The Mechanism by Which Dietary Vitamin a Regulates Skin Stem Cells During Hair Cycling

The mechanism by which dietary vitamin A regulates skin stem cells during hair cycling

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Liye Suo

Graduate Program in Ohio State University Nutrition Program

The Ohio State University

2015

Dissertation Committee:

DR. HELEN B. EVERTS, ADVISOR

DR. AMANDA BIRD

DR. ROBERT CURLEY

DR. EARL HARRISON

DR. TATIANA OBERYSZYN

Copyrighted by

Liye Suo

2015

Abstract

The goal of this study is to identify the mechanisms of vitamin A synthesis and signaling, and its regulation in skin stem cells (SCs) during the cycling hair follicle in health and disease. Precise levels of vitamin A are essential for the health of skin and hair. Both vitamin A deficiency and excess can cause hair loss (alopecia) in animals and humans.

Two previous studies on the role of vitamin A in alopecia areata (AA) and central centrifugal cicatricial alopecia (CCCA) revealed that dietary vitamin A also regulated the hair cycle. We used samples from these two mouse studies to investigate the regulation of dietary vitamin A in skin SCs during hair cycling: C3H/HeJ mice from the AA study and

C57BL/6J (B6) mice from the CCCA study. High levels of vitamin A in a dose-dependent manner increased nuclear localized beta-catenin (CTNNB1; a marker of canonical wingless-type Mouse Mammary Tumor Virus integration site family (WNT) signaling) and levels of WNT7A within the hair follicle bulge in these C3H/HeJ and B6 mice. But excess dietary vitamin A activated bone morphogenetic protein (BMP) signaling in the dermis and silenced WNT signaling. This regulation by vitamin A on BMP/WNT pathways altered hair cycling and could be reversed by lowering vitamin A intake. We also found that RA synthesis protein retinal dehydrogenase 2 (ALDH1A2), cellular RA binding protein 2

(CRABP2), as well as RA degradation enzyme cytochrome p450 26B1 (CYP26B1) co- localized with bulge SCs marker CD34 during telogen by immunofluorescence. These three proteins had different expression patterns during two stages of telogen. We further ii

studied alterations of vitamin A metabolism and signaling in scalp biopsies from CCCA patients. Previously, in B6 mice many key components in vitamin A metabolism and signaling were altered in CCCA including: dehydrogenase reductase member 9 (DHRS9); retinal dehydrogenase 1 (ALDH1A1); cytochrome P450 26A1 (CYP26A1); and retinoic acid receptor beta (RARB). We tested these four proteins in scalp biopsy samples from

CCCA patients. Vitamin A metabolism and signaling decreased as the severity of CCCA increased. Further we developed a potential retinal dehydrogenase (ALDH1A) specific inhibitor, which may have better efficacy and specificity than most of the available

ALDH1As inhibitors in the market. Last but not least, we did IHC against skin SCs markers: CD34, Keratin 15 (Krt 15), CD200 and Blimp1 in the C3H/HeJ and B6 mice; and we reported the different expression levels of skin SCs markers in different mouse models of scarring versus non-scarring alopecia. In conclusion, this study significantly increases our understanding of the potential role of vitamin A in the regulation of skin SCs and its related mechanisms. Taken together, our findings may shed light on skin SCs research and alopecia treatments.

iii

Acknowledgments

It is my great honor to be a buckeye. There are so many great people I need to express my deep and sincere appreciation for their great help and support during my research and life.

Firstly, I would like to thank my mentor and soul guide, Dr. Helen Everts, for her invaluable guidance and support throughout my graduate training. Dr. Everts is the role model for me. Her enthusiasms for research, her passion for teaching and her support for the students continually revealed and reinforced the power of knowledge and research. I deeply appreciate the time I spent together with Dr. Everts, which is and will always be the stimulus and motivation for me to pursue my dream. I was very fortunate to have her as my mentor.

Then, I would like to thank all my other committee members: Dr. Amanda Bird,

Dr. Robert Curley, Dr. Earl Harrison, Dr. Tatiana Oberyszyn. Their help, advice and support also give me the courage and power to solve the problems I had during my Ph.D life and future. Dr. Bird and Dr. Harrison help me a lot in my study and their big smiles are the sunshine for me. I appreciate all the great support from Dr. Curley for the retinal dehydrogenase inhibitor study. His patience and brilliant talent always inspired me to be more dedicated to the research. The last but not the least, I have learned a lot from Dr.

Oberyszyn: not only from the joint skin group meeting she organizes twice a month, but

also from her enthusiasm for science and her caring for the students. It is my great honor to have them together to be my committee members.

Next, I am always proud to be a part of OSUN (OSU nutrition Ph.D program). This

is a big family and we take care of each other. Ph.D training is sometimes tough but always

rewarding; and I consider myself fortunate because I have had the opportunity to work with

so many wonderful colleagues who I can call friends. The time I spent in laboratory is and

will be the invaluable treasure in my mind.

Also I truly appreciate all the great help from Dr. Sundberg and Dr. Wilma Bergfeld

for their support in the mice and human studies.

Finally I need to thank the love and support from my parents, my husband and the

rest of my family and friends.

Vita

1999-2002 ...... Dexing Senior High, China

2002-2010 ...... M.D., Internal Medicine, Peking University

Health Science Center, China

2010 to present ...... Graduate Research Associate, Department

of Nutrition, The Ohio State University

Publications

• Liye Suo, John P. Sundberg and Helen B. Everts. Dietary vitamin A regulates wingless-related MMTV integration site signaling to alter the hair cycle. Exp Biol Med.

Published online 10/30/14 ahead of print. DOI: 10.1177/1535370214557220

• Helen B. Everts, Kathleen A. Silva, Shalise Montgomery, Liye Suo, Monica

Menser, Amy S. Valet, Lloyd E. King, David E. Ong and John P. Sundberg (2013)

Retinoid Metabolism is Altered in Human and Mouse Cicatricial Alopecia. J Invest

Dermatol 133: 325-33.

Fields of Study

Major Field: Ohio State University Nutrition Program

Table of Contents

Abstract ...... ii

Acknowledgments...... iv

Vita ...... vi

Publications………………………………………………………………………….……vi

Fields of Study………………………………………………………………………..…..vi

List of Tables……………………………………………………………………………...x

List of Figures ...... xi

Abbreviations……………………………………………………………...……...……..xiii

Chapter 1: Introduction ...... 1

Chapter 2: Review of literature. ……………………………………………………….....4

2.1 Source of dietary vitamin A and absorption………………………………………..…4

2.2 Vitamin A metabolism and signaling…………………………..……………………..7

2.3 Regulation of vitamin A in skin stem cells (SCs)………………..……………………9

2.4 Effect of dietary vitamin A on hair…………………………..………………………16

2.5 Alopecia Areata (AA)……..……………….…………………………………..…….18

2.6 Central, centrifugal, cicatricial alopecia (CCCA)………………………….………...19

2.7 Retinal dehydrogenase inhibitors………………………………………………….…22

vii

Chapter 3: Dietary vitamin A regulates wingless-related MMTV integration site signaling

to alter the hair cycle ...... 24

3.1 Abstract and Introduction……………………………………………………………24

3.2 Material and Methods………………………………………………………..………27

3.3 Results………………………………………………………………………………..29

3.4 Discussion……………………………………………………………………………31

Chapter 4: Excess dietary vitamin A inhibits anagen initiation by increasing dermal bone morphogenetic protein 4……………………………………………………..…………..34

4.1 Abstract and Introduction……………………………………………………………34

4.2 Material and Methods……………………………………………………………..…38

4.3 Results…………………………………………………………………………….….40

4.4 Discussion……………………………………………………………………………47

Chapter 5: Alterations of Vitamin A metabolism and signaling in central, centrifugal, cicatricial alopecia patients………………………………………………………..….…53

5.1 Abstract and Introduction……………………………………………………………53

5.2 Material and Methods…………………………………………………………..……56

5.3 Results…………………………………………………………………………..……58

5.4 Discussion……………………………………………………………………………64

Chapter 6: Development of specific retinal dehydrogenase inhibitor…………………...67

6.1 Abstract and Introduction……………………………………………………………67

6.2 Material and Methods……………………………………………………………..…69

6.3 Results……………………………………………………………………………..…73

viii

6.4 Discussion……………………………………………………………………………79

Chapter 7: Alteration of skin stem cell markers by dietary vitamin A…………………..84

7.1 Abstract and Introduction……………………………………………………………84

7.2 Material and Methods…………………………………………………………….….86

7.3 Results………………………………………………………………………………..87

7.4 Discussion……...………………………………………………………………….....88

Chapter 8: Conclusions and significance of the study………..……………………….....90

Bibliography……………………………………………………………………………..93

List of Tables

Table 1. Dietary recommendations for vitamin A in different populations...... 6

Table 2.The conversions between RAE to UI...... 6

Table 3. Demographic information of all the patients...... 57

Table 4. IHCL of DHRS9, ALDH1A1, RARB and CYP26A1 in the suprabasal layer, sweat gland, dermis and outer root sheath (ORS) of hair follicle among control and

CCCA patients...... 64

Table 5. Limit of detection (LOD) and limit of quantification (LOQ) of RA analysis in

HPLC ...... 73

List of Figures

Figure 1. Structure of different forms of vitamin A...... 4

Figure 2. Vitamin A metabolism and signaling...... 8

Figure 3. Structure of telogen hair follicle and stem cell markers ...... 11

Figure 4. The standardized central scalp alopecia photographic scale ...... 20

Figure 5. Dietary vitamin A dose dependently altered nuclear localized CTNNB1 and

WNT7A in telogen hair follicles of C3H/HeJ mice...... 30

Figure 6. Dietary vitamin A did not significantly alter BMP4 or TGFB2 in telogen hair follicles of C3H/HeJ mice...... 31

Figure 7. ALDH1A2, CRABP2 and CYP26B1 co-localized with bulge SCs marker

CD34...... 41

Figure 8. ALDH1A2, CRABP2 and CYP26B1 were expressed differently in refractory and competent telogen...... 42

Figure 9. Dietary vitamin A dose dependently altered BMP signaling in telogen hair follicles of C57BL/6J mice...... 44

Figure 10. Dietary vitamin A dose dependently altered the expression of WNT7A but not

WNT7B in telogen hair follicles of C57BL/6J mice...... 45

Figure 11 Dietary vitamin A dose dependently altered nuclear CTNNB1 localization in telogen hair follicles of C57BL/6J mice. ………………………………………………..46

Figure 12. Epidermis expression of DHRS9, ALDH1A1, RARB and CYP26A1 in control and CCCA patients...... 61

Figure 13. Expression of RARB in the sweat gland and dermis among control and CCCA patients...... 62

Figure 14. Expression of DHRS9, ALDH1A1, RARB and CYP26A1 in the sebaceous gland among control and CCCA patients………………………………………………..63

Figure 15. Inhibitory efficacy of different ALDH1As inhibitors in the liver and skin homogenates……………………………………………………………………………..76

Figure 16. Inhibition of RA synthesis by different ALDH1As inhibitors in MDA-MB-468 cells...... 78

Figure 17. Inhibitory percentage (I%) of different ALDH1As inhibitors in MDA-MB-468 cells...... 79

Figure 18. Specificity test of different ALDH1As in other aldehydes in the liver and skin homogenates...... 80

Figure 19. Inhibitory efficacy tests of BIES-DDC and disulfiram in two mouse studies. 82

Figure 20. Expressions of different stem cell markers in B6 mice………………………90

xii

Abbreviations

AA Alopecia areata ABCA1 ATP-binding cassette transporter AIN The American Institute of Nutrition ALDHs Aldehyde dehydrogenases ALDH1A1 Retinal dehydrogenase 1 ALDH1A2 Retinal dehydrogenase 2 ALDH1A3 Retinal dehydrogenase 3 B6 mice C57BL/6J mice BCO1 β-carotene 15,15′-oxygenase BIES-DDC β-ionylidene ethanethiol-diethyldithiocarbamic acid BLIMP1 B lymphocyte-induced maturation protein-1 BMP4 Bone morphogenetic protein 4 BSA Bovine serum albumin CA Cicatricial alopecia CCCA Central centrifugal cicatricial alopecia CD34 Cluster of differentiation 34 CD200 Cluster of differentiation 200 CRABP1 Cellular retinoic acid binding protein 1 CRABP2 Cellular retinoic acid binding protein 2 CTNNB1 Catenin (Cadherin-Associated Protein), Beta 1 CYP26A1 Cytochrome p450 26A1 CYP26B1 Cytochrome p450 26B1 CYP26C1 Cytochrome p450 26C1 DEAB 4-(Diethylamino) Benzaldehyde DGAT1 Diglyceride acyltransferase 1 DHRS9 Dehydrogenase reductase (SDR family) member 9 DISH Diffuse idiopathic skeletal hyperostosis DMEM Dulbecco's Modified Eagle's Medium DMSO Dimethyl sulfoxide EDTA Ethylenediaminetetraacetic acid FBS Fetal bovine serum HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HPLC High-performance liquid chromatography IFA Immunofluorescence IHC Immunohistochemistry IHCL Immunohistochemistry level xiii

IR Immunoreactivity LRIG1 Leucine-rich repeats and immunoglobulin-like domain protein 1 LRG5 Leucine-rich repeat-containing G-protein coupled receptor-5 (LGR5) LGR6 Leucine-rich repeat-containing G-protein coupled receptor6 MTS24+ Mouse thymic stroma 24 NAD+ Nicotinamide adenine dinucleotide NHANES The National Health and Nutrition Examination Survey PCA Primary cicatricial alopecia PPARG Peroxisome proliferator-activated receptors G RA All-trans-retinoic acid RBP4 Retinol binding protein4 RARB Retinoic acid receptor beta RDA Recommended Dietary Allowances SCs Stem cells Scd1 Stearoyl-CoA desaturase-1 SR-BI Scavenger receptor class B type I STRA6 Stimulated by retinoic acid 6 TGFB Transforming growth factor beta UL Upper limit WNT Wingless-related MMTV integration site

xiv

Chapter 1: Introduction

Vitamin A is an essential nutrient for both animals and humans (1-5). Precise levels of vitamin A are required for the health of skin and hair (6-8). Vitamin A and its synthetic forms are called retinoids, which have been widely used to treat many skin diseases, such as acne, psoriasis, and skin cancers (9-11). Also, topical applications of retinoids show promising effects in treating many types of alopecia (12-14). However, the application of retinoids to prevent or treat skin diseases faces limitations, led to retinoid resistance and considerable side effects (15-18). For this reason, it is vital to understand the mechanism

by which vitamin A regulates skin homeostasis.

Vitamin A was found to affect the behavior of the hair follicle and sebaceous gland

in mouse and human studies (8,19,20). Hair growth is a cyclic process, consisting of anagen

(growing phase), catagen (apoptotic phase) and telogen (resting phase). Homeostasis of

skin stem cells (SCs) is essential for hair cycling. All the vitamin A metabolism and

signaling proteins co-localized with skin SCs during the hair cycle (21). Dietary vitamin A levels also dose dependently altered the hair cycle: with high levels (12 & 28 IU/g diet) increasing anagen, but an excessive level (56 IU/g diet) decreasing anagen (6). There is a void of knowledge in the understanding of how dietary vitamin A regulates skin SCs during the hair cycle.

Our long-term goal is to identify the mechanisms of retinoid synthesis and signaling, and

its regulation in skin SCs in hair follicles during cycling and sebaceous glands in health and disease.

Our central hypothesis is that homeostasis of vitamin A metabolism and signaling is

essential for the health of hair; as RA regulates skin SCs during hair cycling, which may

be through the Bone morphogenetic protein (BMP)/ Wingless-related MMTV integration site (WNT) signaling pathways.

We propose three specific aims:

1) To determine the mechanism by which vitamin A affects skin SCs during hair cycle initiation in both C3H/HeJ mice and C57BL/6J (B6) mice. The working hypothesis is that hair follicle SCs are altered in the B6 mice with cicatricial alopecia (CA), not in the

C3H/HeJ mice with alopecia areata (AA); and vitamin A regulates skin SCs during the hair cycle though the BMP/WNT signaling pathways. This was accomplished by examining skin SCs markers in the diseased and non-diseased mice from two studies; and evaluating

BMP and WNT during the hair cycle in C3H/HeJ and B6 mice fed different levels of dietary vitamin A.

2) To evaluate the alterations of vitamin A metabolism and signaling among central, centrifugal, cicatricial, alopecia (CCCA) patients. The working hypothesis is that vitamin A metabolism and signaling are reduced in CCCA patients as severity of the disease increased. This was accomplished by evaluating important components of vitamin

A metabolism and signaling in scalp biopsy samples from CCCA patients with mild, moderate, and severe disease.

3) To develop a specific inhibitor for ALDH1As that does not also inhibit other

ALDHs. This was accomplished by testing the efficacy and specificity of inhibitor

candidates in liver and skin homogenates, cell culture, and mice.

Taken together, our study can significantly increase our understanding of the potential role of vitamin A in the regulation of skin SCs and its related mechanism.

Balancing and restoring normal vitamin A metabolism and signaling may have an essential therapeutic role in treating patients with skin and hair diseases.

Chapter 2: Review of Literature

2.1 Source of dietary vitamin A and absorption

Vitamin A is an essential fat-soluble nutrient for animals and humans. Vitamin A includes retinol, retinal, retinoic acid, retinyl ester, and provitamin A carotenoids (22,23)

(Figure 1). Retinoids is a term used to describe a group of organic compounds that share similar chemical or structural characteristics with retinol. Three generations of retinoids have been widely applied in medicine for more than half a century (15).

Figure 1. Structure of different forms of vitamin A.

The discovery of vitamin A dates back to two hundred years ago when physiologist François Magendie found corneal ulcers and high mortality rates in dogs

deprived of “certain nutrition” (24). Vitamin A regulates many important biological

activities (4,5). Dietary vitamin A has two sources: preformed vitamin A as in foods of

animal origin, such as animal liver, milk, and dairy products, or supplement; and

provitamin A carotenoids as in plant-derived food, such as sweet potato, pumpkin, carrot

and cantaloupe (23).

Preformed vitamin A and provitamin A are digested and absorbed differently (23).

Preformed vitamin A, mainly in the forms of retinyl esters, can be converted into free fatty acids and retinol, which are absorbed by enterocytes. Inside enterocytes, retinol is either re-esterified into retinyl ester to be incorporated into chylomicrons or directly excreted into

the portal circulatory system through the lipid transporter, ATP-binding cassette

transporter (ABCA1) (25). Provitamin A carotenoids are transported into enterocytes

through the scavenger receptor class B type I (SR-BI) (26-28). Inside the enterocytes carotenoids are converted into retinol by β-carotene 15,15′-oxygenase (BCO1) and retinol dehydrogenases (Roldhs). The retinol synthesized from carotenoids is either reassembled into chylomicrons as retinyl ester or directly excreted into the portal circulatory system as mentioned above. Retinyl esters in the chylomicrons are taken up and hydrolyzed back to retinol in the liver (29). The liver is one of the most important sources for vitamin A storage in the body; and balances plasma retinol concentrations (30). Retinol within its binding protein (RBP4) is the most common form of vitamin A in the circulation (31). Plasma retinol concentration is well regulated and maintained within a normal range (32). A plasma retinol level lower than 0.7 mmol/L indicates significant vitamin A deficiency (4).

Intake recommendations for vitamin A are developed by the Food and Nutrition

Board (FNB) at the Institute of Medicine of the National Academies (RDA: recommended

dietary allowance; UL: upper limit; mo: month; y: year; d: day) (4)(Table1).

Age RDA (mcg RAE/d) UL (mcg RAE/d) Infant 0-6 mo 400 600 7-12mo 500 600 Children 1-3 y 300 600 4-8 y 400 900 9-13 y 600 1700 Men (14-70) y 900 3000 Women (14-70) y 700 3000 Pregnant (18-50) y 770 3000 Lactation (18-50) y 1300 3000

Table 1. Dietary recommendations for vitamin A in different populations.

Dietary vitamin A is measured as mcg retinol activity equivalents (RAE) or

international units (IUs): 1 RAE is equivalent to 1 mcg retinol, 12 mcg beta-carotene, and

24 mcg alpha-carotene or beta-cryptoxanthin. The conversions between RAE and UI are listed below (Reference) (Table2).

Retinoids in IU Equivalent conversion in RAE 1 IU retinol 0.3mcg RAE 1 IU beta-carotene from supplements 0.15 mcg RAE 1 IU beta-carotene from food 0.05 mcg RAE 1 IU alpha-carotene or beta-cryptoxanthin 0.025 mcg RAE

Table 2.The conversions between RAE to UI.

Vitamin A deficiency is common health problem; being more frequent in

developing countries. The primary cause of vitamin A deficiency is inadequate intake of

vitamin A. Lower vitamin A intake is commonly seen in certain populations, such as:

children, pregnant women, and patients with impaired lipid absorption (33-37). Dietary vitamin A toxicity is not as common as vitamin A deficiency. Toxicity from vitamin A is usually caused by excessive consumption of supplements containing vitamin A or pharmaceutical doses of synthetic retinoids used for treating certain diseases (38-40).

The recommended dietary vitamin A level for mice is determined by the American

Institute of Nutrition (AIN) as 4 IU/g diet (41,42). Vitamin A levels in the standard unpurified rodent diets are variable (12-28 IU/g diet) and much higher than the recommended level (43).

2.2 Vitamin A metabolism and signaling

Vitamin A metabolism and signaling are well-regulated (Figure 2). Retinol bound to RBP4 is transported into the cells through a multi-transmembrane protein called

Stimulated by Retinoic Acid 6 (STRA6) (31,44,45).

Figure 2. Vitamin A metabolism and signaling.

In the cell, retinol is oxidized to retinoic acid (RA) for vitamin A signaling; and excess retinol can be stored as retinyl ester. RA is the active form of vitamin A and it needs to be maintained at precise levels in the site of action (46,47). Retinol is esterified to retinyl ester for storage by the enzymes lecithin:retinol acyltransferase (LRAT) and diglyceride acyltransferase 1 (DGAT1) (48,49). Alternatively, retinol is reversibly oxidized to retinal by retinol dehydrogenases (50,51). Retinal is further irreversibly oxidized to RA by retinal dehydrogenase 1-3 (ALDH1A1-3) (52). RA bound with different forms of cellular retinoic acid binding proteins (CRABPs) have different fates (53). CRABP1-bound RA is metabolized by cytochrome P450 family members (CYP26A1, CYP26B1 and CYP26C1), 8

which play an important role in protecting cells from vitamin A toxicity (54,55). In contrast, CRABP2-bound RA is transported into the nucleus and is delivered to retinoic acid receptors (RARA, RARB, RARG) (56,57). Activated RARs form the heterodimers

with RXRs, regulating the transcription of over 500 genes involved in proliferation and

differentiation, such as Hoxa1 and RARβ2 (58,59).

2.3 Regulation of vitamin A in skin stem cells (SCs)

As the largest organ in humans, skin plays an essential role in protecting our health against external environmental insults and maintaining internal homeostasis (60). What’s more, it is one of the most important sources for adult stem cells (SCs) (61). There are three identified reservoirs for skin SCs: the interfollicular epidermis (IFE), the hair follicle and the sebaceous gland; the last two combined are called the pilosebaceous unit. The proliferation and differentiation of skin SCs are well balanced, which is important for tissue regeneration and wound healing (62). On the other hand, abnormalities of skin SCs can lead to skin disorders such as impairment of regeneration after injury, alopecia, and skin cancer (63-65). These diseases cause a huge economic burden to patients; in addition, some diseases are fatal (15). Therefore understanding the mechanisms by which skin SCs are regulated is very important.

Currently, these SC pools can be distinguished by the expression of characteristic molecular markers (Figure 3) (61). From the bottom of hair follicle to the top of the epidermis, these SC markers include leucine-rich repeat-containing G-protein coupled receptor-5 (LGR5), cluster of differentiation 34 / keratin 15 (CD34/K15), leucine-rich repeat containing heterotrimeric guanine nucleotide-binding protein-coupled receptor6

(LGR6), B lymphocyte-induced maturation protein-1 (Blimp1), leucine-rich repeat

protein-1/ mouse thymic stroma 24 (LRIG1/MTS24) and β1 integrin. LGR5+ SCs reside in

the hair germ. These cells actively cycle and are long-lived. They are also the precursor of all the cell lineages of hair follicles below the sebaceous glands (66). CD34+/K15+ SCs

localize in the bulge of hair follicle, contribute to the cell lineage of both hair follicle and

sebaceous gland, but not to that of IFE under normal physiological conditions. Above the

bulge area, LGR6+ SCs are distributed at the central portion of isthmus, and contribute to

all the cell lineages of hair follicle. Further upwards are a group of unipotent sebaceous

gland progenitors (Blimp1-expressing cells), derived from the bulge stem cells; and they

give rise to all of the cells of the sebaceous gland (67). The quiescent pool of LRIG1+ stem cells are mainly localized in the infundibulum; and are only responsible for the cell lineage of sebaceous gland and IFE. In addition, β1 integrin-expressing cells represent a group of basal SCs residing in the basal layer of the IFE and play an essential role in the regulation of the homeostasis of the epidermis (62). Together the various skin SC populations in the isthmus specify the upper hair follicle and IFE, while the SCs of the bulge and germ specify the cycling portion of the hair follicle during homeostasis.

Figure 3. Structure of telogen hair follicle and stem cell markers (61). While the signaling pathways that regulate the proliferation and differentiation

behaviors of skin SCs still are waiting for full investigation, many mechanisms have been

discovered. The hair follicle is a good mini-organ to study skin SCs homeostasis due to its self-cycling ability. Hair cycling goes through a regeneration phase (anagen) involving growth and differentiation, an apoptotic phase (catagen) represented by the loss of the

lower parts of the hair follicle, and a resting phase (telogen) (68). The cyclic process of

regeneration and degeneration of the hair follicle is the result of skin SC cycling. SCs in

the hair follicle remain quiescent most of the time during our life (69). These SCs become

activated briefly for the transition from telogen to anagen to start the next new hair cycle

(70,71). SCs in the bulge have been identified as having the potential to regenerate the hair

follicle as shown by lineage tracing experiments in vivo (72). Bone morphogenetic protein

(BMP) and wingless-related MMTV integration site (WNT) signaling are two well-studied

pathways that regulate the hair cycle. Some researchers have divided telogen into two

stages according to the expression of BMP in the dermis: BMP positive or refractory

telogen and BMP negative or competent telogen (73). Hair follicles cannot be activated to

begin anagen when they are in refractory telogen. With decreased BMP4 signaling in the

dermis and BMP2 signaling in the subcutaneous fat, telogen goes from the refractory stage

to the competent stage. BMP4 directly inhibits WNT7A and WNT7B, two important

ligands for the WNT signaling pathways in hair cycling. WNT inhibitors within the SCs

all act to keep hair follicle SCs in the quiescent state of telogen (74-76). Activation of these

SCs occur when transforming growth factor beta 2 (TGFB2) and Noggin inhibit BMP signaling and WNT ligands are secreted from the mesenchyme derived dermal papilla, which sits below the SCs during telogen (77,78). WNT signaling then activates hair follicle

SCs to induce anagen (70,79). At the telogen to anagen transition, BMP pathway inhibition and WNT signaling activation together stabilize nuclear β-catenin (CTNNB1) in the bulge niche, which triggers the SCs to overcome the inhibitory thresholds and start the new hair follicle (70,77,80).

Blimp1 is the specific marker for the SCs in the sebaceous gland (67). Hedgehog signaling is the best-characterized pathway involved in sebocyte proliferation: its activation results in sebocyte proliferation (81) and its inhibition can suppress sebocyte

development (82). The mechanisms that regulate the SCs in the pilosebaceous unit are

complicated and still need further investigation to fully unravel the complexity.

Vitamin A is involved in the regulation of many SC activities. Studies in zebrafish

have shown that vitamin A plays an important role in embryonic development and adult

organ regeneration after injury by inducing SC differentiation (83,84). RA also induces the differentiation of human embryonic and bone marrow SCs both in vitro and in vivo (85,86).

There are a few studies investigating the regulation of vitamin A in the skin SCs (87,88).

Retinoids are one of the most widely used agents in dermatology, and they have

been used successfully to treat many skin diseases, such as acne, psoriasis, and skin cancers

(15). Retinoids can affect the behavior of the hair follicle and sebaceous gland in both mice

and human studies (8,19). The mechanism for how retinoids regulate the skin SCs remains

unclear. Proteins involved in RA synthesis and signaling localized to all sites of skin SCs in a hair cycle-dependent manner, with a peak occurring during mid-anagen through early catagen (21). Research on transgenic mice also provides evidence that RA plays an important role in the homeostasis of skin SCs. Li et al. (2001) (20) selectively disrupted the mouse gene for RXRalpha, the most abundant in the 3 RXR isoforms in the epidermis and hair follicle keratinocytes. RA signaling blockage in these mice led to delayed anagen initiation, progressive alopecia, destruction of HF and an inflammatory reaction in the skin.

Shih et al. (2009) (8) demonstrated that mice with the enzyme acyl-CoA:diacylglycerol

acyltransferase 1 (DGAT1) ablated specifically in the epidermis, had increased levels of

RA and retinol, exhibited accelerated transition from telogen to anagen, and showed

progressive alopecia (8).This evidence suggests that the precise level of RA is essential for

hair follicle SC homeostasis. In the sebaceous gland, RA synthesis was also present in the

SCs with the specific BLIMP1+ marker (21). Retinoids have been successfully applied to

treat acne (11,89,90). Although the mechanism of action of retinoids in acne treatment is still unclear, evidence suggests that regulation of sebaceous SCs by RA may play an

important role. In cultured human sebocytes, both vitamin A deficiency and higher level

of RA led to decreased cell proliferation and lipogenesis (91,92). Precise levels of RA are

required for seboctyes in the sebaceous gland and future studies are needed to understand

the regulation of RA synthesis and signaling in the sebaceous gland.

Studies in human and mouse melanoma cell lines have shown that retinoids could

regulate the behavior of melanoma cells, inducing differentiation and inhibiting

proliferation (93). Retinoids have been found to suppress or reverse epithelial

carcinogenesis in vitro and in vivo, including skin cancers and lung squamous cell

carcinomas (94) in humans and animals models (95-97). Oral retinoids represent one of the

most promising tools for the chemoprevention of non-melanoma skin cancers (NMSCs)

(15). In addition, supplementation of vitamin A was linked to a decreased risk of melanoma

in a recent large prospective cohort study (98). The mechanism of its efficacy on skin

cancer treatment has not been fully understood. Recent studies have provided evidence to

support a hair follicle SC origin for a number of skin cancer types (99-102). Multipotent

tumor-initiating cells (TICs), like normal SCs have unlimited capacity for self-renewal and

proliferation. It is suggested that the efficacy of RA on skin cancers is dependent on its

regulation of SCs in skin.

Despite the promising preclinical and early clinical research, the application of retinoids to prevent or treat skin diseases faces limitations, caused by retinoid resistance and considerable side effects (15). Many studies reported that systemic retinoids have no significant protective effect in patients with multiple NMSCs and melanoma, and more toxicity was noted in the treatment group, such as dryness of the skin and mucous membranes, diffuse idiopathic skeletal hyperostosis (DISH), and teratogenicity (15). For this reason, it is vital to understand the mechanisms by which retinoids regulate the homeostasis of skin SCs, as this information might lead to more specific and less toxic treatment strategies for skin diseases. Recent evidence suggests that the reduction or silencing of RA signaling could be found in many types of carcinoma (103-105). RAR β2 expression is lost early in the process of tumorigenesis in many types of cells (106). RAR

β2 silencing induced by promoter methylation and histone deacetylation are believed to result in retinoid resistance in cancer treatment (103). Many studies have shown that restoring normal RA signaling can greatly increase the efficacy of treatment in many types

of cancer (107-109).

Most of the studies on how vitamin A regulates skin SCs have been done in cancers.

Two previous studies on the role of vitamin A in AA and CCCA revealed that dietary

vitamin A also regulated the hair cycle (6,7). We used samples from these two mouse

studies to investigate the regulation of dietary vitamin A in skin SCs during hair cycling:

C3H/HeJ mice from the AA study and B6 mice from the CCCA study.

2.4 Effect of dietary vitamin A on hair

Dietary vitamin A is important for the health of skin and hair (21), since both dietary

vitamin A deficiency and excess can lead to alopecia (hair loss) in animals and humans (6-

8,110-112). Vitamin A deficiency caused follicular hyperkeratosis and alopecia in rodents,

cattle, and humans (113-116). Although vitamin A deficiency is a global public health problem, the related skin and hair abnormality have seldom been reported because more

attention was paid to other more severe consequences that are a result of vitamin A deficiency, such as blindness and infections. In our previous study, we found that vitamin

A deprivation caused more severe disease in C3H/HeJ mice, a mouse model for AA (7).

Vitamin A toxicity also causes alopecia in animals and humans. Cyclical alopecia in the anagen hair follicle develops in keratinocyte specific acyl-CoA:diacylglycerol acyltransferase 1 (DGAT1) null mice. DGAT1 is one of the enzymes for the storage of excess retinol and protects the body from vitamin A toxicity. Keratinocyte specific DGAT1 null mice had more anagen hair follicles and moderate increase of vitamin A in the skin. A

Vitamin A deficient diet induced more telogen hair follicles and reversed the hair loss in these mice (8). Vitamin A toxicity also induced hair loss in humans. Retinoids are widely used in dermatology (15,117-119). The systemic therapeutic initiation doses of retinoid usually start at 0.2mg/kg/d (120,121), which is much higher than the upper limit (UL) of dietary vitamin A in adults (UL=0.3mg/RAE day). Retinoid-associated alopecia is one of the most psychologically distressing side effects caused by systemic retinoid application

(111,122). Synthetic retinoids acitretin and etretinate caused alopecia at doses between

25mg-75mg/d. Retinoid-induced alopecia is dose-dependent and hair loss is reversible

after significant reduction of retinoid dose or cessation of the systemic therapy (111,122-

125). It is still unknown what the mechanism is involved in the retinoid-induced alopecia.

There are two possible reasons: 1) Retinoid-induced alopecia results in an arrest of hair follicles at the onset of anagen and an anchorage factor defect in the follicle during telogen

(111); 2) Retinoid-induced alopecia also reduced anagen in cultured human hair follicles by inducing TGF- β2 and triggering catagen (126).

Two previous studies on the role of vitamin A in AA and CCCA revealed that dietary vitamin A regulated the hair cycle dose dependently (6,7). The hair cycle consists of a growing phase (anagen), apoptosis phase (categen), and resting phase (telogen); transition from telogen to the next anagen initiates a new hair cycle (127). High levels of vitamin A (12 & 28 IU/g diet) increased the percentage of anagen hair follicles, which are more vulnerable to the inflammatory attack in AA and CCCA; and caused more severe disease. In contrast, excess level of vitamin A (56 IU/g diet) decreased anagen and increased telogen hair follicles, which are resistant to CCCA (6,7). These data indicated that dietary vitamin A dose-dependently regulates the hair cycle, which affects the progression of AA and CCCA.

The interactions between RA and BMP are tissue dependent (128,129). Studies in genetically modified mice demonstrated that RA inhibited BMP4 during inner ear (130) and forelimb development (131), but induced the expression of BMP4 during genital tubercle development (132). RA also required BMP4 up regulation for the induction of interdigital cell death during limb development (133). In addition, RA and BMP4 acted synergistically to regulate differentiation of keratinocytes from both embryonic and adult

skin SCs in vitro (134,135). The interactions between RA and WNT signaling are also tissue dependent. CTNNB1 was shown to be an indirect target gene of RA signaling, and

RA can activate or suppress the WNT canonical signaling pathway (59,136). Byer et al.

(1996) (137) and Ryuto et al. (1997) (138) demonstrated that RA treatment could induce dephosphorylation and accumulation of CTNNB1 in both human breast and renal cancer cell lines. In addition, RA inhibits WNT signaling in both embryonic and adult SCs, and this negative regulation is important for normal embryogenesis and homeostasis of adult

SCs (139-141). In the skin, WNT signaling is also regulated by RA; and RA treatment inhibited WNT signaling and induced the regression of keratoacanthomas (142). These interactions suggest that RA may regulate hair cycling through BMP/WNT signaling.

2.5 Alopecia Areata (AA)

Alopecia areata (AA) is thought to be an autoimmune, non-scarring, hair loss disease. It presents as well defined patches, most commonly appearing on the scalp and often resolving spontaneously (63). AA greatly affects patients’ life quality and self-esteem

(143). Although AA is one of the most common autoimmune diseases, affecting 4.5 million people in the United States (144), the pathogenesis of it is not fully understood, and the recent available therapy is disappointing (145).

With the development of the AA skin graft mouse model, growing evidence of the

AA pathogenesis has been discovered (146). Recently, it was identified that both genetic and environmental factors contribute to the pathogenesis of AA. Many genes are mutated in AA patients. These genes are grouped into Human leukocyte antigen (HLA)-related genes, inflammatory factor and receptor-related genes, and hair follicle specific genes

(147-149). In evolving AA, histological peri- and intrafollicular inflammation of the lower portions of only anagen hair follicles, primarily by T lymphocytes, is a common feature in both human patients and the mouse models (150,151). It is known that in AA, the SCs in the bulge are not affected, and the inflammatory injury involves the bulbar region of the hair follicle and does not lead to permanent hair loss (152). Since the SCs in the bulge are not destroyed in AA, if treated appropriately, some patients could achieve complete remission, although this ideal treatment has not been found.

2.6 Central, centrifugal, cicatricial alopecia (CCCA)

Primary cicatricial alopecias (CA) are a collection of scarring hair loss diseases

(153). CCCA is the most common cause of irreversible alopecia in African American

women (154). Inflammatory cells infiltrate the hair follicle and destroy skin SCs, ultimately leading to permanent hair loss (155). It was first reported as hot comb alopecia by Lopresti et al. half a century ago because hot comb treatments were considered the cause for this alopecia (156). The term finally was changed to CCCA by the North American Hair

Research Society in 2003 (157). CCCA usually starts at the center of the scalp and

symmetrically spreads to the peripheral regions. At the late stage of CCCA, follicular SCs

may be damaged by lymphocytic inflammation and lead to irreversible scarring hair loss

(157). CCCA can be graded as mild, moderate and severe by the standardized central scalp

alopecia photographic scale (158) (Figure 4). Histopathologic characteristics of CCCA

include: perifollicular lymphocyte infiltration, early desquamation of inner root sheath and

fibrosis formation due to follicular rupture (159). The data on the prevalence of CCCA in

women of African descent is variable and dependent on the studied population and how

the participants were recruited: Khumalo et al. reported 1.9% in Africa (160); Olsen et al.

found a prevalence of 5.6% among 3 randomly selected church groups in Ohio, Maryland and North Carolina (153); and a survey conducted by Kyei et al. showed a 17% prevalence

among 326 African American women at 2 churches and a health fair at the Cleveland

Clinic. However, the CCCA diagnosis in the last study was not confirmed by scalp biopsy

(161).

Both environmental and genetic factors may contribute to the disease (162). The

mechanisms that result in such a destructive reaction have not been fully elucidated, thus,

currently there are no effective treatments for CCCA (163). There are many mouse models

for studying the mechanisms of CA (159,164,165).

Figure 4. The standardized central scalp alopecia photographic scale (158). (0: normal; 1 and 2: mild; 3 and 4: moderate; 5: severe)

Asebia mutant mice, with the mutation of stearoyl-CoA desaturase-1 (Scd1), are

one of the animal models used to study CA (166). Conditional (skin specific) Scd1 null

mice have increased skin retinol, retinoic acid and retinoic acid related gene expression

(167). The skin specific Peroxisome proliferator-activated receptor gamma (PPARG) null

mouse, with follicular specific deletion of PPARG, is another animal model for CA.

PPARG is critical for lipid homeostasis and maintaining the normal function of the

sebaceous gland; and specific deletion of PPARG in keratinocytes also cause the atrophy of sebaceous gland, lymphocytic inflammation, and permanent hair loss (165). These animal studies revealed that a primary dysfunction of the sebaceous gland or primary follicular dystrophy caused the secondary inflammatory destruction of hair follicle and

SCs. Evidence from animal studies also showed that CCCA only attacked anagen and

catagen hair follicles and spared telogen hair follicles (6,159). There are other theories of

the pathogenesis of CCCA including: hair shaft twist and penetration of the outer root

sheath, which in turn triggers the peri-follicular inflammatory destruction (159); reduced

sebum in the sebaceous gland (168); and the loss of immune privilege in the hair follicles

(169).

Retinoids play an important role in regulating the replication and differentiation of

skin SCs (170), which regulates hair cycling (98,171). Although retinoids have not yet been

applied to the treatment of scarring alopecia under clinical guidelines, many case reports

haven been published to show the possible potential of retinoids in treating many types of

scarring alopecia (172,173). In addition, we found that Vitamin A plays an important role

in the development of CCCA in the B6 mouse model (6,159). In this study, a 7 times the

recommended level of dietary vitamin A (28 IU/g diet) was associated with more severe

CCCA, but a 14 times the recommended level of dietary vitamin A (56 IU/g diet) caused

more telogen hair follicles and less severe disease (6). These data suggest that a precise

level of vitamin A is essential for the maintenance of healthy hair. Balancing and restoring

normal vitamin A metabolism may have an essential therapeutic role in treating CCCA

patients.

2.7 Retinal dehydrogenase inhibitors

The aldehyde dehydrogenases (ALDHs) are a group of 19 enzymes that oxidize

many aldehydes into related carboxylic acids (174). Retinal dehydrogenases (ALDH1A1,

ALDH1A2, ALDH1A3) specifically convert retinal into all-trans retinoic acid (RA), the

active form of vitamin A (52). The RA signaling pathway is essential for embryogenesis

and other biological activities (59). In addition, ALDH1As have been identified to play a

role in cancer resistance to chemotherapy; and inhibition of ALDH1A1 would be a target

to solve the cancer resistance problem (175-177).

Many chemicals have been identified as effective ALDH inhibitors including: disulfiram (tetraethylthioperoxydicarbonic diamide), DEAB (4-diethylamino

benzaldehyde) and citral (3,7-dimethyl-2,6-octadienal). Disulfiram is a widely used drug

to treat alcohol addiction by inhibiting ALDH2 in the liver (178); also disulfiram can

inhibit ALDH1A1 in vitro and in vivo (179). DEAB is another confirmed ALDH1As

inhibitor (180); but an in vivo study suggests that it would also inhibit ALDH2 (181). Citral

is thought to be a specific ALDH1As inhibitor and it has been used to study the role of RA

in embryogenesis and tissue development (182,183). However, citral failed to reduce RA synthesis in our previous mouse study. In conclusion, none of these chemicals are a specific

ALDH1As inhibitor (178,180,184-186).

Chapter 3: Dietary vitamin A regulates wingless-related

MMTV integration site signaling to alter the hair cycle

Published in Exp Biol Med. Published online 10/30/14 ahead of print. DOI:

10.1177/1535370214557220

3.1 Abstract and Introduction

Abstract

Alopecia areata (AA) is an autoimmune hair loss disease caused by a cell mediated

immune attack of the lower portion of the cycling hair follicle. Feeding mice 3-7 times the

recommended level of dietary vitamin A accelerated the progression of AA in the graft-

induced C3H/HeJ mouse model of AA. In this study we also found that dietary vitamin A,

in a dose dependent manner, activated the hair follicle stem cells (SCs) to induce the growth

phase of the hair cycle, which may have made the hair follicle more susceptible to

autoimmune attack. Our purpose here is to determine the mechanism by which dietary

vitamin A regulates the hair cycle. We found that vitamin A in a dose dependent manner

increased nuclear localized beta-catenin (CTNNB1; a marker of canonical WNT signaling)

and levels of WNT7A (wingless-related MMTV integration site 7A) within the hair follicle bulge in these C3H/HeJ mice. These findings suggest that feeding mice high levels of dietary vitamin A increases WNT signaling to activate hair follicle SCs.

Introduction

Precise level of vitamin A is essential for the health of skin and hair, as both dietary

vitamin A deficiency and excess can lead to alopecia (hair loss) in animals and humans (6-

8,110-112). One hair loss disease Dr. Everts previously examined is alopecia areata (AA).

We found that feeding high levels of (3-7 times recommended level) vitamin A accelerated

disease, while feeding no vitamin A made the disease more severe (7). The hair cycle

consists of a growing phase (anagen), apoptosis phase (catagen), and resting phase

(telogen); transition from telogen to the next anagen initiates a new hair cycle (127). The

cause of AA is debatable, but ultimately anagen hair follicles are destroyed by a cell

mediated immune attack. Inflammatory cells infiltrate and attack predominantly anagen

hair follicles, and cause breakage of defective hair shafts and hair loss (15,187,188).

Vitamin A plays many roles in the immune system (48), which could be contributing to

our previous results. In addition to these immune system effects, we found that feeding

high levels of vitamin A also altered the hair cycle (7).

We found that high vitamin A levels increased the percentage of hair follicles in anagen. The lower cycling portion of the anagen hair follicle is primarily attacked in AA.

Therefore high vitamin A levels may have accelerated the disease by increasing anagen induction or extending the length of anagen and therefore the amount of time the hair follicle could be attacked. This finding needs to be further examined in patients with AA, as most humans with AA have fewer hair follicles in anagen (167). We also reported that

RA synthesis and signaling protein localization within the hair follicle was altered during

the hair cycle (127,189), although few studies have examined how the hair cycle is

regulated by vitamin A.

The hair cycle is maintained through the regulation of SCs in the hair follicle bulge

(69). Several factors regulate hair follicle SCs to ultimately regulate the hair cycle. Bone

morphogenetic protein 2 (BMP2) from the subcutaneous fat, BMP4 from the dermis,

BMP6 from the inner layer of the hair follicle, and several Wingless-related MMTV integration site (WNT) inhibitors within the SCs all act to keep hair follicle SCs in the

quiescent state of telogen (74-76). Activation of these SCs occur when transforming growth factor beta 2 (TGFB2) and Noggin inhibit BMP signaling and WNT ligands are secreted from the mesenchyme derived dermal papilla, which sits below the SCs during telogen

(77,78). WNT signaling then activates hair follicle SCs to induce anagen (70,79). The

canonical WNT signaling pathway involves the translocation of beta-catenin (CTNNB1)

to the nucleus to activate transcription of target genes (190). WNT7A and WNT7B are two

of the 19 members of the WNT secreted protein family that are important for initiation of

anagen and both are directly inhibited by BMP signaling (191,192). RA regulates members

of both the WNT and TGFB/BMP families in other tissues (132,133,193-195), but it is

unknown which of these factors is regulated by RA in the hair follicle.

Our purpose here is to determine the mechanism by which dietary vitamin A

regulates hair follicle SCs to induce anagen. We used control samples from our previous

vitamin A feeding study in C3H/HeJ mice (7). We found that both nuclear localized

CTNNB1 and WNT7A were dose dependently increased by dietary vitamin A. These findings extend the knowledge of vitamin A signaling in the skin SCs.

3.2 Material and Methods

3.21 Animals

Mice: To induce AA, recipient C3H/HeJ mice (The Jackson Laboratory, Bar

Harbor, ME) received a dorsal skin graft as previously described from an older C3H/HeJ mouse that had spontaneously developed AA (sent from John P. Sundberg) (7). Control mice had their own dorsal skin removed and rotated 180 degrees in a sham surgery.

Recipient and control C3H/HeJ mice were fed an AIN93 Maintenance diet containing 0

(n=8), 4 (n=7), 12 (n=9), or 28 (n=8) IU vitamin A/g diet (Research Diets, Inc., New

Brunswick, NJ) starting two weeks before these mice were grafted. The American Institute of Nutrition (AIN) recommends 4 IU vitamin A/g diet (41). The higher doses were based on vitamin A analysis of the diet fed mice in The Jackson Laboratory production facility, which contained 6–28 IU vitamin A/g diet. Samples from the control sham mice collected

15 weeks post grafting were used in this current study. The Ohio State University

Institutional Animal Care and Use Committee (IACUC) approved all procedures

(#2008A0044).

3.22 Immunohistochemistry (IHC)

IHC was performed using a three-antibody system in paraffin slides as described previously (127). The antibodies to the following proteins were used: CTNNB1 (Millipore,

Billerica, MA), WNT7A (Abcam, Cambridge, MA), WNT7B (Novus Biologicals,

Littleton, CO), BMP4 (Vector Lab, Burlingame, CA), and TGFB2 (Santa Cruz,

Biotechnology, Santa Cruz, CA). Biotinylated anti-rabbit and anti-mouse were used as secondary antibodies (Jackson Immunoresearch Laboratories, Inc., West Grove, PA)

followed by a horseradish peroxidase conjugated anti-biotin tertiary antibody (Bethyl

Laboratories, Montgomery TX). The red 3-amino-9-ethylcarbazole plus enhancers (AEC+;

Dako, Carpinteria, CA) substrate chromogen was used followed by counterstaining with

Gils Hematoxylin III (Poly Scientific, Bay Shore, NY) and mounting in aquamount (Dako,

Capinteria, CA). CTNNB1 sections were not counterstained to better visualize nuclear localization. Immunoreactivity (IR) was scored blinded on a 4 point scale with: 0, negative;

1, weak; 2, moderate; 3, strong; and 4, very strong. An IHC level (IHCL) was then calculated for nuclear CTNNB1 for each telogen hair follicle by the following formula:

IHCL= ((# cells with IR of 0/total cells) x 0)

+ ((# cells with IR of 1/total cells) x 1)

+ ((# cells with IR of 2/total cells) x 2)

+ ((# cells with IR of 3/total cells) x 3)

+ ((# cells with IR of 4/total cells) x 4).

The IHCL of all of the telogen hair follicles were then averaged to get one IHCL per mouse. With this scoring method the maximum IHCL is 4. The scoring system is modified from validated protocols (196-198). Blinded IHCLs for WNT7A, WNT7B, and

TGFB were calculated similarly, but each hair follicle received one IR score and the formula summed the fraction of hair follicles with each IR score. Again the maximum

IHCL is 4. A blinded IHC score of BMP4 was measured using the total number of BMP4 positive cells in the dermis around telogen hair follicles divided by the number of telogen hair follicles. Images were captured on an upright Olympus BX51 microscope with the

DP71 camera (Olympus, Tokyo, Japan) and processed in Adobe Photoshop CS4 (San

Francisco, CA).

3.23 Statistical Analysis

Comparison among experimental groups was analyzed by one-way ANOVA, followed by Tukey post hoc tests when appropriate using SPSS, v21 (IBM; Armonk, NY).

P value <0.05 is defined as statistical significance.

3.3 Results

To determine the mechanisms by which vitamin A regulates hair follicle SCs we examined several factors known to regulate the hair cycle by IHC in mice fed various levels of vitamin A. Nuclear CTNNB1 is an important downstream component and marker of the active canonical WNT signaling pathway. We found that dietary vitamin A dose dependently increased the nuclear localization of CTNNB1 in C3H/HeJ mice with a significant increase occurring with as little as 4 IU vitamin A/g diet (p <0.001) and an additional significant increase with 28 IU vitamin A/g diet (Figure 5a-e). WNT7A IR also increased with dietary vitamin A in a dose dependent manner with C3H/HeJ mice fed 28

IU vitamin A/g diet being significantly greater than mice fed no vitamin A (p <0.05, Figure

5f-j). WNT7B IR was not altered by the level of dietary vitamin A (Figure 5k-o). Dietary vitamin A also did not significantly alter either BMP4 or TGFB2 (Figure 6).

Figure 5. Dietary vitamin A dose dependently altered nuclear localized CTNNB1 and WNT7A in telogen hair follicles of C3H/HeJ mice. Immunohistochemistry was performed with antibodies against CTNNB1 (a–e), WNT7A (f–j), and WNT7B (k–o) in dorsal skin from C3H/HeJ mice fed 0 (a, f, k), 4 (b, g, l), 12 (c, h, m), and 28 (d, i, n) IU vitamin A/g diet. Blinded IHC score for CTNNB1 and WNT7A, WNT7B were calculated as the immunoreactivity (IR) measured on a 0–4 scale multiplied by the fraction of cells within the telogen hair follicle and the fraction of telogen hair follicles. Mean ± SD. Different letters are significantly different, P < 0.05. Bar=10.1 µM.

Figure 6. Dietary vitamin A did not significantly alter BMP4 or TGFB2 in telogen hair follicles of C3H/HeJ mice. Immunohistochemistry was performed with antibodies against BMP4 (a–e) and TGFB2 (f–j) in dorsal skin from C3H/HeJ mice fed 0 (a, f), 4 (b, g), 12 (c, h), and 28 (d, i) IU vitamin A/g diet. A blinded IHC score of BMP4 was measured using the total number of BMP4 positive cells in the dermis around telogen hair follicles divided by the number of telogen hair follicles. Blinded IHC score for TGFB2 was calculated as the immunoreactivity (IR) measured on a 0–4 scale multiplied by the fraction of cells within the telogen hair follicle and the fraction of telogen hair follicles. Mean ± SD. Bar=10.1 µM. 3.4 Discussion

Precise levels of vitamin A are important for the development and homeostasis of

many tissues, including skin and hair (5,21,199). Dietary vitamin A has a narrow range

between RDA and UL (4). Recommended dietary vitamin A for mice is 4 IU vitamin A/g

diet, and 28 IU vitamin A/g diet (7 times of the recommended level) is the highest level

measured in the diet of mice in The Jackson Laboratory production facility (6). Our

previous study found that dietary vitamin A (from 0 IU-28 IU vitamin A/g diet) induced anagen in a dose dependent manner in C3H/HeJ mice (7). Increased anagen was also seen

in skin specific Dgat1tm2FarTg (KRT14-cre) 1AMC null mice, which have elevated retinol and RA levels. In our current study, we explored the mechanism by which dietary vitamin

A regulates hair cycling.

We found that dietary vitamin A increased WNT signaling, as indicated by nuclear

CTNNB1, and the ligand WNT7A. We also observed that the alteration of WNT7A

expression by dietary vitamin A and that of CTNNB1 nuclear localization were similar in

C3H/HeJ mice. However a significant increase in CTNNB1 nuclear localization occurred

at a lower dose of dietary vitamin A. This suggests that other signals in addition to WNT7A

may play a role in directing CTNNB1 to the nucleus. While WNT7A and WNT7B are

essential to activate hair follicle bugle SCs (191,192), there are 19 members of the WNT

secretion protein family that play different roles during hair follicle development and

cycling (200). Future studies could analyze different WNTs and WNT inhibitors.

The interactions between RA and WNT signaling are tissue dependent (128,129).

During skeletal myogenesis, osteogenesis, and adipogenesis. RA activates WNT signaling

(193,194). Yet during limb development RA inhibits Wnt7a (195). Our study shows that

RA is activating WNT7A and WNT signaling within the adult skin in vivo. Several Wnts

and WNT inhibitors were also increased in the skin of E16.5 and E18.5 Cyp26b1 null

embryos (201), suggesting that RA regulates WNTs in the developing hair follicle as well.

Future studies are needed to determine whether RA is acting directly on WNT7A within

the hair follicle.

In summary, we demonstrated here that dietary vitamin A activates hair follicle

SCs dose dependently through the activation of WNT signaling. Our next step is to study

the mechanism by which retinoids regulate WNT. We will also establish the healthy range

of retinoids for the hair follicle, which could provide possible information for maintaining

healthy hair in the normal population and determining effective doses of retinoids in treating other hair loss diseases.

Chapter 4: Excess dietary vitamin A inhibits anagen initiation

by increasing dermal bone morphogenetic protein 4

4.1 Abstract & Introduction

Abstract

Precise levels of vitamin A are essential for the health of skin and hair. Central centrifugal cicatricial alopecia (CCCA) is the most common form of scarring hair loss in

African American women. We found that dietary vitamin A impacts the severity of disease in the C57BL/6J (B6) mouse model of CCCA. Systemic retinoid therapy also leads to alopecia. In addition to altering disease, we found that excess dietary vitamin A (56 IU vitamin A/g diet, 14 times of the recommended level for mice) reduced anagen. Our purpose here is to determine the mechanism by which excess dietary vitamin A inhibits anagen initiation. We found that RA synthesis proteins retinal dehydrogenase 2

(ALDH1A2) and cellular RA binding protein 2 (CRABP2), as well as RA degradation enzyme cytochrome p450 26B1 (CYP26B1) co-localized with the bulge SCs marker CD34 during telogen by immunofluorescence. These three proteins had different expression patterns during the two stages of telogen. We also found that B6 mice fed excess vitamin

A had significantly more bone morphogenetic protein 4 (BMP4) positive cells in the dermis compared with the other two diet groups. Lowering dietary vitamin A levels in dams caused lower BMP4 expression, higher WNT7A, CTNNB1 nuclear localization and a higher

percentage of anagen hair follicles in the pups. This study is the first to provide evidence

that excess dietary vitamin A inhibits anagen initiation through, at least partially, increasing

dermal BMP4 expression, which inhibits WNT signaling and prevents anagen induction.

These findings may shed light on alopecia treatments.

Introduction

Dietary vitamin A is important for the health of skin and hair (21); since both

dietary vitamin A deficiency and excess can lead to alopecia (hair loss) in animals and

humans (6-8,110-112). Synthetic retinoids acitretin and etretinate caused alopecia at doses

between 25-75 mg/d (122,123,125). Retinoid-associated alopecia is dose-dependent and hair loss is reversible after significant reduction of retinoid dose or cessation of the systemic therapy (111,122-125). Retinoid-induced alopecia results in an arrest of hair

follicles at the onset of anagen and an anchorage factor defect in the follicle during telogen

(111). Retinoid-induced alopecia also reduced anagen in cultured human hair follicles by

inducing TGF- β2 and triggering catagen (126). The mechanisms by which vitamin A

regulates the transition from telogen to anagen are still unknown.

Vitamin A is an essential nutrient for humans. The active form of vitamin A is all-

trans retinoic acid (RA) (46,47). RA is synthesized at or near the site of action, which is

well regulated by a few key enzymes (43). In the cell, retinol is either esterified to retinyl

esters for storage by the enzymes lecithin:retinol acyltransferase and diglyceride

acyltransferase 1 (DGAT1) (48,49); or retinol is reversibly oxidized to retinal by retinol dehydrogenases (50,51). Retinal is further irreversibly oxidized to RA by retinal dehydrogenase 1-3 (ALDH1A1-3) (52). RA bound with different forms of cellular retinoic

acid binding proteins (CRABPs) have different fates (53). CRABP1-bound RA is metabolized by the cytochrome P450 family members (CYP26A1, CYP26B1 and

CYP26C1), which play an important role in protecting cells from vitamin A toxicity

(54,55). In contrast, CRABP2-bound RA is transported into the nucleus and delivered to retinoic acid receptors (RARA, RARB, RARG) (56,57). RA bound RARs activate the transcription of over 500 genes involved in proliferation and differentiation (58,59).

In our previous study in C57BL/6J (B6) mice, we found that dietary vitamin A dose dependently altered hair cycle: high levels of vitamin A (12-28 IU/g diet) increased the percentage of anagen hair follicles and an excess level of vitamin A (56IU/g diet) decreased anagen and increased telogen hair follicles (6). Here, we used B6 mice to study the mechanism by which dietary vitamin A regulates hair cycle.

The hair follicle has a self-cycling ability. It goes through a regeneration phase

(anagen), a degenerative phase (catagen) involving the apoptosis and loss of the lower two-

thirds of the hair follicle and a resting phase (telogen) (68). The transition from telogen to anagen requires the activation and proliferation of SCs (69). This transition is regulated by interactions between: epithelial SCs of the hair follicle bulge; mesenchymal cells in the dermal papilla and dermis; and subcutaneous fat (75,192,202). Bone morphogenetic protein (BMP) and wingless-related MMTV integration site (WNT) signaling are the two major pathways that regulate the hair cycle, although other factors are also important

(73,79). Plikus et al. (2008) (73) pointed out that according to the expression of BMPs, telogen hair follicles can be divided into two stages: refractory and competent telogen.

With decreased BMP4 signaling in the dermis and BMP2 signaling in the subcutaneous

fat, telogen goes from the refractory stage to the competent stage. Increased WNT signaling

in the hair follicle bulge accompanies this switch to competent telogen. WNT signaling

activates two different pathways: the canonical pathway (beta-catenin, CTNNB1,

dependent) and the non-canonical pathway (CTNNB1 independent). WNT7A and WNT7B are directly inhibited by BMP signaling and relief of this inhibition is important for the initiation of the new hair cycle (191). WNT7A and WNT7B are two of the 19 members of the WNT secreted protein family and transduce their signal by both canonical and non- conical pathways (200,203,204). Together, BMP pathway inhibition and WNT signaling

activation triggers the SCs to overcome the inhibitory thresholds and start the new growth

phase (70,77,202). In other cell types, BMP signaling is regulated by RA (128,129,133),

yet this regulation is still not fully investigated in skin SCs. These interactions provide the

clue that RA may regulate the behavior of hair follicle SCs through BMP signaling.

Our purpose here is to determine how excess dietary vitamin A regulates hair cycling. We used samples from our two previous vitamin A feeding studies in B6 mice (6).

We report here that important vitamin A signaling components co-localized with bulge SCs and dermis in different stages of telogen. In addition, dietary vitamin A increased dermal

BMP signaling. These findings expand the knowledge of why excess vitamin A alters alopecia.

4.2 Material and Methods

4.21 Animals

Mice: The Jackson Laboratory (Bar Harbor, ME) and The Ohio State University

(Columbus, OH) Institutional Animal Care and Use Committees approved all procedures.

C57BL/6J mice (The Jackson Laboratory, Bar Harbor, ME) were used.

Diet study 1 mice: C57BL/6J (B6) Dams and female pups were fed the NIH 31

diet (LabDiet 5K52, Purina Mills, St Louis, MO) until 12 weeks of age, and then switched

to an AIN93 maintenance diet containing 4 (n=4), 28 (n=5), or 56 (n=7) IU vitamin A/g

diet (Research Diets, Inc., New Brunswick, NJ) for 16 weeks (6).

Diet study 2 mice: B6 Dams and their first generation female pups were fed an

AIN93 growth diet containing 4 IU vitamin A/g diet (Research Die, Inc. New Brunswick,

NJ). Second generation female pups were used in this study. Six week old B6 mice were

fed an AIN93 Maintenance diet containing 4 (n=7), 28 (n=15), or 56 (n=6) IU vitamin A/g

diet (Research Diets, Inc., New Brunswick, NJ) for 12 weeks (6).

4.22 Immunohistochemistry (IHC)

IHC was performed in paraffin slides as described previously (127). The following

antibodies were used: primary antibodies to the following proteins were CTNNB1

(Millipore, Billerica, MA), WNT7A (Abcam, Cambridge, MA), WNT7B (Novus

Biologicals, Littleton, CO), BMP4 (Vector Lab, Burlingame, CA), and TGFB2 (Santa

Cruz, Biotechnology, Santa Cruz, CA). Biotinylated anti-rabbit and anti-mouse (Jackson

Immunoresearch Laboratories, Inc., West Grove, PA) were used as secondary antibody.

Tertiary antibody was a horseradish peroxidase conjugated anti-biotin (Bethyl

Laboratories, Montgomery TX). CTNNB1 sections were not counterstained to better

visualize nuclear localization. Blinded IHCLs for nuclear CTNNB1, WNT7A, WNT7B,

and TGFB were calculated as described in Chapter 3 (170). Blinded IHC score for

CTNNB1, WNT7A, WNT7B and TGFB were calculated as the immunoreactivity (IR) measured on a 0–4 scale multiplied by the fraction of cells within the telogen hair follicle and the fraction of telogen hair follicles. A blinded IHC score of BMP4 was measured using the total number of BMP4 positive cells in the dermis around telogen hair follicles divided by the number of telogen hair follicles. Images were captured on an upright

Olympus BX51 microscope with the DP71 camera (Olympus, Tokyo, Japan) and processed in Adobe Photoshop CS4 (San Francisco, CA).

4.23 Immunofluorescence (IFA)

IFA was performed using a two-antibody co-localization system in frozen dorsal skin sections as previously described (7). The following primary antibodies were used with the indicated concentrations: ALDH1A2 (Rabbit, 1:500, produced in the Ong laboratory),

CRABP2 (Rabbit, 1:500, produced in the Ong laboratory), CYP26B1 (Rabbit, 1:1000,

ACRIS, Chapel Hill, NC), BMP4 (Mouse, 1:200, Vector Lab, Burlingame, CA), CD34

(Rat, 1:200, BD Pharmingen, San Jose, CA,), CD207 (Rat, 1:500, eBioscience, San Diego,

CA). Secondary antibodies (anti-rat, anti-rabbit, anti-mouse 1:5000, Invitrogen, Carlsbad,

CA, USA) conjugated with Alexa Fluor 488 or Alexa Fluor 594. Nuclei were stained with

DAPI (0.05μl/ml, Invitrogen, Carlsbad, CA, USA) for 1min and subsequently the sections were washed and coverslipped with Prolong Gold antifade reagent (Molecular Probes,

Eugene, OR, USA). Pictures were taken at 20X and 60X magnification (Nikon ACT-1,

Tokyo) and processed in Adobe Photoshop CS4 (San Francisco, CA).

4.24 Statistical Analysis

Comparison among experimental groups was analyzed by one-way ANOVA, followed by Tukey post hoc tests when appropriate using SPSS, v21 (IBM; Armonk, NY).

P value <0.05 is defined as statistical significance.

4.3 Results

Vitamin A metabolism proteins co-localize to CD34 positive bulge stem cells during

telogen

CD34 is the specific SC marker for the murine bulge (205). It has been used to

show the location of bulge SCs by IFA in frozen sections of B6 murine skin. We found that

the RA synthesis enzyme (ALDH1A2), the binding protein (CRABP2), and the

degradation enzyme (CYP26B1) co-localized with CD34 in B6 mouse skin (Figure 7a-c).

Figure 7. ALDH1A2, CRABP2 and CYP26B1 co-localized with bulge SCs marker CD34. ALDH1A2 (a, green), CRABP2 (b, green) and CYP26B1 (c, green) co-localized with CD34 (red) in the bulge of C57BL/6J mice, yellow arrow. Bar=10.1µM.

BMP4 was used as a marker for refractory telogen to further identify the stage of

telogen hair follicles expressing ALDH1A2, CRABP2, and CYP26B1 (73). ALDH1A2 localized to the bulge when BMP4 was positive in the dermis (Figure 8a). ALDH1A2 also co-localized with BMP4 in the dermis, although ALDH1A2 did not localize to all BMP4 positive cells. CRABP2 localized to the bulge when BMP4 was negative in the dermis

(Figure 8d), while CYP26B1 was localized to the bulge under both situations (Figure 8e

and f). We further demonstrated that CYP26B1 localized to bulge and dermis throughout

telogen. These results were also confirmed by IHC with antibodies against ALDH1A2,

CRABP2, or CYP26B1 versus BMP4 in serial sections of B6 mouse skin 18, 21, and 24

days after mice were depilated to induce and synchronize the hair cycle (data not shown).

Figure 8. ALDH1A2, CRABP2 and CYP26B1 were expressed differently in refractory and competent telogen. ALDH1A2 (a,b red) localized to the bulge and ALDH1A2 immunoreactivity was greatest when BMP4 (green) was positive in the dermis (a, refractory telogen). CRABP2 (c,d, red) localized to the bulge when BMP4 was negative in the dermis (d, competent telogen); CYP26B1 (e, f, red) localized to the bulge and dermis throughout telogen and co-localized with BMP4 . Bar=10.1µM. Green arrow, co- localization with BMP4 in dermis; yellow arrow, bulge SCs.

Dietary vitamin A altered BMP4 expression in the dermis surrounding telogen hair

follicles

Our previous study showed that an excess level of dietary vitamin A (56 IU/g diet)

significantly decreased the percentage of anagen hair follicles in study 1; however, there

was no difference in the percentage of anagen hair follicles among the three groups (4 IU/g

diet; 28 IU/g diet; 56 IU/g diet) in study 2 (6). To better understand the mechanisms by

which dietary vitamin A regulates the hair cycle in B6 mice, we examined BMP signaling

and WNT signaling expression by IHC in the skin of B6 mice fed various levels of vitamin

A. We found that feeding mice excess vitamin A (56 IU vitamin A/g diet) significantly increased BMP4 positive cells in the dermis (p <0.05) (Figure 9a-d); and significantly decreased WNT7A (Figure 10a-d) and nuclear localization of CTNNB1 (Figure 11a-d) in the hair follicle (p <0.05) compared with the other two groups in the study (4 and 28 IU

vitamin A/g diet). The expression of TGFB2 (Figure 9i-l) and WNT7B (Figure 10i-l) was

similar among three groups.

Lowering dietary vitamin A for two generations altered BMP and WNT signaling

expression in the second generation pups in the study 2. IHCL of BMP4 (Figure 9g) and

TGFB2 (Figure 9o) of the 56 IU/g diet group in the study 2 were much less than those in

the study 1 (p <0.05). IHCL of nuclear CTNNB1 (Figure 11g) of the 56 IU/g diet group in

the study 2 was significantly higher than that in the study 1 (p < 0.05). WNT7A (Figure

10e-h) expression increased along with the dietary vitamin A levels. The expression of

WNT7B (Figure 10m-p), nuclear CTNNB1 (Figure 11e-h), BMP4 (Figure 9e-h) and

TGFB2 (Figure 9m-p) were similar among all three groups in study 2.

Figure 9. Dietary vitamin A dose dependently altered BMP signaling in telogen hair follicles of C57BL/6J mice. Immunohistochemistry was performed with antibodies against BMP4 (a– h, green arrow) and TGFB2 (i–p) in dorsal skin from C57BL/6J mice fed 4 (a, e, i, m), 28 (b, f, j, n) and 56 (c, g, k, o) IU vitamin A/g diet in the study 1 (a-d and i-l) and study 2 (e-h and m-p). A blinded IHC score of BMP4 was measured using the total number of BMP4 positive cells in the dermis around telogen hair follicles divided by the number of telogen hair follicles. Blinded IHC score for TGFB2 was calculated as the immunoreactivity (IR) measured on a 0–4 scale multiplied by the fraction of telogen hair follicles. Different letter are significantly different, P < 0.05. Mean ± SD. Bar=25 µM.

Figure 10. Dietary vitamin A dose dependently altered the expression of WNT7A but not WNT7B in telogen hair follicles of C57BL/6J mice. Immunohistochemistry was performed with antibodies against WNT7A (a–h) and WNT7B (i–p) in dorsal skin from C57BL/6J mice fed 4 (a, e, i, m), 28 (b, f, j, n) and 56 (c, g, k, o) IU vitamin A/g diet in the study 1 (a-d and i-l) and study 2 (e-g and m-p). Blinded IHC scores of WNT7A and WNT7B were calculated as the immunoreactivity (IR) measured on a 0–4 scale multiplied by the fraction of telogen hair follicles. Different letter are significantly different, P < 0.05. Mean ± SD. Bar=25 µM. 45

Figure 11 Dietary vitamin A dose dependently altered nuclear CTNNB1 localization in telogen hair follicles of C57BL/6J mice. Immunohistochemistry was performed with antibodies against CTNNB1 (a-h) in dorsal skin from C57BL/6J mice fed 4 (a, e), 28 (b, f) and 56 (c, g) IU vitamin A/g diet in the study 1 (a-d) and study 2 (e-h). Blinded IHC score of CTNNB1 was calculated as the IR measured on a 0–4 scale multiplied by the fraction of cells within the telogen hair follicle. Different letter are significantly different, P < 0.05. Mean ± SD. Bar=10.1 µM.

4.4 Discussion

Health of skin and hair requires a specific level of dietary vitamin A (5,21,199).

