Galactomyces Ferment Filtrate Suppresses

Melanization and Oxidative Stress in Epidermal Melanocytes

A dissertation submitted to the

Division of Research and Advanced Studies of the University of Cincinnati

in partial fulfillment of the requirements for the degree of

DOCTORATE OF PHILOSOPHY (Ph.D.)

in the Division of Pharmaceutical Sciences of the James L. Winkle College of Pharmacy

by

JàNay Karen Woolridge Cooper 2017

B.S. Biological Sciences University of Missouri, Columbia, MO, 2009

Committee Chair: Raymond E. Boissy, Ph.D.

ABSTRACT

Epidermal melanocytes are particularly susceptible to oxidative stress and cellular cytotoxicity in comparison to the other primary skin cell types due to their normal cell function of melanin synthesis which produces reactive intermediates and oxygen species as byproducts. This vulnerability encourages the development of oxidative driven diseases such as vitiligo, a depigmenting skin disorder resulting from the dysfunction or death of melanocytes consumed by biological stress and immune targeting. Galactomyces ferment filtrate (GFF, Pitera™) is a yeast derived extract comprised of a unique blend of vitamins, minerals, small peptides, and oligosaccharides. GFF is currently used as a moisturizing agent in cosmetics and has demonstrated anti-aging, barrier enhancing, and hypopigmenting capacities in skin. GFF may have a therapeutic utility that is uniquely applicable for vitiligo patients through diminishing melanization and coupled reactive oxygen species generation, while concurrently upregulating intracellular antioxidant activity. To commence this initiative, the fundamental biological mechanisms of GFF were investigated in human pigment cell cultures.

The leadoff study assessed the effect of GFF on melanization in cultured human melanoma cells and normal human melanocytes. GFF successfully inhibited melanization in both long- and short- term treatment protocols. The activity of tyrosinase, the primary melanogenic enzyme that commands biochemical melanin synthesis, was also revealed to be diminished by GFF. mRNA expression of novel ion channels and transporters associated with established modulators of cellular and organelle pH were shown to be substantially downregulated in melanocytes treated with GFF; thus, prompting the investigation of pH within the melanosome, the specialized acidic

ii organelle that hosts the melanogenesis process. ABCC9, TRPM6, TRPV1, SLC9A4, and

SLC13A1 were the gene targets selected for further experimentation. Expression of these uncharacterized ion channels and transporters were found to be reduced in GFF treated melanocytes. Three of these pore proteins also exhibited significant colocalization with melanogenic enzymes and proteins, which implies subcellular localization to melanosomes.

Overall, this data demonstrated that GFF efficaciously suppressed melanization in the human pigment cell model in part by deterring tyrosine hydroxylase activity. The new working hypothesis is that GFF significantly alters the pH of the melanosome counter to optimal tyrosine hydroxylase activity via downregulation of ion transport proteins.

The second study evaluated the effect of GFF in dampening oxidative stress in normal epidermal melanocytes. GFF proved cytoprotective capabilities after sustaining the cell viability in melanocytes assaulted with 4-tertiary butylphenol, an apoptotic phenolic compound. GFF also suppressed reactive oxygen species generation in cells individually challenged with 4-tertiary butylphenol and ultraviolet radiation. mRNA expression from gene regulatory elements involved in oxidative stress response were shown to be modulated after treatment; thus, instigating an examination of the Nrf2-ARE pathway. GFF positively regulated the expression of the transcription factor Nrf2 and the established downstream phase II enzyme targets HO1, NQO1, and TXNRD1 in both long and short-term treatment protocols. These results begin to explain the mechanisms of action of GFF in the suppression of melanization and oxidative stress in epidermal melanocytes.

iii

iv ACKNOWLEDGMENTS

For the revelation awaits an appointed time; it speaks of the end and will not prove false.

Though it linger, wait for it; it will certainly come and will not delay.

Habakkuk 2:3, NIV

I give my utmost gratitude to my Heavenly Father for setting this path before me and equipping me with everything I needed to complete this journey. Thank you for picking me up out of every dark moment, fortifying the foundation of our faith relationship, and empowering my witness through this test. It wasn’t pretty, but I stand in awe and await the full completion of your work.

To my parents, Eric and Renée Cooper, I love you and thank you for everything you’ve done throughout my life to get me to this moment. Mommi, you’ve always advised me to be the best in my field and command my destiny through education at the highest level. I am grateful for your vision and the tenacious spirit you instill in your children. Dad, I cannot begin to describe how invaluable your spiritual wisdom and guidance has been to my life and especially through these last six years. I am beyond appreciative for your presence and wholeheartedly share this victory with you by adopting your last name in hopes that I may bring it honor for the rest of my days.

Thank you both for your refreshing encouragement, love, and care. Eric Jr., you are the comedic relief and true natural talent that brings completion to our household. I’m proud being your big sister.

v To my beautifully, divine Purnell family: Pop, Mom, G, and YaYa. You are the light force behind everything. Since my birth, you created an environment that has enriched my deepest nature and personality. I am so proud to be a product of your love and provision. I miss you desperately and will carry your colorful memory with me in all that I do in this life. Thank you for helping me become.

My best friends and family: Christopher Davis, Jessica Winslow, and Kendra Julion, sustaining this great effort would have been impossible without you. Christopher, thank you for supporting my dream from the beginning. Jessica, our spiritual sisterhood and like-mindedness continues to hearten my disposition and I am excited to travel this life path together. Kendra, thank you for being my respite from harsh realities. You each have upheld me through this journey in an essential manner. I am blessed to have you in my circle and will forever cherish our bonds.

I sincerely appreciate the University of Cincinnati, the James L. Winkle College of Pharmacy, and the College of Medicine for this prime educational opportunity and the financial support. I especially thank my supervisory professor and committee chair Dr. Raymond Boissy for welcoming me in to the famed Boissy laboratory and allowing me to develop organically as a scientist. Your critical evaluations and enthusiasm have been motivational throughout the project, especially for my transition into the Doctorate program. I also thank my committee members: Dr.

R. Randall Wickett, Dr. Ana Luisa Kadekaro, Dr. Gary Kelm, and Dr. Yuhang Zhang; collectively, you are an astounding group of scientists with extremely pertinent insight and feedback for this project that has been vital to its successes. I particularly thank Dr. Wickett for accepting into the

Cosmetic Science program in addition to your multifaceted guidance in a variety of my student

vi affairs. I also am inspired by Dr. Kadekaro; your passion for student education and mentorship as well as the thoroughness of your work embodies the scientist I aspire to become. I express gratitude to the entire UC Department of Dermatology for your amassed kindnesses and support throughout my time here. Amy Koshoffer, I am also grateful for your efforts in initiating my advancement as a scientist and teaching me 90% of the techniques and bioassays I know! I appreciate Dr. Zalfa

Abdel-Malek and her staff Renny Starner and Viki Swope, whom I had the pleasure of working alongside in the laboratory, for all their assistance and instruction. Additionally, I wish to acknowledge Procter & Gamble, Co. and Dr. Tomohiro Hakozaki for the opportunity to work on such a dynamic project that has been particularly well suited to my scientific and professional interests in pigmentation biology and cosmetics.

I am eternally thankful for the host of teachers, coaches, and mentors I have encountered throughout my lengthy scholastic journey for your joint contributions, with special thanks to the initiator Mrs. Sharon Dryden. I also have deep appreciation for Dr. Terri Hill, my mentor of twelve years and the staff at Terri L. Hill, M.D., PA. Thank you all for your support of my professional and educational endeavors beginning in my teenage years. To Mrs. Karen Henry, I cannot thank you enough for all your mentorship, care, and advocacy during my time at UC. I admire your intentionality and devotion to higher education and student affairs for both myself and the entire student roster of the Cosmetic Science program. You have been nothing but helpful and encouraging in the utmost manner.

Finally, I thank a host of special friends who have enriched my life during my time in graduate school. Nikkita Womack, my dear friend and former teammate, thank you for your unwavering

vii support, wise words, and precious friendship. Dr. Anne von Koschembahr, thank you for being such a considerate friend and motivator both personally and in the laboratory experience; I am inspired by your own educational journey and the determination you displayed in overcoming difficulties. Dr. Shoná Burkes-Henderson, thank you for your example and the wealth of peer guidance in navigating the Cosmetic Science program and the city of Cincinnati. I especially thank all the prayer warriors, especially LaCheryl Jones, for keeping me elevated throughout this process. Lastly, but not in the least, to the TAKERS, the city’s best kickball team, I enjoyed creating our legacy of winning. Thanks for tolerating my inner child and being a weekly stress reliever.

I am beyond joyful to close this chapter in my life and propel into a fruitful future.

God bless.

viii DEDICATION

This dissertation is dedicated to the lives and loving memory of my grandparents

Marvin Wilmer and Ernestine JoAnn Purnell.

I will love and cherish you forever.

ix TABLE OF CONTENTS

ABSTRACT………………………………………………………………………………………ii

ACKNOWLEDGMENTS………………………………………………………………………...v

DEDICATION……………………………………………………………………………………ix

TABLE OF CONTENTS………………………………………………………………………….1

LIST OF TABLES………………………………………………………………………………...4

LIST OF FIGURES………………………………………………………………………………..5

LIST OF ABBREVIATIONS……………………………………………………………………..7

CHAPTER 1: INTRODUCTION………………………………………………………………..11

1.1 Vitiligo……………………………………………………………………………………….12

1.2 Research Aim………………………………………………………………………………...13

CHAPTER 2: SCIENTIFIC BACKGROUND…………………………………………………..14

2.1 Human Skin…………………………………………………………………………………..15

2.2 Skin Pigmentation……………………………………………………………………………18

2.2.1 Melanin Biochemical Pathway…………………………………………………………19

2.2.2 Melanosome Maturation ……………………………………………………………….20

2.3 Melanosome pH Modulates Pigmentation...…………………………………………………22

2.4 Melanocyte Redox and Inherent Cytotoxicity………………………………………………..26

2.5 Melanocyte Antioxidant Defense…………………………………………………………….29

2.5.1 Enzymatic Antioxidants………………………………………………………………..29

2.5.2 Nonenzymatic Antioxidants……………………………………………………………35

2.5.3 Nrf2-ARE Pathway in Skin…………………………………………………………….35

1 2.6 Oxidative Stress in Vitiligo…………………………………………………………………..39

2.7 Properties of Galactomyces Ferment Filtrate………………………………………………...44

CHAPTER 3: RATIONALE, OBJECTIVE, HYPOTHESES, AND SPECIFIC AIMS…………46

3.1 Rationale and Objective……………………………………………………………………...47

3.2 Hypotheses and Specific Aims……………………………………………………………….49

CHAPTER 4: GALACTOMYCES FERMENT FILTRATE SUPPRESSES MELANIZATION IN

HUMAN PIGMENT CELLS……………………….……………………………………………51

4.1 Abstract………………………………………………………………………………………52

4.1.1 Keywords………………………………………………………………………………53

4.2 Introduction……………………………………………………………………………..……54

4.3 Material and Methods………………………………………………………………….……..56

4.4 Results………………………………………………………………………………………..62

4.5 Discussion…………………………………………………………………………………....66

4.6 Acknowledgments……………………………………………………………………………70

4.7 Author Contributions…………………………………………………………………………71

4.8 Figures and Tables.………………………………………………………………………...…72

4.9 Supplemental Figures and Tables…………………………………………………………….83

CHAPTER 5: GALACTOMYCES FERMENT FILTRATE SUPPRESSES REACTIVE

OXYGEN SPECIES GENERATION AND PROMOTES CELLULAR REDOX BALANCE IN

HUMAN MELANOCYTES VIA NRF2-ARE PATHWAY…………………………………….87

5.1 Abstract………………………………………………………………………………………88

5.1.1 Keywords………………………………………………………………………………88

5.2 Introduction…………………………………………………………………………………..89

2 5.3 Material and Methods………………………………………………………………………...92

5.4 Results………………………………………………………………………………………..97

5.5 Discussion…………………………………………………………………………………..101

5.6 Acknowledgments…………………………………………………………………………..107

5.7 Author Contributions………………………………………………………………………..108

5.8 Figures and Tables…………………………………………………………………………..109

5.9 Supplemental Figures and Tables…………………………………………………………...118

CHAPTER 6: CONCLUSION………………………………………………………………….120

6.1 Major Findings……………………………………………………………………………...121

6.2 Synergy Between Melanization and Redox Properties of Galactomyces Ferment Filtrate….122

6.3 Future Recommendations…………………………………………………………………...127

REFERENCES…………………………………………………………………………………128

3 LIST OF TABLES

Table Page

2.1 Melanocyte antioxidant enzymes 30

4.1 RNA-seq melanogenic protein targets – 24 hours 77

4.2 RNA-seq melanogenic protein targets – 120 hours 77

4.3 RNA-seq pore protein targets – 24 hours 79

4.4 RNA-seq pore protein targets – 120 hours 79

4.5 Colocalization of melanogenic and pore proteins 82

4.S1 Western blot analysis of melanogenic protein expression 86

5.1 RNA-seq antioxidant enzyme targets – 24 hours 113

5.2 RNA-seq antioxidant enzyme targets – 120 hours 113

4 LIST OF FIGURES

Figure Page

2.1 Cross-section of human skin, the integumentary system 15

2.2 Epidermal melanin unit 18

2.3 Biochemical pathway of melanin synthesis and the generation of ROS 20

2.4 Melanosome maturation 21

2.5 Putative melanosomal ion transport proteins 23

2.6 Melanocyte enzymatic antioxidant defense against reactive quinones and

oxygen species 34

2.7 Nrf2-ARE pathway activation and deactivation 38

4.1 GFF effect on melanin content in human SK-MEL and NHEM 72

4.2 GFF effect on melanosome maturity and quantity in NHEM 73

4.3 GFF effect on tyrosine hydroxylase activity in situ 74

4.4 GFF effect on tyrosine hydroxylase activity in vitro 75

4.5 GFF effect on tyrosine hydroxylase activity of mushroom tyrosinase 76

4.6 GFF effect on melanogenic proteins expression 78

4.7 GFF effect on pore proteins expression 80

4.8 Colocalization imaging of melanogenic and pore proteins 81

4.S1 Melanin content and cell proliferation in NHEM after long term dosing 83

4.S2 GFF visual effect on melanization in NHEM 84

4.S3 GFF effect on gene transcription in NHEM 85

5.1 GFF effect on cell viability in 4TBP challenged NHEM 109

5 5.2 GFF effect on 4TBP induced ROS generation 110

5.3 GFF effect on UV induced H2O2 generation 111

5.4 GFF effect on various oxidative stress targets in NHEM 112

5.5 GFF effect on Nrf2-ARE pathway 114

5.6 GFF effect on Nrf2-ARE pathway in the cytoplasm and nucleus 116

5.S1 GFF effect on Nrf2-ARE pathway challenged by UVB 118

5.S2 Cytoplasmic and nuclear HO1 expression 119

6 LIST OF ABBREVIATIONS

•OH Hydroxyl radical

4TBP 4-tertiary butylphenol

α-MSH alpha-Melanocyte-stimulating hormone

ABCC9/SUR2 ATP Binding Cassette Subfamily C Member 9/Sulfonylurea Receptor 2

ARE Antioxidants response element

BVR Biliverdin reductase

Ca2+ Calcium ion

CAT Catalase

CFTR Cystic fibrosis transmembrane conductance regulator

Cl- Chloride ion

ClC Chloride channels

CO Carbon monoxide

CO2 Carbon dioxide

Cul3 Cullin 3

CYBA Cytochrome B-245, alpha polypeptide

DHI 5,6-dihydroxyindole

DHICA 5,6-dihydoxyindole-2-carboxylic acid

DOPA 3,4-dihydroxyphenylalanine

EMC Endogenous melanogenic cytotoxicity

Fe2+ Ferrous Iron

Fe3+ Ferric Iron

7 G6PD Glucose-6-phosphate dehydrogenase

GCLC γ-Glutamate-cysteine ligase, catalytic subunit

GLCM γ-Glutamate-cysteine ligase, modifier subunit

GFF Galactomyces ferment filtrate

GFF-A Galactomyces ferment filtrate A

GFF-B Galactomyces ferment filtrate B

GPX Glutathione peroxidase

GSH Glutathione

GSR Glutathione reductase

GSSG Glutathione disulfide

H+ Hydrogen ion, Proton

H2O2 Hydrogen peroxide

HO Heme oxygenase

IDH Isocitrate dehydrogenase

IL-2 Interleukin-2 iNOS Inducible nitric oxide synthase

K+ Potassium ion

Keap1 Kelch-like ECH-associated protein 1

LPS Lipopolysaccharide

MC1R Melanocortin 1 receptor

MITF Micropthalmia-assocaited transcription factor

Na+ Sodium ion

NADP+ Nicotinamide adenine dinucleotide phosphate

8 NADPH Reduced nicotinamide adenine dinucleotide phosphate

Nrf2 Nuclear factor, erythroid 2-like 2

NHEM Normal human epidermal melanocytes

NO Nitric oxide

NOX NADPH oxidase

NQO NAD(P)H quinone oxidoreductase

O2 Molecular oxygen

- O2 Superoxide anion

OH- Hydroxide

ONOO− Peroxynitrite

PIH Post-inflammatory hyperpigmentation

PMEL Premelanosome protein

PRDX Peroxiredoxin

RNA-seq RNA sequencing

ROS Reactive oxygen species

SK-MEL Human melanoma

SLC Solute carrier family

SLC9A4/NHE4 Solute carrier family 9 member A4/ Na+/H+ exchanger 4

SLC13A1 Solute carrier family 13 member 1

SOD Superoxide dismutase

SSRs Short sequence reads

TRP Transient receptor potential channels

9 TRPM6 Transient receptor potential cation channel subfamily M (melastatin)

member 6

TRPV1/VR1 Transient receptor potential cation channel subfamily V member 1/

Vanilloid receptor 1

TXN Thioredoxin

TXN-S2 Thioredoxin disulfide

TXNRD Thioredoxin reductase

TYR Tyrosinase

TYRP1 Tyrosinase related protein 1

TYRP2/DCT Tyrosinase related protein 2/ Dopachrome tautomerase

UVR Ultraviolet radiation

UVB Ultraviolet B

V-ATPase Vacuolar-type proton ATPase

10

CHAPTER 1:

INTRODUCTION

11 1.1 Vitiligo

Vitiligo is an acquired, depigmentary disease that manifests as unpredictable, colorless patches on skin. These lesions typically occur on sun exposed areas (face, upper chest), the extremities (hands, feet, arms, and legs), body folds (armpits and groin), and around body openings such as the eyes or mouth. Vitiligo affects 1-2% of the world’s population, occurring equally in people of all skin colors, races, and genders [American Academy of Dermatology, 2017; National Vitiligo

Foundation Inc., 2017]. This disease is not immediately life threatening or contagious; however, patients experience both negative physical and psychosocial complications including sunburn, increased skin cancer risk, rapid skin aging, visual or hearing impairment, and severe emotional distress. Treatment options are distinguished by two rudimentary strategies: (i) restoring the natural skin color through repigmentation therapy; or (ii) destruction of the remaining natural skin pigment to achieve an even appearance through depigmentation therapy. Unfortunately, the success rates are highly varied and patient specific. Because of the persisting efficacy and patient satisfaction challenges with current therapeutic options, there is a need for innovation in treatments and skincare prevention methods.

12 1.2 Research Aim

Galactomyces ferment filtrate (GFF), commercially known as Pitera™, is currently used as a cosmetic ingredient in the prestige skincare line SK-II, manufactured by Procter & Gamble, Co.

GFF is a yeast derived extract that contains a unique composition of vitamins, minerals, small peptides, and oligosaccharides. Clinically, Pitera™ demonstrates anti-aging and hypopigmenting effects on skin within 14 days [Procter & Gamble, Co., 2017]. The long-term goal of this research project is to understand the usability of GFF/ Pitera™ as a topically applied therapeutic agent for vitiligo.

13

CHAPTER 2:

SCIENTIFIC BACKGROUND

14 2.1 Human Skin

Skin is the functional outer covering that separates an organism from the external environment.

Skin operates as a protective barrier, homeostatic regulator, and sensory register to promote organism survival. In humans, the skin is the largest organ that accounts for approximately 15% of the total body weight [Khan et al., 2015]. Human skin is comprised of three preeminent cell types: fibroblasts, keratinocytes, and melanocytes. These cells differentiate and organize the skin into two distinct layers denoted as the epidermis and dermis, which attach to the subcutis/hypodermis, constituting the integumentary system (Figure 2.1).

Figure 2.1 – Cross-section of human skin, the integumentary system [Modified from Mescher, 2009]. The stratum lucidium in the epidermis is an additional layer that only occurs on the palms and soles.

15 The epidermis is the topmost layer divisible into four sublayers or strata: stratum basale/germinativum, stratum spinosum, stratum granulosum, and stratum corneum (Figure 2.1).

Keratinocytes are the primary cell type of the epidermis, by more than 95%, which produce the fibrous structural protein keratin [Yaar & Gilchrest, 2001]. These cells form the tangible barrier of skin via differential keratin expression throughout the four sublayers. Keratinocytes are generated in the stratum basale and subsequently transform into corneocytes as they migrate towards the stratum corneum; the visible, outermost sublayer of skin. This conversion fuels a constant renewal cycle of the stratified layers while the uppermost strata are shed in a process known as desquamation; thus, promoting a robust skin barrier. There is also a mild population of melanocytes within the epidermis, 1-2% of the total epidermal cells, that are responsible for propagating and maintaining the pigmentary system or skin color [Yaar & Gilchrest, 2001].

Melanocytes are only found in the stratum basale at the epidermal-dermal junction, the area where the epidermis anchors to the dermis via a laminins, fibrils, and filaments lattice. Additional cellular elements for sensory and immune functions are also present throughout this skin layer. Most notably, the epidermis provides water loss prevention, obstruction to invasive microorganisms, chemical/pollutant resistance, and photoprotection from ultraviolet radiation (UVR) through the pigmentary system. Proper maintenance of the epidermal barrier is critical to survival.

The dermis is the dominant layer in skin, split into two sublayers: papillary and reticular.

Fibroblasts are the principle cell type in the dermis, accountable for collagen and elastin production and secretion; these proteins combine with glycosaminoglycans such as hyaluronan to the form the watery extracellular matrix (ECM) that gives skin tensile strength and connectivity. The papillary dermis participates in metabolic exchange with the epidermis through papillae and blood

16 capillaries at the epidermal-dermal junction. The reticular dermis encompasses most of the functional skin appendages including: hair follicles, sebaceous glands, nerve endings, sensory receptors, apocrine glands, eccrine glands, etc. (Figure 2.1). Altogether, the dermis provides essential skin processes including: thermoregulation; sensatory operations; and skin energy, flexibility, and strength.

The first two layers of skin are affixed to the body through the hypodermis/subcutis, a superficial connective fascia. The hypodermis is chiefly comprised of adipocytes that organize into fatty adipose tissue which form an insulating, cushiony padding for mechanical protection against physical stress and for smooth contour. Fibroblasts also help construct the loosely packed ECM of the areolar tissues found in this layer of skin. The hypodermis physically supports the base of the skin found in the reticular dermis and houses some sensory receptors and a dense network of blood and lymph vasculature.

17 2.2 Skin Pigmentation

Biological anthropology established that human skin color is a protective adaptation of proximity to UVR exposure [Jablonski & Chaplin, 2010]. Cutaneous melanocytes are a modest cell population located in the stratum basale at the epidermal-dermal junction in skin and hair follicles that is charged with the principle role of photoprotection via pigment synthesis and transfer. These cells generate melanin, a chromophore pigment that efficiently absorbs and dissipates UVR, acting as a natural sunscreen. Melanin is constructed and retained in specialized organelles referred to as melanosomes, which undergo a four-stage maturation process within the melanocyte preceding transfer to neighboring keratinocytes. Once inside the keratinocyte, matured melanosomes are distributed above the nucleus forming a protective cap that shields from UVR and scavenges nearby reactive oxygen species (ROS). One melanocyte provides pigment to 36 surrounding keratinocytes via dendritic projections extending from the cell body; this complex is termed an epidermal melanin unit (Figure 2.2) [Cichorek et al., 2013; Nordlund & Boissy, 2001]. The keratinocytes retain the melanin pigment as they transform through the cornification process, producing the visible skin tone seen within the stratum corneum.

Figure 2.2 – Epidermal melanin unit [Cichorek et al., 2013].

18 2.2.1 Melanin Biochemical Pathway

Melanin biosynthesis involves a series of oxidation reactions; beginning with hydroxylation of the amino acid L-tyrosine to dihydroxyphenylalanine (DOPA) and subsequent oxidation of DOPA to

DOPAquinone, mutually catalyzed by the copper containing enzyme tyrosinase (TYR) (Figure

2.3) [Denat et al., 2014]. Following DOPAquinone production, the biochemical pathway bifurcates into subpathways that generate two chemically distinct melanin chromophores: brown/black eumelanin and red/yellow pheomelanin. Each chromophore product is produced separately in individual melanosomes. The eumelanogenesis pathway incorporates three enzymes,

TYR and tyrosinase related proteins -1 and -2 (TYRP1, TYRP2), contrasting pheomelanogenesis, a spontaneous process that will only occur in the presence of glutathione (GSH) or cysteine. The expression of these melanogenic enzymes is primarily regulated by microphthalmia-associated transcription factor (MITF) via multiple signaling cascades, including precedent activation of the melanocortin 1 receptor (MC1R) located on the cell membrane. The resulting pigmentary phenotype is a genetically programmed baseline total quantity of melanin in combination with the ratio of eumelanin to pheomelanin, known as constitutive pigmentation. Melanization that is induced by extraneous factors like UVR, hormones, or stressors is referred to as facultative pigmentation [Nordlund & Boissy, 2001].

19

Figure 2.3 – Biochemical pathway of melanin synthesis and the generation of ROS [Modified from Denat et al. 2014]. DHI = dihydroxyindole, DHICA = dihydroxyindole-2-carboxylic acid; DOPA = dihydroxyphenylalanine; GSH = glutathione; TYRP1 = tyrosinase related protein 1; TYRP2 = tyrosinase related protein 2

2.2.2 Melanosome Maturation

Melanosome biogenesis is not fully characterized; yet, most research alludes to a derivative pathway of endolysosomal origins [Raposo & Marks, 2002; Marks & Seabra, 2001; Orlow 1995].

Melanosomes mature through four distinct stages within the melanocyte during melanogenesis

(Figure 2.4) [Marks & Seabra, 2001]. Stage I premelanosomes are characterized by a spherical shape with ill-defined internal filaments [Orlow, 1995]. Stage II melanosomes develop an elliptical

20 shape filled with well-defined internal striations, signifying the structural premelanosome (PMEL) protein matrix that sequesters incoming melanogenic enzymes [Raposo et al., 2001; Orlow, 1995].

Stage III melanosomes show initiation of polymerized melanin accumulation on the matrix filaments and Stage IV melanosomes have complete opacification of the internal compartment

[Raposo et al., 2001; Orlow, 1995]. Melanosome maturation occurs as the organelles move away from the cell body down into the extending dendrites.

Figure 2.4 – Melanosome maturation [Marks & Seabra, 2001]. (a) Schematic diagram of cell body and dendrite of a skin melanocyte with stage I through IV melanosomes. (b) Electron micrograph of stages II, III, and IV melanosomes. “Courtesy of L. Collinson and C. Hopkins, Imperial College, London, UK.”

21 2.3 Melanosomal pH Modulates Pigmentation

Because melanosomes have a lysosomal lineage, the intramelanosomal pH was initially predicted to be acidic, estimated to range from pH 3.0-5.0 [Orlow, 1995]. Lysosomes are digestive organelles known to be highly acidic with a pH range of 4.0-5.0 [Mindell, 2012; DiCiccio & Steinberg, 2011].

Lysosomes maintain their acidic nature chiefly through vacuolar-type proton ATPase (V-ATPase) activity (Figure 2.5) [Bellono & Oancea, 2014; DiCiccio & Steinberg, 2011]. V-ATPases use ATP to acidify the lysosome interior by internalizing protons (H+). This strategy also requires movement of counterions to dissipate the transmembrane voltage, suggesting that for every proton translocated there is one cation removed or one anion introduced [Mindell, 2012; DiCiccio &

Steinberg, 2011]. Counterion movements for lysosomal acidification can be achieved by various mechanisms; notable proteins involved include cystic fibrosis transmembrane conductance regulator (CFTR); chloride (Cl-) channel 7 (ClC-7); and transient receptor potential cation channel, mucolipin subfamily, member 1 (TRPML1) (Figure 2.5). CFTR is a Cl- channel uniquely activated by ATP and phosphorylation that facilitates downhill anion flux [Mindell, 2012].

Belonging to the ATP-binding cassette (ABC) transporter family, these proteins use ATP hydrolysis to move substrates against their concentration gradients [Jones et al., 2009]. The chloride family (ClC) is comprised of homodimer protein channels and pumps for Cl-. ClC-7 has been identified in the lysosomal membrane and ClC-4 and ClC-5 in the endosome [DiCiccio &

Steinberg, 2011; Graves et al., 2008; Mohammad-Panah et al., 2003; Günther et al., 1998].

TRPML1 is a member of the mucolipin subfamily of transient receptor potential (TRP) channels, which consists of tetrameric cation channels with six transmembrane domains that transport potassium (K+), sodium (Na+), calcium (Ca2+), and magnesium (Mg2+) cations. Comparable acidification proteins (V-ATPases and ion transporters for H+, Na+, K+, Ca2+, Cl-) have been

22 Figure 2.5 – Putative melanosomal ion transport proteins [Bellono & Oancea, 2014]. Schematic diagram of ion channels and transporters on the membranes of endolysosomes (left) and melanosomes (right) “Bidirectional arrows represent anti- or co-transporters and unidirectional arrows represent uniporters or pumps.” V-ATPase = vacuolar-type proton ATPase; TRPML= transient receptor potential cation channels, mucolipin subfamily; TPC = two-pore channels; ClC-7 = chloride channel 7; TYR = tyrosinase; OCA2 = oculocutaneous albinism II; SLC = solute carrier family; OA1 = osteoarthritis QTL 1; ATP7A = ATPase, copper-transporting, A polypeptide.

identified in melanosomes; thus, melanosome acidification is speculated to align with the strategies

used by lysosomes [Bellono & Oancea, 2014; Dooley et al., 2013; Cheli et al., 2009; Raposo et

al., 2002; Raposo et al., 2001]. Melanosomal V-ATPases and solute carrier family (SLC) proteins

demonstrate involvement in pH regulation; SLC24A4, SLC24A5, and SLC45A2 are already

acknowledged in the control of pigmentation in the eye, hair, and skin (Figure 2.5) [Bellono &

Oancea, 2014; Cheli et al., 2009]. Additional members of the TRP family, such as TRPM1 and

23 TRPA1, have been identified in the cell membrane of melanocytes and may contribute to the influx of Ca2+ into the cytosol that also exhibits a regulatory link with pigment synthesis [Bellono et al.,

2013; Devi et al., 2009; Oancea et al., 2009].

Modulation of pigmentation by intramelanosomal pH has been theorized throughout scientific literature. Internal pH changes are predicted to occur during melanosome maturation; initially, the melanosome pH is considered acidic, which is thought to be optimal for tyrosine hydroxylase activity and L-DOPA stabilization [Schallreuter et al., 2008]. Acidic pH of melanosomes is also necessary for ensuring the maximal binding of melanin proteins for the formation of high molecular weight melano-protein complexes [Mani et al., 2001]. However, melanin synthesis in human melanocyte lysates was found maximal at pH 6.8 and optimal TYR activity in pH range

6.5-7.0 [Schallreuter et al. 2008; Ancans et al., 2001]. Melanin formation from commercial substrates increased along with pH and the soluble and insoluble ratios were altered at different pH values [Mani et al., 2001]. The ratio of pheomelanin to eumelanin was also affected by pH demonstrated by treatment with proton pump inhibitors on human melanocyte cultures [Ancans et al., 2001]. α-Melanocyte-stimulating hormone (α-MSH), the MC1R agonist that stimulates melanin production, increased the pH of melanosomes in B16 mouse melanoma cells suggesting optimal melanin formation at a neutral pH [Cheli et al., 2009]. Similar results were established in human melanocyte and melanoma cultures, showing increases in melanogenesis after treatment with V-ATPase inhibitors concanamycin A and bafilomycin A1 [Ancans et al., 2001]. Contrarily,

TYR activity was increased under acidic conditions in neonatal mice melanocyte cultures [Puri et al., 2000]. At this time, specific pH values for the melanosomal lumen have only been reported in

MNT-1 human melanoma cells, at 5.7 ± 0.6, and B16-F1 mouse melanoma cells, at 5.51 ± 0.02

24 [Ambrosio et al., 2016; Bellono et al. 2016]. There are conflicting accounts on the ideal pH for melanin synthesis and optimal TYR activity; nonetheless, it is evident that pH of the melanosome is an appreciable factor in influencing melanin output and there is a need for continued study in human pigment cell models to elucidate this action.

25 2.4 Melanocyte Redox and Inherent Cytotoxicity

Reactive oxygen species (ROS) are defined as volatile oxygen-based molecules derived from molecular oxygen (O2) that contain or can produce unpaired electrons. ROS are integral to numerous cellular functions including signaling, proliferation, differentiation, immune responses, and apoptosis. Moderate generation of ROS occurs during normal metabolic processes, e.g. the mitochondrial electron transport chain; however, ROS become damaging at excessive quantities.

The production of ROS must be equally challenged by inactivation via antioxidants or consumption to maintain a normal redox state within the cell. Oxidative stress is the result of a cellular redox imbalance due to high ROS and low or impaired detoxifying agents, which is manifested as lipid peroxidation, DNA mutagenesis, enzyme dysregulation, and protein oxidation/degradation. An oxidized cellular environment consequently disrupts homeostasis leading to altered metabolism, dysregulated signaling, and/or biomolecular damage. Chronic oxidative stress without repair may cause permanent changes in cell function and/or physiology that can incite pathological conditions [Valko et al., 2007; Trouba et al., 2002].

Melanocytes are particularly at risk for chronic oxidative stress due to the oxidation reactions that occur in melanin synthesis resulting in the generation of ROS as byproducts (Figure 2.3) [Denat et al., 2014]. TYR catalytic activity generates superoxide anion (O2-) in the production of

DOPAquinone [Koga et al., 1992]. The eumelanogenesis pathway also generates additional O2- and hydrogen peroxide (H2O2) via TYRP1 and TYR catalytic activity, respectively [Denat et al.,

2014; Nappi & Vass, 1996]. Furthermore, pheomelanogenesis involves the consumption of the antioxidant GSH. Another cellular redox burden is the reactive and cytotoxic nature of the chemical intermediates and melanin polymers produced in melanogenesis. Metabolites such as

26 5,6-dihydroxyindole (DHI), 5,6-dihydroxyindole-2-carboxylic acid (DHICA), quinones, and indole-quinones are documented to cause cellular damage through oxidative mechanisms

[Suzukawa et al., 2012; Urabe et al., 1994; Graham et al.,1978; Pawelek & Lerner, 1978]. Melanin polymers possess both antioxidant and pro-oxidant properties via direct ROS scavenging and/or interaction with metal ions that influence their redox status [Meyskens Jr. et al., 2001; Jimbow,

1995]. For example, pseudo-superoxide dismutase activity, a O2- deactivation process, was experimentally demonstrated for isolated melanins [Bustamante et al., 1993]. Contrariwise, melanin polymer autoxidation or interaction with oxidants, iron (Fe), or copper (Cu) can lead to redox cycling and/or ROS formation via Fenton chemistry [Novellino et al., 1999; Nappi & Vass,

1996; Sotomatsu et al., 1994]. The Fenton reaction results in the production of hydroxyl radicals

2+ 2+ 3+ - (•OH) from ferrous iron (Fe ) and H2O2 substrates (Fe + H2O2 → Fe + •OH + OH ). Superoxide

- anion participates in the Haber-Weiss reaction (O2- + H2O2 → O2 + •OH + OH ), which combines

3+ 3+ 2+ the Fenton reaction and a ferric iron (Fe ) reduction (Fe + O2- → Fe + O2) [Valko et al., 2007;

Matés et al., 1999]. Lipid peroxidation was accelerated by synthetic melanins in the presence of

Fe ions, postulated as the result of •OH produced from Fenton reactions with melanin derived O2-

[Sotomatsu et al., 1994]. Fenton chemistry is also known to contribute to DNA damage produced under oxidative stress [Imlay, 2013; Park & Imlay, 2003; Luo et al., 1996]. To conclude, the melanosomal confinement of melanin synthesis is understood to protect the cellular environment from direct exposure to the ROS and cytotoxic intermediates [Chen et al., 2009b]. However, leakage of these substances through the melanosome membrane into the cytoplasm, nucleus, and mitochondria induces deleterious cytotoxic effects that are referred to as endogenous melanogenic cytotoxicity (EMC) [Chen et al., 2009a; Chen et al., 2009b].

27 Melanocytes exhibit elevated oxidation compared to the other skin cell types keratinocytes and fibroblasts. In vitro experiments confirmed cultured melanocytes maintained significantly higher basal levels of ROS compared to other skin cell types in conjunction with reduced total cellular antioxidant activity [Jenkins & Grossman, 2013; Yohn et al., 1991]. Pelle et al. [2014] presented melanocyte cultures with increased 8 oxo-2’-deoxyguanosine (8-oxo-dG) DNA lesions and decreased levels of GSH compared to keratinocyte cultures; concluding that the increased ferritin also found in the melanocytes caused higher levels of oxidation through Fenton chemistry.

Melanocytes were determined to be more sensitive to the depletion of p16, a cell cycle and oxidative stress protein regulator, when compared to keratinocytes and fibroblasts [Jenkins &

Grossman, 2013, Jenkins et al., 2011]. Interestingly, the inhibition of pigment synthesis via phenylthiourea (PTU) administration significantly reduced endogenous ROS in melanocytes

[Jenkins & Grossman, 2013]. Other investigators report treatments with or modulation of antioxidants showing a regulatory effect on melanin output, further validating the inherent connection between pigmentation and the redox state of the cell [Shin et al., 2014; Nakajima et al., 2013; Choi et al., 2010].

28 2.5 Melanocyte Antioxidant Defense

Antioxidants, also known as oxidation inhibitors, are defined as compounds that prevent or delay the oxidation of other molecules by impeding the initiation or propagation of oxidizing chain reactions [Panda, 2012]. Their superior molecular stability allows for the personal donation of electrons to neutralize reactive compounds and free radicals with unpaired electrons. Redox balance is achieved through an integrated antioxidant protection system comprised of various factors working in conjunction, which include endogenous enzymatic and nonenzymatic antioxidants, exogenous nutrient derived antioxidants, and metal binding proteins [Baird &

Dinkova-Kostova, 2011; Percival, 1998].

2.5.1 Enzymatic Antioxidant Defense

Enzymatic antioxidants work by decomposing and removing free radicals themselves or regenerating electron donor molecules through catalytic mechanisms. The antioxidant enzymes often convert highly reactive oxidative products to H2O2 and then to water in multi-step processes within the presence of cofactors such as copper (Cu), zinc (Zn), manganese (Mn), and iron (Fe); also co-functioning in metal sequestration [Nimse & Pal, 2015]. Of particular importance are the enzymes that have a neutralizing role for H2O2, O2-, and reactive quinones, the ROS generated during melanin synthesis, e.g. listed in Table 2.1.

Superoxide dismutase (SOD) is the singular enzymatic defense that targets O2- exclusively [Imlay,

2008]. A copper-zinc binding isoform (Cu/Zn-SOD/SOD1) is present throughout the cytosol which inactivates O2- and converts it to the less reactive H2O2 using a ping pong type successive

29 - + SOD oxidation and reduction of the transition metal ion at the active site (2 O2 + 2 H → O2 + H2O2)

[Matés et al., 1999]. This defense strategy relies on other abundant peroxidase enzymes to degrade this reaction product [Trouba et al., 2002].

Table 2.1 – Melanocyte antioxidant enzymes Name Enzyme Redox Function Superoxide dismutase SOD Superoxide anion disproportionation Catalase CAT Hydrogen peroxide degradation Glutathione peroxidase GPX Hydrogen peroxide degradation Glutathione reductase GSR Reduced glutathione regeneration Heme oxygenase HO Biliverdin synthesis Biliverdin reductase BLVR Bilirubin synthesis Peroxiredoxin PRDX Hydrogen peroxide degradation Thioredoxin reductase TXNRD Reduced thioredoxin regeneration γ-Glutamate cysteine ligase GCL C/M Glutathione synthesis Glutathione synthetase GSS Glutathione synthesis NAD(P)H quinone oxidoreductase NQO Quinone detoxification Glucose-6-dehydrogenase G6PD NADPH production Isocitrate dehydrogenase IDH NADPH production

Catalase (CAT) is a tetrameric enzyme, primarily localized in peroxisomes, that effectively

CAT degrades H2O2 to water and molecular oxygen (2 H2O2 → O2 + 2 H2O). CAT levels are low in melanocytes compared to the other epidermal cell types, which is considered a contributing factor in the heightened sensitivity to ROS [Schallreuter et al., 2008; Yohn et al., 1991]. Activity levels are strongly reduced after exposure to UVR, which is likely due to oxidative damage of the enzyme

[Steenvoorden & Beijersbergen van Henegouwen, 1997].

30 Glutathione peroxidase (GPX) is a selenium (Se) dependent enzyme also active against H2O2.

GPX GPX utilizes the thiol GSH as an electron donor to degrade H2O2 to water (2 GSH + H2O2 →

GSSG + 2 H2O). Melanocytes express cytosolic, mitochondrial, and extracellular GPX isoenzymes. The oxidized GSH becomes glutathione disulfide (GSSG) and is regenerated back to

GSH by glutathione reductase (GSR); GSR, also present in the cytosol, consumes NADPH for electron donation to reduce GSSG (GSSG + NADPH GSR→ 2 GSH + NADP+). This systematic metabolism of GSH is one of the most essential antioxidant defense mechanisms [Matés et al.,

1999]. The GSH regenerative capacity is linked to the redox state; and the GSH/GSSG ratio is accepted as a good measure for oxidative stress in an organism [Valko et al., 2007; Matés et al.,

1999].

Heme oxygenase (HO) has indirect antioxidant action that can aid in the defense against both O2- and H2O2. HO catalyzes the degradation of heme into ferrous iron, carbon monoxide (CO), water,

HO 2+ and biliverdin (Heme b + 3 AH2 + 3 O2 → Biliverdin + Fe + CO + 3 A + 3 H2O). The newly synthesized biliverdin is then rapidly converted to bilirubin by biliverdin reductase (BVR) at the expense of NADPH (Biliverdin + NADPH BVR→ Bilirubin + NADP+). Bilirubin, a potent antioxidant, scavenges ROS primarily in lipophilic locations such as bilayer membranes [Briganti

& Picardo, 2003; Wei et al., 2003,]. Melanocytes express two isoenzymes for both HO (HO1,

HO2) and BVR (BLVRA/BLVRB). HO1 is highly inducible by the cellular presence of excessive free radicals and some chemical agents, whereas HO2 is constitutive [Elassiuty et al., 2011; Jian et al., 2011]. Consequently, HO1 mRNA is a useful biomarker for cellular oxidative stress at the

RNA level [Valko et al., 2007]. Both HO and BVR are cytoplasmic proteins with nuclear translocation ability [Tudor et al., 2008; Lin et al., 2007].

31 Peroxiredoxin (PRDX) is an antioxidant enzyme that reduces H2O2 and alkyl hydroperoxides. Six isoenzymes are expressed in melanocytes throughout the intracellular and extracellular space.

PRDX uses thiols, primarily thioredoxin (TXN), as electron donors in their catalytic action

PRDX (2 R'-SH + ROOH → R'-S-S-R' + H2O + ROH). Another enzyme, thioredoxin reductase

(TXNRD), regenerates the oxidized thioredoxin disulfide (TXN-S2) back to thioredoxin (TXN-

TXNRD + (SH)2) using NADPH (TXN-S2 + NADPH → TXN-(SH)2 + NADP ). Melanocytes express the cytoplasmic and mitochondrial isoenzymes (TXNRD1, TXNRD2). This system plays essential roles in a variety of cellular functions including redox control of transcription factors, deoxyribonucleotide synthesis, and cell growth [Karlenius & Tonissen, 2010]. TXNRD is also capable of regenerating other antioxidant compounds such as ascorbic acid, selenium-containing substances, lipoic acid, and ubiquinone; and additionally supports α-tocopherol function

[Nordberg & Arnér, 2001].

γ-Glutamate cysteine ligase (GCLC, catalytic subunit/ GCLM, modifier subunit) is the first rate- limiting enzyme involved in GSH biosynthesis. The GCL complex catalyzes the production of gamma-L-glutamyl-L-cysteine, which becomes the substrate for a second enzyme, glutathione

GCLC/M 3 synthetase (GSS) (ATP + glutamate + cysteine → ADP + PO4 + gamma-glutamylcysteine).

GSS catalytic activity (ATP + gamma-glutamylcysteine + glycine GSS→ ADP + phosphate + GSH) yields the final GSH product. GSH is the most abundant thiol in human tissues and has several protective roles in addition to its cofactor function for detoxifying enzymes [Steenvoorden &

Beijersbergen van Henegouwen, 1997]. These include hydroxyl radical scavenging, singlet oxygen scavenging, nonenzymatic antioxidant regeneration, and amino acid transport through the plasma membrane [Valko et al., 2006].

32 NAD(P)H quinone oxidoreductase 1 (NQO1) is a widely distributed flavoprotein that catalyzes single step 2-electron reductions of quinones, quinoneimines, nitroaromatics, and azo dyes using

NADPH via a ping pong mechanism (NADPH + a quinone NQO1→ NADP+ + a hydroquinone).

This reaction bypasses the toxic semiquinone radical intermediate to produce the less reactive hydroquinone species and can yield substrates for phase II conjugation reactions that promote excretion [Ross et al., 2000]. Melanocytes express both NQO1 and NQO2 isoenzymes. NQO1 is primarily localized in the cytosol, existing as a homodimer with one molecule of flavin adenine dinucleotide (FAD), a redox cofactor, per monomer; NQO2 is found in the nucleus and extracellular space. NQO1 additionally generates antioxidant forms of ubiquinone and vitamin E and is proposed to have a role in O2- scavenging [Siegel et al., 2004; Ross et al., 2000].

Glucose-6-dehydrogenase (G6PD) and isocitrate dehydrogenase (IDH) are two enzymes that produce the electron donor NADPH in their catalytic activity. The primarily cytosolic G6PD catalyzes the rate-limiting step in the oxidative pentose-pathway using the substrates glucose-6- phosphate and NADP+, this being an alternative route to glycolysis for carbohydrate dissimilation

(Glucose 6-phosphate + NADP+ G6PD→ 6-phospho-D-glucono-1,5-lactone + NADPH).

Melanocytes express three IDH isoenzymes including a cytosolic protein that uses isocitrate and

+ NADP to produce NADPH, carbon dioxide (CO2), and 2-oxoglutarate for metabolic processes

+ IDH (Isocitrate + NADP → 2-oxoglutarate + CO2 + NADPH). Enzymatic activity of G6PD and

IDH was demonstrated to be greater in the epidermis compared to the dermis by 111% and 313%, respectively [Shindo et al., 1994]. Expression of both G6PD and IDH was also found in peroxisomes, suggesting a role for NADPH regeneration within the organelle [Frederiks et al.,

33 2007; Geisbrecht et al., 1999]. Figure 2.6 exhibits the key redox functions of these enzymes specifically in melanocytes.

Figure 2.6 – Melanocyte enzymatic antioxidant defense against reactive quinones and oxygen - species. During melanin synthesis, quinones and reactive oxygen species O2 and H2O2 are generated and may escape from the protective confine of the melanosome into the cytosol. Melanocytes express an assortment of cytosolic and organelle specific antioxidant enzymes to - neutralize reactive quinones, O2 , and H2O2. O2 = molecular oxygen, H2O = water. Antioxidant enzymes: SOD = superoxide dismutase; HO = heme oxygenase; BVR = biliverdin reductase; NQO = NAD(P)H quinone oxidoreductase; GCLC/M = γ-glutamate-cysteine ligase, catalytic/modifier; GSS = glutathione synthetase; GSR = glutathione reductase; GPX = glutathione peroxidase; CAT = catalase; PRDX = peroxiredoxin; TXNRD = thioredoxin reductase; G6PD = glucose-6- dehydrogenase; IDH = isocitrate dehydrogenase. Thiols: GSH = glutathione; GSSG = glutathione disulfide; TXN-(SH)2 = thioredoxin; TXN-S2 = thioredoxin disulfide.

34 2.5.2 Nonenzymatic Antioxidant Defense

Nonenzymatic antioxidants function analogously, yet are dispersed in various locations for different ROS targets throughout the cellular environment. These low molecular weight molecules are the electron donors themselves that can disrupt free radical chain reactions. Regeneration by antioxidant enzymes or other proteins is required for continued effectivity. Bilirubin, ubiquinone, nicotinamide adenine dinucleotide phosphate (NADPH), and thiols GSH and TXN-S2 are endogenous nonenzymatic antioxidants/cofactors that work in conjunction with the antioxidant enzymes mentioned previously in section 2.5.1. Exogenous antioxidants include vitamins C, and

E; carotenoids like beta carotene, lycopene, and lutein; and polyphenols such as flavonoids, flavones, flavanols, and proanthocyanidins. These antioxidants cannot be generated through transcription and are consumed through the diet or medicinal administration [Birben et al., 2012;

Panda, 2012; Percival, 1998]. There are also metal binding proteins such as albumin, metallothionein, ferritin, and transferrin that capture reactive metals to prevent redox cycling. As previously mentioned, melanin chromophores can also donate electrons contingent on the redox environment.

2.5.3 Nrf2-ARE Pathway in Skin

The transcription factor nuclear factor, erythroid 2-like 2 (Nrf2) is considered a central regulator for cell protection and survival through activation of the antioxidant response element (ARE) pathway. This pathway is stimulated by changes in the redox state of the cell and functions to restore homeostasis by upregulating antioxidant, xenobiotic metabolizing, and other cytoprotective proteins and enzymes constituting over 600 gene targets [Espinosa-Diez et al. 2015;

35 Baird & Dinkova-Kostova, 2011]. Nrf2 is constitutively expressed for a mild activation response; and in basal conditions, excess Nrf2 remains in the cytosol sequestered to the protein inhibitor

Kelch-like ECH-associated protein 1 (Keap1) (Figure 2.7). Keap1 then assembles with Cullin 3

(Cul3) forming a complex that mediates polyubiquitination of Nrf2 for subsequent proteasomal degradation. In stress conditions, modifications of cysteine residues on Keap1 activate a conformational change that allows dissociation from Nrf2. Nrf2 is then available for activation through phosphorylation on serine and threonine residues by kinases. Active Nrf2 translocates into the nucleus and heterodimerizes with small Maf protein. This complex binds to ARE promoters resulting in gene transcription of cytoprotective proteins such as antioxidant enzymes HO, NQO,

PRDX, TRXRD, GSR, GCLC/M, etc. Post induction, Nrf2 is exported from the nucleus and rejoined with the Keap1/Cul3 complex that leads to proteasomal degradation.

Scientific literature is progressively uncovering a substantial role for Nrf2 in the metabolism of skin cells. Marrot et al. [2007] demonstrated Nrf2-ARE pathway induction with chemical treatments, UVR exposure, and transfection of Keap1 silencing RNA (siRNA) resulting in the stimulatory gene expression of antioxidant enzymes GCLC/M, HO1, and NQO1 in both keratinocyte and melanocyte cell cultures. Jian et al. [2011] treated melanocyte cultures with H2O2 causing an increase in their expression of HO1; this upregulation also prevented H2O2 induced cell death indicating a protective effect. In this study, HO1 exhibited the most inducible behavior after

H2O2 treatment compared to NQO1, GCLC, and GCLM [Jian et al., 2011]. Both 4-tertiary butylphenol (4TBP), the cytotoxic compound, and UVR exposure were confirmed to upregulate

HO1 in cultures of melanocytes and organotypic human skin, respectively [Elassiuty et al., 2011].

In contrast, Kokot et al. [2009] demonstrated reduced expression of Nrf2 and three downstream

36 targets post ultraviolet B (UVB) irradiation in keratinocytes, melanocytes, and ex vivo skin cultures. However, co-treatment of α-MSH and UVB resulted in increased mRNA of Nrf2, HO1,

GCL, and glutathione-S-transferase π (GSTπ) – another xenobiotic detoxifying enzyme [Kokot et al., 2009]. In conclusion, Nrf2 signaling and the expression of ARE targets is now recognized to regulate skin homeostasis under stress conditions.

37

Figure 2.7 – Nrf2-ARE pathway activation and deactivation [Modified from Surh, 2003]. The transcription factor nuclear factor, erythroid 2-like 2 (NRF2) is primarily localized in the cytosol and repressed by a cytosolic inhibitor Kelch-like ECH-associated protein 1 (KEAP1). Keap1 contains numerous cysteine residues (H→R) that are covalently modified by oxidants or chemical inducers resulting in a conformational change that causes the release of Nrf2. Nrf2 is then accessible for phosphorylation (P) on serine (S) and threonine (T) residues by various kinases. This initiates translocation in to the nucleus, where Nrf2 binds with small MAF in heterodimer form on antioxidant response element (ARE) promoters, thus stimulating gene expression of antioxidant and phase II enzymes. Post induction, Nrf2 is exported from the nucleus and recruited into a complex with Keap1 and Cullin 3 (CUL3) leading to ubiquitin (Ub) tagging and proteasomal degradation. Kinases: p38 = p38 mitogen activated protein kinase; ERK = extracellular signal regulated kinase; JNK = c-Jun N-terminal kinase; PKC = protein kinase C; PI3K = phosphatidylinositol 3 kinase. Antioxidants/Phase II Enzymes: HO = heme oxygenase; NQO = NAD(P)H quinone oxidoreductase; TXNRD = thioredoxin reductase; GCLC = γ-glutamate- cysteine ligase, catalytic; GCLM = γ-glutamate-cysteine ligase, modifier; PRDX = peroxiredoxin.

38 2.6 Oxidative Stress in Vitiligo

The etiology of vitiligo is still unclear; yet, it is generally accepted that several biological components work in tandem to initiate the depigmentation process. Patients with vitiligo generally have genetic characteristics that render their melanocytes highly vulnerable to oxidative stress, which leads to immune targeting and apoptosis [Koshoffer & Boissy, 2014]. The immune system is concomitantly sensitive; thus, stressors such as UVR, environmental pollutants, physiological signals, hormones, etc., can provoke onset. Lesional, depigmented skin patches develop as dysfunctional or dying melanocytes amass in quantity.

Most research has established a role for redox imbalance in the pathogenesis of vitiligo [Colucci et al., 2015b; Koshoffer & Boissy, 2014; Laddha et al., 2013b]. Systemic increases in oxidant status with simultaneous decreases in antioxidant status were observed in both localized and general vitiligo patients compared to healthy controls [Akoglu et al., 2013]. Several antioxidant enzymes exhibited diminished expression or activity levels in vitiligo patients including CAT,

GPX, G6PD, and vitamins C and E [Dammak et al., 2009; Khan et al., 2009; Sravani et al., 2009;

Arican & Kurutas, 2008; Jain et al., 2008; Schallreuter et al., 1991]. Reports of increased activity levels of SOD from patients suggest there is a low SOD/CAT ratio, an indicator for cellular oxidative stress due to H2O2 accumulation [Laddha et al., 2013a; Dammak et al., 2009; Sravani et al., 2009, Arican & Kurutas, 2008]. Yet, opposing reports of reduced SOD activity exist [Khan et al., 2009; Koca et al., 2004]. Upregulation of ROS induced oxidative stress via H2O2 and peroxynitrite (ONOO−) has been documented [Akoglu et al., 2013; Salem et al., 2009; Schallreuter et al., 2007].

39 Various scientific studies associate vitiligo with multiple classical manifestations of oxidative stress such as lipid peroxidation, impaired signaling, structural irregularities, and dysfunction of cellular organelles. Patients have exhibited significantly higher lipid peroxidation levels compared to healthy controls in several studies, which were highest in active vitiligo compared to stable vitiligo [Laddha et al., 2014; Laddha et al., 2013a; Dammak et al., 2009; Khan et al., 2009; Jain et al., 2008; Koca et al., 2004]. Dell’Anna et al. [2001] detected mitochondrial alterations in active vitiligo patients correlating to increased ROS generation, imbalanced antioxidant system, and modified transmembrane potential in peripheral blood mononuclear cell samples. Boissy et al.

[1991] reported persistent dilation and abnormal formation of the rough endoplasmic reticulum in cultured melanocytes from vitiligo patients. Similar observations occurred in the immortalized vitiligo melanocyte cell line PIG3V [Le Poole et al., 2000]. Vacuolation was observed in melanocyte cultures established from both lesional and nonlesional epidermis of vitiligo patients

[Schallreuter et al., 1999].

Intriguingly, recent studies indicate impaired nuclear factor, erythroid 2-like 2 – antioxidant response element (Nrf2-ARE) pathway signaling in vitiligo pathophysiology. Amin et al. [2013] confirmed a reduction of both Nrf2 and α-MSH mRNA in the lesional and perilesional skin of vitiligo patients. An allelic variant of Nrf2, A-650, was determined to be significantly associated with vitiligo patients opposed to healthy controls of Han Chinese in a 600-participant study [Guan et al., 2008]. Jian et al. [2014] demonstrated hypersensitivity to H2O2 induced oxidative stress due to decreased Nrf2-ARE signaling in vitiligo melanocyte cultures; cells had reduced Nrf2 nuclear translocation and transcriptional activity that aligned with their diminished heme oxygenase 1

(HO1) expression. Likewise, immortalized PIG3V vitiligo melanocytes showed decreased ARE

40 luciferase activity in contrast to normal PIG1 melanocytes; notably, the overexpression of Nrf2 via plasmid transfection in PIG3V cells restored their viability even against H2O2 assault [Jian et al., 2014]. Jian et al. [2014] also reported low HO1 and high serum levels of the immune cytokine interleukin-2 (IL-2) in both stable and progressive vitiligo patients signifying an inversely proportional relationship (r = -0.4482). This group postulated that the generation of carbon monoxide (CO) byproduct from HO1 catalytic activity inhibits IL-2 secretion thereby suppressing

T cell proliferation [Jian et al., 2014; Pae et al., 2004]. Dampened Nrf2-ARE may account for IL-

2 activation in vitiligo patients, which is speculated as a mechanistic link between oxidative stress and immune activation in vitiligo [Jian et al., 2014; Qiu et al., 2014].

Experimentation on selective induction of Nrf2-ARE signaling is a strategy of interest for many oxidative driven diseases including vitiligo. Treatment with curcumin and santalol, two electrophilic Nrf2 inducers, caused a substantial increase in phase II enzymes synthesis in cultured melanocytes, yet simultaneously resulted in increased apoptosis of keratinocytes at the same concentrations [Natarajan et al., 2010]. Whole skin compatibility is of clinical importance. Non- lesional vitiligo skin biopsies did show upregulation of Nrf2 and phase II enzymes after incubation in curcumin and santalol, but lesional skin did not, illustrating the need for continued discovery of working compounds in this area [Natarajan et al., 2010]. Bear in mind, all inducers do not function the same way, as there are several working models for modulation of Keap1 and Nrf2; hence, newly discovered compounds will require mechanistic analysis [Baird & Dinkova-Kostova, 2011].

Restoration of redox balance has therapeutic value. In the case of vitiligo, an assortment of antioxidants orally administered or adjuvants to phototherapy are recommended [Boissy et al.,

2012]. Both patients and physicians show interest in using naturally derived, complementary

41 treatments such as herbal products or vitamin supplements [Cohen et al., 2015]. Traditional pharmacotherapy is linked with severe side effects and additional oxidative stress; thus, it is valuable to seek compounds that cause selective activation of Nrf2 especially naturally obtained materials similar to curcumin, sulforaphane, green tea polyphenols, or extracts that may serve as therapeutic or preventative options for oxidative driven skin diseases like vitiligo and melanoma

[Gęgotek & Skrzydlewska, 2015; Boissy et al., 2012]. Nonetheless, the efficacies of such materials are a valid concern that will require rigorous study to generate appreciable outcomes.

Normalizing the cellular redox status and increasing protection from excessive ROS buildup is a prime strategy for preventing the aggression of the oxidative driven diseases like vitiligo. In vitiligo, it is postulated that dysfunctional melanocytes cannot efficiently combat the ROS accumulation from melanin synthesis that manifests in detrimental effects caused by chronic oxidative stress [Koshoffer & Boissy, 2014]. Both oral supplementation and topical application of antioxidant blends have improved the clinical efficiency of UVB phototherapy [Colucci et al.,

2015a; Dell’Anna et al., 2007; Schallreuter et al., 1995]. The vacuolation of vitiligo melanocyte cultures reported by Schallreuter et al. [1999] was also reversible upon the addition of catalase.

Antioxidant treatments and the induction of antioxidant genes show protective effects against peroxidative or chemical exposure in in vitro experimentation [Elassiuty et al., 2011; Jian et al.,

2011]. Antioxidants have been proven to prevent oxidative injury of structural lipids and proteins that contribute to barrier integrity, which is critical for a healthy skin condition. This suggests the cellular redox environment plays a fundamental role in skin homeostasis and skin diseases can result from an imbalance between pro-oxidant and antioxidant stimuli [Bickers & Athar, 2006;

Briganti & Picardo, 2003]. The pathological study of vitiligo unambiguously discloses various

42 connections with oxidative stress; therefore, balancing the cellular redox state is a valid approach to prevent or alleviate symptoms.

43 2.7 Properties of Galactomyces Ferment Filtrate

Galactomyces ferment filtrate (GFF) currently functions as the prominent ingredient in the prestige skincare line SK-II that is used for improvement of skin tone, texture, and the appearance of unwanted hyperpigmentation and age-related wrinkles [Procter & Gamble, Co., 2017]. Research and product development of SK-II has continued for over 30 years; still, the biological working mechanisms of GFF associated with the clinical results are not completely understood. Minimal work has been published that identifies the specific biological effects of GFF on melanogenesis and ROS in the human pigment cell model.

Tsai et al. [2006] showed that GFF co-treatment resulted in the downregulation of inducible nitric oxide synthase (iNOS) expression and concurrent upregulation of HO1 expression in RAW264.7 mouse macrophages induced with lipopolysaccharide (LPS). LPS pre-treated RAW264.7 cells that were subsequently incubated in GFF exhibited reduced nitric oxide (NO) production and iNOS enzymatic activity. Large amounts of NO generated by iNOS are known to be toxic and pro- inflammatory. The oxidant potential of NO also increases in the presence of O2-. In the same study,

GFF also prevented dyskeratinization and hypergranulosis in EpiDerm™ human engineered skin compared to the control sample treated with LPS alone. Additionally, UVB irradiation of

EpiDerm™ resulted in irregular keratinocyte morphology, surface roughness, and cell blebbing that was not observed in EpiDerm™ pre-treated with GFF. Concomitant results were achieved in another study with UVB irradiated human skin equivalents; ROS generation was suppressed in the samples treated with GFF comparable to known antioxidants SOD, β-carotene, and ascorbate

[Hakozaki et al., 2008].

44 Skin enhancing assertions regarding barrier function have also been made with GFF experimental data. GFF treatment upregulated expression of tight junction-related proteins and the transepithelial electrical resistance (TER) in epidermal keratinocytes [Wong et al., 2011]. Caspase-

14 expression and activity, a protein biomarker for cornification, was upregulated by GFF in a study using multiple in vitro human skin models including keratinocytes, skin equivalents, and stratified epidermis [Hattori et al., 2010]. GFF also increased the expression of hyaluronan in the epidermal region of human skin equivalents [Osborne et al., 2007]. Takei et al. [2015] demonstrated the activation of the aryl hydrocarbon receptor (AhR), a xenobiotic chemical response receptor, that resulted in the gene upregulation of cytochrome P450-1A1 (CYP1A1) gene expression in GFF treated normal human epidermal keratinocyte (NHEK) cultures. Nuclear translocation of AhR also induced a signaling cascade resulting in the augmentation of skin barrier proteins filaggrin (FLG) and loricrin (LOR). GFF effectively combated the interleukin-4 (IL-4) and interleukin-13 (IL-13) induced downregulation of FLG and LOR mRNA; thus, suggesting a utility in T helper 2 cytokine (Th2) mediated skin disorders [Takei et al., 2015]. This varied experimentation collectively initiates the scientific validation of GFF as a naturally derived, broad- spectrum skin health agent. Further understanding the isolated properties of GFF in the human pigment cell model may reveal inventive strategies to effectively prevent oxidative driven skin disease and promote melanocyte preservation for healthy skin barrier function.

45

CHAPTER 3:

RATIONALE, OBJECTIVE, HYPOTHESES, AND SPECIFIC AIMS

46 3.1 Rationale and Objective

Epidermal melanocytes are especially susceptible to significant amounts of oxidative stress attributable to their purposed function of melanin synthesis, which involves the generation of

- reactive melanin intermediates, superoxide anion (O2 ), and hydrogen peroxide (H2O2) as well as the depletion of glutathione (GSH) [Denat et al., 2014]. There is extensive evidence establishing a significant role for oxidative stress in the etiology of melanocyte death in vitiligo.

Pitera™/galactomyces ferment filtrate (GFF) exhibits substantial skin barrier enhancing, antioxidant, and hypopigmenting properties. The rationale that underlies this research project is that GFF may have a therapeutic utility that is uniquely applicable for vitiligo patients through diminishing melanization and coupled ROS generation, while concurrently upregulating intracellular antioxidant activity. Vitiligo onset is typically instigated by increased facultative melanization that is commonly triggered by precipitating factors such as ultraviolet radiation

(UVR), hormones, cytokines, trauma, etc. This induced melanization elevates cytotoxicity through increased melanin intermediates and ROS generation that is intolerable by the genetically predisposed vitiligo melanocyte, resulting in immune targeting and cell death [Boissy et al., 2012].

Thus, the GFF properties would be useful in a prevention strategy against early onset vitiligo to potentially impede lesion progression, or as an accompaniment in depigmentation therapy. There is no cure currently available; hence, efficient prevention and remedial strategies that can halt the progression of oxidative stress in epidermal melanocytes are desirable. Other hyperpigmentary disorders such as solar lentigines, melasma, or post-inflammatory hyperpigmentation (PIH) may also benefit from the clinical properties of GFF [Lee, 2015; Bickers & Athar, 2006].

47 Another noteworthy consideration is the common adversities associated with the use and manufacture of synthetic medications. A sector of the scientific community is recommitting to natural products research for pharmacological benefits. The production and manufacture of naturally derived and sustainably sourced treatments continues to rise, being well received by the public. Vitiligo patients in particular also seek remedies involving natural compounds and methodology; yet, regulation, efficacy, and potency are still considerable challenges [Cohen et al.,

2015]. GFF is a filtered, yeast derived extract that exhibits comprehensive skin health benefits that may have the potential to address such impediments.

Therefore, to commence the assessment of GFF as a possible medicant for vitiligo, the foremost objective of this research project is to identify the specific biological mechanisms by which GFF functions as a suppressant of constitutive melanin pigmentation and oxidative stress.

48 3.2 Hypotheses and Specific Aims

The central hypotheses of this research project are that GFF inhibits melanin synthesis by modulating the pH of the melanosomal microenvironment, consequently reducing tyrosine hydroxylase activity; Additionally, GFF nullifies oxidative stress by priming the intracellular environment with increased antioxidant enzymes.

The central hypotheses will be tested by pursuing the following two specific aims:

1. How does GFF reduce the constitutive melanin pigmentation produced in human pigment

cells?

Preliminary data demonstrate that GFF treatment significantly reduces melanization and tyrosine hydroxylase activity in human melanocytes and melanoma. GFF treatment also modifies the mRNA expression of ion transporters including vacuolar-type proton ATPases (V-ATPases) and

H+, Na+, K+, Ca2+, Cl- specific pumps, carriers, and channels known to affect both intracellular and organelle pH. The working hypothesis based on preliminary data is that GFF significantly alters the pH environment of the melanosome counter to optimal tyrosine hydroxylase activity. The working hypothesis will be tested by:

a. Evaluating the expression levels of ion transporter proteins in human pigment cells treated

with GFF.

b. Detecting the location of ion transporter proteins on the melanosomal membrane in human

pigment cells.

c. Determining the pH of melanosomes in human pigment cells treated with GFF.

49 2. How does GFF suppress ROS generation and promote cellular redox balance?

Preliminary data demonstrate that GFF suppresses the generation of ROS in human melanocytes challenged with 4-tertiary butylphenol (4TBP). GFF treatment also modifies the mRNA expression of antioxidant enzymes heme oxygenase 1 (HO1), NAD(P)H quinone oxidoreductase

1 (NQO1), and thioredoxin reductase 1 (TXNRD1), downstream targets of the antioxidant response element (ARE) pathway. The working hypothesis based on preliminary data is that GFF induces nuclear translocation of the transcription factor nuclear factor, erythroid 2-like 2 (Nrf2) to activate the Nrf2-ARE pathway; thereby increasing endogenous antioxidant quantity and activity levels in the intracellular environment. The working hypothesis will be tested by:

a. Determining the expression and/or activity levels of enzymatic antioxidant targets from the

Nrf2-ARE pathway in human melanocytes treated with GFF.

b. Evaluating nuclear translocation of Nrf2 in human melanocytes treated with GFF.

c. Evaluating the impact of inducing ROS generation on the Nrf2-ARE pathway in human

melanocytes treated with GFF.

50 CHAPTER 4:

GALACTOMYCES FERMENT FILTRATE SUPPRESSES MELANIZATION IN HUMAN

PIGMENT CELLS

JàNay K. W. Cooper1, Amy Koshoffer2, Toral Vaidya2, Ana Luisa Kadekaro2, Tomohiro

Hakozaki3, and Raymond E. Boissy2

1James L. Winkle College of Pharmacy, University of Cincinnati, Cincinnati, Ohio, USA;

2Department of Dermatology, University of Cincinnati College of Medicine, Cincinnati, Ohio,

USA; 3Procter & Gamble, Co., Cincinnati, Ohio, USA

Presently being edited for submission to Journal of Investigative Dermatology.

Modified for dissertation purposes.

51 4.1 Abstract

Background/Objective: Galactomyces ferment filtrate (GFF, Pitera™) is a yeast derived extract currently used as a moisturizing agent in cosmetics. GFF demonstrates a variety of skin health benefits and is reported to have hypopigmenting capacity. The mechanisms of action underlying

GFF are relatively unknown and therefore are the focus of our study.

Methods: In vitro human pigment cell models, including foreskin derived NHEM and SK-MEL-

188, were treated with GFF 0-10%. Melanin content was quantified by spectrophotometric assay.

NHEM dendrites were evaluated by transmission electron microscopy. Tyrosine hydroxylase activity was determined with tritiated tyrosine and assayed by liquid scintillation. RNA-seq transcriptome profiling was completed with mRNA extracted from NHEM. Protein expression was quantitated by Western blot and densitometric analyses. Expression of ion transporters and their colocalization to melanogenic proteins were assessed by immunofluorescent microscopy.

Results: GFF suppressed pigmentation in NHEM and SK-MEL-188 and reduced tyrosine hydroxylase activity in NHEM both in situ and in vitro. GFF did not significantly alter the mRNA nor protein levels of the primary melanogenic proteins required for melanin synthesis (MITF,

PMEL, TYR, TYRP1, TYRP2/DCT). Notably, GFF did modify the mRNA expression of ion channels and transporters ABCC9, TRPM6, TRPV1, SLC9A4, and SLC13A1 that are associated with other known pore proteins documented to affect intracellular and/or organelle pH. TRPM6,

TRPV1, and SLC9A4 exhibited considerable colocalization to melanogenic proteins.

52 Conclusions: GFF inhibits melanin synthesis by deterring tyrosine hydroxylase activity. The new working hypothesis is that GFF significantly alters the pH of melanosomes counter to optimal tyrosine hydroxylase activity by downregulating ion channel and transporter proteins.

4.1.1. Keywords: Galactomyces ferment filtrate, pigmentation, melanosome pH, SLC9A4/NHE4,

TRPM6, TRPV1/VR1

53 4.2 Introduction

On trend cosmetic skincare provides hyperpigmentation reduction and tone correction.

Galactomyces ferment filtrate (GFF) is a yeast derived extract currently used as a moisturizing agent in cosmetics. GFF demonstrates a variety of skin health benefits and is reported to have hypopigmenting capacity [Takei et al., 2015; Wong et al., 2011; Hattori et al., 2010; Hakozaki et al., 2008; Tsai et al. 2006]. Yet, the mechanisms of action underlying GFF have not been fully characterized in regard to melanization and therefore are the focus of our studies.

Cutaneous melanocytes are a modest cell population located in the stratum basale at the epidermal- dermal junction in skin and hair follicles that is charged with the principle role of photoprotection via melanin synthesis and transfer. Melanin is constructed and retained in specialized organelles, referred to as melanosomes, which house the series of oxidation reactions necessary to produce melanin by melanizing proteins and enzymes: premelanosome protein (PMEL), tyrosinase (TYR), and tyrosinase related proteins -1 and -2 (TYRP1, TYRP2). The expression of these melanogenic genes is primarily regulated by microphthalmia-associated transcription factor (MITF) via multiple signaling cascades, including precedent activation of the melanocortin 1 receptor (MC1R) located on the cell membrane.

Regulation of melanization is of high clinical interest due to a variety of dyspigmentary diseases and cosmetic issues in skin. Melanogenesis is affected by a vast number of gene regulatory mechanisms and signaling molecules including cytokines, hormones, and other inherent biological factors. One of growing importance is the pH within the melanosome. Numerous scientific reports endorse an effectual control of melanization output and TYR activity through alteration of the pH

54 within the melanosomal microenvironment [Ambrosio et al., 2016; Bellono & Oancea, 2014; Cheli et al., 2009; Ancans et al. 2001]. pH is regulated by ion transport proteins on the cell and melanosome membrane. Ion transporters including vacuolar-type proton ATPases (V-ATPases), solute carriers (SLCs), and H+, Na+, K+, Ca2+, Cl- specific pumps and channels are known to affect organelle pH [Bellono et al., 2013; Dooley et al., 2013; Mindell 2012; DiCiccio & Steinberg 2011;

Devi et al., 2009;]. There are conflicting accounts on the ideal pH for melanin synthesis and optimal TYR activity; yet, these organelles have a endolysosomal lineage that suggests their acidification mechanisms are comparable and may operate under similar principles. Nonetheless, a substantial collection of scientific literature presents evidence that pH of the melanosome is an appreciable factor in influencing melanin output and there is a need for continued study in human pigment cell models to elucidate this action.

Our aim is to understand how GFF effects the melanization of human pigment cells. In this study we investigated the biological effects of GFF in human melanoma and primary human melanocytes in vitro. We demonstrate that GFF significantly suppresses constitutive melanization in these cell types, inhibits tyrosine hydroxylase activity, and decreases the expression of ion transporters within the melanocyte and on melanosomes that may potentially alter the pH environment of the melanosome counter to optimal tyrosine hydroxylase activity. Understanding the function of GFF in melanin suppression is clinically relevant and may permit a therapeutic utility for dyspigmentation disorders such as melasma, post-inflammatory hyperpigmentation (PIH), solar lentigines, or vitiligo.

55 4.3 Methods and Materials

Cell culture: Primary cultures of normal human epidermal melanocytes (NHEM) were established from discarded neonatal foreskin obtained from different skin types as previously described by

Chalupa et al. [2015]. Tissues were procured from University Hospital in Cincinnati or from the

Christ Hospital in Cincinnati. Patient consent was not required for experimentation because of

USA laws regarding left over human tissues from surgery. Briefly, foreskins were washed with betadine and phosphate buffered saline (PBS), sectioned and incubated in 0.25% trypsin, then rocked at 4oC overnight. Tissues were vortexed and centrifuged to separate the epidermis and dermis. The epidermal fraction was plated in T25 or T75 flasks with melanocyte growth medium.

Melanocyte growth medium consisted of MCDB-153 (Sigma-Aldrich) supplemented with 4% fetal bovine serum (FBS), 1% antibiotic/antimycotic, 15 µg mL-1 bovine pituitary extract (BPE),

5 µg mL-1 insulin, 8 nM 12-O-tetradecanoyl-phorbol-13-acetate (TPA), and 0.6 ng mL-1 basic fibroblast growth factor (bFGF) to aid in cell proliferation, dendricity, and melanization. Human melanoma (SK-MEL-188) cells were a gift from Dr. Alan Houghton, Sloan Kettering Institute,

New York. Cells were removed from liquid nitrogen storage and plated in T150 flasks with melanoma growth medium. Melanoma growth media consisted of DMEM (Invitrogen) supplemented with 5% FBS and 1% antibiotic/antimycotic. Cultures were maintained in a

o humidified incubator with 5% CO2 at 37 C. Growth medium was changed every three to four days.

GFF treatment: GFF was prepared using two different filtration methods: a routine mechanism

GFF-A, and an enhanced mechanism GFF-B. Filtrates were prepared at Procter & Gamble

Innovation Godo Kaisha, Kobe Technical Center; shipped to the laboratory on dry ice; and stored at 4oC. NHEM cultures were plated and incubated in experimental growth medium consisting of

56 MCDB-153 supplemented with 6% FBS, 1% antibiotic/antimycotic, 15 µg mL-1 BPE, 5 µg mL-1 insulin, 2 nM TPA and 0.15 ng mL-1 bFGF for 48 hours before the start of experiments. SK-MEL-

188 cultures were plated and incubated in the abovementioned melanoma growth medium for 48 hours before the start of the experiment. During experimentation, pigment cell cultures were treated with GFF-A or GFF-B, 0-10% added to the cell experimental growth medium every 48

o hours for 5-22 days. Cultures were maintained in a humidified incubator with 5% CO2 at 37 C until experiment harvest.

Melanin content assay: Melanin content was quantified as previously described in Chalupa et al.

[2015]. Cell pellets were separated from protein lysate using radioimmunoprecipitation assay

(RIPA) buffer in the presence of protease and phosphatase inhibitors and centrifugation (10,000

RPM). Protein supernatant was removed and set aside. The remaining melanin pellet was washed with 1:1 ethanol/ether and dissolved in 2 N sodium hydroxide at 60oC. 75 µL aliquots were spectrophotometrically read at 490 nm absorbance with a synthetic melanin (Sigma-Aldrich) standard in microplate reader (Bio-Rad, 550). Results were calculated against the melanin standard curve and normalized to sample protein content, separately analyzed. Results are reported as ng of melanin/µg of protein and as a percentage of the control. (n=3)

Protein determination: Protein was extracted from cells using RIPA buffer supplemented with protease and phosphatase inhibitors. Cells were centrifuged at 10,000 RPM for 10 min at 4oC. The protein supernatant/lysate was separated from the cell pellet and placed on ice. 2-10 µL aliquots were prepared with Pierce® BCA Protein Assay Kit (Thermo Scientific). The colorimetric assay was spectrophotometrically read at 570 nm absorbance with bovine serum albumin (BSA) standard

57 in microplate reader (Bio-Rad, 550). Protein content results were calculated against BSA standard curve and reported as µg of protein. (n=3)

Cell number: Cells were detached with 0.05% trypsin ethylenediamine tetra acetic acid (EDTA) and 500-1000 µL aliquots were diluted in 10 mL isotonic solution and counted with a Coulter counter (Beckman Coulter, Z1 Single). Results are reported as a percentage of the control. (n=3)

Electron microscopy: NHEM were treated with or without 5% and 10% concentrations of GFF-

B for 20 days. Upon harvest, transmission electron microscopy (TEM) was perform as described previously by Boissy et al., [1991]. Briefly, cells were cultured in Tissue-Tek chamber slides

(Nunc) coated with 1% pig. Cells were fixed with Karnovsky’s fixative at half strength in 0.2 M sodium cacodylate buffer at pH 7.2 for 30 minutes. Fixed cells were then washed and treated with

1.0% osmium peroxide with 1.5% potassium ferrocyanide for 30 minutes. These samples were then washed with 0.5% uranyl acetate, dehydrated, and embedded in Eponate 12. Samples were then sectioned with a RMC MT 6000-XL ultramicrotome, stained with 2% uranyl acetate and

0.3% lead citrate for 15 minutes, and photographed with a JEOL JEM-1000CX (Jeol, Ltd.) transmission electron microscope. (n=6)

Tyrosine hydroxylase activity assay: Tyrosine hydroxylase activity was determined with tritiated tyrosine (3H-tyrosine, 54.2 Ci/mmol specific activity) and assayed by liquid scintillation in similar methods previously described by Zhao & Boissy [1994]. For the in situ assay, NHEM were pretreated with GFF for 6 days and then incubated with 0.7 µCi mL-1 of 3H-tyrosine and a final

o GFF dose for an additional 24 hours in a humidified incubator with 5% CO2 at 37 C. Experimental

58 growth medium was collected and combined 1:1 with 10% charcoal in 0.2 N citric acid to stop the reaction. The reaction mixture was then passed through Dowex-citrate columns into scintillation vials. The radioactivity of the tritiated water produced during enzymatic hydroxylation of 3H/L- tyrosine to L-DOPA was assessed in a liquid scintillation analyzer (Packard, 1900 CA). Cell number, melanin content assay, and protein determination were performed separately for each corresponding sample. For the in vitro assay, NHEM cell lysate and commercial mushroom tyrosinase (Sigma-Aldrich) were individually reacted with 3.5 µCi mL-1 of 3H- tyrosine, L-DOPA, and L-Tyrosine (Sigma-Aldrich), and GFF in Eppendorf tubes for 60 minutes at 37oC. The charcoal-citric acid stop reaction mixture was added 1:1 to each tube and the samples were processed for tritiated water, quantitated by the liquid scintillation analyzer as described in the in situ assay. Enzymatic activity as a function of radioactivity is reported in disintegrations per minute

(DPM) and normalized to protein content. Results are expressed as a percentage of the control.

(n=3-4)

RNA-seq transcriptome profiling: Illumina HiSeq-based next generation sequencing RNA-seq analysis was procured from the Genomics, Epigenomics and Sequencing Core (GESC),

Department of Environmental Health, University of Cincinnati College of Medicine. A

Caucasian/lightly pigmented NHEM cell line with a functional MC1R (determined by cyclic AMP analysis performed by the Dr. Zalfa Abdel-Malek laboratory) was treated with GFF-B at 10% concentration. Cells were harvested at two timepoints: 24 hours and 120 hours (5 days). Cells were scraped and detached from culture dishes with cold PBS EDTA solution, centrifuged, and incubated in RNAlater (Thermo Scientific) solution. Total RNA samples were delivered on ice to the GESC core for RNA isolation and processing. In short, the mRNA from the samples was

59 amplified and converted into a library of cDNA fragments with attached adaptors. Each molecule was sequenced to generate short sequence reads (SSRs) and the SSRs were aligned to a reference genome. A genome-scale transcription map was produced and the total reads of each gene’s exons determined the quantifiable expression level of the gene. From the generated report, only genes that were significantly affected by GFF-B treatment with at least 2-fold upregulation or downregulation and/or p-values <0.05 (probability equation [Anders & Huber, 2010]) compared to control cells were considered for further investigation. (n=1)

Western blot analysis: Whole protein lysates were extracted and separated from the cell pellet after incubation in RIPA buffer supplemented with protease and phosphatase inhibitors. Following harvest and protein extraction, equal stock aliquots were frozen at -80oC until Western blot processing. Western blot analysis was performed with cell protein lysate by analogous methods previously described by Chalupa et al. [2015]. 40-80 µg protein was loaded onto 8-12% sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) or gradient midi gels (Novex, Invitrogen), electrophoresed, and transferred to nitrocellulose membrane. Membranes were blocked with either

5% milk or BSA in tris-buffered saline Tween 20 (TBST). Primary antibodies include: MITF,

PMEL, TYR, TYRP1, and TYRP2/DCT (Covance, Santa Cruz). The appropriate horseradish peroxidase-conjugated anti-rabbit or anti-mouse immunoglobin G secondary antibodies were applied (EMD Millipore). Actin HRP was used as loading control (Santa Cruz). Bands were detected by chemiluminescence with Molecular Imager VersaDoc (Bio-Rad, MP 5000) and then quantified by densitometry using ImageLab (Bio-Rad). Results are expressed as a percentage of the control. (n=3)

60 Immunofluorescent microscopy: For expression analysis, NHEM were fixed with 3% paraformaldehyde on microscopic chamber slides and processed by similar methods as previously described by Chalupa et al. [2015]. Slides were incubated with primary antibodies: SUR-

2/ABCC9, TRPM6, VR1/TRPV1, NHE-4/SLC9A4, and SLC13A1, then with the appropriate fluorescent-conjugated anti-rabbit, anti-mouse, or anti-goat immunoglobin G secondary antibodies

(Abcam, Santa Cruz, Thermo Scientific). Images were captured on a Zeiss LSM710 LIVE Duo

Confocal Microscope (Zeiss) at the Live Microscopy Core (40X-63X), Department of Molecular and Cellular Physiology, University of Cincinnati College of Medicine. Corrected total cell fluorescence (CTCF) was quantitated by Image J (NIH). Results are expressed as a percentage of the control. For colocalization analysis, cells were fixed and processed as abovementioned.

Melanogenic and ion transporter primary antibodies were incubated concurrently for 45 minutes, followed by washing, and incubation with fluorescent labeled secondary antibodies for 30 minutes in the dark, also concurrently. After image processing, split Mander’s coefficients for colocalization were determined using Image J. The average Mander’s coefficient values are reported with standard deviation. (n=7-30)

Statistical analysis: Data was statistically analyzed using one-way ANOVA followed by F-Test and then Student’s t-test. Values were considered significant at p<0.05. Error bars represent standard error mean (SEM).

61 4.4 Results

GFF reduces melanin content in SK-MEL and NHEM

The study began by assessing the ability of GFF to modulate melanization in human pigmentary cells. NHEM and SK-MEL-188 cultures were treated with GFF-A or GFF-B added to their growth medium at concentrations ranging from 0-10% for variable periods. Total melanin content was quantified by spectrophotometric assay. After treatment with GFF-B at 5% and 10% concentrations, melanin quantity was reduced by 48% and 50% within 10 days in SK-MEL-188; and 52% and 55% within 22 days in NHEM (Figure 4.1). Additionally, long-term dosing experiments with 10% GFF-A and 8 day dosing with 10% GFF-A and 10% GFF-B showed significant reduction in melanin content and increase in cell proliferation in NHEM (Figure 4.S1,

Figure 4.S2). To complement the melanin content assays, melanosomes in NHEM dendrites were also evaluated using transmission electron microscopy after treatment with 5% or 10% concentration of GFF-B for 20 days. Microscopic observation revealed less stage IV melanosomes occurring in the dendrites of GFF-B treated NHEM compared to the untreated control (Figure

4.2). Altogether, these results confirm that GFF suppresses constitutive melanization in human pigment cell models.

GFF suppresses tyrosine hydroxylase activity

To understand the effect of GFF on melanogenic enzymes involved in the biochemical process of melanization, we evaluated tyrosine hydroxylase activity: the initial conversion of L-tyrosine to

L-DOPA. The tyrosine hydroxylase activity was determined by a radioassay incorporating tritiated tyrosine and the production of tritiated water quantitated by liquid scintillation. First, we performed

62 an in situ assay by treating darkly pigmented NHEM with GFF-A and GFF-B at 10% concentration for 7 days. Total melanin content and cell number were additionally evaluated. Tyrosine hydroxylase activity was significantly reduced approximately 46% in both GFF treatments (Figure

4.3). Total melanin content was significantly reduced by 40% and 50%, respectively. Cell number was not notably affected by treatment.

Tyrosine hydroxylase activity was also evaluated in an in vitro model using NHEM cell lysate.

Both GFF-A and GFF-B treatments ranging from 2.5-10% effectively suppressed the tyrosine hydroxylase activity in NHEM in a dose-dependent manner (Figure 4.4). To determine if the suppression of tyrosine hydroxylase activity was due to enzymatic competition, we also performed the in vitro radioassay with purified mushroom tyrosinase under the same conditions. Interestingly, mushroom tyrosinase activity was increased with GFF-B treatment (Figure 4.5). These results proclaim that GFF is an operative suppressant of tyrosine hydroxylase activity that reduces the total melanin content in human pigmentary cells without competitive inhibition of tyrosinase.

GFF does not modulate melanogenic proteins

To understand the global gene expression changes induced by GFF treatment in human melanocytes, we procured Illumina HiSeq-based next generation sequencing RNA-seq analysis.

From the generated report, we selected genes that were significantly affected by GFF-B treatment with at least 2-fold upregulation or downregulation compared to untreated control NHEM for further investigation. RNA-seq analysis demonstrated a total of 3,574 genes significantly altered

63 in expression with GFF treatment for 24 hours, and 3,466 genes for 120 hours (Figure 4.S3). We focused our analysis on gene categories that were melanization or ion transport specific.

Initially, we assessed the effect of 10% GFF-B treatment on key melanogenic proteins. mRNA expression of MITF, PMEL, TYR, TYRP1, and TYRP2 was not significantly altered by GFF-B at either time point (Table 4.1) (Table 4.2). We proceeded to validate the mRNA data by evaluating the protein expression of melanogenic proteins in SK-MEL-188 and NHEM models treated with

10% GFF-B for various time periods. The Western blot analysis revealed analogous results

(Figure 4.6, Table 4.S1).

GFF modulates RNA and protein expression of pore proteins for ion transport

Through our RNA-seq analysis, we identified several functional categories of genes involved in ion transporters including V-ATPases, SLCs, and H+, Na+, K+, Ca2+, Cl- specific pumps and channels that potentially influence cellular or organelle pH, a factor which is already acknowledged in regulating melanization. We initiated our investigation on five protein candidates with considerable mRNA expression that were most significantly affected in our preliminary

RNA-seq study: ATP Binding Cassette Subfamily C Member 9/Sulfonylurea Receptor 2

(ABCC9/SUR2), Transient receptor potential cation channel subfamily M (melastatin) member 6

(TRPM6), Transient receptor potential cation channel subfamily V member 1/ Vanilloid receptor

1 (TRPV1/VR1), Solute carrier family 9 member A4/ Na+/H+ exchanger 4 (SLC9A4/NHE4), and

Solute carrier family 13 member 1 (SLC13A1) (Table 4.3, Table 4.4).

64 To evaluate the ability of GFF to modulate the expression of ion transporter pore proteins, we performed immunofluorescent confocal microscopy on lightly pigmented NHEM treated with or without 10% GFF-B for 5 days. Fluorescence intensity, calculated as corrected total cellular fluorescence using Image J, was significantly diminished in GFF-B treated NHEM for each pore protein (Figure 4.7). ABCC9 and TRPV1 expression were the most affected, reduced by 54% and

60%, respectively. TRPM6, SLC9A4, and SLC13A1 were reduced by about 30%. Next, the localization of target ion transporters was investigated also through immunofluorescent microscopic analysis. To understand the ion transporters proximity to melanosomes, lightly pigmented NHEM were fixed and double immunostained for pore proteins and melanosomal enzymes and proteins; PMEL immunostaining represented early staged melanosomes while TYR,

TYRP1, and TYRP2 signified later staged melanosomes (Figure 4.8). Colocalization was evaluated via reported Mander’s split coefficients ranging between 1 and 0, with 1 indicating high colocalization opposed to low with 0. High levels of colocalization were determined for both

SLC9A4 (0.837 ± 0.17 with TYR and 0.828 ± 0.14 with TYRP1) and TRPV1 (0.761 ± 0.15 TYR) in later staged melanosomes (Table 4.5). Moderately high colocalization was determined for

TRPM6 in both early and late stage melanosomes. The high M1 values for SLC9A4 and TRPV1 are indicative of higher melanosomal exclusivity. These results proclaim that GFF successfully represses expression of ion transporter proteins that are associated with the melanosome membrane. This occurrence may induce changes in the melanosomal pH that leads to decreased melanization.

65 4.5 Discussion

The rationale for this study is to acquire an understanding of the role of GFF in influencing cellular and molecular activity that permits inhibition of constitutive pigment synthesis. Understanding the function of GFF in melanin suppression is clinically relevant and may permit a therapeutic utility for dyspigmentation. In this study we demonstrated that GFF effectively suppresses melanization in SK-MEL-188 and NHEM by spectrophotometric melanin content analysis completed after various treatment variations and lengths. Our results show that inhibition of melanogenesis was accomplished in part by the reduction of tyrosine hydroxylase activity, revealed through tritiated tyrosine liquid scintillation radioassays. Conversely, GFF had an enhancing effect on the tyrosine hydroxylase activity of mushroom tyrosinase as discovered in the in vitro radioassay; thus, inferring that GFF does not function as a competitive inhibitor of tyrosinase. RNA-seq transcriptome profiling for mRNA analysis, validated by Western blotting, showed that GFF does not notably alter or diminish the expression of fundamental melanogenic proteins PMEL, TYR,

TYRP1 or TYRP2 nor their regulatory transcription factor MITF. This led the investigation in the pursuit of understanding the effect of GFF on pore proteins due to the notable downregulation of many genes with associations to ion transport that are involved in pH regulation. We selected five initial targets from out RNA-seq data based on substantial read quantity and significance of the downregulation: ABCC9, TRPM6, TRPV1, SLC9A4, and SLC13A1. The expression of these pore proteins was then assessed by fluorescent microscopy which revealed significant downregulation, from 30% to 60%. Three of the pore proteins: TRPM6, TRPV1, and SLC9A4, also showed significant colocalization to melanogenic proteins in NHEM via analysis for split Mander’s coefficients; thus, signifying proximity to the melanosome. The conglomeration of these results causes us to conclude that GFF effectively suppresses melanization in the human pigment cell

66 model; this in part occurs due to a disruption of tyrosine hydroxylase activity. The proven decrease in activity is conjectured to be a result of an alteration in pH within the melanosomes that is counter to optimal tyrosinase activity due to the GFF induced downregulation of pore protein expression.

Hakozaki et al. [2009] demonstrated pigmentation reduction by GFF treatment in multiple experimental models including B16 mouse melanoma, human skin equivalent cultures

(Melanoderm™, MEL300B), and foreskin derived primary human melanocytes. In vitro melanin polymer formation from DOPAchrome and 5,6-dihydroxyindole (DHI) was suppressed in a dose- dependent manner when reacted with GFF; 5,6-dihydroxyindole-2-carboxylic acid (DHICA) remained stable, suggesting selective suppression of the DHI eumelanin pathway. In contrast with our results, Hakozaki et al. [2009] also reported downregulated RNA expression of both TYRP1 and TYRP2 after six days GFF treatment in the B16 mouse melanoma; TYR was not affected.

PMEL protein expression was also inhibited following two days treatment with GFF in B16 mouse melanoma. GFF did not challenge mushroom TYR activity in these studies, which is consistent with our result. Other antimelanogenic fermented broths and yeast derived extracts conceptually comparable to GFF exhibit inhibiting capacity of tyrosinase activity, but some conversely show corresponding downregulation of melanogenic proteins [Chan et al., 2014; Hwang et al., 2013;

Ohgidani et al., 2012; Li et al., 2010].

In this study, we present new evidence for potential pore protein targets in connection with the regulation of melanization. ABCC9 is a subunit of an ATP sensitive K+ channel and relative of the

CFTR protein associated with lysosome acidification [Mindell, 2012]. TRPM6 and TRPV1 are family members of transient receptor potential (TRP) channels, which consists of tetrameric cation

67 channels with six transmembrane domains that transport K+, Na+, Ca2+, and Mg2+ cations. Their relatives TRPM1, TRPM7, and TRPA1, demonstrate roles in Ca2+ homeostasis and melanization

[Bellono et al., 2013; Devi et al., 2009; Oancea et al., 2009; Iuga & Lerner, 2007]. TRPM6 is a

Mg2+ and Ca2+ transporter, while TRPV1 is nonselective. Members of the SLC family such as

SLC24A4, SLC24A5, and SLC45A2 are already recognized to influence skin, hair, and eye pigmentation [Haltaufderhyde & Oancea, 2014; Bellono & Oancea, 2014; Cheli et al., 2009].

+ + + 2- SLC9A4/NHE4 is a Na / H exchanger and SLC13A1 is a Na / SO4 cotransporter. SLC9A4 and related family members NHE1, NHE2 are reported to regulate pH of cells and organelles in other biological models [Arena et al., 2012; Beltrán et al. 2008; Nakamura et al., 2005; Sarangarajan et al., 2001]. The combined M1 Mander’s coefficients for SLC9A4 between TYR, TYRP1, and

TYRP2 and the strength of both M1 and M2 for TRPV1 and TYR suggest high melanosomal exclusivity. M2 coefficients between TRPM6 and PMEL, TYR indicate TRPM6 localizes throughout the cell to other structures in addition to its presence on melanosomes.

Because melanosomes have a endolysosomal lineage, they are speculated to utilize similar pH modulation mechanisms. Lysosomes are digestive organelles known to be highly acidic with a pH range of 4.0-5.0. Lysosomes maintain their acidic nature chiefly through V-ATPase activity, which use ATP to acidify the organelle interior by internalizing protons (H+). This strategy also requires movement of counterions to dissipate the transmembrane voltage, suggesting that for every proton translocated there is one cation removed or one anion introduced [Bellono & Oancea, 2014;

Mindell, 2012; DiCiccio & Steinberg, 2011]. Counterion movements can be achieved by the other putative transporters such as the ones in this study. Takaya [1977] reported the presence of magnesium in premelanosomes and the edges of the mature melanosomes using energy dispersive

68 X-ray microanalysis of melanosomes from black facial hair bulbs. Medium to high levels of calcium, potassium, and sulfur were apparent as well as nominal phosphorus and chlorine [Takaya,

1977]. Through this research, we begin to uncover the specific intracellular events involved in the suppression of pigmentation in melanocytes enacted by GFF. The novel ion transporters identified also augment the growing collection of pore proteins involved with melanosomal pH and the control of melanogenesis for further study.

69 4.6 Acknowledgments

This research was supported by Procter & Gamble, Co. and the National Vitiligo Foundation, Inc.

70 4.7 Author Contributions

JKWC, AK, TV, REB performed the research

JKWC, ALK, TH, REB designed the research study

ALK, TH, REB contributed essential reagents or tools

JKWC, AK, TV, REB analyzed the data

JKWC, REB wrote the paper

71 4.8 Figures and Tables

Figure 4.1 – GFF effect on melanin content in human SK-MEL and NHEM. Treatment with

5% and 10% GFF-B effectively reduced melanin content in human SK-MEL-188 and NHEM, analyzed spectrophotometrically and normalized to protein content. Cell pellets are pictured below. In SK-MEL-188, melanin was reduced by 48% and 50%, respectively, within 10 days (A).

In NHEM, melanin was reduced by 52% and 55% within 22 days (B). *p<0.05, SEM

72

Figure 4.2 – GFF effect on melanosome maturity and quantity in NHEM. NHEM were treated with 5% and 10% GFF-B for 20 days then processed for transmission electron microscopy (TEM).

Reduction of stage IV melanosomes was objectively observed in the dendrites of NHEM. Arrows indicate examples of undermelanized melanosomes.

73

Figure 4.3 – GFF effect on tyrosine hydroxylase activity in situ. Spectrophotometric analysis combined with liquid scintillation radioassay revealed 10% GFF-A and 10% GFF-B treatment reduced the melanin content and tyrosine hydroxylase activity in NHEM cultures in situ within 7 days. Tyrosine hydroxylase activity was reduced 46% by both formulations. Melanin content was suppressed by 40% and 50%, respectively. Cell number, quantified with Coulter counter, was statistically unaffected by treatment. Results are expressed as a percentage of the control. *p<0.05,

SEM

74

Figure 4.4 – GFF effect on tyrosine hydroxylase activity in vitro. NHEM protein lysate was reacted with 3H-tyrosine, L-tyrosine, and L-DOPA substrates with GFF-A or GFF-B at 0-10% concentrations for 60 minutes. Tyrosine hydroxylase activity in vitro was determined by liquid scintillation radioassay. Treatment with concentrations of 5% and 10% GFF-A as well as 2.5%-

10% GFF-B diminished activity in a relatively dose-dependent manner. Results (DPM/µg protein) are expressed as a percentage of the control. *p<0.05, SEM

75

Figure 4.5 – GFF effect on tyrosine hydroxylase activity of mushroom tyrosinase. Mushroom

TYR was reacted with 3H-tyrosine, L-tyrosine, and L-DOPA substrates with 0-10% GFF-B for 60 minutes. Tyrosine hydroxylase activity in vitro was determined by liquid scintillation radioassay.

All concentrations of GFF-B conversely increased activity of mushroom TYR indicating GFF is not a competitive inhibitor of TYR. Results (DPM/U) are expressed as a percentage of the control.

*p<0.05, SEM

76 Table 4.1 – RNA-seq melanogenic protein targets – 24 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

MITF 6441.57 8370.07 1.30 0.458

PMEL 47624.59 44389.67 1.07 0.998

TYR 9665.19 11565.13 1.20 0.632

TYRP1 196163.04 231537.76 1.18 0.846

TYRP2 193.51 276.50 1.43 0.301

Table 4.2 – RNA-seq melanogenic protein targets – 120 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

MITF 5220.92 6392.53 1.22 0.539

PMEL 42258.77 42174.12 1.00 0.998

TYR 6304.03 8929.83 1.42 0.324

TYRP1 167282.25 216509.73 1.29 0.908

TYRP2 418.21 582.18 1.39 0.229

Table 4.1 and 4.2 – RNA-seq melanogenic protein targets. RNA-seq transcriptome profiling analysis was performed with mRNA extracted from NHEM treated with 10% GFF-B treatment for 24 hours and 120 hours (5 days). GFF-B did not significantly alter the mRNA levels of melanogenic proteins that are required for melanin synthesis. Results are expressed as read counts.

Significance: Fold Change > 2, *p<0.05

77

Figure 4.6 – GFF effect on melanogenic proteins expression. Analogous results for the melanogenic proteins were confirmed by Western blotting and densitometric analyses in NHEM and SK-MEL-188 treated with or without 10% GFF-B for varying time periods (5-12 days tested).

Results expressed as a percentage of the control. *p<0.05

78 Table 4.3 – RNA-seq pore protein targets – 24 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

ABCC9 8168.07 3671.77 2.22 0.021*

SLC9A4 3884.39 1402.60 2.77 0.002*

Table 4.4 – RNA-seq pore protein targets – 120 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

ABCC9 7777.41 2929.57 2.66 0.003*

TRPM6 158.99 70.57 2.25 0.029*

TRPV1 197.54 97.55 2.03 0.041*

SLC9A4 3121.18 1230.77 2.54 0.001*

SLC13A1 234.16 88.21 2.66 0.004*

Table 4.3 and 4.4 - RNA-seq pore protein targets. RNA-seq transcriptome profiling analysis was performed with mRNA extracted from NHEM treated with 10% GFF-B treatment for 24 hours and 120 hours (5 days). GFF-B treatment significantly altered the mRNA levels of these ion channels and transporters at estimated significance. Results are expressed as read counts.

Significance: Fold Change > 2, *p<0.05

79

Figure 4.7 – GFF effect on pore proteins expression. NHEM were treated with 10% GFF-B for

5 days. Expression of pore proteins was examined with immunofluorescent microscopy

(corresponding confocal images below). ABCC9 and TRPV1 were reduced by 54 and 60%, respectively. TRPM6, SLC9A4, and SLC13A1 were reduced similarly by 30% as well. Results are expressed as a percentage of the control (corrected total cell fluorescence). *p<0.05, SEM

80

Figure 4.8 – Colocalization imaging of melanogenic and pore proteins. Fluorescent composite and colocalization images from untreated, fixed NHEM immunostained for melanogenic and ion channels and transporter targets. BIP + TYRP2 is the negative control. Melanogenic proteins are stained with Alexa Fluor 488 and pore proteins are stained with Alexa Fluor 546. Nuclear staining with DAPI. White pixels indicate colocalization.

81 Table 4.5 – Colocalization of melanogenic and pore proteins Proteins M1 M2 ABCC9 + PMEL 0.090 ± 0.03 0.315 ± 0.12 ABCC9 + TYR 0.116 ± 0.02 0.102 ± 0.03 ABCC9 + TYRP1 0.183 ± 0.09 0.130 ± 0.03 ABCC9 + TYRP2 0.073 ± 0.02 0.053 ± 0.05 TRPM6 + PMEL 0.274 ± 0.10 0.659 ± 0.23 TRPM6 + TYR 0.463 ± 0.022 0.668 ± 0.20 TRPV1 + PMEL 0.197 ± 0.11 0.175 ± 0.11 TRPV1 + TYR 0.761 ± 0.15 0.740 ± 0.16 SLC9A4 + PMEL 0.267 ± 0.16 0.530 ± 0.25 SLC9A4 +TYR 0.837 ± 0.17 0.453 ± 0.25 SLC9A4 + TYRP1 0.828 ± 0.14 0.498 ± 0.13 SLC9A4 + TYRP2 0.593 ± 0.20 0.542 ± 0.18 BIP + TYRP2 0.023 ± 0.01 0.063 ± 0.04

Table 4.5 – Colocalization of melanogenic and pore proteins. Untreated NHEM were fixed and fluorescently immunostained for melanogenic and ion channels and transporter proteins. BIP +

TYRP2 is the negative control. Mander’s split coefficients with standard deviation are reported for colocalization analysis. M1 signifies Mander’s coefficient for the percentage of pore proteins that colocalize with melanogenic proteins. M2 signifies Mander’s coefficient for the percentage of melanogenic proteins that colocalize with pore proteins.

82 4.9 Supplemental Figures and Tables

Figure 4.S1 – Melanin content and cell proliferation in NHEM after long term dosing. NHEM were treated with or without 10% GFF-A for 8 weeks. Melanin content, analyzed spectrophotometrically, was reduced by 37%. Cell number, quantitated by Coulter counter, was doubled. Results are expressed as ng melanin/ µg protein and as a percentage of the control.

*p<0.05, SEM

83

Figure 4.S2 – GFF visual effect on melanization in NHEM. NHEM were treated with or without

10% concentration of GFF-A or GFF-B for 8 days. Harvested NHEM in suspension with apparent reduction of melanization are pictured on the left. Melanin content, analyzed spectrophotometrically, exhibited a dose-dependent reduction. Results expressed as ng melanin/

µg protein. *p<0.05, SEM

84

Figure 4.S3 – GFF effect on gene transcription in NHEM. RNA-seq heat maps visually illustrate notable quantities of genes in NHEM modulated due to treatment with 10% GFF-B at both 24 and 120 hours (5 days); upregulated (red) or downregulated (green). At 24 hours, 3,574 genes were altered by 2-fold and 533 of those genes had p-values <0.05. At 120 hours, 3,466 genes were modulated and 572 of those genes had p-values <0.05.

85 Table 4.S1 – Western blot analysis of melanogenic protein expression Protein Cell Model Treatment Dosing Length % of Control ± SEM MITF NHEM 10% GFF-B 5 days 102.2% ± 3.1 PMEL NHEM 10% GFF-B 7 days 90.5% ± 9.45 PMEL SK-MEL-188 10% GFF-B 10 days 108.9% ± 5.33 PMEL NHEM 10% GFF-B 23 days 76.2% ± 13.08 TYR NHEM 10% GFF-B 8 hours 109.3% ± 11.24 TYR NHEM 10% GFF-B 8 days 67.4% ± 31.80 TYR SK-MEL-188 10% GFF-B 10 days 79.5% ± 17.58 TYR SK-MEL-188 10% GFF-B 12 days 72.0% ± 4.80* TYR NHEM 10% GFF-B 12 days 136.9% ± 8.60* TYR NHEM 10% GFF-B 12 days 103.0% ± 9.71 TYR NHEM 10% GFF-B 21 days 84.4% ± 4.29 TYRP1 SK-MEL-188 10% GFF-B 10 days 103.1% ± 7.12 TYRP2 NHEM 10% GFF-B 12 days 97.4% ± 15.1

Table 4.S1 – Western blot analysis of melanogenic protein expression. 10% GFF-B did not significantly diminish or alter the expression of melanogenic proteins as evaluated by Western blot analysis in various protocols. Results are reported as a percentage of control with standard error mean (SEM) values. Bold values are additionally depicted in Figure 4.6. *p<0.05

86 CHAPTER 5:

GALACTOMYCES FERMENT FILTRATE SUPPRESSES REACTIVE OXYGEN SPECIES

GENERATION AND PROMOTES CELLULAR REDOX BALANCE IN HUMAN

MELANOCYTES VIA NRF2-ARE PATHWAY

JàNay K. W. Cooper1, Amy Koshoffer2, Ana Luisa Kadekaro2, Tomohiro Hakozaki3, and

Raymond E. Boissy2

1James L. Winkle College of Pharmacy, University of Cincinnati, Cincinnati, Ohio, USA;

2Department of Dermatology, University of Cincinnati College of Medicine, Cincinnati, Ohio,

USA; 3Procter & Gamble, Co., Cincinnati, Ohio, USA

Presently being edited for submission to Pigment Cell and Melanoma Research.

Modified for dissertation purposes.

87 5.1 Abstract

Background/Objective: Galactomyces ferment filtrate (GFF, Pitera™) is a yeast derived extract currently used as a moisturizing agent in cosmetics. GFF demonstrates anti-aging, barrier health, and hypopigmenting effects in skin. The antioxidant properties of GFF are the focus of our study.

Methods: Foreskin derived NHEM of varying skin types were treated with GFF 0-10%. Cell viability was determined by MTT assay. ROS were visualized by fluorescent microscopy. UVR induced H2O2 generation was monitored by luminescence. RNA-seq transcriptome profiling was completed with mRNA extracted from NHEM. Protein expression was quantitated by Western blot and densitometric analyses.

Results: GFF sustained cell viability and combated ROS generation in NHEM challenged with

4TBP. GFF also suppressed H2O2 generation in NHEM irradiated with UVB. RNA-seq and

Western blot analyses revealed that GFF significantly altered the mRNA and protein levels of phase II antioxidant enzymes mediated by Nrf2-ARE pathway (HO1, NQO1, TXRD1, Nrf2). HO1 was highly inducible in both the cytoplasm and nucleus in NHEM (6- and 8-fold).

Conclusions: GFF effectively suppresses generation of ROS in part by upregulating the presence of endogenous antioxidant enzymes via activating the Nrf2-ARE pathway; thus, priming and protecting NHEM against oxidative assaults.

5.1.1. Keywords: Galactomyces ferment filtrate, Nrf2-ARE, heme oxygenase 1, NAD(P)H quinone oxidoreductase 1, thioredoxin reductase 1.

88 5.2 Introduction Vitiligo is an acquired hypopigmentary disease resulting from the loss or dysfunction of epidermal melanocytes. The understanding of the etiology is incomplete, yet various studies affirm implications of cellular cytotoxicity, oxidative stress, and systemic immune dysregulation

[Koshoffer & Boissy, 2014]. Epidermal melanocytes are innately susceptible to excessive free radicals and consequent oxidative stress compared to the other skin cell types [Pelle et al., 2014;

Jenkins & Grossman, 2013; Yohn et al., 1991]. Because melanin synthesis involves the generation of cytotoxic melanin intermediates and reactive oxygen species (ROS), melanocytes maintain an inherent pro-oxidation denoted as endogenous melanogenic cytotoxicity (EMC) [Denat et al.,

2014; Chen et al., 2009b; Nappi & Vass, 1996; Koga et al., 1992]. This vulnerable state in combination with exposure to aggravating factors like ultraviolet radiation (UVR), hormones, stress, or cytotoxic compounds can lead to the progression of both dyspigmentation or oxidative driven skin diseases [Lee, 2015; Boissy et al., 2012; Bickers & Athar, 2006]. Thus, effectively managing pigmentary disorders while simultaneously encouraging melanocyte vitality is of consequence.

The transcription factor nuclear factor, erythroid 2-like 2 (Nrf2) is considered a central regulator for cell protection and survival through activation of the antioxidant response element (ARE) pathway. This pathway is stimulated by changes in the redox state of the cell and functions to restore homeostasis by upregulating antioxidant, xenobiotic metabolizing, and other cytoprotective proteins and enzymes constituting over 600 gene targets [Espinosa-Diez et al. 2015;

Baird & Dinkova-Kostova, 2011]. Scientific literature is also progressively uncovering a substantial role for Nrf2 in the metabolism and protection of skin cells [Jian et al. 2011; Kokot et al. 2009; Marrot et al. 2007].

89 Most research links redox imbalance to the pathogenesis of vitiligo [Colucci et al., 2015b;

Koshoffer & Boissy, 2014; Laddha et al., 2013b]. Redox implications including systemic increases in oxidant status with simultaneous decreases in antioxidant status; downregulation of enzymatic antioxidants like catalase (CAT), glutathione peroxidase (GPX), and glucose-6-phosphate dehydrogenase (G6PD); and upregulation of ROS hydrogen peroxide (H2O2) and peroxynitrite

(ONOO−) have been documented [Akoglu et al., 2013; Dammak et al., 2009; Khan et al., 2009;

Sravani et al., 2009; Salem et al., 2009; Arican & Kurutas, 2008; Jain et al., 2008; Schallreuter et al., 2007; Schallreuter et al., 1991]. Various scientific studies associate vitiligo with multiple classical manifestations of oxidative stress such as lipid peroxidation, impaired signaling, structural irregularities, and dysfunction of cellular organelles [Laddha et al., 2014; Laddha et al.,

2013a; Dammak et al., 2009; Khan et al., 2009; Jain et al., 2008; Koca et al., 2004; Dell’Anna et al. 2001; Le Poole et al., 2000; Schallreuter et al., 1999; Boissy et al. 1991]. Intriguingly, recent studies indicate impaired Nrf2-ARE pathway signaling and functional elements in vitiligo pathophysiology [Jian et al. 2014; Qiu et al., 2014; Amin et al. 2013; Guan et al., 2008; Pae et al.,

2004]. Thus, normalizing the cellular redox status and increasing protection from excessive ROS buildup potentially through selective induction of Nrf2-ARE pathway is a prime strategy for preventing the aggression of the oxidative driven diseases like vitiligo.

Galactomyces ferment filtrate (GFF), commercially known as Pitera™, is currently used as a cosmetic ingredient in skincare. GFF is a yeast derived extract that contains a unique composition of vitamins, minerals, small peptides, and oligosaccharides. Studies with GFF elucidate anti-aging, barrier health, and hypopigmenting effects on skin [Takei et al., 2015; Wong et al., 2011; Hattori

90 et al., 2010; Hakozaki et al., 2008; Tsai et al. 2006]. However, the full characterization of its properties is still in progress.

In this study, we investigated the biological effects of GFF in primary human melanocyte cultures to further understand its antioxidant capacity and cellular health inducing benefits. We demonstrate that GFF has the ability to upregulate downstream phase II antioxidant enzymes through activation of the Nrf2-ARE pathway, thereby effectively priming the cells more readily against oxidative assault. Selective induction of Nrf2-ARE is also a compelling therapeutic strategy for vitiligo prevention and other oxidative driven diseases that impact melanocytes.

91 5.3 Materials and Methods

Cell culture: Primary cultures of normal human epidermal melanocytes (NHEM) were established from discarded neonatal foreskin obtained from different skin types as previously described by

Chalupa et al. [2015]. Tissues were procured from University Hospital in Cincinnati or from the

Christ Hospital in Cincinnati. Patient consent was not required for experimentation because of

USA laws regarding left over human tissues from surgery. Briefly, foreskins were washed with betadine and phosphate buffered saline (PBS), sectioned and incubated in 0.25% trypsin, then rocked at 4oC overnight. Tissues were vortexed and centrifuged to separate the epidermis and dermis. The epidermal fraction was plated in T25 or T75 flasks with melanocyte growth medium.

Melanocyte growth medium consisted of MCDB-153 (Sigma-Aldrich) supplemented with 4% fetal bovine serum (FBS), 1% antibiotic/antimycotic, 15 µg mL-1 bovine pituitary extract (BPE),

5 µg mL-1 insulin, 8 nM 12-O-tetradecanoyl-phorbol-13-acetate (TPA), and 0.6 ng mL-1 basic fibroblast growth factor (bFGF) to aid in cell proliferation, dendricity, and melanization. Cultures

o were maintained in a humidified incubator with 5% CO2 at 37 C. Growth medium was changed every three to four days.

GFF treatment: GFF was prepared using two different filtration methods: a routine mechanism

GFF-A, and an enhanced mechanism GFF-B. Filtrates were prepared at Procter & Gamble

Innovation Godo Kaisha, Kobe Technical Center; shipped to the laboratory on dry ice; and stored at 4oC. NHEM cultures were plated and incubated in experimental growth medium consisting of

MCDB-153 supplemented with 6% FBS, 1% antibiotic/antimycotic, 15 µg mL-1 BPE, 5 µg mL-1 insulin, 2 nM TPA and 0.15 ng mL-1 bFGF for 48 hours before the start of experiments. During experimentation, NHEM were treated with GFF-A or GFF-B, 0-10% added to the cell

92 experimental growth medium every 48 hours for 5-7 days. In single day experiments, NHEM were administered a single dose of GFF-A or GFF-B, 0-10%, added to their experimental growth medium and then harvested at the indicated timepoint. All cultures were maintained in a

o humidified incubator with 5% CO2 at 37 C until experiment harvest.

4TBP administration: 4-tertiary butylphenol (4TBP) was administered to NHEM as previously described by Manga et al. [2006]. In short, 4TBP (Sigma-Aldrich) was dissolved in 70% ethanol and added to the experimental growth medium at indicated final concentrations (0-300 µM).

Control NHEM were treated with 70% ethanol vehicle (maximum of 0.5% final concentration of alcohol).

UV irradiation: Caucasian/lightly-pigmented NHEM were incubated in PBS and irradiated with

UVB 90 or 105 mJ cm-2 using FS 20 lamps, with a peak emission at 313 nM, as previously described by Kadekaro et al. [2010]. Post irradiation, cells were incubated with 10% GFF-B in experimental growth medium until indicated harvest.

MTT cell viability assay: NHEM were treated with 200 µM 4TBP alone or in combination with

0-10% GFF-B for 6 days. Upon harvest, cell viability was measured by MTT assay (BioAssay

Systems) according to manufacturer instructions. In brief, 1 X 104 cells/well were plated in triplicate in a 6 well plate to attach overnight. MTT reagent (tetrazole) was added to the cells and incubated in a 37oC humidified chamber for 4 hours. The tetrazole is converted to formazan in mitochondria of living cells. The formazan crystals formed were solubilized in solubilization buffer and the wavelength was read in a microplate reader (Bio-Rad, 550). Fixation was used as a

93 positive control. Cell viability was calculated from absorbance readout and the results are expressed as a percentage of the control. (n=3)

Fluorescent ROS visualization: NHEM were challenged with 300 µM 4TBP alone or in combination with 10% GFF-B or 30 U Catalase (Sigma-Aldrich) for 90 minutes. ROS generation was visualized using Image-IT® LIVE Green Reactive Oxygen Species Detection Kit (Molecular

Probes) according to the manufacturer’s instructions as previously described by Manga et al.

[2006]. Images in both phase-contrast and fluorescence microscopy were captured on Olympus

IMT-2 inverted microscope (20X).

H2O2 generation luminometer assay: Caucasian/lightly pigmented NHEM were irradiated in

PBS with UVB (105 mJ cm-2) as mentioned above and immediately treated with or without 10%

GFF-B. Samples were harvested at various timepoints (0, 15, 30, and 45 minutes) and then analyzed for H2O2 generation by luminescence of luminol as previously described by Kadekaro et al. [2005]. In short, aliquots of medium were transferred to tubes with respiratory buffer and luminol (Sigma-Aldrich). H2O2 release was determined by luminescence measured with a luminometer (Turner, TD10e). Readings were plotted against a standard curve with known H2O2

5 concentrations. Results are expressed as pMol H2O2 per 1 X 10 cells. (n=3)

RNA-seq transcriptome profiling: Illumina HiSeq-based next generation sequencing RNA-seq analysis was procured from the Genomics, Epigenomics and Sequencing Core (GESC),

Department of Environmental Health, University of Cincinnati College of Medicine. A

Caucasian/lightly pigmented NHEM cell line with a functional MC1R (determined by cyclic AMP

94 analysis performed by the Dr. Zalfa Abdel-Malek laboratory) was treated with GFF-B at 10% concentration. Cells were harvested at two timepoints: 24 and 120 hours (5 days). Cells were scraped and detached from culture dishes with cold PBS ethylenediamine tetra acetic acid (EDTA) solution, centrifuged, and incubated in RNAlater (Thermo Scientific) solution. Total RNA samples were delivered on ice to the GESC core for RNA isolation and processing. In short, the mRNA from the samples was amplified and converted into a library of cDNA fragments with attached adaptors. Each molecule was sequenced to generate short sequence reads (SSRs) and the SSRs were aligned to a reference genome. A genome-scale transcription map was produced and the total reads of each gene’s exons determined the quantifiable expression level of the gene. From the generated report, only genes that were significantly affected by GFF-B treatment with at least 2- fold upregulation or downregulation and/or p-values <0.05 (probability equation [Anders &

Huber, 2010]) compared to control cells were considered for further investigation. (n=1)

Protein determination: Protein was extracted from cells using radioimmunoprecipitation assay

(RIPA) buffer supplemented with protease and phosphatase inhibitors. Cells were centrifuged at

10,000 RPM for 10 min at 4oC. The protein supernatant/lysate was separated from the cell pellet and placed on ice. 2-10 µL aliquots were prepared with Pierce® BCA Protein Assay Kit (Thermo

Scientific). The colorimetric assay was spectrophotometrically read at 570 nm absorbance with bovine serum albumin (BSA) standard in microplate reader (Bio-Rad, 550). Protein content results were calculated against BSA standard curve and reported as µg of protein. (n=3)

Western blot analysis: Whole protein lysates were extracted and separated from the cell pellet after incubation in RIPA-buffer as described above. Cytoplasmic and nuclear protein extraction

95 was performed with NE-PER™ Nuclear and Cytoplasmic Extraction Reagents Kit (Thermo

Scientific) according to the manufacturer’s instructions. Following harvest, lysate extraction, and protein determination; equal stock aliquots were frozen at -80oC until Western blot processing.

