (54x) view of blood vessels composing a normal
brain versus tumor treated vehicle or NEO212. Tissue was counterstained with hematoxylin to show nucleus. On the
bottom right of center and right images for each marker depicted are tumor cells with larger nuclei and detached cells.
Positive staining is seen as brown while no staining is blue A. No staining was observed in vivo for any of the integrin
markers. This could be due to non-specific reactivity of two of the antibodies (αvβ3 Integrin and β1 integrin) or the
coating gel on which cells were grown on. With arrows are pointing at supposed areas of microvessels that stain
negative for integrin marker ( blue hue in stain). B. E-Selectin (CD62; endothelial leukocyte adhesion molecule) is a
cellular adhesion molecule that is involved with binding inflammatory innate cell, particularly neutrophils (Hess,
Thompson, Sprinkle, Carroll, & Smith, 1996) and is enhanced by supernatants released by glioma cells(Tabatabai et
al., 2008).Interestingly, E-selectin does show brown stain in blood vessels ( bottom left row ;arrow) and decreases
staining ( brown) in tumor blood vessels (also arrow) even if treated with NEO212. The overview of the tissue (at 40x)
does not show microvessel staining. However, in tumor, blood vessels (bottom center and bottom right image) shows
faint positively stained for E selectin blood vessels. E selectin is seen stained in vivo but not in vitro because of the
surrounding microenvironment like leukocytes that are ligands to E-selectin for extravasation into adjacent tissue.
Similarily, CD44 is a transmembrane protein known to be a marker for glial stem cells (Chen, Zhao, Karnad, & Freeman,
2018), however, not much is known of the expression in endothelial cells.CD44 is not stained in normal blood vesels
(top and bottom right; arrow), however, the expression is greatly increased in tumor blood vessels and remains even
after treatment with NEO212 (top and bottom center and right;arrows). Normal peri-tumoral tissue (top center and top
right; star) also show an increase of CD44 expression however whether this stain is on endothelial cells is still unclear.
ICAM is also a surface glycoprotein expressed as a ligand for an integrin found on leukocytes. It is also studied and
T
N
N
N
N
Overview
Blood Vessel Overview
Blood Vessel Overview
40x
40x
NB
NB
NB
NB
T
T
Secondary Only
35
found to be significant levels of expression in intra-tumoral vascular endothelial cells(Kuppner, van Meir, Hamou, & de
Tribolet, 1990). Shown in picture is positive staining in normal blood vessel ( bottom left ; arrow) however, 40x
magnification does not show micro blood vessel stain. (Top and bottom center; arrow) show an increase in staining of
ICAM in tumoral blood vessels and this expression decreases when treated with NEO212 (top and bottom right; arrow).
Peritumoral tissue also shows an increased stain from normal to tumor tissue and decreases back when treated with
NEO212 (star). Lastly, ZO1- does not show staining in any of three conditions blood vessels shown with arrows. Even
thou these markers, including claudin 3 show staining differences from normal brain to tumor associated brain blood
vessels, no staining was observed for the in vitro analysis which does not prove of endothelial cell specificity of stain.
This lack of staining could be affected by the absence of microenvironment and inflammatory cells in vitro. NB is normal
brain and T is tumor. Images represent best image of three independent experiments (n=3).
Figure C (below) Immunocytochemistry (ICC) comparing BECs and TuBECs in different treatment conditions
including untreated, vehicle-treated, NEO100 treated, and NEO212 treated. Although some markers expressed
expression in vivo, results were not coherent with ICC results, which could mean the stain is not representative of BEC
or TuBEC specificity. A. All three integrins show no positive staining (brown color) for BEC or TuBEC in three conditions.
