University of Southern California Dissertations and Theses

Maternal inflammation disrupts fetal blood-brain barrier formation via cyclooxygenase activation

(USC Thesis Other)

Maternal inflammation disrupts fetal blood-brain barrier formation via cyclooxygenase activation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

MATERNAL INFLAMMATION DISRUPTS FETAL BLOOD-BRAIN BARRIER
FORMATION VIA CYCLOOXYGENASE ACTIVATION

by

Weiye Dai

A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF PHARMACY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(MOLECULAR PHARMACOLOGY AND TOXICOLOGY)

August 2022

Copyright 2022 Weiye Dai
ii
ACKNOWLEDGEMENTS
There is no way I could have reached this stage in my academic career without the help
and support of many. My appreciation and gratitude towards them go beyond what I can put
down in words in this short section.
I would like to first thank Dr. Alexandre Bonnin and Dr. Qiuying Zhao, from whom I
received the most amount of help throughout the past few years. Dr. Bonnin has been an amazing
mentor for me since 2019 when I was a technician in his lab at University of Southern California
(USC), Keck School of Medicine, and I highly appreciate the opportunity to carry out this
Master thesis project under his guidance. Not only has he always been patient in guiding me
through obstacles in research, but he also trusts and encourages me to come up with my own
solutions and explanations for troubleshooting. Whenever I needed his help for something in the
lab, he always tried his best to help, whether it’s about an experiment, the data, or an equipment
issue. Most importantly, I was greatly inspired and motivated seeing his studying spirit and his
optimistic mindset when it comes to research, which made my graduate school journey full of
excitements. I am also thankful for Dr. Zhao, who has been the best trainer and coworker I could
ever ask for. She was always there when I needed a helping hand and offered me insightful
advice from her past experiences in the scientific field. And I admire her for the hard-working
spirit, high efficiency, endurance, kindness, and consideration for others that she has showed me.
In addition, I would like to thank all the faculty members from USC School of Pharmacy
who helped me along the way from various aspects. I am grateful for all the professors and
scientists who enthusiastically presented in classes to pass on the knowledge and inspire my
desire for solving the unknown. I am also fortunate to have Dr. Sarah Hamm-Alvarez and Dr.
iii
Jean Chen Shih to kindly serve on my thesis committee, providing me with meaningful feedback
and suggestions.
I would also like to thank all the other staff, scientists, and students that I’ve encountered
or worked with for any type of favor they did for me, my family and relatives for supporting me
mentally and financially, my cohort for going through all the stressful, frustrating, and anxious
moments together throughout the curriculum, and all my close friends for backing me up and
entertaining me. Last but certainly not least, I cannot thank my significant other enough for
always encouraging me to find the confidence and courage in me to do what I want to do,
validating my feelings and thoughts when I doubted myself, and trying best to cheer me up and
bring my smile back.
The past two years has not been an easy journey, especially when the pandemic makes
everything much more complicated and inconvenient. It is therefore my great honor to have
received the support I needed, continuing to grow and learn in my life.

iv
TABLE OF CONTENTS
Acknowledgements..........................................................................................................................ii
List of Tables..................................................................................................................................vi
List of Figures................................................................................................................................vii
Abbreviations................................................................................................................................viii
Abstract............................................................................................................................................x
Chapter 1: Introduction and Literature Review...............................................................................1
1.1 Maternal Inflammation and Immune Activation...........................................................1
1.2 Defensive Mechanisms for Fetal Brain Development...................................................6
1.2.1 The blood-brain barrier...................................................................................6
1.2.2 The blood-placenta barrier..............................................................................8
1.3 The Use of Poly(I:C) in Mice and Relevant Issues.....................................................11
Chapter 2: Materials and Methods.................................................................................................14
2.1 Animals........................................................................................................................14
2.1.1 Timed pregnancy..........................................................................................14
2.1.2 Gestational MIA...........................................................................................15
2.1.3 Genotyping....................................................................................................16
2.2 Injections......................................................................................................................18
2.2.1 Poly(I:C).......................................................................................................18
2.2.2 Celecoxib......................................................................................................18
2.3 Histology and Staining.................................................................................................19
2.3.1 Dissection and tissue harvest.........................................................................19
2.3.2 Tissue processing..........................................................................................19
2.3.3 Sectioning.....................................................................................................20
2.3.4 Hematoxylin and eosin staining....................................................................20
2.3.5 Immunohistochemistry staining....................................................................21
2.4 Image Quantification and Analysis..............................................................................23
2.4.1 Microscopy and quantification......................................................................23
2.4.2 Statistics........................................................................................................23
Chapter 3: Presentation of Research Results..................................................................................24
3.1 Gestational MIA Acutely Disrupts Fetal Brain Development......................................24
3.2 Poly(I:C) L/HMW Dosage Determination in C57BL/6J Mouse Strain........................28
3.3 Prevention of Gestational MIA Disruption in Fetal Brain............................................36
Chapter 4: Discussion and Conclusion...........................................................................................40
4.1 Summary of Experimental Data...................................................................................40
4.2 Limitations...................................................................................................................41
4.3 Implications and Future Directions..............................................................................44
v
Bibliography ..................................................................................................................................46

vi
LIST OF TABLES
Table 1. Number of litters and fetuses included for each sub-project...........................................14
Table 2. PCR conditions, primer sequences, and expected band weights.....................................17
Table 3. Constitutions of LMW and HMW poly(I:C) solutions for L/HMW mixture dosages
tested................................................................................................................................18
Table 4. Primary and secondary antibodies used..........................................................................22
Table 5. Summary of number of embryos survived MIA for three poly(I:C) dosages.................29

vii
LIST OF FIGURES
Figure 1. Gestational brain development timeline comparison between mouse and human........2
Figure 2. Demonstration of how the cre/loxP system works in the COX2 conditional
knockout mice................................................................................................................5
Figure 3. Representation of blood-brain barrier............................................................................7
Figure 4. Complex structures of mature placenta.........................................................................10
Figure 5. Representative image of MBS in mouse placenta at GD 15........................................13
Figure 6. Transgenic mice mating scheme..................................................................................15
Figure 7. MIA induced decreased pericyte coverage on BV endothelial cells in CD-1
fetuses at GD 15...........................................................................................................25
Figure 8. MIA induced hyper activation of Iba1+ microglia with or without COX2
expression in CD-1 fetuses at GD 15...........................................................................27
Figure 9. Representative embryo from dam tested with 12.8 mg/kg L/HMW poly(I:C)
mixture.........................................................................................................................29
Figure 10. Comparison of placentae histology from three dosages of L/HMW poly(I:C)
mixtures........................................................................................................................30
Figure 11. Quantification of MIA effect on MBS surface area.....................................................34
Figure 12. Quantification of Iba1+ COX2+ cells in GD 15 fetal brains.......................................35
Figure 13. CX prevents abnormal Iba1 and/or COX2 expression in microglia in fetal brains
of CD-1 mice at GD 15................................................................................................37
Figure 14. Conditional COX2 knockout lowers activation of Iba1+ microglia in fetal cortex
at GD 15.......................................................................................................................38
Figure 15. Quantification of pericyte coverage of BV endothelial cells.......................................39
Figure 16. Quantification of TJ markers in fetal BBB..................................................................42

viii
ABBREVIATIONS
AD
ASD
BBB
BPB
BV
Cll5
CNS
COX
CX
dsRNA
(f)MRI
GD
HMW
Iba1
IHC
IL
LMW
LPS
MBS
MCP-1
MIA
NSAID
Alzheimer’s disease
Autism spectrum disorder
Blood-brain barrier
Blood-placenta barrier
Blood vessel
Claudin-5
Central nervous system
Cyclooxygenase
Celecoxib
Double-stranded RNA
(Functional) magnetic resonance imaging
Gestational day
High molecular weight
Ionized calcium binding adaptor molecule 1
Immunohistochemistry
Interleukin
Low molecular weight
Lipopolysaccharide
Maternal blood space
Monocyte chemoattractant protein 1
Maternal immune activation
Non-steroidal anti-inflammatory drug
ix
PBS
PD
PDGFRβ
PeCAM
PG
Poly(I:C)
SMA
TJ
TNF-α
WT
ZO1
Phosphate-buffered saline
Postnatal day
Platelet-derived growth factor receptor beta
Platelet endothelial cell adhesion molecule
Prostaglandin
Polyriboinosinic-polyribocytidylic acid
Smooth muscle actin
Tight junction
Tumor necrosis factor alpha
Wild type
Zonula occludens-1
x
ABSTRACT
Maternal immune activation (MIA) during pregnancy induces many significant
neuropathologies in the offspring. Using a synthetic double stranded RNA (dsRNA) molecule,
polyriboinosinic-polyribocytidylic acid [poly(I:C)], which mimics virus infection in animal
models, it has been discovered that the abnormal neurodevelopmental phenotypes occur both
acutely during gestational stages and persistently into adulthood. However, the underlying
mechanism for such phenotypes are yet to be uncovered. In our lab, we seek to identify key
pathways and/or molecules responsible for the abnormalities and provide some insight for future
studies on gestational MIA treatment development.
We developed our initial MIA model using CD-1 mice and observed increased
permeability and structural disruption of fetal blood-brain barrier (BBB), increased number of
microglial cells displaying increased expression and activation of cyclooxygenase 2 (COX2)
enzyme. This led to our hypothesis that activation of COX2-expressing microglia plays an
important role in MIA impact on BBB formation. We tested this hypothesis using
pharmacological and genetic approaches to suppress COX2 activity at time of MIA induction.
Injection of celecoxib, a selective COX2 inhibitor, prevented all the abnormalities observed in
MIA fetuses.
To investigate the role of COX2 expression more specifically in microglia, we next used
a conditional COX2 knockout mouse model. Due to these mice being on a C57BL/6J genetic
background, we first determined an optimal poly(I:C) dosage to trigger MIA phenotypes
comparable to those obtained in CD-1 mice. We found that genetically deleting COX2 gene in
microglia prevented all effects on the fetal BBB.
xi
Taken together, our data demonstrate that induction of COX2 expression in microglia
serves a significant part in MIA disruption of fetal BBB. These findings provide promising new
directions for research on early life treatments to prevent neuropsychiatric pathologies stemming
from prenatal MIA.

