University of Southern California Dissertations and Theses

TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

(USC Thesis Other)

TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

TLR4-mediated innate immune response and neuroinflammation:
focus on APOE and obesity

by

Jiahui Liu

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOLOGY OF AGING)

December 2022

Copyright 2022 Jiahui Liu
ii
Acknowledgements

Foremost, I would like to express my deepest gratitude to my advisor Dr. Christian Pike.
Thank you for giving me the opportunity to join your lab before my Ph.D. program. Thank you for
your patience and guidance in all aspects of being a good research scientist. I could not have
imagined having a better mentor for my Ph.D. study.

Special thanks to Dr. Amy Christensen, who not only trained me with all the experiments
and animal works, but also provided valuable advice on many aspects of the research in this
dissertation. I’m thankful to Dr. Alexandra Moser for your training and helpful discussions during
my first few years in Pike Lab. Thanks should also go to present and former lab members,
Cassandra McGill, Alicia Quihuis, Ali Zaidi, Dr. Tomomitsu Iida, Dr. Terri Stephen, Dr. Camille
Sample, Bayla Breningstall, and Wenjie Qian for your help with my research and a joyful
environment that brightened my days in the lab.

I would like to thank my committee members, Dr. Sean Curran, Dr. Bérénice Benayoun,
and Dr. Lisa Ellerby, for your advice and help.

I would like to thank my parents, for your endless love and support throughout my life, for
always believing I could become a scientist ever since I was a child. Also a big thank to my
boyfriend and best friend, Jinsen Li, for your emotional support and encouragement. I would have
not made it without you. Finally, thanks to my snowboard and ski friends for all the fun we had on
the mountains.

iii

Table of Contents

Acknowledgements ........................................................................................................................ ii
List of Figures ............................................................................................................................... iv
Abstract ......................................................................................................................................... vi
Chapter 1: Introduction ................................................................................................................. 1
1. Alzheimer’s disease and innate immune response ............................................................... 1
2. APOE4 and innate immune response ................................................................................... 2
3. Obesity and inflammation ..................................................................................................... 3
4. Microglia and neuroinflammation ........................................................................................ 10
5. TLR4 signaling .................................................................................................................... 12
6. Dissertation objectives ........................................................................................................ 13
Chapter 2: Effects of APOE genotype and estrogen status on innate immune challenge in
female mice ................................................................................................................................ 15
1. Introduction ......................................................................................................................... 16
2. Methods .............................................................................................................................. 17
3. Results ................................................................................................................................ 21
4. Discussion ........................................................................................................................... 34
5. Conclusions ........................................................................................................................ 39
Chapter 3: Role of microglial TLR4 in obesity-induced neural impairments ............................... 40
1. Introduction ......................................................................................................................... 41
2. Methods .............................................................................................................................. 42
3. Results ................................................................................................................................ 53
4. Discussion ........................................................................................................................... 94
5. Conclusion ........................................................................................................................ 101
Chapter 4: Conclusions and future directions ........................................................................... 102
1. Summary of the findings ................................................................................................... 102
2. Other mechanisms underlying obesity .............................................................................. 103
3. Future directions ............................................................................................................... 114
References ............................................................................................................................... 118

iv
List of Figures

Figure 1. Effects of APOE genotype and estrogen status on body and tissue weights. ............. 23
Figure 2. Effects of APOE genotype and estrogen status on sickness behavior. ....................... 25
Figure 3. Effects of APOE genotype and estrogen status on thermoregulatory response
to LPS. ........................................................................................................................................ 27
Figure 4. Effects of APOE genotype and estrogen status on cytokine levels in plasma. ........... 29
Figure 5. Effects of APOE genotype and estrogen status on cytokine expression in liver. ........ 31
Figure 6. Effects of APOE genotype and estrogen status on cytokine expression in brain. ....... 33
Figure 7. Validation of TLR4 deletion efficacy in TLR4-MKO mice. ............................................ 55
Figure 8. Effects of tamoxifen treatment on adiposity. ................................................................ 55
Figure 9. Effects of HFD and microglial TLR4 deletion on body weight and metabolic
outcomes in male mice. .............................................................................................................. 59
Figure 10. Effects of HFD and microglial TLR4 deletion on body weight and metabolic
outcomes in female mice. ........................................................................................................... 63
Figure 11. Effects of HFD and microglial TLR4 deletion on adipose inflammation in male
mice. ........................................................................................................................................... 65
Figure 12. Effects of HFD and microglial TLR4 deletion on adipose inflammation in
female mice. ............................................................................................................................... 67
Figure 13. Effects of HFD and microglial TLR4 deletion on plasma cytokine levels in
male mice. .................................................................................................................................. 69
Figure 14. Effects of HFD and microglial TLR4 deletion on plasma cytokine levels in
female mice. ............................................................................................................................... 70
Figure 15. Effects of HFD and microglial TLR4 deletion on gene expression in the
hypothalamus in male mice. ....................................................................................................... 72
Figure 16. Effects of HFD and microglial TLR4 deletion on gene expression in the
hypothalamus in female mice. .................................................................................................... 74
v
Figure 17. Effects of HFD and microglial TLR4 deletion on Iba1+ cell morphology in the
ARC in male mice. ...................................................................................................................... 76
Figure 18. Effects of HFD and microglial TLR4 deletion on GFAP immunoreactivity in the
ACR in male mice. ...................................................................................................................... 77
Figure 19. Effects of HFD and microglial TLR4 deletion on Iba1+ cell morphology in the
ARC in female mice. ................................................................................................................... 79
Figure 20. Effects of HFD and microglial TLR4 deletion on GFAP immunoreactivity in the
ACR of female mice. ................................................................................................................... 80
Figure 21. Effects of HFD and microglial TLR4 deletion on gene expression in the
hippocampus in male mice. ........................................................................................................ 82
Figure 22. Effects of HFD and microglial TLR4 deletion on gene expression in the
hippocampus in female mice. ..................................................................................................... 83
Figure 23. Effects of HFD and microglial TLR4 deletion on microglia morphology and
GFAP immunoreactivity in the hippocampus of male mice. ....................................................... 85
Figure 24. Effects of HFD and microglial TLR4 deletion on microglia morphology and
GFAP immunoreactivity in the hippocampus of female mice. .................................................... 86
Figure 25. Effects of microglial TLR4 deletion on exploration and anxiety-like behaviors
in male and female mice. ............................................................................................................ 89
Figure 26. Effects of HFD and microglial TLR4 deletion on cognitive performance in male
and female mice. ........................................................................................................................ 91
Figure 27. Effects of HFD and microglial TLR4 deletion on hippocampal neurogenesis in
male and female mice. ................................................................................................................ 93

vi
Abstract

Excessive activation of innate immunity and inflammation are associated with and may
contribute to many age-related conditions, such as Alzheimer’s disease (AD) and obesity. The
toll-like receptor 4 (TLR4) signaling pathway has been proposed as one of the main factors of
disease associated inflammatory responses. In my dissertation, I investigated the relationship
between TLR4-mediated innate immune response and neuroinflammation in the context of the
AD genetic risk factor Apolipoprotein E (APOE) allele e4 (APOE4) and obesity. Chapter 1
provides an introduction to topics relevant to my dissertation. I begin by examining the
association between innate immune response and AD, and how APOE genotype interacts with
innate immunity in driving AD. I then addressed obesity-induced neural dysfunctions as well as
possible underlying mechanisms, including inflammation, microglia activation, and TLR4
signaling pathway. In Chapter 2, I examined the effect of APOE genotype on innate immune
responses in female mice. I also examined whether this relationship is affected by estradiol
levels. Results from this study suggested that APOE4 was associated with relatively protective
outcomes in acute inflammatory response. In Chapter 3, I investigated the role of microglial
TLR4 signaling in modulating obesity-induced neural inflammation and dysfunction using a
newly generated mouse model in male and female mice. I found that microglial TLR4 deletion
yielded protection against diet-induced metabolic disruption, peripheral inflammation, cognitive
impairment, and neurogenesis with sex differences. Chapter 5 is a summary of my key findings.
I also discussed possible improvements that can be made as well as some future directions.

1
Chapter 1: Introduction

1. Alzheimer’s disease and innate immune response
Alzheimer’s disease (AD) is an age-related neurodegenerative disease characterized by
accumulation of amyloid-β (Aβ) plaques, neurofibrillary/tau tangles and neuronal loss. There are
a number of genetic and environmental factors that affect AD risk. Recently, studies have
provided evidence that the innate immune system both in the peripheral and central nervous
system (CNS) is an important regulator in the initiation and progression of the disease
1–3
.
So far, genome-wide association studies and whole exome/genome sequencing have
identified more than 30 AD risk genes. Importantly, many of them are associated with innate
immune responses and microglial function, such as APOE, CD33 and Trem2
4,5
. When
assessing epigenomic changes in an AD mouse model, a recent study showed that there was a
significant enrichment in enhancers and promotors implied in innate immunity, along with an
increase in overlap between these enhances and reported AD-related risk genes
6
.
Besides Aβ and tau pathology, important hallmarks observed in the AD brains include
increased microglial activation and associated neuroinflammation. Many studies have observed
increased levels of inflammatory cytokines in the blood and cerebrospinal fluid (CSF) in AD
patients
7
. In postmortem AD brains, activated microglia were found surrounding Aβ plaques and
tau tangles
8,9
. It has been thought that microglia activation was triggered by Aβ which then led to
the production of pro-inflammatory cytokines such tumor necrosis factor α (TNFα), interleukin-
1β (IL-1β), and interleukin-6 (IL-6)
10
. Microglial activation could drive many of the adverse neural
outcomes under pathological conditions. There was a significant negative correlation between
microglial activation and structural integrity
11,12
and functional connectivity
13
of the brain in AD
patients. Microglia induced neuroinflammation was also associated with impairments in
cognitive performance
13,14
. In animal models of AD, activated microglia produced toxic factors
2
such as reactive oxygen species and nitric oxide that could lead to synapse loss and neuronal
injuries
15
. In addition, reactive microglia could crosstalk with astrocytes, inducing astrocyte
activation which caused neuronal death
16
. Together, these findings further implied that innate
immune response is an essential mechanism underlying AD pathogenesis.
Studies that have manipulated inflammatory status to investigate effects of
neuroinflammation on Aβ pathology yielded mixed results, as some showed it has beneficial
effects while others suggested the opposite. For example, overexpressing pro-inflammatory
cytokines TNFα or IL-6
17,18
, or suppressing anti-inflammatory cytokine interleukin-10 (IL-10)
19
in
AD mice decreased Aβ deposition in the brain. In contrast, other studies showed that increased
level of neuroinflammation induced by central administration of lipopolysaccharide (LPS) or low
levels of anti-inflammatory cytokine interleukin-4 (IL-4) was associated with exaggerated Aβ
pathology
20,21
. Therefore, more studies are needed to better understand the complex role of
neuroinflammation in AD.

2. APOE4 and innate immune response
APOE codes for ApoE protein, which transports cholesterol and other lipids by binding
with its receptors
22,23
. It also affects neuronal signaling involved in synaptic plasticity
22,23
. There
are three isoforms of APOE, APOE2, APOE3 and APOE4. The greatest genetic risk factor of
AD is the e4 allele of APOE. APOE4 carriers showed a much earlier onset to develop AD and
were associated with greater cognitive decline in a gene dose-dependent manner
24
. In animal
models, APOE4 mice showed greater glial activation along with more Aβ accumulation in the
hippocampus after inhibiting the activity of Aβ degrading enzyme
25
. In another AD mouse model
(EFAD mice), APOE4 was also associated with more Aβ deposition and exhibited a more
compact plaque morphology compared with APOE3 mice
26
.
3
Studies have indicated that APOE4 may contribute to AD by altering innate immunity
and inflammatory response
27
. When treating the blood collected from healthy subjects with
lipopolysaccharide (LPS), a major component of the outer membrane wall of Gram-negative
bacteria and a ligand for toll-like receptor 4 (TLR4), APOE4 was associated with higher
production of pro-inflammatory cytokines
28
. In addition, APOE4 carriers showed higher
hyperthermia and increased plasma levels of pro-inflammatory cytokines after challenged with
intravenous administration of LPS
28
.
The observed higher inflammatory response in APOE4 carriers were recapitulated in
animal models. In cultured cells of a macrophage line, LPS-induced higher production of TNFα
and lower levels of IL-10 in APOE4 macrophages
29
. Moreover, APOE4 mice showed more
pronounced hypothermia, accompanied with higher levels of plasma TNFα and tissue injuries
after peripheral administration of LPS
28
. The effects of APOE4 on innate immune response in
the brain have also been studied. In primary culture of microglia, APOE4 was associated with
higher basal levels of inflammatory cytokines
30,31
as well as greater production of TNFα, IL-6,
and IL-1b after LPS treatment
31,32
. In vivo, central administration of LPS induced significant
inflammatory response including glial activation and inflammatory cytokines production that was
also more pronounced in the APOE4 mice
33–35
.
Collectively, growing evidence from human and animal studies shown that APOE is an
important modulator of innate immunity, and generally APOE4 was associated with a negative
effect on peripheral and neuroinflammation after innate immunity activation.

3. Obesity and inflammation
Obesity is recognized as a global health concern. Mean body mass index (BMI) has
remarkably increased in both developed and developing countries in the past 40 years, with
19.5% of the adult are classified as obese
36
. Obesity is significantly associated with a wide
4
range of medical conditions, including metabolic syndrome
37
, diabetes
38
, cardiovascular
disease
39
, and certain cancers
40
. Obesity is also associated with many neural dysfunctions such
as decreased hippocampal volume
41
, changes in gray matter and fiber density
42
, loss of white
matter integrity
43
and blood-brain-barrier (BBB) disruption
44
. Most importantly, obesity is linked
with impaired cognition and increased risk of dementia and developing AD
45–47
. Therefore, it is
important to understand the mechanisms leading to both obesity and its adverse outcomes.
Pro-inflammatory pathways are essential components to many of the damaging
metabolic outcomes of obesity
48,49
. In the brain, obesity drives chronic neuroinflammation that is
also robustly associated with reductions in neurogenesis
50
, cognitive dysfunction
51
and the
development of AD
7
. Thus, neuroinflammatory pathways may be critical in mediating obesity-
induced neural dysfunction.

3.1 Obesity induced peripheral inflammation
Obesity is characterized by chronic, low-grade inflammation
52
. Different than pathogen-
induced inflammation, obesity-related inflammation is triggered by over consumption of nutrients
and can lead to alterations in many organs. So far, it has been mostly described in peripheral
tissues.
The association between obesity and inflammation was firstly reported in adipose
tissues with an increased level of TNFα in obese mice compared to lean mice. Subsequent
studies demonstrate that not only TNFα, but a large number of cytokines and chemokines are
increased in adipose tissues, including IL-6, IL-1β, chemokine ligand 2 (CCL2) and C-X-C motif
chemokine 10 (CXCL10)
53,54
. Accompanied by increased level of proinflammatory mediators,
morphology of individual adipocytes and whole adipose tissues is also altered with obesity.
Under energy imbalance, preadipocytes differentiate and adipocytes store excess energy,
leading to an increase both in number and size
55
.
5
Another hallmark of inflammation in adipose tissues is infiltration of macrophages and
other immune cells. Macrophage population was increased
56
, and expression of many
macrophage-specific genes were strongly increased in either genetic (ob/ob and db/db mice) or
high-fat diet (HFD)-induced obese mice
57
. In a study that used irradiation and followed by bone
marrow transplantation, it was shown that most macrophages in the adipose tissue were
derived from the bone marrow progenitors
56
. Subsequent studies found that these infiltrated
cells were largely monocyte-derived, and that the infiltration was dependent on C-C motif
chemokine receptor 2 (CCR2), as CCR2 knockout reduced macrophage accumulation along
with inflammatory profile in the adipose tissue
58,59
. However, there was also evidence showing
that resident macrophages within adipose tissue was able to proliferate and contributed to
macrophage accumulation associated with obesity
60
. Interestingly, macrophages rather than
adipocytes seem to be the main source of inflammatory cytokines in the adipose tissue. The
expression levels of inflammatory markers such as TNFα, IL-6, Nos2 and IKKβ were
significantly higher in the macrophage-enriched cell fraction compared with other cells in the
adipose tissue
56,57
. Of note, there is a vicious cycle between adipose inflammation and obesity-
induced metabolic dysregulation. Not only does obesity induce inflammation, but also
inflammation in the adipose tissue can disrupt metabolic processes which further drives obesity.
In other words, inflammatory mediators produced by macrophages lead to alterations in
peripheral organs and impairment in insulin signaling transduction, which could further
contribute to the pathogenesis of obesity
57
.
Inflammation in adipose tissue contributes to obesity-induced insulin resistance.
Exposure of TNFα impaired insulin signaling and reduced insulin-induced glucose uptake in the
adipocytes
61,62
. Moreover, genetic mutations in TNFα protected mice from impairment in glucose
dysregulation and insulin sensitivity
63,64
. Mice lacking other inflammatory mediators including c-
Jun N-terminal kinase (JNK), IKKe and TLR2 were also associated with improved metabolic
functions after HFD-feeding
65–67
. Studies have also investigated the effects of macrophage-
6
specific inflammation on obesity induced metabolic dysfunction. Depleting CD11c+
macrophages decreased adipose inflammation and attenuated insulin resistance in obese mice
induced by HFD
68
. Similarly, deletion of inflammatory modulator IKKβ specifically in the myeloid
cells protected mice from glucose intolerance and insulin resistance
69
.
Together, current data suggest a link between adipose inflammation in obesity-induced
insulin resistance and that suppressing macrophage activation and inflammation are beneficial
for metabolic homeostasis.

3.2 Obesity-induced neuroinflammation
3.2.1 Hypothalamic inflammation
Recent research has identified an association between obesity and a similar type of low-
grade inflammation in the central nervous system (CNS). Since the hypothalamus is the site of
central regulation of food intake and energy expenditure
70,71
, studies have largely focused on
hypothalamic inflammation induced by obesity and implicated it as an important mediator of
metabolic diseases.
After a 4-month period of HFD feeding in rats, JNK and nuclear factor-κB (NF-κB)
signaling were strongly activated in the hypothalamus, which resulted in an increased
production of pro-inflammatory cytokines such as TNF-α, IL-1β, and IL-6 along with impairment
in insulin and leptin signaling
72
. Different than inflammation in the adipose tissue, which typically
is only observed several weeks after HFD treatment is initiated
57,73
, some studies have shown
that the inflammation induced by HFD can develop much faster in the hypothalamus than in the
peripheral tissues and is independent of body weight gain, suggesting a causal role of
hypothalamic inflammation in early metabolic alterations in induced by obesity. For example,
expressions of inflammatory cytokines were significantly increased after just 24 hours of HFD
consumption prior to any significant body weight gain
74
. Short-period infusions (6h) of glucose or
oleic acid through third ventricle following fasting were also shown to induce hypothalamic
7
inflammation via NF-κB activation, but a similar effect was not observed in peripheral tissues
75
.
Similarly, central administration with long-chain saturated fatty acids (SFAs) increased
production of inflammatory cytokines in the hypothalamus within 3 days
76
.Even after HFD was
switched back to a chow diet, body weight decreased, and insulin sensitivity improved,
hypothalamic inflammation was still evident
77
. In this context, pharmacologic inhibition or genetic
modification of inflammatory pathways specifically in CNS attenuated the effects of HFD-
induced weight gain and leptin resistance
72,75,76
.
Hypothalamic inflammation has been proposed to induced insulin and leptin resistance
by promoting stress and death of neurons that regulate appetite and food consumption. Indeed,
hypothalamic inflammation is associated with altered firing activity of agouti-related peptide
(AgRP) and proopiomelanocortin (POMC) neurons as well as POMC neuron injuries and
loss
74,78
. Furthermore, treating the animals with TNFα by intracerebroventricular (ICV) injections
alone was able to induce neuronal apoptosis in the hypothalamus
79
.
While a variety of peripheral signals regulate hypothalamic function, inflammation in the
hypothalamus also exerts broad effects on peripheral tissues. For example, after ICV
administration of low-dose TNFα, rats showed increased insulin secretion from isolated
pancreatic islets, disruption of insulin signaling in the liver and skeletal muscle, and finally
reduced thermogenesis without significant changes in body weight
80
. Moreover, induction of
brain ER stress in mice caused glucose intolerance and systemic and hepatic insulin resistance
in the absence of body weight gain, and these effects are diminished by hypothalamic NF-κB
inhibition
81
.
Collectively, hypothalamic inflammation is sufficient to induce metabolic impairments
commonly associated with obesity and therefore may contribute to the development of obesity.
Hypothalamus also interacts with peripheral tissues to maintain energy homeostasis.

8
3.2.2 Hippocampal inflammation and cognitive impairment
Neuroinflammation derived from obesity can extend beyond the hypothalamus. The
hippocampus, an important region in learning and memory, is also subjected to inflammation in
obesity.
Like hypothalamus, obesity-induced inflammation in the hippocampus is characterized
by increased activation of proinflammatory NF-κB signaling pathway, expression of cytokines
such as TNFα, IL-1β and IL-6, and activation of microglia and astrocytes
82–87
. Furthermore, HFD
induced neuronal dysfunction in the hippocampus, as neurons from HFD-fed mice showed
decreases in branch length as well as dendritic complexity compared with controls
84
.
Neurogenesis in the dentate gyrus of both dorsal and ventral hippocampus was also impaired
after HFD exposure
88
. Importantly, obesity-associated hippocampal inflammation requires long
periods of HFD feeding. For example, hippocampal inflammatory markers were not significantly
altered after 2 weeks of HFD exposure in rats
89
. Changes in microglia activation and
microgliosis were not observed after up to 8 weeks of HFD treatment
74
. However, when the
consumption of HFD lasted more than 12 weeks, hippocampal inflammation has been
consistently reported in many studies. For instance, a 3-month HFD exposure was associated
with increased protein levels of hippocampal IL1-β and decreased synaptic protein PSD95 and
synaptophysin, but no substantial changes in these markers was found at earlier time points
90
.
Accumulating evidence has identified that obesity and diabetes are associated with
increased risk of developing mild cognitive impairment
45,91
. Adolescents who were obese or with
metabolic syndrome displayed lower cognitive performance
92,93
. In adults, studies also indicated
that high BMI in midlife was associated with lower overall cognitive function in older age
94,95
.
Interestingly, one study has found that men who are obese are more susceptible than women to
develop MCI
96
. Indeed, there was a significant negative correlation between BMI and brain gray
matter volume only in men in a Japanese population
97
. Obesity is associated with not only an
increased risk of development of MCI, but additionally, late-life dementia and AD. Data from meta-
9
analyses showed that obesity in midlife was associated with over two times increased risk of AD
98
.
Besides, morbidly obese patients show higher levels of hippocampal Aβ and tau when compared
with non-obese patients
99
.
Parallel with human findings, experimental studies using animal models have also
demonstrated that obesity is associated with cognitive decline and development of AD-like
pathology. For example, one study showed that consumption of HFD for 12 weeks caused obesity,
insulin resistance, and poor performance in an operant task, which exams short-term memory
retention and executive function
100
. Furthermore, obese rats showed impaired learning and
memory evaluated by radial arm water maze after 3 months of HFD feeding, as they took longer
and made more errors to locate a hidden platform
101
.
Studies have found that hippocampal inflammation may be responsible for the
compromised cognitive function induced by obesity. Overexpression of TNFα in the
hippocampus increased anxiety and impaired memory consolidation in fear conditioning test
102
.
The effects of HFD on cognitive function also seems to be dependent on dietary treatment
duration. For example, rats fed on HFD for 8 weeks showed similar performance on novel object
recognition compared lean mice. In contrast, 16 weeks of HFD feeding induced cognitive
defects, as obese rats spend significantly less time exploring the novel object over the familiar
one
103
. Such impairment was accompanied by increased expression of inflammatory cytokines
in the hippocampus
103
. In another study, 20 weeks of HFD exposure in mice significantly
impaired performance in spatial learning and memory in Morris water maze test, which was
associated with increased hippocampal TNFα and ionized calcium-binding adapter molecule 1
(Iba1)
84
. Furthermore, mice showed cognitive impairment in after 20 weeks of HFD, which is
linked with activation of endoplasmic reticulum stress and IKKβ/NF-κB-mediated inflammatory
signaling in the hippocampus
82
.
10
Thus, timing of the appearance of neuroinflammation varies in different brain regions.
Given sufficient exposure to HFD, hippocampus displays significantly increased inflammation
which might promote cognitive decline.

4. Microglia and neuroinflammation
Microglial cells are the resident macrophages of the CNS. They represent around 10%
of the non-neuronal cells in the CNS parenchyma
104
. These cells actively surveil the
extracellular environment and are considered as the first line of immune defense in case of any
kind of brain injury since they can phagocytose the toxic products, release cytotoxic factors and
behave as antigen-presenting cells
105
. In addition, they play important roles in regulating
neuronal activities during development and adulthood
106,107
.
Traditionally, microglia have been morphologically characterized to exist in two states,
“resting” and “activated”. In non-pathological conditions, microglia display a ramified morphology
and regularly scan the neighboring regions in efforts to maintain homeostasis
108
. As a
consequence of brain injury or disease, they enter an “activated” state especially for those that
are adjacent to the injury site. Once activated, they go through drastic changes in morphology
and migrate to the site of injury
105
. These changes were widely considered as shortened and
thickened processes and hypertrophy of the cell body
105
. However, recent studies observed an
associated between chronic stress and a hyper-ramified microglial morphology that
characterized as increased process length and branching
109–111
.
Growing evidence from transcriptomic studies have showed that activated microglia
displayed complex phenotypes, such that some are beneficial and some are detrimental
depending on the conditions. During homeostasis, transcriptional profile of microglia was
associated with enrichment in genes related to sensing and neural development compared with
other peripheral macrophages
112–115
. In the context of infections or neurodegeneration, genes
involved in both neuroprotective and neurotoxic pathways were upregulated in microglia
112–115
.
11
In addition, recent studies using single-cell analysis showed heterogeneity of microglia during
aging and disease progression in humans and animal models
116,117
, which further highlight the
complexity of microglial function.
The relationship between diet-induced obesity (DIO) and microglia activation is well-
documented. In primary cell culture, exposure of microglial cells to saturated fatty acids (SFAs)
that are abundant in HFD led to increased inflammatory cytokines expression
118,119
. In animal
models, HFD feeding also induced microglia activation that was generally characterized as
increased expression of macrophage/microglia markers Iba1, CD68 and MHCII
120
. Microglia
proliferation and morphological alterations were observed in the hypothalamus with HFD
feeding
74
.In contrast, inhibition of microglial proliferation in the hypothalamus prevented mice
from HFD-induced body weight gain and adiposity, reduced gliosis and restored hypothalamic
leptin sensitivity
121
. In addition, microglia depletion reduced hypothalamic inflammation and
neuronal stress induced by HFD
121
. Microglia-specific deletion of IKKb attenuated microgliosis in
the hypothalamus after HFD feeding, which was associated with reduced food intake and body
weight gain
119
.
Studies also found microglia activation in the brain regions involved in cognition and
memory. In db/db mice, increased levels of microglial activation markers including MHCII and
IL-1b were observed throughout the hippocampus
122
. Similarly, HFD significantly increased
microgliosis, induced NF-kB activation and elevated production of inflammatory cytokines in the
hippocampus in rats
86
. One the other hand, pharmacological inhibition of microglial activation
improved cognitive performance in obese mice induced by HFD
123
.
Together, microglia are key mediators of HFD-induced neuroinflammation. However, it is
still unknown to which extend microglial activation is beneficial in restoring homeostasis or
detrimental in manifesting pathology. Future studies are needed to determine how microglia
might influence neuronal activity and contribute to DIO.
12
5. TLR4 signaling
TLR4 belongs to the toll-like family of receptors, which has well-established roles in
mediating inflammatory events in many tissues
124,125
. Upon binding with pathogen-associated
molecular patterns, TLR4 together with myeloid differentiation factor 88 (MyD88) induce NFkB
translocation into the nucleus that in turn drives expression of pro-inflammatory cytokines
126
.
Importantly, studies showed that TLR4 signaling in monocytes or macrophages could be
stimulated by SFAs
127
likely via cellular metabolism reprogramming
128
.
TLR4 has been implicated as a central regulator of obesity-related inflammation
129
. In
genetically obese or DIO mouse models, TLR4 signaling was significantly activated in
adipocytes
130
. In addition to increasing inflammation, TLR4 signaling activation was shown to
impair insulin signaling in adipocytes by increasing serine phosphorylation of insulin receptor
substrate proteins
131
. TLR4 activation by HFD-feeding can also induce brain ER stress
76
and
neuronal apoptosis in the hypothalamus
132
. Pharmacological inhibition, constitutive knockout, or
tissue-specific deletion of TLR4 attenuated HFD-induced inflammation in various tissues
including adipose tissue, liver, skeletal muscle
130,133–137
, reduced body weight gain
76,135
,
improved glucose dysregulation and insulin resistance
130,133–136,138,139
.
In the CNS, TLR4 is widely expressed by microglia and to a lesser extent by
astrocytes
125
. It has been reported that microglial cells are the major cell type in the CNS that
responded to peripheral injected LPS in a TLR4-dependent manner. Indeed, LPS did not
directly bind to neurons, but instead, neuronal death caused by LPS required the presence of
microglia
140
. Moreover, SFA-induced inflammatory responses appeared to be mostly contributed
by microglia, as primary astrocytes
141
and neurons
142
did not display inflammatory responses
when treated with long-chain SFAs.
Accumulating evidence has linked activation of inflammatory pathways in the brain to the
development as well as progression of metabolic dysregulation induced by obesity. Central
administration of SFAs or HFD treatment caused hypothalamic inflammation, while TLR4 whole
13
body knockout or brain-specific deletion of MyD88 protected animals from obesity-related body
weight gain and leptin resistance
76,143
. In regard of downstream mediators of TLR4 signaling,
IKKb/NF-kB pathway has been implied to mediate neuroinflammation and contribute to DIO.
Pharmacological inhibition of IKKβ/NF-kB signaling in the brain reduced body weight gain,
adiposity, and improved leptin sensitivity in the hypothalamus
144
. Mice with IKKβ deletion
specifically in microglia showed reduced caloric intake and body weight change, and less
infiltration of peripheral macrophages during HFD-feeding
145
. Microglia-specific IKKβ deficiency
also increased survival of neural stem cell and neurogenesis in the hypothalamus
146
.
Together, recent studies have provided evidence for a role of TLR4 signaling in
peripheral and central inflammation, as well as metabolic dysfunctions induced by obesity.
However, the exact mechanisms of how TLR4 signaling is activated upon obesity require further
investigation. It is also necessary to elucidate the relative contributions of microglial TLR4
signaling in the onset of obesity-induced inflammation.

6. Dissertation objectives
Innate immunity dysfunction and inflammation in the brain is associated with many
diseases. Chronic activation of innate immune cells including macrophages and microglia and
releasing inflammatory factors are thought to be harmful under pathological conditions.
Therefore, understanding of the alterations in innate immune functions and identifying the
inflammatory pathways that contribute to the adverse neural effects may allow the development
of therapeutic strategies devoted to controlling inflammatory status, and therefore restoring
innate immune function. Importantly, recent evidence has implicated TLR4 signaling in
contributing to neuroinflammation linked to disease pathogenesis. The goal of my dissertation is
to elucidate the role of TLR4-mediated innate immune response in neuroinflammation, with a
focus on (i) AD risk gene APOE4 and (ii) obesity.
14
APOE4 is the predominant genetic risk factor for late-onset AD, which has been
implicated in contributing to AD risk in part by increasing susceptibility to inflammation
27
.
Consistent with a regulatory role in innate immune responses, acute challenge with the
endotoxin LPS is reported to yield higher inflammation in both human carriers of APOE4 and
mice with knock-in of human APOE4
28,33,34,147
. Another AD risk factor is age-related depletion of
the estrogens in women
148,149
. Estradiol, the primary bioactive estrogen, is also a regulator of
inflammation
150
. Uncertain are the extent to which APOE genotype may modulate innate
immune responses in females and whether this relationship is affected by estradiol levels. To
study these questions, I studied the peripheral, neural, and behavioral effects of acute LPS
challenge on adult female mice homozygous for human APOE3 or APOE4 in the presence or
absence of estradiol, as discussed in Chapter 2.
Obesity is characterized by a chronic state of low-grade inflammation. The CNS innate
immune cells microglia have been implicated in the regulation neuroinflammation, which is a key
contributor to obesity-related cognitive impairment, neural dysfunction, and dementia
151
. Among
all inflammatory pathways, TLR4 signaling might be particularly important. TLR4 expression
increases during obesity in many tissues
152,153
. In the CNS, TLR4 is mostly expressed by
microglia
76,125,154
. While many studies using pharmacological inhibition or genetic modifications
of TLR4 have demonstrated a role of TLR4 signaling in regulating systemic effects of obesity,
the current knowledge of the effects microglial TLR4 neural obesity outcomes are largely
unknown. Therefore, in Chapter 3 I used a newly generated mouse model with TLR4 deletion
specifically in microglia, to investigate the contribution of TLR4-mediated neuroinflammation to
the adverse neural effects of obesity.
Together, the results of my studies provide new insights into how innate immunity
modulates inflammation to drive both AD and obesity. This investigation into the role of the
innate immune response will help us understand the underlying mechanisms and develop novel
treatment strategies.
15
Chapter 2: Effects of APOE genotype and estrogen status on
innate immune challenge in female mice

Abstract
Risk for Alzheimer’s disease (AD) is multifactorial and includes apolipoprotein E e4
(APOE4) genotype, innate immune pathways, and, in females, depletion of estrogens. Although
female sex and estrogen status are known to affect innate immunity as well as the effects of
APOE4, possible interactions among these factors is not well understood. In this study, we
compared the systemic and neural effects of acute innate immune challenge across APOE3 and
APOE4 genotypes in female mice under varying estrogen conditions (sham ovariectomy and
ovariectomy with vehicle or estradiol). As expected, exposure to the innate immune stimulus
lipopolysaccharide (LPS) resulted rapid and robust increases in sickness behavior,
thermoregulatory responses, and cytokine levels. Both sickness behavior and hypothermia, which
are largely though to represent adaptive response to infection, were greater in APOE4 than
APOE3 females. LPS-induced cytokine levels in plasma and liver were either similar across
genotypes or lower in APOE4 mice, suggesting attenuated cytokine responses with APOE4
genotype. Interestingly, basal cytokine levels also tended to be lower in APOE4 mice. Estrogen
status significantly affected basal body weight, adiposity, and uterine weight in a manner that was
stronger in APOE4 versus APOE3 mice. Perhaps unexpectedly, estrogen status had modest,
largely nonsignificant effects on LPS-induced outcomes across tissues. Collectively, these
findings suggest that, in young adult female mice, APOE4 significantly affects acute immune
challenge in a manner that suggests a more beneficial response relative to APOE3.