Instead no marker is observed cytoplasmically or on cell surfaces (arrow).B. Cell Adhesion Molecules (CD44, E-
Selectin, or ICAM) likely, do not show any staining on the cell surface nor cytoplasmically (arrow) for either BEC or
TuBEC. This could be due to the absence of leukocytes (for E-Selectin OR ICAM), hyaluronan or collagens (for CD44)
- as ligands that enable the expression of these on either tumor or normal endothelial cell. Although claudin 3 is
expressed in the BBB as determined by studies, there is no positive staining (no brown) in vitro. Arrows point at higher
cluster of cells in close contact where tight junction would form. However, there is expression of claudin 3 even in cell-
cell junctions. Similarly, ZO-1 is not stained. The absence of proper tight junction formation could affect the expression
of ZO-1 cytoplasmically (arrows point blue signifying absence of staining). Negative control (far right) show no
staining (all blue).
α6 Integrin αvβ3 Integrin
BEC TuBEC BEC TuBEC
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
Secondary Only
A
40x
40x
36
Claudin 3
β1 Integrin CD44
BEC TuBEC BEC TuBEC
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x
40x 40x
40x
40x
40x
40x
BEC TuBEC
40x
40x
B
Secondary Only
37
E Selectin
ICAM
BEC TuBEC
40x 40x 40x
40x
40x
40x
40x
40x
40x 40x 40x
40x 40x
40x
40x 40x
40x 40x
BEC TuBEC BEC TuBEC
ZO-1
Secondary Only
40x
40x
40x
C
38
References
Abbott, N. J. (2002). Astrocyte-endothelial interactions and blood-brain barrier permeability. J
Anat, 200(6), 629-638. doi:10.1046/j.1469-7580.2002.00064.x
Banks, W. A. (2009). Characteristics of compounds that cross the blood-brain barrier. BMC
Neurol, 9 Suppl 1, S3. doi:10.1186/1471-2377-9-S1-S3
Bauer, H. C., Krizbai, I. A., Bauer, H., & Traweger, A. (2014). "You Shall Not Pass"-tight junctions
of the blood brain barrier. Front Neurosci, 8, 392. doi:10.3389/fnins.2014.00392
Charalambous, C., Hofman, F. M., & Chen, T. C. (2005). Functional and phenotypic differences
between glioblastoma multiforme-derived and normal human brain endothelial cells. J
Neurosurg, 102(4), 699-705. doi:10.3171/jns.2005.102.4.0699
Cho, H. Y., Wang, W., Jhaveri, N., Lee, D. J., Sharma, N., Dubeau, L., . . . Chen, T. C. (2014).
NEO212, temozolomide conjugated to perillyl alcohol, is a novel drug for effective
treatment of a broad range of temozolomide-resistant gliomas. Mol Cancer Ther, 13(8),
2004-2017. doi:10.1158/1535-7163.MCT-13-0964
Dejana, E., Hirschi, K. K., & Simons, M. (2017). The molecular basis of endothelial cell plasticity.
Nat Commun, 8, 14361. doi:10.1038/ncomms14361
Dudley, A. C. (2012). Tumor endothelial cells. Cold Spring Harb Perspect Med, 2(3), a006536.
doi:10.1101/cshperspect.a006536
Gerhardt, H., Golding, M., Fruttiger, M., Ruhrberg, C., Lundkvist, A., Abramsson, A., . . . Betsholtz,
C. (2003). VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell
Biol, 161(6), 1163-1177. doi:10.1083/jcb.200302047
Gumbiner, B. (1987). Structure, biochemistry, and assembly of epithelial tight junctions. Am J
Physiol, 253(6 Pt 1), C749-758. doi:10.1152/ajpcell.1987.253.6.C749
Hanif, F., Muzaffar, K., Perveen, K., Malhi, S. M., & Simjee, S. U. (2017). Glioblastoma Multiforme:
A Review of its Epidemiology and Pathogenesis through Clinical Presentation and
Treatment. Asian Pac J Cancer Prev, 18(1), 3-9. doi:10.22034/APJCP.2017.18.1.3
Kalluri, R., & Weinberg, R. A. (2009). The basics of epithelial-mesenchymal transition. J Clin
Invest, 119(6), 1420-1428. doi:10.1172/JCI39104
Karnati, H. K., Panigrahi, M., Shaik, N. A., Greig, N. H., Bagadi, S. A., Kamal, M. A., & Kapalavayi,
N. (2014). Down regulated expression of Claudin-1 and Claudin-5 and up regulation of β-
catenin: association with human glioma progression. CNS Neurol Disord Drug Targets,
13(8), 1413-1426. doi:10.2174/1871527313666141023121550
Lanevskij, K., Japertas, P., & Didziapetris, R. (2013). Improving the prediction of drug disposition
in the brain. Expert Opin Drug Metab Toxicol, 9(4), 473-486.
