Modification of the human hair: modulators of melanogenesis as agents of colour changing

Universidade do Minho Escola de Engenharia Bruno Pacheco Fernandes setembro de 2020 Modification of the human hair: modulators of melanogenesis as agents of colour changing Bruno Pacheco Fernandes Modification of the human hair: modulators of melanogenesis as agents of colour changing UMinho|2020 setembro de 2020 Trabalho efetuado sob a orientação do Professor Doutor Artur Manuel Cavaco Paulo Tese de Doutoramento Doutoramento em Engenharia Química e Biológica Universidade do Minho Escola de Engenharia Bruno Pacheco Fernandes Modification of the human hair: modulators of melanogenesis as agents of colour changing Universidade do Minho Escola de Engenharia ii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Licença concedida aos utilizadores deste trabalho Atribuição-NãoComercial-SemDerivações CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/ iii AGRADECIMENTOS Terminada esta etapa da minha formação académica, não podia deixar de agradecer a todos aqueles que me acompanharam nesta longa caminhada e contribuíram para a concretização da presente tese. Em primeiro lugar, agradeço ao meu orientador, o Professor Artur Cavaco-Paulo, pela oportunidade de integrar o seu grupo de investigação, pela orientação e incentivo durante estes anos. Quero também agradecer de forma especial à Doutora Teresa Matamá, que não sendo minha orientadora oficial, foi uma mentora excecional. Obrigado Teresa por todos os conhecimentos transmitidos, pelas trocas de ideias, pela dedicação, disponibilidade, apoio incansável, motivação constante e principalmente pela amizade. Agradeço ainda à Professora Andreia Gomes que me acompanhou desde o início do doutoramento, e cujos comentários e sugestões ajudaram sempre a melhorar o trabalho. Á Fundação para a Ciência e a Tecnologia, agradeço o financiamento da minha bolsa de doutoramento (SFRH/BD/131824/2017). Agradeço também à Universidade do Minho, ao Centro de Engenharia Biológica (Departamento de Engenharia Biológica, Escola de Engenharia) e ao Centro de Biologia Molecular e Ambiental (Departamento de Biologia, Escola de Ciências) por terem proporcionado todas as condições técnicas necessárias à realização deste projeto. Obrigado a todos os antigos e atuais colegas do grupo de investigação Bioprocessos e Bionanotecnologia (BBRG), que das mais variadas formas me apoiaram neste percurso. De modo particular, agradeço à Célia, à Joana, à Vanessa, ao Nuno, ao Vadim e ao Zé pela companhia e boa disposição. Não só pelo companheirismo, mas pela valiosa ajuda em diferentes fases do trabalho experimental, agradeço também de modo particular à Diana, à Cristiana e à Madalena. Um reconhecimento igualmente merecido vai para todas as pessoas que de boa vontade se prontificam a colaborar neste projeto: Marisa Passos, Ana Sofia Teixeira e Inês Pinto – assistência no trabalho com os embriões de peixe-zebra; Isabel Pinto (Diretora de Qualidade, Seara S.A.) – ajuda na colheita de amostra de pele de porco; Doutor Francisco X. Real (Centro Nacional de Investigaciones Oncológicas, Espanha) e Doutora Sofia Magina (Faculdade de Medicina da Universidade do Porto, Portugal) – cedência das linhas celulares SK-Mel-23 e SK-Mel-1; todos os voluntários do estudo cosmético ou que cederam amostras de cabelo. Estes agradecimentos não estariam completos sem reconhecer aqui a importância do apoio que recebi de todos os familiares e amigos de longa data no decorrer do meu doutoramento. Contudo, o meu mais sentido reconhecimento não podia deixar de ir para os meus pais pela educação, pela presença constante e por me apoiarem incondicionalmente. São eles a principal razão de eu ter chegado até aqui, e é a eles a quem eu dedico esta tese. A todos, muito obrigado! iv STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. MODIFICAÇÃO DO CABELO HUMANO: MODELADORES DA MELANOGÉNESE COMO AGENTES DE MUDANÇA DA COR v RESUMO A cor do cabelo é um elemento marcante do nosso visual e tem um impacto social irrefutável. Em virtude da importância dada ao aspeto físico, a coloração do cabelo é hoje uma prática bastante comum. Todavia, os efeitos adversos do recurso frequente a tais práticas exigem que novos métodos sejam desenvolvidos para alteração da cor do cabelo. Neste sentido, esta tese teve como objetivo principal demonstrar que, interferindo com o processo fisiológico da pigmentação do cabelo, é possível alterar a sua cor natural a partir do folículo, de forma segura. O tom natural do cabelo é determinado pelo seu conteúdo em melaninas, pigmentos sintetizados por melanócitos do folículo capilar através de um processo conhecido como melanogénese. O presente trabalho foi concebido com base na hipótese de escurecer ou clarear o cabelo via aplicação tópica de modeladores da melanogénese no couro cabeludo. Primeiro, desenvolveu-se um método com base na fluorescência dos produtos de oxidação da melanina, mais expedito que os métodos de quantificação deste pigmento já estabelecidos e com igual ou melhor precisão, exatidão, sensibilidade e especificidade que o método mais usado na literatura. Este novo método foi essencial para a procura de modeladores da produção in vitro de melaninas numa linha celular humana pigmentada. De entre 1200 fármacos com perfil toxicológico bem definido e longo historial de uso seguro em humanos, 23 fármacos apresentaram efeitos significativos. Os quatro com propriedades proou anti-melanogénicas mais acentuadas foram otimizados tendo em vista a sua toxicidade e relação dose-resposta; a atividade da tirosinase foi também avaliada para esclarecer os respetivos mecanismos de ação. Atendendo aos resultados obtidos, um indutor e dois inibidores da melanogénese foram selecionados para testes in vivo em voluntários humanos. Como previsto, o cabelo dos voluntários cresceu mais escuro ou mais claro devido à aplicação tópica do Composto A+ ou Composto B-, respetivamente. O Composto Cdiminuiu o teor de melaninas, mas não alterou a cor visualmente percetível dos cabelos. Nenhum efeito adverso foi reportado. A eficiência de um sistema de entrega folicular particulado face a uma solução não particulada foi ainda avaliada ex vivo para o fármaco ciclosporina A, como modelo de tratamento capilar, neste caso, para alopecia. Os resultados confirmaram que nanopartículas de ácido polilático são um excelente veículo de entrega folicular, em particular, para fármacos com baixa permeabilidade cutânea. De acordo com o objetivo inicial, esta tese institui o uso tópico de moduladores da melanogénese como uma inovadora abordagem segura e eficaz na alteração da cor natural do cabelo, sendo o ponto de partida para um novo tipo de cosméticos de coloração capilar. PALAVRAS-CHAVE: Cabelo; Cor; Cosméticos; Fármacos; Melanogénese MODIFICATION OF THE HUMAN HAIR: MODULATORS OF MELANOGENESIS AS AGENTS OF COLOUR CHANGING vi ABSTRACT Hair colour is an undeniable element of body image with great social impact. Given the importance attributed to hair look, the dyeing of hair fibres is today a common practice. Nonetheless, the adverse effects of long-term usage of such fibre colouring procedures demand the development of other methods for colour change. Therefore, the major purpose of this Ph.D. thesis was the development of a new method for hair colour modification by safely interfering with the physiological process of hair pigmentation at the hair follicle level. The natural colour shade of hair is largely defined by its content in melanins, a class of pigments synthesized by follicular melanocytes through a biochemical process known as melanogenesis. The current work was built upon the hypothesis of hair darkening/lightening through the topical treatment of the scalp with modulators of melanin synthesis. First, a new methodology was developed for melanin quantification based on the fluorescence of its oxidation products. This method was more expeditious than other established assays, performing equally or better than the most used method in terms of sensitivity, precision, accuracy, and specificity. This new method was essential for the in vitro search of melanogenesis modulators. A collection of 1200 generic drugs with well-defined toxicological profile and a long history of safe usage in humans was screened regarding the ability to alter the melanin contents of a melanin-producing human cell line. Among the 23 drugs that showed highly significant effects, the top four inducers and inhibitors were selected for further validation and optimization regarding their cytotoxicities and dose-response activities. Additionally, the activity of tyrosinase was also assessed to gain some insight into the mechanisms of action. Taking into consideration the in vitro data generated, three drugs were chosen for a clinical trial with intervention of cosmetics. As predicted, the hair of human volunteers grew darker or lighter due to the topical treatment of their scalps with Compound A+ or Compound B-, respectively. Compound Calso decreased the melanin content of hairs but without visually perceptible colour changes. No adverse effects were reported. The efficiency of a particulate follicular delivery system was also evaluated ex vivo by comparing it to a non-particulate vehicle. Cyclosporine A was used as a model drug of hair treatment, in this case, for alopecia. The results confirmed that polylactic acid nanoparticles are an excellent vehicle for follicular delivery of drugs with low skin permeability. According to its major purpose, this thesis provides pioneer evidence that melanogenesis modulators, applied topically, can safely and effectively change the colour of human hair from the inside out, opening a new window of opportunities for the development of ground-breaking hair colouring cosmetics. KEYWORDS: Colour; Cosmetics; Drugs; Hair; Melanogenesis. vii TABLE OF CONTENTS AGRADECIMENTOS .................................................................................................... iii RESUMO ..................................................................................................................... v ABSTRACT ................................................................................................................. vi TABLE OF CONTENTS ............................................................................................... vii LIST OF SYMBOLS AND ABBREVIATIONS ................................................................... xii LIST OF FIGURES ..................................................................................................... xix LIST OF TABLES .................................................................................................... xxviii LIST OF EQUATIONS .............................................................................................. xxxi PROLOGUE ..................................................................................................................... 1 Framework of the thesis ............................................................................................. 1 Scope and Goals ......................................................................................................................... 2 Thesis outline .............................................................................................................................. 2 CHAPTER I ..................................................................................................................... 5 Paving the way for innovative hair colour-changing cosmetics: a comprehensive review of human hair pigmentation ....................................................................................... 5 Abstract ...................................................................................................................................... 6 I.1. General overview ................................................................................................................. 7 I.2. Development of the hair follicle pigmentary unit .................................................................... 8 I.3. Hair follicle melanogenesis and the hair growth cycle ......................................................... 11 I.4. Molecular and cellular biology of hair pigmentation ............................................................. 13 I.4.1 Melanosomes biogenesis ............................................................................................ 15 I.4.1.1 Major players in melanosomes biogenesis ............................................................. 17 I.4.1.1.1 Premelanosome protein (PMEL) .................................................................... 17 I.4.1.1.2 Protein melan-a (MLANA) .............................................................................. 18 I.4.1.1.3 G protein-coupled receptor 143 (GPR143) ..................................................... 18 I.4.1.2 Protein trafficking to melanosomes ........................................................................ 19 I.4.2 Biochemical pathway of melanogenesis ...................................................................... 20 I.4.2.1. Major players in melanins synthesis ....................................................................... 24 I.4.2.1.1. Tyrosinase (TYR) .......................................................................................... 24 I.4.2.1.2. Tyrosinase related protein 1 (TYRP1) ............................................................ 25 I.4.2.1.3. Cystine transporters ..................................................................................... 25 I.4.2.2. Control of melanosomal pH ................................................................................... 25 xiv G g GFRP Glu GPER GPR143 GSH GSK3β GTP GTP-CH-I Gram GTP-cyclohydrolase I feedback regulatory protein Glutamate G protein-coupled oestrogen receptor G protein-coupled receptor 143 Glutathione, reduced form Glycogen synthase kinase 3β Guanosine triphosphate GTP-cyclohydrolase I H h H2O2 H2R His Hist hpf HPLC HSP HTS Hour Hydrogen peroxide Histamine H2 receptor Histidine Histamine Hours post fertilization High performance liquid chromatography Heat shock protein High throughput screening I IBMX IC50 IFD ILV IP3 IPM ISO 3-isobutyl-1-methyl-xanthine Half maximal inhibitory concentration Infundibulum Intralumenal vesicle Inositol–triphosphate Isopropyl myristate International organization for standardization J JNK c-Jun N-terminal kinase K KA kDa Ker KIT KITLG Kojic acid Kilodalton Keratinocyte KIT proto-oncogene receptor tyrosine kinase KIT ligand xv L L LAMP L-DOPA LEF lncRNA L-Tyr Litre Lysosomal-associated membrane protein L-3,4-dihydroxyphenylalanine Lymphoid enhancer factor Long non-coding RNA L-tyrosine M M MAPK MC(1/4)R Mel Mel-Mc MelSC mg mGluR6 min miRNA MITF mL MLANA MLPH mm mM M(1-5)R mRNA MTN MTNR MTT Mw MYO Molar Mitogen-activated protein kinase Melanocortin (1/4) receptor Melanocyte Melanogenic melanocyte Melanocyte stem cell Milligram Metabotropic glutamate receptor 6 Minute Micro RNA Microphthalmia-associated transcription factor Millilitre Protein melan-a Melanophilin Millimetre Millimolar Muscarinic (1-5) receptor Messenger RNA Melatonin Melatonin receptor Thiazolyl blue tetrazolium bromide Molecular weight Myosin N NaOH ncRNA ND Sodium hydroxide Non-coding RNA Not detectable xvi N NEpi ng NHE NHM NK1R nm nM NMDA NO NOS Norepinephrine Nanogram Sodium hydrogen exchanger Normal human melanocyte Neurokinin 1 receptor Nanometre Nanomolar N-methyl-D-aspartate Nitric oxide Nitric oxide synthase O OCA2 ORS Oculocutaneous albinism 2 protein Outer root sheath P PA PAH PAR2 PAX3 PBS PC PDE PDI PGC Phe PI3K PKA PKC PKG PLA PLCγ PMEL POMC PPAR Palmitic acid Phenylalanine hydroxylase Protease-activated receptor 2 Paired box 3 Phosphate buffered saline solution Proprotein convertase Phosphodiesterase Polydispersity index Peroxisome proliferator-activated receptor coactivator Phenylalanine Phosphatidylinositol 3-kinase Protein kinase A Protein kinase C Protein kinase G Poly (D,L-lactide) Phospholipase Cγ Premelanosome protein Pro-opiomelanocortin Peroxisome proliferator-activated receptor Q q-BH2 Quinonoid dihydropterin xvii R RNA RNEC ROS Ribonucleic acid Registo nacional de ensaios clínicos Reactive oxygen species S SCF SD Ser SIK2 SLC7A5 SLC7A11 SLC24A5 SLC45A2 SNARE SOX SP SSRI Stem cell factor Standard deviation Serine Salt-inducible kinase 2 Solute carrier family 7, member 5 Solute carrier family 7, member 11 Solute carrier family 24, member 5 Solute carrier family 45, member 2 Soluble N-ethylmaleimide-sensitive-factor attachment protein receptor SRY (sex determining region Y)-box Substance P Selective serotonin reuptake inhibitor T TCF TGF TGN TPC2 TRH TRHR Trp TRPM1 TRPML3 TYR TYRP1 TYRP2 T-cell factor Transforming growth factor trans-Golgi network Two pore segment channel 2 Thyrotropin-releasing hormone Thyrotropin-releasing hormone receptor Tryptophan Transient receptor potential cation channel, subfamily M, member 1 Transient receptor potential cation channel, mucolipin subfamily, member 3 Tyrosinase Tyrosinase related protein 1 Tyrosinase related protein 2 U UV Ultraviolet V v/v VAMP V-ATPase Volume/volume Vesicle-associated membrane protein Vacuolar ATPase xviii W w/w w/v Weight/weight Weight/volume 4a-OH-BH4 4a-OH-BH4 DH 5-HT 5-HTR 6BH4 7-BH4 4a-hydroxy-tetrahydrobiopterin 4a-hydroxy-tetrahydrobiopterin dehydratase Serotonin Serotonin receptor (6R)-L-erythro 5,6,7,8 tetrahydrobiopterin (7R)-L-erythro-5,6,7,8-tetrahydrobiopterin °C α-MSH β-END β-MSH σ ζ-potential λ μ µg µL µm µM Ψm Degree Celsius Alpha melanocyte-stimulating hormone Beta-endorphin Beta melanocyte-stimulating hormone Standard deviation Zeta-potential Wavelength Mean Microgram Microlitre Micrometre Micromolar Membrane potential xix LIST OF FIGURES Figure I.1. Distribution of melanocytes in the different regions of the human anagen scalp hair follicle. Amel-Mc: amelanotic melanocytes; DP: dermal papilla; IFD: infundibulum; Mel-Mc: melanogenic melanocytes; ORS: outer root sheath. ............................................................................................... 10 Figure I.2. Schematic representation of the fate of melanocytes during the anagen, catagen and telogen phases of the hair growth cycle. KIT: KIT proto-oncogene receptor tyrosine kinase; Mel-Mc: melanogenic melanocytes; MelSC: melanocyte stem cells; SCF: stem cell factor. ................................................... 12 Figure I.3. Schematic representation of melanosomes biogenesis. AP-1/3: adaptor-related protein complex 1/3; ATP7A: ATPase copper transporting alpha; BLOC-1/2: biogenesis of lysosomal organelles complex 1/2; ER: endoplasmic reticulum; GPR143: G protein-coupled receptor 143; ILV: intralumenal vesicles; MITF: microphthalmia-associated transcription factor; MLANA: protein melan-a; OCA2: oculocutaneous albinism 2 protein; PMEL: premelanosome protein; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. .............................................................................................................................. 16 Figure I.4. Metabolism of L-phenylalanine to L-tyrosine via phenylalanine hydroxylase (PAH). The activity of PAH depends on the co-factor (6R)-L-erythro 5,6,7,8 tetrahydrobiopterin (6-BH4), which melanocytes have full capacity for de novo synthesis (blue arrows) and recycling (green arrows). 4a-OH-BH4 DH: 4ahydroxy-tetrahydrobiopterin dehydratase; 4a-OH-BH4: 4a-hydroxy-tetrahydrobiopterin; 7-BH4: (7R)-Lerythro-5,6,7,8-tetrahydrobiopterin; DHPR: dihydropteridine reductase; GFRP: GTP-cyclohydrolase I feedback regulatory protein; GTP-CH-I: GTP-cyclohydrolase I; q-BH2: quinonoid dihydropterin; SLC7A5: solute carrier family 24, member 5. .................................................................................................. 21 Figure I.5. Biosynthetic pathways involved in the production of eumelanin and pheomelanin.in human melanosomes. Spontaneous reactions are denoted as black arrows. Enzyme-catalysed reactions are signalized by blue arrows. CD: cysteinyldopa; L-DOPA: L-3,4-dihydroxyphenylalanine; TYR: tyrosinase; DCT: dopachrome tautomerase. ....................................................................................................... 23 Figure I.6. Schematic representation of the cascade of ion channels and transporters with a role in melanogenesis. Cys: cysteine; CySS: cystine; Glu: glutamate; NHE: sodium hydrogen exchanger; OCA2: oculocutaneous albinism 2 protein; SLC(7A11/24A5/45A2): solute carrier family 7/24/45, member 11/5/2; TPC2: two pore segment channel 2; TRPML3: transient receptor potential cation channel, mucolipin subfamily, member 3; V-ATPase: vacuolar ATPase. ............................................................ 26 Figure I.7. Mechanisms of melanosome transport and transfer. Melanosomes bind to kinesin for moving along microtubules from the perinuclear region to the periphery of melanocytes. At the dendrites, the movement of melanosomes in the rich network of actin filament is regulated by the RAB27A-MLPH- xx MYO5A complex. Several models have been proposed regarding the transference of melanin from melanocytes to surrounding keratinocytes. MLPH: Melanophilin; MYO5A:myosin VA; RAB27A: Ras-related protein RAB27A. ............................................................................................................................... 29 Figure I.8. Major molecular pathways involved in the regulation of melanin synthesis in melanocytes. 