Retinoids are widely applied in treating many skin diseases including: skin cancers, acne

vulgaris, and psoriasis (11,15,206). In contrast, retinoid-associated side effects in hair and

skin, such as alopecia, have received more and more attention (111,126,207). Evidence

from both animal and human studies has demonstrated that precise concentrations of

dietary vitamin A are essential to maintain normal hair cycling (6,21,208). Our previous

studies in B6 mice found that an excess level of dietary vitamin A (56 IU/g diet)

significantly reduced anagen; and lowering the dietary vitamin A levels for two generations

reversed this effect in the second generation pups (6). In our current study, we explored the

mechanism by which dietary vitamin A regulates hair cycling. We report that RA biosynthesis and signaling proteins co-localize with hair follicle SCs in different stages of telogen; and excess vitamin A blocks the initiation of the hair cycle by regulating dermal

BMP levels. Excessive levels of dietary vitamin A (56 IU/g diet, 14 times the recommended level) increased dermal BMP signaling, which inhibited WNT signaling in the bulge and arrested the hair follicle in telogen. Lowering vitamin A by breeding mice on low vitamin A reversed this effect. To our knowledge, this is the first study to show that dietary vitamin A alters BMP signaling in the skin.

This is the first report of the localization of RA metabolism proteins in different stages of telogen. Telogen is divided into two stages by the dermal expression of BMPs:

BMP positive telogen or refractory telogen, and BMP negative telogen or competent telogen. (73). Our study suggests that during telogen endogenous RA is synthesized in the bulge SCs and a few dermal cells during refractory telogen. RA signaling occurs within the bulge SCs during competent telogen; since RA induces the expression of CRABP2

(209,210). In addition, the concentration of RA is well controlled in both the bulge and dermis; as CYP26B1 localized to both locations during both stages of telogen. We saw a similar temporal localization pattern of RA synthesis and signaling in the dermal papilla,

which sits at the base of the hair follicle and regulates differentiation, during mid-anagen

(127). The ALDH1A2 positive dermal cells may be dermal dendritic cells; as we

previously co-localized ALDH1A2 with the dendritic cell marker langerin/CD207 (7).

Cyp26b1 mRNA was also seen in the epidermis, dermis, and dermal papilla during embryonic development (201,211), but its localization pattern in the adult has not been reported. The expression of CYP26B1 within the CD34 positive bulge cells in our study may be one reason why dermis specific (En1tm2(cre)Wrst/J) Cyp26b1tm1Hh null mice have a

lower induction of Rarb, no increases in Crabp1, Crabp2, or Stra6, and a less severe

phenotype compared with germ line Cyp26b1tm1Hh null mice (201). These data suggest that:

RA synthesis occurs in refractory telogen; RA signaling occurs in competent telogen since

CRABP2 is induced by RA; and RA is degraded throughout telogen.

Previous studies suggested that the BMP4 signal from the dermis and the WNT7A signal from hair follicle SCs together control the behavior of hair follicle cycling (73,212).

WNT7A is directly inhibited by BMP signaling and relief of this inhibition is important for initiation of the new hair cycle (191). Our localization pattern suggests that when RA exceeds the capacity of CYP26B1 it could act in an autocrine and/or paracrine manner to increase the expression of dermal BMP4 during refractory telogen. This could occur when mice are fed excess dietary vitamin A or when pharmacological doses of retinoids are given orally. This increased BMP4 would then inhibit WNT7A and WNT signaling resulting in an arrest in the hair follicle in refractory telogen, which is what we saw in study 1.

Therefore vitamin A regulates the hair cycle in a spatial and temporal manner.

We also found that the inhibitory effect of excess dietary vitamin A on dermal

BMP4 expression could be reduced when vitamin A levels were reduced by breeding mice on low vitamin A diets. Obrochta et al. (2014) (43) found that when dams were fed lower dietary vitamin A for three generations, pups had significantly lower RA concentrations in liver, kidney, testes and white fat tissue, although alterations of RA concentrations in skin were not measured. These reductions in RA occurred regardless of changes in retinol and/or retinyl ester levels. In our studies, dams in study 2 were fed lower dietary vitamin

A (4 IU vitamin A/g diet) for two generations compared to dams in study 1 (NIH 31 diet,

>20 IU vitamin A/g diet). We did not measure RA in these mice, but we saw an increase in liver retinol after two generations on the low vitamin A diet similar to what Obrochta et al. (2014) (43) saw in B6 mice, suggesting that RA levels in the skin may also be reduced in our mice in study 2. Feeding dams less dietary vitamin A: reduced dermal BMP4 expression; increased CTNNB1 nuclear localization; and increased anagen hair follicles

(6) in pups fed excess dietary vitamin A (56 IU vitamin A/g diet). Moreover, a threshold of RA levels greater than the capacity of CYP26B1 may exist that was not reached in study

skin SCs in vitro (134,135). During skeletal myogenesis RA directly inhibited Bmp4 (194).

Our study shows that RA may also be regulating BMP4 signaling within the adult skin in vivo. Future studies are needed to determine the mechanisms by which RA is regulating

BMP4 in the dermis.

RA also regulates WNT signaling in many tissues. CTNNB1 was proved to be an

indirect target gene for RA signaling, and RA can activate or suppress the WNT canonical

signaling pathway (59,136). Byer et al. (1996) and Ryuto et al. (1997) demonstrated that

RA treatment could induce dephosphorylation and accumulation of CTNNB1 in both

human breast and renal cancer cell lines. In addition, RA inhibits WNT signaling in both

embryonic and adult SCs, and this negative regulation is important for the normal

embryogenesis and homeostasis of adult SCs (139-141). In the skin, RA treatment inhibited

WNT signaling and induced the regression of keratoacanthomas (142). Our study reported that dietary vitamin A dose dependently (0-28 IU/g diet) increased WNT7A and nuclear

CTNNB1 within the skin in C3H/HeJ mice (170); however we did not find these in B6 mice in study 1. This suggests that there is also a strain difference in vitamin A’s regulation of WNT signaling. B6 mice have a polymorphism in alcohol dehydrogenases 7 (Adh7), which may be important for retinoid metabolism (6). Reduced function of Adh7 in B6 mice may have caused them to be more vulnerable to vitamin A toxicity. In addition,

WNT7A, but not nuclear CTNNB1, was increased by dietary vitamin A in B6 mice in study 2. WNT signaling activates two different pathways: canonical pathway (CTNNB1, dependent) and non-canonical pathway (CTNNB1 independent). WNT7A is one of the 19 members of the WNT secreted protein family and transduces its signal by both canonical

and non-canonical pathways (200,203,204). Our findings indicated that some non-

canonical pathways induced by WNT7A may be important for hair cycling in B6 mice.

Future studies are needed to determine the mechanisms involved in this strain difference

of dietary vitamin A regulation of WNT signaling.

Retinoid-induced hair loss is one of the most difficult obstacles for the

pharmaceutical application of retinoids (111,126). This adverse effect is dose-dependent

and hair loss is reversible after significant reduction of retinoid dose or cessation of the systemic therapy (124,125,193). In humans, retinoid-induced hair loss was caused by an arrest of hair follicle in telogen; followed by a defect in follicle anchorage (111). We also found that excess dietary vitamin A (56 IU vitamin A/g diet) in study 1 significantly reduced anagen hair follicles; but the B6 mice had no anchorage defects and this actually protected the mice. Here we found that excessive levels of dietary vitamin A increased dermal BMP expression, which inhibited WNT signaling in the bulge and arrested the hair follicle in telogen. This finding may also provide some knowledge in preventing and treating retinoid-induced alopecia.

In summary, we found here that retinoids regulate initiation of the hair cycle in a spatial and temporal manner by regulating dermal BMP4 (Figure 9). In addition, this effect was dose dependent. The understanding of mechanisms by which excess vitamin A regulates hair cycling can expand the knowledge of the mechanisms involved in hair cycling regulation and also help provide possible interventions for the patients suffering with various types of alopecia when treated with systemic retinoids. It is important to

further study the mechanism by which retinoids regulate BMP and translate it to human research.

Chapter 5: Alterations of Vitamin A metabolism and signaling

in central, centrifugal, cicatricial alopecia patients

5.1 Abstract and Introduction

Abstract

Central, centrifugal, cicatricial alopecia, or CCCA is the most common scarring

hair loss among African American women. Although the pathogenesis of CCCA is

unknown, vitamin A plays an important role in the development of CCCA. Previously, in

mice we found that many key components in vitamin A metabolism and signaling were

altered in CCCA including: DHRS9 (dehydrogenase reductase member 9); ALDH1A1

(retinal dehydrogenase 1); CYP26A1 (cytochrome P450 26A1); and RARB (retinoic acid

receptor beta). Their expression increased in mild disease and decreased in severe disease.

The purpose of this study was to examine the possible alteration of those proteins among

CCCA patients. Reviewed were African American women diagnosed with CCCA at the

Cleveland Clinic in the past eight years. These patients were initially clinically evaluated with a standardized central scalp alopecia photographic grading scale that graded the disease severity. The diagnosis of CCCA and severity was confirmed histologically. The control subjects were African American women diagnosed with pilar cyst. In total, we had

11 mild disease, 11 moderate disease, 5 severe disease and 12 controls.

Immunohistochemistry (IHC) on all the scalp biopsy samples using antibodies against 53

DHRS9, ALDH1A1, CYP26A1, and RARB were examined. The results demonstrated a decrease of all four proteins in the basal layer of the severe group as compared to controls.

Also found was a decrease of RARB expression in the sweat glands and dermis in the severe group as compared to the mild group. These findings suggest that the expression of important vitamin A metabolism and signaling components in the skin decreases as severity of the CCCA increased. These findings expand the knowledge of pathogenesis of

CCCA and emphasize the importance of vitamin A metabolism and signaling in the health of skin and hair.

Introduction

Primary cicatricial alopecia (PCA) is a group of hair loss diseases with unknown causes. They are characterized by permanent damage to hair follicles and scarring

(cicatricial) formation (145). CCCA, one kind of PCA, is most commonly observed in middle-aged African American women (154). It was first termed hot comb alopecia in 1968 because hot comb treatments were thought to be the cause of this disease at that time (156).

In 2003, the “hot comb alopecia” was changed to CCCA by the North American Hair

Research Society because it was identified that hair styling such as hot combs were not associated with the development of CCCA (157). CCCA usually starts at the center of the scalp and symmetrically spreads to the peripheral regions. At the late stage of CCCA, follicular SCs are damaged by lymphocytic inflammation, which leads to irreversible hair loss (157). CCCA can be graded as mild, moderate and severe by the standardized central scalp alopecia photographic scale (158).

Both environmental and genetic factors may contribute to the disease (162). There are many theories about the pathogenesis of CCCA including: hair shaft twist and

penetration out of the outer root sheath, which trigger peri-follicular inflammatory

destruction (159); loss of functional peroxisome proliferator activated receptor gamma and

lipid metabolism imbalance in the hair follicle (165); reduced sebum in the sebaceous gland

(168); and the loss of immune privilege in the hair follicle (169).

Vitamin A also plays an essential role in the development of CCCA (6,159). Both

vitamin A deficiency and toxicity lead to alopecia in mice and humans (8,110-112).

Flowers et al. (2011) (167) found that increased skin retinol, retinoic acid and retinoic acid

related gene expression were observed in mice with a skin-specific deletion of stearoyl-

CoA desaturase-1 (Scd1), which is one of the animal models for PCA. Our lab showed that

a higher intake of dietary vitamin A was associated with more severe CCCA in a different

mouse model, C57BL/6J (B6) mice (6). In addition, the expression of vitamin A

metabolism proteins were associated with disease severity in these B6 mice, increasing in

mild disease and dropping in severe disease. In a small number of human samples we also

saw alterations in the expression of vitamin A metabolism genes, but there were not enough

samples to separate them based on disease severity. These data suggest that a precise level

of dietary vitamin A is essential for the maintenance of healthy hair. Balancing and

restoring normal vitamin A metabolism may have an essential role in treating CCCA

patients.

Vitamin A metabolism and signaling are well-regulated. The majority of retinol

entering keratinocytes is esterified to retinyl esters for storage (213). Retinoic acid (RA) is

the active form of vitamin A and its synthesis occurs at or near the site of action in a

regulated manner (46,47). RA synthesis involves a two-step reaction: in the first step,

retinol is reversibly oxidized to retinal by retinol dehydrogenases (RDH10, dehydrogenase

reductase, SDR family, member 9; DHRS9) (50,51,214,215); and in the second step, retinal is irreversibly converted to RA by retinal dehydrogenase 1-3 (ALDH1A1-3) (52).

Synthesized RA bound to retinoic acid receptors alpha, beta, and gamma (RARA, RARB,

RARG) serve as transcription factors, regulating the expression of over 500 genes (58,59).

Excess RA is degraded by cytochrome P450 family members (CYP26A1, CYP26B1 and

CYP26C1), which play an important role in protecting cells from vitamin A toxicity

(54,55). In our previous study in B6 mice that spontaneously developed CCCA in The

Jackson Laboratory (Bar Harbor, Maine) production facility, we found that key

components in vitamin A metabolism and signaling: DHRS9, ALDH1A1, CYP26A1 and

RARB were increased in mild disease and non-detectable in severe disease (6). But in B6 mice fed a purified diet for 4 months we observed reduced expression of RARB, in the hair follicle and epidermis with increased severity of disease (6). DHRS9 was reduced in the epidermis, but not the hair follicle. In addition, neither protein was changed by the severity of disease in the sebaceous gland. We also noted that the follicular dystrophy in the diet- controlled studies was less severe than what we saw in the mice randomly collected.

Our purpose here is to identify the alterations of those proteins in the skin of CCCA

patients. We used scalp biopsy samples from African American CCCA patients at the

Cleveland Clinic. We report here that expression of important vitamin A metabolism and signaling components in the skin decreases as severity increases. These findings expand

the knowledge of pathogenesis of CCCA and provide clues for the management of this disease.

5.2 Material and Methods

5.21 Human samples

Our study was approved by the Institutional Review Board of the Cleveland Clinic in Cleveland, OH. The standardized central scalp alopecia photographic scale (158) was used to grade the disease severity in the Cleveland Clinic. Reviewed were hematoxylin and eosin-stained sections from those cases with the confirmed pathologic diagnosis of central, centrifugal, cicatricial alopecia (years 2006-2014). Paraffin blocks were sectioned from 11 mild CCCA cases, 11 moderate CCCA cases, and 5 severe CCCA cases. We used scalp biopsies from African American female patients diagnosed with pilar cyst as control

(n=12) (year 2010-2014). All CCCA patients and control were African American women

(Table 3).

Group Number Ethnicity Gender Age Biopsy location

Control 12 African American Female 49.8±14.0 Scalp

Mild 11 African American Female 50.0±10.0 Scalp

Moderate 11 African American Female 51.3±10.8 Scalp

Severe 5 African American Female 58.0±4.4 Scalp

Table 3. Demographic information of all the patients: number, ethnicity, gender, age (Average ± SD), biopsy location.

5.22 Immunohistochemistry (IHC)

IHC was performed using a three-antibody system in paraffin slides as described

previously (189). Antibodies to the following proteins were used: RARB (Santa Cruz,

Biotechnology, Santa Cruz, CA), CYP26A1 (Alpha Diagnostic, San Antonio, TX),

ALDH1A1 and DHRS9 were produced within Dr. Ong’s laboratory (Vanderbilt University

Medical Center) and their specificity was confirmed previously (127). Biotinylated anti-

rabbit was used as the secondary antibody (Jackson Immunoresearch Laboratories, Inc.,

West Grove, PA) followed by a horseradish peroxidase conjugated anti-biotin tertiary

antibody (Bethyl Laboratories, Montgomery TX). The percentage of positive cells was

graded blinded as follows: 0 (none), 1 (1–25%), 2 (26–50%), 3 (51–75%), and 4 (76–

100%). Immunoreactivity (IR) was scored blinded on a 4 point scale with: 0, negative; 1,

weak; 2, moderate; 3, strong; and 4, very strong. An IHC level (IHCL) was calculated as

the sum of the percentage of positive cells plus the IR by the following formula:

IHCL= score of percentage of positive cells + score of IR of positive cells as per

Peng et al (196).

With this scoring method, the maximum IHCL is 8. For each slide, the scores of six locations of the skin were collected: basal layer of epidermis, supra-basal layer of

epidermis, sweat gland, sebaceous gland, hair follicle and dermis. Images were captured

on an upright Olympus BX51 microscope with the DP71 camera (Olympus, Tokyo, Japan)

and processed in Adobe Photoshop CS4 (San Francisco, CA) (170).

5.22 Statistical Analysis

Comparison among experimental groups was analyzed by Kruskal-Wallis nonparametric test followed by Mann-Whitney post hoc tests using SPSS, v22 (IBM;

Armonk, NY). P value <0.05 is defined as statistical significance.

5.3 Results

To determine the alterations of vitamin A metabolism and signaling in CCCA disease in humans, we examined the retinoid metabolism proteins that had been altered in mice with CCCA among control, mild, moderate and severe CCCA patients at the

Cleveland Clinic. The proteins included: DHRS9, ALDH1A1, CYP26A1 and RARB (6).

We found that DHRS9 in the basal layer of the epidermis was significantly decreased in both moderate (n=11) and severe (n=5) groups compared with controls (n=12) (Figure 12 a-d and r). DHRS9 expression was similar among all four groups in other skin locations

(Table 4). ALDH1A1 showed a similar expression pattern, but was only significant in the severe group compared to controls (Figure 12e-h and s) (Table 4). CYP26A1 significantly dropped in all patients with CCCA compared with controls in the basal layer (n=12) (Figure

12m-p and u). CYP26A1 expression was also significantly reduced in severe disease compared to mild disease. CYP26A1 significantly decreased in the severe group compared to controls in the suprabasal layer (Figure 12m-p and Table 4). Expression of RARB in the basal layer dropped as severity increased. RARB in the suprabasal layer significantly decreased in diseased groups compared with controls; but there was no difference among diseased groups (Figure 12i-l and Table 4). Also RARB expression in the sweat gland

(Figure 13a-d and m) and dermis (Figure 13e-l and n) significantly decreased in the severe

disease (n=5) compared with mild disease (n=11) (Table 4). All four retinoid metabolism proteins were strongly expressed in the sebaceous gland regardless of severity, and there was no difference between control (n=1) and CCCA (n=15) (Figure 14). The numbers of sweat glands, hair follicles and sebaceous glands were not significantly altered by the severity of CCCA, although there was a decreasing trend of the number of the sebaceous glands with the progression of CCCA.

Figure 12. Epidermis expression of DHRS9, ALDH1A1, RARB and CYP26A1 in control and CCCA patients. IHC was performed with antibodies against DHRS9 (a-d and r), ALDH1A1(e-h and s), RARB (i-l and t), CYP26A1 (m-p and u) in the scalp biopsies from control (a, e, i and m) and CCCA patients (mild: b, f, j and n; moderate: c, g, k, o; severe: d, h, l and p). A blinded IHCL of all four proteins were calculated as: IHCL= score of percentage of positive cells + score of IR of positive cells. Percentage and IR of positive cells were scored blinded on a 4 point scale. Different letters are significantly different, p <0.05. Mean ± SD. Bar=25.1 µM.

Figure 13. Expression of RARB in the sweat gland and dermis among control and CCCA patients. IHC was performed with antibodies against RARB in the scalp biopsies from control (a, e and i) and CCCA patients (mild: b, f and j; moderate: c, g and k; severe: d, h and i). A blinded IHCL of RARB was calculated as: IHCL= score of percentage of positive cells + score of IR of positive cells. Percentage and IR of positive cells were scored blinded on a 4 point scale. Different letters are significantly different, p <0.05. Mean ± SD. Bar=50.3 µM (e-h); Bar=10.1 µM (a-d and i-l).

Figure 14. Expression of DHRS9, ALDH1A1, RARB and CYP26A1 in the sebaceous gland among control and CCCA patients. IHC was performed with antibodies against DHRS9 (a-d and r), ALDH1A1(e-h and s), RARB(i-l and t), CYP26A1(m-p and u) in the scalp biopsies from control(a, e, i and m) and CCCA patients (mild: b, f, j and n; moderate: c, g, k, o; severe: d, h, l and p). A blinded IHCL of all four proteins were calculated as: IHCL= score of percentage of positive cells + score of IR of positive cells. Percentage and IR of positive cells were scored blinded on a 4 point scale. Different letters are significantly different, p <0.05. Mean ± SD. Bar=25.1 µM.

Suprabasal layer Sweat gland Dermis HF(ORS) DHRS9 control 5.8±0.3 6.0 (n=1) 5.0 (n=1) 6.0 (n=1) mild 5.5±0.3 5.7±0.3 3.9±0.6 5.4±0.3 moderate 6.0±0.3 5.2±0.3 4.3±0.6 5.8±0.3 severe 6.6±0.4 6.3±0.6 3.8±0.9 5.0±0.4 ALDH1A1 control 6.1±0.3 6.0 (n=1) 6.0 (n=1) 5.0 (n=1) mild 6.8±0.4 5.4±0.3 4.9±0.6 5.3±0.4 moderate 6.8±0.3 5.0±0.3 4.6±0.6 4.7±0.3 severe 6.8±0.5 5.0±0.4 4.4±0.8 5.3±0.5 RARB control 7.8±0.2 7.0 (n=1) 7.0 (n=1) 7.0 (n=1) mild 6.2±0.3 5.5±0.2 5.2±0.4 6.7±0.3 moderate 6.1±0.3 5.3±0.2 4.0±0.4 6.3±0.3 severe 5.8±0.4* 4.0±0.3# 2.8±0.6# 5.3±0.5& CYP26A1 control 7.6±0.3 7.0 (n=1) 6.0 (n=1) 6.0 (n=1) mild 5.8±0.3 5.8±0.3 3.4±0.3 5.9±0.4 moderate 6.6±0.3 5.7±0.3 4.0±0.3 6.1±0.4 severe 5.4±0.4* 4.6±0.5 2.6±0.5 5.5±0.5

Table 4. IHCL of DHRS9, ALDH1A1, RARB and CYP26A1 in the suprabasal layer, sweat gland, dermis and outer root sheath (ORS) of hair follicle among control and CCCA patients. (Mean ± SD. * p <0.05, compared to control; # p <0.05, compared to mild disease; & p =0.09, compared to mild disease).

5.4 Discussion

An optimal level of dietary vitamin A is essential for the health of skin and hair.

Both vitamin A deficiency and toxicity can cause hair loss (8,110-112). The B6 mouse is

a good model to study CCCA for they spontaneously develop alopecia with similar features

of CCCA (159). In our examination of human samples, four retinoid metabolism proteins

all decreased in the basal layer of the epidermis; and RARB also decreased in the sweat

gland and dermis in severe disease, which is similar to the findings in B6 mice. However,

we did not find these four proteins differ in the diseased groups compared with controls in

other locations of the skin. These findings suggested that vitamin A metabolism and

signaling are also altered among CCCA patients.

At the late stage of PCA, skin SCs are damaged following inflammation, which

leads to irreversible hair loss (216-219). The basal layer of the epidermis is one of the important locations for skin SCs (220). Decreased vitamin A metabolism and signaling in basal SCs in severe disease may suggest imbalance of vitamin A regulation in basal SCs is involved in the progression of CCCA. Vitamin A controls the proliferation and differentiation of many cell types (221). Topical retinoid application induced epidermal

hyperplasia due to basal SC hyperproliferation (222-224). It is possible that the decreased

vitamin A metabolism and signaling in basal SCs reduced the proliferative ability of skin

SCs. The only SC marker tested in CCCA patients to date is keratin 15 (Krt15). Krt15 was

thought to be a bulge SC marker but is also seen in the basal layer and different regions of

the outer root sheath (219,225). Krt15 decreased in the inflamed basal layers in CCCA

patients; and the desquamation of inner root sheath in CCCA patients changed the

morphology of bulge epithelium, causing the loss of Krt15 expression (226). We also found

that the expression of RARB in the hair follicle outer root sheath, the sweat gland and

dermis dropped as the disease severity increased, where expression of vitamin A

metabolism proteins were not significantly altered. RARB expression was also decreased

with the increased severity of CCCA in the hair follicle and epidermis of B6 mice before

DHRS9 dropped in mice fed purified diets (6). RARB is a target gene for RA (227);

therefore lower RARB expression in hair follicle, sweat gland and dermis suggests

decreased RA signaling. There are SCs located in the outer root sheath of the hair follicle

and sweat gland (228,229); and reduced RA signaling may be caused by the damage of

SCs in severe disease. Also RA signaling regulates activation of skin SCs through the WNT

(wingless-related MMTV integration site) signaling pathway (170), and lower RA

signaling may reduce WNT signaling and inhibit the proliferation of skin SCs in CCCA

patients.