doi:10.1517/17425255.2013.754423
39
Le Rhun, E., Preusser, M., Roth, P., Reardon, D. A., van den Bent, M., Wen, P., . . . Weller, M.
(2019). Molecular targeted therapy of glioblastoma. Cancer Treat Rev, 80, 101896.
doi:10.1016/j.ctrv.2019.101896
Marín-Ramos, N. I., Jhaveri, N., Thein, T. Z., Fayngor, R. A., Chen, T. C., & Hofman, F. M. (2019).
NEO212, a conjugate of temozolomide and perillyl alcohol, blocks the endothelial-to-
mesenchymal transition in tumor-associated brain endothelial cells in glioblastoma.
Cancer Lett, 442, 170-180. doi:10.1016/j.canlet.2018.10.034
Muoio, V., Persson, P. B., & Sendeski, M. M. (2014). The neurovascular unit - concept review.
Acta Physiol (Oxf), 210(4), 790-798. doi:10.1111/apha.12250
Potenta, S., Zeisberg, E., & Kalluri, R. (2008). The role of endothelial-to-mesenchymal transition
in cancer progression. Br J Cancer, 99(9), 1375-1379. doi:10.1038/sj.bjc.6604662
Salvador, E., Burek, M., & Förster, C. Y. (2016). Tight Junctions and the Tumor Microenvironment.
Curr Pathobiol Rep, 4, 135-145. doi:10.1007/s40139-016-0106-6
Schwartzbaum, J. A., Fisher, J. L., Aldape, K. D., & Wrensch, M. (2006). Epidemiology and
molecular pathology of glioma. Nat Clin Pract Neurol, 2(9), 494-503; quiz 491 p following
516. doi:10.1038/ncpneuro0289
Sladojevic, N., Stamatovic, S. M., Johnson, A. M., Choi, J., Hu, A., Dithmer, S., . . . Andjelkovic,
A. V. (2019). Claudin-1-Dependent Destabilization of the Blood-Brain Barrier in Chronic
Stroke. J Neurosci, 39(4), 743-757. doi:10.1523/JNEUROSCI.1432-18.2018
Staehelin, A.L. Structure of Function of Intracellular Junctions. International Review of Cytology,
Volume 39. Edited by Bourne,G.H, Danielli,J.F. New York, New York, Academic Press,
1974, pp. 199-214
Stupp, R., Mason, W. P., van den Bent, M. J., Weller, M., Fisher, B., Taphoorn, M. J., . . . Group,
N. C. I. o. C. C. T. (2005). Radiotherapy plus concomitant and adjuvant temozolomide for
glioblastoma. N Engl J Med, 352(10), 987-996. doi:10.1056/NEJMoa043330
Swisshelm, K., Macek, R., & Kubbies, M. (2005). Role of claudins in tumorigenesis. Adv Drug
Deliv Rev, 57(6), 919-928. doi:10.1016/j.addr.2005.01.006
Tajes, M., Ramos-Fernández, E., Weng-Jiang, X., Bosch-Morató, M., Guivernau, B., Eraso-
Pichot, A., . . . Muñoz, F. J. (2014). The blood-brain barrier: structure, function and
therapeutic approaches to cross it. Mol Membr Biol, 31(5), 152-167.
doi:10.3109/09687688.2014.937468
Thakkar, J. P., Dolecek, T. A., Horbinski, C., Ostrom, Q. T., Lightner, D. D., Barnholtz-Sloan, J.
S., & Villano, J. L. (2014). Epidemiologic and molecular prognostic review of glioblastoma.