Western Blot analysis was performed with cell protein lysate by analogous methods previously described by Chalupa et al. [2015]. 40-80 µg protein was loaded onto 8-12% sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) or gradient midi gels (Novex, Invitrogen), electrophoresed, and transferred to nitrocellulose membrane. Membranes were blocked with either

5% milk or BSA in tris-buffered saline Tween 20 (TBST). Primary antibodies include: HO1,

NQO1, TXNRD1, and Nrf2 (Santa Cruz). The appropriate horseradish peroxidase-conjugated anti-rabbit or anti-mouse immunoglobin G secondary antibodies were applied (EMD Millipore).

Actin HRP, Lamin A, or Lamin A/C were used as loading controls (Santa Cruz). Bands were detected by chemiluminescence with Molecular Imager VersaDoc (Bio-Rad, MP 5000) and then quantified by densitometry using ImageLab (Bio-Rad). Results are expressed as a percentage of the control. (n=1-3)

Statistical analysis: Data was statistically analyzed using one-way ANOVA followed by F-Test and then Student’s t-test. Values were considered significant at p<0.05. Error bars represent standard error mean (SEM).

96 5.4 Results

GFF preserves cell viability against induced cytotoxicity

The first experimental initiative was to determine the effect of GFF on cell viability. We challenged

NHEM with 4TBP, a phenolic depigmenting compound cytotoxic to melanocytes, alone or in combination with 0-10% GFF-B for 6 days. Cell viability was analyzed by MTT assay. All concentrations of GFF-B, 2.5%-10%, maintained cell viability despite 4TBP co-administration compared to the untreated, 4TBP challenged cells and the positive control of fixed cells. This result demonstrated a protective effect against induced 4TBP cytotoxicity (Figure 5.1).

GFF suppresses ROS generation

Next, we evaluated the antioxidant effect of GFF in our cell culture model using a fluorescent assay for ROS visualization by microscopy. We challenged NHEM with 300 µM 4TBP in ethanol vehicle alone or in combination with 10% GFF-B for 90 minutes. Catalase, a known antioxidant that scavenges H2O2, was used as a negative control. As expected, 4TBP alone increased the fluorescent signal corresponding to ROS generation (Figure 5.2). 4TBP and 10% GFF-B co- treatment resulted in a diminished fluorescent signal, suppressing the ROS induction. The result was equivalent to the weakened signal identified in 4TBP and Catalase co-treatment.

We then performed a specific H2O2 generation assay through luminescence analysis. Lightly pigmented NHEM were irradiated with UVB 105 mJ cm-2 and then incubated with or without 10%

GFF-B until harvest for analysis at varying timepoints (0-45 minutes). H2O2 generation was monitored by the addition of luminol to the media immediately following irradiation and then

97 quantitated with luminescence readings captured by a luminometer. At timepoint 15 minutes

(T15), 10% GFF-B suppressed the induced H2O2 by 70% compared to the untreated and irradiated control (Figure 5.3). This suppression continued for the remainder of the experiment. Among the unirradiated NHEM, the 10% GFF-B treated cells showed reduced endogenous H2O2 generation compared to the untreated control at T15. Altogether, these results indicate GFF treatment inhibits the generation of ROS via a protective mechanism in NHEM.

GFF activates melanocyte antioxidant response

To understand the global gene expression changes induced by GFF treatment in human melanocytes, we procured Illumina HiSeq-based next generation sequencing RNA-seq analysis.

From the generated report, we selected genes that were significantly affected by 10% GFF-B treatment at two timepoints (24 hours and 120 hours) with at least 2-fold upregulation or downregulation compared to untreated control cells for further investigation. RNA-seq analysis demonstrated a total of 3,574 genes significantly altered in expression with GFF treatment for 24 hours, and 3,466 genes for 120 hours (Figure 4.S3). We focused our analysis on gene categories related to oxidative stress and endogenous antioxidant response, with initial attention on the 24 hours timepoint. Several gene targets associated with the adaptive oxidative stress response were modulated by at least 2-fold (Figure 5.4). Mitogen-activated protein kinase 10 (MAPK10/JNK3) was upregulated while its potential downstream signaling target Jun B proto-oncogene (Jun B/AP-

1) was downregulated indicating probable transcriptional activation of responsive stress pathways.

Additionally, three antioxidant enzymes including heme oxygenase 1 (HMOX1/HO1), NAD(P)H quinone oxidoreductase 1 (NQO1), and thioredoxin reductase 1 (TXNRD1) from the Nrf2-ARE

98 pathway were significantly upregulated (Table 5.1). However, glutathione peroxidase 3 (GPX3)

- and superoxide dismutase 3 (SOD3), enzymes that neutralize H2O2 and O2 respectively, were both downregulated. Downregulation of cytochrome B-245, alpha polypeptide (CYBA), a protein

- involved in O2 production and phagocytosis, was also observed signifying a protective response of ROS synthesis repression. This antioxidant response was partially maintained at the second timepoint 120 hours (5 days) (Table 5.2). In total, this data indicates that GFF is able to successfully activate antioxidant and protective gene transcription in NHEM.

GFF upregulates Nrf2-ARE downstream targets HO1, NQO1, and TXNRD1

To statistically validate the RNA-seq data, Western blot analysis was performed to elucidate the protein expression of the antioxidant targets. NHEM cultures were evaluated in both long-term and short-term experimental in vitro protocols. In the long-term protocol, NHEM were treated with or without 10% concentration of GFF-A or GFF-B for 5-7 days. At the completion of dosing, whole cell protein lysates were extracted and analyzed. NHEM treated with 10% GFF-B showed significant upregulation of HO1, NQO1, and TXNRD1 protein expression (Figure 5.5). Nrf2 polyubiquitinated protein was also increased after 7 days with both 10% GFF-A and GFF-B treatment.

Short-term, single dose protocols were also performed with 10% GFF-B for investigation of separated cytoplasmic and nuclear protein lysate fractions. Initially time course experiments were performed at 4, 8 and 24 hours to understand the trend of the protein expression and to select optimal timepoints (data not shown). Cytoplasmic and nuclear protein fractions from NHEM

99 treated with 10% GFF-B for 8 hours were subsequently evaluated. Staining for tyrosinase was done in the nuclear fraction to validate adequate separation (not shown). Increases of Nrf2 was demonstrated in both the nuclear and cytoplasmic fractions compared to the untreated control

(Figure 5.6). Notably, HO1 protein was increased 6- and 8-fold in both the cytoplasmic and nuclear fractions, respectively. This collection of experiments confirmed that GFF does upregulate the expression of antioxidant enzymes via activation of the Nrf2-ARE pathway. We supplementally investigated the sustenance of GFF’s effect on Nrf2-ARE in NHEM after irradiation with UVB. NHEM were pretreated with or without 10% GFF-B for 6 days, irradiated with UVB 90 mJ cm-2, and then treated again for an additional 24 hours. NQO1 and HO1 expression was evaluated at harvest by Western blotting. Analysis revealed NQO1 levels were maintained in NHEM post irradiation (Figure 5.S1). HO1 again demonstrated vast inducibility; however, expression was barely sustained over untreated, irradiated NHEM. Altogether, this data illustrates the positive ability of GFF to incite endogenous antioxidant capacity in NHEM through mediation of the Nrf2-ARE pathway.

100 5.5 Discussion The rationale for this experimentation is to acquire an understanding of the role of GFF in influencing the Nrf2-ARE pathway and to comprehend the implications of GFF treatment on the downstream antioxidant targets in association with the protective skin health features exhibited in cosmetic use. Explication of the antioxidant capacity of GFF and its function in ROS suppression is clinically relevant and may permit a therapeutic utility for oxidative driven diseases like vitiligo.

In this study we demonstrated that GFF effectively suppresses ROS generation and successfully executes a protective capacity in NHEM assaulted with the cytotoxic, depigmenting compound

4TBP, analyzed by fluorescent microscopy and MTT assays for cell viability. This conclusion was further supplemented by luminescence analysis on UVB irradiated NHEM and the inhibition of

H2O2 generation in GFF rescued cells. The RNA-seq transcriptome profiling analysis propelled the investigation further, establishing new gene targets for endogenous antioxidant activity which we validated by Western blot and densitometric analyses. HO1, NQO1, and TXNRD1 were all substantially upregulated in NHEM of varying pigmentation that were treated with GFF.

Accumulation of polyubiquitinated Nrf2 also increased after a week of dosing with GFF signifying

Nrf2-ARE mediated activation as the cause of antioxidant upregulation. When we investigated the cytoplasmic and nuclear fractions independently by Western blot after a single treatment of GFF in NHEM, Nrf2 expression was found to be increased in both segments, as well as consistent polyubiquitinated Nrf2 results. Most notable was the 6-fold and 8-fold upregulated HO1 expression revealed in both the cytoplasmic and nuclear fractions. Lastly, the endogenous antioxidant capacity was challenged with UVB. NHEM that were treated with GFF and then irradiated subsequently upheld the increased NQO1 expression. HO1 proved highly inducible in treated, unirradiated NHEM, but levels were not sustained in treated and irradiated NHEM.

Overall, the data from our investigation leads us to conclude that GFF has effective antioxidant

101 capacity in NHEM which is accomplished in part by upregulating antioxidant enzymes through

Nrf2-ARE pathway activation. Results also suggest GFF may have physiological components with functional antioxidant action.

HO1 has indirect antioxidant activity that aids in the defense against ROS by catalyzing the

2+ degradation of heme into Fe , CO, H2O, and biliverdin. Newly synthesized biliverdin is rapidly converted to bilirubin by biliverdin reductase (BVR) at the expense of NADPH. Bilirubin, a potent antioxidant, scavenges ROS primarily in lipophilic locations such as bilayer membranes [Kim &

Park, 2012; Briganti & Picardo, 2003; Wei et al., 2003]. Because of the high inducibility of HO1, the release of Fe2+ is a concern and could be connected with the elevated levels of ferritin reported in melanocytes compared to keratinocytes that correlated with increased DNA oxidative lesions

[Pelle et al., 2014]. Bioavailable iron is a source for cytotoxic Fenton reactions; therefore, more work is needed to illustrate a clear picture of iron and heme homeostasis in melanocytes especially with UVR. HO1 levels weren’t maintained with GFF treatment in NHEM exposed to UVB, which could have been a protective response if the free iron becomes more threatening under irradiated conditions; however, UVA is more effective at inducing HO1 [Nisar et al., 2015; Xiang et al.,

2011; Zhong et al., 2010]. High ferritin levels matched with HO1 is considered protective by other authors [Tyrrell, 2012; Xiang et al., 2011]. Nevertheless, HO1 deficiency in various models does lead to oxidative stress and adverse pathophysiologies [Kovtunovych et al., 2010; Grochot-

Przeczek1 et al., 2009; Poss & Tonegawa, 1997].

NQO1 is primarily localized in the cytosol, existing as a homodimer with one molecule of FAD per monomer. It catalyzes single step 2-electron reductions of quinones using NADPH via a ping

102 pong mechanism. This results in less reactive hydroquinone species and can yield substrates for phase II conjugation reactions that promote excretion; a strategy that would work well against the reactive quinone intermediates generated in melanin synthesis [Ross et al., 2000]. Okubo et al.

[2016] exhibited protection from cytotoxicity induced by phenolic compound rhododendrol (RD) by overexpression of NQO1 in B16BL6 mouse melanoma and NHEM, which parallels to the GFF induced protection from 4TBP in our study. NQO1 additionally generates antioxidant forms of

- ubiquinone and vitamin E and is proposed to have a role in O2 scavenging that would also serve well in this model [Siegel et al., 2004; Ross et al., 2000].

TXNRD1 functions in the thioredoxin (TXN) detoxification system throughout the cytoplasm.

This system is led by peroxiredoxin (PRDX), an antioxidant enzyme that reduces H2O2 and alkyl hydroperoxides by using thiols, primarily TXN, as electron donors in their catalytic action.

TXNRD1 regenerates oxidized thioredoxin disulfide (TXN-S2) back to thioredoxin (TXN-(SH)2) using NADPH. This system plays essential roles in a variety of cellular functions including redox control of transcription factors, deoxyribonucleotide synthesis, and cell growth [Karlenius &

Tonissen, 2010]. TXNRD1 is also capable of regenerating other antioxidant compounds such as ascorbic acid, selenium-containing substances, lipoic acid, and ubiquinone; and additionally supports α-tocopherol function [Nordberg & Arnér, 2001].

GFF induced ARE responses have been reported in additional models. Tsai et al. [2006] showed that GFF co-treatment resulted in the downregulation of inducible nitric oxide synthase (iNOS) expression and concurrent upregulation of HO1 expression in RAW264.7 mouse macrophages induced with lipopolysaccharide (LPS). Finlay et al. [2009] described potential modulation of the

103 ARE pathway in human keratinocytes and fibroblasts after comparing treatments with GFF and olive oil derivatives separately and in combination. HO1 expression was increased in a dose- dependent manner in both the primary cell cultures and human skin explant cultures when treated with GFF and/or olive oil derivatives. Yet, this group concluded that the ARE transcription demonstrated in ARE-32, a luciferase-based reporter cell line, was much lower than expected based on the magnitude of the HO1 upregulation observed in the primary cultures leading them to suspect involvement of hypoxia inducible factor 1 (HIF1). HO1 expression was also less affected by GFF in comparison to olive oil. Furue et al., [2017] reported individual and synergistic inhibition of ROS generation with GFF at 0.1% concentration and green tea flavonoid epigallocatechin gallate (EGCG) in human keratinocytes. Each compound could not overcome

TNFα induced ROS production alone; it was only effectively suppressed in combination [Furue et al., 2017].

GFF may have a special utility in the therapy of oxidative driven dyspigmentation disorders like vitiligo. Epidermal melanocytes are especially susceptible to significant amounts of oxidative stress attributable to their purposed function of melanin synthesis, which involves oxidation reactions and the generation of O2- and H2O2 [Denat et al., 2014]. There is extensive evidence establishing a significant role for oxidative stress in the etiology of melanocyte death in vitiligo.

Vitiligo melanocytes have increased sensitivity to UVB induced cell death; exaggerated sensitivity to chemical and physical oxidative stress; high levels of ROS (H2O2, ONOO-, iNOS) in the epidermis; and reduced levels of expression and activity of catalase [Denat et al., 2014, Boissy et al., 2012]. Recent scientific literature indicates there is impairment of the Nrf2-ARE pathway that contributes to redox imbalance and chronic oxidative stress found in vitiligo patients. Nrf2 is

104 considered an imperative super regulator of cellular homeostasis via transcriptional control of various detoxification genes [Baird & Dinkova-Kostova, 2011]. Nrf2-ARE regulation in skin is becoming a prominent scientific area of investigation for the development of medical strategies against oxidative driven diseases. There are still many ambiguities in the pathway targets, activation, and general metabolism per cellular model.

GFF exhibits substantial antioxidant properties that may potentially fortify epidermal melanocytes against ROS assault, thus thwarting the exacerbation of the oxidative damage component in disease pathophysiology. GFF also has antimelanogenic properties. Because there is a niche for melanin reduction and depigmentation strategies for vitiligo patients, GFF can be purposeful through providing concurrent melanin reduction. There is no cure currently available; hence, efficient prevention and remedial strategies that can halt the progression of oxidative stress in epidermal melanocytes are desirable. Other hyperpigmentary disorders such as solar lentigines, melasma, or post-inflammatory hyperpigmentation (PIH) may also benefit from the clinical properties of GFF [Lee, 2015; Bickers & Athar, 2006].

Another noteworthy consideration is the common adversities associated with the use and manufacture of synthetic medications. A sector of the scientific community is recommitting to natural products research for pharmacological benefits. The production and manufacture of naturally derived and sustainably sourced treatments continues to rise, being well received by the public. Vitiligo patients in particular also seek remedies involving natural compounds and methodology; yet, regulation, efficacy, and potency are still considerable challenges [Cohen et al.,

2015]. GFF is a filtered, yeast derived extract that exhibits comprehensive skin health benefits that

105 may have the potential to address such impediments. Our results have an important positive impact towards developing therapeutic agents that can prime the cellular environment for ROS assault, aiding in the prevention of oxidative driven disease and the sustainability of epidermal melanocytes in skin.

106 5.6 Acknowledgments

This research was supported by Procter & Gamble, Co. and the National Vitiligo Foundation, Inc.

107 5.7 Author Contributions

JKWC, AK, ALK, REB performed the research

JKWC, AK, ALK, TH, REB designed the research study

ALK, TH, REB contributed essential reagents or tools

JKWC, AK, ALK, REB analyzed the data

JKWC, REB wrote the paper

108 5.8 Figures and Tables

Figure 5.1 – GFF effect on cell viability in 4TBP challenged NHEM. NHEM were treated with or without 200 µM 4TBP alone or in combination with 2.5-10% GFF-B for 6 days. Cell viability was analyzed by MTT assay. Only 4TBP alone significantly diminished the cell viability compared to the ethanol control; thus co-treatment with GFF-B at all concentrations demonstrated a protective effect in NHEM. Fixation was used as a positive control. Results are expressed as a percentage of the control. *p<0.05, SEM

109

Figure 5.2 – GFF effect on 4TBP induced ROS generation. NHEM were challenged with 300

µM 4TBP alone or in combination with 10% GFF-B for 90 minutes. ROS were visualized with fluorescent microscopy. GFF-B suppressed ROS generation in NHEM exposed to 4TBP, an effect comparable to catalase, the negative control. Phase-contrast (above) and fluorescent (below) microscopy.

110

-2 5.3 GFF effect on UV induced H2O2 generation. NHEM were irradiated with UVB (105 mJ cm ) and then dosed with 10% GFF-B until harvest at the indicated timepoint (0-45 minutes). GFF-B reduced H2O2 generation in UVB irradiated NHEM at all timepoints. GFF-B also decreased endogenous H2O2 production in unirradiated NHEM at timepoints T0 and T15. Results are expressed

5 in H2O2 pMol/ 1 X 10 cells. *p<0.05 for UV 105 + 10% GFF-B compared to UV 105; ^p<0.05 for

UV 105 + 10% GFF-B compared to Control.

111

Figure 5.4 – GFF effect on various oxidative stress targets in NHEM [Modified from Morel

& Barouki, 1999]. RNA-seq transcriptome profiling analysis was performed with mRNA extracted from NHEM treated with 10% GFF-B treatment for 24 hours and 120 hours (5 days). At the 24 hours timepoint, GFF-B treatment modulated several gene targets associated with oxidative stress with fold change > 2; including MAPK10 = mitogen activated protein kinase 10; JUNB = proto-oncogene; CYBA = cytochrome B-245, alpha polypeptide; HMOX1 = heme oxygenase 1;

TXNRD1 = thioredoxin reductase 1; GPX3 = glutathione peroxidase 3; NQO1 = NAD(P)H quinone oxidoreductase 1; SOD3 = superoxide dismutase 3, extracellular.

112 Table 5.1 – RNA-seq antioxidant enzyme targets – 24 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

HMOX1 94.40 350.65 3.71 0.0003*

NQO1 1623.61 3328.54 2.05 0.024*

TXNRD1 652.28 1307.26 2.004 0.024*

Table 5.2 – RNA-seq antioxidant enzyme targets – 120 hours. Upregulation Downregulation Gene ID Control GFF-B Fold Change p-value

HMOX1 92.51 260.48 2.82 0.002*

NQO1 1327.87 3885.33 2.93 0.0002*

Table 5.1 and 5.2 – RNA-seq antioxidant enzyme targets. 10% GFF-B treatment for 24 hours and 120 hours (5 days) significantly upregulated mRNA levels of antioxidant enzymes from the

Nrf2-ARE pathway that are involved in ROS neutralization. Significance: Fold Change > 2,

*p<0.05

113

114 Figure 5.5 – GFF effect on Nrf2-ARE pathway. NHEM with varying pigmentary contents were treated with or without 10% GFF-A and 10% GFF-B for 5 or 7 days. At harvest, protein expression was analyzed by Western blot and densitometry. Polyubiquitinated Nrf2 was increased by 44% and 57%, respectively. HO1 and NQO1 expression was nearly doubled by 10% GFF-A and 10%

GFF-B. TXNRD1 was increased by 68% after treatment with 10% GFF-B. Results are expressed as a percentage of the control. *p<0.05, SEM

115

116 Figure 5.6 – GFF effect on Nrf2-ARE pathway in the cytoplasm and nucleus. NHEM was administered a single dose of 10% GFF-B for 8 hours. Cytoplasmic and nuclear fractions of the protein lysate were analyzed by Western blot and densitometry. Expression of Nrf2 was increased by 30% in the cytoplasm and the nucleus. Polyubiquitinated Nrf2 was also increased in cytoplasm by 41%. HO1 was highly inducible, showing 6-fold and 8-fold increases between the cytoplasm and nucleus.

117 5.9 Supplemental Figures and Tables

Figure 5.S1 – GFF effect on Nrf2-ARE pathway challenged by UVB. Lightly pigmented

NHEM were pretreated with 10% GFF-B for 6 days. Cells were irradiated with UVB (90 mJ cm-

2) while in PBS and then immediately rescued with or without 10% GFF-B for an additional 24 hours. At harvest, cells were processed with Western blot and densitometric analyses. NQO1 expression was maintained with GFF-B treatment post irradiation 2-fold over untreated, irradiated cells. HO1 was highly inducible with GFF-B treatment alone, yet expression was not maintained post irradiation. Results are expressed as a percentage of the control.

118

Figure 5.S2 – Cytoplasmic and nuclear HO1 expression. HO1 expression was increased in both the cytoplasm and nucleus of NHEM treated with 10% GFF-B for 8 hours. Results are expressed as a percentage of the control. *p<0.05

119

CHAPTER 6:

CONCLUSION

120 6.1 Major Findings

• GFF inhibits melanization in human epidermal melanocytes and melanoma. This is

accomplished in part by suppression of tyrosine hydroxylase activity that is not the result

of enzymatic competition with tyrosinase.

• ABCC9, TRPM6, TRPV1, SLC9A4, and SLC13A1 are putative ion transporters involved

in the regulation of pigmentation.

• TRPM6, TRPV1, and SLC9A4 are novel ion channels and transporters that associate with

the melanosome.

• GFF has effectual antioxidant properties, partly enacted through gene transcriptional

mediation of the Nrf2-ARE pathway in melanocytes.

• HO1, NQO1, and TXNRD1 are the primary antioxidant enzyme gene transcriptional

targets of GFF induced Nrf2 activation in melanocytes.

• GFF has consistent high inducibility of HO1 in various cell types, now including

melanocytes.

• GFF induces global gene downregulation and limited upregulation in melanocytes which

is predicted to be associated with its other regulatory mechanisms and epidermal benefits.

121 6.2 Synergy Between Melanization and Redox Properties of Galactomyces Ferment Filtrate

There is an ongoing research effort to understand how melanin content modulates redox biology.

Pigmentary phenotype is considered both a substantial factor in the cellular redox state and the epidermal protection level against UVR, which is characterized by the Fitzpatrick scale for skin color as phototypes I (light) – VI (dark) [Astner & Anderson, 2004]. Several research groups have published data on the connection between the pigmentary system and enzymatic antioxidant defense. Both CAT and TXNRD protein and/or activity were found to be positively correlated with increased pigmentation [Maresca et al., 2008; Maresca et al., 2006; Picardo et al., 1999;

Schallreuter et al., 1987]. Higher levels of SOD and CAT were also protected in higher pigmented reconstructed epidermal samples post UV irradiation [Maresca et al., 2006]. SOD/CAT ratio, an indicator for cellular susceptibility to external pro-oxidant stress, was determined higher in lower phototype melanocytes, inferring increased H2O2 accumulation [Picardo et al., 1999].

Furthermore, normal melanocytes exposed to peroxides mediated their extracellular H2O2 generation proportional to their melanin content, where melanin rich cells inhibited the initial H2O2 buildup [Meyskens Jr. et al., 1997].