16
1. Introduction

Alzheimer’s disease (AD) is an age-related neurodegenerative disease and the leading
cause of dementia. As a multifactorial disease, a number of genetic and environmental factors
affect risk of AD. Among the many genes significantly associated with increased AD risk, a notably
high proportion are linked with innate immunity
5
. Indeed, studies across several disciplines have
provided strong evidence that innate immune pathways both in the central nervous system (CNS)
and systemically are important regulators of the initiation and progression of the disease
1–3
. The
role of innate immunity in AD is thought to involve, in large part, chronic inflammation resulting
from activation of innate immune cells, particularly brain microglia and peripheral
macrophages
155,156
. Several innate immune pro-inflammatory pathways have been implicated in
AD, with particularly compelling evidence for involvement of toll-like receptor 4 (TLR4) signaling
though its role is still unclear
157
.
The predominant genetic risk factor for AD is the e4 allele of apolipoprotein E (APOE).
Several studies have implicated APOE in the regulation of innate immunity and associated
inflammatory responses, which in turn have been postulated to contribute to APOE4-associated
AD risk
27
. For example, upon treatment with TLR4 agonist lipopolysaccharide (LPS), APOE4
typically has been linked with exaggerated pro-inflammatory response in the brain
33,34,147
and
higher levels of peripheral cytokines and organ injury
28,33
. Moreover, APOE4 is associated with
increased levels of pro-inflammatory cytokines and cell death in macrophages
29,158
and
microglia
32
. Conversely, there is recent evidence that APOE4 is associated with reduced innate
immune markers in certain populations
159
. Collectively, APOE is an important modulator of innate
immunity and inflammatory processes though its influence likely depends upon context.
In women, AD risk is increased by age-related depletion of estrogens
148,149
. Estrogens
exert a number of actions potentially relevant to protection from AD, including regulation of
inflammation
150
. For example, low circulating estrogen is associated with changes in the
peripheral immune response with increases in inflammatory markers
160
. Postmenopausal women
17
have increased immune markers including elevated levels of pro-inflammatory cytokines that can
be attenuated by estrogen-based hormone therapy
161
. Likewise, in animal models, pretreatment
with the estrogen 17b-estradiol (E2) strongly prevents neuroinflammation induced by LPS by
inhibiting microglial activation
162
. In primary glial and neuronal cultures, E2 treatment also reduced
the production of nitrite and cell death after LPS challenge
163,164
. Together, these findings suggest
that estrogen may act as a potential anti-inflammatory factor.
In addition to their independent effects, numerous interactions among APOE, female sex,
and estrogen status have been reported. APOE4 is associated with cognitive decline with a
stronger relationship in females
165–167
. Additionally, APOE4 is associated with altered gene
expression profile in the brain with sex differences
168
. Specifically, expression of genes related to
immune system response and inflammation is higher in female APOE4 carriers compared with
males
168
. However, the extent to which APOE4 and estrogen interact to regulate innate immunity
is still unclear. Interestingly, exogenous E2 treatment reduced the production of inflammatory
mediators induced by the combination of LPS and interferon g (IFNg) in cultures of microglia and
peritoneal macrophages from human APOE3 targeted replacement (TR) mice, but only a modest
reduction from APOE4-TR mice
169
, suggesting APOE-dependent modulation of E2 anti-
inflammatory actions. To study these relationships in animal models, we used human APOE TR
mice to compare the systemic and neural responses to acute LPS challenge in APOE3 versus
APOE4 female mice across low and high estrogen conditions.
Overall, we observed modest effects of estrogen status on LPS-induced responses and,
perhaps unexpectedly, a general trend for APOE4 mice to express lower cytokine expression
under both basal and challenge conditions compared with APOE3 mice, suggesting a protective
role of APOE4 allele in acute inflammatory response.

2. Methods

2.1 Animals
18
All female mice were homozygous for knock-in of human APOE3 or APOE4 and were
generously provided by Mary Jo LaDu (University of Illinois at Chicago). The mice were generated
from a breeding colony of EFAD (APOE
+/+
, 5xFAD
+/-
) mice
26
but were non-carriers of the 5xFAD
Alzheimer’s-related genes (APOE
+/+
, 5xFAD
-/-
). After receipt, mice were maintained at vivarium
facilities at the University of Southern California under 12h light/dark cycle with lights on at 0600h
with ad libitum access to rodent chow and water for the duration of the experiment. All experiments
were conducted under an institution-approved animal protocol (#20617) and in accordance with
the National Research Council’s Guide for the Care and Use of Laboratory Animals.

2.2 Surgery and treatments
At 18-22 weeks of age, APOE3 and APOE4 mice were randomized to treatment groups
in a manner that retained similar initial mean body weights across the three groups: sham-
ovariectomized (Sham), ovariectomized (OVX), and ovariectomized with 17b-estradiol (E2)
treatment (OVX+E2). Under 2-3% isoflurane anesthesia, mice were bilaterally ovariectomized or
sham-ovariectomized and implanted subcutaneously with a Silastic capsule (1.47 mm ID 1.96
mm OD; Dow Corning, Midland, MI). Each capsule had a total length of 7 mm with the inner 3
mm packed with cholesterol (vehicle) or 0.5% (w/w) E2. The E2 dosage was empirically
determined in a separate dose-response study to identify the optimal E2 capsule length needed
to restore uterine weight in OVX female mice (data not shown). Uterine weight is recognized as
an excellent bioassay of endogenous E2 levels
170,171
. Sham and OVX groups received capsules
with only cholesterol, whereas OVX+E2 mice received capsules containing E2 and cholesterol.
At the time of OVX surgery, all mice were implanted subcutaneously with IPTT 300 transponders
(Bio Medic Data Systems, Seaford, DE) to monitor body temperature. Mice were housed
individually following surgery and for the remainder of the experiment. Four weeks following
surgery, mice were weighed and administered a single intraperitoneal (i.p.) injection of sterile
19
saline alone or saline containing 500 µg/kg lipopolysaccharide (LPS; Escherichia coli O55:B5,
Sigma #L4524). All injections were conducted between the hours of 0800 - 1000.

2.3 Behavior
Sickness behavior was monitored 30 min, 1h, 2h, and 4h following saline or LPS injection
in the home cage using a scale consistent with prior reports
172–174
. At each time point, each mouse
was evaluated on the following sickness behavior symptoms on a 3-point (0 - 2) scale of severity:
ptosis (drooping eyelids), lethargy (lack of movement), and huddling (kyphosis; curled body
posture). Then, each cage was removed and the mouse was evaluated on escape from finger
stroke (rapid, normal response = “0”, slow response = “1”, or no response = “2”). Scores were
summed within animals to yield composite sickness behavior.

2.4 Body temperature
Body temperatures were measured 30 min, 1h, 2h, and 4h following saline or LPS injection
using a compatible wireless reader (Bio Medic Data Systems, Seaford, DE) as instructed by the
manufacturer.

2.5 Tissue Collection
Four hours following LPS or saline injection, mice were euthanized. Blood from each
mouse was collected into EDTA-coated tubes following cardiac puncture. Collected blood was
centrifuged to separate plasma, which was aliquoted and stored at -80 ̊ C. The brain was rapidly
removed and hemisected along the sagittal plane. Cortex from one hemibrain was dissected and
immediately frozen at -80 ̊ C for RNA extractions. Liver was also harvested and stored at -80 ̊ C.
Uterus and gonadal and retroperitoneal fat pads were dissected and weighed.

20
2.6 RNA isolation and real-time PCR
For RNA extractions, cortex and liver were homogenized using TRIzol reagent (Invitrogen)
following the manufacturer’s protocol. Purified RNA (1 μg) was used for reverse transcription
using the iScript cDNA synthesis system (Bio-Rad), and the resulting cDNA was used for real-
time quantitative PCR. Real-time quantitative PCR was performed in duplicates using
SsoAdvanced Universal SYBR Green Supermix (BioRad) and the Bio-Rad CFX Connect
Thermocycler. Relative quantification of mRNA was determined using the DD-CT method after
normalizing each sample with its corresponding b-actin expression levels. The following primer
pairs were used in both cortex and liver: tumor necrosis factor alpha (TNFα), forward: 5′-
CCCTCACACTCAGATCATCTTCT-3′, reverse: 5′-GCTACGACGTGGGCTACAG-3′; interleukin-
6 (IL-6), forward: 5′-CTCTGGGAAATCGTGGAAAT-3′, reverse: 5′-CCAGTTTGGTAGCAT
CCATC-3′; interleukin-1β (IL-1β), forward: 5′-GCAACTGTTCCTGAACTCAACT-3′, reverse: 5′-
ATCTTTTGGGGTCCGTCAACT-3′; β-actin, forward: 5′-AGCCATGTACGTAGCCATCC-3′,
reverse: 5′-CTCTCAGCTGTGGTGGTGAA-3′; interferon gamma (IFN-γ), forward: 5′-
ACAGCAAGGCGAAAAAGGATG-3′, reverse: 5′-TGGTGGACCACTCGGATGA-3′; interleukin-12
subunit beta (IL-12b), forward: 5′- TGGTTTGCCATCGTTTTGCTG-3′, reverse: 5′-ACA
GGTGAGGTTCACTGTTTCT-3′.

2.7 Multiplex analysis of plasma cytokines
Plasma cytokine levels were measured using Meso Scale Discovery (MSD) V-Plex
Proinflammatory Panel 1 (mouse) cytokine assay according to the manufacturer protocol. MSD
plates were analyzed on a QuickPlex SQ 120 instrument (MSD). All standards and samples were
measured in duplicate.

21
2.8 Statistical analysis
Body temperature, sickness, and cytokine data were analyzed separately for LPS- and
saline-treated groups. Three-way repeated ANOVA was performed on body temperature and
sickness behavior with repeated measures across time points and genotype and estrogen status
as a between-subjects factors. All other data were analyzed with 2 ´ 3 ANOVAs with genotype
and estrogen status as between-subjects factors. Significant effects were further analyzed with
planned comparisons between groups of interest using the Bonferroni correction. Body
temperature and sickness score were analyzed using SPSS (Version 28). All other data were
analyzed using GraphPad Prism (Version 9). All datasets are expressed as mean ± SEM. The
threshold for statistical significance was set at p < 0.05.

3. Results

3.1 Effects of APOE and estrogen status on body and tissue weights
The efficacy of OVX surgery and E2 treatment on systemic estrogenic function was
assessed by analysis of uterine weights. There was a statistically significant main effect of OVX
group (F(2, 86) = 111.5, p < 0.001) and an interaction between genotype and estrogen status (F(2,
86) = 3.7, p = 0.028) on uterine weights. In both APOE3 and APOE4 female mice, OVX was
associated with significant reduction in uterine weight whereas E2 treatment increased uterine
weight (Fig. 1A). The effects of E2 were more robust in APOE4 mice with uterine weight
significantly higher in the OVX+E2 group relative to the Sham group only in APOE4 females.
Both APOE genotype and estrogen status can affect body weight. Although there were no
significant differences in body weight across groups initially (data not shown), by four weeks
following surgery body weight differed significantly by estrogen status in a manner that was
modulated by APOE genotype (Fig. 1B). Two-way ANOVA for body weight showed significant
main effects of genotype (F(1, 82) = 4.0, p = 0.048) and estrogen status (F(2, 82) = 12.5, p < 0.001)
22
as well as an interaction between the two factors (F(2, 82) = 3.1, p = 0.049). More specifically, OVX
was associated with increased body weight that was prevented by E2 treatment, effects that were
more pronounced in APOE4 mice. Between group comparisons showed statistically significant
differences in Sham vs OVX (p < 0.001) and OVX vs OVX+E2 (p = 0.002) comparisons only in
APOE4 mice (Fig. 1B).
Both APOE genotype and estrogen status also affected weights of adipose depots.
Gonadal fat pad weights significantly differed by estrogen status (F(2, 86) = 19.4, p < 0.001) and
APOE genotype (F(1, 86) = 9.7, p = 0.003) with higher weights associated with OVX and APOE4
(Fig. 1 C). In between group comparisons, the OVX group had significantly increased gonadal fat
relative to Sham only in APOE4 mice (p = 0.001), whereas OVX+E2 exhibited significantly lower
gonadal fat weight than OVX in both APOE3 (p = 0.031) and APOE4 (p < 0.001) mice (Fig. 1C).
For retroperitoneal fat, there were less robust but still significant main effects of estrogen status
(F(2, 86) = 3.3, p = 0.041) and APOE genotype (F(1, 86) = 16.4, p < 0.001) (Fig. 1D).

23

Figure 1. Effects of APOE genotype and estrogen status on body and tissue weights.
A) Uterine weights of APOE3 (white bars) and APOE4 (grey bars) in female mice 4 weeks
following OVX surgery and hormone treatment. B) Body weight gain over 4 weeks following
surgery. C) Weight of the gonadal fat pads and D) retroperitoneal fat pads. Data show mean
(+SEM) values (N = 13-15/group). Statistically significant main effects are denoted by A (APOE
genotype), E (estrogen status) and I (genotype and estrogen status interaction). * denotes p<0.05
relative to APOE3 mice on same estrogen condition; # denotes p<0.05 relative to Sham condition
in matched genotype; $ denotes p <0.05 relative to OVX condition in matched genotype.

Sham OVX OVX+E2
0.00
0.05
0.10
0.15
0.20
Uterine weight (g)
APOE3
APOE4
# #
$ #
$
E I
Sham OVX OVX+E2
0.0
0.2
0.4
0.6
0.8
Gonadal fat (g)
A E
#
*
$
$
Sham OVX OVX+E2
0
1
2
3
4
Body weight gain (g)
$
#
A E I
Sham OVX OVX+E2
0.00
0.05
0.10
0.15
0.20
0.25
RP fat (g)
A E
A B
C D
24
3.2 Effects of APOE and estrogen status on sickness behavior
The immune challenge induced by LPS treatment is known to result in sickness
behavior
175–177
. As expected, LPS was associated with increasing sickness behavior across the 4
h time period, which differed among groups (Fig. 2). Multivariate tests on sickness score across
time for LPS-treated mice showed a significant main effect of time (F(3, 38) = 55.5, p < .001), and
an interaction between time and estrogen status (F(6, 76) = 2.4, p = 0.037). Between-subject tests
showed a significant main effect of genotype (F(1, 40) = 6.9, p = 0.012), and an interaction between
genotype and estrogen status (F(2, 40) = 3.6, p = 0.036). Pairwise comparison revealed that under
OVX condition, APOE4 mice exhibited more sickness behavior than APOE3 mice at 1h (p =
0.041), 2h (p < 0.001) and 4h (p = 0.018) time points. Estrogen treatment slowed the early
development of sickness behavior in APOE4 mice, as APOE4 OVX+E2 group exhibited less
sickness behavior at the 1h time point than their genotype matched APOE4 sham (p = 0.023) and
APOE4 OVX groups (p = 0.003).
For saline control groups, there was a significant effect of time (F(3, 36) = 9.2, p < 0.001),
and an interaction between time and genotype (F(3, 36) = 5.0 , p = 0.005). All groups of animals
showed a modest increase in sickness behavior score that was likely due to low locomotor activity
during the light phase when the experiments were conducted. No group difference was found at
any time point in post hoc pairwise comparisons.

25

Figure 2. Effects of APOE genotype and estrogen status on sickness behavior.
Sickness scores were measured 30 min, 1h, 2h, and 4h following peripheral injection of saline (A-
C) or LPS (D-F) in APOE3 (white circles) and APOE4 (black squares) female mice under Sham
(A,D), OVX (B,E), and OVX+E2 (C,F) estrogen status groups. Data show mean (+SEM) values
(N = 7-8/group). Dotted lines in each graph indicate averaged sickness score over 4h-treatment
period with saline across all genotype and estrogen status groups. * denotes p<0.05 relative to
APOE3 mice on same estrogen condition; # denotes p<0.05 relative to Sham condition in
matched genotype; $ denotes p <0.05 relative to OVX condition in matched genotype.

.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
Sham
APOE3
APOE4
.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
OVX
.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
*
*
*
.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
OVX+E2
.5h 1h 2h 4h
0
2
4
6
8
Post-treatment (h)
Sickness Score
# $
Saline
LPS
A B C
D E F
26
3.3 Effects of APOE and estrogen status on body temperature
Since acute systemic inflammation is typically accompanied by thermoregulatory changes
in body temperature, we assessed the effects of APOE genotype and estrogen status on the
thermoregulatory response to LPS. Multivariate tests yielded a significant main effect of time (F(4,
37) = 38.9, p < 0.001), and an interaction between time and genotype (F(4, 37) = 7.5, p < 0.001).
APOE4 mice showed greater LPS-induced hypothermia than APOE3 mice under Sham and OVX
conditions specifically at the 2h (p = 0.001) and 4h (p < 0.001) time points. Additionally, there
were significant main effects of genotype (F(1, 40) = 23.2, p < 0.001) and estrogen status (F(2, 40) =
4.5, p = 0.017) as well as an interaction between genotype and estrogen status (F(2, 40) = 4.1, p =
0.024). Pairwise comparison revealed that APOE3 OVX mice did not develop hypothermia over
the 4h treatment period and had significantly higher body temperature than APOE3 OVX+E2 mice
at 2h (p = 0.004) and 4h (p = 0.029) time points.
Across all genotype and estrogen status groups, saline controls showed a slight decrease
in temperature at 4h following the injection (36.08 ± .08°C) compared with 0h time point (36.63
± .11°C). Confirming this pattern of temperature changes, multivariate tests showed a significant
effect of main effect of time (F(4, 34) = 4.49, p = 0.005). Estrogen treatment also was associated
with a modest but significantly lower body temperature in APOE4 mice only at the 2h time point
compared with Sham (p = 0.046) and OVX (p = 0.026) mice. Post-hoc pairwise comparisons did
not show any other genotype or estrogen effects among any groups at any time point.

27

Figure 3. Effects of APOE genotype and estrogen status on thermoregulatory response to
LPS.
Body temperatures were measured 0h, 30 min, 1h, 2h, and 4h following peripheral injection of
saline (A-C) or LPS (D-F) in APOE3 (white circles) and APOE4 (black squares) female mice under
Sham (A,D), OVX (B,E), and OVX+E2 (C,F) estrogen status groups. Data show mean (+SEM)
values (N = 7-8/group). Dotted lines in each graph indicate average body temperature at 0h
(36.63°C) across all genotype and estrogen status groups in saline-treated mice. * denotes
p<0.05 relative to APOE3 mice on same estrogen condition; $ denotes p <0.05 relative to OVX
condition in matched genotype.

0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
Sham
APOE3
APOE4
0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
*
*
0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
OVX
0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
*
*
0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
OVX+E2
# $
0h .5h 1h 2h 4h
30
32
34
36
38
Post-treatment (h)
Temperature (°C)
$
# $
Saline
LPS
A B C
D E F
28
3.4 Effects of APOE and estrogen status on plasma cytokine levels
Levels of circulating cytokines in plasma were measured across groups under basal
(saline) and LPS conditions. In the saline groups, there was a significant main effect of APOE
genotype in which APOE4 was associated with significantly lower levels of IL-2 (F(1, 34) = 6.17, p
= 0.018), IL-12p70 (F(1, 34) = 7.32, p = 0.011) and IFN-γ (F(1, 34) = 6.98, p = 0.012) relative to APOE3.
There was a nonsignificant trend of genotype on IL-4 (p = 0.070) with lower levels in APOE4.
Additionally, there was a significant main effect of estrogen status on levels of IL-2 (F(1, 34) = 4.91,
p = 0.013) that reflected significantly lower IL-2 in OVX+E2 mice vs Sham groups (p = 0.021)
specifically within APOE3 mice (Fig. 4A). There was no significant interaction between APOE and
estrogen status and no significant differences across groups in the levels of TNFα, IL-1β, IL-5,
and IL-10.
As expected, LPS treatment yielded greatly increased plasma levels of cytokines. There
were significant main effects for APOE genotype on plasma cytokine levels for seven of the eight
cytokines: TNFα, IL-1β, IL2, IL-4, IL-5, IL-12p70, and IFN-γ. For each of the affected cytokines,
the same relationship was observed with lower levels in APOE4 mice (Fig. 4B). Between group
comparisons of APOE3 vs APOE4 mice matched for estrogen status showed only significant
effects in which APOE4 Sham mice had lower IL-5 (p = 0.010) and IL-12p70 (p = 0.021) levels
than APOE3 Sham mice. Levels of all cytokines were comparable under the different estrogen
status groups.

29

Figure 4. Effects of APOE genotype and estrogen status on cytokine levels in plasma.
Plasma cytokine levels were measured 4h following peripheral injection of saline (A) or LPS (B)
in APOE3 and APOE4 female mice under Sham (white bars), OVX (light grey bars), and OVX+E2
(dark grey bars) estrogen status groups. Data are presented as mean (±SEM) values (N = 6-
7/group for saline treatment; N = 7-8/group for LPS treatment). Statistically significant main effects
are indicated by A (APOE genotype) and E (estrogen status). * denotes p<0.05 relative to APOE3
mice on same estrogen condition; # denotes p<0.05 relative to same genotype on Sham condition.

APOE3 APOE4
0
10
20
30
40
TNFα (pg/ml)
APOE3 APOE4
0
1
2
3
4
IL-1β (pg/ml)
APOE3 APOE4
0
5
10
15
20
25
IL-5 (pg/ml)
APOE3 APOE4
0
10
20
30
IL-12p70 (pg/ml)
A
APOE3 APOE4
0.0
0.5
1.0
1.5
IL-4 (pg/ml)
Genotype p=0.070
APOE3 APOE4
0
10
20
30
40
IL-10 (pg/ml)
APOE3 APOE4
0
2
4
6
IL-2 (pg/ml)
A E
#
Sham
OVX
OVX+E2
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
2.5
IFNγ (pg/ml)
A
A
APOE3 APOE4
0
1000
2000
3000
4000
5000
TNFα (pg/ml)
A
APOE3 APOE4
0
50
100
150
200
250
IL-1β (pg/ml)
A
APOE3 APOE4
0
50
100
150
IL-5 (pg/ml)
A
*
APOE3 APOE4
0
500
1000
1500
2000
IL-12p70 (pg/ml)
A
*
APOE3 APOE4
0
5
10
15
20
IL-4 (pg/ml)
A
APOE3 APOE4
0
1000
2000
3000
IL-10 (pg/ml)
APOE3 APOE4
0
50
100
150
200
250
IL-2 (pg/ml)
A
Sham
OVX
OVX+E2
APOE3 APOE4
0
50
100
150
200
IFNγ (pg/ml)
A
B
30
3.5 Effects of APOE and estrogen status on liver cytokine
To further investigate the effects of APOE genotype and estrogen status on the acute
inflammatory response, we next examined gene expression levels of select pro-inflammatory
cytokines in the liver of LPS-treated and saline control female mice under sham, OVX, and
OVX+E2 conditions. LPS treatment induced, on average, over 10-fold increases in hepatic TNFα,
IL-1β, IL-6, IL-12b, and IFN-γ mRNA expression levels compared to saline treated mice (Fig. 5).
There was a significant main effect of APOE for TNFα ( F(1, 41)= 10.0, p = 0.003) and IL-12b ( F(1,
41)= 7.4, p = .001) expression in liver following LPS in which levels were lower in APOE4 mice.
Estrogen status significantly affected expression of two cytokines. First, the LPS-induced increase
in liver IL-6 expression differed significantly by estrogen status (F(2, 40) = 5.4, p = 0.008). Post hoc
comparisons revealed that OVX group showed significantly lower IL-6 expression than the
OVX+E2 group (p = 0.013) in APOE4 but not APOE3 mice (Fig. 5G). Second, there was a
significant main effect of estrogen status on the expression IFN-γ (F(2, 40) = 3.6, p = 0.037),
however this relationship did not show statistically significant between group differences in post
hoc tests (Fig. 5H). No significant main effects or interactions of APOE genotype or estrogen
status were found in saline-treated groups.

31

Figure 5. Effects of APOE genotype and estrogen status on cytokine expression in liver.
Hepatic mRNA levels of TNFα, IL-1β, IL-6, IFN-γ and IL-12b were compared in APOE3 and
APOE4 female mice under Sham (white bars), OVX (light grey bars), and OVX+E2 (dark grey
bars) estrogen status groups that were treated with saline (A-E) or LPS (F-J). Data show mean
(±SEM) expression levels of qPCR results normalized to the APOE3 Sham saline group for each
cytokine (N = 6-8/group for saline; N = 7-8/group for LPS). Statistically significant main effects are
denoted by A (APOE genotype) and E (estrogen status). $ denotes p <0.05 relative to OVX
condition in matched genotype.

APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
TNFα expression
sham
OVX
OVX+E2
APOE3 APOE4
0
10
20
30
40
TNFα expression
A
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
2.5
IL-1β expression
APOE3 APOE4
0
10
20
30
40
IL-1β expression
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
2.5
IL-6 expression
APOE3 APOE4
0
50
100
150
200
IL-6 expression
E
$
APOE3 APOE4
0.0
0.5
1.0
1.5
IFNγ expression
APOE3 APOE4
0
10
20
30
40
50
IFNγ expression
E
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
IL-12b expression
APOE3 APOE4
0
20
40
60
80
IL-12b expression
A
Saline
LPS
A B C D E
E F G H I
32
3.6 Effects of APOE and estrogen status on brain cytokine expression
Next, we assessed cytokine expression levels in cerebral cortex following peripheral LPS
or saline injections across APOE genotype and estrogen status (Fig. 6). LPS induced robust
increases in cortical mRNA levels of TNFα, IL-1β, IL-6, IL-12b, and IFN-γ. There were no
significant main effects or interactions of APOE genotype and estrogen status on the probed pro-
inflammatory genes in the cortex of LPS-treated mice. However, in saline controls, there was a
significant main effect of APOE on IL-6 expression (F(1, 34) = 18.1, p < 0.001) with ~50% lower
levels in APOE4 mice (Fig. 6C). Between group comparisons of APOE3 vs APOE4 mice from
matched estrogen status groups yielded statistically nonsignificant trends in Sham (p = 0.063),
OVX (p = 0.050), and OVX+E2 (p = 0.083) groups on IL-6 expression. There was no significant
main effect of estrogen status on cerebrocortical expression of any of the five cytokines, however
there were nonsignificant trends for IFN-γ (p = 0.082) and IL-12b (p = 0.074) that were driven by
relatively higher cytokine levels in OVX+E2 groups specifically within APOE3 mice (Fig. 6D, 6E).

33

Figure 6. Effects of APOE genotype and estrogen status on cytokine expression in brain.
Cerebrocortical mRNA levels of TNFα, IL-1β, IL-6, IFN-γ and IL-12b were compared in APOE3
and APOE4 female mice under Sham (white bars), OVX (light grey bars), and OVX+E2 (dark grey
bars) estrogen status groups that were treated with saline (A-E) or LPS (F-J). Data show mean
(±SEM) expression levels of qPCR results normalized to the APOE3 Sham saline group for each
cytokine (N = 6-8/group for saline treatment; N = 7-8/group for LPS treatment). Statistically
significant main effects are denoted by A (APOE genotype) and E (estrogen status).

APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
TNFa expression
sham
OVX
OVX+E2
APOE3 APOE4
0
20
40
60
80
100
TNFα expression
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
IL-1β expression
APOE3 APOE4
0
100
200
300
IL-1β expression
APOE3 APOE4
0.0
0.5
1.0
1.5
IL-6 expression
A
APOE3 APOE4
0
50
100
150
IL-6 expression
APOE3 APOE4
0
1
2
3
IFNγ expression
Estrogen p=0.082
APOE3 APOE4
0
5
10
15
IFNγ expression
APOE3 APOE4
0.0
0.5
1.0
1.5
2.0
2.5
IL-12b expression
Estrogen p=0.074
APOE3 APOE4
0
50
100
150
IL-12b expression
Saline
LPS
A B C D E
E F G H I
34
4. Discussion

The goals of this study are to investigate the effects of APOE genotype in modulating
innate immune responses in females and to determine if this relationship is affected by estradiol
levels. In comparing systemic and neural measures under both basal conditions and following
acute LPS challenge, we found that APOE4 is associated with more pronounced changes in
sickness behavior and body temperature in a manner that differs across estrogen status.
Interestingly, APOE4 showed stronger responses to changes in estrogen status for measures of
uterine weight, body mass, and adiposity. Importantly, although cytokine levels were often similar
across APOE3 vs APOE4 groups matched for experimental treatment, when differences were
observed, APOE4 was generally associated with lower levels of cytokines. Our findings support
the position that APOE is a modulator of immune responses, a relationship that is significantly
modulated by estrogen status on a subset of measures.
Acute LPS treatment induces a range of responses throughout the body in response to
pathogen challenge. One change is the appearance of sickness behaviors, including anorexia,
lethargy, sleepiness, and depression-like behaviors
175–177
. Sickness behavior has been suggested
to be a protective strategy. For example, by reducing locomotor activity and minimizing energy
expenditure, the host may better tolerate the large energy deficits caused by pathogen infection
178
.
The mechanism of LPS-induced sickness behavior involves alterations in neural activity that are
likely modulated by inflammation
179
. In agreement with this idea, peripheral or central
administration of the pro-inflammatory cytokines TNFα or IL-1β has been shown to induce
sickness behavior
180,181
. Further, agouti-related protein (AgRP) neurons in the hypothalamus
regulate appetite by increasing food intake
182
. It has been reported that peripheral administration
of LPS decreases the secretion of AgRP via IL-1β signaling in the hypothalamus
183
, linking LPS-
induced inflammatory signaling with decreased food-seeking, a sickness-related behavior. We
observed that APOE4 was associated with greater LPS-induced sickness behavior, an effect that
was increased in the context of low estrogen. As sickness behavior is generally beneficial, these
35
data suggest that APOE4 was associated with a more favorable behavioral response to LPS than
APOE3.
Acute inflammation induced by LPS is also associated with significant changes in body
temperature that can manifest as either fever or hypothermia. These thermoregulatory responses
depend in part upon environmental temperature and LPS dose
184–186
. For rodents, the
thermoneutral temperature is approximately 30°C
187,188
. In typical vivarium housing environments
(sub-thermoneutral), mice develop fever in response to a low dose of LPS, whereas higher LPS
doses (>100µg/kg) result in a hypothermic response
186
. In our study, exposure to 500 µg/kg LPS
generally yielded reductions in body temperature. The one exception was APOE3 mice in the low
estrogen, OVX condition. APOE4 mice showed greater hypothermia than APOE3 mice under
both Sham and OVX conditions. The hypothermia observed with LPS exposure may represent a
beneficial response to immune challenge. Prior work demonstrated that when rats were placed
on a thermogradient apparatus and allowed to choose their preferred body temperature, LPS-
induced hyperthermia was accompanied by cold-seeking behavior
184
. Rats that developed
hypothermia “naturally” in a sub-neutral environment had higher survival rates compared with rats
unable to regulate their body temperature in a warm environment and experimentally “forced” to
develop fever
189,190
. Moreover, hypothermia instead of “forced” fever induced by E. coli or LPS is
associated with lower levels of lung inflammation and organ dysfunction
190
. Reinforcing the idea
that hypothermia is a favorable response to innate immune challenge, a recent study suggests
that hypothermia caused by LPS likely involves an energetic trade-off between immune
responses and the ability to maintain body temperature, which results in higher resistance and
tolerance towards pathogen infection
191
. Following this logic, the higher level of hypothermia
observed in LPS-treated APOE4 mice may suggest a more protective response to LPS relative
to APOE3, a possibility consistent with the sickness behavior findings and the observed pattern
of generally lower cytokines in APOE4 mice.
36
An important consequence of LPS-induced innate immune challenge is a robust increase
in cytokines and other inflammation-related factors in both neural and peripheral tissues.
Peripherally, LPS triggers inflammation in macrophages and monocytes through binding to the
TLR4/myeloid differentiation factor 2 complex, which activates NF-κB and IRF3 to induce cytokine
expression
192–194
. When examining the effects of APOE on peripheral innate immune response,
we found that APOE4 is associated with lower levels of IL-2, IL12p70 and IFN-γ in plasma under
basal conditions and lower levels of all cytokines except for IL-10 after LPS treatment. These
findings are largely consistent with those reported by Marattoli et al (2017)
195
in which male
APOE4 mice in the presence (EFAD mice) and absence (EFAD non-carrier mice) of AD-related
transgenes exhibited lower basal plasma levels of several cytokines including IL-1 β, IL12p70 and
IFN-γ. Under chronic LPS challenge, there were varied effects of APOE genotype on cytokine
levels with clusters that were unaffected by genotype, higher in APOE3, or higher in APOE4
195
.
In liver, we found no genotype effects on basal levels of expression for five cytokines but, after
LPS challenge, APOE4 mice had significantly lower expression of TNFα and IL-12b. Studies have
shown that TNFα contributes to LPS-induced hepatitis and hepatocyte apoptosis
196,197
.
Interestingly, treating rats with a neutralizing form of the soluble TNF type 1 receptor along with
LPS significantly decreased TNFα levels peripherally and yielded a faster recovery from LPS-
induced hypothermia, suggesting TNFα is disadvantageous in response to LPS challenge
198
.
Therefore, our data are consistent with the theme that APOE4 is associated with relatively
protective outcomes in acute inflammatory response.
We also examined the effects of APOE genotype on basal and LPS-induced cytokine
expression in female brain. Previous work demonstrated that LPS does not cross the brain-blood
barrier (BBB)
199,200
; however a recent study observed LPS infiltration into the brain, likely via a
lipoprotein-mediated mechanism
201
. This recent development opens the possibility that
peripherally administrated LPS could directly activate cells in the CNS including glial cells,
neurons, and endothelial cells. On the other hand, peripheral inflammatory mediators induced by
37
LPS may cross BBB
202
or translocate into brain through indirect mechanisms such as the vagus
nerve
203
or BBB disruption
204
to induce glial activation and production of pro-inflammatory
cytokines. As expected, our study showed that LPS induced robust increases in brain mRNA
levels of TNFα, IL-1b, IL-6, IFN-γ, and IL-12b. Perhaps unexpectedly, there were no statistically
significant differences in expression levels of these factors in APOE3 vs APOE4 female brains.
In the CNS, prior work suggests that LPS-induced responses may be cell type dependent. For
example, in cultured microglia APOE4 was associated with higher basal levels of inflammatory
markers
30,31
and higher secretion of TNFα, IL-6, and IL-1β after LPS exposure
31,32
. Unlike
microglia, primary cultures of astrocytes from APOE4 mice had lower basal levels of IL-6 and
reduced levels of LPS-induced cytokines including TNFα, IL-6, and IL-1β
205
. As our findings
include all brain cells, the contributions of and potential differences among glial and other neural
cell types is uncertain.
In apparent contrast to our findings, some prior studies showed that APOE4 is associated
with increased levels of inflammation after LPS challenge in humans, rodents, and in vitro
models
28,33–35,147
. One possible reason for the discordant observations might relate to LPS
treatment protocols. Indeed, inflammatory response induced by LPS is dynamic and depends on
the dose
206,207
. Low doses of LPS selectively induce a pro-inflammatory response while higher
doses result in activation of both pro- and anti-inflammatory mediators
208,209
. A limitation of our
study is that a single, moderate dose of LPS was used and only a single time point was assessed.
Nonetheless, it is noteworthy that the overall trend of dampened outcomes in APOE4 mice was
also observed under basal conditions. A second key issue is sex. Our studies utilized females,
whereas most prior studies that investigated LPS treatment across APOE genotype examined
only male mice
33–35,147
. Sex significantly impacts many effects of APOE genotype
210,211
and
multiple aspects of immune responses
212,213
. LPS-induced inflammatory responses are modulated
by sex, though the literature is mixed with abundant evidence that females show lower levels of
inflammation
214–216
though other studies suggest the opposite
217,218
. Moreover, APOE may interact
38
with sex in modulating LPS-induced inflammation. Peritoneal macrophages from male APOE4
mice showed higher levels of nitric oxide compared to adult male APOE3 mice after treatment
with LPS combined with IFN-γ, but this genotype difference was not observed in female mice
219
.
Thus, while our findings present new insights into the interactions between innate immunity and
APOE genotype in females, there remains an incomplete understanding of these complex
relationships.
Given the established links between sex and sex hormones with both APOE genotype
and immune responses, a significant modulatory impact of estrogen status on several outcomes
was expected. The strongest effects of estrogen status were predictably on uterine weight, body
weight, and adiposity, measures long-established to be regulated by estradiol
220,221
. Interestingly,
the effects of relatively low vs high estrogens on body weight and fat mass were significantly
stronger in APOE4 females, suggesting an augmented responsiveness to some estrogen-
mediated actions. However, we found only modest effects of estrogen status on sickness behavior,
thermoregulation, and general trends in cytokine levels. LPS-induced hypothermia was not
observed in APOE3 mice under the low estrogen OVX condition. This observation is consistent
with the finding that LPS-induced hypothermia requires the presence of estradiol
222
. However, the
exaggerated hypothermic response observed in APOE4 mice persisted across all estrogen
groups, implying an independence in APOE4 females of the estrogen modulation of
thermoregulation. Estrogen is a regulator of inflammation, a role that can interact with APOE and
is associated with vulnerability to age-related cognitive decline and risk of AD
150
. For example,
17β-estradiol treatment in macrophages derived from OVX APOE mice strongly reduced nitrite
and TNFa levels induced by LPS combined with IFN-γ, an effect that was stronger in APOE3 than
APOE4 mice
169
. In our study, estrogen status generally showed only small effects on cytokine
levels. The clearest result of an estrogen status effect on cytokines that differed by APOE
genotype was in liver, in which estrogen status significantly affected LPS-induced IL-6 expression
in APOE4 but not APOE3 females. Although the significance of estrogen status to immune
39
function has been established in prior literature, the observation of generally small effects in the
present study suggests more limited roles of estrogens in acute inflammatory responses.

