University of Southern California Dissertations and Theses

Characterization of Kupffer cell subpopulation in liver injury model through newly identified Kupffer cell markers

(USC Thesis Other)

Characterization of Kupffer cell subpopulation in liver injury model through newly identified Kupffer cell markers

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
i
Characterization of Kupffer Cell Subpopulation in Liver Injury Model through Newly
Identified Kupffer Cell Markers

By

Handan Hong

A Thesis Presented to the
FACULTY OF THE USC MANN SCHOOL OF PHARMACY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(PHARMACEUTICAL SCIENCES)

May 2023

Copyright 2023. Handan Hong

ii
Acknowledgements
First of all, I want to earnestly thank my mentor, Dr. Bangyan L. Stiles for her consistent
support of my master study. Her passion for science and patience to us encourage me
to further explore this field. She is the best mentor I have ever had.
Then, I would like to thank my committee members, Dr. Martine Culty and Dr. Keigo
Machida. Their support and guidance allow me to finish my master thesis.
I would also like to thank all the members in Dr. Stiles’ lab, including Dr. Lina He, Dr.
Taojian Tu, Mario Alba, Brittney Hua, Ielyzaveta Slarve, Qi Tang, Jared A Khan for
their technical support and suggestions. Especially, I want to thank Dr. Taojian Tu. He
taught me lots of professional techniques and encouraged me a lot during this period.
Finally, I would like to thank my family and friends for their consistent support and
encouragement.

Handan Hong
May 2023
University of Southern California

iii
Table of Contents
Acknowledgements ......................................................................................................ii
List of Figures..............................................................................................................v
Abstract......................................................................................................................vii
Chapter I Introduction and Background.......................................................................1
1. Epidemiology of chronic liver diseases ........................................................... 1
2. Triggers of inflammation in NAFLD ................................................................. 2
3. Hepatic macrophages ..................................................................................... 9
4. Hepatic macrophages in healthy and disease state ..................................... 10
5. Heterogeneity of Kupffer cells ....................................................................... 14
Goal of Project and Scope of the Study ................................................................. 16
Chapter II Methods....................................................................................................18
1. Intracellular immunofluorescent staining ...................................................... 18
2. Isolation of non-parenchymal cells ............................................................... 19
3. Intracellular flow cytometry ........................................................................... 19
4. RNA isolation, reverse transcription and quantitative PCR .......................... 20
5. Biochemical analysis .................................................................................... 22
Chapter III Results.....................................................................................................24

iv
1. Identification of new Kupffer cell markers ..................................................... 24
2. In silico verifications ...................................................................................... 29
3. Validating CD5L expression using liver injury models .................................. 39
4. Specific expression of CD5L in Kupffer cells ................................................ 41
5. Plasma concentration CD5L as potential biomarker for the activation of KCs
......................................................................................................................44
Chapter IV Discussion...............................................................................................45
Bibliography...............................................................................................................50

v
List of Figures
Figure 1. “Two-hit” hypothesis of the development from NAFLD to NASH....................4
Figure 2. Signaling pathways that can activate KCs and an overview of inter-organ
crosstalk during NAFLD/NASH....................................................................................6
Figure 3. An overview of crosstalk between KCs and other hepatic cells during
NAFLD/NASH..............................................................................................................8
Figure 4. Intercellular signals at healthy state.............................................................12
Figure 5. Files and codes used in the project............................................................25
Figure 6. Flow chart of the whole project...................................................................26
Figure 7. Identification of new markers for Kupffer cells............................................28
Figure 8. Correlation analysis of Cd5l and markers of different cell types based on
GSE149863...............................................................................................................31
Figure 9. Correlation analysis of Folr2 and markers of different cell types in
GSE149863...............................................................................................................32
Figure 10. Correlation analysis of markers of different cell types in
GSE149863...............................................................................................................33
Figure 11. Correlation analysis of Cd5l and markers of different cell types in
GSE90497.................................................................................................................34
Figure 12. Correlation analysis of Folr2 and markers of different cell types in

vi
GSE90497................................................................................................................35
Figure 13. Correlation analysis of markers of different cell types based on
GSE90497.................................................................................................................36
Figure 14. Heatmap showing correlation coefficients between selected genes in
three different datasets..............................................................................................38
Figure 15. Validation of Cd5l expression in liver injury mouse models......................40
Figure 16. Flow cytometry analysis to validate Cd5l specific expression in Kupffer
cells............................................................................................................................42
Figure 17. Representative immunofluorescent staining of Cd5l and Clec4f..............43
Figure 18. Changes of Cd5l plasma concentration in liver injury mouse
models.......................................................................................................................44

vii
Abstract
Immune cells are highly versatile and have many subsets. Each subset of the immune
cells is known to exert their unique functions by specifically expressing a range of
receptors capable of recognizing diverse ligands and activating the downstream
signaling pathways. With the realization of heterogeneity of immune cells, more and
more studies are focused on investigating the distinct roles of these subpopulations in
various pathophysiology conditions. However, there is a lack of specific markers to
identify these subpopulations. Meanwhile, liver resident macrophages, known as
Kupffer cells, play significant functions in both physiological and pathophysiological
progresses, including the secretion of TNF-alpha to promote steatosis, the production
of TGF-beta to promote fibrosis, the expression of PD-L1 to inhibit the activation of T
cells. In order to further study the precise role that Kupffer cells play in the chronic liver
disease, specific markers for Kupffer cells are important. Unfortunately, there are very
few identified markers. The most widely used ones for mice are Clec4f, Vsig4 and
Timd4, but no well-defined markers for human Kupffer cells has been reported.
Without characterizing Kupffer cells, the involvement of Kupffer cells in the pathologies
will remain elusive.

Given this consideration, the goal of this project is to characterize Kupffer cell
subpopulations by identifying novel markers. Using bioinformatics screening, we
analyzed the published single cell RNA-seq data as our discovery dataset and ranked
the top 500 genes that are specifically enriched in Kupffer cells. This analysis identified

viii
Cd5l and Folr2 as two promising novel markers for Kupffer cells. We further validate
these markers by using in silicone analysis to demonstrate the correlation between
these two candidates and Clec4f. Both markers show great correlations with Clec4f
and Vsig4, and are not correlated with the markers of other cell types. Using two
steatosis mouse models representing alcoholic liver diseases and non-alcoholic liver
disease, we confirmed the expression correlation of Cd5l and Clec4f in the livers using
qPCR. In addition, we further confirmed the main source of Cd5l is KCs through
immunofluorescent staining and flow cytometry. Collectively, our data shows that Cd5l
is a putative novel marker for Kupffer cells.

1
Chapter I Introduction and Background
1. Epidemiology of chronic liver diseases
Primary liver cancer has become the sixth most commonly occurring cancer and the
third most frequent cause of cancer mortality in the world in recent decades (1).
Hepatocellular carcinoma (HCC) (accounting for approximate 75% cases) and
cholangiocarcinoma (about 15%) are two most common pathological subtypes of liver
cancer (2). The incidence of HCC is different from region to region with the highest in
Eastern and South-Eastern Asian countries. Diverse human sex ratios of different
regions contribute to this phenomenon since the prevalence of HCC among men is
two- to three-fold higher than that in women. Another crucial contribution is the variable
prevalence of etiologies among regions (3). Non-alcoholic fatty liver disease (NAFLD),
chronic infection of hepatitis B (HBV) or hepatitis C (HCV), and chronic alcohol
consumption associated liver disease (ALD) contribute to the major cases of HCC (4-
6). With the high vaccination coverage level and the approval of antiviral medications,
decline of HBV or HCV related HCC is expected to occur (7). Therefore, NAFLD and
ALD are increasingly becoming the major cause of HCC.

In recent years, the prevalence of NAFLD has dramatically increased worldwide (8, 9).
A 2016 meta-analysis estimated that the global prevalence of NAFLD was 25.24%,
ranging from 31.9% in Middle East to 13.48% in Africa (10), while another 2019 meta-
analysis indicated that its global prevalence was 30% with the highest in South
America and North America (11). The increasing prevalence of NAFLD is parallel with

2
the rising number of individuals with metabolic disorders such as type 2 diabetes and
obesity. Noticeably, approximate 51.34% of the patients with NAFLD are obese, while
the prevalence of NAFLD varies from 30% to 37% in obesity individuals (10, 12).
Besides, ALD is another crucial cause of HCC and contributes greatly to alcohol-
related mortality and morbidity in recent decades (13, 14). Alcohol consumption and
abuse are closely associated with the development of ALD (15) and they cause more
than 3 million deaths per year based on a report of the World Health Organization (16).
As drinking patterns vary from country to country, ALD exhibits a greater prevalence
in regions with heavier per capita alcohol consumption (17). Together, NAFLD and
ALD have become leading indications for liver transplantation in the United States (18).

2. Triggers of inflammation in NAFLD
NAFLD and ALD include a board spectrum of disorders, ranging from simple non-
alcoholic and alcoholic fatty liver to more severe non-alcoholic steatohepatitis (NASH
and ASH), which is characterized by hepatocellular injury, infiltrating inflammation and
fibrosis. Simple fatty livers are generally considered as a benign condition and are
often characterized by simple steatosis, while NASH and ASH are more progressive
form and are always companied by inflammation (19). A subset of patients with fatty
livers can progress to NASH, and may further progress to fibrosis and HCC. The
progression from fatty liver to NASH involves multiple steps. The original proposed
molecular mechanism for the development from NAFLD to NASH involves a “two-hit”
hypothesis (Fig.1). With this model, lipid accumulation and insulin resistance provide

3
the “first hit” to the liver, oxidative stress, lipotoxicity and proinflammatory cytokines
(the “second hit”) further aggravate the disease (20). Overconsumption of
carbohydrates and fat resulting from metabolic dysfunction such as type 2 diabetes
mellitus can increase lipolysis of triglyceride in adipose tissue and consequently
increase the delivery of free fatty acids (FFAs) to the liver (21). The presence of high
level of saturated FFAs or their metabolites leads to apoptosis of hepatocytes, the
major parenchymal cells of the liver, via increasing endoplasmic reticulum (ER) stress
and oxidative stress (22, 23). For example, palmitate and stearate, two canonical
saturated long-chain fatty acids, have both been shown to be toxic to the cells. Both
lipids were shown to induce ER stress and lead to the apoptosis of hepatocytes in
tissue culture (24). A mechanism that was dependent of PERK/ATF4/CHOP was
reported to mediate this effect (25). In vivo, intraperitoneal injection of palmitate also
triggered the mRNA expression of inflammatory cytokines, such as TNF-α (Tumor
necrosis factor-α) and IL-1β (Interleukin-1 β). The upregulation of these cytokines were
also companied with the infiltration of neutrophils and macrophages (26),
demonstrating that palmitate exposure indeed induced inflammation. The “lipo-
apoptosis” effects of these lipids on hepatocytes results in the secretion of damage-
associated molecular pattern molecules (DAMPs), that are capable of inducing the
immune response and subsequent inflammation. Emerging evidence suggests that
HMGB1, a widely investigated DAMP, could promote inflammation by stimulating
macrophages to release TNF-α in a TLR4 (Toll-like receptor 4)-dependent manner (27,
28) (Fig.2A).

4

In addition, diet rich in fat and sugar also favors the compositional change of intestinal
microbiota and enhances intestinal permeability (29-31). Recent clinical studies have
shown that compared with healthy controls, the fecal samples of patients with NASH
show higher abundance of Bacteroides and lower abundance of its competitor,
Prevotella (32, 33). As a result of compositional change and gut barrier dysfunction,
translocation of lipopolysaccharides (LPS), double-stranded DNA and RNA (dsDNA,
dsRNA), and other pathogen-associated molecular patterns (PAMPs) to the liver
through portal vein increases (Fig.2A) (34). Excessive PAMPs also bind to pattern
recognition receptors, such as TLR, leading to the activation of immune response and
the release of inflammatory cytokines (35). In mice fed methionine/choline-deficient
(MCD) diet to induce NASH, TLR4-deficient led to attenuated immune cell infiltration,
Figure 1. The “first hit” involves insulin resistance and lipid deposition in
liver, which is caused by unhealthy lifestyle. It promotes the susceptibility of
liver to many factors, such as impaired balance between antioxidants and
oxidants, proinflammatory cytokines and enhanced saturated fatty acids
(SFAs) level. Those triggers further aggravate the disease, which is termed
as “second hit”.
normal steatosis cirrhosis
First hit: calorie-rich food
insufficient exercise
overconsumption of alcohol
Second hit: enhanced ROS/RNS
excessive SFAs
TNF-α, IL-6, IL-8

5
lipid accumulation and collagen deposition compared with mice with control genotype.
This study suggests that the inflammation in NASH is dependent on the TLR4 signal
(36). Furthermore, deletion of TLR9, the canonical receptor for dsDNA, also protected
mice from diet-induced NASH (37). Together, these studies suggest that the
progression of NAFLD to NASH is complex, where lipotoxicity, the gut barrier
dysfunction and other factors contribute to it simultaneously (38) (Fig.2B).

6

Adipokines & Cytokines
Lipolysis
Insulin resistance
Hepatokines
Antimicrobial molecules
Primary bile acids
PAMPs
Enterokines
Bacterial metabolites
Bacterial metabolites
Adipokines
HMGB1 LPS
mtDNA
Bacterial
lipopeptide
Cholesterol
oxLDL
Viral RNA
TNF-α
IL-1β
TGF-β
Activated
KC
TLR2
TLR3
TLR4
TLR4
TLR9
P2X
7
bacterial
Apoptotic
hepatocytes
PAMPs
DAMPs
LPS
Lipoprotein
Peptidoglycan
HMGB1
mtDNA
ATP
Bacterial RNA
A
B
Figure 2. (A) There are various signaling pathways to activate KCs in
NAFLD. PAMPs refer to a set of molecules that are produced by gut bacteria.
DAMPs refers to a bunch of molecules that are release by apoptotic
hepatocytes. (B) It shows overview of inter-organ crosstalk.
Overconsumption of carbohydrates and fat result in metabolic dysfunctions.
Adipokines various pro-inflammatory cytokines and excessive FFAs
released by impaired adipose tissue are involved in the development of
NAFLD. More and more evidence show that hepatokines cell-signaling
factors produced by hepatocytes, play important roles in the regulation of
adipose tissue energy homeostasis. Additionally, diet rich in fat and sugar
also favors the compositional change of gut bacteria and enhances intestinal
permeability, which increases the translocation of microorganism, DAMPs
and bacterial secondary metabolites to liver. Subsequently, immune
responses are activated, and pro-inflammatory triggers are released to
enhance or initiate hepatic inflammation. Bile acids synthesized by liver
enter intestinal lumen via bile duct and induce the production of antimicrobial
peptides to strictly inhibit bacterial overgrowth. Short-chain fatty acids, the
predominant products derived from bacterial fermentation of carbohydrates,
are closely related to many biological process in adipose tissue such as
lipolysis, inflammation and adipogenesis.

7
The crosstalk between liver steatosis and inflammation forms a vicious cycle. On one
hand, metabolic syndrome and hepatocytic lipid accumulation can promote liver
inflammation by increasing hepatic DAMP levels. On the other hand, liver inflammation
aggravate steatosis through inflammatory cytokines ad chemokines. Clinical studies
indicate that, compared with healthy controls, patients with NASH have significantly
higher serum level of TNF-α, IL-6 and IL-1β (39, 40). Meanwhile, their serum levels
show positive correlation with the severity of diseases. Mechanistically, the liver
macrophages secret various chemokines and cytokines to regulate the fate of their
surrounding cells (e.g., hepatocytes, hepatic stellate cells and liver sinusoidal
endothelial cells) in response to DAMPs and PAMPs. It is indicated that, by binding to
TNF receptor-1 (TNFR1), macrophage-derived TNF-α promotes lipid-loaded
hepatocyte apoptosis via caspase 8-mediated signaling pathway (41, 42). In addition,
TNF-α also contributes to insulin resistance. Mice lacking TNF-α were shown to
preserve insulin sensitivity in diet-induced obesity (43). In addition, macrophages
release IL-1β to enhance lipid accumulation in hepatocytes through promoting
lipogenesis pathway instead of suppressing fatty acid oxidation in lipolysis (44).
Another lipogenic cytokine secreted by the liver macrophages is IL-6. IL-6 decreases
the insulin sensitivity of hepatocytes via inhibiting its signal transduction (45). Besides
the crosstalk with hepatocytes, liver macrophages also regulate the fate of hepatic
stellate cells. Via releasing TGF-β (Transforming growth factor β), liver macrophages
promote the trans-differentiation of quiescent HSCs into their active forms (46) which
are proposed to be responsible for the development of liver fibrosis. Moreover, by

8
secreting IL-1β and TNF, macrophages also accelerate the progression of diseases
by enhancing the survival of activated HSCs in a NF-кB mediated manner (47) (Fig.3).

DAMPs
EVs
DAMPs
EVs
DAMPs
EVs
CCL2
CCL11
CXCL3
CCL2
CCL5
CCL2
CCL5
TNF-α
IL-1β
TGF-β
IL-10
IL-23
IL-12
IL-6
TNF-α
IL-1β
IL-6
TNF-α
IL-1β
TGF-β
LSEC
KC
monocyte
MoMF
Quiescent
HSC
Activated
HSC
T cell
B cell
hepatocytes
HMGB1
LPS
mtDNA
Bacterial
lipopeptide
Cholesterol
oxLDL
Viral RNA
TNF-α
IL-1β
TGF-β
Activated
KC
TLR2
TLR3
TLR4
TLR4
TLR9
P2X
7
bacterial
Apoptotic
hepatocytes
PAMPs
DAMPs
LPS
Lipoprotein
Peptidoglycan
HMGB1
mtDNA
ATP
Bacterial RNA
A
B
Figure 3. It shows crosstalk between KCs and other hepatic cells in NAFLD.
In the progression of NAFLD, apoptotic hepatocytes secrete DAMPs (e.g.,
HMGB1) and extracellular vesicles (EVs), both of which mediate the
activation of KCs. Activated KCs release CCL2 and CCL5 to recruit
circulating monocytes which can differentiate into MoMFs. Both activated
KCs and MoMFs release pro-inflammatory mediators (e.g., TNF-α, IL-1β,
IL-6) to aggravate the apoptosis of hepatocytes and further advance hepatic
inflammation. Meanwhile, activated KCs also promote the trans-
differentiation of quiescent HSCs into their active form via secreting TGF-β,
TNF-α and IL-1β. CCL2 and CCL5 produced by activated KCs further
enhance the migration of HSCs to injury sites. Cytokines (e.g., IL-6, IL-10,
IL-12 IL-23) secreted by KCs play crucial roles in the differentiation of T cells.

9
3. Hepatic macrophages
Liver harbors the most abundant population of macrophages in the body. Traditionally,
macrophages are simply classified into two categories based on their phenotypes
induced by specific cytokine treatments in vitro, M1 pro-inflammatory and M2 anti-
inflammatory macrophages. M1 macrophages are phenotypes classically induced by
the treatment of IL-12, IFN-γ (interferon-γ) and LPS. The cytokines includes IL-1 and
IL-6 secreted by these M1 macrophages are pro-inflammatory resembles. M1
macrophages are also phagocytic and secret reactive oxygen species. On the contrary,
M2 macrophages, induced by the treatment of IL-4 and IL-13, are responsible for
tissue remodeling and the release of immune-regulatory mediators that are anti-
inflammatory including IL-10 (48, 49). However, recent single cell RNAseq studies
have uncovered that it was difficult to define the macrophages as M1 or M2 in vivo,
since liver macrophages display a wide spectrum of gene expression profile far
beyond the central dogma of this dichotomous classification (50).

In order to investigate the role that hepatic macrophages play in the homeostasis and
the progression of liver disease, a classification based on cellular origins was also
proposed. Kupffer cells (KCs), the tissue-resident macrophages, and monocyte-
derived macrophages (MoMFs) are two major subsets based on this classification.
Besides, these two subsets can be distinguished from each other by their different
expression of cell surface markers (Table 1) and their functions (51). Murine KCs are
embryonically derived and can generally be distinguished based on their CD45
+
,

10
CD11b
low
, F4/80
high
, Clec4f
+
surface phenotype, while murine MoMFs originate from
circulating bone marrow-derived monocytes and are generally characterized as CD45
+
,
CD11b
hi
, F4/80
int
, Clec4f
-
cells (52, 53). Human macrophages however are less
characterized as there is a lack of specific markers for human KCs or monocytes-
derived macrophages. The most commonly used marker for human macrophages is
CD68. Recently, through the application of new high-throughput approaches, several
new markers are being discovered, shedding new lights on characterization of human
macrophages subsets (54, 55).

4. Hepatic macrophages in healthy and disease state
In healthy livers, KCs are the dominant subset of hepatic macrophages that play
crucial roles in maintaining liver homeostasis. Since KCs reside in sinusoidal lumen
and are attached to endothelial cells, they are able to sample the blood from portal
vein as well as the main circulation. KCs express a range of pattern recognition
receptors, and serve as the first-line defense against translocated gut microbiota and
Human Mouse Type of cells
CD68
+
Marco
+
CD14
+
Vsig4
+
CD11b
low
F4/80
high
Clec4f
+
Vsig4
+
Timd4
+
Kupffer cells
CD14
+
CCR2
+
CD11b
high
F4/80
low
Ly6C
+
CCR2
+
CX3CR1
+
Monocytes-derived macrophages
CD14
+
CD16
+
CCR2
+
Ly6C
+
Monocytes
Table 1. Identified markers of Kupffer cells, monocytes-derived macrophages and monocytes

11
other insoluble macromolecules (56). The Gut-derived endotoxin LPS which is
consisted of a lipid and a polysaccharide is rich in the outer membrane of gram-
negative bacteria and is able to induce potent inflammatory response by binding to
TLR4. Upon LPS binding, KCs produce low level of pro-inflammatory cytokines such
as IL-6 (57) which in turn triggers the hepatocellular production of acute-phase
proteins and promote phagocytosis (58).

In steady state, KCs and other liver immune cells are continuously exposed to gut
derived pathogens that enter the liver sinusoids through the portal vein. To avoid
sustained inflammatory response and subsequent liver damage, KCs have developed
a mechanism to repress the excessive activation of immune system, which is known
as immunological tolerance. Located on the luminal side of the liver sinusoids, KCs
exhibit a spectrum ability of antigen presenting cells (APC) to recognize and
phagocytic pathogens. Uptake of pathogens in APCs normally result in antigen
presentation and T cell activation. However, due to the low expression of MHC-II and
co-stimulatory molecules, KCs do not activate antigen specific T cells [58]. In addition,
KCs can also secret prostaglandins PGE2 and 15-deoxy-delta12,14-PGJ2 (15d-PGJ2)
to suppress effector T cells induced by other APCs (59). Consequently, deletion of KCs
activates antigen-specific T cells and abrogates liver immunological tolerance (60).
Beside their effect on antigen-specific T cells, with high expression of PD-L1, KCs can
simultaneously expand regulatory T cells (Treg) which can secrete immunosuppressive

12
factor interleukin-10 (IL-10) to further promote hepatic tolerance (61, 62) (Fig.4).

The immunological tolerance mediated by KCs is strictly limited to healthy conditions.
With high concentration of LPS, not only the suppressive activity of Treg can be
overcome, but also the proliferation of antigen-specific T cells can be stimulated (63).
Furthermore, LPS also stimulates KCs to secret C-C motif chemokine ligand 2 (CCL2),
also known as MCP-1 (monocyte chemoattractant protein-1). Through TLR/MyD88
signaling pathway, CCL-2 are crucial for the recruitment of CCR2
+
MoMFs into the
liver. Massive recruitment of pro-inflammatory MoMFs changes the composition of the
liver APCs by lowering the percentage of KCs and thus negates the effect of Kupffer
cell mediated immune suppression. In mouse model of NASH induced by choline-
Figure 4. It shows intercellular signals of KCs in healthy state. Due to the
low expression of MHC-II and co-stimulation molecules. KCs do not
activate antigen specific T cells. Meanwhile, PGE2 and 15d-PGJ2 secreted
by KCs can simultaneously inhibit the activity of effector T cells induced by
other APCs. Moreover, with high expression of PD-L1, KCs can expand IL-
10
+
Treg to further promote hepatic tolerance.

13
deficient amino acid-defined (CDAA) diet, less lipid accumulation and ameliorated
inflammation were observed in CCR2
-/-
, TLR4
-/-
and MyD88
-/-
mice compared with wild
type mice. This was also accompanied with reduced infiltrating MoMFs (64). Another
study elucidates that after oral administration of cenicriviroc (CVC), an antagonist of
CCR2, to MCD diet-challenged mice, reduced steatohepatitis and inhibited MoMF
infiltration is observed when compared with the control group (65). In addition to
CCR2-CCL2, other chemokines pathway like CXCR3-CXCL10 and CCR8-CCL1 can
also contribute to KC recruitment of MoMFs (66, 67).

In healthy adults, KCs are maintained by local self-renewal without the contribution of
circulating monocytes (52, 68). Chronic liver diseases often disturb KC homeostasis.
In NASH, there is a concurrent increase of KC death and KC replenishment which
results in increase of total number of KCs (69, 70). Studies using bone-marrow
chimeras (CD45.1
+
donors and CD45.2
+
Ccr2
-/-
recipients) showed that some CD45.1
monocytes developed KC-like characteristics after feeding with MCD-diet (69). In other
words, a large number of monocyte-derived cells acquired “Kupffer cell phenotype”,
which was initiated by the embryo-derived KCs death in NASH condition. Although
these cells have a KC phenotype, they still have some differences in the expression
profile with embryo-derived KCs. Another recent study showed that after injecting DT
to mice whose KCs specifically expressed DTR, the embryo-derived KCs could be
completely eliminated and recruited monocytes-derived cells repopulated KC niches
(71). Based on RNA-seq analysis of these KC-like cells, the results suggested that

14
Clec4f, the common marker of KCs, was rapidly upregulated in the MoMFs after the
depletion of embryo-derived KCs. At the same time, another specific marker of
embryo-derived KCs, Timd4, was also upregulated but it could not reach the same
expression as in embryo-derived KCs (71). Although these KC-like cells exhibited
similar capacity of clearing senescent red blood cells compared to embryo-derived
KCs, they are less efficient in uptake of acetylated low density lipoprotein (72).

5. Heterogeneity of Kupffer cells
With the emergence of single-cell approaches, the heterogeneity of distinct KCs
subpopulations is more comprehensively explored (53-55, 73, 74). Analyzing the
single cell RNA-seq data, the researchers revealed 2 distinct subsets of murine KCs
in healthy state (53). Further investigation of their transcriptomic signatures indicated
that CD206 and endothelial cell adhesion molecule (ESAM) are among the top
differentially expressed genes. Therefore, they defined these two subsets as
CD206
lo
ESAM
-
KC1 and CD206
hi
ESAM
+
KC2. The most noticeable difference
between these two subsets is their expression profile at different pathological stages.
KC2 population is distinct for their upregulated expression of genes involved in lipid
metabolism and oxidative stress and found associated with high-fat diet (HFD) induced
metabolic disorders including fatty liver. KC1 population though is distinct for their
upregulated expression of genes involved in immune response. The lipid metabolism
function exhibited by KC2 was due to their specifically high expression of CD36, a
crucial protein involved in the uptake of lipid. The deletion of KC2 in mice was shown

15
to reduce diet-induced metabolic impairments (53). However, this view is challenged
by other researchers who argue that the newly identified CD206
hi
ESAM
+
KC2 were
cell doublets consisted of liver sinusoidal endothelial cell (LSECs) and KC (75). By
applying the same gating strategy to B cells, a subpopulation of B cells was found and
they highly expressed ESAM and CD206. These ESAM and CD206 expressing B cells
and KC2 both highly expressed other markers of LSECs, such as CD26, CD31 and
CD38, indicating a potential of cell doublets (75).

Via the application of single-cell genomics technologies, a recent study revealed the
heterogeneity of human macrophages as well as other intrahepatic immune cells
including B cells, T cells, NK cells and so on (54). Based on the analysis of t-SNE (t-
distributed stochastic neighbor embedding) plot, two macrophage subsets were also
identified. The CD68
+
MARCO
-
subset was distinct for being enriched in LYZ, CSTA
and CD74, which was similar to the recruited MoMFs in mice, whereas the other
CD68
+
MARCO
+
subset was distinct for being enriched in VSIG4, CD163 and HMOX1,
which resembled the characterized murine KCs. The transcriptional characterization
of these KC populations also indicated their different functionality: one is involved in
inflammation while the other is responsible for hepatic tolerance (54). In addition,
another study characterized the existence of two subpopulations within the tolerogenic
macrophages. Both populations were enriched in the expression of CD163, MARCO
and VCAM1, while the expression of TIMD4 can distinguished them from each other.
Compared with healthy people, patients with cirrhosis showed a selective loss of

16
MARCO
+
TIMD4
-
subpopulation (73). Other studies also tried to define human
macrophage subsets using other sets of markers. A CD68
+
CD32
hi
CD14
int
tolerogenic
subtype and CD68
+
CD32
int
CD14
hi
inflammatory subtypes were defined, but further
investigations of their biological function were still needed (74).

Goal of Project and Scope of the Study
Immune cells are highly versatile and have many subsets. This heterogeneity
characteristic is especially well known in B and T cells. Each subset of the immune
cells are known to exert their unique functions by specifically expressing a range of
receptors capable of recognizing diverse ligands and activating the downstream
signaling pathways. With the realization of heterogeneity of immune cells, more and
more studies are focused on investigating the distinct roles of these subpopulations in
various pathophysiology conditions. However, there is a lack of specific markers to
identify these subpopulations without in depth transcriptomic analysis such as single
cell RNAseq. Meanwhile, as illustrated above, KCs play significant functions in both
physiological and pathophysiological progresses in the liver, including detoxification
and digestion the pathogens or gut microbiota metabolites flowing from portal vein,
clearance of damaged or dead cells as well as participation in chronic inflammatory
response that can lead to NASH, fibrosis and HCC. In order to further study the precise
role that KCs play in the chronic liver disease, specific markers for KCs are important.
Unfortunately, there are very few identified markers. The most widely used ones for
mice are Clec4f, Vsig4 and Timd4 (69, 76, 77), and no well-defined markers for human

17
KCs has been reported. Without characterizing KCs, the involvement of KCs in the
pathologies will remain elusive.

Even though there is still debate on the definition of KC subpopulations, it is known
that there are at least embryonic KCs vs monocyte derived KCs and studies have
shown that monocyte derived KCs replace embryonic KCs during the progression of
NASH (69, 77). The functional differences between these KCs are largely unclear.
Given this consideration, the goal of this project is to characterize KC subpopulations
by identifying novel markers. Using bioinformatics screening, we analyzed the
published single cell RNA-seq data as our discovery dataset and ranked the top 500
genes that are specifically enriched in KCs. This analysis identified Cd5l and Folr2 as
two potential novel markers for Kupffer cells. We further validate these markers by
using in silicone analysis to demonstrate the correlation between these two candidates
and Clec4f, one of the well characterized KC markers. Both markers show great
correlations with Clec4f and Vsig4, and are not correlated with the markers of other
cell types. Using two steatosis mouse models representing ALD/ASH and a
NAFLD/NASH, we confirmed the expression correlation of Cd5l and Clec4f in the livers
using qPCR. In addition, we further confirmed the main source of Cd5l is KCs through
immunofluorescent staining and flow cytometry. Collectively, our data shows that Cd5l
is a putative novel marker for KCs.

18
Chapter II Methods
1. Intracellular immunofluorescent staining
Bake paraffin-embedded liver section slides on slides baker at 63°C (overnight for
mouse slides, 1 hour for human slides). Slides were deparaffinized in xylene twice for
10min each time, then dehydrated in 100%, 95%, 70%, 50% ethanol respectively. Run
slides under tap water for 2min. Slides were retrieved by antigen retrieving buffer (0.1M
citric acid: 4.5mL, 0.1M sodium citrate: 20.5mL, distilled water: 22.5mL) in water bath
for 22.5min (settings of microwave: higher power for 7.5min, power at 20% for 15min).
After heating, let the slides cool down for 20min. Run slides under running water for
20min. Use pap pen to draw a circle around tissue. During the process, ensure the
tissue doesn’t dry out. Wash with PBS for 5min. Add 1% Tween-20 (diluted in PBS) on
each tissue and let it permeabilized for 15min. Shake off solution and wash with PBS
for 3min. Add 1 X universal blocking solution on each tissue and let it sit for only 10min.
Shake off the blocking solution and wash with PBS for 3min. Add 10% goat serum
(diluted in PBS) to block non-specific binding sites for 30min at room temperature.
Shake off goat serum and add diluted primary antibody. Put all slides in humidifier and
incubate at 4°C overnight in the dark. Wash slides with diluted sensitive washing buffer
(20 X Wash buffer: 4mL, distilled water: 76mL, PBS: 160mL) for 10min on shaker. Dip
in PBS for several times and shake off excessive solution. Add diluted fluorescein-
conjugated secondary antibody and incubate at room temperature for 1h in the dark.
Wash again with sensitive washing buffer and PBS as described above. Add the other
primary and secondary antibody as described above. After washing the slide, add

19
diluted DAPI solution (1:10 in PBS) for nuclear staining. Cover the tissue with cover
slips, and then images can be taken with the Zeiss immunofluorescent microscope
using a 20 X 0.75 dry objective lens. Primary antibodies used to stain the slides
included anti-Clec4f, CD5L antibodies. The fluorescein-conjugated secondary
antibodies used in this study were goat anti-rabbit IgG-Cy3 (1:200), goat anti-mouse
IgG AF488 (1:200).

2. Isolation of non-parenchymal cells
Immediately after anesthesia, the superior vena cava is cannulated and the liver is
perfused with BSA buffer (20mg/ml in Hanks’ buffer with 34.2mg/ml EDTA) at 37°C for
3min to wash out blood and loosen cell-cell connection. Then it is digested with
collagenase solution (4mg/ml in Hanks’ buffer with 1mM CaCl2) at 37°C for 2min to
dissociate the extracellular matrix. Dissect out the liver gently and grind it with
borosilicate glass. Remove unblended tissue with filter and let the filtered suspension
settle down for 30min on ice. Transfer the supernatant to fresh tubes and centrifuge at
900 X g for 10min. Add 20ml red blood cell lysis buffer. Resuspend all the cells and
incubate for 15min. Centrifuge at 900 X g for 10min and the cells precipitating at the
bottom of the tube is non-parenchymal cells.

3. Intracellular flow cytometry
Non-parenchymal progenitor cells were obtained by perfusing and digesting the whole
murine liver. Resuspend all the cells in 5ml of FACS buffer (1%FBS + DPBS) and

20
thoroughly mix to avoid any cell clumps. Count the isolated immune cells. For each
1X10
6
cells in 1.5 ml tube, perform centrifugation and discard the supernatant.
Resuspend cells and stain the extracellular antigen (CD45, CD11b, F4/80, Clec4f) with
diluted fluorescence-conjugated antibodies (diluted in FACS buffer) for 30 min on ice
in the dark. Rinse the cells twice with FACS buffer. Resuspend cells and fix them by
2.5% paraformaldehyde solution (25% paraformaldehyde stock solution + double
distilled water) at room temperature for 8 minutes. Centrifuge and gently remove the
supernatant. Fix the cells with the same procedure in step6 again. Rinse them twice
with FACS buffer. For each 1X10
6
cells, resuspend and permeabilize them with 0.1%
Tween-20 solution (0.1% Tween-20 in FACS buffer) at room temperature for 8 min.
Centrifuge and gently remove the supernatant. Permeabilize the cells with the same
procedure above again. Resuspend cells and stain the intracellular antigen with diluted
primary antibody (diluted in FACS buffer contained 0.1% Tween-20) for 30 min at room
temperature in the dark. Rinse the cells twice with FACS buffer. For each 1X10
6
cells,
resuspend and stain the intracellular antigen with diluted fluorescence-conjugated
secondary antibody (diluted in FACS buffer contained 0.1% Tween-20) for 30 min at
room temperature in the dark. Rinse the cells twice with FACS buffer. Resuspend the
cells in 500μL of FACS buffer. At least 30,000 hepatic immune cells are subjected to
intracellular flow cytometry analysis by employ the BD FACSVerse flow cytometer.
Experimental data are analysis by FlowJo.

4. RNA isolation, reverse transcription and quantitative PCR

21
A small section of liver is lysed and homogenized with 1mL of TRIZOL reagent.
Centrifuge to remove unblended tissue and transfer the supernatant to fresh tubes.
Add 200μL of chloroform to each tube for per 1mL of TRIZOL. Vortex samples
vigorously for 15sec and incubate them at room temperature for 15min. Centrifuge the
samples at 12,000 X g for 15min at 4°C. After centrifugation, the mixture separates
into lower red phenol-chloroform phase, an interphase and a colorless upper aqueous
phase. RNA remains exclusively in the aqueous phase. Transfer the upper aqueous
phase carefully without disturbing the interphase into new tubes. Precipitate RNA from
aqueous phase by mixing with isopropyl alcohol. Add 500μL of isopropyl alcohol for
per 1mL of TRIZOL reagent used for the initial homogenization. Incubate samples at -
20°C for 15min. Centrifuge at 12,000 X g for 10min at 4°C. RNA is often visible before
centrifugation, which forms a gel-like pellet at the bottom of the tubes. Remove the
supernatant completely. Wash the RNA pellet once with 75% ethanol. Add 1mL of 75%
cold ethanol (diluted with DEPC water and cooled down to -20°C) for per 1mL of
TRIZOL reagent used for the initial homogenization. Mix samples by vortexing and
centrifuging at 7,500 X g for 5min at 4°C. Air-dry RNA pellet for 5-10 minutes. It is
important not to let the RNA pellet dry completely as this will greatly decrease its
solubility. Dissolve RNA in DEPC-treated water by passing solution a few times
through a pipette tip. The RNA purity and concentration are determined by using a
NanoDrop ND-1000.

cDNA synthesis was performed with M-MLV reverse transcriptase system by using

22
1μg of total RNA for each sample. The reaction was performed in an Eppendorf
Mastercycler Gradient following the designed program. Quantitative PCR was
performed using Syber Green Master Mix and 7900 HT Standard Real-Time PCR
System. Relative gene expression is analyzed by the 2
!∆∆#$
method. Housekeeping
gene 18s was used as endogenous control.

5. Biochemical analysis
Plasma samples were collected from the eyes at the end of animal studies. A mouse
CD5L ELISA kit from RayBiotech was used to quantify CD5L plasma concentration.
Make a 10000-fold dilution for each plasma sample with 1 X Assay Diluent and mix
well. Prepare all the standard solution and use 1 X Assay Diluent as the zero standard.
Add 100μL of each standard and diluted samples into each well. Cover the wells and
incubate for 2.5h at room temperature with gentle shaking. Discard the solution and
wash 4 times with 1 X Wash Solution. Wash each well by 300μL of Wash Buffer.
Complete removal of liquid at each step is essential for good results. After the final
wash, remove any remaining Wash Buffer by inverting the plate and blotting it against
clean paper towels. Add 100μL of 1 X biotinylated antibody to each well and incubate
for 1h at room temperature with gentle shaking. Discard the solution and wash wells
as described above. After washing, add 100μL of Streptavidin solution to each well
and incubate for 45min at room temperature with gentle shaking. Discard the solution
and wash wells as described above. After washing, add 100μL of substrate reagent to
each well and incubate for 30min at room temperature in the dark with gentle shaking.

23
Finally, add 50μL of stop solution to each well and read the absorbance of each well
at 450nm immediately.

24
Chapter III Results
1. Identification of new Kupffer cell markers
In order to identify new Kupffer cell markers, we analyzed a published single cell RNA-
seq dataset which is available at www.livercellatlas.org. Our analysis is based on the
dataset of total CD45
+
liver cells which were isolated from mice fed with western diet
or normal chow. Four files could be acquired from the website: barcodes, features,
matrix and cell annotation. According to single cell RNA-seq workflow, each cell was
labelled with a unique barcode which was comprised of sixteen oligo-nucleotides. The
sequence of those barcodes were stored in the barcode file (Fig.5A). The features file
contained the name of genes detected in the sequencing (Fig.5B). The number of
reads for those detected genes was saved in the matrix file. In matrix file, each column
represented one single cell and each row represented one gene. In other words, each
number in the matrix file indicated the number of reads for a gene in a specific cell
(Fig.5C). After calculating highly variable genes, the researchers classified all the cells
into several groups and identified their cell type based on their expression of
commonly used markers. T cells, B cells, NK cells, KCs and many other cell types
were characterized. The information about the cell type of each single cell could be
obtained from the cell annotation file (Fig.5D).

25

AAACCCAAGGTTCTTG-1 AAACCCAAGGTATCTC-1 AAACCCAAGCTAGAAT-1
8 4 1 Xkr4
12 6 3 Gm1992
3 4 7 Gm37381
20 14 4 Rp1
40 8 6 Sox17
matrix
Figure 3. (A-D) They are parts of four files downloaded from liver cell atlas website. (E) It shows part of codes used for the analysis of human
single cell RNA-seq data. (E) It shows part of codes used for the analysis of murine single cell RNA-seq data.
A B
C
D cell annotation
E
KCs
other
cells
setwd(‘/cd5l’)
annotation=read.csv('annot_human.csv', header=TRUE)
matrix=readMM('matrix.mtx’)
barcodes=read.table('barcodes.tsv’)
features=read.table('features.tsv')
macrophage_columns=grep(‘Macrophages', annotation[,4])
macrophage_barcodes=annotation[macrophage_columns,7]
a=1
for (i in 1:7076){
a[i]=grep(macrophage_barcodes[i],barcodes[,1])
}
macrophage_expression=matrix[,a]
macrophage_mean=1
for (i in 1:32738){
macrophage_mean[i]=mean(macrophage_expression[i,])
}
for (i in 1:32738){
+
cat(c(features[i,1],macrophage_mean[i],'\n'),file='names&mea
ns.csv',sep=',',append=TRUE)
+ }
read files
grab expression of each gene
in macrophages from matrix file
calculate the mean expression
of each gene in macrophages
and export the result into a
spreadsheet
Figure 5. (A-D) They are parts of four files downloaded from liver cell atlas
website. (E) It shows part of codes used for the analysis of human single
cell RNA-seq data.

26
To identify new markers for KC, we developed R code to analyze 2
parameters in these dataset. To do this, we first divided all cells into KCs
group and other cell types group (Fig.3E). We calculated the mean
expression of each gene in the KC group vs. other cell types group and
calculated their expression ratio (Expression
ratio=
!"#$ "&'("))*+$ *$ ,-)
!"#$ "&'("))*+$ *$ +./"( 0"11 .2'")
) to define KC specificity. Top 500 genes
within the highest expression ratio ranking were then selected (Fig.6).

Figure 6. It shows the flow chart of the whole project.
Acquire the barcodes of KCs and other cells from the
annotation file
Extract columns of the KCs and other cells from the matrix of
expression data
Calculate the mean expression of each gene in the KCs vs
other cells.
Calculate the expression ratio of each gene and rank all the
genes bases on that.
Select promising candidates and run t test for them
Figure 4. It shows the flow chart of the whole project.

27
A second criteria we used to define novel KC genes is that they should be detected
easily during the assay. We thus used expression levels as the 2
nd
parameter. The 500
selected genes were thus further analyzed based on their mean expression in KCs
and expression ratio (Fig.7A). Based on these two criteria, we defined an area that
included genes that were among the top 50 in expression ratio ranking and whose
mean expression in KCs was set to be higher than 1. Three widely used KCs markers,
Clec4f, Vsig4 and Timd4 were all found in this area (shown in magenta). Moreover,
Marco and Cd163, two markers that are preferentially expressed by KCs are also
located in this area (shown in red). Three other genes in this quadrant, Slc40al, Fcna,
C6 are also highly expressed by liver endothelial cells, hepatocytes or hepatic stellate
cells based on the single-cell RNA seq data (shown in green). Fabp7, CXCL13, Cd5l
and Flor2 have the same distribution in cell types with Clec4f (blue and yellow).
However, compared with other two markers, we found that Cd5l and Folr2 (yellow)
were expressed not only by murine KCs but also by human KCs (Fig.7B). Taking all
above into consideration, we decided to focus on Cd5l and Folr2, two potential KCs
markers in the subsequent validation experiments.

28
Figure 5. Identification of new markers for KCs. (A) Analysis of public mouse single cell RNA-seq. It shows top 500 genes within the rank of
expression ratio. X axis presents the expression ratio of each gene, while Y axis represents their mean KC expression ratio. On the right side of the
vertical line are the top 50 genes within the expression ratio ranking. On the upper side of the horizonal line are the genes whose mean expression
in KCs is higher than 1. Promising markers of KCs are supposed to be in the upper-right quadrant. Three well-known markers for murine KCs
(Clec4f, Vsig4, Timd4) are marked in magenta. Two newly identified markers (Cd5l, Folr2) are marked in yellow. (B) Analysis of public human single
cell RNA-seq. The quadrant was defined by the same criteria of mouse dataset. Four identified markers for human KCs (MARCO, VCAM1, CETP,
LYVE1) are marked in magenta. Two newly identified markers (CD5L, FOLR2) are marked in yellow.
A
B
8 16 32 64 128 256 512
0.0001
0.001
0.01
0.1
1
10
100
1000
expression ratio
macrophage expression
CD5L
VCAM1
FOLR2
CETP
MARCO
C1QC
C1QB
C1QA
SLC40A1
LYVE1
C17orf102
AXL
CD163
Vsig4
CD68
MARCO
VCAM1
CETP
LYVE1
FOLR2
CD5L
CXCL12
SDC3
Figure 7. Identification of new markers for KCs. (A) Analysis of public mouse
single cell RNA-seq. It shows top 500 genes within the rank of expression
ratio. X axis presents the expression ratio of each gene, while Y axis
represents their mean KC expression ratio. (B) Analysis of public human
single cell RNA-seq. The quadrant was defined by the same criteria of
mouse dataset.

29
2. In silico verifications
In order to validate these two genes as KC markers, we first performed in silico
verification by analyzing two public datasets: GSE90497 and GSE149863. We
developed R code to obtain the detected genes and their expression in each sample.
To validate the expression of Folr2 and Cd5l as markers for KCs, we performed linear
correlation between them and the two widely used murine KC markers, Clec4f and
Vsig4, respectively. As positive controls, Clec4f and Vsig4 showed R
2
of 0.511 and
0.5664 in GSE149863 and GSE90497 respectively (Fig.10A, Fig.13A). Pearson
correlation analysis displayed strong linear correlation of Folr2 (R
2
=0.6268 for
GSE149863; R
2
=0.4902 for GSE90497) and Cd5l (R
2
=0.7107 for GSE149863;
R
2
=0.7094 for GSE90947) with Clec4f (Fig.8-9A, Fig.11-12A). Similarly, both KC
candidate genes also demonstrate strong correlation with Vsig4 (GSE149863:
R
2
=0.5963, R
2
=0.4021, GSE90497: R
2
=0.6582, R
2
=0.7458 for Folr2 and Cdl5
respectively) (Fig.8-9B, Fig.11-12B). Together, these data suggests that Cd5l and
Folr2 may represent a wider population of KC than the two widely used markers. To
further validate the markers are specific to KCs and are not to other cell types, we
performed similar correlation of these two promising markers, Flor2 and CD5l, with
markers of B cells (Cd79a), total macrophage (Adgre1), monocytes (Ly6c1), MoMFs
(Ccr2), neutrophils (Slfn4) and T cells (Cd3g) (Fig.8-13). Validating our initial analysis,
only strong correlation is observed for Adgre1, a macrophage marker in both dataset.
Moderate correlation was observed for T cells in one but not the other dataset. To be
noted, Clec4f also demonstrated correlation with T cells in this dataset, suggesting

30
potential impurity of cell population in the experiment performed to obtain the data.

31

Figure 8. Correlation analysis of Cd5l and markers of different cell types
based on GSE 149863. (A&B) It shows that Cd5l has strong correlation with
Clec4f and Vsig4, two well-characterized KCs markers, respectively. (C) It
show that Cd5l and Adgre1, an identified markers of total macrophages has
strong correlation. (D&E) It show that Cd5l are weakly correlated with Ly6c1
and Ccr2, two known MoMFs markers, respectively. (F) Slfn4 is the marker
of neutrophils. It has weak correlation with Cd5l. (G) It shows that Cd5l is
weakly correlated with Cd3g, the marker of T cells. (H) It reveals that Cd79a,
the marker of B cells, is barely correlated with Cd5l.
A B
C D
E F
G H
R² = 0.6268
0
50
100
150
200
250
300
350
0 50 100 150 200 250
expression amount of Cd5l
expression amount of Clec4f
Correlation between Clec4f and Cd5l
P=4.212e
-20 R² = 0.5963
0
50
100
150
200
250
300
350
0 20 40 60 80 100 120
expression amount of Cd5l
expression amount of Vsig4
Correlation between Vsig4 and Cd5l
P=1.266e
-18
R² = 0.6346
0
50
100
150
200
250
300
350
0 10 20 30 40 50
expression amount of Cd5l
expression amount of Adgre1
Correlation between Adgre1 and Cd5l
P=1.696e
-20
R² = 0.0884
0
50
100
150
200
250
300
350
0 10 20 30 40
expression amount of Cd5l
expression amount of Ly6c1
Correlation between Ly6c1 and Cd5l
P=0.005
R² = 0.1134
0
50
100
150
200
250
300
350
0 2 4 6 8
expression amount of Cd5l
expression amount of Ccr2
Correlation between Ccr2 and Cd5l
P=0.001
R² = 0.2379
0
50
100
150
200
250
300
350
0 1 2 3 4 5
expression amount of Cd5l
expression amount of Slfn4
Correlation between Slfn4 and Cd5l
P=1.433e
-06
R² = 0.4172
0
50
100
150
200
250
300
350
0 2 4 6
expression amount of Cd5l
expression amount of Cd3g
Correlation between Cd3g and Cd5l
P=1.085e
-11
R² = 0.1456
0
50
100
150
200
250
300
350
0 0.5 1 1.5 2
expression amount of Cd5l
expression amount of Cd79a
Correlation between Cd79a and Cd5l
P=0.0002

32

Figure 9. Correlation analysis of Folr2 and markers of different cell types
based on GSE 149863. (A&B) It shows that Folr2 has strong correlation with
Clec4f and Vsig4, two well-characterized KCs markers, respectively. (C) It
show that Folr2 and Adgre1, an identified markers of total macrophages has
strong correlation. (D&E) It show that Folr2 are weakly correlated with Ly6c1
and Ccr2, two known MoMFs markers, respectively. (F) Slfn4 is the marker
of neutrophils. It has weak correlation with Folr2. (G) It shows that Folr2 is
weakly correlated with Cd3g, the marker of T cells. (H) It reveals that Cd79a,
the marker of B cells, is barely correlated with Folr2.
A B
C D
E F
G H
R² = 0.4902
0
10
20
30
40
50
60
0 50 100 150 200 250
expression amount of Folr2
expression amount of Clec4f
Correlation between Clec4f and Folr2
P=3.474e
-14
R² = 0.4021
0
10
20
30
40
50
60
0 20 40 60 80 100 120
expression amount of Folr2
expression amount of Vsig4
Correlation between Vsig4 and Folr2
P=2.351e
-11
R² = 0.5735
0
10
20
30
40
50
60
0 10 20 30 40 50
expression amount of Folr2
expression amount of Adgre1
Correlation between Adgre1 and Folr2
P=1.375e
-17
R² = 0.0409
0
10
20
30
40
50
60
0 10 20 30 40
expression amount of Folr2
expression amount of Ly6c1
Correlation between Ly6c1 and Folr2
P=0.059
R² = 0.1242
0
10
20
30
40
50
60
0 2 4 6 8
expression amount of Folr2
expression amount of Ccr2
Correlation between Ccr2 and Folr2
P=0.0008
R² = 0.1999
0
10
20
30
40
50
60
0 1 2 3 4 5
expression amount of Folr2
expression amount of Slfn4
Correlation between Slfn4 and Folr2
P=1.259e
-05
R² = 0.4136
0
10
20
30
40
50
60
0 1 2 3 4 5 6
expression amount of Folr2
expression amount of Cd3g
Correlation between Cd3g and Folr2
P=1.414e
-11
R² = 0.1233
0
10
20
30
40
50
60
0 0.5 1 1.5 2
expression amount of Folr2
expression amount of Cd79a
Correlation between Cd79a and Folr2
P=0.0008

33

Figure 10. Correlation analysis of markers of different cell types based on
GSE 149863. (A) Clec4f and Vsig4 are two well-characterized KCs markers.
Their correlation is used as a positive control for analysis. (B) Adgre1 is an
identified markers of total macrophages. The correlation between it and
Clec4f is used as a negative control for analysis. (C)The correlation between
Clec4f and Ly6c, a MoMFs marker, is used as a negative control for analysis.
(D) Slfn4 is the marker of neutrophils. The correlation between it and Clec4f
is used as a negative control for analysis. (E) The correlation between Cd3g,
the marker of T cells, and Clec4f is used as a negative control for analysis.
(F) The correlation between Cd79a, the marker of B cells, and Clec4f is used
as a negative control for analysis.

R² = 0.1037
0
50
100
150
200
250
0 0.5 1 1.5 2
expression amount of Clec4f
expression amount of Cd79a
Correlation between Cd79a and Clec4f
R² = 0.3141
0
50
100
150
200
250
0 1 2 3 4 5 6
expression amount of Clec4f
expression amount of Cd3g
Correlation between Cd3g and Clec4f
R² = 0.0718
0
50
100
150
200
250
0 10 20 30 40
expression amount of Clec4f
expression amount of Ly6c1
Correlation between Ly6c1 and Clec4f
R² = 0.1969
0
50
100
150
200
250
0 1 2 3 4 5
expression amount of Clec4f
expression amount of Slfn4
Correlation between Slfn4 and Clec4f
R² = 0.5066
0
50
100
150
200
250
0 10 20 30 40 50
expression amount of Clec4f
expression amount of Adgre1
Correlation between Adgre1 and Clec4f
R² = 0.511
0
20
40
60
80
100
120
0 50 100 150 200 250
expression amount of Vsig4
expression amount of Clec4f
Correlation between Clec4f and Vsig4
A B
C D
E F
P=5.187e
-15 P=7.647e
-15
P=1.492e
-05
P=0.012
P=1.357e
-08
P=0.002

34

Figure 11. Correlation analysis of Cd5l and markers of different cell types
based on GSE 90497. (A&B) It shows that Cd5l has strong correlation with
Clec4f and Vsig4, two well-characterized KCs markers, respectively. (C) It
show that Cd5l and Adgre1, an identified markers of total macrophages has
strong correlation. (D&E) It show that Cd5l are weakly correlated with Ly6c1
and Ccr2, two known MoMFs markers, respectively. (F) Slfn4 is the marker
of neutrophils. It has weak correlation with Cd5l. (G) It shows that Cd5l is
weakly correlated with Cd3g, the marker of T cells. (H) It reveals that Cd79a,
the marker of B cells, is barely correlated with Cd5l.

R² = 0.7107
0
500
1000
1500
2000
2500
0 500 1000 1500 2000
expression amount of Cd5l
expression amount of Clec4f
Correlation between Clec4f and Cd5l
R² = 0.6582
0
500
1000
1500
2000
2500
0 100 200 300 400 500 600
expression amount of Cd5l
expression amount of Vsig4
Correlation between Vsig4 and Cd5l
R² = 0.1132
0
500
1000
1500
2000
2500
0 100 200 300 400 500
expression amount of Cd5l
expression amount of Adgre1
Correlation between Adgre1 and Cd5l
R² = 0.2112
0
500
1000
1500
2000
2500
0 50 100 150 200
expression amount of Cd5l
expression amount of Ccr2
Correlation between Ccr2 and Cd5l
R² = 0.0003
0
500
1000
1500
2000
2500
2 3 4 5 6
expression amount of Cd5l
expression amount of Slfn4
Correlation between Slfn4 and Cd5l
R² = 0.1637
0
500
1000
1500
2000
2500
0 10 20 30 40
expression amount of Cd5l
expresion amount of Cd3g
Correlation between Cd3g and Cd5l
R² = 0.0102
0
500
1000
1500
2000
2500
6 7 8 9 10 11 12
expression amount of Cd5l
expression amount of Cd79a
Correlation between Cd79a and Cd5l
R² = 0.3012
0
500
1000
1500
2000
2500
0 20 40 60 80 100
expression amount of Cd5l
expression amount of Ly6c1
Correlation between Ly6c1 and Cd5l
A B
C D
E F
G H
P=1.654e
-11
P=3.748e
-10
P=0.017
P=0.0003
P=0.003
P=0.844
P=0.011
P=0.542

35

Figure 12. Correlation analysis of Folr2 and markers of different cell types
based on GSE 90497. (A&B) It shows that Folr2 has strong correlation with
Clec4f and Vsig4, two well-characterized KCs markers, respectively. (C) It
show that Folr2 and Adgre1, an identified markers of total macrophages has
strong correlation. (D&E) It show that Folr2 are weakly correlated with Ly6c1
and Ccr2, two known MoMFs markers, respectively. (F) Slfn4 is the marker
of neutrophils. It has weak correlation with Folr2. (G) It shows that Folr2 is
weakly correlated with Cd3g, the marker of T cells. (H) It reveals that Cd79a,
the marker of B cells, is barely correlated with Folr2.

R² = 0.7094
0
20
40
60
80
100
120
140
160
180
0 500 1000 1500 2000
expression amount of Folr2
expression amount of Clec4f
Correlation between Clec4f and Folr2
R² = 0.7458
0
20
40
60
80
100
120
140
160
180
0 100 200 300 400 500 600
expression amount of Folr2
expression amount of Vsig4
Correlation between Vsig4 and Folr2
R² = 0.0011
0
20
40
60
80
100
120
140
160
0 100 200 300 400 500
expression amount of Folr2
expression amount of Adgre1
Correlation between Adgre1 and Folr2
R² = 0.4116
0
20
40
60
80
100
120
140
160
0 50 100 150 200
expression amount of Folr2
expression amount of Ccr2
Correlation between Ccr2 and Folr2
R² = 0.5933
0
20
40
60
80
100
120
140
160
0 20 40 60 80 100
expression amount of Folr2
expression amount of Ly6c1
Correlation between Ly6c1 and Folr2
R² = 0.0482
0
20
40
60
80
100
120
140
160
2 3 4 5 6
expression amount of Folr2
expression amount of Slfn4
Correlation between Slfn4 and Folr2
R² = 0.1845
0
20
40
60
80
100
120
140
160
0 10 20 30 40
expression amount of Folr2
expression amount of Cd3g
Correlation between Cd3g and Folr2
R² = 0.0115
0
20
40
60
80
100
120
140
160
6 8 10 12
expression amount of Folr2
expression amount of Cd79a
Correlation between Cd79a and Folr2
A B
C D
E F
G H
P=1.802e
-11
P=1.480e
-12
P=0.839
P=9.811e
-09
P=1.075e
-05
P=0.179
P=0.006
P=0.516

36

Figure 13. Correlation analysis of markers of different cell types based on
GSE 90497. (A) Clec4f and Vsig4 are two well-characterized KCs markers.
Their correlation is used as a positive control for analysis. (B) Adgre1 is an
identified markers of total macrophages. The correlation between it and
Clec4f is used as a negative control for analysis. (C)The correlation between
Clec4f and Ly6c, a MoMFs marker, is used as a negative control for analysis.
(D) Slfn4 is the marker of neutrophils. The correlation between it and Clec4f
is used as a negative control for analysis. (E) The correlation between Cd3g,
the marker of T cells, and Clec4f is used as a negative control for analysis.
(F) The correlation between Cd79a, the marker of B cells, and Clec4f is used
as a negative control for analysis.

R² = 0.0032
0
200
400
600
800
1000
1200
1400
1600
1800
6 7 8 9 10 11 12
expression amount of Clec4f
expression amount of Cd79a
Correlation between Cd79a and Clec4f
R² = 0.2162
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40
expression aount of Clec4f
expression amount of Cd3g
Correlaton between Cd3g and Clec4f
R² = 0.0385
0
200
400
600
800
1000
1200
1400
1600
1800
2 2.5 3 3.5 4 4.5 5
expression amount of Clec4f
expression amount of Slfn4
Correlation between Slfn4 and Clec4f
R² = 0.5411
0
200
400
600
800
1000
1200
1400
1600
1800
0 20 40 60 80 100
expression amount of Clec4f
expression amount of Ly6c1
Correlation between Ly6c1 and Clec4f
R² = 0.0157
0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300 400 500
expression amount of Clec4f
expression amount of Adgre1
Correlation between Adgre1 and Clec4f
R² = 0.5664
0
100
200
300
400
500
600
0 500 1000 1500 2000
expression amount of Vsig4
expression amount of Clec4f
Correlation between Clec4f and Vsig4 A B
C D
E F
P=3.281e
-08
P=0.447
P=9.577e
-08
P=0.231
P=0.003
P=0.732

37
For each GSE dataset, correlation heatmap with hierarchical clustering was also
drawn based on the Pearson correlation coefficients between those selected genes.
Unsupervised clustering demonstrated that Cd5l and Folr2 are clustered with the
markers of murine KCs, but not markers of other cell types (Fig.14A-B). To increase
the stringency and statistical power, we also used a merged dataset are based on
ranking. The merged dataset is composed of 15 GEO datasets from HFD feeding
experiments with total of 587 samples. The merged data were ranked to take care of
batch effects. We performed Spearman ranking correlation analysis of the selected
genes in the ranked dataset. Consistent with the results from the two single dataset
used above, the result shows that markers of the same cell type were clustered
together (Fig.14C). In particular, Cd5l, Folr2 clustered with Vsig4 and Adgre1 (Clec4f
was not reported in this merged dataset). Collectively, our in-silico validation confirmed
that both Cd5l and Folr2 are likely novel markers of murine KCs.

38

Merged dataset C
Figure 12. Heatmap showing the correlation coefficients between selected genes in three different datasets. (A&B) The pearson correlation
coefficient between two genes is calculated based on their expression in each sample. Each green box represents a cluster of genes that could be
the markers of a same cell type. (C) The spearman ranking correlation coefficient between two genes is calculated based on their expression
ranking in each sample. Each green box represents a cluster of genes that could be the markers of a same cell type.
GSE90497 A
B GSE149863
Figure 14. Heatmap showing the correlation coefficients between selected
genes in three different datasets. (A&B) The pearson correlation coefficient
between two genes is calculated based on their expression in each sample.
Each green box represents a cluster of genes that could be the markers of
a same cell type. (C) The spearman ranking correlation coefficient between
two genes is calculated based on their expression ranking in each sample.
Each green box represents a cluster of genes that could be the markers of
a same cell type.

39
3. Validating CD5L expression using liver injury models
In a recent single cell RNA-seq study, Folr2 was identified to cluster with Vsig4, Cd163
and Cd169. This study also validated the expression of Folr2 in Vsig4 expressing cells
using flow cytometry

(75). Thus, our further validation experiments focused on Cd5l in
two different liver injury mouse models. In the alcoholic liver injury model representing
ALD, steatosis and immune cells infiltration were induced by a long term hybrid feeding
which involved ad libitum consumption of high-calorie food (40% of total calorie) and
intragastric feeding of HFD plus ethanol (60% of total calorie) simultaneously (78). In
a non-alcoholic liver injury model that represents NASH, the steatohepatitis was
induced by the deletion of PTEN (Phosphatase and Tensin Homolog deleted on
Chromosome 10), a phosphatase that inhibits insulin signaling (Pten
LoxP/LoxP
; Alb-Cre
+
,
PM mice) (79). This model mimic the hepatic hyperinsulinemia that occurs in insulin
resistant patients. In both liver injury models, the expression of Cd5l are significantly
increased, concurrent with that of Clec4f, a known marker of KCs (Fig.15A-D).
Furthermore, regardless of genotype or treatment, high correlation (R
2
=0.8689) is
observed for the expression of Cd5l and Clec4f in all mice (Fig.15E). Thus consistent
with our in silico verification results, these data further support that Cd5l is a likely
marker for KC cells and its expression is induced in liver injury.

40

0.00000 0.00001 0.00002 0.00003 0.00004
0.00000
0.00002
0.00004
0.00006
0.00008
Combining correlation of Clec4f and Cd5l
Clec4f expression
Cd5l expression
R
2
=0.8689
9 mon PM (18s)
9 mon WT (18s)
0.00000
0.00001
0.00002
0.00003
0.00004
9 mon PM vs. WT Clec4f
expression amount of Clec4f
✱
9 mon PM (18s)
9 mon WT (18s)
0.00000
0.00002
0.00004
0.00006
0.00008
9 mon PM vs. WT CD5l
expression amount of Cd5l
✱
6 month EtOH
6 month WT
0.00000
0.00001
0.00002
0.00003
6 mon EtOH vs. WT Clec4f
expression amount of Clec4f
✱✱✱
6 mon EtOH
6 mon WT
0.00000
0.00002
0.00004
0.00006
0.00008
6 mon EtOH vs. WT Cd5l
expression amount of Cd5l
✱✱
Clec4f expression
Cd5l expression
Clec4f expression
Cd5l expression
Cd5l expression
Clec4f expression
Combined correlation analysis of Clec4f and Cd5l
A B
C D
E
Figure 13. Validation of Cd5l expression in liver injury mouse models. (A&B) qPCR data of Clec4f and Cd5l mRNAexpression in liver of PM
(n=6) and wild type mice (n=6). Data are presented as mean ± SD. *p < 0.05. (C&D) qPCR data of Clec4f and Cd5l mRNA expression in liver of
mice subjected to chronic ethanol feeding (n=4) and mice fed with normal diet (n=4). Data are presented as mean ± SD. **p < 0.01 and ***p <
0.001. (E) Combined linear correlation analysis of Clec4f and Cd5l in mice, regardless of the genotype or the treatment (n=20).
6 mon EtOH vs. control Clec4f 6 mon EtOH vs. control Cd5l
9 mon PM vs. WT Clec4f 9 mon PM vs. WT Cd5l
9 mon PM
9 mon WT
9 mon PM
9 mon WT
6 mon EtOH
6 mon Control
6 mon EtOH
6 mon Control
Figure 15. Validation of Cd5l expression in liver injury mouse models.
(A&B) qPCR data of Clec4f and Cd5l mRNA expression in liver of PM (n=6)
and wild type mice (n=6). Data are presented as mean ± SD. *p < 0.05.
(C&D) qPCR data of Clec4f and Cd5l mRNA expression in liver of mice
subjected to chronic ethanol feeding (n=4) and mice fed with normal diet
(n=4). Data are presented as mean ± SD. **p < 0.01 and ***p < 0.001. (E)
Combined linear correlation analysis of Clec4f and Cd5l in mice, regardless
of the genotype or the treatment (n=20).

41
4. Specific expression of CD5L in Kupffer cells
Considering the fact that Cd5l was a secreted protein, we conducted intracellular flow
cytometry analysis of Cd5l protein expression in all hepatic immune cells for further
confirmation. Since mice with liver injury harbored a larger population of hepatic
macrophages, we euthanized 9-month-old PM mice where severe injury and
inflammation is expected (79). Using conventional markers, KCs were identified as
CD45
+
CD11b
intermediate
F4/80
high
cells, while infiltrating MoMFs were characterized as
CD45
+
CD11b
high
F4/80
intermediate
cells (Fig.16A). The identity of KCs were further
confirmed with their high expression of Clec4f (Fig.16B). Histogram of Cd5l expression
in all hepatic immune cells identified the existence of two populations: Cd5l
high

subpopulation and Cd5l
low
subpopulation. A complete overlap of the Cd5l expression
with Clec4f
+
KCs is observed in all 3 PM mice suggested that KCs not only highly
express Cd5l but also are the predominant population that express Cd5l among
hepatic immune cells (Fig.16C,E,G). We further analyzed the CD5L expressing
populations in immune cells isolated from two healthy livers. The recovery of total
immune cell counts are lower in these healthy livers as expected. Similar overlap of
CD5L with KC is observed in these none-injured livers (Fig.16D,F). We performed a
quantitative analysis of the median fluorescent intensity of Cd5l in KCs vs. MoMFs and
showed that the expression of Cd5L is significantly higher in KCs vs. MoMFs
regardless of mouse genotype or injury status (Fig.16H). Together with our in silico
verification (Fig.8-14), the analysis revealed that Cd5l is expressed by KCs, not
MoMFs.

42

KC MoMF
0
200
400
600
800
1000
Expression of Cd5l in KCs vs. MoMFs
MFI of Cd5l
✱
Cd5l-BV421
F4/80-PerCP
CD11b-FITC
Clec4f-AF488
CD5L-BV421
Cd5l-BV421
Cd5l-BV421 Cd5l-BV421
Cd5l-BV421
All CD45+ cells in Pten-null mouse Kupffer cells & monocytes-derived macrophages
CD5L expression in cells isolated from Pten-null mouse CD5L expression in cells isolated from wild type mouse
CD5L expression in cells isolated from Pten-null mouse CD5L expression in cells isolated from wild type mouse
CD5L expression in cells isolated from Pten-null mouse
A B
C D
E F
G H
KCs
MoMFs
Immune cells
KCs
MoMFs
Immune cells
KCs
MoMFs
Immune cells
KCs
MoMFs
Immune cells
KCs
MoMFs
Immune cells
KCs
MoMFs
Figure 14. Discriminating Kupffer cells and monocytes-derived macrophages from other immune cells in PM and wild type mice. (A&B)
Flow cytometry analysis of CD45
+
CD11b
intermediate
F4/80
high
Kupffer cells also highly express Clec4f. (C&E&G) Histogram of CD5L expression in
different types of cells which are isolated from the PM mice. (D&F) Histogram of CD5L expression in different types of cells which are isolated from
the wild type mice. (H) Compare median fluorescent intensity of CD5L in Kupffer cells vs. monocytes-derived macrophages. *p < 0.05. The noise
signal has been subtracted.
Figure 16. Discriminating Kupffer cells and monocytes-derived
macrophages from other immune cells in PM and wild type mice. (A&B)
Flow cytometry analysis of CD45
+
CD11b
intermediate
F4/80
high
Kupffer cells also
highly express Clec4f. (C&E&G) Histogram of CD5L expression in different
types of cells which are isolated from the PM mice. (D&F) Histogram of
CD5L expression in different types of cells which are isolated from the wild
type mice. (H) Compare median fluorescent intensity of CD5L in Kupffer
cells vs. monocytes-derived macrophages. *p < 0.05. The noise signal has
been subtracted.

43
To further confirmed the results of flow cytometry, we performed the intracellular
immunofluorescence staining of Clec4f and Cd5l using the paraffin-embedded liver
section of PM and wild type mice (Fig.17). The results indicated that 14.35% Clec4f
+

cells are also positive for Cd5l, while essentially all Cd5l
+
cells are Clec4f
+
.

Figure 15. Immunofluorescent staining to validate the specific expression of Cd5l by Kupffer cells. Detection of dying Clec4f
+
Cd5l
+
Kupffer cells on paraffin-embedded liver section of Pten-null mice. Kupffer cells that also express Cd5l are pointed out by narrows.
Cd5l; Clec4f; DAPI
Figure 17. Immunofluorescent staining to validate the specific expression
of Cd5l by Kupffer cells. Detection of dying Clec4f
+
Cd5l
+
Kupffer cells on
paraffin-embedded liver section of Pten-null mice. Kupffer cells that also
express Cd5l are pointed out by narrows.

44
5. Plasma concentration CD5L as potential biomarker for the activation of KCs
Our qPCR results showed that CD5L mRNA expression significantly increased in the
livers of diet-induced ALD and transgenic NASH mouse model. Since CD5L is a
secreted protein and detectable in plasma at high concentration, we conducted an
enzyme linked immunosorbent assay (ELISA) to investigate CD5L plasma
concentration changes during the progression of NAFLD/NASH. A gradual increase of
CD5L plasma concentration is observed in HFD fed mice with increase diet exposure.
A moderate (approximately 20%) but insignificant increase of CD5L plasma
concentration is observed with 5 mon HFD feeding compared with mice fed with low-
fat diet (LFD). This difference however became striking (2.5-3 fold) and highly
significant at 9 month (Fig.18). The data revealed that CD5L plasma concentration
increases with the severity of liver injury and maybe used as a serum marker for
hepatic inflammation and activation of KC.

Figure 18. Changes of Cd5l plasma concentration in diet-induced
liver injury mouse models. Analysis of Cd5l plasma concentration in
mice subjected to different diet. 5 months high-fat diet (n=3), 5
months low-fat diet (n=4), 9 months high-fat diet (n=4), 9 months low-
fat diet (n=5). Data are presented as mean ± SD. ****p < 0.0001.

9mon HFD
9mon LFD
0
200
400
600
800
9 mon HFD vs. LFD
plasma concentration of CD5L (pg/ml)
✱✱✱✱
5mon HFD
5mon LFD
0
200
400
600
5 mon HFD vs. LFD
plasma concentration of CD5L (pg/ml)
ns
Figure 16. Changes of Cd5l plasma concentration in diet-induced liver injury mouse models. Analysis of Cd5l plasma concentration in mice
subjected to different diet. 5 months high-fat diet (n=3), 5 months low-fat diet (n=4), 9 months high-fat diet (n=4), 9 months low-fat diet (n=5). Data
are presented as mean± SD. ****p < 0.0001.

45
Chapter IV Discussion
Macrophages are part of the innate immunity and well-known for their function as
phagocytes in clearance of endogenous and exogenous pathogens (80). They are
distributed in various organs throughout the body and originated during the embryonic
development (81). Extensive plasticity is a hallmark of macrophages, which enables
them to adapt their function and phenotype dependent on local tissue
microenvironment (82). Therefore, macrophages are highly versatile and adopt
different phenotypes, such as M1 and M2, to make different contributions to
homeostasis and the progression of disease. For instance, as liver is constantly
exposed to bacterial pathogens and dietary antigens, the unique immune-suppressive
characteristic of tissue resident macrophages in the liver allows them to establish the
immune-tolerant environment that is crucial to avoid excessive immune activation and
tissue injury (83). However, during severe injury and chronic exposure to pathogens,
macrophages are reprogrammed to produce a board spectrum of pro-inflammatory
factors involved in the recruitment of other immune cells and are also crucial in
crosstalk with other surrounding cells (84). These responsibilities of macrophages are
increasingly recognized in metabolic diseases including obesity-induced insulin
resistance and type 2 diabetes mellitus in addition to their already recognized
importance in autoimmune disease and cancer (85).

Liver is an important immunological organ in the body that is equipped with a large
population of macrophages. It serves as the frontline to detect, ingest and degrade the

46
immunogenic molecules entering the body from the gastrointestinal tract through
portal vein. Similar to other organs, (81) hepatic macrophages are mainly composed
of two subpopulations: tissue-resident KCs that are embryonic origin and monocyte-
derived macrophages that are infiltrating from the circulating system. Even though liver
macrophages are recognized as crucial component for maintaining liver homeostasis
and driving the development and resolution of inflammation in liver diseases (51), how
they assume each specific functions are unknown due to a lack of effective strategy
to distinguish them from each other. The widely used approach to define different
macrophage population is based on their cell surface expression of different markers.
Regrettably, few markers are available for their discrimination, which limits the ability
to comprehensively understand the diversity of macrophages under different
inflammatory conditions. Clec4f, Vsig4 and Timd4 are the three most commonly used
markers for identifying murine KCs, but there is no specific markers for human KCs.
Therefore, in order to further clarify the heterogeneity of KCs and understand the
functions of each subsets during homeostasis and pathology, it is crucial to identify
new markers for both KCs and its subsets.

With the rapid evolution of single-cell transcriptomic technologies, identifying the
diversity of KCs and the unique functions of each subtype has become possible with
extensive bioinformatics analysis (53, 55, 69). Though secondary analysis of a
published single cell RNA-seq data, we ranked out top genes that are preferentially
expressed by KCs and selected Folr2 and Cd5l as the promising new markers. We

47
further explore their correlation with two identified KC markers by using two published
GEO dataset (GSE149863 and GSE90497) respectively. The results indicated that
Cd5l and Folr2 were strongly correlated with the markers of KCs, but displayed weak
correlation with the markers of other cell types. Folr2 has been confirmed as the
markers for KCs in flow cytometry analysis, but the validation of Cd5l is still lacking.
Therefore, our in vivo verifications were mainly focused on Cd5l. We utilized two
different liver injury mouse models to verify the strong correlation between Cd5l and
Clec4f on the mRNA level. Moreover, intracellular flow cytometry analysis and
immunofluorescent staining were also performed to further validate the unique
expression of Cd5l in KCs on the protein level. Both in silico and in vivo validations
suggested that Cd5l could be a newly characterized markers for KCs.

CD5L (CD5 molecule-like), a glycoprotein mainly synthesized and secreted by
macrophages, is a key regulator of inflammatory responses in infection,
atherosclerosis and cancer (86-88). The expression of it is regulated by a transcription
factor LXRs, a member of the nuclear receptor family that are crucial in lipid
homeostasis (89). The most characterized function of CD5L is supporting the survival
of myeloid and lymphoid cells (90). In an endotoxin-induced fulminant hepatitis mouse
model, more infiltrating macrophages and increased inflammation were observed in
the transgenic mice overexpressing CD5L, which was attributed to the decreasing
apoptosis of macrophages (91). However, the studies of CD5L indicated that it could
also exhibit opposite functions in some certain process where they either enhance or

48
attenuate the inflammation (87). A recent research revealed that CD5L polarized
macrophages towards a specific phenotype similar to that induced by IL-10 not only in
cell marker expression but also in anti-inflammatory function (86). Another in vitro
research also unveiled the anti-inflammatory role of CD5L since it modulated the
cytokine secretion of monocytes whose secretion of TNF and IL-1β were decreased
while the production of IL-10 was promoted (92).

In addition to modulating the functions of leukocytes, CD5L is also reported to engaged
in the regulation of lipid metabolism via a CD36 dependent endocytic manner (93).
Within adipocytes, the endocytosed CD5L binds to cytosolic fatty acid synthase (FAS)
and inhibits its activity. This action of CD5L leads to decreased lipid droplet storage
within adipocytes and results in the rapid release of free fatty acids. Additionally,
analysis of TCGA data shows that HCC patients with higher Cd5l expression survived
better compared to those with lower CD5L. These analyses also showed strong
correlation of CD5L with genes associate with T cells. Therefore, CD5L may play a
role in macrophages/KC cell crosstalk with T cells to influence metabolic states and
liver disease pathogenesis.

Plasma ALT (alanine transaminase) and AST (aspartate transaminase) have been
used in the clinic for diagnosis of liver injury. Like ALS and AST, CD5L is detectable in
circulating system at relatively high concentration, where it is stabilized by combining
with IgM (94, 95). Glycoproteomic analysis characterized CD5L as a novel serum

49
biomarker candidates for the indication of atopic dermatitis (86). Our analysis here
shows that CD5L is uniquely expressed by liver KCs. Thus, we proposed that CD5L
could also be a biomarker for KC activation, and can be used as a potential marker for
liver injury together with ALT and AST to aid further diagnosis and guide treatment of
NAFLD/NASH.

In conclusion, we discovered Cd5l as a new Kupffer cell marker and confirmed its
expression in liver and upregulation in liver injury. Furthermore, as a secreted protein,
CD5L can be detected in plasma and is increased with the severity of NAFLD/NASH,
thus may serve as a segregate serum marker for KC activation.

50
Bibliography:
1. H. Sung et al., Global Cancer Statistics 2020: GLOBOCAN Estimates of
Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a
cancer journal for clinicians 71, 209-249 (2021).
2. S. F. Altekruse, S. S. Devesa, L. A. Dickie, K. A. McGlynn, D. E. Kleiner,
Histological classification of liver and intrahepatic bile duct cancers in SEER
registries. Journal of registry management. 38, 201-205 (2011).
3. J. L. Petrick et al., International trends in hepatocellular carcinoma incidence,
1978–2012. International journal of cancer 147, 317-330 (2020).
4. Z. M. Younossi et al., Association of nonalcoholic fatty liver disease (NAFLD)
with hepatocellular carcinoma (HCC) in the United States from 2004 to 2009.
Hepatology (Baltimore, Md.) 62, 1723-1730 (2015).
5. H. B. El-Serag, Epidemiology of Viral Hepatitis and Hepatocellular Carcinoma.
Gastroenterology (New York, N.Y. 1943) 142, 1264-1273.e1261 (2012).
6. N. Ganne-Carrié, P. Nahon, Hepatocellular carcinoma in the setting of alcohol-
related liver disease. Journal of hepatology 70, 284-293 (2019).
7. L. A. Beste, P. Green, K. Berry, P. Belperio, G. N. Ioannou, Hepatitis C–Related
Hepatocellular Carcinoma Incidence in the Veterans Health Administration After
Introduction of Direct-Acting Antivirals. JAMA : the journal of the American
Medical Association 324, 1003-1005 (2020).
8. X. Ge, L. Zheng, M. Wang, Y. Du, J. Jiang, Prevalence trends in non-alcoholic
fatty liver disease at the global, regional and national levels, 1990-2017: a
population-based observational study. BMJ open 10, e036663-e036663 (2020).
9. Z. Younossi et al., Global burden of NAFLD and NASH: Trends, predictions, risk
factors and prevention. Nature reviews. Gastroenterology & hepatology 15, 11-
20 (2018).
10. Z. M. Younossi et al., Global epidemiology of nonalcoholic fatty liver disease—
Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology
(Baltimore, Md.) 64, 73-84 (2016).
11. M. H. Le et al., 2019 Global NAFLD Prevalence: A Systematic Review and
Meta-analysis. Clinical gastroenterology and hepatology, (2022).
12. G. Vernon, A. Baranova, Z. M. Younossi, Systematic review: the epidemiology
and natural history of non-alcoholic fatty liver disease and non-alcoholic

51
steatohepatitis in adults. Alimentary pharmacology & therapeutics 34, 274-285
(2011).
13. K. D. Kochanek, S. L. Murphy, J. Xu, E. Arias, Deaths: Final data for 2017.
National vital statistics reports 68, 1-18 (2019).
14. J. P. Rehm et al., Global burden of disease and injury and economic cost
attributable to alcohol use and alcohol-use disorders. The Lancet (British edition)
373, 2223-2233 (2009).
15. V. T. Savolainen, K. Liesto, A. Männikkö, A. Penttilä, P. J. Karhunen, Alcohol
Consumption and Alcoholic Liver Disease: Evidence of a Threshold Level of
Effects of Ethanol. Alcoholism, clinical and experimental research 17, 1112-
1117 (1993).
16. F. Idalsoaga, A. V. Kulkarni, O. Y. Mousa, M. Arrese, J. P. Arab, Non-alcoholic
Fatty Liver Disease and Alcohol-Related Liver Disease: Two Intertwined
Entities. Frontiers in medicine 7, 448-448 (2020).
17. J. P. Arab et al., Alcohol-related liver disease: Clinical practice guidelines by the
Latin American Association for the Study of the Liver (ALEH). Annals of
hepatology 18, 518-535 (2019).
18. G. Cholankeril, A. Ahmed, Alcoholic Liver Disease Replaces Hepatitis C Virus
Infection as the Leading Indication for Liver Transplantation in the United States.
Clinical gastroenterology and hepatology 16, 1356-1358 (2018).
19. G. T. Brown, D. E. Kleiner, Histopathology of nonalcoholic fatty liver disease
and nonalcoholic steatohepatitis. Metabolism, clinical and experimental 65,
1080-1086 (2015).
20. C. P. Day, S. Saksena, Non-alcoholic steatohepatitis: Definitions and
pathogenesis. Journal of gastroenterology and hepatology 17, S377-S384
(2002).
21. N. Méndez-Sánchez, M. Arrese, D. Zamora-Valdés, M. Uribe, Current concepts
in the pathogenesis of nonalcoholic fatty liver disease. Liver international 27,
423-433 (2007).
22. R. Lomonaco et al., Effect of adipose tissue insulin resistance on metabolic
parameters and liver histology in obese patients with nonalcoholic fatty liver
disease. Hepatology (Baltimore, Md.) 55, 1389-1397 (2012).
23. H. Malhi, F. J. Barreyro, H. Isomoto, S. F. Bronk, G. J. Gores, Free fatty acids
sensitise hepatocytes to TRAIL mediated cytotoxicity. Gut 56, 1124-1131 (2007).

52
24. Y. Wei, D. Wang, F. Topczewski, M. J. Pagliassotti, Saturated fatty acids induce
endoplasmic reticulum stress and apoptosis independently of ceramide in liver
cells. American Journal of Physiology - Endocrinology And Metabolism 291,
275-281 (2006).
25. J. Cao et al., Saturated fatty acid induction of endoplasmic reticulum stress and
apoptosis in human liver cells via the PERK/ATF4/CHOP signaling pathway.
Molecular and cellular biochemistry 364, 115-129 (2012).
26. Y. Ogawa et al., Palmitate-induced lipotoxicity is crucial for the pathogenesis of
nonalcoholic fatty liver disease in cooperation with gut-derived endotoxin.
Scientific reports 8, 11365-11314 (2018).
27. L. T. Gan et al., Hepatocyte free cholesterol lipotoxicity results from JNK1-
mediated mitochondrial injury and is HMGB1 and TLR4-dependent. Journal of
hepatology 61, 1376-1384 (2014).
28. M. Ganz et al., Progression of non-alcoholic steatosis to steatohepatitis and
fibrosis parallels cumulative accumulation of danger signals that promote
inflammation and liver tumors in a high fat-cholesterol-sugar diet model in mice.
Journal of translational medicine 13, 193-193 (2015).
29. M. Martinez-Medina et al., Western diet induces dysbiosis with increased E coli
in CEABAC10 mice, alters host barrier function favouring AIEC colonisation.
Gut 63, 116-124 (2014).
30. M. Serino et al., Metabolic adaptation to a high-fat diet is associated with a
change in the gut microbiota. Gut 61, 543-553 (2012).
31. K. Rahman et al., Loss of Junctional Adhesion Molecule A Promotes Severe
Steatohepatitis in Mice on a Diet High in Saturated Fat, Fructose, and
Cholesterol. Gastroenterology (New York, N.Y. 1943) 151, 733-746.e712
(2016).
32. J. Boursier et al., The severity of nonalcoholic fatty liver disease is associated
with gut dysbiosis and shift in the metabolic function of the gut microbiota.
Hepatology (Baltimore, Md.) 63, 764-775 (2016).
33. T. Yatsunenko et al., Human gut microbiome viewed across age and geography.
Nature (London) 486, 222-227 (2012).
34. S. Pendyala, J. M. Walker, P. R. Holt, A High-Fat Diet Is Associated With
Endotoxemia That Originates From the Gut. Gastroenterology (New York, N.Y.
1943) 142, 1100-1101.e1102 (2012).
35. C. A. Rivera et al., Toll-like receptor-4 signaling and Kupffer cells play pivotal

53
roles in the pathogenesis of non-alcoholic steatohepatitis. Journal of
hepatology 47, 571-579 (2007).
36. T. Csak et al., Deficiency in myeloid differentiation factor-2 and toll-like receptor
4 expression attenuates nonalcoholic steatohepatitis and fibrosis in mice.
American Journal of Physiology - Gastrointestinal and Liver Physiology 300,
433-441 (2011).
37. I. Garcia-Martinez et al., Hepatocyte mitochondrial DNA drives nonalcoholic
steatohepatitis by activation of TLR9. The Journal of clinical investigation 126,
859-864 (2016).
38. E. Buzzetti, M. Pinzani, E. A. Tsochatzis, The multiple-hit pathogenesis of non-
alcoholic fatty liver disease (NAFLD). Metabolism, clinical and experimental 65,
1038-1048 (2016).
39. A. Hadinia, A. H. Doustimotlagh, H. R. Goodarzi, A. Arya, M. Jafarinia,
Circulating levels of pro-inflammatory cytokines in patients with nonalcoholic
fatty liver disease and non-alcoholic steatohepatitis. Iranian journal of
immunology 16, 327-333 (2019).
40. I. C. Bocsan et al., Cytokines Genotype-Phenotype Correlation in Nonalcoholic
Steatohepatitis. Oxidative medicine and cellular longevity 2017, 4297206-
4297207 (2017).
41. M. Marí et al., Mitochondrial free cholesterol loading sensitizes to TNF- and
Fas-mediated steatohepatitis. Cell metabolism 4, 185-198 (2006).
42. W. Zhang et al., Tumor necrosis factor-α accelerates apoptosis of steatotic
hepatocytes from a murine model of non-alcoholic fatty liver disease.
Biochemical and biophysical research communications 391, 1731-1736 (2010).
43. B. M. De Taeye et al., Macrophage TNF-α contributes to insulin resistance and
hepatic steatosis in diet-induced obesity. American Journal of Physiology -
Endocrinology And Metabolism 293, 713-725 (2007).
44. K. A. Negrin et al., IL-1 signaling in obesity-induced hepatic lipogenesis and
steatosis. PloS one 9, e107265-e107265 (2014).
45. J. J. Senn, P. J. Klover, I. A. Nowak, R. A. Mooney, Interleukin-6 Induces Cellular
Insulin Resistance in Hepatocytes. Diabetes (New York, N.Y.) 51, 3391-3399
(2002).
46. B. Cai et al., Macrophage MerTK promotes liver fibrosis in nonalcoholic
steatohepatitis. Cell metabolism 31, 406-421 (2020).

54
47. J. P. Pradere et al., Hepatic macrophages but not dendritic cells contribute to
liver fibrosis by promoting the survival of activated hepatic stellate cells in mice.
Hepatology (Baltimore, Md.) 58, 1461-1473 (2013).
48. J. P. Edwards, D. M. Mosser, Exploring the full spectrum of macrophage
activation. Nature reviews. Immunology 8, 958-969 (2008).
49. P. J. Murray, T. A. Wynn, Protective and pathogenic functions of macrophage
subsets. Nature reviews. Immunology 11, 723-737 (2011).
50. X. Xiong, H. Kuang, T. Liu, J. D. Lin, A Single-Cell Perspective of the
Mammalian Liver in Health and Disease. Hepatology (Baltimore, Md.) 71, 1467-
1473 (2020).
51. Y. Wen, J. Lambrecht, C. Ju, F. Tacke, Hepatic macrophages in liver
homeostasis and diseases-diversity, plasticity and therapeutic opportunities.
Cellular & molecular immunology 18, 45-56 (2021).
52. E. Gomez Perdiguero et al., Tissue-resident macrophages originate from yolk-
sac-derived erythro-myeloid progenitors. Nature (London) 518, 547-551 (2015).
53. C. Blériot et al., A subset of Kupffer cells regulates metabolism through the
expression of CD36. Immunity (Cambridge, Mass.) 54, 2101-2116.e2106
(2021).
54. S. A. MacParland et al., Single cell RNA sequencing of human liver reveals
distinct intrahepatic macrophage populations. Nature communications 9, 4383-
4321 (2018).
55. J. Zhao et al., Single-cell RNA sequencing reveals the heterogeneity of liver-
resident immune cells in human. Cell discovery 6, 22-22 (2020).
56. B. Gao, W. I. Jeong, Z. Tian, Liver: An organ with predominant innate immunity.
Hepatology (Baltimore, Md.) 47, 729-736 (2008).
57. P. A. Knolle et al., Regulation of endotoxin-induced IL-6 production in liver
sinusoidal endothelial cells and Kupffer cells by IL-10. Clinical and experimental
immunology 107, 555-561 (1997).
58. A. Inatsu et al., Novel mechanism of C-reactive protein for enhancing mouse
liver innate immunity. Hepatology (Baltimore, Md.) 49, 2044-2054 (2009).
59. Q. You, L. Cheng, R. M. Kedl, C. Ju, Mechanism of T cell tolerance induction
by murine hepatic Kupffer cells. Hepatology (Baltimore, Md.) 48, 978-990
(2008).

55
60. F. Heymann et al., Liver inflammation abrogates immunological tolerance
induced by Kupffer cells. Hepatology (Baltimore, Md.) 62, 279-291 (2015).
61. P. Knoll et al., Human Kupffer cells secrete IL-10 in response to
lipopolysaccharide (LPS) challenge. Journal of hepatology 22, 226-229 (1995).
62. E. Breous, S. Somanathan, L. H. Vandenberghe, J. M. Wilson, Hepatic
regulatory T cells and Kupffer cells are crucial mediators of systemic T cell
tolerance to antigens targeting murine liver. Hepatology (Baltimore, Md.) 50,
612-621 (2009).
63. C. Wiegard et al., Murine liver antigen presenting cells control suppressor
activity of CD4+CD25+ regulatory T cells. Hepatology (Baltimore, Md.) 42, 193-
199 (2005).
64. K. Miura, L. Yang, N. van Rooijen, H. Ohnishi, E. Seki, Hepatic recruitment of
macrophages promotes nonalcoholic steatohepatitis through CCR2. American
journal of physiology: Gastrointestinal and liver physiology 302, G1310-G1321
(2012).
65. O. Krenkel et al., Therapeutic inhibition of inflammatory monocyte recruitment
reduces steatohepatitis and liver fibrosis. Hepatology (Baltimore, Md.) 67,
1270-1283 (2018).
66. X. Zhang et al., CXC chemokine receptor 3 promotes steatohepatitis in mice
through mediating inflammatory cytokines, macrophages and autophagy.
Journal of hepatology 64, 160-170 (2015).
67. F. Heymann et al., Hepatic macrophage migration and differentiation critical for
liver fibrosis is mediated by the chemokine receptor C-C motif chemokine
receptor 8 in mice. Hepatology (Baltimore, Md.) 55, 898-909 (2012).
68. S. Yona et al., Fate mapping reveals origins and dynamics of monocytes and
tissue macrophages under homeostasis. Immunity 38, 79-91 (2013).
69. S. Tran et al., Impaired Kupffer cell self-renewal alters the liver response to lipid
overload during non-alcoholic steatohepatitis. Immunity 53, 627-640 (2020).
70. O. Krenkel et al., Myeloid cells in liver and bone marrow acquire a functionally
distinct inflammatory phenotype during obesity-related steatohepatitis. Gut 69,
551-563 (2020).
71. M. Sakai et al., Liver-derived signals sequentially reprogram myeloid enhancers
to initiate and maintain Kupffer cell identity. Immunity 51, 655-670 (2019).
72. L. Beattie et al., Bone marrow-derived and resident liver macrophages display

56
unique transcriptomic signatures but similar biological functions. Journal of
hepatology 65, 758-768 (2016).
73. P. Ramachandran et al., Resolving the fibrotic niche of human liver cirrhosis at
single-cell level. Nature (London) 575, 512-518 (2019).
74. X. Wu et al., Human Liver Macrophage Subsets Defined by CD32. Frontiers in
immunology 11, 2108-2108 (2020).
75. M. Guilliams et al., Spatial proteogenomics reveals distinct and evolutionarily
conserved hepatic macrophage niches. Cell 185, 379-+ (2022).
76. Y. Ait Ahmed et al., Kupffer cell restoration after partial hepatectomy is mainly
driven by local cell proliferation in IL-6-dependent autocrine and paracrine
manners. Cellular & molecular immunology 18, 2165-2176 (2021).
77. C. L. Scott et al., Bone marrow-derived monocytes give rise to self-renewing
and fully differentiated Kupffer cells. Nature communications 7, 10321-10321
(2016).
78. H. Tsukamoto, H. Mkrtchyan, A. Dynnyk, Intragastric Ethanol Infusion Model in
Rodents. Alcohol 447, 33-48 (2008).
79. B. Stiles et al., Liver-specific deletion of negative regulator Pten results in fatty
liver and insulin hypersensitivity [corrected]. Proceedings of the National
Academy of Sciences - PNAS 101, 2082-2087 (2004).
80. E. Barreby, P. Chen, M. Aouadi, Macrophage functional diversity in NAFLD —
more than inflammation. Nature reviews. Endocrinology 18, 461-472 (2022).
81. S. Epelman, Kory J. Lavine, Gwendalyn J. Randolph, Origin and Functions of
Tissue Macrophages. Immunity (Cambridge, Mass.) 41, 21-35 (2014).
82. K. Liddiard, P. R. Taylor, Understanding local macrophage phenotypes in
disease: Shape-shifting macrophages. Nature medicine 21, 119-120 (2015).
83. K. Oura, A. Morishita, J. Tani, T. Masaki, Tumor immune microenvironment and
immunosuppressive therapy in hepatocellular carcinoma: A review.
International journal of molecular sciences 22, 5801 (2021).
84. M. D. Park, A. Silvin, F. Ginhoux, M. Merad, Macrophages in health and disease.
Cell 185, 4259-4279 (2022).
85. A. Chawla, K. D. Nguyen, Y. P. S. Goh, Macrophage-mediated inflammation in
metabolic disease. Nature reviews. Immunology 11, 738-749 (2011).

57
86. L. Sanjurjo et al., CD5L promotes M2 macrophage polarization through
autophagy-mediated upregulation of ID3. Frontiers in immunology 9, 480-480
(2018).
87. L. Sanchez-Moral et al., Multifaceted roles of cd5l in infectious and sterile
inflammation. International journal of molecular sciences 22, 4076 (2021).
88. V. G. Martinez et al., The macrophage soluble receptor AIM/Api6/CD5L displays
a broad pathogen recognition spectrum and is involved in early response to
microbial aggression. Cellular & molecular immunology 11, 343-354 (2014).
89. L. Sanjurjo, G. Aran, N. Roher, A. F. Valledor, M. R. Sarrias, AIM/CD5L: a key
protein in the control of immune homeostasis and inflammatory disease.
Journal of leukocyte biology 98, 173-184 (2015).
90. K. Kuwata et al., AIM Inhibits Apoptosis of T Cells and NKT Cells in
Corynebacterium-Induced Granuloma Formation in Mice. The American journal
of pathology 162, 837-847 (2003).
91. I. Haruta et al., Association of AIM, a Novel Apoptosis Inhibitory Factor, with
Hepatitis via Supporting Macrophage Survival and Enhancing Phagocytotic
Function of Macrophages. The Journal of biological chemistry 276, 22910-
22914 (2001).
92. L. Sanjurjo et al., The human CD5L/AIM-CD36 axis: A novel autophagy inducer
in macrophages that modulates inflammatory responses. Autophagy 11, 487-
502 (2015).
93. J. Kurokawa et al., Macrophage-Derived AIM Is Endocytosed into Adipocytes
and Decreases Lipid Droplets via Inhibition of Fatty Acid Synthase Activity. Cell
metabolism 11, 479-492 (2010).
94. M. R. Sarrias et al., Biochemical characterization of recombinant and circulating
human Spalpha. Tissue antigens 63, 335-344 (2004).
95. J.-D. Tissot et al., IgM are associated to Spα (CD5 antigen-like).
Electrophoresis 23, 1203-1206 (2002).

Abstract (if available)

Abstract Immune cells are highly versatile and have many subsets. Each subset of the immune cells are known to exert their unique functions by specifically expressing a range of receptors capable of recognizing diverse ligands and activating the downstream signaling pathways. Liver resident macrophages, known as Kupffer cells, play significant functions in both physiological and pathophysiological progresses.Unfortunately, there are very few identified markers for Kupffer cells. The most widely used ones for mice are Clec4f, Vsig4 and Timd4, but no well-defined markers for human Kupffer cells has been reported. Without characterizing Kupffer cells, the involvement of Kupffer cells in the pathologies will remain elusive. Given this consideration, the goal of this project is to characterize Kupffer cell subpopulations by identifying novel markers. Using bioinformatics screening, we analyzed the published single cell RNA-seq data and identified Cd5l and Folr2 as two promising novel markers for Kupffer cells. We further validate these markers by using in silicone analysis to demonstrate the correlation between these two candidates and Clec4f. Both markers show great correlations with Clec4f and Vsig4, and are not correlated with the markers of other cell types. Using two steatosis mouse models representing alcoholic liver diseases and non-alcoholic liver disease, we confirmed the expression correlation of Cd5l and Clec4f in the livers using qPCR. In addition, we further confirmed the main source of Cd5l is KCs through immunofluorescent staining and flow cytometry. Collectively, our data shows that Cd5l is a putative novel marker for Kupffer cells.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Pro-inflammatory Kupffer cells secrete CXCL5 to promote liver cancer

PDF

Steatosis induces Wnt/β-catenin pathway to stimulate proliferation of hepatic tumor initiating cells and promote liver cancer development

PDF

Exploring the impact of elevated bile acids in metabolic dysfunction- associated steatotic liver disease in vitro

PDF

Transcriptional regulation of lipid metabolism and exploring the specific roles of AKT isoforms

PDF

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

PDF

Downregulation of osteopontin by PTEN in liver-specific Pten-null mouse model

PDF

Eicosanoids in the progression of liver steatosis to cancer

PDF

Methodology of exploring the role of SOX9 in the human HCC stem-like cells

PDF

Investigating sodium butyrate as a potential treatment for alcohol liver disease through the gut-liver axis

PDF

The roles of endoplasmic reticulum chaperones in regulating liver homeostasis and tumorigenesis

PDF

Downregulation of METTL7A gene in liver-specific Pten-null mouse model

PDF

Ethanol-HIF-1 alpha axis in inflammatory gene expression in liver cells

PDF

The study of DHM effects on counteracting ethanol intoxications

PDF

Role of the NF-κB-CXCL1 signaling pathway in the development of metabolic dysfunction-associated steatohepatitits

PDF

The immunomodulatory effects of midostaurin on T cells

PDF

Inhibition of NR3B1 attenuates the progression of NAFLD and NASH in liver-specific Pten knockout mice

PDF

Engineering B cells to express fully customizable therapeutic antibodies with enhanced Fc functions

PDF

Mechanisms of nuclear translocation of pyruvate dehydrogenase and its effects on tumorigenesis of hepatocellular carcinoma…

PDF

Disruption of immature Sertoli cell functions by endocrine disruptors and analgesic drugs

PDF

Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug

Asset Metadata

Creator Hong, Handan (author)

Core Title Characterization of Kupffer cell subpopulation in liver injury model through newly identified Kupffer cell markers

School School of Pharmacy

Degree Master of Science

Degree Program Pharmaceutical Sciences

Degree Conferral Date 2023-05

Publication Date 05/12/2023

Defense Date 05/11/2023

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

Tag alcoholic liver disease,chronic liver disease,Kupffer cell,liver injury,non-alcoholic liver disease,OAI-PMH Harvest

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Stiles, Bangyan (committee chair), Culty, Martine (committee member), Machida, Keigo (committee member)

Creator Email handanho@usc.edu,xiaanran1138@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113121752

Unique identifier UC113121752

Identifier etd-HongHandan-11835.pdf (filename)

Legacy Identifier etd-HongHandan-11835

Document Type Thesis

Format theses (aat)

Rights Hong, Handan

Internet Media Type application/pdf

Type texts

Source 20230512-usctheses-batch-1043 (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.

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