1
CHAPTER 1: Introduction and Literature Review
1.1 Maternal inflammation and immune activation
Inflammation during pregnancy, such as triggered by viral infections, has never been an
ideal scenario for conceiving women, but in many cases, it turns out to be inevitable. It is also
well-established that such inflammatory events can lead to fetal developmental issues and
increase the risk in the offspring for developing neurological diseases and psychiatric disorders
later in life. Therefore, how maternal immune activation (MIA) induced by inflammation during
pregnancy can affect offspring development has been an important research topic throughout the
years in both clinical and preclinical settings. However, there are still many unanswered
questions such as the precise cellular pathways and mechanisms in the development of MIA
outcomes.
There are many factors that can affect the neurodevelopmental result in the offspring
following an inflammatory event during pregnancy, including the type of MIA-inducing agents,
the duration and severity of MIA, the gestational stage when MIA happens, etc. Studies have
been done to determine the influential factors, but interestingly, the MIA inducing agents per se
doesn't always dictate specific outcomes in the offspring exposed to MIA (Smolders et al., 2018).
By comparing various MIA studies using different immune-activating agents including virus,
polyinosinic:polycytidylic acid [poly(I:C)], and lipopolysaccharide (LPS), Smolders et al.
concluded that only the general maternal and fetal immune responses to the inflammatory events
have determining impact on the abnormalities observed in MIA offspring. For example, like
during any immune responses, all kinds of cytokines are secreted in maternal and fetal
compartments during MIA as well, serving an important role to trigger impaired
neurodevelopment in the fetus (Boulanger-Bertolus et al., 2018). Experiments have shown that
2
regardless of the occurrence of an induced inflammation, the presence of excess interleukins (IL)
such as IL-1β and IL-6 may determine outcomes in offspring (Deverman & Patterson, 2009).
The gestational stage when MIA happens is another influencing factor as the fetal brain
development follows an orderly and genetically driven sequence of events that are affected by
immune responses in different ways. For example, in both mice and humans, neurogenesis in the
fetuses happens during the relatively earlier state of pregnancy, myelination during late gestation
continuing to postnatal development, and microglia invasion starting around early on and lasting
until term (Boulanger-Bertolus et al., 2018; Figure 1). Thus, mouse models have demonstrated
that MIA around gestational day (GD) 10 to 15 results in autism spectrum disorder (ASD)-like
phenotypes, which have been associated with cortical development defects, whereas MIA around
GD 18 to birth leads to cerebral palsy which is associated with myelination defects (Gumusoglu
& Stevens, 2019).

Figure 1. Gestational brain development timeline comparison between mouse and human.
The diagram depicts a simplified generalization of the time frame for each major
neurodevelopmental stages throughout early pregnancy, 3 to 4 weeks (wk) in human and GD 7.5
in mouse, till birth (Boulanger-Bertolus et al., 2018). (Created with BioRender.com)

3
When activated, immune cells express many small molecules and proteins that pass along
the inflammatory signal and help the immune system to combat the source(s) of inflammation.
Cyclooxygenases (COX) are enzymes that convert arachidonic acid to prostaglandins (PG), and
the two isoforms characterized to be activated by inflammatory stimuli, COX1 and COX2, have
different mechanisms in response to inflammation (Phillis et al., 2006). COX1 is believed to be
the primary enzyme responsible for maintaining prostaglandin secretion under homeostasis,
whereas COX2 is mainly induced during an inflammatory event (Choi et al., 2009). Downstream
of COX, PGH2 is synthesized and gets converted to various isomerases depending on the tissue
types. Specifically, in brains, the predominant product is PGE2, which, when bound to
corresponding receptors, mediates many physiological responses in immune regulation. For
example, the arteries are dilated, and vascular permeability is increased to allow increased blood
flow into the region(s) of inflammation, causing the classic inflammatory signs of redness and
swelling (Ricciotti & Fitzgerald, 2011). The multiplicity of PGs effects depends on
combinatorial expression of four different receptors in the target regions (Foudi et al., 2012).
One of the key components in immune responses that influences neurodevelopment is the
activation of microglia, as they are derived from primitive yolk-sac macrophages that invade the
brain during early fetal development and thus are considered to be the “resident immune cells of
the CNS” (Ginhoux et al., 2010). Thus, we considered that COX2 expression in microglia is a
good indicator of neuroinflammation, and quantification of active COX2 positive microglia can
be used as one type of measurements to determine the severity of abnormal fetal
neurodevelopment induced by MIA.
Due to the natural functions of COX1 and COX2 enzymes, it is commonly believed that
selective inhibitors of COX2 are safer yet still effective anti-inflammatory agents, leaving COX1
4
activity unaffected for its primary function (Choi et al., 2009). For instance, celecoxib (CX), a
non-steroidal anti-inflammatory drug (NSAID), is a selective inhibitor of COX2 and can pass on
to the fetus through placental transfer (Takahashi et al., 2000). Therefore, we considered
injecting CX following MIA stimulation to pharmacologically test the developmental result of
suppression of fetal neuroinflammation.
To further test if inhibition of COX2 can prevent the neuropathology observed in MIA
groups, a genetic approach can be used whereby COX2 gene (i.e., Ptgs2) is conditionally deleted
in microglia, without causing disruption to the immune system in other tissues throughout the
body. This is possible using mouse models where the Cre/loxP recombination system is
introduced by insertion of the cre cDNA into the endogenous M lysozyme loci (LysM-cre) that
are expressed specifically in myeloid cells, which transform into microglia in the CNS. The
portion of Ptgs2 sequence flanked by loxP sites (termed flox, denoted fl) gets cut off from the
genome in the specific cell type that expresses recombinase Cre (i.e., microglia) and the
corresponding protein (i.e., COX2) can no longer be expressed (Clausen et al., 1999; Figure 2).
It is expected that MIA effects on neurodevelopment will be prevented in both COX2
conditional knockout (COX2-MKO; LysM
Cre/+
, COX2
fl/fl
) mice and CX-treated mice. Similar
blocking effects of genetic and pharmacological inhibition would strongly support the hypothesis
that COX2 activation is a critical pathway causing MIA neurodevelopmental effects in the
offspring.
5

Figure 2. Demonstration of how the cre/loxP system works in the COX2 conditional
knockout mice. Only in cells with LysM expression will Cre be translated to delete COX2 gene,
resulting in no expression of COX2 from the specific cell type. (Created with BioRender.com)

6
1.2 Defensive mechanisms for fetal brain development
The blood-brain barrier (BBB) selectively filters the blood circulating in the brain to and
from the rest of the body, protecting the brain from potentially harmful molecules, whereas the
blood-placenta barrier (BPB) serves as a barrier between maternal and fetal blood circulations to
prevent foreign cells and molecules from disturbing the balance within the fetus that’s
undergoing crucial developmental stages. Thus, an ideal protection for fetal brain development
requires both an effective fetal BBB and a functional BPB to serve as the primary line of
defense.

1.2.1 The blood-brain barrier
The BBB contains an inner layer of endothelial cells at the cerebral blood vessels (BV) to
form a tight junction (TJ) and outer layers of pericytes embedded within the basal lamina and, in
the adult but not fetal brain, astrocytic end feet that are attaching (Kadry et al., 2020; Figure 3).
There is emerging evidence from mammal animal studies showing that MIA can result in
abnormal brain development leading to various CNS disorders including schizophrenia, ASD,
epilepsy, Alzheimer disease (AD), Parkinson disease (PD), etc. (Knuesel et al., 2014). Since
inflammation is known to alter BBB structure and function in the adult, we hypothesized that
during MIA, the fetal BBB formation and integrity is impaired, which could contribute to the
known increased risk of developing neuropsychiatric disorders in offspring exposed to
inflammation during pregnancy (Bale et al., 2010; Knuesel et al., 2014; Meyer, 2019). This chain
of events is termed “prenatal programming” whereby the “enduring or later-in-life alterations of
the offspring brain structure and function owing to conditions during gestation”, which further
7
suggests links between maternal perturbations and long-term diseases in offspring (Gumusoglu
& Stevens, 2019).

Figure 3. Representation of blood-brain barrier. Zoomed out depiction of blood vessels (or
capillaries) (left) and zoomed in diagram of cross section structures. (Created with
BioRender.com)

Even though the exact mechanism causing the damages still remains unclear, several
studies have proved that loss of BBB integrity can lead to “loss of highly vulnerable neurons,
production of neuroinflammatory sites, and focal white matter injury” (Knuesel et al., 2014). A
study done by Montagne et al. demonstrated that genetically modified pericyte-deficient mice
undergo BBB breakdown, showing phenotypes similar to those observed in dementia patients
(2018). It is therefore concluded that pericytes, or BBB permeability in general, plays a
significant role in neuropathology, and thus, prenatal damage on BBB formation is likely to
potentially result in unwanted neurodevelopmental insults to the developing brain that may have
long-lasting influence postpartum.
8
Since increased BBB permeability leads to the leakage of signaling molecules from
peripheral immune responses into the brain parenchyma, loss of BBB integrity can further
activate the immune cells in the CNS, including microglia, which are critical for the development
of neuropathologies (Knuesel et al., 2014). Previous studies have shown that MIA leads to long-
lasting activation of microglia and neuroinflammation in the adult offspring, but there have been
little studies focusing on the mechanisms by which this happens and how the fetal brain
development is affected, which is one of the main purposes of this research project.