5. Conclusions

In summary, our data show a generally protective effect of APOE4 from acute innate
immune challenge in comparison to APOE3. This finding is consistent with an increasingly
appreciated role of APOE4 exerting protective roles against infection. APOE4 has been
hypothesized to be evolutionarily adaptive by promoting immune responses to pathogens
159,223
.
In humans, the APOE4 allele is associated with lower levels of blood immune markers, especially
in populations under high-infectious environment
159,224–226
. Further, APOE4 is associated with
reduced vulnerability to infection by hepatitis C virus
227
as well as less severe consequences to
hepatitis infection, including liver fibrosis
228
. The extent to which these relationships extend to the
brain is less clear. Interestingly, APOE4 is associated with higher cognitive performance in
populations with high levels of infection or parasitic burden
229–231
. However, APOE4 is also
strongly associated with increased vulnerability to age-related cognitive decline and risk for AD
232–
234
, relationships that are stronger in female APOE4 carriers
165–167
. Our findings are consistent
with the idea that APOE4 may be an example of antagonistic pleiotropy. In this case, APOE4
modulation of innate immune pathways, presumably including macrophage and microglial
activation, can confer relative protection against infection that may yield cognitive advantage
particularly in early life. However, regulation of these same pathways may contribute to its strong
association with risk for late life cognitive decline and AD
1,2
. These relationships are expected to
be affected by sex as both immunity
212,213
and APOE effects
210,211
show sexual dimorphisms and
regulation by sex hormones.

40
Chapter 3: Role of microglial TLR4 in obesity-induced neural
impairments

Abstract
Obesity is associated with adverse effects on the brain and neural function, including
reduced hippocampal volume, acceleration of brain aging, and increased risks for developing
AD and vascular dementia. Obesity is also characterized by chronic and low-grade inflammation
that affects numerous tissues. In the brain, obesity drives chronic neuroinflammation in part by
increasing microglial activation and production of pro-inflammatory cytokines. Among all pro-
inflammatory signaling pathways, Toll-like receptor 4 (TLR4) signaling pathways may be
particularly important in driving obesity-associated inflammation. TLR4 signaling is implicated as
an essential regulator of systemic effects of obesity in metabolic measures and inflammation.
Genetic knockout and pharmacologic inhibition of TLR4 can attenuate glucose dysregulation,
insulin resistance and inflammation in peripheral tissues. TLR4 signaling in microglia has been
suggested to mediate the progression of AD and several other neural disorders but its role in
obesity-induced neural outcomes in is largely unknown. Therefore, we generated and analyzed
mice that carry a microglial-specific deletion of TLR4 following 16 weeks of exposure to either a
control or high fat diet. In male mice, absence of microglial TLR4 had no significant effects on
body weight, adiposity, and metabolic measures. However, we observed significant reduction in
adipose inflammation in TLR4 knockout mice fed on high fat diet. In females, lacking microglial
TLR4 was associated with significant reduction in body weight and improvement of metabolic
function. Furthermore, obese mice with microglial TLR4 deletion exhibited significant protection
against diet-induced cognitive impairments and glial activation both in males and females.
These findings implicate an important role for microglial inflammation in obesity-induced neural
impairments. Additionally, these findings point to the potential for clinical translation of TLR4 as
a therapeutic target for obesity-induced neural injury.
41
1. Introduction
Obesity is a global health concern with increasing prevalence in countries at all
economic levels. Besides manifested as excessive accumulation of body fat, obesity is
significantly associated with a wide range of comorbidities, including metabolic syndrome
37
,
diabetes
38
, cardiovascular disease
39
, and certain cancers
40
. In the brain, obesity is associated
with deleterious neural dysfunction, such as decreased hippocampal volume
41
, impaired
cognition and increased risk of dementia
45–47
. Growing evidence suggested that obesity is also a
risk factor for accelerated tissue aging due to their great similarities in molecular pathway
alterations during pathological process
235,236
.
Obesity is characterized by a chronic state of low-grade inflammation
237
. Toll-like
receptor 4 (TLR4) signaling is crucial in mediating inflammatory events
124,125
. With regards to
metabolic disorder, insulin resistance and obesity were associated with increased TLR4 levels
in many tissues in humans, including peripheral blood mononuclear cells
152,153
, adipose
tissue
153,238
, and muscle
239
. Importantly, a recent study showed that TLR4 was required for
saturated fatty acids (SFA)-induced inflammation through cellular metabolism reprogramming
128
.
The implication of TLR4 in obesity-related outcomes was further accessed by studies showing
that TLR4 loss-of-functional mutation, knockdown or its pharmacological inhibition, protected
mice from diet-induced inflammation and insulin resistance
130,133–135,138,240
. A role of TLR4 in the
central effects of high fat diet (HFD) has also been investigated. Brain specific deletion of
myeloid differentiation factor (MyD88), an adaptor of TLR4 signaling, prevented HFD-induced
obesity and leptin and insulin resistance
143
. In addition, mice with TLR4 mutation were protected
from HFD-induced cerebral vascular change and cognitive impairment
139
. Previously, we
demonstrated that pharmacological inhibition of TLR4 signaling resulted significant protection in
HFD-induced microglial activation and neurogenesis
137
. TLR4-dependent inflammation has also
been suggested to mediate the progression of Alzheimer’s disease and several other neural
disorders
157,241
.
42
Neuroinflammation is widely considered as a key mediator of obesity-induced neural
dysfunction
242–244
. Mechanism wise, obesity drives neuroinflammation in part by increasing
microglial production of pro-inflammatory cytokines in hypothalamus, hippocampus, and other
brain regions
245–247
. Microglia proliferation and activation, as well as altered morphology were
observed after HFD exposure
74,137
. Exposure of microglial cells to SFAs resulted in an increase
in expression of TNF-a and IL-6
118
. In the CNS, TLR4 is mostly expressed by microglia under
basal condition and HFD-feeding
76,125,154
. TLR4/MyD88 signaling was strongly activated in the
hypothalamus after HFD exposure
76
. Moreover, TLR4 signaling is required for SFA-induced
microglia activation
141
. Blocking TLR4 signaling attenuated HFD-induced expression of pro-
inflammatory cytokines
76
. Thus, microglial-dependent TLR4 signaling contributes to HFD-
induced neuroinflammation.
Studies targeting neuroinflammation by inhibiting microglia activation or depletion of
microglia demonstrated that microglia modulated food intake during HFD feeding and
contributed to diet-induced obesity (DIO)
119,121,145
. To investigate the hypothesized role of
microglial TLR4 signaling as the key mediator of obesity-induced neural impairments, we
generated and analyzed mice that carry a microglial-specific deletion of TLR4 following 16
weeks HFD. Comparing metabolic, molecular, and behavioral outcomes, we found that
microglial TLR4 ablation protected HFD-induced adipose inflammation without affecting body
weight in male HFD-fed mice. In female, deletion of microglial TLR4 was associated with lower
body weight and improvement of metabolic functions. Microglial TLR4 deletion exhibited
significant protection against diet-induced cognitive impairments and glial activation in both
sexes.

2. Methods
2.1 Animals and Treatment
43
Microglia specific TLR4 knockout mice were generated by crossing TLR4
flx/flx
mice
(Jackson Labs, Stock No: 024872) with CX3CR1
CreER
mice (Jackson Labs, Stock No: 020940),
which expressing CreER recombinase induced by tamoxifen administration under the control of
CX3CR1 promoter. TLR4 flox-homozygous and Cre-positive mice (TLR4
flx/flx
, CX3CR1
CreER +/-
)
are referred to as the TLR4-MKO model. Their littermates, which are Cre-negative (TLR4
flx/flx
,
CX3CR1
CreER -/-
) referred to as controls (Ctl). At 3 months of age, male and female mice were
treated with tamoxifen (i.p., 150 mg/kg, Sigma Aldrich; solved in safflower oil) at two time points
48hr apart, a regimen previously reported to induce gene deletion
248
. Two weeks following the
last treatment, mice were randomly assigned into two nutrient-matched dietary groups: control
diet (10% fat; #D12450J, Research Diets; CTL) or high fat diet (60% fat; #D12492, Research
Diets; HFD). Animals were kept on experimental diets for 16 weeks. Animals were housed
under 12h light/dark cycle with lights on at 0600h with ad libitum access to food and water for
the duration of the experiment.
At the conclusion the dietary treatment, animals were euthanized via carbon dioxide
exposure following a 16h overnight fasting. Blood samples were collected by cardiac puncture
into EDTA-coated tubes. Aliquots of whole blood and plasma were frozen at −80 °C. Animals
were perfused with ice-cold phosphate-buffered saline (PBS) for 15min. Brain was rapidly
removed and either dissected into cortex, hippocampus and hypothalamus and immediately
frozen at -80°C, or fixed in 4% paraformaldehyde in 0.1 M Sorenson’s phosphate buffer for 72 h
at 4°C. Visceral fat pad was also harvested and stored at -80 ̊C.

2.2 Metabolic Measurements
Bodyweight and food intake were recorded on a weekly basis during the 16-week diet
period. Body composition (lean mass, fat mass and fluid) was determined using a NMR
analyzer (Bruker LF90 Minispec, Bruker Optics).

44
2.3 Glucose Measurement
Fasting glucose was measured at week 0 and every 4 weeks afterwards of the 16-week
diet period following an overnight fasting (16h). Blood was collected from the lateral tail vein
using a glucose test strip and immediately measured for glucose level using a Precision Xtra
Glucose Monitor (Abbott). At week 12, glucose tolerance test (GTT) was performed immediately
after baseline fasting glucose level was taken. Mice were administered with a glucose bolus (2
g/kg, D-glucose) via i.p. injection. Glucose levels were recorded 15-, 30-, 60-, and 120 min
thereafter.

2.4 Behavior Tests
For all behavioral tests, mice were brought into the behavior room and allowed to
acclimate for 30 min prior to testing. After each trial, animals were returned to their home cages.
The testing arena was disinfected with 70% ethanol and air-dried before next trial. Elevated plus
maze test was scored live. All other tests were recorded by a camera mounted above the maze
and analyzed using Noldus Ethovision XT software (Version 14). Both camera and software
were calibrated according to the manufacturer’s instructions. All behavioral tests were
performed during the hours of 0800 - 1600.

2.4.1 Baseline locomotor activity and anxiety
Open field test was performed 1 week following the last tamoxifen treatment. After 30min
acclimation, mice were placed in the middle of the arena (40 ´ 40cm) and allowed to move
freely for 5 min. The arena was equally divided into 16 squares and the middle 4 squares were
identified as center area. Total distance moved, total time the animals are mobile () and time
spent in center were recorded. Average velocity was calculated as total distance moved divided
by total time mobile.
45
Elevated plus maze was performed one day after the open field test. The maze was
elevated 40cm above the floor and consisted of a center area (6cm ´ 6cm), two closed arms
(30cm ´ 6cm) and two open arms (30cm ´ 6cm). Briefly, mice were placed in the center
platform of the maze facing a closed arm and allowed to move freely for 5 min. An arm entry or
exit was defined as both front paws were placed into or outside of the arm. Arm choices and
time spent in open arms were recorded for each animal.

2.4.2 Learning and memory
2.4.2.1 Barnes Maze
Barnes maze was performed at week 14 of dietary treatment using a modified protocol
form
123
. The maze was 90cm above the floor and had a white circular platform (91.5cm in
diameter) with 20 circular holes (5cm in diameter) evenly spaced around the boarder. The maze
was walled four-sided with black curtain. Four different visual cues at the level of the platform
were placed on each side of the walls. On day 1, animals were habituated to the maze. Mice
were placed in the center of the maze and allowed to move freely for 3 min under the red light.
All holes were closed during the habituation trial. From day 2 to day 5, each animal was given
three training trials with a 15 min inter-trial interval per day. One cuboid escape box (11cm L x
5cm W x 5cm H) was hidden beneath the maze. Three decoy boxes (5cm L x 5cm W x 2.5cm
H) were also placed beneath the maze to prevent visual confirmation of the location of the
escape box. All other holes were closed during the training trials. The location of the escape and
decoy boxes varied among animals but were kept constant for a given mouse throughout all
trials. A bright light was placed directly above the maze, and a buzzer located below the maze
was turn on to motivate the mice for the task. During each training trial, mice were placed in an
opaque cylinder in the middle of the maze. After 10s delay the cylinder was lifted, and the
animals were allowed to move freely up to 3 min to locate and enter the escape box. If the
46
animals did not find the escape box within 3 min, they were gently hand-guided into the box.
Once the animals enter the box either by successful escape or guidance, the light and buzzer
were turned off and mice was remained in the escape box for 1 minute before being returned to
their home cages. In each trial, the latency for full body entered the escape box was recorded.
Average latency of three training trials performed on the same day was calculated for each
animal.
48 h after the last training trial (day 7), mice were tested on a probe trial in which the
escape box was switched to a decoy box. The location of other three decoy boxes remained
unchanged from training trials for a given Mouse. The animals were placed in the cylinder for
10s. Then the cylinder was lifted and the mice were allowed to freely explore the maze for 3
min. Correct holes were considered as the target hole where the escape box was previously
located, as well as one adjacent hole on each side. Latency to reach the target hole (nose poke)
for the first time, and number of errors (nose pokes into incorrect holes) were recorded. Two
animals did not show successful escapes over the four days of training were excluded from all
analysis, as they were considered having low motivation to perform the task. The exclusion did
not affect the statistic results.

2.4.2.2 Novel object placement and recognition
Novel object placement and recognition tests were performed two days after the probe
trial in Barnes maze test, using a protocol adapted form
249,250
. Same arena in the open field test
was used. To avoid anxiety caused by the objects during the test, one day before first
habituation trial, a small LEGO piece was placed and remained in the home cage until all test
trials were completed. Two sets of objects (master lock and small stapler) were used in
sampling and test trials. On day 1 and day 2, each animal was given one habituation trial per
day for 2 consecutive days. During habituation, mice were placed in the middle of the empty
open field arena and allowed to freely explore the maze for 5 min. On day 3, mice were given
47
one habituation trial, one sampling trial and one test trial. 24 h after the second habituation trial,
animals were placed in the middle of the empty to freely explore for 2 min. Immediately after
that, the arena was set up in which two identical objects were placed in two adjacent corners
(5cm away from either wall). Mice were placed back in the arena with their heads positioned
opposite the objects. The trial lasted for 20 minutes or until a criterion that the total exploration
time of 30 s for both objects was met. Mice were considered as exploring an object when
showing investigative behavior and their nose is within 1.5 cm of the object. 4 h after the
sampling trial, animals were placed in the arena with one of the identical objects moved to the
diagonally opposite corner. The animals were given 20min or they had explored the objects for
a total of 30 seconds. On day 4 (24h after the sampling trial), animals were placed in the arena
with one of the identical objects replaced by a novel object. Animals were given 20 min or same
criterion of 30 s of total exploration time of both objects was achieved. During sampling and test
trials, the exploration time of each object was recorded. The location and nature of the novel
object were randomized and balanced across all animals.

2.5 Blood and Plasma Measurements
Hba1c levels were measured in whole blood using a commercially available kit (80310;
Crystal Chem). Fasting leptin and insulin were measured in plasma by ELISA kits (EZML-82K,
MilliporeSigma; EZRMI-13K, MilliporeSigma). HOMA-IR was calculated using the formula:
[ fasting blood glucose (mg/dL) × fasting plasms insulin (mU/L) / 405]. Plasma cytokine levels
were measured using Meso Scale Discovery (MSD) V-Plex Proinflammatory Panel 1 (mouse)
cytokine assay and analyzed on a QuickPlex SQ 120 instrument (MSD). Levels of IL-1b, IL-4
and IL-12p70 were below fit curve range in over half of the samples and therefore were
excluded. All measurements were performed according to manufactures protocol. All standards
and samples were measured in duplicate.

48
2.6 Immunohistochemistry and quantification
Fixed brains were transferred into 20% sucrose in PBS for 2 days until they sank down
to the bottom. Brains were then exhaustively sectioned in the coronal plane at 20 μm using a
cryostat (Leica Biosystems). Sections were stored in PBS with 0.03% sodium azide at 4◦C until
immunohistochemistry was performed. Every fourth section from around bregma -1.40mm to -
2.00mm (Paxinos and Franklin’s Mouse Brain Atlas, Edition 4th) were immunostained with
ionized calcium binding adaptor molecule 1 (Iba-1) and glial fibrillary acidic protein (GFAP).
Every fourth section from around bregma -2.40mm to -2.90mm (Paxinos and Franklin’s Mouse
Brain Atlas, Edition 4th) were immunostained with doublecortin (DCX). For Iba1 staining, brain
sections were firstly incubated with 10nM EDTA (pH6) at 95°C for 10 min. For DCX staining,
brain sections were pre-treated with 95% formic acid for 5min at room temperature. No antigen
retrieval was performed for GFAP staining. Next, sections were rinsed in Tris-buffered saline
(TBS) and treated with an endogenous peroxidase blocking solution for 10 min. Sections were
then rinsed in TBS with 0.2% Triton-X before being blocked for 30 min in corresponding
blocking solution. The blocking solution consisted of TBS with 2% bovine serum albumin (BSA)
for Iba1, 2% BSA and 2% normal goat serum for GFAP, and 5% normal horse serum for DCX.
Sections were incubated overnight at 4°C with primary antibody directed against Iba-1 (1:2000;
FUJIFILM Wako), GFAP (1:1000; Dako) and DCX (1:2500; Santa Cruz) diluted in the block
solution. On the following day, sections were rinsed in TBS with 0.1% Triton-X and incubated in
appropriate biotinylated secondary antibody diluted in the blocking solution for 1 h. After rinsing
in TBS with 0.1% Triton-X, sections were incubated in an avidin-biotin complex (Vectastain ABC
Elite kit, Vector Laboratories) for 1 h and immunoreactivity was visualized using
diaminobenzidine tetrahydrochloride (Vector Laboratories).
Microglia morphology was analyzed as previously described
251
.Briefly, brightfield images
contained the hippocampus CA1 (2 fields/section, 4 sections/animal) or arcuate nucleus (ARC)
of the hypothalamus (1 field/section, 6 section/animal) were taken using Keyence BZ-X710
49
microscope. Z series stacks (at least 10μm) with 0.7μm interval were acquired at 40x
magnification and processed with maximum contrast projection using BZ-X Analyzer software
(Version 1.3.1.1, Keyence). Then Images were imported into ImageJ (Version 1.53m) and
region of interest (ROI) of the CA1 and ARC was manually outlined. Microglia process length
and endpoints were analyzed using AnalyzeSkeleton (2D/3D) plugin
252
. Microglia cell somas (5
cells/field, 40 cells/animal) were manually outlined and their size and roundness were
determined using ImageJ (Version 1.53m).
To quantify GFAP immunoreactivity, brightfield images contained the hippocampus (20-
30 fields/section, 4 sections/animal) and ARC (1 field/section, 6 section/animal) were captured
with Keyence BZ-X710 microscope at 20x magnification. For hippocampus, overlapping images
from different fields were merged (uncompressed) using BZ-X Analyzer software (Version
1.3.1.1, Keyence). ROIs of the entire CA1 and ARC was manually outlined and images were
converted to greyscale and thresholded using ImageJ (Version 1.53m). GFAP load was
calculated as the percentage of the positively immunolabeled pixels over the total area of ROI.
DCX+ cells in the sub-granular zone and granule cell layer of the dentate gyrus (4
sections/animal) were counted live using Olympus BX50 microscope under 100× magnification.
DCX+ cells were classified based on morphology consistent with previous studies
253–255
: type 1,
no or one short (shorter than the diameter of the cell body) process; type 2, one process longer
than type 1 but only reached within the granule cell layer; or type 3, one long process or multiple
processes that branched into the molecular layer. The percentage of type 1, 2, and 3 cells in
each animal were calculated.

2.7 Microglia and monocyte isolation
Microglia and peripheral monocytes were collected from 6-7 months old male and
female mice 3 weeks after last tamoxifen treatment using the same regimen.
50
Animals were euthanized via carbon dioxide exposure and perfused with ice-cold PBS
for 5min. Brains were quickly removed and stored in Hanks' balanced salt solution without Ca
2+
,
Mg
2+
(HBSS w/o) on ice. Whole brain from each mouse was cut into small pieces using a sterile
scalpel blade. After centrifuged at 300 × g for 2 min, brain tissues were homogenized using
neural tissue dissociation kit (MiltenyiBiotec) following manufacture’s instruction. Next, cell
suspension was passed through a cell strainer (70 µm; MiltenyiBiotec) and centrifuged at 300g
for 10 minutes. Cells were then resuspended into 30% Percoll (GE Healthcare) and centrifuged
at 700 × g for 15 minutes at room temperature. After centrifugation, upper myelin layer was
carefully removed. Cells were resuspended in MACS buffer (PBS with 0.5% BSA and 2 mM
EDTA, pH 7.2) and incubated with CD11b microbeads (MiltenyiBiotec) for 15 minutes at 4°C.
After washing with MACS buffer, cells were applied onto MS columns (MiltenyiBiotec) and
separated using a magnetic separator (MiltenyiBiotec). Microglia (CD11b+) and rest cells
(CD11b-) were collected and centrifuged at 300g for 10 min. Cell pellets were stored at -80°C.
Peripheral monocytes were isolated using an adapted protocol from
256
. Briefly, for each
mouse blood (0.8-1ml) was collected by cardiac puncture into EDTA-coated tubes. Blood was
then diluted with red blood cell lysis solution (MiltenyiBiotec) and incubated 10 minutes at room
temperature. After centrifuging at 300 × g for 10minutes, cells were resuspended in RPMI and
carefully layered on top of Ficoll (GE Healthcare). Upper layer of plasma was removed, and the
middle layer of mononuclear cells was collected. Cells were then washed and resuspended in
HBSS following centrifuging at 400 × g for 30 min. After centrifuging at 200 × g for 5 min, cells
were resuspended in RPMI and transferred to cell culture plates. After 30min incubation in a 5%
CO2 container at 37°C, nonadherent cells were discarded and adherent cells were collected
after washing with PBS.

51
2.8 RNA isolation and quantitative PCR
RNA from isolated monocytes, microglia and other brain cells were extracted using the
quick-RNA miniprep kit (Zymo Research) according to manufacturer’s instructions. For adipose
and brain tissues, RNA was extracted using TRIzol reagent (Invitrogen) following the
manufacturer’s protocol, except an extra step of centrifuging at 12000 × g for 10 min was
performed after homogenization in TRIzol for adipose tissue to remove excess fat. Purified RNA
(500 ng) was reverse-transcribed using the iScript cDNA synthesis system (Bio-Rad). The
resulting cDNA was used for real-time quantitative PCR using SsoAdvanced Universal SYBR
Green Supermix (Bio-Rad) and the Bio-Rad CFX Connect Thermocycler. All samples were
tested in duplicates. Expression level of each probed gene was compared with averaged Ct
value of succinate dehydrogenase complex, subunit A (SDHA) and hypoxanthine guanine
phosphoribosyl transferase (HPRT) in the adipose tissue, b-actin and phosphoglycerate kinase
1 (pgk1) in the brain tissue, and b-actin in the isolated microglia and monocytes. Relative
quantification of mRNA was determined using the DD-CT method and normalized to Ctl mice or
male Ctl mice fed on CD. Data was presented as log2 fold change and statistical test was run
using the DCT values. Primer pair sequences of all target genes are listed in Table 1.

52
Table 1. Target genes and their respective PCR primer sequences.

Target Gene Primer Sequence
TNFa
Forward: 5’-CCCTCACACTCAGATCATCTTCT-3’
Reverse: 5’-GCTACGACGTGGGCTACAG-3’
IL-1b
Forward: 5’-GCAACTGTTCCTGAACTCAACT-3’
Reverse: 5’-ATCTTTTGGGGTCCGTCAACT-3’
IL-6
Forward: 5’-CTCTGGGAAATCGTGGAAAT-3’
Reverse: 5’-CCAGTTTGGTAGCATCCATC-3’
IL-10
Forward: 5’-CCCATTCCTCGTCACGATCTC-3’
Reverse: 5’- TCAGACTGGTTTGGGATAGGTTT-3’
TLR4
Forward: 5’-ATGGCATGGCTTACACCACC-3’
Reverse: 5’- GAGGCCAATTTTGTCTCCACA-3’
CD68
Forward: 5’-TTCTGCTGTGGAAATGCAAG-3’
Reverse: 5’-AGAGGGGCTGGTAGGTTGAT-3’
F4/80
Forward: 5’-CAAGTACGGCAACATGACCC-3’
Reverse: 5’-GCACTTGGTCAGTACTTTAGGTG-3’
Tmem119
Forward: 5’-GTGTCTAACAGGCCCCAGAA-3’
Reverse: 5’-AGCCACGTGGTATCAAGGAG-3’
Trem2
Forward: 5’-CTGGAACCGTCACCATCACTC-3’
Reverse: 5’- CGAAACTCGATGACTCCTCGG-3’
H2Ab1
Forward: 5’-CAGACGCCGAGTACTGGAAC-3’
Reverse: 5’-CAGCGCACTTTGATCTTGGC-3’
CD74
Forward: 5’-CAAGTACGGCAACATGACCC-3’
Reverse: 5’-GCACTTGGTCAGTACTTTAGGTG-3’
b-actin
Forward: 5’-AGCCATGTACGTAGCCATCC-3’
Reverse: 5’-CTCTCAGCTGTGGTGGTGAA-3’
Pgk1
Forward: 5’-GCCTGTTGACTTTGTCACTGC-3’
Reverse: 5’-GAGTGACTTGGTTCCCCTGG-3’
SDHA
Forward: 5’-ACACAGACCTGGTGGAGACC-3’
Reverse: 5’- GGATGGGCTTGGAGTAATCA-3’
HPRT
Forward: 5’-AAGCTTGCTGGTGAAAAGGA-3’
Reverse: 5’- TTGCGCTCATCTTAGGCTTT-3’
53
2.9 Statistics
Data were analyzed using GraphPad Prism (Version 9) and SPSS (Version 28). RNA
expression in isolated microglia and monocyte, open field test and elevated plus maze were
analyzed by unpaired two-tail t test with Welch’s correction. Novel object placement and
recognition tests were analyzed using one-sample t-test to determine whether the time spent
with the moved or novel object differed from chance level (15s)
249,257
. Three-way repeated
ANOVAs were performed on body weights, fasting glucose, glucose tolerance and escape
latency in the Barnes maze test with repeated measures across time points and genotype and
diet as a between-subjects factors. All other data were analyzed with two-way ANOVAs with
genotype and diet as between-subjects factors. For all ANOVAs, Greenhouse-Geisser
correction was used and significant effects were further analyzed with planned comparisons
between groups of interest using the Bonferroni or Tukey post-hoc testing. All cytokine levels in
plasma except for TNFa showed a highly non-normal distribution by Kolmogorov-Smirnov and
Shapiro-Wilks tests (p < 0.001), and skewness was either greater than 1 or less than -1. Thus,
data were log10 transformed before analysis. All datasets are expressed as mean ± SEM or
from minimum to maximum with quantiles. The threshold for statistical significance was set at p
< 0.05.

3. Results

3.1 Microglia TLR4 deletion improved HFD-induced insulin resistance without affecting adiposity
in male mice
To access the extent to which microglial TLR4 signaling is mediating the effects of obesity,
we created a new transgenic mouse model with induced TLR4 deletion specifically in microglia at
adult age by taking advantage of different turnover rate in different CX3CR1+ cell populations. It
has been demonstrated that CX3CR1 is exclusively expressed by microglia in the adult brain
258
.
54
Due to their longevity and large contribution of self-proliferation during microgliosis
259
, it has been
reported that in gene recombination in the microglia remains stable over several months after
tamoxifen treatment in CX3CR1
CreER
model
248
. Notably, during adulthood CX3CR1 is also
expressed on circulating blood monocytes, as well as other subsets of peripheral mononuclear
phagocytes, including tissue macrophages and dendritic cells
260,261
. However, these cells are
short-lived and replaced by bone marrow–derived progenitors
262
, and therefore will be replenished
by cells expressing unarranged TLR4. To validate TLR4 gene deletion, we isolated microglia and
blood mononuclear myeloid cells of TLR4-MKO and Ctl mice 3 weeks after tamoxifen injection.
After tamoxifen treatment, mRNA level of TLR4 in TLR4-MKO mice was reduced in microglia but
comparable to Ctl mice in other brain cells and peripheral monocytes (Figure 7). Tamoxifen has
been shown to reduce fat mass with an insignificant change in body weight, possibly through
increased reactive oxygen species production and apoptosis and autophagy of adipocytes. We
measured body composition before and after tamoxifen injection. Consistent with previous
findings, tamoxifen injection caused decreases in percentage of body fat in both sexes. But no
differences were found between control and TLR4-MKO mice before and after tamoxifen injection
(Figure 8).

55

Figure 7. Validation of TLR4 deletion efficacy in TLR4-MKO mice.
Quantitative real- time PCR was used to quantify mRNA expression levels of TLR4 in A) brain
cells and B) peripheral monocytes. Data are presented as mean (±SEM) values; n=4/group,
mixed sexes. ** denotes p < 0.01.

Figure 8. Effects of tamoxifen treatment on adiposity.
Percent body fat were measure 2 weeks after tamoxifen treatment in A) male and B) female
mice. Data are presented as mean (±SEM) values; n=20-27/group. Male Ctl mice are shown as
light blue bars with circles, male TLR4-MKO mice are shown as dark blue bars with triangles;
female Ctl mice are shown as pink bars with circles; female TLR4-MKO mice are shown as red
bars with triangles. Statistically significant main effects are denoted by T (tamoxifen treatment).
* denotes p < 0.05, ** denotes p < 0.01, *** denotes p < 0.001.

CD11b- CD11b+
0.0
0.5
1.0
1.5
TLR4 Expression
✱✱
Ctl
TLR4-MKO
Ctl TLR4-MKO
0.0
0.5
1.0
1.5
2.0
TLR4 Expression
A B
Pre-TAM Post-TAM
0
5
10
15
20
25
% Body Fat
M Ctl
M TLR4-MKO
✱✱✱✱
✱✱✱✱
T
Pre-TAM Post-TAM
0
5
10
15
20
% Body Fat
F Ctl
F TLR4-MKO
✱✱✱✱
✱✱✱✱
T
A B
56
In order to investigate the potential role of microglial TLR4 signaling in mediating obesity-
induced outcomes, 3-month-old TLR4-MKO and Ctl mice of both sexes were treated with
tamoxifen to induce TLR4 deletion. Two weeks following the treatment, mice were fed on nutrient-
matched diets with either 10% or 60% fat for a total of 16 weeks. In male mice, HFD was
associated with similar increases in body weight in TLR4-MKO and Ctl mice (Figure 9A). We
found a significant main effect of time (F (2.604, 164.1) = 273.3, p < 0.001), diet (F (1, 63) = 142.6, p <
0.001), and an interaction between time and diet (F (2.604, 164.1) = 152.5, p < 0.001) on body weight.
There was no main effect of genotype or interactions between diet and genotype on measures of
body weight. HFD was also associated with increased adiposity (Figure 9B) in male mice. We
found a significant main effect of diet (F (1, 64) = 382.9, p < 0.001) on percent body fat, and there
were no main effect of genotype or interactions between diet and genotype. Similar pattern was
found in food intake. There was a significant main effect of diet (F (1, 58) = 104.5, p < 0.001), such
that TLR4-MKO and Ctl male mice fed on HFD had higher daily kilocalorie consumption compared
with mice fed on CD (Figure 9C). There were neither significant main effect of genotype nor
interactions between diet and genotype.
To investigate metabolic outcomes of diet-induced obesity, we first measured levels of
fasting glucose during the 16-week diet period in TLR4-MKO and Ctl male mice. Regardless of
genotype, mice fed on HFD showed a significant increase in blood glucose (F (1, 64) = 96.0, p <
0.001; Figure 9D) at 4-, 8-, 12- and 16-week (p<0.05). When comparing Hba1c levels in the blood,
we found a significant main effect of diet (F (1, 64) = 382.9, p < 0.001; Figure 9E), and there were
no main effect of genotype or interactions between diet and genotype. Additionally, we performed
a glucose tolerance test and mice fed on HFD from both genotypes showed higher levels of blood
glucose (F (1, 64) = 77.2, p < 0.001; Figure 9F) specifically at 60- and 120-min time points (p<0.001).
We also calculated AUC (Figure 9G) and change in glucose levels at 120min from baseline
(Figure 9H) to access glucose tolerance. And again, there was a significant main effect of diet on
AUC (F (1, 64) = 102.9, p < 001) and change in glucose (F (1, 64) = 28.1, p < 0.001), such that HFD
57
mice performed significantly worse than mice fed on CD at glucose clearance. No significant main
effect of genotype or interactions between diet and genotype was observed on both
measurements.
We next examined the effects of HFD and microglial TLR4 deletion on levels of fasting
leptin and insulin in male mice. HFD significantly increased plasma leptin (F (1, 28) = 171.3, p <
0.001, Figure 9I) and insulin (F (1, 36) = 48.9, p < 0.001; Figure 9J) in TLR4-MKO and Ctl mice.
Interestingly, we also found significant interactions between diet and genotype treatment on
insulin levels (F (1, 36) = 7.5, p = 0.010), such that HFD significantly increased insulin levels and
microglial TLR4 deletion significantly lowered plasma insulin only in HFD-fed mice. We also
accessed effects of HFD on insulin resistance by calculating HOMA-IR (Figure 9K). There was a
significant main effect of diet (F (1, 36) = 55.7, p < 0.001), genotype (F (1, 36) = 4.8, p = 0.035), and
interaction between diet and genotype (F (1, 36) = 5.8, p = 0.021) on HOMA-IR, such that TLR4-
MKO mice were less insulin resistant compared with Ctl mice when fed on HFD.

58

-2 0 2 4 6 8 10 12 14 16
0
20
30
40
50
Week
Body weight (g)
M Ctl CD
M TLR4-MKO CD
M Ctl HFD
M TLR4-MKO HFD
Diet: p < 0.001
0 4 8 12 16
75
100
125
150
175
Week
Blood Glucose (mg/dL)
Diet: p < 0.001
CD HFD
0
1
2
3
4
Insulin (ng/ml)
✱✱✱ ✱
✱
D I
0 120 15 30 60
0
200
400
600
Minutes
Blood Glucose (mg/dL)
Diet: p < 0.001
CD HFD
0
20
40
60
% Body Fat
✱✱✱
✱✱✱
D
CD HFD
0
4
5
6
7
% Hba1c
✱✱
✱
D
CD HFD
0
20000
40000
60000
AUC
✱✱✱
✱✱✱
D
CD HFD
0
100
200
300
400
Glucose change
from baseline (mg/dL)
✱
✱✱✱
D
CD HFD
0
8
10
12
14
16
Food Intake (kcal / day)
✱✱✱
✱✱✱
D
M Ctl
M TLR4-MKO
CD HFD
0
10
20
30
40
Leptin (ng/ml)
✱✱✱
✱✱✱
D
CD HFD
0
10
20
30
HOMA-IR
✱✱✱ ✱
✱✱
D G I
A B C
D E
I K
G
J
F H
59
Figure 9. Effects of HFD and microglial TLR4 deletion on body weight and metabolic
outcomes in male mice.
A) Body weights of Ctl (black circle) and TLR4-MKO (blue triangle) male mice maintained on CD
(dashed line) and HFD (solid line) during the 16-week dietary treatment; Vertical dotted line
indicates the week dietary treatment started. B) Percent body fat at the conclusion of the
experiment. C) Calculated average daily caloric intake. D) Blood glucose levels following 16h
overnight fasting during the 16-week dietary treatment. E) Levels of blood Hba1c at the end of
the dietary treatment. F) Blood glucose levels over time after a glucose bolus. G) Calculated
AUC for the glucose tolerance test. H) Change in fasting blood glucose levels relative to
baseline at 120 min after a glucose bolus. I) Plasma leptin and J) insulin levels at the end of the
dietary treatment. K) Calculate HOMA-IR score. Data are presented as mean (+SEM) values;
n=15-19/group in A), D) and F). Male Ctl mice are shown as light blue bars with circles, male
TLR4-MKO mice are shown as dark blue bars with triangles. Statistically significant main effects
are denoted by D (diet), G (genotype) and I (diet and genotype interaction). * denotes p < 0.05,
** denotes p < 0.01, *** denotes p < 0.001.

60
3.2 Microglia TLR4 deletion improved HFD-associated metabolic outcomes in female mice
In females, HFD was also associated with increases in body weight (F (1, 58) = 29.8, p <
0.001; Figure 10A). Additionally, we found significant main effect of genotype (F (1, 58) = 9.1, p =
0.004) and interactions among time, diet and genotype (F (16, 928) = 2.6, p = 0.01) on body weight
during the 16-week dietary treatment. Between group comparisons revealed that TLR4-MKO mice
had lower body weight compared with Ctl mice only when fed on HFD (p = 0.006). HFD
significantly increased percent body fat (F (1, 56) = 101.5, p < 0.001; Figure 10B) in TLR4-MKO and
Ctl mice, and there was a non-significant trend towards an effect of genotype (p = 0.053), with
microglial TLR4 deletion was associated with lower body fat percentage in females. The observed
differences in body weight and adiposity were likely contributed by food intake (Figure 10C). There
was a significant main effect of diet (F (1, 57) = 53.9, p < 0.001) and genotype (F (1, 57) = 6.7, p =
0.021) on daily kilocalorie consumption. Post hoc tests showed a non-significant trend towards
TLR4-MKO mice having less food intake compared with Ctl mice when fed on HFD (p = 0.088).
The differences among groups in body weight and adiposity were paralleled by the
effects of diet and microglial TLR4 deletion on metabolic measures. HFD was associated with a
significant increase in fasting glucose levels (F (1, 58) = 21.9, p < 0.001; Figure 10D) in females.
Between group comparisons showed HFD increased fasting glucose at 8-, 12- and 16-week
(p<0.05) in Ctl mice, whereas TLR4-MKO mice fed on HFD showed higher fasting glucose at 4-
and 16-week (p<0.05) compared with CD-fed mice. Similarly, HFD also increased Hba1C levels
regardless of genotype (F (1, 44) = 6.9, p = 0.012; Figure 10E). No significant main effect of
genotype or interactions between diet and genotype was found. When challenging the mice with
a glucose bolus, we found a significant main effect of diet (F (1, 58) = 76.04, p < 0.001), and a
non-significant trend towards a main effect of genotype (p = 0.052) and interaction between diet
and genotype (p = 0.052) on blood glucose (Figure 10F). Between group comparisons revealed
that Ctl HFD mice had higher blood glucose at all time points (p<0.001) compared with Ctl mice
fed on CD, while the effect of diet was only observed at 30- and 60-min time points in TLR4-
61
MKO mice (p < 0.001). Analysis of GTT AUC showed a significant main effect of diet (F (1, 58) =
82.8, p < 0.001), genotype (F (1, 58) = 6.1, p = 0.016) and interactions between diet and genotype
(F (1, 58) = 4.1, p = 0.049), such that HFD increased AUC in both genotypes and microglial TLR4
deletion reduced AUC only in HFD-fed mice (Figure 10G). Additionally, female HFD mice were
impaired at returning to baseline glucose levels compared with CD-fed mice (F (1, 58) = 4.4, p =
0.040; Figure 10H). And there was a non-significant trend towards TLR4-MKO mice performed
better than Ctl mice (p = 0.056).
When examining fasting leptin levels in females, we found a significant main effect of diet
(F (1, 29) = 18.4, p<0.001) and genotype (F (1, 29) = 4.6, p = 0.039) on plasma levels of leptin (Figure
10I). Post hoc test showed that HFD significantly increased plasma leptin levels only in Ctl mice
(p = 0.002). Neither HFD nor microglial TLR4 deletion significant affected fasting insulin levels
(Figure 10J). However, there was a significant main effect of diet in calculated HOMA-IR (F (1, 34)
= 6.1, p = 0.019; Figure 10K) such that HFD induced insulin resistance only in Ctl mice (p = 0.038).

62

-2 0 2 4 6 8 10 12 14 16
0
15
20
25
30
35
Week
Body weight (g)
F Ctl CD
F TLR4-MKO CD
F Ctl HFD
F TLR4-MKO HFD
✱✱
Diet: p < 0.001
Genotype: p = 0.004
0 4 8 12 16
0
75
100
125
Week
Blood Glucose (mg/dL)
Diet: p < 0.001
CD HFD
0.0
0.5
1.0
1.5
Insulin (ng/ml)
0 120 15 30 60
0
200
400
600
Minutes
Blood Glucose (mg/dL)
Diet: p < 0.001
Genotype: p = 0.052
Diet × Genotype: p = 0.052
CD HFD
0
20
40
60
% Body Fat
✱✱✱
✱✱✱
D, Genotype p = 0.053
CD HFD
0
4.0
4.5
5.0
5.5
6.0
% Hba1c
D
CD HFD
0
20000
40000
60000
AUC
✱✱✱ ✱
✱✱✱
D G I
CD HFD
-50
0
50
100
150
200
250
Glucose change
from baseline (mg/dL)
D, Genotype p = 0.056
CD HFD
0
8
10
12
14
16
Food Intake (kcal / day)
✱✱✱
✱✱✱
D G
F Ctl CD
F TLR4-MKO
p = 0.088
CD HFD
0
10
20
30
40
Leptin (ng/ml)
D G
✱✱ p = 0.076
CD HFD
0
5
10
15
HOMA-IR
D
✱
A B C
D E
I K
G
J
F H
63
Figure 10. Effects of HFD and microglial TLR4 deletion on body weight and metabolic
outcomes in female mice.
A) Body weights of Ctl (black circle) and TLR4-MKO (red triangle) female mice maintained on
CD (dashed line) and HFD (solid line) during the 16-week dietary treatment; Vertical dotted line
indicates the week dietary treatment started. B) Percent body fat at the conclusion of the
experiment. C) Calculated average daily caloric intake. D) Blood glucose levels following 16h
overnight fasting during the 16-week dietary treatment. E) Levels of blood Hba1c at the end of
the dietary treatment. F) Blood glucose levels over time after a glucose bolus. G) Calculated
AUC for the glucose tolerance test. H) Change in fasting blood glucose levels relative to
baseline at 120 min after a glucose bolus. I) Plasma leptin and J) insulin levels at the end of the
dietary treatment. K) Calculate HOMA-IR score. Data are presented as mean (+SEM) values;
n=15-16/group in A), D) and F). Female Ctl mice are shown as pink bars with circles; female
TLR4-MKO mice are shown as red bars with triangles. Statistically significant main effects are
denoted by D (diet), G (genotype) and I (diet and genotype interaction). * denotes p < 0.05, **
denotes p < 0.01, *** denotes p < 0.001.

64
3.3 Microglial TLR4 deletion reduced HFD-associated peripheral inflammation
DIO is characterized as low-grade chronic inflammation that affects many organs
including adipose tissue. Though CX3CR1 is expressed only in a small population of adipose
tissue resident macrophages
263
, it has been reported that increased macrophage pool induced
by HFD
56
is contributed by local proliferation
60,264
as well as infiltration of peripheral
macrophage
265
and other immune cells
266
.Therefore, we first assessed TLR4 expression in the
visceral fat pad. In male mice, neither diet nor genotype significantly affected TLR4 expression
levels in the adipose tissue (Figure 11A). To investigate the role of microglial TLR4 on adipose
inflammation, we then examined gene expressions of inflammatory cytokines. In males, there
was a significant main effect of diet on levels of TNFa (F (1, 28) = 4.790, p = 0.037; Figure 11B).
Post hoc test showed that HFD increased TNFa expression only in Ctl (p = 0.038) but not in
TLR4-MKO mice. We also found significant interactions between diet and genotype on
expressions of IL-1b (F (1, 28) = 8.869, p = 0.006; Figure 11C) and IL-6 (F (1, 28) = 6.242, p = 0.019;
Figure 11D). Between-group comparisons revealed that HFD increased IL-b and IL-6 levels in
Ctl mice (p < 0.05), and absence of microglial TLR4 significantly decreased IL-6 only in HFD-fed
mice (p = 0.001). Furthermore, we quantified mRNA expression of the macrophage markers
CD68 and F4/80. We observed significant main effect of diet on CD68 expression (F (1, 28) =
36.71, p < 0.001; Figure 11E), in which HFD increased CD68 expression levels in both Ctl and
TLR4-MKO mice (p < 0.01). HFD also increased F4/80 levels (F (1, 28) = 5.462, p = 0.027; Figure
11F), though this did not reach statistical significance in either Ctl or TLR4-MKO mice. Finally,
we examined gene expression of MHCII protein H2ab1 and MHCII invariant chain peptide
CD74. There was a significant main effect of diet on levels of H2ab1 (F (1, 28) = 6.904, p = 0.014;
Figure 11G) and CD74 (F (1, 28) = 4.731, p = 0.038; Figure 11H). Between-group comparisons
revealed that HFD is associated with increased H2ab1 and CD74 expression only in Ctl mice
but not TLR4-MKO mice.
65

Figure 11. Effects of HFD and microglial TLR4 deletion on adipose inflammation in male
mice.
Quantitative real-time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) TNFa, C) IL-1b and D) IL-6, macrophages marker E) CD68 and F)
F4/80, and MHCII-related marker G) H2Ab1 and H) CD74. Data are represented as log2 fold
change, from minimum to maximum with quantiles; n=8-9/group. Male Ctl mice are shown as
light blue bars with circles, male TLR4-MKO mice are shown as dark blue bars with triangles.
Statistically significant main effects are denoted by D (diet), G (genotype) and I (diet and
genotype interaction). * denotes p < 0.05, ** denotes p < 0.01, *** denotes p < 0.001.

CD HFD
-2
0
2
4
6
TNFα expression
(log2 fold change)
✱
D
CD HFD
-2
0
2
4
6
CD68 Expression
(log2 fold change)
D
✱✱✱
✱✱
CD HFD
-4
-2
0
2
4
6
IL-1β expression
(log2 fold change)
✱
I
CD HFD
-4
-2
0
2
4
F4/80 Expression
(log2 fold change)
D
CD HFD
-4
-2
0
2
4
6
IL-6 expression
(log2 fold change)
✱ ✱✱
I G, Diet p = 0.072
M Ctl
M TLR4-MKO
CD HFD
-4
-2
0
2
4
6
8
H2Ab1 Expression
(log2 fold change)
✱
D, Interaction p = 0.067
CD HFD
-2
0
2
4
6
CD74 Expression
(log2 fold change)
✱
D, Interaction p = 0.073
CD HFD
-2
0
2
4
TLR4 Expression
(log2 fold change)
Genotype p=0.094
A B C D
E F G H
66
In females, HFD significantly increased TLR4 expression in in the adipose tissue (F (1, 28)
= 4.511, p = 0.043; Figure 12A). For expression of inflammatory cytokines and macrophages
markers, there was a general pattern that HFD increased expression levels of TNFa (F (1, 28) =
10.0, p = 0.004; Figure 12B), IL-6 (F (1, 28) = 7.9, p = 0.009; Figure 12D), F4/80 (F (1, 28) = 6.1, p =
0.020; Figure 12F), H2ab1 (F (1, 28) = 6.2, p = 0.020; Figure 12G) and CD74 (F (1, 28) = 8.0, p =
0.009; Figure 12H). Post test showed a statistically non-significant trend towards HFD increased
H2ab1 expression only in Ctl mice. Microglial TLR4 deletion did not significantly altered mRNA
levels of these genes, except there was a non-significant trend of lower IL-6 in TLR-MKO mice.
We found significant interactions between diet and genotype on levels of CD68 (F (1, 28) = 4.8, p
= 0.038; Figure 12E) and a non-significant trend towards an effect of diet (p = 0.055), in which
HFD increased CD68 expression levels only on Ctl mice (p = 0.030). Diet and genotype did not
significantly affect IL-1b expressions in the adipose tissue in female mice (Figure 12C).

67

Figure 12. Effects of HFD and microglial TLR4 deletion on adipose inflammation in female
mice.
Quantitative real- time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) TNFa, C) IL-1b and D) IL-6, macrophages marker E) CD68 and F)
F4/80, and MHCII-related marker G) H2Ab1 and H) CD74. Data are represented as log2 fold
change, from minimum to maximum with quantiles; n=8-9/group. Female Ctl mice are shown as
pink bars with circles, female TLR4-MKO mice are shown as red bars with triangles. Statistically
significant main effects are denoted by D (diet), G (genotype) and I (diet and genotype
interaction). * denotes p < 0.05.

CD HFD
-2
0
2
4
TNFα expression
(log2 fold change)
D
CD HFD
-5
0
5
CD68 Expression
(log2 fold change)
✱
I, Diet p = 0.055
CD HFD
-3
-2
-1
0
1
2
3
IL-1β expression
(log2 fold change)
CD HFD
-3
-2
-1
0
1
F4/80 Expression
(log2 fold change)
D
CD HFD
-4
-2
0
2
4
IL-6 expression
(log2 fold change)
D, Genotype p = 0.057
F Ctl
F TLR4-MKO
CD HFD
-4
-2
0
2
4
6
H2Ab1 Expression
(log2 fold change)
D
p = 0.061
CD HFD
-4
-2
0
2
4
6
CD74 Expression
(log2 fold change)
D
CD HFD
-3
-2
-1
0
1
2
TLR4 Expression
(log2 fold change)
D A B C D
E F G H
68
3.4 Effects of HFD and microglial TLR4 deletion on plasma cytokine levels
We assessed levels of circulating cytokines in plasma as another measurement of
peripheral inflammation. In male mice, HFD significantly increased plasma levels of cytokines.
There was significant main effect of diet on plasma levels of TNFa (F (1, 36) = 4.2, p = 0.047;
Figure 13A), IL-5 (F (1, 36) = 26.4, p < 0.001; Figure 13C), and KC/GRO (F (1, 36) = 8.1, p < 0.008;
Figure 13G). Similarly, we observed a statistically non-significant trend of higher levels of IFNg
in HFD animals (Figure 13F). In contrast, we found a significant main effect of genotype (F (1, 36)
= 4.9, p = 0.034; Figure 13D) on plasma IL-6 levels, with statistically non-significant trend of
TLR4-MKO mice having lower IL-6 than Ctl mice only when fed with HFD.
In females, HFD also significantly increased plasma levels of TNFa (F (1, 32) = 4.6, p =
0.039; Figure 14A), IL-5 (F (1, 32) = 28.5, p < 0.001; Figure 14C), IL-6 (F (1, 32) = 6.2, p = 0.018;
Figure 14D) and IFNg (F (1, 32) = 5.9, p = 0.021; Figure 14F). Additionally, there was a significant
main effect of genotype on IL-6 levels (F (1, 32) = 6.9, p = 0.013), in which microglial TLR4 deletion
reduced plasma IL-6 levels in HFD mice. Microglial TLR4 deletion is associated with increased
circulating levels of IL-2 (F (1, 32) = 4.3, p = 0.045; Figure 14B), such that there was a statistically
non-significant trend of higher IL-2 in TLR4-MKO HFD-fed mice compared with CD mice.
69

Figure 13. Effects of HFD and microglial TLR4 deletion on plasma cytokine levels in male
mice.
Plasma levels of A) TNFa, B) IL-2, C) IL-5, D) IL-6, E) IL-10, F) IFNg and G) KC/GRO were
measured. Data are represented as minimum to maximum with quantiles; n=9-10/group. Male
Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are shown as dark
blue bars with triangles. Statistically significant main effects are denoted by D (diet), and G
(genotype). * denotes p < 0.05, *** denotes p < 0.001.

CD HFD
0.1
1
10
IFNγ (pg/ml)
Diet p = 0.061
CD HFD
0.1
1
10
100
IL-5 (pg/ml)
M Ctl
M TLR4-MKO
D
p = 0.054
✱✱✱
CD HFD
1
10
100
IL-10 (pg/ml)
CD HFD
1
10
100
1000
IL-6 (pg/ml)
p = 0.065
G
CD HFD
10
100
1000
KC/GRO (pg/ml)
D
✱
CD HFD
0
10
20
30
40
TNFα (pg/ml)
D
CD HFD
0.1
1
10
IL-2 (pg/ml)
A B C
D E
F G

70

Figure 14. Effects of HFD and microglial TLR4 deletion on plasma cytokine levels in
female mice.
Plasma levels of A) TNFa, B) IL-2, C) IL-5, D) IL-6, E) IL-10, F) IFNg and G) KC/GRO were
measured. Data are represented as minimum to maximum with quantiles; n=9/group. Female
Ctl mice are shown as pink bars with circles, female TLR4-MKO mice are shown as red bars
with triangles. Statistically significant main effects are denoted by D (diet), and G (genotype). *
denotes p < 0.05, ** denotes p < 0.01.

CD HFD
0.01
0.1
1
10
100
IFNγ (pg/ml)
D
CD HFD
0.1
1
10
100
IL-5 (pg/ml)
F Ctl
F TLR4-MKO
✱✱
✱✱
D
CD HFD
1
10
100
IL-10 (pg/ml)
CD HFD
0.1
1
10
100
1000
IL-6 (pg/ml)
✱✱ ✱
D G
CD HFD
10
100
1000
KC/GRO (pg/ml)
CD HFD
0
5
10
15
20
25
TNFα (pg/ml)
D
CD HFD
0.1
1
10
IL-2 (pg/ml)
p = 0.072
G
A B C
D E
F G
71
3.5 Effects of HFD and microglial TLR4 deletion on gene expression in the hypothalamus
Studies has identified an association between obesity and a similar type of low-grade
inflammation in CNS. To investigate the contribution of microglial TLR4 in regulating HFD
induced neuroinflammation, we assessed gene expression of cytokines, microglia/macrophage
markers in the hypothalamus using quantitative PCR. In male mice, we confirmed that TLR4
expression levels were still significant lower in TLR4-MKO male mice after 16-week of dietary
treatment (F (1, 29) = 16.2, p < 0.001; Figure 15A). There was a significant main effect of
genotype on IL-6 expression (F (1, 29) = 6.4, p = 0.016; Figure 15B), with a statistically non-
significant trend of increased IL-6 expression in TLR4-MKO mice fed on CD. For anti-
inflammatory cytokine IL-10, we found a significant interaction between diet and genotype (F (1,
29) = 4.9, p = 0.035; Figure 15C). Post hoc test revealed that HFD is associated with increased
expression of IL-10 only in Ctl mice. For microglia/macrophage markers, there was a significant
main effect of genotype on levels of CD68 (F (1, 29) = 4.818, p = 0.036; Figure 15D), with TLR4-
MKO mice having higher levels of CD68. In contrast, expression of Tmem119 which is
specifically expressed by microglia in the brain
112,267
was increased by HFD only in Ctl mice
(Figure 15E). There was a significant main effect of diet (F (1, 29) = 6.6, p = 0.016) as well as an
interaction between diet and genotype (F (1, 29) = 7.6, p = 0.010) on Tmem119 expression.
Trem2 is also largely expressed by microglia
112,268
, however neither significant main effects or
diet or genotype, nor an interaction between these factors was found on Trem2 levels (Figure
15F). Upregulated MHCII has been considered as a marker of reactive microglia
269
, thus we
examined gene expression of H2Ab1 and CD74. We found a significant main effect of diet on
levels of H2Ab1 (F (1, 29) = 4.3, p = 0.047; Figure 15G) and Cd74 (F (1, 29) = 8.4, p = 0.008; Figure
15H). Post hoc tests revealed a non-significant trend towards HFD increasing H2Ab1 (p =
0.071) and CD74 (p = 0.058) in Ctl mice.
72

Figure 15. Effects of HFD and microglial TLR4 deletion on gene expression in the
hypothalamus in male mice.
Quantitative real-time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) IL-6, C) IL-10, macrophage marker D) CD68, microglia marker E)
Tmem119 and F) Trem2, and MHCII-related marker G) H2Ab1 and H) CD74. Data are
represented as log2 fold change, from minimum to maximum with quantiles; n = 8-9/group. Male
Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are shown as dark
blue bars with triangles. Statistically significant main effects are denoted by D (diet), G
(genotype) and I (diet and genotype interaction). * denotes p < 0.05, ** denotes p < 0.01.

CD HFD
-2
-1
0
1
2
TLR4 Expression
(log2 fold change)
p = 0.065 ✱
G
CD HFD
-0.5
0
0.5
1
IL-6 Expression
(log2 fold change)
p = 0.055
G
CD HFD
-2
-1
0
1
2
IL-10 Expression
(log2 fold change)
I
CD HFD
-1
-0.5
0
0.5
1
CD68 Expression
(log2 fold change)
G
M Ctl
M TLR4-MKO
A B C D
CD HFD
-1
-0.5
0
0.5
1
1.5
Tmem119 Expression
(log2 fold change)
D I
✱✱
CD HFD
-1
-0.5
0
0.5
1
Trem2 Expression
(log2 fold change)
CD HFD
-2
-1
0
1
2
3
H2Ab1 Expression
(log2 fold change)
D
p = 0.077
CD HFD
-2
-1
0
1
2
CD74 Expression
(log2 fold change)
D
p = 0.058
E F G H
73
When examining hypothalamic gene expression in female mice, we found similar trends
as with male mice. In females, hypothalamic gene expression of TLR4 was significantly affected
by genotype (F (1, 29) = 7.5, p = 0.010; Figure 16A), with TLR4-MKO mice had lower levels of
TLR4 compare with Ctl mice. Moreover, absence of microglial TLR4 increased IL-6 expression
(F (1, 29) = 5.5, p = 0.026; Figure 16B). We found no significant main effects or interactions on
levels of IL-10, CD68, Tmem119 and Trem2 (Figure 16C-F). Furthermore, HFD is associated
with upregulation of H2Ab1 (F (1, 29) = 7.9, p = 0.009; Figure 16G) and CD74 (F (1, 29) = 5.3, p =
0.029; Figure 16H) in the hypothalamus in female mice.

74

Figure 16. Effects of HFD and microglial TLR4 deletion on gene expression in the
hypothalamus in female mice.
Quantitative real- time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) IL-6, C) IL-10, macrophage marker D) CD68, microglia marker E)
Tmem119 and F) Trem2, and MHCII-related marker G) H2Ab1 and H) CD74. Data are
represented as log2 fold change, from minimum to maximum with quantiles; n=8-9/group.
Female Ctl mice are shown as pink bars with circles, female TLR4-MKO mice are shown as red
bars with triangles. Statistically significant main effects are denoted by D (diet) and G
(genotype).

CD HFD
-2
-1
0
1
TLR4 Expression
(log2 fold change)
G
CD HFD
-1
-0.5
0
0.5
1
IL-6 Expression
(log2 fold change)
G
CD HFD
-1
0
1
2
IL-10 Expression
(log2 fold change)
CD HFD
-1
-0.5
0
0.5
1
CD68 Expression
(log2 fold change)
F Ctl
F TLR4-MKO
A B C D
CD HFD
-1
-0.5
0
0.5
1
Tmem119 Expression
(log2 fold change)
CD HFD
-1
-0.5
0
0.5
1
Trem2 Expression
(log2 fold change)
CD HFD
-3
-2
-1
0
1
2
3
H2Ab1 Expression
(log2 fold change)
D
CD HFD
-2
-1
0
1
2
3
CD74 Expression
(log2 fold change)
D
E F G H
75
3.6 TLR4-MKO mice were protected from glial activation in the hypothalamus
HFD has been shown to increase activation states of microglia and astrocytes. To
address possible mechanisms underlying the interactive effects of HFD and microglial TLR4
signaling, we examined glial activation by analyzing Iba-1+ cell morphology and quantifying
GFAP immunoreactivity in the ARC in the hypothalamus. In male mice, neither diet nor
genotype significantly affected averaged process length and endpoints of microglia (Figure 17B,
C). However, there was a non-significant trend towards an effect of genotype on process length,
with TLR4-MKO mice having shorter processes than Ctl mice. Activation phenotypes of
microglia are also associated with increased soma size and reduced soma roundness.
Therefore, we measured microglial soma size and roundness as complementary measures of
microglia activation. We found a significant main effect of diet (F (1, 20) = 5.208, p = 0.034) and
an interaction between diet and genotype (F (1, 20) = 13.41, p = 0.002) on microglia soma size
in male mice (Figure 17D). Between-group comparisons revealed that HFD was associated with
larger soma in Ctl mice (p = 0.002) and microglial TLR4 deletion significantly reduced soma size
in HFD-fed mice (p = 0.003). There was a non-significant trend of a main effect of diet on soma
roundness (Figure 17E).
We next performed reactivity analyses with astrocytes in the ARC. In male mice, the
level of astrocyte activation was significantly increased by HFD (F (1, 20) = 7.7, p = 0.012),
whereas microglial TLR4 deletion was associated with lower levels of astrocyte activation
compared with Ctl mice (F (1, 20) = 5.1, p = 0.036; Figure 18B).

76

Figure 17. Effects of HFD and microglial TLR4 deletion on Iba1+ cell morphology in the
ARC in male mice.
A) Representative images of Iba1 immunoreactive cell in the hypothalamus. Scale bar = 50μm.
B) Average process length of Iba1+ cell. C) Average endpoints of Iba1+ cell. D) Iba1+ cell
soma size. E) Iba1+ cell soma roundness. Data are presented as mean (+SEM) values; n =
6/group. Male Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are
shown as dark blue bars with triangles. Statistically significant main effects are denoted by D
(diet) and I (diet and genotype interaction). ** denotes p < 0.01.
77

Figure 18. Effects of HFD and microglial TLR4 deletion on GFAP immunoreactivity in the
ACR in male mice.
A) Representative images of GFAP immunostaining in the hypothalamus. Scale bar = 100μm.
B) Quantification of GFAP load. Data are presented as mean (+SEM) values; n = 6/group. Male
Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are shown as dark
blue bars with triangles. Statistically significant main effects are denoted by D (diet) and G
(Genotype).

78

When examining microglial activation in female mice, we found a significant main effect
of diet (F (1, 20) = 4.9, p = 0.039) and an interaction (F (1, 20) = 4.8, p = 0.040) on microglia
process length, in which HFD significantly increased process length only in Ctl mice (Figure
19B). However, neither diet nor genotype significantly affected average microglia endpoints
(Figure 19C). We then examined the microglia soma size (Figure 19D) and found a significant
main effect of genotype (F (1, 20) = 4.8, p = 0.041) and an interaction between diet and
genotype (F (1, 20) = 8.2. p = 0.009) such that HFD was associated with increased soma size in
Ctl but not TLR4-MKO mice, and microglial TLR4 deletion recued microglia soma sized when
fed HFD. Moreover, there was a significant main effect of diet on soma roundness (F (1, 20) =
5.820, p = 0.026; Figure 19E), with Ctl HFD-fed mice having lower roundness of microglial soma
compared with mice fed on CD. Lastly, HFD largely induced GFAP immunoreactivity (F (1, 20) =
11.0, p = 0.004; Figure 20B) and this effect was only significant in Ctl mice.
Together, our data suggested that deletion of microglial TLR4 significantly attenuated
microglia and astrocyte activation in the hypothalamus when exposed to HFD in both sexes.

79

Figure 19. Effects of HFD and microglial TLR4 deletion on Iba1+ cell morphology in the
ARC in female mice.
A) Representative images of Iba1 immunoreactive cell in the hypothalamus. Scale bar = 50μm.
B) Average process length of Iba1+ cell. C) Average endpoints of Iba1+ cell. D) Iba1+ cell
soma size. E) Iba1+ cell soma roundness. Data are presented as mean (+SEM) values; n =
6/group. Female Ctl mice are shown as pink bars with circles, female TLR4-MKO mice are
shown as red bars with triangles. Statistically significant main effects are denoted by D (diet)
and I (diet and genotype interaction). ** denotes p < 0.01.
80

Figure 20. Effects of HFD and microglial TLR4 deletion on GFAP immunoreactivity in the
ACR of female mice.
A) Representative images of GFAP immunostaining in the hypothalamus. Scale bar = 100μm.
B) Quantification of GFAP load. Data are presented as mean (+SEM) values; n = 6/group.
Female Ctl mice are shown as pink bars with circles, female TLR4-MKO mice are shown as red
bars with triangles. Statistically significant main effect is denoted by D (diet). * denotes p < 0.05.

81
3.7 Effects of HFD and microglial TLR4 deletion on hippocampal gene expressions
Neuroinflammation derived from obesity can extend beyond the hypothalamus.
Therefore, we accessed gene expression on the same set of markers in the hippocampus by
quantitative PCR. In male mice, we observed that microglial TLR4 deletion was generally
associated with improvement in diet-induced microgliosis and microglial activation. Firstly,
TLR4-MKO mice had significant lower TLR4 expression than Ctl mice on both diet (F (1, 31) =
32.3, p < 0.001; Figure 21A). Secondly, results demonstrated a significant interaction between
diet and genotype on levels of pro-inflammatory cytokine IL-6 (F (1, 30) = 4.8, p = 0.037; Figure
21B), and between-group comparisons revealed a non-significant trend toward higher IL-6
levels in TLR4-MKO mice compared with the Ctl mice when fed with CD. Moreover, mRNA
levels of anti-inflammatory cytokine IL-10 were significantly upregulated by HFD (F (1, 30) = 4.6,
p = 0.041; Figure 21C) but not affected by genotype. Thirdly, no significant diet or genotype
effects was observed on expression of macrophage marker CD68 (Figure 21D). However, HFD
induced the expression of microglia-specific markers Tmem119 (F (1, 31) = 7.4, p = 0.011;
Figure 21E) and Trem2 (F (1, 31) = 4.9, p = 0.034; Figure 21F), and the effect reached
statistical significance only in Ctl mice. Lastly, genes associated with microglia activation
including H2Ab1 (F (1, 30) = 5.4, p = 0.027; Figure 21G) and CD74 (F (1, 30) = 6.7, p = 0.015;
Figure 21H) were upregulated by HFD, and again the effect was only significant Ctl mice.
Female TLR4-MKO mice also had lower expression levels of TLR4 (F (1, 30) = 12.0, p =
0.002; Figure 22A) in the hippocampus than Ctl mice. However, we found no significant effects
of diet, genotype, or their interaction on expression of rest of the markers (Figure 22B-H).

82

Figure 21. Effects of HFD and microglial TLR4 deletion on gene expression in the
hippocampus in male mice.
Quantitative real- time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) IL-6, C) IL-10, macrophage marker D) CD68, microglia marker E)
Tmem119 and F) Trem2, and MHCII-related marker G) H2Ab1 and H) CD74. Data are
represented as log2 fold change, from minimum to maximum with quantiles; n=8-9/group. Male
Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are shown as dark
blue bars with triangles. Statistically significant main effects are denoted by D (diet), G
(genotype) and I (diet and genotype interaction). * denotes p < 0.05, ** denotes p < 0.01.

CD HFD
-2
-1
0
1
TLR4 Expression
(log2 fold change)
G
✱✱ ✱✱
CD HFD
-1
-0.5
0
0.5
1
IL-6 expression
(log2 fold change)
I
CD HFD
-1
-0.5
0
0.5
1
IL-10 Expression
(log2 fold change)
D
CD HFD
-1
-0.5
0
0.5
1
CD68 Expression
(log2 fold change)
M Ctl
M TLR4-MKO
A B C D
CD HFD
-1
-0.5
0
0.5
1
Tmem119 Expression
(log2 fold change)
✱
D
CD HFD
-1
-0.5
0
0.5
1
Trem2 Expression
(log2 fold change)
D, Interaction p = 0.064
✱
CD HFD
-2
-1
0
1
2
3
H2Ab1 Expression
(log2 fold change)
D
✱
CD HFD
-2
-1
0
1
2
3
CD74 Expression
(log2 fold change)
D
✱
E F G H
83

Figure 22. Effects of HFD and microglial TLR4 deletion on gene expression in the
hippocampus in female mice.
Quantitative real- time PCR was used to quantify mRNA expression levels of A) TLR4,
inflammatory cytokines B) IL-6, C) IL-10, macrophage marker D) CD68, microglia marker E)
Tmem119 and F) Trem2, and MHCII-related marker G) H2Ab1 and H) CD74. Data are
represented as log2 fold change, from minimum to maximum with quantiles; n=8-9/group.
Female Ctl mice are shown as pink bars with circles, female TLR4-MKO mice are shown as red
bars with triangles. Statistically significant main effects are denoted by G (genotype). * denotes
p < 0.05.

CD HFD
-2
-1
0
1
TLR4 Expression
(log2 fold change)
✱
G
CD HFD
-1
-0.5
0
0.5
1
IL-6 expression
(log2 fold change)
CD HFD
-1
0
1
2
IL-10 Expression
(log2 fold change)
CD HFD
-1
-0.5
0
0.5
1
CD68 Expression
(log2 fold change)
F Ctl
F TLR4-MKO
A B C D
CD HFD
-1
-0.5
0
0.5
1
Tmem119 Expression
(log2 fold change)
CD HFD
-1
-0.5
0
0.5
1
Trem2 Expression
(log2 fold change)
CD HFD
-2
-1
0
1
2
3
H2Ab1 Expression
(log2 fold change)
CD HFD
-1
0
1
2
3
CD74 Expression
(log2 fold change)
E F G H

84
3.8 Male TLR4-MKO mice were protected from glial activation in the hippocampus
Similar pattern was found when we examined microglia morphology in the hippocampus.
In male mice, there was a significant main effect of diet on averaged process length (F (1, 20) =
14.2, p = 0.001; Figure 23A) and endpoints (F (1, 20) = 5.3, p = 0.032; Figure 23B), with HFD
increased process length and endpoints only in Ctl mice. Neither diet nor genotype affect
microglial soma size in males (Figure 23C). However, we found a significant interaction
between diet and genotype on soma roundness (F (1, 20) = 6.5, p = 0.019; Figure 23D). Post
hoc test showed that TLR4-MKO mice had significantly higher roundness in microglia soma
than Ctl HFD-fed mice.
When examining astrocyte reactivity in the hippocampus, we found that in male mice
absence of microglial of TLR4 was associated with lower GFAP immunoreactivity (F (1, 20) =
4.825, p = 0.040; Figure 23E). Additionally, there was a non-significant trend towards an effect
of diet (p = 0.051).
Regardless of genotype, exposure to HFD did not significantly alter microglia
morphology or astrocyte reactivity in female mice in the hippocampus (Figure 24).

85

Figure 23. Effects of HFD and microglial TLR4 deletion on microglia morphology and
GFAP immunoreactivity in the hippocampus of male mice.
A-D) Microglial cell morphology was measured by Iba1 immunostaining. A) Average process
length of Iba1+ cell. B) Average endpoints of Iba1+ cell. C) Iba1+ cell soma size. D) Iba1+ cell
soma roundness. E) Astrocyte reactivity was accessed by GFAP immunoreactivity load. Data
are presented as mean (+SEM) values; n = 6/group. Male Ctl mice are shown as light blue bars
with circles, male TLR4-MKO mice are shown as dark blue bars with triangles. Statistically
significant main effects are denoted by D (diet), G (genotype) and I (diet and genotype
interaction). * denotes p < 0.05.

CD HFD
0
200
400
600
800
1000
Process length/cell (µm)
D
✱
CD HFD
0
20
40
60
Soma Size (µm
2
)
CD HFD
0
50
100
150
200
Endpoints/cell
M Ctl
M TLR4-MKO
✱
D
CD HFD
0
0.60
0.65
0.70
0.75
0.80
Soma Roundness
I
✱
CD HFD
0
10
20
30
40
50
GFAP Load (% Total Area)
G, Diet p=0.051
A B
C D
E
86

Figure 24. Effects of HFD and microglial TLR4 deletion on microglia morphology and
GFAP immunoreactivity in the hippocampus of female mice.
A-D) Microglial cell morphology was measured by Iba1 immunostaining. A) Average process
length of Iba1+ cell. B) Average endpoints of Iba1+ cell. C) Iba1+ cell soma size. D) Iba1+ cell
soma roundness. E) Astrocyte reactivity was accessed by GFAP immunoreactivity load. Data
are presented as mean (+SEM) values; n = 6/group. Female Ctl mice are shown as pink bars
with circles, female TLR4-MKO mice are shown as red bars with triangles.

CD HFD
0
200
400
600
800
1000
Process length/cell (µm)
CD HFD
0
20
40
60
Soma Size (µm
2
)
CD HFD
0
50
100
150
200
Endpoints/cell
F Ctl
F TLR4-MKO
CD HFD
0
0.60
0.65
0.70
0.75
0.80
Soma Roundness
A B
C D
E
CD HFD
0
10
20
30
40
50
GFAP Load (% Total Area)
87
3.9 Effect of HFD and microglial TLR4 on cognitive performance
Prior work with TLR4 knockouts has demonstrated that TLR4 has important functions in
brain development and behavior. Mice with constitutive knockout of TLR4 have impaired motor
coordination and the deficiency is associated with a reduction in the thickness of the molecular
layer of the cerebellum
270
. Moreover, TLR4 antagonist treated mice have higher levels of anxiety
responses
271
. Therefore, we determined if there was a behavioral phenotype associated with
microglial TLR4 deletion one week prior to dietary treatment. There were no genotype
differences in any of the behavioral outcomes, such as exploratory activity in open field (Figure
25A-C) or anxiety-like behavior in elevated plus maze (Figure 25D-E).
To investigate whether HFD and microglial TLR4 modulate cognitive outcomes, we
examined behavior on the Barnes maze to access hippocampal-dependent spatial learning and
memory. We first evaluated spatial learning by examining the latency to reach the escape box
during training trials. Male mice across all groups showed significant learning with increased
training, as indicated by a significant main effect of time (F (3, 147) = 67.4, p < 0.001; Figure
26A) and shorter latencies to reach the escape box on days 2-4, than on day 1 of training (p <
0.05). However, neither diet nor genotype affected rates of learning. When examining
performance on the probe trial, we found a significant effect of interaction between diet and
genotype on primary latency in male mice (F (1, 49) = 4.8, p = 0.033; Figure 26B), with a
statistically nonsignificant trend toward longer latency in HFD mice with Ctl genotype compared
with mice fed on CD (p = 0.13), and a nonsignificant trend of shorter latency in HFD-fed TLR4-
MKO mice than Ctl mice (p = 0.11). We also found a significant main effect of genotype on
percent errors (F (1, 49) = 9.3, p = 0.004; Figure 26C), in which TLR4-MKO mice approached
the correct hole more frequently than Ctl mice.
Female mice also showed significant learning during training trials, as there was a
significant main effect of time (F (2.440, 131.7) = 75.8, p < 0.001; Figure 26D). Additionally, we
found a significant interaction between time and genotype (F (2.440, 131.7) = 3.6, p = 0.022)
88
and a non-significant tread toward an effect of diet (p = 0.056). Between group comparisons
revealed that Ctl mice fed on HFD were slower at locating the escape box than CD mice
specifically on day 2 of the training (p = 0.01). There was no main effect of diet nor an
interaction between genotype and diet on probe latency or errors in female mice (Figure 26E-F).
We also examined cognitive performance using object placement (OP) and object
recognition (OR) tasks. For male mice, obesity impaired cognitive function such that Ctl mice on
HFD had impaired performance compared with CD mice in OP and OR tasks (Figure 25G-H).
TLR4-MKO mice on HFD performed better than control HFD mice as they spent significantly
more time than chance with moved object (p<0.05), suggesting absence of microglial TLR4
improves cognition in obese male mice. For female mice, HFD was associated with impaired
performance in OP and OR tasks regardless of genotype (Figure 26I-J).

89

Figure 25. Effects of microglial TLR4 deletion on exploration and anxiety-like behaviors
in male and female mice.
A-C) Explorative and anxiety-like behaviors were examined in the open field. A) Average
velocity and B) total distance moved in the open field, and C) the amount of time spent in the
center field. D-E) Anxiety-like behavior was assessed in the elevated plus maze. D) Percentage
of the number of times the animals crossed into the open arm and E) the amount of time spent
in the open arm of the maze. Data are presented as mean (+SEM) values; n=16-19/group. Male
Ctl mice are shown as light blue bars with circles, male TLR4-MKO mice are shown as dark
blue bars with triangles; female Ctl mice are shown as pink bars with circles, female TLR4-MKO
mice are shown as red bars with triangles.

M Ctl
M TLR4-MKO
F Ctl
F TLR4-MKO
0
5
10
15
Velocity (cm/s)
M Ctl
M TLR4-MKO
F Ctl
F TLR4-MKO
0
20
40
60
80
%Open Arm Entries
M Ctl
M TLR4-MKO
F Ctl
F TLR4-MKO
0
1000
2000
3000
4000
Distance moved (cm)
M Ctl
M TLR4-MKO
F Ctl
F TLR4-MKO
0
10
20
30
40
50
%Time in Open Arms
M Ctl
M TLR4-MKO
F Ctl
F TLR4-MKO
0
10
20
30
40
50
% Time in Center
A B C
D E
90

1 2 3 4
0
50
100
150
200
Day
Latency (s)
M Ctl CD
M TLR4-MKO CD
M Ctl HFD
M TLR4-MKO HFD
1 2 3 4
0
50
100
150
200
Day
Latency (s)
F Ctl CD
F TLR4-MKO CD
F Ctl HFD
F TLR4-MKO HFD
Time × Genotype: p < 0.05
Diet: p = 0.056
*
CD HFD
0
10
20
30
Time with Moved Object (sec)
** * *
CD HFD
0
10
20
30
Time with Moved Object (sec)
* *
CD HFD
0
10
20
30
40
Latency (s)
I
CD HFD
0
10
20
30
Latency (s)
CD HFD
0
10
20
30
Time with Novel Object (sec)
* *
CD HFD
0
10
20
30
Time with Novel Object (sec)
* *
CD HFD
0
20
40
60
80
100
% Errors
G
M Ctl
M TLR4-MKO
CD HFD
0
20
40
60
80
100
% Errors
F Ctl
F TLR4-MKO
A B C
D E F
G H
I J

91
Figure 26. Effects of HFD and microglial TLR4 deletion on cognitive performance in male
and female mice.
Barnes maze test was performed in A-C) male and D-F) female mice. A, D) Average escape
latency during the 4 days of training. B, E) Prime latency in probe trial. C, F) Percent errors
made during the probe trial. Novel object placement and recognition was performed in G-F)
male and I-J) female mice. G, I) Time spent with moved object. H, J) Time spent with novel
object. Dotted lines in G-J) indicate exploration at chance level (15s). Data are presented as
mean (±SEM) values; n=12-16/group. Male Ctl mice are shown as light blue bars with circles,
male TLR4-MKO mice are shown as dark blue bars with triangles; female Ctl mice are shown as
pink bars with circles, female TLR4-MKO mice are shown as red bars with triangles. Statistically
significant main effects are denoted by G (genotype) and I (diet and genotype interaction). *
denotes p < 0.05, ** denotes p < 0.01.

92
3.10 Microglial TLR4 deletion improved hippocampal neurogenesis in obese male mice
To examine neurogenesis, we quantified DCX-labeled cell numbers in the dentate gyrus
region of the hippocampus. In male mice, we found a significant main effect of diet (F (1, 20) =
6.1, p = 0.023) and an interaction between diet and genotype (F (1, 20) = 5.8, p = 0.026) on
DCX+ cell numbers (Figure 27E), such that HFD was associated with impaired neurogenesis
only in Ctl mice. We also assessed the maturation of DCX-positive cells by examining their
morphology. No main effects or interactions was found on the proportion of immature (type 1)
and less immature (type 2) cells. However, HFD-fed mice had significantly less mature cells
(type 3) compared with mice on CD (F (1, 20) = 6.6, p = 0.018; Figure 27F).
When examining DCX-labeled cell numbers in female, no difference was found across
groups on numbers or relative maturity of DCX+ cells (Figure 27G-H).

93

Figure 27. Effects of HFD and microglial TLR4 deletion on hippocampal neurogenesis in
male and female mice.
A) Representative image of DCX+ cell in the dentate gyrus. Scale bar = 100μm. B-D)
Representative images of DCX+ cell morphology. B) Type 1 cell C) type 2 cell and D) type 3
cell. Scale bar = 15μm. E, G) Numbers of DCX immunoreactive cells per dentate gyrus in E)
male and G) female mice. F, H) Percentages of each type of cells in F) male and H) female
mice. Data are presented as mean (+SEM) values; n=6/group. Statistically significant main
effects are denoted by D (diet) and I (diet and genotype interaction). * denotes p < 0.05.
94
4. Discussion
Numerous studies have examined interactions between CNS and systemic
inflammation. In the context of obesity, it was suggested that inflammatory signaling in the brain
contributes to HFD-induced metabolic disorders and mediates peripheral inflammation
245,272–275
.
Pharmacological inhibition of IKKβ/NF-kB signaling in the CNS
144
, genetic knockout of IKKβ in
the neurons
75
or knockout of Myd88 in the CNS
143
reduces HFD-induced weight gain and levels
of insulin resistance. Additionally, intracerebroventricular (ICV) infusion of TNFa increased
insulin secretion and impaired insulin sensitivity in peripheral tissues and blocking hypothalamic
TNFα signaling prevented HFD-induced weight gain and insulin resistance
80
. Similarly, ICV
administration of IL-4 exacerbated the metabolic outcomes of HFD and hypothalamic
inflammation, and these effects were abolished by pre-treating the mice with an IKKβ
inhibitor
276
. A role of microglia has been implied as the underlying mechanism that links central
inflammation and metabolic function
151,273,277
. Microglia depletion was associated with reduced
food intake induced by high dietary SFAs consumption
119,145
. Likewise, blocking microglia
proliferation prevented HFD-induced weight gain, improved hypothalamic leptin sensitivity and
decreased hypothalamic and peripheral inflammation
121
. Effects of inflammatory signaling in
microglia on DIO has also been accessed. Mice with IKKβ deficiency specifically in microglia
showed reduced caloric intake and body weight change, and less infiltration of peripheral
macrophages during HFD-feeding
145
. Together, these studies highlight the importance of
microglial inflammatory signaling in regulating DIO.
Recent studies have provided evidence for a role of TLR4 signaling in obesity-induced
inflammation and insulin resistance. The goal of our study is to access the extent to which
microglial TLR4 signaling contributes to the obesity-related adverse outcomes. We examined
systemic and neural changes after 16-week HFD exposure in the presence or absence of
microglial TLR4 in male and female mice. Here, we demonstrated that TLR4-MKO mice were
95
protected from diet-induced metabolic disruption, peripheral inflammation, cognitive impairment
and neurogenesis with sex differences.
In male mice, absence of microglial TLR4 did not affect body weight, adiposity and
glucose metabolism. Similarly, no genotype effect was found on macrophage accumulation in
the adipose tissue. However, TLR4-MKO mice had significantly lower fasting insulin levels as
well as reduced peripheral inflammation, as dementated by lower IL-b, IL-6, H2ab1 and CD74
levels in the visceral fat pad. There was a general pattern of HFD-associated increase in plasma
cytokines. However, paralleled with our findings in the adipose tissue we observed lower levels
of plasma IL-6 in TLR4-MKO male mice. This is in line with previous findings that circulating IL-6
is largely contributed by adipose tissue
278
. In female mice, lacking microglial TLR4 was
associated with significant lowering in body weight, improvement of metabolic function and
reduction in systemic inflammation. This is consistent with previous findings showing microglial
TLR4 signaling pathway is critical for mediating neuronal activity and feeding behavior in the
ARC
78
.
The association between lower insulin levels and reduced peripheral inflammation status
in TLR4-MKO obese mice is consistent with research in human populations. Several studies
have reported a subgroup of individuals that was described as “metabolically healthy obese”
(MHO). More specifically, despite being obese (BMI ≥30.0) they had lower insulin levels and
appeared to be protected from obesity related insulin resistance compared with obese but not
MHO subjects
279
. Peripheral inflammation has been implied as the underlying mechanisms of
the existence of MHO individuals
280,281
. In rodent models, inflammation in adipose tissue is
strongly associated with insulin resistance
282–284
. MHO individuals had significant lower IL-b and
IL-6 in the visceral fat pad
285
, as well as lower circulating C-reactive protein (CRP), TNFa and
IL-6
286
. Interestingly, peripheral inflammatory profile might also contribute to the existence of
96
lean but metabolically unhealthy individuals, as they were reported having higher levels of CRP,
TNFa and IL-6 in the blood
286
.
Collectively, our data showed that microglial TLR4 signaling is a key mediator of energy
balance and systemic inflammatory responses. The decreased peripheral inflammation in TLR4-
MKO female mice might be due to reduced food intake and fat mass. However, our findings in
male mice provide evidence for potential links between central and peripheral inflammatory
responses. Specifically, inhibiting microglia activation by TLR4 deletion improved HFD-induced
insulin resistance and reduced adipose inflammation without affecting macrophage expansion.
In our mouse model, tamoxifen triggered TLR4 deletion in the brain is microglia specific
and was sustained after the 16-week diet exposure. At adult age, Kupffer cells in the liver,
peritoneal, splenic and lung macrophages do not express Cx3CR1
262
. Although circulating
monocytes and a subpopulation of macrophages residing in the intestine has been reported
expressing Cx3CR1 during adulthood, they are short-lived and have a half-life of 2 days and 3
weeks respectively
262,287
. In lean mice, there was a limited population of macrophages in the
adipose tissue (~6%) is positive for CX3CR1
263
. After 8 weeks of HFD feeding, CX3CR1
+
cell
population was largely expanded, but was mostly contributed by infiltrating macrophage
263
. In
our study, tamoxifen treatment did not reduce TLR4 expression in the adipose tissue in TLR4-
MKO female mice, and unexpectedly non-significantly increased TLR4 expression in males. The
latter might be a result of compensatory upregulation of TLR4 in other cell types, such as cells
in stromal vascular fraction
134
, after transient deletion of TLR4 in circulating
monocytes/macrophages. Nonetheless, we cannot complete rule out the effects of peripheral
tissue-resident macrophages on HFD-induced metabolic function and systemic inflammation.
In contrast to lower peripheral IL-6 levels in TLR4-MKO mice, we observed that absence
of microglial TLR4 was generally associated with increased expression of IL-6 in the
hippocampus and hypothalamus. Besides a role as immune modulator, studies have shown that
97
IL-6 signaling in the CNS plays an important role in regulating food intake and metabolic
functions. For example, central administration of IL-6 significantly reduced food intake, improved
glucose tolerance and increased brown adipose tissue thermogenesis in lean as well as obese
mice
288,289
likely by suppressing hypothalamic IKKβ activation and ER stress
290
. On the other
hand, inhibit IL-6 expression in the lateral parabrachial nucleus increased food intake,
accompanied with increased body weight and adiposity
289
. IL-6 in the CNS is also involved in
the control of neuronal functions such as neurogenesis
291
, synaptic plasticity
292
and learning and
memory
293
. Because of the divergent pattern of IL-6 levels in the plasma and brain, our findings
suggest that IL-6 signaling might have different functions in the CNS than peripheral. Indeed, IL-
6 signaling promotes obesity induced adipogenesis and macrophage accumulation in the
adipose tissue
294
. Moreover, chronic subcutaneous infusion of IL-6 impaired hepatic insulin
signaling and induced insulin resistance
295
. Interestingly, a recent paper showed that IL-6
secreted from adipocytes promoted, while IL6 expressed by myeloid cells (including
macrophages and microglia) suppressed HFD-induced metabolic impairment and adipose
tissue inflammation
296
. This suggested that the role of IL-6 signaling is cell type dependent,
which might help explain the potential discrepancy of anti- vs pro-inflammatory characteristic of
IL-6 in the CNS than peripheral. In the CNS, IL-6 is expressed by many cell types including
neurons, microglia, astrocytes and endothelial cells
297
. Thought the exact cell contribution of
observed increases in IL-6 expression is unknow, it is possible that elevated IL-6 signaling is
responsible for improved metabolic function in TLR4-MKO mice.
Our study supports the hypothesis that microglial TLR4 signaling modulates HFD-
induced microgliosis and glial activation. First, we observed that HFD increased expression of
microglial specific marker Tmem119 and Trem2 in male mice, with a general pattern that
trended towards a significant increase only in Ctl animals. The only exception was that no diet
effect was found on Trem2 levels in the hypothalamus, which might be due to microglia
heterogeneity and a much lower Trem2 expression in the hypothalamus than other brain
98
regions
268
. A transcriptome study using RNA-seq has demonstrated downregulation of
Tmem119 and Trem2 in isolated microglia after HFD exposure
298
. However, different expression
pattern associated with microglial activation and inflammation was observed when examining
mRNA changes on the level of whole tissue. HFD is associated with elevated expression of
Trem2 in the hippocampus
299
. In addition, expression of Tmem119 and Tem2 was upregulated
in the hippocampus and was further increased by HFD feeding in Alzheimer’s disease mouse
models
300,301
. Though the exact function and gene expression pattern of Tmem119 and Trem2
under pathological conditions need to be further elucidated, HFD-associated upregulation of
these genes observed in our study reflects more likely an increase in microglia cell number
rather than altered activation state. Second, we found attenuated HFD-induced microglial
activation in TLR4-MKO mice in both sexes, demonstrated by partial reductions in markers of
microglia activation (H2Ab1 and CD74), as well as changes in microglia morphology. Third,
TLR4-MKO mice had significant less astrocyte activation in the hippocampus and hypothalamus
in male mice and in the hypothalamus in female mice, suggesting microglia-astrocyte
interactions in modulating neuroinflammation.
The role of microglia in diet-induced cognitive deficits has been evaluated. Rats with
HFD feeding fed showed significant cognitive impairments along with altered microglial
morphology
302
. Inhibition of microglial activation by minocycline treatment or blocking microglial
phagocytosis did not affect HFD-induced body weight gain, however, improved performance in
hippocampal-dependent behavioral tasks
123
. Moreover, a study showed that switching back to a
low-fat diet reduced microglial activation and attenuated the synaptic plasticity in obese mice
without complete reversing body weight
90
. These findings suggested a stronger relationship
between cognitive function and microglial activation than with adiposity. In our study, we
observed that HFD was associated with cognitive impairments in both sexes, and microglial
TLR4 deletion was able to improve some of the behavioral measures.
99
Activation of TLR4 and NF-κB pathways has been demonstrated to have negative
effects on proliferation of neural progenitor cells and neurogenesis
146,303
. Moreover, microglia
have an important role in regulating generation of new neurons both under physiological
condition and stress
304–306
. After eliminating most of the microglia in the adult brain by PLX5622
treatment, microglia were able to rapidly regenerate
307
. Interesting, repopulated microglia
exhibited different gene expression profile than resident microglia, with significant upregulation
of genes related to cell cycle and downregulation of genes associated with pattern/damage
recognition
308
. Functionally, these repopulated microglia adopted a neuroprotection phenotype
that increased neurogenesis and improved cognition after traumatic brain injury through
increasing IL-6 expression in the neuron and an elevation of IL-6 trans-signaling
308
. When
examining neurogenesis in the hippocampus, we found that DCX+ cell numbers in the dentate
gyrus was not significantly affected by HFD in female mice. In contrast, HFD impaired
hippocampal neurogenesis in male mice, an effect was only found in Ctl mice. No genotype
difference was observed in the percentage of type 3 cells, suggesting protective effect of
microglial TLR4 signaling in generation of new neurons rather than neuron maturation.
In additional to effects of microglial TLR4, we also evaluated sex differences in DIO-
related outcomes. We found that microglial TLR4 deletion did not ameliorate majority of the
metabolic dysregulation induced by HFD feeding in male mice. In contrast, female obese mice
with microglial TLR4 deficiency were associated with improved metabolic function which was
likely contributed by lower caloric intake. Many studies have found that female sex is generally
protected from the development of obesity and HFD-induced metabolic alterations
309–311
,
however, as the dietary treatment prolonged female mice showed similar levels of body weight
gain, insulin resistance and adipose inflammation as males
312,313
. Although the exact
mechanisms remain unknown, hypothalamic neural circuits controlling energy balance have
been proposed to attribute to the delayed response to HFD in female mice. Activation of the
POMC neurons supresses food intake and increases energy expenditure
314,315
. Female mice
100
had higher expression of POMC in the hypothalamus
316–318
, and their POMC neurons exhibited
higher neural activity
317
. The protection against obesity in female mice might also be due to
inherent sex differences in microglial transcriptional profiles. Male sex is significantly associated
with upregulation of NF-κB and inflammatory processes
319
. Female mice with CX3CR1
deficiency displayed a ‘masculinized’ response to HFD, where they gained significantly more
weight and displayed higher microglial activation and inflammation than their control
littermates
320
. Given the findings from the literatures, we speculated that a diet with less fat
content, such as western diet (45% fat), would expose the probable effects of microglial TLR4
on metabolic functions in male mice, while a more chronic treatment of HFD in female mice
would diminish the protective effect of microglial TLR4 on HFD-induced increases in food intake
and body weight, and glucose dysregulation. Additionally, HFD affected gene expression and
glial activation in hippocampus as well as hypothalamus in male mice, while in female mice HFD
did not affect any of the neuronal outcomes in the hippocampus. This suggested that diet-
induced neuroinflammation was not generalized to the whole brain in females, at least with 16-
week HFD treatment. Indeed, inflammation induced by HFD can develop much faster in the
hypothalamus than other brain regions. In male rodents, expression of inflammatory cytokines in
the hypothalamus was elevated within 24 hours of HFD consumption, while changes were not
observed in the hippocampus after up to 12 weeks of HFD
74,89,90
. Even though female sex is
more resistant to diet-induced metabolic impairment and inflammation, they are not protected
from obesity-related cognitive impairment
321,322
. We found sex differences on some of the
behavioral outcomes. HFD significantly impaired learning behavior during Barnes maze training
phase only in female Ctl HFD-fed mice. Additionally, microglial TLR4 deletion improved
performance in OP test in male but not female mice. Previous studies showed that the male sex
was associated with better performance on hippocampal-dependent spatial navigation or
learning tasks in rodent models
323–325
. Similarly, sex differences on spatial learning and memory
performance were found in humans, with men outperformed female on many of the spatial
101
test
326–329
. Thus, our observed difference in behavioral measures were likely contributed by
inherent sex differences in capability of spatial tasks. Additional research is needed to further
define how sex interacts with microglial inflammation in regulating DIO as well as the underlying
mechanisms of these interactions.

5. Conclusion
In summary, our data showed a sex-dependent protective effect of microglial TLR4
signaling against HFD-induced adverse outcomes. In male mice, absence of microglial TLR4
improved insulin resistance, inhibited adipose inflammation and increased neurogenesis. The
reduced adiposity and an overall improvement of metabolic functions in female TLR4-MKO mice
might be resulted from lowered food intake, which suggested links between microglial TLR4
signaling and hypothalamic neuronal circuits in regulating energy balance. Obese mice with
microglial TLR4 deletion exhibited significant protection against diet-induced cognitive
impairments and glial activation in both male and female mice. These findings add to the
growing literature on implying the important role of microglial inflammation in mediating obesity-
related outcomes. Clinical approaches targeting microglial TLR4 signaling may allow for the
intervention of neuroinflammation and obesity therapeutics.

102
Chapter 4: Conclusions and future directions

1. Summary of the findings
The role of innate immune activation has been proposed in chronic inflammation
associated with obesity, AD and many other age-related diseases. Among all pro-inflammatory
pathways, TLR4 signaling may be particularly important in mediating activation of innate
immune cell including microglia and macrophages. In this dissertation, I focused on TLR4-
dependent innate immune response in the context of two examples of diseases, AD and
obesity.
AD is an age-related neurodegenerative disease with multiple risk factors. The
predominant genetic risk factor for late-onset AD is the e4 allele of APOE. Importantly, prior
studies showed that APOE4 is associated with increased levels of innate immune response and
inflammation
28,33–35,147
. In addition, APOE-related risk of developing AD is modified by sex, with
women showing greater risk than men
210,211
. Extensive human and animal model literatures
indicate that women’s risk for AD is also affected by the normal, age-related depletion of
estrogens at menopause. For example, in mouse models, estrogen depletion by ovariectomy
accelerates the onset and progression of AD-like neuropathology and cognitive impairment,
whereas estrogen-based hormone therapies can partially prevent and or reverse these
effects
330,331
. Potential interactions between APOE4 genotype and estrogen status in regulating
innate immune response have not been well elucidated. In Chapter 2, I examined the
behavioral, peripheral, and neural effects of acute LPS challenge on adult female mice
homozygous for human APOE3 or APOE4 in the presence or absence of estradiol. We found
that overall APOE4 can exert relatively protective outcomes in acute inflammatory response.
Estrogen status showed generally small effects in all of the outcomes we measured, suggesting
absence of estrogen might not be sufficient to expose the interaction of female sex and APOE4
in acute inflammatory responses induced by LPS.
103
Obesity is significantly associated with a range of serious, age-related medical
conditions. Pro-inflammatory pathways have been proposed as essential components to many
of the adverse metabolic outcomes of obesity
48,49
. Peripheral macrophages residing in
metabolically active tissues regulate lipid metabolism and cytokine production in state in
response to stimuli including circulating SFAs that are enriched in HFD
332
. Neuroinflammation,
another well-established consequence of obesity, is also robustly associated with increased
age
333
, cognitive impairment
51
, and the development of dementias as well as AD
7
. Therefore,
neuroinflammation has been considered as a key mediator of obesity-induced neural
dysfunction
242–244
. Specifically, TLR4 signaling regulates inflammatory events in various cell and
tissue types
124,125
. Prior work has demonstrated that TLR4 signaling is an important mediator of
obesity-induced inflammation
134,137,138,240
. However, its role in neural obesity effects is largely
unknown. In Chapter 3, I investigated the hypothesis that microglial TLR4 signaling regulates
the neural dysfunction caused by obesity. Our results demonstrated links between microglial
TLR4 signaling and diet-induced peripheral inflammation, neuroinflammation, neurogenesis and
cognitive dysfunction in a sex-dependent manner.
Collectively, my research highlights the importance of innate immune alterations in
diseases associated with chronic inflammation. Moreover, my studies indicated that non-
neuronal cells such as microglia are essential regulators of neuroinflammation and other
adverse neural outcomes induced by obesity. Future therapeutics targeting microglial TLR4
signaling and modulating microglial functions could offer new approaches to treat obesity and
other age-related diseases.

2. Other mechanisms underlying obesity
Neuroinflammation has been implied in the development and progression of obesity, AD
and other neurodegenerative diseases. Many factors such as inflammatory stimuli, cytokines,
chemokines, brain cell types are all linked to neuroinflammation. A number of them are relevant
104
to but not addressed in my study. Here, I discussed some of the underlying mechanisms, with a
focus on obesity, that contribute to metabolism alterations and neuroinflammation during
pathological diseases.

2.1 Microglia-mediated neuroinflammation
Generally, activated microglia in response to brain insults are classified into two
phenotypes, a M1 (pro-inflammatory) and a M2 (anti-inflammatory) phenotype. The classical M1
microglia is associated with production of pro-inflammatory mediators such as TNF-α, IL-1β, IL-
6 and reactive oxygen species, and nitric oxide
334
. M2 microglia are characterized as releasing
anti-inflammatory cytokines such as IL-4, IL-10, and TGF-β that antagonist the pro-inflammatory
effects of M1 microglia
335
. Therefore, it has been proposed that the M1/M2 ratio might be a
relevant factor in the prediction of disease severity/stage and therapeutics targeting switching
microglia to a more M2 phenotype would be important
336
.
However, recent findings showed that the function of microglial cells under physiological
or pathological conditions are more perplexed than just M1/M2 categorization
337
. In primary cell
culture, pro-inflammatory microglia (M1) were associated with increases in glucose and
glutamine metabolism and suppression of fatty acid oxidation and synthesis
338
. In contrast, M2
microglia exhibited no changes in glucose metabolism but increased fatty acid oxidation and
synthesis
338
. During HFD-feeding, activated microglia were associated with M1 phenotype as
demonstrated by upregulation of TNFα and IL-1β
74
. However, they also displayed
characteristics of M2 phenotype on genes related to metabolism. For example, an increased
expression of lipoprotein lipase (LPL), a transport protein for lipids, was observed in the reactive
microglia of obese mice brain
256
. Genetic knockout of microglial LPL exaggerated body weight
gain from a high-carbohydrate HFD by inhibiting microglial uptake of lipid and increasing
mitochondrial utilization of glutamine
256
. Importantly, LPL deficiency in microglia also
accelerated loss of POMC neurons in the hypothalamus
256
. Another metabolic mediator,
105
mitochondrial uncoupling protein 2 (UCP2), is also important in regulating microglia metabolism
during activation. In animal models of DIO, HFD significantly increased UCP2 expression in the
microglia, along with increased mRNA levels of inflammatory cytokines and remodeling of
mitochondrial dynamics
339
. On the other hand, deletion of microglial UCP2 increased
mitochondrial respiration and ATP production and prevented mitochondrial morphology
alterations in microglia, which further protected mice from metabolic dysregulation induced by
HFD
339
.
Besides effects on immunometabolism, HFD also impairs microglial phagocytosis. In
genetically diabetic mouse models (ob/ob and db/db mice), researchers found significantly
fewer CD68-immunoreactive microglia in the hypothalamus
340
. Likewise, HFD feeding in rats
induced a decrease in the expression levels of CD68 in microglia
341
. LPL also regulates
microglial immune reactivity. Mice with microglial depletion of LPL showed decreased microglial
cell density and activation, as well as decreased phagocytic capacity in the hypothalamus of
obese mice
256
. Together, these data suggest that microglial immunometabolism and phagocytic
capacity are crucial components of reactive microglia in mediating pathogenesis of DIO. It would
be interesting to examine immunometabolism and phagocytic related factors in my study, as it
has been reported that activation of TLR4 signaling increased glycolysis and impaired
phagocytic capacity in macrophages and microglia
342,343
.
It is still unclear whether microglial activation has an adaptive significance to normal
physiology, since reactive microglia could exert both beneficial and detrimental effects in the
CNS
335,344,345
. Interestingly, a recent study reported a role of microglial activation in maintaining
glucose homeostasis. When mice were pair-fed and weight matched, microglial IKKb deletion
was associated with larger impairment in glucose regulation, insulin sensitivity, and neuronal
glucose sensing compared with wildtype controls
346
. In contrast, increasing microglial
inflammation improved glucose regulation in mice fed with either regular chow or HFD without
affecting body weight
346
. These data support the hypothesis that microglial activation induced
106
by HFD may be beneficial in some aspects. More studies are needed to continue exploring the
function of microglia activation in the context of normal physiology as well as obesity.

2.2 Other cellular contributions to neuroinflammation
2.2.1 Neurons
Neurons were shown to be affected during DIO. In the hypothalamus, HFD increased
expression as well as protein levels of Hsp72, indicating increased neuronal stress and
injuries
74,119
. Additionally, HFD feeding induced hypothalamic IKKβ/NF-κB activation which
further promoted neuronal ER stress
75
.
AgPR and POMC are two types of neurons that controls food intake and energy
expenditure. Activation of AgPR neurons suppresses, while POMC neurons induces food
intake
182,315
. Obesity induced by HFD attenuated the ability of AgRP neurons in sensing an array
of food-related stimuli such as intragastric fat, cholecystokinin and ghrelin, which consequently
impaired the performance on food-motivated behavior that was mediated by AgRP neurons
347
.
Obesity was also associated with increased autophagy in POMC neurons and eventual loss of
POMC neurons
74
.In the hippocampus, HFD decreased dendritic complexity and length, and
impaired synaptic plasticity
84,90
. HFD is also associated with decreased neurogenesis in the
dorsal and ventral of the dentate gyrus
88
.
When examining obesity-induced neuronal inflammation, investigators have been
heavily focused on the mediobasal hypothalamus (MBH), likely due to its important role in
regulating energy balance and glucose metabolism
70,71
. MyD88 deletion in nestin-expressing
cells prevented diet-induced weight gain and restored leptin sensitivity
143
. Studies targeting
IKKβ/NF-kB signaling pathway further validated the critical role of neuron-mediated
inflammation in the metabolic dysregulation caused by obesity. In normal physiology condition,
IKKβ was largely expressed in the hypothalamus specifically by neuronal cells compared with
other peripheral tissues
75
. Upon HFD-feeding, IKKβ was further upregulated in the neurons
107
located in the MBH
75
. Mice with conditional deletion of IKKb in all brain cells or AgRP neurons
exhibited lower body weight, improved glucose tolerance, and preserved central leptin/insulin
signaling sensitivity after HFD-feeding
75
. The protective effect against HFD-induced obesity of
IKKb deletion in AgPR neurons was mediated by SOCS3 signaling. A similar anti-obesity
phenotype was observed in mice with neuron specific SOCS3 deletion
348
. That is, they showed
significant protect against body weight gain and leptin and insulin resistance induced by HFD
348
.
Exogenous over-expression of SOCS3 in the MBH neurons diminished the anorexia effect of
IKKb deletion in AgPR neurons induced by HFD
75
.
On the other hand, IKKβ signaling might have different functions in POMC neurons.
First, IKKβ deficiency in POMC neurons did not prevent body weight gain, however it protected
mice from development of hypertension induced by HFD feeding
349
. Second, central
administration of TNFα differentially activated AgPR and POMC neurons, and IKKβ
phosphorylation levels was significantly higher in POMC neurons than AgPR neurons
349
. Third,
POMC neurons had a strong response to LPS- or leptin-induced NF-kB activation, and mice
with IKKβ deficiency in POMC neurons were protected from LPS- or leptin-induced anorexia
and subsequent decreases in body weight
350
.
How do neurons get activated during HFD-induced obesity? Treating cultured
hypothalamic neurons with TNFa was sufficient to induced neuronal insulin resistance
142
.
However, short-term exposure of SFA mixture or palmitate acid alone did not cause increased
expression of IκBa and IL-6 nor impairment in insulin signaling
142
. Moreover, constitutive
activation of the IKKβ/NF-κB in hypothalamic neurons did not trigger the production of
proinflammatory cytokines
75
. These findings suggested that SFAs do not directly affect
hypothalamic neurons. Instead, neuronal insulin resistance induced by obesity is developed via
stimulation of inflammatory mediators secreted by other cells, such as microglia. Indeed,
medium from activated BV-2 cell by SFAs strongly induced neuronal apoptosis in primary
108
neuronal culture
141
. In addition, although functionally hypothalamic IKKβ/NF-κB has a role in
diet-induced leptin and insulin resistance, obesity-associated increases in inflammatory
cytokines are more likely contributed by other non-neuronal cells.
It would be interesting to examine some the neuronal changes in my studies. In Chapter
3, we observed microglial TLR4 deletion was associated with reduced food intake in HFD-fed
female mice. We also found an association between HFD-feeding and impaired neurogenesis,
which is regulated by microglial TLR4 signaling in male mice. Besides hypothalamus, neurons in
the hippocampus also have a role in sensing endocrine and modulate food intake
351
. Thus,
assessing alterations in appetitive-associated neurol circuits and other markers related to
neuronal damage in both hypothalamus and hippocampus would help define the underlying
mechanism of microglial TLR4 signaling in protection of DIO.

