University of Southern California Dissertations and Theses

Characterizing the role of Dnmt1 and its downstream factors in progressive alopecia

(USC Thesis Other)

Characterizing the role of Dnmt1 and its downstream factors in progressive alopecia

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Characterizing the Role of Dnmt1 and its Downstream Factors in Progressive Alopecia
by
Wen-Chien Jea
A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(EXPERIMENTAL AND MOLECULAR PATHOLOGY)
December 2022
Copyright 2022 Wen-Chien Jea
ii
Acknowledgements
During these two years, I have been supported by many people, and I would like to show
my appreciation for each of them. First, I would like to thank my mentor, Dr. Cheng-Ming
Chuong, for welcoming me to join this big family and guiding me on how to think critically as a
scientist. He has provided us with many learning opportunities, and always kindly reminds me to
read more, learn more, and think more. His passion for science motivates me to push myself
moving forward and encourages me to think like an independent researcher. I also want to show
my gratitude to Dr. Randall Widelitz who has given me advice on my project. He always teaches
me with patience when I asked him about course materials and any project-related questions.
Thanks to Dr. Joseph Carlson who kindly agree to be one of my committee members. He is
friendly and willing to give me many suggestions for my project, and I have learned a lot from
him. I would like to appreciate Dr. Ya-Chen Liang, who is the main supervisor to teach me the
ideas and experimental techniques of this project. She has guided me with her expertise and
passion for science, and I am so grateful that I can have this great opportunity to work with her.
In addition, she is very organized and always plans well in advance when conducting an
experiment, and that attitude has deeply influenced me. Her mind has so many interesting ideas,
so I can always learn new things from her after having discussions.
For the experimental support, I would also like to thank Dr. Zhou Yu who taught me the
experimental techniques of cloning and lentivirus packaging. Besides, she often gives me advice
for improving my data quality. She taught me those experiments from the very beginning, and
she always guides me with patience. When I encounter difficulties or fail the experiments, she
would lead me to think and find resources to solve the problems. Thanks to Dr. TingXing Jiang
for helping and teaching me how to perform lentivirus injections and giving me advice on how to
iii
approve my immunostaining skill. I also want to show my appreciation to all the Chuong lab
members who are so friendly and will always give me support. I enjoy the atmosphere in Chuong
lab where people help each other and improve together, and I am glad to be a part of them.
During the first semester, I was taking online courses in Taiwan because of the pandemic.
Thanks to Dr. Chuong’s recommendation, I had an opportunity to join Dr. Michael Hughes' lab
at National Cheng Kung University. I would like to thank Dr. Hughes for warmly welcoming me
to join his lab and giving me the chance to practice some basic experimental techniques. It was a
great time for me to access a different research field from my previous experiences. I would also
like to thank my academic advisor, Michele King, who is so friendly and easy to talk to, and
kindly provided me with many resources and helps during these two years.
Finally, I would like to thank my family and friends who have supported me over the sea
or here beside me. I am so thankful to my parents who fully encouraged me to study abroad, so I
can have this amazing experience at USC. Although my family is not accompanying me
physically, their support through video calls provided me with energy and courage.

iv
Table of Contents
Acknowledgements .......................................................................................................................ii
List of Tables ................................................................................................................................v
List of Figures ..............................................................................................................................vi
Abstract .......................................................................................................................................vii
Chapter 1: Introduction ................................................................................................................1
Chapter 2: Results .........................................................................................................................8
2.1 Identity differentially expressed genes via comparing WT and Dnmt1-cKO mice
skin samples.........................................................................................................................8
2.2 Evaluate the expression level of the downstream factors with the focus on
immune cell infiltration......................................................................................................13
Chapter 3: Discussion and Future Directions............................................................................26
3.1 Significance..................................................................................................................26
3.2 Lentivirus-mediated epidermal overexpression of Klf2 and other molecules...............26
3.3 Future directions...........................................................................................................27
Intrinsic factors
3.3.1 Analyze whether stem cells property decline with increased age...................27
3.3.2 Methylome analysis.......................................................................................28
Extrinsic factors
3.3.3 Fibroblast heterogeneity................................................................................29
3.3.4 Infiltration of other types of leukocytes ........................................................30
3.3.5 Immunosuppressive drugs.............................................................................31
3.4 Similarity to a human disease.......................................................................................32
3.5 Conclusion....................................................................................................................33
Chapter 4: Materials and Methods.............................................................................................34
References ....................................................................................................................................41
v
List of Tables

Table 1. Summary of K14-Cre Dnmt1
fl/fl
mouse phenotype shows the alterations related to
hair structure, hair cycling, and hair regeneration ability..................................................................5
Table 2. Pathological defects of the K14-Cre Dnmt1
fl/fl
mouse at different ages..............................5
Table 3. The candidate genes’ functions related to immune regulation..........................................17
Table 4. Primers for genotyping.....................................................................................................34
Table 5. Primary and secondary antibodies used in this study........................................................36
Table 6. Fluorescence-labeled antibodies used in this study...........................................................37

vi
List of Figures

Figure 1. K14-Cre Dnmt1
fl/fl
mouse shows progressive alopecia phenotype....................................6
Figure 2. Hypothetical graph of potential intrinsic and extrinsic factors influences hair loss
phenotype.........................................................................................................................................7
Figure 3. Transcriptome and accessible chromatin profiling analyses using sorted BSCs
and the whole HFs from different aged Dnmt1-cKO and WT mice show the candidate DEGs
and inflammatory-related upstream regulators and enriched pathways..........................................10
Figure 4. Immunostaining results of the upregulated DEGs show higher expression intensity
in the bulge region, HFs, and microenvironment of Dnmt1-cKO mouse skin................................18
Figure 5. Immunostaining results of the upstream regulators TGF-β and TNF-α...........................21
Figure 6. Immunostaining images and quantification results of leukocyte marker CD45,
and macrophage markers F4/80 and MSR1....................................................................................22
Figure 7. FACS results of macrophage and dendritic cell quantification show macrophage
infiltration in the epidermis and dermis of Dnmt1-cKO mice with a mild or severe phenotype......24
Figure 8. The experimental flow of lentivirus-mediated epidermal gene overexpression...............27
Figure 9. The sorted hair follicle stem cell number shows total HFSC percentage is decreased
in Dnmt1-cKO mice compared to WT mice also decreased with increased age..............................28
Figure 10. The RRBS-Seq results by analyzing Bock et al.’s published data show the
potential regions of the candidates can be regulated by DNA methylation.....................................29
Figure 11. FACS results of WT and Dnmt1-cKO mice epidermis and dermis show that other
types of leukocytes need to be further confirmed............................................................................31
Figure 12. Map of plasmid construct for lentivirus-mediated epidermal Klf2 overexpression.......39

vii
Abstract
DNA methylation is an essential epigenetic modification that regulates the proliferation
or differentiation of epidermal progenitor and bulge stem cells (BSCs). DNA methyltransferase 1
(Dnmt1) is the major enzyme for maintaining DNA methylation patterns after cell replication.
Previously, Li et al. generated K14-Cre-Dnmt1
fl/fl
(Dnmt1-cKO) mouse to investigate the role of
Dnmt1 in hair cycling, and they concluded that Dnmt1 is crucial for regulating epidermal
progenitors and hair follicle (HF) homeostasis. However, the molecular mechanism of this
process remains unclear. Here, we utilize transcriptomic and accessible chromatin profiling using
whole HFs and sorted BSCs from different aged Dnmt1-cKO and wild-type (WT) mice. We
identify cKO-specific differentially expressed genes which also have enriched binding motifs on
cKO-specific differentially accessible chromatin regions, including Klf2, Cebpd, Snai1, Bhlhe41,
Pparg, and Lpin1. We also identify significantly enriched inflammation-related pathways in
BSCs from Dnmt1-cKO mice. Moreover, we observe the upregulation of these candidate genes
and inflammation in surrounding fibroblasts. Therefore, we hypothesize that intrinsic alterations
of epidermal progenitors and BSCs induce the secondary changes as extrinsic factors that cause
inflammatory cell infiltration and fibroblast heterogeneity resulting in progressive alopecia. The
immunostaining of candidate genes and FACS-sorting of inflammation-related cells suggest that
more macrophages infiltrate in Dnmt1-cKO mouse skin. Together, we suggest that losing Dnmt1
not only decreases stem cell numbers in epidermal progenitors and BSCs but also causes DNA
hypomethylation and activation of the candidate genes. These genes trigger the secondary
changes in the skin microenvironment, resulting in inflammation, fibroblast heterogeneity, and
impairing hair cycling.
1
Chapter 1: Introduction

