University of Southern California Dissertations and Theses

Characterization of new stem/progenitor cells in skin appendages

(USC Thesis Other)

Characterization of new stem/progenitor cells in skin appendages

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CHARACTERIZATION OF NEW STEM/PROGENITOR CELLS
IN SKIN APPENDAGES

by

Yvonne Leung

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

August 2013

Copyright 2013 Yvonne Leung
ii
Table of Contents

Acknowledgements .......................................................................................................... iv

List of Tables .................................................................................................................... vi

List of Figures.................................................................................................................. vii

Abbreviations .....................................................................................................................x

Abstract........................................................................................................................... xiv

Chapter 1: Introduction
1.1 The Skin and its Appendages.............................................................................1
1.2 Sweat Gland Structure and Function .................................................................7
1.3 Sweat Gland in Wound Healing ......................................................................11
1.4 Sweat Gland Disorders ....................................................................................13
1.5 Nail Structure and Function.............................................................................15
1.6 Nail Differentiation..........................................................................................17
1.7 Nail Disorders ..................................................................................................19
1.8 Identifying Label Retaining Cells....................................................................22
1.9 LRCs and Hair Follicle Stem Cells..................................................................24
1.10 Identifying LRCs as Putative Stem Cells in Various Tissues........................28

Chapter 2: Label Retaining Cells (LRCs) with Myoepithelial Characteristic from
the Proximal Acinar Region Define Stem Cells in Sweat Glands.
2.1 Abstract............................................................................................................31
2.2 Introduction......................................................................................................33
2.3 Slow Cycling LRCs are Localized to the Proximal Acinar Region
of Sweat Glands ...............................................................................................37
2.4 Sweat Gland LRCs are Attached to the Basement Membrane and
Demonstrate Myoepithelial Characteristics.....................................................41
2.5 Sweat Gland LRCs Express Keratin 15 (K15) in the Acinar Region and
K15 Marked Cells Contribute Long-Term to the Sweat Gland Structure,
but not to Epidermal Homeostasis ...................................................................44
2.6 Isolating LRCs from Sweat Glands .................................................................48
2.7 Defining Sweat Gland LRCs Characteristics...................................................52
2.8 BMP Signaling is Required for Sweat Gland Morphogenesis.........................58
2.9 Acinar Sweat Gland Cells Do Not Contribute to the Epidermis
during Wound Healing.....................................................................................61
2.10 Sweat Gland LRCs can Trans-Differentiate into Epidermis under
Prolonged Isolated Wound Healing...............................................................64
2.11 Dissociated Sweat Gland Cells can Regenerate both Sweat Glands and
Hair Follicles..................................................................................................69
iii
2.12 Discussion......................................................................................................74

Chapter 3: Localization, Characterization and Isolation of Label-Retaining Cells
(LRCs) from Nails as Putative New Skin Stem Cells.
3.1 Abstract............................................................................................................84
3.2 Introduction......................................................................................................86
3.3 Identification of Nail LRCs in the Ventral Proximal Fold at the Root
of the Nail Plate ...............................................................................................92
3.4 Characterizing the Specific Localization and Organization of LRCs
in Nails.............................................................................................................95
3.5 Dissection and Isolation Strategy for H2B-GFP positive LRCs
from Nails ........................................................................................................99
3.6 Identifying the Transcriptional Profile of Nail LRCs via
Microarray Analysis.......................................................................................101
3.7 BMP Signaling is Required for Proper Nail Formation.................................104
3.8 Quiescent Nail LRCs can become Activated Upon Nail Removal ...............108
3.9 Assessment of Nail LRCs Differentiation Capabilities after
Engraftment in vivo.......................................................................................110
3.10 Discussion....................................................................................................112

Chapter 4: Comparison of Hair Follicle, Sweat Gland, and Nail Skin Appendages
4.1 Dermal-Epidermal Interactions are Important in the Development of
Skin Appendages ...........................................................................................119
4.2 LRCs in Skin Appendages.............................................................................121
4.3 Sweat Glands and Hair Follicles Contain Multiple Stem Cell Populations
and can Both Contribute to Epidermal Wound Healing ................................124
4.4 Hair Follicles and Nails are Morphologically Similar...................................127
4.5 BMP Signaling is Required for Hair Follicle, Sweat Gland, and
Nail Development ..........................................................................................129

Chapter 5: Concluding Remarks..................................................................................130

Chapter 6: Materials and Methods
6.1 Generation of Transgenic Mouse Lines.........................................................136
6.2 Immunohistochemistry and Immunofluorescence Staining...........................138
6.3 Isolation of Skin Stem Cells ..........................................................................142
6.4 Transplantations.............................................................................................145
6.5 Microscopy....................................................................................................147
6.6 RNA Isolation and qPCR...............................................................................149
6.7 Microarray Analysis.......................................................................................150

References.......................................................................................................................151

iv
Acknowledgements

First and foremost, I would like to thank my parents for their love and unconditional
support in everything I choose to pursue, even if they think it’s a bad idea. A special and
big thank you to my mentor Dr. Kris Kobielak for teaching me everything I know in stem
cell research, now we just need to work on the swimming and biking. Thank you for
making me do all those talks and poster presentations, now I actually have something to
put in my CV. As scary as it was, it felt great afterwards especially after the International
Investigative Dermatology (IID) meeting in Scotland, UK (2013). Thank you for always
having confidence in me, I could not have done it without your trust and encouragement!
Thank you for Scotland, it was a great trip to end my PhD career here! I would also like
to thank Dr. Agnes Kobielak for her help, ideas, advice, sharing of reagents and samples,
and always thinking of how we can improve or further my projects. Thank you Kris and
Agnes for putting up with my ups and downs throughout the years (I know I can get a bit
obnoxious sometimes), inviting me into your home, the countless birthday cakes, parties,
chocolates, BBQs, and of course the polish sausages! I love that we can hangout, drink
and have fun together, and talk about random things outside of work and science. I am
glad that we have more that just a work relationship and can actually have fun together
outside of the lab.

In addition, I would like to thank Dr. Cheng Ming Chuong for taking the time to be my
dissertation committee chair and always being so passionate and excited to hear about my
v
projects. Thank you for all the fruitful discussions we’ve had and all the positive
encouragement you’ve given me. I would also like to thank Dr. Gregor Adams for always
having his door open for me, listening to me, being straight with me, and giving me
advice even when he is busy. Thank you for inviting me to all the parties you host at your
house. I miss being lab neighbors and hearing all the stories you used to tell during lunch.

Finally, I would also like to thank all the current and past lab members especially Dr. Eve
Kandyba for the countless hours she has spent helping me troubleshoot my projects,
isolate cells, take care of my work when I am sick or out of town, and even pulling an all
nighter with me for my project. Thank you for listening and keeping me sane with all
those tea breaks. I am very glad to have met you here in LA. Thank you Justin Yu for
feeding me, taking care of me, and putting up with my crankiness when I am stressed,
especially these past few months when I was working on this dissertation. To my fellow
classmates, Ben (aka Freddie), Christine, Lauren, and Sarah, I am sad that we are going
our separate ways, but I am very happy to have shared this journey with you guys! We
did it! I hope to see everyone again in the future!

vi
List of Tables

Table 2.1 Categorization of the sweat gland gene expression profiles
based on function .......................................................................................55

Table 3.1 Categorization of the nail gene expression profiles based
on function ...............................................................................................103

Table 4.1 Comparison of hair follicle, sweat gland, and nail skin
appendages...............................................................................................122

Table 6.1 List of primary antibodies........................................................................140

Table 6.2 List of secondary antibodies ....................................................................141

Table 6.3 List of primers for qPCR..........................................................................149

vii
List of Figures

Figure 1.1 Structures of the haired and hairless skin ....................................................2

Figure 1.2 Hair follicle morphogenesis is regulated by dermal-epidermal
interactions...................................................................................................4

Figure 1.3 Cyclic regeneration of hair follicles.............................................................5

Figure 1.4 Structure and components of the sweat glands..........................................10

Figure 1.5 Structure of the human and mouse nail......................................................16

Figure 1.6 Pustular eruptions in hallopeau’s acrodermatitis and ulceration
in nail melanoma........................................................................................20

Figure 1.7 K5 driven, tetracycline controlled H2B-GFP label retaining system ........26

Figure 2.1 Sweat gland LRCs are localized in the acinar gland region of SGs ..........38

Figure 2.2 Sweat gland LRCs possess slow cell cycle dynamics but are non
post-mitotic cells........................................................................................39

Figure 2.3 Dynamics of sweat gland LRCs.................................................................40

Figure 2.4 Sweat gland LRCs are attached to the basement membrane and possess
myoepithelial characteristics......................................................................42

Figure 2.5 Wholemount staining with 3D reconstruction confirms that sweat gland
LRCs are attached to the basement membrane..........................................43

Figure 2.6 Sweat gland LRCs express Keratin 15.......................................................45

Figure 2.7 K15-GFP reporter marks both the basal and luminal layers......................46

Figure 2.8 Keratin 15 marked sweat gland cells contribute long-term to the
acinar SG structure.....................................................................................47

Figure 2.9 Isolation strategy for sweat gland LRCs....................................................48

Figure 2.10 FACS isolation of SG LRCs and non-LRCs..............................................49

Figure 2.11 Culturing of H2B-GFP sweat gland LRCs ................................................50

viii
Figure 2.12 FACS isolation and cell culture of K15-GFP sweat gland cells................51

Figure 2.13 Determining the molecular characteristics of sweat gland LRCs..............53

Figure 2.14 Validation of genes identified in the microarray analysis by qPCR..........54

Figure 2.15 Validation of genes expressed and identification of active BMP
signaling in sweat glands ...........................................................................57

Figure 2.16 Sweat gland appendages develop from K14 marked progenitor cells.......58

Figure 2.17 BMP signaling is required for sweat gland formation...............................59

Figure 2.18 Unlike hair follicles, ID2 is not required for sweat gland development....60

Figure 2.19 Acinar sweat gland cells do not contribute to the epidermis during
typical wound healing................................................................................63

Figure 2.20 Sweat ducts are proliferative under normal homeostasis...........................63

Figure 2.21 Sweat gland cells can trans-differentiate into the epidermis under
prolonged isolated wound healing conditions............................................66

Figure 2.22 Sweat gland LRCs can trans-differentiate into the epidermis under
prolonged isolated wound healing conditions............................................68

Figure 2.23 Kidney capsule transplantation of dissociated unsorted 4 week chased
SG LRCs ....................................................................................................69

Figure 2.24 Dissociated sweat gland cells can regenerate sweat glands, hair follicles,
and the epidermis .......................................................................................71

Figure 2.25 Subcutaneously transplanted dissociated sweat gland cells can regenerate
sweat glands, hair follicles, and the epidermis .........................................73

Figure 3.1 Localization of nail LRCs in the ventral proximal fold.............................94

Figure 3.2 Nail LRCs are localized in the basal layer at the granular layer border ....96

Figure 3.3 Nail LRCs express K15 and K15 marked cells contribute long-term
to the nail....................................................................................................97

Figure 3.4 The nail matrix contains actively proliferating cells, not LRCs ................98

ix
Figure 3.5 Isolation of live nail LRCs.........................................................................99

Figure 3.6 Purifying live nail LRCs for gene expression profiling...........................102

Figure 3.7 BMP signaling is required for proper nail development..........................105

Figure 3.8 The nail plate differentiation is compromised in Bmpr1a KO mice........106

Figure 3.9 K1 and Loricrin epidermal markers are ectopically expressed in nail
in the absence of BMP signaling .............................................................107

Figure 3.10 Nail LRCs are not lost upon nail removal ...............................................108

Figure 3.11 Normally quiescent nail LRCs can be activated upon injury ..................109

Figure 3.12 Transplanted nail LRCs can contribute to the nail structure....................111

Figure 4.1 Alkaline phosphatase activity is found in the dermis of sweat gland
and nail skin appendages .........................................................................120

Figure 4.2 Different stem cell populations in hair follicles and sweat glands ..........125

Figure 4.3 Sca-1 is expressed in sweat ducts and not the acinar region....................126

Figure 4.4 Comparison of hair follicle and nail morphologies .................................127

Figure 5.1 Schematic of skin appendages and its LRCs localization........................133
x
Abbreviations

3D 3-dimensional
3
H tritiated
Ab antibody
AP alkaline phosphatase
BMP bone morphogenetic protein
Bmpr1a bone morphogenetic protein receptor 1a
BrdU bromodeoxyuridine
BSA bovine serum albumin
Ctr control
DAPI 4’,6-diamidino-2-phenylindole
DAVID Database for Annotation, Visualization and Integrated Discovery
DEG differentially expressed genes
DMF dimethylformamide
doxy doxycycline
DP dermal papilla
du duct
eYFP enhanced yellow fluorescent protein
ECM extracellular matrix
ED ectodermal dysplasia
EDs ectodermal dysplasias
EDTA ethylene diamine tetraacetic acid
xi
FACS fluorescence activated cell sorting
FITC fluorescein isothiocyanate
GFP green fluorescent protein
H&E hematoxylin and eosin
H2B histone 2B
H2B-GFP histone 2B conjugated green fluorescent protein
Het heterozygous
HF hair follicle
IP intraperitoneally
IRS inner root sheath
K1 keratin 1
K14 keratin 14
K15 keratin 15
K5 keratin 5
KO knockout
KZ keratogenous zone
LacZ β-galactosidase
LOF loss of function
LRCs label retaining cells
MEF mouse embryonic fibroblasts
MSG mouse salivary gland
Mx matrix
xii
N/A not applicable
Na
+
/K
+
sodium potassium
NPS nail-patella syndrome
ORS outer root sheath
PAS periodic acid schiff
PBS phosphate buffered saline
PC pachyonychia congenita
PCR polymerase chain reaction
PE phycoerythrin
Phospho, P phosphorylated
PR progesterone receptor
qPCR quantitative polymerase chain reaction
R26 rosa26
RU RU486, mifepristone
Sca1 stem cell antigen 1
SG sweat gland
SGs sweat glands
SMA smooth muscle actin
TA transit amplifying
tdTomato tandem dimer tomato
TNS tooth and nail syndrome
TPA tetradecanoylphorbol acetate
xiii
TRITC tetramethyl rhodamine isothiocyanate
X-gal 5-bromo-4-chloro-indoyl- β-D-galactopyranoside
YFP yellow fluorescent protein

xiv
Abstract

Stem cells have long term self-renewal potential and are capable of differentiating into a
variety of different cell types. This persistence and multipotency is crucial for the
maintenance of tissue during homeostasis and repair. Due to its extraordinary
regenerative potential, understanding its maintenance and regulation will prove its
usefulness in tissue regeneration and treating various disorders. Thus, scientists have
been intensely searching for stem cells in various organs in hopes to understand and take
advantage of their regenerative abilities. Among the different organs, the skin is the most
easily accessible making it a good model to study stem cells and wound healing.
Moreover, the skin is a complex organ containing a number of different “mini-organs”,
skin appendages that are likely maintained by independent stem cell populations.
Understanding the similarities and differences between these stem cell populations may
not only reveal how they maintain and regulate independent appendages, but also how
stem cells maintain their own regenerative potential.

The skin contains a number of different skin appendages including hair follicles, sweat
glands, and nails. Among these appendages, hair follicles and their stem cells have been
the most well characterized while relatively little is known about the presence of stem
cells in the sweat gland and nail appendages. Since stem cells must persist throughout
life, they have been proposed to be slow cycling cells in order to preserve their self-
renewing potential and minimize DNA replication errors during cell division. Decades
xv
ago, a population of slow cycling label retaining cells (LRCs) was identified in the hair
follicle bulge. Numerous subsequent studies have collectively shown that these bulge
LRCs are the hair follicle stem cells required for hair follicle regeneration. Moreover,
these hair follicle stem cells have been shown to participate in epidermal wound healing
during injury. Given that sweat glands are also abundantly distributed in the skin, there is
great interest and speculation on whether sweat gland cells can also contribute to wound
healing.

Using a K5 driven tetracycline regulated H2B-GFP transgenic system, we have identified
distinct LRCs in the nails and sweat glands as putative new skin stem cells. More
specifically, LRCs were localized to the ventral proximal fold of the nail and the acinar
secretory region of sweat glands. Both nail and sweat gland LRCs were found in the basal
layer expressing K15, a known hair follicle stem cells marker. Lineage tracing
experiments demonstrate that the K15 cells can contribute long-term to their respective
appendages, suggesting long-term self-renewal capabilities found in stem cells. Isolation
of live nail and sweat gland LRCs allowed for gene expression profiling to reveal their
molecular characteristics, where BMP pathway signaling genes were identified in both
sweat gland and nail appendages. Furthermore, we demonstrate the functional
requirement of Bmpr1a-mediate signaling in both appendages. Transplanting strips of
nail LRCs showed a contribution of H2B-GFP cells to the nail structure. Similar to hair
follicle stem cells, sweat gland LRCs can differentiate and contribute to the epidermis
xvi
under prolonged wound healing conditions. In addition, transplantation of sweat gland
cells suggests its plasticity in regenerating sweat glands as well as hair follicles.

In conclusion, we have identified slow cycling LRCs in the sweat gland and nail skin
appendages that showed stem cell characteristics. Isolation and characterization of these
LRCs show that they share similarities to hair follicle stem cells, but are also unique in
possessing their own features. Although more studies are needed for further
characterization of these LRCs, these data may shed some light on putative stem cells
markers for human nails and sweat glands in the future.

1
Chapter 1:
Introduction

1.1 The Skin and its Appendages
Skin is the largest organ of the human body serving as a protective barrier against
environmental insults. It consists of an epidermis, underlying dermis, and associated skin
appendages: hair follicles, sebaceous glands, sweat glands, and nails. The components of
the skin vary at different regions of the body. Sweat glands, for example, are abundantly
found all over the body whereas nails are restricted to the digit tips. Hair follicles and
sebaceous glands are predominantly found together forming the pilosebaceous unit.
However, sebaceous glands can also exist independent of hair follicles opening directly
onto the surface of the skin at the lips, corner of the mouth, glans penis, labia minora, and
mammary nipple. Although hair follicles are also abundantly found in most areas, there
are regions of the skin that contain only sweat glands and no hairs such as the palms and
soles. This hairless epidermis is often thicker and is called palmar or planter epidermis
while the epidermis containing hair is referred to as interfollicular epidermis (Figure 1.1).
2

Figure 1.1. Structures of the haired and hairless skin. The palms and soles consist of
thick and hairless skin containing only sweat gland appendages (left). Most other parts of
the body are covered in haired skin containing hair follicle, sebaceous gland, and sweat
gland skin appendages (right). - Wikipedia

The epidermis of the skin is composed of several layers forming a stratified epithelium
with basal, spinous, granular, and cornified layers all expressing distinct differentiation
markers. Stem and progenitor cells responsible for epidermal homeostasis reside in the
basal layer and its progeny becomes progressively differentiated as they migrate through
the epidermal layers. When these cells reach the outermost cornified layer (stratum
corneum), they become terminally differentiated and sloughed off (Blanpain and Fuchs,
2006).

3
The epidermis and its associated appendages are all formed from the ectoderm during
embryogenesis where reciprocating ectodermal-mesenchymal interactions determine the
fate of the appendages (Dhouailly, 1973; Hamrick, 2001). Induction of hair follicle
formation is dependent on crosstalk between signals emitted by the epithelium and the
underlying dermis (Figure 1.2). Although the initial signals are still unknown, hair
follicle morphogenesis begins with a thickening of the epidermis called the placode.
Next, epithelial signals are believed to orchestrate the condensation of underlying dermal
cells to form the dermal papilla (DP). Similar to its role in activating hair growth, the DP
promotes downgrowth and differentiation of epithelial cells to form the hair follicle skin
appendage. Such intricate crosstalk between the dermis and epidermis is required for both
proper hair follicle formation and regeneration (Millar, 2002). Similar epithelial-dermal
interactions are likely required for sweat gland and nail morphogenesis.
4

Figure 1.2. Hair follicle morphogenesis is regulated by dermal-epidermal
interactions. Hair follicle development begins with placode formation, which is induced
by dermal signals. The epithelial placode subsequently promotes dermal condensation
and dermal papilla formation important for hair follicle development and regeneration.
Adapted from (Millar, 2002).

Among the different skin appendages, hair follicles have been the best characterized. The
hair follicle undergoes cycles of growth (anagen), degeneration (catagen), and quiescence
(telogen), which are controlled by its crosstalk with the DP signaling center in the dermis
(Figure 1.3) (Alonso and Fuchs, 2006; Millar, 2002; Rendl et al., 2008; Schmidt-Ullrich
and Paus, 2005). This cyclic activation is made possible by the presence of hair follicle
stem cells in the bulge region. At the onset of anagen, progenitor cells of the hair germ,
the structure repopulated by proliferating bulge stem cells localized between the bulge
and DP, proliferate and differentiate into transit amplifying (TA) cells of the matrix
(Greco et al., 2009). These matrix cells then proceed to further differentiate into the
5
different layers of the hair follicle including the outer root sheath (ORS), inner root
sheath (IRS), and the cuticle, cortex, and medulla of the hair shaft. Under normal
homeostasis, the epidermis and hair follicles are maintained by independent stem cells
(Ito et al., 2005).

Figure 1.3. Cyclic regeneration of hair follicles. Hair follicle skin appendages undergo
cycles of growth (anagen), degeneration (catagen), and rest (telogen). In the anagen
phase, proliferative matrix cells at the base of the hair follicle differentiate into the
different layers of the hair follicle. Hair follicle stem cells are located in the bulge region,
which is the permanent portion of the hair follicle. Adapted from (Millar, 2002).

Attached to the hair follicles are sebaceous glands. Sebaceous glands secrete an oily
substance called sebum through the hair canal onto the skin’s surface for lubrication and
waterproofing. Sebum secreted from these glands is also responsible for acne when
infected by bacteria. Although sebaceous glands are part of the pilosebaceous unit, they
are maintained by Blimp1+ unipotent progenitor cells distinct from hair follicle stem cells
(Horsley et al., 2006). More recently, lineage tracing analysis with a new K15CreERT2
6
system demonstrated that hair follicle bulge cells contribute to sebaceous gland renewal
during skin homeostasis (Petersson et al., 2011). In addition, recent studies have also
identified a Lgr6+ stem cell population capable of maintaining the hair follicle
infundibulum and interfollicular epidermis in addition to the sebaceous glands (Snippert
et al., 2010). However, much less is known about the sweat gland and nail skin
appendages, which will be discussed in further detail below.
7
1.2 Sweat Gland Structure and Function

In humans, there are three different types of sweat glands: eccrine, apocrine and
apoeccrine. Eccrine glands are abundantly found all over the skin secreting mainly water
and electrolytes through the surface of the skin vital for thermoregulation. Apocrine
glands on the other hand are only found in skin containing hair since they secrete their
oily substances containing lipids, proteins, and steroids through the hair canals (Labows
et al., 1979; Wilke et al., 2007). Apocrine glands are larger than eccrine glands and are
restricted to the armpits, mammary, anal, and genital areas (Wilke et al., 2007). Rather
than responding to temperature, they often respond to emotional stimuli. For decades, it
was believed that there are only two types of sweat glands. In 1987, however, Sato et al.
have reported the existence of apoeccrine glands that was found in areas of apocrine
glands and secreted watery fluids similar to eccrine glands (Sato et al., 1987). More work
is required for elucidating the mechanisms of these apoeccrine glands.

In contrast to humans, animals such as dogs and mice have eccrine sweat glands only in
their paws. The majority of their skin lack sweat glands as they have evolved a different
method of thermoregulation, namely panting. However, sweat glands are present in the
paws as a mean to provide friction for running and climbing. In our research, we have
focused our studies to eccrine sweat glands only and for the remainder of this
dissertation, the use of “sweat glands” will refer to eccrine sweat glands.

8
Sweat glands are coiled tubular appendages covering almost the entire human skin.
Approximately 1.6 to 5 million sweat glands is found in the skin where its principle
function is thermoregulation (Sato et al., 1989a, b). The amount of sweat glands varies
between people as well as anatomical sites. The region with greatest sweat gland density
is believed to be the palms and soles containing 600-700 sweat glands per square
centimeter (Sato et al., 1989a). The main function of sweat glands is to keep the core
body temperature at a cool ambient temperature of approximately 37°C in a hot
environment or during physical activity (Sato et al., 1989a; Wilke et al., 2007). Sweat
glands are innervated by neurons so the process of sweating is controlled by the central
nervous system. It has been reported that thermosensitive neurons in the brain can detect
the internal body temperature as well as external skin temperature and respond
accordingly to keep the core body temperature constant (Sato et al., 1989a; Wilke et al.,
2007). When temperatures are increased, our bodies are induced to sweat in order to cool
down. The skin and internal body temperature decreases when the water in our sweat
evaporates at the skin’s surface. Sweat glands are therefore vital in keeping our
temperature constant. A core body temperature of 40 °C and up can lead to protein
denaturation and apoptosis at the molecular level (Wilke et al., 2007). Physically, it can
lead to hyperthermia, also known as heat exhaustion or heat stroke, which can be fatal. In
addition to temperature, sweat glands can also respond to emotional stimulus including
anxiety and fear. Under these circumstances, sweating is often observed in the armpits,
palms, and soles (Allen et al., 1973; Chalmers and Keele, 1952; Eisenach et al., 2005).

9
Sweat is a dilute electrolyte solution containing 99% water and mostly sodium chloride,
potassium, and bicarbonate. In addition, it also contains calcium, magnesium, lactate,
ammonia, and urea (Shirreffs and Maughan, 1997). Upon sweating, some of the ions are
reabsorbed as it passes through the sweat duct. This is made possible by sodium
potassium (Na
+
/K
+
) ATPases on the membrane (Sato et al., 1989a; Wilke et al., 2007). In
addition to the Na
+
/K
+
pumps, chloride channels are also found.

Sweat glands are composed of a coiled acinar secretory structure in the dermis and a
straight duct connecting the acinar structure to the surface of the epidermis. The secretory
coil consists of a basal layer composed of 2 distinct cell types, clear cells and
myoepithelial cells, in addition to a luminal layer composed of dark cells (Figure 1.4)
(Kierszenbaum, 2007). Dark cells of the luminal layer secrete glycoproteins. They
contain granules that can be identified with Periodic Acid Schiff (PAS) staining, which
labels glycoproteins. Clear cells are rich in mitochondria and contain basolateral
infoldings where water and ions are secreted. This sweat subsequently travels through
small intercellular canals where it reaches the lumen, travels through the sweat duct, and
becomes secreted at the surface of the skin (Kierszenbaum, 2007). Myoepithelial cells are
located at the periphery of sweat glands and are believed to provide support for the sweat
gland structure. Recently, Lu et al. have reported the presence of unipotent stem cells in
adult sweat glands located in the basal and luminal layers (Lu et al., 2012).

10

Figure 1.4. Structure and components of sweat glands.
Sweat glands are coiled tubular structures connected to the epidermis through a straight
excretory duct. The coiled secretory portion is composed of 3 distinct cell types: clear
and myoepithelial cells of the basal layer and dark cells of the luminal layer. Adapted
from (Kierszenbaum, 2007).

11
1.3 Sweat Glands in Wound Healing
Although hair follicle stem cells do not participate in epidermal homeostasis, it can
migrate up to the epidermis and contribute to wound healing in the event of an injury.
This has been shown using a K15CrePR driven R26-LacZ reporter lineage tracing system
(Ito et al., 2005). The human skin contains both hair follicles and sweat glands in most
regions of the body; however, there are areas of the skin that do not have hair and only
contain sweat glands such as the palms and soles. Therefore, there has been much interest
and speculation regarding the ability of sweat glands in participating in epidermal wound
healing, especially in areas where hair follicles are absent. To test whether sweat glands
have the potential to contribute to the epidermis, Miller et al. have generated wounds on
porcine skin since its structure is very similar to the human skin. Deep wounds were
generated in order to remove the hair follicles along with the epidermis and since sweat
glands reside deeper in the dermis, occasional sweat glands were left. Although shallow
wounds containing hair follicles healed faster, the deep wounds have also re-
epithelialized in the absence of hair follicles. This newly formed skin was distinctly
different from the surrounding skin and did not contain any hair follicles (Miller et al.,
1998). Presumably, sweat gland cells have migrated up to re-epithelialize the epidermis
in the absence of hair follicles. In an independent study, human sweat gland cells were
isolated to generate a skin graft in culture and transplanted onto the backs of
immunocompromised rat where they formed a fully stratified epidermis (Biedermann et
al., 2010). More recently, the contribution of sweat gland cells in wound healing has also
been reported in the human skin (Rittie et al., 2013). These reports suggest that sweat
12
gland cells also have the potential to differentiate into cells of the epidermis and
contribute to wound healing after injury. However, it remains unclear which sweat gland
cells have this potential and whether they are the stem cells that maintain the normal
homeostasis of this appendage.