2.2.2 Astrocytes
Astrocytes are involved in many essential brain functions, such as synaptic activities,
neuronal metabolism and neurovascular functions
352,353
. More specifically, astrocytes are able to
oxidize circulating fatty acids and synthesize ketone bodies, and preferably depend on ketone
bodies and lipids as their energy source
354
. Moreover, astrocytes express nutrient sensing
receptors and lipid regulators that modulate hypothalamic neuronal circuits related to feeding
bebavior
354
. For example, hypothalamic astrocytes express leptin receptors, and deletion of
leptin receptors specifically in astrocytes significantly altered astrocyte morphology and synaptic
plasticity in POMC and AgRP neurons, which resulted in increased food intake after fasting or
ghrelin treatment
355
. APOE, a lipid transporter, is mainly expressed by astrocytes in the
hypothalamus and has been shown to interact with leptin receptor in regulating food
intake
356
.Upon HFD-feeding, APOE was upregulated in astrocytes which required the presence
of leptin receptor; on the other hand, fasting was associated with decreased expression of
APOE as well as leptin levels
356
. Interestingly, the effect of decreased food intake caused by
109
leptin was also dependent on APOE, as APOE knockout mice showed lower sensitivity
regarding central leptin administration
356
.
TLR4 expression levels in astrocytes are significantly less compared with microglia in
normal adult brain in humans and many animal models
125
. However, there were studies that
reported increased astrocytic TLR4 expression during brain injuries or inflammation
357–360
.
Interestingly, this increased expression was not found in all astrocytes
360
, suggesting astrocytic
heterogeneity in response to exogenous inflammatory stimuli. In vitro evidence from cultured
primary astrocytes showed that they were able to respond to LPS, as LPS underwent binding
and was transported into astrocytes demonstrated by fluorescent staining
361
. Consequently,
LPS-induced alterations in gene expression profiles including pathogen recognition receptors in
a NF-κB-dependent fashion
361,362
. However, there was a minimal increase in the expressions of
pro-inflammatory cytokines, such as IL-1β, IL-6, and TNFα
361
. Treating the astrocytes with long-
chain fatty acid also did not induce pro-inflammatory cytokines secretion in primary culture
119
.
On the other hand, TLR4 overexpression enhanced LPS-induced responses in cultured primary
astrocytes
360
. They showed increased NF-κB activation as well as the expressions of IL-1β and
TNFα
360
. Together, these findings suggested an active function of TLR4 signaling in astrocytes
that is associated with astrocyte activation and production of pro-inflammatory cytokines. The
mechanism of astrocytic TLR4 upregulation during pathological conditions remains to be
resolved, however current literature suggests a role of microglia in increasing the sensitivity of
astrocytes to activation.
It has been proposed that the presence of microglia is required for increased production
of pro-inflammatory cytokines in astrocytes induced by LPS and other TLR4 ligands
361
. In
addition, treating mice with minocycline, a drug has been shown to inhibit microglia but not
astrocyte activation
363,364
, significantly suppressed nuclear translocation of STAT3 in astrocytes
induced by LPS in the brain
359
. These findings suggest a cross talk between microglia and
astrocyte in regulation of neuroinflammation; specifically, stimuli firstly trigger the inflammatory
110
responses in microglia, and astrocyte activation and astrogliosis may be a secondary response
to the microglia activation. In line with this finding, in Chapter 3 we found decreased astrocyte
activation in the hippocampus as well as hypothalamus that was associated with microglial
TLR4 deletion.
Astrocyte activation was observed within one week of HFD feeding, as shown by
increased expression of markers related to astrocyte activation and altered astrocyte
morphology
74,365
. Long term HFD exposure also affected astrocytic process plasticity and
induced NF-κB activation in the hypothalamus
366,367
. Besides increased expression of pro-
inflammatory cytokines such as TNF-α and IL-6, astrocyte reactivity was strongly associated
with elevated extracellular levels of the neurotransmitter GABA during DIO. In turn, this
suppressed the excitability of a sub-population of GABAergic neurons involved in energy
balance and finally led to increased body weight gain and adiposity
366,368
.
Studies targeting inflammatory signaling IKKβ/NF-kB implied that astrocytes are relevant
to metabolic dysregulation associated with DIO. Inhibition of IKKβ/NFκB signaling in astrocytes
prevented astrocyte activation after acute HFD treatment, accompanied with a slight increase in
food intake
365
. Forced activation of IKKβ in astrocytes increased body weight and fat mass in
mice fed with regular diet
366
. In addition, mice with astrocytes IKKβ deficiency exhibited lower
food intake when fed either regular chow or HFD, together with lower body weight gain in obese
mice
366
. In slight contrast, another group using a tamoxifen-inducible model showed that
ablation of IKKβ in astrocytes prior to HFD feeding did not alter the metabolic function of obese
mice
369
. However, astrocyte IKKβ deletion induced 6 weeks after HFD exposure protected mice
from HFD-induced glucose dysregulation by inhibiting food intake and increasing energy
expenditure, and reduced hypothalamic astrocyte activation along with pro-inflammatory
cytokine expressions
369
. Nonetheless, these findings collectively support an important role of
astrocytic immune response and inflammatory signaling in obesity-induced metabolic
dysregulation.
111
2.2.3 BBB integrity and infiltrating macrophages
The blood–brain barrier (BBB) consists of many types of cells preventing influx of
circulating toxic substances and controlling the transport of nutrients such as fatty acids and
hormones into the brain
370
. However, evidence from humans and animal models suggest an
association between loss of BBB integrity and obesity-induced chronic inflammation. For
example, the ratio of CSF:serum levels of albumin was increased in overweight or obese
women, suggesting alterations in BBB integrity
371
. Likewise, long-term HFD treatment caused a
significant increase of Evans blue dye transportation into the brain
372
. Moreover, HFD feeding
induced decreases of the levels of tight junction proteins such as claudin-5 and occludin in the
brain
373
. BBB integrity is also related to cognitive impairment induced by HFD. In a mouse
model of obesity (db/db mouse), suppressing BBB leakage by a protein kinase C beta inhibitor
reduced protein levels of TNFα, IL1-β and IL-6 in the forebrain, and improved spatial memory
374
.
As mentioned, HFD-induced obesity is associated with microglia activation and
microgliosis. One interesting question remaining to be answered is what cells are responsible
for the microglia proliferation when inflammatory stimuli are presented. In rodents,
approaches that transiently deplete microglia from the brain showed that resident microglia
were able to proliferate
307
. Meanwhile, HFD-induced inflammation in peripheral tissues is
associated with infiltration of monocytes; a similar process also occurs in the CNS. In normal
conditions, monocytes derived from the periphery cannot enter the brain due to impermeability
of BBB. Therefore, immune surveillance in the CNS mostly relies on the resident microglia.
However, monocytes can contribute to the population of CNS myeloid cells in case of BBB
disruption and leakage. Using a bone marrow chimeric mouse model with green-fluorescent
protein-labeled peripheral immune cells, studies have observed infiltration of peripheral
monocytes during high-fat feeding
375,376
. Such infiltration primarily occurs in circumventricular
regions but eventually spread throughout the whole brain
377
. Once entered into the CNS, these
infiltrated cells display some characteristics of activated microglia, such as expressing Iba-1 and
112
exhibiting a microglia-like morphology with enlarged soma and thickened processes
376
.
However, recent studies showed that they still maintained a unique transcriptional and
functional identity compared with resident microglia
377,378
, though the differences were less
evident when they were challenged with LPS
378
.
It is still uncertain what factors trigger the infiltration of peripheral monocytes into the
CNS. Microglia depletion or deleting microglial IKKβ protected mice from HFD-induced body
weight gain as well as peripheral myeloid cell infiltration into the CNS
145
. By contrast, activation
of microglia by deleting microglial A20, a negative regulator of NF-kB, exaggerated metabolic
dysregulation and increased peripheral cells infiltration in obese animals
145
. These data
reinforce the strong association between neuroinflammation and peripheral cell infiltration, but
also suggest a contribution of infiltrated macrophages to the pathogenesis of DIO.
When examining TLR4 expression levels in the brain at the end of the experiment, we
found that TLR4-MKO animals were still showing reduced TLR4 mRNA levels but to a lesser
extent compared with TLR4 levels in microglial fraction before dietary treatment. It would be
interesting to access the TLR4 expression by cell types. We speculate contributions of infiltrated
macrophages, but it is also possibly due to the loss of recombination efficiency in the microglia
and astrocyte activation as discussed above. This would allow us to determine the relative
impact of microglia, infiltrated cells, and astrocytes on HFD-induced neuroinflammation.

2.3 Sex differences in diet-induced obesity
It is well documented that sex differences exist in energy homeostasis and effects of
DIO. Specifically, in rodent models males are generally more susceptible to obesity induced
body weight gain and metabolic dysregulation compared with females
309–311,320,379–382
. The
relative protection of females in DIO may be contributed by their higher estrogen levels.
Estrogen is an essential regulator of body weight and adiposity. In female mice,
depleting endogenous estrogen by OVX increased body weight and body fat
220,221,383,384
.In
113
Chapter 2, we observed effects of estrogen status on body weight and weights of adipose fat
pads. OVX was associated with increased body weight that was prevented by E2 treatment.
Likewise, OVX mice showed significantly higher weights of gonadal and retroperitoneal fat pads.
In humans, lack of estrogen was also associated with visceral fat accumulation in
postmenopausal females
385
. Estrogen receptor ERα has been shown to regulate estrogen
signaling and modulate body weight gain in obesity. Mice with ERα deficiency displayed
increased body fat and adipocyte size by suppressing energy expenditure
386
. In the CNS, ERα
is widely expressed in different brain regions including hypothalamus, hippocampus and cortex.
Long term HFD-feeding was associated with alterations in the expression of ERα in brain,
suggesting a role of ERα in controlling the adverse neural effects induced by obesity
387
. Mice
with ERα deletion in nestin-expressing cells exhibited increased body weight and fat
accumulation, along with increased food intake and decreased energy expenditure
388
.
Moreover, mice lacking ERα in specifically in POMC neurons showed increased food intake and
impaired leptin sensitivity
388
. Importantly, a recent study reported that ERα signaling also
contributed to hypothalamic inflammation induced by HFD. In vitro, E2 treatment or
overexpression of ERα in mouse hypothalamic neuronal cell cultures significantly decreased the
expression of TNFα and IL-6 induced by palmitic acid
389
.
Of note, estrogen and ERα also exhibited protective effect against obesity in male mice.
Decreased levels of ERα protein was observed in the hypothalamus of male mice after long-
term HFD exposure
389
. In addition, male mice with ERα deficiency showed increased body
weight gain as well as body fat mass
390
. Further, estrogen treatment in male mice was
associated with decreased body weight, adiposity as well as food intake
391
.
Interestingly, likely due to decreased estrogen levels and or responsiveness by age, the
protection against obesity in female mice is diminished in late adulthood. Indeed, HFD feeding
in middle-aged mice was associated with greater increases in body weight gain, adiposity and
impairments in glucose tolerance in female compared with male mice
392,393
. Moreover, HFD
114
suppressed immune functions in peritoneal macrophage in both male and female middle-aged
mice
394
.
Collectively, high levels of circulating estrogen at least partially contribute to protection
against obesity in females at young age. Future work will need to fully determine the extent to
which neuroinflammation induced by excessive nutrition is regulated by sex.

3. Future directions
Microglia are involved in numerous normal functions in brain as well as in immune
responses to injury, infections and CNS diseases
395,396
. TLR4 is an important contributor to
microglial activation, known to initiate an inflammatory cascade in response to various CNS
stimuli, but may also participate in autoimmunity, neurodegeneration, and brain injuries
397
. Thus,
there are applications of our mouse model with microglial TLR4 deletion to many other
neurobiology areas beyond obesity.
Age-related activation of microglia is observed in normal aging. In human autopsy tissue,
aged microglia were associated with a dystrophic morphology characterized as de-ramification,
shortened processes, and cytoplasmic fragmentation
398
. Besides morphological changes,
microglial functions in maintaining CNS homeostasis decline by age. For example, aged
microglia have been found to have increased expression of activation markers and levels of
inflammatory cytokines
399–402
. These findings were verified in a recent transcriptional study.
Aged brain was associated with expansions of two distinct clusters of microglial cells, with one
cluster characterized by upregulation of IL-1b and chemokines Ccl3 and Ccl4, and the other
cluster displayed enrichment in genes related to interferon-response
403
. Moreover, aged
microglia showed significantly lower motility in response to injury, and were associated with
more sustained activation at injury site
404
. Phagocytosis was also impaired in aged microglia.
Aged microglia showed elevated ROS generation at baseline and decreased phagocytosis after
115
stimulated with phorbol 12-myristate 13-acetate and ionomycin
405
. Evidence suggested that this
impaired phagocytic ability was highly relevant to inflammatory statues. In BV-2 cell cultures,
treating the microglia with LPS inhibited microglial phagocytosis of Aβ
406
. On the other hand,
treatment with anti-inflammatory cytokine IL-4 strongly blocked the LPS-induced inhibition of Aβ
phagocytosis, suggesting that increased inflammation limits the phagocytic ability of
microglia
406
.
Age-related microglial changes have also been identified in neurodegenerative diseases.
Taking AD as an example, microglia that surround Aβ plaques were associated with activated
morphology as well as increased production of proinflammatory and activation markers
including MHCII, TNFa, IL-1b, and IL-6
407
. Importantly, this phenotypic change of microglial
activation is age dependent. In primary microglia culture isolated from an AD mouse model
(APPswe/PS1dE9), there was a significant decrease in internalized Aβ42 in aged microglia
compared with young or neonatal microglia
402
. Using the same animal model, another group
reported that microglia shifted to a pro-inflammatory phenotype only at advanced stage of the
disease, demonstrated by increased expression of inflammatory cytokines and iNOS induced by
Aβ
408
.
It has also been reported that activated microglia contribute to cognitive impairment
associated with aging. In wild-type mice, age-related increases in production of IL-1β by
microglia was associated with impairments in synaptic plasticity and learning and memory
409
.
Inhibition of inflammatory responses in myeloid cells by suppressing prostaglandin E2 signaling
improved synaptic plasticity in the hippocampus and spatial memory in aged mice
410
. In 3xTg-
AD mice, treatment of low does PLX5622 induced a significant reduction in the number of
microglial cells associated with Aβ plaques, without affecting total Aβ or Tau levels in the
brain
411
. In addition, low does PLX5622 treatment improved cognitive performance in aged
3xTg-AD mice
411
.
116
Recent studies provided evidence suggesting TLR4 involvement in neurodegenerative
diseases
412
. The TLR4 SNP Asp299Gly has been found to significantly associated with sporadic
AD in an Italian population
413,414
. TLR4 expression was increased in an AD-transgenic mice
overexpressing APP
415
. In AD patients, TLR4 immunoreactivity was also increased in areas
adjacent to Aβ plaques
415
. Importantly, TLR4 signaling is required for the inflammatory
response induced by Aβ, as microglia isolated from mice with TLR4 mutations showed lower
expression levels of nitrite and IL-6 compared with wild-type mice after treatment with Aβ
peptides
415
. Likewise, central administration of oligomeric Aβ induced microglia activation
through TLR4 signaling
416
. In APPswe/PS1dE9 mice, TLR4 mutations significantly reduced
levels of inflammatory cytokines in the brain compared with wild type AD mice
417
. Together,
these finding support a detrimental role of TLR4 signaling in AD pathogenesis and cognitive
impairment.
On the other hand, there are other studies suggesting TLR4 activation can also be
neuroprotective. In AD transgenic mice TLR4 mutations did not affect Aβ deposition in the brain
at early age, but were associated with increased Aβ load
418,419
as well as stronger impairments
in cognitive performance as pathology progress
419
. In line with this finding, mild stimulation of
TLR4 by monophosphoryl lipid A, an agonist that is less pyrogenic than LPS, was associated
with improved cognitive function and decreased Aβ pathology in APP transgenic mice
420
. These
conflicting results highlight the complexity of TLR4 activation as it relates to AD pathology.
One possible explanation for these opposing results is that TLR4 signaling is altered
during different disease stages. Indeed, microglia phagocyte Aβ via TLR4 signaling, as
activation of TLR4 LPS increased Aβ uptake in cultured microglia
112
. Moreover, one study found
that microglia in APP/PS1 mice showed stronger TLR4 activation by LPS than wild type mice at
young age, while aged AD microglia displayed less TLR4 activation than aged wild type
microglia
421
. Together, it is possible that activation TLR4 signaling may help with Aβ clearance
117
at early stage of the diseases, however, continuous exposure of Aβ as diseases progress may
diminish microglial capability to clear Aβ and further induce neuronal loss.
Although age-associated alteration in microglial functions and TLR4 signaling have been
implied in normal aging as well as age-related neurodegenerative diseases, it is still unclear
whether they directly contribute to this process. Ongoing work in the lab is using our newly
generated mouse model to examine role of microglial TLR4 in age-related neural outcomes.
APOE4 is associated with altered immune responses that are TLR4-dependent
412,422
, however
the direct link between APOE and TLR4 in AD pathogenesis is still unknown. Therefore, it would
also be interesting to use our mouse model in the context of AD to examine TLR4 and APOE4
interactions in AD pathogenesis.