All of the cells in an organism share the same genetic information; however, during
development, the cells could differentiate into diverse tissues and organs through gene
regulation. Gene regulation is a process that can decide whether a gene should be turned on or
off and then create distinct gene expression profiles in different cell types [1]. The expression of
genes can be regulated by various epigenetic mechanisms which could modify the function and
activity of genes but would not change the sequence of the DNA. One of the important chemical
DNA modifications is DNA methylation. In mammals, DNA methylation is critical for
embryonic development, gene transcription, genomic imprinting, and more [2]. In skin studies,
DNA methylation plays an important role in tissue self-renewal and progenitor maintenance in
the mammalian epidermis.
The biochemical process of DNA methylation is that the fifth position of cytosine will be
added to a methyl group and become the 5-methylcytosine (5mC), and the key enzyme that
catalyzes a methyl group transfer to DNA is called DNA methyltransferase (DNMT). S-adenosyl
methionine (SAM) is the methyl donor in this process. Most of the cytosines are methylated for
transcription silencing in mammals [3]. However, at CpG islands (CGIs), the genome region has
a high density of CpG dinucleotide repeats and is usually associated with gene promoters, less
than 10% of DNA methylation happens. CGIs usually involve transcription start sites of
housekeeping and developmental regulator genes, and it is important for enhancing DNA
accessibility and transcription factor binding [1, 4]. The well-known canonical DNMT enzymes,
which have catalytic DNMT activity, are DNMT1, DNMT3a, and DNMT3b. DNMT2 and
DNMT3L are two other human genomes encoded DNMT enzymes that do not have
2
methyltransferase activity [5]. There are two types of DNMT, de novo and maintenance
methyltransferase. The main de novo DNMTs, DNMT3a and DNMT3b, can newly methylate
cytosines in early embryo development, and they help to establish a whole new methylation
pattern. In addition, after cell replication, the methylation pattern is maintained by DNMT1,
which is the major maintenance DNMT and the most abundant DNMT in mammals [6].
Contrary to DNA methylation, the process that removing or modifying the methyl group from
5mC is called DNA demethylation. The enzymes belonging to the ten-eleven translocation
(TET) family are the main players in DNA demethylation [7].
In skin-related literature, Sen et al.’s study stated that DNMT1 is essential for
maintaining proliferation and suppressing differentiation in mammalian skin epidermal
progenitor cells [8]. Some studies stated that abnormal epigenetic signaling could be a reason for
cellular senescence and organismal aging [9]. For example, DNA methylation defects, a kind of
abnormal epigenetic signaling, play an important role in aging. Back in 1995, Courtois et al. had
already used the phototrichogram technique to observe the human scalp’s hair cycling during
aging. Their finding was that the hair growth’s duration and hair shaft diameter decreased, and
the interval between telogen and anagen was prolonged [10]. Ciccarone et al.’s results stated that
aging will affect the DNMTs expression; in detail, from their large-scale population research,
they confirmed that the expression of DNMT1 and DNMT3b linearly decreased with age [11].
Liu et al.’s study that used Dnmt1 heterozygous mouse model concluded that decreased genomic
methylation activity would badly affect the healthy aging process but not influence survival and
mortality [9].
According to the paper “Progressive Alopecia Reveals Decreasing Stem Cell Activation
Probability during Aging of Mice with Epidermal Deletion of DNA Methyltransferase 1,” Li et
3
al. study the role of DNMT1 for maintaining skin homeostasis and normal hair cycle [12]. The
authors presented the DNMT1 expression in the anagen (growing phase) and telogen (resting
phase) phases of normal hair cycling by immunohistochemistry images. They stated that
DNMT1 is expressed in the outer root sheath, inner root sheath, and hair matrix at the anagen
phase while in the hair germ at the telogen phase. They summarized that DNMT1 expression is
higher in proliferating cells and lower in differentiating cells. In order to investigate the role of
Dnmt1, they generated the transgenic mouse, K14-Cre Dnmt1
fl/fl
(Dnmt1-cKO), by conditionally
knocking out Dnmt1 in the K14-expressing epidermal progenitors and BSCs. They observed the
Dnmt1-cKO mouse show the alopecia phenotype that the hair follicles show a prolonged telogen
phase and fail to reenter the anagen phase. Moreover, they found that the hair size and number
reduced. The cell proliferation and differentiation are decreased while the apoptosis is increased
in the Dnmt1-cKO mouse. They stated that the activation process of BSCs is less efficient, and
the hair cycle would become more and more difficult to achieve anagen reentry (Table 1). They
claimed that the alopecia phenotype they observed is a progressive change, and we can find the
clues from the overview of WT and Dnmt1-cKO mice from young to aged (Table 2). Hair loss of
the Dnmt1-cKO mice becomes more and more obvious (Figure 1A). We find that the scalp
region will start showing hair loss around postnatal day 10 (P10), and most of the Dnmt1-cKO
mice have this mild hair loss phenotype that can be observed on the scalp and back regions.
Interestingly, we can find some mice present severe phenotype which show developmental delay,
movement difficulty, and hair loss. We can clearly see that the mice that have severe phenotype
have stronger K14-Cre and Dnmt1-recombinant bands based on our genotyping results (Figure
1B). That suggests that the phenotype would become more serious based on the genotype
difference. The study indicated the important role of DNA methylation in maintaining skin and
4
hair follicle homeostasis during development, hair cycling, and regeneration. Although they
presented the epidermis and hair phenotype of the Dnmt1-cKO mouse, the molecular mechanism
is still under investigation.
In this study, we assume that both intrinsic and extrinsic factors could contribute to this
mechanism. Intrinsic factors might be the alteration of stem cell properties and the alteration of
influences on adjacent cells. Li et al. stated that the hair follicle stem cells gradually lose their
property to activate hair cycling [12]. In addition, Zhang et al. indicated that reducing stem cells
could be related to epithelial cells escaping from the niche [13]. Furthermore, during aging, hair
follicle stem cell number is reduced cyclically because of epidermal differentiation, and they
found that the loss of stemness is because of the COL17A1 deficiency [14]. These studies all
addressed how intrinsic alteration, or alteration of the stem cell population, could lead to
outcomes that similar to the aging process. Here we also presume that these intrinsic influences
might trigger secondary changes induced by extrinsic factors which might modulate signals from
dermal niches within HFs. The signal can from the dermis, fibroblast, immune cells, adipose
tissue, vasculature, nerve... and so on (Figure 2). In this project, we will focus on whether the
alteration of epigenetic status, which means losing Dnmt1 in epidermal progenitors and BSCs,
could trigger extrinsic changes, such as immune cell attack, and finally resulting in alopecia
phenotype.

5

Table 1. Summary of K14-Cre Dnmt1
fl/fl
mouse phenotype shows the alterations related to hair
structure, hair cycling, and hair regeneration ability [12].