Eumelanin/pheomelanin ratio is also a considerable factor in understanding pigmentation, its properties, and its effect on the redox state of the cell. The controversy of the pro-oxidant and antioxidant tendencies of each chromophore continue to be of high scientific interest.

Pheomelanogenesis is repeatedly associated with antioxidant depletion and high phototoxicity particularly upon UVR exposure [Napolitano et al., 2014]. Yet, eumelanin synthesis is biochemically favored and generates 2-fold greater levels of peroxide in addition to the DHI and

DHICA intermediates [Meyskens Jr. et al., 2001; Urabe et al., 1994; Pawelek & Lerner, 1978].

122 Higher eumelanin/pheomelanin ratios are correlated with resistance to oxidative stress [Kim et al.,

2015; Nasti & Timares, 2015; Diehl, 2014, Roulin et al., 2011]. Solving the conundrum of how the melanocytes maintain viability despite the nature of the pigment metabolism is a prodigious pursuit.

Regarding our study, GFF was investigated as a construct without specific knowledge of its full composition, removing that factor from influencing the research strategy. Nevertheless, inquisitive theorizing would lead to the question of how the composition works together to affect both the melanization and oxidative stress and/or how the melanization suppression impacts the redox state of the melanocytes and vice versa. Indubitably, there must be an apparent degree of synergy, especially considering the biological nature of the melanocyte.

Today, many flavonoids, polyphenolic compounds, are being pursued as nutraceuticals because of their ease of availability and broad range of health functions. A generous subset is also characterized as anti-melanogenic and operate mainly through transcriptional downregulation of

MITF and TYR, TYRP1, TYRP2 via cAMP and protein kinase A (PKA) signaling or demonstrate anti-proliferative mechanisms [Liu-Smith & Meyskens, 2016]. Other yeast derived or fermented broth extracts also report analogous results [Chan et al., 2014; Hwang et al., 2013]. Despite supposed functional parallels with GFF, the downregulation of melanogenic proteins was not observed in our studies, validated by various analytical methods and adjusted protocols. Yet, reduction in tyrosinase activity is a common thread [Chan et al., 2014; Ohgidani et al., 2012; Li et al., 2010]. Consequently, it is logical to speculate the effect of Nrf2-ARE response and its impact on melanogenesis.

123 Shin et al. [2014] proposed Nrf2-ARE activation negatively regulates melanogenesis through modulating PI3K/Akt signaling. In this study, primary NHEM cultures were transfected with a recombinant adenovirus expressing Flag-tagged Nrf2 to accomplish overexpression; this led to increased downstream targets HO1 and thioredoxin 1 (Trx1). Pigment quantity was decreased by

20%; TYR activity was reduced by 50-60%; and TYR and TYRP1 protein levels showed 10-30% reduction while TYRP2 was unaffected [Shin et al., 2014]. Nrf2 overexpression also inhibited

TYR promoter activity. The reverse was demonstrated by Keap1 overexpression mediated with recombinant adenovirus transfection, which resulted in inhibited Nrf2 nuclear translocation; increased melanin content by 18%; increased TYR activity by 31%; and upregulation of TYR and

TYRP1 protein expression. TYR promoter activity was still decreased but minimally at 11% [Shin et al., 2014]. Co-transduction relatively balanced the effects in melanin content, TYR expression, and TYR promoter activity. Further work was done to elucidate the signaling, which was found to involve PI3K/Akt pathway; PI3K inhibitors and Akt knockdown exhibited melanogenic effects.

Overall, this work deduced Nrf2 induced depigmentation is not modulated directly, but through intracellular signaling and the downstream effect of Nrf2 activation on the intracellular ROS that may impact the Akt signaling, similarly corroborated by other authors [Li et al., 2014; Chang,

2012; Tu et al., 2012; Lee et al., 2010; Oh et al., 2010; Kim et al., 2008; Lee et al., 2007; Khaled et al., 2003; Oka et al., 2000]. Shin et al. [2014] recommended the Nrf2-Keap1 pathway as a target for whitening agents.

Other connections between melanin synthesis and the antioxidant targets of this study have also been established. Lim et al. [2016] reported modulation of melanogenesis by HO1 via p53 in

NHEM cultures; these investigators used a supposed HO1 activator, protoporphyrin IX cobalt

124 chloride (CoPP), and HO1 inhibitor, protoporphyrin IX zinc (ZnPP), to modulate HO1 levels and melanogenic factors. Contradictorily, treatment with the inhibitor ZnPP increased HO1 expression and decreased MITF and TYR in a dose-dependent manner determined by the western blot analysis

[Lim et al., 2016]. Yet, the authors concluded HO1 induction increases MITF and TYR expression and inhibition does the reverse in their proposed model. Choi et al. [2010] claims that NQO1 positively regulates pigmentation. In this study, the NQO1 protein and activity levels were demonstrated to be proportional to TYR protein and pigmentation in normal melanocytes and several melanoma lines. NQO1 inhibitors decreased MITF, TYR, and TYRP1 expression in

NHEM and the pigmentation in zebrafish embryos; NQO1 overexpression increased tyrosinase protein and activity [Choi et al., 2010]. Even so, Yamaguchi et al. [2010] postulated that NQO1 could negatively regulate skin pigmentation by its antioxidant action via the increased production of intracellular hydroquinones, which are also synthetically mass produced as a hypopigmenting agent. This theory would be congruent with our results.

CYBA was an additional gene target identified in the RNA-seq data that may be worth further investigation. CYBA gene encodes for the alpha subunit of NADPH oxidases (NOXs). NOX can produce O2- via NADPH consumption. CYBA was downregulated with GFF treatment suggesting the decrease of NOX and intracellular O2- that would be supplementary to the antioxidant functions. Nakano et al. [2008] and Cross et al. [2004] suggested a connection with CYBA and

NOX activity in the regulation of pH. NOX on phagocytic vacuoles in the cytoplasm regulate pH

+ via the production of O2- that consumes the internal protons and encourages internal K accumulation in order to alkalize the internal pH [Cross et al., 2004]. Vacuoles also may have an indirect involvement in intracellular pH in plant and yeast cells [Gout et al., 1992; Klionsky et al.,

125 1990]. The ability of GFF to downregulate this protein would presumably disrupt NOX functionality leading to consequent pH alterations that may impact melanization. The effect of cellular pH of the melanocyte on melanin synthesis is another inquiry worthy of attention, especially in connection with ion gradients created by cell membrane transporters that work in consort with melanosomes [Bellono & Oancea, 2014; Sarangarajan et al., 2001]. Although no strong predictions can be brought forth at this time, these studies altogether encourage the suppositions that an increase of endogenous antioxidants is likely to have an impact on the melanization in melanocytes. CYBA downregulation may also be another factor included in the assembly of inhibitory mechanisms achieved by GFF in both melanization and ROS generation.

126 6.3 Future Recommendations

• Determine by various mechanisms the effect of GFF on the pH of melanosomes and clearly

delineate alterations between treated and untreated groups. Further characterize the

collective manner of inhibition and how GFF impacts the eumelanin/pheomelanin ratio.

• Test the influence of the new pore protein targets in modulation of the melanosome through

silencing and other controlled or specific modifications to understand the extent of their

impact or regulatory influence in melanin synthesis. Expand the investigation to include

examination of CYBA in potential relation to cellular/organelle pH.

• Assess similar bioassays and outcomes in vitiligo melanocyte cultures for comparability of

antioxidant rescue mechanisms in order to validate the project objectives and continued

study.

• Analyze the molecular components of GFF to map specific ingredients with the effects

observed in the melanocyte, particularly HO1 inducibility. Additionally, advance to tissue

models for complexity and to understand how skin barrier filtering impacts delivery of the

activating components within the composition and the strength of impact compared to

results uncovered in the direct delivery cellular culture system.

• Initiate clinical trials with GFF to assess its effectivity in various dyspigmentary diseases

such as melasma, post-inflammatory hyperpigmentation (PIH), solar lentigines, and

vitiligo.

127

REFERENCES

128 Akoglu G, Metin E, Akbas A, Yorulmaz A, Isikoglu S, Sener S, Kilinc F. (2013). Evaluation of

total oxidant and antioxidant status in localized and generalized vitiligo. Clin Exp

Dermatol, 38, 701-706.

Ambrosio AL, Boyle JA, Aradi AE, Christian KA, Di Pietro SM. (2016). TPC2 controls

pigmentation by regulating melanosome pH and size. Proc Natl Acad Sci, 113(20), 5622-

5627.

American Academy of Dermatology, Inc. (2017, October). Vitiligo: who gets and causes.

Retrieved from American Academy of Dermatology:

https://www.aad.org/public/diseases/color-problems/vitiligo#causes

Amin IM, AbuZeid OM, Rashed LA. (2013). Tissue level of nuclear factor-erythroid2-related

factor2 and melanocyte-stimulating hormone in vitiligo patients. J Egypt Women Dermatol

Soc, 10, 89-93.

Ancans J, Tobin DJ, Hoogduijn MJ, Smit NP, Wakamatsu K, Thody AJ. (2001). Melanosomal pH

controls rate of melanogenesis, eumelanin/pheomelanin ratio and melanosome maturation

in melanocytes and melanoma cells. Exp Cell Res, 268, 26-35.

Anders S, Huber W. (2010). Differential expression analysis for sequence count data. Genome

Biol, 11, R106.

Arena EA, Longo WE, Roberts KE, Geibel P, Nateqi J, Brandstetter M, Geibel JP. (2012).

Functional role of NHE4 as a pH regulator in rat and human colonic crypts. Am J Physiol

Cell Physiol, 302, C412-C418.

129 Arican O, Kurutas EB. (2008). Oxidative stress in the blood of patients with active localized

vitiligo. Acta Dermatoven APA, 17, 12-16.

Astner S, Anderson RR. (2004). Skin phototypes 2003. J Invest Dermatol, 122, xxx-xxxi.

Baird L, Dinkova-Kostova AT. (2011). The cytoprotective role of Keap1-Nrf2 pathway. Arch

Toxicol, 85, 241-272.

Bellono NW, Escobar IE, Oancea E. (2016). A melanosomal two-pore sodium channel regulates

pigmentation. Sci Rep, 6(26570).

Bellono NW, Kammel LG, Zimmerman AL, Oancea E. (2013). UV light phototransduction

activates transient receptor potential A1 ion channels in human melanocytes. Proc Natl

Acad Sci USA, 110, 2383-2388.

Bellono NW, Oancea EV. (2014). Ion transport in pigmentation. Arch Biochem Biophys, 563, 35-

41.

Beltrán AR, Ramírez MA, Carraro-Lacroix LR, Hiraki Y, Rebouças NA, Malnic G. (2008). NHE1,

NHE2, and NHE4 contribute to regulation of cell pH in T84 colon cancer cells. Pflugers

Arch, 455(5), 799-810.

Bickers DR, Athar M. (2006). Oxidative stress in the pathogenesis of skin disease. J Invest

Dermatol, 126, 2565-2575.

Birben E, Sahiner UM, Sackesen C, Erzurum S, Kalayci O. (2012). Oxidative stress and

antioxidant defense. World Allergy Organ J, 5, 9-19.

Boissy RE, Dell'Anna ML, Picardo M. (2012). On the pathophysiology of vitiligo: possible

treatment options. Indian J Dermatol Venereol Leprol, 78, 24-29.

130 Boissy RE, Liu YY, Medrano EE, Nordlund JJ. (1991). Structural abberation of the rough

endoplasmic reticulum and melanosome compartmentalization in long-term cultures of

melanocytes from vitiligo patients. J Invest Dermatol, 97, 395-404.

Briganti S, Picardo M. (2003). Antioxidant activity, lipid peroxidation and skin diseases. What's

new. J Eur Acad Dermatol Venereol, 17, 663-669.

Bustamante J, Bredeston L, Malanga G, Mordoh J. (1993). Role of melanin as a scavenger of

active oxygen species. Pigment Cell Res, 6, 348-353.

Chalupa A, Koshoffer A, Galan E, Yu L, Liu FT, Boissy RE. (2015). Melanocytic galectin-3 is

associated with tyrosinase-related protein-1 and pigment biosynthesis. J Invest Dermatol,

135, 202-211.

Chan C-F, Huang C-C, Lee M-Y, Lin Y-S. (2014). Fermented broth in tyrosinase- and

melanogenesis inhibition. Molecules, 19, 13122-13135.

Chang T-S. (2012). Natural melanogenesis inhibitors acting through the down-regulation of

tyrosinase activity. Materials, 5, 1661-1685.

Cheli Y, Luciani F, Khaled M, Beuret L, Bille K, Gounon P, Ortonne J-P, Bertolotto C, Ballotti

R. (2009). aMSH and cyclic AMP elevating agents control melanosome pH through a

protein kinase A-independent mechanism. J Biol Chem, 284, 18699-18706.

Chen KG, Leapman RD, Zhang G, Lai B, Valencia JC, Cardarelli CO, Vieira WD, Hearing VJ,

Gottesman MM. (2009a). Influence of melanosome dynamics on melanoma drug

senstivity. J Natl Cancer Inst, 101, 1259-1271.

131 Chen KG, Valencia JC, Gillet J-P, Hearing VJ, Gottesman MM. (2009b). Involvement of ABC

transporters in melanogenesis and the development of multidrug resistance of melanoma.

Pigment Cell Melanoma Res, 22, 740-749.

Choi T-Y, Sohn K-C, Kim J-H, Kim S-M, Kim C-H, Hwang J-S, Lee J-H, Kim CD, Yoon T-J.

(2010). Impact of NAD(P)H:quinone oxidoreductase-1 on pigmentation. J Invest

Dermatol, 130, 784-792.

Cichorek M, Wachulska M, Stasiewicz A, Tyminska A. (2013). Skin melanocytes: biology and

development. Postep Derm Alergol, XXX, 30-41.

Cohen BE, Elbuluk N, Mu EW, Orlow SJ. (2015). Alternative systemic treatments for vitiligo: a

review. Am J Clin Dermatol, 463-474.

Colucci R, Dragoni F, Conti R, Pisaneschi L, Lazzeri L, Moretti S. (2015a). Evaluation of an oral

supplement containing Phyllanthus emblica fruit extracts, vitamin E, and carotenoids in

vitiligo treatment. Dermatol Ther, 28, 17-21.

Colucci R, Dragoni F, Moretti S. (2015b). Oxidative stress and immune system in vitiligo and

thyroid diseases. Oxid Med Cell Longev, 2015, 631927-631933.

Cross AR, Segal AW. (2004). The NADPH oxidase of professional phagocytes-prototype of the

NOX electron transport chain systems. Biochim Biophys Acta, 1657, 1-22.

Dammak I, Boudaya S, Ben Abdallah F, Turki H, Attia H, Hentati B. (2009). Antioxidant enzymes

and lipid peroxidation at the tissue level in patients with stable and active vitiligo. Int J

Dermatol, 48, 467-480.

132 Dell'Anna ML, Maresca V, Briganti S, Camera E, Falchi M, Picardo M. (2001). Mitochondrial

impairment in peripheral blood mononuclear cells during the active phase of vitiligo. J

Invest Dermatol, 117, 908-913.

Dell'Anna ML, Mastrofrancesco A, Sala R, Venturini M, Ottaviani M, Vidolin AP, Leone G,

Calzavara PG, Westerhof W, Picardo M. (2007). Antioxidant and narrow band-UVB in the

treatment of vitiligo: a double-blind placebo controlled trial. Clin Exp Dermatol, 32, 631-

636.

Denat L, Kadekaro AL, Marrot L, Leachman SA, Abdel-Malek ZA. (2014). Melanocytes as

instigators and victims of oxidative stress. J Invest Dermatol, 134, 1512-1518.

Devi S, Kedlaya R, Maddodi N, Bhat KM, Weber CS, Valdivia H, Setaluri V. (2009). Calcium

homeostasis in human melanocytes: role of transient receptor potential melastatin 1

(TRPM1) and its regulation by ultraviolet light. Am J Physiol Cell Physiol, 297, C679-

C687.

DiCiccio JE, Steinberg BE. (2011). Lysosomal pH and analysis of the counter ion pathways that

support acidification. J Gen Physiol, 137, 385-390.

Diehl C. (2014). Melanocytes and oxidative stress. J Pigment Disord, 1(4), 127-132.

Dooley CM, Schawarz H, Mueller KP, Mongera A, Konantz M, Neuhauss SC, Nüsslein-Volhard

C, Geisler R. (2013). Slc45a2 and V-ATPase are regulators of melanosomal pH

homeostasis in zebrafish, providing a mechanism for human pigment evolution and

disease. Pigment Cell Melanoma Res, 26, 205-217.

133 Elassiuty YE, Klarquist J, Speiser J, Yousef RM, Refaee AAEL, Hunter NS, Shaker OG, Gundeti

M, Nieuweboer-Krobotova L, Le Poole IC. (2011). Heme oxygenase-1 expression protects

melanocytes from stress-induced cell death: implications for vitiligo. Exp Dermatol, 20,

496-501.

Espinosa-Diez C, Miguel V, Mennerich D, Kietzmann T, Sánchez-Pérez P, Cadenas S, Lamas S.

(2015). Antioxidant responses and cellular adjustments to oxidative stress. Redox Biol,

6(2015), 183-197.

Finlay DR, Date A, Matheny HE, Mullins LA, Jarrold BB, Osborne R. (2009). Skin biomarkers

confirm the antioxidant activity of olive derivaties and yeast ferment filtrate. J Am Acad

Dermatol, 60(3), AB81.

Frederiks WM, Vreeling-Sindelárová H, Van Noorden CJF. (2007). Loss of peroxisomes causes

oxygen insensitivity of the histochemical assay of glucose-6-phosphate dehydrogenase

activity to detect cancer cells. J Histochem Cytochem, 55, 175-181.

Furue M, Uchi H, Mitoma C, Hashimoto-Hachiya A, Chiba T, Ito T, Nakahara T, Tsuji G. (2017).

Antioxidants for healthy skin: the emerging role of aryl hydrocarbon receptors and nuclear

factor-erythroid 2-related factor-2. Nutrients, 223(9), 1-11.

Gęgotek A, Skrzydlewska E. (2015). The role of transcription factor Nrf2 in skin cells metabolism.

Arch Dermatol Res, 307, 385-396.

Geisbrecht BV, Gould SJ. (1999). The human PICD gene encodes a cytoplasmic and peroxisomal

NADP+-dependent isocitrate dehydrogenase. J Biol Chem, 274, 30527-30533.

134 Gout E, Bligny R, Douce R. (1992). Regulation of intracellular pH values in higher plant cells.

Carbon-13 and phosphorus-31 nuclear magnetic resonance studies. J Biol Chem, 267(20),

13903-13909.

Graham DG, Tiffany SM, Vogel FS. (1978). The toxicity of melanin precursors. J Invest Dermatol,

70, 113-116.

Graves AR, Curran PK, Smith CL, Mindell JA. (2008). The Cl-/H+ antiporter ClC-7 is the primary

chloride permeation pathway in lysosomes. Nature, 453, 788-792.

Grochot-Przeczek A, Lach R, Mis J, Skrzypek K, Gozdecka M, Sroczynska P, Dubiel M,

Rutkowski A, Kozakowska M, Zagorska A, Walczynski J, Was H, Kotlinowski J, Drukala

J, Kurowski K, Kieda C, Herault Y, Dulak J, Jozkowicz A. (2009). Heme oxygenase-1

accelerates cutaneous wound healing in mice. PLos ONE, 4(6), e5803.

Guan C-P, Zhou M-N, Xu A-E, Kang K-F, Liu J-F, Wei X-D, Li Y-W, Zhao D-K, Hong W-S.

(2008). The susceptibility to vitiligo is associated with NF-E2-related factor2 (Nrf2) gene

polymorphisms: a study on Chinese Han population. Exp Dermatol, 17, 1059-1062.

Günther W, Lüchow A, Cluzeaud F, Vandewalle A, Jentsch TJ. (1998). ClC-5, the chloride

channel mutated in Dent's disease, colocalizes with the proton pump in endocytotically

active kidney cells. Pro Natl Acad Sci, 95, 8075-8080.

Hakozaki T, Date A, Yoshii T, Toyokuni S, Yasui H, Sakurai H. (2008). Visualization and

characterization of UVB-induced reactive oxygen species in a human skin equivalent

model. Arch Dermatol Res, 300, S51-S56.

135 Hakozaki T, Laughlin T, Wakamatsu K, Boissy RE. (2009). The effect of yeast ferment filtrate on

melanogenesis via suppressing expression of tyrosinase related proteins, pmel-17 and

melanin polymerization. J Invest Dermatol, 129, S138.

Haltaufderhyde KD, Oancea E. (2014). Genome-wide transcriptome analysis of human epidermal

melanocytes. Genomics, 104, 482-489.

Hattori K, Date A, Nakajima A, Ishida K, Yan X, Yoshii T, Tamura H. (2010). Effects of

galactomyces ferment filtrate on epidermal barrier marker caspase-14 in human skin cells.

J Am Acad Dermatol, 62(3), AB54.

Hwang J-A, Goh M-J, Kim E-J, Lee M-R, Park N-H, Na Y-J, Cho J-C, Lee H-K. (2013).

Identification of sake extract as a new anti-melanogenic ingredient by in vitro and clinical

trials. Nat Prod Commun, 8(11), 1605-1608.

Imlay JA. (2008). Cellular defenses against superoxide and hydrogen peroxide. Ann Rev Biochem,

77, 755-776.

Imlay JA. (2013). The molecular mechanisms and physiological consequences of oxidative stress:

lessons from a model bacterium. Nat Rev Microbiol, 11, 443-454.

Iuga AO. Lerner EA. (2007). TRP-ing up melanophores: TRPM7, melanin synthesis, and pigment

cell survival. J Invest Dermatol, 127, 1855-1856.

Jablonksi NG, Chaplin G. (2010). Human skin pigmentation as an adaptation to UV radiation.

PNAS, 107, 8962-8968.

136 Jain D, Misra R, Kumar A, Jaiswal G. (2008). Levels of malondialdehyes and antioxidants in the

blood of patients with vitiligo of age group 11-20 years. Indian J Physiol Pharmacol, 52,

297-301.

Jenkins NC, Grossman D. (2013). Role of melanin in melanocyte dysregulation of reactive oxygen

species. BioMed Res Int, 2013, 1-3.

Jenkins NC, Tong L, Cassidy P, Leachman SA, Boucher KM, Goodson AG, Samadashwily G,

Grossman D. (2011). The p16(INK4A) tumor suppressor regulates cellular oxidative

stress. Oncogene, 30, 265-274.

Jian Z, Li K, Liu L, Zhang Y, Zhou Z, Li C, Gao T. (2011). Heme oxygenase-1 protects human

melanocytes from H2O2-induced oxidative stress via the Nrf2-ARE pathway. J Invest

Dermatol, 131, 1420-1427.

Jian Z, Li K, Song P, Zhu G, Zhu L, Cui T, Liu B, Tang L, Wang X, Wang G, Gao T, Li C. (2014).

Impaired activation of the Nrf2-ARE signaling pathway undermines H2O2-induced

oxidative stress response: a possible mechanism for melanocyte degeneration in vitiligo. J

Invest Dermatol, 134, 2221-2230.

Jimbow K. (1995). Current update and trends in melanin pigmentation and melanin biology. Keio

J Med, 44, 9-18.

Jones PM, O'Mara ML, George AM. (2009). ABC transporters: a riddle wrapped in a mystery

inside an enigma. Trends Biochem Sci, 34, 520-531.

Kadekaro AL, Kavanagh R, Kanto H, Terzieva S, Hauser J, Kobayashi N, Schwemberger S,

Cornelius J, Babcock G, Shertzer HG, Scott G, Abdel-Malek ZA. (2005). α-Melanocortin

137 and endothelin-1 activate antiapoptotic pathways and reduce DNA damage in human

melanocytes. Cancer Res, 65(10), 4292-4299.

Kadekaro AL, Leachman S, Kavanagh RJ, Swope V, Cassidy P, Supp D, Sartor M, Schwemberger

S, Boissy RE, Manga P, Sturm RA, Abdel-Malek ZA. (2010). Melanocortin 1 receptor

genotype: an important determinant of the damage response of melanocytes to ultraviolet

radiation. FASEB J, 24, 3850-3860.

Karlenius TC, Tonissen KF. (2010). Thioredoxin and cancer: a role for thioredoxin in all states of

tumor oxygenation. Cancers, 2, 209-232.

Khaled M, Larribere L, Bille K, Ortonne J-P, Ballotti R, Bertolotto C. (2003). Microphthalmia

associated transcription factor is a target of the phosphatidylinositol-3-kinase pathway. J

Invest Dermatol, 121, 831-836.