13
1.4 Sweat Gland Disorders
Sweat gland disorders range from excessive sweating (hyperhidrosis), decreased sweating
(hypohidrosis), to no sweating (anhidrosis). Hyperhidrosis is generally not a serious
condition, but anhidrosis can lead to death from hyperthermia. Patients with hypohidrosis
or anhidrosis often have symptoms of heat intolerance that may lead to fatigue, weakness,
dizziness, and difficulty in breathing. Hyperhidrosis is a social problem leading to
embarrassment, which most commonly affects the armpits, palms, and soles (Chan et al.,
1985). Depending on its severity, it can be treated with topical aluminum salts or
anticholinergic oral medications (Wenzel and Horn, 1998).

Hypohidrosis and anhidrosis are most commonly caused by obstruction of the sweat
pores and ducts as seen in patients with psoriasis, dermatitis, sclerosis, and miliaria.
Some patients with miliaria (also known as a sweat rash) feel a stinging sensation in the
affected areas due to sweat retention from the ductal occlusion (Wenzel and Horn, 1998).
In some cases, controlling the temperature and humidity of the environment to reduce
sweating can relieve the obstruction in a few days to weeks. Hypohidrosis and anhidrosis
can also be caused by dysfunctional sweat glands, as in Fabry’s disease of systemic
sclerosis, or absent sweat glands in anhidrotic ectodermal dysplasia (Cheshire and
Freeman, 2003; Wenzel and Horn, 1998). Anhidrotic ectodermal dysplasia is a
dermatological disorder that also affects other skin appendages in addition to sweat
glands (Kere et al., 1996). It is caused by a mutation in the ED1-gene encoding
ectodysplasin-A (EDA) ligand, its EDAR receptor, or EDARDD adaptor protein and can
14
be life threatening for children due to their inability to sweat (Cluzeau et al., 2011;
Vincent et al., 2001; Wisniewski et al., 2002). Hypohidrosis can also be a result of
injuries caused by burns, irradiation, and trauma that damage sweat glands. In general, all
described conditions vary in severity and can either be localized to a specific region of
the body or more globally affect a patient.

Many times, hyperhidrosis, hypohidrosis, or anhidrosis can be associated with more
serious underlying diseases and processes. For example, people undergoing anxiety,
menopause, or drug withdrawal often experience excessive sweating. Since sweat glands
are innervated by neurons and are controlled by the central nervous system, diseases
affecting the central nervous system (such as Parkinson’s disease) or spinal cord often
cause abnormal sweating (Sato et al., 1989b; Turkka and Myllyla, 1987).

15
1.5 Nail Structure and Function
The nail skin appendage is composed of a hard and keratinized structure that serves as a
protective covering by preventing trauma to the toes and fingers. Additionally, nails are
used as a tool to pick up small objects and are important for fine manipulations and subtle
finger functions. Within the nail unit, the nail plate is the hard keratinized structure
composed of flattened and anucleated cells overlying the distal phalange. The nail plate
exerts counter pressure at the fingertips for protection and enhances sensitivity. This nail
plate is attached to the finger by interdigiting with the underlying nail bed. Over the
proximal end of the nail plate, we have the proximal nail fold which is a continuation of
the epidermis. The cuticle extends from the edge of the proximal fold onto the nail plate
sealing the proximal end of the nail and protecting it from toxins and foreign substances.
Similarly, at the distal end of the nail, the hyponychium seals the nail plate to the nail bed
to prevent infection. As a continuation of the ventral proximal fold wrapping around the
proximal end of the nail plate, we have the nail matrix composed of actively proliferating
cells (de Berker and Baran, 2012). Above the matrix lies the keratogenous zone where
matrix cells are believed to differentiate, flatten out, die, and deposit into the overlying
nail plate. Distal to the matrix is the nail bed, which is only 2-3 cell layers thick.
Comparing the human and mouse nail, it is apparent that the shape is very different likely
owing to its difference in evolution and function. The human nail is more flat whereas the
claw is a near conical structure with greater curvature (Figure 1.5). This shape is
influenced by the underlying distal phalanx and the distribution of the matrix (Dawber et
al., 2001; de Berker and Baran, 2012).
16

Figure 1.5. Structure of the human and mouse nail. (A) Schematic of the human nail.
Adapted from (de Berker and Baran, 2012). (B) Section through a mouse nail.

17
1.6 Nail Differentiation
So far, it is believed that the matrix is largely responsible for nail plate production.
However, it remains unclear whether the matrix is the sole source of nail differentiation
or if other parts of the nail unit, such as the nail bed, also contribute to the nail plate.
Previously, Lewis observed that the nail plate is composed to 3 different layers and have
proposed that they arise from 3 different parts of the nail unit. It was theorized that the
superficial layer of the dorsal nail originates from the proximal nail fold, the intermediate
nail arises from the matrix, and the deep ventral nail is produced by the nail bed (Lewis,
1954). However, the 3 separate nail layers could not be demonstrated in all specimens
and lineage tracing experiments were still needed. About a decade later, Zaias and
Alvarez used radioactive tritiated glycine to mark and follow nail cells in the squirrel
monkey demonstrating that the uptake in label have moved from the matrix into the nail
plate over time (Zaias and Alvarez, 1968). Although, the proximal nail fold had also
incorporated the labeled glycine, it was claimed to have only contributed to the stratum
corneum of the proximal fold and not the nail plate. Only small amounts of glycine were
incorporated in the nail bed and were therefore dismissed as a source of nail production
due to its inactivity. Taken together, the nail matrix was proposed to be solely responsible
for nail plate formation (Zaias and Alvarez, 1968). Similar studies were also conducted
on human volunteers using both tritiated thymidine and glycine where matrix cells were
proposed to migrate into the nail bed in addition to the nail plate (Norton, 1971). This
hypothesis that the nail is produced by the matrix only was further supported by Berker
and Angus who demonstrated that matrix cells were highly proliferative when compared
18
to the relatively inactive nail bed (de Berker and Angus, 1996). In other studies, it has
been proposed that although the matrix produces the bulk of the nail plate, the nail bed
also contributes to the nail plate based on nail thickness and mass (Johnson et al., 1991;
Johnson and Shuster, 1993). It is argued that the contribution of nail bed cells to the
ventral nail plate allows distal movement of the nail plate as it grows. Although there
appears to be a general consensus that the nail plate is produced by matrix cells, it
remains unclear whether these matrix cells are the only source of progenitor cells
required for nail production or if the nail bed and proximal fold also contributes to nail
differentiation. Understanding the mechanisms and origin of nail differentiation may
prove useful in treating nail disorders in the future.

19
1.7 Nail Disorders
Nail abnormalities are often associated with defects in the nail matrix or nail bed that
either arises from inherited disorders, acquired diseases, or environmental causes such as
infection or trauma. There is a large variety of nail abnormalities ranging from
hyperplasia, hypoplasia, clubbing, splitting, discoloration, inflammation, to detachment
of the nails (Tosti and Piraccini, 2003). The characteristics of the nail defects are often
helpful in clinically diagnosing the associated disorder.

Although some nail disorders may affect nails alone, many also affect other tissues and
skin appendages. Examples of disorders affecting only nails include inherited anonychia
and isolated congenital nail dysplasia. Affected families with inherited anonychia have
severe hypoplasia of the nails where mutations in R-spondin 4 (Rspo4), which is
implicated in the Wnt signaling pathway, was identified suggesting an important role of
Rspo4 in nail development (Blaydon et al., 2006). Isolated nail dysplasia display nail
plate thinning with longitudinal streaks (Hamm et al., 2000). Although this appears to be
more of a cosmetic problem, some nail diseases are associated with more severe
problems. For example, 40% of patients with nail patella syndrome, which is
characterized by Lmx1b mutations leading to absent or hypoplastic nails, develop kidney
disease that can potentially lead to kidney failure (Chen et al., 1998). Patients with
pachyonychia congenita have mutations in keratin 6a, 6b, 16, or 17 causing thickening of
the nail making it extremely difficult to trim (Liao et al., 2007; McLean et al., 1995;
Terrinoni et al., 2001). Often time, abnormal nails are also associated with defects in
20
other ectodermal appendages including hair follicles, sweat glands, and teeth as in
ectodermal dysplasia, psoriasis, and other dermatological diseases. For example, patients
with Witkop syndrome, a form of ectodermal dysplasia, have nail dysplasia and missing
teeth caused by mutations in the Msx-1 gene (Jumlongras et al., 2001). Although many
nail defects are asymptomatic causing only cosmetic problems, some conditions are more
severe and can lead to severe pain or even death. Some patients with psoriasis may
experience painful pustular inflammation under the nail plate, as seen in Hallopeau’s
acrodermatitis (Figure 1.6) (Tosti and Piraccini, 2003). Melanoma of the nail is rare, but
can be fatal with a 5-year survival rate of only 15% (Figure 1.6) (Banfield et al., 1998).

Figure 1.6. Pustular eruptions in hallopeau’s acrodermatitis and ulceration in nail
melanoma. Patients with pustular or palmoplantar psoriasis may experience painful
pustular eruptions under the nail plate, as seen in Hallopeau’s acrodermatitis (left).
Patient with nail melanoma (right). Adapted from (Tosti and Piraccini, 2003).

Similar to sweat glands, nail defects can also be symptoms of more serious systematic
diseases. For example, acquired nail clubbing is often associated with pulmonary disease
and yellow nail syndrome is associated with respiratory problems (Tosti and Piraccini,
21
2003). Leukonychia, white discoloration of the nail, is a common sign of liver cirrhosis,
whereas autoimmune connective tissue disorders display abnormal capillaries is the
proximal nail fold (Cutolo et al., 2007). Taken together, nail disorders themselves are
rarely fatal, but can often cause discomfort and pain. Moreover, nail abnormalities are
often associated with more severe conditions and can be used as a tool for diagnosis.

22
1.8 Identifying Label Retaining Cells
Adult stem cells found in various organs throughout the body are responsible for
maintaining the normal turnover of organs as well as tissue repair in the event of an
injury. These are made possible by its ability to self-renew and differentiate into the
different specialized cell types in its respective organs. Since stem cells are required for
tissue homeostasis and regeneration throughout life, they must have long-term self-
renewal capacity. Moreover, it has been proposed that stem cells divide infrequently to
avoid incorporation of mutations associated with cell division. In each organ, stem cells
reside in specialized microenvironments called niches, which helps maintain its
quiescence until it is needed. Upon activation, these stem cells divide and leave the niche
to form more differentiated transit amplifying (TA) progenitor cells. In contrast to the
relatively quiescent stem cells, TA cells rapidly proliferate and differentiate into cells
needed for regeneration or repair. This has led scientists to believe that stem cells are
slow cycling cells.

Adult stem cells are theoretically present in all regenerative tissues, but its localization
and characterization is often difficult when little is known about those organs. Therefore,
many scientists have used the slow cycling property to label and identify putative stem
cells in various organs. In early pulse-chase studies, animals were injected with tritiated
thymidine (
3
H-thymidine), which gets incorporated into the newly synthesized DNA of
dividing cells. After efficient labeling, the animal undergoes a period of chase where
dividing cells dilute out the
3
H-thymidine label and slow cycling cells retain this
23
radioactive label. This method allowed for the identification of slow cycling label
retaining cells (LRCs); however, researchers were unable to use this to probe for co-
localization of potential markers through immunological histology. Soon after, scientists
began to use a new synthetic thymidine analog: bromodeoxyuridine (BrdU). In contrast
to
3
H-thymidine, BrdU is not radioactive. In addition, antibodies against BrdU allows for
immunohistochemistry to probe for markers that co-localizes with BrdU marked LRCs.
Pulse-chase studies with BrdU have been widely used to identify LRCs in numerous
tissues. More recently, a new method for identifying LRCs using transgenic mice with
tetracycline induced histone 2B-green fluorescent protein (H2B-GFP) has been described
(Tumbar et al., 2004). With this system, a specific or ubiquitous promoter can be used to
drive inducible H2B-GFP expression in the tissues of interest. One advantage of this
system is that it can more efficiently label the cells prior to chase while the BrdU method
can only label cells that are actively dividing in the synthesis phase during pulse.
Moreover, this H2B-GFP transgenic label retaining system allows for the isolation of live
LRCs for further characterization and analysis.

24
1.9 LRCs and Hair Follicle Stem Cells
In many different species, the skin is abundantly covered in hair follicle skin appendages.
Decades ago, hair follicle matrix cells were believed to be hair follicle stem cells due to
its multipotent nature. However, since stem cells are supposed to persist throughout the
lifetime of an animal, it was hypothesized to divide infrequently in order to conserve its
proliferative potential and minimize accumulation of replication errors during cell
division. In 1990, Cotsarelis et al. have identified a unique population of slow cycling
LRCs in the bulge region of the hair follicle using
3
H-thymidine pulse chase experiments.
Moreover, he had demonstrated that these bulge LRCs can be stimulated to proliferate
upon treatment with a tetradecanoylphorbol acetate (TPA) tumor promoter and went on
to hypothesize that these LRCs are stem cells of the hair follicle (Cotsarelis et al., 1990).
Several years later, Morris and Potten have reported that these LRCs in the bulge can
persist for over 1 year and can be activated after hair plucking (Morris and Potten, 1999).
This data supports the theory that stem cells persist throughout the lifetime of an animal
to contribute to the tissue during regeneration and wound repair. In 2001, Oshima et al.
have shown that transplanted LacZ labeled vibrissae bulge cells are multipotent and can
contribute to all the different lineages of the hair follicle further supporting the hypothesis
that bulge LRCs are stem cells of the hair follicle (Oshima et al., 2001).

Aside from its label retaining characteristic, no markers have been identified for these
bulge LRCs for years. In 1998, however, hair follicle bulge cells were reported to
specifically express keratin 15 (K15) in the human scalp (Lyle et al., 1998, 1999). In
25
2003, Liu et al. demonstrated that bulge cells in mice also expressed K15. Consequently,
they have generated a K15 driven LacZ reporter mouse to target these putative hair
follicle stem cells in the bulge where 92% of the targeted cells were LRCs (Liu et al.,
2003). At the same time, Trempus et al. have identified another marker, CD34, which
specifically co-localizes with these K15-expressing LRCs in the bulge. This CD34
marker was used in conjunction with α6-integrin to isolate live keratinocytes from the
bulge (Trempus et al., 2003). Finally in 2004, Morris et al. used an inducible K15 driven
CrePR recombinase crossed onto a ubiquitous LacZ reporter to mark and trace K15
expressing cells in vivo (Morris et al., 2004). In this lineage tracing experiment, they have
shown that K15 marked bulge cells are mutipotent and can differentiate into all cell
lineages of the hair follicle at the onset of anagen during normal hair follicle
regeneration. Moreover, isolated K15-GFP bulge cells display higher proliferative
potential in vitro and can regenerate the entire skin including sebaceous glands and the
epidermis (Morris et al., 2004). Taken together, this shows that these K15-expressing
cells, enriched in LRCs, are hair follicle stem cells that regenerate this appendage during
normal homeostasis and wound repair. Isolation of K15-GFP cells have also allowed for
gene expression profiling of hair follicle stem cells to identify candidate genes
responsible for maintaining these stem cells in their undifferentiated state.

In 2004, a new method of identifying and isolating LRCs in the skin using a keratin 5
(K5) driven tetracycline repressor-VP16 fusion transgene crossed onto tetracycline
controlled H2B-GFP mice was developed (Figure 1.7) (Tumbar et al., 2004). In this
26
system, GFP positive progeny expressing both transgenes were selected for pulse chase
experiments. The K5 promoter specifically targets the skin labeling all its cells including
hair follicles with H2B-GFP during embryogenesis. Upon tetracycline treatment using
doxycycline, the transcription of H2B-GFP is inhibited. The existing H2B-GFP label is
subsequently diluted out of dividing cells during prolonged doxycycline treatment. Over
time, only slow cycling LRCs will remain H2B-GFP positive.

Figure 1.7. K5 driven, tetracycline controlled H2B-GFP label retaining system. K5
driven tetracycline repressor-VP16 fusion protein transgenic mice is crossed with
tetracycline controlled H2B-GFP reporter to drive the tetracycline controlled H2B-GFP
expression specifically in the skin. Upon tetracycline (doxycycline) treatment, H2B-GFP
is turned off. The existing H2B-GFP labels gets diluted out of dividing cells and only
slow cycling, LRCs remain H2B-GFP positive. Adapted from (Tumbar et al., 2004).

27
For the first time, this method allowed the isolation of live hair follicle label retaining
stem cells for gene expression profiling to determine the molecular characteristics of
these stem cells independent of any other markers such as K15 or CD34 (Tumbar et al.,
2004).

28
1.10 Identifying LRCs as Putative Stem Cells in Various Tissues
Prior to identifying LRCs in the hair follicle bulge, Cotsarelis et al. have also reported
LRCs in the cornea limbus using similar pulse chase experiments (Cotsarelis et al., 1989).
It was demonstrated that these limbal LRCs were preferentially activated in response to
injury and can also be stimulated by TPA. Moreover, the limbus epithelium doesn’t
express the 64kd keratin differentiation marker found in the rest of the cornea and is now
believed to house the corneal stem cells. With the successful identification of hair follicle
stem cells using its label retaining property, it has been hypothesized that this slow
cycling property may be a common characteristic of stem cell populations in other
tissues. In the absence of known markers for adult stem cells residing in various tissues,
looking for LRCs has become a common method to identify populations of putative stem
cells.

Over the years, populations of LRCs have been identified in a number of different tissues
where some have been demonstrated to possess stem cell like characteristics such as the
ability to respond to injury or tissue regeneration and enhanced proliferative ability in
vitro. Previously, LRCs have been identified in the +4 position of intestinal crypts and
was proposed to be stem cells (Potten et al., 2002). A decade later, two independent
intestinal stem cell populations were also found at the bottom of the crypts expressing
Lgr5 and Bmi1 (Barker et al., 2007; Sangiorgi and Capecchi, 2008), both capable of
differentiating into all intestinal lineages. Bmi1 marked stem cells are predominantly
found in the +4 position where LRCs are localized; however, more work is needed to
29
confirm whether Bmi1 cells are LRCs (Sangiorgi and Capecchi, 2008). Recently, Tian et
al. have shown that in the absence of Lgr5+ stem cells, Bmi1 stem cells can maintain the
normal turnover of the intestines while repopulating the Lgr5+ stem cell population (Tian
et al., 2011). This suggests that Bmi1+ cells, localized where LRCs are found, act as a
reserve stem cell population. In other examples, LRCs of the neural tissue and kidney
have been identified and demonstrated to proliferate in response to wounding similar to
hair follicle and limbal stem cells (Oliver et al., 2004; Tavazoie et al., 2008). In
regenerative tissues that undergo extensive remodeling such as the mammary gland,
ovaries, and endometrium, LRCs have been shown to activate in response to pregnancy
and the estrous cycle, respectively (Chan and Gargett, 2006; Szotek et al., 2008; Welm et
al., 2002). Although less well characterized, LRCs have also been identified in the heart,
tracheal epithelium, and bladder (Borthwick et al., 2001; Kurzrock et al., 2008; Urbanek
et al., 2006). More work is needed to check whether LRCs in many of these tissues
represent bona fide stem cells. Due to the varying dynamics and characteristics of
different tissues, LRCs may not always necessarily mark stem cells as in the case of
hematopoietic stem cells (Kiel et al., 2007). Although LRCs are not specific for
hematopoietic stem cells, they can be used in conjunction with other markers to enrich
for these stem cells (Foudi et al., 2009). Nevertheless, it serves as a helpful tool in
locating putative stem cell populations where we can subsequently test its regenerative
potential and look for specific markers.

30
In the skin, the label retaining property has helped identify hair follicle stem cells
important for regeneration in normal homeostasis and repair during injury (Cotsarelis et
al., 1990; Ito et al., 2005). With its high regenerative potential, understanding the
molecular characteristics and mechanisms of these stem cells may prove useful in
regenerative medicine. Over the years, piles of data have emerged on understanding how
the hair follicle and its stem cells are regulated. However, little is known about the other
skin appendages and how they are regulated by their stem cells. More recently, LRCs
have been reported in the nails and sweat glands; however, further characterization is
required (Nakamura and Ishikawa, 2008; Nakamura and Tokura, 2009). Therefore, we
have used pulse chase experiments to identify, isolate, and characterize LRCs from the
sweat gland and nail as putative stem cells of these skin appendages.

31
Chapter 2:
Label Retaining Cells (LRCs) with Myoepithelial
Characteristic from the Proximal Acinar Region Define
Stem Cells in Sweat Glands.

2.1 Abstract
Adult stem cells possess the ability to reconstitute different tissue lineages leading to
interest in their therapeutic potential. Although hair follicle stem cells are one of the best
characterized adult stem cell systems, very little is known about other stem cells in
different skin appendages like sweat glands (SGs). Here, we used a H2B-GFP label
retaining cells (LRCs) system for in vivo detection of infrequently dividing cells. This
system allowed us to localize and isolate stem cells with label-retention and
myoepithelial characteristics restricted to the sweat gland (SG) proximal acinar region
with basal layer localization, absent in the SG distal ducts. Using an alternative genetic
approach, we also demonstrated that SG LRCs express Keratin 15 (K15) in the acinar
region and lineage tracing determined that K15 labeled cells contributed long-term to the
SG structure, but not to epidermal homeostasis. Surprisingly, wound healing experiments
did not activate proximal acinar SG cells to participate in epidermal healing. Instead,
predominantly non-LRCs in the SG duct actively divided while the majority of SG LRCs
remained quiescent. However, when we further challenged the system under more
favorable isolated wound healing conditions, we were able to trigger normally quiescent
32
acinar LRCs to trans-differentiate into the epidermis and adopt its long-term fate. In
addition, dissociated SG cells were able to regenerate SGs and, surprisingly, hair follicles
demonstrating their in vivo plasticity. By determining the gene expression profile of
isolated SG LRCs and non-LRCs in vivo, we identified several Bone Morphogenetic
Protein (BMP) pathway genes to be up-regulated and confirmed a functional requirement
for BMP receptor 1A (Bmpr1a)-mediated signaling in SG formation. Our data highlight
the existence of SG stem cells and their primary importance in SG homeostasis. It also
emphasizes SG stem cells as an alternative source of cells in wound healing and their
plasticity for regenerating different skin appendages.

33
2.2 Introduction
The skin contains a number of different skin appendages including hair follicles,
sebaceous glands, nails, and sweat glands (SGs). In adult skin, each hair follicle contains
a reservoir of stem cells localized in the bulge (Morris et al., 2004; Oshima et al., 2001;
Rochat et al., 1994). Stem cells reside in niches that provide a specialized environment to
regulate their proliferation and differentiation, which is important for tissue homeostasis
and repair (Fuchs et al., 2004; Lin, 2002; Spradling et al., 2001). Not only are hair follicle
stem cells important for hair follicle homeostasis and regeneration, they have also been
shown to differentiate into epidermal cells during wound healing (Gurtner et al., 2008; Ito
et al., 2005; Taylor et al., 2000). So far, hair follicle stem cells and their niches are one of
the best characterized systems in skin; however, much less is known about whether stem
cells and their niches exist in other skin appendages such as SGs.

In humans, eccrine SGs are abundantly present all over the body with an important
function to regulate body temperature. Improper thermoregulation can result in
hyperthermia, which could lead to death. Eccrine SGs are coiled tubular glands which
release their secretions through the straight duct at the distal part of SGs connected to the
surface of the skin (Lobitz and Dobson, 1961). The secretory acinar part of SGs contains
three distinct cell types: large basal clear cells rich in glycogen without secretory
granules, small apical dark cells with Schiff-reactive granules, and myoepithelial cells
localized between the basement membrane and basal part of clear cells. Clear cells are
responsible for sweat secretion (water and electrolytes) through intercellular canaliculi,
34
which reaches the lumen through intercellular spaces between apical dark cells. Apical
dark cells secrete glycoproteins into the lumen by exocytosis – a merocrine type of
secretion (Lee, 1960; Lobitz and Dobson, 1961). The normal sweat composition contains
water, sodium, potassium, chloride, urea, creatine, creatinine, lactate, and phosphate with
small amounts of mucoprotein (Lobitz and Dobson, 1961). SG density varies in different
regions of the body with its highest density on the sole, palm, and scalp in humans (Sato
et al., 1989a).

Furthermore, it has been shown that SG cells can contribute to and reconstitute a
functional epidermis leading to great interest in this skin appendage as an alternative
source of cells apart from hair follicle stem cells in skin regeneration (Biedermann et al.,
2010; Miller et al., 1998). Previously, results suggested by Lobitz et al. in human skin
showed that removal of the epidermis induced proliferation predominantly in basal cells
of the SG straight ducts (Lobitz and Holyoke, 1954). Subsequent experiments where SG
ducts were injured showed an initial migration of luminal cells after 24h followed by
proliferation of basal cells at 48h giving rise to new spiraling luminal cells as well as
tongues of prickle cells at both ends of the cut duct (Lobitz et al., 1956). Recently, Ritte
et al. confirmed that eccrine SGs are the most abundant appendage in human skin and are
major contributors of keratinocyte outgrowth to re-epithelialize the human epidermis
after injury (Rittie et al., 2013). However, these studies have not addressed precisely
which cell compartments or populations, if any, in SGs possess this regenerative capacity
with stemness characteristic.
35

Infrequently, slow dividing cells found in hair follicles and the corneal limbus have been
previously described as stem cells (Cotsarelis et al., 1989; Cotsarelis et al., 1990).
Therefore, this slow cycling characteristic has been suggested to be a common feature to
identify putative stem cells in different organs. In skin, pulse chase experiments with
labeled nucleotides designated the bulge as the residence of slow cycling, hair follicle
label retaining cells (LRCs) (Cotsarelis et al., 1990; Morris and Potten, 1999; Taylor et
al., 2000). More recently, BrdU pulse chase experiments identified the presence of LRCs
in the SG skin appendage (Nakamura and Tokura, 2009); however, these methods were
not useful to isolate live LRCs. A genetic approach that overcomes this obstacle was
developed using a double transgenic mouse model with inducible expression of a histone
2B (H2B) conjugated to green fluorescent protein (GFP, H2B-GFP) driven by a tissue
specific keratin 5 (K5) promoter (Diamond et al., 2000; Tumbar et al., 2004). In this
double transgenic line, we turn on H2B-GFP expression (“pulse”, no doxycycline
treatment) from early embryogenesis. By feeding the mice doxycyline (doxy), H2B-GFP
expression is turned off for the duration of the treatment (“chase”). During 4 weeks of
“chase”, slow cycling cells retain H2B-GFP expression (“label retaining cells”), whereas
rapidly dividing transit amplifying (TA) cells dilute out the H2B-GFP label upon each
division. This approach was previously used to detect H2B-GFP marked hair follicle
LRCs with slow cycling characteristic in vivo allowing isolation and characterization of
live hair follicle stem cells (Tumbar et al., 2004).

36
In this study, we exploited this H2B-GFP LRCs system for the in vivo detection of
infrequently dividing cells in SGs. This system allowed us to localize and isolate SG stem
cells with label retaining characteristics. We observed that SG LRCs were restricted to
the proximal acinar gland region and were absent in the SG ductal region. More
specifically, LRCs were localized to the basal layer of the glandular/acinar region and
displayed myoepithelial characteristics. Transcriptional analysis of SG LRCs and non-
LRCs allowed us to define common and unique features of these populations in vivo. We
also demonstrated that SG LRCs co-expressed keratin 15 (K15) in the acinar part and
K15 labeled cells were able to survive long-term and participate in SG homeostasis, but
not in epidermal homeostasis. Epidermal injury did not activate these SG cells in the
proximal acinar region to participate in epidermal healing. However, when the system
was further challenged under more favorable isolated wound healing conditions, we were
able to activate quiescent SG LRCs to trans-differentiate into the epidermis. Furthermore,
we demonstrated the plasticity of dissociated SG cells to regenerate SGs and surprisingly
hair follicles in vivo. Understanding the biology of SG stem cells may have great
implications in developing potential treatments for patients with hyperhidrosis (increased
sweating) and patients with inherited anhidrotic or hypohydrotic ectodermal dysplasias
who either lack or have underdeveloped appendages similar to burn victims (Kobielak et
al., 2001; Mikkola, 2009; Sato et al., 1989b; Wisniewski et al., 2002).