Table 2. Pathological defects of the K14-Cre Dnmt1
fl/fl
mouse at different ages [12].
6

Figure 1. K14-Cre Dnmt1
fl/fl
mouse shows progressive alopecia phenotype.
The hair loss phenotype starts from the scalp region of young Dnmt1-cKO mice, and the
phenotype becomes much more severe with increasing age (A). Based on the genotype
difference, the Dnmt1-cKO mice can either show mild or severe phenotypes (The three mice are
P27-P28 old) (B). Scale bar = 1 cm.
7

Figure 2. Hypothetical graph of potential intrinsic and extrinsic factors influences hair loss
phenotype.
The figure is created with BioRender.com.

8
Chapter 2: Results

2.1 Identify differentially expressed genes via comparing WT and Dnmt1-cKO mice skin
samples

In order to investigate which genes play important roles in the mechanism of losing
Dnmt1 in epidermal progenitors and BSCs results in hair loss, we collect different ages of WT
and Dnmt1-cKO mice skin samples to do the bulk RNA sequencing and further identify the
DEGs. There are two kinds of samples we collect (Figure 3A). One kind of sample is specifically
sorting the BSCs (Sca
-
, CD34
+
, and alpha6
+
) where intrinsic changes could happen. The other
one is using hair follicle-enriched fraction to collect the whole hair follicles and the adjacent
cells by low-speed centrifuge (20 rcf), and those cells represent the intrinsic and extrinsic
alteration contributors. After analyzing the bulk RNA-seq data, we would like to focus on the
transcription factors in the DEG list. To narrow down the targets, we further perform the Assay
for Transposase-Accessible Chromatin using sequencing (ATAC-seq) to investigate whether the
top enriched motifs overlap with the DEGs. We can find that many Klf family members and Batf
appear in 2-month-old and 1-year-old mice skin samples in which the hair cycle is at the telogen
phase (Figure 3B). Next, we narrow down the DEG list to 22 DEGs from the combination of
transcriptomic and accessible chromatin profiling data. We also can find that Batf, Cebpd, and
Klf2 appear in the lists from both BSC-specific and the whole HF samples (Figure 3C). Next, we
search the published information of the candidate genes in the Ingenuity Pathways Analysis
(IPA) database, and we found that four Gene Ontology (GO) terms frequently appear: (1)
immune and inflammatory responses, (2) cell proliferation, (3) cell differentiation, and (4)
9
apoptosis. Therefore, we grouped these genes into those four groups by their biological and
molecular functions, roles, and processes (Figure 3D) to further narrow the candidate gene list.
The upstream regulator analyses from the BSC-specific samples of different aged mice
reveal some inflammatory-related genes, such as MYC and TP53, and some cytokines, TGF-β
and TNF, can be found in the results (Figure 3E). In addition, many enriched pathways related to
inflammatory responses appear in canonical pathway analysis data. In detail, in 1-month-old
BSC-specific data, the inducible nitric oxide synthase (iNOS) signaling pathway shows up and it
is related to the macrophage inflammatory response which can trigger proinflammatory
responses [15]. In addition, EIF2 signaling was suggested to regulate proinflammatory cytokine
expression which is important for antiviral and antibacterial responses [16]. We also can find
some pathways associated with innate immune response, such as interleukin, interferon, and
chemokine. In the 2-month-old BSC-specific results, there are some pathways relevant to
adaptive immune responses including T cell and B cell pathways. Moreover, by comparing 1-
year-old data from BSC-specific and the whole HF groups, the enriched pathways, like defense
and immune responses, can be found in the results (Figure 3F).
Based on our upstream regulator and canonical pathway analyses, we suggest that the
DEGs grouped in immune and inflammatory responses are promising to do further evaluation.
On one hand, Klf2, Cebpd, and Batf, from both BSC-specific and the whole HF data, their
upregulation could possibly contribute to intrinsic and extrinsic influence. On the other hand, the
upregulated factors Snai1, Bhlhe41, Pparg, Lpin1, and Bcl6b which appear in the whole HF
result may involve the extrinsic influence.

10

11

12

Figure 3. Transcriptome and accessible chromatin profiling analyses using sorted BSCs and the
whole HFs from different aged Dnmt1-cKO and WT mice show the candidate DEGs and
inflammatory-related upstream regulators and enriched pathways.
Sorted BSCs and the whole HFs from 1-month, 2-month, and 1-year-old WT and Dnmt1-cKO
mice are used for bulk RNA-seq and ATAC-seq (A). The enriched Motifs list from the ATAC-
seq results (B) is used to combine with the DEGs data to narrow the candidate genes (the genes
highlighted in red represent the ones that overlap with the DEG list). The first narrowed target
transcription factors from the whole hair follicle sample (all) and BSC-specific sample
(highlighted in red) (C). Grouping candidates by the four GO terms that frequently appear in the
published information in the IPA database of these genes (D). The upstream regulator analysis
shows some inflammatory-related genes and cytokines (highlighted in red color) in different
aged BSC-specific results (E). Canonical pathway analyses also show several inflammatory-
related pathways (highlighted in red color) in BSC-specific and whole HFs samples at different
aged (F). Sample collection flow (A) is created with BioRender.com.

13
2.2 Evaluate the expression level of the downstream factors with the focus on immune cell
infiltration