Cancer Epidemiol Biomarkers Prev, 23(10), 1985-1996. doi:10.1158/1055-9965.EPI-14-
0275
Zlokovic, B. V. (2008). The blood-brain barrier in health and chronic neurodegenerativedisorders.
Neuron, 57(2), 178-201. doi:10.1016/j.neuron.2008.01.003
40
Akizuki, R., Shimobaba, S., Matsunaga, T., Endo, S., & Ikari, A. (2017). Claudin-5, -7, and -18
suppress proliferation mediated by inhibition of phosphorylation of Akt in human lung
squamous cell carcinoma. Biochim Biophys Acta Mol Cell Res, 1864(2), 293-302.
doi:10.1016/j.bbamcr.2016.11.018
Berndt, P., Winkler, L., Cording, J., Breitkreuz-Korff, O., Rex, A., Dithmer, S., . . . Haseloff, R. F.
(2019). Tight junction proteins at the blood–brain barrier: far more than claudin-5.
Cellular and Molecular Life Sciences, 76(10), 1987-2002. doi:doi:10.1007/s00018-019-
03030-7
Butler, B., Williams, M. P., & Blystone, S. D. (2003). Ligand-dependent activation of integrin
alpha vbeta 3. J Biol Chem, 278(7), 5264-5270. doi:10.1074/jbc.M206997200
Charalambous, C., Hofman, F. M., & Chen, T. C. (2005). Functional and phenotypic differences
between glioblastoma multiforme-derived and normal human brain endothelial cells. J
Neurosurg, 102(4), 699-705. doi:10.3171/jns.2005.102.4.0699
Chen, C., Zhao, S., Karnad, A., & Freeman, J. W. (2018). The biology and role of CD44 in
cancer progression: therapeutic implications. Journal of Hematology & Oncology, 11(1),
1-23. doi:doi:10.1186/s13045-018-0605-5
Cho, H. Y., Wang, W., Jhaveri, N., Lee, D. J., Sharma, N., Dubeau, L., . . . Chen, T. C. (2014).
NEO212, temozolomide conjugated to perillyl alcohol, is a novel drug for effective
treatment of a broad range of temozolomide-resistant gliomas. Mol Cancer Ther, 13(8),
2004-2017. doi:10.1158/1535-7163.MCT-13-0964
Cho, H. Y., Wang, W., Jhaveri, N., Torres, S., Tseng, J., Leong, M. N., . . . Chen, T. C. (2012).
Perillyl alcohol for the treatment of temozolomide-resistant gliomas. Mol Cancer Ther,
11(11), 2462-2472. doi:10.1158/1535-7163.MCT-12-0321
Guo, M., Breslin, J. W., Wu, M. H., Gottardi, C. J., & Yuan, S. Y. (2008). VE-cadherin and beta-
catenin binding dynamics during histamine-induced endothelial hyperpermeability. Am J
Physiol Cell Physiol, 294(4), C977-984. doi:10.1152/ajpcell.90607.2007
Hess, D. C., Thompson, Y., Sprinkle, A., Carroll, J., & Smith, J. (1996). E-selectin expression on
human brain microvascular endothelial cells. Neurosci Lett, 213(1), 37-40.
doi:10.1016/0304-3940(96)12837-8
Kalluri, R., & Weinberg, R. A. (2009). The basics of epithelial-mesenchymal transition. J Clin
Invest, 119(6), 1420-1428. doi:10.1172/JCI39104
Kuppner, M. C., van Meir, E., Hamou, M. F., & de Tribolet, N. (1990). Cytokine regulation of
intercellular adhesion molecule-1 (ICAM-1) expression on human glioblastoma cells. Clin
Exp Immunol, 81(1), 142-148. doi:10.1111/j.1365-2249.1990.tb05305.x
Liebner, S., Fischmann, A., Rascher, G., Duffner, F., Grote, E. H., Kalbacher, H., & Wolburg, H.