1.2.2 The blood-placenta barrier
The placenta plays a complicated yet crucial role in fetal development. Like other organs,
human placenta and mice placenta share a similar function but differ in their structures (Figure 4
A, B). Although the complete functions of each placental compartment and cell type is not yet
fully understood, the labyrinth is a crucial layer for exchanging molecules between the fetal and
maternal circulations. Blood vessels and capillaries grow separately from the maternal and fetal
blood circulations to form two circulatory systems that intertwine closely within the labyrinth
layer to allow the exchange of nutrients, waste, and gases (Croy, 2014).
In a well-developed mouse placenta, the blood-placenta barrier (BPB) (i.e., the
fetomaternal vasculature interface within the labyrinth) constitutes of several layers of different
cell types to allow selective permeability, somewhat like the BBB (Figure 4C). During mid to
late gestation, the yolk sac nutrient is no longer sufficient for further growth of the fetus, and
thus it is important for a functional BPB formation to support transferring of nutrients form the
maternal circulation. Consistent with this function, it is known that failure of the BPB to become
permeable enough (i.e., placental insufficiency) leads to underdeveloped fetuses (Woods et al.,
9
2018). Conversely, an overly permeable BPB may also be deleterious to normal fetal
development by allowing fetal exposure to maternal blood-borne unwanted and toxic molecules.
Consistent with this idea, observations in the lab show that an acute and severe MIA leads to a
leaky BPB that disrupts fetal homeostasis and likely over activates the fetal immune system.

10
A

B

C

Figure 4. Complex structures of mature placenta. Cross section diagrams of mouse (A) and
human (B) placentae. (C) Zoomed in representation of fetomaternal vasculature interface within
the mouse labyrinth. EC, endothelial cell; FBS, fetal blood space; MBS, maternal blood space;
Per, pericyte; S-TGC, syncytiotrophoblast giant cell; SynT, syncytiotrophoblast.

Decidua
Junctional Zone
Labyrinth
Chorionic Plate
Umbilical Cord
Trophoblastic Giant Cells
Trophoblastic Glycogen Cells
Fetal Blood Vessels
Decidua
Chorionic Plate
Umbilical Cord
Fetal Blood Vessels
Chorionic Villi
Extravillous Trophoblast
Intervillous Space (Maternal Blood)
MBS
FBS
EC
S-TGC
SynT-I SynT-II
Per
Labyrinth
11
1.3 Use of poly(I:C) in mice and relevant issues
To safely induce MIA in pregnant mice, poly(I:C), a synthetic dsRNA viral mimetic
molecule, was used to mimic an inflammatory response from viruses, such as influenza
(Gumusoglu & Stevens, 2019). However, the types and dosages of poly(I:C) generate variable
responses. Factors including the strain of the mice used and the how different batches of
poly(I:C) are manufactured directly influence the severity of the MIA induced in the animals
(Chen, 2021). It is especially tricky to reach an ideal strength of inflammatory response in our
case. This is because we need to model a severe inflammatory event, but the mother receiving
poly(I:C) needs to survive, and we also require a few surviving fetuses to provide meaningful
experimental data.
It has been well studied that the inflammatory effects in both the mothers and the fetuses
are generally attributable to MIA itself rather than specific pathogens, with some exceptions
(e.g., viruses such as Zika that can infect the fetus through the placenta) (Knuesel et al., 2014).
Thus, literature data showing that poly(I:C) injection during pregnancy mimics common viral
infections by inducing increase in inflammatory markers validates the use of the viral-mimetic
molecule in this project. In addition, previous lab work measured significant increases in
maternal blood tumor necrosis factor alpha (TNF-α), monocyte chemoattractant protein 1 (MCP-
1), and interleukin-6 (IL-6) to further demonstrate the effectiveness of poly(I:C) in our MIA
mouse model (Chen, 2021).
Interestingly, a review article by Boulanger-Bertolus et al. compared research data across
several studies on the effect of IL-6 increase induced by MIA, concluding that IL-6 specifically
is highly likely to be responsible for mediating the neurobehavioral outcomes in MIA offspring
(2018). In recent years, more human studies started to investigate this topic to fill in the gap with
12
follow-ups using longitudinal study designs. Functional magnetic resonance imaging (fMRI) was
used for observation and measurements of the developing human brains from in utero to toddler
or adolescent age. The increased levels of inflammatory markers in maternal blood including IL-
6 during gestational stage was shown to have a correlation with abnormal network development
in the offspring (Boulanger-Bertolus et al., 2018).
Previous experiments in the lab were done using the CD-1 mouse strain, mostly because
they are outbred and very fertile. However, most mice used for genetic studies are on a different
background, usually a substrain of C57BL/6. In fact, the mice used in this project to
conditionally knockout Ptgs2 gene expression in microglia were on the C57BL/6J background,
which is more amenable to genetic manipulations than CD-1 background. They also present the
advantage in genetic studies to be inbred and therefore genetically identical to one another
(except modified genes of interest). Yet, various studies have shown that different strains of mice
respond differently to immune activators (Gumusoglu & Stevens, 2019). Therefore, it was
necessary to test out an appropriate dosage of poly(I:C) for MIA in the transgenic mice strain,
C57BL/6J, to ensure the effectiveness of the MIA while still allowing enough fetuses to survive
the acute inflammation. Previous studies in the lab have shown that the easiest phenotype to
characterize and tell apart MIA and control groups is that placentae from MIA groups are
bloodier, sometimes even swollen, which is likely a result of decreased or impaired maternal
blood flow due to inflammation in the pregnant animal. Therefore, the size increase of placental
maternal blood spaces (MBS; Figure 5) can be used for determining the efficacy of MIA
induction, a phenotypic marker we used in C57BL/6J mice in this thesis project.

13

Figure 5. Representative image of MBS in mouse placenta at GD 15. Two MBS surface areas
in the section are circled in blue. They contain non-nucleated, maternal red blood cells (arrows).
Fetal villi containing nucleated and non-nucleated fetal red blood cells surround MBS (Scale bar,
50 µm.)

MBS
Fetal villi
14
CHAPTER 2: Materials and Methods
2.1 Animals
Timed pregnant mice from CD-1 strain were ordered from Charles River Laboratory and
delivered at gestational day (GD) 12. C57BL/6J was the background strain for WT (LysM
+/+
,
COX2
+/+
), COX2-MKO (LysM
Cre/+
, COX2
fl/fl
), and COX2-FLOX (LysM
+/+
, COX2
fl/fl
) mice. WT
mice were purchased from The Jackson Laboratory. The original transgenic mating pair were
gifted from Dr. David Meriwether and Dr. Srinivasa Reddy (UCLA, Los Angeles, CA)
(Meriwether et al., 2019). All mice stayed in ventilated cages in a pathogen-free condition in an
animal facility under a 12 h-12 h dark-light cycle with temperature- and humidity- control and
access to the same water and food throughout the entire study. Animals were housed in groups
with a maximum of five same-sex adult mice per cage (except for males that were previously
used in mating pairs in order to avoid aggressive behaviors). All procedures were performed
according to the NIH Guide for the Care and Use of Laboratory Animals and approved by the
Institutional Animal Care and Use Committee at University of Southern California (Zhao et al.,
2022).
Table 1. Number of litters and fetuses included for each sub-project.
Experiment Group Number of dams per experiment group
CD-1 prenatal IHC 33 (1 to 2 fetuses from each dam)
C57BL/6J WT poly(I:C) dosage testing 10 (1 to 3 fetuses from each dam)
Transgenic prenatal IHC 3 (1 to 2 fetuses from each dam)

2.1.1 Timed pregnancy
Mice between the ages of 9 weeks and 8 months were used for timed pregnancy mating.
Mating trios (for WT mice) or pairs (for transgenic mice, see Figure 6 for the scheme used) were
put into one cage between 4 PM to 6 PM and separated between 8 AM to 10 AM. Time of
15
separation was considered as GD 0.5, and the weight of the female mice were recorded
regardless of whether copulatory plug is observed. Repeated weight measuring of the mated
females were done every three to five days until around GD 10. Those with a weight increase of
over 4 g were considered pregnant and were assigned to experimental groups (Heyne et al.,
2015). Female mice that gained less than 4 g over two weeks were left for observation for
another week to ensure that they were not pregnant with a small litter before they were mated
again.

Figure 6: Transgenic mice mating scheme. Animals were put in mating pairs according to their
genotypes. We mated one COX2-FLOX (LysM
+/+
, COX2
fl/fl
) female with one COX2-MKO
(LysM
cre/+
, COX2
fl/fl
) male as it is the most efficient way to obtain the highest percentage of
COX2-MKO offspring (50% in theory). (Created with BioRender.com)

2.1.2 Gestational MIA
At GD 13, pregnant mice were weighed and intraperitoneally (i.p.) injected once with one
dose of poly(I:C) or 0.9% sterilized saline solution according to the weight. All animals were
returned to their home cages immediately after injection.

16
2.1.3 Genotyping
To extract DNA, tissue samples were heated in solution 1 containing 0.2 mM
ethylenediaminetetraacetic acid (EDTA) and 26 mM NaOH for a total of 2 h at 95 °C. Every 30
mint during heating, tubes were vortexed or flicked to ensure thorough digestion and were placed
on ice for 10 min to cool down after heating. To each sample tube was added solution 2
containing 40 mM tris(hydroxy-methyl)aminomethane [(Tris)×HCl] (volume equal to solution 1)
to neutralize the DNA solutions.
LysM-cre PCR and sex-determining region of the Y chromosome (SRY) PCR were run
with Taq Plus 2X Master Mix Red (D124R; Lamda Biotech), and COX2-FLOX PCR was run
using Kapa 2G HotStart Taq 2X Ready Mix with dye (KK5609; Roche Sequencing and Life
Science, Kapa Biosystems). In COX2-FLOX PCR reaction mixture, additional MgCl2 stock
solution and glycerol were added to obtain a final concentration of 25 mM MgCl2 and 6.5%
glycerol. Listed in Table 2 are the expected bands to get from agarose gel (2%) electrophoresis,
specific PCR temperature and cycling conditions, and primer information.