In Li et al.’s study, they mainly focused on how intrinsic changes could contribute to hair
loss phenotype. They counted the BrdU-positive cells to confirm that cell proliferation is
decreased in Dnmt1-cKO mouse HFs within the anagen phase. In addition, they conducted the
CldU and IdU labeling during the anagen phase and stated decreased cell differentiation in
Dnmt1-cKO HFs. Furthermore, the TUNEL assay was performed and indicated that apoptosis is
increased in the anagen phase, too [12]. However, they had not provided evidence showing how
extrinsic factors play a role in the mechanism. In this study, we would like to explore more about
intrinsic influence, and also bring in the idea of extrinsic contribution resulting in hair loss.
We would like to further evaluate our candidate genes, Klf2, Cebpd, Batf, Snai1,
Bhlhe41, Pparg, Lpin1, and Bcl6b which are in the group of “immune and inflammatory
responses (Figure 3D),” and from previous publishment, these genes have been addressed in
their functions related to immune regulation (Table 3). Here, immunostaining is performed to
investigate these protein expression patterns and compare Dnmt1-cKO skin to WT skin. Since
they are all upregulated DEGs in Dnmt1-cKO mouse skin, we expected that their expression
level would be higher in the Dnmt1-cKO skin sample than in WT. Since Klf2 and Cebpd are the
DEGs from both BSC-specific and the whole HF samples, we double-stained them with the BSC
marker, K15, so we can check the expression level in the bulge region as well. From our
immunostaining images of Klf2 and K15 dual staining, Klf2 expression can be found within HF,
especially the bulge region, and some positive signals appear near the HF. For Cebpd and K15
double staining, we can observe that Cebpd expression is higher within HF, including the inner
14
root sheath and bulge region. However, no positive cells can be found in the HF niche (Figure
4A). Therefore, we suggest that upregulating Klf2 might play a role in changing stem cells or the
cells in the microenvironment in an intrinsic and extrinsic manner while Cebpd may contribute
more to intrinsic changes in Dnmt1-cKO mouse skin. Because Snai1, Bhlhe41, Pparg, and Lpin1
only appear in the whole HF data, we first check their expression by single staining. For the
Snai1 image, stronger signals are not only within the HF but also in the dermis. There are two
possibilities, one is that the cells in the dermis themselves have more Snai1 expressions in the
Dnmt1-cKO situation. The other is the Snai1-positive cells might migrate from the epidermis to
the dermis since Snai1 is a well-known EMT marker that plays an important role in controlling
cell-cell adhesion, cell migration, and ECM degradation [25]. In order to dig deep into this,
double staining with other cell markers is needed, such as immune cells or fibroblasts. Lpin1 also
has expression in both HF and dermis while Bhlhe41 and Pparg are mainly showing higher
expression levels within the HFs (Figure 4B). The single staining data suggest that higher
expression intensity of Snai1, Bhlhe41, Pparg, and Lpin1 can be observed in Dnmt1-cKO mouse
skin than in WT which matches our bulk RNA-seq data.
Based on our bulk RNA-seq results and IPA analysis, TGF-β and TNF are the upstream
regulators in Dnmt1-cKO mouse skin. Therefore, immunostaining is performed to check the
expression level and pattern of the cytokine markers, TGF-β (3 isoforms) and TNF-α. In the WT
mouse skin images, the most of positive signals appear within HF. On contrary, the expression
intensity of TGF-β and TNF-α is not higher within HF in Dnmt1-cKO mouse skin, but the
positive signals in the dermis are much more in Dnmt1-cKO than in WT mouse skin (Figure 5).
We assume that more cytokines are released into the HF niche and dermis in Dnmt1-cKO mouse
skin, but further quantification by using q-PCR or other methods needs to be conducted.
15
Since we assume knocking out Dnmt1 in epidermal progenitors and BSCs could trigger
secondary chances; in other words, some extrinsic factors, such as immune cells and other cells
within the dermal niche, could play important roles in leading to hair loss. We also identify
several upstream regulators and canonical pathways that are inflammatory-related. Therefore, we
would like to investigate whether immune cell infiltration or abnormal immune activation could
be observed in the Dnmt1-cKO mouse skin microenvironment. First, we choose the leukocyte
marker CD45 to check whether immune activation happens in general. The images show that
more leukocytes infiltrate in Dnmt1-cKO skin, especially in the dermis compared to WT mouse
skin. Next, we select the macrophage markers F4/80 and MSR1 to do the immunostaining
because of the clues from canonical pathway analysis. The data suggest that more F4/80, and
MSR1 positive cells in the Dnmt1-cKO mouse skin than WT mouse, especially in the dermal
niche as well (Figure 6A). In order to answer the question of whether Dnmt1-cKO mouse skin
shows more inflammatory cell infiltration, we further conduct two experiments to quantify these
positive cells to confirm the increasing level. On one hand, we simply quantify the CD45, F4/80,
and MSR1 positive cells around hair follicles manually from immunostaining images. The
results show that more leukocytes, especially macrophages, are infiltrating in Dnmt1-cKO mice
skin (Figure 6B). Since immunostaining is not considered a reliable quantification method, on
the other hand, we conducted fluorescence-activated cell sorting (FACS) to directly quantify the
number of macrophages and dendritic cells (DCs), which are the key immune cells in innate
immunity, in WT and Dnmt1-cKO mice skin. The Dnmt1-cKO mice with mild phenotype and
severe phenotype, which are about 1 month old and whose hair cycle is at the anagen phase, have
been used in the FACS experiment (Figure 1B). The sample preparation is following the protocol
from Lou et al.’s group (Figure 7A) [26], and we adjust their immune cell sorting strategy
16
(Figure 7B). First, the CD45
+
leukocytes are separated by CD64 which is a macrophage-specific
marker. Next, the CD64
-
cells will be grouped by F4/80 which some DCs will express this
surface marker. Finally, CD64
+
cells will be further gated by F4/80 and CD11b which are both
macrophage markers [27]. The data suggest that more macrophages are infiltrating in Dnmt1-
cKO epidermis and dermis compared to WT, especially significant in severe phenotype cKO
mice. However, the DC number does not show a significant enhancement in Dnmt1-cKO mice
(Figure 7C). So far, we have done 2 repeats of Dnmt1-cKO mice with mild phenotype and WT.
However, we only have done one experiment by using a severe phenotype Dnmt1-cKO mouse
because the mouse with severe phenotype is rarer than the mild one. Therefore, we will need to
conduct more repeats to ensure the result is consistent.
To sum up, in Aim 2, we explore more about intrinsic factors and involve extrinsic
factors. Our immunostaining data of Klf2 and Cebpd with K15 suggest that their upregulation
could be an intrinsic and extrinsic factor, and other genes would probably contribute to extrinsic
influences. In addition, the extrinsic factor we focus on in this study is abnormal immune
activation, our immunostaining and FACS data suggest that there is more inflammatory-related
cell infiltration in the Dnmt1-cKO mice skin.

17

Table 3. The candidate genes’ functions related to immune regulation.
18

19

20
Figure 4. Immunostaining results of the upregulated DEGs show higher expression intensity in
the bulge region, HFs, and microenvironment of Dnmt1-cKO mouse skin.
Klf2 and K15 double staining shows that the expression intensity of Klf2 is higher in the bulge
region of Dnmt1-cKO mouse skin than in WT mouse, and some positive signals can be found in
the hair follicle niche. Double staining of Cebpd and K15 also reveals Cebpd is higher expressed
in the bulge region (A). Immunostaining images of Snai1, Bhlhe41, Lpin1, and Pparg show that
the expression is mostly located within HF in WT mouse skin while significantly more positive
signals showing in the dermis of Dnmt1-cKO mouse than WT (B). Scale bar = 100 µm, *
indicates autofluorescence. Skin samples from 6 to 8 months old WT and Dnmt1-cKO mice are
used.

21

22
Figure 5. Immunostaining results of the upstream regulators TGF-β and TNF-α.
In WT mouse skin, TGF-β and TNF-α positive signals mainly locate within HFs, but much more
positive signals can be found in the Dnmt1-cKO mouse skin dermis. Scale bar = 200 µm. Skin
samples from 6 to 8 months old WT and Dnmt1-cKO mice are used.

23

Figure 6. Immunostaining images and quantification results of leukocyte marker CD45, and
macrophage markers F4/80 and MSR1.
Immunostaining results of CD45, F4/80, and MSR1 suggest more leukocytes and macrophages
infiltrate in Dnmt1-cKO mouse skin, significantly increased in the dermis, than in WT mouse
(A). The quantification results by calculating the positive cells near each HFs within a distance
of up to 50 mm show more leukocytes and macrophages are infiltrating in Dnmt1-cKO mouse
skin (B). n =10. Scale bar = 200 µm, * indicates autofluorescence. Skin samples from 6 to 8
months old WT and Dnmt1-cKO mice are used.

24

25

Figure 7. FACS results of macrophage and dendritic cell quantification show macrophage
infiltration in the epidermis and dermis of Dnmt1-cKO mice with a mild or severe phenotype.
Sample preparation is following the protocol of Lou et al.’s group (A) [15]. The sorting strategy
is adjusted from the protocol of Lou et al. (B). The bar chart shows the normalized macrophage
and DC number from the epidermis and dermis of Dnmt1-cKO mice with either mild or severe
phenotype and WT mice (C). The macrophage and DC numbers are normalized with the total
live cells. WT and cKO-mild phenotype, n = 2; cKO-severe phenotype, n = 1. The data were
analyzed by GraphPad Prism software. The experimental flow is created with BioRender.com.

26
Chapter 3: Discussion and Future Directions

3.1 Significance
In this study, we investigated more about how the intrinsic change that epidermal
methylation level is altered by knocking out Dnmt1 in epidermal progenitors and BCSs could
trigger secondary extrinsic changes that inflammatory-related components become more in the
dermal niche. From our transcriptomic and accessible chromatin profiling analyses, we not only
identify the candidate transcription factors but also observe the activation of inflammatory
responses. Therefore, we suggest that epidermal progenitors and BSCs with abnormal
methylation patterns may induce chronic inflammatory changes, including immune cell
infiltration and aberrant activation of fibroblasts.