Khan R, Satyam A, Gupta S, Sharma VK, Sharma A. (2009). Circulatory levels of antioxidants

and lipid peroxidation in Indian patients with generalized and localized vitiligo. Arch

Dermatol Res, 301, 731-737.

Khan SA, Bank J, Song DH, Choi EA. (2015). The Skin and Subcutaneous Tissue. In A. D.

Brunicardi F (Ed.), Schwart'z Principles of Surgery (10th ed.). New York, New York:

McGraw-Hill.

Kim E, Panzella L, Micillo R, Bentley WE, Napolitano A, Payne GF. (2015). Reverse engineering

applied to red human hair pheomelanin reveals redox-buffering as a pro-oxidant

mechanism. Sci Rep, 5(18447), 1-14.

138 Kim JH, Baek SH, Kim DH, Choi TY, Yoon TJ, Hwang JS, Kim MR, Kwon HJ, Lee CH. (2008).

Downregulation of melanin synthesis by haginin A and its application to in vivo lightening

model. J Invest Dermatol, 128(5), 1227-1235.

Kim SY, Park SC. (2012). Physiological antioxidative network of the bilirubin system in aging

and age-related diseases. Front Pharmacol, 3, 1-9.

Klionsky DJ, Herman PK, Emr SD. (1990). The fungal vacuole: composition, function, and

biogenesis. Microbiol Rev, 54(3), 266-292.

Koca R, Armutcu F, Altinyazar HC, Gürel A. (2004). Oxidant-antioxidant enzymes and lipid

peroxidation in generalized vitiligo. Clin Exp Dermatol, 29, 406-409.

Koga S, Nakano M, Tero-Kubota S. (1992). Generation of superoxide during the enzymatic action

of tyrosinase. Arch Biochem Biophys, 292, 570-575.

Kokot A, Metze D, Mouchet N, Galibert MD, Schiller M, Luger TA, Böhm M. (2009). α-

Melanocyte-stimulating hormone counteracts the suppressive effect of UVB on Nrf2 and

Nrf-dependent gene expression in human skin. Endocrinology, 150, 3197-3206.

Koshoffer A, Boissy RE. (2014). Current understanding of the etiology of vitiligo. Curr Derm

Rep, 3, 1-5.

Kovtunovych G, Eckhaus MA, Ghosh MC, Ollivierre-Wilson H, Rouault TA. (2010). Dysfunction

of the heme recycling system in heme oxygenase 1-deficient mice: effects on macrophage

viability and tissue iron distribution. Blood, 116(26), 6054-6062.

139 Laddha NC, Dwivedi M, Gani AR, Shajil EM, Rasheedunnisa B. (2013a). Involvement of

superoxide dismutase isoenzymes and their genetic variants in progression of and higher

susceptibility to vitiligo. Free Radic Biol Med, 65, 1110-1125.

Laddha NC, Dwivedi M, Mansuri MS, Gani AR, Ansarullah M, Ramachandran AV, Dalai S,

Begum R. (2013b). Vitiligo: interplay between oxidative stress and immune system. Exp

Dermatol, 22, 245-250.

Laddha NC, Dwivedi M, Mansuri MS, Singh M, Gani AR, Yeola AP, Panchal VN, Khan F, Dave

DJ, Patel A, Madhavan SE, Gupta R, Marfatia Z, Marfatia YS, Begum R. (2014). Role of

oxidaitve stress and autoimmunity in onset and progression of vitiligo. Exp Dermatol, 23,

352-354.

Le Poole IC, Boissy RE, Sarangarajan R, Chen J, Forristal JJ, Sheth P, Westerhof W, Babcock G,

Das PK, Saelinger CB. (2000). PIG3V, an immoratlized human vitiligo melanocyte cell

line, expresses dilated endoplasmic reticulum. In Vitro Cell Dev Biol Anim, 36, 309-319.

Lee A-Y. (2015). Recent progress in melasma pathogenesis. Pigment Cell Melanoma Res, 28, 648-

660.

Lee J, Jung K, Kim YS, Park D. (2007). Diosgenin inhibits melanogenesis through the activation

of phosphatidylinositol-3-kinase pathway (PI3K) signaling. Life Sci, 81(3), 249-254.

Lee J-H, Jang J-Y, Park C, Kim B-W, Choi Y-H, Choi B-T. (2010). Curcumin suppresses α-

melanocyte stimulating hormone-stimulated melanogenesis in B16F10 cells. Int J Mol

Med, 26, 101-106.

140 Li L, Dong H, Song E, Xu X, Liu L, Song Y. (2014). Nrf2/ARE pathway activation, HO-1 and

NQO1 induction by polychlorinated biphenyl quinone is associated with reactive oxygen

species and PI3K/AKT signaling. Chem Biol Interact, 209, 56-67.

Li Z, Lee J, Cho MH. (2010). Antioxidant, antibacterial, tyrosinase inhibitory, and biofilm

inhibitory activities of fermented rice bran broth with effective microorganisms.

Biotechnol Bioprocess Eng, 15, 139-144.

Lim H-S, Jin S, Yun SJ. (2016). Modulation of melanogenesis by heme oxygenase-1 via p53 in

normal human melanocytes. Chonnam Med J, 52(1), 45-52.

Lin Q, Weis S, Yang G, Weng Y-H, Helston R, Rish K, Smith A, Bordner J, Polte T, Gaunitz F,

Dennery PA. (2007). Heme oxygenase-1 protein localizes to the nucleus and activates

trancription factors important in oxidative stress. J Biol Chem, 282, 20621-20633.

Liu-Smith F, Meyskens FL. (2016). Molecular mechanisms of flavonoids in melanin synthesis and

the potential for the prevention and treatment of melanoma. Mol Nutr Food Res, 60, 1264-

1274.

Luo Y, Henle ES, Linn S. (1996). Oxidative damage to DNA constituents by iron-mediated fenton

reactions: the deoxycytidine family. J Biol Chem, 271(35), 21167-21176.

Manga P, Sheyn D, Yang F, Sarangarajan R, Boissy RE. (2006). A role for tyrosinase-related

protein 1 in 4-tert-butylphenol-induced toxicity in melanocytes: implications for vitiligo.

Am J Pathol, 169(5), 1652-1662.

Mani I, Sharma V, Tamboli I, Raman G. (2001). Interaction of melanin with proteins-the

importance of an acidic intramelanosomal pH. Pigment Cell Res, 14, 170-179.

141 Maresca V, Flori E, Briganti S, Camera E, Cario-André M, Taieb A, Picardo M. (2006). UVA-

induced modification of catalase charge properties in the epidermis is correlated with the

skin phototype. J Invest Dermatol, 126, 182-190.

Maresca V, Flori E, Briganti S, Mastrofrancesco A, Fabbri C, Mileo AM, Paggi MG, Picardo M.

(2008). Correlation between melanogenic and catalase activity in in vitro human

melanocytes: a synergic strategy against oxidative stress. Pigment Cell Melanoma Res, 21,

200-205.

Marks MS, Seabra MC. (2001). The melanosome: membrane dynamics in black and white. Nat

Rev Mol Cell Biol, 2, 1-11.

Marrot L, Jones C, Perez P, Meunier J-R. (2007). The significance of Nrf2 pathway in (photo)-

oxidative stress response in melanocytes and keratinocytes of the human epidermis.

Pigment Cell Melanoma Res, 21, 79-88.

Matés JM, Pérez-Gómez C, Núñez de Castro I. (1999). Antioxidant enzymes and human diseases.

Clin Biochem, 32, 595-603.

Mescher, AL. (2009). Skin. In Junqueira's Basic Histology: Text and Atlas (p. 316). New York,

New York: McGraw-Hill Medical.

Meyskens Jr. FL, Farmer P, Fruehauf JP. (2001). Redox regulation in human melanocytes and

melanoma. Pigment Cell Res, 14, 148-154.

Meyskens Jr. FL, van Chau H, Tohidian N, Buckmeier J. (1997). Luminol-enhanced

chemiluminescent response of human melanocytes and melanoma cells to hydrogen

peroxide stress. Pigment Cell Res, 10, 184-189.

142 Mindell JA. (2012). Lysosomal acidification mechanisms. Annu Rev Physiol, 74, 69-86.

Mohammad-Panah R, Harrison R, Dhani S, Ackerley C, Huan L-J, Wang Y, Bear CE. (2003). The

chloride channel ClC-4 contributes to endosomal acidification and trafficking. J Biol

Chem, 278, 29267-29277.

Morel Y, Barouki R. (1999). Repression of gene expression by oxidative stress. Biochem J, 342,

481-496.

Nakajima H, Nagata T, Koga S, Imokawa G. (2013). Reduced glutathione disrupts the intracellular

trafficking of tyrosinase and tyrosinase-related protein-1 but not dopachrome tautomerase

and Pmel17 to melanosomes, which results in the attenuation of melanization. Arch

Dermatol Res, 306, 37-49.

Nakamura N, Tanaka S, Teko Y, Mitsui K, Kanazawa H. (2005). Four Na+/H+ exchanger isoforms

are distributed to golgi and post-golgi compartments and are involved in organelle pH

regulation. J Biol Chem, 280(2), 1561-1572.

Nakano Y, Longo-Guess CM, Bergstrom DE, Nauseef WM, Jones SM, Bánfi B. (2008). Mutation

of the Cyba gene encoding p22phox causes vestibular and immune defects in mice. J Clin

Invest, 118(3), 1176-1185.

Napolitano A, Panzella L, Monfrecola G, d'Ischia M. (2014). Pheomelanin-induced oxidative

stress: bright and dark chemistry bridging red hair phenotype and melanoma. Pigment Cell

Melanoma Res, 27, 721-733.

143 Nappi AJ, Vass E. (1996). Hydrogen peroxide generation associated with the oxidations of the

eumelanin precursors 5,6-dihydroxyindole and 5,6-dihydroxyindole-2-carboxylic acid.

Melanoma Res, 6, 341-349.

Nasti TH, Timares L. (2015). Invited review MC1R, eumelanin and pheomelanin: their role in

determining the susceptibility to skin cancer. Photochem Photobiol, 91(1), 188-200.

Natarajan VT, Singh A, Kumar AA, Sharma P, Kar HK, Marrot L, Meunier JR, Natarajan K, Rani

R, Gokhale RS. (2010). Transcriptional upregulation of Nrf2-dependent phase II

detoxification genes in the involved epidermis of vitiligo vulgaris. J Invest Dermatol, 130,

2781-2789.

National Vitiligo Foundation, Inc. (2017, October). Vitiligo basics. Retrieved from National

Vitiligo Foundation Inc (NVFI) HOME:

http://nationalvitiligofoundation.wildapricot.org/about_vitiligo#What_Is_Vitiligo

Nimse SB, Pal D. (2015). Free radicals, natural antioxidants, and their reaction mechanisms. RSC

Adv, 5, 27986-28006.

Nisar ME, Parsons KS, Bian CX, Zhong JL. (2015). UVA irradiation induced heme oxygenase-1:

a novel phototherapy for morphea. Photochem Photobiol, 91(1), 210-220.

Nordberg J, Arnér ESJ. (2001). Reactive oxygen species, antioxidants, and the mammalian

thioredoxin system. Free Radic Biol Med, 31, 1287-1312.

Nordlund JJ, Boissy RE. (2001). The biology of melanocytes. In D. Woodley, & R. Freindel (Eds.),

The Biology of Skin (pp. 113-131). Pearl River, NY: Parthenon Publishing.

144 Novellino L, Napolitano A, Prota G. (1999). 5,6-Dihydroxyindoles in the fenton reaction: a model

study of the role of melanin precursors in oxidative stress and hyperpigmentary processes.

Chem Res Toxicol, 12, 985-992.

Oancea E, Vriens J, Brauchi S, Jun J, Splawski I, Clapham DE. (2009). TRPM1 forms ion channels

associated with melanin content in melanocytes. Sci Signal, 2, ra21.

Oh EY, Jang JY, Choi YH, Choi YW, Choi BT. (2010). Inhibitory effects of 1-O-methyl-

fructofuranose from Schisandra chinensis fruit on melanogenesis in B16F0 melanoma

cells. J Enthnopharmacol, 132(1), 219-224.

Ohgidani M, Komizu Y, Goto K, Ueoka R. (2012). Antimelanogenic and antioxidative effects of

residual powders from Shochu distillation remnants. Food Chem, 132, 2140-2143.

Oka M, Nagai H, Ando H, Fukunaga M, Matsumura M, Araki K, Ogawa W, Miki T, Sakaue M,

Tsukamoto K, Konishi H, Kikkawa U, Ichihashi M. (2000). Regulation of melanogenesis

through phosphatidylinositol 3-kinase-Akt pathway in human G361 melanoma cells. J

Invest Dermatol, 115(4), 699-703.

Okubo A, Yasuhira S, Shibazaki M, Takahashi K, Akasaka T, Masuda T, Maesawa C. (2016).

NAD(P)H dehydrogenase, quinone 1 (NQO1), protects melanin-producing cells from

cytotoxicity of rhododendrol. Pigment Cell Melanoma Res, 29, 309-316.

Orlow SJ. (1995). Melanosomes are specialized members of the lysosomal lineage of organelles.

J Invest Dermatol, 105, 3-7.

145 Osborne R, Hakozaki T, Jarrold B, Mullins L. (2007). Stimulation of epidermal hyaluronan by

yeast ferment. 21st World Congress of Dermatology, Buenas Aires, September 30-October

5.

Pae H-O, Oh G-S, Choi B-M, Chae S-C, Kim Y-M, Chung K-R, Chung H-T. (2004). Carbon

monoxide produced by heme oxygenase-1 suppresses T cell proliferation via inhibition of

IL-2 production. J Immunol, 172, 4744-4751.

Panda SK. (2012). Assay guided comparison for enzymatic and non-enzymatic activities with

special reference to medicinal plants. In E.-M. MA (Ed.), Antioxidant enzyme (pp. 381-

400). Intech.

Park S & Imlay JA. (2003). High levels of intracellular cysteine promote oxidative DNA damage

by driving the fenton reaction. J Bacteriol, 185(6), 1942-1950.

Pawelek JM, Lerner AB. (1978). 5,6-Dihydroxyindole is a melanin precursor showing potent

cytotoxicity. Nature, 276, 627-628.

Pelle E, Huang X, Zhang Q, Pernodet N, Yarosh DB, Frenkel K. (2014). Increased endogenous

DNA oxidation correlates to increased iron levels in melanocytes relative to keratinocytes.

J Cosmet Sci, 65, 277-284.

Percival M. (1998). Antioxidants. Clinical Nutrition Insights, 30, 1-4.

Picardo M, Maresca V, Eibenschutz L, De Bernardo C, Rinaldi R, Grammatico P. (1999).

Correlation between antioxidants and phototypes in melanocytes cultures. A possible link

of physiologic and pathologic relevance. J Invest Dermatol, 113, 424-425.

146 Poss KD, Tonegawa S. (1997). Heme oxygenase 1 is required for mammalian iron reutilization.

Proc Natl Acad Sci USA, 94, 10919-10924.

Procter & Gamble, Co. (2017, August). High end skin care & beauty products. Retrieved from

SK-II Official Site: https://www.sk-ii.com

Puri N, Gardner JM, Brilliant MH. (2000). Aberrant pH of melanosomes in pink-eyed dilution (p)

mutant melanocytes. J Invest Dermatol, 115, 607-613.

Qiu L, Song Z, Setaluri V. (2014). Oxidative stress and vitiligo: the Nrf2-ARE signaling

connection. J Invest Dermatol, 134, 2074-2076.

Raposo G, Marks MS. (2002). The dark side of lysosome-related organelles: specialization of the

endocytic pathway for melanosome biogenesis. Traffic, 3, 237-248.

Raposo G, Tenza D, Murphy DM, Berson JF, Marks MS. (2001). Distinct protein sorting and

localization to premelanosomes, melanosomes, and lysosomes in pigmented melanocytic

cells. J Cell Biol, 152, 809-823.

Ross D, Kepa JK, Winski SL, Beall HD, Anwar A, Siegel D. (2000). NAD(P)H:quinone

oxidoreductase 1 (NQO1): chemoprotection, bioactivation, gene regulation and genetic

polymorphisms. Chem Biol Interact, 129, 77-97.

Roulin A, Almasi B, Meichtry-Stier KS, Jenni L. (2011). Eumelanin- and pheomelanin-based

colour advertise resistance to oxidative stress in opposite ways. J Evol Biol, 24(10), 2441-

2447.

Salem MM, Shalbaf M, Gibbons NC, Chavan B, Throton JM, Schallreuter KU. (2009). Enhanced

DNA binding capacity on up-regulated epidermal wilde-type p53 in vitiligo by H2O2-

147 mediated oxidation: a possible repair mechanism for DNA damage. FASEB J, 23, 3790-

3807.

Sarangarajan R, Shumaker H, Soleimani M, Le Poole C, Boissy RE. (2001). Molecular and

functional characterization of sodium-hydrogen exchanger in skin as well as cultured

keratinocytes and melanocytes. Biochim Biophys Acta, 1511, 181-192.

Schallreuter KU, Gibbons NCJ, Zothner C, Elloof MMA, Wood JM. (2007). Hydrogen peroxide-

mediated oxidative stress disrupts calcium binding on calmodulin: more evidence for

oxidative stress in vitiligo. Biochem Biophys Res Commun, 360, 70-75.

Schallreuter KU, Hordinsky MK, Wood JM. (1987). Thioredoxin reductase: role in free radical

reduction in different hypopigmentation disorders. Arch Dermatol, 123, 615-619.

Schallreuter KU, Kothari S, Chavan B, Spencer JD. (2008). Regulation of melanogenesis -

controversies and new concepts. Exp Dermatol, 17, 395-404.

Schallreuter KU, Moore J, Wood JM, Beazley WD, Gaze DC, Tobin DJ, Marshall HS, Panske A,

Panzig E, Hibberts NA. (1999). In vivo and in vitro evidence for hydrogen peroxide

(H2O2) accumulation in the epidermis of patients with vitiligo and its successful removal

by a UVB-activated pseudocatalase. J Investig Dermatol Symp Proc, 4, 91-96.

Schallreuter KU, Wood JM, Berger J. (1991). Low catalase levels in the epidermis of patients with

vitiligo. J Invest Dermatol, 97, 1081-1085.

Schallreuter KU, Wood JM, Lemke KR, Levenig C. (1995). Treatment of vitiligo with a topical

application of pseudocatalase and calcium in combination with short-term UVB exposure:

a case study on 33 patients. Dermatology, 190, 223-229.

148 Shin J-M, Kim MY, Sohn K-C, Jung S-Y, Lee H-E, Lim JW, Kim S, Lee Y-H, Im M, Seo Y-J,

Kim CD, Lee J-H, Lee Y, Yoon T-J. (2014). Nrf2 negatively regulates melanogenesis by

modulating PI3K/Akt signaling. PLoS One, 9, e96035.

Shindo Y, Witt E, Han D, Epstein W, Packer L. (1994). Enzymatic and non-enzymatic antioxidants

in epidermis and demis of human skin. J Invest Dermatol, 102, 122-124.

Siegel D, Gustafson DL, Dehn DL, Han JY, Boonchoong P, Berliner LJ, Ross D. (2004).

NAD(P)H:quinone oxidoreductase 1: role as a superoxide scavenger. Mol Pharmacol, 65,

1238-1247.

Sotomatsu A, Tanaka M, Hirai S. (1994). Synthetic melanin and ferric ions promote superoxide

anion-mediated lipid peroxidation. FEBS Lett, 342, 105-108.

Sravani PV, Babu NK, Gopal KV, Rao GR, Rao AR, Moorthy B, Rao TR. (2009). Determination

of oxidative stress in vitiligo by measuring superoxide dismutase and catalase levels in

vitiliginous and non-vitiliginous skin. Indian J Dermatol Venereol Leprol, 75, 268-271.

Steenvoorden DPT, Beijersbergen van Henegouwen GMJ. (1997). The use of endogenous

antioxidants to improve photoprotection. J Photochem Photobiol B, 41, 1-10.

Surh Y-J. (2003). Cancer chemoprevention with dietary phytochemicals. Nat Rev Cancer, 3, 768-

780.

Suzukawa AA, Vieira A, Winnischofer SM, Scalfo AC, Di Mascio P, Ferreira AM, Ravanat JL,

Martins Dde L, Rocha ME, Martinez GR. (2012). Novel properties of melanin include

promotion of DNA strand breaks, impairment of repair, and reduced ability to damage

DNA after quenching of singlet oxygen. Free Radic Biol Med, 52, 1945-1953.

149 Takaya K. (1977). Electron microscopy of human melanosomes in unstained, fresh air-dried hair

bulbs and their examination by electron probe microanalysis. Cell Tissue Res, 178(2), 169-

173.

Takei K, Mitoma C, Hashimoto-Hachiya A, Takahara M, Tsuji G, Nakahara T, Furue M. (2015).

Galactomyces fermentation filtrate prevents T helper 2-mediated reduction of filaggrin in

an aryl hydrocarbon receptor-dependent manner. Clin Exp Dermatol, 40, 786-793.

Trouba KJ, Hamadeh HK, Amin RP, Germolec DR. (2002). Oxidative stress and its role in skin

disease. Antioxid Redox Signal, 4, 665-673.

Tsai H-H, Chen Y-C, Lee W-R, Hu C-H, Hakozaki T, Yoshii T, Shen S-C. (2006). Inhibition of

inflammatory nitric oxide production and epidermis damages by saccharomycopsis

ferment filtrate. J Dermatol Sci, 42, 249-257.

Tu CX, Lin M, Lu SS, Qi XY, Zhang RX, Zhang YY. (2012). Curcumin inhibits melanogenesis

in human melanocytes. Phytother Res, 26(2), 174-179.

Tudor C, Lerner-Marmarosh N, Engelborghs Y, Gibbs PEM, Maines MD. (2008). Biliverdin

reductase is a transporter of heme into the nucleus and is essential to regulation of HO-1

gene expression by hematin. Biochem J, 413, 405-416.

Tyrrell RM. (2012). Modulation of gene expression by the oxidative stress generated in human

skin cells by UVA radiation and the restoration of redox homeostasis. Photochem

Photobiol, 11, 134-147.

150 Urabe K, Aroca P, Tsukamoto K, Mascagna D, Palumbo A, Prota G, Hearing VJ. (1994). The

inherent cytotoxicity of melanin precursors: a revision. Biochim Biophys Acta, 1221, 272-

278.

Valko M, Leibfritz D, Moncol J, Cronin MTD, Mazur M, Telser J. (2007). Free radicals and

antioxidants in normal physiological functions and human disease. Int J Biochem Cell Biol,

39, 44-84.

Wei Y, Ji X, Dang X, Hu S. (2003). Studies on electrochemical properties and scavenge of

superoxide anion in aprotic media by using carbon nanotubes powder microelectrode.

Bioelectrochemistry, 61, 51-56.

Wong W-R, Hakozaki T, Yoshii T, Chen T-Y, Pang J-HS. (2011). Up-regulation of tight junction-

related proteins and increase of human epidermal keratinocytes barrier function by

saccharomycosis ferment filtrate. J Cosmet Dermatol Sci Appl, 1, 15-24.

Xiang Y, Liu G, Yang L, Zhong JL. (2011). UVA-induced protection of skin through the induction

of heme oxygenase-1. Biosci Trends, 5(6), 239-244.

Yaar N & Gilchrest M. (2001). Aeging and photoageing of kerartinocytes and melanocytes. Clin

Exp Dermatol, 26, 583-591.

Yamaguchi Y, Hearing VJ, Maeda A, Morita A. (2010). NADPH:quinone oxidoreductase-1 as a

new regulatory enzyme that increases melanin synthesis. J Invest Dermatol, 130, 645-647.

Yohn JJ, Norris DA, Yrastorza DG, Buno IJ, Leff JA, Hake SS. (1991). Disparate antioxidant

enzyme activities in cultured human cutaneous fibroblasts, keratinocytes, and melanocytes.

J Invest Dermatol, 97, 405-409.

151 Zhao H & Boissy RE. (1994). Distinguishing between the catalytic potential and apparent

expression of tyrosinase activities. Am J Med Sci, 308(6), 322-330.

Zhong JL, Edawrds GP, Raval C, Li H, Tyrrell RM. (2010). The role of Nrf2 in ultraviolet A

mediated heme oxygenase 1 induction in human skin fibroblasts. Photochem Photobiol

Sci, 9(1), 18-24.

152