37
2.3 Slow Cycling LRCs are Localized to the Proximal Acinar Region of
Sweat Glands.
We employed the recently developed H2B-GFP system, composed of two transgenic
mouse lines: keratin 5-driven tetracycline repressor mice (K5-tTA) (Diamond et al.,
2000) and tetracycline response element-driven histone H2B-GFP transgenic mice
(pTRE-H2B-GFP) (Tumbar et al., 2004), to detect live, slow cycling, LRCs in vivo. In
these animals, H2B-GFP expression was uniformly detected in all cells of the epidermis,
hair follicles, and SGs (ducts and glands) prior to doxy treatment (Figure 2.1A and B).
After 4 weeks of “chase” by switching off H2B-GFP expression with doxy treatment, we
demonstrated the presence of infrequently dividing LRCs in SGs (Figure 2.1C and D).
Histological localization of LRCs in SGs was confirmed on serial sections of the paw
region by hematoxylin and eosin staining (H&E) (Figure 2.1E and F). The H2B-GFP
expression of LRCs was restricted to the proximal acinar gland region with no expression
in the ductal region (distal part of the SGs connected to the overlying epidermis) (Figure
2.1C and D). This region was previously reported to contain LRCs in human eccrine
glands (Nakamura and Tokura, 2009).
38

Figure 2.1. Sweat gland LRCs are localized in the acinar gland region of SGs. (A,B)
H2B-GFP is expressed in the epidermis and sweat glands before doxycycline treatment.
(C,D) Sweat gland LRCs are found in the acinar gland region after a 4 week chase with
doxycycline. (E,F) Tissue histology on sections with H&E staining.

These SG LRCs are extremely slow cycling with some of them persisting for more than
20 weeks of chase (Figure 2.2). By using the same exposure times, we demonstrate that
the H2B-GFP intensity decreases over time, presumably with each cell division (Figure
2.2). To probe the rate of cell division, the number of H2B-GFP marked LRCs was
quantified in relation to the amount of non-LRCs exclusively in the SG acinar secretory
region (not including sweat ducts). Preliminary studies show that at 4 weeks of chase,
approximately 43% of the acinar SG cells retained H2B-GFP expression. At 10 weeks of
chase, the amount of H2B-GFP LRCs in the acinar region only decreased to roughly 27%
demonstrating that about 15% of LRCs have diluted out its label through cell division.
Little to insignificant changes were observed at 15 and 20 weeks of chase where the
percentage of LRCs in the secretory coils remained around 25% (Figure 2.3). Such low
39
cellular turnover is expected since sweat glands do not undergo physiological tissue
remodeling or shedding of cells during sweat secretion. However, repeating this
experiment with additional animals is required for confirmation of these values. As an
internal control, we observed H2B-GFP expression of LRCs restricted to hair follicle
stem cells (bulge) after doxy treatment (Figure 2.2A).

Figure 2.2. Sweat gland LRCs possess slow cell cycle dynamics, but are non post-
mitotic cells. Hair follicle bulge at (A) 4 weeks, (B) 10 weeks, (C) 15 weeks and (D) 20
weeks of chase with doxycycline. 10x magnification of sweat glands at (E) 4 weeks, (F)
10 weeks, (G) 15 weeks and (H) 20 weeks of chase with doxycycline. 20x magnification
of sweat glands at (I) 4 weeks, (J) 10 weeks, (K) 15 weeks and (L) 20 weeks of chase
with doxycycline. Exposure time for the GFP channel of all 10x images were 800ms and
20x images were 160ms.

40
Dynamics of Sweat Gland LRCs
0
10
20
30
40
50
0 5 10 15 20 25
Weeks of Chase
% of H2B-GFP+ cells

Figure 2.3 Dynamics of sweat gland LRCs. Percentage of H2B-GFP labeled sweat
gland LRCs relative to unlabeled DAPI marked cells exclusively within the acinar
secretory region at 4 weeks, 10 weeks, 15 weeks, and 20 weeks of chase with
doxycycline. All images used for quantification were taken at a constant 100ms exposure
time.

41
2.4 Sweat Gland LRCs are Attached to the Basement Membrane and
Demonstrate Myoepithelial Characteristics.
SGs are composed of three different cell types, dark apical cells of the lumen, clear and
myoepithelial cells of the basal layer; therefore, we used immunofluorescence staining
with a number of different markers to determine where SG LRCs are localized. We
demonstrated that SG LRCs are attached to the basement membrane expressing β4
integrin (Figure 2.4A). In addition, SG LRCs also co-expressed the basal layer marker
keratin 14 (K14) (Figure 2.4B), whereas luminal layer markers, keratin 8 (K8) and
keratin 18 (K18), did not overlap with SG LRCs (Figure 2.4C & D, respectively). Next,
by performing p63 antibody (Ab) staining for a mammary gland myoepithelial cell
marker (Barbareschi et al., 2001), which was shown to be important for epidermal self-
renewal and differentiation (Koster et al., 2004), we distinguished that SG LRCs
specifically co-localize with myoepithelial cells of the basal layer (Figure 2.4E). In
addition, we show co-localization of smooth muscle actin (SMA) with SG LRCs further
supporting that SG LRCs are myoepithelial cells localized to the basal layer (Figure
2.4F).

42

Figure 2.4. Sweat gland LRCs are attached to the basement membrane and possess
myoepthelial characteristics. (A) Sweat gland LRCs are attached to the basement
membrane positive for β4 integrin (red) and are (B) found in the basal layer co-localizing
with K14. (C, D) Sweat gland LRCs do not express luminal layer markers K8 and K18,
respectively. (E) Within the basal layer, LRCs co-localize with myoepithelial cell marker
p63, inset denotes p63 single channel, arrows, and (F) with myoepithelial cell marker
SMA, arrows. Arrows denote co-localization of markers with the H2B-GFP LRCs.

SGs exist as 3-dimensional (3D) structures; therefore, we performed whole mount
staining of 4 week chased SGs with a basement membrane marker, laminin
(counterstained with DAPI), to examine how these cells are organized within the
appendage (Figure 2.5A and B). We used two-photon confocal microscopy to acquire
serial Z-stacks and performed 3D reconstruction of enzymatically purified whole SGs
(see Material and Method for details). This allowed us to generate a 3D model of SGs to
43
visualize the 3D organization of LRCs attached to the basement membrane (Figure 2.5A
and B, arrows).

Figure 2.5. Wholemount staining with 3D reconstruction confirms that sweat gland
LRCs are attached to the basement membrane. (A) Whole mount staining of SG
LRCs with laminin and (B) DAPI. Arrows denote co-localization of markers with H2B-
GFP marked LRCs.

44
2.5 Sweat Gland LRCs Express Keratin 15 (K15) in the Acinar Region
and K15 Marked Cells Contribute Long-Term to the Sweat Gland
Structure, but not to Epidermal Homeostasis.
It has been previously shown that hair follicle LRCs in the bulge specifically express K15
(Morris et al., 2004). Therefore, we examined whether K15 expression is present in SGs.
First, using K15 Ab staining, we demonstrated that mouse SG LRCs co-localized with
K15 (Figure 2.6A) confirming previously published data for human eccrine gland LRCs
(Nakamura and Tokura, 2009). Next, we established a genetic lineage tracing approach to
monitor K15 expressing cells in SGs by generating a K15 driven recombinase (Cre)
conjugated with a truncated progesterone receptor (PR), (K15CrePR), (Morris et al.,
2004) with three different Cre-dependent Rosa26 reporter mice: eYFP (enhanced Yellow
Fluorescent Protein, R26eYFP) (Srinivas et al., 2001), tdTomato (tandem dimer Tomato,
R26tdTom) (Madisen et al., 2010), and LacZ ( β-galactosidase, R26LacZ) (Soriano,
1999). After RU486 (RU) treatment for 16 days beginning at P43, YFP expression was
detected in SGs of the palm in both the foot pads and fingertips (Figure 2.6B, arrows).
This system allowed us to mark initially K15 positive cells and permanently label all of
its progeny in SG structures. To demonstrate whether K15 marked cells in SGs co-
localized with LRCs, we used a K15CrePR system with Rosa26-tdTomato reporter
mouse crossed onto the K5TetOff/TreH2BGFP background. After 4 weeks of chase with
doxy treatment (at ~P49) when LRCs (green) were present in SGs, we labeled the K15
expressing cells using a short RU treatment for 2 days to mark K15 positive cells in SGs
45
with tdTomato expression (Figure 2.6C). These results revealed that tdTomato expression
(red) was restricted to the acinar regions of SGs and co-localized with LRCs (Fig. 2.6C).

Figure 2.6. Sweat gland LRCs express Keratin 15. (A) K15 staining of sweat gland
LRCs indicate positive K15 expression. (B) Fluorescent photo of K15CrePR/R26eYFP
RU

palm containing YFP positive sweat glands after long term YFP activation. (C) Section of
K15CrePR/R26tdTom
RU
crossed onto K5TetOff/TreH2BGFP sweat glands after a 4
week chase with doxycycline followed by 2 days of RU treatment.

We also used a K15-GFP (Green Fluorescent Protein) reporter system (Morris et al.,
2004) to visualize cells that actively expressed K15 (without permanently labeling their
progenies) and confirmed that K15-GFP expressing cells were localized in the SG acinar
region (Figure 2.7A and B) where LRCs are found (Figure 2.6A and C). Moreover, we
stained K15-GFP SGs for basal layer marker, K14, and luminal layer marker, K18, to
demonstrate that K15 positive SG cells marked both layers of the SG structure (Figure
2.7A and B). Additionally, fluorescence activated cell sorting (FACS) of K15-GFP SGs
demonstrated that less than 6% of the total number of isolated cells from SG regions were
GFP positive (Figure 2.7C). Approximately less than half of the K15-GFP positive cells
(2.6% out of 5.8%) were labeled by α6-integrin, a basal layer marker, with the remaining
46
half (3.2% out of 5.8%) of K15-GFP positive SG cells negative for α6-integrin (Figure
2.7C).

Figure 2.7. K15-GFP reporter marks both basal and luminal layers.
(A) K14 basal layer staining co-localizes with GFP expression in K15-GFP transgenic
sweat glands. (B) K18 luminal marker staining co-localizes with K15-GFP expression in
sweat glands. (C) FACS analysis of K15-GFP sweat glands demonstrate that
approximately half of the K15-GFP positive cells are localized to the basal layer
expressing α6 integrin.

Finally, to address the long-term contribution of K15 expressing cells in SG and
epidermal homeostasis, we used K15CrePR mice crossed with a R26LacZ reporter. After
induction of LacZ by 16 days of RU treatment in adult mice, we were able to detect LacZ
positive cells (with X-gal staining) from the time of RU treatment to more than 6 months
after initial transgene activation in the acinar region (Figure 2.8A). This demonstrated
that K15 positive SG cells could survive long-term and contribute to SG homeostasis.
However, LacZ positive cells were not present in surrounding epidermis or the distal SG
duct (Figure 2.8A) suggesting that these cells do not contribute towards physiological
epidermal homeostasis behaving similar to hair follicles (Ito et al., 2005).
47

Figure 2.8. Keratin 15 marked sweat gland cells contribute long-term to the acinar
SG structure.
(A) Histology of X-Gal-treated K15CrePR/R26LacZ transgenic mice, blue stain indicates
transgene expression for more than 6 months after transgene activation in sweat glands.
(B) K15 expression co-localizes with LRCs in the acinar sweat gland region.

48
2.6 Isolating LRCs from Sweat Glands.
To isolate pure fractions of SG LRCs, we used a combination of surgical dissection with
subsequent enzymatic digestions. To avoid contamination from hair follicle LRCs, we
collected the whole paw with the toes and dissected out the SGs with the surrounding
sole’s epidermis (Figure 2.9A, yellow dissection line and Figure 2.9B). To purify SGs
from the attached sole’s epidermis, we digested the dissected SGs with collagenase for 1
hour at 37°C and mechanically separated the epidermis (Figure 2.9C). Purified SGs were
further digested with collagenase and hyaluronidase at 37°C for 1 hour and then
trypsinized for 20 min at 37°C to obtain a single cell suspension ready for FACS (Figure
2.9D).

Figure 2.9. Isolation strategy for sweat gland LRCs. (A) Whole
K5TetOff/TreH2BGFP toe tip under the stereomicroscope, dotted yellow line marks
sweat gland dissection. (B) Dissected sweat glands with surrounding sole’s epidermis.
(C) Separation of sweat glands from the attached sole’s epidermis after collagenase
treatment. (D) A series of further enzymatic digestions result in a single cells suspension
of sweat gland cells for FACS.

Since SG LRCs were attached to the basement membrane, we stained these cells with a
FACS specific antibody against α6 integrin to help purify these live LRCs and adjacent
non-LRC basal cells. The specific FACS gates to sort double positive SG LRCs (H2B-
49
GFP+ and α6 integrin+) and single positive SG non-LRCs (H2B-GFP- and α6 integrin+)
were setup according to negative (unstained), single GFP positive, and single α6 integrin
positive control cells (Figure 2.10). The majority of SG LRCs (H2B-GFP+) was positive
for α6 integrin (Figure 2.10). In parallel, we isolated adjacent SG non-LRC basal cells
( α6 integrin+) (Figure 2.10). FACS analysis revealed that SG LRCs accounted for
approximately 1-6% of the whole dermal fraction containing the SG skin appendage
(after detachment of sole’s epidermis).

Figure 2.10. FACS isolation of SG LRCs and non-LRCs. Sweat gland cells sorted for
H2B-GFP and α6 integrin-PE double positive LRCs as well as α6 integrin-PE single
positive surrounding basal cells (non-LRCs).

In an attempt to propagate these SG LRCs in vitro, we plated unsorted and sorted cells
with and without feeder. Cells seeded without feeder was cultured in mouse salivary
gland (MSG) media as previously described for sphere formation (Lombaert et al., 2008).
Under these conditions, SG cells formed colonies but grew slowly (Figure 2.11 A-E’). In
addition, we have also plated cells on mouse embryonic fibroblasts (MEF) grown in
media used for hair follicle stem cells propagation (Blanpain et al., 2004). Again,
50
formation of colonies was observed, but growth was inefficient (Figure 2.11 F-G’). In
both cases, the cells survived in culture for months; however, we were unable to passage
them.

Figure 2.11. Culturing of H2B-GFP SG LRCs.
(A, A’) Culturing of unsorted 4 week chased sweat gland cells in mouse salivary gland
(MSG) media at 3 days, (B, B’) 5 days, (C, C’) 7 days, (D, D’) 10 days and (E, E’) 14
days. (F, F’) Sorted GFP α6+ and (G, G’) GFP+ SG LRCs also formed colonies on
irradiated mouse embryonic fibroblasts (MEF).

So far there have been no reports of successful in vitro propagation of these tetracycline
regulated H2B-GFP cells. Therefore, we took an alternative approach and isolated K15
positive SG cells from K15-GFP reporter mice for culturing since SG LRCs also express
K15. These K15-GFP marked sweat gland cells were sorted through FACS and plated on
mitomycin treated 3T3 fibroblasts. Unlike the H2B-GFP LRCs, we were successful in
passaging these cells using media for hair follicle stem cells. However, the K15 driven
51
GFP expression was turned off in culture (Figure 2.12). Nevertheless, it will be
interesting to test the regenerative potential and plasticity of these K15 SG cells in the
future.

Figure 2.12. FACS isolation and cell culture of K15-GFP sweat gland cells.
(A) FACS isolation of K15-GFP+ sweat gland cells. (B-B’) Culture of K15GFP+ α6+ and
(C-C’) K15GFP+ cells at 1 week, (D-D’) Culture of K15GFP+ α6+ and (E-E’) K15GFP+
cells at 2 weeks. (F-F’) Culture of K15GFP+ α6+ and (G-G’) K15GFP+ cells at 3 weeks.
52
2.7 Defining Sweat Gland LRCs Characteristics.
To identify the transcriptional gene expression profile of SG basal layer cells, total RNA
was extracted from GFP+/ α6+ population (SG LRCs), GFP-/ α6+ adjacent basal layer
cells (SG non-LRCs), and α6+ basal layer cells of the sole’s epidermis for microarray
analysis from two independent experiments. Purification and microarray hybridization
(Affymetrix Mouse Gene 1.0ST) of each fraction were performed in duplicate. Then,
each population, SG LRCs and non-LRCs, were independently compared to the basal
layer of the sole’s epidermis (Figure 2.13A). By comparing SG LRCs to the sole’s basal
layer epidermis ( α6+), we identified 2426 (845+1581) genes to be consistently up-
regulated and 1342 (718+624) genes to be down-regulated by at least 2-fold in two
independent microarray analyses. Subsequently, we compared non-LRCs to the sole’s
basal epidermal layer ( α6+) and identified 1877 (296+1581) genes that were consistently
up-regulated and 998 (374+624) genes down-regulated by at least 2-fold (Figure 2.13B).
We then examined how many of the gene changes identified in SG LRCs and SG non-
LRCs were commonly or uniquely expressed between both populations. We identified
1581 up- and 624 down-regulated genes (out of 2,205 total) to be commonly up-regulated
or down-regulated by at least 2-fold in two independent microarray analyses from both
GFP+/ α6+ and GFP-/ α6+ basal layer cells of the SG (Figure 2.13A). Moreover, 1563
genes (845 up- and 718 down-regulated) were uniquely expressed in SG LRCs (Figure
2.13A), whereas 670 (296 up and 374 down-regulated) genes were uniquely expressed in
SG non-LRCs (Figure 2.13A). All these genes were consistently regulated in the two
independent experiments conducted.
53

Figure 2.13. Determining the molecular characteristics of sweat gland LRCs. (A)
GFP+/ α6+ sweat gland LRCs and GFP-/ α6+ sweat gland non-LRCs basal cells were
independently compared to the basal layer of the sole’s epidermis. (B) The resulting gene
expression profiles of these sweat gland LRCs and basal cells were then compared to
each other to determine common and unique genes expressed.

Some of the genes were validated through real-time PCR using separately isolated
biological samples from 2-3 different experiments independent from the microarray
analyses. In particular, genes expressed in the SG GFP- α6+ non-LRCs, Mmp2, Tcf4, and
Timp2, were confirmed to be up-regulated through real-time PCR when compared to the
SG LRCs fraction (Figure 2.14). In addition, we also performed real-time PCR for
biglycan (Bgn), which was found to be up-regulated in both SG LRCs and non-LRCs in
the microarray analyses. Although biglycan was found to be commonly expressed in both
fractions, it was up-regulated by at least 2-fold more in the non-LRCs fraction compared
to the LRCs population. Accordingly, our real-time PCR data validates that biglycan is
up-regulated in SG GFP- α6+ non-LRCs when compared to SG LRCs. All real-time data
was performed in triplicate using cDNA from either 2 or 3 independent biological
samples.
54

Figure 2.14. Validation of genes identified in the microarray analysis by qPCR.
Using SG LRCs as the baseline, we confirmed an up-regulation of Bgn, Mmp2, Tcf4, and
Timp2 in SG non-LRCs ( α6+ basal layer cells) when compared to SG LRCs GFP+/ α6+
SG LRCs in either 2 or 3 independent biological samples. Representative data from one
is shown.

Since SG LRCs showed myoepithelial characteristics, we further probed how the gene
expression profile of this population corresponded to its function when compared to SG
non-LRCs basal cells. To this end, we performed functional annotations (grouped
according to the DAVID software) enabling us to categorize a number of identified genes
in LRCs and non-LRCs (Table 2.1). We found that a number of these genes were
involved in cell adhesion, signaling and transcription. Moreover, a number of transporters
and ion channels, important for SG function, were identified in the microarray analyses
(Table 2.1).
55

Table 2.1. Categorization of the sweat gland gene expression profiles based on
function. Gene expression profiles consistently found in two independent microarray
analyses from independent biological samples were categorized into ion and protein
transport, signaling, transcription, extracellular matrix (ECM) and cell adhesion based on
function.

In particular, we identified the expression of sodium potassium (Na
+
/K
+
) ATPase pumps
(Atp1a1, Atp1b1, Atp1b3) in both GFP+/ α6+ and GFP-/ α6+ cells of the SGs. This
expression was confirmed by Na
+
/K
+
ATPase Ab staining in the SG skin appendage after
a 4 week chase with doxy (Figure 2.15A). In addition, vimentin was found to be
commonly up-regulated in LRCs and non-LRCs, which was previously reported in the
human SG myoepithelium (Schon et al., 1999). Expression of Gja1 (also known as
connexin 43) was up-regulated in both LRCs and non-LRCs. Staining for Gja1, we
observed positive expression mainly in the sweat ducts and not the acinar region;
however, some LRCs especially those at the coil-duct junction did express Gja1 (Figure
2.15B, arrow). Moreover, Smad2 was specifically up-regulated in the SG LRCs
population. To validate its expression, we stained 4 week chased sections with a
56
phosphorylated (phosho) Smad2 antibody, which indicates active signaling. All SG LRCs
showed positive nuclear staining for phospho-smad2 confirming the microarray data.
However, phospho-smad2 expression was not specific to LRCs only. It was also found in
other parts of the SG including luminal cells as illustrated by K8 co-localization (Figure
2.15C and D). Finally, in an attempt to validate the Smad5 expression in LRCs, we used a
phospho-Smad1/5/8 antibody where we observed positive expression in some weak
LRCs among positively marked luminal cells. Although the antibody was not specific, it
confirms that either smad 1, 5, 8, or a combination thereof, is expressed in the SG (Figure
2.15E and F).

57

Figure 2.15. Validation of genes expressed and identification of active BMP
signaling in sweat glands. (A) Sodium Potassium ATPase expression is confirmed in
sweat glands. (B) Gja1 (Connexin 43) was mainly expressed in the sweat ducts and not
the secretory glands. LRCs at the gland-duct junction expressed Gja1. (C) Phospho-
Smad2 expression was found in all LRCs. (D) P-smad2 and K8 channels only
corresponding to panel C. Arrows denote LRCs. (E) Positive phospho-smad1/5/8 staining
indicates active BMP signaling in sweat glands. (F) Arrows denote LRCs corresponding
to panel E.

58
2.8 BMP Signaling is Required for Sweat Gland Morphogenesis.
A number of genes important for Bone Morphogenetic Protein (BMP) signaling,
including Bmpr1a, Bmpr2, Smad5, Id2, Id3, and Decorin were up-regulated in the SG
when compared to the sole’s epidermis (Table 2.1). Therefore, we tested for BMP
signaling activity through phospho-smad1/5/8 staining and observed positive nuclear
phospho-Smad activity in adult mouse SGs suggesting that BMP signaling may be
important in this appendage (Figure 2.16A). Since the SG skin appendage develops from
K14 marked precursors as indicated by X-gal staining of a keratin 14 driven Cre
recombinase (K14Cre) R26LacZ reporter (Figure 2.16B), we decided to ablate Bmpr1a
during development using K14Cre (Vasioukhin et al., 1999) crossed onto a K14-
H2BGFP reporter (Rendl et al., 2005) to functionally examine the effects of BMP
signaling in SGs.

Figure 2.16. Sweat gland appendages develop from K14 marked progenitor cells.
(A) Active BMP signaling is found in the sweat gland indicated by phospho-smad1/5/8
staining. (B) Lineage tracing show that sweat glands develop from K14 precursors.

At P1 and P8, we observed downgrowth and SG development in control mice (Figure
2.17A and C). However, in Bmpr1a KO mice, SGs were absent and failed to form
59
suggesting that BMP signaling is required for normal SG morphogenesis (Figure 2.17B
and D). Ki67 staining indicated that the basal layer of Bmpr1a KO epidermis appeared
capable of proliferation, but failed to acquire the SG fate (Figure 2.17E and F).

Figure 2.17. BMP signaling is required for sweat gland formation. (A) Downgrowth
of sweat glands is observed in P1 control paws but is (B) absent in Bmpr1a/K14Cre/K14-
H2BGFP KO paws. (C) Similarly, more developed sweat glands are observed in P8
control paws but are (D) still absent in KO mice. (E) The basal layer of the epidermis as
well as the sweat glands is proliferative at P1. (F) Although sweat glands are absent in
Bmpr1a/K14Cre/K14-H2BGFP KO paws, the epidermal basal layer is still capable of
division.

Taken together, BMP signaling plays an important role in SG development. Previous
reports have demonstrated that BMP signaling was also required for proper hair follicle
formation since Bmpr1a deletion resulted in abnormal hair follicle differentiation where
the IRS and hair shaft structures were missing (Andl et al., 2004; Kobielak et al., 2003;
Yuhki et al., 2004). One of the suggested in vitro downstream target of BMP signaling is
the ID2 gene (Nakahiro et al., 2010), which was consistently up-regulated in both
60
microarrays. Therefore, we tested if ID2 is also important in SG morphogenesis by
analyzing ID2 null SGs. Similar to Bmpr1aK14Cre KO mice, ID2 null mice contained
smaller hair follicles than controls (data not shown). Looking more closely, we observed
a compromised hair shaft in ID2 KO animals lacking AE15 expression in the medulla
(Figure 2.18A and B). Although the ID2 phenotype is similar to Bmpr1a KOs, the hair
follicle defects are less severe. Interestingly, however, SG development in the ID2 null
mice appears to be unaffected. Downgrowth with proper formation of SGs is observed in
both heterozygous and KO animals at P1 as indicated by K5 (Figure 2.18 C and D) and
H&E staining (Figure 2.18 E and F).

Figure 2.18. Unlike hair follicles, ID2 is not required for sweat gland development.
(A) AE15 stains the IRS and medulla (arrows) layer of the hair shaft. (B) Hair shaft is
compromised in ID2 null mice as indicated by the absence of the medulla, arrow. (C) K1
and K5 staining demonstrating normal downgrowth of sweat glands in heterozygous and
(D) ID2 null mice. (E) H&E staining of serial sections of ID2 heterozygous and (F) ID2
KO sweat glands at P1.
61
2.9 Acinar Sweat Gland Cells Do Not Contribute to the Epidermis
during Wound Healing.
Previously, it has been reported that bulge hair follicle stem cells with LRCs
characteristic do not participate in epidermal homeostasis; however, they can actively
deliver cells to the epidermal wound during skin injury (Ito et al., 2005). In addition,
keratinocytes in the region directly above bulge LRCs, marked by Lgr6, can postnatally
maintain sebaceous gland and interfollicular epidermis homeostasis and execute long
term wound repair (Petersson et al., 2011; Snippert et al., 2010). However, little is known
about the role of SGs in active epidermal regeneration initiated upon wounding. In human
SGs, it has been reported that basal cells of the straight duct undergo division when
provoked by skin injury (Lobitz and Holyoke, 1954). Since we have demonstrated that
the SG acinar cells marked by K15CrePR/R26LacZ do not participate in epidermal
keratinocyte lineages during homeostasis, we next examined if these acinar SG cells
could respond and actively contribute to epidermal wound repair upon injury. Wounds
where the epidermis was effectively scraped off were performed on K15CrePR/R26LacZ
mice in order to trace K15 positive SG cells and their progeny. Wounds were allowed to
heal for 24h, 48h, and 72h when samples were collected for analysis. X-gal staining for
LacZ enabled visualization of K15 positive SG cells and their progeny (blue). At all time
points, no blue cells were detected in the regenerating epidermis (Figure 2.19A to C). To
investigate whether SG LRCs were responding to the injury, we performed similar
wound healing experiments on 4 week chased K5TetOff/TreH2BGFP mice. In this
experiment, we pulsed the animals with BrdU to probe for proliferating cells in the SGs.
62
BrdU was detected sporadically in a few LRCs, but most SG cells remained quiescent
(Figure 2.19D). To examine which cells were responding to the injury, we stained the
wound healing samples from K15CrePR/R26LacZ mice with a Ki67 Ab and found that
the basal layer of the epidermis and SG duct cells were active in the cell cycle, whereas
the SG acinar region remained quiescent (Figure 2.19E and F). Under physiological
homeostasis, SG duct cells appear to be more active than the SG acinar region similar to
the basal layer of the epidermis (Figure 2.20A and B).
63

Figure 2.19. Acinar sweat gland cells do not contribute to the epidermis during
typical wound healing. (A) K15CrePR/R26LacZ
RU
marked sweat gland cells do not
contribute to the epidermis at 24h, (B) 48h, and (C) 72h after wounding. (D) BrdU pulse
shows that a few SG cells are activated upon injury (inset, arrows) while most SG LRCs
remain quiescent. (E) Ki67 staining confirms that the acinar sweat gland region is
quiescent while the SG duct and epidermal basal layer is proliferative at 24h and (F) 48h.
Abbreviations: du, duct.

Figure 2.20. Sweat ducts are proliferative under normal homeostasis. (A) Under
normal homeostasis, cells of the SG duct and epidermal basal layer are active in the cell
cycle. (B) Corresponding single Ki67 (red) channel. Abbreviations: du, duct.

64
2.10 Sweat Gland LRCs can Trans-Differentiate into Epidermis under
Prolonged Isolated Wound Healing.
Although the acinar cells of SGs did not contribute to wound healing under normal
circumstances, we further challenged the system using more favorable conditions. For
this, we isolated 4 week chased H2B-GFP labeled whole SGs through collagenase
digestion (as described in Figure 2.9 and Material and Methods). Then, we transplanted
the dermal portion of these SGs into a silicon chamber implanted onto the back of an
immunocompromised “nude” mouse (Weinberg et al., 1993). The advantage of this
chamber graft experiment was that the transplanted silicon dome containing the SGs
physically separated and initially prevented the surrounding “nude” epidermis from
closing the wound, thereby increasing the chance for the transplanted SG epithelial cells
to respond to wound regeneration. The chamber was removed approximately 14 days
after transplantation and the wound was allowed to heal for about one month (Figure
2.21A). At 34 days after transplantation, H2B-GFP labeled cells were observed in the
transplanted region of the regenerated skin (Figure 2.21B). To determine the participation
of these marked SG cells in epidermal regeneration, we sectioned and stained these
transplants with a number of different markers at 38 and 46 days after transplantation.
The majority of the transplanted H2B-GFP marked SGs were found in the dermis (Figure
2.21C, arrows). However, when we closely examined the epidermis and increased
exposure time for the GFP channel on the same section, we found that a number of H2B-
GFP expressing cells with lower intensity contributed to the newly formed epidermis
(Figure 2.21C vs D). These H2B-GFP cells co-localized with the basal layer marker
65
keratin 5 (K5) at 38 and 46 days (Figure 2.21D, E and I, respectively). Furthermore, we
demonstrated that these H2B-GFP labeled SG cells could proliferate along the basal layer
using Ki67 staining (Figure 2.21F and J). Next, we checked if the H2B-GFP marked SG
cells found in the regenerating epidermis were able to adopt epithelial characteristic by
performing immunofluorescence staining for different epidermal differentiation markers.
At both 38 days and 46 days, we showed that the H2B-GFP positive SG cells could
contribute to the suprabasal layer of the epidermis indicated by co-localization with K1, a
suprabasal layer marker (Figure 2.21G and K). Similarly, we observed that H2B-GFP
labeled cells could differentiate into cells of the granular layer indicated by loricrin
staining (Figure 2.21H and L). Thus, these lineage tracing experiments demonstrated that
H2B-GFP marked SG cells were able to contribute to epidermal regeneration following
injury. However, since these experiments were performed “off doxy” we were not able to
rule out whether the SG H2B-GFP LRCs themselves proliferated and contributed to this
newly formed epidermis or whether other non-LRC SG cells “turned on” H2B-GFP
expression in the absence of doxy.

66

Figure 2.21. Sweat gland cells can trans-differentiate into the epidermis under
prolonged isolated wound healing conditions. (A) DIC photo of the back skin area
containing transplanted H2B-GFP labeled sweat glands at 34 days. (B) Corresponding
fluorescent image displaying the presence of H2B-GFP positive expression, arrows.
Asterisk (*) marks autofluorescence of wound area. (C) Biopsy was taken at 38 days
where H2B-GFP sweat glands were observed in the dermis, arrow. (D) Higher exposure
time for the GFP channel on the same section from (C) detected marked cells with lower
H2B-GFP intensities in the epidermis. (E, I) K5 basal layer staining demonstrating the
contribution of sweat gland cells to the newly formed epidermis at 38 and 46 days,
respectively. (F, J) These H2B-GFP labeled cells are still able to proliferate as indicated
by Ki67 staining. (G, K) Sweat gland cells can differentiate into cells of the suprabasal
layer marked by K1 at 38 and 46 days, respectively. (H, L) These sweat gland cells can
also contribute to the granular layer as marked by loricrin. White dotted lines mark
dermal-epidermal interfaces.
67

To address this, we repeated this experiment using 4 week chased H2B-GFP labeled SGs.
In this case, the host mouse with the transplanted SG dermis was kept on doxy treatment
for the entire experiment; thus, only SG LRCs and their direct descendent would be
marked by H2B-GFP. At 30 and 40 days after transplantation, the H2B-GFP label
appeared to have been diluted out of some acinar SG structures (confirmed by K8 luminal
layer staining), but not all (Figure 2.22A to C, arrows). Moreover, some of these
transplanted SGs appeared to be fused with the newly regenerated epidermis (Figure
2.22A and B). In some instances, SGs closer to the SG-epidermal connections were the
ones lacking H2B-GFP marked SG LRCs, suggesting that the LRCs of these SGs had
been actively dividing and diluted out the nuclear H2B-GFP label (Figure 2.22B and C,
arrows). This observation was confirmed using Ki67 staining where we observed Ki67
positive cells present in SGs lacking LRCs as marked by green K5 membrane staining
(Figure 2.22D). Finally, we show that SG LRCs themselves can also contribute to the
regeneration of the epidermis as marked with K5 co-localization (Figure 2.22F, arrows).
Staining for Ki67, we illustrate that SG LRCs near the basal layer of the newly
regenerated epidermis, marked by basement membrane marker β4 intergrin (CD104), can
proliferate (Figure 2.22E). In addition, SG LRCs can differentiate into cells of the
suprabasal and granular layers of the epidermis as marked by K1 and loricrin co-
localization, respectively (Figure 2.22G and H). In conclusion, we have shown that under
more favorable conditions, SG LRCs can divide and contribute to the different layers of
the epidermis.
68

Figure 2.22. Sweat gland LRCs can trans-differentiate into the epidermis under
prolonged isolated wound healing conditions. (A-C) When 4 week chased sweat glands
are transplanted and kept on doxycycline treatment, the H2B-GFP label gets diluted out
of some sweat glands (arrows), as confirmed by K5 and K8 staining for sweat glands. (B)
Some sweat glands are connected to the newly formed epidermis lacking visible H2B-
GFP LRCs. (D) Ki67 staining show that these sweat glands lacking visible H2B-GFP
LRCs (green - nuclear) defined by K5 positive staining (green membrane staining)
contains Ki67 positive dividing cells (arrows). (E) Sweat gland LRCs contributing to the
epidermis are proliferative near the CD104 marked epidermal basal layer as marked by
Ki67, inset shows magnification of co-localization. (F) H2B-GFP sweat gland LRCs can
contribute to the newly formed epidermis as indicated by K5 basal layer staining
(arrows). (G) Sweat gland LRCs can contribute to the suprabasal layer marked by K1 as
well as the (H) granular layer expressing loricrin. Abbreviations: CD104, β4 integrin.
69
2.11 Dissociated Sweat Gland Cells can Regenerate both Sweat Glands
and Hair Follicles.
To test the regenerative potential of all SG cells, we dissociated 4 weeks chased, H2B-
GFP labeled, SGs into a single cell suspension after separation from the sole’s epidermis
(as in Figure 2.9C and D). These unsorted cells were then injected under the membrane
of the kidney capsule of immunocompromised “nude” mice (Figure 2.23A). In two
independent experiments, these dissociated SG cells formed SG-like structures at 13 days
and 25 days after transplantation (Figure 2.23B and C).

Figure. 2.23. Kidney capsule transplantation of dissociated unsorted 4 week chased
SG LRCs. (A) Injection of unsorted 4 week chased LRCs single cells suspension. (B)
Formation of sweat gland-like structures was observed at 13 days and (C) 25 days in 2
independent experiments.

To further probe the regenerative potential, we performed chamber graft transplantation
by mixing these unsorted H2B-GFP marked dissociated SG cells with unmarked freshly
isolated back skin dermal fibroblasts. Surprisingly, 29 days after transplantation, we
observed several GFP positive areas under the epidermis with some of them connected to
GFP positive hair-like fibers sticking out of the graft region (Figure 2.24A). Indeed,
70
analysis of sections from the graft area confirmed the presence of GFP positive hair
follicles, likely originating from the transplanted unsorted H2B-GFP labeled SG single
cells suspension (Figure 2.24B). These newly formed hair follicles were further
characterized by immunofluorescence staining with several hair specific markers
including K5 positive expression in the ORS (Figure 2.24C), AE15 expression in the IRS
and medulla (Figure 2.24C’), and AE13 expression in the cortex of the hair shaft (Figure
2.24C”). Interestingly, when we analyzed the graft 70 days after transplantation, we also
found coexisting (on the same section) H2B-GFP positive SG structures expressing SG
markers: Na
+
/K
+
ATPase and luminal layer marker K8 (Figure 2.24D, inset, and 2.24F,
respectively) in addition to GFP marked hair follicles (Figure 2.24D to E’). This
demonstrated that dissociated single SG cells suspension could still be flexible in their
fate decision choices between SGs or hair follicles. Both these structures were found to
coexist in the same region, which is not physiologically observed in mice. Consistent
with the rest of the SG graft data presented here, the single SG cells suspension can also
contribute to the regeneration of differentiated layers of the epidermis marked by K5, K1,
and loricrin (Figure 2.24G to I).

71

Figure 2.24. Dissociated sweat gland cells can regenerate sweat glands, hair follicles,
and the epidermis. (A) Chamber graft of sweat glands dissociated into a single cell
suspension labeled with H2B-GFP (chased for 4 weeks) mixed with newborn unmarked
dermal fibroblasts yields GFP positive hair-like fibers at 29 days. (B) Section through
graft confirms the presence of H2B-GFP positive hair follicles. (C) K5 staining marks the
outer root sheath of this H2B-GFP+ hair follicle. (C’) AE15 stains the inner root sheath
and medulla, arrows. (C”) AE13 stains the hair shaft. (D) 70days after transplantation,
H2B-GFP positive sweat gland structures were found, as confirmed by Na
+
/K
+
ATPase
expression (inset), in addition to H2B-GFP positive hair follicle. (E) Magnification of the
hair follicle in (D) with (E’) GFP single channel. (F) The sweat gland structures found
also expressed K5 basal and K8 luminal layer markers. (G) H2B-GFP positive cells were
also found in the K5 basal layer, (H) K1 suprabasal layer, and (I) Loricrin marked
granular layer of the newly regenerated epidermis.

In an independent experiment, we also subcutaneously injected unsorted dissociated cells
from H2B-GFP labeled SGs (chased for 4 weeks) after separation from the sole’s
epidermis in combination with freshly isolated dermal fibroblasts under the back skin of
72
an immunocompromised “nude” mouse. This transplant was harvested 39 days after
subcutaneous injection where we observed a cluster of hair follicles and GFP positive
structures at the injection sites (Figure 2.25A and B). Similar to the chamber graft
experiment with dermal fibroblasts, we observed the presence of GFP positive hair
follicles differentiated from the injected H2B-GFP positive SG cells (Figure 2.25C). In
addition, we also found H2B-GFP labeled SG structures containing a basal and luminal
layer marked by K5 and K8, respectively (Figure 2.25D). These injected cells have again
differentiated into cells of the epidermis marked by K5, K1, and loricrin (Figure 2.25E
and F). As a control, unsorted epidermal cells simultaneously isolated from the sole’s
epidermis were also subcutaneously injected with freshly isolated dermal fibroblasts;
however, no signs of H2B-GFP marked hair follicle formation were observed (data not
shown).

73

Figure 2.25. Subcutaneously transplanted dissociated sweat gland cells can
regenerate sweat glands, hair follicles, and the epidermis. (A, B) In an independent
experiment, 39 days after subcutaneous injections of 4 week chased unsorted dissociated
cells from H2B-GFP labeled SGs with unmarked newborn dermal fibroblasts, a cluster of
GFP positive cells containing hair follicles were observed. (C) Magnification of a GFP+
hair follicle from panel (B). (D) Sections show the presence of H2B-GFP labeled SG
structures expressing K8 luminal layer marker. (E) H2B-GFP positive cells are again also
found in the basal, suprabasal, and (F) granular layers of the epidermis.

74
2.12 Discussion
Here, we demonstrate that cells with slow cycling characteristic exist in SGs as a
scattered population localized in the SG basal layer of the proximal acinar region. As hair
follicle LRCs have been previously described as stem cells (Tumbar et al., 2004), we
asked if these newly identified SG LRCs also possess stem cell characteristics in vivo.
Although LRCs have been reported in both mouse and human SGs (Nakamura and
Tokura, 2009), their characterization, precise localization and function have not been
addressed so far. The K5TetOff/TreH2BGFP approach allows us to mark and isolate live
SG LRCs in vivo for further characterization. We were able to localize LRCs in the basal
layer of the proximal acinar part of SGs and demonstrate their myoepithelial
characteristic by SMA co-expression. In addition, we demonstrated that SG LRCs
specifically co-localize with p63 expression, which has been shown to be specifically
expressed in mammary gland myoepithelial cells (Barbareschi et al., 2001). Previous
studies illustrated that p63 is not only essential for epithelial development, but is also
important for epidermal self-renewal and differentiation (Koster et al., 2004; Yang et al.,
1999). In addition, p63 is believed to be a marker of corneal and epidermal stem cells
(Pellegrini et al., 2001). Due to its specificity for SG LRCs, p63 may potentially also be
used as a marker for SG LRCs in other systems. Recently, some similar findings
regarding SG LRCs were reported by Lu et al. (Lu et al., 2012). However, they did not
further characterize this SG LRC population per se or address their in vivo function.
Instead, they employed elegant systems, previously published in mammary glands (Van
Keymeulen et al., 2011), to identify and characterize distinct stem cell populations in the
75
basal and luminal layers of SGs (Lu et al., 2012). In our study, we used a different
approach and focused predominantly on basal, myoepithelial LRCs (GFP+/ α6+) after 4
weeks of chase. Although we were able to characterize this population of SG LRCs, this
genetic approach did not allow us to study the luminal layer of SGs in more detail.

The Proximal Acinar SG Region only Contributes to its Own Structure during
Homeostasis and Wound Healing
As an alternative and parallel approach, we used genetic K15CrePR in vivo systems to
mark cells specifically in the proximal acinar part of SGs including SG LRCs in the basal
layer (Figure 2.6C). However, these K15 labeled cells did not specifically mark SG LRCs
but had also marked cells of the luminal layer in the proximal acinar part of SGs as
illustrated with K15-GFP reporter mice (Figure 2.7A and B). This data was confirmed by
FACS analysis of SGs from adult K15-GFP reporter mice, where we observed an
approximate 1:1 proportion of GFP marked basal and luminal cells (Figure 2.7C).
Although our finding differs from the data recently published by Lu et al., this
discrepancy could be attributed to the use of different time points for labeling these K15
progenitors. In our case, we marked them by RU treatment for K15CrePR systems, or
analyzed them in K15-GFP reporter mice during adulthood at/or after P21 (after SG
morphogenesis was completed), whereas Lu et al. labeled them at an early postnatal time
point between P5 to P9 (during SG morphogenesis) and analyzed them at P22 (Lu et al.,
2012). Thus, it appears that this K15 promoter has different specificity during and after
SG morphogenesis. Since the K15CrePR system permanently marks K15 expressing cells
76
and its progeny, we used it to evaluate the contribution of K15 marked acinar cells in
overall SG and skin homeostasis. Our results demonstrate that K15 labeled cells,
localized exclusively in the acinar part of SGs, contributed only the proximal glandular
part long term, but not to the homeostasis of SG ducts or the surrounding epidermis
(Figure 2.8A). Furthermore, we also show that typical wound stimulation surprisingly did
not activate these proximal acinar SG cells to participate in epidermal healing (Figure
2.19A to C). Instead, we observed that only SG duct cells proliferated (Figure 2.19E). In
general, these results support previously published wound healing data in human SGs by
Lobitz et al. (Lobitz and Holyoke, 1954) as well as mouse SGs by Lu et al. (Lu et al.,
2012), which showed the contribution of SG duct cells in wound healing. However, since
Lu et al. used a different system, Sox9CreER/RosaLacZ or YFP, they could not rule out
fully if these contributing cells are in fact coming only from the duct or from the upper
acinar part of SGs as well. In addition, they also used the K15CrePR/R26LacZ approach;
however, they specifically labeled only luminal SG cells in the glandular part after early
postnatal RU treatment and could not address whether basal layer myoepithelial cells
could respond to injury. Although our results are consistent, using later adult postnatal
RU applications in K15CrePR/R26LacZ mice allowed us to mark both basal and luminal
layers of the proximal acinar part of SGs. Therefore, we were able to extend this
conclusion to both SG layers and make the statement that not only luminal but basal cells
as well from the SG proximal acinar region do not contribute to wound healing of the
epidermis. Collectively, previous reports together with our findings emphasize that both
layers of SG cells remain quiescent in the proximal acinar region during wound healing
77
and only SG duct cells were able to proliferate. It still remains to be addressed in the
future which duct layer, luminal or basal, can contribute to wound healing and cells from
which layers are able to fully trans-differentiate into epidermal cells long term.

SG LRCs Possess Multipotency and Stem Cells Characteristic in vivo and has
Potential to Trans-Differentiate into Epidermis under Prolonged Isolated Wound
Healing
As we demonstrated here, K15CrePR/R26LacZ labeled cells in the acinar part of SGs
appear to be generally slow cycling, but these cells were able to selectively maintain and
participate in the long-term homeostasis of the glandular part of SGs (Figure 2.8A). Thus,
it suggests that at least part of this SG structure contains cells with stem cell
characteristics that can maintain this glandular portion. Although K15 labeled cells
overlap with SG LRCs in the basal layer of SGs, it also marked cells of the luminal layer
(Figure 2.6C and 2.7B) preventing us from determining whether SG LRCs themselves
possess stem cells characteristic. In addition, we show that SG LRCs survive long term
and persists for more than 20 weeks of chase illustrating their extreme slow cycling
property. Although the H2B-GFP label persists for such a long period of chase, we
demonstrate that its intensity slowly diminishes as these SG LRCs slowly divide over
time (Figure 2.2).

To assess the stem cell properties of SG LRCs in vivo, we had to challenge our system
further since regular wound healing conditions failed to provoke SG acinar cells to
78
contribute to wound healing of the epidermis. Under this special wound condition where
we gave SG cells an advantage by preventing wound closure from the surrounding
epidermis, we observed that 4 week chased H2B-GFP labeled SG LRCs can proliferate
and trans-differentiate into all epidermal layers (Figure 2.22F to H). In this experiment,
animals were kept on continuous doxy treatment allowing us to distinguish SG LRCs
from other cells and confirm that SG LRCs themselves are multipotent. Together, our
results remain in agreement with previously published results on human, mouse and
porcine (Biedermann et al., 2010; Lobitz et al., 1956; Lobitz et al., 1954; Lu et al., 2012;
Miller et al., 1998; Rittie et al., 2013), demonstrating that in general, SG cells can
respond and re-epithelialize the skin after wounding. However, for the first time, we
show that under more favorable, isolated, and prolonged wound healing condition,
normally quiescent myoepithelial SG LRCs can contribute to and reconstitute a stratified
epidermis. Thus, we demonstrate that under favorable conditions, these relatively
quiescent SG LRCs can be activated and work as an alternative source of cells
confirming that these cells are multipotent with SC characteristics in vivo.

Purification and Characterization of Basal Layer Myoepithelial SG LRCs from the
Acinar Sweat Gland Region
To further characterize these SG stem cells, we used the K5TetOff/TreH2BGFP approach
to localize and isolate SG myoepithelial LRCs from the proximal acinar part of SGs. We
purified SG LRCs and adjacent basal layer cells representing SG non-LRCs, which were
predominantly composed of basal layer cells from the acinar and ductal regions.
79
Although all SG LRCs showed myoepithelial characteristics co-expressing SMA and
p63, only a fraction of SMA positive cells were LRCs, while the remaining majority of
SMA positive cells did not display label retaining characteristics (Figure 2.4F). Thus, in
our experimental setup, we specifically targeted and purified SG LRCs, which
represented approximately 30 to 40% of all basal SG cells attached to the basement
membrane. Consequently, our SG LRCs purification strategy was different from the
strategy recently published by Lu et al., where they purified all myoepithelial cells
including those without LRC characteristic as well as luminal cells in both glandular and
ductal parts to identify and characterize distinct unipotent stem cell populations in SGs.
They used Abs against α6 (CD49) and β1 (CD29) integrins, markers previously published
for mammary gland SCs purification (Shackleton et al., 2006; Stingl et al., 2006), in
conjunction with K14-H2BGFP mice which marked basal cells with high GFP intensity
and luminal or suprabasal layer cells with lower GFP intensity, whereas we focused more
selectively on basal myoepithelial cells with LRCs and non-LRCs characteristic. Our
microarray data revealed gene expression changes in both basal populations, SG LRCs
and SG non-LRCs, when compared to the sole’s epidermis. This allowed us to identify
the expression of various ion channels important for SG function including the Na
+
/K
+

ATPase pumps, Atp1a1, Atp1b1, and Atp1b3. In addition, vimentin, which was
previously published as a marker of the human SG myoepithelium (Schon et al., 1999),
was found up-regulated in SG LRCs and SG non-LRCs populations. Interestingly, we
found components of BMP signaling including Bmpr1a, Bmpr2, Smad5, Id2, Id3 and
Decorin to be up-regulated in the SG when compared to the sole’s epidermis (Table 2.1)
80
suggesting a quiescent behavior of those populations in vivo. This is consistent with a
well-known role of BMP signaling in maintaining quiescence in several adult stem cell
populations including hair follicle, hematopoietic, intestinal and neural stem cells
(Haramis et al., 2004; He et al., 2004; Kandyba et al., 2013; Kobielak et al., 2007; Mira et
al., 2010; Zhang et al., 2003). Although the direct role of BMP signaling in adult SG stem
cells has not been investigated yet, we were able to confirm the functional requirement of
Bmpr1a during SG formation. In the absence of Bmpr1a, no SGs are formed. In contrast,
ablation of a BMP signaling target gene, ID2, which was also identified in our
microarray, was not sufficient in preventing SG morphogenesis. This may be explained
by potential compensation from other inhibition of differentiation genes such as ID3,
which was also identified in our microarray. This difference in phenotype was similar to
that observed in Bmpr1a KO versus ID2 KO hair follicles where ID2 KO hair follicles
displayed less severe defects than Bmpr1a KOs. In the future, it will be important to
develop new genetic tools to specifically address further questions about the function of
several newly identified genes in SGs to better understand their role in SG biology both
during development and in adult SG homeostasis.

Plasticity of Sweat Gland Cells to Reconstitute Sweat Gland and Hair Follicles in
vivo
Finally, since we were unable to passage and expand these K5TetOff/TreH2BGFP cells
in culture, we could not probe their in vitro potential and subsequently use them for in
vivo reconstitution assays. Instead, we used unsorted dissociated SG cells isolated
81
directly from whole SGs (after removing the surrounding epidermis) to demonstrate their
regenerative potential. In our reconstitution assay, we showed that these cells were able
to generate H2B-GFP positive SG structures expressing SG markers K8 and Na
+
/K
+

ATPase (Figure 2.24D and F; Figure 2.25D). Surprisingly, we also observed the
formation of a few hair follicles with fully differentiated hair shaft fibers (Figure 2.24A
to E’). In addition, both SG and hair follicle appendages were found to coexist in the
same region, which is physiologically not observed in mouse sole’s epidermis (only SGs)
or back skin (only hair follicles). Since this finding was unexpected, we tested for
possible contamination from surrounding paw’s epidermis, which could potentially be a
source of keratinocytes capable of generating hair under the inductive properties of
freshly isolated GFP negative newborn dermal fibroblasts. For our control, we used
unsorted cells isolated from the attached 4 week chased sole’s epidermis mixed with
freshly isolated unmarked dermal fibroblasts and did not observe any H2B-GFP positive
hair follicle formation. Thus, it is unlikely that the H2B-GFP marked hair follicles
resulted from contaminating keratinocytes; however, fewer cells were used in this control
experiment when compared to the SG reconstitution assay. Instead, an alternative
explanation would be that some SG cells were able to adopt this new hair follicle fate.
This observation might suggest that some SG cells still possess plasticity to choose
between their final fate decisions dependent on surrounding environmental cues. This
raises additional questions to be addressed in the future about signals that promote SGs
but not hair fate. Surprisingly in this assay, even when we used freshly isolated dermal
fibroblasts from back skin, which normally induce hair follicle formation, the inductive
82
signals from these fibroblasts still induced SG formation in addition to hair follicles.
Thus, it is possible that inductive signals are similar during the development of these
appendages. In fact, Plikus et al. published that overexpression of a BMP signaling
inhibitor, Noggin, resulted in the trans-differentiation of SGs into hair follicles (Plikus et
al., 2004). Since we have shown here as well that BMP signaling is critical for SG
formation, it will be interesting to address how fine tuning of this pathway might
modulate its fate decision in the future.

Taken together, we have explored the role of SG LRCs in SGs and were able to localize
them to the basal layer myoepithelial cells of the proximal acinar region. We were able to
isolate these SG LRCs allowing us to further characterize them and determine their gene
expression profile. Among these genes, a number of BMP signaling genes were identified
and we demonstrated the requirement of this signaling pathway in SG formation. We
propose that SG LRCs are the stem cell population required for the maintenance and
homeostasis of the SG skin appendage. This suggests that at least one distinct stem cell
population exists in the proximal acinar region of SGs, which contains relatively
quiescent cells contributing only to their own glandular structures during homeostasis and
typical wound healing. In fact, our results are in agreement with previously published
observations in human SGs, where sweat ducts were completely or partially injured in the
dermis (Lobitz et al., 1956). Interestingly, they observed that the deep portion of SGs
maintained their quiescence and survived similar to the acinar part containing SG stem
cells in our study. In contrast, the ductal part of human SGs were not able to rebuild the
83
lower acinar part of SGs in vivo, but had instead slowly disappeared (Lobitz et al., 1956).
Moreover, we demonstrated that SG LRCs in the acinar compartment in fact possess
multipotency and stem cell characteristics in vivo having the potential to trans-
differentiate into the epidermis under prolonged isolated wound healing conditions.
Finally, our data also suggest plasticity of SG cells to reconstitute both SGs and hair
follicles in vivo. As an alternative approach, we have also successfully cultured K15
marked SG cells. Although these K15 cells are not specific for SG LRCs, it will be
interesting to test their plasticity and regenerative potential in the future since it includes
slow cycling LRCs.

84
Chapter 3:
Localization, Characterization and Isolation of
Label Retaining Cells (LRCs) from Nails as
Putative New Skin Stem Cells

3.1 Abstract
Due their astonishing abilities and promise in regenerative medicine, scientists have long
been trying to unravel the fundamental mechanisms regulating stem cells. Recognizing
the similarities and differences between different stem cell populations may help identify
common requirements for stem cell maintenance as well as unique aspects controlling
tissue regeneration. Since the identification of hair follicle stem cells in the skin, piles of
data have quickly accumulated in characterizing their regulation and behavior; however,
reports have yet to describe the identification of stem cells in the nail skin appendage.
Given the common origin and localization of nails and hair follicles in the skin, they
presumably share similar features. Therefore, we have used the slow cycling label
retaining property observed in hair follicle stem cells to search for putative stem cells in
nails.

Using a K5-driven tetracycline induced histone2B-GFP system, we have localized a
novel population of label retaining cells (LRCs) in the ventral proximal fold of mouse
85
nails. Through immunofluorescence staining, we demonstrate that these nail LRCs share
some similarities with hair follicle stem cells such as their basal layer localization and
K15 expression. Moreover, lineage tracing experiments have shown that these K15 cells
can contribute to the nail and persist for over 6 months suggesting self-renewal abilities.
Although LRCs are generally quiescent, we have shown that nail LRCs can be activated
in response to injury. Transplanting strips of nail LRCs highlights its potential to
contribute to the nail structure. These abilities to persist throughout the lifetime of an
animal, respond to injury, and contribute to its organ are characteristics of stem cells. In
addition, this H2B-GFP label retaining system allowed the isolation of live nail LRCs for
gene expression profiling, which may help identify new candidate markers in the future.
Finally, we have shown that Bone Morphogenetic Protein (BMP) signaling is also
required for nail differentiation similar to hair follicles. In conclusion, we have identified
a new population of nail LRCs that display stem cell like characteristics; however, more
work is required to demonstrate that these LRCs are bona fide nail stem cells.
Nevertheless, we have presented a new population of putative stem cells that have not
been reported before.

86
3.2 Introduction
Adult stem cells are undifferentiated cells found in various organs throughout the body
capable of self-renewal and responsible for the maintenance and repair of their respective
tissues. Among known adult stem cell populations, hair follicle stem cells in the skin are
one of the best characterized so far. In addition to the hair follicle (HF), the skin contains
a number of other different appendages like sebaceous glands, sweat glands, and nails.
Recently sweat gland stem cells have been discovered (Lu et al., 2012); however, much
less is known about other skin appendages like nails and whether they are maintained by
their own stem cells.

The nail skin appendage contains a hard keratinized structure called the nail plate, which
is composed of terminally differentiated cells called corneocytes. These cells contribute
to the layers of the nail plate, which serves as a protective cover for the digit tip to
prevent trauma to the toes and fingers. In addition, nails are used as a tool to pick up
small objects and are important for fine manipulations and subtle finger functions.
Corneocytes are abundant with hard keratins and are formed during nail differentiation
where nail cells (onychocytes) flatten and condensate, ultimately eliminating its nuclei
and others organelles. Underneath the distal end of the nail plate is the nail bed, which is
sealed together at the fingertip by the hyponychium (Figure 1.5B). The nail bed is
composed of a basal layer and one to two layer of suprabasal postmitotic keratinocytes,
contributing a few horn cells to the undersurface of the distal nail plate (Mecklenburg et
al., 2004). At the “root” of nail plate, below the proximal end, is the nail matrix (Mx),
87
which is composed of actively proliferating cells. Above the matrix lies the keratogenous
zone (KZ) where matrix cells are believed to differentiate, flatten out, die, and deposit
into the overlying nail plate during the nail differentiation process, which is still not fully
understood. However, pulse chase studies in squirrel monkey confirmed that matrix cells
move superficially into the nail plate and distally into the nail bed (Zaias and Alvarez,
1968).

From the top, the proximal end of the nail plate is surrounded by a structure called the
nail proximal fold, which is a continuation of the epidermis that folds inward to cover the
nail plate (Figure 1.5B). At the distal edge of the proximal fold is the eponychium,
responsible for forming superficial layers of the cuticle, which extends onto the nail plate
to seal the top of the nail plate protecting it from toxins and foreign substances.
Interestingly, normal epidermal differentiation is “switched off” just after the proximal
fold folds inward (at the ventral side); thus, nail layers continuing beyond this point (the
remaining proximal fold cells, Mx, KZ, and nail bed) do not differentiate and form the
granular layer typically observed in normal epidermis (Figure 1.5B). Further along as a
morphological continuation of the nail proximal fold are nail matrix cells, which attaches
to each other wrapping around the proximal end of the nail plate (Figure 1.5B). Due to
their adjacent localization, the relationship and distinction between cells in proximal fold
and matrix has not yet been examined or established. In addition, although it is currently
believed that the nail matrix contains progenitor cells responsible for nail differentiation,
it remains unclear whether matrix cells alone possess stem cell characteristics and can
88
generate the whole nail appendage or if there are other cells contributing to nail
differentiation. In general, still little is known about this appendage and how it is
regulated. So far, the mechanism of nail differentiation remains unclear and we don’t
know yet if this structure is maintained by independent skin stem cells.

These are very important questions that might help fill the gaps in our current
understanding of nails. In the future, this knowledge may potentially be applied to treat
patients with nail defects. In humans, there are different types of ectodermal dysplasias
(EDs), which manifest symptoms associated with lack of skin appendages including hair,
teeth, sweat glands as well as nails (Kobielak et al., 2001; Wisniewski et al., 2002). An
example of an autosomal dominant group of ED is pachyonychia congenita (PC), which
is characterized by hypertrophic nail dystrophy accompanied by other ectodermal
changes (OMIM no. 167200 and OMIM no. 167210) (Jackson and Lawler, 1951;
McLean et al., 1995). In addition, nails are also affected in patients with nail-patella
syndrome (NPS, MIM 161200) (Chen et al., 1998; Dreyer et al., 1998) and Witkop
syndrome, which is also known as tooth and nail syndrome (TNS) (Jumlongras et al.,
2001). In addition to EDs, PC, NPS, and TNS, there are also rare autosomal recessive
diseases including isolated anonychia and hyponychia congenita (OMIM 206800) where
the only presenting phenotype is the absence or severe hypoplasia of all fingernails and
toenails (Bergmann et al., 2006; Blaydon et al., 2006). More than two decades ago, a
population of slow cycling label retaining cells (LRCs) has been identified in the hair
follicle skin appendage (Cotsarelis et al., 1990). These LRCs were later characterized to
89
be the bona fide stem cell population responsible for hair follicle homeostasis and
regeneration (Morris and Potten, 1999). This label retaining characteristic has also been
used to identify stem cells in other organs including the corneal limbus and recently
sweat glands (Cotsarelis et al., 1989; Lu et al., 2012). In addition, a recent report
suggested the presence of LRCs in the basal layer of the nail matrix adjacent to the nail
bed in the mouse nail (Nakamura and Ishikawa, 2008). More recently, Sellheyer et al.
used immunohistochemical staining to show that the ventral proximal fold in human nails
expresses hair follicle stem cells markers during embryogenesis suggesting that the
ventral proximal fold may represent the human nail stem cell niche (Sellheyer and
Nelson, 2012). Although the existence of LRCs has been reported in the mouse nail
matrix, its localization did not correlate with the position of putative nail stem cells
identified in human nails (Nakamura and Ishikawa, 2008; Sellheyer and Nelson, 2012).
Therefore, these discrepancies in the location of putative nail stem cells between the
matrix and ventral proximal nail fold need further investigation. In addition, it will be
interesting to address whether nail LRCs localized in the mouse nail matrix possess “per
se” nail stem cell characteristics. However, this recently published identification of nail
LRCs employed BrdU pulse-chase experiments. This BrdU pulse-chase method did not
allow further isolation and characterization of live nail LRCs.

Here, to address these questions, we used a double transgenic keratin 5 (K5) driven,
tetracycline inducible histone 2B green fluorescent protein (K5TetOff/TreH2BGFP)
mouse model to identify and isolate live LRCs from the nail skin appendage (Diamond et
90
al., 2000; Tumbar et al., 2004). This approach was previously used to detect H2B-GFP
marked hair follicle LRCs with slow cycling characteristic in vivo and allowed the
isolation and characterization of live stem cells for the first time (Tumbar et al., 2004).
Therefore, we utilized this same H2B-GFP system for in vivo detection of infrequently
dividing cells in nails to further check their putative stem cell characteristics.

Using this K5TetOff/TreH2BGFP system, we have identified a new population of LRCs
localized in the basal layer of the nail ventral proximal fold. Similar to hair follicle stem
cells, we show that these nail LRCs express keratin 15 (K15) and these K15 cells can
survive and contribute to the nail structure long-term. Moreover, the nail structure can be
regenerated upon removal and the nail LRCs are not lost in the process. When
transplanted, strips of nail LRCs can contribute to the nail structure while some cells still
retain its slow cycling characteristic. With this K5TetOff/TreH2BGFP system, we were
also able to isolate live nail LRCs allowing us to determine its transcriptional gene
expression profile to analyze the common and unique features of these LRCs. Finally,
since Bone Morphogenetic Protein (BMP) signaling have been previously reported to be
important for both hair follicle differentiation and stem cell maintenance, we probed for
the presence of active BMP signaling in the nail and have shown that this signaling
pathway is required for proper nail formation. Understanding the differences and
similarities among different stem cell populations in the same tissue (like skin) is an
important basic science question in stem cell biology. This knowledge might be highly
instructive to study the general mechanisms that underlie stem cells homeostasis as well
91
as stem cell plasticity. This data would also be very helpful in understanding how
different signals in the different stem cell populations determine tissue-specific
regeneration. Moreover, these putative new adult skin stem cells from nails may be useful
in translating these basic discoveries to novel forms of stem cell therapy with applications
in human diseases.

92
3.3 Identification of Nail LRCs in the Ventral Proximal Fold at the Root
of the Nail Plate
We used a transgenic system composed of keratin 5-driven tetracycline repressor mice
(K5-tTA) (Diamond et al., 2000) crossed with tetracycline response element-driven
histone H2B-GFP transgenic mice (pTRE-H2B-GFP) (Tumbar et al., 2004)
(K5TetOff/TreH2BGFP) to identify, localize, and isolate new label retaining cells
(LRCs) in mouse nails. In this system, histone 2B conjugated green fluorescent protein
(H2B-GFP) expression is turned on starting from early embryogenesis in K5 specific
cells efficiently labeling the entire adult skin and its appendages, including nails, hair
follicles, sweat glands, and the epidermis with H2B-GFP (Figure 3.1A, D, G and J). By
feeding these animals 1 mg/g doxycycline food, H2B-GFP expression is turned off and
existing H2B-GFP labels are diluted out with each cell division. In this study, we
performed a 4 week “chase” with doxycycline treatment to turn off H2B-GFP expression
where rapidly dividing cells dilute out the existing H2B-GFP label and slow cycling
infrequently dividing cells retain the H2B-GFP label. After 4 weeks of “chase”, we
identified a population of LRCs surrounding the proximal nail plate in digits sectioned
along the proximal-distal axis (Figure 3.1B and E, arrows, top view). Analyzing sections
from the dorsal-ventral axis parallel to the finger, this population of nail LRCs was found
above the nail plate at the ventral side of the proximal fold, which is a continuation of the
epidermis folded inwards above the proximal end of the nail plate (Figure 3.1H). Since
the mouse nail exists as a 3-dimensional near conical structure, a cross section from the
side often reveals 2 clusters of LRCs: one directly in the ventral proximal fold above the
93
nail plate structure (upper LRCs) (Figure 3.1H, side) and one on the lower edge from the
proximal fold which appears to be localized below the nail plate (lower LRCs) (Figure
3.1K, side, see as well Figure 3.2D) as this layer of LRCs wrap around the exterior
proximal end of the nail plate. As an internal control, we observed restricted H2B-GFP
expression in hair follicle stem cells after doxycycline treatment (Figure 3.1E, bulge). For
better topological localization around the proximal nail plate structure, we performed
hematoxylin and eosin (H&E) staining on the same sections (Figure 3.1C and F, top view
and Figure 3.1I and L, side view).
94

Figure 3.1. Localization of nail LRCs in the ventral proximal fold. (A and D) Top
view sections before and (B and E) after 4 weeks of chase with doxycycline identifying a
population of H2B-GFP marked LRCs surrounding the nail structure. (C and F)
Corresponding H&E stainings. (G and J) Side view sections before and (H) after 4 weeks
of chase with doxycycline demonstrating the presence of LRCs in the ventral proximal
fold. (K) As the strip of LRCs wrap around the nail, some sections show LRCs at the
lower edge of the nail. (I and L) Corresponding H&E stainings. Dotted white lines
indicate dermal-epidermal junctions.

95
3.4 Characterizing the Specific Localization and Organization of LRCs
in Nails
Since nails exist as 3-dimensional (3D) structures, we would like to know how these
LRCs are organized in this appendage. Here, we observed that LRCs form a “semi-ring
shape” surrounding the nail plate when visualized from the top view (proximal-distal)
sections (Figure 3.1B and E). However, from the side view (dorsal-ventral), nail LRCs
were found at the proximal end of the nail plate (Figure 3.1H and K; Figure 3.2D).
Therefore, we used confocal microscopy (Zeiss LSM5 confocal microscope) with serial
Z-stack 3D re-construction to visualize the 3D organization of these nail LRCs. In order
to expose the nail LRCs for imaging, the epidermis above the ventral proximal nail fold
was carefully cut open and peeled apart to directly reveal and visualize underlying LRCs.
At the ventral proximal fold, we observed multiple rows of LRCs organized in a
concaved fashion around the exterior of the nail plate (Figure 3.2A) after a 3 week chase
with doxycycline. This confirms our results from nail sections that nail LRCs wrap
around the outside of the nail plate structure and are indeed localized in the ventral
proximal fold. To determine the precise topological localization of LRCs, we performed
immunofluorescence staining with a number of different markers. Since hair follicle
LRCs are localized to the basal layer, we stained 4 week chased nail LRCs with basal
layer marker keratin 5 (K5) and confirmed that nail LRCs are also localized to the basal
layer (Figure 3.2B). Furthermore, we confirmed that these basal layer LRCs are attached
to the β4 integrin marked basement membrane (Figure 3.2C). Unlike the epidermis, the
nail structure does not have a stratum granulosum. Although the proximal nail fold is a
96
continuation of the epidermis containing a stratum granulosum, the granular layer fades
away in the ventral proximal fold. A previous report have defined this border to be the
boundary between the proximal nail fold and nail matrix where the granular layer is
present in the proximal fold and absent in the matrix (Zaias and Alvarez, 1968). To more
specifically localize nail LRCs, we stained 4 week chased nails with loricrin and found
the LRCs at the region where the granular layer fades away (Figure 3.2D and E). Looking
at p63, a protein important in the self-renewal and proper development of skin, we show
that nail LRCs express p63 similar to the basal layer of the epidermis and sweat gland
LRCs; however, p63 expression was also detected in the nail matrix and is therefore not
specific for nail LRCs (Figure 3.2F).

Figure 3.2. Nail LRCs are localized in the basal layer at the granular layer border.
(A) 3D reconstruction of nail LRCs after 3 weeks of chase with doxycycline. (B) Nail
LRCs are localized to the K5 expressing basal layer (C) attached to the β4 integrin
marked basement membrane. (D) Nail LRCs are found at the border where loricrin
expression ends. (E) Magnification of panel D. (F) Nail LRCs co-express p63.

97
Previously, hair follicle stem cells have been shown to specifically express keratin 15
(K15) in the bulge region; therefore, we stained nail LRCs for K15 and demonstrated that
they are also positive for K15 expression (Figure 3.3A). Subsequently, we used a
K15CrePR/Rosa26LacZ transgenic mouse line to label these nail LRCs and perform
lineage tracing analysis in vivo (Figure 3.3B). This K15 promoter specifically targeted
nail LRCs and contributed to the nail structure for more than 6 months after transgene
activation (similar to previously described hair follicles, see Figure 3.3B). In summary,
we have identified a population of K15 positive nail LRCs at the basal layer of the ventral
proximal fold where the epidermal granular layer fades away (Figure 3.3C).

Figure 3.3. Nail LRCs express K15 and K15 marked cells contribute long-term to
the nail. (A) Similar to hair follicle stem cells, nail LRCs express K15. (B) Using a
K15CrePR/R26LacZ system, we show that these K15 expressing cells survive long-term
for more than 6 months contributing to the nail structure. (C) Schematic summarizing
markers expressed.

As a continuation of the proximal fold beyond the LRCs, we have the nail matrix
wrapping around the proximal end of the nail plate. The matrix consists of actively
proliferating progenitor cells that eventually differentiate into cells of the keratogenous
zone to form the nail plate. Previous reports have suggested the presence of LRCs in the
nail matrix (Nakamura and Ishikawa, 2008). However, we observed a clear distinction
98
between our slow cycling nail LRCs from the actively dividing matrix cells with as little
as a 1 week chase with doxycycline (Figure 3.4A). Staining with Ki67 to mark actively
proliferating matrix cells, we observed a population of weakly positive H2B-GFP cells
expressing Ki67 directly adjacent to our bright H2B-GFP+, Ki67 negative, nail LRCs
(Figure 3.4A, arrows). This Ki67 co-localization with weakly positive H2B-GFP
expression illustrates the dilution of the H2B-GFP label with cell division. Further along
the matrix away from the nail LRCs, we show an abundance of Ki67+ matrix cells no
longer expressing H2B-GFP, suggesting that these cells have undergone more rounds of
proliferation than the weakly expressing H2B-GFP cells close to the nail LRCs.
Analyzing the nail after a 2 week chase, the H2B-GFP label has been further diluted in
the matrix while the LRCs remain brightly H2B-GFP+ (Figure 3.4B).

Figure 3.4. The nail matrix contains actively proliferating cells, not LRCs. (A) H2B-
GFP label becomes progressively diluted out of proliferating matrix cells with a 1 week
chase with doxycycline and (B) 2 weeks chase with doxycycline. Arrows denote H2B-
GFP co-localization with Ki67.

99
3.5 Dissection and Isolation Strategy for H2B-GFP positive LRCs from
Nails
To analyze the functional properties of nail LRCs, we first needed to develop methods to
isolate them without contamination from other LRCs of surrounding tissues. We achieved
this through surgical microdissections with subsequent enzymatic digestion. The
epidermis on the dorsal surface of each finger was cut open along the finger and pulled
apart to reveal the strip of nail LRCs at the ventral proximal fold (Figure 3.5A and B).
Next, nail LRCs were isolated through careful dissection to obtain a strip of cells
enriched in H2B-GFP marked nail LRCs (Figure 3.5C). These nail LRC strips were
collected and digested in 0.25% Trypsin-EDTA overnight at 4°C to obtain a single cell
suspension of nail LRCs (Figure 3.5D). The next day, the trypsin was neutralized and the
cells filtrated for fluorescence activated cell sorting (FACS) (Figure 3.6A).

Figure 3.5. Isolation of live nail LRCs. (A) The epidermis is cut open along proximal
fold of 4 week chased mice to reveal the (B) underlying nail LRCs. (C) Strips of nail
LRCs are isolated through microdissection and (D) trypsinized to obtain a single cell
suspension for FACS.

Since we demonstrated that nail LRCs were attached to the basement membrane, we
stained these cells with a PE-conjugated FACS specific antibody against α6 integrin
100
(CD49f) to purify these live LRCs together with adjacent basal cells (DAPI staining was
used to sort live cells, data not shown). The GFP+ α6+ double positive fraction of nail
LRCs and the GFP- α6+ fraction of surrounding basal cells were collected (Figure 3.6A).
The specific gates to sort the double positive nail LRCs were setup according to negative
(unstained), single GFP positive, and single α6 integrin-PE positive stained control cells
(not shown). We have collected GFP+ α6+ double and GFP- α6+ single positive sorted
cells for RNA isolation and microarray analyses. A small portion of these cells have been
taken to check the post-sorting purity where we confirmed by cytospin analysis that
almost all double positive sorted nail LRCs expressed H2B-GFP (data not shown).

101
3.6 Identifying the Transcriptional Profile of Nail LRCs via Microarray
Analysis
The GFP+α6+ nail LRCs and GFP- α6+ non-LRCs were collected for microarray
analyses. In addition, we have isolated interfollicular epidermal basal layer cells from the
sole’s epidermis where α6+ cells were collected as a control. Total RNA from two
separate double positive populations (GFP+/ α6+, nail LRCs) and two single positive
(GFP-/ α6+, adjacent basal cells) from independent experiments were extracted and
compared to the α6+ basal layer of the sole’s epidermis to identify the nail gene
expression profile (Figure 3.6B). In this analysis, 665 genes were consistently up-
regulated and 859 genes consistently down-regulated in the nail LRCs by at least 2-fold
in the 2 independently isolated nail LRCs fractions. Similarly, 505 genes were
consistently up-regulated and 709 genes were down-regulated in the non-LRCs. Next, we
compared the gene expression profile of the nail LRCs with the surrounding non-LRCs
and found 171 genes to be commonly up-regulated and 492 genes commonly down-
regulated (Figure 3.6C).
102

Figure 3.6. Purifying live nail LRCs for gene expression profiling. (A) GFP+ α6+ Nail
LRCs and GFP- α6+ surrounding non-LRCs were purified through FACS. (B) Nail LRCs
and non-LRCs were then independently compared to the sole’s epidermis to identify
genes expressed in the nail unit. (C) Gene expression profiles of nail LRCs and non-
LRCs were then compared to each other.

The genes identified were categorized based on function including signaling,
transcription, cell adhesion, and keratinization using the DAVID software (Table 3.1).
From the microarray analysis, we unexpectedly observed an up-regulation of the
epidermal granular layer marker, loricrin, in both LRCs and non-LRCs. Looking more
closely at the loricrin staining, we observed a co-localization of some LRCs with loricrin
especially at the border where loricrin expression disappears (Figure 3.2E, arrows). This
explains the expression of loricrin identified in the microarray since loricrin expression is
absent in the basal layer of the epidermis, which was used as the baseline. Among the
common genes between nail LRCs and non-LRCs, we identified 2 Bone Morphogenetic
Protein (BMP) signaling inhibitors, Bambi and Decorin, to be consistently down-
regulated in the nail when compared to the sole’s epidermis suggesting that BMP
signaling may be active in the nail skin appendage (Table 3.1).
103

Table 3.1. Categorization of the nail gene expression profiles based on function.
Gene expression profiles consistently found in two independent microarray analyses from
independent biological samples were categorized based on function. Genes in green are
down-regulated by at least 2-fold.

104
3.7 BMP Signaling is Required for Proper Nail Formation
In addition to the down-regulation of BMP inhibitors identified in the gene expression
analysis, BMP signaling is also required for proper hair follicle differentiation (Andl et
al., 2004; Kobielak et al., 2003) and sweat gland development (see data in Chapter 2).
Therefore, we wanted to see if it is also important in the nail skin appendage. To check
whether BMP signaling is found in the nail, we stained 4 week chased
K5TetOff/TreH2BGFP nail sections for phospho-smad 1/5/8, which indicates active
BMP signaling, and observed positive expression in nail LRCs as well as the nail matrix
(Figure 3.7A). Using a keratin14 (K14) driven Cre recombinase (K14Cre), which is
activated during early morphogenesis at E10, crossed onto a Rosa26LacZ reporter
(K14Cre/R26LacZ) mouse line to mark common skin progenitors cells of the basal layer
and its progeny, including but not limited to nail LRCs, we show that K14 expressing
cells can contribute to the entire nail appendage during morphogenesis as labeled by
LacZ (blue) (Figure 3.7B). To study the role of BMP signaling in nail differentiation
during morphogenesis, we ablated Bmpr1a during embryogenesis in the basal layer of the
skin using K14Cre. In KO animals, the nail matrix, keratogenous zone, nail bed, and nail
plate appears to be affected (Figure 3.7E and F) when compared to control animals
(Figure 3.7C and D). The keratogenous zone above the matrix, where cells flatten and
differentiate into cells of the nail plate, appears to be absent in Bmpr1a deficient nails
(Figure 3.7F). In addition, we observed hyperplasia of the nail bed underneath the nail
plate at the distal end in KO animals (Figure 3.7E, vs control Figure 3.7C, insets). The
nail plates in Bmpr1a KOs (Figure 3.7E) are thinner and appear to be improperly
105
differentiated when compared to the controls (Figure 3.7C). In addition, the tip of the nail
plate beyond the hyponychium is absent or broken in Bmpr1a KO nails (Figure 3.7E,
arrow, vs control Figure 3.7C, arrow). Finally, the stratum granulosum of the epidermis
invaginates into the proximal fold and stops there in control animals (Figure 3.7D);
however, it appears to ectopically extend through the entire nail of KO animals (Figure
3.7F).

Figure 3.7. BMP signaling is required for proper nail development. (A) Active BMP
signaling is found in nail LRCs and the nail matrix indicated by positive phospho-
Smad1/5/8 expression. (B) The entire nail structure arises from K14 progenitor cells. (C)
H&E staining of a Bmpr1a/K14Cre/K14-H2BGFP heterozygous nail at P8, inset shows
magnification of nail bed. (D) 20x magnification of the P8 Bmpr1a Het nail. (E)
Corresponding Bmpr1a KO nails display abnormal nail plate, lack of keratogenous zone,
hyperplasia of the nail bed (inset), and a broken nail tip beyond the hyponychium, arrow.
(F) KO nail highlighting the absence of the keratogenous zone and an ectopic granular
layer underneath the nail plate, arrow.

Looking at H&E stainings of P8 nails, the matrix region is smaller in KO animals when
compared to the control. Staining for Ki67, which marks proliferative cells in the matrix,
106
we demonstrate that KO animals contain fewer proliferative matrix cells than controls
(Figure 3.8A and B). Finally, we show that Bmpr1a KO mice lack AE13 expression in
the nail plate found in control mice confirming that the nail plate has been compromised
(Figure 3.8C and D).

Figure 3.8. The nail plate differentiation is compromised in Bmpr1a KO mice. (A)
Ki67 staining marks actively proliferating cells in the matrix of control animals. (B)
Decreased numbers of Ki67+ proliferating cells in Bmpr1a KO mice. (C) AE13 is a
marker of the nail plate. (D) AE13 expression is absent in the Bmpr1a KO nail plate.

To further characterize these Bmpr1a deficient nails, we stained for some epidermal
markers, namely K5, K1, and loricrin. In control animals, K1 and loricrin expression was
detected in the surrounding epidermis including the proximal fold and foot pad. In
addition, positive K1 and loricrin staining was observed in the ventral nail region, but
was absent in the dorsal nail plate (Figure 3.9A, C, and E). In Bmpr1a KO mice, K1 and
loricrin was ectopically expressed in the dorsal nail (Figure 3.9B, D, and F, arrows)
107
suggesting that the nail has adopted the phenotype of an epidermis. This ectopic loricrin
expression, which marks the granular layer, confirms that the stratum granulosum have
extended through the nail instead of stopping at the proximal fold.

Figure 3.9. K1 and Loricrin epidermal markers are ectopically expressed in nail in
the absence of BMP signaling. (A and C) Positive K1 staining is found in the proximal
fold and ventral nail of control mice. (B and D) Ectopic K1 expression is observed in the
Bmpr1a KO nail plate, arrows. (E) Loricrin expression is found in the proximal fold and
ventral nail, and is absent in the dorsal nail plate of control mice. (F) Ectopic loricrin
expression is observed in the Bmpr1a KO nail plate, arrow.

108
3.8 Quiescent Nail LRCs can become Activated Upon Nail Removal
To determine whether nail LRCs are removed with the nail structure, we sectioned 4
week chased K5TetOff/TreH2BGFP paws upon nail removal and demonstrated that some
nail LRCs remain attached to the finger upon nail removal (Figure 3.10A and B).
Corresponding H&E staining show topological localization (Figure 3.10C and D).

Figure 3.10. Nail LRCs are not lost upon nail removal. (A) H2B-GFP marked nail
LRCs are still found in the finger after nail removal. (B) Magnification of panel A. (C
and D) Corresponding H&E stainings after nail removal.

Next, we checked the proliferation status of LRCs in normal nail homeostasis and after
injury. We demonstrated that indeed under normal conditions, nail LRCs do not overlap
with actively proliferating cells (24h after BrdU pulse) in the nail matrix (Figure 3.11A).
However, at 24h after nail removal, normally quiescent nail LRCs started to proliferate as
marked by Ki67 (Figure 3.11B) illustrating that LRCs in the nails are not post-mitotic
109
cells and may become activated in response to injury. To test the regenerative potential of
the nail skin appendage, we removed the keratinized nail structure from mouse paws and
demonstrated that the nail can be fully regenerated within 2 weeks (Figure 3.11C and D).

Figure 3.11. Normally quiescent nail LRCs can be activated upon injury. (A) Nail
LRCs are normally quiescent while the nail matrix contains actively proliferating cells as
marked by BrdU incorporation after BrdU pulse. (B) Upon nail removal, these LRCs
become activated as indicated by Ki67 co-expression. (C) Upon nail removal, (D) the nail
appendage can be regenerated within 2 weeks.

110
3.9 Assessment of Nail LRCs Differentiation Capabilities after
Engraftment in vivo.
To test whether nail LRCs can differentiate into nail cells, we isolated H2B-GFP labeled
nail LRC strips from 4 week chased animals and transplanted them underneath the
proximal fold of immunocompromised nod scid mice (Figure 3.12A and B). About 2
weeks after transplantation, we show that these H2B-GFP labeled cells have contributed
to the hard nail structure in multiple transplants for up to 3 weeks (Figure 3.12C and D).
To determine whether LRCs still existed in this transplant, we performed a 22day chase
with doxycycline treatment on the transplanted nod scid mouse. Weakly positive H2B-
GFP cells were observed through the epidermis of the proximal fold (Figure 3.12E) and
to confirm that there are indeed nail LRCs remaining, we sectioned this chased transplant
and found H2B-GFP positive nail LRCs underneath the proximal fold of the nod scid
mouse where they were placed (Figure 3.12F). These transplantation assays demonstrated
that these strips of cells enriched in H2B-GFP labeled nail LRCs can contribute to the
nail skin appendage. Moreover, we show that although many cells undergo cell division
to contribute to the nail as shown by the dilution of the H2B-GFP label during the 22day
chase, some cells remains quiescent (Figure 3.12F).

111

Figure 3.12. Transplanted nail LRCs can contribute to the nail structure. (A) Strips
of cells possessing nail LRCs are isolated and transplanted underneath the proximal fold
of immunocompromised nod scid mice. (B) H2B-GFP marked nail cells contributing to
the nail structure at 10 days (C) 17 days and (D) 24 days after transplantation. (E) 22 day
chased transplant and (F) section demonstrating the presence of LRCs remaining from the
transplant.

112
3.10 Discussion
Various pulse-chase methods have been used to identify populations of infrequently
dividing LRCs as putative stem cells in different organs. Among the different organs,
clusters of LRCs have been identified in the hair follicle bulge region as stem cells of this
skin appendage (Cotsarelis et al., 1990). More recently, new methods of pulse-chase
experiments, namely the K5TetOff/TreH2BGFP system, have been used to isolate and
characterize live hair follicle label retaining stem cells for the first time.

Hair follicle stem cells are one of the most well characterized adult stem cell populations,
but little is known about the presence of stem cells in other skin appendages such as the
nail. Although the existence of LRCs in the nail matrix has been recently reported in
mice, further isolation and characterization of these cells have not yet been addressed. In
addition, whether these LRCs possess stem cells characteristics have not been
investigated or established. Thus, we used the K5TetOff/TreH2BGFP system in
combination with K15CrePR/R26LacZ genetic lineage tracing to address these questions,
probe for the presence of LRCs and trace K15 labeled progenies as putative stem cells in
the nail.

The ventral proximal fold contains slow cycling LRCs while the nail matrix contains
highly proliferative cells.
In recent years, studies have reported the presence of LRCs in the sweat gland and nail
skin appendages (Lu et al., 2012; Nakamura and Ishikawa, 2008; Nakamura and Tokura,
113
2009). For the nail, one report has identified LRCs in the basal layer of the mouse nail
matrix using BrdU pulse chase experiments (Nakamura and Ishikawa, 2008); however,
we did not observe any LRCs in the matrix after 4 weeks of chase. Instead, we found a
cluster of LRCs in the basal layer of the ventral proximal fold (Figure 3.1H). In further
analysis, we have shown that the nail matrix contains highly proliferative cells as marked
by Ki67, BrdU incorporation, and the rapid dilution of the H2B-GFP label with a 1 and 2
week chase when compared to the proximal fold LRCs (Figure 3.4A and B). These
discrepancies regarding LRCs localization (in our results) could be explained by the
different methods and time points used for labeling. In the previous report, BrdU pulse-
chase experiments were performed in adult mice; therefore, it is very likely that the
proximal fold LRCs found in our studies were not targeted and labeled properly by the
BrdU pulse due to its slow cycling nature. In contrast, the matrix contained actively
proliferating cells allowing incorporation of BrdU during pulse experiments leading to
the localization of LRCs in the matrix. In our study, we used the K5TetOff/TreH2BGFP
system where all nail cells were uniformly marked by tetracycline regulated H2B-GFP
expression from embryogenesis (Figure 3.1A). Moreover, this system did not require the
proliferation of cells for label incorporation allowing us to probe for the label retaining
property in all K5 targeted cells.

Using the K5TetOff/TreH2BGFP system, we did not observe any overlap between LRCs
and the Ki67+ proliferative nail matrix after 4 weeks of chase (LRCs were distinctly
separate from matrix cells). By performing shorter chases, we have demonstrated that the
114
nail matrix contains rapidly proliferating cells as marked by Ki67 in addition to weakly
H2B-GFP positive cells resulting from the dilution of the H2B-GFP label upon cell
division after a 1 week chase with doxycycline (Figure 3.4A). Furthermore, we have
demonstrated that these matrix cells continue to divide as the H2B-GFP label becomes
even more diluted and difficult to detect in the matrix after a 2 weeks chase (Figure
3.4B). Looking more closely, we have noticed that the nail LRCs closer to the matrix
contains weaker H2B-GFP expression suggesting that they have divided more than the
LRCs, but less than that of matrix cells no longer expressing H2B-GFP (Figure 3.4A). It
is tempting to speculate that these weaker LRCs closer to the matrix have divided more in
order to fuel the rapidly proliferating transit amplifying matrix cells that will eventually
further differentiate into cells of the nail plate. Since the nail is a continuously growing
appendage unlike the cycling hair follicle, it is expected that the H2B-GFP label of the
slow cycling LRCs will become diluted out faster as these putative stem cells divide to
fuel nail differentiation. In fact, we observed that these H2B-GFP LRCs become
increasingly difficult to locate with longer chases beyond 6-8 weeks (data not shown).
Such turnover of these nail LRCs also shows that these nail LRCs are not post-mitotic
cells.

Recently, in agreement to our study, the ventral proximal fold was proposed to be the nail
stem cell niche in the developing embryonic human nail based on the expression of bulge
stem cell markers: CK15, CK19, and PHLDA1. In addition, it has been demonstrated that
the human nail matrix contains highly proliferative cells marked by Ki67 similar to the
115
mouse nail matrix and hair follicle matrix, whereas the ventral proximal fold contains
very few Ki67+ cells making them more quiescent similar to hair follicle stem cells. This
report further supports our findings of nail LRCs as putative stem cells in the ventral
proximal fold. Therefore, for the first time, we have identified, characterized, and isolated
this unique population of slow-cycling LRCs at the ventral proximal fold consistent with
the postulated localization of nail stem cells in humans.

BMP signaling is required for proper nail formation similar to hair follicle
differentiation.
This K5TetOff/TreH2BGFP system has also permitted the isolation of live nail LRCs
allowing us to determine their gene expression profile. From the microarray analysis, we
observed a down-regulation of two BMP signaling inhibitors, Bambi and Decorin. In hair
follicles, it was previously demonstrated that BMP signaling is important in maintaining
quiescent stem cell homeostasis and regulating the differentiation of all hair layers
required for the regeneration of this appendage. More specifically, previous studies have
shown that deletion of Bmpr1a in the basal layer of the skin results in a failure to form
the inner root sheath and hair shaft demonstrating that BMP signaling is required for hair
follicle differentiation. In addition, previous studies have reported visually abnormal nails
in BMP deficient mice; however, its specific defects were not characterized (Andl et al.,
2004; Yuhki et al., 2004). Similar to the hair follicle and consistent with previous reports,
our detailed analysis revealed that BMP signaling is also required for proper nail
differentiation by ablating Bmpr1a in the basal layer using a K14Cre recombinase.
116
Moreover, it has been previously shown that Msx2 and Foxn1, both downstream targets
of BMP signaling, regulate nail differentiation (Cai and Ma, 2011). Compound mutations
of Msx2 and Foxn1 resulted in more severe nail abnormalities than the individual KO
lines and it has been hypothesized that an ablation of BMP signaling will yield similar
effects. Similar to the Msx2 and Foxn1 double mutant, we also observed hyperplastic nail
beds, broken nails at the tip beyond the hyponychium, and irregular nail plates in Bmpr1a
KO mice. In addition, the keratogenous zone appears to be absent or reduced in Bmpr1a
deficient nails making its phenotype more severe than the Msx2 and Foxn1 double
mutant.

Moreover, we have demonstrated that the Bmpr1a KO nail plate is compromised lacking
proper AE13 expression (Figure 3.8D). Instead, we found ectopic K1 and loricrin
expression in the dorsal nail where the nail plate should be (Figure 3.9B, D and F). In
addition, similar to Foxn1 and Hoxc13 single mutants (Godwin and Capecchi, 1998;
Mecklenburg et al., 2004), the stratum granulosum of the epidermis have extended into
the nail structure of Bmpr1a KO mice instead of stopping at the proximal fold (Figure
3.7F). This was confirmed by the ectopic loricrin expression in the KO nail (Figure 3.9F).
The ectopic expression of epidermal markers demonstrates that BMP signaling may be
required for the nail fate and without it the nail structure acquires an epidermal like
phenotype.

117

Nail LRCs can contribute to the nail structure, self-renew, and become activated in
response to injury.
Similar to the human ventral proximal nail fold as well as hair follicle stem cells, we
show that these nail LRCs also expresses K15 in the mouse nail (Figure 3.3A). Thus, by
using a K15CrePR/R26LacZ lineage tracing system we were able to mark K15
expressing cells and trace its progeny. In these animals, we observed K15 targeted LacZ
expression in the ventral proximal fold contributing to layers of the nail plate for more
than 6 months after initial transgene activation in addition to the LacZ marked hair
follicle stem cells (Figure 3.3B). This demonstrates that these K15 expressing cells are
capable of self renewal and can survive long-term. To test the regenerative ability of the
nail appendage, we removed the keratinized nail structure and have shown that the mouse
nail can be fully regenerated (Figure 3.11D). In order to see whether nail LRCs remained
in the finger after nail removal, we sectioned samples of 4 week chased nails after nail
removal and found weakly positive H2B-GFP LRCs left in the finger (Figure 3.10A and
B). Although these nail LRCs are generally quiescent (Figure 3.11A), they can be
activated in the event of nail removal as indicated by positive Ki67 staining (Figure
3.11B) illustrating that they can respond to injury and further confirms that they are not
post-mitotic cells. By transplanting strips of dissected H2B-GFP marked nail LRCs
underneath the proximal fold of an unmarked host, we further demonstrate that these nail
LRCs can contribute to the nail structure (Figure 3.12C and D). Furthermore, we have
shown that LRCs still persist in these transplants after 3 weeks of chase with doxycycline
118
(Figure 3.12F). Taken together, we show that these K15 expressing nail LRCs in the
ventral proximal fold is capable of self renewal, responding to nail injury, and
contributing to the nail structure. These are all characteristics of stem cells supporting our
hypothesis that these newly identified LRCs are putative nail stem cells.

In conclusion, we have identified a novel population of LRCs in the basal layer of the
ventral proximal nail fold. Similar to hair follicle stem cells, these nail LRCs express K15
and is capable of self renewal and contributing to the nail structure. In addition, we have
determined its gene expression profile and demonstrated the requirement of BMP
signaling in nail differentiation. Taken together, we propose that these LRCs localized in
the proximal fold are putative nail stem cells similar to the recently proposed nail stem
cells in humans. Further knowledge about nail stem cells should be helpful to develop
cell-based therapy for esthetical and functional regenerations of this appendage in human.

119
Chapter 4:
Comparison of Hair Follicle, Sweat Gland, and
Nail Skin Appendages

4.1 Dermal-Epidermal Interactions are Important in the Development
of Skin Appendages.
Hair follicle, sweat gland, and nail skin appendages all arise from the ectoderm during
morphogenesis. All three appendages begin their development as invaginations of the
basal layer during embryogenesis and continue its development postnatally. More
specifically, we demonstrate that the nail develops from K14 marked progenitors through
lineage tracing, which has been previously reported to also be the common precursor for
the hair follicle and sweat gland skin appendages (Lu et al., 2012; Vasioukhin et al.,
1999). Moreover, these appendages may potentially also be regulated by reciprocal
interactions with dermal signals such as those secreted by the dermal papilla in regulating
hair follicle regeneration. Staining with alkaline phosphatase (AP) strongly marks the
dermal papilla of hair follicles (Figure 4.1A); however, its role in the dermal papilla
remains unclear (Handjiski et al., 1994). Similar to hair follicles, the underlying dermis of
sweat glands and nails are also rich in AP positive cells, but the pattern of AP staining is
more diffuse in the sweat gland and nail appendages (Figure 4.1B and C). Nevertheless,
further studies are required to elucidate the role of the underlying mesenchyme in
regulating the nail or sweat gland appendages.
120

Figure 4.1 Alkaline phosphatase activity is found in the dermis of sweat gland and
nail skin appendages. (A) The hair follicle dermal papilla can be identified through
positive AP staining. (B) Positive AP staining is also observed in the dermis directing
surrounding sweat glands. (C) Detection of AP activity in the underlying dermis of the
nail.

121
4.2 LRCs in Skin Appendages
Using a K5 driven tetracycline regulated H2B-GFP label retaining system, we have
identified populations of slow cycling LRCs in all 3 appendages. After 4 weeks of chase
with doxycycline, LRCs were found in the hair follicle bulge (Tumbar et al., 2004),
myoepithelial cells of the SG acinar basal layer described in chapter 2 (Lu et al., 2012),
and ventral proximal nail fold. Moreover, these distinct LRCs populations are all
commonly localized to the K5 expressing basal layer attached to the basement membrane
of their respective appendages. However, sweat gland LRCs are distributed in a scattered
fashion whereas nail and hair follicle LRCs are clustered together. Similar to hair follicle
stem cells, sweat gland and nail LRCs also express K15. In sweat glands, however, K15
expression is not restricted to LRCs, but is also found in the non-LRCs population of
sweat glands. On the other hand, p63 expression was found to be specific for sweat gland
LRCs, but not nail LRCs where its expression was also found in the matrix.
122

Hair Follicles Sweat Glands Nails
K14 progenitors (+) (+) (+)
LRCs Location Bulge Myoepithelial cells Ventral Proximal
Fold
LRCs organization Clustered Scattered Clustered
Appendage
dynamics
Cyclic Regeneration Extremely low
turnover
Continuous growth
LRCs turnover Moderate Slowest Fastest
LRC specific K15+ p63+ K15+
Markers expressed K15+, K5+, β4+ K15+, K5+, β4+,
p63+, SMA, Smad2
K15+, K5+, β4+,
p63+, Smad1/5/8
Proliferative TA
cells/ Matrix
Yes Quiescent Yes
Bmpr1a LOF
(K14Cre-driven)
No terminal
differentiation
No morphogenesis No terminal
differentiation
LRCs in epidermal
homeostasis?
No No N/A
LRCs in wound
healing?
Yes Yes N/A
Table 4.1. Comparison of hair follicle, sweat gland, and nail skin appendages.
Similarities and differences in LRCs characteristic and behavior between the different
skin appendages. Abbreviations: LOF – loss of function; N/A – not applicable

The hair follicle undergoes cycles of growth, degeneration, and quiescence while nails
undergoes continuous growth. Accordingly, we observed a faster turnover of LRCs in the
nails when compared to hair follicles. With longer chases, it becomes increasingly
difficult to detect H2B-GFP marked LRCs in the proximal nail fold after 6-8 weeks of
chase, whereas hair follicle LRCs still brightly express H2B-GFP in the bulge. While hair
follicles grow periodically and nails grow continuously under normal homeostasis in
adults, sweat gland growth is complete once morphogenesis is finished. The main
function of sweat glands is to secrete water and electrolytes and it does so in a manner
where sweat gland cells do not have to undergo massive apoptosis. Unlike the growing
123
hair follicles and nails that get trimmed or the continuous turnover of the epidermis that
gets sloughed off at the outermost layer, sweat glands require little cellular turnover.
Consistent with this, we observed H2B-GFP marked sweat gland LRCs beyond 20 weeks
of chase. Among these three skin appendages, nail LRCs have the fastest turnover while
sweat glands are the most quiescent.

124
4.3 Sweat Glands and Hair Follicles Contain Multiple Stem Cell
Populations and can Both Contribute to Epidermal Wound Healing
The hair follicle and sweat gland skin appendages are both abundantly dispersed over the
human body, whereas nails are only found at the digit tips. Consequently, sweat glands
and nails have greater implications in epidermal wound healing. Here, we have shown
that K15 marked sweat gland cells do not contribute to the normal homeostasis of the
epidermis similar to K15 marked hair follicle stem cells. However, under wound healing
conditions, we have shown that sweat gland LRCs have the potential to differentiate into
cells of the various epidermal layers similar to hair follicle stem cells. Given that sweat
gland LRCs reside deeper in the dermis and only contributed to the epidermis under
favorable conditions, more readily available hair follicle label retaining stem cells appear
to be the preferred choice for wound healing. This epidermal wound healing potential
remains to be tested in nail LRCs although its interest is diminished due to its limited
localization at the digits tips.

Although the hair follicle bulge contains the stem cells required for hair follicle growth
and regeneration, other populations of stem cells have been reported in the hair follicle
isthmus (Jaks et al., 2010). Distinct from the K15 and Lgr5 marked bulge stem cells that
forms the cycling portion of the hair follicle, there are 2 stem cells populations in the
isthmus that maintains the isthmus and infundibulum permanent portions of the upper
hair follicle. Lrig1 and MTS24 marks stem cells in the upper isthmus and is responsible
for the maintenance of the infundibulum (Jensen et al., 2009; Nijhof et al., 2006). More
125
recently, scientists have discovered a Lgr6+ population in the lower isthmus. In addition
to maintaining the infundibulum, it also contributes to the sebaceous gland and
interfollicular epidermis under normal homeostasis (Snippert et al., 2010). Upon injury,
all lineage traced hair follicle stem cells can contribute to wound healing and all of them
can form the entire hair follicle in reconstitution assays (Jaks et al., 2010; Jensen et al.,
2009; Morris et al., 2004; Nijhof et al., 2006; Snippert et al., 2010). Similar to hair
follicles, recent studies have suggested multiple populations of stem cells in the sweat
glands (Lu et al., 2012). Analogous to the hair follicle isthmus, it was suggested that the
sweat duct also contains cells with stem cell characteristics. Lu et al. have reported that
the sweat duct can contribute to epidermal wound healing, be propagated in culture, and
reconstitute sweat gland structures upon transplantation (Lu et al., 2012).

Figure 4.2. Different stem cell populations in hair follicles and sweat glands. (A)
Multiple stem cell populations have been reported in the hair follicle. Stem cells in the
bulge, including the LRCs, K15+, and Lgr5+ stem cells, are responsible for regenerating
the cyclic portion of the hair follicle and differentiation of the hair shaft. Meanwhile,
stem cells found in the isthmus is reported to maintain the upper permanent portion of the
hair follicle, the infundibulum. Adapted from (Jaks et al., 2010). (B) Similarly, sweat
glands were recently reported to have multiple progenitor cell populations, both in the
lower acinar region and the upper sweat duct. Adapted from (Lu et al., 2012).

126
In terms of organization, sweat glands resemble hair follicles such that LRCs and K15
promoter expression is restricted to the lower portion of the appendage: the acinar region
and bulge, respectively. In addition, the K15 marked sweat gland acinar region does not
appear to contribute to the sweat duct under normal homeostasis. Similarly, K15 marked
hair follicle bulge stem cells do not contribute to the isthmus or infundibulum. Looking at
Sca-1 expression, it is restricted to the infundibulum of the hair follicle and epidermis
(Jensen et al., 2008). Likewise, we observed positive Sca-1 expression in the sweat duct
and epidermis, which confirms previously published results (Lu et al., 2012). Both the
acinar sweat gland region and hair follicle bulge do not express Sca1 (Figure 4.3).

Figure 4.3. Sca-1 is expressed in sweat ducts and not the acinar region. Similar to
hair follicles where it is expressed only in the infundibulum, Sca-1 is generally expressed
in the sweat ducts (du) close to the epidermis and not the acinar secretory region
containing LRCs.
127
4.4 Hair Follicles and Nails are Morphologically Similar.
As for the nails, they are similar to hair follicles such that both appendages produce a
keratinized structure, the hair shaft and nail plate, which grows throughout life.
Furthermore, both of these terminally differentiated structures are generated from rapidly
proliferating cells of the matrix in their respective appendages. Sweat glands, on the other
hand, do not produce any physical structures similar to the nail plate or hair shaft and do
not have any morphological structures resembling a well defined proliferative matrix.
Decades ago, the morphological similarities between the hair follicle and nail have been
reported where the nail is like a modified version of hair follicles rotated 90 ° (de Berker
and Baran, 2012). Since then, there have been multiple models highlighting the analogy
between these two appendages (Figure 4.4).

Figure 4.4. Comparison of hair follicle and nail morphologies. Adapted from (de
Berker and Baran, 2012).

128
More importantly, a recent report demonstrated that the ventral proximal fold, the region
of nail LRCs localization, expresses hair follicle bulge stem cell markers in humans.
More specifically, K15, K19, and PHLDA1 bulge stem cell markers were all found in the
ventral proximal fold suggesting that this region contains the stem cells of the nail
appendage (Sellheyer and Nelson, 2012). Due to the similarities of the nail and hair
follicles, they may potentially be regulated by some same mechanisms where certain
mutations affect both appendages including Foxn1 and Hoxc13 deletion (Godwin and
Capecchi, 1998; Mecklenburg et al., 2004; Potter et al., 2011).

129
4.5 BMP Signaling is Required for Hair Follicle, Sweat gland, and Nail
Development.
Previously, it has been shown that BMP signaling is required for proper hair follicle
differentiation during morphogenesis (Andl et al., 2004; Kobielak et al., 2003). More
specifically, Bmpr1a deletion in the basal layer of the skin using a K14Cre resulted in
smaller hair follicles with no inner root sheath (IRS) or hair shafts. Using the same
transgenic system crossed onto a K14-H2BGFP reporter, we demonstrate that BMP
signaling is also required for proper sweat gland and nail morphogenesis. In sweat
glands, the initiation of morphogenesis was inhibited as indicated by the complete
absence of sweat glands in Bmpr1a KO mice. Nail structures were still apparent in the
absence of Bmpr1a; however, a range of defects including nail bed hyperplasia, absence
of keratogenous zone, compromised nail plate, and ectopic expression of epidermal
markers were observed. This demonstrates that BMP signaling, particularly through
Bmpr1a, is essential for the proper development of all three appendages.
130
Chapter 5:
Concluding Remarks

Among the skin appendages, hair follicles and their stem cells have been the best
characterized thus far. Hair follicles undergo cyclic regeneration governed by signals
emitted by the mesenchymal dermal papilla. Upon activation of the growth phase
(anagen), hair follicle stem cells proliferate and differentiate into TA cells of the matrix
where they terminally differentiate into the different layers of the hair follicle. With the
use of current transgenic and gene targeting technology, the molecular mechanisms
regulating this process is quickly unraveling. Unlike hair follicles, reports of the
dynamics and regulations of sweat glands and nails have been limited. Furthermore, the
existence of sweat gland or nail stem cells has been poorly characterized in comparison to
the hair follicles. A recent report was the first to identify and characterize stem cells in
the sweat glands (Lu et al., 2012) while scientists have yet to identify the presence of
stem cells in nails.

Almost two decades ago, a distinct population of slow-cycling LRCs was identified in the
hair follicle bulge region (Cotsarelis et al., 1990). Numerous subsequent studies have
demonstrated that these bulge cells are the bona fide stem cell population required for the
maintenance and regeneration of hair follicles. Since then, researchers have searched for
LRCs in various tissues in hopes to identify new putative stem cell populations. In 2008
and 2009, Nakamura et al. have reported the presence of slow cycling LRCs in the nails
131
and eccrine sweat glands, respectively (Nakamura and Ishikawa, 2008; Nakamura and
Tokura, 2009). However, these newly identified LRCs were not isolated or thoroughly
characterized.

Here, we used a K5 driven tetracycline regulated H2B-GFP system to localize, isolate,
and characterize infrequently dividing LRCs of the sweat glands and nails as putative
new skin stem cells. This system was previously used to isolate and characterize live
label retaining hair follicle stem cells (Tumbar et al., 2004). At the time we began our
study, there have been no reports on the identification and characterization of sweat gland
or nail stem cells. In 2012, however, Lu et al. characterized the presence of unipotent
sweat gland stem cell populations in the basal and luminal layers responsible for the
regeneration of their respective layers in adult sweat glands (Lu et al., 2012). Although
they have also used this same K5TetOff/TreH2BGFP system to confirm the existence of
LRCs in the acinar region, they did not isolate these specific cells or demonstrate their
regenerative potential as putative stem cells. Instead, they had a different strategy using
K14-H2BGFP reporter mice to isolate basal and luminal layer cells for further
characterization. Their molecular characterization of the sweat gland basal layer cells did
not distinguish clear cells from myoepithelial cells of the basal layer; therefore, the
analysis of its regenerative potential was not specific to the LRCs population, which is
the focus of this dissertation.

132
In sweat glands, we have localized slow cycling LRCs to the myoepithelial cells of the
basal layer, restricted to the coiled secretory region only (Figure 5.1C). We have shown
that these sweat gland LRCs specifically expresses p63, which has been demonstrated to
be essential for epidermal differentiation and self renewal. Consistent with previous
reports, we have also shown that sweat glands LRCs express K15, which is a known
marker of hair follicle stem cells (Nakamura and Tokura, 2009). Similar to hair follicle
stem cells, these K15 marked sweat gland cells contribute to its own appendage long-
term, but do not contribute to epidermal homeostasis. We have successfully isolated the
SG LRCs for gene expression analysis and have identified a number of BMP signaling
genes. By knocking out Bmpr1a expression in the basal layer, we demonstrated that
Bmpr1a-mediated signaling is essential for sweat gland development. In regards to
wound healing, we have shown that under favorable conditions, sweat gland LRCs
themselves can trans-differentiate and contribute to the different layers of a newly
regenerated stratified epidermis after wounding. Moreover, preliminary studies have
demonstrated that unsorted dissociated sweat gland cells transplanted along with freshly
isolated newborn dermal fibroblasts have the plasticity to regenerate the sweat gland
structure, contribute to the epidermis, and trans-differentiate into hair follicles. However,
it remains unclear which specific sweat gland cells types are responsible for the
contribution. Nevertheless, it highlights the plasticity and regenerative potential of the
sweat gland cells, which are characteristics of stem cells.

133
In the nails, we were the first to identify LRCs in the basal layer of the ventral proximal
nail fold at the border where the epidermal granular layer ceases (Figure 5.1A). A
previous study has reported LRCs in the nail matrix (Nakamura and Ishikawa, 2008), but
we have demonstrated that the nail matrix contains actively proliferating cells, quickly
diluting out the H2B-GFP label. Instead, we found a cluster of nail LRCs organized in a
strip surrounding the nail structure. Similar to hair follicle stem cells and sweat gland
LRCs, we have shown that nail LRCs also express K15. Moreover, these K15 marked
cells can contribute long-term to the nail structure. Although nail LRCs are slow cycling,
they can be activated upon nail removal. In addition, transplanted strips of nail LRCs can
contribute to the nail structure. We have isolated and determined the transcriptional gene
expression profile of these nail LRCs where the down-regulation of BMP signaling
inhibitors suggested active BMP signaling in the nails. Similar to hair follicles and sweat
glands, we demonstrated the importance of BMP signaling in proper nail differentiation.

Figure 5.1. Schematic of skin appendages and its LRCs localization. Summary of
markers expressed in the (A) nail, (B) hair follicle, (C) sweat glands and their LRCs
localization.
134

Taken together, we have identified two distinct populations of LRCs in the sweat glands
and nails. Our data demonstrates that these two LRCs population possess stem cell-like
characteristics suggesting that they may be stem cells of these appendages. Although
distinct in their own ways, these newly identified LRCs carry many similarities to the
well-characterized hair follicle stem cells, which are also LRCs. For example, both sweat
gland and nail LRCs are also localized to the K5 expressing basal layer similar to hair
follicle stem cells. Moreover, they express K15, which is an established hair follicle stem
cell marker. For the first time, we have also demonstrated that BMP signaling is
important for proper sweat gland and nail development similar to the hair follicle
appendage.

Functionally, we have shown that K15 marked cells can contribute to the sweat gland and
nail structures long-term demonstrating its ability to self renew, which is a criteria for
stem cells. In addition, both sweat gland and nail LRCs can be activated when stimulated
by injury. Transplanted strips of nail LRCs can contribute to the nail structure while
dissociated sweat gland cells can regenerate sweat gland structures. Although further
characterization and lineage tracing studies are required to determine whether these LRCs
are bona fide stem cells, the current data demonstrates that these newly identified sweat
gland and nail LRCs have stem cell characteristics.

The hair follicle, sweat gland, and nail are distinct skin appendages with unique
structures, functions, and localization. Therefore, they are presumably maintained by
135
distinct stem cells populations within their own appendages. These stem cells are likely
regulated by distinct signals and mechanisms to meet the needs of their respective
appendages. At the same time, they may very well share many similarities given their
common origin and localization to the skin. Unveiling their similarities and differences
will further our knowledge on how different stem cell populations can be regulated in the
same organ. More importantly, understanding the common mechanisms regulating these
appendages may be useful in treating dermatological disorders affecting multiple skin
appendages such as ectodermal dysplasia. Similarly, understanding the unique
mechanisms regulating each of these distinct appendages may prove important in treating
isolated disorders affect individual appendages alone including anonychia, isolated nail
dysplasia, and hyperhidrosis. In conclusion, we have identified, isolated, and
characterized distinct LRCs in the sweat glands and nails as putative stem cells. Further
characterization and understanding of these potential stem cells may have great
therapeutic potential in the future.

136
Chapter 6:
Materials and Methods

6.1 Generation of Transgenic Mouse Lines
Keratin 5-driven tetracycline repressor mice (K5-tTA) (Diamond et al., 2000) was mated
with tetracycline response element-driven histone H2B-GFP transgenic mice (pTRE-
H2B-GFP) (Tumbar et al., 2004) to obtain H2B-GFP positive double transgenic animals.
GFP negative pups were sacrificed while GFP positive pups were kept for analysis.
Groups of 25-30 GFP positive mice were collected for chase with doxycycline at 3 to 4
weeks of age after weaning.

Bmpr1a-floxed mice generated by (Mishina et al., 2002) was mated with a K14Cre
recombinase generated by (Vasioukhin et al., 1999) to generate homozygous knockout
mice for Bmpr1a in the skin as previously described. This transgenic line was mated with
a K14 driven H2BGFP reporter to label K14 cells of the skin. Bmpr1a KO animals were
identified by its lack of whisker phenotype and confirmed with genotyping.

137
Genotyping of transgenic mice
Animals were tagged and tail clips were collected for genotyping. Tails were digested in
tail lysis buffer and 1mg/ml proteinase K overnight at 56 °C. Digested samples were
vortexed, diluted with 500ul H
2
O, and centrifuged for 5min at 15,000rpm. 1.5ul of DNA
template was subsequently used for genotyping. Store at 4 °C.
Tail Lysis Buffer
50mM Tris pH 8.0
2mM NaCl
10mM EDTA
0.1% SDS

138
6.2 Immunohistochemistry and Immunofluorescence Staining
Hematoxylin and Eosin Staining
10-12um thick frozen sections were thawed and fixed in 4% Paraformaldehyde for 10
min at room temperature. Sections are then washed with 1x phosphate buffered saline
(PBS) 2 times for 2 minutes each and stained with hematoxylin for 4 minutes. The
hematoxylin was then aspirated and the samples washed with distilled water twice. Next,
samples were dipped in 0.1% NH
4
OH for bluing and again washed with distilled water
twice for 2 minutes. Meanwhile, a freshly mixed 0.5% Eosin solution containing 1ml
Eosin, 1ml 100% EtOH, and 12.5ul glacial acetic acid was prepared and used for staining
for 3 minutes at room temperature. Sections were then washed with 95% EtOH twice for
1 minute each followed by a quick rinse with PBS. Slides are closed with 80% glycerol,
sealed with nail polish and stored at room temperature.

X-gal staining for β-galactosidase detection
All frozen sections were fixed in 4% Paraformaldehyde. For LacZ visualization, frozen
sections were fixed in 0.2% Glutaraldehyde for 1min, thoroughly washed with PBS, and
stained with 1mg/ml X-gal overnight at 37°C. Mount sections in 80% glycerol. X-gal
must be freshly added to the following staining solution:
X-gal Staining Solution:
100mM Na Phosphate (pH 7.3)
1.3mM MgCl
2

3mM K
3
Fe(CN)
6

139
3mM K
4
Fe(CN)
6
1mg/ml X-gal in DMF*
H
2
O

BrdU pulse
Animals are weighed and 10mg/ml 5-Bromo-2’deoxyuridine (BrdU) in sterile DPBS was
injected intraperitoneally (IP) for a final concentration of 1.5mg BrdU/g mouse. Animals
were sacrificed 2-3 hours afterwards for BrdU pulse.

Immunofluorescence staining
For immunofluorescence staining, frozen sections were fixed in 4% paraformaldehyde
for 10min, washed 3 times with PBS for 5 minutes each, and permeabilized with 0.1%
Triton X-100 for 10min. Sections are blocked in 0.1% Triton-PBS, 0.5% goat serum, and
0.1% BSA blocking buffer for 1h at room temperature. Primary antibodies were
incubated in the blocking buffer overnight at 4°C and wash with PBS 3 times the
following morning. Secondary antibodies were incubated in 0.1% BSA for 1h at room
temperature and washed with PBS twice. 1:2000 DAPI (10mg/ml stock) was used for
sections stained with DAPI for 5 minutes in H
2
O. Finally, sections were mounted in
fluoromount (Sigma), imaged and temporarily stored at 4 °C.

140
For BrdU Staining, sections were incubated in pre-warmed 1N HCl for 1h at 37°C
followed by 10min at 4 °C after paraformaldehyde fixation to denature the DNA for BrdU
detection.

The following primary antibodies were used:

Antibody Species Dilution Company Catalog
number
AE13 Mouse 1:100 Santa Cruz Biotech sc80607
AE15 Mouse 1:100 Santa Cruz Biotech sc57012
BrdU Rat 1:200 Abcam ab6326
CD104 ( β4) Rat 1:100 BD Pharmingen 553745
Gja1 Rabbit 1:400 Abcam ab11370
Keratin1 Rabbit 1:300 Gift from C. Jamora N/A
Keratin14 Rabbit 1:200 Gift from E. Fuchs N/A
Keratin15 Mouse 1:100 Thermo Scientific MS-1068-P1
Keratin18 Rabbit 1:200 Gift from E. Fuchs N/A
Keratin5 Chicken 1:300 Gift from C. Jamora N/A
Keratin8 Rat 1:100 DSHB TROMA-I
Ki67 Rabbit 1:200 Leica NCL-Ki67p
Laminin Rabbit 1:100 Thermo Scientific RB-082-A1
Loricrin Rabbit 1:300 Gift from C. Jamora N/A
Na
+
/K
+
ATPase Rabbit 1:300 Abcam ab58475
p63α (H-129) Rabbit 1:100 Santa Cruz Biotech sc-8344
P-Smad1/5/8 Rabbit 1:50 Cell Signaling 9511
P-Smad2 Rabbit 1:200 Cell Signaling 3101
SMA Mouse 1:200 Sigma A5228
Table 6.1. List of primary antibodies. List of primary antibodies and dilutions used for
immunofluorescence stainings.

141

The following Secondary antibodies were used:

Antibody Dilution Company Catalog
Number
Goat anti-Chicken Alexa 594 1:500 Invitrogen A11042
Goat anti-Mouse TRITC 1:300 Sigma T6528
Goat anti-Rabbit TRITC 1:300 Sigma T6778
Rabbit anti-Rat TRITC 1:300 Sigma T4280
Goat anti-Chicken Alexa 488 1:500 Invitrogen A11039
Goat anti-Rabbit FITC 1:300 Sigma F9887
Donkey anti-Rabbit Alexa 350 1:150 Invitrogen A10039
Goat anti-Rat Alexa 350 1:150 Invitrogen A21093
Table 6.2. List of secondary antibodies. List of secondary antibodies and dilutions used
for immunofluorescence stainings.

142
6.3 Isolation of Skin Stem Cells
Isolation of SG LRCs and Sole’s Epidermal Basal Cells
K5TetOff/TreH2BGFP animals were fed 1mg/g doxycycline food for 4 weeks starting at
P21-28. GFP+ sweat glands were dissected out with its surrounding sole’s epidermis
from the fingertips of 20-30 mice and treated with 1000U/ml Collagenase type I for 1h at
37°C with shaking. Sweat glands were mechanically separated from its epidermis and
treated further with 1000U/ml Collagenase type I and 500ug/ul Hyaluronidase (Sigma)
for 1h at 37°C with shaking. Purified sole’s epidermis and SGs were independently
washed with DPBS and digested with 0.25% Trypsin-EDTA for 20min at 37°C with
shaking. Neutralize and filter cells through a 40um cell strainer.

Isolation of Nail LRCs
K5TetOff/TreH2BGFP animals were fed 1mg/g doxycycline food for 4 weeks starting at
P21-28. The paws were collected and their epidermis was carefully cut open at the
proximal fold to reveal the nail LRCs. Next, strips of nail LRCs were individually
isolated through microdissection and collected in PBS on ice. When completed, isolated
nail LRC strips were digested with 0.25% Trypsin-EDTA overnight at 4°C with shaking.
The trypsin is neutralized and single cells were filtered through a 40µm cell strainer. For
the isolation of the sole’s epidermal cells, GFP+ sweat glands were dissected out with its
surrounding sole’s epidermis from the fingertips of 20-30 mice and treated with
1000U/ml Collagenase type I for 1h at 37°C with shaking. Purified sole’s epidermis and
143
was washed with DPBS and digested in 0.25% Trypsin-EDTA for 30min at 37°C with
shaking. Neutralize and filter cells through a 40um cell strainer.

Isolation of Newborn Dermal Fibroblasts
Backskins from newborn mice were removed and incubated in 2.4U/ml dispase, floating
dermis side down, overnight at 4 °C. The epidermis was carefully removed from the
dermis and discarded. Mince dermis tissues and incubate in 1000U/ul collagenase type I
(1ml 4U/ul collagenase stock + 3ml media) at 37 °C for 30min with shaking. Next, pipet
dermal solution in a 25ml serological pipet for 20sec to disaggregate cells and add 20ml
fibroblast media. Filter dermal solution through a 70um cell strainer and centrifuge at
800rpm for 5min. Transfer supernatant to a new tube and centrifuge at 1400rpm for 5min,
save the pellet. Resuspend the 800rpm pellet in 20ml fibroblast media and centrifuge at
300rpm for 5min. Transfer supernatant to a new tube and centrifuge at 1000rpm for
5mine, save the pellet. Resuspend the 300rpm pellet in 20ml fibroblast media and
centrifuge at 1000rpm for 5min. Transfer supernatant into a new tube and centrifuge at
1000rpm for 5min, save the pellet. Finally, resuspend and pool the 3 pellets of dermal
fibroblasts in 15ml media each. Centrifuge at 1000rpm for 5min, decant supernatant,
resuspend in 25ml media, and repeat centrifugation. Resuspend pellet in 20ml media and
filter through a 40um cells strainer.

144
Fluorescence Activated Cell Sorting (FACS)
For FACS, isolated 4 week chased H2B-GFP labeled sweat gland cells were stained with
a primary antibody: anti- α6 integrin (CD49f) conjugated to PE (1:200; BD Pharmingen)
for 30min and sorted using the FACS Aria II cell sorter (BD, Bioscience) for
H2BGFP+/ α6+ and H2BGFP-/ α6+ populations. Cells were collected in RNAprotect Cell
Reagent (Qiagen) for later RNA isolation. Similarly, FACS analysis were performed on
isolated K15-GFP labeled sweat gland cells, stained with the primary antibody against
anti- α6 integrin as described above.

145
6.4 Transplantations
Kidney Capsule Transplantation
The overlying skin and muscle was cut open over the kidney of immunocompromised
nude mice anesthetized under isoflurane. A single cells suspension of 4 week chased
unsorted sweat gland cells were carefully injected underneath the membrane of the
kidney. The kidney was then placed back into the body cavity and the muscle was sutured
shut. The skin was stapled together and the mice were allowed to heal until the indicated
times.

Chamber Graft
A silicon chamber was implanted onto the backs of immunocompromised “nude” mice
with a full-thickness skin wound as previously described (Weinberg et al., 1993). 4 week
chased whole H2B-GFP sweat glands were dissected out with its surrounding sole’s
epidermis from the fingertips and treated with 1000U/ml Collagenase type I for 1h at
37°C with shaking. After separation from the sole’s epidermis, the purified dermis with
remaining sweat glands was transplanted into the humidified silicone chamber. The upper
chamber was removed 2 weeks after transplantation and the bottom half is removed 3
weeks after transplantation as the skin is healing. The nude mice were either fed regular
mouse diet or doxycycline food after transplantation for the duration of experiments.
Similarly, dissociated unsorted 4 week chased SG single cell suspension labeled with
H2B-GFP after separation from the sole’s epidermis was mixed with freshly isolated
unmarked newborn dermal fibroblasts (approximately 6 million cells total) at a 1:1
146
proportion and injected into the chamber. Mice were sacrificed and samples from the
graft regions were taken for GFP+ expression and tissue analysis. All mice work was
conducted according to Institutional Animal Care and Use Committee at University of
Southern California (IACUC). The protocols (No. 11306 and 11325) were approved by
the IACUC Committee. All surgery was performed under either isoflurane or ketamine
anesthesia, and all efforts were made to minimize suffering with analgesics (Buprenex
prior and postsurgery was administrated).

Subcutaneous Injections
Dissociated unsorted SG single cell suspension labeled with H2B-GFP after 4 weeks of
chase after separation from the sole’s epidermis was mixed with freshly isolated
unmarked newborn dermal fibroblasts at a 1:1 proportion and injected subcutaneously
underneath the back skin of an immunocompromised “nude” mouse (approximately 1.5
million cells per spot).

Tranplantation of Nail LRCs
Nail LRCs enriched strips of epithelium were isolated through microdissection after a 4
week chase with doxycycline as described above. Isolated strips of nail LRCs were then
transplanted under the proximal fold of immunocompromised nod scid mice.

147
6.5 Microscopy
Confocal Microscopy of sweat gland LRCs and 3D reconstruction of sweat glands
Whole sweat glands were imaged on a Zeiss (Carl Zeiss, LLC) Axiovert 200 inverted
microscope with an LSM 510 meta confocal scan head using a 40x/1.2NA water
immersion lens. Tissue was dissected, stained and placed in a 35mm glass bottom tissue
culture dish (MatTek corporation, Asland, MD). Tissues were submerged in deionized
water to minimize refractivity. Histone 2B GFP and TRITC stained laminin (1:100;
Thermo Scientific, RB-082-A1), were visualized using conventional confocal imaging
using argon laser lines at 488nm and 543nm, respectively. DAPI was imaged with 2-
photon excitation using a Coherent Chameleon (Coherent Inc, Santa Clara, CA), pulsed
laser tuned to 800nm. Images were collected at 0.22 μm in plane (xy) and optically
sectioned at 2 μm (in z). 3D reconstruction and visualizations were performed in ImageJ
(http://rsbweb.nih.gov/ij/), Fiji (http://fiji.sc/wiki/index.php/Fiji), Avizó 6.3 (VSG,
Burlinton, MA) and Vaa3D (http://www.vaa3d.org/).

Confocal Microscopy and 3D reconstruction of nail LRCs.
The epidermis of a 3 week chased paw was cut open along the proximal fold to reveal the
underlying nail LRCs. Nail LRCs were subsequently imaged on a Zeiss (Carl Zeiss,
LLC) AxioImager.M1 upright microscope with an LSM5 Pascal confocal scan head
using a 40x/1.3NA lens. H2B-GFP marked nail LRCs were visualized using conventional
confocal imaging using argon laser lines at 488nm. Images were and optically sectioned
at 2 μm (in z). 3D reconstruction and visualizations were performed in ImageJ
148
(http://rsbweb.nih.gov/ij/), Fiji (http://fiji.sc/wiki/index.php/Fiji), Avizó 6.3 (VSG,
Burlinton, MA) and Vaa3D (http://www.vaa3d.org/).

149
6.6 RNA isolation and qPCR
Total RNAs were purified from FACS-sorted SG LRCs, SG non-LRCs, and the basal
layer of the sole’s epidermis using Qiagen’s RNeasy Micro kit according to the
manufacturer’s instructions. Equal amounts of RNA were reverse transcribed using the
Superscript III First-Strand Synthesis System (Invitrogen) according to the
manufacturer’s instructions. cDNAs were amplified by PCR and used in triplicate for
each qPCR sample primer set with all primer sets designed to work under the same
conditions. Real-time PCR amplification of particular genes of interest was performed
using an Applied Biosystems 7900HT Fast Real-Time PCR System and the fold
difference between samples and controls were calculated based on the 2- ΔΔCT method,
normalized to β-actin levels.

Gene Primer Sequence (5’ to 3’)
Bgn-F TCTCACCTGACACCACACTGCT
Bgn-R GTTCAAAGCCACTGTTCTCCAG
Mmp2-F CCGGCGATGTCGCCCCTAAA
Mmp2-R ACCTGTCTGGGGCAGCCCAA
Tcf4-F CACCCGGCCATCGTCACAC
Tcf4-R GCCACCTGCGCCCGAGAAT
Timp2-F CCAGAAGAAGAGCCTGAACCA
Timp2-R GTCCATCCAGAGGCACTCATC
β-actin-F GGATGCAGAAGGAGATTACTG
β-actin-R CCGATCCACACAGAGTACTTG
Table 6.3. List of primers for qPCR. Primers used for real-time PCR (qPCR).

150
6.7 Microarray analysis
Total RNAs from FACS of LRCs (GFP+/ α6 integrin+), non-LRCs (GFP-/ α6 integrin+)
either from sweat glands or nails, and sole’s epidermis ( α6 integrin+) were purified using
a RNeasy Micro Kit (Qiagen, Valencia California, United States), and quantified
(Nanodrop, United States) for two separate microarray analysis from two independent
biological samples. RNA 6000 Pico Assay (Agilent Technologies, Palo Alto, California,
United States) was used for RNA quality check. Amplification/labeling were performed
on 50 nanogram (ng) and 250ng to obtain biotinylated cRNA (Ovation™ RNA
Amplification System; Nugen, San Carlos, California, United States and Ambion Kit;
Affymetrix, Santa Clara, California, United States, respectively), and either 3.75 μg or
5.5 μg ssDNA were used for fragmentation, labeling and hybridization. Hybridization
was performed at 45°C for 18 h to Mouse Gene 1.0 ST array (Affymetrix, Santa Clara,
California, United States). Processed chips were read by GeneChip Scanner 3000 7G
(Genomics Core Facility, Children’s Hospital Los Angeles, Los Angeles, California,
United States). The raw expression intensity data was imported into Partek Genomic
Suite v6 (Partek Inc., St. Louis, MO, United States). The data was pre-processed using
the RMA algorithm with the default Partek setting. Following fold change calculations,
differentially expressed gene (DEG) lists containing probe sets with 2-fold intensity
changes in either direction were generated. Common DEG list was generated by
comparing the DEG list of the sweat gland LRCs experiment and GFP- α6+ sweat glands
experiment to the sole’s epidermis. Functional annotation of the DEG list was carried out
using the “Database for Annotation, Visualization and Integrated Discovery” (DAVID).
151
References

Allen, J.A., Armstrong, J.E., and Roddie, I.C. (1973). The regional distribution of
emotional sweating in man. J Physiol 235, 749-759.

Alonso, L., and Fuchs, E. (2006). The hair cycle. J Cell Sci 119, 391-393.

Andl, T., Ahn, K., Kairo, A., Chu, E.Y., Wine-Lee, L., Reddy, S.T., Croft, N.J., Cebra-
Thomas, J.A., Metzger, D., Chambon, P., et al. (2004). Epithelial Bmpr1a regulates
differentiation and proliferation in postnatal hair follicles and is essential for tooth
development. Development 131, 2257-2268.

Banfield, C.C., Redburn, J.C., and Dawber, R.P. (1998). The incidence and prognosis of
nail apparatus melanoma. A retrospective study of 105 patients in four English regions.
Br J Dermatol 139, 276-279.

Barbareschi, M., Pecciarini, L., Cangi, M.G., Macri, E., Rizzo, A., Viale, G., and
Doglioni, C. (2001). p63, a p53 homologue, is a selective nuclear marker of
myoepithelial cells of the human breast. Am J Surg Pathol 25, 1054-1060.

Barker, N., van Es, J.H., Kuipers, J., Kujala, P., van den Born, M., Cozijnsen, M.,
Haegebarth, A., Korving, J., Begthel, H., Peters, P.J., et al. (2007). Identification of stem
cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003-1007.

Bergmann, C., Senderek, J., Anhuf, D., Thiel, C.T., Ekici, A.B., Poblete-Gutierrez, P.,
van Steensel, M., Seelow, D., Nurnberg, G., Schild, H.H., et al. (2006). Mutations in the
gene encoding the Wnt-signaling component R-spondin 4 (RSPO4) cause autosomal
recessive anonychia. Am J Hum Genet 79, 1105-1109.

Biedermann, T., Pontiggia, L., Bottcher-Haberzeth, S., Tharakan, S., Braziulis, E.,
Schiestl, C., Meuli, M., and Reichmann, E. (2010). Human eccrine sweat gland cells can
reconstitute a stratified epidermis. J Invest Dermatol 130, 1996-2009.

Blanpain, C., and Fuchs, E. (2006). Epidermal stem cells of the skin. Annu Rev Cell Dev
Biol 22, 339-373.

Blanpain, C., Lowry, W.E., Geoghegan, A., Polak, L., and Fuchs, E. (2004). Self-
renewal, multipotency, and the existence of two cell populations within an epithelial stem
cell niche. Cell 118, 635-648.

Blaydon, D.C., Ishii, Y., O'Toole, E.A., Unsworth, H.C., Teh, M.T., Ruschendorf, F.,
Sinclair, C., Hopsu-Havu, V.K., Tidman, N., Moss, C., et al. (2006). The gene encoding
R-spondin 4 (RSPO4), a secreted protein implicated in Wnt signaling, is mutated in
inherited anonychia. Nat Genet 38, 1245-1247.
152

Borthwick, D.W., Shahbazian, M., Krantz, Q.T., Dorin, J.R., and Randell, S.H. (2001).
Evidence for stem-cell niches in the tracheal epithelium. Am J Respir Cell Mol Biol 24,
662-670.

Cai, J., and Ma, L. (2011). Msx2 and Foxn1 regulate nail homeostasis. Genesis 49, 449-
459.

Chalmers, T.M., and Keele, C.A. (1952). The nervous and chemical control of sweating.
Br J Dermatol 64, 43-54.

Chan, P., Kao, G.F., Pierson, D.L., and Rodman, O.G. (1985). Episodic hyperhidrosis on
the dorsum of hands. J Am Acad Dermatol 12, 937-942.

Chan, R.W., and Gargett, C.E. (2006). Identification of label-retaining cells in mouse
endometrium. Stem Cells 24, 1529-1538.

Chen, H., Lun, Y., Ovchinnikov, D., Kokubo, H., Oberg, K.C., Pepicelli, C.V., Gan, L.,
Lee, B., and Johnson, R.L. (1998). Limb and kidney defects in Lmx1b mutant mice
suggest an involvement of LMX1B in human nail patella syndrome. Nat Genet 19, 51-55.

Cheshire, W.P., and Freeman, R. (2003). Disorders of sweating. Semin Neurol 23, 399-
406.

Cluzeau, C., Hadj-Rabia, S., Jambou, M., Mansour, S., Guigue, P., Masmoudi, S., Bal,
E., Chassaing, N., Vincent, M.C., Viot, G., et al. (2011). Only four genes (EDA1, EDAR,
EDARADD, and WNT10A) account for 90% of hypohidrotic/anhidrotic ectodermal
dysplasia cases. Hum Mutat 32, 70-72.

Cotsarelis, G., Cheng, S.Z., Dong, G., Sun, T.T., and Lavker, R.M. (1989). Existence of
slow-cycling limbal epithelial basal cells that can be preferentially stimulated to
proliferate: implications on epithelial stem cells. Cell 57, 201-209.

Cotsarelis, G., Sun, T.T., and Lavker, R.M. (1990). Label-retaining cells reside in the
bulge area of pilosebaceous unit: implications for follicular stem cells, hair cycle, and
skin carcinogenesis. Cell 61, 1329-1337.

Cutolo, M., Sulli, A., Secchi, M.E., Olivieri, M., and Pizzorni, C. (2007). The
contribution of capillaroscopy to the differential diagnosis of connective autoimmune
diseases. Best Pract Res Clin Rheumatol 21, 1093-1108.

Dawber, R.P., De Berker, D., and Baran, R. (2001). Science of the nail apparatus. Baran
and Dawber's diseases of the nails and their management 3, 1-47.

153
de Berker, D., and Angus, B. (1996). Proliferative compartments in the normal nail unit.
Br J Dermatol 135, 555-559.

de Berker, D.A., and Baran, R. (2012). Science of the Nail Apparatus. Baran & Dawber's
Diseases of the Nails and their Management, Fourth Edition, 1-50.

Dhouailly, D. (1973). Dermo-epidermal interactions between birds and mammals:
differentiation of cutaneous appendages. J Embryol Exp Morphol 30, 587-603.

Diamond, I., Owolabi, T., Marco, M., Lam, C., and Glick, A. (2000). Conditional gene
expression in the epidermis of transgenic mice using the tetracycline-regulated
transactivators tTA and rTA linked to the keratin 5 promoter. J Invest Dermatol 115, 788-
794.

Dreyer, S.D., Zhou, G., Baldini, A., Winterpacht, A., Zabel, B., Cole, W., Johnson, R.L.,
and Lee, B. (1998). Mutations in LMX1B cause abnormal skeletal patterning and renal
dysplasia in nail patella syndrome. Nat Genet 19, 47-50.

Eisenach, J.H., Atkinson, J.L., and Fealey, R.D. (2005). Hyperhidrosis: evolving
therapies for a well-established phenomenon. Mayo Clin Proc 80, 657-666.

Foudi, A., Hochedlinger, K., Van Buren, D., Schindler, J.W., Jaenisch, R., Carey, V., and
Hock, H. (2009). Analysis of histone 2B-GFP retention reveals slowly cycling
hematopoietic stem cells. Nat Biotechnol 27, 84-90.

Fuchs, E., Tumbar, T., and Guasch, G. (2004). Socializing with the neighbors: stem cells
and their niche. Cell 116, 769-778.

Godwin, A.R., and Capecchi, M.R. (1998). Hoxc13 mutant mice lack external hair.
Genes Dev 12, 11-20.

Greco, V., Chen, T., Rendl, M., Schober, M., Pasolli, H.A., Stokes, N., Dela Cruz-
Racelis, J., and Fuchs, E. (2009). A two-step mechanism for stem cell activation during
hair regeneration. Cell Stem Cell 4, 155-169.

Gurtner, G.C., Werner, S., Barrandon, Y., and Longaker, M.T. (2008). Wound repair and
regeneration. Nature 453, 314-321.

Hamm, H., Karl, S., and Brocker, E.B. (2000). Isolated congenital nail dysplasia: a new
autosomal dominant condition. Arch Dermatol 136, 1239-1243.

Hamrick, M.W. (2001). Development and evolution of the mammalian limb: adaptive
diversification of nails, hooves, and claws. Evol Dev 3, 355-363.

154
Handjiski, B.K., Eichmuller, S., Hofmann, U., Czarnetzki, B.M., and Paus, R. (1994).
Alkaline phosphatase activity and localization during the murine hair cycle. Br J
Dermatol 131, 303-310.