3.2 Lentivirus-mediated epidermal overexpression of Klf2 and other molecules
In order to evaluate the function of these selected factors, lentivirus-mediated epidermal
gene overexpression is performed. Our first investigation is to overexpress Klf2 in the B6 mice
epidermis. We package the Klf2 cDNA insertion plasmid (Figure 12) into lentivirus and inject it
into E9 mouse embryos. After the babies are born, we check the GFP signal on their skin at P1
(Figure 8), and hair loss phenotype is expected to be observed when they grow up. So far, we
keep observing the phenotype; at the same time, we will shave half of their dorsal hairs to
investigate whether the regrow ability of hairs is influenced.
27

Figure 8. The experimental flow of lentivirus-mediated epidermal gene overexpression.
Here we use Klf2 as one example, the lentivirus for overexpressing Klf2 in mouse epidermis is
injected into E9 embryos, and the GFP signal will be checked at P1. The figure is created with
BioRender.com.

3.3 Future directions
Intrinsic factors
3.3.1 Analyze whether stem cells property decline with increased age
Since intrinsic changes might still occur during the aging process, we will analyze our
RNA-seq results by comparing the different ages of the Dnmt1-cKO mice. From our lab's
previous hair follicle stem cell sorting results, we first observe that the stem cell percentage of
Dnmt1-cKO is significantly lower than the stem cell percentage of WT mice. Interestingly, the
percentage of stem cells is lower in 2-year-old Dnmt1-cKO mice by comparing to 1-year-old
(Figure 9). Therefore, we assume that stem cell numbers and properties could still be lost during
the aging process, and that might be another intrinsic influence.
28

Figure 9. The sorted hair follicle stem cell number shows total HFSC percentage is decreased in
Dnmt1-cKO mice compared to WT mice also decreased with increased age.
This figure was analyzed by Dr. Ya-Chen Liang.

3.3.2 Methylome analysis
As our mouse model is conditionally knocking out Dnmt1 in epidermal progenitors and
BSCs, we would like to investigate whether the methylation patterns of the candidate genes
would be changed. Bock et al.’s work published in 2012 used four different skin cells, telogen
(quiescent) BSC (TBSC), anagen (active) BSC (ABSC), matrix transient amplifying cell
(MTAC), and companion layer differentiated cell (CLDC), to perform Reduced-representation
bisulfite sequencing (RRBS-Seq) [28]. We check our candidate genes from their published
sequencing data, and we found that the candidate transcription factors’ promoter regions could
be regulated by DNA methylation (Figure 10). In the future, we plan to conduct Methylated
DNA immunoprecipitation sequencing (MeDIP-Seq) to get more information about how
methylation patterns could be changed after losing Dnmt1 in basal layer cells.
29

Figure 10. The RRBS-Seq results by analyzing Bock et al.’s published data [28] show the
potential regions of the candidates can be regulated by DNA methylation.
Use Klf2 as one example, and this figure was analyzed by Dr. Ya-Chen Liang.

Extrinsic factors
3.3.3 Fibroblast heterogeneity
From our immunostaining images, we found that there are many fibroblasts in the dermal
niche that might also play an important role as an extrinsic factor to alter epidermal and stem cell
homeostasis. In the review paper by Plikus et al., they addressed that abnormal activation of
activated fibroblast could be the cause of uncontrolled extracellular matrix synthesis and
increased fibrosis and scarring during chronic injury or inflammation. Furthermore, they stated
that some special fibroblast subpopulations may appear during unstabled states, such as
30
inflammation [29]. Therefore, during inflammation, fibroblasts would be stimulated to work for
fibrosis or even scarring. For future work, we are going to investigate more on the role of
fibroblast as an extrinsic inducer because fibrosis may strangulate hair follicles and lead to hair
loss.

3.3.4 Infiltration of other types of leukocytes
Our FACS data indicate that macrophage infiltration in Dnmt1-cKO mice epidermis and
dermis. Based on the analysis, we found that other types of leukocytes can also infiltrate the skin,
but we need to further do immunostaining and FACS to check the percentage of each type of
leukocyte (Figure 11). The enriched pathways relevant to Th1/Th2 activation and B cell
signaling can be found in our canonical pathway analyses, as well as natural killer cell signaling.
Therefore, the investigation of those leukocytes could help us understand what types of
inflammatory cell infiltration are more significant in the Dnmt1-cKO mouse skin
microenvironment.

31

Figure 11. FACS results of WT and Dnmt1-cKO mice epidermis and dermis show that other
types of leukocytes need to be further confirmed.
The percentages shown in red represent the remaining leukocytes after subtracting the
percentages of macrophage and DC from total leukocyte.

3.3.5 Immunosuppressive drugs
In order to confirm immune responses have a direct impact on causing hair loss, we
would like to treat immunosuppressive drugs for Dnmt1-cKO mice at a young age to see whether
the drugs can rescue the alopecia phenotype. The current therapeutic option for LPP is PPARg
agonists and FFA is hormone modulators [30]. Since PPARg is one of our candidates, treating
the Dnmt1-cKO mouse with PPARg agonists might be taken into consideration. Furthermore,
32
Corrêa-Moreira et al.’s study, they treated Dexamethasone, an immunosuppressive agent that can
induce CD4
+
T lymphocyte depletion, in mice drinking water. The treated mice showed lower
percentages of macrophages and neutrophils [31].

3.4 Similarity to a human disease
In general, alopecia can be classified into two groups, noncicatricial and cicatricial
(scarring) alopecia [32]. Scarring alopecias, or primary cicatricial alopecias (PCA), are chronic
inflammatory disorders caused by inflammatory responses breaking the ability of epithelial hair
follicle stem cells to process hair cycling and regeneration, then causing permanent hair loss.
Several studies indicated that inflammatory infiltration could be observed around the bulge area
in PCA cases [33]. The three most common PCA are Lichen Planopilaris (LPP), Frontal
Fibrosing Alopecia (FFA), and Central Centrifugal Cicatricial Alopecia (CCCA). Wang et al.
published their bulk RNA-seq data showing the enriched pathways, and they found that mast
cell, fibrosis, and immune signaling genes are upregulated [30]. Interestingly, in LPP pathway
analysis, four enriched pathways are overlapping with our Dnmt1-cKO analysis, including EIF2
signaling, FXR/RXR activation, LPS/IL-1 mediated inhibition of RXR function, and fatty acid-
oxidation I pathways. According to Chen et al.’s study, Th1-biased cytotoxic T cell infiltration
around the bulge region can be observed in LPP cases. Our bulk RNA-seq analysis also shows
that the Th1 pathway is one of the canonical pathways in Dnmt1-cKO mouse skin. Therefore, we
hope our research could provide more information for future alopecia research, skin disease
studies, or therapeutic research and development.

33
3.5 Conclusion
In this project, we used the Dnmt1-cKO mouse to investigate the mechanism of how
losing Dnmt1 in epidermal progenitors and BSCs leads to progressive alopecia phenotype. We
hypothesize that alopecia can be caused by intrinsic and/or extrinsic factors. Here, for intrinsic
influence, we focus on the alteration of epigenetic status, and for extrinsic influence, we examine
the activation of immune cells. In order to explore intrinsic factors deeper, we identify the
upregulated transcription factors from transcriptome analyses. The expression intensity in
Dnmt1-cKO mouse skin by immunostaining represents the upregulation of these genes. In
addition, upstream regulator and canonical pathway analyses reveal the activation of
inflammatory-related cells, among other changes. Therefore, we further identify immune
infiltration as an extrinsic component that is evaluated by immunostaining and FACS. Those data
indicate that more macrophages infiltrate Dnmt1-cKO mice's skin from a young age. In
summary, we suggest that losing Dnmt1 not only reduces the numbers of epidermal progenitors
and BSCs but also leads to the candidate genes' upregulation. These intrinsic factors trigger the
extrinsic factors, such as inflammatory cell infiltration and fibroblast heterogeneity, in the
Dnmt1-cKO skin microenvironment, and those changes would result in hair loss phenotype.