17
Table 2. PCR conditions, primer sequences, and expected band weights. All PCR programs
started with 2-5 min of preheating at 95 °C and used 105 °C as lid temperature. FP, forward
primer; RP, reverse primer; bp, base pairs.
LysM
cre

WT: 513 bp
Cre: 250 bp
FP
5'- CTT GGG CTG CCA
GAA TTT CTC -3'
94 °C, 20 sec; 60 °C, 20 sec;
70°C, 30 sec; repeat 39 cycles.
RP-WT
5'- CCT CAC CCC AGC
ATC TCT AAT TC -3'
RP-Cre
5'- ATC ACT CGT TGC
ATC GAC CGG TAA -3'
COX2
fl

WT: 253 bp
Flox: 192 bp
FP
5’- TGC CCT TGT TGT
TGT TGT TG -3’
94 °C, 30 sec; 65 °C 1 min;
68 °C, 30 sec; repeat 10 cycles,
-0.5 °C per cycle.
94 °C, 30 sec; 60 °C, 30 sec;
72 °C, 2 min; repeat 28 cycles.
RP-WT
5’- GTT GGG CAG TCA
TCT GCT AC -3’
RP-Flox
5’- TGG ACG TAA ACT
CCT CTT CAG AC -3’
SRY
Male: 250 bp
Female: No band
FP
5’- AGA GAT CAG CAA
GCA GCT GG -3’
94 °C, 1 min; 60 °C 3 min;
70 °C, 3 min; repeat 33 cycles.
RP
5’- TCT TGC CTG TAT
GTG ATG GC -3’

18
2.2 Injections
2.2.1 Poly(I:C)
For poly(I:C) dosage testing experiments, a mixture of two products (tlrl-picw_#PIW 41-
03, tlrl-pic_#PIC 40-07; InvivoGen) were used in all dosages given throughout the study to
ensure a consistent ratio of high and low molecular weight (HMW, LMW) dsRNA fragments.
Stock solutions were prepared according to the instructions given by the manufacturer
(InvivoGen) and diluted with sterile saline solutions accordingly to reach the final
concentrations. Shown in Table 3 is the constitution of the three dosages of poly(I:C) mixtures
(L/HMW) used in the study.
Table 3. Constitutions of LMW and HMW poly(I:C) solutions for L/HMW mixture
dosages tested. Mixtures were prepared according to the estimated total weight of mice to be
used no earlier than a month before the injection.
6.4 mg/kg 9.6 mg/kg 12.8 mg/kg
LMW Poly(I:C)
(Stock: 10 mg/mL)
Dilute to 1 mg/mL,
use 5.2 mg/kg.
Dilute to 1 mg/mL,
use 5.2 mg/kg.
Dilute to 5 mg/mL,
use 10.4 mg/kg.
HMW Poly(I:C)
(Stock: 0.5 mg/mL)
Use 1.2 mg/kg. Use 1.2 mg/kg. Use 2.4 mg mg/kg.
Injected volume of mixture 7.6 mL/kg 11.4 mL/kg 6.88 mL/kg

2.2.2 Celecoxib
Celecoxib (PZ0008; Sigma) was dissolved in saline solution with 20% DMSO and
injected i.p. once to the dams at around 24 h post initial injection of saline or poly(I:C). Control
groups were only given saline and DMSO vehicle injection i.p. (Zhao et al., 2022).
19
2.3 Histology and staining
2.3.1 Dissection and tissue harvest
At GD 15, dams were anesthetized, and fetal tails, fetal brains, and placentae were
collected. Fetal tails were stored at -20 °C until used for DNA extraction. Fetal brains and
placenta were fixed with 4% paraformaldehyde (PFA) dissolved in phosphate-buffered saline
(PBS). Samples were left overnight at 4 °C, and a shaker with gentle agitation was used to obtain
thorough fixation.

2.3.2 Tissue processing
After PFA fixation, a subset of fetal brains and placentae were dehydrated in 15%
sucrose solution (prepared using PBS) for 48 h and 30% sucrose solution for another 24 h.
Samples were kept at 4 °C with gentle agitation on a shaker throughout the dehydration
processes. Afterwards, the tissues were then embedded in cryomolds with optimal cutting
temperature compound (OCT) and stored frozen at -80 °C until ready for the sectioning
procedure.
Other tissues were prepared for formalin-fixed paraffin-embedding (FFPE) approach.
Fetal brains and placentae were transferred to 70% ethanol and stored at 4 °C at least overnight.
Labeled cassettes were used to for the dehydration series of 3 times of 10 min incubation in 80%
ethanol, 3 times of 20 min incubation in 90% ethanol, 4 times of 20 min incubation in 100%
ethanol, and 2 times of 20 min incubation in Neo-Clear xylene substitute. Overnight incubation
in liquid paraffin in an oven maintaining 65 °C to ensure infiltration of paraffin before
embedding. Finished paraffin blocks are stored at room temperature (RT).

20
2.3.3 Sectioning
Frozen tissues in OCT were transferred to -20 °C for at least 24 h and were then
sectioned with a cryostat (CM3050S; Leica) at around -21 °C. Fetal brains and placentae were
sliced into 20-μm-thick sections and collected on superfrost-plus slides. All samples were kept
frozen throughout the procedure, and slides were stored at -80 °C.
FFPE samples were sectioned with a microtome (RM2125 RTS; Leica) into 5-μm thick
and were mounted on superfrost-plus slides, which were left on a rack to fully dry overnight and
then stored at RT until further analyses.

2.3.4 Hematoxylin and eosin staining (H&E)
Frozen slides were allowed to thaw and dry for 15 min inside a chemical fume hood at
RT. They were then placed in the following series of staining solutions: hematoxylin stain for 30
sec (followed by gentle rinsing with tap water), differentiation ready-to-use solution (RTU; citric
acid aqueous solution) for 30 sec, Scott’s tap water substitute (MgSO4 solution buffered with
NaHCO3) for 10 sec (followed by gentle rinsing with tap water), 70% ethanol for 30 sec, eosin
stain for 2-3 sec. Slides were then washed with 70% ethanol briefly before moving to a
dehydration sequence for duplicate washes of 1 min in 70%, 95%, 100% ethanol, and limonene
(total of 8 min).
Slides with FFPE sections were first placed in the oven at 65 °C for 25 min to melt
paraffin before treated with the following (1) rehydration sequence of fresh xylene for 10 min
twice, 100% ethanol for 5 min twice, 95% ethanol for 5 min once, 70% ethanol for 5 min once,
and distilled water for 5 min once, (2) H&E staining sequence of 3 min in hematoxylin (followed
by gentle rinsing with distilled water), 1 min in differentiation RTU, 1 min in Scott’s tap water
21
substitute (followed by gentle rinsing with distilled water), 1 min in 70% ethanol, and 1 min in
eosin stain, and (3) dehydration sequence of 70% ethanol for 1 min once, 95% ethanol for 1 min
once, 100% ethanol for 1 min twice, and fresh xylene for 2 min twice.
Both approaches were followed by gentle removal of excess liquids from the slides, and
dibutylphthalate polystyrene xylene (DPX) mounting medium was used with the coverslips.
Slides were left undisturbed to be fully dried for at least 48 h before imaging.

2.3.5 Immunohistochemistry staining (IHC)
Frozen slides were thawed and dried for 15 min inside a chemical fume hood at RT
before washed and permeabilized for 15 min with agitation in PBS-Triton (PBS-T) solution
consisting of PBS with 0.1% Triton X-100. Afterwards, slides underwent a 2-h incubation at RT
inside a humid chamber with blocking solution consisting of 2% fetal bovine serum diluted in
PBS-T. The same setup and conditions were also used for overnight incubation of primary
antibodies and 2 h incubation of secondary antibodies. An additional step of amplification
incubation for 1.5 h was added to COX2 staining. Slides underwent 5-min-long washes in PBS-T
four times after both secondary antibody incubation and amplification incubation steps, followed
by 5 min of DAPI incubation and additional three PBS-T washes. Light exposure was limited
when fluorescence molecules were involved at any time point during the procedures. Finally,
sections were glued with coverslips using Prolong-Gold (Vector) and underwent imaging after at
least 24 h of drying in dark.

22
Table 4. Primary and secondary antibodies used. Ab, antibodies.
Primary Ab
Anti-PeCAM Hamster 1:1000 Millipore MAB1398Z
Anti-SMA Mouse 1:100 Dako M0851
Anti-PDGFRb Goat 1:100 R&D Systems, AF-1042
Anti-TER119 Rat 1:1000 R&D Systems, MAB1125
Anti-Iba1 Rabbit 1:500 WAKO, 019-19741
Anti-Iba1 Goat 1:200 Novus Biological, NB100-1028
Anti-COX2 Rabbit 1:200 Abcam, Ab15191
Anti-COX2 Rabbit 1:150 Cayman, 160126
Anti-Cll5 (Alexa 488 conjugated) Mouse 1:100 ThermoFisher, 352588
Anti-ZO1 Rabbit 1:100 ThermoFisher, 40-2200
Secondary Ab
Anti-hamster, Alexa 488 Goat 1:800 Jackson ImmunoResearch
Anti-rat, Alexa 488 Donkey 1:800 Jackson ImmunoResearch
Anti-mouse, horseradish
peroxidase (HRP)
Rabbit 1:800 Jackson ImmunoResearch
Anti-goat, Alexa 488 Donkey 1:800 Jackson ImmunoResearch
Anti-goat, Rhodamine red Donkey 1:800 Jackson ImmunoResearch
Anti-rabbit, Rhodamine red Donkey 1:800 Jackson ImmunoResearch
DyLight 549 Streptavidin - 1:500 Vector Labs SA-5549

23
2.4 Image quantification and analysis
2.4.1 Microscopy and quantification
Images of H&E staining of the placentae were acquired with a Zeiss Axioimager II
microscope. In each placenta, three regions of interest from the left, middle, and right portions of
the labyrinth were imaged in three adjacent sections (20-100 μm apart). The surface areas of the
largest three maternal blood spaces in each image were measured using ImageJ and averaged per
placenta.
IHC staining of fetal brains were imaged with a Zeiss Axioimager II microscope and a
Zeiss LSM800 confocal microscope. Images of the parietal cortex regions in both left and right
hemispheres of the same brain section were acquired, and three adjacent sections (80-100 μm
apart) were included per brain. The numbers of cells considered positive for the markers used
were determined by manually counting them, normalized to the surface area of each section
imaged, and averaged per brain. Quantification was done using either ImageJ or the AxioVision
software (version 4.8.2; Zeiss).

2.4.2 Statistics
Statistical analyses were all performed using GraphPad Prism software (version 8.0).
Experimental results achieved were analyzed using the unpaired two-tailed t test and/or one-way
analysis of variance (ANOVA) followed by Tukey’s multiple comparison analysis. All data are
presented as mean ± standard error of the mean (SEM), and statistical significance was set at p <
0.05 (*).

24
CHAPTER 3: Presentation of Research Results
3.1 Gestational MIA acutely disrupts fetal brain development
Group work in the lab showed that pericyte coverage (platelet-derived growth factor
receptor-β, PDGFRβ; smooth muscle actin, SMA) of BV endothelial cells (platelet endothelial
cell adhesion molecule, PeCAM) is significantly decreased in the brain regions including
thalamus, striatum, and cortex of the GD 15 fetal brains in MIA groups compared to saline
controls (Figure 7). Consistent with studies of BBB in the adult brain, such decreased coverage
led to increased BBB permeability in the fetal brain, which was confirmed using live fMRI tests
(Zhao et al., 2022).
To understand the molecular mechanism causing such disruption in fetal BBB
development, we measured the concentrations of various immune-activation-related molecules in
fetal brains from GD 13 to GD 15. We found that cytokines concentrations in the fetal brain are
not affected by maternal inflammation, but we also observed a significant increase in the
inflammatory mediator prostaglandin E2 (PGE2) in the GD 14 and 15 MIA cortex. Since PGE2
is one of the major metabolic products of arachidonic acid by the COX2 enzyme, this finding led
us to the hypothesis that COX2 activation by MIA is contributing to abnormal BBB development
(Zhao et al., 2022).

25
A PeCAM SMA DAPI B PeCAM PDGFRb DAPI
Saline

Saline

Poly(I:C)

Poly(I:C)

C D

Figure 7. MIA induced decreased pericyte coverage on BV endothelial cells in CD-1 fetuses
at GD 15. (A, B) Representative IHC staining images of GD 15 fetal brain. PeCAM, green: BV
marker. SMA (A) and PDGFRb (B), red: pericyte markers. DAPI, blue: nuclear marker. (Scale
bar, 20 µm.) Quantified colocalization in fetal cortex of SMA and PeCAM (C), PDGFRb and
PeCAM (D). n=6 (C) or 3 (D) dams from each group, 1 fetus from each dam. Unpaired two-
tailed t test: **** P < 0.0001 (C), *** P < 0.001 (D). Data presented as mean ± SEM.
Saline Poly(I:C)
30
40
50
60
70
80
% SMA/PeCAM
✱✱✱✱
Saline Poly(I:C)
30
40
50
60
70
80
% PDGFRβ/PeCAM
✱✱✱
Saline Poly(I:C)
30
40
50
60
70
80
% PDGFRβ/PeCAM
✱✱✱
26
Due to lack of studies on COX2 activation in fetal neurodevelopment, we next explored
COX2 expression in various cell types in fetal brains that can be induced by gestational MIA.
We found that microglia with ionized calcium binding adaptor molecule 1 (Iba1) expression play
an important role in COX2 activation. Not only did we measure increased number of Iba1+
microglia in GD 15 MIA fetal brains (Figure 8B), but there was also a significantly higher
percentage of Iba+ microglia expressing COX2 (Figure 8C), suggesting that gestational MIA
induces the activation of COX2 expression in Iba+ microglial cells (Zhao et al., 2022).
In order to test the hypothesis that increased number of microglia expressing COX2 is
driving fetal BBB disruption, we next planned to examine the outcomes of gestational MIA after
pharmacological inhibition on COX2 activation or genetic elimination of COX2 expression in
microglia.

27
A PeCAM Iba1 COX2 DAPI

B C

Figure 8. MIA induced hyper activation of Iba1+ microglia with or without COX2
expression in CD-1 fetuses at GD 15. (A) Representative IHC staining image of GD 15 fetal
brain. PeCAM, green: BV marker. Iba1, white. COX2, red. DAPI, blue: nuclear marker. (Scale
bar, 20 µm.) Quantification of Iba1+ (B) and fraction of Iba1+COX2+ microglia (C) in fetal
cortex. n=6 (B) or 3 (C) dams from each group, 1 fetus from each dam. Unpaired two-tailed t
test: *** P < 0.001 (C), ** P< 0.01 (B). Data presented as mean ± SEM.

Saline Poly(I:C)
0
50
100
150
Iba1+ cells (per mm
2
)
✱✱
0
20
40
60
80
100
120
Saline Poly(I:C)
Iba1+COX2+/Iba1+(%)
✱✱✱
28
3.2 Poly(I:C) L/HMW dosage determination in C57BL/6J mouse strain
Literature studies have concluded that poly(I:C) from different batches with various
molecular weights can lead to different results in terms of maternal inflammation and its effect
on embryos. Therefore, experiments were run in our lab prior to this project to come up with a
recipe using only poly(I:C) stock products of known molecular weights (LMW and HMW) and
result in a mixture (L/HMW) that can produce consistent and stable MIA in CD-1 mice.
As the research transitioned to the use of transgenic mice with C57BL/6J strain as their
genetic background, we started testing MIA response from poly(I:C) in C57BL/6J WT mice to
evaluate their tolerance to the poly(I:C) mixture that produced reproducible MIA outcomes in
CD-1 strain. WT C57BL/6J mice were mated following the timed pregnancy method inside our
facility the same way we would do to transgenic mice. Following published studies using higher
poly(I:C) dosage in C57BL/6J mice than what was used in CD-1 mice for i.p. injection
(Gumusoglu & Stevens, 2019), we started by administering a double dosage of 12.8 mg/kg
poly(I:C) mixture at GD 13. To avoid the confounding effect of having too much fluid injected
into the peritoneum of the pregnant mice, we used a different dilution from the stock to have the
mixture solution approximately twice as concentrated instead of injecting twice the volume of
the original formula (Table 3). However, the dosage appeared to be too strong for our
experimental purpose as in two of the three dams, all embryos were found dead at GD 15. Even
though all embryos of the third dam were found to be still alive by the time of dissection, the
fetuses all had an abnormal appearance with subcutaneous bleeding evident throughout the body
(Figure 9). Since mating was overnight, it is possible that the fetuses in this dam were a few
hours older than those from the other two dams, explaining their survival up to GD 15, but the
severe subcutaneous bleeding suggested that they would likely mostly be aborted before birth.
29
Table 5. Summary of number of embryos survived MIA for three poly(I:C) dosages. Only
two dams were included in groups tested with 6.4 and 9.6 mg/kg. Total embryos counting does
not include the ones whose appearance clearly indicated an abortion before MIA.
L/HMW Poly(I:C)
Dosage (mg/kg)
Number of dead embryos/total embryos
Dam 1 Dam 2 Dam 3
6.4 1/7 0/7 -
9.6 1/2 10/10 -
12.8 4/4 9/9 0/7

Figure 9. Representative embryo from dam tested with 12.8 mg/kg L/HMW poly(I:C)
mixture. Image captured from dissecting microscope at GD 15. Note hemorrhaging spots
distributed throughout the fetal body.

Given these results, we decreased the dosage and tested out the L/HMW poly(I:C)
mixture using what was administered originally in CD-1 mice (i.e., 6.4 mg/kg) and 2/3 higher,
midway to the previously tested dosage (i.e., 9.6 mg/kg). The higher dosage turned out still too
strong as by time of dissection only one fetus out of all twelve embryos survived, and we were
only able to collect a total of three embryos from two dams. On the other hand, the lower dosage,
which was the same one used for CD-1 mice in previous studies, resulted in only one embryo
death (Table 5). Furthermore, staining of the placental sections obtained from the three dosage
groups showed increasing levels of red blood cells number in the labyrinth, which correlated
30
with increased volume of maternal blood present in the placentae; these are phenotypic markers
of severe MIA that we aimed to obtain (Figure 10).
A TER119 PeCAM DAPI
Saline

Poly(I:C) 6.4 mg/kg

31
Poly(I:C) 9.6 mg/kg

Poly(I:C) 12.8 mg/kg

32
B
Saline

Poly(I:C) 6.4 mg/kg

33
Poly(I:C) 9.6 mg/kg

Poly(I:C) 12.8 mg/kg

Figure 10. Comparison of placentae histology from three dosages of L/HMW poly(I:C)
mixtures. Representative images of IHC stainings (A) and H&E stainings (B) of placenta
labyrinth regions from saline, poly(I:C) 6.4, 9.6, and 12.8 mg/kg dams at GD 15. (A) TER119,
red: red blood cell marker. PeCAM, green: BV marker. DAPI, blue: nuclear marker. (Scale bar,
100µm.)
34
Since our studies require multiple fetuses per dam, we would not be able to move forward
with the higher dosages. Thus, to assess the severity of MIA induced by the lower dosage of
L/HMW poly(I:C), we randomly picked three placentae from each of the two dams that received
6.4 mg/kg dosage and compared the surface areas of maternal blood space (MBS) in cross
sections with those of the vehicle control dams. Previously, only a trend of increased MBS was
observed, and the difference was not statistically significant (Chen, 2021). However, we found
that the 6.4 mg/kg dosage of L/HMW mixture injected i.p. at GD 13 led to a significantly
enlarged MBS in the placentae (Figure 11). Therefore, we concluded that this dosage was
appropriate and feasible for our experimental purpose to study the effect of MIA on fetal brain
development.