Haramis, A.P., Begthel, H., van den Born, M., van Es, J., Jonkheer, S., Offerhaus, G.J.,
and Clevers, H. (2004). De novo crypt formation and juvenile polyposis on BMP
inhibition in mouse intestine. Science 303, 1684-1686.

He, X.C., Zhang, J., Tong, W.G., Tawfik, O., Ross, J., Scoville, D.H., Tian, Q., Zeng, X.,
He, X., Wiedemann, L.M., et al. (2004). BMP signaling inhibits intestinal stem cell self-
renewal through suppression of Wnt-beta-catenin signaling. Nat Genet 36, 1117-1121.

Horsley, V., O'Carroll, D., Tooze, R., Ohinata, Y., Saitou, M., Obukhanych, T.,
Nussenzweig, M., Tarakhovsky, A., and Fuchs, E. (2006). Blimp1 defines a progenitor
population that governs cellular input to the sebaceous gland. Cell 126, 597-609.

Ito, M., Liu, Y., Yang, Z., Nguyen, J., Liang, F., Morris, R.J., and Cotsarelis, G. (2005).
Stem cells in the hair follicle bulge contribute to wound repair but not to homeostasis of
the epidermis. Nat Med 11, 1351-1354.

Jackson, A.D., and Lawler, S.D. (1951). Pachyonychia congenita; a report of six cases in
one family, with a note on linkage data. Ann Eugen 16, 142-146.

Jaks, V., Kasper, M., and Toftgard, R. (2010). The hair follicle-a stem cell zoo. Exp Cell
Res 316, 1422-1428.

Jensen, K.B., Collins, C.A., Nascimento, E., Tan, D.W., Frye, M., Itami, S., and Watt,
F.M. (2009). Lrig1 expression defines a distinct multipotent stem cell population in
mammalian epidermis. Cell Stem Cell 4, 427-439.

Jensen, U.B., Yan, X., Triel, C., Woo, S.H., Christensen, R., and Owens, D.M. (2008). A
distinct population of clonogenic and multipotent murine follicular keratinocytes residing
in the upper isthmus. J Cell Sci 121, 609-617.

Johnson, M., Comaish, J.S., and Shuster, S. (1991). Nail is produced by the normal nail
bed: a controversy resolved. Br J Dermatol 125, 27-29.

Johnson, M., and Shuster, S. (1993). Continuous formation of nail along the bed. Br J
Dermatol 128, 277-280.

Jumlongras, D., Bei, M., Stimson, J.M., Wang, W.F., DePalma, S.R., Seidman, C.E.,
Felbor, U., Maas, R., Seidman, J.G., and Olsen, B.R. (2001). A nonsense mutation in
MSX1 causes Witkop syndrome. Am J Hum Genet 69, 67-74.

155
Kandyba, E., Leung, Y., Chen, Y.B., Widelitz, R., Chuong, C.M., and Kobielak, K.
(2013). Competitive balance of intrabulge BMP/Wnt signaling reveals a robust gene
network ruling stem cell homeostasis and cyclic activation. Proc Natl Acad Sci U S A.

Kere, J., Srivastava, A.K., Montonen, O., Zonana, J., Thomas, N., Ferguson, B., Munoz,
F., Morgan, D., Clarke, A., Baybayan, P., et al. (1996). X-linked anhidrotic
(hypohidrotic) ectodermal dysplasia is caused by mutation in a novel transmembrane
protein. Nat Genet 13, 409-416.

Kiel, M.J., He, S., Ashkenazi, R., Gentry, S.N., Teta, M., Kushner, J.A., Jackson, T.L.,
and Morrison, S.J. (2007). Haematopoietic stem cells do not asymmetrically segregate
chromosomes or retain BrdU. Nature 449, 238-242.

Kierszenbaum, A.L. (2007). Histology and Cell Biology: An Introduction to Pathology
(Mosby Elsevier).

Kobielak, K., Kobielak, A., Roszkiewicz, J., Wierzba, J., Limon, J., and Trzeciak, W.H.
(2001). Mutations in the EDA gene in three unrelated families reveal no apparent
correlation between phenotype and genotype in the patients with an X-linked anhidrotic
ectodermal dysplasia. Am J Med Genet 100, 191-197.

Kobielak, K., Pasolli, H.A., Alonso, L., Polak, L., and Fuchs, E. (2003). Defining BMP
functions in the hair follicle by conditional ablation of BMP receptor IA. J Cell Biol 163,
609-623.

Kobielak, K., Stokes, N., de la Cruz, J., Polak, L., and Fuchs, E. (2007). Loss of a
quiescent niche but not follicle stem cells in the absence of bone morphogenetic protein
signaling. Proc Natl Acad Sci U S A 104, 10063-10068.

Koster, M.I., Kim, S., Mills, A.A., DeMayo, F.J., and Roop, D.R. (2004). p63 is the
molecular switch for initiation of an epithelial stratification program. Genes Dev 18, 126-
131.

Kurzrock, E.A., Lieu, D.K., Degraffenried, L.A., Chan, C.W., and Isseroff, R.R. (2008).
Label-retaining cells of the bladder: candidate urothelial stem cells. Am J Physiol Renal
Physiol 294, F1415-1421.

Labows, J.N., Preti, G., Hoelzle, E., Leyden, J., and Kligman, A. (1979). Steroid analysis
of human apocrine secretion. Steroids 34, 249-258.

Lee, M.M. (1960). Histology and histochemistry of human eccrine sweat glands, with
special reference to their defense mechanisms. Anat Rec 136, 97-105.

156
Lewis, B.L. (1954). Microscopic studies of fetal and mature nail and surrounding soft
tissue. AMA Arch Derm Syphilol 70, 733-747.

Liao, H., Sayers, J.M., Wilson, N.J., Irvine, A.D., Mellerio, J.E., Baselga, E., Bayliss,
S.J., Uliana, V., Fimiani, M., Lane, E.B., et al. (2007). A spectrum of mutations in
keratins K6a, K16 and K17 causing pachyonychia congenita. J Dermatol Sci 48, 199-205.

Lin, H. (2002). The stem-cell niche theory: lessons from flies. Nat Rev Genet 3, 931-940.

Liu, Y., Lyle, S., Yang, Z., and Cotsarelis, G. (2003). Keratin 15 promoter targets
putative epithelial stem cells in the hair follicle bulge. J Invest Dermatol 121, 963-968.

Lobitz, W.C., Jr., and Dobson, R.L. (1961). Dermatology: the eccrine sweat glands. Annu
Rev Med 12, 289-298.

Lobitz, W.C., Jr., and Holyoke, J.B. (1954). The histochemical response of the human
epidermis to controlled injury; glycogen. J Invest Dermatol 22, 189-198.

Lobitz, W.C., Jr., Holyoke, J.B., and Brophy, D. (1956). Response of the human eccrine
sweat duct to dermal injury. J Invest Dermatol 26, 247-259; discussion, 259-262.

Lobitz, W.C., Jr., Holyoke, J.B., and Montagna, W. (1954). Responses of the human
eccrine sweat duct to controlled injury: growth center of the epidermal sweat duct unit. J
Invest Dermatol 23, 329-344.

Lombaert, I.M., Brunsting, J.F., Wierenga, P.K., Faber, H., Stokman, M.A., Kok, T.,
Visser, W.H., Kampinga, H.H., de Haan, G., and Coppes, R.P. (2008). Rescue of salivary
gland function after stem cell transplantation in irradiated glands. PLoS One 3, e2063.

Lu, C.P., Polak, L., Rocha, A.S., Pasolli, H.A., Chen, S.C., Sharma, N., Blanpain, C., and
Fuchs, E. (2012). Identification of stem cell populations in sweat glands and ducts reveals
roles in homeostasis and wound repair. Cell 150, 136-150.

Lyle, S., Christofidou-Solomidou, M., Liu, Y., Elder, D.E., Albelda, S., and Cotsarelis,
G. (1998). The C8/144B monoclonal antibody recognizes cytokeratin 15 and defines the
location of human hair follicle stem cells. J Cell Sci 111 ( Pt 21), 3179-3188.

Lyle, S., Christofidou-Solomidou, M., Liu, Y., Elder, D.E., Albelda, S., and Cotsarelis,
G. (1999). Human hair follicle bulge cells are biochemically distinct and possess an
epithelial stem cell phenotype. J Investig Dermatol Symp Proc 4, 296-301.

Madisen, L., Zwingman, T.A., Sunkin, S.M., Oh, S.W., Zariwala, H.A., Gu, H., Ng, L.L.,
Palmiter, R.D., Hawrylycz, M.J., Jones, A.R., et al. (2010). A robust and high-throughput
157
Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13,
133-140.

McLean, W.H., Rugg, E.L., Lunny, D.P., Morley, S.M., Lane, E.B., Swensson, O.,
Dopping-Hepenstal, P.J., Griffiths, W.A., Eady, R.A., Higgins, C., et al. (1995). Keratin
16 and keratin 17 mutations cause pachyonychia congenita. Nat Genet 9, 273-278.

Mecklenburg, L., Paus, R., Halata, Z., Bechtold, L.S., Fleckman, P., and Sundberg, J.P.
(2004). FOXN1 is critical for onycholemmal terminal differentiation in nude (Foxn1)
mice. J Invest Dermatol 123, 1001-1011.

Mikkola, M.L. (2009). Molecular aspects of hypohidrotic ectodermal dysplasia. Am J
Med Genet A 149A, 2031-2036.

Millar, S.E. (2002). Molecular mechanisms regulating hair follicle development. J Invest
Dermatol 118, 216-225.

Miller, S.J., Burke, E.M., Rader, M.D., Coulombe, P.A., and Lavker, R.M. (1998). Re-
epithelialization of porcine skin by the sweat apparatus. J Invest Dermatol 110, 13-19.

Mira, H., Andreu, Z., Suh, H., Lie, D.C., Jessberger, S., Consiglio, A., San Emeterio, J.,
Hortiguela, R., Marques-Torrejon, M.A., Nakashima, K., et al. (2010). Signaling through
BMPR-IA regulates quiescence and long-term activity of neural stem cells in the adult
hippocampus. Cell Stem Cell 7, 78-89.

Mishina, Y., Hanks, M.C., Miura, S., Tallquist, M.D., and Behringer, R.R. (2002).
Generation of Bmpr/Alk3 conditional knockout mice. Genesis 32, 69-72.

Morris, R.J., Liu, Y., Marles, L., Yang, Z., Trempus, C., Li, S., Lin, J.S., Sawicki, J.A.,
and Cotsarelis, G. (2004). Capturing and profiling adult hair follicle stem cells. Nat
Biotechnol 22, 411-417.

Morris, R.J., and Potten, C.S. (1999). Highly persistent label-retaining cells in the hair
follicles of mice and their fate following induction of anagen. J Invest Dermatol 112,
470-475.

Nakahiro, T., Kurooka, H., Mori, K., Sano, K., and Yokota, Y. (2010). Identification of
BMP-responsive elements in the mouse Id2 gene. Biochem Biophys Res Commun 399,
416-421.

Nakamura, M., and Ishikawa, O. (2008). The localization of label-retaining cells in
mouse nails. J Invest Dermatol 128, 728-730.

158
Nakamura, M., and Tokura, Y. (2009). The localization of label-retaining cells in eccrine
glands. J Invest Dermatol 129, 2077-2078.

Nijhof, J.G., Braun, K.M., Giangreco, A., van Pelt, C., Kawamoto, H., Boyd, R.L.,
Willemze, R., Mullenders, L.H., Watt, F.M., de Gruijl, F.R., et al. (2006). The cell-
surface marker MTS24 identifies a novel population of follicular keratinocytes with
characteristics of progenitor cells. Development 133, 3027-3037.

Norton, L.A. (1971). Incorporation of thymidine-methyl-H3 and glycine-2-H3 in the nail
matrix and bed of humans. J Invest Dermatol 56, 61-68.

Oliver, J.A., Maarouf, O., Cheema, F.H., Martens, T.P., and Al-Awqati, Q. (2004). The
renal papilla is a niche for adult kidney stem cells. J Clin Invest 114, 795-804.

Oshima, H., Rochat, A., Kedzia, C., Kobayashi, K., and Barrandon, Y. (2001).
Morphogenesis and renewal of hair follicles from adult multipotent stem cells. Cell 104,
233-245.

Pellegrini, G., Dellambra, E., Golisano, O., Martinelli, E., Fantozzi, I., Bondanza, S.,
Ponzin, D., McKeon, F., and De Luca, M. (2001). p63 identifies keratinocyte stem cells.
Proc Natl Acad Sci U S A 98, 3156-3161.

Petersson, M., Brylka, H., Kraus, A., John, S., Rappl, G., Schettina, P., and Niemann, C.
(2011). TCF/Lef1 activity controls establishment of diverse stem and progenitor cell
compartments in mouse epidermis. EMBO J 30, 3004-3018.

Plikus, M., Wang, W.P., Liu, J., Wang, X., Jiang, T.X., and Chuong, C.M. (2004).
Morpho-regulation of ectodermal organs: integument pathology and phenotypic
variations in K14-Noggin engineered mice through modulation of bone morphogenic
protein pathway. Am J Pathol 164, 1099-1114.

Potten, C.S., Owen, G., and Booth, D. (2002). Intestinal stem cells protect their genome
by selective segregation of template DNA strands. J Cell Sci 115, 2381-2388.

Potter, C.S., Pruett, N.D., Kern, M.J., Baybo, M.A., Godwin, A.R., Potter, K.A.,
Peterson, R.L., Sundberg, J.P., and Awgulewitsch, A. (2011). The nude mutant gene
Foxn1 is a HOXC13 regulatory target during hair follicle and nail differentiation. J Invest
Dermatol 131, 828-837.

Rendl, M., Lewis, L., and Fuchs, E. (2005). Molecular dissection of mesenchymal-
epithelial interactions in the hair follicle. PLoS Biol 3, e331.

Rendl, M., Polak, L., and Fuchs, E. (2008). BMP signaling in dermal papilla cells is
required for their hair follicle-inductive properties. Genes Dev 22, 543-557.
159

Rittie, L., Sachs, D.L., Orringer, J.S., Voorhees, J.J., and Fisher, G.J. (2013). Eccrine
sweat glands are major contributors to reepithelialization of human wounds. Am J Pathol
182, 163-171.

Rochat, A., Kobayashi, K., and Barrandon, Y. (1994). Location of stem cells of human
hair follicles by clonal analysis. Cell 76, 1063-1073.

Sangiorgi, E., and Capecchi, M.R. (2008). Bmi1 is expressed in vivo in intestinal stem
cells. Nat Genet 40, 915-920.

Sato, K., Kang, W.H., Saga, K., and Sato, K.T. (1989a). Biology of sweat glands and
their disorders. I. Normal sweat gland function. J Am Acad Dermatol 20, 537-563.

Sato, K., Kang, W.H., Saga, K., and Sato, K.T. (1989b). Biology of sweat glands and
their disorders. II. Disorders of sweat gland function. J Am Acad Dermatol 20, 713-726.

Sato, K., Leidal, R., and Sato, F. (1987). Morphology and development of an apoeccrine
sweat gland in human axillae. Am J Physiol 252, R166-180.

Schmidt-Ullrich, R., and Paus, R. (2005). Molecular principles of hair follicle induction
and morphogenesis. Bioessays 27, 247-261.

Schon, M., Benwood, J., O'Connell-Willstaedt, T., and Rheinwald, J.G. (1999). Human
sweat gland myoepithelial cells express a unique set of cytokeratins and reveal the
potential for alternative epithelial and mesenchymal differentiation states in culture. J
Cell Sci 112 ( Pt 12), 1925-1936.

Sellheyer, K., and Nelson, P. (2012). The ventral proximal nail fold: stem cell niche of
the nail and equivalent to the follicular bulge--a study on developing human skin. J Cutan
Pathol 39, 835-843.

Shackleton, M., Vaillant, F., Simpson, K.J., Stingl, J., Smyth, G.K., Asselin-Labat, M.L.,
Wu, L., Lindeman, G.J., and Visvader, J.E. (2006). Generation of a functional mammary
gland from a single stem cell. Nature 439, 84-88.

Shirreffs, S.M., and Maughan, R.J. (1997). Whole body sweat collection in humans: an
improved method with preliminary data on electrolyte content. J Appl Physiol 82, 336-
341.

Snippert, H.J., Haegebarth, A., Kasper, M., Jaks, V., van Es, J.H., Barker, N., van de
Wetering, M., van den Born, M., Begthel, H., Vries, R.G., et al. (2010). Lgr6 marks stem
cells in the hair follicle that generate all cell lineages of the skin. Science 327, 1385-1389.

160
Soriano, P. (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain.
Nat Genet 21, 70-71.

Spradling, A., Drummond-Barbosa, D., and Kai, T. (2001). Stem cells find their niche.
Nature 414, 98-104.

Srinivas, S., Watanabe, T., Lin, C.S., William, C.M., Tanabe, Y., Jessell, T.M., and
Costantini, F. (2001). Cre reporter strains produced by targeted insertion of EYFP and
ECFP into the ROSA26 locus. BMC Dev Biol 1, 4.

Stingl, J., Eirew, P., Ricketson, I., Shackleton, M., Vaillant, F., Choi, D., Li, H.I., and
Eaves, C.J. (2006). Purification and unique properties of mammary epithelial stem cells.
Nature 439, 993-997.

Szotek, P.P., Chang, H.L., Brennand, K., Fujino, A., Pieretti-Vanmarcke, R., Lo Celso,
C., Dombkowski, D., Preffer, F., Cohen, K.S., Teixeira, J., et al. (2008). Normal ovarian
surface epithelial label-retaining cells exhibit stem/progenitor cell characteristics. Proc
Natl Acad Sci U S A 105, 12469-12473.

Tavazoie, M., Van der Veken, L., Silva-Vargas, V., Louissaint, M., Colonna, L., Zaidi,
B., Garcia-Verdugo, J.M., and Doetsch, F. (2008). A specialized vascular niche for adult
neural stem cells. Cell Stem Cell 3, 279-288.

Taylor, G., Lehrer, M.S., Jensen, P.J., Sun, T.T., and Lavker, R.M. (2000). Involvement
of follicular stem cells in forming not only the follicle but also the epidermis. Cell 102,
451-461.

Terrinoni, A., Smith, F.J., Didona, B., Canzona, F., Paradisi, M., Huber, M., Hohl, D.,
David, A., Verloes, A., Leigh, I.M., et al. (2001). Novel and recurrent mutations in the
genes encoding keratins K6a, K16 and K17 in 13 cases of pachyonychia congenita. J
Invest Dermatol 117, 1391-1396.

Tian, H., Biehs, B., Warming, S., Leong, K.G., Rangell, L., Klein, O.D., and de Sauvage,
F.J. (2011). A reserve stem cell population in small intestine renders Lgr5-positive cells
dispensable. Nature 478, 255-259.

Tosti, A., and Piraccini, B.M. (2003). Nail disorders. Dermatology 1st ed London:
Mosby, 1061-1078.

Trempus, C.S., Morris, R.J., Bortner, C.D., Cotsarelis, G., Faircloth, R.S., Reece, J.M.,
and Tennant, R.W. (2003). Enrichment for living murine keratinocytes from the hair
follicle bulge with the cell surface marker CD34. J Invest Dermatol 120, 501-511.

161
Tumbar, T., Guasch, G., Greco, V., Blanpain, C., Lowry, W.E., Rendl, M., and Fuchs, E.
(2004). Defining the epithelial stem cell niche in skin. Science 303, 359-363.

Turkka, J.T., and Myllyla, V.V. (1987). Sweating dysfunction in Parkinson's disease. Eur
Neurol 26, 1-7.

Urbanek, K., Cesselli, D., Rota, M., Nascimbene, A., De Angelis, A., Hosoda, T., Bearzi,
C., Boni, A., Bolli, R., Kajstura, J., et al. (2006). Stem cell niches in the adult mouse
heart. Proc Natl Acad Sci U S A 103, 9226-9231.

Van Keymeulen, A., Rocha, A.S., Ousset, M., Beck, B., Bouvencourt, G., Rock, J.,
Sharma, N., Dekoninck, S., and Blanpain, C. (2011). Distinct stem cells contribute to
mammary gland development and maintenance. Nature 479, 189-193.

Vasioukhin, V., Degenstein, L., Wise, B., and Fuchs, E. (1999). The magical touch:
genome targeting in epidermal stem cells induced by tamoxifen application to mouse
skin. Proc Natl Acad Sci U S A 96, 8551-8556.

Vincent, M.C., Biancalana, V., Ginisty, D., Mandel, J.L., and Calvas, P. (2001).
Mutational spectrum of the ED1 gene in X-linked hypohidrotic ectodermal dysplasia. Eur
J Hum Genet 9, 355-363.

Weinberg, W.C., Goodman, L.V., George, C., Morgan, D.L., Ledbetter, S., Yuspa, S.H.,
and Lichti, U. (1993). Reconstitution of hair follicle development in vivo: determination
of follicle formation, hair growth, and hair quality by dermal cells. J Invest Dermatol 100,
229-236.

Welm, B.E., Tepera, S.B., Venezia, T., Graubert, T.A., Rosen, J.M., and Goodell, M.A.
(2002). Sca-1(pos) cells in the mouse mammary gland represent an enriched progenitor
cell population. Dev Biol 245, 42-56.

Wenzel, F.G., and Horn, T.D. (1998). Nonneoplastic disorders of the eccrine glands. J
Am Acad Dermatol 38, 1-17; quiz 18-20.

Wilke, K., Martin, A., Terstegen, L., and Biel, S.S. (2007). A short history of sweat gland
biology. Int J Cosmet Sci 29, 169-179.

Wisniewski, S.A., Kobielak, A., Trzeciak, W.H., and Kobielak, K. (2002). Recent
advances in understanding of the molecular basis of anhidrotic ectodermal dysplasia:
discovery of a ligand, ectodysplasin A and its two receptors. J Appl Genet 43, 97-107.

Yang, A., Schweitzer, R., Sun, D., Kaghad, M., Walker, N., Bronson, R.T., Tabin, C.,
Sharpe, A., Caput, D., Crum, C., et al. (1999). p63 is essential for regenerative
proliferation in limb, craniofacial and epithelial development. Nature 398, 714-718.
162

Yuhki, M., Yamada, M., Kawano, M., Iwasato, T., Itohara, S., Yoshida, H., Ogawa, M.,
and Mishina, Y. (2004). BMPR1A signaling is necessary for hair follicle cycling and hair
shaft differentiation in mice. Development 131, 1825-1833.

Zaias, N., and Alvarez, J. (1968). The formation of the primate nail plate. An
autoradiographic study in squirrel monkey. J Invest Dermatol 51, 120-136.

Zhang, J., Niu, C., Ye, L., Huang, H., He, X., Tong, W.G., Ross, J., Haug, J., Johnson, T.,
Feng, J.Q., et al. (2003). Identification of the haematopoietic stem cell niche and control
of the niche size. Nature 425, 836-841.

Abstract (if available)

Abstract Stem cells have long term self-renewal potential and are capable of differentiating into a variety of different cell types. This persistence and multipotency is crucial for the maintenance of tissue during homeostasis and repair. Due to its extraordinary regenerative potential, understanding its maintenance and regulation will prove its usefulness in tissue regeneration and treating various disorders. Thus, scientists have been intensely searching for stem cells in various organs in hopes to understand and take advantage of their regenerative abilities. Among the different organs, the skin is the most easily accessible making it a good model to study stem cells and wound healing. Moreover, the skin is a complex organ containing a number of different “mini-organs”, skin appendages that are likely maintained by independent stem cell populations. Understanding the similarities and differences between these stem cell populations may not only reveal how they maintain and regulate independent appendages, but also how stem cells maintain their own regenerative potential. ❧ The skin contains a number of different skin appendages including hair follicles, sweat glands, and nails. Among these appendages, hair follicles and their stem cells have been the most well characterized while relatively little is known about the presence of stem cells in the sweat gland and nail appendages. Since stem cells must persist throughout life, they have been proposed to be slow cycling cells in order to preserve their self renewing potential and minimize DNA replication errors during cell division. Decades ago, a population of slow cycling label retaining cells (LRCs) was identified in the hair follicle bulge. Numerous subsequent studies have collectively shown that these bulge LRCs are the hair follicle stem cells required for hair follicle regeneration. Moreover, these hair follicle stem cells have been shown to participate in epidermal wound healing during injury. Given that sweat glands are also abundantly distributed in the skin, there is great interest and speculation on whether sweat gland cells can also contribute to wound healing. ❧ Using a K5 driven tetracycline regulated H2B-GFP transgenic system, we have identified distinct LRCs in the nails and sweat glands as putative new skin stem cells. More specifically, LRCs were localized to the ventral proximal fold of the nail and the acinar secretory region of sweat glands. Both nail and sweat gland LRCs were found in the basal layer expressing K15, a known hair follicle stem cells marker. Lineage tracing experiments demonstrate that the K15 cells can contribute long-term to their respective appendages, suggesting long-term self-renewal capabilities found in stem cells. Isolation of live nail and sweat gland LRCs allowed for gene expression profiling to reveal their molecular characteristics, where BMP pathway signaling genes were identified in both sweat gland and nail appendages. Furthermore, we demonstrate the functional requirement of Bmpr1a-mediate signaling in both appendages. Transplanting strips of nail LRCs showed a contribution of H2B-GFP cells to the nail structure. Similar to hair follicle stem cells, sweat gland LRCs can differentiate and contribute to the epidermis under prolonged wound healing conditions. In addition, transplantation of sweat gland cells suggests its plasticity in regenerating sweat glands as well as hair follicles. ❧ In conclusion, we have identified slow cycling LRCs in the sweat gland and nail skin appendages that showed stem cell characteristics. Isolation and characterization of these LRCs show that they share similarities to hair follicle stem cells, but are also unique in possessing their own features. Although more studies are needed for further characterization of these LRCs, these data may shed some light on putative stem cells markers for human nails and sweat glands in the future.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Identification and characterization of adult stem cells in the oral cavity

PDF

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

PDF

Role of the bone marrow niche components in B cell malignancies

PDF

The mechanism by which extracellular Hsp90α promotes cell migration: implications in wound healing and cancer

PDF

Regional localization and regulation of hematopoietic stem cells in the bone marrow stem cell niche

PDF

Role of STAT3 phosphorylation in mouse embryonic stem cell self-renewal and differentiation

PDF

Adaptive patterning during skin formation: assembly of vasculature and adipose tissue

PDF

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

PDF

The calcium-sensing receptor in the specification of normal and malignant hematopoietic cell localization in the bone marrow microenvironment

PDF

Macroenvironmental regulation of hair follicle stem cells

PDF

Pro-regenerative effects of fasting/fasting mimicking diet (FMD) on hematopoietic system and pancreatic islet

PDF

The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

PDF

The function of BS69 in mouse embryogenesis and embryonic stem cell differentiation

PDF

Characterization of the progenitor cell zone in feather follicles

PDF

Shifting topology of stem cells and molecular expression during feather regenerative cycling

PDF

Pleotropic potential of Stat3 in determining self-renewal, apoptosis, and differentiation in mouse embryonic stem cells

PDF

Skin appendage growth control by hormones and morphogens

PDF

Determination of the source of myofibroblasts and therapeutic strategy against fibrosis

PDF

Extracellular Hsp90α in diabetic wound healing: structure, function, and mechanism of action

PDF

The role of HES/HEY transcriptional repressors in specification and maintenance of cell fate in the mouse organ of Corti

Asset Metadata

Creator Leung, Yvonne (author)

Core Title Characterization of new stem/progenitor cells in skin appendages

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 07/19/2015

Defense Date 06/20/2013

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

Tag label retaining cells,nail,OAI-PMH Harvest,sweat glands

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chuong, Cheng-Ming (committee chair), Adams, Gregor B. (committee member), Kobielak, Krzysztof (committee member)

Creator Email leung.yvonne@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-293682

Unique identifier UC11288192

Identifier etd-LeungYvonn-1797.pdf (filename),usctheses-c3-293682 (legacy record id)

Legacy Identifier etd-LeungYvonn-1797.pdf

Dmrecord 293682

Document Type Dissertation

Rights Leung, Yvonne

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

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