34
Chapter 4: Materials and Methods

Mice
WT (Dnmt1
fl/fl
) and Dnmt1-cKO (K14-Cre Dnmt1
fl/fl
homozygote) mice were used in this
study. Their tail skins were harvested and genotyped by using the primers (Table 4) referred to
Li et al. [12]. DreamTaq Green PCR Master Mix (Thermo Scientific, K1082) was used for PCR
amplification. The PCR amplifications were performed under the following conditions: initial
denaturation at 94°C for 2 minutes, followed by 30 cycles of denaturation at 94°C for 15
seconds, primer annealing at 60°C for 30 seconds, and extension at 72°C for 40 seconds, with a
final extension at 72°C for 5 minutes. Both WT and Dnmt1-cKO samples can be amplified with
a 368-bp DNMT1
fl/fl
fragment. Dnmt1-cKO samples can also be amplified with a 335-bp K14-
Cre and a 250-bp recombinant fragments.
Primer Name Primer Sequence (5’  3’)
Dnmt1-1 GGG CCA GTT GTG TGA CTT GG
Dnmt1-2 CTT GGG CCT GGA TCT TGG GGA TC
Dnmt1-3 ATG CAT AGG AAC AGA TGT GTG C
Cre-F TTG CCC CTG TTT CAC TAT CCA G
Cre-R ATG GAT TTC CGT CTC TGG TG
Table 4. Primers for genotyping.

RNA-seq
The procedures of sample preparation, library establishment, and data analysis are
referred to Liang et al. [34].

35
ATAC-seq
The procedures of sample preparation, library establishment, and data analysis are
referred to Liang et al. [34].

Immunofluorescence staining
Mouse dorsal skins were harvested and then fixed with 4% paraformaldehyde in PBS at
4°C overnight. After embedding tissues in paraffin, cut the 7-μm thickness sections with a
microtome (Leica). After deparaffinization and rehydration, sections were treated with sodium
citrate or ethylenediaminetetraacetic acid (EDTA) buffers to unmask the antigens and epitopes.
After antigen retrieval, sections were submerged in 30% hydrogen peroxide in methanol for 10
minutes to reduce autofluorescence. After washing in TBS with 0.1% Tween 20 (TBST),
sections were blocked by using Zeller’s buffer (10mM Tris, 100mM MgCl2, 5% FBS, 1% BSA,
and 0.5% Tween 20 in ddH2O) for 30 minutes. Incubated sections with primary antibodies
(Table 5) at 4°C overnight. Secondary antibodies were used at room temperature for 2 hours,
then stained nuclei with DAPI (1:1000 in TBST) at room temperature for 30 minutes. Sections
were mounted with 50% glycerol in ddH2O and then stored at 4°C. The images were taken with
KEYENCE (BZ-X710) microscope.

36
Primary Antibody Ratio Company Catalog Number
DNMT1 1:200 GeneTex GTX116011
KLF2 1:200 Bioss BS-2772R
CEBPD 1:200 Rockland 600-401-A61S
SNAI1 1:200 Proteintech 13099-1-AP
BHLHE41 1:200 MyBioSource MBS8817280
PPARG 1:200 Proteintech 16643-1-AP
LPIN1 1:200 Proteintech 27026-1-AP
K15 1:200 Thermo Fisher MA1-90929
CD45 1:200 Proteintech 20103-1-AP
F4/80 1:200 Proteintech 28463-1-AP
MSR1 1:200 Proteintech 17858-1-AP
TGF-β1 1:200 GeneTex GTX110630
TGF-β2 1:200 Santa Cruz [35]
TGF-β3 1:200 Santa Cruz Sc-83
TNF-α 1:200 Santa Cruz sc-52746
Secondary Antibody Ratio Company Catalog Number
Goat anti-Rabbit IgG,
Alexa Fluor™ 594
1:100-200 Invitrogen A-11072
Donkey anti-Goat IgG,
Alexa Fluor™ 488
1:100-200 Invitrogen A-11055
Goat anti-Mouse IgG,
Alexa Fluor™ 594
1:100-200 Invitrogen A-11005
Table 5. Primary and secondary antibodies used in this study.

37
Immune cell quantification
The immune cell quantification method adjusted from Xie et al. [36] is performed by
manually quantifying the number of positive signals (Figure 6) in the area around the hair
follicles within a distance of up to 50 mm. The data were analyzed by GraphPad Prism software.

Fluorescence-activated Cell Sorting (FACS)
WT and Dnmt1-cKO mice dorsal skin samples were harvested. The sample digestion and
antibody staining procedures referred to the protocol from Lou et al. [26]. The antibodies used
were listed in Table 6. The macrophages and dendritic cells were sorted on the BD Aria II flow
cytometer at USC Stem Cell Facility.

Antibody Fluorophore Ratio Company Catalog Number
CD45 Alexa Fluor™ 700 1:200 eBioscience 56-0451-80
CD64 FITC 1:10 SinoBiological 50086-R027-F
F4/80 PE 1:200 BD Pharmingen 565410
CD11b APC 1:200 eBioscience 17-0112-81
DAPI BV421 1:1000 BD Pharmingen 564907
Table 6. Fluorescence-labeled antibodies used in this study.

Cell culture
293T cells for lentivirus packaging were cultured at 37°C with 5% CO2 cell incubator
and maintained in DMEM medium (DMEM (1X) + GlutaMAX™, Gibco, 10569-010)
supplemented with 10% (v/v) FBS (Gibco, 10082-147), 1% (v/v) Pen-Strep (Gibco, 15070-063),
38
and 500 µg/mL Geneticin (Gibco, 10131-027).
DF-1 cells for lentivirus titration were cultured at 38°C with 5% CO2 cell incubator and
maintained in DMEM medium supplemented with 10% (v/v) FBS, 1% (v/v) Pen-Strep, and 2%
chicken serum.

Lentivirus construct production
Here we use Klf2 as a candidate gene example. In order to overexpress Klf2 in the
mouse’ epidermis, the full length of Klf2 cDNA was amplified from B6 mouse cDNA. To get
the B6 mouse cDNA, we first collected the lung tissue that expressed higher Klf2 than other
organs. Next, the RNA was extracted by using the AllPrep DNA/RNA Micro Kit (QIAGEN®,
80284), then the cDNA was reverse transcribed from the RNA by using the All-In-One 5X RT
MasterMix (Applied Biological Materials Inc., G592). For preparing the lentiviral vector for
cloning, we first modified the LV-RFP [37] by introducing T2A self-cleaving peptide between
hPGK promoter and reporter sequence, and also replaced RFP with our wanted reporter gene
EGFP. Next, Klf2 cDNA was amplified by using PrimeSTAR® GXL Premix (Takara Bio USA,
Inc., R052). Finally, Klf2 cDNA was inserted into the vector by using the In-Fusion Snap
Assembly cloning kits (Takara Bio USA, Inc., ST2320). The final plasmid construct is shown in
Figure S5. For the control group, we generated the construct that only replaced RFP with EGFP
without introducing T2A or other genes between hPGK promotor and EGFP.

39

Figure 12. Map of plasmid construct for lentivirus-mediated epidermal Klf2 overexpression.
This map is created with SnapGene.