Figure 11. Quantification of MIA effect on MBS surface area. WT mice on C57BL/6J
background were used. MBS size was quantified at GD 15 after treatment at GD 13. n=2 dams
from each group, 2 placentae from each dam. Unpaired two-tailed t test: ** P < 0.01. Data
presented as mean ± SEM.

Before testing the effects of MIA in transgenic mice, it was important to also verify that
the 6.4 mg/kg L/HMW mixture of poly(I:C) disrupts neurodevelopment in WT C57BL/6J mice,
as it did in CD-1 mice (Zhao et al., 2022). Using IHC, we found the number of Iba1/COX2
Saline Poly(I:C)
40000
50000
60000
70000
80000
Maternal Blood Space (µm
2
)
✱✱
35
double positive microglia in GD 15 fetuses from MIA dams significantly increased compared to
those from the saline group (Figure 12), confirming that this dosage is effective and MIA
outcomes are consistent in fetal brains across the two mouse strains.

Figure 12. Quantification of Iba1+ COX2+ cells in GD 15 fetal brains. WT mice with
C57BL/6J background were used. n=2 dams from each group, 3 placentae from each dam.
Number of COX2+ microglia was quantified at GD 15, 48 h after treatment. Unpaired two-tailed
t test: * P < 0.05. Data presented as mean ± SEM.

Saline Poly(I:C)
0
5
10
15
20
Iba1+ COX2+ cells (per mm
2
)
✱
36
3.3 Prevention of gestational MIA disruption in fetal brain
To further understand the involvement of COX2 in gestational MIA outcomes in the
fetuses, we administered celecoxib (CX) i.p. 24 h after the initial injection at GD 13.
Pharmacologically inhibiting COX2 activity selectively (i.e., COX1 was not affected) enabled us
to test whether the effect of MIA on fetal neurodevelopment depends on COX2 pathways
activation. Using the same IHC methodology described above, we observed that within 24 h of
poly(I:C) administration, CX treatment prevented the MIA-induced increase in total Iba1+
microglia number (Figure 13A), the increase in fraction expressing COX2 (Figure 13B), and the
pathology of increased BBB permeability in the GD 15 fetal brain (Zhao et al., 2022; Figure
15A).
This finding confirmed the important role COX2+ microglia play in causing the
neuropathology observed in fetuses from the MIA dams. To further test if MIA effect require
COX2 activation in microglia, our next step consisted of testing gestational MIA outcomes in
conditional knockout mice whose COX2 gene (Ptgs2) is selectively deleted in myeloid cells
prenatally, which includes microglia (COX2-MKO). We performed timed pregnancy mating as
demonstrated in the scheme from chapter 2.1.1 to obtain both embryos with normal expression of
COX2 (COX2-FLOX) and MKO fetuses within one dam (Figure 6). We confirmed absence of
COX2 expression in cortex microglia of MKO fetal brains (Figure 14A) and compared Iba1+
microglial cell numbers in response to MIA between the two genotypes. The quantitative data
showed that at GD 15, 48 h after inducing MIA in the moms, the COX2-MKO embryos had a
normal Iba1+ microglia counting in the cortex of the fetal brains, similar to what was seen in
saline dams in previous experiments, whereas the number was still elevated in COX2-FLOX
littermates (Figure 14B).
37
A

B

Figure 13. CX prevents abnormal Iba1 and/or COX2 expression in microglia in fetal brains
of CD-1 mice at GD 15. Quantification of Iba1+ microglia (A) and the fraction of Iba1+
microglia expressing COX2 (B). Quantifications were done in the fetal cortex at GD15, 48h after
treatment. n=3 dams from each group, 1 fetus from each dam. Data presented as mean ± SEM.
One-way ANOVA with Tukey’s multiple comparison test: *** P < 0.001 (A, B), ** P < 0.01
(A), ns, not significant (A, B).
0
40
80
120
160
Iba1+ cells (% of Vehicle)
Saline: + - + -
Poly(I:C): - + - +
DMSO: + + - -
CX: - - + +
✱✱ ✱✱✱
ns
0
20
40
60
80
100
120
Iba1+COX2+/Iba1+(%)
✱✱✱
ns
ns
Saline: + - + -
Poly(I:C): - + - +
DMSO: + + - -
CX: - - + +
38
A PeCAM Iba1 COX2 DAPI B
FLOX

MKO

Figure 14. Conditional COX2 knockout lowers activation of Iba1+ microglia in fetal cortex
at GD 15. (A) Representative IHC staining images of GD 15 fetal brain. Blood vessel marker,
PeCAM, green. Iba1, white. COX2, red. Nuclear marker DAPI, blue. Double positive
COX2+/Iba1+ microglia are observed in FLOX but not MKO tissue (arrow). (Scale bar, 20 µm.)
(B) Quantified statistics of Iba1+ microglia counting in cortex. n=3 dams from each group, 2
fetuses from each dam. Unpaired two-tailed t test: ** P < 0.01. Data presented as mean ± SEM.

We also demonstrated that injecting CX prenatally or conditionally knocking out Ptgs2
prevented the increased permeability in fetal BBB induced by MIA. Quantification of
SMA/PeCAM colocalization showed that pericyte coverage returned to normal level in
poly(I:C)+CX groups (Figure 15A), and PDGFRβ/PeCAM staining in COX2-MKO fetuses rose
to 1.6-fold that of COX2-FLOX fetuses (Figure 15B). These results, combined with the
FLOX MKO
0
10
60
80
100
120
Iba1+ cells (per mm
2
)
✱✱
39
microglia counting data, strongly suggested that the suppressing COX2+ microglia activation
directly prevents MIA-induced BBB structural disruption and leakage.
A

B

Figure 15. Quantification of pericyte coverage of BV endothelial cells. Percentage of
SMA/PeCAM colocalization in CD-1 mice fetal cortex and PDGFRb/PeCAM colocalization in
C57BL/6J mice fetal brain at GD 15. n=3 dams from each group, 1 (A) or 2 (B) fetuses from
each dam. (A) One-way ANOVA with Tukey’s multiple comparison test: ** P < 0.01; ns, not
significant. (B) Unpaired two-tailed t test: *** P < 0.001. Data presented as mean ± SEM.
0
20
40
60
80
% SMA/PeCAM
Saline: + - + -
Poly(I:C): - + - +
DMSO: + + - -
CX: - - + +
✱✱
ns
✱✱
FLOX MKO
30
40
50
60
70
80
% PDGFRβ/PeCAM
✱✱✱
40
CHAPTER 4: Discussion and Conclusion
4.1 Summary of experimental data
We discovered that maternal inflammation leads to acute disruption of fetal BBB
development with increased permeability caused by decreased pericyte coverage of BV
endothelial cells. Measurements of various inflammatory molecules showed a significant
increase in PGE2 concentration in fetal brains from the MIA groups, leading us to hypothesize
that COX2 pathway activation is an important driver of MIA outcomes in fetal brains.
Focusing on the COX2 pathway to understand how the fetal immune system responds to
gestational MIA, we used CX to selectively inhibit COX2 after MIA induction. Our data showed
that CX treatment 24 h after poly(I:C) injection prevents MIA pathological effects. These
observations were made in outbred CD-1 mice. In order to genetically test COX2 pathway
involvement in MIA effects using a conditional KO mouse model, we first experimented with
MIA effects in C57BL6/J mouse strain. By testing out multiple dosages of L/HMW poly(I:C)
mixture in WT C57BL/6J mice, we determined the most suitable dosage of 6.4 mg/kg for our
experimental purposes. We used this L/HMW poly(I:C) mixture to introduce MIA in COX2-
MKO fetuses, which lack COX2 expression in myeloid cells including microglia, and their
COX2-FLOX littermates. We found that knocking out COX2 expression in Iba1+ microglial
cells prevented the fetal BBB from becoming hyper permeabilized by MIA. Evidence of
successful prevention included absence of COX2 expression in fetal microglia and normal
number of Iba1+ microglia in COX2-MKO fetuses exposed to MIA compared to their COX2-
FLOX littermates.
To summarize, we developed a gestational MIA mouse model that can be consistently
replicated using a set ratio of HMW and LMW poly(I:C) mixture at the dosage of 6.4 mg/kg, and
41
we demonstrated pharmacologically and genetically that activating the COX2 pathway in Iba1+
microglial cells drives the gestational MIA effects in fetal brain, involving disruption of early
BBB formation. Overall, the data we obtained supports our hypothesis in spite of several
limitations and imperfections in the experimental design due to time and efficiency concerns,
especially considering minimizing the use of animals.