118
References

1. Heneka, M. T., Golenbock, D. T. & Latz, E. Innate immunity in Alzheimer’s disease. Nat
Immunol 16, 229–236 (2015).
2. Ennerfelt, H. E. & Lukens, J. R. The role of innate immunity in Alzheimer’s disease. Immunol
Rev 297, 225–246 (2020).
3. Page, A. L. et al. Role of the peripheral innate immune system in the development of
Alzheimer’s disease. Exp Gerontol 107, 59–66 (2018).
4. Pimenova, A. A., Raj, T. & Goate, A. M. Untangling Genetic Risk for Alzheimer’s Disease.
Biol Psychiat 83, 300–310 (2018).
5. Griciuc, A. & Tanzi, R. E. The role of innate immune genes in Alzheimer’s disease. Curr Opin
Neurol 34, 228–236 (2021).
6. Gjoneska, E. et al. Conserved epigenomic signals in mice and humans reveal immune basis
of Alzheimer’s disease. Nature 518, 365–369 (2015).
7. Wyss-Coray, T. & Rogers, J. Inflammation in Alzheimer Disease—A Brief Review of the Basic
Science and Clinical Literature. Csh Perspect Med 2, a006346 (2012).
8. Nagele, R. G. et al. Contribution of glial cells to the development of amyloid plaques in
Alzheimer’s disease. Neurobiol Aging 25, 663–674 (2004).
9. Serrano-Pozo, A. et al. Reactive Glia not only Associates with Plaques but also Parallels
Tangles in Alzheimer’s Disease. Am J Pathology 179, 1373–1384 (2011).
10. Minter, M. R., Taylor, J. M. & Crack, P. J. The contribution of neuroinflammation to amyloid
toxicity in Alzheimer’s disease. J Neurochem 136, 457–474 (2016).
11. Femminella, G. D. et al. Does Microglial Activation Influence Hippocampal Volume and
Neuronal Function in Alzheimer’s Disease and Parkinson’s Disease Dementia? J Alzheimer’s
Dis 51, 1275–1289 (2016).
12. Melah, K. E. et al. Cerebrospinal Fluid Markers of Alzheimer’s Disease Pathology and
Microglial Activation are Associated with Altered White Matter Microstructure in Asymptomatic
Adults at Risk for Alzheimer’s Disease. J Alzheimer’s Dis 50, 873–886 (2016).
13. Yokoi, T. et al. Involvement of the Precuneus/Posterior Cingulate Cortex Is Significant for
the Development of Alzheimer’s Disease: A PET (THK5351, PiB) and Resting fMRI Study. Front
Aging Neurosci 10, 304 (2018).
14. Edison, P. et al. Microglia, amyloid, and cognition in Alzheimer’s disease: An
[11C](R)PK11195-PET and [11C]PIB-PET study. Neurobiol Dis 32, 412–419 (2008).
119
15. Colonna, M. & Butovsky, O. Microglia Function in the Central Nervous System During
Health and Neurodegeneration. Annu Rev Immunol 35, 1–28 (2016).
16. Liddelow, S. A. et al. Neurotoxic reactive astrocytes are induced by activated microglia.
Nature 541, 481–487 (2017).
17. Chakrabarty, P. et al. Massive gliosis induced by interleukin-6 suppresses Abeta deposition
in vivo: evidence against inflammation as a driving force for amyloid deposition. Faseb J Official
Publ Fed Am Soc Exp Biology 24, 548–59 (2009).
18. Chakrabarty, P., Herring, A., Ceballos-Diaz, C., Das, P. & Golde, T. E. Hippocampal
expression of murine TNFα results in attenuation of amyloid deposition in vivo. Mol
Neurodegener 6, 16 (2011).
19. Guillot-Sestier, M.-V. et al. Il10 Deficiency Rebalances Innate Immunity to Mitigate
Alzheimer-Like Pathology. Neuron 85, 534–548 (2015).
20. Qiao, X., Cummins, D. J. & Paul, S. M. Neuroinflammation-induced acceleration of amyloid
deposition in the APPV717F transgenic mouse. Eur J Neurosci 14, 474–482 (2001).
21. Kiyota, T. et al. CNS expression of anti-inflammatory cytokine interleukin-4 attenuates
Alzheimer’s disease-like pathogenesis in APP+PS1 bigenic mice. Faseb J 24, 3093–3102
(2010).
22. Bu, G. Apolipoprotein E and its receptors in Alzheimer’s disease: pathways, pathogenesis
and therapy. Nat Rev Neurosci 10, 333–344 (2009).
23. Holtzman, D. M., Herz, J. & Bu, G. Apolipoprotein E and Apolipoprotein E Receptors:
Normal Biology and Roles in Alzheimer Disease. Csh Perspect Med 2, a006312 (2012).
24. Liu, C.-C., Liu, C.-C., Kanekiyo, T., Xu, H. & Bu, G. Apolipoprotein E and Alzheimer disease:
risk, mechanisms and therapy. Nat Rev Neurol 9, 106–118 (2013).
25. Belinson, H. & Michaelson, D. M. ApoE4-dependent Abeta-mediated neurodegeneration is
associated with inflammatory activation in the hippocampus but not the septum. J Neural
Transm Vienna Austria 1996 116, 1427–34 (2009).
26. Youmans, K. L. et al. APOE4-specific Changes in Aβ Accumulation in a New Transgenic
Mouse Model of Alzheimer Disease*. J Biol Chem 287, 41774–41786 (2012).
27. Parhizkar, S. & Holtzman, D. M. APOE mediated neuroinflammation and neurodegeneration
in Alzheimer’s disease. Semin Immunol 101594 (2022) doi:10.1016/j.smim.2022.101594.
28. Gale, S. C. et al. APOε4 is associated with enhanced in vivo innate immune responses in
human subjects. J Allergy Clin Immun 134, 127-134.e9 (2014).
29. Jofre-Monseny, L. et al. Effects of apoE genotype on macrophage inflammation and heme
oxygenase-1 expression. Biochem Bioph Res Co 357, 319–324 (2007).
120
30. Lanfranco, M. F., Sepulveda, J., Kopetsky, G. & Rebeck, G. W. Expression and secretion of
apoE isoforms in astrocytes and microglia during inflammation. Glia 69, 1478–1493 (2021).
31. Mhatre-Winters, I., Eid, A., Han, Y., Tieu, K. & Richardson, J. R. Sex and APOE Genotype
Alter the Basal and Induced Inflammatory States of Primary Microglia from APOE Targeted
Replacement Mice. Int J Mol Sci 23, 9829 (2022).
32. Maezawa, I., Nivison, M., Montine, K. S., Maeda, N. & Montine, T. J. Neurotoxicity from
innate immune response is greatest with targeted replacement of ε4 allele of apolipoprotein E
gene and is mediated by microglial p38MAPK. Faseb J 20, 797–799 (2006).
33. Lynch, J. R. et al. APOE Genotype and an ApoE-mimetic Peptide Modify the Systemic and
Central Nervous System Inflammatory Response*. J Biol Chem 278, 48529–48533 (2003).
34. Ophir, G. et al. Apolipoprotein E4 enhances brain inflammation by modulation of the NF-κB
signaling cascade. Neurobiol Dis 20, 709–718 (2005).
35. Zhu, Y. et al. APOE genotype alters glial activation and loss of synaptic markers in mice.
Glia 60, 559–569 (2012).
36. Abarca-Gómez, L. et al. Worldwide trends in body-mass index, underweight, overweight,
and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement
studies in 128·9 million children, adolescents, and adults. Lancet 390, 2627–2642 (2017).
37. McGill, A.-T. Causes of metabolic syndrome and obesity-related co-morbidities Part 1: A
composite unifying theory review of human-specific co-adaptations to brain energy
consumption. Archives Public Heal 72, 30 (2014).
38. Bray, G. A. Obesity increases risk for diabetes. Int J Obes Relat Metabolic Disord J Int
Assoc Study Obes 16 Suppl 4, S13-7 (1992).
39. Rocha, V. Z. & Libby, P. Obesity, inflammation, and atherosclerosis. Nat Rev Cardiol 6,
399–409 (2009).
40. Bianchini, F., Kaaks, R. & Vainio, H. Overweight, obesity, and cancer risk. Lancet Oncol 3,
565–574 (2002).
41. Isaac, V. et al. Adverse Associations between Visceral Adiposity, Brain Structure, and
Cognitive Performance in Healthy Elderly. Front Aging Neurosci 3, 12 (2011).
42. Gupta, A. et al. Patterns of brain structural connectivity differentiate normal weight from
overweight subjects. Neuroimage Clin 7, 506–517 (2015).
43. Kullmann, S., Schweizer, F., Veit, R., Fritsche, A. & Preissl, H. Compromised white matter
integrity in obesity. Obes Rev Official J Int Assoc Study Obes 16, 273–81 (2014).
44. Tucsek, Z. et al. Obesity in Aging Exacerbates Blood–Brain Barrier Disruption,
Neuroinflammation, and Oxidative Stress in the Mouse Hippocampus: Effects on Expression of
121
Genes Involved in Beta-Amyloid Generation and Alzheimer’s Disease. Journals Gerontology
Ser 69, 1212–1226 (2014).
45. O’Brien, P. D., Hinder, L. M., Callaghan, B. C. & Feldman, E. L. Neurological consequences
of obesity. Lancet Neurology 16, 465–477 (2017).
46. Christensen, A. & Pike, C. J. Menopause, obesity and inflammation: interactive risk factors
for Alzheimer’s disease. Front Aging Neurosci 7, 130 (2015).
47. Moser, V. A. & Pike, C. J. Obesity and sex interact in the regulation of Alzheimer’s disease.
Neurosci Biobehav Rev 67, 102–118 (2016).
48. Hotamisligil, G. S. Inflammation and metabolic disorders. Nature 444, 860–867 (2006).
49. Yaffe, K. et al. The Metabolic Syndrome, Inflammation, and Risk of Cognitive Decline. Jama
292, 2237–2242 (2004).
50. Ekdahl, C. T., Claasen, J.-H., Bonde, S., Kokaia, Z. & Lindvall, O. Inflammation is
detrimental for neurogenesis in adult brain. Proc National Acad Sci 100, 13632–13637 (2003).
51. Bischof, G. N. & Park, D. C. Obesity and Aging: Consequences for Cognition, Brain
Structure, and Brain Function. Psychosom Med 77, 697–709 (2015).
52. Gregor, M. F. & Hotamisligil, G. S. Inflammatory Mechanisms in Obesity. Annu Rev Immunol
29, 415–445 (2011).
53. Berg, A. H. & Scherer, P. E. Adipose Tissue, Inflammation, and Cardiovascular Disease.
Circ Res 96, 939–949 (2005).
54. Yao, L., Herlea-Pana, O., Heuser-Baker, J., Chen, Y. & Barlic-Dicen, J. Roles of the
Chemokine System in Development of Obesity, Insulin Resistance, and Cardiovascular
Disease. J Immunol Res 2014, 181450 (2014).
55. Tandon, P., Wafer, R., Minchin, J. E. N., Suarez, R. K. & Hoppeler, H. H. Adipose
morphology and metabolic disease. J Exp Biol 221, jeb164970 (2018).
56. Weisberg, S. P. et al. Obesity is associated with macrophage accumulation in adipose
tissue. J Clin Invest 112, 1796–1808 (2003).
57. Xu, H. et al. Chronic inflammation in fat plays a crucial role in the development of obesity-
related insulin resistance. J Clin Invest 112, 1821–1830 (2003).
58. Chen, A. et al. Diet Induction of Monocyte Chemoattractant Protein-1 and its Impact on
Obesity. Obes Res 13, 1311–1320 (2005).
59. Weisberg, S. P. et al. CCR2 modulates inflammatory and metabolic effects of high-fat
feeding. J Clin Invest 116, 115–124 (2006).
122
60. Amano, S. U. et al. Local Proliferation of Macrophages Contributes to Obesity-Associated
Adipose Tissue Inflammation. Cell Metab 19, 162–171 (2014).
61. Hotamisligil, G. S., Murray, D. L., Choy, L. N. & Spiegelman, B. M. Tumor necrosis factor
alpha inhibits signaling from the insulin receptor. Proc National Acad Sci 91, 4854–4858 (1994).
62. Engelman, J. A., Berg, A. H., Lewis, R. Y., Lisanti, M. P. & Scherer, P. E. Tumor necrosis
factor alpha-mediated insulin resistance, but not dedifferentiation, is abrogated by MEK1/2
inhibitors in 3T3-L1 adipocytes. Mol Endocrinol Baltim Md 14, 1557–69 (2000).
63. Uysal, K. T., Wiesbrock, S. M., Marino, M. W. & Hotamisligil, G. S. Protection from obesity-
induced insulin resistance in mice lacking TNF-α function. Nature 389, 610–614 (1997).
64. Uysal, K. T., Wiesbrock, S. M. & Hotamisligil, G. S. Functional Analysis of Tumor Necrosis
Factor (TNF) Receptors in TNF-α-Mediated Insulin Resistance in Genetic Obesity**This work is
supported in part by a grant from the NIH (DK-52539). Endocrinology 139, 4832–4838 (1998).
65. Hirosumi, J. et al. A central role for JNK in obesity and insulin resistance. Nature 420, 333–
336 (2002).
66. Chiang, S.-H. et al. The protein kinase IKKepsilon regulates energy balance in obese mice.
Cell 138, 961–75 (2009).
67. Himes, R. W. & Smith, C. W. Tlr2 is critical for diet-induced metabolic syndrome in a murine
model. Faseb J 24, 731–739 (2010).
68. Patsouris, D. et al. Ablation of CD11c-Positive Cells Normalizes Insulin Sensitivity in Obese
Insulin Resistant Animals. Cell Metab 8, 301–309 (2008).
69. Arkan, M. C. et al. IKK-β links inflammation to obesity-induced insulin resistance. Nat Med
11, 191–198 (2005).
70. Morton, G. J., Cummings, D. E., Baskin, D. G., Barsh, G. S. & Schwartz, M. W. Central
nervous system control of food intake and body weight. Nature 443, 289–295 (2006).
71. Sandoval, D., Cota, D. & Seeley, R. J. The Integrative Role of CNS Fuel-Sensing
Mechanisms in Energy Balance and Glucose Regulation. Annu Rev Physiol 70, 513–535
(2008).
72. Souza, C. T. D. et al. Consumption of a fat-rich diet activates a proinflammatory response
and induces insulin resistance in the hypothalamus. Endocrinology 146, 4192–9 (2005).
73. Mori, M. A. et al. A Systems Biology Approach Identifies Inflammatory Abnormalities
Between Mouse Strains Prior to Development of Metabolic Disease. Diabetes 59, 2960–2971
(2010).
74. Thaler, J. P. et al. Obesity is associated with hypothalamic injury in rodents and humans. J
Clin Invest 122, 153–162 (2012).
123
75. Zhang, X. et al. Hypothalamic IKKβ/NF-κB and ER Stress Link Overnutrition to Energy
Imbalance and Obesity. Cell 135, 61–73 (2008).
76. Milanski, M. et al. Saturated Fatty Acids Produce an Inflammatory Response Predominantly
through the Activation of TLR4 Signaling in Hypothalamus: Implications for the Pathogenesis of
Obesity. J Neurosci 29, 359–370 (2009).
77. Wang, X. et al. Increased Hypothalamic Inflammation Associated with the Susceptibility to
Obesity in Rats Exposed to High-Fat Diet. Exp Diabetes Res 2012, 847246 (2012).
78. Reis, W. L., Yi, C.-X., Gao, Y., Tschöp, M. H. & Stern, J. E. Brain Innate Immunity Regulates
Hypothalamic Arcuate Neuronal Activity and Feeding Behavior. Endocrinology 156, 1303–1315
(2015).
79. Degasperi, G. R. et al. UCP2 protects hypothalamic cells from TNF-α-induced damage.
Febs Lett 582, 3103–3110 (2008).
80. Arruda, A. P. et al. Low-Grade Hypothalamic Inflammation Leads to Defective
Thermogenesis, Insulin Resistance, and Impaired Insulin Secretion. J Clin Endocrinol
Metabolism 96, 869–869 (2011).
81. Purkayastha, S. et al. Neural dysregulation of peripheral insulin action and blood pressure
by brain endoplasmic reticulum stress. Proc National Acad Sci 108, 2939–2944 (2011).
82. Lu, J. et al. Ursolic acid improves high fat diet-induced cognitive impairments by blocking
endoplasmic reticulum stress and IκB kinase β/nuclear factor-κB-mediated inflammatory
pathways in mice. Brain Behav Immun 25, 1658–1667 (2011).
83. Dinel, A.-L. et al. Cognitive and Emotional Alterations Are Related to Hippocampal
Inflammation in a Mouse Model of Metabolic Syndrome. Plos One 6, e24325 (2011).
84. Jeon, B. T. et al. Resveratrol Attenuates Obesity-Associated Peripheral and Central
Inflammation and Improves Memory Deficit in Mice Fed a High-Fat Diet. Diabetes 61, 1444–
1454 (2012).
85. MIAO, Y. et al. α-lipoic acid attenuates obesity-associated hippocampal neuroinflammation
and increases the levels of brain-derived neurotrophic factor in ovariectomized rats fed a high-
fat diet. Int J Mol Med 32, 1179–1186 (2013).
86. Kang, E. -B. et al. Neuroprotective Effects of Endurance Exercise Against High-Fat Diet-
Induced Hippocampal Neuroinflammation. J Neuroendocrinol 28, (2016).
87. Cai, M. et al. The signaling mechanisms of hippocampal endoplasmic reticulum stress
affecting neuronal plasticity-related protein levels in high fat diet-induced obese rats and the
regulation of aerobic exercise. Brain Behav Immun 57, 347–359 (2016).
88. Murata, Y. et al. A high fat diet-induced decrease in hippocampal newly-born neurons of
male mice is exacerbated by mild psychological stress using a Communication Box. J Affect
Disorders 209, 209–216 (2017).
124
89. Beilharz, J. E., Kaakoush, N. O., Maniam, J. & Morris, M. J. The effect of short-term
exposure to energy-matched diets enriched in fat or sugar on memory, gut microbiota and
markers of brain inflammation and plasticity. Brain Behav Immun 57, 304–313 (2016).
90. Hao, S., Dey, A., Yu, X. & Stranahan, A. M. Dietary obesity reversibly induces synaptic
stripping by microglia and impairs hippocampal plasticity. Brain Behav Immun 51, 230–239
(2016).
91. Samaras, K. & Sachdev, P. S. Diabetes and the elderly brain: sweet memories? Ther Adv
Endocrinol Metabolism 3, 189–196 (2012).
92. Yau, P. L., Castro, M. G., Tagani, A., Tsui, W. H. & Convit, A. Obesity and Metabolic
Syndrome and Functional and Structural Brain Impairments in Adolescence. Pediatrics 130,
e856–e864 (2012).
93. Liang, J., Matheson, B. E., Kaye, W. H. & Boutelle, K. N. Neurocognitive correlates of
obesity and obesity-related behaviors in children and adolescents. Int J Obesity 38, 494–506
(2014).
94. Hassing, L. B., Dahl, A. K., Pedersen, N. L. & Johansson, B. Overweight in Midlife Is
Related to Lower Cognitive Function 30 Years Later: A Prospective Study with Longitudinal
Assessments. Dement Geriatr Cogn 29, 543–552 (2010).
95. Dahl, A. K. et al. Body mass index across midlife and cognitive change in late life. Int J
Obesity 37, 296–302 (2013).
96. Elias, M. F., Elias, P. K., Sullivan, L. M., Wolf, P. A. & D’Agostino, R. B. Obesity, diabetes
and cognitive deficit: The Framingham Heart Study. Neurobiol Aging 26, 11–16 (2005).
97. Taki, Y. et al. Relationship Between Body Mass Index and Gray Matter Volume in 1,428
Healthy Individuals. Obesity 16, 119–124 (2008).
98. Anstey, K. J., Cherbuin, N., Budge, M. & Young, J. Body mass index in midlife and late-life
as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 12, e426–e437
(2011).
99. Mrak, R. E. Alzheimer-type neuropathological changes in morbidly obese elderly individuals.
Clin Neuropathol 28, 40–45 (2009).
100. McNeilly, A. D., Williamson, R., Sutherland, C., Balfour, D. J. K. & Stewart, C. A. High fat
feeding promotes simultaneous decline in insulin sensitivity and cognitive performance in a
delayed matching and non-matching to position task. Behav Brain Res 217, 134–141 (2011).
101. Alzoubi, K. H., Khabour, O. F., Salah, H. A. & Hasan, Z. Vitamin E prevents high-fat high-
carbohydrates diet-induced memory impairment: The role of oxidative stress. Physiol Behav
119, 72–78 (2013).
125
102. Klaus, F. et al. Differential effects of peripheral and brain tumor necrosis factor on
inflammation, sickness, emotional behavior and memory in mice. Brain Behav Immun 58, 310–
326 (2016).
103. Dutheil, S., Ota, K. T., Wohleb, E. S., Rasmussen, K. & Duman, R. S. High-Fat Diet
Induced Anxiety and Anhedonia: Impact on Brain Homeostasis and Inflammation.
Neuropsychopharmacol 41, 1874–1887 (2016).
104. Lawson, L. J., Perry, V. H., Dri, P. & Gordon, S. Heterogeneity in the distribution and
morphology of microglia in the normal adult mouse brain. Neuroscience 39, 151–170 (1990).
105. Hanisch, U.-K. & Kettenmann, H. Microglia: active sensor and versatile effector cells in the
normal and pathologic brain. Nat Neurosci 10, 1387–1394 (2007).
106. Béchade, C., Cantaut-Belarif, Y. & Bessis, A. Microglial control of neuronal activity. Front
Cell Neurosci 7, 32 (2013).
107. Schafer, D. P. et al. Microglia Sculpt Postnatal Neural Circuits in an Activity and
Complement-Dependent Manner. Neuron 74, 691–705 (2012).
108. Nimmerjahn, A., Kirchhoff, F. & Helmchen, F. Resting Microglial Cells Are Highly Dynamic
Surveillants of Brain Parenchyma in Vivo. Science 308, 1314–1318 (2005).
109. Ladeby, R. et al. Microglial cell population dynamics in the injured adult central nervous
system. Brain Res Rev 48, 196–206 (2005).
110. Beynon, S. B. & Walker, F. R. Microglial activation in the injured and healthy brain: What
are we really talking about? Practical and theoretical issues associated with the measurement of
changes in microglial morphology. Neuroscience 225, 162–171 (2012).
111. Ziebell, J. M., Adelson, P. D. & Lifshitz, J. Microglia: dismantling and rebuilding circuits
after acute neurological injury. Metab Brain Dis 30, 393–400 (2015).
112. Hickman, S. E. et al. The Microglial Sensome Revealed by Direct RNA Sequencing. Nat
Neurosci 16, 1896–1905 (2013).
113. Gosselin, D. et al. Environment Drives Selection and Function of Enhancers Controlling
Tissue-Specific Macrophage Identities. Cell 159, 1327–1340 (2014).
114. Zhang, Y. et al. An RNA-Sequencing Transcriptome and Splicing Database of Glia,
Neurons, and Vascular Cells of the Cerebral Cortex. J Neurosci 34, 11929–11947 (2014).
115. Wes, P. D., Holtman, I. R., Boddeke, E. W. G. M., Möller, T. & Eggen, B. J. L. Next
generation transcriptomics and genomics elucidate biological complexity of microglia in health
and disease. Glia 64, 197–213 (2016).
116. Masuda, T., Sankowski, R., Staszewski, O. & Prinz, M. Microglia Heterogeneity in the
Single-Cell Era. Cell Reports 30, 1271–1281 (2020).
126
117. Wang, H. Microglia Heterogeneity in Alzheimer’s Disease: Insights From Single-Cell
Technologies. Frontiers Synaptic Neurosci 13, 773590 (2021).
118. Duffy, C. M. et al. Role of orexin A signaling in dietary palmitic acid-activated microglial
cells. Neurosci Lett 606, 140–144 (2015).
119. Valdearcos, M. et al. Microglia Dictate the Impact of Saturated Fat Consumption on
Hypothalamic Inflammation and Neuronal Function. Cell Reports 9, 2124–2138 (2014).
120. Milanova, I. V., Correa-da-Silva, F., Kalsbeek, A. & Yi, C.-X. Mapping of Microglial Brain
Region, Sex and Age Heterogeneity in Obesity. Int J Mol Sci 22, 3141 (2021).
121. André, C. et al. Inhibiting Microglia Expansion Prevents Diet-Induced Hypothalamic and
Peripheral Inflammation. Diabetes 66, 908–919 (2016).
122. Erion, J. R. et al. Obesity Elicits Interleukin 1-Mediated Deficits in Hippocampal Synaptic
Plasticity. J Neurosci 34, 2618–2631 (2014).
123. Cope, E. C. et al. Microglia Play an Active Role in Obesity-Associated Cognitive Decline. J
Neurosci 38, 8889–8904 (2018).
124. Crack, P. J. & Bray, P. J. Toll-like receptors in the brain and their potential roles in
neuropathology. Immunol Cell Biol 85, 476–480 (2007).
125. Vaure, C. & Liu, Y. A Comparative Review of Toll-Like Receptor 4 Expression and
Functionality in Different Animal Species. Front Immunol 5, 316 (2014).
126. Kawasaki, T. & Kawai, T. Toll-Like Receptor Signaling Pathways. Front Immunol 5, 461
(2014).
127. Lee, J. Y., Sohn, K. H., Rhee, S. H. & Hwang, D. Saturated Fatty Acids, but Not
Unsaturated Fatty Acids, Induce the Expression of Cyclooxygenase-2 Mediated through Toll-like
Receptor 4*. J Biol Chem 276, 16683–16689 (2001).
128. Lancaster, G. I. et al. Evidence that TLR4 Is Not a Receptor for Saturated Fatty Acids but
Mediates Lipid-Induced Inflammation by Reprogramming Macrophage Metabolism. Cell Metab
27, 1096-1110.e5 (2018).
129. Fessler, M. B., Rudel, L. L. & Brown, J. M. Toll-like receptor signaling links dietary fatty
acids to the metabolic syndrome. Curr Opin Lipidol 20, 379–385 (2009).
130. Shi, H. et al. TLR4 links innate immunity and fatty acid–induced insulin resistance. J Clin
Invest 116, 3015–3025 (2006).
131. Song, M. J., Kim, K. H., Yoon, J. M. & Kim, J. B. Activation of Toll-like receptor 4 is
associated with insulin resistance in adipocytes. Biochem Bioph Res Co 346, 739–745 (2006).
132. Moraes, J. C. et al. High-Fat Diet Induces Apoptosis of Hypothalamic Neurons. Plos One 4,
e5045 (2009).
127
133. Poggi, M. et al. C3H/HeJ mice carrying a toll-like receptor 4 mutation are protected against
the development of insulin resistance in white adipose tissue in response to a high-fat diet.
Diabetologia 50, 1267–1276 (2007).
134. Jia, L. et al. Hepatocyte Toll-like Receptor 4 Regulates Obesity-Induced Inflammation and
Insulin Resistance. Nat Commun 5, 3878–3878 (2014).
135. Tsukumo, D. M. L. et al. Loss-of-function mutation in Toll-like receptor 4 prevents diet-
induced obesity and insulin resistance. Diabetes 56, 1986–98 (2007).
136. Saberi, M. et al. Hematopoietic Cell-Specific Deletion of Toll-like Receptor 4 Ameliorates
Hepatic and Adipose Tissue Insulin Resistance in High-Fat-Fed Mice. Cell Metab 10, 419–429
(2009).
137. Moser, V. A., Uchoa, M. F. & Pike, C. J. TLR4 inhibitor TAK-242 attenuates the adverse
neural effects of diet-induced obesity. J Neuroinflamm 15, 306 (2018).
138. Zhang, N. et al. Pharmacological TLR4 Inhibition Protects against Acute and Chronic Fat-
Induced Insulin Resistance in Rats. Plos One 10, e0132575 (2015).
139. Obadia, N. et al. TLR4 mutation protects neurovascular function and cognitive decline in
high-fat diet-fed mice. J Neuroinflamm 19, 104 (2022).
140. Lehnardt, S. et al. Activation of innate immunity in the CNS triggers neurodegeneration
through a Toll-like receptor 4-dependent pathway. Proc National Acad Sci 100, 8514–8519
(2003).
141. Wang, Z. et al. Saturated fatty acids activate microglia via Toll-like receptor 4/NF-κB
signalling. Brit J Nutr 107, 229–241 (2012).
142. Choi, S. J., Kim, F., Schwartz, M. W. & Wisse, B. E. Cultured hypothalamic neurons are
resistant to inflammation and insulin resistance induced by saturated fatty acids. Am J Physiol-
endoc M 298, E1122–E1130 (2010).
143. Kleinridders, A. et al. MyD88 Signaling in the CNS Is Required for Development of Fatty
Acid-Induced Leptin Resistance and Diet-Induced Obesity. Cell Metab 10, 249–259 (2009).
144. Benzler, J. et al. Central Inhibition of IKKβ/NF-κB Signaling Attenuates High-Fat Diet–
Induced Obesity and Glucose Intolerance. Diabetes 64, 2015–2027 (2015).
145. Valdearcos, M. et al. Microglial Inflammatory Signaling Orchestrates the Hypothalamic
Immune Response to Dietary Excess and Mediates Obesity Susceptibility. Cell Metab 26, 185-
197.e3 (2017).
146. Li, J., Tang, Y. & Cai, D. IKKβ/NF-κB Disrupts Adult Hypothalamic Neural Stem Cells to
Mediate Neurodegenerative Mechanism of Dietary Obesity and Pre-Diabetes. Nat Cell Biol 14,
999–1012 (2012).
128
147. Vitek, M. P., Brown, C. M. & Colton, C. A. APOE genotype-specific differences in the
innate immune response. Neurobiol Aging 30, 1350–1360 (2009).
148. Manly, J. J. et al. Endogenous estrogen levels and Alzheimer’s disease among
postmenopausal women. Neurology 54, 833–837 (2000).
149. Pike, C. J. Sex and the development of Alzheimer’s disease. J Neurosci Res 95, 671–680
(2016).
150. Uchoa, M. F., Moser, V. A. & Pike, C. J. Interactions between inflammation, sex steroids,
and Alzheimer’s disease risk factors. Front Neuroendocrin 43, 60–82 (2016).
151. Mendes, N. F., Kim, Y.-B., Velloso, L. A. & Araújo, E. P. Hypothalamic Microglial Activation
in Obesity: A Mini-Review. Front Neurosci-switz 12, 846 (2018).
152. Jialal, I., Huet, B. A., Kaur, H., Chien, A. & Devaraj, S. Increased Toll-Like Receptor
Activity in Patients With Metabolic Syndrome. Diabetes Care 35, 900–904 (2012).
153. Ahmad, R. et al. Elevated expression of the toll like receptors 2 and 4 in obese individuals:
its significance for obesity-induced inflammation. J Inflamm Lond Engl 9, 48–48 (2012).
154. Rehli, M. Of mice and men: species variations of Toll-like receptor expression. Trends
Immunol 23, 375–378 (2002).
155. Mandrekar-Colucci, S. & Landreth, G. E. Microglia and Inflammation in Alzheimers
Disease. Cns Neurological Disord - Drug Targets 9, 156–167 (2010).
156. Yang, Q., Wang, G. & Zhang, F. Role of Peripheral Immune Cells-Mediated Inflammation
on the Process of Neurodegenerative Diseases. Front Immunol 11, 582825 (2020).
157. Calvo-Rodriguez, M., García-Rodríguez, C., Villalobos, C. & Núñez, L. Role of Toll Like
Receptor 4 in Alzheimer’s Disease. Front Immunol 11, 1588 (2020).
158. Cash, J. G. et al. Apolipoprotein E4 Impairs Macrophage Efferocytosis and Potentiates
Apoptosis by Accelerating Endoplasmic Reticulum Stress*. J Biol Chem 287, 27876–27884
(2012).
159. Garcia, A. R. et al. APOE4 is associated with elevated blood lipids and lower levels of
innate immune biomarkers in a tropical Amerindian subsistence population. Elife 10, e68231
(2021).
160. Straub, R. H. The Complex Role of Estrogens in Inflammation. Endocr Rev 28, 521–574
(2007).
161. Abu-Taha, M. et al. Menopause and Ovariectomy Cause a Low Grade of Systemic
Inflammation that May Be Prevented by Chronic Treatment with Low Doses of Estrogen or
Losartan. J Immunol 183, 1393–1402 (2009).
129
162. Vegeto, E. et al. Estrogen receptor-alpha mediates the brain antiinflammatory activity of
estradiol. P Natl Acad Sci Usa 100, 9614–9 (2003).
163. Vegeto, E. et al. Estrogen Prevents the Lipopolysaccharide-Induced Inflammatory
Response in Microglia. J Neurosci 21, 1809–1818 (2001).
164. Tenenbaum, M., Azab, A. N. & Kaplanski, J. Effects of estrogen against LPS-induced
inflammation and toxicity in primary rat glial and neuronal cultures. J Endotoxin Res 13, 158–
166 (2007).
165. Holland, D., Desikan, R. S., Dale, A. M. & McEvoy, L. K. Higher Rates of Decline for
Women and Apolipoprotein E ε4 Carriers. Am J Neuroradiol 34, 2287–2293 (2013).
166. Wang, X. et al. Sex Difference in the Association of APOE4 with Memory Decline in Mild
Cognitive Impairment. J Alzheimer’s Dis 69, 1161–1169 (2019).
167. Beydoun, M. A. et al. Sex differences in the association of the apolipoprotein E epsilon 4
allele with incidence of dementia, cognitive impairment, and decline. Neurobiol Aging 33, 720-
731.e4 (2012).
168. Hsu, M., Dedhia, M., Crusio, W. E. & Delprato, A. Sex differences in gene expression
patterns associated with the APOE4 allele. F1000research 8, 387 (2019).
169. Brown, C. M., Choi, E., Xu, Q., Vitek, M. P. & Colton, C. A. The APOE4 genotype alters the
response of microglia and macrophages to 17β-estradiol. Neurobiol Aging 29, 1783–1794
(2008).
170. Korach, K. S. & McLachlan, J. A. Techniques for detection of estrogenicity. Environ Health
Persp 103, 5–8 (1995).
171. Skarda, J. Sensitivity and specificity of bioassay of estrogenicity on mammary gland and
uterus of female mice. Physiol Res 51, 407–12 (2002).
172. Toapanta, F. R. & Ross, T. M. Impaired immune responses in the lungs of aged mice
following influenza infection. Respir Res 10, 112 (2009).
173. Silverman, M. N. et al. Glucocorticoid receptor dimerization is required for proper recovery
of LPS-induced inflammation, sickness behavior and metabolism in mice. Mol Psychiatr 18,
1006–1017 (2013).
174. Clark, S. M. et al. Dissociation between sickness behavior and emotionality during
lipopolysaccharide challenge in lymphocyte deficient Rag2−/− mice. Behav Brain Res 278, 74–
82 (2015).
175. Hart, B. L. Biological basis of the behavior of sick animals. Neurosci Biobehav Rev 12,
123–137 (1988).
176. Dantzer, R. et al. Molecular basis of sickness behavior. Ann Ny Acad Sci 856, 132–8
(1998).
130
177. Biesmans, S. et al. Systemic Immune Activation Leads to Neuroinflammation and Sickness
Behavior in Mice. Mediat Inflamm 2013, 271359 (2013).
178. Dantzer, R. Cytokine-Induced Sickness Behavior: Mechanisms and Implications. Ann Ny
Acad Sci 933, 222–234 (2001).
179. Odoj, K. et al. In vivo mechanisms of cortical network dysfunction induced by systemic
inflammation. Brain Behav Immun 96, 113–126 (2021).
180. Bluthé, R. M. et al. Synergy between tumor necrosis factor α and interleukin-1 in the
induction of sickness behavior in mice. Psychoneuroendocrino 19, 197–207 (1994).
181. Bluthé, R. et al. Role of interleukin-1β and tumour necrosis factor-α in lipopolysaccharide-
induced sickness behaviour: a study with interleukin-1 type I receptor-deficient mice. Eur J
Neurosci 12, 4447–4456 (2000).
182. Ilnytska, O. & Argyropoulos, G. The Role of the Agouti-Related Protein in Energy Balance
Regulation. Cell Mol Life Sci 65, 2721 (2008).
183. Scarlett, J. M. et al. Regulation of Agouti-Related Protein Messenger Ribonucleic Acid
Transcription and Peptide Secretion by Acute and Chronic Inflammation. Endocrinology 149,
4837–4845 (2008).
184. Almeida, M. C., Steiner, A. A., Branco, L. G. S. & Romanovsky, A. A. Cold-seeking
behavior as a thermoregulatory strategy in systemic inflammation. Eur J Neurosci 23, 3359–
3367 (2006).
185. Wanner, S. P. et al. Lipopolysaccharide-Induced Neuronal Activation in the Paraventricular
and Dorsomedial Hypothalamus Depends on Ambient Temperature. Plos One 8, e75733
(2013).
186. Rudaya, A. Y., Steiner, A. A., Robbins, J. R., Dragic, A. S. & Romanovsky, A. A.
Thermoregulatory responses to lipopolysaccharide in the mouse: dependence on the dose and
ambient temperature. Am J Physiology-regulatory Integr Comp Physiology 289, R1244–R1252
(2005).
187. Gordon, C. J. Thermal physiology of laboratory mice: Defining thermoneutrality. J Therm
Biol 37, 654–685 (2012).
188. Gordon, C. J. Thermal biology of the laboratory rat. Physiol Behav 47, 963–991 (1990).
189. Romanovsky, A. A., Shido, O., Sakurada, S., Sugimoto, N. & Nagasaka, T. Endotoxin
shock-associated hypothermia. How and why does it occur? Ann Ny Acad Sci 813, 733–7
(1997).
190. Liu, E. et al. Naturally occurring hypothermia is more advantageous than fever in severe
forms of lipopolysaccharide- and Escherichia coli-induced systemic inflammation. Am J
Physiology-regulatory Integr Comp Physiology 302, R1372–R1383 (2012).
131
191. Ganeshan, K. et al. Energetic Trade-Offs and Hypometabolic States Promote Disease
Tolerance. Cell 177, 399-413.e12 (2019).
192. Chow, J. C., Young, D. W., Golenbock, D. T., Christ, W. J. & Gusovsky, F. Toll-like
Receptor-4 Mediates Lipopolysaccharide-induced Signal Transduction*. J Biol Chem 274,
10689–10692 (1999).
193. Raetz, C. R. H. & Whitfield, C. LIPOPOLYSACCHARIDE ENDOTOXINS. Biochemistry-us
71, 635–700 (2002).
194. Godowski, P. J. A smooth operator for LPS responses. Nat Immunol 6, 544–546 (2005).
195. Marottoli, F. M. et al. Peripheral Inflammation, Apolipoprotein E4, and Amyloid-β Interact to
Induce Cognitive and Cerebrovascular Dysfunction. Asn Neuro 9, 1759091417719201 (2017).
196. Schmich, K. et al. Tumor necrosis factor α sensitizes primary murine hepatocytes to
Fas/CD95-induced apoptosis in a Bim- and Bid-dependent manner. Hepatology 53, 282–292
(2011).
197. Faletti, L. et al. TNFα sensitizes hepatocytes to FasL-induced apoptosis by NFκB-mediated
Fas upregulation. Cell Death Dis 9, 909 (2018).
198. Töllner, B. et al. The role of tumor necrosis factor (TNF) in the febrile and metabolic
responses of rats to intraperitoneal injection of a high dose of lipopolysaccharide. Pflügers
Archiv 440, 925–932 (2000).
199. Singh, A. K. & Jiang, Y. How does peripheral lipopolysaccharide induce gene expression in
the brain of rats? Toxicology 201, 197–207 (2004).
200. Banks, W. A. & Robinson, S. M. Minimal penetration of lipopolysaccharide across the
murine blood–brain barrier. Brain Behav Immun 24, 102–109 (2010).
201. Vargas-Caraveo, A. et al. Lipopolysaccharide enters the rat brain by a lipoprotein-mediated
transport mechanism in physiological conditions. Sci Rep-uk 7, 13113 (2017).
202. Qin, L. et al. Systemic LPS causes chronic neuroinflammation and progressive
neurodegeneration. Glia 55, 453–462 (2007).
203. Goehler, L. E. et al. Interleukin-1beta in immune cells of the abdominal vagus nerve: a link
between the immune and nervous systems? J Neurosci Official J Soc Neurosci 19, 2799–806
(1999).
204. Xaio, H., Banks, W. A., Niehoff, M. L. & Morley, J. E. Effect of LPS on the permeability of
the blood–brain barrier to insulin. Brain Res 896, 36–42 (2001).
205. Maezawa, I., Maeda, N., Montine, T. J. & Montine, K. S. Apolipoprotein E-specific innate
immune response in astrocytes from targeted replacement mice. J Neuroinflamm 3, 10–10
(2006).
132
206. Copeland, S. et al. Acute Inflammatory Response to Endotoxin in Mice and Humans. Clin
Vaccine Immunol 12, 60–67 (2005).
207. Vedder, H. et al. Dose-dependence of bacterial lipopolysaccharide (LPS) effects on peak
response and time course of the immune-endocrine host response in humans. Inflamm Res 48,
67–74 (1999).
208. Maitra, U. et al. Molecular Mechanisms Responsible for the Selective and Low-Grade
Induction of Proinflammatory Mediators in Murine Macrophages by Lipopolysaccharide. J
Immunol 189, 1014–1023 (2012).
209. Morris, M. C., Gilliam, E. A., Button, J. & Li, L. Dynamic Modulation of Innate Immune
Response by Varying Dosages of Lipopolysaccharide (LPS) in Human Monocytic Cells*. J Biol
Chem 289, 21584–21590 (2014).
210. Riedel, B. C., Thompson, P. M. & Brinton, R. D. Age, APOE and sex: Triad of risk of
Alzheimer’s disease. J Steroid Biochem Mol Biology 160, 134–147 (2016).
211. Gamache, J., Yun, Y. & Chiba-Falek, O. Sex-dependent effect of APOE on Alzheimer’s
disease and other age-related neurodegenerative disorders. Dis Model Mech 13, dmm045211
(2020).
212. Klein, S. L. & Flanagan, K. L. Sex differences in immune responses. Nat Rev Immunol 16,
626–638 (2016).
213. Jaillon, S., Berthenet, K. & Garlanda, C. Sexual Dimorphism in Innate Immunity. Clin Rev
Allerg Immu 56, 308–321 (2019).
214. Cai, K. C. et al. Age and sex differences in immune response following LPS treatment in
mice. Brain Behav Immun 58, 327–337 (2016).
215. Kuo, S.-M. Gender Difference in Bacteria Endotoxin-Induced Inflammatory and Anorexic
Responses. Plos One 11, e0162971 (2016).
216. Santos-Galindo, M., Acaz-Fonseca, E., Bellini, M. J. & Garcia-Segura, L. M. Sex
differences in the inflammatory response of primary astrocytes to lipopolysaccharide. Biol Sex
Differ 2, 7–7 (2011).
217. Erickson, M. A. et al. Genetics and sex influence peripheral and central innate immune
responses and blood-brain barrier integrity. Plos One 13, e0205769 (2018).
218. Dockman, R. L., Carpenter, J. M., Diaz, A. N., Benbow, R. A. & Filipov, N. M. Sex
differences in behavior, response to LPS, and glucose homeostasis in middle-aged mice. Behav
Brain Res 418, 113628 (2022).
219. Brown, C. M. et al. Apolipoprotein E isoform mediated regulation of nitric oxide release.
Free Radic Biology Medicine 32, 1071–5 (2002).
133
220. D’Eon, T. M. et al. Estrogen Regulation of Adiposity and Fuel Partitioning EVIDENCE OF
GENOMIC AND NON-GENOMIC REGULATION OF LIPOGENIC AND OXIDATIVE
PATHWAYS*. J Biol Chem 280, 35983–35991 (2005).
221. Leeners, B., Geary, N., Tobler, P. N. & Asarian, L. Ovarian hormones and obesity. Hum
Reprod Update 23, 300–321 (2016).
222. Velez-Perez, A., Holder, M. K., Fountain, S. & Blaustein, J. D. Estradiol Increases
Microglial Response to Lipopolysaccharide in the Ventromedial Hypothalamus during the
Peripubertal Sensitive Period in Female Mice. Eneuro 7, ENEURO.0505-19.2020 (2020).
223. Trumble, B. C. & Finch, C. E. THE EXPOSOME IN HUMAN EVOLUTION: FROM DUST
TO DIESEL. Q Rev Biology 94, 333–394 (2019).
224. Martiskainen, H. et al. Decreased plasma C-reactive protein levels in APOE ε4 allele
carriers. Ann Clin Transl Neur 5, 1229–1240 (2018).
225. Lumsden, A. L., Mulugeta, A., Zhou, A. & Hyppönen, E. Apolipoprotein E (APOE)
genotype-associated disease risks: a phenome-wide, registry-based, case-control study utilising
the UK Biobank. Ebiomedicine 59, 102954 (2020).
226. Vasunilashorn, S. et al. Inflammatory Gene Variants in the Tsimane, an Indigenous
Bolivian Population with a High Infectious Load. Biodemogr Soc Biol 57, 33–52 (2011).
227. Mueller, T. et al. Apolipoprotein E allele frequencies in chronic and self-limited hepatitis C
suggest a protective effect of APOE4 in the course of hepatitis C virus infection. Liver Int 36,
1267–1274 (2016).
228. Nascimento, J. C. R. et al. Apolipoprotein E polymorphism influences orthotopic liver
transplantation outcomes in patients with hepatitis C virus-induced liver cirrhosis. World J
Gastroentero 27, 1064–1075 (2021).
229. ORIÁ, R. B. et al. APOE4 Protects the Cognitive Development in Children with Heavy
Diarrhea Burdens in Northeast Brazil. Pediatr Res 57, 310–316 (2005).
230. Oriá, R. B. et al. ApoE polymorphisms and diarrheal outcomes in Brazilian shanty town
children. Braz J Med Biol Res 43, 249–256 (2010).
231. Trumble, B. C. et al. Apolipoprotein E4 is associated with improved cognitive function in
Amazonian forager-horticulturalists with a high parasite burden. Faseb J 31, 1508–1515 (2017).
232. Caselli, R. J. et al. Longitudinal modeling of frontal cognition in APOE ε4 homozygotes,
heterozygotes, and noncarriers. Neurology 76, 1383–8 (2011).
233. Rawle, M. J. et al. Apolipoprotein-E (Apoe) ε4 and cognitive decline over the adult life
course. Transl Psychiat 8, 18 (2018).
134
234. Gharbi-Meliani, A. et al. The association of APOE ε4 with cognitive function over the adult
life course and incidence of dementia: 20 years follow-up of the Whitehall II study. Alzheimer’s
Res Ther 13, 5 (2021).
235. Salvestrini, V., Sell, C. & Lorenzini, A. Obesity May Accelerate the Aging Process. Front
Endocrinol 10, 266 (2019).
236. Nunan, E. et al. Obesity as a premature aging phenotype — implications for sarcopenic
obesity. Geroscience 44, 1393–1405 (2022).
237. Monteiro, R. & Azevedo, I. Chronic Inflammation in Obesity and the Metabolic Syndrome.
Mediat Inflamm 2010, 289645 (2010).
238. Vitseva, O. I. et al. Inducible Toll-like Receptor and NF-κB Regulatory Pathway Expression
in Human Adipose Tissue. Obesity 16, 932–937 (2008).
239. Reyna, S. M. et al. Elevated Toll-Like Receptor 4 Expression and Signaling in Muscle From
Insulin-Resistant Subjects. Diabetes 57, 2595–2602 (2008).
240. Vila, I. K. et al. Immune Cell Toll-like Receptor 4 Mediates the Development of Obesity-
and Endotoxemia-Associated Adipose Tissue Fibrosis. Cell Reports 7, 1116–1129 (2014).
241. Fiebich, B. L., Batista, C. R. A., Saliba, S. W., Yousif, N. M. & Oliveira, A. C. P. de. Role of
Microglia TLRs in Neurodegeneration. Front Cell Neurosci 12, 329 (2018).
242. Stranahan, A. M. Models and mechanisms for hippocampal dysfunction in obesity and
diabetes. Neuroscience 309, 125–139 (2015).
243. Pugazhenthi, S., Qin, L. & Reddy, P. H. Common neurodegenerative pathways in obesity,
diabetes, and Alzheimer’s disease. Biochimica Et Biophysica Acta Bba - Mol Basis Dis 1863,
1037–1045 (2017).
244. Guillemot-Legris, O. & Muccioli, G. G. Obesity-Induced Neuroinflammation: Beyond the
Hypothalamus. Trends Neurosci 40, 237–253 (2017).
245. Thaler, J. P. & Schwartz, M. W. Minireview: Inflammation and Obesity Pathogenesis: The
Hypothalamus Heats Up. Endocrinology 151, 4109–4115 (2010).
246. Pistell, P. J. et al. Cognitive impairment following high fat diet consumption is associated
with brain inflammation. J Neuroimmunol 219, 25–32 (2010).
247. Boitard, C. et al. Juvenile, but not adult exposure to high-fat diet impairs relational memory
and hippocampal neurogenesis in mice. Hippocampus 22, 2095–2100 (2012).
248. Goldmann, T. et al. A new type of microglia gene targeting shows TAK1 to be pivotal in
CNS autoimmune inflammation. Nat Neurosci 16, 1618–1626 (2013).
135
249. Fortress, A. M., Kim, J., Poole, R. L., Gould, T. J. & Frick, K. M. 17β-Estradiol regulates
histone alterations associated with memory consolidation and increases Bdnf promoter
acetylation in middle-aged female mice. Learn Memory 21, 457–467 (2014).
250. Tuscher, J. J., Fortress, A. M., Kim, J. & Frick, K. M. Regulation of object recognition and
object placement by ovarian sex steroid hormones. Behav Brain Res 285, 140–157 (2015).
251. Young, K. & Morrison, H. Quantifying Microglia Morphology from Photomicrographs of
Immunohistochemistry Prepared Tissue Using ImageJ. J Vis Exp Jove 57648 (2018)
doi:10.3791/57648.
252. Arganda-Carreras, I., Fernández-González, R., Muñoz-Barrutia, A. & Ortiz-De-Solorzano,
C. 3D reconstruction of histological sections: Application to mammary gland tissue. Microsc Res
Techniq 73, 1019–1029 (2010).
253. Plümpe, T. et al. Variability of doublecortin-associated dendrite maturation in adult
hippocampal neurogenesis is independent of the regulation of precursor cell proliferation. Bmc
Neurosci 7, 77 (2006).
254. Oomen, C. A. et al. Severe early life stress hampers spatial learning and neurogenesis, but
improves hippocampal synaptic plasticity and emotional learning under high-stress conditions in
adulthood. J Neurosci Official J Soc Neurosci 30, 6635–45 (2010).
255. Saaltink, D.-J., Zwet, E. W. van & Vreugdenhil, E. Doublecortin-Like Is Implicated in Adult
Hippocampal Neurogenesis and in Motivational Aspects to Escape from an Aversive
Environment in Male Mice. Eneuro 7, ENEURO.0324-19.2020 (2020).
256. Gao, Y. et al. Lipoprotein Lipase Maintains Microglial Innate Immunity in Obesity. Cell
Reports 20, 3034–3042 (2017).
257. Fortress, A. M., Fan, L., Orr, P. T., Zhao, Z. & Frick, K. M. Estradiol-induced object
recognition memory consolidation is dependent on activation of mTOR signaling in the dorsal
hippocampus. Learn Memory 20, 147–155 (2013).
258. Cardona, A. E. et al. Control of microglial neurotoxicity by the fractalkine receptor. Nat
Neurosci 9, 917–924 (2006).
259. Ajami, B., Bennett, J. L., Krieger, C., Tetzlaff, W. & Rossi, F. M. V. Local self-renewal can
sustain CNS microglia maintenance and function throughout adult life. Nat Neurosci 10, 1538–
1543 (2007).
260. Jung, S. et al. Analysis of Fractalkine Receptor CX 3 CR1 Function by Targeted Deletion
and Green Fluorescent Protein Reporter Gene Insertion. Mol Cell Biol 20, 4106–4114 (2000).
261. Geissmann, F., Jung, S. & Littman, D. R. Blood Monocytes Consist of Two Principal
Subsets with Distinct Migratory Properties. Immunity 19, 71–82 (2003).
262. Yona, S. et al. Fate Mapping Reveals Origins and Dynamics of Monocytes and Tissue
Macrophages under Homeostasis. Immunity 38, 79–91 (2013).
136
263. Serbulea, V. et al. Macrophage phenotype and bioenergetics are controlled by oxidized
phospholipids identified in lean and obese adipose tissue. P Natl Acad Sci Usa 115, E6254–
E6263 (2018).
264. Zheng, C. et al. Local proliferation initiates macrophage accumulation in adipose tissue
during obesity. Cell Death Dis 7, e2167–e2167 (2016).
265. Surmi, B. K. & Hasty, A. H. Macrophage infiltration into adipose tissue: initiation,
propagation and remodeling. Future Lipidol 3, 545–556 (2008).
266. Kiran, S., Kumar, V., Murphy, E. A., Enos, R. T. & Singh, U. P. High Fat Diet-Induced
CD8+ T Cells in Adipose Tissue Mediate Macrophages to Sustain Low-Grade Chronic
Inflammation. Front Immunol 12, 680944 (2021).
267. Bennett, M. L. et al. New tools for studying microglia in the mouse and human CNS. Proc
National Acad Sci 113, E1738–E1746 (2016).
268. Schmid, C. D. et al. Heterogeneous expression of the triggering receptor expressed on
myeloid cells-2 on adult murine microglia. J Neurochem 83, 1309–1320 (2002).
269. Jurga, A. M., Paleczna, M. & Kuter, K. Z. Overview of General and Discriminating Markers
of Differential Microglia Phenotypes. Front Cell Neurosci 14, 198 (2020).
270. Zhu, J.-W., Li, Y.-F., Wang, Z.-T., Jia, W.-Q. & Xu, R.-X. Toll-Like Receptor 4 Deficiency
Impairs Motor Coordination. Front Neurosci-switz 10, 33 (2016).
271. Okun, E. et al. Evidence for a Developmental Role for TLR4 in Learning and Memory. Plos
One 7, e47522 (2012).
272. Dorfman, M. D. & Thaler, J. P. Hypothalamic inflammation and gliosis in obesity. Curr Opin
Endocrinol Diabetes Obes 22, 325–330 (2015).
273. Valdearcos, M., Xu, A. W. & Koliwad, S. K. Hypothalamic Inflammation in the Control of
Metabolic Function. Annu Rev Physiol 77, 131–160 (2015).
274. Jais, A. & Brüning, J. C. Hypothalamic inflammation in obesity and metabolic disease. J
Clin Invest 127, 24–32 (2017).
275. Cai, D. & Khor, S. “Hypothalamic Microinflammation” Paradigm in Aging and Metabolic
Diseases. Cell Metab 30, 19–35 (2019).
276. Oh-I, S. et al. Central administration of interleukin-4 exacerbates hypothalamic
inflammation and weight gain during high-fat feeding. Am J Physiol-endoc M 299, E47–E53
(2010).
277. Maldonado-Ruiz, R., Montalvo-Martínez, L., Fuentes-Mera, L. & Camacho, A. Microglia
activation due to obesity programs metabolic failure leading to type two diabetes. Nutr Diabetes
7, e254–e254 (2017).
137
278. Mohamed-Ali, V. et al. Subcutaneous adipose tissue releases interleukin-6, but not tumor
necrosis factor-alpha, in vivo. J Clin Endocrinol Metabolism 82, 4196–200 (1997).
279. Stefan, N. et al. Identification and Characterization of Metabolically Benign Obesity in
Humans. Arch Intern Med 168, 1609–1616 (2008).
280. Phillips, C. M. Metabolically healthy obesity: Definitions, determinants and clinical
implications. Rev Endocr Metabolic Disord 14, 219–227 (2013).
281. Muñoz-Garach, A., Cornejo-Pareja, I. & Tinahones, F. J. Does Metabolically Healthy
Obesity Exist? Nutrients 8, 320 (2016).
282. Zhu, Q. et al. Suppressing adipocyte inflammation promotes insulin resistance in mice. Mol
Metab 39, 101010 (2020).
283. Kunz, H. E. et al. Adipose tissue macrophage populations and inflammation are associated
with systemic inflammation and insulin resistance in obesity. Am J Physiol-endoc M 321, E105–
E121 (2021).
284. Zatterale, F. et al. Chronic Adipose Tissue Inflammation Linking Obesity to Insulin
Resistance and Type 2 Diabetes. Front Physiol 10, 1607 (2020).
285. Barbarroja, N. et al. The obese healthy paradox: is inflammation the answer? Biochem J
430, 141–149 (2010).
286. Phillips, C. M. & Perry, I. J. Does Inflammation Determine Metabolic Health Status in
Obese and Nonobese Adults? J Clin Endocrinol Metabolism 98, E1610–E1619 (2013).
287. Jaensson, E. et al. Small intestinal CD103+ dendritic cells display unique functional
properties that are conserved between mice and humans. J Exp Medicine 205, 2139–2149
(2008).
288. Timper, K. et al. IL-6 Improves Energy and Glucose Homeostasis in Obesity via Enhanced
Central IL-6 trans-Signaling. Cell Reports 19, 267–280 (2017).
289. Mishra, D. et al. Parabrachial Interleukin-6 Reduces Body Weight and Food Intake and
Increases Thermogenesis to Regulate Energy Metabolism. Cell Reports 26, 3011-3026.e5
(2019).
290. Ropelle, E. R. et al. IL-6 and IL-10 Anti-Inflammatory Activity Links Exercise to
Hypothalamic Insulin and Leptin Sensitivity through IKKβ and ER Stress Inhibition. Plos Biol 8,
e1000465 (2010).
291. Bobbo, V. C. et al. Interleukin-6 actions in the hypothalamus protects against obesity and is
involved in the regulation of neurogenesis. J Neuroinflamm 18, 192 (2021).
292. Gruol, D. L. IL-6 regulation of synaptic function in the CNS. Neuropharmacology 96, 42–54
(2015).
138
293. Chen, L. et al. MicroRNA-146a protects against cognitive decline induced by surgical
trauma by suppressing hippocampal neuroinflammation in mice. Brain Behav Immun 78, 188–
201 (2019).
294. Kraakman, M. J. et al. Blocking IL-6 trans-Signaling Prevents High-Fat Diet-Induced
Adipose Tissue Macrophage Recruitment but Does Not Improve Insulin Resistance. Cell Metab
21, 403–416 (2015).
295. Klover, P. J., Zimmers, T. A., Koniaris, L. G. & Mooney, R. A. Chronic exposure to
interleukin-6 causes hepatic insulin resistance in mice. Diabetes 52, 2784–9 (2003).
296. Han, M. S. et al. Regulation of adipose tissue inflammation by interleukin 6. P Natl Acad
Sci Usa 117, 2751–2760 (2020).
297. Erta, M., Quintana, A. & Hidalgo, J. Interleukin-6, a Major Cytokine in the Central Nervous
System. Int J Biol Sci 8, 1254–1266 (2012).
298. Baufeld, C., Osterloh, A., Prokop, S., Miller, K. R. & Heppner, F. L. High-fat diet-induced
brain region-specific phenotypic spectrum of CNS resident microglia. Acta Neuropathol 132,
361–375 (2016).
299. Wu, M. et al. Hippocampal overexpression of TREM2 ameliorates high fat diet induced
cognitive impairment and modulates phenotypic polarization of the microglia. Genes Dis 9, 401–
414 (2020).
300. Reilly, A. M. et al. Metabolic Defects Caused by High-Fat Diet Modify Disease Risk through
Inflammatory and Amyloidogenic Pathways in a Mouse Model of Alzheimer’s Disease. Nutrients
12, 2977 (2020).
301. Nam, K. N. et al. Effect of high fat diet on phenotype, brain transcriptome and lipidome in
Alzheimer’s model mice. Sci Rep-uk 7, 4307 (2017).
302. Bocarsly, M. E. et al. Obesity diminishes synaptic markers, alters microglial morphology,
and impairs cognitive function. Proc National Acad Sci 112, 15731–15736 (2015).
303. Rolls, A. et al. Toll-like receptors modulate adult hippocampal neurogenesis. Nat Cell Biol
9, 1081–1088 (2007).
304. Gemma, C. & Bachstetter, A. D. The role of microglia in adult hippocampal neurogenesis.
Front Cell Neurosci 7, 229 (2013).
305. Sato, K. Effects of Microglia on Neurogenesis. Glia 63, 1394–1405 (2015).
306. Pérez-Rodríguez, D. R., Blanco-Luquin, I. & Mendioroz, M. The Participation of Microglia in
Neurogenesis: A Review. Brain Sci 11, 658 (2021).
307. Huang, Y. et al. Repopulated microglia are solely derived from the proliferation of residual
microglia after acute depletion. Nat Neurosci 21, 530–540 (2018).
139
308. Willis, E. F. et al. Repopulating Microglia Promote Brain Repair in an IL-6-Dependent
Manner. Cell 180, 833-846.e16 (2020).
309. Huang, K.-P. et al. Sex differences in response to short-term high fat diet in mice. Physiol
Behav 221, 112894 (2020).
310. Casimiro, I., Stull, N. D., Tersey, S. A. & Mirmira, R. G. Phenotypic sexual dimorphism in
response to dietary fat manipulation in C57BL/6J mice. J Diabetes Complicat 35, 107795
(2021).
311. Maric, I. et al. Sex and Species Differences in the Development of Diet-Induced Obesity
and Metabolic Disturbances in Rodents. Frontiers Nutrition 9, 828522 (2022).
312. Elzinga, S. E. et al. Sex differences in insulin resistance, but not peripheral neuropathy, in
a diet-induced prediabetes mouse model. Dis Model Mech 14, dmm.048909 (2021).
313. Medrikova, D. et al. Sex differences during the course of diet-induced obesity in mice:
adipose tissue expandability and glycemic control. Int J Obesity 36, 262–272 (2012).
314. Millington, G. W. The role of proopiomelanocortin (POMC) neurones in feeding behaviour.
Nutr Metabolism 4, 18–18 (2007).
315. Anderson, E. J. P. et al. 60 YEARS OF POMC: Regulation of feeding and energy
homeostasis by α-MSH. J Mol Endocrinol 56, T157–T174 (2016).
316. Nohara, K. et al. Early-Life Exposure to Testosterone Programs the Hypothalamic
Melanocortin System. Endocrinology 152, 1661–1669 (2011).
317. Wang, C. et al. TAp63 contributes to sexual dimorphism in POMC neuron functions and
energy homeostasis. Nat Commun 9, 1544 (2018).
318. Oraha, J., Enriquez, R. F., Herzog, H. & Lee, N. J. Sex-specific changes in metabolism
during the transition from chow to high-fat diet feeding are abolished in response to dieting in
C57BL/6J mice. Int J Obesity 46, 1749–1758 (2022).
319. Villa, A. et al. Sex-Specific Features of Microglia from Adult Mice. Cell Reports 23, 3501–
3511 (2018).
320. Dorfman, M. D. et al. Sex differences in microglial CX3CR1 signalling determine obesity
susceptibility in mice. Nat Commun 8, 14556 (2017).
321. Underwood, E. L. & Thompson, L. T. A High-Fat Diet Causes Impairment in Hippocampal
Memory and Sex-Dependent Alterations in Peripheral Metabolism. Neural Plast 2016, 7385314
(2016).
322. Underwood, E. L. & Thompson, L. T. High-fat diet impairs spatial memory and
hippocampal intrinsic excitability and sex-dependently alters circulating insulin and hippocampal
insulin sensitivity. Biol Sex Differ 7, 9 (2016).
140
323. Frick, K. M. & Gresack, J. E. Sex Differences in the Behavioral Response to Spatial and
Object Novelty in Adult C57BL/6 Mice. Behav Neurosci 117, 1283–1291 (2003).
324. Faraji, J., Metz, G. A. & Sutherland, R. J. Characterization of spatial performance in male
and female Long-Evans rats by means of the Morris water task and the ziggurat task. Brain Res
Bull 81, 164–172 (2010).
325. Keeley, R. J., Bye, C., Trow, J. & McDonald, R. J. Strain and sex differences in brain and
behaviour of adult rats: Learning and memory, anxiety and volumetric estimates. Behav Brain
Res 288, 118–131 (2015).
326. Voyer, D., Voyer, S. & Bryden, M. P. Magnitude of Sex Differences in Spatial Abilities: A
Meta-Analysis and Consideration of Critical Variables. Psychol Bull 117, 250–270 (1995).
327. Peters, M., Lehmann, W., Takahira, S., Takeuchi, Y. & Jordan, K. Mental Rotation Test
Performance in Four Cross-Cultural Samples (N = 3367): Overall Sex Differences and the Role
of Academic Program in Performance. Cortex 42, 1005–1014 (2006).
328. Rizk-Jackson, A. M. et al. Effects of sex on object recognition and spatial navigation in
humans. Behav Brain Res 173, 181–190 (2006).
329. Silverman, I., Choi, J. & Peters, M. The Hunter-Gatherer Theory of Sex Differences in
Spatial Abilities: Data from 40 Countries. Arch Sex Behav 36, 261–268 (2007).
330. Carroll, J. C. et al. Progesterone and Estrogen Regulate Alzheimer-Like Neuropathology in
Female 3xTg-AD Mice. J Neurosci 27, 13357–13365 (2007).
331. Christensen, A., Liu, J. & Pike, C. J. Aging Reduces Estradiol Protection Against Neural
but Not Metabolic Effects of Obesity in Female 3xTg-AD Mice. Front Aging Neurosci 12, 113
(2020).
332. Kratz, M. et al. Metabolic Dysfunction Drives a Mechanistically Distinct Proinflammatory
Phenotype in Adipose Tissue Macrophages. Cell Metab 20, 614–625 (2014).
333. Barrientos, R. M., Kitt, M. M., Watkins, L. R. & Maier, S. F. Neuroinflammation in the
normal aging hippocampus. Neuroscience 309, 84–99 (2015).
334. Block, M. L., Zecca, L. & Hong, J.-S. Microglia-mediated neurotoxicity: uncovering the
molecular mechanisms. Nat Rev Neurosci 8, 57–69 (2007).
335. Cherry, J. D., Olschowka, J. A. & O’Banion, M. K. Neuroinflammation and M2 microglia:
the good, the bad, and the inflamed. J Neuroinflamm 11, 98 (2014).
336. Tang, Y. & Le, W. Differential Roles of M1 and M2 Microglia in Neurodegenerative
Diseases. Mol Neurobiol 53, 1181–1194 (2016).
337. Ransohoff, R. M. A polarizing question: do M1 and M2 microglia exist? Nat Neurosci 19,
987–991 (2016).
141
338. Geric, I. et al. Metabolic Reprogramming during Microglia Activation. Immunometabolism 1,
(2019).
339. Kim, J. D., Yoon, N. A., Jin, S. & Diano, S. Microglial UCP2 Mediates Inflammation and
Obesity Induced by High-Fat Feeding. Cell Metab 30, 952-962.e5 (2019).
340. Gao, Y. et al. Hormones and diet, but not body weight, control hypothalamic microglial
activity. Glia 62, 17–25 (2014).
341. Milanova, I. V. et al. Diet-Induced Obesity Disturbs Microglial Immunometabolism in a
Time-of-Day Manner. Front Endocrinol 10, 424 (2019).
342. Lauterbach, M. A. et al. Toll-like Receptor Signaling Rewires Macrophage Metabolism and
Promotes Histone Acetylation via ATP-Citrate Lyase. Immunity 51, 997-1011.e7 (2019).
343. Lee, J.-W. et al. TLR4 (toll-like receptor 4) activation suppresses autophagy through
inhibition of FOXO3 and impairs phagocytic capacity of microglia. Autophagy 15, 753–770
(2019).
344. Ekdahl, C. T., Kokaia, Z. & Lindvall, O. Brain inflammation and adult neurogenesis: The
dual role of microglia. Neuroscience 158, 1021–1029 (2009).
345. Haruwaka, K. et al. Dual microglia effects on blood brain barrier permeability induced by
systemic inflammation. Nat Commun 10, 5816 (2019).
346. Douglass, J. D. et al. Microglial inflammatory activation paradoxically improves glucose
tolerance during diet-induced obesity. Biorxiv 2022.04.19.488819 (2022)
doi:10.1101/2022.04.19.488819.
347. Beutler, L. R. et al. Obesity causes selective and long-lasting desensitization of AgRP
neurons to dietary fat. Elife 9, e55909 (2020).
348. Mori, H. et al. Socs3 deficiency in the brain elevates leptin sensitivity and confers
resistance to diet-induced obesity. Nat Med 10, 739–743 (2004).
349. Purkayastha, S., Zhang, G. & Cai, D. Uncoupling the mechanisms of obesity and
hypertension by targeting hypothalamic IKK-β and NF-κB. Nat Med 17, 883–887 (2011).
350. Jang, P.-G. et al. NF-κB Activation in Hypothalamic Pro-opiomelanocortin Neurons Is
Essential in Illness- and Leptin-induced Anorexia*. J Biol Chem 285, 9706–9715 (2010).
351. Kanoski, S. E. & Grill, H. J. Hippocampus Contributions to Food Intake Control: Mnemonic,
Neuroanatomical, and Endocrine Mechanisms. Biol Psychiat 81, 748–756 (2017).
352. Sofroniew, M. V. & Vinters, H. V. Astrocytes: biology and pathology. Acta Neuropathol 119,
7–35 (2010).
142
353. Kim, Y., Park, J. & Choi, Y. K. The Role of Astrocytes in the Central Nervous System
Focused on BK Channel and Heme Oxygenase Metabolites: A Review. Antioxidants 8, 121
(2019).
354. Yi, C.-X., Habegger, K. M., Chowen, J. A., Stern, J. & Tschöp, M. H. A Role for Astrocytes
in the Central Control of Metabolism. Neuroendocrinology 93, 143–149 (2011).
355. Kim, J. G. et al. Leptin signaling in astrocytes regulates hypothalamic neuronal circuits and
feeding. Nat Neurosci 17, 908–910 (2014).
356. Shen, L. et al. Up-regulation of apolipoprotein E by leptin in the hypothalamus of mice and
rats. Physiol Behav 98, 223–228 (2009).
357. Mishra, B. B., Mishra, P. K. & Teale, J. M. Expression and distribution of Toll-like receptors
in the brain during murine neurocysticercosis. J Neuroimmunol 181, 46–56 (2006).
358. Zhang, Z., Zhang, Z.-Y., Wu, Y. & Schluesener, H. J. Immunolocalization of Toll-Like
Receptors 2 and 4 as well as Their Endogenous Ligand, Heat Shock Protein 70, in Rat
Traumatic Brain Injury. Neuroimmunomodulat 19, 10–19 (2011).
359. Nakano, Y. et al. Astrocytic TLR4 expression and LPS-induced nuclear translocation of
STAT3 in the sensory circumventricular organs of adult mouse brain. J Neuroimmunol 278,
144–158 (2015).
360. Rosciszewski, G. et al. Toll-Like Receptor 4 (TLR4) and Triggering Receptor Expressed on
Myeloid Cells-2 (TREM-2) Activation Balance Astrocyte Polarization into a Proinflammatory
Phenotype. Mol Neurobiol 55, 3875–3888 (2018).
361. Marinelli, C. et al. Ligand engagement of Toll-like receptors regulates their expression in
cortical microglia and astrocytes. J Neuroinflamm 12, 244 (2015).
362. Gorina, R., Font-Nieves, M., Márquez-Kisinousky, L., Santalucia, T. & Planas, A. M.
Astrocyte TLR4 activation induces a proinflammatory environment through the interplay
between MyD88-dependent NFκB signaling, MAPK, and Jak1/Stat1 pathways. Glia 59, 242–
255 (2011).
363. Kobayashi, K. et al. Minocycline selectively inhibits M1 polarization of microglia. Cell Death
Dis 4, e525–e525 (2013).
364. Garrido-Mesa, N., Zarzuelo, A. & Gálvez, J. Minocycline: far beyond an antibiotic:
Minocycline: far beyond an antibiotic. Brit J Pharmacol 169, 337–352 (2013).
365. Buckman, L. B. et al. Evidence for a novel functional role of astrocytes in the acute
homeostatic response to high-fat diet intake in mice. Mol Metab 4, 58–63 (2015).
366. Zhang, Y., Reichel, J. M., Han, C., Zuniga-Hertz, J. P. & Cai, D. Astrocytic Process
Plasticity and IKKβ/NF-κB in Central Control of Blood Glucose, Blood Pressure, and Body
Weight. Cell Metab 25, 1091-1102.e4 (2017).
143
367. Popov, A. et al. A high-fat diet changes astrocytic metabolism to promote synaptic plasticity
and behavior. Acta Physiol 236, e13847 (2022).
368. Sa, M. et al. Hypothalamic GABRA5-positive Neurons Control Obesity via Astrocytic
GABA. Biorxiv 2021.11.07.467613 (2022) doi:10.1101/2021.11.07.467613.
369. Douglass, J. D., Dorfman, M. D., Fasnacht, R., Shaffer, L. D. & Thaler, J. P. Astrocyte
IKKβ/NF-κB signaling is required for diet-induced obesity and hypothalamic inflammation. Mol
Metab 6, 366–373 (2017).
370. Kadry, H., Noorani, B. & Cucullo, L. A blood–brain barrier overview on structure, function,
impairment, and biomarkers of integrity. Fluids Barriers Cns 17, 69 (2020).
371. Gustafson, D. R. et al. Mid-life adiposity factors relate to blood–brain barrier integrity in late
life. J Intern Med 262, 643–650 (2007).
372. Nerurkar, P. V. et al. Momordica charantia (bitter melon) attenuates high-fat diet-
associated oxidative stress and neuroinflammation. J Neuroinflamm 8, 64 (2011).
373. Pepping, J. K., Freeman, L. R., Gupta, S., Keller, J. N. & Bruce-Keller, A. J. NOX2
deficiency attenuates markers of adiposopathy and brain injury induced by high-fat diet. Am J
Physiol-endoc M 304, E392–E404 (2013).
374. Stranahan, A. M., Hao, S., Dey, A., Yu, X. & Baban, B. Blood–brain barrier breakdown
promotes macrophage infiltration and cognitive impairment in leptin receptor-deficient mice. J
Cereb Blood Flow Metabolism 36, 2108–2121 (2016).
375. Morari, J. et al. Fractalkine (CX3CL1) Is Involved in the Early Activation of Hypothalamic
Inflammation in Experimental Obesity. Diabetes 63, 3770–3784 (2014).
376. Buckman, L. B. et al. Obesity induced by a high-fat diet is associated with increased
immune cell entry into the central nervous system. Brain Behav Immun 35, 33–42 (2014).
377. Cronk, J. C. et al. Peripherally derived macrophages can engraft the brain independent of
irradiation and maintain an identity distinct from microglia. J Exp Med 215, 1627–1647 (2018).
378. Shemer, A. et al. Engrafted parenchymal brain macrophages differ from microglia in
transcriptome, chromatin landscape and response to challenge. Nat Commun 9, 5206 (2018).
379. Grove, K. L., Fried, S. K., Greenberg, A. S., Xiao, X. Q. & Clegg, D. J. A microarray
analysis of sexual dimorphism of adipose tissues in high-fat-diet-induced obese mice. Int J
Obesity 34, 989–1000 (2010).
380. Benz, V. et al. Sexual Dimorphic Regulation of Body Weight Dynamics and Adipose Tissue
Lipolysis. Plos One 7, e37794 (2012).
381. Yang, Y., Smith, D. L., Keating, K. D., Allison, D. B. & Nagy, T. R. Variations in body
weight, food intake and body composition after long-term high-fat diet feeding in C57BL/6J
Mice. Obes Silver Spring Md 22, 2147–2155 (2014).
144
382. Morselli, E. et al. A sexually dimorphic hypothalamic response to chronic high-fat diet
consumption. Int J Obesity 40, 206–209 (2016).
383. Wallen, W. J., Belanger, M. P. & Wittnich, C. Sex Hormones and the Selective Estrogen
Receptor Modulator Tamoxifen Modulate Weekly Body Weights and Food Intakes in Adolescent
and Adult Rats. J Nutrition 131, 2351–2357 (2001).
384. Rogers, N. H., Perfield, J. W., Strissel, K. J., Obin, M. S. & Greenberg, A. S. Reduced
Energy Expenditure and Increased Inflammation Are Early Events in the Development of
Ovariectomy-Induced Obesity. Endocrinology 150, 2161–2168 (2009).
385. Blouin, K., Boivin, A. & Tchernof, A. Androgens and body fat distribution. J Steroid
Biochem Mol Biology 108, 272–280 (2008).
386. Heine, P. A., Taylor, J. A., Iwamoto, G. A., Lubahn, D. B. & Cooke, P. S. Increased
adipose tissue in male and female estrogen receptor-alpha knockout mice. P Natl Acad Sci Usa
97, 12729–34 (2000).
387. Scudiero, R. & Verderame, M. Gene expression profile of estrogen receptors alpha and
beta in rat brain during aging and following high fat diet. C R Biol 340, 372–378 (2017).
388. Xu, Y. et al. Distinct Hypothalamic Neurons Mediate Estrogenic Effects on Energy
Homeostasis and Reproduction. Cell Metab 14, 453–465 (2011).
389. Morselli, E. et al. Hypothalamic PGC-1α Protects Against High-Fat Diet Exposure by
Regulating ERα. Cell Reports 9, 633–645 (2014).
390. Callewaert, F. et al. Differential regulation of bone and body composition in male mice with
combined inactivation of androgen and estrogen receptor-α. Faseb J 23, 232–240 (2009).
391. Finan, B. et al. Targeted estrogen delivery reverses the metabolic syndrome. Nat Med 18,
1847–1856 (2012).
392. Salinero, A. E., Anderson, B. M. & Zuloaga, K. L. Sex differences in the metabolic effects
of diet-induced obesity vary by age of onset. Int J Obesity 42, 1088–1091 (2018).
393. Arcones, A. C., Cruces-Sande, M., Ramos, P., Mayor, F. & Murga, C. Sex Differences in
High Fat Diet-Induced Metabolic Alterations Correlate with Changes in the Modulation of GRK2
Levels. Cells 8, 1464 (2019).
394. Hunsche, C., Toda, I. M. de & Fuente, M. D. la. Impacts of the late adulthood diet-induced
obesity onset on behavior, immune function, redox state and life span of male and female mice.
Brain Behav Immun 78, 65–77 (2019).
395. Morris, G. P., Clark, I. A., Zinn, R. & Vissel, B. Microglia: A new frontier for synaptic
plasticity, learning and memory, and neurodegenerative disease research. Neurobiol Learn
Mem 105, 40–53 (2013).
145
396. Wu, Y., Dissing-Olesen, L., MacVicar, B. A. & Stevens, B. Microglia: Dynamic Mediators of
Synapse Development and Plasticity. Trends Immunol 36, 605–613 (2015).
397. Trotta, T., Porro, C., Calvello, R. & Panaro, M. A. Biological role of Toll-like receptor-4 in
the brain. J Neuroimmunol 268, 1–12 (2014).
398. Streit, W. J., Sammons, N. W., Kuhns, A. J. & Sparks, D. L. Dystrophic microglia in the
aging human brain. Glia 45, 208–212 (2004).
399. Kullberg, S., Aldskogius, H. & Ulfhake, B. Microglial activation, emergence of ED1-
expressing cells and clusterin upregulation in the aging rat CNS, with special reference to the
spinal cord. Brain Res 899, 169–186 (2001).
400. Ye, S.-M. & Johnson, R. W. An Age-Related Decline in Interleukin-10 May Contribute to
the Increased Expression of Interleukin-6 in Brain of Aged Mice. Neuroimmunomodulat 9, 183–
192 (2002).
401. Sierra, A., Gottfried-Blackmore, A. C., McEwen, B. S. & Bulloch, K. Microglia derived from
aging mice exhibit an altered inflammatory profile. Glia 55, 412–424 (2007).
402. Njie, eMalick G. et al. Ex vivo cultures of microglia from young and aged rodent brain
reveal age-related changes in microglial function. Neurobiol Aging 33, 195.e1-195.e12 (2012).
403. Hammond, T. R. et al. Single-Cell RNA Sequencing of Microglia throughout the Mouse
Lifespan and in the Injured Brain Reveals Complex Cell-State Changes. Immunity 50, 253-
271.e6 (2019).
404. Damani, M. R. et al. Age-related alterations in the dynamic behavior of microglia. Aging
Cell 10, 263–276 (2011).
405. Ritzel, R. M. et al. Age- and location-related changes in microglial function. Neurobiol
Aging 36, 2153–2163 (2015).
406. Koenigsknecht-Talboo, J. & Landreth, G. E. Microglial phagocytosis induced by fibrillar
beta-amyloid and IgGs are differentially regulated by proinflammatory cytokines. J Neurosci
Official J Soc Neurosci 25, 8240–9 (2005).
407. Glass, C. K., Saijo, K., Winner, B., Marchetto, M. C. & Gage, F. H. Mechanisms Underlying
Inflammation in Neurodegeneration. Cell 140, 918–934 (2010).
408. Jimenez, S. et al. Inflammatory response in the hippocampus of PS1M146L/APP751SL
mouse model of Alzheimer’s disease: age-dependent switch in the microglial phenotype from
alternative to classic. J Neurosci Official J Soc Neurosci 28, 11650–61 (2008).
409. Patterson, S. L. Immune dysregulation and cognitive vulnerability in the aging brain:
Interactions of microglia, IL-1β, BDNF and synaptic plasticity. Neuropharmacology 96, 11–18
(2015).
146
410. Minhas, P. S. et al. Restoring metabolism of myeloid cells reverses cognitive decline in
ageing. Nature 590, 122–128 (2021).
411. Rice, R. A. et al. Elimination of Microglia Improves Functional Outcomes Following
Extensive Neuronal Loss in the Hippocampus. J Neurosci 35, 9977–9989 (2015).
412. Zhou, Y., Chen, Y., Xu, C., Zhang, H. & Lin, C. TLR4 Targeting as a Promising Therapeutic
Strategy for Alzheimer Disease Treatment. Front Neurosci-switz 14, 602508 (2020).
413. Minoretti, P. et al. Effect of the functional toll-like receptor 4 Asp299Gly polymorphism on
susceptibility to late-onset Alzheimer’s disease. Neurosci Lett 391, 147–149 (2006).
414. Balistreri, C. et al. Association between the Polymorphisms of TLR4 and CD14 Genes and
Alzheimers Disease. Curr Pharm Design 14, 2672–2677 (2008).
415. Walter, S. et al. Role of the Toll-Like Receptor 4 in Neuroinflammation in Alzheimer’s
Disease. Cell Physiol Biochem 20, 947–956 (2007).
416. Balducci, C. et al. Toll-like receptor 4-dependent glial cell activation mediates the
impairment in memory establishment induced by β-amyloid oligomers in an acute mouse model
of Alzheimer’s disease. Brain Behav Immun 60, 188–197 (2016).
417. Jin, J.-J., Kim, H.-D., Maxwell, J. A., Li, L. & Fukuchi, K. Toll-like receptor 4-dependent
upregulation of cytokines in a transgenic mouse model of Alzheimer’s disease. J Neuroinflamm
5, 23–23 (2008).
418. Tahara, K. et al. Role of toll-like receptor signalling in Aβ uptake and clearance. Brain 129,
3006–3019 (2006).
419. Song, M. et al. TLR4 mutation reduces microglial activation, increases Aβ deposits and
exacerbates cognitive deficits in a mouse model of Alzheimer’s disease. J Neuroinflamm 8, 92
(2011).
420. Michaud, J.-P. et al. Toll-like receptor 4 stimulation with the detoxified ligand
monophosphoryl lipid A improves Alzheimer’s disease-related pathology. Proc National Acad
Sci 110, 1941–1946 (2013).
421. Go, M., Kou, J., Lim, J.-E., Yang, J. & Fukuchi, K. Microglial response to LPS increases in
wild-type mice during aging but diminishes in an Alzheimer’s mouse model: Implication of TLR4
signaling in disease progression. Biochem Bioph Res Co 479, 331–337 (2016).
422. Tai, L. M. et al. APOE-modulated Aβ-induced neuroinflammation in Alzheimer’s disease:
current landscape, novel data, and future perspective. J Neurochem 133, 465–488 (2015).