32 Figure I.8A. Highlight of molecular pathways involved in the regulation of melanin synthesis by proopiomelanocortin (POMC)- derived peptides. α/β-MSH: α/β-melanocyte-stimulating hormone; β-END: βendorphin; AC: adenylyl cyclase; Akt: serine/threonine kinase; cAMP: cyclic adenosine monophosphate; CBP: CREB-binding protein; CREB: cAMP response element-binding protein; CRE: cAMP responsive element; CRF(R): corticotropin-releasing factor (receptor); CRTC1: CREB-regulated transcription coactivator 1; GSK3β: glycogen synthase kinase 3β; Ker: keratinocytes; MC1/4R: melanocortin 1/4 receptor; Mel: melanocytes; MITF: microphthalmia-associated transcription factor; PC1/2: proprotein convertase 1/2; PI3K: phosphatidylinositol 3-kinase; PKA/C: protein kinase A/C; SIK2: salt-inducible kinase 2; TRH(R): thyrotropin-releasing hormone (receptor); TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ......................................................................................................................................... 34 Figure I.8B. Highlight of molecular pathways involved in the regulation of melanin synthesis by WNT proteins. GSK3β: glycogen synthase kinase 3β; LEF: lymphoid enhancer-binding factor; MITF: microphthalmia-associated transcription factor; PKA: protein kinase A; TCF: T-cell factor; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ................................................................................................. 37 Figure I.8C. Highlight of molecular pathways involved in the regulation of melanin synthesis by stem cell factor (SCF). CREB: cAMP response element-binding protein; ERK: extracellular signal-regulated kinases; JNK: c-Jun N-terminal kinases; KIT: KIT proto-oncogene, receptor tyrosine kinase; PI3K: phosphatidylinositol 3-kinase; PKA/C: protein kinase A/C; MAPK: mitogen-activated protein kinases; MITF: microphthalmia-associated transcription factor; RSK: p90 ribosomal S6 kinase; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ................................................................................................. 39 Figure I.8D. Highlight of molecular pathways involved in the regulation of melanin synthesis by endothelin 1 (ET-1). CREB: cAMP response element-binding protein; DAG: diacylglycerol; EDNRB: endothelin receptor type B; IP3: inositol–triphosphate; MAPK: mitogen-activated protein kinases; MITF: microphthalmia-associated transcription factor; PKC: protein kinase C; PLCγ: phospholipase Cγ; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ................................................................................ 40 Figure I.8E. Highlight of molecular pathways involved in the regulation of melanin synthesis by neurotransmitters. 5-HT(R): serotonin (receptor); AC: adenylyl cyclase; Ach: acetylcholine; AR: adrenergic receptor; cAMP: cyclic adenosine monophosphate; CREB: cAMP response element-binding protein; DAG: diacylglycerol; (N)Epi: (nor)epinephrine; H2R: histamine H2 receptor; Hist: histamine; IP3: inositol– xxi triphosphate; Ker: keratinocytes; PKA/C: protein kinase A/C; PLCγ: phospholipase Cγ; MAPK: mitogenactivated protein kinases; Mel: melanocytes; M2/4R: muscarinic 2/4 receptor. m(Glu)R6: metabotropic (glutamate) receptor 6; MITF: microphthalmia-associated transcription factor; MTN(R): melatonin (receptor); TRPM1: transient receptor potential cation channel, subfamily M, member 1; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ................................................................................................. 42 Figure I.8F. Highlight of molecular pathways involved in the regulation of melanin synthesis by bone morphogenetic proteins (BMP). BMPR1/2: bone morphogenetic proteins receptors 1/2; CREB: cAMP response element-binding protein; ERK: extracellular signal-regulated kinases; Ker: keratinocytes; MAPK: mitogen-activated protein kinases; Mel: melanocytes; MITF: microphthalmia-associated transcription factor; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. .............................................................. 44 Figure I.8G. Highlight of molecular pathways involved in the regulation of melanin synthesis by oestrogens. cAMP: cyclic adenosine monophosphate; CREB: cAMP response element-binding protein; GPER: G protein-coupled oestrogen receptor; MITF: microphthalmia-associated transcription factor; PKA: protein kinase A; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. ......................................................... 45 Figure I.9. Aged hair follicle, presenting grey/white hair .The age-related pigment loss in hair fibres may be caused may be caused by 1) depletion of melanotic melanocytes (Mel-Mc) in the hair matrix, 2) exhaustion of the melanocyte stem cell (MelSC) reservoir and 3) MelSC defective activation or migration during the onset of anagen phase of the hair growth cycle. ................................................................ 50 Figure II.1. Schematic overview of the proposed protocol for melanin quantification by fluorescence spectroscopy in in vitro assays. ......................................................................................................... 61 Figure II.2. Effect of forskolin (Fsk) and kojic acid (KA) on intracellular melanin content of non-melanotic (BJ-5ta, NCTC2544), amelanotic (A-375, melanoma cells with no visible pigmentation), or melanized cells (SK-Mel-1 and SK-Mel-23, pigmented melanoma cells), after 72 h of treatment. Melanin content was calculated by interpolating the results with standard curves, generated by the fluorescence (after oxidation) (A) or absorbance (B) of Sepia melanin solutions of known concentration, prepared in 1 M NaOH containing 10% (v/v) of DMSO. The results were normalized by total protein levels in each sample. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. EtOH: ethanol. .................................................................................. 63 Figure II.3. Quantification of melanin in biological samples by fluorescence spectroscopy. (A) Developmental changes in the accumulation of melanin in zebrafish embryos. (B) Contents of total melanin in human hair samples of various colours (visual phenotypes). Melanin contents were calculated by interpolating the results with standard curves, generated by the fluorescence (after oxidation) of Sepia melanin solutions of known concentration, prepared in 1 M NaOH containing 10% (v/v) of DMSO. The xxii results were normalized by number of embryos (A) or amount of hair (B) in each sample. Data were analysed by one-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. hpf: hours post fertilization. ............................................................................. 65 Figure II.S1. Time response study of the fluorescence of synthetic (A) and Sepia (B) melanin solutions, oxidized at 25 °C with different amounts of hydrogen peroxide (λexcitation= 470 nm; λemission = 550 nm). Melanin standard solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. After addition of hydrogen peroxide, the volume of samples was adjusted to normalize the melanin concentration at 100 µg/mL. The data was analysed by two-way ANOVA, followed by post hoc Tukey’s test. ............... 68 Figure II.S2. Fluorescence and absorbance values of solutions containing different concentrations of synthetic (A) and Sepia (B) melanins. Solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO, by two-fold serial dilutions. Absorbance of solutions was measured at 405 nm. Fluorescence intensity was measured (λexcitation= 470 nm; λemission = 550 nm) after incubation of melanin solutions with 20% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by one-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, compared to the respective blank sample (0 µg/mL). ........................................................................................ 69 Figure II.S3. Plots of standard curves of synthetic (A) and Sepia (B) melanins obtained by fluorescence and absorption spectroscopy. Solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO, by two-fold serial dilutions. Absorbance of solutions was measured at 405 nm. Fluorescence intensity was measured (λexcitation= 470 nm; λemission = 550 nm) after incubation of melanin solutions with 20% (v/v) hydrogen peroxide solution for 4 h. ........................................................................................... 70 Figure II.S4. Fluorescence (A) and absorbance (B) intensities of lysates containing different amounts of non-melanotic (BJ-5ta, NCTC2544), amelanotic (A-375, melanocytes with no visible pigmentation) or melanized (SK-Mel-1 and SK-Mel-23) cells grown in culture. Cell lysates were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. Fluorescence intensity was measured after incubation of cell lysates with 30% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by two-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, when melanocytes concentrations were compared to the respective blank sample (0 cells/mL) or against the same concentration of non-melanotic and amelanotic cells. ........................................................................ 72 Figure II.S5. Fluorescence intensities of lysates containing different amounts of zebrafish embryos at specific time points of development (8 – 120 hours post fertilization). The lysates were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. Fluorescence intensity was measured after incubation of lysates with 30% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by two-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, when samples xxiii concentrations were compared to the respective blank sample (0 embryos/mL) or against the same concentration of amelanotic samples (8 and 24 hpf). ........................................................................ 75 Figure II.S6. Fluorescence intensities of Sepia melanin-containing solutions in the presence of variable amounts (0 – 5 µM) of sepiapterin (A), guanine (B) or tetrahydrobiopterin (BH4) (C). Solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. Fluorescence intensity was measured after incubation with 30% (v/v) hydrogen peroxide solution for 4 h. Data was analysed by two-way ANOVA, followed by post hoc Dunnett’s test. .................................................................................................. 76 Figure II.S7. Fluorescence intensities of lysates containing different amounts of human hair of variable colours. The lysates were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. Fluorescence intensity was measured after incubation of lysates with 30% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by two-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, when samples concentrations were compared to the respective blank sample (0 mg of hair/mL) or against the same concentration of amelanotic samples (white hair). ..... 77 Figure II.S8. Effect of forskolin (Fsk) and kojic acid (KA) on intracellular melanin content of non-melanotic (BJ-5ta, NCTC2544), amelanotic (A-375, melanoma cells with no visible pigmentation) or melanized cells (SK-Mel-1 and SK-Mel-23, pigmented melanoma cells), after 72 h of treatment. Melanin content was calculated by interpolating the results with standard curves, generated by the fluorescence (after oxidation) (A) or absorbance (B) of synthetic melanin solutions of known concentration, prepared in 1 M NaOH containing 10% (v/v) of DMSO. The results were normalized by total protein levels in each sample. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ......................................................................................................... 78 Figure III.1. High Throughput Screening (HTS) of Prestwick Chemical Library® regarding the in vitro modulation of melanin. SK-Mel-23 cells were treated with compounds for 72 h and then, melanin quantification was performed using a fluorescence-based method. The melanin contents were normalized by total protein levels in each sample and expressed as a percentage of the vehicle control, 1% (v/v) DMSO. HTS-Corrector software (version 2.0) was used to examine HTS data and hit selection. The cutoff for hit melanogenesis inducers and inhibitors was set at μ + 3σ and μ - 2σ, respectively. μ: whole mean assay. σ: standard deviation. .................................................................................................. 83 Figure III.2. SK-Mel-23 cell viabilities (MTT assay) determined at 24, 48 and 72 h of exposure to different concentration of hit inducers. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define cytotoxicity ranges: non-cytotoxicity > 80%; weak 80-60%; moderate 60-40%; strong < 40%. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ......................................................................................................... 86 xxx Table V.12. Antipsoriatics with in vitro effects on melanogenesis. Melanin contents and intracellular tyrosinase activities are presented as a percentage of the control .................................................... 159 Table V.13. Antithrombotic and antihemorrhagic agents with in vitro effects on melanogenesis. Melanin contents and intracellular tyrosinase activities are presented as a percentage of the control ............. 160 Table V.14. Drugs used in erectile dysfunction with in vitro effects on melanogenesis. Melanin contents and intracellular tyrosinase activities are presented as a percentage of the control ........................... 161 Table V.15. Immunosuppressants with in vitro effects on melanogenesis. Melanin contents and intracellular tyrosinase activities are presented as a percentage of the control. NHM: Normal human melanocytes ................................................................................................................................... 162 Table V.16. Psychoanaleptics with in vitro effects on melanogenesis. Melanin contents and intracellular tyrosinase activities are presented as a percentage of the control. IBMX: 3-isobutyl-1-methyl-xanthine. NHM: Normal human melanocytes ................................................................................................. 163 Table V.17. Psycholeptics with in vitro effects on melanogenesis. Melanin contents and intracellular and tyrosinase activity are presented as a percentage of the control. NHM: Normal human melanocytes 165 Table V.18. Sex hormones and modulators of the genital system with in vitro effects on melanogenesis. Melanin contents and tyrosinase activity are presented as a percentage of the control ..................... 167 Table V.19. Other drugs with in vitro effects on melanogenesis. Melanin contents and tyrosinase activity are presented as a percentage of the control. IBMX: 3-isobutyl-1-methyl-xanthine. NHM: Normal human melanocytes ................................................................................................................................... 168 xxxi LIST OF EQUATIONS Equation IV.1. Yield of nanoparticles obtained in the production of cyclosporin A-loaded poly (D,L-lactide) nanoparticles. PLA: poly (D,L-lactide). ............................................................................................. 125 Equation IV.2. Entrapment efficiency of cyclosporin A in poly (D,L-lactide) nanoparticles. CsA: Cyclosporin A. ................................................................................................................................................... 125 Equation IV.3. Loading efficiency of cyclosporin A in the poly (D,L-lactide) nanoparticles. CsA: Cyclosporin A. ................................................................................................................................................... 126 PROLOGUE FRAMEWORK OF THE THESIS Framework of the Thesis 2 SCOPE AND GOALS Hair colour is one of the most distinctive traits of our physical appearance, greatly affecting the way we are perceived by others and ourselves. Today, more than ever, our looks are a predicate for social success and career opportunities, with millions of individuals routinely changing the colour of hair to enhance their own idea of self-beauty and/or to follow fashion trends. In this context, the hair colouring market has thrived by supplying a variety of hair fibre dyeing systems that comply with distinct aesthetic requirements of consumers. On the downside, most of those systems are known to cause extensive damage to the hair fibre structure and their long-term use has been associated with several adverse health effects like acute to mild dermatitis and hair loss. Consequently, the development of strategies for hair colour modification more focused on healthy hair and consumers wellbeing and safety is an important issue to be addressed. While most of the upfront research on safer solutions for colour modification is devoted to act on the external portion of hair fibres, the innovative factor of this Ph.D. thesis consists in demonstrating the cosmetic potential of transiently interfering with the biological process of pigmentation, that occurs at hair follicle, to change the natural colour of hair from the inside out. To do so, two major goals were defined: 1. Discover new modulators of the in vitro synthesis of melanins, the natural pigments responsible for hair colour, among pharmaceutical drugs already approved for human usage; 2. Provide evidence that topical application of those drugs to the human scalp can safely and effectively modulate the in vivo production of melanin by follicular melanocytes, changing the colour of hair fibres as they are produced at the bulb. The journey for achieving these goals is reported over the next chapters, as outlined. THESIS OUTLINE This thesis is divided into five chapters, with chapters II to IV detailing the experimental work. The content of each chapter is summarized below. Chapter I. Paving the way for innovative hair colour-changing cosmetics: a comprehensive review of human hair pigmentation The introductory chapter of this thesis provides a comprehensive overview of the natural process of human hair pigmentation. Additionally, current cosmetic procedures employed in hair colour modification Modification of the human hair: modulators of melanogenesis as agents of colour changing 3 are briefly described. At last, case reports of hair darkening, lightening or repigmentation as a side effect of the therapeutic usage of many drugs are mentioned, substantiating the feasibility to tune hair colour by interfering with the follicular pigmentary unit. This review gives an overall perspective from which it becomes clear that being able to influence melanogenesis (the synthesis of melanins) is crucial for changing the hair colour phenotype. Outstandingly, the biochemical pathway of melanin synthesis presents countless molecular targets for doing so. Chapter II. Fluorescent quantification of melanin The basis of many studies dealing with the discovery of melanogenesis modulators involves the in vitro measurement of melanins in cultured cells. Although absorption spectroscopy is the most widely used method, it does not always provide the sensitivity and specificity required to evaluate the significance of such modulators. As a result, in this chapter, the fluorometric method is revisited to establish a protocol for easy and accurate quantification of melanins. Besides cell samples, the usefulness of this protocol is assessed in more complex pigmented matrices like zebrafish embryos and human hair. Chapter III. Discovery of new melanogenesis modulators for hair colour modification: an in vitro to in vivo study The core chapter of this Ph.D. thesis starts by reporting the quest for new modulators of melanogenesis in drugs already approved for human usage. Given that animal testing for cosmetic purposes is prohibited in the E.U. since 2013, looking for new cosmetic ingredients among compounds with known toxicological profile is, today, a reliable approach to ensure that they pose no to low risk for humans. Thus, the data gathered in the screening of Prestwich Chemical Library® regarding the ability to change the in vitro production of melanin is presented. Then, the results of validation and optimization assays performed with the top hit inducers and inhibitors are discussed. Lastly, to gain some insight regarding their mechanism of action, the effects of the top hit modulators on the intracellular activity of tyrosinase (ratelimiting enzyme of the biochemical pathway of melanin synthesis) are analysed. In the second part of this chapter, the feasibility of using some selected drugs as topically applied ingredients for the darkening or lightening of hair is debated, considering the findings of a pilot study performed in human volunteers. Chapter IV. Cyclosporin A-loaded Poly (D, L-lactide) nanoparticles: a promising tool for treating alopecia Changing the natural colour of hair from the inside out is a biphasic challenge. After finding the best modulators of melanin production, we still have to efficiently deliver them to their cellular targets. Follicular Framework of the Thesis 4 drug delivery is a hot topic in dermatology and hair cosmetic sciences. Cyclosporin A (CsA), an immunosuppressant drug, has hair growth-inducing properties but its poor cutaneous absorption undermines its topical use in the treatment of alopecia. This chapter addresses the ability of Poly (D,Llactide) (PLA) nanoparticles to improve the dermal permeation of CsA. Conceptually, this work is aligned with the main scope of the present thesis because it reinforces the suitability of nanoparticles-based systems for the efficient and selective delivery of drugs into hair follicles upon topical application. In this context, the PLA nanoparticles here reported will certainly be of interest in future studies that aim to improve the hair colour-changing effects of the modulators of melanogenesis found in the previous chapter. Chapter V. Pharmacological modulation of melanogenesis: where do we stand now and which drugs to repurpose next? Drugs with melanogenic effects have long been studied in view of their implications for skin pigmentationrelated disorders. Thus, as a way of concluding, the last chapter places together the findings of the current thesis work with existing literature to generate an up-to-date overview on pharmacological modulators of pigmentation. In doing so, not only the important contribution of this thesis to the field of innovative hair colouring cosmetics are highlighted, but also the profound connotations between the data generated and the development of new therapeutic approaches to deal with diseases related with abnormal production of melanin in the skin are pointed out. CHAPTER I PAVING THE WAY FOR INNOVATIVE HAIR COLOUR-CHANGING COSMETICS: A COMPREHENSIVE REVIEW OF HUMAN HAIR PIGMENTATION Chapter I: A comprehensive review of human hair pigmentation 6 Chapter I Paving the way for innovative hair colour-changing cosmetics: a comprehensive review of human hair pigmentation ABSTRACT Hair colour is one of the most distinctive human phenotypic traits, contributing significantly to our visual perception of other individuals. The natural colouration of hairs takes place in the bulb of hair follicles and it is strictly couple to the hair growth cycle. Three critical steps must proceed in perfectly synchrony for an efficient pigmentation of hair shafts: 1) melanosomes biogenesis in neural crest-derived melanocytes, 2) biochemical synthesis of melanins (melanogenesis) inside melanosomes, and 3) the transfer of melanin granules to surrounding pre-cortical keratinocytes for incorporation into the forming fibres. All these steps are under complex genetic control, with the array of natural hair shades (ranging from black, brown, and blond to red) being ascribed to polymorphisms in several pigmentary genes. A myriad of factors acting via autocrine, paracrine, and endocrine mechanisms also contributes greatly for the pigmentation of hair and its diversity. By scrutinizing hair pigmentation, this review pinpoint key targetable processes for the development of innovative cosmetics that can safely change the hair colour from the inside out.  This chapter is based on the following scientific paper: Bruno Fernandes, Teresa Matamá and Artur Cavaco-Paulo. Paving the way for innovative hair colourchanging cosmetics: a comprehensive review of human hair pigmentation. To be submitted. Modification of the human hair: modulators of melanogenesis as agents of colour changing 7 I.1. GENERAL OVERVIEW Hair colour is one of the most distinctive human phenotypes, with various selective pressures having contributed to the dissemination of an array of natural shades ranging from black, brown, and blond to red. Additionally, grey/white hair is often seen, harbingering the loss of youth. As much of the perceived variation between human sub-groups can be ascribed to the different types of hair pigmentation, the colour of hair has an enormous social and cosmetic impact [1–5]. Hair and skin pigmentation are a manifestation of the presence of pigments called melanins. Although melanins of the epidermis had important evolutionary implications (acting as filter for ultraviolet (UV) light, they protect the skin against UV-induced damage), the reasons behind the evolutionary selective pressures for the development of pigmented scalp hair are less clear. Due to the dominant position of fish in the diet of early humans, pigmented scalp hair might have been important to prevent the build-up of toxic metals from fish species. Since toxins/metals have the ability to selective bind to melanins and hair fibres present a fast growth and turnover, pigmented hair shafts allow the body to rapidly get rid of harmful substances, limiting their access to the living tissue of the highly vascularized scalp. Hair analyses corroborate this theory as significant amounts of various heavy metal ions bound to melanin have been reported. Moreover, as our nakedness draws much attention to our facial and scalp hair, the development of pigmented scalp hairs could also have been favoured as a non-verbal mean of social communication and mate selection [1,3,5,6]. The colouration of hair shafts results from precise and sequential interactions between different cell populations in the hair follicle, the mini-organ responsible for the hair shaft production. First, melanins are synthesized by melanocytes, inside cytoplasmic membrane-bound organelles called melanosomes. Melanins are the product of a complex biochemical pathway called melanogenesis and they can undertake two chemically distinct forms: the black to brown eumelanic form and the reddish-brown to yellow pheomelanic form. Melanins are then transferred to the surrounding keratinocytes that will proliferate and differentiate originating hair shafts. During the differentiation process, the growing hairs become pigmented. The diversity in hair pigmentation relies mainly on the quantity and ratio of the two types of melanin incorporated into the hair shafts, with other physical aspects of hair fibres intervening only as minor or segmental colour modifiers [1–8]. The genetic basis of hair colour variety has been the subject of many studies and led to the identification of several genes involved in the normal variation of human hair colour. The protein products depending on these loci and their polymorphisms exhibit different patterns of transcription, translation, and functional Chapter I: A comprehensive review of human hair pigmentation 8 activity during the pigmentation of hairs. Melanogenesis per se presents a wide range of potential targets with different functions: enzymes, structural proteins, transcription regulators, transporters, receptors, and their ligands. Besides, many intrinsic hair follicle signals, as the products of keratinocytes, have also been uncovered as intervenient in many aspects of hair pigmentation and its diversity [3,8]. I.2. DEVELOPMENT OF THE HAIR FOLLICLE PIGMENTARY UNIT Melanocytes that constitute the follicular and epidermal pigmentary units derive from the pluripotent cells of the neural crest, a transient component of the ectoderm located between the neural tube and the epidermis. Besides melanocytes, the neural crest gives rise to several other types of cells including neurons and glia cells of the peripheral nervous system as well as bone and cartilage cells of the head skeleton. The commitment of neural crest cells to the melanocyte lineage and the entire journey of melanoblasts (precursors of melanocytes) during embryogenesis to produce the pigmentary unit is regulated by several factors [6,8–11]. Microphthalmia-associated transcription factor (MITF) appears to be the master regulator of melanocyte identity at the neural crest. MITF regulates the expression of genes that confer melanocyte characteristics to the neural crest cells such as tyrosinase ( TYR ), tyrosinase related protein 1 ( TYRP1 ) and dopachrome tautomerase ( DCT ). The transcription of MITF is synergistically activated by paired box 3 (PAX3) and SRYbox 10 (SOX10) transcription factors. In turn, forkhead box D3 (FOXD3) and SOX2 transcription factors repress MITF expression and thus, their downregulation is a critical initial step in the specification of melanoblasts; the mechanism by which their expression is suppressed is still unclear, but histone deacetylase 1 (HDAC1) seems to be involved. The upregulation of WNT proteins (namely, WNT1 and WNT3A) and downregulation of bone morphogenetic protein 4 (BMP4) also appear to be involved in the commitment of neural crest cells to the melanocyte lineage [12,13,22,23,14–21]. Once melanoblasts begin to express specific markers, they migrate dorsoventrally through the developing embryo and start to proliferate. The migration of melanoblasts is directed by chemotactic signals and patterns of cell surface molecules and receptors expressed by the mobilized melanoblasts or present within the extracellular matrix through which they move. The migration and proliferation of melanoblasts require endothelin receptor type B (EDNRB) and its ligand endothelin 3 (ET-3). The KIT proto-oncogene receptor tyrosine kinase (KIT) and its ligand stem cell factor (SCF, also known as KITLG) are also essential for migration and proliferation as well as for survival of melanoblasts; other factors as B-cell lymphoma 2 Modification of the human hair: modulators of melanogenesis as agents of colour changing 15 MITF is the regulator of many genes associated with the pigmentary process; additionally, it also controls the expression of many other genes involved in differentiation, proliferation, and apoptosis [66–68]. MITF is a basic helix-loop-helix leucine zipper (bHLH-ZP) transcription factor that belongs to the MYC superfamily. At least 10 isoforms of MITF have been described in humans, with MITF-M being the specific isoform of the melanocyte lineage. As other bHLH transcription regulators, it recognizes the nucleotide consensus sequence CANNTG (E-box), with the association being maximal if this motif is 5’-flanked with T on either strand. At the N-terminus (NTAD), MITF exhibits the transcription activation domain, essential for interaction with transcriptional coactivators p300/CBP (CREB-binding protein). In turn, the transcription of MITF is controlled by a variety of regulators capable of binding to specific DNA sequences in its M promoter region: SOX9, SOX10, CREB (cAMP response element-binding protein), PAX3, LEF1 (lymphoid enhancer-binding factor 1), ZEB2 (zinc finger E-box binding protein 2), one cut domain 2 (ONECUT2) and MITF itself [61,69–75]. I.4.1 MELANOSOMES BIOGENESIS Melanosomes are intracellular membrane-bound organelles, produced through the action of the Golgi and rough endoplasmic reticulum, inside which melanin synthesis takes place. These organelles are part of the secretory/endocytic pathway and, by sharing many common features with lysosomes [acidic luminal pH, presence of lysosomal-associated membrane proteins (LAMPs) in their outer membrane, among others], they were thought to be assembled via a similar process. However, despite the many common features, melanosomes also express an array of exclusive proteins (melanogenesis-related proteins, as well as certain LAMPs) and they present distinctive morphological characteristics (manifestation of an ellipsoid shape and internal structures composed of lamellae or regularly striated filaments). Therefore, it is presumed that the biochemical process of melanosome biogenesis forms a separate lineage [76– 79]. The development and maturation of melanosomes proceeds through four stages, characterized by unique ultrastructural morphology – Figure I.3. These phases can be divided into two main steps: unpigmented and pigmented. The unpigmented step comprises the stages I and II during which the melanosome matrix is formed; at this point, melanosomes are referred to as immature or pre-melanosomes compartments. The pigmented phase comprises stages III and IV, in which the synthesis of melanin starts upon the formed matrix; at this point, melanosomes are denoted as mature/late compartments. Melanosomes are assembled in the perinuclear region, near the Golgi stacks, receiving all enzymatic and structural proteins required for melanogenesis [80,81]. Chapter I: A comprehensive review of human hair pigmentation 16 Figure I.3. Schematic representation of melanosomes biogenesis. AP-1/3: adaptor-related protein complex 1/3; ATP7A: ATPase copper transporting alpha; BLOC-1/2: biogenesis of lysosomal organelles complex 1/2; ER: endoplasmic reticulum; GPR143: G protein-coupled receptor 143; ILV: intralumenal vesicles; MITF: microphthalmia-associated transcription factor; MLANA: protein melan-a; OCA2: oculocutaneous albinism 2 protein; PMEL: premelanosome protein; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. In Stage I, immature melanosomes are spherical vacuolar domains of early endosomes that harbour intralumenal vesicles (ILVs) formed by invagination of the limiting membrane. The irregular arrays of amyloid fibrils that emanate from the ILVs make them distinct from early endosomes in other cells. Those fibrils are mainly composed of fragments of PMEL (premelanosome protein) [69,76,80–84]. In Stage II, Modification of the human hair: modulators of melanogenesis as agents of colour changing 17 pre-melanosomes are characterized by a fully formed melanosome matrix. PMEL fibrils are organized into arrays of parallel sheets, transforming the spherical Stage I melanosomes into elongated, elliptical organelles. Although no active melanogenesis takes place in stage II, they already contain melanogenic proteins (TYR, TYRP1, among others). The formation of PMEL fibrils segregates the melanosomal from the endosomal pathways. Although the mechanism regulating the separation are unclear, it seems to involve GPR143 (G protein-coupled receptor 143) and MLANA (protein melan-a) [82–87]. In Stage III melanosomes, melanogenesis starts, with the pigment being regularly and uniformly synthesized on the fibrillar matrix. This causes the darkening and thickening of the matrix fibrils. By Stage IV, the melanosomes are saturated with melanin and the internal fibrillar structure becomes completely masked by the pigment; mature melanosomes are ready to be transferred to adjacent keratinocytes [76,78– 80,85]. I.4.1.1 MAJOR PLAYERS IN MELANOSOMES BIOGENESIS As stated above, the three major players in melanosomes development and maturation are PMEL, MLANA and GPR143; altogether, they form a complex in early stages melanosomes assuring their proper composition and structure. I.4.1.1.1 PREMELANOSOME PROTEIN (PMEL) PMEL (also known as PMEL17, gp100, SILV and ME20) is a type I transmembrane glycoprotein, composed of a short signal peptide, a transmembrane domain, a long luminal N-terminal domain, and a short cytoplasmic C-terminal domain. It is the major functional component of melanosomes, and the only melanosomal protein necessary for the formation of very stable, β-sheet rich oligomeric structures that act as scaffold upon which melanin is polymerized, condensed, and stored. PMEL fibrils takes part in the maintenance of melanosome integrity by preventing highly reactive (cytotoxic) melanin intermediates to freely diffuse within the organelle. The sequestering of melanin intermediates has also been shown to accelerate melanin synthesis and its condensation in PMEL fibrils may even optimize the transfer from melanocytes to keratinocytes. Despite its functional role, the amyloid nature of such fibrils is challenging for melanocytes since an incorrect formation and organization leads to toxicity. Thus, the processing, trafficking, sorting, and fibrillation of PMEL is highly regulated to avoid aggregation in the wrong compartments [83,85,87,88]. PMEL is synthesized in the ER and firstly modified there by removal of the signal peptide, the addition of Chapter I: A comprehensive review of human hair pigmentation 18 N-glycosylations and the formation of disulphide bonds. Then, PMEL is exported to the Golgi complex and further modified by the addition of O -glycosylations. PMEL also undergoes multiple proteolytic processing in the Golgi apparatus and post Golgi compartment by a proprotein convertase (PC), forming a large fibrillogenic Mα fragment which remains linked to the Mβ fragment by a disulphide bond [85,89–92]. From the trans-Golgi network (TGN), PMEL is targeted to endosomes and then to pre-melanosomes where it is cleaved by the β secretase BACE2 (β-site APP-cleaving enzyme 2); the adaptor-related protein complex 2 (AP-2) is required for the transport of PMEL from the TGN to endosomes and for melanosomal accumulation. The cleavage releases the Mα fragment associated to a portion of the Mβ fragment (MβN) and a C-terminal fragment (CTF). The latter is rapidly degraded in the lysosomes while the MβN fragments suffers another proteolytic processing, required for proper fibril formation. This processing is known to be mediated by ADAM17 (ADAM metallopeptidase domain 17), a disintegrin and metalloprotease, from a family of proteases known to be involved in ectodomain shedding and for playing a role in cellular processes as adhesion and migration; besides ADAM17, other proteins are also thought to be involved in the processing of MβN fragments [84,85,93–95]. Finally, fully processed PMEL fibrillogenic fragments are sorted to the ILVs that, allowing the loading and concentration of such fragments at their surface, provide a favourable environment for the nucleation of fibrils. Unlike most proteins, the sorting of PMEL to ILVs does not require ubiquitylation or the activity of the ESCRT (endosomal sorting complexes required for transport) machinery, being regulated by Apolipoprotein E (APOE) and tetraspanin CD63. Along with the maturation of PMEL fibrils into sheets, the ILVs disappear, probably by fusing with the membrane of melanosomes or degradation by lipases [81,85,96,97]. I.4.1.1.2 PROTEIN MELAN-A (MLANA) MLANA (also known as melanoma antigen recognized by T cells 1 or MART-1) is an integral membrane protein. Besides melanosomes, it can be found in late endosomes, lysosomes, among others. MLANA forms a complex with PMEL, affecting its expression, trafficking, stability, and processing. MLANA also interacts biochemically with GPR143, providing stability to this receptor [81,86,98]. I.4.1.1.3 G PROTEIN-COUPLED RECEPTOR 143 (GPR143) GPR143 (also known as oculocutaneous albinism 1 protein or OA1) is a pigment-cell specific transmembrane G protein-coupled receptor found in the membrane of melanosomes and other organelles like late endosomes and lysosomes [99,100]. This receptor acts as a biosensor of melanosome Modification of the human hair: modulators of melanogenesis as agents of colour changing 19 maturation and provides regulation of their size and number by triggering the activation of a transcription cascade, involving MITF, that culminates in sustained PMEL expression [101]. It has also been proposed that GPR143 activity delays the delivery of PMEL-containing endosomes to the lysosomes in order to allow time for commitment with melanosome biogenesis [100]. Like many membrane GPCR, GPR143 might also couple to melanosomal ion channels and modulate ion gradients and membrane potential, important in fusion and fission events that regulate protein trafficking and organelle biogenesis [102]. I.4.1.2 PROTEIN TRAFFICKING TO MELANOSOMES The maturation of melanosomes requires the delivery of several melanogenic proteins. Most of them are integral membrane glycoproteins, synthesized by ribosomes associated with the ER and become integrated in its membrane across which they translocate. In the ER, such proteins are N-glycosylated, folded, and assembled by enzymes and chaperones. Melanosomal proteins that are properly folded in the ER transit to the Golgi complex for functional modifications and from there to the endosomes; failure to be properly processed results in retro translocation of these melanosomal proteins in the ER and degradation by the proteasome [77,81]. The molecular intervenient on the transition of melanogenic proteins from the Golgi to endosomes are mainly unknown, but the interaction between GIPC (GAIP interacting protein COOH-terminus) and APPL (adaptor protein containing pleckstrin homology, phosphotyrosine binding domain, and leucine zipper motif) seems to be important, at least in the trafficking of TYRP1 [103,104]. Within endosomes, melanogenic proteins are usually loaded into vesicular carriers for delivery into melanosomes through a process involving the fusion of their limiting membranes. Two transport pathways involving multisubunit protein complexes are known to allow the trafficking of proteins from the endosomes to melanosomes – Figure I.3. One pathway is mediated by adaptor-related protein complexes, namely AP-1 and AP-3 [81]. The AP-1 complex is involved in the trafficking of TYR and TYRP1. The function of AP-1 requires the recruitment of microtubule motor protein KIF13A (kinesin family member 3A) to position endosomes near maturing melanosomes to facilitate the formation of transient connections between them for cargo transfer; two other molecular motor proteins, myosin VI (MYO6) and MYO1B were also shown to be part of the machinery that regulates the delivery of melanogenic enzymes to maturing melanosomes [81,105–108]. The AP-3 complex is required for the efficient delivery of TYR and OCA2 (oculocutaneous albinism 2 protein), a protein involved in the regulation of melanosomal pH and a major determinant of mammalian pigmentation; OCA2 itself have also been pointed out as being Chapter I: A comprehensive review of human hair pigmentation 20 implicated in the trafficking and processing of tyrosinase and related proteins [81,108–115]. The other transport pathway is mediated by biogenesis of lysosomal organelles complexes (BLOC). BLOC-1 allows the delivery TYRP1 and ATP7A to melanosomes; ATP7A is a copper transporting ATPase that supplies Cu2+ to the Cu2+-dependent enzyme TYR, thus sustaining melanin synthesis The trafficking of TYRP1 by this pathway was shown to be dependent on the ESCRT-I complex. BLOC-2 is implicated in the delivery of TYR and TYRP1 to the melanosome [81,116–118]. SNAREs (Soluble N-ethylmaleimide-sensitive-factor Attachment protein Receptor) are also reported to be involved in the delivery of proteins into the melanosomes due to their role in the membranes fusion process The interaction of melanosomal SNAREs syntaxin 3 and SNAP23 (synaptosomeassociated protein 23) with SNARE VAMP7 (vesicle associated membrane protein 7) on TYRP1-containing vesicles regulates its trafficking to melanosomes. Additionally, the interaction of melanosomal VAMP7 with endosomal syntaxin 13 or VARP (VPS9-domain ankyrin repeated protein) is also accountable for proper delivery of TYR and TYRP1 to melanosomes [81,119,120]. The Rab family of small GTPases are also known to facilitate the trafficking of proteins to the melanosomes. By recruiting tethering/docking and fusion factors, as well as actinand microtubule-based proteins, they establish molecular interaction between melanosomes and cargo vesicles and coordinate their positioning for cargo delivery. RABs that have been involved in the trafficking of melanogenic proteins, but also in other aspects of melanosomes biogenesis, are: RAB38/RAB32, RAB11, RAB9, RAB7, RAB33A, among others [81,121–128]. I.4.2 BIOCHEMICAL PATHWAY OF MELANOGENESIS Melanogenesis in vivo produce mixtures of two chemically distinct types of melanins through a process often referred to as mixed melanogenesis. Eumelanin is a black-to-brown highly polymerized pigment while pheomelanin is reddish-brown to yellow and less polymerized. Pheomelanin is reported to be synthesized first and eumelanin is subsequently deposited on the preformed pheomelanin, giving melanin granules. In hair follicles, the synthesis of pheomelanin can be completely switched-off during eumelanogenesis, but pheomelanogenesis never occurs with total inhibition of eumelanin synthesis. The total amount of melanins produced is proportional to the availability of L-Tyr and the dopaquinone (DQ) formed, which in turn is proportional to the activity of tyrosinase. The ratio of eumelanin to pheomelanin is also determined by the activity of tyrosinase, but mainly influenced by the availability of cysteine [2,3,129–131]. Modification of the human hair: modulators of melanogenesis as agents of colour changing 21 The initiation of melanogenesis requires L-Tyr that can be directly transported from the extracellular space or synthesized inside melanocytes through hydroxylation of L-phenylalanine by PAH (EC 1.14.16.1). The activity of PAH depends on the essential cofactor/electron donor 6-BH4, that melanocytes have full capacity for de novo synthesis and recycle – Figure I.4. Figure I.4. Metabolism of L-phenylalanine to L-tyrosine via phenylalanine hydroxylase (PAH). The activity of PAH depends on the co-factor (6R)-L-erythro 5,6,7,8 tetrahydrobiopterin (6-BH4), which melanocytes have full capacity for de novo synthesis (blue arrows) and recycling (green arrows). 