4.2 Limitations
Despite the data shown in earlier chapters that proved the hypothesis of increased BBB
permeability in MIA fetuses, there are still a few shortcomings in the research design and
inconsistent data that are worth a discussion.
First, besides pericyte coverage, we also studied the difference in fetal brain tight
junctions (TJ) between saline and poly(I:C) groups in pursuit of further understanding the
phenotypes of increased BBB permeability. However, staining of claudin-5 (Cll5) and zonula
occludens-1 (ZO1) did not show significant changes in their fluorescence intensity (Figure 16).
Since the two proteins are responsible for TJ integration and BBB permeability regulation (Jia et
al., 2014; Tornavaca et al., 2015), we were expecting to see a decrease in the fetal brains of MIA
groups, which would account for the increased permeability. Yet, it is also noteworthy to
mention that the TJ is an extremely complicated feature in the brains with a lot more proteins
involved (e.g., occludins and junction adhesion molecules), so it is possible that the proteins
responsible for the BBB abnormality we observed were not tested in our experiments (Kadry et
al., 2020). Another possible factor is the developmental stage of the fetal brains we experimented
on at GD 15. Since TJ in mice start to be expressed around GD 10-11 (Saili et al., 2017), the
density of baseline TJ proteins in saline groups may still be low at GD 15, making it difficult to
42
measure differences induced by gestational MIA, especially since fluorescence intensity
measurements from IHC staining is not highly sensitive. In addition, since most MIA effects
were observed in the cortex, that was the only brain region we measured in the fetal brains using
this approach. Therefore, we could be missing the brain regions that show more significant
impacts of TJ disruption contributing to the general increased BBB permeability we observed
throughout the fetal brain. A more accurate and quantitative measure of TJ expression disruption
by MIA could be obtained using western blotting experiments on micro-dissected fetal brain
regions.
A B

Figure 16. Quantification of TJ markers in fetal BBB. Cll5 (A) and ZO1 (B)
immunofluorescence intensity measured in fetal cortex at GD 15. n=3 dams from each group,
one fetal brain from each dam. Unpaired two-tailed t test: ns, not significant. Data presented as
mean ± SEM.

Besides, when it comes to the transgenic mice study, an ideal experiment setup would be
to include multiple male and female embryos from the same dam for various genotypes
(LysM
Cre/+
, COX2
fl/fl
; LysM
Cre/+
, COX2
fl/+
; LysM
Cre/+
, COX2
+/+
; LysM
+/+
, COX2
fl/fl
; LysM
+/+
,
COX2
fl/+
; LysM
+/+
, COX2-+/+) to have a complete set of control groups. However, due to the
Saline Poly(I:C)
0
2
4
6
8
10
Cll5 intensity (AU)
ns
Saline Poly(I:C)
0
10
20
30
ZO1 intensity (AU)
ns
43
nature of C57BL/6J mouse strain having small dams (i.e., typical litter size is around six pups)
(MGI - Inbred Strains: C57BL, n.d.), the approach was neither feasible nor efficient. In addition,
although previous studies in our lab using CD-1 mice did not show sex-related differences in
MIA outcomes between male and female offspring, we still decided to only use male embryos
for all the quantifications to avoid possible inconsistencies in sex differences between strains.
Therefore, the two limitations combined left us with a small number of embryos available for
experiments and analyses. Nevertheless, a significant blunting of MIA effects in COX2-MKO
embryos compared to COX2-FLOX littermates was measured despite the small sample size,
supporting the hypothesis that COX2 expression in microglia is driving MIA effects in fetal
brain.
Finally, it is still questionable whether COX2 activity in the pregnant moms play a role in
how gestational MIA affects the fetal brain development of their offspring regardless of the
expression level of COX2 in the embryos. Specifically, we cannot tell if fetuses from COX2-
MKO dams would suffer abnormal outcomes in brain development due to an insufficient
immune system activation in the moms rather than the embryos. Therefore, in this project, we
decided to mate COX2-FLOX females with COX2-MKO males and not the other way around.
Yet, it might provide some additional insight of mechanism for MIA by including COX2-MKO
females in future studies to see if preventing microglial COX2 activity in the moms would also
prevent MIA effects in the fetuses.

44
4.3 Implications and future directions
In this project, we used pharmacological and genetic approaches in an animal model to
show that COX2 expression in Iba1+ microglia acutely induced by gestational MIA is driving
abnormal fetal brain development such as altered BBB formation and permeability. However, the
precise mechanism behind such influence is still unclear. Therefore, in addition to making the
effort to include more control groups and rule out potential unclarity due to the limitations
mentioned in the previous section, there are a few other aspects we are interested in for future
studies.
Apart from the acute effect prenatally, recent studies in our lab have also found
neurodevelopmental abnormalities in adults from the MIA dams as well, aging from postnatal
day (PD) 30 to 180. Specifically, we observed the phenotype of BBB leakage together with the
presence of COX2 expression and increased microglia activity. We also performed behavior tests
on these animals which showed anxiety- and memory-associated behavioral deficits. Fortunately,
like what was seen with the fetuses, CX treatment 24 h after MIA induction also prevented the
postnatal anomalies as well as the altered behaviors in adult offspring. Given the fact that the
effects of maternal inflammation on BBB leakage and brain inflammation are long-lasting
throughout offspring life, the data suggest that inflammation during pregnancy could lead to
increased risk of age-related neurodegenerative diseases in which BBB function is altered, such
as AD in mice and human patients (Li et al., 2016; Montagne et al., 2015; van de Haar et al.,
2016). In order to get a more thorough understanding of the connection between gestational MIA
and development of non-genetic AD, we are now studying brain pathology in one-year-old
offspring from MIA dams.
45
Furthermore, it will be informative to perform long-term postnatal studies in the
transgenic mice as well. Literature shows that COX2 deletion from all cell types can be
detrimental during a neuroinflammatory response (Choi et al., 2010). However, in our
conditional COX2 knockout model, we did not observe any fatal signs due to the gene
modification in the fetuses or in the transgenic mice colony. Thus, we plan to extend the research
presented in this thesis project to PD 30, PD 180, and one-year-old COX2-MKO mice.
In conclusion, this project provides an important finding that activation of the COX2
pathway mediates neurodevelopmental impacts of MIA via disruption of development in fetal
BBB formation. Additional unpublished results from the lab suggest that MIA-induced BBB
disruption remains dependent on COX2 activation in microglia throughout the offspring lifespan,
potentially leading to premature brain aging and associated dementia. Thus, although very
speculative at this point, our research may uncover novel etiological factors leading to AD-like
diseases in aging, which could also have the potential for novel preventative treatments of these
devastating diseases.