Abstract (if available)

Abstract Excessive activation of innate immunity and inflammation are associated with and may contribute to many age-related conditions, such as Alzheimer’s disease (AD) and obesity. The toll-like receptor 4 (TLR4) signaling pathway has been proposed as one of the main factors of disease associated inflammatory responses. In my dissertation, I investigated the relationship between TLR4-mediated innate immune response and neuroinflammation in the context of the AD genetic risk factor Apolipoprotein E (APOE) allele ε4 (APOE4) and obesity. Chapter 1 provides an introduction to topics relevant to my dissertation. I begin by examining the association between innate immune response and AD, and how APOE genotype interacts with innate immunity in driving AD. I then addressed obesity-induced neural dysfunctions as well as possible underlying mechanisms, including inflammation, microglia activation, and TLR4 signaling pathway. In Chapter 2, I examined the effect of APOE genotype on innate immune responses in female mice. I also examined whether this relationship is affected by estradiol levels. Results from this study suggested that APOE4 was associated with relatively protective outcomes in acute inflammatory response. In Chapter 3, I investigated the role of microglial TLR4 signaling in modulating obesity-induced neural inflammation and dysfunction using a newly generated mouse model in male and female mice. I found that microglial TLR4 deletion yielded protection against diet-induced metabolic disruption, peripheral inflammation, cognitive impairment, and neurogenesis with sex differences. Chapter 5 is a summary of my key findings. I also discussed possible improvements that can be made as well as some future directions.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Innate immune response to amyloid-beta: relevance to Alzheirmer’s disease and neuroinflammation

PDF

Identifying and measuring cell-intrinsic and cell-extrinsic factors influencing aging

PDF

Synaptic transmission, nutrient sensors, and aging in Drosophila melanogaster

PDF

Defining the functional roles of neurotransmitters and neuropeptides in neural circuits

PDF

Investigating brain aging and neurodegenerative diseases through omics data

PDF

Neuroinflammation and ApoE4 genotype in at-risk female aging: implications for Alzheimer's disease

PDF

The role of inflammation in mediating effects of obesity on Alzheimer's disease

PDF

Roles of epithelial-mesenchymal transition and niche in tumorigenesis of tumor-initiating cells

PDF

Modeling neurodegenerative diseases using induced pluripotent stem cells and identifying therapeutic targets

PDF

Role of TLR4 and AID in lymphomagenesis induced by obesity and hepatitis C virus

PDF

Synergism between TLR4-c-JUN axis and surface receptor stimulation translocate c-MYC into Ig loci, through AID leading to lymphomagenesis

PDF

Longitudinal assessment of neural stem-cell aging

PDF

Mediating the homeostatic response during light & cytotoxic induced retinal damage

PDF

Phenotypic and multi-omic characterization of novel C. elegans models of Alzheimer's disease

PDF

Discovery of mitochondria-DNA-encoded microproteins in neurodegeneration

PDF

Natural divergence of traits in species of mice reveals novel molecular mechanisms of cellular senescence

PDF

Genomic and phenotypic novelties in the Southeast Asian house mouse

PDF

Defining novel molecular mechanisms to restore balance of a stress response homeostat across spatial and temporal boundaries

PDF

The effects of hormone therapy on carotid artery intima-media thickness and serum lipids by ApoE4 genotype

PDF

The effects of a fasting mimicking diet (FMD) on mouse models of Alzheimer's and Parkinson's disease

Asset Metadata

Creator Liu, Jiahui (author)

Core Title TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Biology of Aging

Degree Conferral Date 2022-12

Publication Date 12/08/2023

Defense Date 11/30/2022

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

Tag APOE,microglia,neuroinflammation,OAI-PMH Harvest,obesity,TLR4

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pike, Christian (committee chair), Benayoun, Bérénice (committee member), Curran, Sean (committee member), Ellerby, Lisa (committee member)

Creator Email jiahuil@usc.edu,ljhsnoopy@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC112619796

Unique identifier UC112619796

Identifier etd-LiuJiahui-11350.pdf (filename)

Legacy Identifier etd-LiuJiahui-11350

Document Type Dissertation

Format theses (aat)

Rights Liu, Jiahui

Internet Media Type application/pdf

Type texts

Source 20221213-usctheses-batch-995 (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