4a-OH-BH4 DH: 4ahydroxy-tetrahydrobiopterin dehydratase; 4a-OH-BH4: 4a-hydroxy-tetrahydrobiopterin; 7-BH4: (7R)-Lerythro-5,6,7,8-tetrahydrobiopterin; DHPR: dihydropteridine reductase; GFRP: GTP-cyclohydrolase I feedback regulatory protein; GTP-CH-I: GTP-cyclohydrolase I; q-BH2: quinonoid dihydropterin; SLC7A5: solute carrier family 24, member 5. The rate-limiting enzyme in de novo synthesis of 6-BH4 is GTP-cyclohydrolase I (GTP-CH-I, EC 3.5.4.16). GTP-CH-I is stimulated by GFRP and, in turn, GFRP can be stimulated by L-phenylalanine or inhibited by Chapter I: A comprehensive review of human hair pigmentation 22 6B-H4. [132–135]. The recycle of 6B-H4 commences with the formation of L-Tyr from L-phenylalanine via PAH. This leads to the generation of 4a-hydroxy-tetrahydrobiopterin (4a-OH-BH4) which through a nonenzymatically process can generate (7R)-L-erythro-5,6,7,8-tetrahydrobiopterin (7-BH4). On the other hand, its enzymatic metabolization by 4a-hydroxy-tetrahydrobiopterin dehydratase (4a-OH-BH4 DH, EC 4.2.1.96) originates quinonoid dihydropterin (q-BH2) that it is reduced through the NADH-dependent dihydropteridine reductase (DHPR, EC 1.5.1.34) or reduced glutathione (GSH) to 6-BH4. Additionally, thioredoxin reductase (TR, EC 1.8.1.9) can controls the redox status of 6-BH4/6-biopterin, which has implications in the regulation of melanin production; reduced 6-BH4 can bind to tyrosinase and inhibit the enzyme while oxidized 6-biopterin has no effect on it [134,136–138]. For the synthesis of melanin to occur, L-Tyr must be taken into melanosomes. SLC7A5 (solute carrier family 7, member 5) encodes for a member of the L-type aminoacids transporter family that it is specialized in the transport of histidine, tryptophan, tyrosine and neutral aminoacids. Although the inhibition of this transporter causes in vitro loss of pigmentation, its presence in melanosomes could not yet be validate and thus, the nature of tyrosine transporter inside melanosomes still remains elusive [139]. In melanosomes, melanogenesis begin with synthesis of L-3,4-dihydroxyphenylalanine (L-DOPA) from L-Tyr and subsequent oxidation to DQ – Figure I.5. The hydroxylation of L-Tyr to L-DOPA is catalysed by the multi-functional enzyme tyrosinase (EC 1.14.18.1), the most important enzyme in the melanogenic pathway. Alternative mechanisms for L-DOPA formation include the reduction of DQ back to L-DOPA or even direct hydroxylation of L-Tyr by tyrosine hydroxylase isoform I (TH I, EC 1.14.16.2) [8,130,140,141]. From here, the mixed melanogenesis proceeds, using the same precursor (DQ) for the synthesis of pheomelanin and eumelanin. Pheomelanin is spontaneously produced from DQ if cysteine is present in concentrations above 1 μM – Figure I.5. The first step is the reductive addition of cysteine to DQ, giving rise to cysteinyldopa (CD) isomers 5-S-cysteinyldopa (5-S-CD) and 2-S-cysteinyldopa (2-S-CD) in a ratio of 5 to 1; an alternative route for the CD isomers formation is the conjugation of DQ with glutathione followed by the hydrolysis of the resultant glutathionyldopa by glutamyltranspeptidase. The second step is the redox exchange between CD isomers and DQ to produce CD-quinones (and DOPA), followed by their cyclization through dehydration to form ortho-quinonimine (QI). Then, QI is rearranged with or without decarboxylation to form 1,4-benzothiazine intermediates. The ring closure to yield such intermediates may involve a peroxidase/H2O2 reaction or TYR-catalysed oxidation; in this context, one enzyme indirectly affecting the production of pheomelanin is catalase (EC 1.11.1.6) that regulates hydrogen peroxide removal. The last Modification of the human hair: modulators of melanogenesis as agents of colour changing 23 step involves the polymerization of benzothiazine intermediates to pheomelanin. The production of pheomelanins is preferred over the production of eumelanin in the presence of CD isomers concentrations above 10 μM [130,142–144]. Figure I.5. Biosynthetic pathways involved in the production of eumelanin and pheomelanin.in human melanosomes. Spontaneous reactions are denoted as black arrows. Enzyme-catalysed reactions are signalized by blue arrows. CD: cysteinyldopa; L-DOPA: L-3,4-dihydroxyphenylalanine; TYR: tyrosinase; DCT: dopachrome tautomerase. Chapter I: A comprehensive review of human hair pigmentation 24 The production of eumelanin begins after most CD isomers and cysteine are depleted – Figure I.5. As a highly reactive ortho-quinone intermediate, DQ undergoes spontaneous intramolecular cyclization to give rise to cyclodopa. Then, cyclodopa rapidly undergoes a redox exchange with another DQ molecule to produce one molecule of dopachrome, a relatively stable intermediate, and one molecule of DOPA. In the absence of additional factors, dopachrome undergoes spontaneous decarboxylative rearrangement to form 5,6-dihydroxyindole (DHI) [145]. In the presence of DCT (EC 5.3.3.12), the tautomerization of dopachrome can also forms 5,6-dihydroxyindole-2-carboxylic acid (DHICA) [146–148]. However, although DCT can be found in the human epidermis and follicular melanocytes of many mammalians, human bulbar melanocytes (at least those of eumelanic phenotypes) do not express this protein and human hairs contain low yield of DHICA [149,150]. Both dihydroxyindoles (DHI and DHICA) are further oxidized and assembled by cross-linking reactions into eumelanin polymers. Oxidative polymerization of DHI is catalysed directly by tyrosinase and indirectly by DQ. Oxidation of DHICA in mice is catalysed by TYRP1, but the human homologue does not act the same. Although its function is not completely clear, in humans, TYPR1 appears to ensure appropriate processing and stabilization of tyrosinase, to maintain melanosomal structure integrity and to function as tyrosine hydroxylase at low concentrations of substrate. Thus, tyrosinase seems to be responsible for the oxidation and incorporation of DHICA units into human eumelanins, along with peroxidase (EC 1.11.1.7) and PMEL [88,151–158]. I.4.2.1. MAJOR PLAYERS IN MELANINS SYNTHESIS As already been exposed, the synthesis of melanins cannot occur without the activity of TYR and TRP1. In turn, the ratio of pheomelanin to eumelanin is determined by the availability of cysteine, which correlates with the activity of cystine transporters. I.4.2.1.1. TYROSINASE (TYR) Tyrosinase is a type 3 copper containing enzyme, involved in several steps of melanin synthesis. This membrane glycoprotein can be divided into three domains: a large N-terminal intra-melanosomal domain (that possesses a catalytic subdomain), a single transmembrane α-helix domain (anchoring tyrosinase in melanosome membrane) and a small C-terminal cytoplasmic tail. These three regions are conserved among TYR and TYR-related proteins and they follow quite similar processing and trafficking pathways. The maturation of tyrosinase includes a signal sequence cleavage, heavy glycosylation (N-glycosylations at Asn272, 319 and 353 are particularly important for proper maturation and stability), disulphide bond Modification of the human hair: modulators of melanogenesis as agents of colour changing 31 I.4.3.1.3. MEMBRANE FUSION-MEDIATED TRANSFER The fusion model assumes that melanocytes and keratinocytes fuse their membranes to connect their cytoplasms. This fusion appears to involve filopodia that extend from the dendrite of melanocytes and adhere to keratinocytes, resulting in the formation of conduits that admit unidirectional transport of melanosomes [229]. Although this kind of organelle transport exist in other cells, and transfer of melanosomes at filopodia locations have been observed, definitive proof of membrane fusion and melanosome transport across the channels formed has yet to be given [222]. I.4.3.1.4. MEMBRANE VESICLE-MEDIATED TRANSFER The membrane vesicle-mediated transfer model reports that melanosome-containing vesicles are released into the extracellular space, adhere to keratinocytes and then, they are internalized by phagocytosis [222,223] It has also been hinted that multiple melanosomes can be concentrated at the filopodia for being released inside globules from various areas of the dendrites [230,231]. Although the exact mechanism (or combination of mechanisms) of melanin transfer remains unclear, phagocytosis (a process of cellular engulfment of particles with a diameter of more than 0.5 μM) is a mostly necessary step. The phagocytic nature of keratinocytes is widely known and, regarding melanin transfer, phagocytosis has been reported to be modulate mainly by the protease-activated receptor 2 (PAR2, also known as coagulation factor II receptor-like 1 or F2RL1) [232–237]. PAR2 is a seventransmembrane G protein-coupled receptor, related to the thrombin receptors. The proteolytic cleavage at the extracellular N-terminal moiety exposes a new N-terminus that acts as a tethered ligand that binds the receptor leading to its activation. This activation induces the secretion of serine proteases by keratinocytes that are responsible in the first place for the cleavage of PAR2, thus creating a positive feedback loop [232,233]. Despite PAR2 performs the leading role in phagocytosis associated with melanin transfer, it is not expected to be the only receptor that modulates the process as suggested by the incomplete inhibition of melanin transfer upon treatment with serine protease inhibitors. Important to mention, the modulation of pigmentation by PAR2 goes beyond the direct activation of the phagocytic pathway in keratinocytes: PAR2 activation also increases melanocyte dendricity (upon stimulation of prostaglandins PGE2 and PGF2α release by keratinocytes) and the expression of SCF, a paracrine regulator of melanogenesis [238,239]. Chapter I: A comprehensive review of human hair pigmentation 32 I.5. REGULATION OF FOLLICULAR MELANOGENESIS Melanogenesis is under a complex control involving several positive and negative regulators/factors that act via autocrine, paracrine, and endocrine mechanisms – Figure I.8. Although much of the available data regarding the control of melanogenesis pertains to human epidermal or murine follicular melanocytes, it is thought that such regulators contribute in a similar way to human hair pigmentation; given that active follicular melanocytes are located beyond the reach of its direct stimulation, one notable exception is the UV radiation, the principle regulator of melanin synthesis in the epidermis [69,240]. Figure I.8. Major molecular pathways involved in the regulation of melanin synthesis in melanocytes. Modification of the human hair: modulators of melanogenesis as agents of colour changing 33 I.5.1. PRO-OPIOMELANOCORTIN (POMC)-DERIVED PEPTIDES The hair follicles are local sources and targets for pro-opiomelanocortin (POMC)-derived peptides; although originally discovered in the anterior pituitary, these peptides are also expressed and secreted by melanocytes and keratinocytes [69,240–242]. POMC transcription and translation is hair cycledependent and increases significantly in the anagen phase. The proteolytic processing of POMC by PC1 produces the peptides adrenocorticotropin (ACTH) and β-Lipotropin (β-LPH), with the latter being further processed to β-endorphin (β-END) and β-melanocyte-stimulating hormone (β-MSH). PC2 cleaves the first 14 amino acids of the ACTH sequence to generate ACTH(1-14)OH, the precursor of α-MSH [243–245]. POMC-derived products are the main regulators of follicular melanogenesis, and even unprocessed POMC is reported to stimulate melanin production – Figure I.8A. Additionally, POMC-derived peptides also influence the hair growth [242,246–250]. ACTH, α-MSH and β-MSH share an essential core peptidic sequence of four amino acids (His-Phe-ArgTrp) that allows the binding to melanocortin receptors. While the latter acts on melanocortin 4 receptor (MC4R), ACTH and α-MSH bind to MC1R, a G protein-coupled membrane receptor abundantly expressed in melanocytes of hair follicles. Nonetheless, activation of both receptors leads to the stimulation of adenylyl cyclase (AC), an enzyme that catalyses the conversion of ATP to cAMP (cyclic adenosine monophosphate) [69,240,251–254]. Then, cAMP binds to the regulatory subunit of PKA (protein kinase A), allowing the catalytic subunit to be liberated and activated; termination of this cellular signalling is provided by phosphodiesterases (PDEs), which hydrolyse free and PKA-bounded cAMP, leading to reassociation of the two subunits [255]. Activated PKA is translocated to the nucleus where it induces the phosphorylation of CREB at Ser133. Phosphorylated CREB activates the expression of specific genes containing the consensus CRE (cAMP responsive element) sequences in their promoters, as MITF ; PKA also phosphorylates the nuclear transcriptional coactivator CBP that interacts with CREB family proteins for PKA-dependent gene expression. Consequently, the increased transcription of MITF upregulates the expression of several genes involved in synthesis of melanin, but also in the regulation of melanogenesis (as MC1R ), melanocyte proliferation and dendrite formation [69,240,252,256,257]. Besides direct phosphorylation of CREB, PKA also increases the transcription of MITF by inhibiting (via phosphorylation at Ser587) salt-inducible kinase 2 (SIK2) activity. SIK2 promotes the phosphorylation of CREB-regulated transcription coactivator 1 (CRTC1), preventing its translocation from the cytoplasm to the nucleus, an essential step for CREB-mediated gene expression. CRTC1 and SIK-2 have been shown to be fundamental determinants of melanogenic program in mice [258,259]. Chapter I: A comprehensive review of human hair pigmentation 34 Figure I.8A. Highlight of molecular pathways involved in the regulation of melanin synthesis by proopiomelanocortin (POMC)- derived peptides. α/β-MSH: α/β-melanocyte-stimulating hormone; β-END: βendorphin; AC: adenylyl cyclase; Akt: serine/threonine kinase; cAMP: cyclic adenosine monophosphate; CBP: CREB-binding protein; CREB: cAMP response element-binding protein; CRE: cAMP responsive element; CRF(R): corticotropin-releasing factor (receptor); CRTC1: CREB-regulated transcription coactivator 1; GSK3β: glycogen synthase kinase 3β; Ker: keratinocytes; MC1/4R: melanocortin 1/4 receptor; Mel: melanocytes; MITF: microphthalmia-associated transcription factor; PC1/2: proprotein convertase 1/2; PI3K: phosphatidylinositol 3-kinase; PKA/C: protein kinase A/C; SIK2: salt-inducible kinase 2; TRH(R): thyrotropin-releasing hormone (receptor); TYR: tyrosinase; TYRP1: tyrosinase related protein 1. Modification of the human hair: modulators of melanogenesis as agents of colour changing 35 The peroxisome proliferator-activated receptor γ coactivator-1α (PGC-1α) and PGC-1β are other transcriptional regulators that are modulated by the cAMP/PKA pathway. α-MSH signalling increase the expression of PGC-1α and stabilize both PGC-1α and PGC-1β proteins, probably through phosphorylation mediated by PKA, with consequent stimulation of MITF, tyrosinase expression and synthesis of melanin [260]. PGC-1α and PGC-1β coactivate many transcription factors, including the nuclear hormone peroxisome proliferator-activated receptor γ (PPARγ), that also has shown to be involved in the 𝛼-MSHinduced melanogenesis; other PPARs subtypes are reported to be expressed in melanocytes and PPARα, but not PPARβ/δ, is known to contribute to the control of melanogenesis [261–264]. PKA also inhibits the activity of phosphatidylinositol 3-kinase (PI3K) and one of its key effectors, the serine/threonine kinase Akt; the activation of Akt depends on the phosphorylation of its Thr308 and Ser473 residues upon binding to the PI3K phospholipid products. Inactive Akt is incapable of phosphorylate the glycogen synthase kinase 3β (GSK3β) at Ser9 which promotes its inactivation. Thus, GSK3β can phosphorylate MITF at Ser298, which facilitates its binding to the TYR promotor and stimulates melanogenesis. Curiously, PKA can promote the degradation of GSK3β (phosphorylation at Ser9) in a cross talk with the Wnt/β-catenin signalling pathway, addressed below (Section I.5.2 and Figure I.8B) [265–268]. The inhibition of PI3K/Akt pathway by PKA not only avoids the degradation of GSK3β, but also prevents the activation of serine/threonine kinase p70S6K1, and mammalian target of rapamycin (mTOR) along with its complex (mTORC1). Both factors have been implicated in the negative regulation of melanogenesis and the cAMP-induced melanogenesis has shown to be, at least in part, mediated by their inhibition [269–274]. The remaining POMC-derived peptide, β-END, has shown to positively regulate melanogenesis, proliferation and dendricity in melanocytes of the epidermis and hair follicles by binding to its μ-opiate receptor. Contrarily to other POMC-derived peptides, β-END is not expected to promote its melanogenic modulatory effect through the cAMP/PKA signalling pathway; in fact, signalling through μ-opiate downregulates the level of cAMP. A PKC-dependent pathway (addressed bellow) have been suggested as a downstream mechanism for stimulation of melanogenesis by β-END [69,240,246,275]. I.5.1.1. CORTICOTROPIN-RELEASING FACTOR (CRF) Corticotropin-releasing factor (CRF, also known as corticotropin-releasing hormone or CRH) is the most proximal element of the hypothalamic-pituitary-adrenal axis (HPA). The binding of CRF to its receptors mediates the POMC expression and subsequent production of the derived peptides. Both skin and hair Chapter I: A comprehensive review of human hair pigmentation 36 follicles contain an equivalent to the HPA axis, with keratinocytes and melanocytes expressing CRF and the corresponding G protein-coupled corticotropin-releasing factor receptors (CRFR); isoforms of the CRF1R subtype are expressed in the epidermis while isoforms of the CRF2R subtype are expressed in the hair follicle. CRF can induce the synthesis of POMC through signalling involving PKA and PKC – Figure I.8A [69,240,276,277]. I.5.1.2. THYROTROPIN-RELEASING HORMONE (TRH) Thyrotropin-releasing hormone (TRH) is the most proximal regulatory element of the hypothalamicpituitary-thyroid axis (HPT), controlling the production of thyroid hormone. TRH is recognized as a potent stimulator of POMC expression, with human skin and hair follicles being reported to transcribe it along with its receptor (TRHR). The expression of TRH stimulates growth, melanin synthesis and melanocyte dendricity in hair follicles; in the epidermis, TRH does not appear to affect melanogenesis. In hair follicles, the expression of TRHR is confined to the inner root sheath, being absent in the follicular pigmentary unit. This suggests that THR stimulate the production of POMC in keratinocytes, with the derived peptides posteriorly acting on melanocortin receptors of melanocytes – Figure I.8A. Alternatively, TRH have also been proposed to bind directly to MC1R, which could explain the reported cases of normal hair pigmentation in the absence of melanocortin synthesis [278,279]. Besides TRH, the related thyroid hormones have also shown to directly affect human hair pigmentation [280]. I.5.2. WNT PROTEINS The Wingless-type MMTV integration site family (Wnt) consist of cysteine-rich lipoglycoproteins members that act on G protein-coupled frizzled (FZD) receptors. The binding to FDZ receptors activates an intracellular cascade of events involving (canonical form) or not (noncanonical form) the transcription regulator β-catenin [69]. The non-canonical pathways are diverse but poorly characterized; melanin suppression by WNT5 is known to be dependent on noncanonical Wnt/Ror2 pathway [281]. In the canonical pathway, the binding of WNTs to FZD receptors and low density lipoprotein-related protein (LRP) 5/6 coreceptor inhibits the constitutive degradation of β-catenin by GSK3β, promoting its accumulation; further stability of β-catenin is provided by PKA through phosphorylation of the Ser675 – Figure I.8B [268,282]. β-catenin is then translocated into the nucleus, where it interacts with T-cell factor (TCF)/LEF transcription factors to upregulate the transcription of MITF . β-catenin also interacts directly with the MITF (phosphorylating the Modification of the human hair: modulators of melanogenesis as agents of colour changing 37 Ser298 residue) and increase binding of MITF to the M box of the TYR promotor [69,283,284]. Figure I.8B. Highlight of molecular pathways involved in the regulation of melanin synthesis by WNT proteins. GSK3β: glycogen synthase kinase 3β; LEF: lymphoid enhancer-binding factor; MITF: microphthalmia-associated transcription factor; PKA: protein kinase A; TCF: T-cell factor; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. In the absence of WNTs, β-catenin is phosphorylated (Ser33, Ser37, Ser45 and Thr41) by a multiprotein complex containing axin, adenomatous polyposis coli (APC), casein kinase Iα (CKIα) and GSK3β; these phosphorylations lead to ubiquitination and proteasomal degradation [268,285,286]. The activation of Chapter I: A comprehensive review of human hair pigmentation 38 the Wnt/β-catenin signalling pathway by WNT1, WNT3a, WNT7A/B or WNT10B have been implicated in the upregulation of melanogenesis, melanocyte differentiation and proliferation and hair follicle regeneration [287–292]. Many other ligands have also shown to modulate the Wnt signalling pathway, and consequent production of melanin, by interacting with WNT proteins: Dickkopf 1 (DDK1) [293,294] WNT inhibitory factor 1 (WIF1) [295] and secreted frizzled-related protein 2 (SFRP2) [296]. I.5.3. STEM CELL FACTOR (SCF) SCF is a growth factor secreted by keratinocytes and fibroblasts. SCF is the specific ligand of KIT, which is expressed in various types of cells including melanocytes. The binding of SCF to the extracellular cellular domain of KIT prompts its dimerization, activation of intrinsic tyrosine kinase activity and autophosphorylation. Then, KIT can phosphorylate various substrates that lead to the activation of the mitogen-activated protein kinases (MAPK) signalling pathways and modulation of melanocyte proliferation, differentiation and melanogenesis. Of note, SCF/KIT signalling can also activate PI3K, which effect on melanogenesis was discussed before (Section I.5.1 and Figure I.8A) [69,240,297–300]. There are three well-characterized subfamilies of the MAPK superfamily known to have crucial roles in melanin synthesis: the extracellular signal-regulated kinases (ERK, mainly 1/2), the c-Jun N-terminal kinases (JNKs, mainly 1/2, also known as stress-activated protein kinases or SAPKs) and the p38-MAPK – Figure I.8C [69,301,302]. Activation of ERK, which can additionally be achieved through cAMP/PKA or PKC-mediated signalling, is known to cause phosphorylation of MITF at Ser73, allowing the recruitment of the transcriptional coactivator CBP/P300 of CREB and consequent transcription of melanogenic enzymes. However, such phosphorylation, along with another one at Ser409 induced by RSK (p90 ribosomal S6 kinase, ERK downstream activation), also leads to MITF destabilization via increased ubiquitination, followed by proteasome-dependent degradation; in due course, the production of melanin is decreased [69,71,303–309]. Several lipid second messenger have shown to modulate the production of melanin, mainly through the ERK signalling pathway: sphingosylphosphorylcholine [272,310–313], ceramide [314–316], sphingosine-1-phosphate [317,318] and lysophosphatidic acid [319]. Activation of JNK has also been reported to inhibit melanogenesis via degradation of MITF. In addition, the antimelanogenic effect of JNK was linked to the blockage of nuclear translocation of CRTC3 (CREBregulated transcription coactivator 3) and further impairment of the transcriptional activity of CREB [320]. Activation of p38-MAPK induces phosphorylation of CREB, via downstream MSK1 (mitogenand stressactivated protein kinase 1), activating MITF that promotes the transcription of melanogenesis-related genes and the increase in melanin synthesis. However, there are also some evidence demonstrating the Modification of the human hair: modulators of melanogenesis as agents of colour changing 39 involvement of p38-MAPK in the inhibition of melanin production by proteasome-dependent degradation of tyrosinase and other related proteins [300,321–323]. Figure I.8C. Highlight of molecular pathways involved in the regulation of melanin synthesis by stem cell factor (SCF). CREB: cAMP response element-binding protein; ERK: extracellular signal-regulated kinases; JNK: c-Jun N-terminal kinases; KIT: KIT proto-oncogene, receptor tyrosine kinase; PI3K: phosphatidylinositol 3-kinase; PKA/C: protein kinase A/C; MAPK: mitogen-activated protein kinases; MITF: microphthalmia-associated transcription factor; RSK: p90 ribosomal S6 kinase; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. Chapter I: A comprehensive review of human hair pigmentation 40 I.5.4. ENDOTHELIN 1 (ET-1) ETs are peptides present in a broad range of tissues, including the skin. The three known ETs isopeptides (ET-1,2 and 3) are obtained through cleavage of preproendothelin by prohormone convertases. ET-1 is synthesized and secreted by keratinocytes, acting on G protein-coupled receptor EDNRB of melanocytes with ensuing stimulation of proliferation, melanogenesis and dendricity – Figure I.8D [69,240,324–328]. Figure I.8D. Highlight of molecular pathways involved in the regulation of melanin synthesis by endothelin 1 (ET-1). CREB: cAMP response element-binding protein; DAG: diacylglycerol; EDNRB: endothelin receptor type B; IP3: inositol–triphosphate; MAPK: mitogen-activated protein kinases; MITF: microphthalmia-associated transcription factor; PKC: protein kinase C; PLCγ: phospholipase Cγ; TYR: tyrosinase; TYRP1: tyrosinase related protein 1. Modification of the human hair: modulators of melanogenesis as agents of colour changing 47 [373,377,386,387,378–385]. Many regulators of pigmentation, with unknown targets, have also been disclosed: miR-9, miR-21, miR-28, miR-130b, miR-139-5p, miR-155, miR-182, miR-193, miR-206, miR221, miR-222, miR330-5p, miR-335, miR-365 and miR-455 [384,388–390]. Of note, miR-230, miR3305p, miR-675 miR-3196 are not produced by melanocytes, being delivered from keratinocytes via exosomes [373,374,389]. LncRNAs are greater than 200bp in length and regulate diverse biological functions and also some microRNAs [69]. Some lncRNAs (CD1C-2:1, H19, TCONS_00049140, UCA1) have already been shown to have an important role in melanogenesis, although their role in not yet fully described [391–393]. I.5.10. OTHER REGULATORS Zinc α-2-glycoprotein is produced by keratinocytes and it is proposed that it may play a part in the negative regulation of melanin production [394]. An undisclosed member of the heat shock proteins (HSP)70 family have been shown to suppress melanin production in vitro and in vivo through direct interaction with MITF; HSP are constitutively expressed in the skin and confer protection against stressors [395]. Conversely, HSP70-1A expression in dark skin melanocytes was found to be higher compared to those of light skin phenotypes, contributing to skin colour diversity [396]. Ligands of the purinergic receptor type 2 X7 (P2X7) [397], odorant receptor 51E2 [398], olfactory receptor OR2A4/7 [399], aryl hydrocarbon receptor (AHR) [400] and Type I cannabinoid receptor (CB1) [401] are also expect to contribute to the regulation of melanogenesis; the expression of these receptors by human melanocytes have been proved, and their stimulation increased the production of melanin. I.6. DIVERSITY OF HUMAN HAIR NATURAL COLOUR The diversity in human hair colour arises mostly from the quantity and ratio of eumelanin and pheomelanin produced. Mostly eumelanin is the default hair pigmentation, being predominant in more than 90% of the human population. Eumelanic phenotypes range from black, dark brown, medium brown, light brown to blond. All these phenotypes contain small, nearly constant, amounts of pheomelanin (0.850.99 μg melanin/mg of hair), with the varying contents of eumelanin being accountable for the visual differences in hair colour. Dark brown hair reportedly contains eumelanin at 66% (14.6 μg/mg) the level of black hair (22.2 μg/mg), medium brown at 47% (10.4 μg/mg), light brown at 40% (8.7 μg/mg) and blond hair at 23% (4.7 μg/mg). Red hair is the only phenotype that contains comparable amounts of Chapter I: A comprehensive review of human hair pigmentation 48 eumelanin and pheomelanin: 3.8 and 4.7 μg/mg, respectively [2,3,402]. The activity of TYR (which, as discussed before, is greatly affected by melanosomal pH) and cysteine content of melanosomes have been proposed to play critical roles in determining the amount and ratio of the two human hair pigments. Those are influenced by several polymorphisms in a wide range of pigmentary genes. The red hair colour phenotype, characterized by abnormal high pheomelanin contents, is caused mostly by polymorphisms in MC1R , the master regulator of pigment-type switching. The most penetrant MC1R polymorphisms are R142H, R151C, R160W and D294H [2,403–408]. Such polymorphisms cause low levels of MC1R signalling, leading to downregulation of various pigmentary genes: TYR , TYRP1 , OCA2 , SLC24A5 and SLC45A2 ; CTNS is also expected to be downregulated, leaving high levels of cystine (and consequently, cysteine) inside melanosomes. Consequently, the melanosomes are cysteine-rich, acidic and with low tyrosinase activity, leading to the production of low to medium levels of pheomelanin and eumelanin [2,3,409,410]. Polymorphisms in ASIP also show strong association with red hair [2,3,403,404,411]. ASIP encodes for the agouti signalling protein, an antagonist of MC1R, that competes with α-MSH in binding to the receptor [252,256,257,412–415]. By impacting on the acidification of melanosomes, polymorphisms in SLC45A2 , SLC24A5 and OCA2 may additionally contribute to the red hair phenotype [2,3]. Furthermore, polymorphism in cystine/glutamate exchanger SLC7A11 have also been proposed to be important for pheomelanin production [403]. The blond phenotype is characterized by high levels of MC1R signalling, reduced activities of ion transporters, and full activity of CTNS. These suppressed, but pro-eumelanogenic conditions make melanosomes acidic and cysteine deficient, leading to the production of trace amounts of pheomelanin and low levels of eumelanin [3]. No single gene has been reported as the main responsible for this phenotype, with polymorphisms in a number of genes as ASIP , KITLG (encoding for SCF), MC1R , TPCN2 (encoding for TPC2) and TYRP1 being associated with blond hair and found to differentiate it from brown hair [2,403,423,424,411,416–422]. The dark hair phenotypes (brown to black) are characterized by high levels of MC1R signalling and full activities of ion transporters and CTNS. Under these conditions, melanosomes become neutral and cysteine deficient, leading to the production of trace amounts of pheomelanin and high levels of eumelanin [3]. Several studies have confirmed an association between dark hair and polymorphisms in genes such as SLC24A5 , SLC45A2 , HERC2 (HECT and RLD domain containing E3 ubiquitin protein ligase 2) , and IRF4 (interferon regulatory factor 4) [2,403,423,425,426]. Modification of the human hair: modulators of melanogenesis as agents of colour changing 49 I.7. AGE-INDUCED HAIR GREYING The greying of hair (canities) is one of the most obvious and common signs of aging, occurring to a varying degree in all individuals, regardless of gender or race. A rule of thumb for hair greying is that by 50 years of age, 50% of people have 50% grey hair. The term grey hair widely refers to an admixture of pigmented and non-pigmented (white) hairs, although sometimes a single hair fibre can show a progressive dilution from black, through grey to white over several hair cycles or within the anagen phase of a single cycle. Hair greying usually appears first at the temples, spreading to the vertex and then to the remainder of the scalp, affecting the occiput last. At its simplest, the pigment loss in greying hair follicles is due to a marked reduction in melanogenically active melanocytes in the hair bulb. Although it has not yet been completely understood why melanocytes are lost with aging, some mechanisms have been proposed. – Figure I.9 [1,5,6,8,36,427,428]. The traditional view proposes that depletion of hair follicle bulbar melanocytes correlates with oxidative stress. The melanogenic activity of bulbar melanocytes is likely to generate large amounts of reactive oxygen species (ROS), via the oxidation of tyrosine and DOPA to melanin [429]. Although melanocytes possess an efficient antioxidant system, it appears to become impaired with age causing the accumulation of ROS that may generate significant oxidative stress in both melanocytes and anagen hair bulb epithelium. In this context, melanogenic bulbar melanocytes are best suited to assume a postmitotic, terminally differentiated (pre)senescence status to prevent cell malignant transformation [1,36,427,430– 433]. The involvement of ROS in the onset of canities have been supported by several observations. Some melanosomes from grey hair bulbs were identified within auto-phagolysosomes, suggesting that they are defective and perhaps leaking reactive metabolites. Melanosomes in greying and white hair bulbs are highly vacuolated, a common cellular response to increased oxidative stress. Moreover, the common mitochondrial DNA deletion, also a marker of oxidative stress, occurs more prominently in grey hair follicles compared to pigmented ones [1,5,36,433]. A depletion of the hair follicle MelSC reservoir has additionally been associated with the onset of greying. In average, a scalp hair follicle experience 7-15 seedings of melanocytes over the grey-free lifespan. It is possible that canities may reflect an exhaustion of MelSC reservoir seeding potential; in fact, there is some experimental evidence suggesting that this potential is limited [1,36,434]. A defective, age-related, maintenance of MelSC, probably due to an imbalance between the antiapoptotic protein BCL2, MITF (its transcriptional regulator) and other factors, may also be accountable for the depletion of the MelSC reservoir [1,49,427,433,435–437]. Although many studies have shown the absence of MelSC from white Chapter I: A comprehensive review of human hair pigmentation 50 hair follicles, there are reports of hair greying reversibility. In this context, the greying seems to occur due to some defect in the MelSC activation or migration to the hair bulb at the beginning of a new hair cycle [5,36,427]. Figure I.9. Aged hair follicle, presenting grey/white hair .The age-related pigment loss in hair fibres may be caused may be caused by 1) depletion of melanotic melanocytes (Mel-Mc) in the hair matrix, 2) exhaustion of the melanocyte stem cell (MelSC) reservoir and 3) MelSC defective activation or migration during the onset of anagen phase of the hair growth cycle. I.8. MODIFICATION OF THE HAIR FIBRE COLOUR Throughout human history, people have changed the hair colour to segregate the social status. Nowadays, Modification of the human hair: modulators of melanogenesis as agents of colour changing 51 regardless of economical and education background, millions of individuals worldwide commonly dye the hair to enhance youth and beauty and to follow fashion trends. Because hair colouration became popular among the general population, hair colouring products now represent one of the most rapidly growing beauty and personal care markets [438,439]. Regarding the year 2019, the global hair colour market was valued at approximately 22.2 billion USD, and it is expected to generate revenue of around 37.4 billion USD by 2026 (Zion Market Research, NY, USA). Hair dyeing systems can be divided into oxidative and non-oxidative. Additionally, according to the colour durability, they are classified into temporary, semi-permanent, and permanent. Temporary non-oxidative dyes only deposit on the hair surface, leaving the fibre after the first washing. The semi-permanent nonoxidative dyes also interacts predominantly with the cuticle (the most external part of the hair strand), with a small penetration of dyes into the hair cortex occurring; this kind of colouration resists a few washes. Temporary and semi-permanent products do not require chemical reactions to impart colour as they rely on van der Waals forces for adhesion of direct dyes to the hair fibre. Regarding permanent oxidative products, hair colouration happens upon reaction between colourless precursors (developer and coupler) in the presence of an oxidizing agent (H2O2), and under alkaline conditions (ammonia). The combination of oxidizing and alkaline agents causes swelling of the hair cuticle, facilitating the diffusion of the small precursors into the fibre and the bleaching of the natural melanin pigments. Then, the colourless precursors undergo oxidation to form large, coloured molecules that become trapped inside the fibre cortex. The permanent oxidative hair colour products provide the greater efficacy of dyeing and are the most used, representing about 80% of the hair colouring market [438–440]. Due to the destructive nature of the permanent dye products, the hair fibre structure suffers cumulative damages that lead to split ends, dry and dull hair [441]. More important, those products have been identified as the source of various adverse health effects as chemical and allergic reactions that many times result in acute or mild dermatitis with consequent hair loss [439,442–445]. Even more alarming, it is the fact that some studies raised the possibility that long-term usage of permanent dyes can ensue serious and systemic side-effects as the increased risk of developing certain cancers [439,446,447]. Despite the inconvenience and liability, in the absence of other ways, many people continue to dye the hair for cosmetic purposes. Thus, development of safer ways for hair colour modification is more than ever a pertinent issue. Changes in hair colour have been reported as a side-effect of many drugs used in the treatment of several Chapter I: A comprehensive review of human hair pigmentation 52 diseases – Table I.2. Those changes are either as hair lightening or darkening, and even repigmentation of grey/white hair have been reported. Since such alterations are mostly transient, with the hair returning to its original colour after drug withdrawal, these findings raise the attractive possibility of using known, safe, and already approved drugs in hair colour modification as an alternative or in addition to the conventional methods. In fact, as showed here, the process of hair pigmentation offers countless druggable targets for modification of the colour phenotype. However, the cosmetic feasibility of a drugbased approach for modification of the hair colour has yet to be proved. Table I.2. Drugs reported to change the colour of hair as a side effect of medical treatments Drug Reason for use Hair change Ref. α-Interferon Melanoma Lightening [448] Acitretin Psoriasis Repigmentation [449,450] Atezolizumab Lung cancer Repigmentation [451] Chloroquine Malaria prophylaxis Lightening [452] Dermatomyositis [453] Cisplatin Metastatic germ cell neoplasm of the testis Darkening [454] Lightening Cyclosporin Psoriasis Darkening [455] Dabrafenib Metastatic melanoma Repigmentation [456] Dasatinib Chronic myeloid leukaemia Lightening [457] Defibrotide Deep venous thrombosis Darkening [458] Erlotinib Lung adenocarcinoma Repigmentation [459] Etretinate Psoriasis Lightening [460] Repigmentation [461] Pityriasis rubra pilaris [462] Hydroxychloroquine Discoid lupus erythematosus Lightening [463] Imatinib mesylate Gastrointestinal stromal tumour Lightening [464] Chronic myeloid leukaemia [465] L-Thyroxine Hypothyroidism Repigmentation [466] Modification of the human hair: modulators of melanogenesis as agents of colour changing 53 Table I.2. (Continued) Drug Reason for use Hair change Ref. Latanoprost Open-angle glaucoma Repigmentation [467] Lenalidomide Multiple myeloma Repigmentation [468] Levodopa Parkinson’s disease Repigmentation [469] Nivolumab Lung cancer Repigmentation [451] paraaminobenzoic acid Lymphoblastoma cutis Darkening [470] Dermatomyositis Dermatitis herpetiformis Scaly erythroderma Scleroderma Pazopanib Hürthle cell carcinoma Lightening [471] Myofibroblastic sarcoma [472] Pembrolizumab Lung cancer Repigmentation [451] Prednisone Bullous pemphigoid Repigmentation [473] Sunitinib malate Gastrointestinal stromal tumour Lightening [474] Tamoxifen Breast cancer Repigmentation [475] Thalidomide Multiple myeloma Repigmentation [476] Triptorelin Precocious puberty Greying [477] Valproic acid Seizures Lightening [478] Vemurafenib Metastatic melanoma Repigmentation [456] Verapamil Hypertension Repigmentation [479] I.9. CONCLUDING REMARKS Being one of the most distinguishing human features, hair colour has been the subject of extensive studies over the last decades. Here, an overview of the major findings is provided. The identification and comprehension of the events underlying the pigmentation of hair is an important starting point for the development of innovative hair cosmetics to change colour from inside out. CHAPTER II FLUORESCENT QUANTIFICATION OF MELANIN Modification of the human hair: modulators of melanogenesis as agents of colour changing 55 Chapter II Fluorescent quantification of melanin ABSTRACT Melanin quantification is reportedly performed by absorption spectroscopy, commonly at 405 nm. Here, we propose the implementation of fluorescence spectroscopy for melanin assessment. In a typical in vitro assay to assess melanin production in response to an external stimulus, absorption spectroscopy clearly overvalues melanin content. This method is also incapable of distinguish non-melanotic/amelanotic control cells from those that are actually capable of performing melanogenesis. Therefore, fluorescence spectroscopy is the best method for melanin quantification as it proved to be highly specific and accurate, detecting even small variations in the synthesis of melanin. This method can also be applied to the quantification of melanin in more complex biological matrices like zebrafish embryos and human hair.  This chapter is based on the following scientific paper: Bruno Fernandes, Teresa Matamá, Diana Guimarães, Andreia Gomes and Artur Cavaco-Paulo. Fluorescent quantification of melanin. Pigment Cell Melanoma Res. 2016, 29(6): 707-712. Chapter II: Fluorescent quantification of melanin 56 II.1 INTRODUCTION The in vitro measurement of melanin is the basis of many studies involving differentiation of pigmented malignant melanoma cells, the protective role of compounds against UV light and the development of treatments for pigmentation disorders such as vitiligo [480]. Also, in the cosmetic field, the development of new skin whitening agents created the necessity to accurately measure melanin in order to evaluate its significance [481]. There are several methods reported in the literature to assess the amount of melanin in biological samples. Electron spin resonance spectrometry (ESR) allows the measurement of electron spin resonance signals based on free radicals derived from melanin; this technique is highly specific, but it lacks sensitivity. Regarding photoacoustic spectroscopy, it has the advantage of being non-destructive and efficiently applicable to solid, gel and solution samples [481]. High performance liquid chromatography (HPLC) is also used in the quantification of melanin, particularly when it is necessary to distinguish between eumelanin and pheomelanin. This method is highly sensitive and accurate, allowing the quantification of different types of melanin present in samples, through the detection of their specific degradation products [482]. As these techniques are not easily performed in most laboratories (they require very specific equipment and expertise), absorption spectroscopy is the most extensively used method to quantify melanin [481]. Following solubilization of melanin pigments from cells or tissue samples in hot strong alkali (Soluene-350 or 1 M NaOH), total amount of melanin can be estimated spectrophotometrically by analysing absorbance and comparing the data obtained with a standard curve of synthetic melanin or melanin isolated from Sepia officinalis [483]. Although it is the most popular method for melanin quantification, absorption spectroscopy presents major drawbacks and does not always provide the sensitivity and specificity required [483]. Thus, the need for a reliable, simple, and economical method for melanin quantification is still a pertinent issue. Many years ago, Rosenthal et al. proposed the use of fluorescence spectroscopy to quantify melanin in cell cultures and tumours [480]. Melanins do not fluoresce but, once they are subjected to oxidative conditions (heating in alkaline hydrogen peroxide solution), they acquire fluorescence. Sachs was the first to report that pigments of various origins become fluorescent after oxidation with hydrogen peroxide. Recently, other authors confirmed that degradation of melanin is accompanied by the development of strong fluorescence [484]. The method originally proposed by Rosenthal et al. has the potential to overcome the disadvantages of using absorption spectroscopy; according to the authors, soluble products are formed and the fluorescent signal of oxidized melanin is not affected by proteinaceous or lipid Modification of the human hair: modulators of melanogenesis as agents of colour changing 63 Figure II.2. Effect of forskolin (Fsk) and kojic acid (KA) on intracellular melanin content of non-melanotic (BJ-5ta, NCTC2544), amelanotic (A-375, melanoma cells with no visible pigmentation), or melanized cells (SK-Mel-1 and SK-Mel-23, pigmented melanoma cells), after 72 h of treatment. Melanin content was calculated by interpolating the results with standard curves, generated by the fluorescence (after oxidation) (A) or absorbance (B) of Sepia melanin solutions of known concentration, prepared in 1 M NaOH containing 10% (v/v) of DMSO. The results were normalized by total protein levels in each sample. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. EtOH: ethanol. Chapter II: Fluorescent quantification of melanin 64 vehicle control. However, the pronounced depigmentation effect of KA (assessed by visual inspection) was only perceptible after oxidation of cell lysates and quantification of melanin by fluorescence. Concomitantly, the effect of the treatments on the melanin content of SK-Mel-1 was only detectable by fluorescence. Contrarily to SK-Mel-23 cells, visual inspection of these cells does not provide any insight regarding the effect of the compounds on melanogenesis. Nevertheless, the sensitivity of the method is so powerful that even small variations in the synthesis of melanogenesis can be detected and accurately quantified by fluorescence spectroscopy. Using absorption spectroscopy, these differences were neglected. On the one hand, according to the calibration curves generated (Supplementary Material, Figure II.S3), absorbance-based quantification is much less sensitive to small variations of melanin. On the other hand, the contribution of non-melanotic material to absorbance background signal is so pronounced that melanin quantifications are overvalued, and some variations can be masked. The contribution of non-melanotic material to melanin quantification is also clear by the analysis of the results for other cell lines. As mentioned before, NCTC2544 and A-375 cells have melanin contents similar to SK-Mel-1 despite of being unable to produce melanin. II.3.2 MELANIN QUANTIFICATION IN ZEBRAFISH EMBRYOS AND HUMAN HAIR In order to test the applicability of the proposed methodology on biological samples more complex than in vitro cultured cells, fluorescence spectroscopy was used to quantify melanin in zebrafish embryos and human hair – Figure II.3. The protocol presented in Figure II.1 (for cultured cells) was slightly adjusted in order to be applied to more complex samples, mainly in terms of NaOH concentration and incubation time during tissue disruption and melanin solubilization. Zebrafish Danio rerio is a powerful model system to study genetic mechanisms of vertebrate development, including the earliest events of pigment cells biology [490]. More recently, zebrafish was also established as an in vivo model to evaluate the activity of melanogenic regulatory compounds, making the quantification of melanin in those in vivo models of the highest importance [491,492]. Using fluorescence spectroscopy, a melanin production profile during zebrafish embryonic development was successfully drawn up to 120 hpf – Figure II.3A. For each measurement, only ten embryos were used (Supplementary. Material, Section II.5.1.6). According to the literature, much higher number of embryos (at least 100 embryos/mL of NaOH) are commonly used to perform melanin quantification by absorbance spectroscopy [491,493]. As expected, until 24 hpf, the embryos did not present quantifiable pigmentation (non-melanogenic control embryos). From 48 to 120 hpf, there is a steady increase in melanin Modification of the human hair: modulators of melanogenesis as agents of colour changing 65 production; the amount of melanin shows significant statistical difference in embryos collected at the three chosen time points. Also important to mention is that other pigments produced by zebrafish do not interfere in the quantification of melanin by this method (Supplementary Material, Figure II.S6). Figure II.3. Quantification of melanin in biological samples by fluorescence spectroscopy. (A) Developmental changes in the accumulation of melanin in zebrafish embryos. (B) Contents of total melanin in human hair samples of various colours (visual phenotypes). Melanin contents were calculated by interpolating the results with standard curves, generated by the fluorescence (after oxidation) of Sepia melanin solutions of known concentration, prepared in 1 M NaOH containing 10% (v/v) of DMSO. The results were normalized by number of embryos (A) or amount of hair (B) in each sample. Data were analysed by one-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. hpf: hours post fertilization. Chapter II: Fluorescent quantification of melanin 66 Fluorescence spectroscopy was also used to determine total melanin contents in hair samples of different colours – Figure II.3B. Using hair concentrations above the detection limit (Supplementary Material, Section II.5.1.6), the measured melanin contents are according to the expected and shows an excellent correlation with the visual phenotypes: black hair has the highest amount of melanin followed by darkbrown, reddish brown, medium blonde and light blonde hairs. The total melanin content in red hair (described as containing similar amounts of eumelanin and pheomelanin) is similar to that of medium blonde hair and comparable to what was reported by others, using HPLC [402]. HPLC is highly sensitive and, contrarily to absorbance and fluorescence spectroscopy, it also allows the analysis of melanins composition through the specific detection and quantification of the different degradation products of eumelanin and pheomelanin. According to the literature, H2O2 is efficient in the alkaline oxidation of both eumelanin and pheomelanin [402]. However, it is still unknown if their oxidation products exhibit similar spectroscopic properties, being properly accounted for the quantification of total melanin by the fluorimetric method here proposed. Although the results obtained offer some insight regarding the usefulness of fluorescence spectroscopy in the quantification of samples containing pheomelanin, more studies are needed to understand the contribution of each type of melanin to the fluorescent signal. Therefore, the fluorescence-based method should be used in essentially eumelanic systems. II.4 FINAL REMARKS The quantification of melanin in cell cultures by fluorescence spectroscopy was successfully validated. The protocol proposed is easy to perform and highly reproducible. Compared to the traditional widespread methodology based on absorbance, fluorescence measurement of melanin oxidation products is more sensitive and accurate in the quantification of melanin in biological samples. However, in the absence of non-melanogenic components, absorption spectroscopy is also expected to perform well but the methods employed in the extraction of melanin are exhaustive and time consuming. Melanin oxidation followed by the fluorescence measurement do not require the previous isolation of melanin from samples. This work supports fluorescence spectroscopy as the best choice for routine total melanin quantification in complex biological matrices. Modification of the human hair: modulators of melanogenesis as agents of colour changing 67 II.5 SUPPLEMENTARY MATERIAL II.5.1 OPTIMIZATION OF THE FLUORESCENCE-BASED PROTOCOL II.5.1.1. MELANIN OXIDATION The conditions for complete oxidation of melanins, amount of hydrogen peroxide solution used and time of reaction, were optimized using the two most frequent melanin standards: synthetic and Sepia melanins – Figure II.S1. Being a natural melanin, Sepia melanin resembles more closely the composition of melanin in biological tissues and mimics its fluorescent and optical properties. Synthetic melanin was also tested because, being cheaper than Sepia melanin, it is widely used as a standard for melanin quantification though its adequacy is very controversial. For both melanins, low fluorescent signals were detected in the samples before the oxidation with hydrogen peroxide. This autofluorescence is probably caused by some degradation/air oxidation of melanins (in an uncontrolled fashion way) due to the highly alkaline media and heat treatment used for its dissolution [494,495]. Then, the fluorescence intensity increases with increasing concentrations of hydrogen peroxide for up to 20% (v/v); synthetic melanin requires two hours to be completely oxidized while Sepia melanin requires four hours. After complete oxidation, fluorescence intensities of both solutions are stable for at least four hours (eight hours was the maximum reaction time tested by us). Unexpectedly, fluorescence intensity of synthetic melanin was found to be almost three times higher when compared to Sepia melanin. The differences in composition among these two pigments can be responsible for the differences found. Natural melanins (as Sepia melanin) are composed of a chromophoric fraction covalently linked to proteins that act as scaffolding matrix for melanin deposition or as enzymes catalysing various steps of melanogenesis [496]. By acid hydrolysis analysis, Sepia melanin from Sigma-Aldrich® was found to contain protein contents as high as 40%. On the other hand, the protein content of synthetic melanins is negligible [150,497]. Since only the indole residues of melanin are responsible for the fluorescence detected after oxidation, this can explain the results obtained [498]. Also, synthetic melanins, composed mainly of 5,6-dihydroxyindole units (DHI), are usually prepared by oxidizing tyrosine with hydrogen peroxide. In vivo melanogenesis occurs in the presence of tyrosinase and natural melanins are heteropolymers of DHI and 5,6-dihydroxyindole-2-carboxylic acid (DHICA) in a ratio of approximately 1:1 [494]. The differences in the structural composition of these melanins can affect the oxidation process and the type/amount of degradation products, which can also account for the results obtained. Chapter II: Fluorescent quantification of melanin 68 Figure II.S1. Time response study of the fluorescence of synthetic (A) and Sepia (B) melanin solutions, oxidized at 25 °C with different amounts of hydrogen peroxide (λexcitation= 470 nm; λemission = 550 nm). Melanin standard solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. After addition of hydrogen peroxide, the volume of samples was adjusted to normalize the melanin concentration at 100 µg/mL. The data was analysed by two-way ANOVA, followed by post hoc Tukey’s test. II.5.1.2. LIMITS OF DETECTION The limit of detection for a method is commonly defined as the lowest analyte concentration likely to be reliably distinguished from the blank sample (solution containing no analyte) [499]. We measured the absorbance and the fluorescence of solutions containing different amounts of synthetic or Sepia melanins and compared the values with the blank sample – Figure II.S2. The graphical representation is categorical due to a data visibility purpose only. Regarding synthetic melanin, Figure II.S2A, even the lowest concentration tested (0.098 µg/mL) provides a fluorescent signal significantly different from that obtained Modification of the human hair: modulators of melanogenesis as agents of colour changing 69 for the blank. On the other hand, measuring its absorbance signal, melanin can only be reliably detected at concentrations above 0.391 µg/mL. Using this melanin standard, fluorescence spectroscopy proved to be at least four times more sensitive than the traditional absorption spectroscopy. For Sepia melanin, Figure II.S2B, the sensitivity of both methods was set at 1.563 µg/mL. Figure II.S2. Fluorescence and absorbance values of solutions containing different concentrations of synthetic (A) and Sepia (B) melanins. Solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO, by two-fold serial dilutions. Absorbance of solutions was measured at 405 nm. Fluorescence intensity was measured (λexcitation= 470 nm; λemission = 550 nm) after incubation of melanin solutions with 20% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by one-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, compared to the respective blank sample (0 µg/mL). Chapter II: Fluorescent quantification of melanin 70 II.5.1.3. LINEARITY We studied the correlation between the quantity of pigment and signal intensity (absorbance or fluorescence) in the concentration range 0.391 – 100 µg/mL for synthetic melanin and 1.563 – 100 µg/mL for Sepia melanin – Figure II.S3. Figure II.S3. Plots of standard curves of synthetic (A) and Sepia (B) melanins obtained by fluorescence and absorption spectroscopy. Solutions were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO, by two-fold serial dilutions. Absorbance of solutions was measured at 405 nm. Fluorescence intensity was measured (λexcitation= 470 nm; λemission = 550 nm) after incubation of melanin solutions with 20% (v/v) hydrogen peroxide solution for 4 h. Modification of the human hair: modulators of melanogenesis as agents of colour changing 71 As shown in Figure II.S3, plots of the fluorescence and absorbance intensities against melanin concentration indicate good linearity in the range of 1 – 100 µg/mL for synthetic melanin (Figure II.S3A) and 5 – 100 µg/mL for Sepia melanin (Figure II.S3B). The correlation coefficients (R2) obtained by florescence spectroscopy are insignificantly lower than the ones obtained by absorption spectroscopy. This is caused by the apparent loss of linearity near the highest standard for the fluorimetric method that do not compromise the goodness-of-fit of linear regression for this method; run tests confirmed that deviations from linearity for both methods are not significant. Moreover, statistically, the slopes obtained by both methods are considered extremely different (p ≤ 0.0001). As fluorimetric method presents the higher values, this method is expected to be more sensitive to small changes in melanin concentration. II.5.1.4. OXIDATION OF CELL LYSATES Given the complexity of biological samples, we have repeated the optimization process to guarantee the complete oxidation of cell lysates. The best conditions were defined as 30% (v/v) hydrogen peroxide and 4 h of reaction (data not shown). Using these conditions, we have determined the detection limits for both methods in terms of cells number – Figure II.S4. The minimal concentration of pigmented melanocytes (SK-Mel-1 and SK-Mel-23) to produce fluorescent signals significantly different from the blank sample (no cells) and from the lysates of non-melanotic cells (BJ-5ta and NCTC2544) and amelanotic melanoma cells (A-375) is 310000 cells/mL – Figure II.S4A. Regarding to absorbance measurement, the limit of detection increases to 630000 cells/mL when SKMel-23 cells are used – Figure II.S4B. This decrease in sensitivity is cell-type dependent, therefore it should be verified for all the cell lines to be used. II.5.1.5. SELECTIVITY AND REPRODUCIBILITY Using cell concentrations above the limits of detection for melanin, we have studied the selectivity (degree to which a method can quantify the analyte accurately in the presence of interferences) and reproducibility (agreement between independent test results obtained under stipulated conditions) of both methods. To evaluate the interference of other cellular components present in the lysates, we added known amounts of synthetic and Sepia melanins (above the detection limit previously determined) to A-375 cell lysates and analysed the samples by both methodologies – Table II.S1 and Table II.S2. A-375 cells lysates were selected as these cells do not produce melanin, but the respective lysates can reflect the proteinaceous or lipidic composition of melanized melanocytes. Chapter II: Fluorescent quantification of melanin 72 Figure II.S4. Fluorescence (A) and absorbance (B) intensities of lysates containing different amounts of non-melanotic (BJ-5ta, NCTC2544), amelanotic (A-375, melanocytes with no visible pigmentation) or melanized (SK-Mel-1 and SK-Mel-23) cells grown in culture. Cell lysates were prepared in 1 M aqueous NaOH containing 10% (v/v) of DMSO. Fluorescence intensity was measured after incubation of cell lysates with 30% (v/v) hydrogen peroxide solution for 4 h. Data were analysed by two-way ANOVA, followed by post hoc Dunnett’s test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 or ****p ≤ 0.0001, when melanocytes concentrations were compared to the respective blank sample (0 cells/mL) or against the same concentration of non-melanotic and amelanotic cells. CHAPTER III DISCOVERY OF NEW MELANOGENESIS MODULATORS FOR HAIR COLOUR MODIFICATION: AN IN VITRO TO IN VIVO STUDY Chapter III: Discovery of new melanogenesis modulators for hair colour modification 80 Chapter III Discovery of new melanogenesis modulators for hair colour modification: an in vitro to in vivo study ABSTRACT The colour of hair is mostly defined by its content in melanins, a class of natural pigments synthesized by melanocytes in the follicle, through a highly regulated biochemical process known as melanogenesis. In theory, the darkening or lightening of hair could be achieved by the manipulation of follicular melanogenesis. The aim of the work described in this chapter was to find molecular candidates for follicular melanogenesis modulation via topical delivery. By screening a collection of 1200 generic drugs with a long history of human usage, twenty-three compounds showed to significantly alter the melanin content of in vitro cultured SK-Mel-23 cells. The top four inducers and inhibitors of melanogenesis were further validated regarding their melanogenic effect, and their cytotoxicity and dose-response activity was studied. Taking into consideration the data generated, and the already known toxicological profile in humans, three compounds were chosen for further in vivo testing on human volunteers: Compound A+ was chosen as an agent of hair darkening while Compound Band Compound Cwere chosen as agents of hair lightening. The results obtained in a clinical study with intervention of cosmetics support the idea of changing hair colour by topical treatment of the scalp with these new cosmeceutical ingredients. In the future, this approach is expected to serve as a safer and less fibre damaging alternative to the current cosmetic procedures employed in hair colouration.  This chapter is based on the following scientific paper: Bruno Fernandes, Cristiana Costa, Teresa Matamá, Andreia C. Gomes and Artur Cavaco-Paulo. To be submitted. Modification of the human hair: modulators of melanogenesis as agents of colour changing 81 III.1. INTRODUCTION Hair colour has always had an enormous social and cosmetic impact, as exemplified by the ancient trend of dying the hair to appear more attractive. Currently, the concern with hair colour (along with length, shape, and amount) is higher than ever as society focuses more and more on beauty and youthfulness [438,439]. Although hair dyes are used by millions of individuals worldwide, they have the potential to induce chemical and allergic reactions that can result in acute or chronic mild dermatitis with consequent hair loss [439,442–445]. Some studies have also raised the possibility that long-term usage of permanent hair dyes may be associated with an increased risk of developing certain cancers [439,446,447]. Also, the colouration procedures cause split, dry, and dull hair due to cumulative damages on fibre structure [441]. Therefore, the development of safer ways for hair colour modification is a very pertinent issue. The therapeutic usage of many drugs has shown different side effects on hair colour, texture, and shape. Drug-induced changes in hair colour are either as lightening or darkening (including repigmentation of grey/white hair) [448–479]; in most cases, the hair returned to its original colour after drug withdrawal. These findings raised the attractive possibility of using known, safe, and already approved drugs as an alternative or in addition to conventional cosmetic methods of colouration. A few patents have claimed the topical usage of some drugs with melanogenic properties to promote a change in the natural colour of hair fibres [505,506]. However, the lack of unequivocal in vivo cosmetic relevance for those compounds still makes the proof of a follicular-based approach for hair colour modulation a need in the art; this is the main goal of the work presented in this chapter. In this study, Prestwich Chemical Library® (an off-patent collection of small drugs, mostly approved by FDA, EMA and other agencies) was screened regarding the effect on in vitro melanin production. Then, the melanogenic activity of the top hit inducers and inhibitors was confirmed, and the changes is the activity of tyrosinase (rate-limiting enzyme of melanogenesis) was investigated as a preliminary study of the mechanisms of action involved in their melanogenic effects. Finally, a clinical study with intervention of cosmetics (RNEC No.: 92938) was conducted to validate the in vivo potential of selected drugs as cosmeceutical ingredients for the follicular modulation (darkening or lightening) of hair colour. III.2. RESULTS AND DISCUSSION The goal of this work was to identify novel agents for hair colour modulation in well characterized drugs, already in use for other applications. Drug repurposing is an increasingly employed strategy for Chapter III: Discovery of new melanogenesis modulators for hair colour modification 82 researchers in industry and academia: general pharmacology and toxicology are already established as they are the structure, chemical properties and biological functions of most of the repurposed drugs, sparing the cost and time needed to develop new products and facilitating further analysis of the novel functionality. For cosmetic applications, this approach became particularly relevant since animal testing was prohibited in the U.E.; today, one of the most reliable ways to assure that newly developed cosmetic products present in vivo low risk is to formulate them using compounds with already known toxicological profile. III.2.1. SCREENING OF MELANOGENESIS MODULATORS In this work, Prestwick Chemical Library® was screened to identify new modulators of melanogenesis among 1200 small molecules. Several aspects were considering for choosing this library. First, most of the compounds in the library have a long history of human usage, with their toxicological profiles being well-characterized. Second, the original patents of compounds have expired, allowing industrial protection of repurposing. Finally, some of the most represented targets of this library have direct implication in melanogenesis: adrenergic receptors, histamine receptors, oestrogen receptors, phosphodiesterases (PDEs), serotonin receptors (5-HTR), cholinesterases, among others. Notedly, high throughput screenings (HTS) of melanogenesis modulators are not widely explored, and they usually target tyrosinase activity instead of the direct evaluation of cellular melanin contents. This is supposed to be happen because, until the development of the fluorescence-based approach documented in this thesis (Chapter II), the existing methods for quantification of melanin were mostly impracticable at large scale assays compared to the ones employed in the assessment of tyrosinase activity [507]. The major problem with screening melanogenesis modulators based on their actions on tyrosinase is the reported existence of several compounds that change the pigmentation status of melanocytes without affecting the catalytic activity of the enzyme. In this sense, for more highly inclusive screenings, a direct quantification of melanin must be favoured. The screening of Prestwick Chemical Library®, regarding the ability to alter cellular levels of melanin was performed on SK-Mel-23 cells. Average melanin production obtained with each compound is shown in Figure III.1, as a percentage of vehicle control (cells treated with 1% DMSO). The method used for the identification of hits has involved selecting a standard deviation (σ) threshold relative to the whole assay mean (μ). The cut-off for hit melanogenesis inducers was set at μ + 3σ, a commonly used threshold in high throughput screenings corresponding to a false positive error rate of 0.00135. The cut-off for Modification of the human hair: modulators of melanogenesis as agents of colour changing 83 melanogenesis inhibitors was adjusted to μ - 2σ due to a very low incidence of hits when using a more conservative approach [508–510]. Figure III.1. High Throughput Screening (HTS) of Prestwick Chemical Library® regarding the in vitro modulation of melanin. SK-Mel-23 cells were treated with compounds for 72 h and then, melanin quantification was performed using a fluorescence-based method. The melanin contents were normalized by total protein levels in each sample and expressed as a percentage of the vehicle control, 1% (v/v) DMSO. HTS-Corrector software (version 2.0) was used to examine HTS data and hit selection. The cutoff for hit melanogenesis inducers and inhibitors was set at μ + 3σ and μ - 2σ, respectively. μ: whole mean assay. σ: standard deviation. About 2% of the tested compounds were identified as hit modulators of melanin synthesis, which is higher than the typical rate of < 0.1% reported for many screenings of large libraries of small molecules [511]. The rate obtained is presumably related to the fact that the most representative cell targets of the drugs included in this library have direct implications on melanogenesis. From the 23 targeted hits, 10 increase the cellular level of melanin (hit inducers) and 13 decrease the cellular level of melanin (hit inhibitors) – Supplementary Material, Table III.S1. To the best of our knowledge, apart from protriptyline hydrochloride and trimipramine maleate salt, the melanogenic effects presented by those drugs are reported here for the first time. The top four hit inducers and inhibitors were selected for further studies – Table III.1. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 84 Table III.1. Top hit compounds and their effect on intracellular melanin content of SK-Mel-23 Compound Melanin Content (% of Control ± SD) Screening Validation t-test Inducers Compound A+ 327.2 ± 11.48 198.5 ± 13.93 t(2) = 10.08 p = 0.0097 Compound B+ 243.5 ± 7.686 169.8 ± 15.49 t(2) = 6.031 p = 0.0264 Compound C+ 232.2 ± 33.22 218.8 ± 2.263 t(2) = 0.5696 p = 0.6264 Compound D+ 191.9 ± 4.158 202.1 ± 34.44 t(2) = 0.4134 p = 0.7194 Inhibitors Compound A22.22 ± 7.757 71.65 ± 2.899 t(2) = 8.442 p = 0.0137 Compound B33.56 ± 16.82 78.00 ± 4.243 t(2) = 3.623 p = 0.0685 Compound C35.98 ± 14.94 31.10 ± 6.930 t(2) = 0.4186 p = 0.7162 Compound D39.75 ± 13.15 64.60 ± 9.758 t(2) = 2.146 p = 0.0825 III.2.1.1. VALIDATION OF SELECTED HIT COMPOUNDS The validation of the melanogenic effect was first attempted using re-supplies of selected hits from sources other than Prestwick Chemical; the only exception was Compound C+, due its low commercial availability – Table III.1. The melanin contents of SK-Mel-23 cells treated with 10 μM of Compound C+, Compound D+ and Compound Cfor 72 h were not statistically different from those obtained in the primary screening assay (p > 0.05). Contrarily, significantly lower melanin contents were obtained, after incubation with melanogenesis inducers Compound A+ and Compound B+ (p ≤ 0.05). Lower than expected was also the inhibition of melanogenesis by Compound A-, Compound Band Compound D-; however, only melanins contents of cells treated with Compound Awere considered statistically different. Despite some variations in the extent of melanogenesis modulation, for all compounds, the type of modulatory effect in SK-Mel-23 cells was according to the expected. III.2.2. DOSE-RESPONSE ASSAY OF SELECTED HIT COMPOUNDS As a second step in the confirmatory screening, the dose-response of selected hit compounds was tested, Modification of the human hair: modulators of melanogenesis as agents of colour changing 85 and their performance compared with well-established agents of melanogenesis modulation. This analysis was performed in SK-Mel-23 and in SK-Mel-1 (a less pigmented melanoma cell line). A cell viability assay was included in this study to avoid envisaged outcomes due to possible cytotoxicity effects. III.2.2.1. DEFINING THE NON-CYTOTOXIC RANGE OF DRUG CONCENTRATIONS The extent of cytotoxicity from every condition tested was quantified as a percentage of cell viability, compared to vehicle control (DMSO 1%) – Figures III.2 to III.5. Treatment conditions where the mean viability was higher or not statistically lower than 80% (t test; p > 0.05) were considered as non-cytotoxic. Bellow that, the cytotoxicity was ranked as follow: within 80-60% as weak, 60-40% as moderate and below 40% as strong, respectively (ranges according to ISO 10993-5). The treatments with 1 to 10 μM of hit inducers (Compound A+, Compound B+, Compound C+ and Compound D+) were mostly well tolerated by SK-Mel-23 cells – Figure III.2 (also Supplementary Material, Table III.S2). The exceptions were those using 10 μM of Compound C+ for 48 and 72 hours which caused weak cytotoxicity. For the two highest concentrations tested (50 and 100 μM), the cytotoxicity was evident for all inducers, though Compound A+ and especially Compound B+ were less toxic than Compound C+ and Compound D+. Under the same range of concentrations, SK-Mel-1 cells were in general more sensitive to the tested inducers – Figure III.3 (also Supplementary Material, Table III.S2). Regarding Compound C+, besides the decrease in cell viability verified for the incubation with 10 μM for 48 h and 72 h, moderate cytotoxicity was already observed at 72 hours in cells treated with 5 μM of this compound (64.1 ± 1.5%). The treatment with 10 μM of Compound A+ for 72 h also caused a slight decrease in cell viability (60.8 ± 1.5%; weak to moderate cytotoxicity). Besides, Compound D+ was particularly toxic to SK-Mel-1 (5 μM, 48 h: 73.0 ± 2.3%; 5 μM, 72 h: 44.1 ± 4.2%; 10 μM, 48 h: 44.8 ± 2.3%; 10 μM, 72 h: 18.0 ± 1.9%). Among the four inducers tested, only Compound B+ was better tolerated by SK-Mel-1 cells, with mild cytotoxicity just being noticed at 72 h for the treatment with 50 (69.9 ± 3.3%) and 100 μM (73.5 ± 4.5%). The inherent differences of adherent and suspension melanoma cells can justify the higher overall sensitivity of SK-Mel-1 to low concentrations of hit inducers (Supplementary Material, Table III.S2). Apart from this deferent sensitivity, the toxicity of the drugs was very consistent between the two cell models; the inducers less toxic were Compound B+ and Compound A+, followed by Compound C+ and Compound D+. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 86 Figure III.2. SK-Mel-23 cell viabilities (MTT assay) determined at 24, 48 and 72 h of exposure to different concentration of hit inducers. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define cytotoxicity ranges: non-cytotoxicity > 80%; weak 80-60%; moderate 60-40%; strong < 40%. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. Modification of the human hair: modulators of melanogenesis as agents of colour changing 87 Figure III.3. SK-Mel-1 cell viabilities (MTT assay) determined at 24, 48 and 72 h of exposure to different concentration of hit inducers. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define cytotoxicity ranges: non-cytotoxicity > 80%; weak 80-60%; moderate 60-40%; strong < 40%. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 88 Figure III.4. SK-Mel-23 cell viabilities (MTT assay) determined at 24, 48 and 72 h of exposure to different concentration of hit inhibitors. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define cytotoxicity ranges: non-cytotoxicity > 80%; weak 80-60%; moderate 60-40%; strong < 40%. Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. Modification of the human hair: modulators of melanogenesis as agents of colour changing 95 Figure III.8. Effect of various concentrations of hit inhibitors on melanin production of SK-Mel-23, over time. KA: Kojic acid. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal production of melanin (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 96 counteract the anti-melanogenic effects of these compounds. Alternatively, it is also possible that, depending on the dose, those compounds act on different (conflicting) pathways of melanogenesis regulation; although this kind of cellular response towards in vitro melanin production is very uncommon, it has been reported before [514]. Consequently, 41.45%, 33.64% and 61.81% of all variance in melanin content was attributable to the treatment of cells with different concentrations of Compound A-, Compound Band Compound D-, respectively (Supplementary Material, Table III.S4). Nonetheless, the modulation of melanogenesis by Compound Aand Compound Bwas not affected only by the concentration of compounds used. In the case of Compound A-, the interaction between concentration and time of incubation was accountable for 31.54% of all variance in melanin content (Supplementary Material: Table III.S4, interaction as a source of variation). As seen in Figure III.8, for concentrations that cause a depigmenting effect (5 and 10 μM), melanin production over time was lower, the higher the concentration used. In the case of Compound B-, the time of incubation was accountable for 32.15% of all variance in melanin content (Supplementary Material, Table III.S4). For most concentrations, and especially those causing the more pronounced depigmenting effect (10 and 50 μM), a decrease in melanin production occurred from 24 to 48 hours followed by a recover of pigment at 72 hours of incubation with Compound B-. Compound Cwas the only hit inducer that did not show signs of a dose-dependent dual effect on the production of melanin – Figure III.8. It inhibited melanogenesis across the whole range of non-cytotoxic concentrations used (1-10 μM), but mostly over the time of incubation (66.27% of all variance on melanin content; Supplementary Material, Table III.S4). Considering the data presented, the in vitro conditions to obtain the lower melanin contents in SK-Mel-23 are as follow: Compound A-, 10 μM, 72 h (71.7 ± 2.9%); Compound B-, 10 μM, 48 h (74.3 ± 6.5%); Compound C-, 10 μM, 72 h (31.1 ± 6.9%); Compound D-, 10 μM, 72h: 64.6 ± 9.8%). Notably, those compounds had a depigmenting effect in SK-Mel-23 that it is better or at least comparable to KA; the treatments were considered comparable when no statistical significant differences (two-way ANOVA, p > 0.05) were found between melanin contents of cell treated with the hit inhibitors and KA. Kojic acid is a competitive inhibitor of tyrosinase activity during L-DOPA oxidation, and widely used as an inhibitor of in vitro and in vivo (skin) melanin production [515,516]. Compared to SK-Mel-23, the melanogenic effects of hit inhibitors in SK-Mel-1 were quite different, and a U-shaped dose-response relationship was not verified for any of the compounds. – Figure III.9. Despite the use of different concentrations of Compound Aor Compound Dhad a medium effect (melanin Modification of the human hair: modulators of melanogenesis as agents of colour changing 97 Figure III.9. Effect of various concentrations of hit inhibitors on melanin production of SK-Mel-1, over time. KA: Kojic acid. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal production of melanin (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 98 contents obtained with 100 μM were slight lower, compared to the other concentrations), their modulation of melanogenesis was mostly time-dependent: 71.76% and 77.57% of all variance is explained by the use of different times of incubation, respectively – Supplementary Material, Table III.S4. For Compound A-, after an increase at 24 h of incubation, melanin contents decreased over time with a depigmenting effect being noticed at 72 h (78.3 ± 5.4%). For Compound D-, a stimulation of melanogenesis also occurred within the first 24 h, but the depigmenting effect was verified earlier, at 48 h (76.5 ± 7.3%); at 72 h, the effect did not disappear, but becomes less evident with cells increasing melanin production. For both compounds, this behaviour over time was shown for every concentration tested. Regarding Compound B- (1-100 μM), concentration and time also had a significant effect, but contrarily to Compound Aand Compound D-, their interaction explains most of all variance in melanin content (44.22%) – Supplementary Material, Table III.S4. This is well representative of the fact that noteworthy modulation of melanogenesis only occurs with high concentrations (especially 100 μM), with the kind of modulatory activity being dependent on the time of incubation: stimulation of melanogenesis at 24 h and inhibition of melanin production at longer periods of incubation (minimal melanin content at 72 h: 74.0 ± 5.7%). Finally, in the case of Compound C- (1-10 μM), the melanin content of cells treated with 5 μM for 48 hours was 70.1 ± 12.8%. This was the only case of effective melanogenesis inhibition; the other treatments failed to promote a relevant decrease in melanin content (1 μM for 24, 48 and 72 h; 5 μM for 24 h) or induced moderate to strong cytotoxicity. Consequently, the data obtained with the two-way ANOVA is presented in Supplementary Material (Table III.S4), but it was not used to infer about the impact of Compound Cin the production of melanin in SK-Mel-1. Irrespective of the differences in melanogenesis inhibition amongst the two cell lines, in SK-Mel-1, the depigmenting effect of all compounds was still comparable (two-way ANOVA, p > 0.05) to the effect of KA (2 mM, 72 h: 75.8 ± 2.1%). III.2.2.3. EFFECT ON THE ACTIVITY OF TYROSINASE Tyrosinase is the rate-limiting enzyme in the canonical pathway of melanin biosynthesis, catalysing the two initial non-spontaneous steps: hydroxylation of L-tyrosine (L-Tyr) to L-DOPA, and the oxidation of the latter to dopaquinone [2,3,129,131]. The catalytic activity of tyrosinase can be directly regulated by allosteric modulators. Additionally, intracellular activity of tyrosinase can be regulated indirectly by interfering with its transcription, translation, processing, maturation, or degradation rate. Being closely Modification of the human hair: modulators of melanogenesis as agents of colour changing 99 related to melanin production, the study of tyrosinase activity is a good starting point to disclose the cellular targets of melanogenic modulators. In the first step of this study, the activity of mushroom tyrosinase was measured following incubation with hit compounds – Supplementary Material, Figure III.S1; mushroom tyrosinase is often used as a substitute model for the human homologue. None of the inducers or inhibitors under test influenced the rate of L-DOPA oxidation, suggesting that their melanogenic effect is not related to a direct interaction with the model enzyme. However, it is important to point out that, although mushroom tyrosinase has been used in many studies attempting to identify new melanogenesis modulators or their mechanism of action, there is emerging evidence that the modulation of its activity may require molecular motifs distinct from those of the human counterpart [517]. Thus, a direct interaction with human tyrosinase cannot be entirely ruled out as the mechanism leading to stimulation or inhibition of melanin production in SK-Mel23 and SK-Mel-1 by the hit compounds. In a second experiment, the intracellular tyrosinase activity was examined following the treatment of SKMel-23 and SK-Mel-1 cells with the hit compounds. As expected, under conditions of maximal in vitro stimulation of melanogenesis, the activity of tyrosinase from cells treated with Compound A+, Compound C+ and Compound D+ was found to be increased in a significant way – Figure III.10 (also Supplementary Material, Figure III.S2 and Figure III.S3). In contrast, mechanism other than the modulation of intracellular tyrosinase activity seems to govern the melanogenic effect of Compound B+. Despite some pro-melanogenic conditions are associated with increased activity (Supplementary Material, Figure III.S2 and Figure III.S3), maximal stimulation of melanogenesis by Compound B+ is attained without changes in the activity of enzyme or even under conditions that promotes its inhibition – Figure III.10. Although it is not common, this phenomenon had been reported before. Nakajima et al. showed that the treatment of in vitro cultured normal human melanocytes with arbutin cause an increase in melanin production along with the inhibition of intracellular tyrosinase activity; however, the possible mechanisms responsible for the stimulation of melanin production by arbutin were not further explored [518]. Since the mechanism of action involved in the therapeutic usage of Compound B+ is also unknown, without further studies, it is impossible at this moment to predict how Compound B+ may act on melanocytes to stimulate melanogenesis. Other mechanism than the modulation of intracellular tyrosinase activity also seems to rule the inhibition of melanin production in SK-Mel-23 by Compound A-, Compound Band Compound D-. For those compounds, none of the conditions that cause maximal inhibition of melanin production were Chapter III: Discovery of new melanogenesis modulators for hair colour modification 100 Figure III.10. Effect of hit inducers on melanin production and tyrosinase activity of SK-Mel-23 and SKMel-1, cultured under in vitro conditions for maximal melanogenesis stimulation. SK-Mel-23: 10 μM Compound A+, 72 h; 50 μM Compound B+, 24 h; 10 μM Compound C+, 24 h; 10 μM Compound D+, 24 h; 20 μM Fsk, 48 h. SK-Mel-1: 10 μM Compound A+, 48 h; 50 μM Compound B+, 24 h; 5 μM Compound C+, 48 h; 5 μM Compound D+, 24 h; 20 μM Fsk, 72 h. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal production of melanin or tyrosinase activity (100% of vehicle control). Data were analysed by one-sample t test. *p ≤ 0.05, when conditions were compared to the 1% DMSO control (100%). Fsk: forskolin. accompanied by significant changes in the intracellular tyrosinase activity – Figure III.11 (also Supplementary Material, Figure III.S4). In SK-Mel-1, the inhibition of melanogenesis by Compound Dwas also largely independent of the intracellular tyrosinase activity, and a statistically significant increase was Modification of the human hair: modulators of melanogenesis as agents of colour changing 101 Figure III.11. Effect of hit inhibitors on melanin production and tyrosinase activity of SK-Mel-23 and SKMel-1, cultured under in vitro conditions for maximal melanogenesis inhibition. SK-Mel-23:10 μM Compound A-, 72 h; 10 μM Compound B-, 48 h; 10 μM Compound C-, 72 h; 10 μM Compound D-, 72 h; 2 mM KA, 72 h. SK-Mel-1: 100 μM Compound A-, 72 h; 100 μM Compound B-, 72 h;5 μM Compound C-, 48 h; 100 μM Compound D-, 48 h; 2 mM KA, 72 h. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal production of melanin or tyrosinase activity (100% of vehicle control). Data were analysed by one-sample t test. *p ≤ 0.05, when conditions were compared to 1% DMSO control (100%). KA: kojic acid. even observed after the treatment of cells with 100 μM of the compound for 48 h – Figure III.11 (also Supplementary Material, Figure III.S5). Contrarily, the conditions for maximal inhibition of melanogenesis in SK-Mel-1 by Compound Aand Compound Binduced a significant decrease in the activity of the Chapter III: Discovery of new melanogenesis modulators for hair colour modification 102 enzyme activity, as expected – Figure III.11. Amongst the hit inhibitors, the use of Compound Cpresented the most curious variations in tyrosinase activity. Attending to Figure III.11, melanin content obtained following incubation of SK-Mel-23 with 10 μM of Compound Cfor 72 h (maximal inhibition of melanin production) do not seem to be related to a decrease in tyrosinase activity. Nonetheless, the incubation with low concentrations (1 or 5 μM) for the same period cause decreases in melanin contents that are escorted by inhibition of tyrosinase activity (Supplementary Material, Figure III.S4). Moreover, inhibition of melanogenesis is also attained under conditions that cause levels of intracellular tyrosinase activity (10 μM, 24 and 48 h) comparable to those presented by hit inducers (Supplementary Material, Figure III.S4). A considerable increase in tyrosinase activity was also found for the only case of significant melanogenesis inhibition by Compound Cin SK-Mel-1 cells – Figure III.11. In the literature, there are several cases of inconsistencies between melanin content and intracellular tyrosinase activity that could help to explain the results here obtained with the hit inhibitors. Inhibition of melanogenesis without relatable changes in the measurements of intracellular tyrosinase activity was attained with some tricyclic antidepressants [519] and quinolines [520]. The depigment effects verified were found to be associated with altered trafficking of tyrosinase and other related proteins. In melan-a cells treated with these compounds, although tyrosinase was normally expressed and maturated (assuring functionality), at some point, its normal trafficking was disrupted, and the enzyme ended trapped in organelles other than melanosomes. Homochlorcyclizine has likewise been reported to inhibit melanogenesis in B16 melanoma 4A5 cells without directly or indirectly inhibit tyrosinase activity [521]. Despite the particular drug mechanism had not been disclosed, the authors proposed that it could be related to: the blockage of L-phenylalanine transporters on the membrane of melanocytes, inhibition of phenylalanine hydroxylase (converts L-phenylalanine into L-Tyr) or inhibition of L-tyrosine hydroxylase isoform I (along with tyrosinase, also catalyses the conversion of L-Tyr to L-DOPA). The inhibition of solute carrier family 7 member 5 (SLC7A5) by JPH203 (now in clinical trials as an anti-cancer drug) or 2-amino2-norbornanecarboxylic acid (BCH) had also been shown to cause a decrease in melanin content of B16F10 cells without affecting the intracellular tyrosinase activity [139]. SLC7A5 is a member of the Ltype amino-acid transporter family, specialized in the transport of several aminoacids as L-Tyr; consequently, the levels of this tyrosinase substrate inside melanocytes might be reduced by inhibiting SLC7A5, causing a decrease in the production of melanin. Finally, the simplest reason for the discrepancies obtained in our study can be due to the measurement procedure itself. If tyrosinase is misplaced in the intact treated cells but not inactivated by the hit inhibitor, or if the enzyme is reversibly Modification of the human hair: modulators of melanogenesis as agents of colour changing 103 inactivated, depending on the inhibitor concentration, when cell disruption occurs and the exogenous substrate is supplied to the cell lysate, the enzyme activity measured will not be representative of the antimelanogenic effect of the compounds. Taxifolin and luteolin, which inhibit the in vitro production of melanin in B16F10 cells, were found to increase the protein levels of tyrosinase. When the lysates of cells, previously treated with those compounds, were used in the enzymatic assay, the rate of L-DOPA oxidation was found to be increased. On the other hand, when lysates of non-treated cells were mixed with the tested compounds, the formation of dopachrome was inhibited, suggesting that taxifolin and luteolin inhibit the production of melanin by directly inhibiting the enzyme, despite activating simultaneously some cellular pathways that cause an increase in the total enzyme content [522]. III.2.2.4. SELECTION OF COMPOUNDS FOR THE IN VIVO STUDY The data available in the literature gathered with the data obtained in the dose-response assay were considered for the selection of drugs to be further tested in an in vivo study. Amongst the top hit inducers, Compound C+ and Compound D+ were the most cytotoxic compounds and especially harmful to SK-Mel1 cells. Moreover, as a carcinogenic substance (category 1B), Compound C+ has a very limited applicability in cosmetic products, as stated in the Regulation (EC) No 1223/2009 of the European Parliament and of the Council of 30 November 2009 on cosmetic products. To avoid possible safety issues, none of them were selected for in vivo testing. Despite having been well tolerated and the most effective activator of melanogenesis, Compound B+ also promoted an almost complete loss of intracellular pigment content at the highest non-cytotoxic concentration tested. This phenomenon was quite unexpected, and it needs to be extensively explored before a cosmetic use in humans can be attempted. Consequently, Compound A+ was selected as the only agent of hair darkening for the in vivo study; this compound provided a sustained stimulation of in vitro melanogenesis across the whole range of non-cytotoxic concentrations. In a clinical context, Compound A+ exerts its therapeutic effects by inhibiting PDEs, the enzymes responsible for normal degradation of second messengers cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP). In melanocytes, cAMP and cGMP are known to increase the synthesis of melanin through a Protein Kinase A (PKA) [12,56,58,256,523,524] and Protein Kinase G (PKG) [525,526] dependent pathway, respectively; cGMP has been primarily associated with the UVBinduced pigmentation of skin. A cAMPand/or cGMP-dependent stimulation of melanogenesis correlates well with the increased intracellular tyrosinase activity verified upon the treatment of SK-Mel-23 and SK- Chapter III: Discovery of new melanogenesis modulators for hair colour modification 104 Mel-1 cells with Compound A+, supporting the inhibition of PDEs as the trigger event responsible for its melanogenic effect. Several PDEs inhibitors have already shown a pro-melanogenic effect: cilostazol [527], sildenafil [526], vardenafil [526], theophylline and 3-Isobutyl-1-methylxanthine (IBMX) [528]. Of note, Compound A+ has been shown to act as a free radical scavenger and antioxidant. In the context of stimulation of melanin production in hair, these properties are of the highest importance since the inability of melanocytes to deal with the large amounts of oxidative species generated during the synthesis of melanin have been associated with the age-related depigmentation of hair [427,432,433]. Regarding the hit inhibitors, Compound Aand Compound Dwere not included in the in vivo study due to safety concerns. Although none of the compounds have shown in vitro cytotoxicity in the melanotic cell lines used, Compound Ais a substance prohibited in cosmetic products, Regulation (EC) No 1223/2009 of the European Parliament and of the Council of 30 November 2009 on cosmetic products, and Compound Dhave the potential to cause acute dermal toxicity. Thus, only Compound Band Compound Cwere selected as agents of hair lightening. The pharmacological activity of Compound Bis mainly associated with the inhibition of acetylcholinesterase (AChE) and butyrylcholinesterase (BChE), the enzymes responsible for the hydrolysis of neurotransmitter acetylcholine (ACh). The inhibition of cholinesterases can also explain the antimelanogenic effect of Compound B-, at least in SK-Mel-1 cells where a decrease in intracellular tyrosinase activity was verified. By acting on the muscarinic receptors (M2R and M4R) of melanocytes, ACh is thought to inhibit AC and impair the synthesis of cAMP, causing a subsequent decrease in the expression of tyrosinase and synthesis of melanin. The negative modulation of pigment production by increased levels of ACh is reinforced by the observations of low activity of AChE in individuals with vitiligo, a disease characterized by the occurrence of depigmented spots (no melanin) in the skin [332,334,529–531]. Other known mechanisms of action of Compound Bcan also account for the inhibition of melanin production verified in SK-Mel-1. Compound Bis a pharmacological activator of the heat shock factor 1 (HSF1), a facilitator of transcription, production, and accumulation of heat shock proteins (HSP) as HSP70. HSP70 provides negative feedback on melanogenesis as demonstrated by the impossibility to stimulate melanin production in vitro (mice melanocyte cell cultures) and in vivo (transgenic mice) models of HSP70 overexpression [395]. The upregulation of HSP70 has already been proposed as the action mechanism of some anti-melanogenic compounds, causing a decrease in tyrosinase protein levels [532,533]. In studies regarding its anti-inflammatory properties, Compound Balso inhibited the gene/protein Modification of the human hair: modulators of melanogenesis as agents of colour changing 111 treatment according to age, gender, and basal hair pigmentation; different hair phenotypes seemed to respond differently to the two melanogenesis inhibitors under test, but there is no robustness in the data that can confirm those observations. Some cases of proven melanogenesis modulation could have been heightened or even translated into visual perceptible hair colour changes with a more prolonged study. The in vitro effect of Compound Band Compound Cwas time-dependent which can be indicative that a pronounced in vivo effect will only be achieved with long periods of exposure to these drugs. Additionally, a different dose regimen can also accomplish even more satisfactory results. The doses used in our study were far below the usual therapeutic range for those drugs and largely determined by the solubility in the formulation used. The frequency of application intended to improve the recruitment process and adherence of volunteers to the entire study. In a real-life situation, both could easily be increased for an improved effect of compounds; for instance, Compound A+ showed in vitro dose-dependent increase of melanin production so, the use of higher concentrations would certainly boost the darkening effect. Finally, the use of particulate vehicles (polymeric nanoparticles, liposomes, among others) can also increase the hair follicle targeting and improve the effect of these compounds. In testing those hypotheses, more exhaustive safety studies will have to be considered. III.3. CONCLUSION This work proves the feasibility of pharmacological modulation of hair colour. In a clinical study at the pilot scale, the hair of volunteers became darker or lighter following the treatment of the scalp with formulations containing different agents of hair colour modulation: Compound A+, Compound Bor Compound C-. These molecules are repurposed drugs, targeted as inducers or inhibitors of melanin biosynthesis in an in vitro screening of Prestwick Chemical Library®. This screening is a comprehensive survey of melanogenesis modulation by existing drugs and, besides the reported cosmetic implications, the data generated has the potential to reveal new therapeutic approaches to deal with diseases related with abnormal production of melanin. Based on tyrosinase activity assays and the known clinical function of these compounds, some hypothesis regarding the onset of their melanogenesis modulatory effect are presented, but wide-ranging follow up studies will need to be carried out to uncover their cellular targets in melanocytes and to confirm whether the clinical and cosmetic functions of these compounds share the same mechanisms. All treatments with hair colour modulation agents were well tolerated, and no adverse event was reported. Further larger trials will be performed in wider contexts, but data generated already supports the topical use of those drugs as a safe and effective approach to change the natural hair colour. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 112 III.4. SUPPLEMENTARY MATERIAL Table III.S1. Hit compounds and their effect on intracellular melanin content of SK-Mel-23 Inducers Inhibitors Compound Melanin Content (% of control) Compound Melanin Content (% of control) Compound A+ 327.16 Compound A22.22 Compound B+ 243.48 Compound B33.55 Compound C+ 232.21 Compound C35.98 Compound D+ 191.91 Compound D39.75 Compound E+ 178.19 Compound E49.27 Compound F+ 171.58 Compound F53.59 Compound G+ 170.45 Compound G54.79 Compound H+ 169.90 Compound H55.83 Compound I+ 167.09 Compound I56.79 Compound J+ 163.50 Compound J57.16 - - Compound K57.57 Modification of the human hair: modulators of melanogenesis as agents of colour changing 113 Table III.S2. IC 20 and 50 of hit melanogenesis modulators for 24, 48 and 72 hours of incubations with SK-Mel-23 and SK-Mel-1 cells SK-Mel-23 SK-Mel-1 24h 48h 72h 24h 48h 72h Compound A+ IC20 25.48 35.69 30.47 13.70 16.82 5.914 IC50 71.33 52.57 47.43 48.23 40.21 22.21 R2 0.909 0.957 0.959 0.980 0.984 0.921 Compound B+ IC20 36.00 33.04 20.12 96.63 >100 28.10 IC50 >100 96.28 66.76 >100 >100 >100 R2 0.860 0.966 0.907 0.840 - 0.724 Compound C+ IC20 9.767 7.188 6.029 11.17 5.993 2.809 IC50 16.32 13.88 12.23 24.76 11.75 7.033 R2 18.96 16.82 15.04 31.24 14.31 9.198 Compound D+ IC20 15.30 9.969 9.785 10.33 3.618 1.194 IC50 21.59 16.08 15.45 23.85 8.998 3.649 R2 0.989 0.982 0.983 0.982 0.989 0.986 Compound AIC20 >100 >100 >100 >100 >100 >100 IC50 >100 >100 >100 >100 >100 >100 R2 - - - - - - Compound BIC20 >100 >100 >100 >100 >100 >100 IC50 >100 >100 >100 >100 >100 >100 R2 - - - - - - Compound CIC20 10.74 11.14 9.610 8.493 5.782 3.192 IC50 16.59 16.28 15.67 15.33 8.532 5.071 R2 0.983 0.994 0.978 0.997 0.995 0.994 Compound DIC20 >100 >100 >100 >100 >100 >100 IC50 >100 >100 >100 >100 >100 >100 R2 - - - - - - Chapter III: Discovery of new melanogenesis modulators for hair colour modification 114 Table III.S3. Output of the two-way ANOVA analyses performed, regarding the effect of using different hit inducers concentrations and different times of incubations in the production of melanin by SK-Mel-23 and SK-Mel-1 cells SK-Mel-23 Compound Source of Variation F (DFn, DFd) P value % of variation Compound A+ Concentration F (2,18) = 255.9 < 0.0001**** 81.42 Time F (2,18) = 28.30 < 0.0001**** 9.005 Interaction F (4,18) = 10.54 0.0001*** 6.710 Compound B+ Concentration F (3,19) = 335.6 < 0.0001**** 93.64 Time F (2,19) = 8.019 0.0030** 1.492 Interaction F (6,19) = 5.564 0.0018** 3.105 Compound C+ Concentration F (2,16) = 404.4 < 0.0001**** 90.38 Time F (2,16) = 12.99 0.0004*** 2.903 Interaction F (4,16) = 11.02 0.0002*** 4.925 Compound D+ Concentration F (2,12) = 121.6 < 0.0001**** 73.93 Time F (2,12) = 24.97 < 0.0001**** 15.18 Interaction F (4,12) = 5.966 0.0070** 7.252 SK-Mel-1 Compound Source of Variation F (DFn, DFd) P value % of variation Compound A+ Concentration F (2,15) = 83.42 < 0.0001**** 80.60 Time F (2,15) = 0.7571 0.4861ns 0.7313 Interaction F (4,15) = 5.915 0.0046** 11.43 Compound B+ Concentration F (4,22) = 249.5 < 0.0001**** 87.56 Time F (2,22) = 29.27 < 0.0001**** 5.137 Interaction F (8,22) = 7.657 < 0.0001**** 5.374 Compound C+ Concentration F (2,15) = 34.48 < 0.0001**** 50.90 Time F (2,15) = 11.20 0.0011** 16.54 Interaction F (4,15) = 7.277 0.0018** 21.49 Compound D+ Concentration F (2,16) = 32.31 < 0.0001**** 32.93 Time F (2,16) = 54.19 < 0.0001**** 55.23 Interaction F (4,16) = 1.807 0.1768ns 3.684 Modification of the human hair: modulators of melanogenesis as agents of colour changing 115 Table III.S4. Output of the two-way ANOVA analyses performed, regarding the effect of using different hit inhibitors concentrations and different times of incubations in the production of melanin by SK-Mel-23 and SK-Mel-1 cells SK-Mel-23 Compound Source of Variation F (DFn, DFd) P value % of variation Compound AConcentration F (4,24) = 18.60 < 0.0001**** 41.45 Time F (2,24) = 12.23 0.0002*** 13.63 Interaction F (8,24) = 7.077 < 0.0001**** 31.54 Compound BConcentration F (4,28) = 14.06 < 0.0001**** 33.64 Time F (2,30) = 26.88 < 0.0001**** 32.15 Interaction F (8,28) = 3.652 0.0050** 17.46 Compound CConcentration F (2,14) = 21.57 < 0.0001**** 15.31 Time F (2,14) = 93.36 < 0.0001**** 66.27 Interaction F (4,14) = 9.480 0.0006*** 13.46 Compound DConcentration F (4,28) = 18.16 < 0.0001**** 61.81 Time F (2,28) = 0.2355 0.7918ns 0.4006 Interaction F (8,28) = 2.053 0.0763ns 13.97 SK-Mel-1 Compound Source of Variation F (DFn, DFd) P value % of variation Compound AConcentration F (4,24) = 5.507 0.0027** 8.591 Time F (2,24) = 92.01 < 0.0001**** 71.76 Interaction F (8,24) = 3.298 0.0110* 10.29 Compound BConcentration F (4,30) = 4.269 0.0075** 11.03 Time F (2,30) = 19.62 < 0.0001**** 25.36 Interaction F (8,30) = 8.556 < 0.0001**** 44.22 Compound CConcentration F (1,11) = 10.57 0.0077** 28.07 Time F (2,11) = 6.432 0.0141* 34.17 Interaction F (2,11) = 1.607 0.2442 ns 8.538 Compound DConcentration F (4,23) = 3.998 0.0132* 5.091 Time F (2,23) = 121.8 < 0.0001**** 77.57 Interaction F (8,23) = 3.934 0.0046** 10.02 Chapter III: Discovery of new melanogenesis modulators for hair colour modification 116 Figure III.S1. Effect of hit inducers and inhibitors on mushroom tyrosinase activity. Dotted lines define basal levels of enzyme activity (100% of control). Data were analysed by one-sample t test. *p ≤ 0.05, when conditions were compared to the hypothetical value 100%. PA: palmitic acid. KA: kojic acid. Modification of the human hair: modulators of melanogenesis as agents of colour changing 117 Figure III.S2. Effect of various concentrations of hit inducers on tyrosinase activity of SK-Mel-23, over time. Fsk: Forskolin. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal tyrosinase activity (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 118 Figure III.S3. Effect of various concentrations of hit inducers on tyrosinase activity of SK-Mel-1, over time. Fsk: Forskolin. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal tyrosinase activity (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable. Modification of the human hair: modulators of melanogenesis as agents of colour changing 119 Figure III.S4. Effect of various concentrations of hit inhibitors on tyrosinase activity of SK-Mel-23, over time. KA: Kojic acid. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal tyrosinase activity (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable. Chapter III: Discovery of new melanogenesis modulators for hair colour modification 120 Figure III.S5. Effect of various concentrations of hit inhibitors on tyrosinase activity of SK-Mel-1, over time. KA: Kojic acid. Values are presented as a percentage of vehicle control (1% DMSO). Dotted lines define basal tyrosinase activity (100% of vehicle control). Cytotoxic conditions are noted as . Data were analysed by two-way ANOVA, followed by post hoc Tukey’s test. Means that do not share a letter are significantly different. ND: not detectable.