46
BIBLIOGRAPHY
Bale, T. L., Baram, T. Z., Brown, A. S., Goldstein, J. M., Insel, T. R., McCarthy, M. M.,
Nemeroff, C. B., Reyes, T. M., Simerly, R. B., Susser, E. S., & Nestler, E. J. (2010). Early
Life Programming and Neurodevelopmental Disorders. Biological Psychiatry, 68(4), 314–
319. https://doi.org/10.1016/J.BIOPSYCH.2010.05.028
Boulanger-Bertolus, J., Pancaro, C., Mashour, G. A., Stanwood, G., Woodruff, T. M., & Robson,
M. J. (2018). Increasing Role of Maternal Immune Activation in Neurodevelopmental
Disorders. https://doi.org/10.3389/fnbeh.2018.00230
Chen, H. Y. (2021). Improving the reproducibility of a maternal immune activation model in
mice.
Choi, S. H., Aid, S., & Bosetti, F. (2009). The distinct roles of cyclooxygenase-1 and -2 in
neuroinflammation: implications for translational research. Trends in Pharmacological
Sciences, 30(4), 174–181. https://doi.org/10.1016/J.TIPS.2009.01.002
Choi, S. H., Aid, S., Choi, U., & Bosetti, F. (2010). Cyclooxygenases-1 and -2 differentially
modulate leukocyte recruitment into the inflamed brain. The Pharmacogenomics Journal,
10(5), 448–457. https://doi.org/10.1038/TPJ.2009.68
Clausen, B. E., Burkhardt, C., Reith, W., Renkawitz, R., & Förster, I. (1999). Conditional gene
targeting in macrophages and granulocytes using LysMcre mice. In Transgenic Research
(Vol. 8).
Croy, A. (2014). The Guide to Investigation of Mouse Pregnancy. http://www.denisetse-
portfolio.blogspot.ca/.
Deverman, B. E., & Patterson, P. H. (2009). Cytokines and CNS Development. Neuron, 64(1),
61–78. https://doi.org/10.1016/J.NEURON.2009.09.002
Foudi, N., Gomez, I., Benyahia, C., Longrois, D., & Norel, X. (2012). Prostaglandin E2 receptor
subtypes in human blood and vascular cells. European Journal of Pharmacology, 695(1–3),
1–6. https://doi.org/10.1016/J.EJPHAR.2012.08.009
Ginhoux, F., Greter, M., Leboeuf, M., Nandi, S., See, P., Gokhan, S., Mehler, M. F., Conway, S.
J., Ng, L. G., Stanley, E. R., Samokhvalov, I. M., & Merad, M. (2010). Fate mapping
analysis reveals that adult microglia derive from primitive macrophages. Science,
330(6005), 841–845. https://doi.org/10.1126/science.1194637
Gumusoglu, S. B., & Stevens, H. E. (2019). Maternal Inflammation and Neurodevelopmental
Programming: A Review of Preclinical Outcomes and Implications for Translational
Psychiatry. Biological Psychiatry, 85(2), 107–121.
https://doi.org/10.1016/J.BIOPSYCH.2018.08.008
47
Heyne, G. W., Plisch, E. H., Melberg, C. G., Sandgren, E. P., Peter, J. A., & Lipinski, R. J.
(2015). A Simple and Reliable Method for Early Pregnancy Detection in Inbred Mice. In
Journal of the American Association for Laboratory Animal Science (Vol. 54, Issue 4).
Jia, W., Lu, R., Martin, T. A., & Jiang, W. G. (2014). The role of claudin-5 in blood-brain barrier
(BBB) and brain metastases (Review). Molecular Medicine Reports, 9(3), 779–785.
https://doi.org/10.3892/MMR.2013.1875/HTML
Kadry, H., Noorani, B., & Cucullo, L. (2020). A blood–brain barrier overview on structure,
function, impairment, and biomarkers of integrity. In Fluids and Barriers of the CNS (Vol.
17, Issue 1). BioMed Central Ltd. https://doi.org/10.1186/s12987-020-00230-3
Knuesel, I., Chicha, L., Britschgi, M., Schobel, S. A., Bodmer, M., Hellings, J. A., Toovey, S., &
Prinssen, E. P. (2014). Maternal immune activation and abnormal brain development across
CNS disorders. Nature Reviews. Neurology, 10(11), 643–660.
https://doi.org/10.1038/NRNEUROL.2014.187
Li, X. Y., Wang, F., Chen, G. H., Li, X. W., Yang, Q. G., Cao, L., & Yan, W. W. (2016).
Inflammatory insult during pregnancy accelerates age-related behavioral and
neurobiochemical changes in CD-1 mice. Age (Dordrecht, Netherlands), 38(3).
https://doi.org/10.1007/S11357-016-9920-3
Meriwether, D., Sulaiman, D., Volpe, C., Dorfman, A., Grijalva, V., Dorreh, N., Sergio
Solorzano-Vargas, R., Wang, J., O’Connor, E., Papesh, J., Larauche, M., Trost, H.,
Palgunachari, M. N., Anantharamaiah, G. M., Herschman, H. R., Martin, M. G., Fogelman,
A. M., & Reddy, S. T. (2019). Apolipoprotein A-I mimetics mitigate intestinal
inflammation in a COX2-dependent inflammatory disease model. The Journal of Clinical
Investigation, 129(9), 3670–3685. https://doi.org/10.1172/JCI123700
Meyer, U. (2019). Neurodevelopmental Resilience and Susceptibility to Maternal Immune
Activation. Trends in Neurosciences, 42(11), 793–806.
https://doi.org/10.1016/J.TINS.2019.08.001
MGI - Inbred Strains: C57BL. (n.d.). Retrieved May 19, 2022, from
http://www.informatics.jax.org/inbred_strains/mouse/docs/C57BL.shtml
Montagne, A., Barnes, S. R., Sweeney, M. D., Halliday, M. R., Sagare, A. P., Zhao, Z., Toga, A.
W., Jacobs, R. E., Liu, C. Y., Amezcua, L., Harrington, M. G., Chui, H. C., Law, M., &
Zlokovic, B. v. (2015). Blood-brain barrier breakdown in the aging human hippocampus.
Neuron, 85(2), 296–302. https://doi.org/10.1016/J.NEURON.2014.12.032
Montagne, A., Nikolakopoulou, A. M., Zhao, Z., Sagare, A. P., Si, G., Lazic, D., Barnes, S. R.,
Daianu, M., Ramanathan, A., Go, A., Lawson, E. J., Wang, Y., Mack, W. J., Thompson, P.
M., Schneider, J. A., Varkey, J., Langen, R., Mullins, E., Jacobs, R. E., & Zlokovic, B. v.
(2018). Pericyte degeneration causes white matter dysfunction in the mouse central nervous
system. Nature Medicine, 24(3), 326–337. https://doi.org/10.1038/nm.4482
48
Phillis, J. W., Horrocks, L. A., & Farooqui, A. A. (2006). Cyclooxygenases, lipoxygenases, and
epoxygenases in CNS: Their role and involvement in neurological disorders. In Brain
Research Reviews (Vol. 52, Issue 2, pp. 201–243).
https://doi.org/10.1016/j.brainresrev.2006.02.002
Ricciotti, E., & Fitzgerald, G. A. (2011). ATVB in Focus Inflammation Prostaglandins and
Inflammation. https://doi.org/10.1161/ATVBAHA.110.207449
Saili, K. S., Zurlinden, T. J., Schwab, A. J., Silvin, A., Baker, N. C., Sidney, E., Iii, H., Ginhoux,
F., & Knudsen, T. B. (2017). Blood-Brain Barrier Development: Systems Modeling and
Predictive Toxicology EPA Public Access. Birth Defects Res, 109(20), 1680–1710.
https://doi.org/10.1002/bdr2.1180
Smolders, S., Notter, T., Smolders, S. M. T., Rigo, J. M., & Brône, B. (2018). Controversies and
prospects about microglia in maternal immune activation models for neurodevelopmental
disorders. Brain, Behavior, and Immunity, 73, 51–65.
https://doi.org/10.1016/J.BBI.2018.06.001
Takahashi, Y., Roman, C., Chemtob, S., Tse, M. M., Lin, E., Heymann, M. A., & Clyman, R. I.
(2000). Cyclooxygenase-2 inhibitors constrict the fetal lamb ductus arteriosus both in vitro
and in vivo. American Journal of Physiology. Regulatory, Integrative and Comparative
Physiology, 278(6). https://doi.org/10.1152/AJPREGU.2000.278.6.R1496
Tornavaca, O., Chia, M., Dufton, N., Almagro, L. O., Conway, D. E., Randi, A. M., Schwartz,
M. A., Matter, K., & Balda, M. S. (2015). ZO-1 controls endothelial adherens junctions,
cell-cell tension, angiogenesis, and barrier formation. Journal of Cell Biology, 208(6), 821–
838. https://doi.org/10.1083/JCB.201404140/VIDEO-2
van de Haar, H. J., Burgmans, S., Jansen, J. F. A., van Osch, M. J. P., van Buchem, M. A.,
Muller, M., Hofman, P. A. M., Verhey, F. R. J., & Backes, W. H. (2016). Blood-Brain
Barrier Leakage in Patients with Early Alzheimer Disease. Radiology, 281(2), 527–535.
https://doi.org/10.1148/RADIOL.2016152244
Woods, L., Perez-Garcia, V., & Hemberger, M. (2018). Regulation of Placental Development
and Its Impact on Fetal Growth—New Insights From Mouse Models. Frontiers in
Endocrinology, 9, 570. https://doi.org/10.3389/FENDO.2018.00570/BIBTEX
Zhao, Q., Dai, W., Chen, H. Y., Jacobs, R. E., Zlokovic, B. v, Lund, B. T., Montagne, A., &
Bonnin, A. (2022). Prenatal disruption of blood-brain barrier formation via
cyclooxygenase activation leads to lifelong brain inflammation. https://doi.org/10.1073/pnas

Abstract (if available)

Abstract Maternal immune activation (MIA) during pregnancy induces many significant neuropathologies in the offspring. Using a synthetic double stranded RNA (dsRNA) molecule, polyriboinosinic-polyribocytidylic acid [poly(I:C)], which mimics virus infection in animal models, it has been discovered that the abnormal neurodevelopmental phenotypes occur both acutely during gestational stages and persistently into adulthood. However, the underlying mechanism for such phenotypes are yet to be uncovered. In our lab, we seek to identify key pathways and/or molecules responsible for the abnormalities and provide some insight for future studies on gestational MIA treatment development.
We developed our initial MIA model using CD-1 mice and observed increased permeability and structural disruption of fetal blood-brain barrier (BBB), increased number of microglial cells displaying increased expression and activation of cyclooxygenase 2 (COX2) enzyme. This led to our hypothesis that activation of COX2-expressing microglia plays an important role in MIA impact on BBB formation. We tested this hypothesis using pharmacological and genetic approaches to suppress COX2 activity at time of MIA induction. Injection of celecoxib, a selective COX2 inhibitor, prevented all the abnormalities observed in MIA fetuses.
To investigate the role of COX2 expression more specifically in microglia, we next used a conditional COX2 knockout mouse model. Due to these mice being on a C57BL/6J genetic background, we first determined an optimal poly(I:C) dosage to trigger MIA phenotypes comparable to those obtained in CD-1 mice. We found that genetically deleting COX2 gene in microglia prevented all effects on the fetal BBB.
Taken together, our data demonstrate that induction of COX2 expression in microglia serves a significant part in MIA disruption of fetal BBB. These findings provide promising new directions for research on early life treatments to prevent neuropsychiatric pathologies stemming from prenatal MIA.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Insulin sensitivity in cognition, Alzheimer's disease and brain aging

PDF

Regulation of mitochondrial bioenergetics via PTEN (phosphatase and tensin homolog deleted on chromosome 10)/estrogen-related receptor alpha (ERRα) signaling

PDF

Role of purinergic P2X7 receptors in inflammatory responses in the brain and liver: a study using a mouse model of chronic ethanol and high-fat diet exposure

Asset Metadata

Creator Dai, Weiye (author)

Core Title Maternal inflammation disrupts fetal blood-brain barrier formation via cyclooxygenase activation

School School of Pharmacy

Degree Master of Science

Degree Program Molecular Pharmacology and Toxicology

Degree Conferral Date 2022-08

Publication Date 07/23/2022

Defense Date 07/22/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag BBB hyper permeability,celecoxib,cre/loxP conditional knockout,cyclooxygenase 2,maternal immune activation,microglial cells,OAI-PMH Harvest,over activation of fetal immune response,prenatal neurodevelopment

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bonnin, Alexandre (committee chair), Hamm-Alvarez, Sarah (committee chair), Shih, Jean Chen (committee member)

Creator Email weiyedai@g.ucla.edu,weiyedai@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111374356

Unique identifier UC111374356

Legacy Identifier etd-DaiWeiye-10936

Document Type Thesis

Format application/pdf (imt)

Rights Dai, Weiye

Type texts

Source 20220728-usctheses-batch-962 (batch), 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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

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

Repository Email cisadmin@lib.usc.edu