(2000). Claudin-1 and claudin-5 expression and tight junction morphology are altered in
blood vessels of human glioblastoma multiforme. Acta Neuropathol, 100(3), 323-331.
doi:10.1007/s004010000180
41
Luissint, A. C., Artus, C., Glacial, F., Ganeshamoorthy, K., & Couraud, P. O. (2012). Tight
junctions at the blood brain barrier: physiological architecture and disease-associated
dysregulation. Fluids Barriers CNS, 9(1), 23. doi:10.1186/2045-8118-9-23
Malric, L., Monferran, S., Gilhodes, J., Boyrie, S., Dahan, P., Skuli, N., . . . Lemarié, A. (2017).
Interest of integrins targeting in glioblastoma according to tumor heterogeneity and
cancer stem cell paradigm: an update. Oncotarget, 8(49), 86947-86968.
doi:10.18632/oncotarget.20372
Marín-Ramos, N. I., Jhaveri, N., Thein, T. Z., Fayngor, R. A., Chen, T. C., & Hofman, F. M.
(2019). NEO212, a conjugate of temozolomide and perillyl alcohol, blocks the
endothelial-to-mesenchymal transition in tumor-associated brain endothelial cells in
glioblastoma. Cancer Lett, 442, 170-180. doi:10.1016/j.canlet.2018.10.034
Reis, M., Czupalla, C. J., Ziegler, N., Devraj, K., Zinke, J., Seidel, S., . . . Liebner, S. (2012).
Endothelial Wnt/β-catenin signaling inhibits glioma angiogenesis and normalizes tumor
blood vessels by inducing PDGF-B expression. J Exp Med, 209(9), 1611-1627.
doi:10.1084/jem.20111580
Schnell, O., Krebs, B., Wagner, E., Romagna, A., Beer, A. J., Grau, S. J., . . . Goldbrunner, R.
H. (2008). Expression of integrin alphavbeta3 in gliomas correlates with tumor grade and
is not restricted to tumor vasculature. Brain Pathol, 18(3), 378-386. doi:10.1111/j.1750-
3639.2008.00137.x
Schrade, A., Sade, H., Couraud, P. O., Romero, I. A., Weksler, B. B., & Niewoehner, J. (2012).
Expression and localization of claudins-3 and -12 in transformed human brain
endothelium. Fluids Barriers CNS, 9, 6. doi:10.1186/2045-8118-9-6
Stamatovic, S. M., Johnson, A. M., Keep, R. F., & Andjelkovic, A. V. (2016). Junctional proteins
of the blood-brain barrier: New insights into function and dysfunction. Tissue Barriers,
4(1), e1154641. doi:10.1080/21688370.2016.1154641
Stamatovic, S. M., Keep, R. F., & Andjelkovic, A. V. (2008). Brain endothelial cell-cell junctions:
how to "open" the blood brain barrier. Curr Neuropharmacol, 6(3), 179-192.
doi:10.2174/157015908785777210
Stupack, D. G., & Cheresh, D. A. (2004). Integrins and angiogenesis. Curr Top Dev Biol, 64,
207-238. doi:10.1016/S0070-2153(04)64009-9
Sullivan, L. M., Yankovich, T., Le, P., Martinez, D., Santi, M., Biegel, J. A., . . . Judkins, A. R.
(2012). Claudin-6 is a nonspecific marker for malignant rhabdoid and other pediatric
tumors. Am J Surg Pathol, 36(1), 73-80. doi:10.1097/PAS.0b013e31822cfa7e
Tabatabai, G., Herrmann, C., von Kurthy, G., Mittelbronn, M., Grau, S., Frank, B., . . . Wick, W.
(2008). VEGF-dependent induction of CD62E on endothelial cells mediates glioma
tropism of adult haematopoietic progenitor cells. Brain, 131(Pt 10), 2579-2595.
doi:10.1093/brain/awn182
Takada, Y., Ye, X., & Simon, S. (2007). The integrins. Genome Biol, 8(5), 215. doi:10.1186/gb-
2007-8-5-215
42
Tietz, S., & Engelhardt, B. (2015). Brain barriers: Crosstalk between complex tight junctions and
adherens junctions. In J Cell Biol (Vol. 209, pp. 493-506).