High titer lentiviral production and titration
For high-titer lentivirus packaging, the protocol was adjusted from Yu et al. [38]. We
conducted calcium phosphate transfection of 293T cells with the Klf2-expressing plasmid and
helper plasmids pMD2.G and pPAX2. 46 hours after transfection, we collected and filtered viral
supernatant through 0.45 µM Millipore low-protein binding filter units (VWR, 10040-462). We
further performed a two-step centrifuge for an ~2000-fold concentration. First, we concentrated
the virus by using 100 kDa MW cutoff Millipore Centricon 70 Plus (Millipore, UFC701008).
Second, we ultracentrifuged the virus by using Beckman Optima UltraCentrifuge with the SW55
rotor.
40
For lentivirus titration, we infected the concentrated lentivirus to DF-1 cells which were
cultured on coverslips. We would manually count the infected cells expressing GFP signal under
a microscope (KEYENCE, BZ-X710) of a few random 20x fields. The titer of the lentivirus is
around 10
8
-10
9
cfu/mL.

In utero lentivirus injection
We used C57BL/6J mice (Strain #:000664) which were ordered from Jackson Laboratory
to do the lentivirus injection. Female mice at E9 of gestation were analgesized with 0.3 mg/kg
Meloxicam and anesthetized with 0.25 mg Ketamin. We injected each embryo with 1-2 μl of
lentivirus, and around half of the total embryos were injected. The GFP signal expressed on the
skin was checked under a fluorescence microscope at P1. All the surgery was performed with
protocols approved by the USC IACUC.

41
References

1. Moore LD, Le T, Fan G. DNA Methylation and Its Basic Function.
Neuropsychopharmacology. 2012;38(1):23-38. doi:10.1038/npp.2012.112

2. Robertson KD. DNA methylation and human disease. Nature Reviews Genetics.
2005;6(8):597-610. doi:10.1038/nrg1655

3. Greenberg MVC, Bourc’his D. The diverse roles of DNA methylation in mammalian
development and disease. Nature Reviews Molecular Cell Biology. 2019;20(10):590-607.
doi:10.1038/s41580-019-0159-6

4. Smith ZD, Meissner A. DNA methylation: roles in mammalian development. Nature Reviews
Genetics. 2013;14(3):204-220. doi:10.1038/nrg3354

5. Lyko F. The DNA methyltransferase family: a versatile toolkit for epigenetic regulation.
Nature Reviews Genetics. 2017;19(2):81-92. doi:10.1038/nrg.2017.80

6. Bhattacharyya M, De S, Chakrabarti S. Origin and Evolution of DNA methyltransferases
(DNMT) along the tree of life: A multi-genome survey. Published online April 9, 2020.
doi:10.1101/2020.04.09.033167

7. Kohli RM, Zhang Y. TET enzymes, TDG and the dynamics of DNA demethylation. Nature.
2013;502(7472):472-479. doi:10.1038/nature12750

8. Sen GL, Reuter JA, Webster DE, Zhu L, Khavari PA. DNMT1 maintains progenitor function
in self-renewing somatic tissue. Nature. 2010;463(7280):563-567. doi:10.1038/nature08683

9. Liu L, van Groen T, Kadish I, et al. Insufficient DNA methylation affects healthy aging and
promotes age-related health problems. Clinical Epigenetics. 2011;2(2):349-360.
doi:10.1007/s13148-011-0042-6

10. COURTOIS M, LOUSSOUARN G, HOURSEAU C, GROLLIER JF. Ageing and hair
cycles. British Journal of Dermatology. 1995;132(1):86-93. doi:10.1111/j.1365-
2133.1995.tb08630.x

11. Ciccarone F, Malavolta M, Calabrese R, et al. Age-dependent expression
ofDNMT1andDNMT3Bin PBMCs from a large European population enrolled in the MARK-
AGE study. Aging Cell. 2016;15(4):755-765. doi:10.1111/acel.12485

12. Li J, Jiang TX, Hughes MW, et al. Progressive alopecia reveals decreasing stem cell
activation probability during aging of mice with epidermal deletion of DNA methyltransferase 1
(DNMT1). The Journal of investigative dermatology. 2012;132(12):2681-2690.
doi:10.1038/jid.2012.206
42

13. Zhang C, Wang D, Wang J, et al. Escape of hair follicle stem cells causes stem cell
exhaustion during aging. Nature Aging. 2021;1(10):889-903. doi:10.1038/s43587-021-00103-w

14. Matsumura H, Mohri Y, Binh NT, et al. Hair follicle aging is driven by transepidermal
elimination of stem cells via COL17A1 proteolysis. Science. 2016;351(6273).
doi:10.1126/science.aad4395

15. McNeill E, Crabtree MJ, Sahgal N, et al. Regulation of iNOS function and cellular redox
state by macrophage Gch1 reveals specific requirements for tetrahydrobiopterin in NRF2
activation. Free Radical Biology and Medicine. 2015;79:206-216.
doi:10.1016/j.freeradbiomed.2014.10.575

16. Shrestha N, Bahnan W, Wiley DJ, Barber G, Fields KA, Schesser K. Eukaryotic Initiation
Factor 2 (eIF2) Signaling Regulates Proinflammatory Cytokine Expression and Bacterial
Invasion. Journal of Biological Chemistry. 2012;287(34):28738-28744.
doi:10.1074/jbc.m112.375915

17. Turpaev KT. Transcription Factor KLF2 and Its Role in the Regulation of Inflammatory
Processes. Biochemistry (Moscow). 2020;85(1):54-67. doi:10.1134/s0006297920010058

18. Lan X, Wang H, Qi R, et al. Shikonin inhibits CEBPD downregulation in IL-17-treated
HaCaT cells and in an imiquimod-induced psoriasis model. Molecular Medicine Reports.
2020;22(3):2263-2272. doi:10.3892/mmr.2020.11315

19. Zhang P, Liu L, Lai X, et al. Ablation of Basic Leucine Zipper Transcription Factor ATF-
Like Potentiates Estradiol to Induce Atopic Dermatitis. Ali Sheikh MS, ed. Oxidative Medicine
and Cellular Longevity. 2022;2022:1-11. doi:10.1155/2022/7024669

20. Zhang F, Wang H, Wang X, et al. TGF-β induces M2-like macrophage polarization via
SNAIL-mediated suppression of a pro-inflammatory phenotype. Oncotarget. 2016;7(32):52294-
52306. doi:10.18632/oncotarget.10561

21. Rauschmeier R, Gustafsson C, Reinhardt A, et al. Bhlhe40 and Bhlhe41 transcription factors
regulate alveolar macrophage self‐renewal and identity. The EMBO Journal. 2019;38(19).
doi:10.15252/embj.2018101233

22. Karnik P, Tekeste Z, McCormick TS, et al. Hair Follicle Stem Cell-Specific PPARγ Deletion
Causes Scarring Alopecia. Journal of Investigative Dermatology. 2009;129(5):1243-1257.
doi:10.1038/jid.2008.369

23. Dai W, Zheng P, Luo D, et al. LPIN1 Is a Regulatory Factor Associated With Immune
Response and Inflammation in Sepsis. Frontiers in Immunology. 2022;13.
doi:10.3389/fimmu.2022.820164

43
24. Manders PM, Hunter PJ, Telaranta AI, et al. BCL6b mediates the enhanced magnitude of the
secondary response of memory CD8+ T lymphocytes. Proceedings of the National Academy of
Sciences. 2005;102(21):7418-7425. doi:10.1073/pnas.0501585102

25. Yang J, Antin P, Berx G, et al. Guidelines and definitions for research on epithelial–
mesenchymal transition. Nature Reviews Molecular Cell Biology. 2020;21(6):341-352.
doi:10.1038/s41580-020-0237-9

26. Lou F, Sun Y, Wang H. Protocol for Flow Cytometric Detection of Immune Cell Infiltration
in the Epidermis and Dermis of a Psoriasis Mouse Model. STAR Protocols. 2020;1(3):100115.
doi:10.1016/j.xpro.2020.100115

27. Lecoultre M, Dutoit V, Walker PR. Phagocytic function of tumor-associated macrophages as
a key determinant of tumor progression control: a review. Journal for ImmunoTherapy of
Cancer. 2020;8(2):e001408. doi:10.1136/jitc-2020-001408