Sebaceous gland loss and inflammation were thought to play a potential role in the

pathogenesis of PCA (230). This is highlighted in two mouse models of CCCA: mice with

a spontaneous mutation in stearoyl-CoA desaturase-1(Scd1ab2J)(164) and SC conditional

peroxisome proliferator-activated receptors G (PPARG) null mice (165). In cultured

human sebocytes, both vitamin A deficiency and higher levels of RA led to decreased cell

proliferation and lipogenesis (91,92), which indicated that precise levels of RA are required for seboctyes in the sebaceous gland. Our previous studies showed that the sebaceous gland

had increased expression of DHRS9, ALDH1A1, CYP26A1 and RARB in mild disease in

mice and humans with CCCA; yet had decreased expression in severe disease in mice and

pseudopelade (6). These B6 mice had spontaneously developed CCCA in Jackson

Laboratory, and the severity of alopecia in the severe group of these B6 mice was worse than that of B6 mice fed with the purified diet in our subsequent study. Also we observed that the expression of DHRS9 and RARB were not changed in the sebaceous gland by the severity of disease in B6 mice fed with purified diet ((6) figure s4). Here we report that there was no difference of the expression of the above four proteins in the sebaceous glands between the control (n=1) and CCCA groups, similar to what we found in the B6 mice fed a purified diet. It is possible that we did not collect the most severe disease samples because those patients usually do not have biopsies. This suggests that vitamin A metabolism and signaling were maintained in the sebaceous gland through the progression of CCCA, and we still need to further examine the alterations of those proteins in the very last stage of

CCCA patients. Pseudopelade is defined as a permanent progressive scarring alopecia

(231), however it cannot be considered as a distinct disease (232). Our finding suggests that vitamin A metabolism in the sebaceous gland is different between pseudopelade and the severe stage of CCCA, although pseudopelade may represent a disease state most like our most severe B6 mice. In conclusion, we demonstrated that vitamin A metabolism and signaling was not altered in the sebaceous gland in CCCA patients, which was similar to what we found in B6 mice.

In summary, we reported here that the expression of important vitamin A metabolism and signaling components in the skin decreased as severity increased. In the future, we need to further investigate the mechanisms involved in the alterations of vitamin

A metabolism and signaling in CCCA. These findings expand the knowledge of

pathogenesis of CCCA and emphasize the importance of vitamin A metabolism and signaling in the health of skin and hair.

Chapter 6: Development of specific retinal dehydrogenase

inhibitor

6.1 Abstract and Introduction

Abstract

Development of a specific retinal dehydrogenase (ALDH1As) inhibitor is important for better understanding of the mechanism of RA signaling and cancer research.

Today, there are no specific ALDH1As inhibitors available. In collaboration with Dr.

Robert Curley, here we tested the efficacy and specificity of candidate ALDH1As inhibitors. A new compound, termed as β-ionylidene ethanethiol-diethyldithiocarbamic acid (BIES-DDC), had better inhibitory effects compared with disulfiram (in the skin homogenate) and DEAB (in MDA-MB-468 cell line); and it did not inhibit the oxidation reactions of other aldehydes (propionaldehyde (C3), benzaldehyde (C7), cinnamaldehyde

(C9)) in vitro. Topical BIES-DDC treatment also significantly increased retinyl esters in murine skin, although it did not significantly reduce the synthesis of RA. This study suggested that BIES-DDC had better inhibitory efficacy and specificity for ALDH1As than other identified inhibitors. The discovery of a new ALDH1As inhibitor would be an important reagent for the future vitamin A research.

Introduction

The aldehyde dehydrogenases (ALDHs) are a group of 19 enzymes that oxidize many aldehydes into related carboxylic acids (174). Retinal dehydrogenases (ALDH1A1,

ALDH1A2, ALDH1A3) specifically convert retinal into RA (52). The RA signaling pathway is essential for embryogenesis and other biological activities, regulating around

500 genes (59). In addition, ALDH1As have been identified to play a role in cancer resistance to chemotherapy; and inhibition of ALDH1A1 could be a target to solve the cancer resistance problem (175-177). In order to better understand the mechanism by which

RA regulates these biological activities, we collaborated with Dr. Robert Curley, a professor from the College of Pharmacy, to develop and test the specific ALDH1s inhibitor.

Many chemicals have been identified as effective ALDH inhibitors: disulfiram

(tetraethylthioperoxydicarbonic diamide), DEAB (4-diethylamino benzaldehyde) and citral (3,7-dimethyl-2,6-octadienal). Disulfiram is a widely used drug to treat alcohol addiction due to inhibiting ALDH2 in the liver (178); also disulfiram can inhibit

ALDH1A1 in vitro and in vivo (179). DEAB is another confirmed ALDH1As inhibitor

(180); but an in vivo study suggests that it would also inhibit ALDH2 (181). Citral is thought to be a specific ALDH1As inhibitor and has been used to study the role of RA in the embryogenesis and tissue development (182,183). However, citral failed to reduce the

RA synthesis in our previous mouse study. In conclusion, none of these chemicals are a specific ALDH1As inhibitor.

We collaborated with Dr. Robert Curley in the development of ALDH1As inhibitors. He developed a chemical named β-ionylidene ethanethiol-

diethyldithiocarbamic acid (BIES-DDC) from retinal and disulfiram. We tested the specificity and efficacy of this compound in vitro and in vivo. Our hypothesis is that BIES-

DDC is as effective as disulfiram and DEAB yet can specifically block retinal as the only substrate.

6.2 Material and Methods

6.21 Liver and skin homogenate preparation

Tissues (liver or skin) were cut and rinsed with an ice-cold solution of 0.9% NaCl.

Tissues were homogenized in reaction buffer (0.24 g/2 ml) containing: 20 mM HEPES (4-

(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 150 mM KCl, 2mM EDTA

(Ethylenediaminetetraacetic acid) and 2mM beta-mercaptoethanol (233). Cytosolic homogenates (containing cytosolic ALDHs) were prepared by differential centrifugations.

Homogenates were centrifuged at 13,000 rpm at 4 °C for 15 minutes and the supernatants were kept and centrifuged at 49,000 rpm 4 °C for 60 minutes. The protein concentration of supernatants of the second centrifugation was measured with Pierce BCA Protein Assay

Kit-Reducing Agent Compatible (Pierce Biotechnology, Rockford, IL) (234).

6.22 In vitro Retinal dehydrogenase inhibitor efficacy assay

For the assay of retinoic acid synthesis activity and efficacy of ALDH1As inhibitors

in vitro, liver cytosolic homogenates (200μg protein) were incubated with 1 μM, 10 μM

and 100 μM of each inhibitor (BIES-DDC and disulfiram in DMSO) for 30 minutes. Skin

cytosolic homogenates (144 μg protein) were incubated with 25 μM, 50 μM and 100 μM

of each inhibitor (BIES-DDC, disulfiram and DEAB in DMSO) for 30 minutes. Then 200

μM NAD+ and 100 μM retinal in DMSO were added and incubated at 37°C for 40 minutes.

The reaction was stopped by adding 3ml ice cold ethanol. Retinoic acid was extracted after

the reaction and measured with high-performance liquid chromatography (HPLC). The

total incubation volume was 200 μl and reaction buffer was described before (233).

6.23 Cell culture

MDA-MB-468 cells were purchased from American Type Culture Collection

(Rockville, MD) and were cultured in Dulbecco's Modified Eagle's medium (DMEM) high

glucose supplemented with 10% FBS (fetal bovine serum) and 1% penicillin /streptomycin

(Life Technologies, Grand Island, NY) (235). Cells were grown in 60-mm dishes until just confluent. Normal medium (described above) was removed and replaced with 3 ml of UV- treated medium containing 10 μM BSA. Inhibitor groups were treated with 2 μM different inhibitors (BIES-DDC, disulfiram and DEAB) dissolved in DMSO and incubated for 1hr.

Then 2 μM retinol (provided in DMSO) were added to each dish. Cells were incubated with retinol for 12 or 24 hours, and then the cells were scraped into the medium. Cells and medium were extracted together (236).

6.24 Retinoic acid (RA) extraction and HPLC analysis

In vitro reactions were mixed with 3 ml of 4.25 M NaCl, 0.25 M KOH, and 3 ml of

100% ethanol. Cell culture medium and cells were mixed with 6 ml of 4.25 M NaCl, 0.25

M KOH, and 6 ml of 100% ethanol. Neutral retinoids were extracted twice with a volume of hexane equal to the aqueous phase and discarded. The aqueous phase was then acidified

with 180 μl of 4N HCl, and acidic retinoids were extracted once with an equal volume of hexane. Extracted fractions were dried under nitrogen and dissolved in 200 ul of mobile phase (hexane:dioxane:acetic acid, 92:8:0.1) for Agilent 1200 Series HPLC separation

(Agilent Technologies, Santa Clara, CA). Retinoids were separated on an Agilent

ZORBAX 5 silica column (4.6 mm × 250 mm) running at 2 ml/min. RA was monitored at

352 nm. Quantitation was performed by relating the area of the peak to areas obtained by

the analysis of known quantities of RA standards (236). Limit of detection (LOD) and

Limit of quantification (LOQ) are shown in Table 5.

LOD LOQ RA

Area 1.838 4.857 149.264

Concentration (µM) 0.05158 0.03738 4.292 Table 5. Limit of detection (LOD) and limit of quantification (LOQ) of RA analysis in HPLC

6.25 In vitro Retinal dehydrogenase inhibitor specificity assay

For the assay of the specificity of RALDH inhibitors in vitro, liver homogenates

(200 μg protein) were incubated with 1 μM, 10 μM and 100 μM of each inhibitor (BIES-

DDC, disulfiram and DEAB in DMSO) for 30 minutes, and 1 mM NAD+ and 2 mM

aldehydes (benzaldehyde, cinnamaldehyde and farnesal in DMSO) were added and

incubated at 37°C for 40 minutes in a procedure modified from Klyosov (237). Skin

homogenates (144 μg protein) were incubated with 25 μM, 50 μM and 100 μM of each

inhibitor (BIES-DDC, disulfiram, DEAB dissolved in DMSO) for 30 minutes before 1 mM

NAD+ and 2 mM aldehyde (propionaldehyde, benzaldehyde, cinnamaldehyde and farnesal

in DMSO) were added and incubated at 37°C for 40 minutes. The total incubation volume

was 1 ml and reaction buffer was described before. The reaction systems (2 mM aldehydes,

200 μg proteins from liver homogenate or 144 μg proteins from skin homogenate, 1mM

NAD+) without inhibitors and incubation were used as blank. The absorbance at 340 nm

of NADPH was measured by the Agilent 8453 UV-Visible spectrophotometer (Agilent

Technologies, Santa Clara, CA).

6.26 Animals

The Ohio State University (Columbus, OH) Institutional Animal Care and Use

Committees approved all procedures.

First Animal study: Ten C57BL/6J female mice (The Jackson Laboratory, Bar

Harbor, ME) were used. Topical treatments started 5 days after wax depilation. Five mice were topically treated with 2.4 mmol/kg body weight (BW) BIES-DDC in acetone (400 μl)

(the average body weight of mice is 16.85 g, and the BIES-DDC concentration in acetone is 100 mM); and the other five were topically treated with the same volume of acetone. All the mice were sacrificed after 24 hours. Dorsal skin was collected for the measurement of retinol, retinyl ester and retinoic acid.

Second Animal study: Forty-four C3H/HeJ female mice (The Jackson Laboratory,

Bar Harbor, ME) were used. Topical treatments started 5 days after wax depilation. Nine mice were topically treated with 3.2 mmol/kg BW BIES-DDC in acetone (400 μl) (the

average body weight of mice was 18.0 g, and the BIES-DDC concentration in acetone was

144 mM); eight mice with 4.8 mmol/kg BW BIES-DDC in acetone (400 μl) (the average

body weight of mice was 18.0 g, and the BIES-DDC concentration in acetone was 216

mM); nine mice with 3.2 mmol/kg BW disulfiram in acetone (400 μl) (the average body

weight of mice was 18.0 g, and disulfiram concentration in acetone is 144 mM); eight mice with 4.8 mmol/kg BW disulfiram in acetone (400 μl) (the average body weight of mice was 18.0 g, and disulfiram concentration in acetone was 216 mM); and ten mice with the

same volume of acetone. All the mice were sacrificed after 24 hours. Dorsal skin was

collected for the measurement of retinol, retinyl ester and retinoic acid.

All the skin samples were sent to Dr. Napoli’s lab (University of California,

Berkeley, CA) for analysis.

6.27 Statistical Analysis

Comparison among experimental groups was analyzed by univariant analyses,

followed by Tukey’s post hoc tests when appropriate using Statistical Package for the

Social Sciences, v21 (IBM; Armonk, NY). P value <0.05 is defined as statistical

significance.

6.3 Results

6.31 Efficacy of BIES-DDC in vitro

Efficacy of BIES-DDC in liver homogenate

We tested the inhibitory efficacy of BIES-DDC in liver homogenates compared with disulfiram. Both BIES-DDC and disulfiram significantly decreased RA synthesis at all the concentrations (1 μM, 10 μM and 100 μM) (Figure 15a). Inhibitory percentage (I

%) of both inhibitors increased with their concentrations (Figure 15b). BIES-DDC and

disulfiram had similar I % at both 1 μM and 10 μM; and disulfiram had a higher I % than

BIES-DDC at 100 μM.

Concentration of RA in control - Concentration of RA with inhibitor I % = Concentration of RA in control

Efficacy of BIES-DDC in skin homogenate

We tested the inhibitory efficacy of BIES-DDC in skin homogenate compared with disulfiram and DEAB. Both BIES-DDC and DEAB significantly decreased RA synthesis at 50 μM and 100 μM, and disulfiram only significantly decreased RA synthesis at 100 μM

(Figure 15c). BIES-DDC and DEAB had similar I % at 50 μM; and all the three inhibitors had similar I % at 100 μM (Figure 15d).

Figure 15. Inhibitory efficacy of different ALDH1As inhibitors in the liver and skin homogenates. Inhibition of RA synthesis by BIES-DDC and disulfiram at 1 μM, 10 μM and 100 μM were measured in the liver homogenate (a); and at 25 μM, 50 μM and 100 μM were measured in the skin homogenate (c). Inhibitory percentage (I%) was individually calculated (liver: b, skin: d) as I%=(Concentration of RA in control - Concentration of RA with inhibitor)/ Concentration of RA in control. Different letters are significantly different, P <0.05. Mean ± SD.

Efficacy of BIES-DDC in cell culture

We tested the inhibitory efficacy of BIES-DDC in MDA-MB-468 cells (a breast cancer cell line) that can synthesize RA (235). We performed the first three repeats on different days and found that there was a large variation in RA synthesis in the control groups. The last three repeats were performed on the same day, and there was much smaller variation of the RA synthesis in control groups among the last three repeats. Within the same repeat, BIES-DDC significantly decreased the synthesis of RA at 12 hr and 24 hr incubations (Figure 16). We compared I % of BIES-DCC and DEAB within the same repeat, and found that I % of BIES-DCC was more effective than DEAB in the first three repeats. We also found that I % of BIES-DCC at 24 hr was more effective than that at 12 hr (Figure 17). Then we compared I % of BIES-DCC and DEAB within all the repeats, and we found that I % of BIES-DCC and DEAB at 24 hr were similar.

Figure 16. Inhibition of RA synthesis by different ALDH1As inhibitors in MDA-MB-468 cells. BIES-DDC and DEAB inhibitory efficacy were tested at 2 μM in MDA-MB-468 cells for 12 and 24 hours. First three repeats were done in different dates; and the last three repeat were done within the same day. (* P <0.05; Mean ± SD).

Figure 17. Inhibitory percentage (I%) of different ALDH1As inhibitors in MDA-MB-468 cells. BIES-DDC and DEAB inhibitory percentage (I%) were calculated at 2 μM in MDA-MB-468 cells for 12 and 24 hours. First three repeats were done in different dates; and the last three repeat were done within the same day. (* P <0.05; Mean ± SD).

6.32 Specificity of BIES-DDC in vitro

We tested the specificity of BIES-DDC and citral in liver homogenate. We used

benzaldehyde (C7) and cinnamaldehyde (C9) as the reaction substrates. Both BIES-DDC

and citral at all concentrations (1 μM, 10 μM and 100 μM) did not inhibit the reactions of

benzaldehyde (Figure 18a) and cinnamaldehyde (Figure 18b).

We tested the specificity of BIES-DDC, disulfiram and DEAB in skin homogenate.

We used propionaldehyde (C3), benzaldehyde (C7), cinnamaldehyde (C9) as the reaction substrates. Both disulfiram and BIES-DDC at all concentrations (25 μM, 50 μM and 100

μM) did not inhibit the reactions of propionaldehyde (Figure 18c), benzaldehyde (Figure

18d), cinnamaldehyde (Figure 18e), however DEAB inhibited all the reactions.

Figure 18. Specificity test of different ALDH1As in other aldehydes in the liver and skin homogenates. BIES-DDC and citral at 1 μM, 10 μM and 100 μM were tested with benzaldehyde (a) and cinnamaldehyde (b) in the liver homogenate; and disulfiram, BIES-DDC and DEAB at 25μM, 50μM and 100μM were tested with propionaldehyde (c), benzaldehyde (d), cinnamaldehyde (e) in the skin homogenate. (* P <0.05 compared with no inhibitor; Mean ± SD).

6.33 Efficacy of BIES-DDC in vivo

Efficacy of BIES-DDC in animal studies

First mouse study: BIES-DDC (2.4 mmol/kg) treated mice (N = 5) had a trend toward lower RA concentrations (p = 0.1) (Figure 19a) and higher retinyl ester (p <0.01) in the skin compared to control group (N = 5) (Figure 19c). There was no difference of retinol in the two groups (Figure 19b).

Second mouse study: There was no difference of RA among all groups (Figure

19d). BIES-DDC (4.8 mmol/kg) treated mice (N = 8) significantly increased retinol and retinyl esters in the skin compared with control skin. There was larger variability in the retinol measurement, and one of the eight samples was an outlier (2.73 nmol/g tissue).

There was no difference in retinol among all the groups after excluding the outlier.

Disulfiram treated mice (N = 9 in 3.2 mmol/kg group and N = 10 in 4.8 mmol/kg group) had similar retinyl ester and retinol levels as control group (N = 10) (Figure 19e and f).

Figure 19. Inhibitory efficacy tests of BIES-DDC and disulfiram in two mouse studies. RA (a), retinol (b) and retinyl ester (c) were measured in the control and BIES-DDC (compound 8) 2.4 mmol/kg treated groups in B6 mice; and RA (d), retinol (e) and retinyl ester (f) were measured in the control, BIES- DDC (8(1): BIES-DDC 3.2 mmol/kg; 8(2): BIES-DDC 4.8 mmol/kg), and disulfiram (D(1): disulfiram 3.2 mmol/kg; D(2): disulfiram 4.8 mmol/kg) treated groups in C3H/HeJ mice. Different letters are significantly different, P <0.05. Mean ± SD.

6.4 Discussion

Development of specific retinal dehydrogenase (ALDH1As) inhibitors is important

for a better understanding of the mechanism of RA signaling. In addition, ALDH1As have

been identified to play a role in the cancer resistance to chemotherapy; and inhibition of

ALDH1A1 could be a target to solve the cancer resistance problem (175-177). Many of

the inhibitors available can inhibit the activity of ALDH1As, but they also inhibit other

aldehyde dehydrogenases (ALDHs) (174). Here, we collaborated with Dr. Robert Curley;

and we tested the efficacy and specificity of the possible inhibitor candidates designed by

him. A new compound, termed as BIES-DDC, showed a promising specific inhibitory

effect in vitro, however, it did not significantly reduce the synthesis of RA in our mouse

studies.

BIES-DDC showed similar inhibitory effects compared with disulfiram in liver

homogenate. Disulfiram (tetraethylthioperoxydicarbonic diamide) is a confirmed ALDHs

inhibitor: it has been widely used to treat alcohol addiction because it inhibits ALDH2

(178); and it also inhibits the activity of ALDH1As (184). Our data in the liver homogenate

suggested that BIES-DDC would be a potential specific inhibitor for ALDH1As. ALDH1 and ALDH2 are both localized in the liver and skin of mice, but the concentrations of the enzymes in the skin are much lower than in the liver (238). BIES-DDC had a similar

inhibitory effect compared with DEAB in skin homogenate: both significantly reduced the

synthesis of RA in skin homogenate at the concentration of 50 μM. DEAB is another

confirmed ALDH1 inhibitor (180) but an in vivo study suggests that it would inhibit

ALDH1 and ALDH2 (181). Our data indicates that BIES-DDC was a more effective inhibitor in vitro in the skin compared with disulfiram, which only had inhibitory effects at 100 μM in skin homogenate. Also, the inhibitory efficacy of BIES-DDC in liver and skin homogenate is dose dependent.

In cell experiments, the inhibitory efficacy of both BIES-DDC and DEAB after 24 hours were better than 12 hours, suggesting that the effect of BIES-DDC is time dependent.

Although we did not find significant difference of I% between BIES-DDC and DEAB in cells, we observed the decline of I% of BIES-DDC efficacy as time passed. In total, we had six repeats for inhibitor efficacy test in cells, and the efficacy of I% of BIES-DDC in the first test was significantly higher than that in the 4th-6th repeat. This might suggest that

BIES-DDC degraded over time. We did not observe this possible degradation of BIES-

DDC in liver and skin homogenate tests. In liver and skin homogenate experiments, for each repeat the BIES-DDC was diluted from a 10 mM stock solution. In the cell culture experiments, 100 μL of the 200 μM stock solution was then aliquoted into separate vials.

A vial of 200 μM BIES-DDC was then taken from the freezer for each repeat of the cell culture experiment. It is possible that BIES-DDC is unstable for storage at lower concentration (200 μM) and is relatively stable in the higher concentrations (10 mM).

BIES-DDC did not inhibit the oxidation of other aldehydes in both liver and skin homogenates, Thus it had better specificity compared to DEAB. However, we did not directly test the inhibitory efficacy of BIES-DDC on individual ALDH. Disulfiram, an inhibitor for both ALDH1As and ALDH2 (178,184), also did not affect the oxidation of the other aldehydes tested. ALDH2 converts acetaldehyde to acetic acid. However, acetaldehyde was not used as a substrate to test the specificity of BIES-DDC because the boiling point of acetaldehyde is 20.2 °C. A better experiment would be to use pure ALDH2 to test the specificity of BIES-DDC instead of liver or skin homogenates.

Although BIES-DDC specifically inhibited ALDH1As activity in vitro, we did not observe a significant reduction of RA synthesis in the skin of mice. There are three possible explanations for this result: dose, technical difficulties and compensation by other vitamin

A metabolism proteins. Vitamin A metabolism is well regulated in the skin (6). There are three ALDH1As: ALDH1A1, ALDH1A2, and ALDH1A3 that can oxidize retinal to RA.

It is possible that BIES-DDC can only inhibit one or two of ALDH1As, and the compensatory increase of the uninhibited enzymes could balance the synthesis of RA in

the skin. It would be better if we could test the inhibitory efficacy of BIES-DDC on each

ALDH1As individually in the future. Also, the RA degradation enzymes cytochrome P450

family members (CYP26A1, CYP26B1 and CYP26C1) in the skin would decrease to

balance the RA concentration. In our first in vivo study (N = 10), RA was decreased in the

BIES-DDC treated group (100 mM) (p = 0.1) and retinyl ester was significantly increased

in BIES-DDC treated group. These data suggested that BIES-DDC might have a potential

inhibitory effect in vivo, and the inhibition of ALDH1As caused an increase in free retinol,

which can be converted to retinyl ester. But the dose of BIES-DDC was too low for RA to be reduced. Based on these findings, we designed a second mouse study where the sample size (N = 44) and dosage of BIES-DDC was increased (144 mM and 216 mM). We also included disulfiram (144 mM and 216 mM) as a positive control. However, higher concentrations of BIES-DDC in the skin samples caused a blockage in the LC column for

RA measuring (reported by Dr. Napoli’s lab), resulting in the loss of RA data from one of the two repeats. We did not observe a significant difference in RA among all five groups.

There was also no difference between the positive control (disulfiram group) and negative control group, which indicated that the measurement of RA in our second mouse study might not be accurate. In addition, retinyl ester was dose dependently increased only in the

BIES-DDC treated groups, which was similar to the first mouse study. We also changed

the strain of mice between the two studies, thus part of the differences could be strain

dependent. Study 1 used C57BL/6J mice, while the second study used C3H/HeJ mice. This

was done because future studies with this inhibitor would be done in C3H/HeJ mice to

avoid the spontaneous CCCA that occurs in C57BL/6J mice (see chapters 4 and 5), which

would complicate the analysis of hair defects caused by inhibiting RA synthesis.

In conclusion, inhibitory efficacy of BIES-DDC is time and dose dependent. BIES-

DDC had better inhibitory effect compared to disulfiram in vitro and in vivo; and also

BIES-DDC had better specificity than DEAB. Our finding supports that BIES-DDC is a potential ALDH1As specific inhibitor, although we did not observe the reduction of RA in the skin of mice.

Chapter 7: Alteration of skin stem cell markers by dietary

vitamin A

7.1 Abstract and Introduction

Abstract

Skin is one of the important reservoirs for adult SCs. There are seven distinct

locations for skin SCs, characterized by different markers. Skin SCs play different roles in

the pathogenesis of scarring and non-scarring hair loss diseases. Dietary vitamin A altered

the severity of AA (non-scarring alopecia) and CCCA (scarring alopeica). Here, we used

C3H/HeJ mice from the AA study and C57BL/6J (B6) mice from the CCCA study 1 to

examine the alterations of skin SCs in these two strains of mice by IHC. We found that

CD34 was only expressed in the bulge of telogen hair follicles in B6 mice. Leucine-rich

repeat-containing G-protein coupled receptor-5 (LGR5) localized to both the companion layer in the lower supra-bulbar region and inner root sheath in the center of the supra-bulbar region in the anagen hair follicles of non-diseased mice. However, it only localized to the

inner root sheath in the anagen hair follicles of CCCA mice. We could not find CD34 or

LGR5 IR in the anagen or telogen hair follicles of control or diseased C3H/HeJ mice; but

we observed LGR5 expression in the anagen hair follicle in the C3H/HeJ wax stripped

mice. These data suggest there is a strain difference in the expression of these SCs markers.

In addition, LGR5 may be a good SC marker to study the skin SCs in both strains of mice.

Introduction

Primary cicatricial alopecias (CA) are a collection of scarring hair loss diseases.

Alopecia areata (AA) is one of the most common autoimmune, non-scarring, hair loss

disease, affecting 4.5 million people in the United States (National Alopecia Areata

Foudation). Although the pathogenesis of CA and AA are unknown, dysregulation of skin

stem cells (SCs) is thought to play an important role in both diseases (63,145,239). Vitamin

A, an essential regulator for SCs, plays an important role in the development of these

alopecias (6,7).

There are 7 identified locations of skin SCs with different SC markers (61). The

bulge of the hair follicle is one of the most important and well recognized niches for skin

SCs (61), and is marked by cluster of differentiation 34 (CD34), cluster of differentiation

200 (CD200) and Keratin 15 (Krt 15). Skin SCs in the bulge produce all of the cells in the

hair follicle during the normal hair cycle and repopulate the interfollicular epidermis (IFE)

during wound healing (69,240,241). Another important location for skin SCs is in the hair

germ, marked by Leucine-rich repeat-containing G-protein coupled receptor-5 (LGR5).

LGR5+ cells are now thought to be actively cycling, long-lived, and generate all the cell lineages of hair follicles below sebaceous glands (66). These two groups of skin SCs have different fates in CA (scarring alopecia) and AA (non-scarring alopecia). Hair germ SCs

are attacked in both CA and AA; but hair follicle bulge SCs are only damaged in CA, and

spared in AA (63,226,242).

In our mouse models for CA (C56BL/6J mice) and AA (C3H/HeJ mice), we

observed that dietary vitamin A regulated the progression of both diseases (6,7). Higher

levels (12-28 IU/g diet) of vitamin A caused more anagen hair follicle, which are more

vulnerable to the inflammatory attack in both diseases; and excess level (56 IU/g diet) of

vitamin A for B6 mice inhibited the anagen initiation and protected B6 mice from CA. Our

purpose here is to study skin SCs markers in these two strains of mice by IHC to examine

the alterations of skin SCs by dietary vitamin A in two different alopecia mouse models.

This is important for the understanding of how dietary vitamin A regulates skin SCs.

7.2 Material and Methods

We used C3H/HeJ mice from the AA study and C57BL/6J (B6) mice from the

CCCA study 1 as we described in chapter 3 and 4. We also used C3H/HeJ wax stripped

mice (post wax strip 0-24 days). We tested the expression of SC markers: CD34, Krt15,

LGR5, CD200 and Blimp1 in both strains of mice by IHC.

IHC was performed using a three-antibody system in paraffin slides as described

previously (127). The following antibodies were used: CD34 (BD Pharmigen, San Jose,

CA), Krt15 (Biolegend, Dedham, MA), LGR5 (Novus Biologicals, Littleton, CO), CD200

(BD Pharmigen, San Jose, CA) and B lymphocyte-induced maturation protein-1 (Blimp1)

(Novus Biologicals, Littleton, CO). Biotinylated anti-rabbit and anti-rat were used as secondary antibodies (Jackson Immunoresearch Laboratories, Inc., West Grove, PA) followed by a horseradish peroxidase conjugated anti-biotin tertiary antibody (Bethyl

Laboratories, Montgomery TX).

7.3 Results

CD34 was strongly expressed in the bulge of telogen hair follicles in B6 mice

(Figure 20a). We could not find CD34 IR in anagen hair follicles of B6 mice (Figure 20b).

Krt15 was thought to be a bulge SCs marker but is also seen in the basal layer and other

regions of hair follicle (226). In our studies, Krt 15 was strongly expressed in the bulge

area of anagen hair follicles, but it was not specific for the bulge (Figure 20c and d). LGR5

localized to both the companion layer in the lower supra-bulbar region and inner root

sheath in the center of the supra-bulbar region in the anagen hair follicles of non-diseased mice (Figure 20e), and it only localized in the inner root sheath in the anagen hair follicles of CCCA mice (Figure 20f). In both groups, there was no difference of LGR5 expression within the groups fed with different levels of vitamin A.

Figure 20. Expressions of different stem cell markers in B6 mice. IHC was performed against CD34 (a-b), Krt15 (c-d), and LGR5 (e-f) in dorsal skin from B6 mice. Bar = 100uM in a, b and c; 50 uM in d, e and f. Black arrow: bulge; orange arrow: bulb; purple arrow: inner root sheath; green arrow: companion layer. Telogen: a and c; anagen: b,d,e and f. In the AA study, we could not find CD34 or LGR5 IR in the anagen or telogen hair

follicles of control or diseased C3H/HeJ mice; but we observed LGR5 expression in the

anagen hair follicle in the C3H/HeJ wax stripped mice (Data not shown). We also did IHC against CD 200, Blimp1 in both strain of mice, but we did not find any IR for either antibody. 90

7.4 Discussion

Skin is a very good model to study the behavior of SCs. There are 7 identified

locations of skin SCs with different SC markers (61). We wanted to choose a specific bulge

SC marker to investigate the alterations of SCs by dietary vitamin A. We therefore

performed IHC against CD34, Krt15, CD200, but none of them was specifically altered by

vitamin A. We also performed IHC against Blimp1 to label SCs in sebaceous gland, and

we did not find IR of Blimp1 in our samples.

CD34 was identified as a specific hair follicle bulge SC marker in mice and dogs

(205,243). CD34 expression co-localized with both Krt15 and label retaining cells (LRCs)

in the bulge; CD34+ bulge cell can be captured by flow cytometer and cultured in vitro.

LGR5+ SCs in the hair germ has the potential to generate the whole hair follicle (66). In

our studies, CD34 was only expressed in the telogen hair follicle of B6 mice and LGR5 was only expressed in the anagen hair follicle of B6 mice. Neither CD34 nor LGR5 was expressed in C3H/HeJ mice. These data suggest there is a strain difference in the expression of these SCs markers. In addition, we found that wax stripping induced LGR5 expression in the anagen hair follicle of C3H/HeJ mice. Wax stripping is a method to induce and synchronize the hair cycle in animal studies, and it causes minor trauma to the skin (159).

Induced LGR5 expression could be due to the wound healing processes following wax stripping, which suggests that LGR5 is important for skin and hair follicle homeostasis after trauma in C3H/HeJ mice.

Although we did not find a good SC marker useful in IHC to study the behavior of

skin SCs, the alteration of LGR5 labeling cells in B6 mice provided some clues for the

damage of skin SCs by the cicatricial alopecia. LGR5 localized to both the companion

layer in the lower supra-bulbar region and inner root sheath in the center of the supra-bulbar region in anagen hair follicles of non-diseased mice; and it only localized in the inner root sheath in anagen hair follicles of CCCA mice. This finding suggests that there is an alteration in location of LGR5+ SCs in CCCA mice compared with non-diseased mice.

In conclusion, CD34 was a specific bulge SC marker for telogen hair follicles in

B6 mice. LGR5+ SCs were altered in CCCA mice compared to non-diseased B6 mice; and

LGR5+ SCs were induced by wax stripping in C3H/HeJ mice. LGR5 may be a good SCs

marker to study the skin SCs in both strains of mice.

Chapter 8: Conclusions and Significance of the study

This study demonstrates that normal vitamin A metabolism and signaling are

important for the health of skin and hair in mice and humans. Our study provides evidence

that dietary vitamin A regulates skin SCs through the BMP/WNT signaling pathway. High levels of dietary vitamin A increased WNT signaling in the bulge SCs to activate hair cycle initiation, but excess dietary vitamin A inhibited anagen initiation, at least partially, potentially through increasing dermal BMP4 expression, which inhibits WNT signaling and prevents anagen induction. We also found differences between two strains of mice.

Understanding this mechanism involved in hair cycle regulation is essential for studying the pathogenesis of hair loss diseases and providing related knowledge for curing these diseases. In addition, WNT signaling overactivation has been found in many types of cancer including skin cancers; and some WNT inhibitors have shown some promising anticancer effects (244-249). Retinoids have been widely applied as a chemotherapy drug for many cancers, such as skin cancers (15,250). Their systemic therapeutic initiation doses start from 0.2 mg/kg/d (120,121), which is much higher than the upper limit (UL) of dietary vitamin A in adults (UL = 0.3 mg/RAE day). The association of topical retinoids and skin cancer is dose dependent: lower concentration retinoid creams (0.001% RA) increased skin cancers in SKH-1 hairless mice (251); while a much higher concentration (0.1% RA) has

been successfully used to treat many skin cancers (15). Our findings could explain that the dose dependent effect of retinoids in skin cancer occurs through the dose dependent

regulation of vitamin A on WNT signaling. This finding may be applied in studying of mechanisms by which retinoids prevent and treat cancers including skin cancers. In the future, it is important to further investigate the mechanisms of how dietary vitamin A regulates the BMP/WNT signaling and confirm the findings in human studies.

Our study also bridged the gap between mouse and human research. This study is the first to report that for CCCA patients, the expression of important vitamin A metabolism and signaling components in the skin decreased as severity increased. In addition, the progression and pathogenesis of CCCA may be different between humans and mice. These findings expand the knowledge of the pathogenesis of CCCA and emphasize the importance of vitamin A metabolism and signaling in the health of skin and hair. In the future, it will be important to understand whether there is causal relationship between vitamin A intake or the alteration of vitamin A metabolism and signaling and the severity of CCCA. Food frequency questionnaires would be a good method to evaluate the vitamin

A and other hair-related nutrient intake among CCCA patients. In addition, randomized controlled trials (RCT) could be done to evaluate the possible effects of vitamin A supplement or topical retinoids for CCCA patients.

In order to further study the function of vitamin A signaling, we collaborated with

Dr. Curley to design a specific ALDH1As inhibitor. BIES-DDC is a potential specific

ALDH1As inhibitor, although further studies still need to be done to confirm the specificity and efficacy of BIES-DDC in vivo. BIES-DDC may have better efficacy and specificity than any ALDH1As inhibitors available in the market, and can be widely applied in the vitamin A research.

This study also shed light on the study of skin SCs and alopecia treatment. The understanding of the regulatory mechanisms by vitamin A controlling the proliferation and differentiation of skin SCs would help develop potential molecular tools to maintain the health of our most important protective shield, the skin.

Bibliography

1. Maden, M. (2008) Retinoids in nonmammalian embryos. Methods Mol Biol 461, 541-559 2. Liu, H. X., Ly, I., Hu, Y., and Wan, Y. J. (2014) Retinoic acid regulates cell cycle genes and accelerates normal mouse liver regeneration. Biochem Pharmacol 91, 256-265 3. Fernandez, I., Tiago, D. M., Laize, V., Leonor Cancela, M., and Gisbert, E. (2014) Retinoic acid differentially affects in vitro proliferation, differentiation and mineralization of two fish bone-derived cell lines: different gene expression of nuclear receptors and ECM proteins. J Steroid Biochem Mol Biol 140, 34-43 4. Trumbo, P., Yates, A. A., Schlicker, S., and Poos, M. (2001) Dietary reference intakes: vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. J Am Diet Assoc 101, 294-301 5. Wolbach, S. B., and Howe, P. R. (1925) Tissue changes following deprivation of fat-soluble A vitamin. J Exp Med 42, 753-777 6. Everts, H. B., Silva, K. A., Montgomery, S., Suo, L., Menser, M., Valet, A. S., King, L. E., Ong, D. E., and Sundberg, J. P. (2013) Retinoid metabolism is altered in human and mouse cicatricial alopecia. J Invest Dermatol 133, 325-333 7. Duncan, F. J., Silva, K. A., Johnson, C. J., King, B. L., Szatkiewicz, J. P., Kamdar, S. P., Ong, D. E., Napoli, J. L., Wang, J., King, L. E., Jr., Whiting, D. A., McElwee, K. J., Sundberg, J. P., and Everts, H. B. (2013) Endogenous retinoids in the pathogenesis of alopecia areata. J Invest Dermatol 133, 334-343 8. Shih, M. Y., Kane, M. A., Zhou, P., Yen, C. L., Streeper, R. S., Napoli, J. L., and Farese, R. V., Jr. (2009) Retinol Esterification by DGAT1 Is Essential for Retinoid Homeostasis in Murine Skin. J Biol Chem 284, 4292-4299 9. Sokolowska-Wojdylo, M., Lugowska-Umer, H., and Maciejewska-Radomska, A. (2013) Oral retinoids and rexinoids in cutaneous T-cell lymphomas. Postepy Dermatol Alergol 30, 19-29 10. Dogra, S., and Yadav, S. (2014) Acitretin in psoriasis: an evolving scenario. Int J Dermatol 53, 525-538 11. Gregoriou, S., Kritsotaki, E., Katoulis, A., and Rigopoulos, D. (2014) Use of tazarotene foam for the treatment of acne vulgaris. Clin Cosmet Investig Dermatol 7, 165-170 12. Kolesnik, M., Gollnick, H., and Bonnekoh, B. (2013) Complete remission of alopecia areata under treatment of chronic hand eczema with alitretinoin. Eur J Dermatol 23, 110-111

13. Melkote, S., Dhurat, R. S., Palav, A., and Jerajani, H. R. (2009) Alopecia in congenital hidrotic ectodermal dysplasia responding to treatment with a combination of topical minoxidil and tretinoin. Int J Dermatol 48, 184-185 14. Shin, H. S., Won, C. H., Lee, S. H., Kwon, O. S., Kim, K. H., and Eun, H. C. (2007) Efficacy of 5% minoxidil versus combined 5% minoxidil and 0.01% tretinoin for male pattern hair loss: a randomized, double-blind, comparative clinical trial. Am J Clin Dermatol 8, 285-290 15. Carr, D. R., Trevino, J. J., and Donnelly, H. B. (2011) Retinoids for chemoprophylaxis of nonmelanoma skin cancer. Dermatol Surg 37, 129-145 16. Teo, K., and Yazdabadi, A. (2014) Isotretinoin and night blindness. Australas J Dermatol 55, 222-224 17. Gungor, S., and Gokdemir, G. (2013) Anal fissure and rectal bleeding as a complication of systemic isotretinoin therapy: dermatologists know this side- effect, what about proctologists? Colorectal Dis 15, 1187-1188 18. Sarkar, R., Chugh, S., and Garg, V. K. (2013) Acitretin in dermatology. Indian J Dermatol Venereol Leprol 79, 759-771 19. Boelsma, E., van de Vijver, L. P., Goldbohm, R. A., Klopping-Ketelaars, I. A., Hendriks, H. F., and Roza, L. (2003) Human skin condition and its associations with nutrient concentrations in serum and diet. Am J Clin Nutr 77, 348-355 20. Li, M., Chiba, H., Warot, X., Messaddeq, N., Gerard, C., Chambon, P., and Metzger, D. (2001) RXR-alpha ablation in skin keratinocytes results in alopecia and epidermal alterations. Development (Cambridge, England) 128, 675-688 21. Everts, H. B. (2012) Endogenous retinoids in the hair follicle and sebaceous gland. Biochim Biophys Acta 1821, 222-229 22. O'Byrne, S. M., and Blaner, W. S. (2013) Retinol and retinyl esters: biochemistry and physiology. J Lipid Res 54, 1731-1743 23. Harrison, E. H. (2012) Mechanisms involved in the intestinal absorption of dietary vitamin A and provitamin A carotenoids. Biochim Biophys Acta 1821, 70- 77 24. Semba, R. D. (2012) On the 'discovery' of vitamin A. Ann Nutr Metab 61, 192- 198 25. Reboul, E., and Borel, P. (2011) Proteins involved in uptake, intracellular transport and basolateral secretion of fat-soluble vitamins and carotenoids by mammalian enterocytes. Prog Lipid Res 50, 388-402 26. Reboul, E., Abou, L., Mikail, C., Ghiringhelli, O., Andre, M., Portugal, H., Jourdheuil-Rahmani, D., Amiot, M. J., Lairon, D., and Borel, P. (2005) Lutein transport by Caco-2 TC-7 cells occurs partly by a facilitated process involving the scavenger receptor class B type I (SR-BI). Biochem J 387, 455-461 27. van Bennekum, A., Werder, M., Thuahnai, S. T., Han, C. H., Duong, P., Williams, D. L., Wettstein, P., Schulthess, G., Phillips, M. C., and Hauser, H. (2005) Class B scavenger receptor-mediated intestinal absorption of dietary beta- carotene and cholesterol. Biochemistry 44, 4517-4525 28. Moussa, M., Landrier, J. F., Reboul, E., Ghiringhelli, O., Comera, C., Collet, X., Frohlich, K., Bohm, V., and Borel, P. (2008) Lycopene absorption in human 97

intestinal cells and in mice involves scavenger receptor class B type I but not Niemann-Pick C1-like 1. J Nutr 138, 1432-1436 29. Harrison, E. H., Gad, M. Z., and Ross, A. C. (1995) Hepatic uptake and metabolism of chylomicron retinyl esters: probable role of plasma membrane/endosomal retinyl ester hydrolases. J Lipid Res 36, 1498-1506 30. Senoo, H., Kojima, N., and Sato, M. (2007) Vitamin A-storing cells (stellate cells). Vitam Horm 75, 131-159 31. Paik, J., Vogel, S., Quadro, L., Piantedosi, R., Gottesman, M., Lai, K., Hamberger, L., Vieira Mde, M., and Blaner, W. S. (2004) Vitamin A: overlapping delivery pathways to tissues from the circulation. J Nutr 134, 276S-280S 32. Gerster, H. (1997) Vitamin A--functions, dietary requirements and safety in humans. Int J Vitam Nutr Res 67, 71-90 33. Akhtar, S., Ahmed, A., Randhawa, M. A., Atukorala, S., Arlappa, N., Ismail, T., and Ali, Z. (2013) Prevalence of vitamin A deficiency in South Asia: causes, outcomes, and possible remedies. J Health Popul Nutr 31, 413-423 34. Chagas, C. B., Saunders, C., Pereira, S., Silva, J., Saboya, C., and Ramalho, A. (2013) Vitamin a deficiency in pregnancy: perspectives after bariatric surgery. Obes Surg 23, 249-254 35. McLaren, D. S., and Kraemer, K. (2012) Xerophthalmia. World Rev Nutr Diet 103, 65-75 36. McLaren, D. S., and Kraemer, K. (2012) Assessment of vitamin A status. World Rev Nutr Diet 103, 52-64 37. Sherwin, J. C., Reacher, M. H., Dean, W. H., and Ngondi, J. (2012) Epidemiology of vitamin A deficiency and xerophthalmia in at-risk populations. Trans R Soc Trop Med Hyg 106, 205-214 38. Yim, C. Y., Mao, P., and Spinella, M. J. (2014) Headway and hurdles in the clinical development of dietary phytochemicals for cancer therapy and prevention: lessons learned from vitamin A derivatives. Aaps j 16, 281-288 39. Penniston, K. L., and Tanumihardjo, S. A. (2006) The acute and chronic toxic effects of vitamin A. Am J Clin Nutr 83, 191-201 40. Russell, R. M. (2000) The vitamin A spectrum: from deficiency to toxicity. Am J Clin Nutr 71, 878-884 41. Reeves, P. G., Nielsen, F. H., and Fahey, G. C., Jr. (1993) AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J Nutr 123, 1939-1951 42. Reeves, P. G. (1997) Components of the AIN-93 diets as improvements in the AIN-76A diet. J Nutr 127, 838s-841s 43. Obrochta, K. M., Kane, M. A., and Napoli, J. L. (2014) Effects of diet and strain on mouse serum and tissue retinoid concentrations. PloS one 9, e99435 44. Kawaguchi, R., Yu, J., Honda, J., Hu, J., Whitelegge, J., Ping, P., Wiita, P., Bok, D., and Sun, H. (2007) A membrane receptor for retinol binding protein mediates cellular uptake of vitamin A. Science 315, 820-825

45. Sun, H., and Kawaguchi, R. (2011) The membrane receptor for plasma retinol- binding protein, a new type of cell-surface receptor. Int Rev Cell Mol Biol 288, 1- 41 46. Duester, G. (2013) Retinoid signaling in control of progenitor cell differentiation during mouse development. Semin Cell Dev Biol 24, 694-700 47. Gudas, L. J. (2012) Emerging roles for retinoids in regeneration and differentiation in normal and disease states. Biochim Biophys Acta 1821, 213-221 48. MacDonald, P. N., and Ong, D. E. (1988) A lecithin:retinol acyltransferase activity in human and rat liver. Biochem Biophys Res Commun 156, 157-163 49. Orland, M. D., Anwar, K., Cromley, D., Chu, C. H., Chen, L., Billheimer, J. T., Hussain, M. M., and Cheng, D. (2005) Acyl coenzyme A dependent retinol esterification by acyl coenzyme A: diacylglycerol acyltransferase 1. Biochim Biophys Acta 1737, 76-82 50. Rexer, B. N., and Ong, D. E. (2002) A novel short-chain alcohol dehydrogenase from rats with retinol dehydrogenase activity, cyclically expressed in uterine epithelium. Biol Reprod 67, 1555-1564 51. Duester, G. (2001) Genetic dissection of retinoid dehydrogenases. Chem Biol Interact 130-132, 469-480 52. Napoli, J. L. (1999) Interactions of retinoid binding proteins and enzymes in retinoid metabolism. Biochim Biophys Acta 1440, 139-162 53. Ong, D. E. (1987) Cellular retinoid-binding proteins. Arch Dermatol 123, 1693- 1695a 54. Chen, A. C., Yu, K., Lane, M. A., and Gudas, L. J. (2003) Homozygous deletion of the CRABPI gene in AB1 embryonic stem cells results in increased CRABPII gene expression and decreased intracellular retinoic acid concentration. Arch Biochem Biophys 411, 159-173 55. Boylan, J. F., and Gudas, L. J. (1992) The level of CRABP-I expression influences the amounts and types of all-trans-retinoic acid metabolites in F9 teratocarcinoma stem cells. J Biol Chem 267, 21486-21491 56. Dong, D., Ruuska, S. E., Levinthal, D. J., and Noy, N. (1999) Distinct roles for cellular retinoic acid-binding proteins I and II in regulating signaling by retinoic acid. J Biol Chem 274, 23695-23698 57. Sessler, R. J., and Noy, N. (2005) A ligand-activated nuclear localization signal in cellular retinoic acid binding protein-II. Mol Cell 18, 343-353 58. Giguere, V., Ong, E. S., Segui, P., and Evans, R. M. (1987) Identification of a receptor for the morphogen retinoic acid. Nature 330, 624-629 59. Balmer, J. E., and Blomhoff, R. (2002) Gene expression regulation by retinoic acid. J Lipid Res 43, 1773-1808 60. Koster, M. I., and Roop, D. R. (2007) Mechanisms regulating epithelial stratification. Annu Rev Cell Dev Biol 23, 93-113 61. Jaks, V., Kasper, M., and Toftgard, R. (2010) The hair follicle-a stem cell zoo. Exp Cell Res 316, 1422-1428 62. Fuchs, E., and Horsley, V. (2008) More than one way to skin. Genes Dev 22, 976- 985 99

63. Al-Refu, K. (2012) Stem cells and alopecia: a review of pathogenesis. Br J Dermatol 167, 479-484 64. Plikus, M. V., Van Spyk, E. N., Pham, K., Geyfman, M., Kumar, V., Takahashi, J. S., and Andersen, B. (2015) The Circadian Clock in Skin: Implications for Adult Stem Cells, Tissue Regeneration, Cancer, Aging, and Immunity. J Biol Rhythms 65. Becker, J. C., and zur Hausen, A. (2014) Cells of origin in skin cancer. J Invest Dermatol 134, 2491-2493 66. Jaks, V., Barker, N., Kasper, M., van Es, J. H., Snippert, H. J., Clevers, H., and Toftgard, R. (2008) Lgr5 marks cycling, yet long-lived, hair follicle stem cells. Nature genetics 40, 1291-1299 67. 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 68. Schneider, M. R., Schmidt-Ullrich, R., and Paus, R. (2009) The hair follicle as a dynamic miniorgan. Curr Biol 19, R132-142 69. 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 70. 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 71. Zhang, Y. V., Cheong, J., Ciapurin, N., McDermitt, D. J., and Tumbar, T. (2009) Distinct self-renewal and differentiation phases in the niche of infrequently dividing hair follicle stem cells. Cell stem cell 5, 267-278 72. 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 73. Plikus, M. V., Mayer, J. A., de la Cruz, D., Baker, R. E., Maini, P. K., Maxson, R., and Chuong, C. M. (2008) Cyclic dermal BMP signalling regulates stem cell activation during hair regeneration. Nature 451, 340-344 74. Silva, K. A., and Sundberg, J. P. (2013) Surgical methods for full-thickness skin grafts to induce alopecia areata in C3H/HeJ mice. Comp Med 63, 392-397 75. O'Shaughnessy, R. F., Christiano, A. M., and Jahoda, C. A. (2004) The role of BMP signalling in the control of ID3 expression in the hair follicle. Exp Dermatol 13, 621-629 76. Hsu, Y. C., and Fuchs, E. (2012) A family business: stem cell progeny join the niche to regulate homeostasis. Nat Rev Mol Cell Biol 13, 103-114 77. Oshimori, N., and Fuchs, E. (2012) Paracrine TGF-beta signaling counterbalances BMP-mediated repression in hair follicle stem cell activation. Cell stem cell 10, 63-75 78. Xi, H. Q., Cai, A. Z., Wu, X. S., Cui, J. X., Shen, W. S., Bian, S. B., Wang, N., Li, J. Y., Lu, C. R., Song, Z., Wei, B., and Chen, L. (2014) Leucine-rich repeat- 100

containing G-protein-coupled receptor 5 is associated with invasion, metastasis, and could be a potential therapeutic target in human gastric cancer. Br J Cancer 79. Choi, Y. S., Zhang, Y., Xu, M., Yang, Y., Ito, M., Peng, T., Cui, Z., Nagy, A., Hadjantonakis, A. K., Lang, R. A., Cotsarelis, G., Andl, T., Morrisey, E. E., and Millar, S. E. (2013) Distinct Functions for Wnt/beta-Catenin in Hair Follicle Stem Cell Proliferation and Survival and Interfollicular Epidermal Homeostasis. Cell stem cell 13, 720-733 80. Van Mater, D., Kolligs, F. T., Dlugosz, A. A., and Fearon, E. R. (2003) Transient activation of beta -catenin signaling in cutaneous keratinocytes is sufficient to trigger the active growth phase of the hair cycle in mice. Genes Dev 17, 1219- 1224 81. Niemann, C., Unden, A. B., Lyle, S., Zouboulis Ch, C., Toftgard, R., and Watt, F. M. (2003) Indian hedgehog and beta-catenin signaling: role in the sebaceous lineage of normal and neoplastic mammalian epidermis. Proc Natl Acad Sci U S A 100 Suppl 1, 11873-11880 82. Allen, M., Grachtchouk, M., Sheng, H., Grachtchouk, V., Wang, A., Wei, L., Liu, J., Ramirez, A., Metzger, D., Chambon, P., Jorcano, J., and Dlugosz, A. A. (2003) Hedgehog signaling regulates sebaceous gland development. Am J Pathol 163, 2173-2178 83. Costaridis, P., Horton, C., Zeitlinger, J., Holder, N., and Maden, M. (1996) Endogenous retinoids in the zebrafish embryo and adult. Dev Dyn 205, 41-51 84. Mathew, L. K., Sengupta, S., Franzosa, J. A., Perry, J., La Du, J., Andreasen, E. A., and Tanguay, R. L. (2009) Comparative expression profiling reveals an essential role for raldh2 in epimorphic regeneration. J Biol Chem 284, 33642- 33653 85. Werner, B., Gallagher, R. E., Paietta, E. M., Litzow, M. R., Tallman, M. S., Wiernik, P. H., Slack, J. L., Willman, C. L., Sun, Z., Traulsen, A., and Dingli, D. (2014) Dynamics of leukemia stem-like cell extinction in acute promyelocytic leukemia. Cancer research 74, 5386-5396 86. Fu, X., Liu, W., Xie, F., and Xiao, R. (2014) [Functions of retinoic acid on the biological characteristics of human embryonic stem cells]. Zhonghua zheng xing wai ke za zhi = Zhonghua zhengxing waike zazhi = Chinese journal of plastic surgery 30, 191-196 87. Lian, X., Selekman, J., Bao, X., Hsiao, C., Zhu, K., and Palecek, S. P. (2013) A small molecule inhibitor of SRC family kinases promotes simple epithelial differentiation of human pluripotent stem cells. PloS one 8, e60016 88. Kidwai, F. K., Liu, H., Toh, W. S., Fu, X., Jokhun, D. S., Movahednia, M. M., Li, M., Zou, Y., Squier, C. A., Phan, T. T., and Cao, T. (2013) Differentiation of human embryonic stem cells into clinically amenable keratinocytes in an autogenic environment. J Invest Dermatol 133, 618-628 89. Baldwin, H. E., Nighland, M., Kendall, C., Mays, D. A., Grossman, R., and Newburger, J. (2013) 40 years of topical tretinoin use in review. J Drugs Dermatol 12, 638-642

101

90. Dawson, A. L., and Dellavalle, R. P. (2013) Acne vulgaris. BMJ (Clinical research ed.) 346, f2634 91. Zouboulis, C. C., Korge, B., Akamatsu, H., Xia, L. Q., Schiller, S., Gollnick, H., and Orfanos, C. E. (1991) Effects of 13-cis-retinoic acid, all-trans-retinoic acid, and acitretin on the proliferation, lipid synthesis and keratin expression of cultured human sebocytes in vitro. J Invest Dermatol 96, 792-797 92. Zouboulis, C. C., Korge, B. P., Mischke, D., and Orfanos, C. E. (1993) Altered proliferation, synthetic activity, and differentiation of cultured human sebocytes in the absence of vitamin A and their modulation by synthetic retinoids. J Invest Dermatol 101, 628-633 93. Passeron, T., Valencia, J. C., Namiki, T., Vieira, W. D., Passeron, H., Miyamura, Y., and Hearing, V. J. (2009) Upregulation of SOX9 inhibits the growth of human and mouse melanomas and restores their sensitivity to retinoic acid. J Clin Invest 119, 954-963 94. Cheepala, S. B., Yin, W., Syed, Z., Gill, J. N., McMillian, A., Kleiner, H. E., Lynch, M., Loganantharaj, R., Trutschl, M., Cvek, U., and Clifford, J. L. (2009) Identification of the B-Raf/Mek/Erk MAP kinase pathway as a target for all-trans retinoic acid during skin cancer promotion. Mol Cancer 8, 27 95. Anforth, R., Blumetti, T. C., Clements, A., Kefford, R., Long, G. V., and Fernandez-Penas, P. (2013) Systemic retinoids for the chemoprevention of cutaneous squamous cell carcinoma and verrucal keratosis in a cohort of patients on BRAF inhibitors. Br J Dermatol 169, 1310-1313 96. Garcia-Fernandez, R. A., Perez-Martinez, C., and Garcia-Iglesias, M. J. (2014) In vivo long-term effects of retinoic acid exposure in utero on induced tumours in adult mouse skin. Vet Dermatol 25, 538-546, e593-534 97. Nolting, S., Giubellino, A., Tayem, Y., Young, K., Lauseker, M., Bullova, P., Schovanek, J., Anver, M., Fliedner, S., Korbonits, M., Goke, B., Vlotides, G., Grossman, A., and Pacak, K. (2014) Combination of 13-Cis retinoic acid and lovastatin: marked antitumor potential in vivo in a pheochromocytoma allograft model in female athymic nude mice. Endocrinology 155, 2377-2390 98. Asgari, M. M., Brasky, T. M., and White, E. (2012) Association of vitamin A and carotenoid intake with melanoma risk in a large prospective cohort. J Invest Dermatol 132, 1573-1582 99. Bongiovanni, L., Di Diodoro, F., Della Salda, L., and Brachelente, C. (2014) On the role of survivin as a stem cell biomarker of canine hair follicle and related tumours. Vet Dermatol 25, 138-141, e139-140 100. Keyes, W. M., Pecoraro, M., Aranda, V., Vernersson-Lindahl, E., Li, W., Vogel, H., Guo, X., Garcia, E. L., Michurina, T. V., Enikolopov, G., Muthuswamy, S. K., and Mills, A. A. (2011) DeltaNp63alpha is an oncogene that targets chromatin remodeler Lsh to drive skin stem cell proliferation and tumorigenesis. Cell stem cell 8, 164-176 101. Lapouge, G., Youssef, K. K., Vokaer, B., Achouri, Y., Michaux, C., Sotiropoulou, P. A., and Blanpain, C. (2011) Identifying the cellular origin of squamous skin tumors. Proc Natl Acad Sci U S A 108, 7431-7436 102

102. White, A. C., Tran, K., Khuu, J., Dang, C., Cui, Y., Binder, S. W., and Lowry, W. E. (2011) Defining the origins of Ras/p53-mediated squamous cell carcinoma. Proc Natl Acad Sci U S A 108, 7425-7430 103. Olasz, J., Juhasz, A., Remenar, E., Engi, H., Bak, M., Csuka, O., and Kasler, M. (2007) RAR beta2 suppression in head and neck squamous cell carcinoma correlates with site, histology and age. Oncol Rep 18, 105-112 104. Fan, J., Eastham, L., Varney, M. E., Hall, A., Adkins, N. L., Chetel, L., Sollars, V. E., Georgel, P., and Niles, R. M. (2010) Silencing and re-expression of retinoic acid receptor beta2 in human melanoma. Pigment Cell Melanoma Res 23, 419- 429 105. Fazi, F., Zardo, G., Gelmetti, V., Travaglini, L., Ciolfi, A., Di Croce, L., Rosa, A., Bozzoni, I., Grignani, F., Lo-Coco, F., Pelicci, P. G., and Nervi, C. (2007) Heterochromatic gene repression of the retinoic acid pathway in acute myeloid leukemia. Blood 109, 4432-4440 106. Faria, T. N., Mendelsohn, C., Chambon, P., and Gudas, L. J. (1999) The targeted disruption of both alleles of RARbeta(2) in F9 cells results in the loss of retinoic acid-associated growth arrest. J Biol Chem 274, 26783-26788 107. Pili, R., Salumbides, B., Zhao, M., Altiok, S., Qian, D., Zwiebel, J., Carducci, M. A., and Rudek, M. A. (2012) Phase I study of the histone deacetylase inhibitor entinostat in combination with 13-cis retinoic acid in patients with solid tumours. Br J Cancer 106, 77-84 108. Tavares, T. S., Nanus, D., Yang, X. J., and Gudas, L. J. (2008) Gene microarray analysis of human renal cell carcinoma: the effects of HDAC inhibition and retinoid treatment. Cancer Biol Ther 7, 1607-1618 109. Tang, X. H., Albert, M., Scognamiglio, T., and Gudas, L. J. (2009) A DNA methyltransferase inhibitor and all-trans retinoic acid reduce oral cavity carcinogenesis induced by the carcinogen 4-nitroquinoline 1-oxide. Cancer Prev Res (Phila) 2, 1100-1110 110. Anzano, M. A., Lamb, A. J., and Olson, J. A. (1979) Growth, appetite, sequence of pathological signs and survival following the induction of rapid, synchronous vitamin A deficiency in the rat. J Nutr 109, 1419-1431 111. Berth-Jones, J., and Hutchinson, P. E. (1995) Novel cycle changes in scalp hair are caused by etretinate therapy. Br J Dermatol 132, 367-375 112. Bergfeld, W. F. (1998) Retinoids and hair growth. J Am Acad Dermatol 39, S86- 89 113. Baldwin, T. J., Rood, K. A., Kelly, E. J., and Hall, J. O. (2012) Dermatopathy in juvenile Angus cattle due to vitamin A deficiency. J Vet Diagn Invest 24, 763-766 114. Klein-Szanto, A. J., Martin, D. H., and Pine, A. H. (1980) Cutaneous manifestations in rats with advanced vitamin A deficiency. J Cutan Pathol 7, 260- 270 115. Ocon, J., Cabrejas, C., Altemir, J., and Moros, M. (2012) Phrynoderma: a rare dermatologic complication of bariatric surgery. JPEN J Parenter Enteral Nutr 36, 361-364

103

116. Bleasel, N. R., Stapleton, K. M., Lee, M. S., and Sullivan, J. (1999) Vitamin A deficiency phrynoderma: due to malabsorption and inadequate diet. J Am Acad Dermatol 41, 322-324 117. Blasiak, R. C., Stamey, C. R., Burkhart, C. N., Lugo-Somolinos, A., and Morrell, D. S. (2013) High-dose isotretinoin treatment and the rate of retrial, relapse, and adverse effects in patients with acne vulgaris. JAMA Dermatol 149, 1392-1398 118. Ghasri, P., Yentzer, B. A., Dabade, T. S., and Feldman, S. R. (2011) Acitretin for the treatment of psoriasis: an assessment of national trends. J Drugs Dermatol 10, 873-877 119. Alizadeh, F., Bolhassani, A., Khavari, A., Bathaie, S. Z., Naji, T., and Bidgoli, S. A. (2014) Retinoids and their biological effects against cancer. Int Immunopharmacol 18, 43-49 120. Harwood, C. A., Leedham-Green, M., Leigh, I. M., and Proby, C. M. (2005) Low-dose retinoids in the prevention of cutaneous squamous cell carcinomas in organ transplant recipients: a 16-year retrospective study. Arch Dermatol 141, 456-464 121. Pearce, D. J., Klinger, S., Ziel, K. K., Murad, E. J., Rowell, R., and Feldman, S. R. (2006) Low-dose acitretin is associated with fewer adverse events than high- dose acitretin in the treatment of psoriasis. Arch Dermatol 142, 1000-1004 122. Gupta, A. K., Goldfarb, M. T., Ellis, C. N., and Voorhees, J. J. (1989) Side-effect profile of acitretin therapy in psoriasis. J Am Acad Dermatol 20, 1088-1093 123. Katz, H. I., Waalen, J., and Leach, E. E. (1999) Acitretin in psoriasis: an overview of adverse effects. J Am Acad Dermatol 41, S7-S12 124. Koo, J., Nguyen, Q., and Gambla, C. (1997) Advances in psoriasis therapy. Adv Dermatol 12, 47-72; discussion 73 125. Murray, H. E., Anhalt, A. W., Lessard, R., Schacter, R. K., Ross, J. B., Stewart, W. D., and Geiger, J. M. (1991) A 12-month treatment of severe psoriasis with acitretin: results of a Canadian open multicenter study. J Am Acad Dermatol 24, 598-602 126. Foitzik, K., Spexard, T., Nakamura, M., Halsner, U., and Paus, R. (2005) Towards dissecting the pathogenesis of retinoid-induced hair loss: all-trans retinoic acid induces premature hair follicle regression (catagen) by upregulation of transforming growth factor-beta2 in the dermal papilla. J Invest Dermatol 124, 1119-1126 127. Everts, H. B., Sundberg, J. P., King, L. E., Jr., and Ong, D. E. (2007) Immunolocalization of enzymes, binding proteins, and receptors sufficient for retinoic acid synthesis and signaling during the hair cycle. J Invest Dermatol 127, 1593-1604 128. Beildeck, M. E., Gelmann, E. P., and Byers, S. W. (2010) Cross-regulation of signaling pathways: an example of nuclear hormone receptors and the canonical Wnt pathway. Exp Cell Res 316, 1763-1772 129. Tanese, K., Fukuma, M., Yamada, T., Mori, T., Yoshikawa, T., Watanabe, W., Ishiko, A., Amagai, M., Nishikawa, T., and Sakamoto, M. (2008) G-protein-

104

coupled receptor GPR49 is up-regulated in basal cell carcinoma and promotes cell proliferation and tumor formation. Am J Pathol 173, 835-843 130. Thompson, D. L., Gerlach-Bank, L. M., Barald, K. F., and Koenig, R. J. (2003) Retinoic acid repression of bone morphogenetic protein 4 in inner ear development. Mol Cell Biol 23, 2277-2286 131. 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., Stange, D. E., Toftgard, R., and Clevers, H. (2010) Lgr6 marks stem cells in the hair follicle that generate all cell lineages of the skin. Science 327, 1385-1389 132. Liu, L., Suzuki, K., Nakagata, N., Mihara, K., Matsumaru, D., Ogino, Y., Yashiro, K., Hamada, H., Liu, Z., Evans, S. M., Mendelsohn, C., and Yamada, G. (2012) Retinoic acid signaling regulates sonic hedgehog and bone morphogenetic protein signalings during genital tubercle development. Birth Defects Res B Dev Reprod Toxicol 95, 79-88 133. Rodriguez-Leon, J., Merino, R., Macias, D., Ganan, Y., Santesteban, E., and Hurle, J. M. (1999) Retinoic acid regulates programmed cell death through BMP signalling. Nat Cell Biol 1, 125-126 134. Metallo, C. M., Ji, L., de Pablo, J. J., and Palecek, S. P. (2008) Retinoic acid and bone morphogenetic protein signaling synergize to efficiently direct epithelial differentiation of human embryonic stem cells. Stem cells (Dayton, Ohio) 26, 372- 380 135. Bilousova, G., Chen, J., and Roop, D. R. (2011) Differentiation of mouse induced pluripotent stem cells into a multipotent keratinocyte lineage. J Invest Dermatol 131, 857-864 136. Mulholland, D. J., Dedhar, S., Coetzee, G. A., and Nelson, C. C. (2005) Interaction of nuclear receptors with the Wnt/beta-catenin/Tcf signaling axis: Wnt you like to know? Endocr Rev 26, 898-915 137. Byers, S., Pishvaian, M., Crockett, C., Peer, C., Tozeren, A., Sporn, M., Anzano, M., and Lechleider, R. (1996) Retinoids increase cell-cell adhesion strength, beta- catenin protein stability, and localization to the cell membrane in a breast cancer cell line: a role for serine kinase activity. Endocrinology 137, 3265-3273 138. Ryuto, M., Jimi, S., Ono, M., Naito, S., Nakayama, Y., Yamada, Y., Komiyama, S., and Kuwano, M. (1997) All-trans-retinoic acid-dependent inhibition of E- cadherin-based cell adhesion with concomitant dephosphorylation of beta-catenin in metastatic human renal carcinoma cells. Jpn J Cancer Res 88, 982-991 139. Zhao, X., and Duester, G. (2009) Effect of retinoic acid signaling on Wnt/beta- catenin and FGF signaling during body axis extension. Gene Expr Patterns 9, 430-435 140. Verani, R., Cappuccio, I., Spinsanti, P., Gradini, R., Caruso, A., Magnotti, M. C., Motolese, M., Nicoletti, F., and Melchiorri, D. (2007) Expression of the Wnt inhibitor Dickkopf-1 is required for the induction of neural markers in mouse embryonic stem cells differentiating in response to retinoic acid. J Neurochem 100, 242-250

105

141. Memezawa, A., Takada, I., Takeyama, K., Igarashi, M., Ito, S., Aiba, S., Kato, S., and Kouzmenko, A. P. (2007) Id2 gene-targeted crosstalk between Wnt and retinoid signaling regulates proliferation in human keratinocytes. Oncogene 26, 5038-5045 142. Zito, G., Saotome, I., Liu, Z., Ferro, E. G., Sun, T. Y., Nguyen, D. X., Bilguvar, K., Ko, C. J., and Greco, V. (2014) Spontaneous tumour regression in keratoacanthomas is driven by Wnt/retinoic acid signalling cross-talk. Nat Commun 5, 3543 143. Picardi, A., Pasquini, P., Cattaruzza, M. S., Gaetano, P., Baliva, G., Melchi, C. F., Papi, M., Camaioni, D., Tiago, A., Gobello, T., and Biondi, M. (2003) Psychosomatic factors in first-onset alopecia areata. Psychosomatics 44, 374-381 144. McMichael, A. J., Pearce, D. J., Wasserman, D., Camacho, F. T., Fleischer, A. B., Jr., Feldman, S. R., and Balkrishnan, R. (2007) Alopecia in the United States: outpatient utilization and common prescribing patterns. J Am Acad Dermatol 57, S49-51 145. Harries, M. J., and Paus, R. (2010) The pathogenesis of primary cicatricial alopecias. Am J Pathol 177, 2152-2162 146. King, L. E., Jr., McElwee, K. J., and Sundberg, J. P. (2008) Alopecia areata. Curr Dir Autoimmun 10, 280-312 147. Duvic, M., Welsh, E. A., Jackow, C., Papadopoulos, E., Reveille, J. D., and Amos, C. (1995) Analysis of HLA-D locus alleles in alopecia areata patients and families. J Invest Dermatol 104, 5s-6s 148. Shellow, W. V., Edwards, J. E., and Koo, J. Y. (1992) Profile of alopecia areata: a questionnaire analysis of patient and family. Int J Dermatol 31, 186-189 149. Biran, R., Zlotogorski, A., and Ramot, Y. (2015) The genetics of alopecia areata: New approaches, new findings, new treatments. J Dermatol Sci 150. Perret, C., Wiesner-Menzel, L., and Happle, R. (1984) Immunohistochemical analysis of T-cell subsets in the peribulbar and intrabulbar infiltrates of alopecia areata. Acta Derm Venereol 64, 26-30 151. Sundberg, J. P., Cordy, W. R., and King, L. E., Jr. (1994) Alopecia areata in aging C3H/HeJ mice. J Invest Dermatol 102, 847-856 152. Whiting, D. A. (2003) Histopathologic features of alopecia areata: a new look. Arch Dermatol 139, 1555-1559 153. Olsen, E. A., Callender, V., McMichael, A., Sperling, L., Anstrom, K. J., Shapiro, J., Roberts, J., Durden, F., Whiting, D., and Bergfeld, W. (2011) Central hair loss in African American women: incidence and potential risk factors. J Am Acad Dermatol 64, 245-252 154. Dlova, N. C., Jordaan, F. H., Sarig, O., and Sprecher, E. (2014) Autosomal dominant inheritance of central centrifugal cicatricial alopecia in black South Africans. Journal of the American Academy of Dermatology 70, 679-682 e671 155. Ogunleye, T. A., McMichael, A., and Olsen, E. A. (2014) Central centrifugal cicatricial alopecia: what has been achieved, current clues for future research. Dermatol Clin 32, 173-181

106

156. LoPresti, P., Papa, C. M., and Kligman, A. M. (1968) Hot comb alopecia. Arch Dermatol 98, 234-238 157. Olsen, E. A., Bergfeld, W. F., Cotsarelis, G., Price, V. H., Shapiro, J., Sinclair, R., Solomon, A., Sperling, L., Stenn, K., Whiting, D. A., Bernardo, O., Bettencourt, M., Bolduc, C., Callendar, V., Elston, D., Hickman, J., Ioffreda, M., King, L., Linzon, C., McMichael, A., Miller, J., Mulinari, F., and Trancik, R. (2003) Summary of North American Hair Research Society (NAHRS)-sponsored Workshop on Cicatricial Alopecia, Duke University Medical Center, February 10 and 11, 2001. Journal of the American Academy of Dermatology 48, 103-110 158. Olsen, E. A., Callender, V., Sperling, L., McMichael, A., Anstrom, K. J., Bergfeld, W., Durden, F., Roberts, J., Shapiro, J., and Whiting, D. A. (2008) Central scalp alopecia photographic scale in African American women. Dermatologic therapy 21, 264-267 159. Sundberg, J. P., Taylor, D., Lorch, G., Miller, J., Silva, K. A., Sundberg, B. A., Roopenian, D., Sperling, L., Ong, D., King, L. E., and Everts, H. (2011) Primary follicular dystrophy with scarring dermatitis in C57BL/6 mouse substrains resembles central centrifugal cicatricial alopecia in humans. Vet Pathol 48, 513- 524 160. Khumalo, N. P., Jessop, S., Gumedze, F., and Ehrlich, R. (2007) Hairdressing and the prevalence of scalp disease in African adults. Br J Dermatol 157, 981-988 161. Kyei, A., Bergfeld, W. F., Piliang, M., and Summers, P. (2011) Medical and environmental risk factors for the development of central centrifugal cicatricial alopecia: a population study. Arch Dermatol 147, 909-914 162. Dlova, N. C., and Forder, M. (2012) Central centrifugal cicatricial alopecia: possible familial aetiology in two African families from South Africa. Int J Dermatol 51 Suppl 1, 17-20, 20-13 163. Ohyama, M. (2012) Primary cicatricial alopecia: recent advances in understanding and management. J Dermatol 39, 18-26 164. Stenn, K. S. (2001) Insights from the asebia mouse: a molecular sebaceous gland defect leading to cicatricial alopecia. J Cutan Pathol 28, 445-447 165. Karnik, P., Tekeste, Z., McCormick, T. S., Gilliam, A. C., Price, V. H., Cooper, K. D., and Mirmirani, P. (2009) Hair follicle stem cell-specific PPARgamma deletion causes scarring alopecia. The Journal of investigative dermatology 129, 1243-1257 166. Sundberg, J. P., Boggess, D., Sundberg, B. A., Eilertsen, K., Parimoo, S., Filippi, M., and Stenn, K. (2000) Asebia-2J (Scd1(ab2J)): a new allele and a model for scarring alopecia. Am J Pathol 156, 2067-2075 167. Flowers, M. T., Paton, C. M., O'Byrne, S. M., Schiesser, K., Dawson, J. A., Blaner, W. S., Kendziorski, C., and Ntambi, J. M. (2011) Metabolic changes in skin caused by Scd1 deficiency: a focus on retinol metabolism. PloS one 6, e19734 168. Stenn, K. S., Sundberg, J. P., and Sperling, L. C. (1999) Hair follicle biology, the sebaceous gland, and scarring alopecias. Arch Dermatol 135, 973-974

107

169. Harries, M. J., Meyer, K. C., and Paus, R. (2009) Hair loss as a result of cutaneous autoimmunity: frontiers in the immunopathogenesis of primary cicatricial alopecia. Autoimmun Rev 8, 478-483 170. Suo, L., Sundberg, J. P., and Everts, H. B. (2014) Dietary vitamin A regulates wingless-related MMTV integration site signaling to alter the hair cycle. Exp Biol Med (Maywood) 171. Petersson, M., and Niemann, C. (2012) Stem cell dynamics and heterogeneity: implications for epidermal regeneration and skin cancer. Curr Med Chem 19, 5984-5992 172. Mundi, J. P., Marmon, S., Fischer, M., Kamino, H., Patel, R., and Shapiro, J. (2012) Dissecting cellulitis of the scalp. Dermatol Online J 18, 8 173. Udayashankar, C., Nath, A. K., and Anuradha, P. (2013) Extensive Darier's disease with pityriasis amiantacea, alopecia and congenital facial nerve palsy. Dermatol Online J 19, 18574 174. Koppaka, V., Thompson, D. C., Chen, Y., Ellermann, M., Nicolaou, K. C., Juvonen, R. O., Petersen, D., Deitrich, R. A., Hurley, T. D., and Vasiliou, V. (2012) Aldehyde dehydrogenase inhibitors: a comprehensive review of the pharmacology, mechanism of action, substrate specificity, and clinical application. Pharmacol Rev 64, 520-539 175. Song, Y. H., Zhong, M. Z., Gan, P. P., Yi, P. Y., Tang, Y. H., Liu, Y. P., Jiang, J. Q., and Li, L. (2014) ALDH1A1 mediates resistance of diffuse large B cell lymphoma to the CHOP regimen. Tumour Biol 35, 11809-11817 176. Li, X. S., Xu, Q., Fu, X. Y., and Luo, W. S. (2014) ALDH1A1 overexpression is associated with the progression and prognosis in gastric cancer. BMC cancer 14, 705 177. Meng, E., Mitra, A., Tripathi, K., Finan, M. A., Scalici, J., McClellan, S., Madeira da Silva, L., Reed, E., Shevde, L. A., Palle, K., and Rocconi, R. P. (2014) ALDH1A1 maintains ovarian cancer stem cell-like properties by altered regulation of cell cycle checkpoint and DNA repair network signaling. PloS one 9, e107142 178. Bell, R. G., and Smith, H. W. (1949) Preliminary report on clinical trials of antabuse. Can Med Assoc J 60, 286-288 179. Kim, S. K., Kim, H., Lee, D. H., Kim, T. S., Kim, T., Chung, C., Koh, G. Y., Kim, H., and Lim, D. S. (2013) Reversing the intractable nature of pancreatic cancer by selectively targeting ALDH-high, therapy-resistant cancer cells. PloS one 8, e78130 180. van den Hoogen, C., van der Horst, G., Cheung, H., Buijs, J. T., Pelger, R. C., and van der Pluijm, G. (2011) The aldehyde dehydrogenase enzyme 7A1 is functionally involved in prostate cancer bone metastasis. Clin Exp Metastasis 28, 615-625 181. Mahmoud, M. I., Potter, J. J., Colvin, O. M., Hilton, J., and Mezey, E. (1993) Effect of 4-(diethylamino)benzaldehyde on ethanol metabolism in mice. Alcohol Clin Exp Res 17, 1223-1227

108

182. Kronmiller, J. E., Beeman, C. S., Nguyen, T., and Berndt, W. (1995) Blockade of the initiation of murine odontogenesis in vitro by citral, an inhibitor of endogenous retinoic acid synthesis. Arch Oral Biol 40, 645-652 183. Le Bouffant, R., Guerquin, M. J., Duquenne, C., Frydman, N., Coffigny, H., Rouiller-Fabre, V., Frydman, R., Habert, R., and Livera, G. (2010) Meiosis initiation in the human ovary requires intrinsic retinoic acid synthesis. Hum Reprod 25, 2579-2590 184. Moore, S. A., Baker, H. M., Blythe, T. J., Kitson, K. E., Kitson, T. M., and Baker, E. N. (1998) Sheep liver cytosolic aldehyde dehydrogenase: the structure reveals the basis for the retinal specificity of class 1 aldehyde dehydrogenases. Structure (London, England : 1993) 6, 1541-1551 185. Boyer, C. S., and Petersen, D. R. (1991) The metabolism of 3,7-dimethyl-2,6- octadienal (citral) in rat hepatic mitochondrial and cytosolic fractions. Interactions with aldehyde and alcohol dehydrogenases. Drug Metab Dispos 19, 81-86 186. Russo, J. E., Hauguitz, D., and Hilton, J. (1988) Inhibition of mouse cytosolic aldehyde dehydrogenase by 4-(diethylamino)benzaldehyde. Biochem Pharmacol 37, 1639-1642 187. Xing, L., Dai, Z., Jabbari, A., Cerise, J. E., Higgins, C. A., Gong, W., de Jong, A., Harel, S., DeStefano, G. M., Rothman, L., Singh, P., Petukhova, L., Mackay- Wiggan, J., Christiano, A. M., and Clynes, R. (2014) Alopecia areata is driven by cytotoxic T lymphocytes and is reversed by JAK inhibition. Nat Med 188. Gilhar, A., Etzioni, A., and Paus, R. (2012) Alopecia areata. N Engl J Med 366, 1515-1525 189. Everts, H. B., King, L. E., Jr., Sundberg, J. P., and Ong, D. E. (2004) Hair cycle- specific immunolocalization of retinoic acid synthesizing enzymes Aldh1a2 and Aldh1a3 indicate complex regulation. J Invest Dermatol 123, 258-263 190. Niehrs, C. (2012) The complex world of WNT receptor signalling. Nat Rev Mol Cell Biol 13, 767-779 191. 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 110, 1351-1356 192. Plikus, M. V. (2012) New activators and inhibitors in the hair cycle clock: targeting stem cells' state of competence. J Invest Dermatol 132, 1321-1324 193. Boer, J., and van Gemert, M. J. (1999) Long-term results of isotretinoin in the treatment of 68 patients with hidradenitis suppurativa. J Am Acad Dermatol 40, 73-76 194. 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., and Clevers, H. (2007) Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003-1007 195. Watanabe, M. (2014) Adult tissue stem cell therapy for gastrointestinal diseases. J Gastroenterol Hepatol

109

196. Peng, C., Zhang, X., Wang, Y., Li, L., Wang, Q., and Zheng, J. (2012) Expression and prognostic significance of wnt7a in human endometrial carcinoma. Obstet Gynecol Int 2012, 134962 197. Shamonki, M. I., Kligman, I., Shamonki, J. M., Schattman, G. L., Hyjek, E., Spandorfer, S. D., Zaninovic, N., and Rosenwaks, Z. (2006) Immunohistochemical expression of endometrial L-selectin ligand is higher in donor egg recipients with embryonic implantation. Fertility and sterility 86, 1365- 1375 198. Wong, S. C., Lo, E. S., Chan, A. K., Lee, K. C., and Hsiao, W. L. (2003) Nuclear beta catenin as a potential prognostic and diagnostic marker in patients with colorectal cancer from Hong Kong. Molecular pathology : MP 56, 347-352 199. Zouboulis, C. C. (2001) Retinoids--which dermatological indications will benefit in the near future? Skin Pharmacol Appl Skin Physiol 14, 303-315 200. Reddy, S., Andl, T., Bagasra, A., Lu, M. M., Epstein, D. J., Morrisey, E. E., and Millar, S. E. (2001) Characterization of Wnt gene expression in developing and postnatal hair follicles and identification of Wnt5a as a target of Sonic hedgehog in hair follicle morphogenesis. Mech Dev 107, 69-82 201. Okano, J., Levy, C., Lichti, U., Sun, H. W., Yuspa, S. H., Sakai, Y., and Morasso, M. I. (2012) Cutaneous retinoic acid levels determine hair follicle development and downgrowth. J Biol Chem 287, 39304-39315 202. Ito, M., Yang, Z., Andl, T., Cui, C., Kim, N., Millar, S. E., and Cotsarelis, G. (2007) Wnt-dependent de novo hair follicle regeneration in adult mouse skin after wounding. Nature 447, 316-320 203. Sassi, N., Laadhar, L., Allouche, M., Zandieh-Doulabi, B., Hamdoun, M., Klein- Nulend, J., Makni, S., and Sellami, S. (2014) The roles of canonical and non- canonical Wnt signaling in human de-differentiated articular chondrocytes. Biotech Histochem 89, 53-65 204. von Maltzahn, J., Bentzinger, C. F., and Rudnicki, M. A. (2012) Wnt7a-Fzd7 signalling directly activates the Akt/mTOR anabolic growth pathway in skeletal muscle. Nat Cell Biol 14, 186-191 205. 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, 501-511 206. Monfrecola, G., and Baldo, A. (2009) Retinoids and phototherapy for psoriasis. J Rheumatol Suppl 83, 71-72 207. Kmiec, M. L., Pajor, A., and Broniarczyk-Dyla, G. (2013) Evaluation of biophysical skin parameters and assessment of hair growth in patients with acne treated with isotretinoin. Postepy Dermatol Alergol 30, 343-349 208. Ruzicka, T., Sommerburg, C., Goerz, G., Kind, P., and Mensing, H. (1992) Treatment of cutaneous lupus erythematosus with acitretin and hydroxychloroquine. Br J Dermatol 127, 513-518 209. Astrom, A., Tavakkol, A., Pettersson, U., Cromie, M., Elder, J. T., and Voorhees, J. J. (1991) Molecular cloning of two human cellular retinoic acid-binding 110

proteins (CRABP). Retinoic acid-induced expression of CRABP-II but not CRABP-I in adult human skin in vivo and in skin fibroblasts in vitro. J Biol Chem 266, 17662-17666 210. Durand, B., Saunders, M., Leroy, P., Leid, M., and Chambon, P. (1992) All-trans and 9-cis retinoic acid induction of CRABPII transcription is mediated by RAR- RXR heterodimers bound to DR1 and DR2 repeated motifs. Cell 71, 73-85 211. Liao, X. H., and Nguyen, H. (2014) Epidermal expression of Lgr6 is dependent on nerve endings and Schwann cells. Exp Dermatol 23, 195-198 212. Kandyba, E., and Kobielak, K. (2013) Wnt7b is an important intrinsic regulator of hair follicle stem cell homeostasis and hair follicle cycling. Stem cells (Dayton, Ohio) 213. Kurlandsky, S. B., Duell, E. A., Kang, S., Voorhees, J. J., and Fisher, G. J. (1996) Auto-regulation of retinoic acid biosynthesis through regulation of retinol esterification in human keratinocytes. J Biol Chem 271, 15346-15352 214. Matsuzaka, Y., Okamoto, K., Tsuji, H., Mabuchi, T., Ozawa, A., Tamiya, G., and Inoko, H. (2002) Identification of the hRDH-E2 gene, a novel member of the SDR family, and its increased expression in psoriatic lesion. Biochem Biophys Res Commun 297, 1171-1180 215. Lee, S. A., Belyaeva, O. V., Wu, L., and Kedishvili, N. Y. (2011) Retinol dehydrogenase 10 but not retinol/sterol dehydrogenase(s) regulates the expression of retinoic acid-responsive genes in human transgenic skin raft culture. J Biol Chem 286, 13550-13560 216. Hiroi, A., Ito, T., Seo, N., Uede, K., Yoshimasu, T., Ito, M., Nakamura, K., Ito, N., Paus, R., and Furukawa, F. (2006) Male New Zealand Black/KN mice: a novel model for autoimmune-induced permanent alopecia? Br J Dermatol 155, 437-445 217. 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 218. Pozdnyakova, O., and Mahalingam, M. (2008) Involvement of the bulge region in primary scarring alopecia. J Cutan Pathol 35, 922-925 219. Al-Refu, K., Edward, S., Ingham, E., and Goodfield, M. (2009) Expression of hair follicle stem cells detected by cytokeratin 15 stain: implications for pathogenesis of the scarring process in cutaneous lupus erythematosus. Br J Dermatol 160, 1188-1196 220. Blanpain, C., and Fuchs, E. (2009) Epidermal homeostasis: a balancing act of stem cells in the skin. Nat Rev Mol Cell Biol 10, 207-217 221. Mongan, N. P., and Gudas, L. J. (2007) Diverse actions of retinoid receptors in cancer prevention and treatment. Differentiation 75, 853-870 222. Chapellier, B., Mark, M., Messaddeq, N., Calleja, C., Warot, X., Brocard, J., Gerard, C., Li, M., Metzger, D., Ghyselinck, N. B., and Chambon, P. (2002) Physiological and retinoid-induced proliferations of epidermis basal keratinocytes are differently controlled. EMBO J 21, 3402-3413

111

223. Kang, S., Duell, E. A., Fisher, G. J., Datta, S. C., Wang, Z. Q., Reddy, A. P., Tavakkol, A., Yi, J. Y., Griffiths, C. E., Elder, J. T., and et al. (1995) Application of retinol to human skin in vivo induces epidermal hyperplasia and cellular retinoid binding proteins characteristic of retinoic acid but without measurable retinoic acid levels or irritation. J Invest Dermatol 105, 549-556 224. Griffiths, C. E., Finkel, L. J., Tranfaglia, M. G., Hamilton, T. A., and Voorhees, J. J. (1993) An in vivo experimental model for effects of topical retinoic acid in human skin. Br J Dermatol 129, 389-394 225. Bose, A., Teh, M. T., Mackenzie, I. C., and Waseem, A. (2013) Keratin k15 as a biomarker of epidermal stem cells. Int J Mol Sci 14, 19385-19398 226. Sperling, L. C., Hussey, S., Wang, J. A., and Darling, T. (2011) Cytokeratin 15 expression in central, centrifugal, cicatricial alopecia: new observations in normal and diseased hair follicles. J Cutan Pathol 38, 407-414 227. de The, H., Marchio, A., Tiollais, P., and Dejean, A. (1989) Differential expression and ligand regulation of the retinoic acid receptor alpha and beta genes. EMBO J 8, 429-433 228. Leung, Y., Kandyba, E., Chen, Y. B., Ruffins, S., and Kobielak, K. (2013) Label retaining cells (LRCs) with myoepithelial characteristic from the proximal acinar region define stem cells in the sweat gland. PloS one 8, e74174 229. Hsu, Y. C., Li, L., and Fuchs, E. (2014) Emerging interactions between skin stem cells and their niches. Nature medicine 20, 847-856 230. Al-Zaid, T., Vanderweil, S., Zembowicz, A., and Lyle, S. (2011) Sebaceous gland loss and inflammation in scarring alopecia: a potential role in pathogenesis. J Am Acad Dermatol 65, 597-603 231. Braun-Falco, O., Imai, S., Schmoeckel, C., Steger, O., and Bergner, T. (1986) Pseudopelade of Brocq. Dermatologica 172, 18-23 232. Amato, L., Mei, S., Massi, D., Gallerani, I., and Fabbri, P. (2002) Cicatricial alopecia; a dermatopathologic and immunopathologic study of 33 patients (pseudopelade of Brocq is not a specific clinico-pathologic entity). Int J Dermatol 41, 8-15 233. Napoli, J. L., and Race, K. R. (1987) The biosynthesis of retinoic acid from retinol by rat tissues in vitro. Arch Biochem Biophys 255, 95-101 234. Walker, J. M. (1994) The bicinchoninic acid (BCA) assay for protein quantitation. Methods Mol Biol 32, 5-8 235. Mira, Y. L. R., Zheng, W. L., Kuppumbatti, Y. S., Rexer, B., Jing, Y., and Ong, D. E. (2000) Retinol conversion to retinoic acid is impaired in breast cancer cell lines relative to normal cells. J Cell Physiol 185, 302-309 236. Rexer, B. N., Zheng, W. L., and Ong, D. E. (2001) Retinoic acid biosynthesis by normal human breast epithelium is via aldehyde dehydrogenase 6, absent in MCF- 7 cells. Cancer Res 61, 7065-7070 237. Klyosov, A. A. (1996) Kinetics and specificity of human liver aldehyde dehydrogenases toward aliphatic, aromatic, and fused polycyclic aldehydes. Biochemistry 35, 4457-4467

112

238. Cheung, C., Davies, N. G., Hoog, J. O., Hotchkiss, S. A., and Smith Pease, C. K. (2003) Species variations in cutaneous alcohol dehydrogenases and aldehyde dehydrogenases may impact on toxicological assessments of alcohols and aldehydes. Toxicology 184, 97-112 239. Sellheyer, K., Atanaskova-Mesinkovska, N., Nelson, P., and Bergfeld, W. F. (2011) Differential expression of stem cell markers in lichen planopilaris and alopecia areata. Br J Dermatol 165, 1149-1151 240. 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, 411-417 241. Ohyama, M., Terunuma, A., Tock, C. L., Radonovich, M. F., Pise-Masison, C. A., Hopping, S. B., Brady, J. N., Udey, M. C., and Vogel, J. C. (2006) Characterization and isolation of stem cell-enriched human hair follicle bulge cells. J Clin Invest 116, 249-260 242. Amoh, Y., Maejima, H., Niiyama, S., Mii, S., Hamada, Y., Saito, N., and Katsuoka, K. (2011) Hair follicle stem cell marker nestin expression in regenerating hair follicles of patients with alopecia areata. Eur J Dermatol 21, 209-212 243. Pascucci, L., Mercati, F., Gargiulo, A. M., Pedini, V., Sorbolini, S., and Ceccarelli, P. (2006) CD34 glycoprotein identifies putative stem cells located in the isthmic region of canine hair follicles. Vet Dermatol 17, 244-251 244. Fonar, Y., and Frank, D. (2011) FAK and WNT signaling: the meeting of two pathways in cancer and development. Anticancer Agents Med Chem 11, 600-606 245. Su, T. R., Lin, J. J., Tsai, C. C., Huang, T. K., Yang, Z. Y., Wu, M. O., Zheng, Y. Q., Su, C. C., and Wu, Y. J. (2013) Inhibition of melanogenesis by gallic acid: possible involvement of the PI3K/Akt, MEK/ERK and Wnt/beta-catenin signaling pathways in B16F10 cells. Int J Mol Sci 14, 20443-20458 246. Brinkhuizen, T., van den Hurk, K., Winnepenninckx, V. J., de Hoon, J. P., van Marion, A. M., Veeck, J., van Engeland, M., and van Steensel, M. A. (2012) Epigenetic changes in Basal Cell Carcinoma affect SHH and WNT signaling components. PloS one 7, e51710 247. Kang, M. I., Baker, A. R., Dextras, C. R., Cabarcas, S. M., Young, M. R., and Colburn, N. H. (2012) Targeting of Noncanonical Wnt5a Signaling by AP-1 Blocker Dominant-Negative Jun When It Inhibits Skin Carcinogenesis. Genes Cancer 3, 37-50 248. Ren, X., Zheng, D., Guo, F., Liu, J., Zhang, B., Li, H., and Tian, W. (2015) PPARgamma suppressed Wnt/beta-catenin signaling pathway and its downstream effector SOX9 expression in gastric cancer cells. Med Oncol 32, 536 249. Hua, H. W., Jiang, F., Huang, Q., Liao, Z., and Ding, G. (2015) MicroRNA-153 promotes Wnt/beta-catenin activation in hepatocellular carcinoma through suppression of WWOX. Oncotarget 250. Lens, M., and Medenica, L. (2008) Systemic retinoids in chemoprevention of non-melanoma skin cancer. Expert Opin Pharmacother 9, 1363-1374

113

251. (2012) Photocarcinogenesis study of retinoic acid and retinyl palmitate [CAS Nos. 302-79-4 (All-trans-retinoic acid) and 79-81-2 (All-trans-retinyl palmitate)] in SKH-1 mice (Simulated Solar Light and Topical Application Study). Natl Toxicol Program Tech Rep Ser, 1-352

114