Yeruva, L., Pierre, K. J., Elegbede, A., Wang, R. C., & Carper, S. W. (2007). Perillyl alcohol and
perillic acid induced cell cycle arrest and apoptosis in non small cell lung cancer cells.
Cancer Lett, 257(2), 216-226. doi:10.1016/j.canlet.2007.07.020
Zeng, L., Fee, B. E., Rivas, M. V., Lin, J., & Adamson, D. C. (2014). Adherens junctional
associated protein-1: a novel 1p36 tumor suppressor candidate in gliomas (Review). Int
J Oncol, 45(1), 13-17. doi:10.3892/ijo.2014.2425
Abstract (if available)
Abstract
Glioblastoma Multiforme (GBM) is an astrocytoma with high aggressiveness and resistance known to have extensive number of vessels. The increase in vessel formation during GBM happens via the dissociation of cell-cell junction proteins which maintain the blood-brain-barrier (BBB) and maintain vessel integrity from incoming macromolecules, toxins, and xenobiotics. The loss of tight junctions and adherens junctions contributes to this dissociation while integrins allow for endothelial cell migration and angiogenesis. Part of the transition from normal to leaky vessels is due to an increase in mesenchymal markers in endothelial cells, thus rendering this transition as endothelial to mesenchymal transition (EndMT) from brain endothelial cells (BECs) to tumor-associated endothelial cells (TuBECs). This study uses immunocytochemistry and immunohistochemistry as methods to study a panel of markers comparing BEC versus TuBEC. Additionally, the study uses NEO212 as a tool to understand if this novel conjugate drug can normalize phenotypic differences from TuBEC to BEC. Hence, the study finds that claudin 6, a marker not yet associated with the BBB increases in tumor endothelial cells, while Claudin 5 and 3 increase rather than decrease as expected from other studies. Further β-Catenin, a cytoplasmic protein associated with cadherin adherens junctions, increases cytoplasmically. However, when treated with NEO212, this expression as well as the expression of claudin 6 decreases back to similar yet not same intensity to BECs. This shows promising results to employ NEO212 in further studies for GBM treatment, as well as focus on novel markers to better understand EndMT transition from BEC to TuBEC.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme
PDF
The role of tumor angiogenesis in glioblastoma multiforme and implications for anticancer therapy
PDF
Molecular targets for treatment of glioblastoma multiforme
PDF
Interleukin-11: a study of its effects in glioblastoma multiforme
PDF
The opening of the blood brain barrier by homogenized perillyl alcohol
PDF
Current models of non-homologous end joining and their implications in gene therapy
PDF
Glioblastoma treatment with 2,5-dimethyl-celecoxib (DMC) in vitro: - effects of additional chemotherapeutic drugs and tumor microenvironment, - inconsistencies among commonly used in vitro cell g...
PDF
Tight junction protein CLDN18.1 attenuates malignant properties and related signaling pathways of human lung adenocarcinoma in vivo and in vitro
Asset Metadata
Creator
Martinez, Aida
(author)
Core Title
Analysis of human brain endothelial cells versus human tumor associated brain endothelial cell tight junction and adherens junction expression differences in glioblastoma multiforme
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Experimental and Molecular Pathology
Publication Date
05/07/2020
Defense Date
05/05/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
brain cancer,GBM,glioblastoma multiforme,OAI-PMH Harvest,tight junctions
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Hofman, Florence (
committee chair
), Ouelette, Andre (
committee member
), Widelitz, Randall (
committee member
)
Creator Email
aidamart@usc.edu,mrtza0993@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-299036
Unique identifier
UC11665707
Identifier
etd-MartinezAi-8446.pdf (filename),usctheses-c89-299036 (legacy record id)
Legacy Identifier
etd-MartinezAi-8446.pdf
Dmrecord
299036
Document Type
Thesis
Rights
Martinez, Aida
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
brain cancer
GBM
glioblastoma multiforme
tight junctions