28. Bock C, Beerman I, Lien WH, et al. DNA Methylation Dynamics during In Vivo
Differentiation of Blood and Skin Stem Cells. Molecular Cell. 2012;47(4):633-647.
doi:10.1016/j.molcel.2012.06.019

29. Plikus MV, Wang X, Sinha S, et al. Fibroblasts: Origins, definitions, and functions in health
and disease. Cell. 2021;184(15):3852-3872. doi:10.1016/j.cell.2021.06.024

30. Wang EHC, Monga I, Sallee BN, et al. Primary cicatricial alopecias are characterized by
dysregulation of shared gene expression pathways. Brenner D, ed. PNAS Nexus. 2022;1(3).
doi:10.1093/pnasnexus/pgac111

31. Corrêa-Moreira D, dos Santos A, Menezes RC, Morgado FN, Borba CM, Oliveira-Ferreira J.
Skin Immune Response of Immunocompetent and Immunosuppressed C57BL/6 Mice After
Experimental Subcutaneous Infection Caused by Purpureocillium lilacinum. Frontiers in
Microbiology. 2021;12. doi:10.3389/fmicb.2021.615383

32. Žnidarič M, Žurga ŽM, Maver U. Design of In Vitro Hair Follicles for Different
Applications in the Treatment of Alopecia—A Review. Biomedicines. 2021;9(4):435.
doi:10.3390/biomedicines9040435

33. Harries MJ, Meyer K, Chaudhry I, et al. Lichen planopilaris is characterized by immune
privilege collapse of the hair follicle’s epithelial stem cell niche. The Journal of Pathology.
2013;231(2):236-247. doi:10.1002/path.4233

34. Liang YC, Wu P, Lin GW, et al. Folding Keratin Gene Clusters during Skin Regional
Specification. Developmental Cell. 2020;53(5):561-576.e9. doi:10.1016/j.devcel.2020.05.007

35. Ting-Berreth SA, Chuong CM. Local Delivery of TGF β2 Can Substitute for Placode
Epithelium to Induce Mesenchymal Condensation during Skin Appendage Morphogenesis.
Developmental Biology. 1996;179(2):347-359. doi:10.1006/dbio.1996.0266
44

36. Xie Y, Chen D, Jiang K, et al. Hair shaft miniaturization causes stem cell depletion through
mechanosensory signals mediated by a Piezo1-calcium-TNF-α axis. Cell Stem Cell.
2022;29(1):70-85.e6. doi:10.1016/j.stem.2021.09.009

37. Beronja S, Livshits G, Williams S, Fuchs E. Rapid functional dissection of genetic networks
via tissue-specific transduction and RNAi in mouse embryos. Nature Medicine. 2010;16(7):821-
827. doi:10.1038/nm.2167

38. Yu Z, Jiang K, Xu Z, et al. Hoxc-Dependent Mesenchymal Niche Heterogeneity Drives
Regional Hair Follicle Regeneration. Cell Stem Cell. 2018;23(4):487-500.e6.
doi:10.1016/j.stem.2018.07.016

Abstract (if available)

Abstract DNA methylation is an essential epigenetic modification that regulates the proliferation or differentiation of epidermal progenitor and bulge stem cells (BSCs). DNA methyltransferase 1 (Dnmt1) is the major enzyme for maintaining DNA methylation patterns after cell replication. Previously, Li et al. generated K14-Cre-Dnmt1-2lox (Dnmt1-cKO) mouse to investigate the role of Dnmt1 in hair cycling, and they concluded that Dnmt1 is crucial for regulating epidermal progenitors and hair follicle (HF) homeostasis. However, the molecular mechanism of this process remains unclear. Here, we utilize transcriptomic and accessible chromatin profiling using whole HFs and sorted BSCs from different aged Dnmt1-cKO and wild-type (WT) mice. We identify cKO-specific differentially expressed genes which also have enriched binding motifs on cKO-specific differentially accessible chromatin regions, including Klf2, Cebpd, Snai1, Bhlhe41, Pparg, and Lpin1. We also identify significantly enriched inflammation-related pathways in BSCs from Dnmt1-cKO mice. Moreover, we observe the upregulation of these candidate genes and inflammation in surrounding fibroblasts. Therefore, we hypothesize that intrinsic alterations of epidermal progenitors and BSCs induce the secondary changes as extrinsic factors that cause inflammatory cell infiltration and fibroblast heterogeneity resulting in progressive alopecia. The immunostaining of candidate genes and FACS-sorting of inflammation-related cells suggest that more macrophages infiltrate in Dnmt1-cKO mouse skin. Together, we suggest that losing Dnmt1 not only decreases stem cell numbers in epidermal progenitors and BSCs but also causes DNA hypomethylation and activation of the candidate genes. These genes trigger the secondary changes in the skin microenvironment, resulting in inflammation, fibroblast heterogeneity, and impairing hair cycling.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Implications of constitutively active bone morphogenetic protein (BMP) signaling in skin morphogenesis and skin postnatal homeostasis

PDF

DNA methylation in the mouse cochlea promotes maturation of supporting cells and contributes to the failure of hair cell regeneration

PDF

The role of Fgfr2 within Scx-expressing cells of the hair follicle

PDF

Transcriptional and epigenetic mechanisms underlying sensory hair cell differentiation and regeneration

PDF

The epigenetic landscapes underlying differentiation and plasticity in the developing organ of Corti

PDF

The organ of Corti sensory progenitor cell competence is controlled by SoxC transcription factors

PDF

Canonical and non-canonical Wnt signaling in the patterning of multipotent stem cells during feather development

PDF

Characterization of new stem/progenitor cells in skin appendages

PDF

Molecular aspects of skin early morphogenesis

PDF

The role of ERK1/2 in mouse embryonic stem cell fate control

PDF

Control of endogenous gene expression in the mouse

PDF

Role of FGFR2b signaling pathway in the development of ectodermal derivatives

PDF

Mechanisms that dictate beta cells’ response to stress in the context of genetic mutation, pregnancy, and infection

PDF

Homologous cell systems for the study of progression of androgen-dependent prostate cancer to castration-resistant prostate cancer

PDF

Development of an “in cubo” culture system for the study of early avian development using dynamic imaging

PDF

The role of Prkci in stem cell maintenance and cell polarity using a 3-D culture system

PDF

Transcriptomic maturation of developing human cone precursors in fetal and 3D hESC-derived tissues

PDF

Exploration of the roles of cancer stem cells and survivin in the pathogenesis and progression of prostate cancer

PDF

IPS and CNS cell models of peroxisomal disorders

PDF

Identification and characterization of adult stem cells in the oral cavity

Asset Metadata

Creator Jea, Wen-Chien (author)

Core Title Characterizing the role of Dnmt1 and its downstream factors in progressive alopecia

School Keck School of Medicine

Degree Master of Science

Degree Program Experimental and Molecular Pathology

Degree Conferral Date 2022-12

Publication Date 01/05/2024

Defense Date 12/14/2022

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

Tag bulge stem cells,DNMT1,hair follicles,inflammatory cell infiltration,inflammatory responses,OAI-PMH Harvest,progressive alopecia

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chuong, Cheng-Ming (committee chair), Carlson, Joseph (committee member), Liang, Ya-Chen (committee member), Widelitz, Randall (committee member)

Creator Email winniejea@gmail.com,wjea@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC112710938

Unique identifier UC112710938

Identifier etd-JeaWenChie-11401.pdf (filename)

Legacy Identifier etd-JeaWenChie-11401

Document Type Thesis

Format theses (aat)

Rights Jea, Wen-Chien

Internet Media Type application/pdf

Type texts

Source 20230111-usctheses-batch-1000 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu