Thi Thu Tram NGUYEN

ANNÉE 2014

THÈSE / UNIVERSITÉ DE RENNES 1 sous le sceau de l’Université Européenne de Bretagne

pour le grade de DOCTEUR DE L’UNIVERSITÉ DE RENNES 1 Mention : Chimie Ecole doctorale Sciences De La Matière Thi Thu Tram NGUYEN

Préparée dans l’unité de recherche UMR CNRS 6226 Equipe PNSCM (Produits Naturels Synthèses Chimie Médicinale) (Faculté de Pharmacie, Université de Rennes 1)

Screening of Thèse soutenue à Rennes le 19 décembre 2014 mycosporine-like devant le jury composé de : compounds in the Marie-Dominique GALIBERT Professeur à l’Université de Rennes 1 / Examinateur Dermatocarpon genus. Holger THÜS Conservateur au Natural History Museum Londres / Phytochemical study Rapporteur Erwan AR GALL of the lichen Maître de conférences à l’Université de Bretagne Occidentale / Rapporteur Dermatocarpon luridum Kim Phi Phung NGUYEN Professeur à l’Université des sciences naturelles (With.) J.R. Laundon. d’Hô-Chi-Minh-Ville Vietnam / Examinateur Marylène CHOLLET-KRUGLER Maître de conférences à l’Université de Rennes1 / Co-directeur de thèse Joël BOUSTIE Professeur à l’Université de Rennes 1 / Directeur de thèse

Remerciements

En premier lieu, je tiens à remercier Monsieur le Dr Holger Thüs et Monsieur le Dr Erwan Ar Gall d’avoir accepté d’être les rapporteurs de mon manuscrit, ainsi que Madame la Professeure Marie-Dominique Galibert d’avoir accepté de participer à ce jury de thèse.

J’exprime toute ma gratitude au Dr Marylène Chollet-Krugler pour avoir guidé mes pas dès les premiers jours et tout au long de ces trois années. Je la remercie particulièrement pour sa disponibilité et sa grande gentillesse, son écoute et sa patience. Sa parfaite maîtrise scientifique et ses idées qui m’ont si souvent sortie de l’embarras et parfois des impasses.

Je remercie très sincèrement le Professeur Joël Boustie, mon directeur de thèse, de m’avoir accueillie dans son laboratoire et de m’avoir permis d’accéder à l’univers passionnant de la chimie des lichens. Je lui suis reconnaissante de m’avoir laissée une grande liberté afin que je puisse développer mes propres idées au cours de cette thèse. Je le remercie vivement de sa disponibilité et de sa rigueur scientifique qui m'ont permis de travailler dans les meilleures conditions.

Je remercie chaleureusement Madame la Professeure Kim Phi Phung Nguyen, qui a été ma directrice de master recherche au Vietnam. C’est auprès d’elle que j’ai acquis mes premières connaissances en chimie des produits naturels et en chimie moléculaire. C’est également grâce à son aide que j’ai pu rencontrer le Professeur Joël Boustie.

Je tiens également à exprimer ma gratitude à l’ensemble de l’équipe qui m’a soutenue ces trois années, pour leurs conseils avisés, leurs remarques pertinentes, leur écoute, leur aide et leur gentillesse. Ils ont fait de ce travail le fruit d’une véritable réflexion collective. Je voudrais adresser plus particulièrement tous mes remerciements à Françoise pour m’avoir guidée dans la réalisation des tests biologiques afin de valoriser les molécules que j'avais isolées. Merci à Aurélie de m’avoir expliqué le fonctionnement des appareillages les premiers jours et de m’avoir procuré les produits chimiques dont j’avais besoin. Isabelle et Solenn, merci pour votre aide dans la réalisation des tests biologiques. Merci à Sophie et à Béatrice pour vos relectures. Merci à Audrey de m’avoir donné l’accès aux herbiers (H. des Abbayes et L. J-C. Massé). Anne-Cécile, Claudia, David, Maryse, Marie-Laurence, Patricia, Philippe, Pierre,

Michèle, Gilles, Nicolas, Jacques, Myriam, Mickaël, Jeff, Jean-Charles un grand merci à tous d’avoir été là pour moi.

J’adresse mes remerciements sincères à Monsieur Jean-Yves Monnat, Biologiste à l’Université de Bretagne Occidentale (Brest, France) pour m’avoir aidée à collecter et identifier les lichens. Je remercie également les autres membres de l’Association Française de Lichénologie pour m’avoir fourni des échantillons de Dermatocarpon.

Je remercie Isabelle Gallais et le Pr. Odile Sergent (UMR Inserm 1085, IRSET, Université de Rennes 1) pour la réalisation des tests antioxydants portant sur la peroxydation lipidique.

Je tiens également à exprimer ma reconnaissance à tout le personnel du CRMPO pour les analyses de masse effectuées sur les produits que j'avais isolées.

Merci à tous les étudiants et post-doctorants pour les discussions et les agréables moments: Pierre, Delphine, Nathalie, Friardi, Maïwenn, Sarah Komaty et Vianney.

Je remercie tous mes amis vietnamiens, surtout Binh, Duy, Linh, Tu, Hung, Huyen et Ha, qui ont partagé ma vie au quotidien pendant mon séjour en France. Un grand merci à Tam pour ton aide sur les analyses statistiques.

J’exprime toute ma gratitude envers le Gouvernement vietnamien pour l'attribution de la bourse qui m’a permis de mener cette thèse. Je remercie également le laboratoire du professeur Joël Boustie pour le complément de financement qui a permis la réalisation de ce travail en France.

Je remercie du fond du cœur ma famille pour leurs encouragements et leur soutien au quotidien. Merci à mes parents qui ont tant fait pour moi. Merci à mon frère aîné et à ma sœur pour notre complicité. J’ai une pensée toute particulière pour mon petit garçon Khanh, qui m’a fait sourire dans les moments difficiles; à ma belle-sœur Thuy qui m’a aidée dans les démarches administratives surtout quand c’était en français; à mon mari Thien pour ses encouragements, son dévouement auprès de notre fils pendant mon absence et son amour inconditionnel.

Enfin merci à tous ceux qui liront cette thèse pour l’intérêt qu’ils porteront à mon travail…

Table of Contents

Remerciements ...... i Table of Contents ...... iii List of abbreviations ...... vii List of Tables ...... ix List of Figures...... xi Résumé en français ...... xv Introduction ...... 1 CHAPTER 1: LICHENS AND PHOTOPROTECTION ...... 3 1 Presentation of lichens ...... 5 1.1 Definition ...... 5 1.2 Thallus morphology and anatomy ...... 5 1.3 Symbiosis ...... 8 1.4 Chemical compounds ...... 9 2 Photoprotective responses of lichens ...... 11 2.1 Harmful effects of UV radiation ...... 11 2.2 Photoprotective mechanisms ...... 12 2.2.1 Screening mechanism ...... 13 2.2.2 Quenching mechanism ...... 13 2.2.3 Repair mechanism ...... 15 2.3 Chemical features of UV absorbing compounds ...... 15 3 Mycosporines and MAAs – natural photoprotectants ...... 18 3.1 History ...... 18 3.2 Characteristics ...... 19 3.3 Biosynthesis ...... 22 3.4 Role of mycosporine-like compounds in nature ...... 23 3.5 Mycosporine-like compounds in lichens ...... 24 CHAPTER 2: SCREENING AND QUANTIFICATION OF MYCOSPORINE-LIKE COMPOUNDS IN THE DERMATOCARPON GENUS ...... 29 Purpose of this study ...... 31 1 Screening of mycosporine-like compounds ...... 33 1.1 Materials and methods ...... 33 1.1.1 Lichen material ...... 33 1.1.2 Extraction and purification of mycosporine-like compounds ...... 40 1.1.3 HPTLC-UV spectrophotodensitometry and HPLC-DAD-MS analysis ....42 1.1.4 Esterification conditions of mycosporine glutamicol ...... 43 1.2 Results and discussion ...... 44 1.2.1 Extraction and purification of mycosporine-like compounds ...... 44 1.2.2 HPTLC-UV spectrophotodensitometry ...... 45 1.2.3 HPLC-DAD-MS analysis ...... 46

iii

1.2.4 Esterification conditions of mycosporine glutamicol ...... 48 2 Quantification of mycosporines ...... 53 2.1 Materials and methods ...... 53 2.1.1 Lichen materials ...... 53 2.1.2 Extraction of mycosporines ...... 54 2.1.3 Quantification of mycosporines ...... 55 2.2 Results and discussion ...... 56 3 Experimental ...... 60 3.1 Identification of mycosporines and MAAs ...... 60 3.1.1 Extraction and purification of mycosporines and MAAs ...... 60 3.1.2 Analysis ...... 60 3.2 Quantification of mycosporines ...... 61 3.2.1 Extraction of mycosporines ...... 61 3.2.2 Mycosporines standards ...... 61 3.2.3 Internal standard ...... 61 3.2.4 Validation of analytical procedure ...... 62 3.2.5 HILIC-HPLC-DAD analysis ...... 62 CHAPTER 3: PHYTOCHEMICAL STUDY ON DERMATOCARPON LURIDUM (WITH.) J.R. LAUNDON ...... 63 Purpose of this study ...... 65 1 Description of Dermatocarpon luridum (With.) J.R. Laundon ...... 66 1.1 Ecology ...... 66 1.2 Botany and description ...... 67 1.3 Chemical studies ...... 68 2 Materials and methods ...... 70 2.1 Lichen material...... 70 2.2 Extraction ...... 70 2.3 Purification ...... 70 2.3.1 Mycosporines ...... 70 2.3.2 Other isolated compounds ...... 72 2.4 Determination of physico-chemical properties of mycosporines...... 73 2.4.1 The molar extinction coefficient ...... 73 2.4.2 pKa values ...... 73 3 Results and discussion ...... 74 3.1 Mycosporines ...... 76 3.1.1 Extraction and purification ...... 76 3.1.2 Structure elucidation ...... 78 3.1.3 Determination of physico-chemical properties ...... 83 3.2 Other isolated compounds ...... 87 3.2.1 Isolation ...... 87 3.2.2 Structure elucidation ...... 89 4 Experimental ...... 96

4.1 Extraction ...... 96 4.2 Purification and identification of mycosporines ...... 96 4.2.1 Purification on cation exchange resin...... 96 4.2.2 Purification by CPC ...... 96 4.2.3 A new purification protocol ...... 97 4.2.4 Sources, physico-chemical and spectroscopic data ...... 99 4.3 Other isolated compounds ...... 107 4.3.1 Purification and identification ...... 107 4.3.2 Structure elucidation ...... 109 4.4 Determination of the (ε) and pKa values of the isolated mycosporines ...... 122 4.4.1 The molar absorption coefficient...... 122 4.4.2 pKa values ...... 122 CHAPTER 4: BIOLOGICAL TESTS AND PHOTOPROTECTIVE EVALUATION 125 Purpose of this study ...... 127 1 Cytotoxic activity ...... 129 2 Photoprotective activity ...... 130 2.1 UV-filters ...... 131 2.2 Cytotoxic and phototoxic activities ...... 134 2.3 Antioxidant activity ...... 135 2.3.1 DPPH and NBT assays ...... 136 2.3.2 Lipid peroxidation assay...... 137 2.4 Photostability under UVA and UVB...... 140 3 Experimental ...... 142 3.1 Cytotoxic activity ...... 142 3.2 UV filters ...... 142 3.2.1 Emulsion preparation...... 142 3.2.2 Sample preparation ...... 143 3.2.3 Measurements and calculations ...... 143 3.3 Antioxidant activity ...... 145 3.3.1 DPPH assay ...... 145 3.3.2 NBT assay ...... 145 3.3.3 Lipid peroxidation assay...... 146 3.4 Cytotoxic and phototoxic activities ...... 148 3.5 Photostability under UVA and UVB...... 149 Discussion–Conclusion ...... 151 References ...... 157 General experimental procedures ...... 169 Appendix ...... 173

List of abbreviations

1D one dimensional 2D two dimensional Ac Acetone ACN Acetonitrile AcOH Acetic acid aq. aqueous AUC Area Under the Curve brs broad singlet BuOH n-Butanol calcd calculated CC Silica gel Column Chromatography

CHCl3 Chloroform

CH2Cl2 Dichloromethane COSY Correlation Spectroscopy CPC Centrifugal Partition Chromatography d doublet dd doublet of doublets

D2O Deuterium Oxide DEPT Distortionless Enhancement by Polarisation Transfer DL1 crude aq. extract of D. luridum DL2 semi-purified aq. extract of D. luridum DM1 crude aq. extract of D. miniatum DM2 semi-purified aq. extract of D. miniatum DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid dw. dry weight Eq. Equation EtOAc Ethyl Acetate EtOH Ethanol F.C Flash Chromatography h hours HILIC Hydrophilic Interaction Liquid Chromatography HMBC Heteronuclear Multiple Bond Correlation Spectroscopy HPLC High Performance Liquid Chromatography HPLC-DAD-MS High Performance Liquid Chromatography coupled to a Diode Array Detector and a Mass Spectrometer HPTLC-UV High Performance Thin Layer Chromatography coupled to a spectrophotodensitometer HRMS-ESI High Resolution Mass Spectrum Electrospray Ionization HSQC Heteronuclear Single Quantum Correlation spectroscopy IR Infrared Spectrophotometry IS Internal Standard m multiplet MAAs Mycosporine-like Amino Acids

vii

MDA Malondialdehyde Me Methyl MeOH Methanol min minutes m.p melting point MS Mass Spectrum NMR Nuclear Magnetic Resonance PE Petroleum Ether ppm parts per million (chemical shift value) Prep. HPLC Preparative High Performance Liquid Chromatography Prep. TLC Preparative Thin-Layer Chromatography q quartet ROS Reactive Oxygen Species rpm round per minute rt room temperature s singlet t triplet TLC Thin Layer Chromatography Tol Toluene UV ultraviolet

viii

List of Tables

Table 1: Mycosporine-like compounds distribution in lichens ...... 26 Table 2: Collected samples data for screen mycosporine-like compounds ...... 34 Table 3: Morphological comparison of the four Dermatocarpon species ...... 36 Table 4: Physico-chemical properties and mass fragmentation of 1, 2 and 3 ...... 51 Table 5: Linearity validation results of mycosporines 1 and 2 ...... 57 Table 6: List of metabolites isolated from D. luridum ...... 75 Table 7: pKa and molar extinction coefficient values of mycosporines 1 and 3 ...... 86 Table 8: Summary of the collected tubes corresponding to the mode used in CPC experiment ...... 97 Table 9: Preparation of buffer solutions with pH>2 ...... 122 Table 10: Preparation of buffer solutions with pH<2 ...... 123 Table 11: Measured absorbance at 310 nm depending on pH of two mycosporines 1 and 3 (n=6) ...... 124 Table 12: IC50 values for compound 7 on the eight cell lines and selectivity indexes (SI) relative to HaCaT and Fibroblast ...... 130 Table 13: Absolute, relative indexes and predicted results of the aq. extracts along with the three mycosporines comprared to three commercialized UV filters ...... 133 Table 14: Cytotoxicity and phototoxic activities of aq. extracts and mycosporines....135

List of Figures

Figure 1: Major growth forms of lichen ...... 5 Figure 2: Structure thallus homoeomerous (a) and heteromerous (b) ...... 7 Figure 3: Reproductive organs of lichens ...... 7 Figure 4: The nutritional exchange between partners of lichens ...... 8 Figure 5: Probable pathways leading to the major groups of lichen products ...... 10 Figure 6: Electromagnetic spectrum of UV radiation and biologic effects ...... 12 Figure 7: Defence strategies adopted by lichens to counteract the harmful effects of UV radiation ...... 12 Figure 8: Two major xanthophyll cycles involved in the dissipation of excess light energy ...... 14 Figure 9: UV absorption characteristics of common functional groups...... 16 Figure 10: Chemical structures of compounds from lichens reported as UV screens ....17 Figure 11: Organisms contain mycosporines and MAAs...... 19 Figure 12: Structure characteristics of oxo- and imino-mycosporines ...... 20 Figure 13: Common fungal mycosporines (a) and MAAs from marine organisms (b): structures, chemical and spectroscopic characteristics ...... 21 Figure 14: Postulated convergent pathways of MAAs biosynthesis in cyanobacteria ...22 Figure 15: The two different habitats of the four Dermatocarpon species investigated 33 Figure 16: Purification process of lichen crude aq. extract by cation exchange resin Dowex chromatography ...... 42 Figure 17: TLC profile of D. miniatum extracts before and after using NaCl in Dowex purification chromatography ...... 44 Figure 18: Comparative amounts (dry residue weight) of aq. extracts obtained from 100 mg lichen material ...... 45 Figure 19: The obtained UV absorption profiles of each extract screened by spectrophotodensitometry (C = 1 mg/mL) ...... 46 Figure 20: (a) and (b): PDA chromatogram of crude aq. extract and semi-purified aq. extract of D. miniatum; (c), (d), (e): UV spectra of mycosporines 1, 2, and 3; (f), (g), and (h): MS spectra of mycosporines 1, 2, and 3, respectively ..49 Figure 21: The proposed fragmentation patterns of mycosporines 1, 2 and 3 ...... 50 Figure 22: UV absorption profiles obtained in different conditions investigating esterification by spectrophotodensitometry (C = 1 mg/mL) ...... 52 Figure 23: View of lichens samples analysed from H. des Abbayes herbarium (a) and Natural History Museum (London, UK) (b) for mycosporines quantification ...... 54 Figure 24: Biphasic solvent distribution at silica surface in HILIC mode ...... 55 Figure 25: HILIC-HPLC chromatogram of D. miniatum (UV detector at 310 nm) and inset absorbance spectra of compounds 1 and 2 ...... 56 Figure 26: Calibration curves of mycosporine 1 (a) and mycosporine 2 (b) ...... 57

Figure 27: Information of Dermatocarpon species in H. des Abbayes herbarium before and after analysis ...... 59 Figure 28: Distribution of D. luridum in France ...... 66 Figure 29: D. luridum on submerged rock (a), macroscope observation of dried thallus (b), microscope observation of thallus transverse cut stained with cotton blue (c) ...... 67 Figure 30: Polyol and steroid constituents in the Dermatocarpon genus ...... 68 Figure 31: Free amino acids in the Dermatocarpon genus ...... 69 Figure 32: Carotenoids in the Dermatocarpon genus ...... 69 Figure 33: Kromaton Technologies FCPC 50mL apparatus ...... 72 Figure 34: Extraction procedure of Dermatocarpon luridum ...... 75 Figure 35: The TLC profile of semi-purified aq. extract of D. luridum using CPC in multiple dual mode ...... 77 Figure 36: A novel isolation protocol of mycosporines from aq. extract of D. luridum 80 Figure 37: IR spectra of mycosporines 1 (a), 2 (b) and 3 (c) ...... 81 1 Figure 38: H NMR spectrum of 3 (D2O, 300 MHz) ...... 82 13 Figure 39: C NMR spectrum of 3 (D2O, 75 MHz) ...... 82 Figure 40: Key 1H-1H COSY (bold line) and HMBC (1H→13C) correlations of 3 ...... 83 Figure 41: UV spectra of compounds 1, 2 and 3 at C= 2.5×10-5M in water ...... 83 Figure 42: UV spectra of mycosporine 3 at pH range from -1 to 7 (C= 1.56×10-5M) ...84 -5 Figure 43: UV spectra of mycosporine 1 at pH range from -1 to 3.84 (C0= 6.82×10 M) ...... 85 -5 Figure 44: UV spectra of mycosporine 3 at pH range from -1 to 3.62 (C0= 1.56×10 M) ...... 85 Figure 45: Resonance delocalization of the base form B ...... 86 Figure 46: Distribution diagramm of mycosporines 1 or 3 ...... 87 Figure 47: Isolation procedure of ten compounds from D. luridum...... 88 Figure 48: Three possible structures (a): α-L-glutamylglycine, (b): γ-L-glutamylglycine and (c) of dipeptide 6 ...... 91 Figure 49: The similar fragmentation patterns observed of two structures (a) and (b) ..91 Figure 50: Key 1H–1H COSY (bold line) and HMBC (1H→13C) correlations of compound 7 ...... 92 Figure 51: ESI-MS fragmentation of 8 ...... 94 Figure 52: Galactosylceramides isolated in lichen Ramalina celastri (Gal: galactose) .95 1 Figure 53: H NMR spectrum of 2 (D2O, 500 MHz) ...... 103 13 Figure 54: C NMR spectrum of 2 (D2O, 125 MHz) ...... 103 Figure 55: 1H-1H COSY NMR spectrum of 3 ...... 106 Figure 56: HSQC NMR spectrum of 3 ...... 106 Figure 57: HMBC NMR spectrum of 3...... 107 1 Figure 58: H NMR spectrum of 4 (D2O, 300 MHz) ...... 110 13 Figure 59: C NMR spectrum of 4 (D2O, 75 MHz) ...... 110 1 Figure 60: H NMR spectrum of 6 (D2O, 300 MHz) ...... 113 13 Figure 61: C NMR spectrum of 6 (D2O, 75 MHz) ...... 113

xii

1 Figure 62: H NMR spectrum of 7 (CDCl3, 300 MHz) ...... 116 13 Figure 63: C NMR spectrum of 7 (CDCl3, 75 MHz)...... 116 1 Figure 64: H NMR spectrum of 8 (pyridine-d5, 500 MHz) ...... 119 13 Figure 65: C NMR spectrum of 8 (pyridine-d5, 125 MHz) ...... 119 Figure 66: HSQC NMR spectrum of 8 ...... 120 Figure 67: Cytotoxicity of compounds 2, 3, 4, 6, 7 and 8 against eight cell lines at C = 25µM ...... 129 Figure 68: Cross section of the skin and penetration of sunlight radiation into the tissues ...... 131 Figure 69: Principal component analysis (PCA) of the tested compounds and extracts discriminating UV absorbing properties ...... 133 Figure 70: Mycosporine contents in crude aq. extract and semi-purified aq. extract of D. luridum and D. miniatum ...... 134 Figure 71: Antioxidant activity of the aq. extracts and three mycosporines 1, 2, 3 through DPPH and NBT methods compared to the positive control quercetin ...... 137 Figure 72: Proposed model for the role of early ROS-induced fluidizing effect of ethanol in the amplification of oxidative stress ...... 138 Figure 73: Effect of DM2 extract (5.20 µg/mL) and mycosporine 3 (4.00 µg/mL ~ 12 µM) on WIF-B9 cell line by ethanol-induced lipid peroxidation assay ....139 Figure 74: HPLC chromatograms of trolox (left) and mycosporine 1 (right) ...... 141 Figure 75: Solution preparation for UV-filters test ...... 143 Figure 76: Reaction leads to the formation of superoxide anion ...... 145

xiii

xiv

Résumé en français

INTRODUCTION

Notre étude s’inscrit dans l’une des thématiques de l’équipe PNSCM, consacrée aux substances naturelles lichéniques et à leurs applications dans le domaine de la photoprotection. Pour en comprendre le lien, il faut s’intéresser d’un peu plus près aux lichens. Les lichens sont des organismes symbiotiques constitués de l’association entre une algue verte et/ou une cyanobactérie (photosymbiote), un champignon (mycosymbiote) et des bactéries associées [1]. Ils sont apparus il y a environ 680 millions d’années, primitivement aquatiques ils ont réussi à s’adapter au milieu terrestre. Notre attention s’est particulièrement portée sur le fait qu’ils résistent bien à des conditions écologiques extrêmes, tel un ensoleillement intense, des températures extrêmes, une dessiccation importante... En effet cette association leur confère une structure unique : le thalle lichénique. Cette symbiose leur confère aussi une reproduction différente comparée à celle de chaque organisme seul, un métabolisme spécifique avec la production de métabolites secondaires généralement originaux. Ils sont aussi capables de stopper leur métabolisme lorsqu’ils sont à l’état sec pour reprendre leur activité lorsque les conditions hydriques redeviennent favorables. En réponse aux radiations UV, ils ont notamment développé au cours du temps des systèmes de protection efficaces tels que la production de composés aromatiques polyfonctionalisés [2]. Récemment, une autre classe de photoprotecteurs, connus dans d’autres branches du monde vivant, a été décelée au sein des lichens. Il s’agit des mycosporines et des MAAs (Mycosporine-like Amino Acids). Le chapitre 1 sera donc consacré à la description des lichens, aux mécanismes de photoprotection qu’ils ont mis en place et à la présentation des mycosporines et MAAs.

Récemment au laboratoire, un protocole de criblage de ces molécules particulières a été mis au point dans les lichens [3] et a permis de mettre en évidence la présence de ces composés uniquement dans des cyanolichens, organismes issus de la symbiose entre un champignon et une cyanobactérie. Alors que les chlorolichens (cyanobactérie remplacée par une algue verte) correspondent à 90% des espèces de lichens, aucun composé de type mycosporine n’y a été jusqu’ici détecté. La poursuite de ce criblage nous a donc semblé importante pour plusieurs raisons : recherche de nouvelles mycosporines, confirmation de leur intérêt en tant que marqueurs chimiotaxonomiques de différents genres et élucidation de leur biogenèse au sein

de l’association lichénique. Une étude préliminaire sur deux chorolichens et deux cyanolichens a permis de repérer la présence de composés de type mycosporine dans les deux cyanolichens mais aussi dans un des deux chlorolichens. Ce repérage inédit de mycosporines dans un chlorolichen : Dermatocarpon luridum nous a amené à focaliser notre étude sur ce genre et tout particulièrement cette espèce qu’il était possible de collecter en quantité suffisante en Bretagne pour une étude phytochimique. Il s’agit d’un lichen foliacé saxicole hydrophile appartenant au genre Dermatocarpon (Verrucariaceae), vivant sur les rochers retrouvés dans des milieux fréquemment inondés comme les cours d’eau. Nous avons aussi décidé de poursuivre ce criblage sur les trois autres Dermatocarpon aquatiques/semi-aquatiques présents en Bretagne : D. meiophyllizum, D. leptophyllodes et D. miniatum. Si D. miniatum a un habitat un peu différent [4] elles contiennent toutes l’algue verte Diplosphaera chodatii (Trebouxiophyceae) comme photobionte.

Dans le chapitre 2, nous présentons le résultat du criblage. Celui-ci doit nous permettre de confirmer ou non la présence de composés types mycosporines dans les trois autres espèces, de confirmer aussi leur intérêt en tant que marqueurs chimiotaxonomiques et d’avancer d’éventuelles hypothèses sur leur biogénèse. Nous espérons pouvoir aussi faire un lien entre l’écologie du lichen, la date de la collecte et leur contenu en mycosporines. Pour cela nous avons cherché à mettre au point une méthode de quantification des mycosporines à partir d’une faible biomasse. Nous avons eu l’opportunité de collaborer avec le Dr. Holger Thüs, conservateur de l’herbier lichen au Natural History Museum de Londres, spécialiste des Verrucariaceae qui nous a fourni 49 échantillons mis en herbier depuis plus de 170 ans. Cette étude a été complétée avec 30 échantillons provenant de l’herbier de Rennes (herbiers des Pr. Henry des Abbayes et Dr. Louis, Jean-Claude Massé) et 15 échantillons récoltés depuis moins de 10 ans. Nous avons aussi visé à démontrer qu’un herbier peut être une source importante d’informations pour nos études, au-delà de ses fonctions habituelles.

Suite à ce criblage, Dermatocarpon luridum (With.) J.R. Laundon 1984, en quantité suffisante dans le Finistère et espèce non menacée [69], est choisi afin de l’étudier plus en détail d’un point de vue phytochimique (chapitre 3). Nous espérons ainsi isoler les mycosporines détectées au cours du criblage, préciser leur structure et déterminer de nouvelles propriétés spectroscopiques et physicochimiques. L’étude phytochimique doit aussi nous permettre d’identifier d’autres métabolites originaux, étant donné qu’aucun composé polyphénolique habituellement retrouvé dans les lichens n’a été à ce jour décrit dans le genre Dermatocarpon.

xvi

Dans le cadre d’une évaluation dans le domaine de la photoprotection, le chapitre 4 est consacré aux activités biologiques des composés isolés en quantité suffisante ainsi que d’extraits aqueux de deux espèces de Dermatocarpon. Un test préliminaire consistera à évaluer leur activité cytotoxique sur 8 lignées cellulaires sur la plateforme ImPACcell à l’Université de Rennes 1. Outre leur intérêt possible dans le traitement du cancer pour les composés hautement cytotoxiques, le profil cytotoxique sur un panel de lignées cellulaires est une première étape importante pour sélectionner les composés ainsi que la dose appropriée pour d'autres tests. En ce qui concerne l’évaluation des capacités photoprotectrices, un premier test consistera à évaluer leur capacité en tant que filtre UV. Ensuite leur phototoxicité sur cellules non tumorales de kératinocytes humains, soumises à des radiations UVA, sera étudiée. L’oxydation au sein des systèmes biologiques fait appel à différents processus. Trois tests in vitro et in cellulo, mettant en jeu des mécanismes différents, seront donc réalisés pour déterminer l’activité antioxydante des composés et extraits. Enfin, dans le but de confirmer leur potentielle utilisation en tant photoprotecteur d’origine naturelle, leur photostabilité sous irradiations UVA et UVB sera évaluée.

Indépendamment de notre contribution au profilage chimique du genre Dermatocapon, avec l'accent mis sur la présence de mycosporines, nous espérons que ce travail apportera de nouvelles perspectives dans le métabolisme des lichens et qu’il soulignera l'intérêt général des lichens comme une source prometteuse de substances biologiquement actives.

CHAPITRE 1: LICHENS ET PHOTOPROTECTION

Afin de mieux présenter les espèces sur lesquelles nous avons travaillé et le cadre de nos travaux, ce chapitre donne quelques précisions sur les lichens et leurs métabolites en mettant l’accent sur les propriétés photoprotectrices de quelques métabolites. Dans la symbiose lichénique, le mycosymbiote est le partenaire englobant le photosymbiote dans une structure originale : le thalle lichénique. Il présente une grande diversité de formes et de couleurs et peut aussi porter un certains nombres d’organes divers. On distingue 3 grands groupes de thalles pour les macrolichens en fonction de leur forme : crustacés, foliacés et fruticuleux [5]. Le mycosymbiote appartient au groupe des ascomycètes (98%) et plus rarement au groupe des basidiomycètes (2%). Quant aux photosymbiotes, les plus fréquents sont les algues vertes des

xvii

genres Trebouxia et Trentepohlia, et la cyanobactérie du genre Nostoc. On décrit environ 18500 espèces de lichens dans le monde colonisant tous les milieux terrestres [6]. Certains lichens appelés tripartites (moins de 5%) incorporent les deux photosymbiotes : algue verte et cyanobactérie. Au sein de l’association, le champignon assure la protection physique à l’ensemble et fournit au photosymbiote l’eau, les sels minéraux et la vitamine C. Le photosymbiote apporte de son côté la matière organique carbonée sous forme de glucose ou polyols et fait du lichen un organisme autotrophe (Figure 4). Les métabolites lichéniques secondaires résultent de différentes voies de biogénèse, la majorité d’entre eux dérivent de la voie des acétates polymalonates, les voies de biogénèse de l’acide shikimique et de l’acide mévalonique sont également présentes (Figure 5). Le nombre de substances lichéniques identifiées à ce jour dépasserait les 1050 [14].

En réponse aux radiations UV, les lichens ont développé des systèmes de protection efficaces mettant en œuvre des mécanismes physiologiques et biochimiques (Figure 7). Parmi ceux-ci on peut citer la production et l’accumulation de composés aromatiques polyfonctionnalisés dans la couche corticale du lichen [2]. Ces métabolites appartiennent à la famille des depsidones, depsides, diphenyl éther, anthraquinones, xanthones, acides pulviniques et scytonémine. Leurs structures ainsi que leurs caractéristiques spectrales sont résumées dans la Figure 10.

D’autres métabolites présentant un intérêt dans la photoprotection, ont été récemment décelés au sein des lichens. Il s’agit des mycosporines et des MAAs (Mycosporine-like Amino Acids). Ils sont formés par un noyau cyclohexènone (oxo-mycosporine) ou cyclohexènimine (imino-mycosporine) (Figure 12). Leur voie de biosynthèse dans les lichens mérite d’être précisée. Chez les champignons et les cyanobactéries, ils sont dérivés de la voie de l’acide shikimique, cependant une nouvelle voie de biosynthèse à partir de la voie des pentoses phosphates a été décrite chez des cyanobactéries [54] (Figure 14). Ils ont des coefficients d’extinction molaires compris entre ε=28 000 M-1.cm-1 et ε=50 000 M-1.cm-1 (310-365 nm), sont très polaires, hydrosolubles et de faible poids moléculaire (<400 g/mol). Outre leur principale propriété d’être de bons filtres UV, de nombreuses études ont montré leur implication dans le stress osmotique, la reproduction fongique, la dessiccation et enfin leur rôle en tant qu’antioxydants. Enfin, un bilan de la distribution des composés de type mycosporine dans 40 lichens montre que seuls les cyanolichens et quelques lichens tripartites produisent des mycosporines (Tableau 1). La poursuite du criblage sur deux chorolichens et deux cyanolichens

xviii

nous a permis de mettre en évidence pour la première fois des composés types mycosporines dans un chlorolichen : Dermatocarpon luridum.

CHAPITRE 2: CRIBLAGE ET QUANTIFICATION DE COMPOSES DE TYPE MYCOSPORINE DANS LE GENRE DERMATOCARPON

Afin de confirmer ce résultat inédit, nous avons choisi de cribler les quatre espèces appartenant au genre Dermatocarpon présentes en Bretagne par une collecte dans le Finistère : D. luridum, D. meiophyllizum, D. leptophyllodes et D. miniatum. Après une description botanique des quatre espèces, un protocole d’extraction aqueuse, de purification et d’analyse par HPTLC-UV et HPLC-DAD-MS est appliqué. La comparaison des profils d’absorption en spectrophotométrie UV et en spectrométrie de masse des extraits (Figures 19 et 20) avec les données bibliographiques (Appendice 2), met en évidence pour la première fois dans les quatre chlorolichens la présence de deux composés de type oxo-mycosporine. Il s’agit de la mycosporine glutaminol 1 et de la mycosporine glutamicol 2 (Tableau 4). Ces deux mycosporines sont présentes dans l’extrait brut mais on peut noter qu’au cours de l’étape de purification de l’extrait, la mycosporine glutaminol 1 est instable et s’hydrolyse en mycosporine 2. Ces deux substances (y compris leur forme glycosylée) ont été observées pour la première fois dans des ascomycètes terrestres, puis dans une cyanobactérie terrestre pour la mycosporine 1 et un cyanolichen pour la mycosporine 2. Un troisième composé, la mycosporine 3 a été détecté dans deux extraits D luridum et D. miniatum, il correspond à l’ester éthylique de la mycosporine glutamicol. Nous avons démontré qu’il n’était pas d’origine naturelle et qu’il se formait au cours de l’étape de purification. Néanmoins il s’agit d’un composé nouveau.

La présence systématique des mycosporines 1 et 2 dans les quatre espèces de Dermatocarpon apporte un argument supplémentaire quant à leur rôle en tant que marqueurs chimiotaxonomiques de genre. Cette découverte apporte aussi des arguments en faveur d’une biogenèse où le mycosymbiote serait impliqué dans la biosynthèse des mycosporines au sein des lichens.

Afin de confirmer ces résultats, nous avons eu l’opportunité d’étendre le criblage sur plusieurs échantillons d’herbier en provenance d’Amérique du Nord et d’Europe et appartenant

xix

à l’ordre des Verrucariales dont 7 taxa de Dermatocarpon (89 échantillons), 3 taxa de Placidium (4 échantillons) et un taxum de Catapyrenium, soit un total de 94 échantillons (Appendice 3). Notre travail a consisté dans un premier temps, à partir d’une courbe de calibration, à valider une méthode de dosage rapide et sensible étant donné la faible biomasse mise à disposition (entre 15 et 50 mg de lichen) (Tableau 5). Ainsi, une méthode comprenant une extraction aqueuse suivie d’un dosage par HPLC-DAD basée sur une séparation en mode HILIC est appliquée pour la première fois, elle a permis de séparer et de déterminer le contenu en mycosporines 1 et 2 avec une séquence d’analyse de 16 min (Figure 25). Sur le panel d’échantillons, cette étude a montré que le contenu global en mycosporine variait entre 4.56 et 46.77 mg/g d’extrait sec correspondant à 0.1 et 1.4 mg/g de lichen. L’étude a montré une instabilité de la mycosporine glutaminol 1 dont la teneur devient nulle dans les échantillons collectés depuis plus de 15 ans alors que la teneur en mycosporine glutamicol 2 est mesurable et importante dans la plupart des échantillons Il est difficile de faire une cinétique de dégradation au cours du temps car nous ne connaissons pas le contenu initial en mycosporine mais notre travail apporte un point de départ à une analyse de ces échantillons sur le long terme. Concernant les quatre lichens frais collectés le même jour à des endroits différents (D. luridum, D. meiophyllizum, D. leptophyllodes et D. miniatum), nous remarquons une différence notable en contenu en mycosporine totale ; il est beaucoup plus élevé dans le D. miniatum. Ceci peut être corrélé à leur écologie différente : exposition au soleil sur rochers secs pour D. miniatum et une exposition à l’ombre sur rochers immergés pour les 3 autres. Une étude complémentaire à partir des quatre espèces collectées dans la même zone devra être mise en place pour confirmer l’influence de l’exposition ou des saisons sur le contenu en mycosporines.

CHAPITRE 3: ETUDE PHYTOCHIMIQUE DU DERMATOCARPON LURIDUM (WITH.) J.R. LAUNDON

Une étude phytochimique plus approfondie de D. luridum a été entreprise dans le but d’isoler dans un premier temps les mycosporines décrites précédemment. Le chapitre débute par une présentation de l’écologie et de la morphologie du D. luridum suivie d’un bilan bibliographique concernant les constituants retrouvés dans le genre Dermatocarpon. Il apparaît que ce genre est très pauvre en métabolites secondaires polyfonctionnalisés. En revanche, il est riche en polyols, en acides aminés et en caroténoïdes.

A partir de 150 g de lichen séché et broyé, nous avons procédé à une extraction successive par des solvants de polarité différente en commençant par l’eau puis le chloroforme (après séchage du lichen) et enfin l’acétone (Figure 34). L’extrait aqueux, le plus riche quantitativement et contenant les mycosporines a été traité en premier. La purification a consisté en une première étape de chromatographie échangeuse de cations afin d’éliminer les polyols. Puis, après quelques essais non concluants de chromatographie de partage centrifuge (CPC) pour éliminer les acides aminés, nous avons adopté une nouvelle combinaison pour la purification de ces composés : une flash chromatographie en mode HILIC suivie d’une chromatographie en phase inversée sur colonne ouverte et enfin une chromatographie semi- préparative en phase inversée (Figure 36). Cela a conduit à l’isolement de 5 mg de mycosporine glutaminol 1, 6,5 mg de mycosporine glutamicol 2 et 6 mg de la mycosporine ester 3. Leur spectre UV dans l’eau a été enregistré permettant ainsi de calculer leur coefficient d’extinction molaire à λmax = 310 nm. Une détermination de la constante d’acidité pKa des mycosporines 1 et 3 par spectrophotométrie a ensuite été entreprise pour la première fois. Les pKa de l’ordre de 1.2 montrent que les deux mycosporines se comportent comme des acides très forts. L’intérêt de cette étude est d’une part, de connaître sous quelle forme se trouve la mycosporine en fonction du pH du milieu et d’autre part, de mettre en évidence une forte délocalisation des électrons au sein du noyau cyclohexenone (Figures 45 et 46).

Trois autres composés polaires 4, 5 et 6 ont été isolés à partir de l’extrait aqueux (Figure 47). Il s’agit de trois substances azotées. Le composé 4 (acide 2-amino-3- acétylaminopropionique) est un composé appartenant à la famille des acides aminés non protéiques qui n’ont jusqu’ici jamais été isolés dans les lichens, ils auraient un rôle défensif contre les insectes herbivores. Le composé 5, la L-Proline est un acide aminé classique déjà décrit dans un Dermatocarpon, il joue un rôle de composé osmotiquement efficace dans les lichens et les algues vertes. Le composé 6 (γ-L-Glutamylglycine) est un dipeptide composé d’une unité glycine associé à un acide glutamique, la structure exacte a été élucidée grâce à la RMN et à la fragmentation en masse. Ce type de composé n’a jamais été décrit dans un lichen. A partir des extraits acétone et chloroforme quatre composés 7, 8, 9 et 10 ont été isolés. Le cerevisterol 7 est un stérol comportant 3 hydroxyles, la configuration absolue a pu être déterminée en comparant son pouvoir rotatoire avec celui de la littérature, il a déjà été décrit dans les lichens mais pas dans le genre Dermatocarpon. Le composé 8 est un céramide ou un sphingolipide constitué d’une sphingosine formant une liaison amide avec un acide gras. La

xxi

longueur des deux chaînes alkyles a pu être déterminée à partir de la fragmentation MS. C’est la première fois qu’un sphingolipide non glycosylé est décrit dans un lichen. Les céramides sont des composés très hydrophobes, par conséquent ils jouent un rôle d’imperméabilisant en empêchant la perte d’eau et pourraient aussi empêcher l’entrée de microorganismes. Les composés 9 et 10 ont été isolés en mélange 85:15, il s’agit de deux polyols, respectivement le D-volémitol et le D-mannitol.

Au total, 10 composés ont été identifiés à partir du D. luridum. Il est intéressant de souligner que ce lichen est pauvre en constituants polyphénoliques. Son profilage chimique est à rapprocher de celui d’un cyanolichen connu pour accumuler des polysaccharides et des substances azotées telles que les mycosporines.

CHAPITRE 4: TESTS BIOLOGIQUES ET EVALUATION DE LA PHOTOPROTECTION

Les différents composés obtenus en quantité suffisante ainsi que des extraits aqueux ont pu être testés pour leurs activités cytotoxiques et certains évalués pour leurs propriétés photoprotectrices. Ainsi la cytotoxicité des six composés isolés 2, 3, 4, 6, 7 et 8 a été évaluée sur six lignées cellulaires cancéreuses: Huh7, CaCo-2, MDA-MB-231, HCT116, PC3, NCI- H727, et deux lignées cellulaires de la peau non cancéreuses: HaCaT et Fibroblastes. La plupart des composés testés ne sont pas toxiques à une concentration de 25 µM, à l'exception du stéroïde cérevisterol 7 (Figure 67). Le composé 7 est en effet toxique sur sept lignées cellulaires

avec des valeurs de CI50 comprises entre 5 à 24 µM, à l'exception de la lignée cellulaire de cancer du sein humain MDA-MB- 231. Ce résultat est en accord avec l’activité cytotoxique du stéroïde décrite il y a peu [138].

L’absence de toxicité des mycosporines sur les cellules de la peau est un résultat très encourageant pour une éventuelle application dans le domaine de la photoprotection. Aussi l’évaluation des capacités photoprotectrices des trois mycosporines a été entreprise ainsi que celle des extraits aqueux bruts et semi-purifiés de D. luridum (respectivement DL1 et DL2) et D. miniatum (respectivement DM1 et DM2). Tout d’abord, nous avons commencé à évaluer leur capacité en tant que filtre UV étant donné que les mycosporines 1-3 possédaient un coefficient d’extinction moléculaire supérieur à 10000 M-1.cm-1 et une absorption maximale

xxii

dans les UV à 310 nm. L’activité photoprotectrice a été déterminée in vitro en enregistrant le spectre d’absorption dans une émulsion huile/eau contenant 10% de chaque produit. En utilisant une analyse en composantes principales (ACP), il apparaît que les 3 mycosporines ainsi que l’extrait DM2 (extrait semi-purifié de D. miniatum) s’avèrent être de bons candidats comme filtres UVB en comparaison du 4-methylbenzylidene Camphor (4-MBC), un filtre commercial (Figure 69). Nous avons aussi démontré que l’activité des extraits était corrélée avec leur teneur en mycosporines.

Suite à ces observations, leur phototoxicité sur cellules non tumorales de kératinocytes humains, soumises à des radiations UVA a été évaluée dans l’hypothèse d’une application cosmétique. Les résultats suggèrent que les extraits lichéniques et les composés purs sont non- phototoxiques (Tableau 14).

Un filtre solaire est d’autant plus intéressant qu’il possède une capacité antioxydante. En effet, les dommages dus aux UV passent par des mécanismes d’oxydation directs ou indirects des constituants de la peau et plus particulièrement de l’ADN, ce qui peut mener à terme l’apparition de cancers cutanés. Après un bilan bibliographique des propriétés antioxydantes des oxo-mycosporines et imino-mycosporines décrites à ce jour (Appendice 4), nous avons testé leur activité antioxydante en effectuant deux tests in vitro (DPPH et NBT) et un test in cellulo (peroxydation lipidique) mettant en jeu des mécanismes différents. Le test DPPH consiste à mesurer la capacité des composés à réduire un composé radicalaire (par transfert d’électrons) alors que le test NBT concerne principalement la capacité des molécules à piéger

.- l’anion superoxyde O2 . Les mycosporines 1-3 ainsi que les extraits semblent présenter peu d’activité réductrice sur le test DPPH alors que la mycosporine 3 et les extraits DL2 et DM2

.- présentent une activité notable dans le piégeage de O2 comparée au témoin positif, la quercétine (Figure 68). On note également une activité modérée des mycosporines 1 et 2 suggérant une interaction synergiste entre les constituants présents dans les extraits DL2 et DM2. Le dernier test de peroxydation lipidique a été effectué en collaboration avec Mme Isabelle Gallais et le Pr. Odile Sergent au sein de l’UMR Inserm 1085, IRSET à l’Université de Rennes 1. Ce test consiste à mesurer la quantité de malondialdehyde (MDA) formée suite à la production de ROS sous l’effet de l’éthanol sur des cellules WIF-B9. Seul l’extrait DM2 et la mycosporine 3 ont été testés car ils présentaient la meilleure activité antioxydante sur l’essai NBT. Le résultat de ce test préliminaire (car seulement issu de deux expériences) semble

xxiii

confirmer une activité antioxydante des deux produits basée sur la diminution de la peroxydation lipidique lorsqu’elle est induite par l’éthanol (Figure 73). Ce résultat très prometteur devra être confirmé par une troisième expérience. En effet, il montre pour la première fois une activité des deux produits comparable à la vitamine E à la différence près que la Vit E est lipophile et que l’extrait et la mycosporine 3 sont hydrophiles.

Enfin, dans le but de confirmer la potentielle utilisation des mycosporines 1-3 et de l’extrait DM2 en tant photoprotecteur d’origine naturelle, leur photostabilité sous irradiations UVA et UVB a été évaluée par HPLC-UV. La comparaison des chromatogrammes obtenus après irradiations ou non des solutions, montre une absence de dégradation des 4 produits après une exposition de 5 J/m2 alors qu’une dégradation du trolox apparaît dès 0.5 J/m2.

CONCLUSION

Le profil chimique intrigant du lichen dulcicole Dermatocarpon luridum correspondant à un chlorolichen de l’ordre des Verrucariales mais ne possédant pas les composés phénoliques habituels, nous a incité à approfondir une étude phytochimique sur ce lichen ainsi que sur quelques espèces voisines. Parmi les composés polaires, nous avons repéré un profil UV caractéristique de mycosporines, connues pour être des filtres UV naturels. Il s’agit ici de deux oxomycosporines dont l’une avait été signalée dans un cyanolichen, Degelia plumbea [3]. Ce résultat étant inattendu, nous avons d'abord vérifié la présence de mycosporines sur une variété d'échantillons de D. luridum et étendu notre prospection à une dizaine d’espèces proches. En employant des méthodes HPTLC-UV et HPLC-DAD-MS, deux oxomycosporines (Mycosporine glutaminol 1 et mycosporine glutamicol 2) ont été caractérisées dans les extraits aqueux des quatre espèces fraîches de Dermatocarpon (D. luridum, D. meiophyllizum, D. leptophyllodes, et D. miniatum) collectées en Bretagne. Un ester éthylique de la mycosporine glutamicol a également été identifié dans quelques extraits. Il correspond à un composé nouveau mais il est en réalité un artefact formé pendant le processus de purification.

Les taux de mycosporine dans les Dermatocarpons semblent être influencés par des intensités de rayonnements solaires puisque un contenu plus élevé en mycosporines est trouvé dans les espèces plus exposées (D. miniatum présente une teneur au moins deux fois supérieure en mycosporines par rapport à D. leptophyllodes et D. meiophyllizum qui sont avec D. luridum

xxiv

au niveau des cours d’eau et dans une exposition plus ombragée). Cependant, cette observation doit être validée par une expérience dédiée et des résultats statistiquement confortés.

Grâce à la possibilité d’avoir un protocole simple et sensible pour le dosage de ces mycosporines, une quantification de tels métabolites a pu être réalisée sur de très petites quantités de lichen et sur un grand panel de Dermatocarpon (89 échantillons). Cela nous a permis d’inclure des échantillons provenant de collections personnelles et institutionnelles. Les herbiers du muséum d'histoire naturelle de Londres (NHM) et de la collection Des Abbayes (REN-Abb) et Massé ont fourni l’essentiel des échantillons provenant d’Europe et d’Amérique du Nord. A partir d’un échantillon récolté en 1841 et conservé au NHM, 14 mg de mycosporine glutamicol par gramme de résidu sec d’un extrait aqueux de ce lichen (15 mg de thalle) ont été mesurés. Excepté pour quelques échantillons, la teneur totale des mycosporines se situe entre 0,1 et 1,4 mg/g de lichen sec avec une bonne reproductibilité pour des échantillons anciens mais conservés dans des conditions similaires. La diminution, dépendante du temps (baisse puis disparition totale en moins de 15 ans), de la mycosporine glutaminol alors que la mycosporine glutamicol s'accumule dans la plupart des échantillons étudiés est à rapprocher de la conversion probable du premier en second. L’absence de l'ester éthylique de la mycopsorine glutamicol dans ce protocole simplifié montre bien que ce n’est pas un produit originel. Le caractère stable de la mycosporine glutamicol dans les échantillons de Dermatocarpon conservés en herbier pourra sans doute être mis à profit pour une cinétique de datation à partir des données de référence que nous avons établies.

Il reste à vérifier que ces mycosporines - qui n’avaient jusqu’alors pas été recherchées compte tenu de la focalisation faite sur l’intérêt chimiotaxonomique des composés phénoliques - est bien réel en étendant plus largement le criblage. Si le repérage de ces composés est aisé, leur isolement l’est beaucoup moins car ces composés polaires sont peu abondants et mélangés aux acides aminés et aux sucres. L'isolement et l'identification structurale par spectroscopie ont permis d’établir la structure des deux oxomycosporines 1 et 2 et de l’artéfact 3 à partir de l'extrait aqueux de D. luridum.

Concernant leur détection dans les Dermartocarpon, l'ordre des Verrucariales doit maintenant être ajouté aux ordres des Peltigerales, des Lichinales et des Lecanorales dans lesquels des mycosporines avaient été caractérisées pour quelques espèces de lichens. Des iminomycosporines sont suspectées dans des lichens tripartites du genre Stereocaulon [190]

xxv

mais toutes les mycosporines identifiées jusqu’alors dans les lichens possèdent un noyau cyclohexenone. Ce noyau est caractéristique des mycosporines fongiques, ce qui nous pousse à émettre l’hypothèse que le mycobiote est impliqué dans cette biosynthèse. La mycosporine glutaminol et la mycosporine glutamicol avec leurs dérivés glycosylés ont d’ailleurs été décrits dans des Ascomycètes non lichenisés [38], [90], [91], [94]. D'autres expériences doivent être entreprises et en particulier la recherche des gènes de la Synthase DeHydroQuinique récemment caractérisés pour être impliqué dans la biosynthèse de mycosporines [54]. La mycosporine glutamicol pourrait être un marqueur taxonomique des Dermatocarpon et probablement des Verrucariales comme la mycosporine serinol qui est principalement trouvée dans les Peltigerales (8 Peltigeraceae) et la mycosporine glycine dans les Lichinales (7 espèces de Lichinaceae et de Peltulaceae) (Tableau 1). Etant donné le peu d’études phytochimiques sur les Verrucariales, il serait intéressant dans les perspectives à donner à ce travail d’étendre cette étude à d’autres espèces appartenant à cet ordre, la chimie pouvant venir en aide à la taxonomie des Verrucariales [215].

L’étude phytochimique sur le lichen D. luridum (150 g de thalle) a mené à l'isolement et à l'identification de dix composés. Un nouveau protocole de purification de mycosporines comprenant la chromatographie en mode HILIC, suivie d’une chromatographie en phase inverse avec HPLC-DAD a permis d’isoler une douzaine de mgs des deux mycosporines natives 1 et 2 avec l’artefact 3. Trois autres composés polaires ont été isolés dans l'extrait aqueux, y compris l'acide 2 amino-3-acetylaminopropionique 4, la L-proline 5, et la γ-L- glutamylglycine 6. À partir des extraits au chloroforme et à l’acétone, le cérevistérol 7, un céramide 8 et un mélange de D-volémitol 9 et de D-mannitol 10 (85:15) ont été également obtenus. Les composés 4, 6 et 8 n'ont pas été mentionnés dans les lichens jusqu'ici et le composé 7 est ici décrit pour la première fois dans D. luridum. Le céramide 8 est le premier représentant non glycosylé décrit dans les lichens et son rôle dans le lichen reste à élucider (stabilisation de la membrane ou médiateur de signalisation). Ces composés, qui sont généralement trouvés dans la nature sous une forme plus complexe, pourrait jouer un rôle important dans la rétention hydrique et la plasticité du thalle

Pour les quelques composés isolés en quantité suffisante, des essais biologiques ont été réalisés. La cytotoxicité de six d’entre eux 2, 3, 4, 6, 7 et 8 a été évaluée sur six lignées de cellules cancéreuses et deux lignées de Keratinocytes HaCaT et Fibroblastes qui sont informatives pour apprécier la tolérance possible sur la peau. La plupart des composés

xxvi

examinés ne sont pas toxiques à une concentration de 25 µM, excepté un stéroïde, le cérevistérol 7. L’absence de toxicité des mycosporines sur les cellules de la peau est un résultat très encourageant pour une éventuelle application dans le domaine de la photoprotection. De plus, la mycosporine 3 et l’extrait DM2 n'induisent pas de phototoxicité sur les Kératinocytes HaCaT exposées à des rayonnements UVA et ont également une bonne photostabilité sous irradiation UV. Ainsi ils ont les conditions requises pour être développés comme produits de protection solaire. Cependant, la synthèse ou l'approche biotechnologique doivent être envisagées en priorité pour leur obtention car le lichen ne peut pas être envisagé comme une source viable. Nous notons avant tout qu’une combinaison de composés semble être plus efficace que le composé pur mais le mélange optimal reste à étudier. Les activités antioxydantes obtenues avec ces composés semblent également être synergiques et la confirmation de résultats in cellulo pour la peroxydation lipidique de la mycosporine 3 est attendue avec grand intérêt car il est inhabituel d’obtenir de tels résultats avec des composés hydrophiles.

Ces tests antioxydants et photoprotecteurs ont toutefois leurs limites et nous préjugeons que ces composés sont considérablement plus actifs au sein des lichens comme le révèlent des expériences publiées en 2004 par Kranner et al [194]. Les effets photoprotecteurs et antioxydants du cycle des xanthophylles doivent certainement aussi y participer comme observés par Clother Coste pour des lichens subaquatiques [109].

En conclusion de ce travail, nos résultats ont permis de revoir un concept qui mettait jusqu'à maintenant en avant la production de mycosporine uniquement dans des cyanolichens [189]. Une prolongation du travail par des analyses génétiques pour repérer les clusters de gènes formant les mycosporines devra être entreprise dans les Dermatocarpons. De la même manière, d’autres lichens doivent être étudiés dans ces approches de génome-mining et en particulier des lichens tripartites formant des céphalodies. La distribution des mycosporines dans le thalle lichénique serait aussi importante à étudier, tant pour connaitre leur distribution que leur rôle. On notera ici que la plupart des organismes produisant des mycosporines sont des organismes très hydrophiles ou évoluant dans un milieu aqueux, cela soulève la question en lien avec l'adaptation terrestre des espèces marines. Comment interpréter d’autre part la relation qu’il peut y avoir entre un métabolisme de chlorolichen aboutissant à la production massive de composés phénoliques et celle d’un chlorolichen ayant un métabolisme plus proche de celui d’un cyanolichen?

xxvii

Par ailleurs, l’élargissement de l’étude à des échantillons issus de collections historiques de lichens en herbiers a été un défi passionnant. La valeur de ces collections pour la science est fortement sous-estimée. À notre connaissance, peu de dosages ont été réalisés sur les métabolites secondaires à partir de collections [195], [196], [197], [198]. Le développement des techniques analytiques nous permet de travailler sur des quantités d’échantillon très réduites ce qui nous ouvre un accès raisonné à ces ressources d’une valeur inestimable. En dehors de leur but principal de taxonomie et de référencement, ces collections d'herbier ont des implications dans les domaines environnementaux, pour l'analyse génétique, la régénération d’espèces mais aussi la constitution de banques de diversité chimique tout à fait remarquables, [206]. Bien que notre travail ait précisé la labilité de certains composés au cours du temps, nous avons pu mesurer toute la nécessité de maintenir et d’enrichir de telles collections, en y apportant notamment des données qui ouvrent la voie à de futurs travaux.

xxviii

Introduction

Over the past ten years, the team involved in natural products research in Rennes has a special interest in lichens. These symbiotic organisms are classified in the fungi kindom and result from the association between algae and/or cyanobacteria (the photobiont) and fungi (the mycobiont) and dedicated bacterial communities [1]. They appeared about 680 million years ago, primarily aquatic they were able to adapt to the terrestrial environment Based on this association, lichens are able to survive under extreme ecological conditions, such as intense radiation, extreme temperatures, dessication, etc. In response to environmental factors and taking into account their specific metabolism, lichens produce a variety of secondary metabolites, most of them being unique. With regard to UV radiation, lichens have developed various mechanisms of photoprotection such as the production of a series of photoprotective compounds [2]. Most of them are typical polyphenolic compounds but lichens also produce non-aromatic photoprotective compounds such as mycosporine-like compounds. Chapter 1 of the manuscript is devoted to the presentation of lichens and to their photoprotective responses, drafting an overview of mycosporine-like compounds as natural photoprotectants.

To now, the distribution of mycosporine-like compounds has been reported in cyanolichens (symbiotic organisms resulting from an association between a fungus and a cyanobacterium) [3]. Although chlorolichens correspond to 90% of lichen species, mycosporine-like compounds have not yet been described in chlorolichens in which cyanobacteria are replaced by green algae. From our interest in such metabolites in lichens, we performed a new screening in four lichen species, including two chlorolichens and two cyanolichens. As expected, mycosporines were detected in both cyanolichens and undetected in one chlorolichen. Interestingly, we observed the possible occurrence of such compounds in the second chlorolichen Dermatocarpon luridum - a foliose, freshwater aquatic chlorolichen belonging to the Dermatocarpon genus (Verrucariaceae).

Members of this genus exhibit rather diverse morphologies and habitats. Some species are aquatic or semi-aquatic lichens. Other members are not associated with watercourses and live on rocks or on soil [4]. In chapter 2, four available Dermatocarpon species located in Brittany (D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum containing Diplosphaera chodatii (Trebouxiophyceae) photobionts), were screened for mycosporines to

confirm the presence of such compounds in chlorolichens and to have a better understanding of their biogenesis in lichens. Due to occurrence of mycosporines in the four fresh species investigated and a suitable physico-chemical profile of these compounds for a UV dosage, quantification of such metabolites in ancient preserved herbarium samples (even more than 170 years) were then performed: 49 samples kindly offered by Dr. Holger Thüs from Natural History Museum (London, UK) and 30 samples originated from our herbariums at Rennes, France (H. des Abbayes and L.J.-C. Massé). Our work aimed to respectfully use a little part of the herbarium samples to generate data which could be useful for a possible survey of mycosporines over years and chemotaxonomy.

After this mycosporine screening and quantification, chapter 3 presents their isolation from D. luridum lichen material collected in Finistère (Brittany). With a suitable quantity of the purified mycosporines, we confirmed their structures, and determined some of their characteristic spectroscopic and physico-chemical properties. Moreover, this study contributed to investigate the other chemical constituents of D. luridum as a previous phytochemical screening revealed that chlorolichens belonging to the Dermatocarpon genus had no usual polyphenolic lichen substances.

Chapter 4 reports some biological tests carried on Dermatocarpon aqueous extracts and all the isolated compounds. The preliminary cytotoxicity screening on eight cell lines (ImPACcell Plateform, University of Rennes 1) provides an overview on cytotoxic activity. Beside the possible interest of highly cytotoxic compounds on cancer cells to be use in cancer therapy, the cytotoxic profile on a panel of cell lines is an important initial step to select compounds and an appropriate dose for further tests. In term of photoprotective activity, samples were evaluated as UV filters by the calculation of the absolute (UV-PF, λc, UVA-PF) and relative index (SUI, ISP). Their phototoxicity were then tested on HaCaT cells exposed to UVA radiations. Their antioxidant activity was evaluated in vitro assays based on electron transfer and superoxide radicals scavenging as well as in cellulo on WIF-B9 cell line by using ethanol-induced lipid peroxidation assay. Finally, to confirm their possible applications as natural photoprotectants, we evaluated their photostability under UVA and UVB exposure.

Beside the contribution to the chemical profile in the Dermatocapon genus along with the focus on lichen mycosporines, we hope this work will give new insights in lichen metabolism and also enlight the broad interest of lichens as a source of valuable compounds.

CHAPTER 1: LICHENS AND PHOTOPROTECTION

1 Presentation of lichens

1.1 Definition

Lichens are stable, consistent and identifiable mutualistic associations between algae and/or cyanobacteria (the photobiont) and fungi (the mycobiont) and dedicated bacterial communities. The plant body of lichen is called “thallus” which had been considered a single plant till 1867 when Schwendener and De Bary demonstrated the lichen thallus to be a symbiotic organism composed of a fungus and an alga. Nowadays, about 18500 lichen species have been described all over the world in which 85% of lichens have green algae as symbionts, approximately 10% have blue-green algae or cyanobacteria, and less than 5% of lichens have both green algae as primary symbionts and blue-green algae as secondary symbionts [5]. Trebouxia is the most common genus of green algae while Nostoc is the most commonly occurring cyanobacteria genus.

The mycobiont is unique in the symbiotic association and usually dominates the association, therefore, lichens are traditionally classified as a life-form of fungi and the scientific names used for lichens are actually those applied to the fungal partner. The photosynthetic partner has a scientific name related to some dozens of the algae or cyanobacteria genera and only given when independent lichen description. Due to an apparently successful form of fungal symbioses, the lichen-forming habit is maintained by one- fifth of all fungi, which includes more than ca. 40% of ascomycetes, but only a few basidiomycetes [6].

1.2 Thallus morphology and anatomy

Lichens are traditionally divided into three growth morphological forms: crustose (representing 90% of the lichen), foliose and fruticose [5] (Figure 1).

Figure 1: Major growth forms of lichen

Crustose lichens may be subdivided as leprose, endolithic, endophloeodic, squamulose, peltate, pulvinate, lobate, effigurate and suffruticose crusts. Majority of crustose lichens grows directly on the surface of the substrate and is referred to as episubstratic, while a small minority grows inside the substratum is called endosubstratic. The episubstratic thallus consists of a crust-like growth adherent or attached to the substratum throughout its underside by hyphae and cannot be detached without destruction.

The foliose or leafy thallus of lichens is typically flattened, dorsiventral, spreading and expanding horizontally outwards and usually attached to the substratum by rhizines arising from the lower surface. Foliose lichens develop a great range of thallus size, color and shape diversity.

Squamulose lichens are intermediate between foliose and crustose thalli in structure. The thallus is composed of numerous small lobes or squamules up to 1 cm long. The internal structure is similar to that of foliose lichens but without any lower cortex or rhizines.

Fruticose type of lichen is either erect or pendent shrubby growth attached to the substratum at the base by basal disc or holdfast (formed by mycobiont hyphae). Some groups have dorsiventrally arranged thalli but most of them exhibit a radial symmetric.

Among the different growth forms of lichens in the evolutionary series, leprose are considered pioneers followed by crustose, squamulose, foliose, and fruticose being the latest. Leprose, crustose and squamulose lichens are called microlichens as they are smaller in size and mostly require a microscopic study for identification. Foliose and fruticose lichens on the other hand are called macrolichens. Macrolichens have a comparatively larger thallus and a hand lens and stereozoom microscope are generally sufficient for identification of the most common species.

Anatomically, if the algae are scattered throughout the medulla of the thallus, the structure is homoeomerous or unstratified (Figure 2 a). A more complex structure found in the most widespread lichens thalli is heteromerous, where different layers are recognizable: an upper cortex, a medulla layer and a lower cortex, constituted by fungal pseudo-tissues; and a photobiont layer housed between the upper cortex and the medulla (Figure 2 b).

Lichens have two fundamentally different sorts of reproductive bodies: sexually reproducing fruiting bodies (apothecia, perithecia and lirella which are exclusively produced by the fungal partner) and asexually reproducing bodies (soredia, isidia containing both fungal hyphae and algal cells) (Figure 3).

Figure 2: Structure thallus homoeomerous (a) and heteromerous (b)

Figure 3: Reproductive organs of lichens [7]

1.3 Symbiosis

In lichens, the fungus plays a role in water and mineral supply and provides a mechanical protection of the organism. The algal partner realizes photosynthesis and provides carbohydrates which are then supplied to the fungus (Figure 4). The fungus consequently produces a variety of secondary metabolites. According to the type of carbohydrate provided to the fungus and fungal metabolism, the final secondary metabolite produced have different patterns. Green algae produce polyol (ribitol, erythritol, sorbitol) while cyanobacteria produce simple glucose. Besides, cyanobionts are capable of fixing atmospheric nitrogen and to provide fixed-nitrogen to the mycobionts. Therefore, cyanolichens have higher nitrogen content than the lichens with only green algae as photobionts [8]. As a general trend, these chlorolichens are producing a variety of polyphenolic metabolites while cyanolichens are considered to be devoided of these substances, accumulating high contents of polar compounds including large amounts of polysaccharides.

Figure 4: The nutritional exchange between partners of lichens [9]

The mycobiont is firmly connected to the photobiont through haustoria and cannot be easily separated. Thalli are generally closely attached to their support but fairly not dependent from the support for nutrients. Due to this type of relationship, lichens are eminently successful and enjoy worldwide distribution being encountered in every conceivable habitat and growing on a variety of substrata: on rock (saxicolous), trees (corticolous), soil (terricolous), wood (lignicolous), leaves (foliicolous), rubber, plastic, glass, etc.

1.4 Chemical compounds

Compounds produced by lichens fall into two classes: primary metabolites and secondary metabolites. Primary metabolites are intracellular products which are directly involved in the metabolic activities of the lichens such as growth, development and reproduction. They include proteins, amino acids, polyols and polysaccharides, which are bound to the cell walls and protoplasts. They are often water soluble and can be extracted with boiling water. The primary metabolites are either of fungal or algal origin or both. They are also nonspecific and present in free-living alga, fungus, higher plants and other organisms.

Secondary metabolites are produced by utilising the primary metabolites and are not involved in the direct metabolism of the lichen. They are the byproducts of primary metabolism and biosynthetic pathways act as storage substances, some having an important ecological role. Secondary metabolites are known to protect lichens against increasing environmental stresses such as light exposure, water potential changes, microbial and herbivore interactions and other changes associated with changes in environmental conditions. The secondary metabolites are deposited on the surface of the hyphae rather than within the cells. They are called extrolites as extracellular compounds and popularly described as lichen acids or lichen substances.

Probable hypothesis of secondary metabolites production is that a fungus undergoes maximum growth when all required nutrients are available in optimal quantities and proportions. If one nutrient becomes altered, then primary metabolism is affected and fungal growth is slowed, which may be due to abiotic factors including environmental changes [10]. The intermediates of primary metabolism that are no longer needed in the quantity in which they are produced may be shifted to another pathway. It is thought that the intermediates may be used in the secondary metabolic pathways serving as an alternative sink for the extra products of primary metabolism while allowing nutrient uptake mechanisms to continue to operate [11]. Thus secondary metabolites are mainly products of an unbalanced primary metabolism resulting from slowed growth, including metabolites that are no longer needed for growth.

The detoxification of primary metabolites is another hypothesis that has been proposed to explain the production of secondary metabolites. If growth of the fungus slows down, but metabolism is still very active, toxic products of primary metabolism may accumulate. The

transformation of these into secondary metabolites may be one method to prevent toxic accumulation of byproducts. This hypothesis may be integrated within the first hypothesis on slow growth rates to explain the production of secondary compounds by fungi [5].

The chemical aspect of lichen substances was published by Zopf in early 19th century. The structure of most lichen substances remained unknown till the studies of Asahina and Shibata in early 20th century. The development of TLC and HPLC in 1960s, together with modern spectroscopic methods led to the isolation and identification of many new lichen

substances [12], [13]. Most lichen substances are phenolic compounds, dibenzofuranes, depsides, depsidones, depsones, lactones, quinones, and pulvinic acid derivatives with more than 1050 secondary metabolites known. The lichen substances are synthesised mainly via three metabolic pathways (Figure 5):

- Acetyl-polymalonyl pathway: includes most common lichen substances (anthraquinones, xanthones, usnic acid, depsidones, depsides, etc.) and about 80% of the secondary metabolites are synthesised through this pathway.

- Mevalonic acid pathway: produces about 17% of total compounds including steroids and triterpenes.

- Shikimic acid pathway: very small portion of secondary metabolites are synthesised by this pathway including terphenylquinones and pulvinic acid derivatives.

Figure 5: Probable pathways leading to the major groups of lichen products [14]

The pharmacological activities of lichen substances can be divided into: antibiotic activity, antitumour and antimutagenic activity, antiviral activity, enzyme inhibitory activity [6]. Recently, UV protectant metabolites from lichens and their symbiotic partners have been remarked [2], [15], [16]. In fact, these organisms can produce unique and/or efficient UV filters to counteract the damaging effects of UV radiations.

2 Photoprotective responses of lichens

2.1 Harmful effects of UV radiation

The increase in solar UV radiation on the earth’s surface is now recognized as a major environmental factor deleterious to all sun-exposed organisms [17]. The UV spectrum is conveniently divided into three groups based on wavelength: UVC (100–290 nm), UVB (290– 320 nm) and UVA (320–400 nm). Solar UV irradiation at the earth’s surface is approximately 90–99% UVA and 1–10% UVB while all the UVC having completely absorbed by the stratospheric ozone [18] (Figure 6). UVB is a small but highly active component of the solar spectrum and strongly absorbed by DNA. Cyclobutane pyrimidine dimers (CPDs) and pyrimidine-(6-4)-pyrimidone photo-products (6-4PPs) are the two major types of mutagenic and cytotoxic DNA lesions induced by UV radiation. These lesions can cause severe structural distortions in the DNA molecule and upset several essential cellular processes, such as DNA replication and transcription, leading to mutagenesis, carcinogenesis and even apoptosis or cell

death [19], [20] . Radiation in UVA range is associated with lower energy but has the ability to penetrate deeper into the cell. In contrast to UVB, UVA is poorly absorbed by DNA, but excites numerous endogenous chromophores, generating reactive oxygen species (ROS) such as singlet oxygen, hydroxyl radical, superoxide radical, and hydrogen peroxide. The predominant ROS-induced lesions formed are oxidized bases, such as 8-oxo-dG with DNA single and double strand breaks [21].

Figure 6: Electromagnetic spectrum of UV radiation and biologic effects

2.2 Photoprotective mechanisms

Lichens have a variety of physiological and biochemical mechanisms available for reducing the damage incurred by exposure to UV radiation, including screening, quenching and repair (Figure 7).

Figure 7: Defence strategies adopted by lichens to counteract the harmful effects of UV radiation

2.2.1 Screening mechanism

Screening can eliminate or at least reduce exposure to UV radiation by absorbing or reflecting damaging wavelengths before they reach to sensitive cellular components. Screening may consist of the production of physical barriers such as morphological or structural features and chemical barriers by production of chemical compounds that absorb UV radiation.

In physical screening, protective strategies are dependent on the anatomy of the vegetative structures (thallus shape, cortex thickness, presence of refractive structure or compounds, etc.) and the state of hydration of the lichens. The heteromerous thallus is optimized for the protection of the photobiont against excessive light. The cortex of the thallus could absorb light radiation 2- to 10-fold more than the epidermis of a leaf. Anatomical adaptation of the upper cortex can appear to significantly decrease the transmission of incident light at the algal layer. Physical light reflection may also be involved in “pruina”, the calcium oxalate accumulation in the externa surface [2].

An important chemical protective mechanism is the production of UV screening compounds such as depsidones, depsides, diphenyl ether, anthraquinones, xanthones or shikimate derivatives (scytonemin, pulvinic acids). Such compounds can absorb the damaging wavelengths and dissipate the absorbed energy as heat to the external environment in a short time, without the production of a reactive oxidating species. (See Figure 10 for chemical structures).

2.2.2 Quenching mechanism

The interaction of UV radiation with some photosensitizing molecules, organic molecules and oxygen, results in the production of toxic photoproducts both intracellularly and in the external environment. Toxic photoproducts have the ability to cause more damage than the UV radiation exposure itself. Toxic photoproducts are neutralized by various agents including antioxidants such as ascorbate, quenchers such as carotenoids and various scavenging enzymes: superoxide dismutase, catalase, peroxidases and ascorbate peroxidases. -. Superoxide dismutase generates H2O2, removing the superoxide anion via the reaction: 2O2 + + 2H → H2O2 + O2, and catalase is one of the main enzymes that scavenges hydrogen peroxide.

Peroxidases detoxify harmful H2O2 present in almost every organism. Ascorbate peroxidases

catalyzes the reduction of H2O2 to water using the reducing power of ascorbate. These enzymes

counteract damage caused by UVB stress and thus contribute to the survival of lichen under UV stress [22].

Carotenoids, perhaps best known as photosynthetic accessory pigments, are also radical- 1 trapping antioxidants, which scavenge oxygen radicals, neutralize the singlet oxygen ( O2), quench the triplet state of chlorophyll a which occurs under excess light exposure, and inhibit lipid peroxidation. Excess energy is dissipated as heat in the cyclic transformations of carotenoids known as xanthophyll cycles. Two major xanthophyll cycles are involved, such as the violaxanthin cycle and the diadinoxanthin cycle. During high light exposure, violaxanthin is converted into antheraxanthin and subsequently into zeaxanthin, which then accumulates. Until there is a sufficient decrease in light exposure, zeaxanthin reconverts into violaxanthin. Likewise, diatoxanthin was accumulated under exposure to high photon flux densities and was then reverted back into diadinoxanthin under dark conditions with no new pigment synthesis involved [23] (Figure 8).

Figure 8: Two major xanthophyll cycles involved in the dissipation of excess light energy

2.2.3 Repair mechanism

In fact exposure to UV radiation cannot be avoided completely and therfore all organisms require some capacity to repair damage caused by these wavelengths. Little is known about the repair mechanism induced by UV radiation in lichens. Most of such knowledge come from research on bacteria, which may be applied to lichens. The repair mechanisms include photoreactivation, excision repair, recombinational repair, SOS response and de novo protein synthesis [24].

2.3 Chemical features of UV absorbing compounds

One of the important protective mechanism of lichen is the production of UV absorbing compounds. A molecule is capable of absorbing UV radiation due to electronic transitions within the molecule, from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO). The electrons in a molecule can be one of three types, namely σ (single bond), π (multiple-bond), or n (non-bonding).

- σ-bond electrons have the lowest energy level and are the most stable electrons. These would require a lot of energy to be displaced to higher energy levels. As a result these electrons generally absorb light in the lower wavelengths of the UV light and these transitions are rare. - π-bond electrons have much higher energy levels for the ground state. These electrons are therefore relatively unstable and can be excited more easily and would require lesser energy for excitation. These electrons would therefore absorb energy in the UV and visible light radiations. - n-electrons or non-bonding electrons are generally electrons belonging to lone pairs of atoms. These are of higher energy levels than π-electrons and can be excited by UV and visible light as well.

The highest energy electrons in the ground state may occupy σ, π or n orbitals and the corresponding excited state of the lowest electronic energy correspond to π* or σ* levels. Most of the absorption in the UV spectroscopy of organic compounds are based on transitions for n or π electrons to the π* excited state because the energies required for these processes bring the absorption bands into the UV-visible region (200 to 700 nm). In addition, the compounds containing heteroatoms such as halogens, nitrogen, oxygen or sulfur with unpaired electrons will give rise to n → σ* transitions or n→ π* transitions in the case of the presence of a heteroatom associated with a double bond. Thus the non-bonding electrons cause shifts in the absorption maxima (λmax) to longer wavelengths and the intensity (ε) [2]. Molecules containing functional groups which are capable of absorbing UV-visible radiation are called chromophores. The characteristics of some chromophores are presented in Figure 9.

Figure 9: UV absorption characteristics of common functional groups [2]

The π-electron system is one of the most effective UV radiation absorbers. They are primarily found in conjugated bond structures that may be represented both in a linear chained molecule with alternating single and double bonds and in many aromatic and cyclic compounds containing electron resonance. Such chemical features are found in some lichen metabolites classes (depsides, depsidones, dibenzofurans, xanthone, etc.) which are highly polyfunctionalized allowing them to efficiently absorb UV radiation (Figure 10).

Lichens and cyanobacteria also produce non-aromatic absorbing UV compounds such as mycosporines and MAAs. Due to their physico-chemical properties (absorption pattern), mycosporines and MAAs are considered as highly efficient natural UV blockers in the sunscreen formulations.

Figure 10: Chemical structures of compounds from lichens reported as UV screens

3 Mycosporines and MAAs – natural photoprotectants

3.1 History

The existence of these UV protecting molecules had been known for some time before their exact chemical nature was revealed. For example, Wittenburg (1960) was first to report a

strong UVB absorbing agent (λmax 305 nm) in the gas gland of an epipelagic jellyfish, but the

isolation of substance was performed nine years later [25], [26]. Similarly, Tsujino and Saito (1961) and Shibata (1969) had reported the existence of molecules capable of strongly absorbing UV radiation in the 310–360 nm region in red algae, in aqueous coral extracts, respectively, but without characterization [27], [28]. Independently, other research groups working on fungi reproduction mechanisms noticed the production of unidentified substances in the mycelia of several fungi (e.g., Pyronema omphalodes, Alternaria chrysanthemi, and Ascochyta pisi), when sporulation was induced by UV light [29]. Molecules having a UV absorption maximum at 310 nm were designated P-310. However, no correlation was made at the time between those studies and previously reported marine organisms observations. Consequently, fungi continued to be explored, and UV-absorbing compounds similar to P310 continued to be observed. Only several years later the structure of the first P310 was finally elucidated, after its isolation from the basidiomycete Stereum hirsutum. This compound became known mycosporine I, as its biosynthesis was largely observed in fungi and thought to be related to sporulation and reproduction mechanisms [30]. Subsequently, Ito and Irata first isolated and characterized mycosporine glycine from the tropical zoanthid Palythoa tuberculosa [31], a compound with the same basic chemical structure of fungal metabolites and this was followed by other similar molecules [32].

From 1977 onwards, this type of molecule began to be more largely isolated and identified from many organisms, such as cyanobacteria, algae, phytoplankton and even animals

[33], [34], [35], [36], [37], [38] (Figure 11).

Figure 11: Organisms contain mycosporines and MAAs

3.2 Characteristics

Mycosporines are a group of fungal compounds with a strong absorption at 310 or 320 nm resulting from the linking of a cyclohexenone ring (oxo-mycosporine) with the nitrogen substituent of an amino acid or amino alcohol. Two different absorption maxima correspond to two different chromophores: mycosporines with 2-OMe (λmax at 310 nm) and nor- mycosporines with 2-OH (λmax 320 nm). Various mycosporines have been identified depending on the attached substituent with the only amino acids involved being serine, glutamine and glutamic acid or their corresponding amino alcohols, serinol, glutaminol and glutamicol, respectively, or the amino acid, alanine [33] (Figure 13 a).

In contrast to fungal metabolites, and with two exceptions (mycosporine glycine and mycosporine taurine are oxo-mycosporines), MAAs from marine organisms are imine derivatives of mycosporines which contain a cyclohexenimine ring (imino-mycosporine) linked to an amino acid, amino alcohol or amino group, having absorption maxima between 310 and 360 nm. Generally a glycine subunit is present on the C3 of the cyclohexenimine ring [33]. However, in some corals glycine has been replaced by methyl amine [39]. Some MAAs isolated from corals also contains sulfate esters [40]. Recently a rare novel MAA, containing the amino acid alanine: euhalothece-362, was isolated from the unicellular cyanobacterium

Euhalothece sp [41]. Another novel MAA tentatively identified as dehydroxyl usujirene was isolated from the cyanobacteria Synechocystis sp.[42] Several forms of atypical MAAs have been isolated from the dinoflagellate Alexandrium species; one of them (M-333) appears to be a mono-methyl ester of shinorine [43]. To date, at least 21 MAAs have been described from various marine organisms (Figure 13 b).

Figure 12: Structure characteristics of oxo- and imino-mycosporines

In our study, the use of “mycosporine-like compounds” is preferred to mycosporines and MAAs as these terms imply to know the origin of the molecules. Mycosporine-like compounds are low-molecular-weight (generally ˂ 400 Da) water-soluble molecules absorbing UV radiation, with the maximum absorbance between 310 and 365 nm and a high molar extinction coefficient (ɛ= 28000 to 50000 M-1.cm-1) [44] (Figure 12). Altogether, up to 40 mycosporine-like compounds have been described from both terrestrial and marine sources some bearing functional groups or being covalently linked with saccharidic units [45], [46],

[47], [48] .

OXO-MYCOSPORINES O Absorbance in nm Mycosporine glutamine R OCH 2 3 Extinction coefficient ε H in M-1.cm-1 HO COOH 310 HO NH Molecular weight in Da R1 Molecular formula 316 C H N O C H NO 13 20 2 7 R2 8 11 4 CONH2

R1 Mycosporine glutamic acid Mycosporine glycine H 310 COOH 311 H 28100 20892 245 317 C10H15NO2 C13H19NO8 COOH COOH

Mycosporine serinol Mycosporine alanine (Mycosporine I) 310 H 309 H 25516 OH Me 259 261 C H NO C H NO 11 17 6 OH 11 19 6 COOH Mycosporine glutamicol Collemine A H 310 HO OH OH 303 O C13H21NO7 HO HO O 310 COOH HO 34000 496 Mycosporine hydroxyglutamicol NH O C19H32N2O13 O H O OH OH 310 HO NH2 319 HO OH C13H21NO8 COOH Nor mycosporine glutamine Mycosporine glutaminol O OH 320 H 33000 302 310 HO NH OH HO C12H18N2O7 302 COOH C13H22N2O6 CONH 2 H NOC 2 (a) (b) Figure 13: Common fungal mycosporines (a) and MAAs from marine organisms (b): structures, chemical and spectroscopic characteristics [49]

3.3 Biosynthesis

Multiple investigations have demonstrated the role of the shikimic acid pathway in the biosynthesis of 4-deoxygadusol (4-DG), a precursor common to both fungal mycosporines and cyanobacterial MAAs (Figure 14). In fungi, evidence was initially derived from the metabolic incorporation of a 14C-labeled 3-dehydroquinate intermediate (DHQ) by the fungal parasite Trichothecium roseum, which subsequently transformed DHQ into structurally related mycosporines [50]. Radiolabeling experiments on the cyanobacterium Chlorogloeopsis sp. PCC 6912 later confirmed that the shikimic acid pathway is involved in MAA biosynthesis, through the incorporation of 14C-labeled pyruvate, an early and obligate precursor of this pathway, and its selective transformation into radiolabeled MAAs [51]. In the same study, it was also shown that the polyketide pathway was not involved, as 14C-labeled acetate was not traceable in the MAA cytoplasmic pool of metabolites. Additional evidence was obtained via the use of exogenous tyrosine (a shikimic acid pathway repressor), while the use of glyphosate (Roundup) as a shikimic acid pathway inhibitor was shown to block MAA biosynthesis in the cyanobacterium Nostoc commune [52], and also in the scleractinian coral Stylophora pistillata [53].

Figure 14: Postulated convergent pathways of MAAs biosynthesis in cyanobacteria [54]

However, unambiguous elucidation of a different MAA biosynthetic pathway in cyanobacteria was recently reported [54]. Biosynthesis of shinorine in Anabaena variabilis (ATCC 29413) and Nostoc punctiforme (ATCC 29133) via the intermediates 4-deoxygadusol and mycosporine-glycine was found to proceed from the sedoheptulose 7-phosphate (SH 7-P)

precursor of the pentose phosphate pathway, rather than 3-deoxy-D-arabinoheptulosonate 7- phosphate (DAHP) of the shikimic acid pathway. Instead of DHQ, 2-epi-5-epi-valiolone (EV, a product of the pentose phosphate pathway) was found to be the precursor of 4-DG in this cyanobacterium. Then it appears that 4-DG can be produced in a convergent manner as the general MAA precursor, via two different biosynthetic pathways.

3.4 Role of mycosporine-like compounds in nature

The major function of mycosporine-like compounds is their capacity to act as photoprotective UV filters in various organisms. However, more and more evidences support that such compounds may have multiple roles: as antioxidant molecules, they are also involved in salt stress, desiccation or fungal reproduction [44].

� Fungal reproduction

Light, and more specifically UV light, is often a requirement for the formation of reproductive organs in fungi. It is sensed by light-induced pigments that specifically absorb between 240 and 310 nm. Colonies grown in the dark without such UV-absorbing pigments do not sporulate. Mycosporines cannot be considered as direct photoproducts, but appear to be connected to a type of metabolism that occurs at a very low level during mycelial growth. Their production increases during the reproductive morphogenesis, and they can thus mediate changes in the cell membranes and the structure of the cell wall [33].

� Photoprotective UV filters

A correlation between mycosporine-like compound content and in situ irradiance levels has been noticed in many locations worldwide and in a wide variety of organisms. Indeed, a high light exposure leds to such metabolites accumulation. For example, epilithic cyanobacteria are known to possess MAAs to protect them from UV radiation [55]. Recently, it was found that rock-inhabiting fungi also produce considerable amounts of mycosporines absorbing in the UV range with wavelength maxima centered around 310nm [56]. The role of mycosporine-like compounds in the protection against UV-induced damage has been demonstrated in a number of studies [33], [35], [57].

� Antioxidant

Mycosporine-like compounds exhibit a high antioxidant activity, scavenging superoxide anions and inhibiting lipid peroxidation resulting from UV induced production of ROS [44]. Antioxidant activities of these molecules have been related to the presence of the cetonic group of oxo-mycosporines, while imino-mycosporines, were oxidatively robust but they were presented as strong antioxidants in lipid medium (phosphatidylcholine peroxidation inhibition assay) [58]. We summarize mostly antioxidant assays of pure and mixture of mycosporine-like compounds regarding to their structures (oxo- and imino-mycosporines) and mechanism of each assay in Appendix 4.

� Salt stress

Oren reported the accumulation of large concentrations of MAAs in a community of unicellular cyanobacteria inhabiting on the bottom of a hypersaline saltern pond. It suggests role of MAAs in osmotic stabilization of the cells [44]. Being polar, highly soluble, uncharged or zwitterionic amino acid derivatives, mycosporine-like compounds fulfill at least part of the criteria for osmotic solutes [59].

� Desiccation

There are many reports on the occurrence of high concentrations of mycosporine-like

compounds in microorganisms exposed to desiccation [60], [56]. For example colonies of black melanized fungi (Sarcinomyces, Coniosporium, Phaeotheca) growing on desert rock surfaces, where they are exposed to desiccation, UV radiation, and nutrient scarcity, contain high concentrations of mycosporine glutaminol glucoside. The presence of this compound may be related to the survival potential and longevity of the vegetative hyphae of these fungi [56].

Finally, mycosporine-like compounds are true “multipurpose” secondary metabolites, which have many additional functions in the cell beyond their well-known UV sunscreen role.

3.5 Mycosporine-like compounds in lichens

To date, the screening of mycosporine-like compounds in 3 chorolichens (green alga photobiont), 31 cyanolichens (cyanobacterium photobiont) and in 5 tripartite lichens (green alga + cyanobacterium photobionts) has demonstrated that the distribution of mycosporine-like

compounds was only reported in cyanolichens and in tripartite lichens [3], [55], [61], [62]. Table

1 shows the occurrence of collemin A, mycosporines glycine, serinol, hydroxyglutamicol and glutamicol in different cyanolichens and tripartite lichen species. Other mycosporines-like compounds have been detected but not yet identified.

To achieve a better comprehension about occurence of mycosporines-like compounds in lichens, a new screening of such metabolites were performed on four lichen species, including two chlorolichens (Dermatocarpon luridum and Lepraria membranacea) and two cyanolichens (Leptogium cyanescens and Nephroma parile). Interestingly, mycosporines were detected for the first time in the chlorolichen Dermatocarpon luridum. Therefore, other species of genus Dermatocarpon should be investigated to confirm the presence of such compounds and to understand more their biogenesis in lichens.

Table 1: Mycosporine-like compounds distribution in lichens

Lichens Order Family Alga partner Cyanobacterium Mycosporine-like compounds λmax Ref.

partner (nm) Cladonia portentosa* Lecanorale Cladoniaceae Trebouxia nd [62] Cetraria islandica* Lecanorale Parmeliaceae Trebouxia nd [62] Lepraria membranacea* Lecanorale Stereocaulaceae Asterochloris nd ps Lichina confinis Lichinales Lichinaceae Calothrix mycosporine serinol 310 [3] Lichina pygmaea Lichinales Lichinaceae Calothrix mycosporine serinol 310 [3] Ephebe lanata Lichinales Lichinaceae Stigonema + 310 [62] Gonohymenia nigritella Lichinales Lichinaceae Gloeocapsa mycosporine glycine 310 [55] Peltula clavata Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Cyanosarcina Peltula euplaca Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Myxosarcina Peltula patellata Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Peltula tortuosa Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Peltula umbilicata Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Peltula impressula Lichinales Peltulaceae Chroococcidiopsis mycosporine glycine 310 [55] Myxosarcina Peltigera membranacea Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [62] Peltigera collina Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [62] Peltigera rufescens Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [62] Peltigera horizontalis Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [3] Peltigera polydactylon Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [3] Peltigera praetextata Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [3] Peltigera leucophlebia Peltigerales Peltigeraceae Nostoc mycosporine serinol 310 [3]

Peltigera aphtosa** Peltigerales Peltigeraceae Coccomyxa Nostoc + 310, 330 [62] Solorina saccata** Peltigerales Peltigeraceae Coccomyxa Nostoc mycosporine serinol 310 [62] Lobaria scrobiculata Peltigerales Lobariaceae Nostoc + 310 [3] Pseudocyphellaria crocata Peltigerales Nephromataceae Nostoc + 310 [3] Nephroma laevigatum Peltigerales Pannariaceae Nostoc mycosporine hydroxyglutamicol 310 [3] Degelia plumbea Peltigerales Pannariaceae Nostoc mycosporine glutamicol 310, 320 [3] Leptogium lichenoides Peltigerales Collemataceae Nostoc nd [62] Leptogium cyanescens Peltigerales Collemataceae Nostoc + 310 ps Nephroma parile Peltigerales Nephromataceae Nostoc + 310 ps Collema flaccidum Lecanorales Collemataceae Nostoc + 310 [3] Collema sp1 and sp2 Lecanorales Collemataceae Nostoc + [3] Collema cristatum Lecanorales Collemataceae Nostoc collemin A 310 [61] Collema cf. coccophorum Lecanorales Collemataceae Nostoc mycosporine glycine 310 [55] Stereocaulon dactylophyllum Lecanorales Stereocaulaceae Stigonema + 315 [3] Stereocaulon grande Lecanorales Stereocaulaceae Stigonema + 310 [3] Stereocaulon vesuvianum Lecanorales Stereocaulaceae Stigonema + 310 [3] Stereocaulon graminosum** Lecanorales Stereocaulaceae ? Nostoc + [3] Stereocaulon montagneanum** Lecanorales Stereocaulaceae ? Scytonema + 335 [3] Stereocaulon halei** Lecanorales Stereocaulaceae ? Scytonema + 310 [3]

*: chlorolichen having a green alga photobiont; **: tripartite lichen having a green alga and a cyanobacterium photobionts; Other lichens without asterisk are cyanolichens having cyanobacteria photobionts; nd: not detected; ps: present study; + : detected by HPTLC-UV spectrophotodensitometry and HPLC-DAD-MS.

CHAPTER 2: SCREENING AND QUANTIFICATION OF MYCOSPORINE-LIKE COMPOUNDS IN THE DERMATOCARPON GENUS

Purpose of this study

Previous studies showed that some mycosporines can be regarded as chemotaxonomic markers such as mycosporine serinol in Lichina genus (two species) and Peltigera genus (seven species) and mycosporine glycine in Peltula genus (six species). However, the presence of mycosporine serinol seems not to be restricted to a given photobiont since the cyanobacteria photobionts of Lichina and Peltigera are different, i.e. Calothrix and Nostoc respectively. Likewise, the photobiont Nostoc is not involved only in the production of a unique mycosporine molecule since mycosporines glycine, serinol, hydroxyglutamicol, glutamicol and collemin A were detected in lichens having the same Nostoc photobiont such as Collema cf. coccophorum, Peltigera species, Nephroma laevigatum, Degelia plumbea and Collema cristatum, respectively. As a result, the involvement of the mycobiont in the mycosporine biosynthesis has to be considered. Moreover, all of the mycosporines identified so far in cyanolichens have the cyclohexenone pattern. This cyclohexenone pattern gives maximum UV absorption at around 310 nm and is characteristic of fungal mycosporines [3]. However, one can raise the question of the cyanobacterium photobiont role in the biosynthesis of mycosporines, as only cyanolichens and tripartite lichens are described to contain mycosporines. The cyanobacteria photobiont can absorb atmospheric nitrogen and incorporate it into their metabolism whereas the chloroalgae photobiont are not designed for. Therefore, the aim of this study is to investigate the accumulation of such compounds in other chlorolichen species to allow a better comprehension of their biogenesis in lichens.

Chlorolichens metabolites profile is generally characterized by a variety of specific phenolic compounds and most of them have UV absorbing properties [63] (Figure 10). A previous phytochemical screening revealed that chlorolichens belonging to the Dermatocarpon genus did not contain the usual lichen substances such as depsides, depsidones, dibenzofurans,

xanthones, anthraquinones which are considered as photoprotectants [64], [65], [66], [67], [68]. It was hypothesized that mycosporine-like compounds may distribute and play an UV- protective role in Dermatocarpon species.

- The first part of our work is screening of mycosporine-like compounds in four Dermatocarpon species (D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum) collected in Brittany, France by using HPTLC-UV and HPLC-DAD-MS methods.

- In the second part, since an unexpected occurrence of mycosporines in the four fresh Dermatocarpon species was revealed, we went on studying their contents in ancient preserved herbarium samples including 49 samples from Natural History Museum (London, UK) which kindly provided by Dr. Holger Thüs (Curator of Lichens and Myxogastria–Life Science Department) and 30 samples from our herbariums at Rennes, France (H. des Abbayes and L.J.- C. Massé). It was found of interest to confirm the distribution of such metabolites in samples stored in herbarium even more than 100 years.

With current apparatus in our laboratory and mycosporine standards, we therefore optimized for the first time a method using HPLC-DAD in HILIC condition (Hydrophilic Interaction Liquid Chromatography) with an internal calibration curve for quantification mycosporines in the Dermatocarpon genus. The impact of species, geographic areas and storage conditions on mycosporine concentrations were here in discussed.

1 Screening of mycosporine-like compounds

1.1 Materials and methods

1.1.1 Lichen material

We investigated four of the five Dermatocarpon species found in Brittany (France), including D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum var. miniatum.

Classification:

Embranchement: Ascomycota

Class: Eurotiomycetes

Order: Verrucariales

Familly: Verrucariaceae [69]

Members of the genus Dermatocarpon are foliose, grey to brown, attached by one or more holdfasts, lower surface without rhizines. Perithecia are immersed in the thallus with ostiole usually clearly visible. Photobiont is a coccoid green alga (Diplosphaera). Ascospores are simple, colourless, ellipsoid, smooth, without perispore, 8-25 µm long, usually irregularly arranged in the ascus [4], [70].

Figure 15: The two different habitats of the four Dermatocarpon species investigated

The four investigated lichens have a very different habitat and ecology as D. luridum, D. meiophyllizum and D. leptophyllodes are sub-aqueous freshwater lichens while D. miniatum stands on dry rock surfaces closed to the sea (Figure 15). Detailed concerning habitat, voucher species, date and place collection were showed in Table 2.

Each species investigated was identified following the key determination (Appendix 1) by macroscopic and microscopic characteristics, and their reaction with Melzer’s reagent [71], [72], [73]. Here, Melzer’s reagent is employed to test for chemical differences in cell wall composition. The chemicals in the cell walls that react with the iodine Melzer’s reagent are not fully understood. However, the colours that result from the reaction are consistent. If the colour changes to blue-black, violet black or black, the lichen is called amyloid; if the colour changes

to reddish brown, it is called dextrinoid‐ (incomplete degradation products of polysaccharides); if little or no colour change occurs, that is inamyloid lichen [74].

Table 2: Collected samples data for screen mycosporine-like compounds

Lichens Collection place Habitat Date Voucher species

D. luridum Kerscollier (N 47 56.629’ on submerged 09/2013 JB/001/09/2013 W 3 28.440’), rocks Querrien, Finistère, Brittany, France D. meiophyllizum Kerscollier (N 47 56.629’ on submerged 09/2013 JB/002/09/2013 W 3 28.440’), rocks Querrien, Finistère, Brittany, France D. leptophyllodes Kerscollier (N 47 56.629’ on submerged 09/2013 JB/003/09/2013 W 3 28.440’), rocks Querrien, Finistère, Brittany, France D. miniatum var. Porz Lamat (N 47 46.329’ on rocks close 09/2013 JB/004/09/2013 miniatum W 3 37.246’), to the sea Moëlan-sur-Mer, Finistère, Brittany, France

Macroscopic and microscopic characteristics

Dermatocarpon luridum (With.) J.R. Laundon 1984

Thallus: upper surface brownish to greyish, green when wet, lower surface light to dark brown; multilobed, with several holdfasts, lobes 4–25 mm; 150–440 µm thick when wet; epinecral layer with air filled hyphae or compressed hyphae; medulla Melzer’s I+ reddish brown; photobiont coccoid green alga (Diplosphaera); perithecia: 210–490 × 150–520 µm; ascospores: 10.5–20 × 5.5–8.5 µm; pycnidia: frequent, pycnospores 3.5–6 × 1 µm.

Dermatocarpon meiophyllizum Vain. 1921

Thallus: 3–12 (20) mm in diameter, umbilicate, epruinose due to compressed hyphae in the epinecreal layer; upper surface brownish, green when wet, lower surface dark brown, smooth or finely granulose, rarely with a few wrinkles or folds; 390–640 µm thick; medulla Melzer’s I-; photobiont coccoid green alga; perithecia: 200–420 × 120–420 µm; ascospores 11–20.5 × 5–10.5 µm.

Dermatocarpon leptophyllodes (Nylander) Zahlbruckner 1921

Thallus: upper surface grey, lower surface brown to dark brown, umbilicate; lobes small, 1–4 (7) mm in diameter, 170–420 µm when wet; medulla Melzer’s I-; photobiont coccoid green alga; perithecia: 200–290 × 160–260 µm; ascospores 13–21 × 5–8 µm.

Dermatocarpon miniatum (L.) W. Mann var. miniatum 1825

Thallus: up to 10–70 mm diameter, pale grey to light brown, umbilicate, 300–500 µm thick; dark brown spots cover the surface, these are the openings from the reproductive perithecia; medulla Melzer’s I-; photobiont coccoid green alga (Diplosphaera); perithecia immersed, about 300 µm diameter; ascopores 10–14 × 5.0–6.0 µm.

Table 3 shows morphological comparison of the four Dermatocarpon species investigated in terms of macroscopic and microscopic characteristics.

Table 3: Morphological comparison of the four Dermatocarpon species

Species Macroscopic characteristic Thallus upper surface Thallus lower surface Medulla melzer’s D. luridum

Multilobed + Thallus when wet : bright green Several holdfasts I reddish brown Lobe : 6-8 mm in diameter

D. meiophyllizum

I- reddish brown Unilobed Single holdfast Lobe : 8-11 mm in diameter

Species Macroscopic characteristic Thallus upper surface Thallus lower surface Medulla melzer’s D. leptophyllodes

Multilobed Single holdfast Lobe : 1-4 mm in diameter - I reddish brown D. miniatum var. miniatum

Unilobed Single holdfast Lobe : 20-23 mm in diameter I- reddish brown

Microscopic characteristic Species Thallus transversal cut Thallus transversal cut with Melzer Ascospore

D. luridum

15 × 5-7.5µm

thick: 350-450 µm Perithecium Medulla I+ perithecia: 100-380 × 90-310 µm

D. meiophyllizum

Medulla I- 15.7 × 7-7.5µm thick: 420-450 µm perithecia: 100-300 × 100-350 µm

Microscopic characteristic Species Thallus transversal cut Thallus transversal cut with Melzer Ascospore

D. leptophyllodes

thick: 350-450 µm Medulla I- 15 × 5-7.5µm perithecia: 150-220 × 150-200 µm

D. miniatum var. miniatum

thick: 550-600 µm Medulla I- 12.5-15 × 5.5-7.5µm perithecia: 200-450 × 220-650 µm

1.1.2 Extraction and purification of mycosporine-like compounds

� Extraction of mycosporine-like compounds

Polar solvents including water [3], [75], [76], ethanol [77], methanol [78], [22], water-

methanol [76], [78], [79], [80], [81], acetonitrile [82], [83], and aq. phosphate buffer (20 mM, pH 7.0) [84] have been successfully used to extract mycosporine-like compounds from algae, cyanobacteria, fungi and cyanolichens. Recently, the HPLC quantification of mycosporine serinol from the lichen Lichina pygmea was performed following different extraction conditions, such as [3]

- Solvent: various proportion of water and methanol

- Extraction temperature (4 °C or 40 °C)

- With or without mechanical crushing of the raw material

The mycosporine yield was about 2–3% of the crude extract (obtained from 50 mg of raw material with 1 mL of solvent, 3 h). For this lichen, water was found to be the best extraction solvent and, somewhat surprisingly, the greatest quantity of mycosporines extracted (0.26 mg) was achieved with pure water at +4 °C, without crushing the lichen material, during 3 h. This optimized method was applied to extract mycosporines in the four Dermatocarpon species from 100 mg of each lichen sample.

� Purification of mycosporine-like compounds

Different purification methods were previously tested to purify mycosporine serinol in two cyanolichens Lichina pygmaea and Peltigera horizontalis, such as precipitation of sugars, gel permeation and ion exchange chromatography. Among them, cation exchange resin Dowex revealed to be the most efficient technique to separate mycosporine from crude aq. extract [3]. Dowex is polysterene resin, manufactured by the reaction of styrene with divinylbenzene. Variable amounts of divinylbenzene result in a range of polymers with different degrees of crosslinking indicated by the terms X-2, X-4 and X-8. Dowex 50W-X8 with the functional

groups –SO3H was a suitable stationary phase for concentrating mycosporines. By this technique, mycosporines appear to be retained on the column much stronger than polyols and non-cationic compounds, which eluted in the first fraction with water eluent. Elution of bound

solute from the resin may be achieved by adjustment of ion strength. The use of a high salt concentration causes displacement of solute by a shift in the equilibrium in favor of the bound counterion. In this study, to recover the retained compounds in a time- and solvent-efficient manner, saturated sodium chloride solution (200 mg/mL) was used, followed by water as eluent. Consequently, all the mycosporines were eluted in the second fraction. To remove the excess of NaCl in the purified fraction, the dry residue was dissolved in EtOH and centrifuged. The supernatant was collected and evaporated to give mycosoprine-enriched fraction. Then water was added to afford semi-purified aq. extract which was ready for HPTLC-UV and HPLC-DAD-MS analysis.

In our study, we purified the four crude lichen aq. extracts by Dowex 50W-X8 (H+ form) resin chromatography. The purification process was summarized in Figure 16.

Figure 16: Purification process of lichen crude aq. extract by cation exchange resin Dowex chromatography

1.1.3 HPTLC-UV spectrophotodensitometry and HPLC-DAD-MS analysis

Mycosporine-like compounds are usually identified based on their retention time in the HPLC spectra and their characteristic UV absorption spectra obtained via diode array detection (DAD). Although DAD allows the fast acquisition of UV visible absorption spectra, the lack of fine spectral absorption, the influence of pH and for some specific mycosporine-like

compounds, the wavelength absorption maxima differs only a few nm apart, makes it very difficult to distinguish such compounds based on absorption spectra only. Therefore, these techniques are not informative enough and may lead inadequate conclusions. For more accurate characterization of these compounds in lichens, the use of mass data was suggested by some

recent studies [45], [46], [85], [86], [87], [88].

In this study, we applied HPTLC-UV spectrophotodensitometry and HPLC–DAD–MS for screening mycosporine-like compounds in crude aq. extracts and semi-purified aq. extracts of the four Dermatocarpon species (total 8 extracts).

As usual, HPTLC-UV spectrophotodensitometry can allow a rapid and simultaneous comparative check of UV-absorbing compounds in 12–20 extracts within 2 h. Mycosporine- like compounds were detected through their symmetrical UV profile centered on a single UV

absorption maximum wavelength (λmax) between 310 and 360 nm. However, this method only provide a basic information on the presence of such constituents in a purified lichen extract and suitable for rapid selection of lichen extracts containing mycosporine-like compounds.

HPLC-DAD-MS provided a more accurate chromatographic separation and allowed a

better characterization on the basis of the retention time, λmax and MS fragmentation data. Indeed, the application of MS/MS methods represents a powerful tool, which is useful for structural elucidation using characteristic fragmentation patterns.

1.1.4 Esterification conditions of mycosporine glutamicol

In fact, in our study, an ethyl ester of mycosporine glutamicol was detected due to a possible reaction of mycosporine glutamicol and ethanol used to remove NaCl. In order to have a better understanding of reaction conditions, some parameters such as the temperature and reaction time for esterification of mycosporine glutamicol with ethanol were investigated. We tested four conditions:

- Ethanol: 4 °C, reaction time: 50 min

- Ethanol: rt, reaction time: 50 min (as given above protocol)

- Ethanol: rt, reaction time: 1h50 min

- Ethanol: rt, reaction time: 2h50 min

Each condition was performed in duplicate and the extracts were investigated by HPTLC- UV spectrophotodensitometry.

1.2 Results and discussion

1.2.1 Extraction and purification of mycosporine-like compounds

After maceration of 100 mg of each lichen with 2 mL of water at 4 °C, we obtain four crude aq. extracts of D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum i.e. 3.20, 2.85, 1.90 and 2.20 mg, respectively. These crude extracts are then purified by cation exchange resin Dowex 50W-X8 chromatography to give four semi-purified extracts (0.35, 0.30, 0.10, 0.40 mg, respectively). The TLC of D. miniatum extracts show that purification on Dowex gave a good result since most of sugars and non-cationic compounds are eluted in the first fraction (fraction B) while mycosporines remain on column to be eluted with NaCl addition (fraction C) (Figure 17).

Figure 17: TLC profile of D. miniatum extracts before and after using NaCl in Dowex purification chromatography

The percentage of semi-purified extracts which contain mycosporine-like compounds range from 5% (in D. leptophyllodes) to 18% (in D. miniatum) of crude extracts indicating the high content of sugars and non-cationic compounds in the four species [64], [67] (Figure 18). The eight aq. extracts (4 crude aq. extracts + 4 semi-purified aq. extracts, C = 1 mg/mL) are

then investigated for mycosporine-like compounds by HPTLC-UV spectrophotodensitometry and HPLC-DAD-MS.

4,00

3,50

3,00 0,35 0,30 2,50

2,00 0,40 mg 0,10 1,50 2,85 2,55 1,00 1,80 1,80

0,50

0,00 D.luridum D.meiophyllizum D.leptophyllodes D.miniatum

Residue Semi-purified extract

Figure 18: Comparative amounts (dry residue weight) of aq. extracts obtained from 100 mg lichen material

1.2.2 HPTLC-UV spectrophotodensitometry

Mycosporine-like compounds are detected by their symmetrical UV profile centered on a single UV absorption maximum wavelength between 310 and 360 nm. HPTLC-UV allows a rapid and simultaneous comparative check of the eight aq. extracts for UV-absorbing compounds within 40 min. The HPTLC-UV analysis shows the possible occurrence of two mycosporine-like compounds (λmax# 310 nm) in aq. extracts of the four species with Rf at 0.23

and 0.30 (Figure 19). The compound with Rf = 0.30 shows a higher signal intensity in crude extracts than in semi-purified extracts, except for D. miniatum. Conversely the compound with

Rf at 0.23 exhibits a significant increased signal in semi-purified extracts suggesting a transformation between these two mycosporines during purification process. We also observe

another mycosporine-like compound (Rf = 0.77) but only in the semi-purified aq. extract of D.

luridum. Interestingly, with λmax around 310 nm, most of mycosporine-like compounds possesse exclusively the aminocyclohexenone ring system [33].

Figure 19: The obtained UV absorption profiles of each extract screened by spectrophotodensitometry (C = 1 mg/mL) 1, 3, 5, and 7: crude aq. extracts; 2, 4, 6, and 8: semi-purified aq. extracts of D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum, respectively

However, this method can only provide basic information on the presence of such constituents in extracts due to the lack of mycosporine-like compound standards. Indeed, the preparation of standards from natural samples is costly and time consuming. On the other hand, some specific MAAs have a few difference in λmax so it makes very difficult to distinguish

MAAs based on absorption spectra and Rf. Since the full diversity of this compound class is yet unknown, Rf and wavelength maximum are not enough informative for the accurate identification of various mycosporine-like compounds.

1.2.3 HPLC-DAD-MS analysis

For a better access to the identification of mycosporine-like compounds, aq. extracts are then analysed by HPLC-DAD-MS using a reverse-phase HPLC column. This method provides a better chromatographic separation than HPTLC-UV and allows a better characterization on the basis of retention time, λmax and MS data. The molecular fragmentation pattern produced when a compound is bombarded by an ion source is key to identify of mycosporine-like compounds in the absence of commercial standards [89]. Then, MS data can be utilized to predict the structure of compound without complete resolution or with a very small amount of the crude extract.

The results of HPLC–DAD–MS2 in the ESI+ mode, using ion trap for mass detection confirm the presence of three mycosporines (Figure 20). Under the experimental conditions, + + compound 1 (Rt = 2.95 min, m/z [M+H] 303) and compound 2 (Rt = 4.73 min, m/z [M+H] 304) would be mycosporine glutaminol and mycosporine glutamicol, respectively, by comparison with MS data fragmentation of known mycosporines [46] (Appendix 2). These mycosporines were isolated for the first time from two terrestrial Ascomycetous fungi Trichothecium roseum Pers. (Link.) (Hypocreomycetidae, Hypocreales, Incertae sedis) and Gnomonia leptostyla (Sordariomycetideae, Diaporthales, Gnomoniaceae) respectively [90], [91]. These Sordariomycetes fungi produce mycosporines in the presence of light during reproduction and accumulate them mainly in spores. Pittet et al demonstrated that by HCl treatment during the cationic exchange chromatography, compound 1 is readily hydrolyzed into compound 2 which is more stable [90] and up to now, in the laboratory (not in natura) the amide type is more widespread mycosporines from fungal cultures [92]. These evidences partially explain the increase in signal intensity of compound 2 in semi-purified extracts as analysed by spetrophotodensitometry (Figure 19).

In the semi-purified extracts of D. luridum and miniatum, we also observe another + mycosporine at Rt = 15.75 min with m/z [M+H] 332 (compound 3). Such compound is not corresponding to any previously described mycosporine and could be an ethyl ester derivative of mycosporine glutamicol since it shows 28 mass units more than that of mycosporine glutamicol. Indeed, the fragmentation patterns of compounds 1, 2 and 3 are very similar (Figure 21). MS2 spectra of the three mycosporines reveal the presence of two product ions corresponding to a loss of one water molecule and a loss of 68 amu, although the mechanism generating this loss is not still understood. A neutral loss of 46 amu is observed as well in compound 2 and 3 (at m/z 258 and 286, respectively).

Compound 3 is supposed to be formed by esterification between mycosporine glutamicol and ethanol with HCl traces used to remove NaCl after the semi-purification step. Indeed, many MAAs were converted to their methyl esters on the treatment with HCl-methanol [93]. The comparison of the crude and the semi-purified aq. extracts using HPLC–DAD–MS2 analysis shows that compound 3 is a byproduct, only found in the semi-purified extract (Figure 20). The same phenomenon has also been observed in the previous screening of cyanolichens using the same purification process [3]. Indeed, four mycosporines have been detected possessing an

additional ethyl group but at that time there was no explanation provided for their possible origin as artefact derivatives.

Table 4 summarized physico-chemical properties and mass fragmentation of mycosporines 1, 2 and 3.

1.2.4 Esterification conditions of mycosporine glutamicol

As presented in Figure 22, no ethyl ester product occurs when using ethanol at 4 °C (track

5). In contrast, such product (Rf = 0.71) is well observed when using ethanol at rt (track 6, 7 and 8). Its intensity is time reaction dependent. The higher intensity it shows, the more time contact is between mycosporine glutamicol and ethanol. The results again confirm ethyl ester mycosporine glutamicol as an artefact formed by reaction of mycosporine glutamicol and ethanol (with HCl traces). It is noticed that ethyl ester derivative of mycosporine containing carboxylic group could occur in this purification protocol. To avoid such a by-product formation, we recommend using cool ethanol (at 4°C) with short contact time between mycosporines and ethanol solution.

Figure 20: (a) and (b): PDA chromatogram of crude aq. extract and semi-purified aq. extract of D. miniatum; (c), (d), (e): UV spectra of mycosporines 1, 2, and 3; (f), (g), and (h): MS spectra of mycosporines 1, 2, and 3, respectively

Figure 21: The proposed fragmentation patterns of mycosporines 1, 2 and 3

Table 4: Physico-chemical properties and mass fragmentation of 1, 2 and 3

Structure λmax Rt m/z ESI-MS Lichens + (nm) (min) [M+H] fragmentation pattern

310 2.95 303 235 [M+H-68]+, D. luridum D. meiophyllizum 267 [M+H-2H O]+, 2 D. leptophyllodes + 285 [M+H-H2O] D. miniatum

310 4.73 304 236 [M+H-68]+, D. luridum D. meiophyllizum 258 [M+H-H O-CO]+ 2 D. leptophyllodes + 268 [M+H-2H2O] , D. miniatum + 286 [M+H-H2O]

+ 310 15.75 332 218 [M+H-C2H6O-68] , D. luridum D. miniatum 264 [M+H-68]+, + 286 [M+H-H2O-CO] + 314 [M+H-H2O]

2 Rt: retention time on HPLC-DAD-MS using a Zorbax Eclipse XDB-C18 (150 mm × 2.1 mm) column

Figure 22: UV absorption profiles obtained in different conditions investigating esterification by spectrophotodensitometry (C = 1 mg/mL) 1: mycosporine glutamicol; 2: mycosporine glutaminol; 3: ethyl ester of mycosporine glutamicol; 4: crude aq. extract of D. miniatum; 5: EtOH 4°C, 50 min; 6: EtOH rt, 50 min; 7: EtOH rt, 1h50 min; 8: EtOH rt, 2h50 min.

As lichens are symbiotic organisms, involvement of the respective partners in the biogenesis of mycosporines is to be elucidated. To now, cyanobacteria were supposed to produce mycosporine-like compounds in cyanolichens [3]. Thus so far, mycosporines 1 and 2 and their glycosylated derivates had been described in various terrestrial Ascomycetous fungi

[38], [90], [91], [94]. In addition, compound 1 has also been detected in the terrestrial cyanobacteria Leptolyngbya sp. [46] and compound 2 in the cyanolichen Degelia plumbea [3]. As no green algae are described to date to synthesize these mycosporines, the fungal partner is priorly considered to be involved in the biosynthesis of mycosporines 1 and 2. This hypothesis is reinforced since these mycosporines described in lichens have the characteristic carbonyl moiety of fungal mycosporines. In this work, only non-glycosylated mycosporines 1 and 2 have been isolated in Dermatocarpon species. However, it would be interesting in further studies to also screen lichen species for their glycolysed derivatives because the glycosylation of these compounds is known to protect the aglycones against hydrolases and decrease their chemical reactivity [94].

In summary, two mycosporine like compounds 1 and 2 have been detected for the first time in the four chorolichen species Dermatocarpon luridum, meiophyllizum, leptophyllodes and miniatum. Mycosporine glutaminol 1 and mycosporine glutamicol 2 are genuine compounds while the ethyl ester of mycosporine glutamicol 3 is a by-product formed during the purification process. In order to confirm their structures, isolation and structure elucidation of these three compounds are initiated and presented in chapter 3.

2 Quantification of mycosporines

Looking for a large panel of samples to be informative, we considered 7 Dermatocarpon taxa corresponding to a variety of geographical and habitat issues, taking opportunity of available Dermatocapon samples in herbaria collections. We also extended the study to 4 others taxa belonging to the Verrucariales order: 3 Placidium taxa and 1 Catapyrenium taxum. This study had therefore to consider also the stability of the analysed compounds according to some long term keeping conditions.

2.1 Materials and methods

2.1.1 Lichen materials

Quantification of mycosporines was performed on a total of 94 samples including 11 taxa: Dermatocarpon arnoldianum, D.intestiniforme, D. leptophyllodes, D. luridum, D. meiophyllizum, D. miniatum var. miniatum, D. rivulorum, Placidium lachneum, Placidium rufescens, Placidium sp and Catapyrenium cinereum, in which 49 samples come from Natural History Museum (London, UK), 26 samples from H. des Abbayes herbarium (Rennes, France), 4 samples from L.J.-C. Massé herbarium (Rennes, France), and 15 fresh samples in France (Figure 23). The detailed location, collector, date collection, habitat and reference numbers are in Appendix 3.

Figure 23: View of lichens samples analysed from H. des Abbayes herbarium (a) and Natural History Museum (London, UK) (b) for mycosporines quantification

2.1.2 Extraction of mycosporines

The limited quantities of thallus amount available from herbarium samples is a great challenge to deal with as the material is destroyed for the mycosporine extraction. Indeed, 100 mg of material as the above extraction protocol is excessive for historic collections. It prompted us to determine an efficient minimum quantity of thallus in which mycosporines can be detected in aq. extract. By this way, six small pieces of dry D. luridum thallus corresponding to 5, 10, 15, 20, 25 and 30 mg amounts were extracted with 300 µL of water at 4 °C for 2 h, in duplicate. The HPLC-DAD analysis showed that mycosporines were well detected from at least 15 mg of thallus. So, variable quantities (15–50 mg) from 94 samples were extracted with 300 µL of water to give 94 crude aq. extracts. These extracts were dissolved in water, mixed with an internal standard (IS) to give a final concentration at 1 mg/mL and further analysed by HPLC–DAD for mycosporines contents.

2.1.3 Quantification of mycosporines

So far, the common quantification method used was HPLC–DAD with the reverse phase C-18 column. For this purpose, a series of known weights of pure MAAs standards were injected and the resulted chromatographic peak areas were related to injected masses to yield a response factor for each MAA. The masses injected of each compound were determined by UV spectroscopy using their specific extinction coefficients (ε) and the dilution factor. However, the extinction coefficients of some MAAs were unknown and in this case, the ε of another MAA, which had the closest match, was used to estimate concentration [93]. In the literature, Whitehead and Hedges published an approach by HPLC-MS analysis to better access to the concentration of such metabolites [85]. Here, the reverse-phase C18 was replaced by bare silica particles, a highly polar stationary phase in HILIC mode. When a polar phase is utilized with a mobile phase high in organic concentration, the more polar water will preferentially adsorb on the surface creating a semi-stagnant, water-rich stationary phase and a water depleted mobile phase. Polar analytes can then partition in to aqueous enriched phase (Figure 24).

In the present study, bare silica HILIC-HPLC-DAD using an internal calibration curve with cytosine as an internal standard was employed to separate and to quantify the two mycosporines in the 94 investigated aq. extracts. Mycosporine glutaminol and mycosporine glutamicol used as mycosporines standards were previously isolated from D. luridum (see chapter 3). The method was validated with respect to linearity, accuracy, limit of quantification (LOQ) and limit of detection (LOD).

Figure 24: Biphasic solvent distribution at silica surface in HILIC mode

2.2 Results and discussion

Various columns have been previously utilised for MAAs detection and quantification as normal phase [95], reverse phases C-8, C-18 [45], [89], [96], [97], [98], [99], [100], [101],

[102], [103], anion exchange NH2 [82] and amino-modified silica gel [104]. This study presents for the first time the use of HILIC conditions to quantify mycosporines. HILIC was introduced by Alpert in 1990 for separation of very polar and hydrophilic compounds [105]. This technique allows analytes to interact with hydrophilic stationary phases such as bare silica particles, amine-, hydroxy-, amide-bounded or zwitterionic (ZIC-HILIC) particles by applying a polar mobile phase with a high content of organic solvent. Indeed, two mycosporines 1 and 2 were well separated on HILIC column with retention times of 5.04 min and 5.67 min, respectively (Figure 25). In comparison to the run time of HPLC-DAD-MS (50 min), the HILIC-HPLC-DAD showed a shorter run time (total 16 min) and appeared be a better choice to study of numerous samples.

Figure 25: HILIC-HPLC chromatogram of D. miniatum (UV detector at 310 nm) and inset absorbance spectra of compounds 1 and 2

Cytosine was chosen as IS in this analysis because it was not co-eluted (Rt 2.50 min) with mycosporines as well as it was soluble in water and having an absorption maximum at 254 nm.

The quantification method was validated with respect to the specificity, linearity, accuracy, limit of quantification (LOQ) and limit of detection (LOD).

Linearity was evaluated by determining six different concentrations of mycosporines standards solution in triplicate. The peak area and concentration were subjected to the regression analysis to calculate the calibration equation and correlation coefficients. The slope

and intercept were investigated with a Student t-test. A visual inspection of the regression line and residuals plot show that the method is linear (Figure 26). The linear range was validated for the entire domain (5−40 µg/mL). Moreover, the linearity was also confirmed by verification of homoscedasticity of variance. The Levene test was applied to test the hypothesis whether the residual variance of a variable, in a regression model, is constant (homoscedasticity condition) [106]. The result showed that the residual variability across all concentration levels was not significantly different (p>0.05), indicating the homoscedasticity for both mycosporines 1 and 2.

The LODs and LOQs were determined as 1.41 µg/mL and 2.93 µg/mL for compound 1, 4.49 µg/mL and 8.99 µg/mL for compound 2. The linearity data was showed in Table 5. Repeatability was evaluated as 4.21% and 2.27% for compound 1 and compound 2, respectively.

Figure 26: Calibration curves of mycosporine 1 (a) and mycosporine 2 (b) Table 5: Linearity validation results of mycosporines 1 and 2

Mycosporine glutaminol 1 Mycosporine glutamicol 2

Concentration range (µg/mL) 5.00 – 40.00 5.00 – 40.00 Slope 0.04 0.03 Intercept -0.03 0.09 Correlation coefficient (R2) 0.9984 0.9922 LOD (µg/mL) 1.41 4.49 LOQ (µg/mL) 2.93 8.99 LOD: limit of detection; LOQ: limit of quantitation

List of Dermatocarpon species with their voucher barcode, collector, collection date, habitat and corresponding mycosporines contents were showed in Appendix 3.

The results showed that mycosporine glutamicol was found to be a widespread mycosporine in Dermatocarpon genus as characterised in 89 investigated samples with the concentration from 4.56 ± 0.84 to 32.85 ± 5.41 mg/g of dried crude aq. extract, corresponding to 0.14−1.00 mg mycosporines per gr of dried lichen (with yield extraction is about 3%). Interestingly, there was no evidence for a significant degradation of mycosporine glutamicol concentration over time since this compound can be clearly detected within specimens stored for more than 150 years (Nº 6, 8, 23, and 44).

Contrasting to mycosporine glutamicol, mycosporine glutaminol was less common as it was only found in 18 samples. Its content ranged from 2.01 ± 1.06 to 23.58 ± 1.38 mg/g of dried crude aq. extract, corresponding to 0.06−0.70 mg mycosporines per gr of dried lichen. Mycosporine glutaminol was noticeably found in rather recent collections (mostly post 2005) while not detected in all pre 1900 collections. These results confirmed an instability state of mycosporine glutaminol suggesting a long time storage condition to be responsible of its degradation or transformation to mycosporine glutamicol.

For samples from Natural History Museum (Nº 1−49), there is no clear pattern related to the environmental conditions of the sampling sites as elevation or proximity to the seashore and mycosporine glutamicol content. Duplicate collections from the same sampling sites, kept under identical storage conditions were found to have near identical mycosporine contents even after decades of storage. For example, there was no significant difference in mycospsorine glutamicol contents between duplicate samples: N° 6 and N° 44, N° 20 and N° 49. In fact, sample N° 6 was duplicate to N° 44 but this specimen was transfered to the London Natural History Museum (NHM) after being stored for several decades at Kew Gardens. Similarly, N° 49 was duplicate to N° 20, but it came directly to NHM, N° 20 instead was kept in the herbarium Holl for several decades before transferred to NHM. Mycosporine glutamicol content appears however to be influenced by storage conditions in different herbaria but is relatively stabilised under good storage conditions. This limits the comparability of results obtained from different herbaria but allows screening studies within a single herbarium if the age of the samples does not differ too much. Samples of D. arnoldianum collected at the same geographic area has a lower content of mycosporines comparing to D. rivulorum (N° 46 and N° 47). However levels

between collections from different sites show larger differences than those between the two taxa. This indicates clearly that mycosporine levels in Dermatocarpon corresponds to an unknown environmental factor (possibly UV radiation). A possible explanation is that at the micros-site UV-levels did indeed differ although the samples are from the same area. D. rivulorum is a species that is often found in fully exposed micro-sites while D. arnoldianum tends to prefer slightly shaded micro-sites.

For samples from H. des Abbayes and L.J.-C. Massé herbaria (Nº 61−90), the mycosporine glutamicol content was ranged from 5.09 ± 0.15 (Nº 71) to 26.40 ± 5.58 (Nº 86) mg/g of dried crude aq. extract, corresponding to 0.15−0.78 mg mycosporines per gr of dried lichen whereas mycosporine glutaminol was not detected. Our work contributed to improve

the respective herbarium data with new information about unexpected mycosporine indicated with a label on each envelop containing the genuine lichen material (Figure 27). These data could be useful for a possible survey of mycosporines over period of time in these samples and could be useful for chemotaxonomy.

Figure 27: Information of Dermatocarpon species in H. des Abbayes herbarium before and after analysis

For 15 fresh samples collected from 2005 to 2014 (Nº 50−60, 91−94), most of them contained both mycosporines, except three samples Nº 51−53. Quantification of mycosporines in the four samples (Nº 90−94) which were screened mycosporines above showed that D. miniatum (Nº 94) possessed a higher level of such metabolites (46 mg/g of dried crude aq. extract). While the others contained 20−30 mg of mycosporines per gr of dried crude aq. extract. This may be related to its different habitat condition. Indeed, we collected D. miniatum

on dry rock surfaces where they were exposed to higher levels of UV radiation whereas the three others were collected on submerged rocks (Figure 15).

3 Experimental

3.1 Identification of mycosporines and MAAs

3.1.1 Extraction and purification of mycosporines and MAAs

Unpowdered dry material of each lichen (100 mg) was placed in a test tube. Water (2 mL) was added and the tube was stored at 4 °C for 4 h. Each supernatant recovered by filtration through a cotton piece was evaporated to dryness in SpeedVac. Each crude aq. extract was purified on cation exchange resin DOWEX. Cation exchange resin DOWEX 50W-X8 (4×2 g) was washed thoroughly with MeOH then with water, and degassed before being poured into a 10×50 mm column. The resin was acidified by eluting with five volumes of 1 M HCl solution and then was washed thoroughly with ultra-pure water until the effluent had a pH higher than 5. The dried aq. lichen extract was dissolved in water (200 µl) and applied on top of the column. Water (3 mL) was eluted. Then an aq. solution of NaCl (200 µl, 200mg/mL) was added on the top of the column and a fraction of 3 mL was collected. This fraction was then evaporated to dryness in SpeedVac at 35°C. To remove the excess of NaCl out of the purified fraction, the dried residue was dissolved in EtOH (500 µl) and centrifuged (1500 g for 5 min). The supernatant was evaporated to dryness in SpeedVac at 35 °C to give semi-purified extract.

3.1.2 Analysis

HPTLC-UV analysis: each crude extract and semi-purified extract was dissolved in water at the concentration 1 mg/mL. Each solution (10 µL) was applied on a plate of silica gel 60 ® F254, Merck by automatic TLC sampler ATS 4 CAMAG . TLC-plate was chromatographed with a mixture CHCl3/MeOH/H2O 6:4:1. The plate was then submitted to a densitometric evaluation by a TLC scanner 3 CAMAG® at 250, 290, 310, 320, 330 and 360 nm, by the measurement of reflection in the absorbance mode.

HPLC-DAD-MS analysis: Each solution of crude extract and semi-purified extract at concentration 1 mg/mL was subjected to HPLC-DAD-MS2 with a Zorbax Eclipse XDB-C18 column (150 mm × 2.1 mm, Agilent Technologies). The mobile phase consisted of (H2O +

0.1% HCOOH) as solvent A and (ACN + 0.1% HCOOH) as solvent B with gradient: 99% of A and 1% of B 0 - 7 min, 1–30% of B during 15 min, 30–50% of B during 6 min, 50–100% of B during 2 min, 100% of B during 3 min, 0–99% of A during 2 min and a 15 min raw of reequilibration. The flow rate was 200 µL/min and 4 µL of each sample was injected. DAD data were recorded at 310 and 330 nm and absorption spectra (210–400 nm) were recorded each second. MS data were obtained in the ESI positive mode using the following parameters: capillary temperature 240 °C; sheath gas 67 arb. U; auxiliary gas 5 arb. U; ion spray voltage 4.5 kV; capillary voltage 45 V. The Xcalibur 1.0 software was used for the data evaluation.

3.2 Quantification of mycosporines

3.2.1 Extraction of mycosporines

Aq. extracts were performed as described in section 4.1.1 with some modifications due to the minute thallus amounts in herbarium. Briefly, different small amounts of materials (15−50 mg) were extracted with 300 µL of water at 4 °C for 2 h. Each supernatant was filtered (0.45 µm) and lyophilized to dryness in Eppendorf tubes. The crude aq. extracts were dissolved in water and IS to give a final concentration at 1 mg/mL.

3.2.2 Mycosporines standards

Mycosporine glutaminol and mycosporine glutamicol were isolated from D. luridum. The standards were dissolved in water to give 1 mg/mL of stock solutions. A wide range of calibration standards (5 – 40 µg/mL) were obtained by suitable dilution in water. Quantification of mycosporines was done using a calibration curve determined by injecting 10 µL of mycosporine standards at known concentrations.

3.2.3 Internal standard

The stock solution (250 µg/mL) was prepared by dissolving 25 mg of cytosine in 100 ml of water and then a suitable dilution was carried out to obtain the concentration of cytosine in each sample corresponding to 25 µg/mL. Thus, the use of the same concentration of cytosine as an IS in each sample will provide a better quantification and limit errors caused by variations from the instruments.

3.2.4 Validation of analytical procedure

The accuracy was investigated by performing six consecutive replicate injections of the same standard solution and peak area measured was expressed in terms of the percentage of relative standard deviation. LOD and LOQ were calculated following the formulas b + 3×σ(b) LOD = (Eq. 1) a b + 10×σ(b) LOQ = (Eq. 2) a

a: slope, b: intercept of calibration curve, σ: standard deviation Statistical analysis were computed by using MS Excel (Microsoft 2013) Levene’s test was applied on calibration data for evaluation of homoscedasticity of variance.

3.2.5 HILIC-HPLC-DAD analysis

Each crude extract and mycosporines standards (10 µL) were injected to HPLC (Shimadzu LC20A) using a Kinetex HILIC 100 Å (2.6 µm, 100×4.60 mm) column, with mobile phase A (ACN: CH3COONH4 50 mM 90:10, pH 5.36) and mobile phase B (ACN: H2O:

CH3COONH4 50 mM 50:40:10, pH 5.36). The elution followed this gradient: 100% of A between 0 and 2 min, 0–100% of B between 2 and 4 min, 100% of B between 4 and 12 min, 0–100% of A between 12 and 14 min and the column was reequilibrated with the mobile phase A for 2 min. The flow rate was 1 mL/min. Peak detection was carried out online using a diode array detector at 254, 310, 330 and 360 nm, and absorption spectra (200−400 nm) were recorded each second directly on the HPLC-separated peaks. Compounds 1 and 2 in extracts were identified by comparison of their retention time, 5.04 min for 1 and 5.67 min for 2, with those of reference standards. The calibration curve was constructed by plotting response factor against mycosporine concentration. Response factor was corresponding to the ratio of AUC at

310 nm for mycosporine and at 254 nm for IS (RtIS = 2.50 min).

CHAPTER 3: PHYTOCHEMICAL STUDY ON DERMATOCARPON LURIDUM (WITH.) J.R. LAUNDON

Purpose of this study

As presented in chapter 2, two genuine mycosporines (1 and 2) with an artifact mycosporine 3 were identified for the first time in D. luridum and D. miniatum by HPTLC-UV spectrophotodensitometry and HPLC-DAD-MS techniques. Isolation of a suitable quantity of these metabolites was necessary, not only to confirm their structures but also to determine some spectroscopic and physico-chemical properties and further to investigate their photobiological activities (see chapter 4).

Mycosporine isolation was undertaken from aq. extract of D. luridum species, an available source in Brittany (France). A low mycosporine content in this species is quite challenging for the isolation (22.6 mg/g of dried crude aq. extract) implying to collect 14 g of fresh material to obtain 10 mg of mycosporine (optimistic hypothesis for a full recovery from the crude extract). After testing several purification techniques (CPC, TLC preparative, gel filtration…), we applied a novel combination method including bare silica-HILIC flash chromatography, followed by open reverse phase chromatography, and semi-preparative HPLC-DAD to isolate three pure mycosporines 1, 2 and 3 from the aq. extract. Their spectroscopics properties (NMR, IR, MS) and their molar extinction coefficient (ε) not all yet reported were also measured. Moreover, the acidity constants (pKa) of mycosporines 1 and 3 were for the first time determined, indeed they contain one acido-basic site in their structure (RR’NH). The knowledge of the pKa value of a compound is fundamental to understand many chemical and biochemical processes and also to determine the presence under the acidic or the basic form of mycosporine in water or in a buffer. Different analytic methods can be used to determine the acidity constant such as potentiometry and spectrophotometry [107]. In this study, their UV absorption properties allowed us to use the spectrophotometry method.

Beyond mycosporines, we are of interest in other chemical constituents of D. luridum since such research has not been published so far. Acetone is recommended for lichen subtances extraction because most metabolites are soluble in this solvent [12]. Due to the very poor solubility of the lichen subtances in water, chloroform and acetone were used after the polar aqueous extraction to recover the remaining metabolites in dried thalli. After purification according to a phytochemical process, structures of isolated compounds were elucidated by spectroscopic methods (HRMS, NMR-1D & -2D, IR, UV-vis) and compared their data to those reported in the literature.

1 Description of Dermatocarpon luridum (With.) J.R. Laundon

1.1 Ecology

Dermatocarpon is a unique foliose representative of the Verrucariaceae, other members being crustose or squamulose or even fructicose [108]. Dermatocarpon genus occurs in aquatic or semi-aquatic environments, but some species are found in dry habitats. D. luridum is amphibious in the splash water zone of streams and rivers on stable siliceous rocks, often in moderate shade [109]. The term “amphibious” is used for species that are best developed in the splash water zone or at micro-sites where the desiccation lengthens for at least some weeks during the year. Amphibious lichens have very specific responses to the pH value of their habitat, sometimes not only of the surrounding water but also of the substratum and the precipitation. The underlying processes for the different ranges of pH tolerance are still poorly understood. Specific acidity levels may have different effects on the fungal or algal partner of a lichen and certain developmental stages may also have different tolerance limits. Some general patterns however are obviously related to the kind of associated photobionts: freshwater lichens with cyanobacterial photobionts are best developed in neutral or slightly alkaline waters while they are usually absent in acidic waters. The same pattern is observed for lichens with symbiose Diplosphaera chodatii (most species of Dermatocarpon) [71].

D. luridum is generally not encountered from strongly acidic water bodies (pH < 5) and possibly also sensitive to air-borne acidic emission. It appears to be much tolerant to periods of desiccation than other aquatic lichens, perhaps because it receives the high humidity during summer nights at high elevations. In France, D. luridum was found mostly in the west, northwest and sufficient moistured mountains and common in humid regions [69] (Figure 28). Figure 28: Distribution of D. luridum in France (from Claude Roux kind supply)

So far fungi, lichens, tree bark, tree rings and leaves of higher plants have been used to detect the accumulation and distribution of metal pollution. Specially, lichens show sensitivity and tolerance to high levels of pollutants which make them as ideal and reliable indicators species. Monnet et al (2005) used the lichen D. luridum as bioindicator of copper pollution [110]. The results showed that linear regression was recorded between the increase in copper concentration in the medium and the total concentration of copper in lichen thalli. The malondialdehyde (MDA) concentrations in thalli can be used as a bioindicator of the excess copper in the medium and its phytotoxicity, similar membrane degradation was observed for 0.25 and 0.50 mM copper and for 0.75 and 1.00 mM copper.

1.2 Botany and description

Basionym: Lichen luridus Dillenius ex Withering. Selected synonyms: Dermatocarpon aquaticum (Weiss) Zahlbruckner, Dermatocarpon fluviatile (Weber) Th. Fries, Dermatocarpon weberi (Acharius) W. Mann. Family: Verrucariaceae (Ascomycota) Genus: Dermatocarpon Photobiont: Diplosphaera chodatii (Trebouxiophyceae)

Figure 29: D. luridum on submerged rock (a), macroscope observation of dried thallus (b), microscope observation of thallus transverse cut stained with cotton blue (c)

Macroscope and microscope view of thallus were observed in Figure 29 and described in chapter 2.

Reproductive biology: D. luridum reproduces sexually by producing fungal spores in perithecia. It reproduces vegetatively by fragmentation of thallus lobes which could drift to new sites and reattach to rocks. Aquatic invertebrates that graze on this lichen and ingest spores could also be considered as dispersal vectors.

1.3 Chemical studies

Investigation on chemical constituents of D. luridum species has not been noticed so far as only few studies have paid attention on it. Previous studies revealed that lichen species of the genus Dermatocarpon did not contain common lichen subtances such as depsides, depsidones, dibenzofurans, xanthones, anthraquinones, etc. However, almost all these species included several sugar alcohols as mannitol, sorbitol, volemitol and arabitol [68]. In 1955, Bengt isolated three low-molecular carbohydrates from D. miniatum, including D-mannitol (0.4%), D-volemitol (4%), trehalose (1.1%), and sucrose (0.17%) [111] (

Figure 30). D-volemitol, D-arabitol and ergosterol were determined as well in D. moulinsii collected on rocks near Nandaprayag, India. This lichen contained about 20 % of crude protein (Kjeldahl estimation) and 12 % of carbohydrate. From another analysis of the mineral constituents, the lichen was found to be rich in iron (ca. 1%), calcium (0.5 %), and phosphorus [112]. The higher proportion of mineral constituents may be attributed to a close association of lichen and its habitat (rocks).

Figure 30: Polyol and steroid constituents in the Dermatocarpon genus

Qualitative and quantitative analyses of free amino acids in D. miniatum collected in Japan showed that this lichen contained 18 kinds of amino acids with the highest amount of

glycine, aspartic acid, proline, arginine, lysine, methionine, and cysteic acid and then leucine, serine, threonine, glutamic acid, histidine, and taurine [113] (Figure 31)

Figure 31: Free amino acids in the Dermatocarpon genus

In addition, the total carotenoid content was determined by visible absorption spectra (0.93 mg/g dw.) in which six carotenoids were characterised by column chromatography and TLC, namely β-carotene, β-cryptoxanthin, lutein epoxide (dominant carotenoid containing 55%), zeaxanthin, neoxanthin, and mutatochrome [65] (Figure 32). Acetone extracts of three Dermatocarpon species collected from Anatolia in Turkey (D. intestiniforme, luridum, and miniatium) were investigated by HPLC for their usnic acid contents. Results showed there were devoided from usnic acid as authors concluded that contents were under detection limit [66].

Figure 32: Carotenoids in the Dermatocarpon genus

2 Materials and methods

2.1 Lichen material

D. luridum was collected on submerged rocks in Huelgoat ‘rivière d’Argent”, Brittany, France in April 2012 by J Boustie and myself. Less 50 % of the lichen population in a given place was collected. Its morphological characteristics were observed at macroscopic and microscopic levels and guided by keys described in Appendix 1. The lichen identfication was confirmed by Dr Jean-Yves Monnat (Biologist, University of Brest, France). A voucher specimen was deposited in the herbarium of Pharmacognosy and Mycology, University of Rennes 1, France with the reference number JB/001/04/2012. After collecting, the lichen was sorted out after being cleaned of bits of soil and mosses by rapid contacting with distilled water. It was then dried at ambient temperature and crushed to give 150 g of dried powder material.

2.2 Extraction

We focused on mycosporines components so water was used first in the extraction procedure. Water is the best solvent due to a complete solubility of mycosporines in it. The residue was then extracted with chloroform and acetone to give chloroform and acetone extracts, respectively, which were considered containing apolar and intermediate polar metabolites.

2.3 Purification

2.3.1 Mycosporines

After removing sugars and polyols in the crude aq. extract by cation exchange resin Dowex 50W-X8 chromatography, the semi-purified aq. extract showed a difficult challenging mycosporines isolation due to the presence of numerous amino acids inside. Recently, a Centrifugal Partition Chromatography (CPC) was used to separate efficiently two polar compounds, namely mycosporine serinol and a glutamic acid derivative from the methanolic extract of a cyanolichen Lichina pygmaea [114]. This technique was assessed on the semi- purified extract to isolate the mycosporines and described further. However, poor results obtained in this case prompted us to develop a new purification protocol, including bare silica- HILIC flash chromatography, followed by open reverse phase chromatography, and finally

semi-preparative HPLC on C18 column to isolate three mycosporines from the semi-purified aq. extract.

� Purification on cation exchange resin Dowex

Chromatographic techniques involving gel permeation, ion exchange resins, carbon, norite A, cellulose columns, preparative TLC on silica gel, semi-preparative or preparative HPLC and UV detection [33], [93] have all been used in different combinations in the purification of aq. methanolic or ethanolic extracts of fresh or freeze dried samples containing mycosporine-like compounds. Among the techniques, purification on Dowex 50W-W8 (H+ form) was the most convenient method and gave the best results [49], [3]. In this study, we applied cation exchange resin chromatographic technique as the first step of purification process on crude aq. extract to give semi-purified aq. extract which was expected to remove sugars, polyols and to concentrate mycosporines.

� Purification by CPC

CPC is a novel form of the well-known countercurrent chromatographic technique that is based on the difference in partitioning behavior of components over two immiscible liquid phases [115]. In CPC, one of these phases is kept stationary, whereas the mobile phase flows through. To retain the stationary phase in the chromatographic column, the column has a tortuous internal geometry and is operated in a centrifugal field. A CPC column consists of channels engraved in plates of an inert polymer. The channels are connected by narrow ducts. Several plates are put together to form a cartridge. The cartridges are placed in the rotor of a centrifuge and are connected to form the chromatographic column (Figure 33). CPC can be used either in a normal mode with a polar stationary phase, but also in an inverse mode with an apolar stationary phase. This technique also allows fractionation to be carried out in a normal-phase mode followed by a reversed-phase mode, by a switching valve between descending and ascending modes, called the “dual” mode. CPC allows the use of a wide range of biphasic systems, and provides important benefits for natural compound purification, such as no sample loss on solid support and high recovery.

Due to many advantages of this method, we attempted to adapt the CPC technique previously described by Roullier et al [114] with some modifications to separate three mycosporines from the semi-purified aq. extract.

Figure 33: Kromaton Technologies FCPC 50mL apparatus

� New purification protocol

Because the CPC technique failed to fully separate the three mycosporines in the extract, we looked for a more efficient method to solve this problem. After testing several techniques, we chose a combination of purification methods in which bare silica-HILIC flash chromatography, open reverse phase chromatography, and semi-preparative HPLC-DAD using reverse phase techniques were employed.

2.3.2 Other isolated compounds

Flash chromatography, open column silica gel chromatography and preparative TLC techniques were employed to isolate three other polar compounds from the aq. extract.

The chloroform extract contained a high quantity of several pigments leading to a difficult separation of other metabolites. So far, Sephadex LH-20 has been used routinely to remove chlorophyll from non-polar or mid-polar natural product extracts [116]. LH-20 gel swells sufficiently in organic solvents and allows handling of natural products which are soluble in organic solvents. Here, Sephadex LH-20 chromatography using 100% chloroform as eluent was performed to remove pigments, dominant components in the chloroform extract.

Open column chromatography (SiO2) and recrystallization techniques were also used to isolate compounds from chloroform and acetone extracts.

2.4 Determination of physico-chemical properties of mycosporines

2.4.1 The molar extinction coefficient

The molar extinction coefficient ε at λmax of each mycosporine 1, 2 and 3 in water is determined by spectrophotometry.

2.4.2 pKa values

The pKa determination of mycosporines 1 and 3 using a spectrophotometric method could be done through measurement of absorbance of the mycosporine aq. solution at different pH values. The strategy followed is:

� the selection of the wavelength λ and the pH range endowed with the larger difference of absorbance between the base and its conjugated acid; � the absorbance recording of the mycosporine aq. solution at the selected wavelength th λ at different pH values. Solutions at C0 were prepared by diluting (1/10 ) a concentrated mycosporine aqueous solution in the appropriate buffer; � the pKa value “extraction” from the pH and absorbance using a non-linear least- squares procedure with the NLREG program [117]. This regression analysis also called “curve fitting” determines the values of parameter (here pKa and molar extinction coefficient of each form) that cause the best fit between calculated and measured absorbances. To obtain meaningful statistical parameters, six independent determinations were performed for each mycosporine at 25 °C.

In aqueous solution, the base form of mycosporine (symbolized by B or neutral form) is in equilibrium with its conjugated acid BH+ (or protonated form) as follows:

By assimilating activity and concentration (diluted solutions) and considering the laws of matter conservation and of equilibria in water (Eq. 3-5), Eq. 6 gives the mathematical model allowing the calculation of absorbances

= (Eq. 3) � �� + �� = (Eq. 4) � �=�� +�� (Eq. 5)

� � � � � �� = � � � (Eq. 6) � �� =� 10� ��#$ (Eq. 7) � � �� = 10� #$ (Eq. 8) �� + λ where Ka is the acid dissociation�� constant of the conjugate acid BH ; Ai is the measured

λ λ absorbance at pHi and at the wavelength λ; εBH+ and εB are the molar extinction coefficients

+ at the wavelength λ of each form: conjugated acid BH and base B respectively; C0 is the total concentration of mycosporine in mol/L.

From Eq. 6, the regression analysis consists in calculating for each pHi, the calculated

λ Ai with arbitrary set of values (Ka and molar absorptivities) and then to compare the calculated

λ Ai with the experimental one. The agreed values of Ka and molar extinction coefficients are those giving the best fit between calculated and measured absorbances.

3 Results and discussion

The dried lichen material (150 g) was powdered and extracted with water and then the dried residue was successively extracted with chloroform and acetone to give aqueous (7.5 g, ρ= 5.0 %), chloroform (4.5 g, ρ= 3.0 %) and acetone (3.2 g, ρ= 2.1 %) extracts respectively (Figure 34). From the aqueous extract, six compounds were isolated including three mycosporines 1, 2 and 3, a non-protein amino acid 4, L-proline 5, and a dipeptide γ-L- glutamylglycine 6. Besides, four others compounds (a sterol 7, a ceramide 8 and a mixture of D-volemitol 9 and D-mannitol 10) were also isolated from chloroform and acetone extracts. Table 6 summarizes the isolated compounds from D. luridum according to the solvent which the compounds were obtained.

Figure 34: Extraction procedure of Dermatocarpon luridum Table 6: List of metabolites isolated from D. luridum

N° Compounds Extract Classification 1 Mycosporine glutaminol aqueous mycosporine 2 Mycosporine glutamicol aqueous mycosporine 3 Ethyl ester of mycosporine glutamicol aqueous mycosporine 4 2-Amino-3-acetylaminopropionic acid aqueous amino acid 5 L-Proline aqueous amino acid 6 γ-L-Glutamylglycine aqueous amino acid 7 (22E,24R)-Ergosta-7,22-diene-3β,5α,6β-triol chloroform steroid 8 (2S,3S,4R,2’R)-2-(2’- chloroform ceramide Hydroxytetracosanoylamino)octadecan-1,3,4-triol acetone 9 and 10 A mixture of D-Volemitol and D-Mannitol acetone sugar

As mycosporine identification was the main objective, results are separately presented:

Section 1: extraction, purification and structure elucidation of mycosporines along with their physico-chemical properties determination.

Section 2: isolation and structure elucidation of other compounds.

3.1 Mycosporines

3.1.1 Extraction and purification

Mycosporines were extracted with pure water at +4 °C from 150 g of crushed dried lichen. The crude aq. extract (7.5 g) was a complex mixture of polar compounds corresponding to sugars, polyols, amino acids that prevented the isolation of mycosporines. Hence, an additional purification step using cation exchange resin Dowex chromatography was performed to remove sugars as well as polyols and to concentrate mycosporines. Firstly, non- cationic compounds were removed with 100% of water and then mycosporines were eluted with a solution of 200 mg/mL sodium chloride. After evaporating to dryness (SpeedVac 35°), salt removal was achieved by precipitation in ethanol. Finally, the ethanol solution was evaporated to give a mycosporine-enriched fraction (1.5 g).

The possible presence of three mycosporines in the semi-purified aq. extract, namely mycosporine glutaminol 1, mycosporine glutamicol 2 and ethyl ester of mycosporine glutamicol 3 was previously indicated through the HPLC-DAD-MS analysis (see chapter 2).

First of all, we tried to separate these metabolites by using CPC technique in multiple dual mode following the experimental conditions:

- Bisphasic solvent system BuOH/AcOH/H2O 4:1:5 v/v/v - Rotor speed 1500 rpm - Flow rate 4 mL/min - Multiple dual mode with four iterative elution modes (normal-reverse-normal- reverse modes)

Usually, the more is the number of compounds between the upper and the lower phase, the better is the partition. Therfore, the solvent mixture (BuOH/AcOH/H2O 4:1:5) was chosen according to the most equilibrated repartition of each mycosporine between each phase of the

biphasic solvent system. The best retention value was obtained with a rotor speed of 1500 rpm and a flow rate of 4mL/min. Under these experimental conditions, mycosporines were only eluted in descending mode (reverse-phase mode) indicating a high polarity of such compounds (Figure 35). However, most of other polar compounds such as amino acids were co-eluted in the first change from ascending mode to descending mode (tubes 11−20). Although mycosporines were partially separated from amino acids as switch again to descending mode (tubes 45−50), a low recovery yield along with a poor separation of each mycosporine limit the use of this method for isolation of the three mycosporines in the crude aq. extract of D. luridum.

In fact, mycosporines are polyfunctional compounds with several ionizable groups (carboxyl, secondary amine) therefore their solubility should vary according to the polarity and pH of the biphasic solvent system. So the pH-zone refining CPC method should be a valuable method to separate mycosporines from the extract [118], but we preferred to develop the Hilic method as CPC preliminary optimisation is tedious.

Figure 35: The TLC profile of semi-purified aq. extract of D. luridum using CPC in multiple dual mode, F2: semi-purified aq. extract of D. luridum, green circles correspond to UV recognised mycosporines, TLC was visualized by ninhydrin

As an alternative option to CPC, we proposed a novel combination purification method which efficiently separate the three mycosporines from semi-purified aq. extract (Figure 36). In this method, the aq. layer which formed on the surface of HILIC bare silica particles promoted a stronger interaction with amino acids comparing to mycosporines therefore they were eluted later. A collected fraction containing mycosporines was then purified on open reverse phase column using water as eluent. By this way, mycosporine 3, the least polar, was

isolated (6.0 mg). Then, the two mycosporines 1 and 2 (5.0 and 6.5 mg, respectively) were separated by semi-preparative HPLC on a Prevail C18 column. Such a combined method allowed isolation of the three mycosporines 1, 2 and 3 from the semi-purified aq. extract of D. luridum. This method was successfully used to purify the same mycosporines in D. miniatum species (data not shown). However, the resolution and sensitivity of this method have not yet been tested for a variety of mycosporines samples.

3.1.2 Structure elucidation

After isolation of the three mycosporines 1, 2 and 3 from the aq. extract (5.0 mg, 6.5 mg and 6.0 mg, respectively), their structures were elucidated by ESI-HRMS in positive ionisation mode, NMR in D2O, IR and UV-vis spectra. The IR spectra of mycosporines 1 and 2 display the presence of hydroxyl and amine groups (at 3141 and 3354 cm-1, respectively) as well as an absorption band corresponding to an α,β- unsaturated ketone group (1661 and 1650 cm-1, respectively). IR spectrum of compound 1 shows a peak at 1652 cm-1 corresponding to a primary amide carboxyl group whereas compound 2 shows the presence of an carboxylate ion by two bands at 1557 and 1404 cm-1 (Figure 37 a,b). IR spectrum of compound 3 exhibits a large band in the range of 3200-3650 cm-1 indicating the presence of hydroxyl and amino groups, two strong bands at 1665 and 1712 cm-1 corresponding to an α,β-unsaturated ketone group and an ester carboxyl group, respectively (Figure 37 c).

HR-MS ESI spectrum of compound 1 displays a [M+Na]+ ion peak at m/z = 325.1379, leading to the molecular formula C13H22N2O6Na. Compound 2 gives a [M+Na]+ ion peak at

m/z = 326.1198 (calcd. for C13H21NO7Na 326.1216).

1D and 2D-NMR analysis confirm that compound 2 is mycosporine glutamicol from correspondence between the first NMR spectra data we report in D2O and NMR spectra data in pyridine previously published [90], [91]. Compound 1 exactly matches the published chemical shifts to mycosporine glutamicol (data not shown) as well as the fragmentation pattern similarities indicating compound 1 as mycosprine glutaminol. The 1H and 13C-NMR analysis show that compound 3 has the mycosporine glutamicol substructure and an ethyl group

(Figure 38, Figure 39). The HMBC correlation of H-1’ (δH 4.15) to the carbonyl carbon at δC 175.89 ppm locates the ethyl group at C-13. The COSY and HMBC correlations are in full agreement with the structure of 3 (Figure 40).

Although the three mycosporines possess optical activity (at C5 and C9), their absolute configuration are not determined in this study due to the limited quantities of the purified compounds. Only two synthetic mycosporines, namely mycosporine serinol and mycosporine glycine, were determined to have the S absolute configuration at C5 [119].

Figure 36: A novel isolation protocol of mycosporines from aq. extract of D. luridum

Figure 37: IR spectra of mycosporines 1 (a), 2 (b) and 3 (c)

1 Figure 38: H NMR spectrum of 3 (D2O, 300 MHz)

13 Figure 39: C NMR spectrum of 3 (D2O, 75 MHz)

Figure 40: Key 1H-1H COSY (bold line) and HMBC (1H→13C) correlations of 3

3.1.3 Determination of physico-chemical properties

� The molar extinction coefficient

The UV spectra in Figure 41 show that the three mycosporines 1, 2 and 3 have a strong absorption at 310 nm in the region of UVB (290–320 nm) with high molar extinction coefficients (12540, 17250, and 21300 M-1.cm-1, respectively) using a 2.5 × 10-5 M concentration in water. Indeed, the value of ε ranges from 10000 to 70000 M-1.cm-1 considered for good photoprotection. The obtained results suggest the potential application of the three mycosporines as natural photoprotectants.

UV spectra (in water) 0,6 0,5 Mycosporine 1

Mycosporine 2 0,4 Mycosporine 3 0,3 Absorbance 0,2

0,1

0 200 220 240 260 280 300 320 340 360 380 400 Wavelength nm Figure 41: UV spectra of compounds 1, 2 and 3 at C= 2.5×10-5M in water

� pKa values

pKa determination of the mycosporines to identify the protonated form or the neutral form is useful to discuss the UV absorption. In a first time, we focus only on the pKa determination of mycosporine 1 and 3 which exhibit only one acido-basic site (RR’NH site). The determination of the ionisation constants of 2 will be further investigated. An another mathematical model will be considered taking into account the two ionisation sites (RR’NH and COOH sites).

A preliminary study was performed to select the wavelength λ and the pH range endowed with the largest difference of spectra due to a different ratio between the base and its conjugated acid. The UV spectra of mycosporine 1 and 3 at pH between -1 to 7 have been recorded at 25°C in the range of 200–400 nm (see UV spectra of mycosporine 3 in Figure 42). No change is observed in the range of pH from 3 to 7 for the two mycosporines, the largest difference of absorbance occurring between pH -1 to 3 at 310 nm. In these acidic media, the maximum

wavelength λmax of mycosporine shows a hypsochromic shift from 310 nm at pH 3 to 300 nm in the pH -1 solutions. We also observe a hypochromic shift with the increase of acidity (A = 0.35 at pH 3 and 0.28 at pH 0). The same behaviour has been observed for porphyra-334 containing an aminocycloheximine ring, however the hypsochromic shift is smaller (334 to 330 nm) than that of mycosporines 1 and 3 [120].

Preliminary study

0,4 0,35 pH = -1.0 pH = 0 0,3 pH = 1.1 0,25 pH = 2.2

0,2 pH = 3.0 pH = 4.5 0,15 pH = 5.2 absorbance A 0,1 pH = 6.3

0,05 pH = 7.1 0 220 240 260 280 300 320 340 360 380 400 wavelength λ (nm) Figure 42: UV spectra of mycosporine 3 at pH range from -1 to 7 (C= 1.56×10-5M)

Then, the study is performed for 13 and 12 working pH values in the restricted range - 1.0 ≤ pH ≤ 3.8, respectively for mycosporines 1 and 3 (Figure 43, Figure 44).

UV spectra of mycosporine 1 1 pH = -1 0,9 pH = 0 0,8 pH = 0.5 0,7 pH = 1 pH = 1.2 0,6 pH = 1.35 0,5 pH = 1.71 0,4 pH = 2.09 Absorbance A Absorbance pH = 2.53 0,3 pH = 2.79 0,2 pH = 3.02 0,1 pH = 3.32 0 pH = 3.84 250 260 270 280 290 300 310 320 330 340 350 Wavelenght λ (nm)

-5 Figure 43: UV spectra of mycosporine 1 at pH range from -1 to 3.84 (C0= 6.82×10 M)

0,4 UV spectra of mycosporine 3 pH = -1 0,35 pH = 0.14

A pH = 0.56 0,3 pH = 0.99 0,25 pH = 1.29 0,2 pH = 1.66 pH = 1.86 Absorbance Absorbance 0,15 pH = 2.14 0,1 pH = 2.48 pH = 2.70 0,05 pH = 3.09

0 pH = 3.62 250 260 270 280 290 300 310 320 330 340 350 Wavelenght λ (nm)

-5 Figure 44: UV spectra of mycosporine 3 at pH range from -1 to 3.62 (C0= 1.56×10 M)

310 From the set of absorbance Ai for each pHi, the regression analysis determines the calculated values of pKa and molar extinction coefficient of each form (Table 7). The molar extinction coefficient of each form is systematically considered as unknown. This strategy allows a comparison of molar extinction coefficients determined throught the fitting process

310 with the experimental ones (here only εBexp is available). Results show a good consistence

310 between experimental and calculated molar extinction coefficients εB , so it provides an indirect proof of accuracy of the experiments and of the calculations. Moreover, we observe

the presence of an isosbestic point at approximately 297 nm which provides evidence of only one equilibrium in solution.

Table 7: pKa and molar extinction coefficient values of mycosporines 1 and 3

a 310 -1 -1 310 -1 -1 310 -1 -1 pKa εBH+ (M .cm ) εB (M .cm ) εBexp (M .cm )

Mycosporine 1 1.29 8200 12640 12540 [1.28-1.31] [8187-8213] [12593-12691]

Mycosporine 3 1.13 15210 21670 21300 [1.10-1.17] [15140-15280] [21520-21830]

pKaa determined in aqueous buffer at 25°C

The two mycosporines appear to be rather weak bases or strong acids with a pKa values very low and quite similar to 1.29 and 1.13 for mycosporine 1 and 3 respectively. The conjugated acid (BH+ form) of mycosporines 1 and 3 exhibit pKa values much weaker than the + N-Methyl-anilinium ion (CH3-NH2 -C6H5, pKa = 4.85) [121]. We can explain these results by the occurrence of both withdrawing inductive and a strong resonance effect of the cyclohexenone group which let to the global electron density decrease (Figure 45). The protonation of the unbounded lone pair electrons of the nitrogen atom in the BH+ form would prevent the resonance delocalization of the molecule. So the degree of resonance delocalization is higher in the base form B than in the acid conjudated form BH+ [120] and could explain the hypsochromic shift in strong acidic solutions and the decrease of the molar extinction coefficients between B and BH+ forms.

Figure 45: Resonance delocalization of the base form B

The diagram of the species distribution (Figure 46) allows to conclude that only the neutral form B is present and is responsible of the UV absorption in water at pH 7.

Distribution diagramm of mycosporines 1 or 3 100 90 80 70 60 acid form BH+ 50 40 base form B 30 20

% of predominance 10 0 012345678 pH

Figure 46: Distribution diagramm of mycosporines 1 or 3

3.2 Other isolated compounds

3.2.1 Isolation

Continuing phytochemical study on the aq. extract of D. luridum, we isolated three other polar metabolites, including a non-protein amino acid 4, L-proline 5, and a dipeptide γ-L- glutamylglycine 6. From the chloroform extract, a sterol 7 was obtained by separation on

Sephadex LH-20 column and followed by column chromatography (SiO2) with isocractic

elution (CH2Cl2/MeOH 10:0.5). In addition, a precipitate occurred during evaporation, it was filtered and recrystallized to give ceramide 8. This metabolite also precipitates from acetone upon cooling from 60 °C to rt. When evaporating the remaining acetone solution, another white powder precipitated and was filtered, then washed repeatedly with EtOH, to give a mixture of D-volemitol 9 and D-mannitol 10. The general isolation procedure of the ten compounds is given in Figure 47.

A mixture Comp. 9 + Comp. 10 (50.0 mg)

Figure 47: Isolation procedure of ten compounds from D. luridum

Comp.1: mycosporine glutaminol, Comp.2: mycosporine glutamicol, Comp.3: ethyl ester of mycosporine glutamicol, Comp.4: 2-amino-3-acetylaminopropionic acid, Comp.5: L-proline, Comp.6: γ-L-glutamylglycine, Comp.7: cerevisterol, Comp.8: ceramide, Comp.9: D-volemitol, Comp.10: D-mannitol

3.2.2 Structure elucidation

Identification of the isolated compounds was mainly based on comparison of spectral data with those of the high resolution mass measurement indicated possible structures.

� 2-Amino-3-acetylaminopropionic acid (4)

The HRMS-ESI of compound 4 exhibits molecular formula of C5H10N2O3. The compound reacts with ninhydrin reagent to form a purple spot on TLC plate, suggesting the presence of a primary amino group. Its 1H-NMR spectral features in D2O indicates the presence of a methyl group at δH 2.01 ppm (s), a methylene group

with two non-equivalent protons at δH 3.61 ppm (1H, dd, 6.7, 15.0 Hz) and 3.81 ppm (1H, dd,

3.6, 14.9 Hz), and a methine group at δH 3.91 ppm (dd, 3.6, 6.6 Hz). The low field chemical shift of this methine suggests that it is next to an oxygen or a nitrogen atom. The 13C-NMR

spectrum of 4 shows five signals, which are assigned by DEPT as two carboxyl carbons (δC

175.6 and 172.1), a methine (δC 55.1), a methylene (δC 39.8) and a methyl (δC 21.8) carbons.

The COSY correlations between proton H-2 (δH 3.91) and proton H-3 (δH 3.61 and 3.81) and ’ ’ the HMBC correlations from H-2 to carbon C-1 (δC 175.6) as well as comparison with the reported data [122], [123] confirm the structure of 4 as 2-amino-3-acetylaminopropionic acid. The configuration of carbon C-2 is not determined.

This compound belongs to the family of non-protein amino acids which is a group of secondary nitrogenous compounds. Non-protein amino acids now include about 250 compounds derived from the plant world [124]. Some of these metabolites are accumulated in seeds of neotropical species of the genus Acacia and have been investigated to distinguish seed amino acid patterns [125]. It has been suggested that many non-protein amino acids are toxic

to the larvae of various seed-eating beetles and leaf-eating moths [126], [127], [128]. In fact, compound 4 is found inhibitory of the feeding of both insect species Anacridium melanorhodon and Locusta migratoria [129]. Such non-protein amino acids have never been described in lichens, it would be also interesting to investigate further whether this metabolite protect lichen D. luridum against potential predators.

� L-Proline (5)

The HRMS-ESI establishs the molecular formula of

compound 5 is C5H9NO2. The compound shows a positive light yellow reaction to ninhydrin reagent on TLC plate, suggesting the presence of a secondary amine moiety [130]. Its spectroscopic data confirm that 5 is L-proline, a typical amino acid. Such a primary metabolite was already found in D. moulinsii [113]. Proline was considered as the major osmotic solute in lichen and green algae [131].

� γ-L-Glutamylglycine (6)

Compound 6 has a molecular formula of

C7H12N2O5 determined by HRMS-ESI. This metabolite possess at least a primary amino group as reaction with ninhydrin reagent to form a purple spot 1 on TLC plate. The H-NMR spectrum in D2O shows signals of a methine group at δH 3.80 ppm

(dd. 5.6, 5.6 Hz) and three methylene group at δH 3.75 ppm (s), 2.48 ppm (m), and 2.18 ppm 13 (m). Its C and DEPT NMR show seven carbon signals including three carboxyl groups (δC

177.4, 175.1 and 174.7), a methine group (δC 54.9) and three methylene groups (δC 44.0, 32.1 and 26.9). These spectroscopics data suggest that 6 would be a dipeptide composing of a glycine unit and a glutamic acid unit. These two units could combine together by the three following ways (Figure 48):

(a) glutamic acid is the N-terminal component and the glutamyl residue is linked at α- position.

(b) the glutamic acid is the N-terminal component and the glutamyl residue is linked at γ-position.

Figure 48: Three possible structures (a): α-L-glutamylglycine, (b): γ-L-glutamylglycine and (c) of dipeptide 6

Previously, three isomeric dipeptides containing glutamic acid can be readily distinguished by MS-FAB (fast atom bombardment) [132]. Here, the structure of 6 is determined by HRMS-ESI+ and NMR spectroscopic methods. The HMBC correlation between the methylene protons at δH 3.75 ppm and two carboxyl carbons at δC 174.7 ppm and 177.4 ppm as well as the fragmentation patterns observed at m/z 209, 191, 152, 98 (Figure 49) indicate that structure (c) is irrelevant.

By comparing of its NMR data with those of dipeptide in literature in which the glutamic acid is the N-terminal component, the chemical shift value of glutamyl α-CH in α-glutamyl dipeptides appear at lower field (4.07 ppm), and those of glutamyl α-CH in γ-glutamyl dipeptides at higher field (less than 3.9 ppm) [133]. Compound 6 exhibits its α-CH signal at δH 3.80 ppm suggesting 6 to be γ-L-glutamylglycine (Figure 48 b).

Figure 49: The similar fragmentation patterns observed of two structures (a) and (b)

Many kinds of glutamyl peptides were isolated from higher plants as soybean seed and seedling, green gram seed, ladino clover seed, azuki bean seed, buckwheat seed and broad bean seed [133]. The derivatives of γ -glutamyl dipeptide play as antagonists of excitatory amino acids and as probes of the various receptor binding site [134]. Recently, γ-glutamyl dipeptide has been found to induce persistent signaling from endosomes by a nutrient transceptor [135].

� (22E,24R)-Ergosta-7,22-diene-3β,5α,6β-triol (Cerevisterol) (7)

The HRMS-ESI spectrum of compounds 7 shows an ion peak at m/z 453.3350 [M+Na]+ 13 (calcd. for C28H46O3Na 453.3344). The C-

NMR spectra in CDCl3 shows the existence of 28 skeleton carbons: three sp3 carbons attached

to oxygen atom, including two secondary at δC

68.88 and 73.81ppm and one tertiary at δC 76.11 ppm, four olefinic carbons at δC 117.7, 132.3, 135.5 and 144.2 ppm, six methyls, seven methylenes, six methines and two quaternary carbons. The 1H-NMR spectrum displays two methyl singlets at δH 0.53 and 1.02 ppm, four methyl doublets at δH 0.74 (d, 6.7 Hz), 0.78 (d,

6.7 Hz), 0.86 (d, 6.8 Hz) and 0.97 ppm (d, 6.6 Hz), three olefinic protons at δH 5.10 (1H, dd, 7.5, 15.2 Hz), 5.20 (1H, dd, 6.6, 15.2 Hz) and 5.30 ppm (1H, m) and two oxymethine protons

at δH 3.57 (d, 5.0 Hz) and 4.05 ppm (m). These data as well as the COSY and HMBC correlations (Figure 50) suggest an ergostadiene-type sterol structure. Comparison of the

chemical shift values of H-19, H-6, and H-7 of 7 (δH 1.02, 3.57 and 5.30) with those of ergosta-

7, 22-diene-3β,5α,6α-triol (δH 1.02, 3.93 and 5.00) and ergosta-7,22-diene-3β,5α,6β-triol (δH 1.09, 3.57 and 5.30) establish the ∆7-3β,5α,6β-triol substructure for 7 [136]. Moreover, its D optical rotation [α] 20 = - 34.0 (c= 0.16, pyridine) is suitable with reported data [137]. Thus, the structure of 7 is (22E,24R)-ergosta-7,22-diene-3β,5α,6β-triol or cerevisterol.

Figure 50: Key 1H–1H COSY (bold line) and HMBC (1H→13C) correlations of compound 7

Polyhydroxyl ergostane-type sterols and their derivatives occurring in plant, fungi, soft corals, and sponges possess a variety of biological activities such as antimicrobial, anti- inflammatory, antifouling, cytotoxic, 5α-reductase inhibitiory, and farnesoid X-activated receptor effects [138]. In lichen, ergosterol metabolite is specific to fungi and plays the same role as cholesterol in human cells. Cerevisterol has previously been described in the lichen species Ramalina hierrensis [139] and Stereocaulon azoreum [140] but it is the first report in D. luridum.

� (2S,3S,4R,2’R)-2-(2’-Hydroxytetracosanoylamino)octadecan-1,3,4-triol (8)

The HRMS-ESI spectrum of 8 indicates a molecular formula of C42H85NO5 (m/z = 706.6324 + [M+Na] calcd. for C42H85NO5Na706.6325). Its IR spectrum reveals absorption bands corresponding to hydroxyls at 3329 and 3203 cm-1, a secondary amide at 1620 and 1545 cm-1, and long aliphatic chains at 723 cm-1. The 1H-NMR spectrum in pyridine of 8 shows the presence of two terminal methyls at δH 0.87 (3H, t, 7.1 Hz) and 0.88

ppm (3H, t, 7.0 Hz) and methylenes at δH 1.28–1.44 ppm (ca. 56H, brs), an amide proton signal

at δH 8.60 ppm (1H, d, 8.8), five characteristic signals of protons geminal to hydroxyl groups at δH 4.32 (1H, m), 4.39 (1H, dd, 6.6, 4.4 Hz), 4.46 (1H, dd, 10.9, 4.9 Hz), 4.54 (1H, dd, 10.5, 4.4 Hz) and 4.65 ppm (1H, dd, 7.7, 3.9 Hz). A sixth signal at low field appears as a multiplet 13 at δH 5.15 ppm and is assigned as a methine proton vicinal to the nitrogen atom. The C-NMR

and HSQC spectra of 8 show the presence of one quaternary carbon at δC 175.4 ppm (CONH,

C-1′), four methines at δC 53.1 (CHNH, C-2), 72.6 (CHOH, C-2′), 73.2 (CHOH, C-4) and 77.0 ppm (CHOH, C-3), a methylene at δC 62.2 ppm (CH2OH, C-1). All of the above spectral data reveal that 8 should be a phytosphingosine-type ceramide containing a 2-hydroxy fatty acid. The lengths of the long chain sphingoid base and the fatty acid are determined by positive electrospray ionization ESI-MS/MS. The fragmentation patterns observed at m/z 307 and 340 indicate that the long chain sphingoid base and the fatty acid of 8 must be 2-amino-1,3,4- octadecanetriol and 2-hydroxytetracosanoic acid, respectively (Figure 51). The stereochemistry at C-2, C-3, C-4 and C-2’ is determined as the 2S, 3S, 4R, 2’R configuration D by comparing the optical rotation value of 8 [α] 20 = + 8.0, (c= 0.2, pyridine) with the one D reported in the literature [α] 20 = + 9.6, (c= 0.25, pyridine) [141].

Figure 51: ESI-MS fragmentation of 8

The history of this interesting group of natural products called sphingolipids or ceramides was initially born out of studies on the chemical constitution of the brain cells as early as 1880 [142]. They were involved in diverse cellular processes including cell growth, survival,

differentiation, and adhesion [143], [144]. Such metabolites have been reported to exhibit anti- ulcerogenic, antihepatotoxic, antitumor and immunostimulatory activities [145], and more particularly for compound 8, a significant anti inflammatory activity [146] and a quinone reductase-inducing activity [147].

Their structures consist of a fatty acyl of variable chain lengths bound to an amino group of a sphingoid base, typically sphingosine [148]. The fatty acyl chains are in general saturated or mono-unsaturated and might contain an OH group linked to C2 or to the terminal carbon atom (α- and ω-hydroxy fatty acids, respectively). Among ceramides, those containing long (C16–20) and very long (C22–24) acyl chains are the most abundant in mammalian cells, but ceramides with longer acyl chains (C26–36) can also be found in epidermal keratinocytes and male germ cells during their differentiation and maturation [149]. As a general trend, ceramides present very low polarity and are highly hydrophobic. So, they are poorly soluble in water and cannot exist in solution in biological fluids or in the cytosol. Interestingly, ceramides are the major constituents of the stratum corneum, and this might account for the permeability barrier properties of the skin [150].

More than 300 sphingolipids have been identified from various organisms including

plant, fungi, algae, sponges, sea anemones, sea stars, tunicates, soft corals etc. [142], [146], [151]. Compound 8 has been fisrtly isolated from the edible Basidiomycete Russula cyanoxantha [141], then from Leccinum extremiorientale [152] and Paxillus panuoides [153]. The characterization of a glycosphingolipid in a lichen was previously reported by Marcos in

1997 [154]. This study pointed out the occurrence of galactosylceramides in the lichen Ramalina celastri species. Its main lipid components are (4E)-sphingenine (code a), sphinganine (code b) and icosa-sphinganine (code c), esterified with palmitic (code x), oleic (code y) and 2-hydroxypalmitic acids (code z) (Figure 52). To our knowledge, compound 8 is isolated for the first time in lichens.

Figure 52: Galactosylceramides isolated in lichen Ramalina celastri (Gal: galactose)

� Mixture of D-Volemitol (9) and D-Mannitol (10) While the acetone solution was evaporated, a precipitate occurred. After filtration, the solid was washed repeatedly with EtOH to give a white amorphous powder. The HRMS-ESI spectrum presents two ion peaks at m/z = 235.0792 [M+Na]+ + (calcd. for C7H16O7Na 235.0794) and m/z = 205.0699 [M+Na] 1 (calcd. for C6H14O6Na 205.0688). Together with its H-NMR and 13C-NMR data, the mixture is identified containing two common sugars: D-volemitol 9 and D-mannitol 10 with a ratio 85:15.

To summarize, through the seven “others compounds” found in D luridum, compounds 4, 6 and 8 have been isolated for the first time in lichens and compound 7 has been described for the first time in D. luridum.

4 Experimental

4.1 Extraction

Crushed sorted out and air-dried lichen material (150 g) was macerated with 500 mL of pure water at + 4 °C for 15 h. The supernatant was filtrated and the extraction was repeated until mycosporines were not detected (by TLC under UV detection at 312 nm). The combined extracts were partly concentrated under reduced pressure and then lyophilized to give the crude aq. extract (7.5 g). The dried residue was then extracted by stirring with chloroform at rt, for 4 h (500 mL×3) and acetone at 60 ºC for 4 h (500 mL × 3) to yield crude 4.5 g and 3.2 g of chloroform and acetone extracts, respectively.

4.2 Purification and identification of mycosporines

4.2.1 Purification on cation exchange resin

The crude aq. extract (7.5 g) was dissolved in 10 mL of water and injected on a cation exchange resin column (DOWEX 50W-X8, 80 g). The resin was previously washed with

MeOH (100 × 2 mL) then with H2O (100 × 2 mL) and degassed before being poured into a column (30 × 2.5 cm). The resin was converted to H+ with 100 mL of HCl 1 M, and washed thoroughly with water until the effluent attained a pH > 5. After injection, 200 mL of water were then added to remove unwanted compounds, including sugars and polyols (fraction A1). Then 80 mL of NaCl 200 mg/mL were added, 50 mL fractions were collected with water as eluant. The analytical monitoring of the elution was carried out by TLC (mobile phase

CHCl3/MeOH/H2O 6:4:1) and UV detection at 312 nm. Fractions 1–6 (300 mL) were combined and evaporated to dryness. To remove the excess of NaCl, the dried residue was dissolved in 100 mL of EtOH and centrifuged at 2500 rpm for 5 min. After filtration of the insoluble NaCl, EtOH was evaporated to dryness to obtaine 1.5 g of a semi-purified aq. extract (fraction A2).

4.2.2 Purification by CPC

The separation was performed with a biphasic solvent system BuOH/AcOH/H2O 4:1:5 v/v/v, in the isocratic mode. The rotor was first filled with the lower phase of the solvent system which was corresponding to the stationary phase. The apparatus was rotated at 1500 rpm and the upper mobile phase of the solvent mixture was then pumped into the inlet of the column at

a flow rate of 4 mL/min in the ascending mode. The semi-purified aq. extract (500 mg) was diluted in a mixture of 2 mL of the upper phase and 2 mL of the lower phase. It was loaded in the 5 mL injection-loop, and injected in the column in a “sandwich” mode, i.e. at the same time with the mobile phase. The back pressure was 25 bars. The stationary phase retention at the end of the separation represented 52% of the column volume (57 mL). The content of the outgoing organic phase was offline monitored by TLC analysis.

Elution first occurred in the ascending mode (normal-phase mode): the rotor was filled with the lower polar phase of the solvent mixture, and the pumped mobile phase was the apolar upper phase. The summary of the collected tubes corresponding to the mode used was presented in Table 8. Extrusion was performed after the 60th tube: the upper phase was pumped in the descending mode to eject the totally of the lower phase out of the rotor.

Table 8: Summary of the collected tubes corresponding to the mode used in CPC experiment

Tube Mode Mobile phase Volumn collection

1–10 ascending upper phase

11–20 descending lower phase 2 mL/tube 21–37 ascending upper phase

38–60 descending lower phase

4.2.3 A new purification protocol

The semi-purified aq. extract (fraction A2, 1.5 g) was subjected to flash chromatography. The stationary phase was a bare silica column (Chromabond® Flash RS 15 g SiOH Ref. 732801.

Macherey-Nagel) with mobile phase A (ACN/ CH3COONH4 50 mM 90:10, pH 5.36) and

mobile phase B (ACN/ H2O/ CH3COONH4 50 mM 50:40:10, pH 5.36). The gradient elution was: 100% of A during 5 min, 0–100% of B during 20 min, and 100% of B for 15 min with the flow rate at 10 mL/min. Fractions of 10 mL were collected. Fractions 5–9 containing mycosporines were combined (fraction A2.1, 210 mg) and further purified on an open reverse phase column (C-18 Hydro Chromabond, 4.3 g, Ref. 732810, Macherey-Nagel) using water as the mobile phase with the flow rate at 1.5 mL/min to give compound 3 (m = 6.0 mg) and a sub-

fraction with two other mycosporines 1 and 2. The sub-fraction was subjected to semi- preparative HPLC on a Prevail C18 column (5 µm, 250×10.0 mm) with a guard column Prevail

C18 (10 µm, 33×7.0 mm). The mobile phase of solvent A (H2O) and solvent B (ACN) was regulated at 2.0 mL/min by using the following gradient program: 100% of A for 15 min, 100– 50% of A over 1 min, 50% of A for 10 min, 50–100% of A over 1 min, and 5 min raw of reequilibration. Peak detection was carried out online using a diode array detector at 254, 310, 330 and 360 nm, and absorption spectra (200-400 nm) were recorded each second directly on the HPLC separated peaks. Compound 2 (m = 6.5 mg) was eluted at Rt = 7.23 min, followed by compound 1 (m = 5.0 mg) at Rt = 10.47 min.

4.2.4 Sources, physico-chemical and spectroscopic data

Mycosporine glutaminol 1

Structure:

Sources: Dermatocarpon luridum, D. miniatum, D. leptophyllodes, D. meiophyllizum, D. arnoldianum

Aspect: colorless powder

TLC: Rf = 0.3 (CHCl3/MeOH/H2O 6:4:1)

HPLC: • Kinetex 2.6 µm HILIC 100 Å (100×4.60 mm) column; mobile phase A (ACN:

CH3COONH4 50 mM 90:10, pH 5.36) and mobile phase B (ACN: H2O: CH3COONH4 50 mM 50:40:10, pH 5.36); the gradient: 100% A (0–2 min), 0–100% B (2–4 min), 100% B (4–12 min), 0–100% A (12–14 min) and hold 100% A for 2 min; the flow rate

= 1 mL/min; Rt = 4.87 min.

• Zorbax Eclipse XDB–C18 (150 mm×2.1 mm) column; mobile phase A (H2O + 0.1% formic acid) and mobile phase B (ACN + 0.1% formic acid); the gradient: 99% A and 1% B (0–7 min), 99%–70% A and 1%–30% B (7–22 min), 70%–50% A and 30%–50% B (22–28 min), 50%–0% A and 50%–100% B during 2 min, 100% B during 3 min, 0%–99% of A during 2 min and a 15–min raw of reequilibration; the flow rate = 200

µL/min; Rt = 3.02 min.

• Prevail C18 5 µ (250×4.60 mm) column; mobile phase A (H2O) and mobile phase B (ACN); the gradient: 100% A (0–15 min), a linear increase to 50% B (15–16 min), remaining the condition during 16 and 26 min, a linear decrease to 0% B (26–27 min),

reequilibration 5 min with the moblie phase A; the flow rate = 1 mL/min; Rt = 7.23 min.

MS- ESI: m/z = 303 [M+H]+. Fragmentation patterns: 235, 285

+ HRMS-ESI: m/z = 325.1379 [M+Na] (calcd. for C13H22N2O6Na 325.1376). Fragmentation patterns: 310, 307, 292, 280, 279, 252, 238, 225, 220, 210, 195, 179, 137, 123

-1 IR νmax cm : 3141, 2987, 1661, 1652, 1531, 1403, 1066

UV: λmax 310 (H2O, ε = 12540)

100

Mycosporine glutamicol 2

Structure:

Sources: Dermatocarpon luridum, D. miniatum, D. leptophyllodes, D. meiophyllizum, D. arnoldianum, D. rivulorum, D. hepaticum, D. intestiniforme, D. cinereum, D. rufescens

Aspect: colorless powder

TLC: Rf = 0.23 (CHCl3/MeOH/H2O 6:4:1)

HPLC: • Kinetex 2.6 µm HILIC 100 Å (100×4.60 mm) column; mobile phase A (ACN:

= 1 mL/min; Rt = 5.62 min.

µL/min; Rt = 4.75 min.

• Prevail C18 5 µ (250×4.60 mm) column; mobile phase A (H2O) and mobile phase B (ACN); the gradient: 100% A (0– 15 min), a linear increase to 50% B (15–16 min), remaining the condition during 16 and 26 min, a linear decrease to 0% B (26–27 min),

reequilibration 5 min with the moblie phase A; the flow rate = 1 mL/min; Rt = 10.47 min.

101

MS- ESI: m/z = 304 [M+ H]+. Fragmentation patterns: 236, 258, 286

HRMS-ESI: m/z = 326.1198 [M+Na]+ (calcd. for C13H21NO7Na 326.1216). Fragmentation patterns: 308, 286, 244, 226, 180

-1 IR νmax cm : 3354, 2935, 1650, 1557, 1404, 1120, 1024

UV: λmax 310 (H2O, ε = 17250)

1 1 13 NMR H (500 MHz, D2O), H (300 MHz, pyridine-d5), C (125 MHz, D2O), COSY, HMBC:

1 13 Position δH , J (Hz) δH , J (Hz) δC COSY HMBC ( H→ C)

D2O pyridine-d5 D2O 1 185.5 2 130.1 3 158.9 4 2.89 (1H, d, 17.1) 3.53 (1H, d, 16.6) 33.5 4, 6 2, 3, 5, 6, 7 2.77 (1H, d, 17.3) 3.23 (1H, d, 16.2) 5 72.1 6 2.67 (1H, d, 16.9) 3.23 (1H, d, 16.2) 42.6 4, 6 1, 2, 4, 5, 7 2.41 (1H, d, 16.9) 3.04 (1H, d, 16.9) 7 3.53 (2H, s) 3.96-4.00 (2H, m) 67.6 4, 5, 6 8 3.57 (3H, s) 3.82 (3H, s) 59.1 2 9 3.73 (1H, m) 3.96-4.00 (1H, m) 55.1 10, 11 11 10 3.70 (1H, d, 3.9) 3.96-4.00 (2H, m) 64.2 9, 10, 11 9, 11 3.56 (1H, d, 3.8 ) 11 1.88 (1H, m) 2.38 (2H, m) 27.4 9, 10, 11, 9, 12, 13 1.75 (1H, m) 12 12 2.29 (2H, m) 2.73 (2H, m) 31.4 11, 12 9, 11, 13 13 181.8 NH 6.87 (1H, d, 9.9)

102

1 Figure 53: H NMR spectrum of 2 (D2O, 500 MHz)

13 Figure 54: C NMR spectrum of 2 (D2O, 125 MHz)

103

Ethyl ester of mycosporine glutamicol 3

Structure:

Sources: Dermatocarpon luridum, Dermatocarpon miniatum

Aspect: brown viscous liquid

TLC: Rf = 0.77 (CHCl3/MeOH/H2O 6:4:1)

HPLC: • Kinetex 2.6 µm HILIC 100 Å (100×4.60 mm) column; mobile phase A (ACN:

CH3COONH4 50 mM 90:10, pH 5.36) and mobile phase B (ACN: H2O: CH3COONH4 50 mM 50:40:10, pH 5.36); the gradient: 100% A (0 - 2 min), 0–100% B (2 - 4 min), 100% B (4 - 12 min), 0–100% A (12 - 14 min) and hold 100% A for 2 min; the flow

rate= 1 mL/min; Rt = 1.91 min.

µL/min; Rt = 15.72 min.

reequilibration 5 min with the moblie phase A; the flow rate = 1 mL/min; Rt = 5.64 min.

104

MS- ESI: m/z = 332 [M+ H]+. Fragmentation patterns: 218, 264, 286, 314

+ HRMS-ESI: m/z = 354.1522 [M+Na] (calcd. for C15H25NO7Na 354.1523). Fragmentation patterns: 339, 308, 322, 321, 306, 304, 290, 279, 254, 252, 238, 234, 220, 208, 195, 166

-1 IR νmax cm : 3317, 2938, 1712, 1665, 1531, 1399, 1258, 1196, 1114, 1044

UV: λmax 309 (H2O, ε = 21300)

1 13 NMR H (300 MHz, D2O), C (75 MHz, D2O), COSY, HMBC:

Position δH , J (Hz) δC COSY HMBC (1H→ 13C) 1 185.7 2 130.2 3 158.8 4 2.89 (1H, d, 17.2) 33.4 4, 6 2, 3, 5 2.70 (1H, d, 16.8) 5 72.0 6 2.68 (1H, d, 17.0) 42.8 4, 6 1, 2, 5 2.41 (1H, d, 17.1) 7 3.52 (2H, s) 67.5 4, 5, 6 8 3.54 (3H, s) 59.1 2 9 3.75 (1H, m) 54.7 10, 11 10 3.67 (1H, dd, 4.2, 11.6) 64.2 9, 10 3.57 (1H, m) 11 1.92 (1H, m) 25.8 9, 12 1.81 (1H, m) 12 2.48 (2H, t, 7.1) 30.5 11 9, 11, 13 13 175.9 1’ 4.15 (2H, q, 7.1) 61.8 2’ 2’, 13 2’ 1.22 (3H, t, 7.1) 13.2 1’ 1’

105

Figure 55: 1H-1H COSY NMR spectrum of 3

Figure 56: HSQC NMR spectrum of 3

106

Figure 57: HMBC NMR spectrum of 3

4.3 Other isolated compounds

4.3.1 Purification and identification

The semi-purified aq. extract A2 (1.5g) obtained through cation exchange resin Dowex chromatography was subjected to flash chromatography using bare silica-gel as stationary phase to give two subfractions A2.1 and A2.2 (see condition chromatography in section 4.2.3 A new purification protocol). The residue subfraction A2.2 (1.0 g) was applied on flash chromatography using column Chromabond® Flash RS 15 g SiOH Ref. 732801. Macherey-

Nagel. Sample was run with a mobile phase of CHCl3 and MeOH in linear gradient mode (100-

0% of CHCl3 over 90 min, flow rate at 8.0 mL/min). Fractions of 10 mL were collected. Fractions 30-65 containing a precipitate were combined and filtered to provide 90 mg of a solid. One part of this precipitate (15 mg) was subjected to CC and eluted isocratically with

CHCl3/MeOH 6:4 to obtain compound 4 (7.0 mg). The residue was then purified by CC using

CHCl3/MeOH 6:4 as eluent. By combining the fractions with TLC monitoring, four fractions were obtained. Fraction 2 was separated by preparative TLC (CHCl3/MeOH/H2O 6:4:1) to yield compound 5 (5.8 mg) and compound 6 (3.0 mg).

107

A part of the chloroform extract (1.0 g) was chromatographed over Sephadex LH-20 with

CHCl3 to obtain five fractions. Fraction 3 (18 mg) was subjected to CC eluted with

CH2Cl2/MeOH 20:1 to give compound 7 (6.0 mg). Fraction 4 was concentrated to small volume and then a white solid was precipitated from the solution. The precipitate was filtered (30 mg) and recrystallized by PE/Ac 1:1 to give compound 8 (11.0 mg).

From the acetone extract, the first precipitate was obtained as the temperature of solvent decrease from 60 °C to rt. After filtration and recrystallization (PE/Ac 1:1), we obtained again compound 8 (5.0 mg). While the acetone solution was evaporated, the second precipitate occurred and after filtration followed by thorough washing with EtOH, compound 9 was obtained (50 mg).

108

4.3.2 Structure elucidation

2-Amino-3-acetylaminopropionic acid (4)

Structure:

Sources: Dermatocarpon luridum

Aspect: colorless solid

TLC: Rf = 0.28 (CHCl3/MeOH/H2O 6:4:1)

+ HRMS- ESI: m/z = 169.0592 [M+Na] (calcd. for C5H10N2O3Na 169.0589)

1 13 NMR H (300 MHz, D2O), C (75 MHz, D2O), COSY, HMBC:

1 13 Position δH, J (Hz) δC COSY HMBC ( H→ C) 1 172.1 2 3.91 (1H, dd, 3.6, 6.6) 55.1 3 3 3.81 (1H, dd, 3.6, 14.9) 39.8 2 3.61 (1H, dd, 6.7, 15.0) 1’ 175.6 2’ 2.01 (3H, s) 21.8 1’

109

1 Figure 58: H NMR spectrum of 4 (D2O, 300 MHz)

13 Figure 59: C NMR spectrum of 4 (D2O, 75 MHz)

110

L-Proline (5)

Structure:

Sources: Dermatocarpon luridum

Aspect: white amorphous powder

M.p : 224 – 226 °C

TLC: Rf = 0.49 (CHCl3/MeOH/H2O 6:4:1)

+ HRMS-ESI: m/z = 138.0529 [M+Na] (calcd. for C5H9NO2Na 138.0531)

+ m/z = 160.0358 [M-H+2Na] (calcd. for C5H8NO2Na2 160.03504)

1 13 NMR H (300 MHz, CD3OD), C (75 MHz, CD3OD), COSY, HMBC:

1 13 Position δH, J (Hz) δC COSY HMBC ( H→ C) 2 3.96 (1H, dd, 6.1, 8.6 ) 62.5 3 3, 4, 5, 1’ 3 2.11 (1H, m) 30.4 2, 4 2, 4, 5, 1’ 2.33 (1H, m) 4 1.94 (2H, m) 25.1 3, 5 2, 3, 5 5 3.21 (1H, m) 46.9 4 2, 3, 4 3.39 (1H, m) 1’ 174.2

111

γ-L-Glutamylglycine (6)

Structure:

Sources: Dermatocarpon luridum

Aspect: white amorphous powder

TLC: Rf = 0.11 (n-BuOH/AcOH/H2O 40:20:10)

+ HRMS-ESI: m/z = 227.0645 [M+Na] (calcd. for C7H12N2O5Na 227.0644)

+ m/z = 249.0461 [M-H+2Na] (calcd. for C7H11N2O5Na2 249.0463)

+ m/z = 271.0280 [M-2H+3Na] (calcd. for C7H10N2O5Na3 271.0283)

Fragmentation patterns: 209, 191, 152, 98

1 13 NMR H (300 MHz, D2O), C (75 MHz, D2O), COSY, HMBC:

1 13 Position δH , J (Hz) δC COSY HMBC ( H→ C) 1 175.1 2 3.80 (1H, dd, 5.6, 5.6) 54.9 3 1, 3, 4 3 2.18 (2H, m) 26.9 2, 4 1, 2, 4, 5 4 2.48 (2H, m) 32.1 3 2, 3, 5 5 174.7 1’ 3.75 (2H, s) 44.0 2’, 5 2’ 177.4

112

1 Figure 60: H NMR spectrum of 6 (D2O, 300 MHz)

13 Figure 61: C NMR spectrum of 6 (D2O, 75 MHz)

113

(22E,24R)-Ergosta-7,22-diene-3β,5α,6β-triol (Cerevisterol) (7)

Structure:

Sources: Dermatocarpon luridum

Aspect: white amorphous powder

M.p: 260- 263 °C

TLC: Rf = 0.52 (EtOAc/CH2Cl2/MeOH 15:12:3)

D [α] 20 : -34.0 (c = 0.16, pyridine)

HRMS-ESI: m/z = 453.3350 [M+Na]+ (calcd. for C28H46O3Na 453.3344)

1 13 NMR H (300 MHz, CDCl3), C (75 MHz, CDCl3), COSY, HMBC:

114

1 13 Position δH, J (Hz) δC COSY HMBC ( H→ C) 1 29.9 2 33.2 3 4.05 (1H, m) 68.9 2, 4ax, 4eq

4 2.12 (1H, dd, 11.4, 13.0, H4ax) 40.6 3, 4eq

1.80 (1H, dd, 5.9, 12.6, H4eq) 3, 4ax 5 - 76.1 6 3.57 (1H, d, 5.0) 73.8 7 7 5.30 (1H, m) 117.7 6 8 - 144.2 9 43.6 10 - 37.3 11 22.2 12 39.4 13 - 43.9 14 54.9 15 23.0 16 28.1 17 56.1 18 0.53 (3H, s) 12.5 12, 13, 17 19 1.02 (3H, s) 19.0 2, 5, 9, 10 20 39.6 17, 21, 22 21 0.97 (3H, d, 6.6) 21.3 20 17, 20, 22 22 5.20 (1H, dd, 6.6, 15.2) 135.5 20, 23 23 5.10 (1H, dd, 7.5, 15.2) 132.3 22, 24 24 43.0 25 33.1 26 0.78 (3H, d, 6.7)** 20.1* 25 24, 25, 27 27 0.74 (3H, d, 6.7)** 19.8* 25 24, 25, 26 28 0.86 (3H, d, 6.8) 17.7 24 23, 24, 25

* , ** : shifts may be exchangeable

115

1 Figure 62: H NMR spectrum of 7 (CDCl3, 300 MHz)

13 Figure 63: C NMR spectrum of 7 (CDCl3, 75 MHz)

116

(2S,3S,4R,2’R)-2-(2’-Hydroxytetracosanoylamino)octadecan-1,3,4-triol (8)

Structure: O 5' 23' 24' 1' 3' (CH2)19CH3 HN 2' 4' OH OH 6 HO 2 4 7 17 18

1 3 5 (CH2)11CH3 OH

Sources: Dermatocarpon luridum

Aspect: white amorphous powder

M.p: 120 –121°C

TLC: Rf = 0.15 (Tol/EtOAc/AcOH 70:25:5)

D [α] 20 : +8.0 (c = 0.2, pyridine)

HRMS-ESI: m/z = 706.6324 [M+Na]+ (calcd. for C42H85NO5 706.6325)

Fragmentation patterns: 307, 340

-1 IR (KBr) νmax (cm ): 3329 and 3203 (hydroxyl), 1620 and 1545 (amide), 723 (aliphatic)

1 13 NMR H (500 MHz, pyridine-d5), C (125 MHz, pyridine-d5):

117

Position δH (J Hz) δC 1 4.46 (1H, dd, 10.9, 4.9) 62.2 4.54 (1H, dd, 10.5, 4.4) 2 5.15 (1H, m) 53.1 3 4.39 (1H, dd, 6.6, 4.4) 77.0 4 4.32 (1H, m) 73.2 5 2.29, 1.96 (2H, m) 34.3 6 1.73 (2H, m) 26.0 7-17 1.28 – 1.44 23.1–32.3 18 0.87 (3H, t, 7.1) 14.4 1’ - 175.4 2’ 4.65 (1H, dd, 7.7, 3.9) 72.6 3’ 2.27 (2H, m) 35.9 4’ 1.97 (2H, m) 26.8 5’-23’ 1.28 – 1.44 23.1–32.3 24’ 0.88 (3H, t, 7.0) 14.4 NH 8.61 (1H, d, 8.8) -

118

1 Figure 64: H NMR spectrum of 8 (pyridine-d5, 500 MHz)

13 Figure 65: C NMR spectrum of 8 (pyridine-d5, 125 MHz)

119

Figure 66: HSQC NMR spectrum of 8

120

A mixture of D-Volemitol (9) and D-Mannitol (10) (85:15)

Structure:

Sources: Dermatocarpon luridum

Aspect: white amorphous powder

M.p: 152 – 153 °C

TLC: Rf = 0.49 (CHCl3/ MeOH/ H2O 6:4:1)

+ HRMS-ESI: D-Volemitol m/z = 235.0792 [M+Na] (calcd. for C7H16O7Na 235.0794)

+ D-Mannitol m/z = 205.0699 [M+Na] (calcd. for C6H14O6Na 205.0688)

1 NMR H (500 MHz. D2O): δ 3.62 – 3.96 (m)

13 NMR C (125 MHz, D2O): δ 64.7 (C-1), 72.3 (C-2), 71.0 (C-3), 71.1 (C-4), 73.0 (C-5), 74.4 (C-6), 63.5 (C-7); δ 64.76 (C-1’, C-6’), 70.79 (C-2’, C-5’), 72.36 (C-3’, C-4’)

121

4.4 Determination of the (ε) and pKa values of the isolated mycosporines

4.4.1 The molar absorption coefficient

The UV spectra of three mycosporines 1, 2 and 3 were performed in water at concentration 2.5×10-5M. The molar absorption coefficient (ε) was calculated according to the Beer-Lambert law:

A ε = l × C (Eq. 9)

A: absorbance value at λmax l: the pathlength of curve (cm) (l = 1) C: the concentration of solution (M)

4.4.2 pKa values

� Preparation of buffer solutions

The buffer solutions are the Britton and Robinson ones for pH>2 [155], [156]. They are composed of a mixture of phosphoric acid acetic acid and boric acid which constitute the stock solution at 0.08 M. Then sodium hydroxide 0.4 M and sodium chloride 1M (to set the ionic strength at 0.1 M) are added to obtain the required pH according to Table 9.

Table 9: Preparation of buffer solutions with pH>2

pH V stock (mL)* V NaOH (mL) V NaCl (mL) pH measured

2.0 12.5 0.6 2.0 1.85 2.2 12.5 1.2 1.875 2.09 2.5 12.5 1.7 1.75 2.53 2.7 12.5 1.9 1.65 2.79

3.0 12.5 2.0 1.6 3.02 3.3 12.5 2.2 1.6 3.32 3.5 12.5 2.55 1.45 3.84

122

*Stock solution 0.08M composition: 1.844 g phosphoric acid + 0.96 g acetic acid + 0.99 g boric acid + 200 mL water.

For pH<2, Bascombe and Bell’s acidity functions are chosen, based on the use of concentrated solution of sulfuric acid [157]. Buffer between pH 1 and 2 are prepared by dilution in water of the buffer pH 1 according to Table 10.

Table 10: Preparation of buffer solutions with pH<2

th pH V water (mL) m H2SO4 conc (g) pH solution (dilution 9/10 ) 1.00 50 0.738 1.08 0.50 15 0.444 0.59 0.00 15 0.883 0.12 -1.00 15 3.342 nd

pH V water (mL) V solution pH1 (mL) pH solution (dilution 9/10th) 1.20 15 9.463 1.23 1.35 15 5.971 1.43 1.71 15 2.376 1.76

Overall twelve or thirteen buffer solutions at pH range from -1 to 3.8 were prepared. Before using the buffer solutions, each pH buffer (pH>0) was measured with a pH meter that was calibrated daily with four commercial buffers (manufactured according to the NIST recommendations).

� Preparation of mycosporine solutions and absorbance measurement

Ten mL of a stock solution of mycosporine in water was prepared (6.8×10-4 M and 1.56×10-4 M for mycosporine 1 and 3, respectively). 100 µL was taken and then diluted with 900 µL of the appropriate buffer. Then we recorded the absorbance between 200 and 400 nm in each buffer and six independent measures were performed for each mycosporine (Table 11). Absorbances were taken at 310 nm.

123

Table 11: Measured absorbance at 310 nm depending on pH of two mycosporines 1 and 3 (n=6) Mycosporine 1

1 2 3 4 5 6 average pH i Abs ( λ = 310 nm) -1.00 0.5489 0.5440 0.5539 0.5566 0.5374 0.5812 0.5537 0.12 0.5686 0.5804 0.5899 0.5720 0.5760 0.5824 0.5782 0.59 0.6021 0.6152 0.6125 0.6089 0.6108 0.6209 0.6117 1.08 0.6658 0.6786 0.6704 0.6705 0.6756 0.6670 0.6713 1.23 0.6851 0.6444 0.6988 0.7355 0.7051 0.7154 0.6974 1.43 0.7280 0.7330 0.7390 0.7398 0.7379 0.7413 0.7365 1.76 0.7711 0.7890 0.7913 0.7920 0.8166 0.7890 0.7915 1.85 0.7777 0.7953 0.7838 0.7874 0.7873 0.8043 0.7893 2.09 0.8276 0.8180 0.8179 0.8209 0.8210 0.8220 0.8212 2.53 0.8440 0.8506 0.8536 0.8512 0.8527 0.8525 0.8507 2.79 0.8466 0.8581 0.8708 0.8570 0.8729 0.8580 0.8606 3.02 0.8473 0.8500 0.8595 0.8515 0.8547 0.8461 0.8515 3.32 0.8410 0.8707 0.8461 0.8504 0.8475 0.9020 0.8596 3.84 0.8463 0.8512 0.8654 0.8547 0.8509 0.8514 0.8533

Mycosporine 3

1 2 3 4 5 6 average pH i Abs ( λ = 310 nm) -1.00 0.2300 0.2346 0.2393 0.2572 0.2384 0.2063 0.2343 0.14 0.2480 0.2504 0.2483 0.2467 0.2448 0.2373 0.2459 0.56 0.2594 0.2559 0.2654 0.2590 0.2734 0.2510 0.2607 0.99 0.2745 0.2745 0.2678 0.2769 0.2811 0.2877 0.2771 1.29 0.2919 0.2877 0.2998 0.2908 0.3073 0.2845 0.2937 1.49 0.3106 0.3102 0.3041 0.3063 0.3101 0.3124 0.3090 1.66 0.3071 0.3197 0.3072 0.3081 0.3284 0.2863 0.3095 1.86 0.3069 0.3206 0.3218 0.3192 0.3298 0.3186 0.3195 2.14 0.3200 0.3233 0.3274 0.3455 0.3342 0.3228 0.3289 2.48 0.3210 0.3326 0.3336 0.3325 0.3439 0.3288 0.3321 2.70 0.3429 0.3399 0.3410 0.3335 0.3395 0.3356 0.3387 3.09 0.3302 0.3410 0.3382 0.3378 0.3450 0.3373 0.3383 3.62 0.3250 0.3402 0.3433 0.3389 0.3362 0.3329 0.3361

124

CHAPTER 4: BIOLOGICAL TESTS AND PHOTOPROTECTIVE EVALUATION

125

126

Purpose of this study

As secondary metabolites have a good profile to be designed as drugs we took the opportunity to test some isolated compounds available in suitable quantity for some biological activities. Wathever the physico-chemical properties of mycosporines to be valued in photoprotection, some information about their possible cytotoxicity is a pre-requisite. Therefore, we performed a screening for the cytotoxic activities on eight cell lines available on a testing Platform: ImPACcell – University of Rennes 1. As mycosporine glutaminol 1 was not available in a pure form and in sufficient quantity at the testing period, we only tested mycosporines 2 and 3, two original amino-acid (4 and 6), the steroid 7 (cerevisterol) and the ceramide 8.

The considered cell lines corresponded to two skin related cell lines i. e. HaCaT: human immortalized keratinocytes and a Fibroblast cell line which synthesize the extracellular matrix and collagen but also to six organ-related cancer cell lines. These human derived cancer cell lines correspond to:

- Huh7: a hepato cellular carcinoma,

- CaCo-2: a colorectal adenocarcinoma,

- MDA-MB-231: a breast cancer,

- HCT116: a colon cancer,

- PC3: a prostate cancer,

- NCI-H727: a lung cancer,

With regard to the Fibroblast cell line, a selective and a high cytotoxicity on any of these cancer cell lines would give useful indication to go on further for new anticancer compounds. In an other way, absence of cytotoxicity, particularly on HaCat and Fibroblasts cell lines would open the way for a safe use of these compounds provided that further toxicological tests will validate this. To date, lack of evidence about cytotoxicity of mycosporine-like compounds on

HaCat and Fibroblasts cell lines [158], [159], [160].

Following this cytotoxic assesment, we focused on photoprotective activities of the three isolated mycosporines and lichen aq. extracts containing such compounds. It is widely accepted that the primary role of mycosporine-like compounds is to serve as sunscreens which confer

127

protection against harmful exposure to UV radiation [2], [44], [161]. Indeed, these mycosporines which possess strong absorption at 310 nm with high molar extinction coefficients in the UVB range (290-320 nm) suggest their potential interest as natural photoprotectants.

� We firstly determined the absolute (UV-PF, critical wavelength λc, UVA-PF) and relative indexes (SUI- Spectral Uniformity Index and ISP-Ideal Spectral Profile) of the three available mycosporines along with four lichen aq. extracts (crude and semi- purified aqueous extracts of D. luridum and D. miniatum) to predict their photoprotective activity.

� The Photo-Irritancy Factors (PIF) of mycosporines and extracts were then determined on HaCaT cells. In fact, it is well evidenced through clinical and experimental studies that non-cytotoxic doses of many chemicals could cause

phototoxic responses when exposed to non-phototoxic doses of UV radiation [162], [163].

� Modern sunscreen products frequently combine UV filters with one or more biologically active molecules, a popular example corresponding to antioxidants added in sunscreen products [164]. Up to now, most of mycosporine-like compounds tested for their antioxidant activities are either pure imino-mycosporines or a mixture of several imino-mycosporines. Oxo-mycosporines are less reported, only mycosporine glycine and mycosporine glutaminol glycoside have been tested yet (Appendix 4). So, the antioxidant properties of the three pure oxo-mycosporines along with lichen aq. extracts were firstly investigated under two in vitro assays DPPH and NBT. Samples with interesting antioxidant properties were then selected for their total antioxidant activities with an in cellulo assay. The principle is the determination of malondialdehyde (MDA) level after incubation with ethanol induced oxidative stress and compound on rat hepatoma/human fibroblast hybrid cell line WIF-B9.

� Finally, we evaluated photodegradation of the mycosporines and extracts under UV exposure. Indeed, it was considered that the degradation of compounds following solar irradiation will lead to a reduction in its activity against skin damage caused by UV light. Additionally, this can generate possible hazardous compounds.

128

1 Cytotoxic activity

The cytotoxic activity of six compounds, including compounds 2, 3, 4, 6, 7 and 8, was performed on the ImPACcell Plateform at the University of Rennes1. In a first step, compounds were tested on eight cell lines at a unique concentration of 25 µM. Then for the most promising

compounds, their IC50 were determined by testing the compound at various concentrations.

Cytotoxic activity

140 120 100 80 60

Cell % viability 40 20 0 Huh7 Caco2 MDA-MB HCT116 PC3 NCI HaCat Fibroblast Cell lines 2 3 4 6 7 8

Figure 67: Cytotoxicity of compounds 2, 3, 4, 6, 7 and 8 against eight cell lines at C = 25µM

(Huh7: human hepato cellular carcinoma, CaCo-2: human colorectal adenocarcinoma, MDA-MB-231: human breast cancer, HCT116: human colon cancer, PC3: human prostate cancer, NCI-H727: human lung cancer, HaCaT: human immortalized keratinocytes, Fibroblast: cell which synthesize the extracellular matrix and collagen)

Most of tested compounds are inactive against the eight cell lines at concentration 25 µM since the percentage of cell viability was over 80%, except cerevisterol 7 (Figure 67).

The cytotoxicity of compound 7 was further studied by determing the IC50 on the eight cell

lines. The compound 7 exhibited no IC50 on human breast cancer line MDA-MB-231 (IC50 >

25 µM) but IC50 values range from 5 to 24 µM on other cell lines. A selectivity index (SI) could

be calculated as a ratio between the IC50 calculated on normal cell lines (Keratinocytes and

Fibroblasts) and the IC50 found for cancer cell lines. Higher is the SI, higher the compound is safe for cells. The better SI was obtained from the human colon cancer HCT 116 relative to Fibroblast (SI = 4.8) while it was weak for the other cancer cell lines (Table 12). In respect to

the NCI criteria (IC50 < 20 µM) [165], cerevisterol 7 is a valuable compound even its selectivity is weak (SI < 5).

129

Table 12: IC50 values for compound 7 on the eight cell lines and selectivity indexes (SI) relative to HaCaT and Fibroblast

Cell lines Huh 7 CaCo2 MDA-MB HCT 116 PC3 NCI HaCaT Fibroblast

IC50 (µM) 10 18 >25 5.0 11 16 14 24 SI (HaCaT) 1.4 0.8 - 2.8 1.3 0.9 1.0 0.6 SI (Fibroblast) 2.4 1.3 - 4.8 2.2 1.5 1.7 1.0

The growth inhibition activity of cerevisterol toward the tumor cell lines of a human lung adenocarcinoma (A549) and a human osteosarcoma (MG63) was previously evaluated [138].

It displayed a weaker activity than that of positive control (Adriamycin) with IC50 25.8 and 2.8 µM for A549; 37.5 and 3.4 µM for MG63, respectively. Our results confirm cytotoxic activity of compound 7 on NCI-H727 (human lung cancer cell line) with IC50 16 µM.

Both mycosporines 2 and 3 exhibit non cytotoxic activities on HaCaT and Fibroblast cells. Recently, the protective effects of mycosporine-like compounds against UV-induced on human fibroblast (WI-38), HaCaT and mouse fibroblasts (3T3) cells were indeed observed [166]. These results supported their ability to be used as photoprotective agents and prompted us to further evaluate their photoprotective properties.

2 Photoprotective activity

Acute UV radiation can cause harmful effects to the skin such as immediate sunburn or photocarcinogenesis, photoimmunosuppression and photoaging in longer term [167]. Indeed, when UV radiation reaches the skin, some radiation is reflected away from the surface. The remaining radiation is scattered into the tissues just beneath the skin's surface. The more energetic UVB (290–320 nm) are absorbed by the epidermis causing acute sunburn, DNA mutation or even cancer. Although less energetic, the longer wavelengths in the UVA range (320–400 nm) can penetrate much deeper into the skin. UVA reaches the dermis where it is responsible for the formation of ROS which can cause damage to cellular proteins, lipids, and carbohydrates. Like UVB, it can cause structural damage to the DNA, impair the immune system, and possibly lead to cancer (Figure 68). Due to the harmful effects of UV radiation to the skin, it is necessary to discover and to develop new sunscreens, particularly UVA protectants which have been neglected as UVA radiations are ony considered as harmful since the early 21st century [168]. Recently, MAAs have been used in pharmaceutical and cosmetic

130

applications such as the product Helioguard 365 (a mixture of MAAs isolated from the red alga Porphyra umbilicalis, containing two imino-mycosporines porphyra-334 and shinorine, in a ratio of 11.5:1) [166].

Figure 68: Cross section of the skin and penetration of sunlight radiation into the tissues

2.1 UV-filters

Our study focused on the photoprotective capacities of the three mycosporines 1, 2 and 3 as well as four aq. extracts which contain the two genuine mycosporines 1 and 2, namely DL1, DL2, DM1 and DM2. In this, DL1 (resp. DM1) corresponds to the crude aq. extract of D. luridum (resp. D. miniatum), whereas, DL2 (resp. DM2) corresponds to the semi-purified aq. extract of D. luridum (resp. D. miniatum). The semi-purified aq. extracts were obtained by Dowex 50W-X8 (H+ form) resin chromatography (presented in section 2.1.2, chapter 2).

For this study we use a new method to be published and developed in the laboratory [169]. Photoprotective activities are evaluated through the calculation of the absolute (UV-PF, λc, UVA-PF) and relative indexes (SUI, ISP) by recording the UV spectrum of an oil/water (O/W) emulsion containing 10 % of the product (Table 13). Three commercial UV filters 4-

131

Methylbenzylidene Camphor (4-MBC), Avobenzone and Octocrylene are used as positive controls.

According to the values obtained, the products are divided into five categories: 1 strict UVA filter, 2 UVA booster, 3 UVA+UVB filter, 4 strict UVB filter and 5 rejected.

• If UV-PF > 1.5, λc > 370 nm, UVA-PF > 2, products are suggested as strict UVA filters. • If UV-PF > 1.5, λc > 370 nm, UVA-PF < 2, products are suggested as UVA boosters. • If UV-PF > 1.5, λc < 370 nm, UVA-PF < 2, SUI > 1.2, ISP < 90, products are suggested as UVA+UVB filters. • If UV-PF > 1.5, λc < 370 nm, UVA-PF < 2, SUI < 1.2, ISP > 90, products are suggested as strict UVB filters. • If UV-PF < 1.5, products are suggested to be rejected.

In order to identify the similarity between tested samples from their absolute and relative indexes and also to identify the relationships between these values, a Principal Component Analysis (PCA) was performed on the data given in Table 13. Two dimensions of the PCA captured 93.08 % of total variance (Dim 1 and Dim 2 accounted for 67.06 % and 26.02 %, respectively). The samples which have high (or low) coordinates on a dimension take high (or low) value for the variables which are highly correlated to this dimension. The results point out that the three compounds and extract DM2 could be considered as strict UVB filters and their profile fitted well with those of 4-MBC while three extracts (DM1, DL1, and DL2) are rejected (Figure 69).

The DM2 extract is shown to be a better UVB filter compared to DM1, DL1, and DL2. We propose that its mycosporines contents may be involved. Therefore, the total concentration of mycosporines, including mycosporines 1 and 2, in each extract was determined by using HILIC-HPLC-DAD conditions as reported above in section 3.2.5, chapter 2. As expected, DM2 possess a higher content of mycosporines (68.22) when compared to DM1 (46.04), DL1 (22.59), and DL2 (37.46 mg/g dw. crude aq. extract) (Figure 70).

The accumulation of such metabolites seem to correlate with UV exposure. In fact, D. miniatum was collected on dry rocks surfaces where they were exposed to higher levels of UV

132

radiations while D. luridum was collected on rocks standing in watercourses which periodically inundate the lichen.

Table 13: Absolute, relative indexes and predicted results of the aq. extracts along with the three mycosporines comprared to three commercialized UV filters

Extract/ Mycosporine UV-PF λc UVA-PF SUI ISP DL1a 1.1 374 1.0 1.8 59 DL2a 1.1 365 1.0 1.7 66 DM1a 1.1 375 1.0 1.7 62 DM2b 1.6 332 1.1 1.1 96 1 b 10.9 347 2.9 1.0 111 2 b 6.1 327 1.3 1.0 110 3 b 5.5 327 1.3 1.0 110 4-MBCc 3.5 324 1.2 0.9 117 Avobenzoned 1.8 377 3.9 2.3 47 Octocrylenee 7.4 336 1.7 1.3 84

DL1: crude aq. extract, DL2: semi-purified aq. extract of D. luridum DM1: crude aq. extract, DM2: semi-purified aq. extract of D. miniatum 1: mycosporine glutaminol, 2: mycosporine glutamicol, 3: ethyl ester of mycosporine glutamicol 4-MBC: 4-methylbenzylidene camphor, UV-PF: UV Protection Factor, UVA-PF: UVA Protection Factor

λc: critical wavelength, SUI: Spectral Uniformity Index, ISP: Ideal Spectral Profile a: rejected (UV-PF < 1.5) b: strict UVB filter (UV-PF > 1.5, λc < 370 nm, UVA-PF < 2, SUI < 1.2, ISP > 90) c, d, and e: commercial UVB, UVA and (UVA+UVB) filters, respectively

Figure 69: Principal component analysis (PCA) of the tested compounds and extracts discriminating UV absorbing properties

133

Mycosporine contents 80,0 70,0 60,0 50,0 40,0 30,0

mg/g mg/g dw. extract 20,0 10,0

0,0 DL1 DL2 DM1 DM2

Figure 70: Mycosporine contents in crude aq. extract and semi-purified aq. extract of D. luridum and D. miniatum

DL1: crude aq. extract, DL2: semi-purified aq. extract of D. luridum DM1: crude aq. extract, DM2: semi-purified aq. extract of D. miniatum

Previously, Collemin A, an oxo-mycosporine isolated from the lichenized ascomycete Collema cristatum, partially prevented pyrimidine dimer formation and completely prevented UVB induced erythema when applied (10% concentation) on the skin before an erythema induced [61]. We conclude that the three mycosporines and the semi-purified aqueous extract DM2 can be considered as eligible UV filters. Incorporating them into an emulsion according to the cosmetic guidelines [170] would be of interest to confirm their protection towards UVB radiations.

2.2 Cytotoxic and phototoxic activities

According to clinical and experimental studies, non-cytotoxic doses of many chemicals could however induce phototoxic responses when exposed to non-phototoxic doses of UV radiation [162], [163]. To determine the safety before and after UVA irradiation of the aq. extracts and the three mycosporines, the Neutral Red Uptake phototoxicity test was performed. It allowed the calculation of a ratio called Photo-Irritancy Factors (PIF). A PIF > 5 means that the compound is phototoxic as did chlorpromazine used as positive control in this experiment (Table 14).

134

Table 14: Cytotoxicity and phototoxic activities of aq. extracts and mycosporines

Extract/ Phototoxic activities on HaCaT cells Photo-Irritancy Factor Mycosporine (PIF) IC50 ± SD (µg/mL) Without irradiation With irradiation DL1 > 40 > 40 *1e DL2 > 40 > 40 *1 e DM1 > 40 > 40 *1 e DM2 > 40 > 40 *1 e 1 > 40 > 40 *1 e 2 30.00 ± 6.00 27.00 ± 4.00 1.11 f 3 > 40 > 40 *1e Chlorpromazined 22.00 ± 0.50 3.50 ± 0.00 6.29 f

Except a slight PIF= 1.1 calculated from IC50 = 30 and 27 µM (-UV and +UV) on HaCaT cells for compound 2, no cytotoxicity with or without UVA irradiation could be observed for the other tested compounds and extracts. These PIF values <<5 suggest these lichen extracts and mycosporines as non-phototoxic.

2.3 Antioxidant activity

Nowadays, more and more antioxidant ingredients are found in sunscreens because they scavenge the ROS generated through UVA and UVB radiations [44], [171], [172]. Mycosporine-like compounds have been well recognized for their high antioxidant activity, scavenging superoxide anions and inhibiting lipid peroxidation [78]. Mycosporine-like compounds tested for their antioxidant activities are either pure imino-mycosporines or a mixture of several imino-mycosporines while oxo-mycosporines are less reported (Appendix 4). So, the antioxidant activities of the three oxo-mycosporines 1, 2, 3 and lichen aq. extracts which contain 1 and 2 are investigated firstly using colorimetric assays, the 1,1-diphenyl-2- picrylhydrazyl free radical (DPPH) and superoxide anion scavenging activity (NBT) tests. Mycosporine 3 and the semi-purified aq. extract of D. miniatum which exhibit high antioxidant activity in NBT assay are now evaluated in cellulo (WIF-B9 cell line) by using ethanol-induced lipid peroxidation assay.

135

2.3.1 DPPH and NBT assays

Both assays DPPH and NBT are colorimetric assays based on the decolorization of a former solution. Higher is the decolorization, higher is the antioxidant efficacy of the tested solutions. Several concentrations of the aq. extracts and compounds were tested in a 96-wells plates comparatively to the standard quercetin. An IC50 value could be calculated when a 50% decolorization or more is induced by a compound.

The DPPH assay is based on an electron transfer (ET) mechanism and reveals the in vitro reducing power while the superoxide anion scavenging assay (or NBT assay) is a specific assay which reveals the ability of a compound to quench the superoxide anions.

� DPPH results

According to the DPPH assay, the semi-purified aq. extracts DL2 and DM2 show a slightly stronger activity than DL1 and DM1 (the crude aq. extracts), while the three

mycosporines display a weak electron transfer activity; hence, no IC50 value can be calculated (Figure 71).

� NBT results

As shown in Figure 71, at a concentration 80 µg/mL, the semi-purified extracts DL2 and DM2 having a higher mycosporine concentration (Figure 70) scavenge respectively 2-fold and

4-fold more superoxide anion than the crude extracts DL1 and DM1. DM2 (IC50 = 5.20 ± 2.20

µg/mL) is more active than DL2 (IC50 = 11.00 ± 5.00 µg/mL) which appears to correlate with the mycosporine concentration. However, the DM1 mycosporines content is higher than that of DL1 and DL2, but it exhibits a lowest antioxidant activity. Experiments with DM1 should be repeated and attention should be paid on mycosporine ratios and to the other accompanying compounds in each extract.

The activity of the pure mycosporines 1 and 2 are found moderate (no IC50 could be determined) suggesting a synergistic interaction among several components in the extracts DL2 and DM2. Esterification of the mycosporine glutamicol carboxyl group results in a better scavenging activity (IC50 = 4.00 µg/mL) of compound 3 compared to the standard quercetin

(IC50 = 6.75 µg/mL) towards the superoxide anion.

136

100

inhibition% 40

0 DL1 DL2 DM1 DM2 1 2 3 Quercetin

DPPH (400 µg/mL) Superoxide anion scavenging (80 µg/mL)

Figure 71: Antioxidant activity of the aq. extracts and three mycosporines 1, 2, 3 through DPPH and NBT methods compared to the positive control quercetin. Activity was expressed as inhibition % at the maximum dose (400 and 80 µg/mL, respectively)

DL1: crude aq. extract, DL2: semi-purified aq. extract of D. luridum DM1: crude aq. extract, DM2: semi-purified aq. extract of D. miniatum 1: mycosporine glutaminol 2: mycosporine glutamicol 3: ethyl ester of mycosporine glutamicol

2.3.2 Lipid peroxidation assay

According to previous antioxidant results, the DM2 extract and mycosporine 3 which possess high antioxidant activity are chosen for lipid peroxidation assay in hepatic cells (WIF- B9 cell line). This test has been performed in collaboration with Isabelle Gallais and Odile Sergent (UMR Inserm 1085, IRSET, University of Rennes 1).

So far, oxidative damage to lipids, proteins, and DNA after ethanol intoxication of the liver have been well established [173], [174]. In 2005, Odile Sergent et al proposed a model for the role of early ROS-induced fluidizing effect of ethanol in the amplification of oxidative stress (Figure 72). Briefly, in primary hepatocytes, ethanol metabolism can very rapidly lead to ROS production, which in turn increased membrane fluidity. This increase can promote the elevation of low molecular weight (LMW) iron content and therefore the enhancement of ROS production and lipid peroxidation to finally trigger cell death [175]. Here, we study on the effect of tested samples on ethanol-induced lipid peroxidation in the WIF-B9 cell line. Lipid peroxidation was evaluated by quantification of free MDA, a secondary end-product of

137

oxidized polyunsaturated fatty acid degradation, using size exclusion chromatography (SEC- HPLC) as a specific and sensitive method. Ethanol exhibited a prooxidant activity corresponding to an increase of free MDA recovery in the cells.

Figure 72: Proposed model for the role of early ROS-induced fluidizing effect of ethanol in the amplification of oxidative stress [175]

As shown in Figure 73, a 25 mM ethanol addition induces lipid peroxidation in WIF-B9 cell line measured through an increase in MDA content (variation of about 125 ng/mg protein) as observed in control.

The DM2 extract and compound 3 increase slightly the formation of free MDA in cells in absence of ethanol (about 25–35 ng MDA/mg protein). However, when the cells were treated with ethanol and the tested samples, compared to control: a decrease of 75 and 99 ng MDA/mg protein was observed for DM2 extract and compound 3, respectively.

138

Figure 73: Effect of DM2 extract (5.20 µg/mL) and mycosporine 3 (4.00 µg/mL ~ 12 µM) on WIF-B9 cell line by ethanol-induced lipid peroxidation assay (n=2)

Vitamin E is well known as a lipid-soluble chain-breaking antioxidant in human blood cells and rat liver [176]. Previous study showed that vitamin E at concentration 250 µM prevented the increase in membrane fluidity causing cell death due to ethanol [175], [177]. Our preliminary results suggest that the extract DM2 and mycosporine 3 possess a notable anti- lipoperoxidation activity. Interestingly, the tested concentration of water-soluble mycosporines 3 (12 µM) was much lower than vitamin E. We should make clear that the obtained results were duplicated; therefore, the experiments should be repeated for significant values.

In conclusion, the three mycosporines and the aq. extracts have been tested on three types of antioxidant experiments based on three different mechanisms: electron transfer (ET) mechanism through DPPH assay, superoxide anion scavenging through NBT assay and lipid peroxidation through the ethanol-induced peroxidation in cells. The mycosporine 3 was the most potent and a supposed synergism between the mycosporines 1 and 2 resulted in a noticeable antioxidant activity for the semi-purified extract DM2. The mechanism of this antioxidant activity appears to involve a specific activity on superoxide anion scavenging and lipid peroxidation. Similar results were obtained when looking for the antioxidant activity of mycosporine serinol isolated from the marine lichen Lichina pygmaea. This oxo-mycosporine exhibited a weak activity in DPPH assay and a moderate activity in NBT assay (IC50 = 350 µM compared to trolox IC50 = 15 µM) [178]. Similarly, mycosporine glycine displayed a moderate activity in the phosphatidylcholine peroxidation assay [58]. However, it did not exhibit antioxidant activity in Pyrogallol autooxidation based on ROS scavenging mechanism (at pH 8.2). A highest activity was however recorded in ABTS (2,2´-azinobis (3-ethylbenzothiazoline-

139

6-sulfonate)) assay based on ET mechanism (at pH 8.5) [171]. Thus, the antioxidant activity of compounds tested in different assays is difficult to be compared and real antioxidant effect in living cells needs a more sophisticated experiment.

However, our work supports the hypothesis that mycosporine-like compounds play a role as antioxidants, particularly as a mixture. Interestingly, antioxidant activities obtained in cellulo suggest that membrane transport is possible. As DM2 is more active and easier to obtain than the mycosporines which are costly and time consuming for isolation, a development of extracts, possibly enriched with chemically synthetic analogs is though to be priviledged option.

2.4 Photostability under UVA and UVB

Looking for UV protectants, photostability under UVA and UVB and low toxicity on cells before and after UVA irradiation are prerequisite. Aqueous extract DM2 and the three mycosporines at 40 µg/mL were irradiated under six UV-doses covering the UVA and UVB range. Quantification of the former compounds before and after UV irradiation was determined by using UV-HPLC and compared to those of the positive control trolox.

Byproducts of the positive control trolox appear at 0.5 J/m2 exposures and a 10% and 19% degradation rate are observed at 5 J/m2 exposures to UVA and UVB, respectively. Concerning DM2 and the three mycosporines, no degradation occurs even at 5 J/m2 (Figure 74). These results confirm the absence of photoproducts observed by Moliné et al. when mycosporine glutaminol glycolysed was subjected to UVB irradiation [179].

140

Figure 74: HPLC chromatograms of trolox (left) and mycosporine 1 (right) (A): without UV irradiation (B): at 5J/m2 of UVA irradiation (C): at 5J/m2 of UVB irradiation

141

3 Experimental

3.1 Cytotoxic activity

Codes and origins of the tested cell lines

- Huh7: human hepato cellular carcinoma - CaCo-2: human colorectal adenocarcinoma - MDA-MB-231: human breast cancer - HCT116: human colon cancer - PC3: human prostate cancer - NCI-H727: human lung cancer - HaCaT: human immortalized keratinocytes - Fibroblast: cell which synthesize the extracellular matrix and collagen

The cell lines were obtained from the ECACC collection. Skin diploid fibroblastic cells were provided by BIOPREDIC International Company (Rennes, France). Cells were grown according to ECACC recommendations. The toxicity test of the compounds on these cells was as follows: 2.103 cells/well for HCT116 cell line or 4.103 cells/well for the other cell lines were seeded in 96 well plates. After 24h of seeding, cells were exposed to the compounds at concentration 25µM. After 48h of treatment, the cells were washed in PBS and fixed in ethanol/acetic acid (90:5 v:v) for 20 min. Then, the nuclei were stained with Hoechst 3342 (Sigma). Image acquisition and analysis was performed using a Cellomics ArrayScan VTI/ HCS Reader (Thermo Scientific). When compounds exhibited a toxicity > 50%, a range of

concentrations was tested on cells and the IC50 are determined using Xlfit software.

3.2 UV filters

3.2.1 Emulsion preparation

The oil/water (O/W) emulsion was prepared by dissolving 10 g of a sodium dodecyl sulfate solution (SDS) 10% in 20 g of distilled water. After vigorous mixing in a blender, 10 g of liquid paraffin was incorporated to obtain a homogeneous emulsion.

142

3.2.2 Sample preparation

The lichen compounds and extracts were added to O/W emulsion at 10% (w/w) concentration. This percentage was corresponding to the maximum concentration authorized for many filters by the European regulation. A final test sample concentration of 20 µg/mL was observed in solution S4 by dilution in ethanol. Three commercial UV filters (4- Methylbenzylidene camphor (4-MBC), Avobenzone and Octocrylene were used as positive controls. The detail of the preparation was showed in Figure 75.

P* = 10: percentage’s value of filter in the emulsion

Figure 75: Solution preparation for UV-filters test

3.2.3 Measurements and calculations

Absorbance (Aλ) measurements between 290 and 400 nm were recorded using a double beam spectrophotometer. The experimental data were transferred into an Excel spreadsheet specially arranged for calculation. Transmittance (Tλ) was calculated from absorbance. The experimental values resulted directly in both UV-PF and UVA-PF values, the critical

wavelength (λc) [180], the UVA-PF/UV-PF ratio, the Spectral Uniformity Index (SUI) [181] and the Ideal Spectral Profile (ISP) [182]. UV-PF values are the means of different values calculated from (Eλ × Iλ ) relative to Mexico, Melbourne, COLIPA International 2006 and COLIPA UVA 2007. All tests were done in duplicate and the results averaged.

143

-T A(λ) = 10 (Eq. 10)

400

∑ IE ∆λλλ 290 (Eq. 11) UV − PF = 400

∑ TIE ∆λλλλ 290

400

∑ IE ∆λλλ 320 (Eq. 12) UVA − PF = 400

∑ TIE ∆λλλλ 320

Eλ is the spectral irradiation of terrestrial sunlight at λ, Iλ is the erythemal action spectrum at λ and Tλ is the spectral transmittance of the sample at λ.

= 0.9 A( )d (Eq. 13) λ C ∫ λ λ ∫A(λ)dλ/ ∫dλ UVA/UVB ratio= (Eq. 14) ∫A()λ dλ/ ∫dλ

380

∑ Aλ 290 (Eq. 15) SUI = 380

∑ |Aλ − A | 290

400 A Â ∑ |λ − λ | 290 (Eq. 16) ISP = 400 x100 A ∑ λ 290

where Aλ is the spectral absorbance at λ and Âλ is the ideal spectral absorbance at λ. Âλ is equal to the mean absorbance between 290 and 385 nm for all wavelengths in this spectral

interval. Between 385 and 400 nm. Âλ is given as:

400 − λ Aˆ = A x (Eq. 17) λ 385 15

PCA was performed on the data collected, by using R software.

144

3.3 Antioxidant activity

3.3.1 DPPH assay

Because of its stability in the free radical form, DPPH assay is often used to evaluate the antioxidant activity of natural compounds in an easy and rapid way [183]. The principle behind this assay is the color changes from purple to yellow of the DPPH solution because the radical is quenched by the antioxidant according to electron transfer (ET) mechanism [184].

A reaction mixture containing 100 µL of DPPH (0.5 mM) in methanol and 10 µL of the sample solutions in DMSO to give final concentrations of 400, 133.3, 44.4, 14.8 µg/mL per well, was distributed in each microplate well. The mixture was incubated at rt for 15 min. The decrease in absorbance due to DPPH was measured at 540 nm. Quercetin was used as a positive control on each plate. Each concentration and all tests were done in triplicate and the results averaged. The percentage inhibition at steady state for each dilution was used to determine the

IC50 values graphically [162].

3.3.2 NBT assay

NBT assay based on superoxide radicals scavenging mechanism. In this N N assay, phenazine methosulfate in the N N 2Cl N N N N presence of NADH under aerobic O2N NO2 conditions (+ O2) produces superoxide H3CO OCH3 anions (Figure 76) which are NBT appreciated by a color: NBT (Nitro Blue Tetrazolium chloride).

-. + NADH + O2 + O2

Phenazine methosulfate Figure 76: Reaction leads to the formation of superoxide anion

145

.- NBT belongs to the family of tetrazolium salts and can selectively react to ions O2 resulting in the form of insoluble formazan salt in water. It is then determined the concentration of the staining of formazan salts, visible at 560 nm, after addition of NADH and phenazine methosulfate in the absence and the presence of a tested substance. This coloration is .- .- proportional to the amount of anions O2 . If the substance chelates O2 anions, the color of the solution will be less intense than in the control (without substance). The lower in the violine color appears, the higher the antioxidant activity is.

For the same extracts and compounds, measurements of superoxide anion scavenging activity in 96-well microplates previously described with some modifications were performed [162]. The reaction mixture in the sample wells consisted of NADH (78 µM), nitro-blue tetrazolium (NBT) (50 µM), phenazine methosulfate (PMS) (10 µM), and lichen samples (80, 40, 20, 10, 1 µg/mL). The reagents were dissolved in 16 mM tris–hydrochloride buffer, at pH = 8 except for all the lichen samples which were dissolved in DMSO. After 5 min of incubation at rt, the spectrophotometric measurement was performed at 560 nm against a blank without PMS and sample. Quercetin was used as a positive control. Each concentration and all tests were done in triplicate and the results averaged. The percentage inhibition at steady state for

each dilution was used to determine the IC50 values.

3.3.3 Lipid peroxidation assay

� Tested samples

The semi-purified aq. extract of D. miniatum DM2 and mycosporine 3 were tested in lipid peroxidation with and without ethanol at concentration 5.20 and 4.00 µg/mL, respectively.

These concentrations correspond to their IC50 values in NBT assay.

� Cell isolation and culture

Lipid peroxidation was tested on WIF-B9 cells, a hybrid cell line obtained by fusion of rat hepatoma (Fao) and human fibroblasts (WI38). Some proteins of rat and human, especially in liver, such as albumin are co-expressed. However, the human phenotype is dominant over that of rat regarding metabolic pathways of biliary acids. These lines WIF-B9 cells have the enzymes necessary for the metabolism of ethanol.

146

Cells were seeded at density of 12500 cells per cm2 in petri dishes 60 mm diameter and cultured in a nutrient medium F12 Ham Coon's modified (Sigma-Aldrich, France) supplemented with 5% Fœtal Calf serum (SVF) (Gibco, France), 1% of sodium bicarbonate 22 g/L solution, 1% an antibiotic and antimycotic containing 10000 units per ml penicillin, 10 mg streptomycin and 25 µg amphotericin B (Sigma-Aldrich, France), 1% of glutamine 200 mM solution (Gibco, France) and 0.2% of hypoxanthine sodium (5 mM), aminopterin (20 µM) and thymidine (0.8 mM) (HAT) (Gibco, France) solution. The cells were kept at 37 °C in an

atmosphere of 5% CO2 and 95% air. The medium was changed 3 h after seeding and replaced with the same medium as above but deprived of serum.

The tested samples were added to the cultures at the final concentration 5.20 and 4.00 µg/mL for DM2 and compound 3, respectively. After a 1 h incubation time at 37 °C, some cultures were supplemented with 25 mM ethanol and untreated cultures were used as controls. Cultures were then incubated 48 h for free MDA analysis.

� Free MDA evaluation HPLC procedure [185]

HPLC condition: MDA quantification was performed on the HPLC system (Agilent 1260 Infinity) using a TSK-Gel G 1000 PW column 7.5 mm×30 cm (Tosoh Bioscience, Germany). The eluent was composed of 0.1 M disodiumphosphate buffer, pH 8. Flow rate: 0.7 mL/min, at ambient temperature. The absorbance was monitored at 267 nm using detector UV Agilent Infinity 1260 (Agilent Technologies, France). The automatic injector set at a volume of 250 µL. Each sample was analysed 100 min.

Preparation of free MDA standard: 10 µL of 1,1,3,3-tetramethoxypropane were hydrolysed in 10 mL of 0.1 N HCl during 5 min in boiling water. This solution was then diluted

5000 times in 0.01 M Na2HPO4 buffer pH 7.45, corresponding to a 1.22 µM or 87.8 ng/mL MDA solution. This solution was then ultrafiltered using nitrogen gas (3 bars) through Amicon 1000 daltons membrane before injection HPLC.

Preparation of the samples for HPLC analysis: free MDA was quantified from culture medium and from cells. The samples (culture media and cell homogenates) were filtered through 1000 daltons membrane ultrafilter (Millipore, France) in a 10 mL Amicon (U.S.A.)

147

cell pressurized at 3 bars with nitrogen gas. The filtrate was used for the HPLC procedure. All experiments were performed at least on triplicate cultures.

Determination of protein: protein content was determined on defrozen cell homogenates according to the Bradford reaction [186], using the Bio-Rad reagent, bovine serum albumin serving as standard and performed by a Cobas-Bio automatic analyser. Culture media were collected and cells were washed twice with 0.01 M phosphate buffer, pH 7.45 and resuspended in 1 mL of the same buffer. The cells were lysed using an ultrasonic homogenizer. An aliquot was stored at -17°C until the protein content was estimated.

Calculation of free MDA: quantity of free MDA in samples was calculated by ng of MDA / mg of protein in a flask, followed by Eq.:

Asample ×88×V×1000 (Eq. 18) Astandard×Prot.

Asample: peak area of MDA in sample 88: 87.8 ng of MDA / mL of medium V: volumn of medium in flask (mL) 1000: to convert quantity of protein from µg to mg

Astandard: peak area of MDA standard Prot.: total quantity of protein in sample (µg)

3.4 Cytotoxic and phototoxic activities

Cytotoxicity and phototoxic activities of the four extracts and three mycosporines was evaluated according to the OECD guideline 432 [187] with some modifications [162]. Briefly, 100 µL per well of a cell suspension of HaCaT cells (ATCC) at 8 × 105 cells/mL were 0 maintained in a RPMI culture medium with 5% of calf serum at 37 C under 5% CO2 for 24 h for formation of monolayers. Two 96-well plates per test chemical were preincubated for 1 h with a range of tested compound concentrations (0.01–200 µg/mL) and chlorpromazine (0.01– 100 µg/mL) as the positive control. One set of plates was irradiated (+UV) for 50 min with 5 J/m2 whereas another plate (-UV) was kept in the dark. In both plates the treatment medium (PBS) was replaced by a culture medium and after 24 h of incubation, cell viability was

148

determined by Neutral Red Uptake. Cell viability was expressed as percentage of untreated cell controls and was calculated for each test concentration.

To predict the phototoxic potential, we determined the photo-irritancy factor (PIF) depending on the concentration response curves obtained in the presence and in the absence of irradiation ICUV(− ) PIF = 50 (Eq. 19) IC50 () +UV

IC50: the concentration reducing cell viability to 50% compared to the untreated controls. If a chemical is only cytotoxic in the presence of UV (+UV) and not cytotoxic when tested without irradiation (-UV), the PIF is calculated as followed formula: C (−UV) PIF = max (Eq. 20) IC50 () +UV

If the value PIF < 5, no phototoxic potential is predicted.

If both, IC50 (-UV) and IC50 (+UV) cannot be calculated due to the fact that a chemical does not show any cytotoxicity up to the highest tested concentration, a formal PIF (PIF = *1) is used predicting no phototoxic potential. ICUV(− ) PIF = * 1 = 50 (Eq. 21) IC50 () +UV

3.5 Photostability under UVA and UVB

Extracts and compounds were dissolved in 1 mL of distilled water and were then added to petri dishes containing also 2 mL of distilled water to make a final concentration of 40 µg/mL. Then, after removing the lids, petri dishes were exposed to UVA or UVB at several doses of irradiation: 0.5 J/m2, 1J/m2, 2J/m2, 3J/m2, 4J/m2, and 5J/m2. The source of UVA light (320-400 nm) and UVB light (290–315 nm) were delivered by 5 bulbs of 15 watts each and the internal UV sensor measures the UV light emitted by each UV bulb. Before the first and after each irradiation, 80 µL of the samples were collected and stored in a vial at 0°C until analysis. Trolox was used as a positive control [188]. Thirty µL of irradiated solution was

149

subjected to HPLC to measure the extent of degradation by a decrease in peak area at 290 nm for Trolox and 310 nm for mycosporines. For Trolox: HPLC conditions using a column Kinetex C18 100 Å (2.6 µm, 100 × 4.60 mm) with isocratic elution 50:50 v/v of solvent A (water + 1% acetic acid) and solvent B (ACN) flowing at 0.5 mL/min. For mycosporines: HPLC conditions using the same conditions of mycosporine quantification experiments (see section 3.2.5, chapter 2). All tests were done in duplicate and the results averaged. Percentage of residue is calculated according to (Eq. 22 and Eq. 23)

Ax Atotal x % Residue (Trolox)= × 100 (Eq. 22) A0 A total 0

A %Residue (Mycosporines)=x × 100 (Eq. 23) A0

2 Ax: peak area at x J/m 2 A0: peak area at 0 J/m 2 Atotal x: peak area of Trolox and byproducts at x J/m 2 Atotal 0: peak area of Trolox and byproducts at 0 J/m

150

Discussion–Conclusion

The intriguing chemical profile of the aquatic river lichen Dermatocarpon luridum corresponding to a chlorolichen but not exhibiting the usual phenolic compounds found in lichens associated with green algae prompted us to carry on a phytochemical study on this lichen and related species. Among the polar compounds, we found the characteristic UV shape of mycosporines which are recognized to be natural UV filters. The preserved cyclohexenimine or cyclohexenone skeletons corresponding to iminomycosporines and oxomycosporines respectively are distributed in a limited number of organisms (some mushrooms, cyanobacteria, algae) and correspond to some dozens of structures described to now [49].

While mycosporines have been yet only described in cyanolichens [3], our work first demonstrates the occurrence of mycosporines in chlorolichens [189]. This finding has been confirmed on a variety of Dermatocarpon luridum samples and extended to ten additional related taxa. By using HPTLC-UV and HPLC-DAD-MS methods, two oxo-mycosporines i. e. mycosporine glutaminol 1 and mycosporine glutamicol 2 were characterized in aq. extracts of the four fresh Dermatocarpon species (D. luridum, D. meiophyllizum, D. leptophyllodes, and D. miniatum) collected in Brittany. This was the first description of mycosporine glutaminol in lichens as mycosporine glutamicol was only reported in a cyanolichen: Degelia plumbea belonging to Collemataceae. An additional ethyl derivative was also identified in some extracts but found to be an artifact obtained during the purification process, possibly from mycosporine glutamicol. Mycosporine levels in Dermatocarpon appear to be influenced by UV radiation intensities since the higher mycosporines contents are found in species more exposed to direct solar radiations (D. miniatum is at least two fold more accumulating mycosporines than D. leptophyllum and D. luridum). However, this observation needs to be validated by a designed experiment and statistically supported results.

Thanks to the opportunity given by a simple protocol developed for the dosage of these mycosporines from very small lichen amounts, a quantification of such metabolites on a large Dermatocarpon panel (89 samples) could be extended to a variety of geographic areas, as we included samples from personal and institutional herbaria collections. The London Natural History Museum (BM) and the Des Abbayes (REN Abb) and L.J.-C. Massé lichen herbaria were the major providers for samples mainly collected in Europe and in North-America, the oldest tested specimen being collected in 1841. From 15 mg of this lichen sample, 14 mg of

151

mycosporine glutamicol per gram of dry lichen aq. extract were quantified, corresponding to 0.4 mg mycosporine glutamicol per gram of dried lichen. Except some samples, the dozen of Dermatocarpon species represented in this panel show a total content of mycosporines in a

range 4.56–46.77 mg/g dry lichen aq. extract, corresponding to 0.1–1.4 mg per gram of dried lichen. Analysis of the duplicate collections kept under similar storage conditions reveals a similar mycosporine content even after decades of storage. It is noteworthy that mycosporine glutaminol is found in fresh collections and not detected in all pre-2000 collections whereas mycosporine glutamicol accumulates in most of the investigated samples. These results also confirm the mycopsorine glutamicol ethyl ester to be an artefact as non present in any sample and the instability state of mycosporine glutaminol droping down within approximately 15 years in the dried herbarium samples. A long time storage condition appears to be involved in its degradation or transformation to mycosporine glutamicol. At the opposite, mycosporine glutamicol appears to be a very long lasting compound and a chemical marker of Dermatocarpon samples kept in herbarium conditions. Although it is still found in a 173 years old lichen sample, it was difficult do draw kinetics for stability or degradation as we didn’t know the initial mycosporine content from the herbarium samples. Moreover, they were kept in different conditions but our work gives a first mark which allows further analysis to be compared in some years on selected samples. Additional sampling of the four Dermatocarpon species found in the same area will be scheduled to have the suitable material amount for and also to look for seasonal possible influence on the mycosporine content.

To now, the mycosporine screening is quite limited as not systematically carried on lichen species, particularly because the focus on more apolar and abundant lichen phenols which are also considered to be useful chemotaxonomic markers. Moreover isolation of these polar compounds mixed to amino-acids and sugars is tricky and often disappointing as they are easy to visualize through UV irradiation but in low amounts. Isolation and structure identification with spectrospcopic analysis ascertained the structure of two genuine oxomycosporines 1 and 2 from the aqueous extract of D. luridum. As HPLC UV characterisation could lead to misidentification with such compounds, comparison with these standards and MS-hyphenation is a robust support for identification of these compounds in fresh and dried specimens. In this case the protocol was optimised to avoid artifact generation as the ethyl ester of mycosporine glutamicol 3 obtained in the separation process. However,

152

this compound was unprecedently reported and was of intererest to be tested in our biological tests.

Regarding the described occurrence in lichens, the Verrucariales order is now to be added to Peltigerales, Lichinales and Lecanorales orders in which mycosporines had been characterised in some lichen species. Iminomycosporines are suspected to occur in tripartite Stereocaulon lichens [190] but all mycosporines identified to now in lichen possess a cyclohexenone pattern which is characteristic of fungal mycosporines and have maximum UV absorption around 310 nm. The partner involved in this biosynthesis is not yet clearly identified but it is likely to be related to the fungus instead to the alga. As a support, mycosporine glutaminol and mycosporine glutamicol along with their glycosylated derivates had been described in non-lichenised Ascomycetous fungi [38], [90], [91], [94]. Considering mycosporine glutaminol has also been detected in the terrestrial cyanobacteria Leptolyngbya sp. [46] and mycosporine glutamicol in the cyanolichen Degelia plumbea [3], a microscopic observation of the Dermatocarpon samples do not revealed any significant presence of cyanobacteria. Further experiments have to be undertaken and particularly looking for the presence of DHQ Synthase genes recently characterised to be involved in mycosporine biosynthesis [54]. We can also look for mycosporines in free Diplosphaera samples along with other lichen specimens containing this phycobiont. Although it will be interesting to extend the mycosporine screening to a wider panel of Verrucariales, these lichens should be chosen out from the Verrucariales order, however Diploshaera seems to be strictly related for symbiosis to this order [191]. A discussion was recently opened from phylogenetic studies carried on photobionts of Dermatocarpon luridum and related species. Diplosphaera chodatti is confirmed to be the algal partner of D. miniatum, D. arnoldianum and D. luridum var. luridum while the D. rivulorum is placed in a separated clade both from fungal and algal phylogeny data [4]. The question is still pending [192] but the mycosporine profile found in our study does not allow to chemotaxonomically distinguish these lichen species. Looking for a higher classification level, mycosporine glutamicol could be a taxonomic marker of Dermatocapon and possibly of Verrucariales as mycosporine serinol is mainly found in Peltigerales (8 Peltigeraceae) and mycosporine glycine in Lichinales (7 Lichinaceae and Peltulaceae species) (see Table 1 p. 25). However exceptions limit full discrimination with these compounds as mycosporine glutamicol is also described in Degelia plumbea (Peltigerales, Pannariaceae) and mycosporine glycine in Collema coccophorum (Lecanorales, Collemataceae).

153

Phytochemical study on the lichen D. luridum collected in Brittany led to the isolation and identification of ten compounds. A novel combination for mycosporine purification including HILIC flash chromatography, followed by open reverse phase chromatography, and semi-preparative HPLC-DAD were applied on aqueous extract to isolate a dozen of mgs of the two genuine mycosporines 1 and 2 along with an artefact corresponding to the ethyl ester of mycosporine glutamicol 3 which structure has not been described yet. Mycosporine 1 and 3 appear to be rather weak bases or strong acids with pKa values at 1.29 and 1.13, respectively. Therefore, only the neutral form exists and responds to the UV absorption in water at pH 7.

Three other polar compounds were isolated from aqueous extract, including 2-amino-3- acetylaminopropionic acid 4, L-proline 5, and γ-L-glutamylglycine 6. From chloroform and acetone extracts, four compounds cerevisterol 7, a ceramide 8 and a mixture of D-volemitol 9 and D-mannitol 10 (with a ratio 85:15) were also obtained. Originally, compounds 4, 6 and 8 have not been mentioned in lichens so far and compound 7 has been described for the first time in D. luridum. The ceramide 8 is the first non glycosylated representative described in lichens and its role in Dermatocarpon remains to be elucidated. Ceramides are generally intermediates for more complex forms and involved in membrane stabilisation or as signalling mediators. It is likely ceramides correspond in the lichen to water-impermeable components preventing excessive water loss, helping the tissues to cope with dehydration-rehydration cycles as well as participating as barrier against the entry of microorganisms.

Interestingly, D. luridum appears to be very poor in polyphenolic lichen constituents. The metabolic profile of such a chlorolichen is more related to the ones observed in cyanolichens wich are mainly found to accumulate polysaccharides and nitrogen-containing compounds such as amino acids and mycosporines. As very poor information regarding the chemistry in Verrucariales is available, an extensive profiling of these secondary metabolites is though to be informative. Beside the challenge of discovering new structures and bioactive compounds, it could be useful for taxonomy of this order which is still discussed by specialists [193]. These studies are mainly base on morphological, genetic and ecological criteria and chemistry has been ignored since the lichen profile differs from the usual phenols giving indicative spot test reactions.

As some compounds were isolated in a suitable quantity for biological testing, cytotoxicity of six isolated compounds 2, 3, 4, 6, 7 and 8 were evaluated on six cancer cell lines: Huh7,

154

CaCo-2, MDA-MB-231, HCT116, PC3, NCI-H727, including the Keratinocyte HaCaT and Fibroblast cell lines which are informative for skin tolerance. Most of the tested compounds are not toxic at a 25 µM concentration, except a steroid, cerevisterol 7. Compound 7 is toxic

on seven cell lines with IC50 values range from 5 to 24 µM, except the human breast cancer cell line MDA-MB-231.

Absence of skin toxicity is a favourable result for compounds possibly used for a photoprotective activity. Mycosporine 3 and the semi-purified aqueous extract of D. miniatum, containing a mixture of mycosporine glutaminol and mycosporine glutamicol in a ratio 1.1:1.0, are sufficiently stable and do not induce any phototoxicity on HaCaT cells exposed to UVA radiation having also a good photostability under UV exposure. So they have the main requirements to be developed further as suncare products. However, synthesis or biotechnical approach should be considered as a privileged option as the lichen source cannot be a envisaged in a commercial way. Moreover, a combination of compounds appear to be more efficient than isolated compound but the optimum mixture and ratio remain to be studied. Antioxidant activities obtained with these compounds also appear to be synergistic and in cellulo confirmation for lipid peroxidation of mycosporine 3 is expected with high interest as this is unusal to obtain such results with hydrophilic compounds.

Wathever the antioxidant and photoprotective tests performed, we assume performance of these compounds in lichens is considerably higher as a multiplicity of regulating systems with synergistic mechanisms are involved, as proved by experiments carried on by Kranner et al [194]. Beside the mycosporine photoprotective and antioxidant effects the protective effect of the xanthophyll cycle is certainly to be added as observed by Clother Coste in inundate lichens [109], particularly the involved carotenoids pigments have been described in Dermatocarpon species collected in Anatolia.

As a personal feeling about this work, I realized experiment overcomes any dogma as just an example is sufficient to balance a general trend. Production of mycosporine in lichens was associated till now to cyanolichens only [3], [61], [114] but our results allowed to revise the concept [189]. An extensive work in the way given by the biogenetic studies carried on mycosporines has to be carried on with lichens. The DHQS and O-MT genes recognized to be involved in the oxo-mycosporine synthesis [54] have to be confirmed in Dermatocarpon species. In the same way, additidional lichens have to be studied in this genome mining

155

approach and particularly tripartite lichens forming cephalodia where the distribution of mycosporines in the lichen thallus would be informative, both for biogenis and for a better understanding of their role in lichens. As noticed by Clother Coste in his PhD dissertation on aquatic lichens [109] many has to be discovered with these poorly studied orgamisms. This is also striking, that most of the organisms producing mycosorines are water-related organisms, rising up an evolutionary question about the terrestrial adaptation of marine species and about the metabolism shift between phenol producing organisms and other organisms.

Moreover, this work was a great opportunity to illustrate how an interdisciplinaty cooperative work is necessary for science because our initial entrance through phytochemistry was supported by a variety of skills. We are indebted to numerous scientists for kind and priceless help from botany to biology. Extention of the work from collected samples in the field to historical herbarium lichen collection was an exciting challenge. The value of these collections for science is highly underestimated. To our knowledge, few dosages have been carried on lichen metabolites finding four reports from 1986 [195], [196], [197], [198]. We have to deal with small amounts of these priceless samples. Beside the taxomic major interest and main purpose of these herbarium collections, some examples of their contribution in lichenology are reported in the field of environmental surveys [199], [200], [201], of genetic [202], [203] and chemical analysis [204], [205], with some recent interesting reports testifying they are an effective system of conservation of secondary metabolites [206]. Although our work pointed out the time dependent lability of some compounds, we stress on the need to maintain and enrich such collections for future works, with extensive informations as the technical barriers are always overcame, opening unexpected ways for science contribution.

156

References

[1] M. Cardinale, A. M. Puglia, M. Grube, Molecular analysis of lichen-associated bacterial communities, FEMS Microbiol. Ecol. 57 (2006) 484–95. [2] K.-H. Nguyen, M. Chollet-Krugler, N. Gouault, S. Tomasi, UV-protectant metabolites from lichens and their symbiotic partners, Nat. Prod. Rep. 30 (2013) 1490–1508. [3] C. Roullier, M. Chollet-Krugler, E.-M. Pferschy-Wenzig, A. Maillard, G. N. Rechberger, B. Legouin-Gargadennec, et al., Characterization and identification of mycosporines-like compounds in cyanolichens. Isolation of mycosporine hydroxyglutamicol from Nephroma laevigatum Ach., Phytochemistry. 72 (2011) 1348–57. [4] K. M. Fontaine, A. Beck, E. Stocker-Wörgötter, M. D. Piercey-Normore, Photobiont relationships and Phylogenetic history of Dermatocarpon luridum var. luridum and related Dermatocarpon species, Plants. 1 (2012) 39–60. [5] V. Shukla, D. K. Upreti, R. Bajpai, Lichens to biomonitor the environment, Springer New Delhi Heidelberg New York Dordrecht London, (2014). [6] J. Boustie, M. Grube, Lichens-a promising source of bioactive secondary metabolites, Plant Genet. Resour. 3 (2005) 273–287. [7] http://www.lichens.ie/ [8] N. R. Amar, Cyanolichens: Nitrogen metabolism. In: Cyanobacteria in symbiosis, Kluwer academic, (2003). [9] C. Van-Haluwyn, J. Asta, Guide des lichens de France : lichens des arbres. Paris, (2009). [10] C. Deduke, B. Timsina, M. D. Piercey-Normore, Effect of environmental change on secondary metabolite production in lichen-forming fungi. In: International perspectives on global environmental change, Intech publisher, Europe, (2012). [11] D. Moore, Fungal morphogenesis, Cambridge University Press, New York, (1998). [12] S. Huneck, I. Yoshimura, Identification of lichen substances, Springer, (1996). [13] J. A. Elix, A catalogue of standardized chromatographic data and biosynthetic relationships for lichen substances. Third edition, Canberra, (2014). [14] J. A. Elix, E. Stocker-Wörgötter, Biochemistry and secondary metabolites. In: Lichen biology second edition, Cambridge University Press, (2008). [15] T. Rezanka, M. Temina, A. G. Tolstikov, V. M. Dembitsky, Natural microbial UV radiation filters- mycosporine-like amino acids, Folia Microbiol. 49 (2004) 339–52. [16] M. Millot, F. Di Meo, S. Tomasi, J. Boustie, P. Trouillas, Photoprotective capacities of lichen metabolites: a joint theoretical and experimental study, J. Photochem. Photobiol. B. 111 (2012) 17–26. [17] R. P. Rastogi, A. Incharoensakdi, UV radiation-induced biosynthesis, stability and antioxidant activity of mycosporine-like amino acids ( MAAs ) in a unicellular cyanobacterium Gloeocapsa sp. CU2556, J. Photochem. Photobiol. B Biol. 130 (2014) 287–292. [18] L. Verschooten, L. Declercq, M. Garmyn, Adaptive response of the skin to UVB damage: role of the p53 protein, Int. J. Cosmet. Sci. 28 (2006) 1–7. [19] J. Cadet, S. Courdavault, J.-L. Ravanat, T. Douki, UVB and UVA radiation-mediated damage to isolated and cellular DNA, Pure Appl. Chem. 77 (2005) 947–961. [20] R. P. Rastogi, R. P. Sinha, Apoptosis : molecular mechanisms and pathogenicity, Excli J. 8 (2009) 155–181. [21] J. A. D. Orazio, S. Jarrett, A. Marsch, J. Lagrew, L. Cleary, Melanoma- epidemiology, genetics and risk factors. In: Recent advances in the biology, therapy and management of melanoma, InTech, (2013).

157

[22] G. Singh, P. K. Babele, R. P. Sinha, M. B. Tyagi, A. Kumar, Enzymatic and non-enzymatic defense mechanisms against ultraviolet-B radiation in two Anabaena species, Process Biochem. 48 (2013) 796–802. [23] A. T. Banaszak, Photoprotective physiological and biochemical responses of aquatic organisms. In: UV Effects in aquatic organisms and ecosystems, The Royal society of chemistry, (2003). [24] R. P. Rastogi, R. P. Sinha, Solar ultraviolet radiation-induced DNA damage and protection / repair strategies in cyanobacteria, Int. J. Pharma Bio Sci. 2 (2011) 271–288. [25] J. B. Wittenberg, The source of carbon monoxide in the float of the Portuguese man-of-war, Physalia Physalis L., J. Exp. Biol. 37 (1960) 698–705. [26] H. P. James, H. S. Forrest, 310 mµ absorbance in Physalia physalis: Distribution of the absorbance and isolation of a 310 mμ absorbing compound, Comp. Biochem. Physiol. 30 (1969) 879–888. [27] I. Tsujino, T. Saito, Studies on the compounds specific for each group of marine algae. I. Presence of characteristic ultraviolet absorbing material in Rhodophyceae, Bull. Fac. Fish. Hokkaido Univ. 7 (1961) 49–57. [28] K. Shibata, Pigments and a UV-absorbing substance in corals and a blue-green alga living in the Great Barrier Reef, Plant Cell Physiol. 10 (1969) 325–335. [29] C. Leach, Ultraviolet-absorbing substances associated with light-induced sporulation in fungi, Can. J. Bot. 43 (1965) 185–200. [30] J. Favre-Bonvin, N. Arpin, C. Brevard, Chemotaxonomic research on mushrooms. XXXI. Structure of mycosporine (P 310), Can. J. Chem. 54 (1976) 1105–1113. [31] S. Ito, Y. Hirata, Isolation and structure of a mycosporine from the zoanthidian Palythoa tuberculosa, Tetrahedron Lett. 18 (1977) 2429–2430. [32] Y. Hirata, D. Uemura, K. Ueda, S. Takano, Several compounds from Palythoa Tuberculosa (Coelenterata), Pure Appl. Chem. 51 (1979) 1875–1883. [33] W. M. Bandaranayake, Mycosporines : are they nature's sunscreens?, Nat. Prod. Rep. 15 (1998) 159–172. [34] D. Karentz, Chemical defenses of marine organisms against solar radiation exposure: UV- absorbing mycosporine-like amino acids and scytonemin. In: Marine chemical ecology, CRC Press, (2001). [35] J. M. Shick, W. C. Dunlap, Mycosporine-like amino acids and related Gadusols: biosynthesis, acumulation, and UV-protective functions in aquatic organisms, Annu. Rev. Physiol. 64 (2002) 223–62. [36] R. P. Rastogi, R. P. Sinha, S. P. Singh, D.-P. Häder, Photoprotective compounds from marine organisms, J. Ind. Microbiol. Biotechnol. 37 (2010) 537–558. [37] A. Gröniger, R. P. Sinha, M. Klisch, D. P. Häder, Photoprotective compounds in cyanobacteria, phytoplankton and macroalgae-a database, J. Photochem. Photobiol. B. 58 (2000) 115–22. [38] R. P. Sinha, S. P. Singh, D.-P. Häder, Database on mycosporines and mycosporine-like amino acids (MAAs) in fungi, cyanobacteria, macroalgae, phytoplankton and animals, J. Photochem. Photobiol. B. 89 (2007) 29–35. [39] J. J. W. Won, J. A. Rideout, B. E. Chaiker, Isolation and structure of a novel mycosporine-like amino acid from the reef-building corals Pocillopora damicornis and Stylophora pistillata, Tetrahedron Lett. 36 (1995) 5255–5256. [40] J. J. Wu Won, B. E. Chalker, J. A. Rideout, Two new UV-absorbing compounds from Stylophora pistillata: sulfate esters of mycosporine-like amino acids, Tetrahedron Lett. 38 (1997) 2525– 2526. [41] M. Volkmann, A. A. Gorbushina, L. Kedar, A. Oren, Structure of euhalothece-362, a novel red- shifted mycosporine-like amino acid, from a halophilic cyanobacterium (Euhalothece sp.), FEMS Microbiol. Lett. 258 (2006) 50–54.

158

[42] L. Zhang, L. Li, Q. Wu, Protective effects of mycosporine-like amino acids of Synechocystis sp. PCC 6803 and their partial characterization, J. Photochem. Photobiol. B. 86 (2007) 240–5. [43] J. I. Carreto, M. O. Carignan, N. G. Montoya, Comparative studies on mycosporine-like amino acids, paralytic shellfish toxins and pigment profiles of the toxic dinoflagellates Alexandrium tamarense , A . catenella and A . minutum, Mar. Ecol. Prog. Ser. 223 (2001) 49–60. [44] A. Oren, N. Gunde-Cimerman, Mycosporines and mycosporine-like amino acids: UV protectants or multipurpose secondary metabolites?, FEMS Microbiol. Lett. 269 (2007) 1–10. [45] M. Volkmann, K. Whitehead, H. Rütters, J. Rullkötter, A. A. Gorbushina, Mycosporine- glutamicol-glucoside: a natural UV-absorbing secondary metabolite of rock-inhabiting microcolonial fungi, Rapid Commun. Mass Spectrom. 17 (2003) 897–902. [46] M. Volkmann, A. A. Gorbushina, A broadly applicable method for extraction and characterization of mycosporines and mycosporine-like amino acids of terrestrial, marine and freshwater origin, FEMS Microbiol. Lett. 255 (2006) 286–95. [47] K. Matsui, E. Nazifi, S. Kunita, N. Wada, S. Matsugo, T. Sakamoto, Novel glycosylated mycosporine-like amino acids with radical scavenging activity from the cyanobacterium Nostoc commune, J. Photochem. Photobiol. B. 105 (2011) 81–9. [48] E. Nazifi, N. Wada, M. Yamaba, T. Asano, T. Nishiuchi, S. Matsugo, et al., Glycosylated porphyra- 334 and palythine-threonine from the terrestrial cyanobacterium Nostoc commune, Mar. Drugs. 11 (2013) 3124–54. [49] S. La Barre, C. Roullier, J. Boustie, Mycosporine-like amino acids (MAAs) in biological photosystems. In: Outstanding marine molecules chemistry, Biology, Analysis, (2014). [50] J. Favre-Bonvin, J. Bernillon, N. Salin, N. Arpin, Biosynthesis of mycosporines: mycosporine glutaminol in Trichothecium Roseum, Phytochemistry. 26 (1987) 2509–2514. [51] A. Portwich, F. Garcia-Pichel, Biosynthetic pathway of mycosporines (mycosporine-like amino acids) in the cyanobacterium Chlorogloeopsis sp. strain PCC 6912, Phycologia. 42 (2003) 384– 392. [52] R. P. Sinha, N. K. Ambasht, J. P. Sinha, M. Klisch, D.-P. Häder, UV-B-induced synthesis of mycosporine-like amino acids in three strains of Nodularia (cyanobacteria), J. Photochem. Photobiol. B Biol. 71 (2003) 51–58. [53] J. M. Shick, S. R.- Lioud, C. F.-P. S, J.-P. Gattuso, Ultraviolet-B radiation stimulates shikimate pathway-dependent accumulation of mycosporine-like amino acids in the coral Stylophora pistillata despite decreases in its population of symbiotic dinoflagellates, Limnol. Oceanogr. 44 (1999) 1667–1682. [54] E. P. Balskus, C. T. Walsh, The genetic and molecular basis for sunscreen biosynthesis in cyanobacteria, Science. 329 (2011) 1653–1656. [55] B. Büdel, U. Karsten, F. Garcia-Pichel, Ultraviolet-absorbing scytonemin and mycosporine-like amino acid derivatives in exposed, rock-inhabiting cyanobacterial lichens, Oecologia. 112 (1997) 165–172. [56] A. A. Gorbushina, K. Whitehead, T. Dornieden, A. Niesse, A. Schulte, J. I. Hedges, Black fungal colonies as units of survival : hyphal mycosporines synthesized by rock-dwelling microcolonial fungi, Can. J. Bot. 81 (2003) 131–138. [57] R. P. Rastogi, A. Incharoensakdi, UV radiation-induced accumulation of photoprotective compounds in the green alga Tetraspora sp. CU2551., Plant Physiol. Biochem. 70 (2013) 7–13. [58] W. C. Dunlap, Y. Yamamoto, Small-molecule antioxidants in marine organisms antioxidant activity of mycosporine-glycine, Comp. Biochem. Physiol. 112B (1995) 105–114. [59] E. A. Galinski, Compatible solutes of halophilic eubacteria: molecular principles, water-solute interaction, stress protection, Experientia. 49 (1993) 487–496. [60] J. Tirkey, S. P. Adhikary, Cyanobacteria in biological soil crusts of India, Curr. Sci. 89 (2005) 515– 521.

159

[61] A. Torres, M. Hochberg, I. Pergament, R. Smoum, V. Niddam, V. M. Dembitsky, et al., A new UV-B absorbing mycosporine with photo protective activity from the lichenized ascomycete Collema cristatum, Eur. J. Biochem. 271 (2004) 780–784. [62] Guillevin R-M., Criblage de mycosporines dans des lichens a algues vertes et a cyanobacteries interets pour la photoprotection, Thèse de Docteur en pharmacie, l'Université de Rennes 1, France, (2012). [63] J. Boustie, S. Tomasi, M. Grube, Bioactive lichen metabolites: alpine habitats as an untapped source, Phytochem. Rev. 10 (2010) 287–307. [64] S. Ramakrishnan, S. Sankara, Amino-acids of Dermatocarpon moulinsii, Curr. Sci. 35 (1966) 284–285. [65] B. Czeczuga, M. Olech, Investigations on carotenoids in lichens. XXV. Studies of carotenoids in lichens from Spitsbergen, Phyton (Horn, Austria). 30 (1990) 235–245. [66] D. Duman, A. Sumer, A. Orhan, Determination of usnic acid content in some lichen species found in Anatolia, J. Appl. Biol. Sci. 2 (2008) 41–44. [67] B. Lindberg, A. Misiorny, C. A. Wachtmeister, Studies on the chemistry of lichens, Acta Chem. Scand. 7 (1953) 591–595. [68] A. Elanur, T. Hasan, Antioxidant and genotoxicity screening of aqueous extracts of four lichens collected from North East Anatolia, Fresenius Environ. Bull. 20 (2011) 2085–2091. [69] C. Roux and coll., Catalogue des lichens et champignons lichénicoles de France métropolitaine. Des Abbayes, Fougères (Ille et Vilaine). Copyright: Claude Roux and AFL. In press (2014). [70] F. S. Dobson, Lichens an illustrated guide to the Bristish and Irish species, The Richmond publishing Co. LTD, (2005). [71] H. Thüs, M. Schultz, Freshwater flora of central Europe. 1st Part: Lichens, (2009). [72] S. Heidmarsson, The genus Dermatocarpon (Verrucariales, lichenized Ascomycotina) in the Nordic countries, Nord. J. Bot. 20 (2001) 605–639. [73] C. Roux, Likenoj de Okcidenta Eu ̆ropo Suplemento 4a : eltiraj ô 3 ( extrait 3 ), Bull. Soc. Linn. Provence. 65 (2014) 41–43. [74] http://www2.ac-lille.fr/myconord/micro/produits07.htm [75] T. Misonou, J. Saitoh, S. Oshiba, UV-absorbing substance in the red alga Porphyra yezoensis (Bangiales, Rhodophyta) block thymine photodimer production, J. Mar. Biotechnol. 5 (2003) 194–200. [76] D. Schmid, C. Schürch, F. Zülli, H.-P. Nissen, H. Prieur, Mycosporine-like amino acids: natural UV-screening compounds from red algae to protect the skin against photoaging, SOFW- Journal. 129 (2003) 2–5. [77] C. Leach, E. Trione, An action spectrum for light induced sporulation in the fungus Ascochyta pisi, Plant Physiol. 40 (1965) 808–812. [78] W. Dunlap, K. Masak, Y. Yamamoto, R. Larsen, I. Karube, A novel antioxidant derived from seaweed. In: New developments in marine biotechnology, (1998). [79] P. Huovinen, I. Gòmez, F. Figueroa, N. Ulloa, V. Morales, C. Lovengreen, Ultraviolet-absorbing mycosporine-like amino acids in red macroalgae from Chile, Bot. Mar. 47 (2004) 21–29. [80] A. Torres, C. D. Enk, M. Srebnik, Porphyra-334, a potential natural source for UVA protective sunscreens, Photochem. Photobiol. Sci. 5 (2006) 432–435. [81] S. Scoglio, Y. Benedetti, F. Benvenuti, S. Battistelli, F. Canestrari, S. Benedetti, Selective monoamine oxidase B inhibition by an Aphanizomenon flos-aquae extract and by its constitutive active principles phycocyanin and mycosporine-like amino acids, Phytomedicine. 21 (2014) 992–7. [82] C. A. Llewellyn, R. L. Airs, Distribution and abundance of MAAs in 33 species of microalgae across 13 classes, Mar. Drugs. 8 (2010) 1273–91. [83] H. Yu, R. Liu, Effect of UV-B radiation on the synthesis of UV-absorbing compounds in a terrestrial cyanobacterium, Nostoc flagelliforme, J. Appl. Phycol. 25 (2013) 1441–1446.

160

[84] M. Yoshiki, K. Tsuge, Y. Tsuruta, T. Yoshimura, K. Koganemaru, Production of new antioxidant compound from mycosporine-like amino acid , porphyra-334 by heat treatment, Food Chem. 113 (2009) 1127–1132. [85] K. Whitehead, J. I. Hedges, Analysis of mycosporine-like amino acids in plankton by liquid chromatography electrospray ionization mass spectrometry, Mar. Chem. 80 (2002) 27–39. [86] K. Whitehead, J. I. Hedges, Electrospray ionization tandem mass spectrometric and electron impact mass spectrometric characterization of mycosporine-like amino acids, Rapid Commun. Mass Spectrom. 17 (2003) 2133–8. [87] K. H. M. Cardozo, R. Vessecchi, V. M. Carvalho, E. Pinto, P. J. Gates, P. Colepicolo, et al., A theoretical and mass spectrometry study of the fragmentation of mycosporine-like amino acids, Int. J. Mass Spectrom. 273 (2008) 11–19. [88] K. Cardozo, V. Vessecchi, S. Galembeck, G. T, P. J. Gates, P. E, et al., Fragmentation study of di- acidic mycosporine-like amino acids in electrospray and nanospray mass spectrometry, J. Braz. Chem. Soc. 20 (2009) 1625–1631. [89] Y. V. Yuan, N. D. Westcott, C. Hu, D. D. Kitts, Mycosporine-like amino acid composition of the edible red alga, Palmaria palmata (dulse) harvested from the west and east coasts of Grand Manan Island, New Brunswick, Food Chem. 112 (2009) 321–328. [90] J.-L. Pittet, M.-L. Bouillant, J. Bernillon, N. Arpin, Sur la presence de mycosporines-glutamine reduites, nouvelles molecules, chez plusieurs Deuteromycetes, Tetrahedron Lett. 24 (1983) 65–68. [91] J. Fayret, J. Bernillon, M.-L. Bouillant, J. Favre-Bonvin, N. Arpin, Open and ring forms of mycosporine-2 from the Ascomycete Gnomonia Leptostyla, Phytochemistry. 20 (1981) 2709– 2710. [92] J. Bernillon, M.-L. Bouillant, J.-L. Pittet, J. Favre-Bonvin, N. Arpin, Mycosporine glutamine and related mycosporines in the fungus Pyronema omphalodes, Phytochemistry. 23 (1984) 1083– 1087. [93] J. I. Carreto, M. O. Carignan, Mycosporine-like amino acids: relevant secondary metabolites. Chemical and ecological aspects, Mar. Drugs. 9 (2011) 387–446. [94] J. Pittet, L. Robert, J. Frot-coutaz, N. Arpin, Soluble uridine diphospho-D-glucose: mycosporin glucosyltransferase from spores of Ascochyta fabae Speg, Planta. 159 (1983) 159–164. [95] D. Karentz, F. S. McEuen, M. C. Land, W. C. Dunlap, Survey of mycosporine-like amino acid compounds in Antarctic marine organisms: Potential protection from ultraviolet exposure, Mar. Biol. 108 (1991) 157–166. [96] R. Przeslawski, K. Benkendorff, A. R. Davis, A quantitative survey of mycosporine-like amino acids (MAAS) in intertidal egg masses from temperate rocky shores, J. Chem. Ecol. 31 (2005) 2417–38. [97] D. Mason, F. Schafer, J. Shick, W. Dunlap, Ultraviolet radiation-absorbing mycosporine-like amino acids (MAAs) are acquired from their diet by Medaka fish (Oryzias latipes) but not by SKH-1 hairless mice, Comp. Biochem. Physiol. 120 (1998) 587–598. [98] J. I. Carreto, M. O. Carignan, N. G. Montoya, A high-resolution reverse-phase liquid chromatography method for the analysis of mycosporine-like amino acids (MAAs) in marine organisms, Mar. Biol. 146 (2004) 237–252. [99] E. Helbling, E. Barbieri, R. Sinha, V. Villafañe, D. Häder, Dynamics of potentially protective compounds in Rhodophyta species from Patagonia (Argentina) exposed to solar radiation, J. Photochem. Photobiol B Biol. 75 (2004) 63–71. [100] M. Klisch, D. Häder, Wavelength dependence of mycosporine-like amino acid synthesis in Gyrodinium dorsum, J. Photochem. Photobiol B Biol. 66 (2002) 60–66. [101] F. Conde, M. Churio, C. Previtali, The photoprotector mechanism of mycosporine-like amino acids. Excited-state properties and photostability of porphyra-334 in aqueous solution, J. Photochem. Photobiol B Biol. 56 (2000) 139–144.

161

[102] U. Karsten, T. Sawall, D. Hanelt, An inventory of UV-absorbing mycosporine-like amino acids in macroalgae from polar to warm-temperate regions, Bot. Mar. 41 (1998) 443–453. [103] U. Karsten, T. Sawall, D. Wiencke, A survey of distribution of UV-absorbing substances in tropical macroalgae, Phycol. Res. 46 (1998) 271–279. [104] B. Dehorter, J. Bernillon, Photoinduction des périthèces du Nectria galligena: Production et activité photomorphogéne des mycosporines, Can. J. Bot. 61 (1983) 1435–1442. [105] A. J. Alpert, Hydrophilic-interaction chromatography for the separation of peptides, nucleic acids and other polar compounds, J. Chromatogr. A. 499 (1990) 177–196. [106] S. V. C. de Souza, R. G. Junqueira, A procedure to assess linearity by ordinary least squares method, Anal. Chim. Acta. 552 (2005) 25–35. [107] M. Chollet, J. L. Burgot, pKa values of 3-thioxo-1,2-dithiole-4- and -5-carboxylic acids and of 3- oxo-1,2-dithiole-4- and -5-carboxylic acids in aqueous solution at 298 K, J. Chem. Soc. Perkin Trans. 2. (1998) 387–392. [108] S. Heidmarsson, Molecular study of Dermatocarpon miniatum (Verrucariales) and allied taxa, Mycol. Res. 107 (2003) 459–468. [109] C. Coster, Écologie et fonctionnement des communautés lichéniques saxicoles-hydrophiles, Thèse de Docteur en Écologie, l'Université de Toulouse, France, (2011). [110] F. Monnet, F. Bordas, V. Deluchat, P. Chatenet, M. Botineau, M. Baudu, Use of the aquatic lichen Dermatocarpon luridum as bioindicator of copper pollution: accumulation and cellular distribution tests, Environ. Pollut. 138 (2005) 455–61. [111] B. Lindberg, Studies on the chemistry of lichens, Acta Chem. Scand. 9 (1955) 917–919. [112] P. Rao, K. Sarma, T. Seshadri, Chemical investigation of Indian lichens: Part XXVII- Chemical components of some high altitude lichens from Western Himalayas, (1966) 177–179. [113] Z. Yakugaku, Free amino acids in lichens of Japan III, Pharmaceutical Society of Japan. 93 (1973) 1558–1563. [114] C. Roullier, M. Chollet-Krugler, A. Bernard, J. Boustie, Multiple dual-mode centrifugal partition chromatography as an efficient method for the purification of a mycosporine from a crude methanolic extract of Lichina pygmaea, J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 877 (2009) 2067–73. [115] P. F. Alain, Centrifugal partition chromatography. Chromatographic science series, Marcel Dekker, INC. 68, (1995). [116] G. R. Raymond, D. S. Satyajit, Isolation of natural products by low-presure column chromatography. In: Natural products isolation, Second edition, Humana Press, (2006). [117] P. H. Sherrod, NLREG program and documentation. Copyright ©1991-2004. [118] Y. Ito, pH-zone-refining counter-current chromatography: origin, mechanism, procedure and applications, J. Chromatogr. A. 1271 (2013) 71–85. [119] J. D. White, J. H. Cammack, K. Sakuma, G. W. Rewcastle, R.K. Widener, Transformations of quinic acid. Asymmetric synthesis and absolute configuration of mycosporin I and mycosporin- gly, J. Org. Chem. 60 (1995) 3600–3611. [120] Z. Zhang, Y. Tashiro, S. Matsukawa, H. Ogawa, Short paper influence of pH and temperature on the ultraviolet-absorbing properties of porphyra-334, (2005) 1382–1384. [121] E. Serjeant, B. Dempsey, In IUPAC – Ionization constants of organic acids in aqueous solution, Pergamon Press, Oxford, (1979). [122] J. E. Baldwin, R. M. Adlington, L. C. Mellor, Synthesis of chiral isoxazolidin-5-ones and their applications to the synthesis of β-amino-alanines and β-(N-hydroxyamino)-alanines, Tetrahedron. 50 (1994) 5049–5066. [123] J. M. Scholtzand, P. A. Bartlett, Synthesis and evaluation of inhibitors for escherichia carbamyl phosphate synthetase, Bioorg. Chem. 17 (1989) 422–433. [124] L. Fowden, R. E. Station, G. Britain, Non-protein amino acids of plants, Food Chem. 6 (1980) 201–211.

162

[125] C. S. Evans, M. Y. Qureshi, E. A. Bell, Free amino acids in the seeds of Acacia species, Phytochemistry. 16 (1977) 565–570. [126] E. A. Bell, “Uncommon” amino acids in plants, FEBS Lett. 64 (1976) 29–35. [127] D. Dahlamn, A. R. Gerald, Non-protein amino acid-insect interactions-I. Growth effects and symtomology of L-canavanine consumption by tobacco hornworm, Manduca sexta (L.), Comp. Biochem. Physiol. 51A (1975) 33–36. [128] H. J. Daniel, B. J. Harvey, E. A. Bell, Toxicity of secondary compounds to the seed-eating larvae of the bruchid beetle Callosobruchus Maculatus, Phytochemistry. 16 (1977) 223–227. [129] C. S. Evans, E. A. Bell, Non-protein amino acids of Acacia species and their effect on the feeding of the acridids Anacrldium melanorhodon and Locusta migratoria, Phytochemistry. 18 (1979) 1807–1810. [130] M. Friedman, Applications of the ninhydrin reaction for analysis of amino acids, peptides, and proteins to agricultural and biomedical sciences, J. Agric. Food Chem. 52 (2004) 385–406. [131] M. B. Lewis, J. A. Hellebust, The contribution of organic solutes to osmotic balance in some green and eustigmatophyte algae, J. Phycol. 16 (1980) 265–270. [132] A. G. Harrison, Characterization of alpha- and gamma-glutamyl dipeptides by negative ion collision-induced dissociation, J. Mass Spectrom. 39 (2004) 136–44. [133] T. Kasai, S. Sakamura, NMR spectra of glutamic acid-containing dipeptides in relation to sequence determination, Agr. Biol. Chem. 37 (1973) 2155-57. [134] N. Tal, O. Goldberg, A. Luini, V. I. Teichberg, An evaluation of γ-glutamyl dipeptide derivatives as antagonists of amino acid-induced Na+ fluxes in rat striatum slices, J. Neurochem. 39 (1982) 574–576. [135] M. Rubio-Texeira, G. Van Zeebroeck, J. M. Thevelein, Peptides induce persistent signaling from endosomes by a nutrient transceptor, Nat. Chem. Biol. 8 (2012) 400–8. [136] M. Iorizzi, L. Minale, R. Riccio, J. Soo Lee, T. Yasumoto, Polar steroids from the marine scallop Patinopecten yessoensis, J. Nat. Prod. 6 (1988) 1098–1103. [137] Vincenzo Piccialli, Donato Sica, Four new trihydroxylated sterols from the sponge Spongionella Gracilis, J. Nat. Prod. 50 (1987) 915–920. [138] T.-F. Liu, X. Lu, H. Tang, M.-M. Zhang, P. Wang, P. Sun, et al., 3β,5α,6β-Oxygenated sterols from the South China Sea gorgonian Muriceopsis flavida and their tumor cell growth inhibitory activity and apoptosis-inducing function, Steroids. 78 (2013) 108–14. [139] A. G. Gonzalez, J. Bermejo Barrera, E. M. Rodriguez Perez, C. E. Hernandez Padron, Chemical constituents of the lichen Ramalina hierrensis, Planta Med. 58 (1992) 214–218. [140] A. G. Gonzalez, M. R. P. Elsa, E. H. P. Consuelo, B. B. Jaime, Chemical constituents of the lichen Stereocaulon azoreum, Z. Naturforsch. 47c (1992) 503–507. [141] G. JinMing, Z. L. Dong, J. K. Liu, A new ceramide from the basidiomycete Russula cyanoxantha, Lipids. 36 (2001) 175–80. [142] P. Muralidhar, P. Radhika, N. Krishna, D. V. Rao, Sphingolipids from marine organisms : A Review, Nat. Prod. Sci. 9 (2003) 117–142. [143] T. A. Springer, L. A. Lasky, Sticky sugars for selectins, Nature. 349 (1991) 196–197. [144] T. Feizi, Carbohydrate differentiation antigens : probable ligands for cell adhesion molecules, Trends Biochem Sci. 16 (1991) 84–86. [145] D. Pei, J.-X. Liu, D.-L. Di, Cerebroside and ceramide from the pollen of Brassica napus L., Fitoterapia. 81 (2010) 838–43. [146] D.-W. Huang, C.-P. Chung, Y.-H. Kuo, Y.-L. Lin, W. Chiang, Identification of compounds in adlay (Coix lachryma-jobi L. var. ma-yuen Stapf) seed hull extracts that inhibit lipopolysaccharide- induced inflammation in RAW 264.7 macrophages, J. Agric. Food Chem. 57 (2009) 10651–7. [147] S. Bao-ning, M. Rosana, J. P. Eun, B. D. Santarsiero, A. D. Mesecar, H. H. S. Fong, et al., Isolation and characterization of bioactive principles of the leaves and stems of Physalis philadelphica, Tetrahedron. 58 (2002) 3453–3466.

163

[148] E. Fahy, S. Subramaniam, H. A. Brown, C. K. Glass, A. H. Merrill, R. C. Murphy, et al., A comprehensive classification system for lipids, J. Lipid Res. 46 (2005) 839–61. [149] R. Sandhoff, Very long chain sphingolipids: tissue expression, function and synthesis, FEBS Lett. 584 (2010) 1907–13. [150] B. M. Castro, M. Prieto, L. C. Silva, Ceramide: a simple sphingolipid with unique biophysical properties, Prog. Lipid Res. 54 (2014) 53–67. [151] C. Buré, J.-L. Cacas, S. Mongrand, J.-M. Schmitter, Characterization of glycosyl inositol phosphoryl ceramides from plants and fungi by mass spectrometry, Anal. Bioanal. Chem. 406 (2014) 995–1010. [152] G. JinMing, S. Jie, Z. Anling, Z. Wei, Z. Xin, J. Lu, Chemical constituents of the fungus Leccinum extremiorientale, Chin. J. Org. Chem. 23 (2003) 853–857. [153] G. JinMing, A. Zhang, C. Zhang, J. Liu, Paxillamide : a novel phytosphingosine derivative from the fruiting bodies of Paxillus panuoides, Helv. Chim. Acta. 87 (2004) 1483–1487. [154] J. M. Marcos, G. Marco, A. J. G. Phillip, T. Gianggiacomo, L. Marcello, A galactosphingolipid from the lichen, Ramalina celastri, Phytochemistry. 45 (1997) 651–653. [155] D. Perrin, B. Dempsey, In: Buffers for pH and metal ion control, Chapman and Hall, London, (1974). [156] H. T. S. Britton, R. A. Robinson, The use of the antimony-antimonous oxide electrode in the determination of the concentration of hydrogen ions and in potenliometric tilrations. The Pridenux-Ward Universal Buffer Mixture, J. Chem. Soc. (1931) 458–473. [157] K. N. Bascombe, R. P. Bell, Acidity functions of some aqueous acids, J. Chem. Soc. (1959) 1096– 1104. [158] C. Oyamada, M. Kaneniwa, K. Ebitani, M. Murata, K. Ishihara, Original article mycosporine-like amino acids extracted from scallop (Patinopecten yessoensis) ovaries : UV protection and growth stimulation activities on human cells, Mar. Biotechnol. 10 (2008) 141–150. [159] B. Ryu, Z.-J. Qian, M.-M. Kim, K.W. Nam, and S.-K. Kim, Anti-photoaging activity and inhibition of matrix metalloproteinase (MMP) by marine red alga, Corallina pilulifera methanol extract, Radia. Phys. Chem. 78 (2009) 98–105 [160] D. Schmid, C. Schürch, F. Zülli, Mycosporine-like amino acids from red algae protect against premature skin-aging, Euro Cosmet. 9 (2006) 1–4. [161] R. P. Rastogi, A. Incharoensakdi, Analysis of UV-absorbing photoprotectant mycosporine-like amino acid (MAA) in the cyanobacterium Arthrospira sp. CU2556, Photochem. Photobiol. Sci. 13 (2014) 1016–24. [162] F. Lohézic-Le Dévéhat, B. Legouin, C. Couteau, J. Boustie, L. Coiffard, Lichenic extracts and metabolites as UV filters, J. Photochem. Photobiol. B. 120 (2013) 17–28. [163] P. Jones, In vitro phototoxicity assays. In: Principles and practice of skin toxicology, (2008). [164] H. L. Gensler, M. Magdaleno, Topical vitamin E inhibition of immunosuppression and tumorigenesis induced by ultraviolet irradiation, Nutr. Cancer. 15 (1991) 97–106. [165] G. Cordell, D. Kinghorn, J. Pezzuto, Separation, structure elucidation, and bioassay of cytotoxic natural products. In: Bioactive natural products. Press C., Colegate, Boca raton, (1993). [166] Y. V Yuan, Y. Athukorala, Red algal mycosporine-like amino acids (MAAs) as potential cosmeceuticals. In: Marine cosmeceuticals trends and prospects, (2012). [167] M. Burnett, S. Wang, Current sunscreen controversies: a critical review, Photodermatol Photoimmunol Photomed. 27 (2011) 58–67. [168] J. Hulsman, Enzymes involved in the metabolism of valproate, Pharm Weekbl Sci. 14 (1992) 98– 100. [169] F. Lohézic-Le Dévéhat, B. Legouin, A. Malargé, C. Couteau, L. Coiffard, Evaluation of a new UV- filter screening method. 4èmes journées de l’AFERP « Biodiversité, chimie des substances naturelles et médicaments" et 3ème Symposium franco-pakistanais, Besançon (France) 16-18 July, (2010).

164

[170] COLIPA guidelines : In vitro method for the determination of the UVA protection factor and “critical wavelength” values of sunscreen products, (2009). [171] F. Coba, J. Aguilera, F. L. Figueroa, M. V. Gálvez, E. Herrera, Antioxidant activity of mycosporine- like amino acids isolated from three red macroalgae and one marine lichen, J Appl Phycol. 21 (2009) 161–169. [172] R. P. Rastogi, A. Incharoensakdi, UV radiation-induced accumulation of photoprotective compounds in the green alga Tetraspora sp. CU2551, Plant Physiol. Biochem. 70 (2013) 7–13. [173] R. Nordmann, C. Ribière, H. Rouach, Implication of free radical mechanisms in ethanol-induced cellular injury, Free Radic. Biol. Med. 12 (1992) 219–40. [174] S. M. Bailey, C. C. Cunningham, Contribution of mitochondria to oxidative stress associated with alcoholic liver disease, Free Radic. Biol. Med. 32 (2002) 11–16. [175] O. Sergent, M. Pereira, C. Belhomme, M. Chevanne, L. Huc, D. Lagadic-Gossmann, Role for membrane fluidity in ethanol-induced oxidative stress of primary rat hepatocytes, J. Pharmacol. Exp. Ther. 313 (2005) 104–11. [176] E. Niki, Antioxidants in relaton to lipid peroxidation, Chem. Phys. Lipids. 44 (1987) 227–253. [177] A. Collin, K. Hardonnière, M. Chevanne, J. Vuillemin, Cooperative interaction of benzo[a]pyrene and ethanol on plasma membrane remodeling is responsible for enhanced oxidative stress and cell death in primary rat hepatocytes, Free Radic. Biol. Med. 72 (2014) 11–22. [178] C. Roullier, Recherche de mycosporines et de dérivés aminés lichéniques d’intérêt pour les cancers photoinduits. Etude phytochimique d’un lichen marin: Lichina pygmaea (Lightf.) C. Agardh. Thèse de docteur mention Chimie. L'Université de Rennes 1, France, (2010). [179] M. Moline, E. M. Arbeloa, R. Flores, D. Libkind, E. Farı, S. G. Bertolotti, et al., UVB Photoprotective role of mycosporines in yeast : photostability and antioxidant activity of mycosporine-glutaminol-glucoside, Radiat. Res. 175 (2011) 44–50. [180] B. L. Diffey, J. Robson, A new substrate to measure sunscreen protection factors throughout the ultraviolet spectrum, J. Soc. Cosmet. Chem. 133 (1989) 127–133. [181] B. L. Diffey, Spectral uniformity: a new index of broad spectrum (UVA) protection, Int. J. Cosmet. Sci. 31 (2009) 63–68. [182] B. L. Diffey, M. W. Brown, The ideal spectral profile of topical sunscreens, Photochem. Photobiol. 88 (2012) 744–7. [183] I. I. Koleva, T. a van Beek, J. P. H. Linssen, A. de Groot, L. N. Evstatieva, Screening of plant extracts for antioxidant activity: a comparative study on three testing methods, Phytochem. Anal. 13 (2002) 8–17. [184] F. Le Devehat, H. Thüs, M.-L. Abasq, D. Delmail, J. Boustie, Oxidative stress regulation in lichens and its relevance for survival in coastal habitats. In: Advances in Botanical Research, Vol. 71, Elsevier, (2014). [185] I. Morel, G. Lescoat, J. Cillard, N. Pasdeloup, P. Brissot, P. Cillard, Kinetic evaluation of free malondialdehyde and enzyme leakage as indices of iron damage in rat hepatocyte cultures. Involvement of free radicals, Biochem. Pharmacol. 39 (1990) 1647–55. [186] M. M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72 (1976) 248–54. [187] OECD Guideline for testing of chemicals. In vitro 3T3 NRU phototoxicity test, 432 (2004). [188] M. E. Carlotti, S. Sapino, S. Marino, E. Ugazio, F. Trotta, D. Vione, et al., Influence of hydroxypropyl-β-cyclodextrin on the photostability and antiradical activity of Trolox, J. Incl. Phenom. Macrocycl. Chem. 61 (2008) 279–287. [189] T. Nguyen, M. Chollet-Krugler, F. Lohézic-Le Dévéhat, I. Rouaud, J. Boustie, Mycosporine-like compounds in chlorolichens: isolation and photoprotective properties from Dermatocarpon luridum and Dermatocarpon miniatum, Planta Medica Lett. In press (2014).

165

[190] F. Ismed, Phytochimie de lichens du genre Stereocaulon: étude particulière de S. halei Lamb et S. montagneanum Lamb, deux lichens recoltés en Indonésie. Thèse de docteur mention Chimie. L'Université de Rennes 1, France, (2012). [191] H. Thüs, L. Muggia, S. Pérez-ortega, S. E. Favero-longo, S. Joneson, H. O. Brien, et al., Revisiting photobiont diversity in the lichen family Verrucariaceae ( Ascomycota ), Eur J Phycol. 46 (2011) 399–415. [192] K. M. Fontaine, E. Stocker-Wörgötter, T. Booth, M. D. Piercey-Normore, Genetic diversity of the lichen-forming alga, Diplosphaera chodatii, in North America and Europe, Lichenol. 45 (2013) 799–813. [193] C. Gueidan, S. Savić, H. Thüs, C. Roux, C. Keller, L. Tibell, Generic classification of the Verrucariaceae (Ascomycota) based on molecular and morphological evidence : recent progress and remaining challenges, Taxon. 58 (2009) 184–208. [194] I. Kranner, W.J. Cram, M. Zorn, S. Wornik, I. Yoshimura, E. Stabentheiner, et al., Antioxidants and photoprotection in a lichen as compared with its isolated symbiotic partners, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 3141–6. [195] D. Fahselt, M. Krol, V. Alstrup, N. Huner, Detection of pigments in specimens of recent and subfossil Umbilicaria from North Greenland, Bryologist. 104 (2001) 593–599. [196] C. Leuckert, V. Ahmadjian, C. F. Culberson, A. Johnson, Xanthones and depsidones of the lichen Lecanora dispersa in nature and of its mycobiont in culture, Mycologia. 82 (1990) 370–378. [197] C. F. Culberson, W. L. Culberson, S. Gowan, A. Johnson, New depsides from lichens: microchemical methodologies applied to the study of new natural product discovered in Herbarium specimens, Am. J. Bot. 74 (1987) 403–414. [198] K. Huovinen, T. Ahti, The composition and contents of aromatic lichen substances in the genus Cladina, Ann. Bot. Fenn. 23 (1986) 93–106. [199] Y. Agnan, N. Séjalon-Delmas, A. Probst, Origin and distribution of rare earth elements in various lichen and moss species over the last century in France, Sci. Total Environ. 487 (2014) 1–12. [200] T. Zschau, S. Getty, C. Gries, Y. Ameron, A. Zambrano, T. H. Nash, Historical and current atmospheric deposition to the epilithic lichen Xanthoparmelia in Maricopa county, Arizona, Environ. Pollut. 125 (2003) 21–30. [201] E. I. Friedmann, H. J. Sun, Communities adjust their temperature optima by shifting producer- to-consumer ratio, shown in lichens as models: I. Hypothesis, Microb. Ecol. 49 (2005) 523–527. [202] O. F. Cubero, A. Crespo, J. Fatehi, P. D. Bridge, DNA extraction and PCR amplification method suitable for fresh, herbarium-stored, lichenized, and other fungi, Plant Syst. Evol. 216 (1999) 243–249. [203] M. Grube, P.T. Depriest, A. Gargas, J. Hafellner, DNA isolation from lichen Ascomata, Mycol. Res. 99 (1995) 1321–1324. [204] R. Halvorsen, E. Bendiksen, The chemical variation of Pseudevernia furfuracea in Norway, Nord. J. Bot. 2 (1982) 371–380. [205] P. L. Nimis, N. Skert, Lichen chemistry and selective grazing by the coleopteran Lasioderma serricorne, Environ. Exp. Bot. 55 (2006) 175–182. [206] A. A. Burkin, T. Y. Tolpysheva, G. P. Kononenko, Safety of fungal secondary metabolites in herbarial lichen specimens, Vestn. Mosk. Univ. 3 (2012) 28–32. [207] W. R. Stochaj, W. C. Dunlap, J. M. Shick, Two new UV-absorbing mycosporine-like amino acids from the sea anemone Anthopleura elegantissima and the of zooxanthellae and spectral irradiance on chemical composition and content effects, Mar. Biol. 118 (1994) 149–156. [208] M. O. Carignan, K. H. M. Cardozo, D. Oliveira-Silva, P. Colepicolo, J. I. Carreto, Palythine- threonine, a major novel mycosporine-like amino acid (MAA) isolated from the hermatypic coral Pocillopora capitata, J. Photochem. Photobiol. B. 94 (2009) 191–200.

166

[209] H.-J. Suh, H. Lee, J. Jung, Mycosporine glycine protects biological systems against photodynamic damage by quenching singlet oxygen with a high efficiency, Photochem. Photobiol. 78 (2003) 109–113. [210] I. Yakovleva, R. Bhagooli, A. Takemura, M. Hidaka, Differential susceptibility to oxidative stress of two scleractinian corals : antioxidant functioning of mycosporine-glycine, Comp. Biochem. Physiol. Part B. 139 (2004) 721–730. [211] W. C. Dunlap, J. M. Shick, Review-ultraviolet radiation-absorbing mycosporine-like amino acids in coral reef organisms: a biochemical and environmental perspective, J. Phycol. 34 (1998) 418– 430. [212] C. Tao, T. Sugawara, S. Meada, X. Wang, T. Hirata, Antioxidative activities of a mycosporine-like amino acid, porphyra-334, Fish. Sci. 74 (2008) 1166–1172. [213] R. Nakayama, Y. Tamura, H. Kikuzaki, N. Nakatani, Antioxidant effect of the constituents of Susabinori (Porphyra yezoensis), JAOCS. 76 (1999) 649–653. [214] L. Zhang, L. Li, Q. Wu, Protective effects of mycosporine-like amino acids of Synechocystis sp. PCC 6803 and their partial characterization, J. Photochem. Photobiol B Biol. 86 (2007) 240–245. [215] L. Muggia, C. Gueidan, M. Grube, Phylogenetic placement of some morphologically unusual members of Verrucariales, Mycologia. 102 (2010) 835–846.

167

168

General experimental procedures

• Melting points were measured on a Kofler LEICA VMHB.

• UV spectra were recorded on an Uvikon 931 UV–vis spectrophotometer.

• IR spectra were recorded on a PerkinElmer UATR Two infrared spectrophotometer.

• The NMR experiments were performed on a Bruker DMX 300 and 500 spectrometer.

• Optical rotations were measured on a Perkin Elmer Model 341 polarimeter at 20°C

using thermostable optical glass cell (1 dm path length).

• HPLC was performed on Shimadzu LC-20AD consisting of a system controller CBM-

20A, solvent delivery unit LC-20A, auto-sample SIL-20A, column oven CTO-20. An

SPD-M20A photo-diode array detector was used for detection and quantification

(wavelength range 190 to 800 nm).

• HPLC–DAD–MS2 were recorded on a Thermo Finnigan Surveyor liquid

chromatograph interfaced with a LCQ Deca ion trap mass spectrometer (Thermo

Finnigan, Villebon sur Yvette, France).

• ESI-HRMS were carried out on a MICROMASS ZabspecTOF spectrometer for

electrospray ionization.

• Flash chromatography was performed on a SPOT Flash liquid chromatography (Armen

Instrument).

• Open column chromatography was performed on normal phase silica gel (40-63 µm,

Keselgel 60, Merck 7667), reverse phase silica gel C-18 (C-18 Hydro Chromabond,

Macherey-Nagel), gel sephadex LH-20 (Sigma-Aldrich), cation exchange resin Dowex

50W-X8.

• TLC was performed on Kieselgel 60F254 plates (Merck) and spots were visualized

under UV light or sprayed with anisaldehyde (a solution of 0.5 Ml anisaldehyde in 50

169

mL glacial acetic acid and 1 mL 97% sulfuric acid) or thymol (a solution of 0.5 g thymol

in 95 mL ethanol and add 5 mL 97% sulfuric acid) or ninhydrine (dissolve 0.3 g

ninhydrine in 100 mL 1-butanol and add 3 mL glacial acetic acid), then heated.

• UV stratalinker 2400 (Stratagene, USA) for UV irradiation.

• pH were measured by LPH 430T pH - METER

• Bioblock IKA Yellow Line OST Eurostar basic overhead stirrer, Fisher Scientific,

Illkirch, France to stir the O/W emulsion in UV-filter assay.

• Multiskan FC (Thermo Scientific) spectrophotometer for measure the decrease in

absorbance at 540 nm for DPPH assay and at 570nm for NBT assay.

• Indentification of lichens:

Lichens were examined using a binocular microscope (Laboratory Humeau) that can also perform precise cuts. Cuts, impregnated or not with dye, are then mounted between slide and slip cover and observed under a microscope equipped with a digital camera (Olympus CX41 Microscope). Lactophenol cotton blue was used for staining of thallus (0.1 g of methyl blue + 20 g of lactic acid + 20 mL of water + 40 g glycerine + 2 g phenol). This dye allows to highlight the spore and chitin present in the walls of the hyphae Melzer’s reagent was used for thallus reaction. Melzer’s solution is made by dissolving 1.0 g of iodine, 1.5 g of potassium iodide in 50 mL of distilled water plus 50 g of chloral hydrate.

• Solvents and chemicals: solvents for extraction and for chromatography were purchased

from Carlo Erba Reactifs (Val de Reuil, France). Distilled water was obtained by an

EasyPure (Barnstead, USA) water purification system. Deuterated solvents were

purchased from Euriso-top (Gif-sur-Yvette, France), Quercetin 98% (Sigma Q-0125),

cytosine 99% (Sigma C-3506), trolox 97% (Sigma 238813/1), chlorpromazine 98%

(Sigma C-8138), nitro blue tetrazolium (NBT, Sigma N-6876), 1,1-diphenyl-2-

170

picrylhydrazyl (DPPH, Fluka 43180), NADH (Sigma N-8129), phenazine methosulfate

(PMS, Fluka 68600), neutral red (NR, 209-035-8) and phosphate buffer saline (PBS, P-

5493), dimethylsulfoxyde, sodium dodecyl sulfate (SDS), paraffinum liquidum were obtained from Sigma (St Quentin-Fallavier, France), Avobenzone, 4- methylbenzylidene camphor and octocrylene were purchased from Merck (Fontenay

Sous Bois, France).

171

172

Appendix

Appendix 1: Key to Dermatocarpon Eschw [73] 11 Face inférieure du thalle avec rhizines, de brun clair à brun sombre. Face supérieure plus ou moins pruineuse. Thalle au stade mature avec un ou peu de lobes, avec un seul point de fixation central. 22 Rhizines longues (2 ou 3 mm), nombreuses et denses, brun sombre à brun clair, avec une pointe fine ou s’amincissant […]. – S des Alpes et Pyrénées 1. D. moulinsii (Mont) Zahlbr. 2 Rhizines (en fait papilles) courtes (jusqu’à 0.2 mm). 13. D. miniatum var. cirsodes 1 Thalle sans rhizines 22 Thalle au stade mature plurilobé, avec de nombreux points de fixation disséminés sur toute la partie inférieure du thalle. 33 Médule du thalle I+ (rougeâtre) [utiliser une solution d’Iode concentrée ou mieux le réactif de Melzer] 44 Thalle grand (jusqu’à 30 cm) verdissant avec l’humidité, brun foncé à brun clair dessus et dessous, lisse à la face inférieure ou légèrement ridé. Spores pour la plupart de 13.5- 18 µm de long [(10.5)13.5-18(20) × 5.5-7(8.5) µm]. – Europe. De l’étage collinéen à l’étage alpin. En milieu peu ou moyennement humide, calcaire ou non.— 2. D. luridum (With.) J. R. Laundon 4 Thalle petit (Jusqu’à 5 cm), ne verdissant pas à l’humidité, gris ou brun à la face supérieure, a la face inférieure, au moins partiellement, réticulé, brun légèrement sombre ou clair. Spores pour la plupart de 10-13.5 de long [(8.5)10-13.5(16) × (4.5) 5.5-7(8.5) µm]. –N de l’Europe.—En milieu très humide, dans les suintements ou en milieu peu humide, calcaire.— 3. D. polyphyllizum 3 Médulle du thalle I-. Europe— 44 Spores (10)11-15.5(16) × (4)5-7(8) µm. Long/larg (1.5)1.8-2.7(3.2), ellipsoïdales. Face supérieure du thalle pruineuse, rarement non pruineuse; face inférieure lisse ou peu ridée. Du mésoméditerranéen à l’étage alpin. À noter dans les suintements ou les milieux peu humides, non calcaires—Syn. D. miniatum var. complicatum (Lightf.) W. Mann., D. miniatum var. complicatissimum (Nyl.) Lettau, D. polyphyllum (Wulfen) Dalla Torre et Sarnth., D. miniatum

173

var. compactum (Lamy) Zahlbr., D. miniatum var. panniforme (Lamy) Zahlbr.—Le plus souvent considéré comme variété de D. miniatum— 4. D. complicatum (Lightf.) W. Mann. 4 Spores (6)6.5-9(10.5) × (4.5)5-6.5(7) µm, L/l (1.0)1.1-1.6(2.0), généralement largement ellipsoïdales ou subglobuleuses. La face supérieure du thalle est très pruineuse, la face inférieure lisse ou presque— Hautes montagnes. De l’étage montagnard à l’étage alpin. Pas ou a peine dans les suintements, dans des milieux peu ou très calcaires.—Syn (?) D. decipiens (A. Massal.) Dalla Torre et Sarnth. Non auct., (?) D. miniatum var. crispum (A. Massal.) Zahlbr. non auct.— Attention: D. intestiniforme sensu Orange 2009 [non (Körb.) Hasse] est D. complicatum (Lightf.) W. Mann! 5. D. intestiniforme (Korb.) Hasse 2 Thalle au stade mature avec un ou peu de lobes, avec un seul point de fixation au centre. 33 Spores pour la plupart > 15 µm de long. 44 Face supérieure du thalle brune, non pruineuse (sauf très rares exceptions). Cortex supérieur avec une couche morte de cellules horizontales plates. 55 Face inférieure du thalle lisse ou finement granuleuse en surface, brun sombre ou brun noir. Thalle épais (0.3-0.6 mm) formé de petits lobes (6-12 mm de large), plus ou moins arrondis. Spores (11)14-18(20.5) × (5)6-8(10.5) µm.—N de l’Europe et montagnes d’Europe centrale. Dans des milieux plus ou moins humides, voire dans des suintements, Calcicole ou non— 6. D. meiophyllizum Vain 5 Face inférieure du thalle réticulée, brun sombre à brun clair. Thalle grand (lobes de 13- 30 mm) plus ou moins irrégulier, mince (0.2-0.4 mm. Spores (14)16-21(26) × (5.5)6-8(10.5) µm.-Europe. En haute montagne jusqu’à l’étage nival. En milieu humide et dans les suintements. 7. D. rivulorum (Arnold) Dalla Torre et Sarnth. 4 Face supérieure du thalle ordinairement grise, généralement pruineuse (au moins les jeunes lobes). Cortex supérieur avec couche morte de cellules non plates, gonflées. 55 Thalle composé de petits lobes (1-4 mm de large), disposés en petits groupements denses. Face inférieure du thalle brune ou brun sombre. Spores 13-21 × 5-8 µm.— Europe. De l’étage montagnard à l’étage alpin. En milieu peu humide, non calcaire. 8. D. leptophyllodes (Nyl.) Zahlbr. 5 Thalles plus grands ne se regroupant pas densément.

174

66 Thalle petit (formé de lobes 7-13 mm de large), mince (0.2-0.4 mm), de lisse à inégal ou mȇme ridé, brun sombre à la face inférieure. Spores (11.5)15-20(23) × (4.5)5.5- 7(8.5) µm.—N de l’Europe. Milieux peu humides ou exposés. 9. D. deminuens Vain. 6 Thalle grand (formé de lobes de 13-24 mm de large), un peu plus épais (0.25-0.5 mm), ordinairement réticulé à la face inférieure, brun sombre. Spores (13)16-20(25.5) × (4.5)6-8(9) µm.— Centre et Nord de l’Europe. Étage montagnard— 10. D. bachmannii Anders. 3 Spores la plupart < 15 µm de long. 44 Médule du thalle I+ (rougeâtre) [Utilisez une solution concentrée d’Iode ou mieux, le réactif de Melzer]. Thalle petit (lobes 7-15 mm de large), gris ou brun à la partie supérieure, de brun clair à brun sombre à la partie inférieure, au moins en partie veiné. Spores (8.5)10- 13.5(16) × (4.5)5-7(8.5) µm. — N de l’Europe. Milieu calcaire ou non, neutre ou subneutre, peu humide ou suintements— 11. D. polyphyllizum (Nyl.) Blomb, et Forssell 4 Médulle du thalle I-. 55 Spores largement ellipsoïdes ou subglobuleuses, dans l’asque généralement sur un rang, (7)8-10.5(11.5) × (4.5)6-6.5(7) µm. Thalle mince (0.15-0.4 mm), petit (ordinairement de 5-20 mm de diamètre), brun ou brun sombre à la face inférieure, lisse ou verruqueux en surface.— Europe. Étages montagnard et subalpin, Calcicole. 12. D. leptophyllum (Ach.) K.G.W. Lang 5 Spores ellipsoïdes, rarement subglobuleuses, dans l’asque sur deux rangs, (7)10-14.5(18) ×(4)5-6.5(7) µm. L/l : (1.4)1.8-2.6(3). Thalle diversement épais, plus grand (ordinairement de 10-40(60) mm de diamètre), de brun clair à brun noirâtre à la face inférieure, lisse ou pas.-- Calcicole ou non.— 666 Face inférieure du thalle avec papilles denses, d’environ 0.2 mm de haut. Thalle 0.5- 1.3 mm d’épaisseur. Thalle unilobé, avec une face inférieure brun clair, rarement noirâtre. De l’étage montagnard à l’étage alpin, sur roches calcaires ou basiques non calcaires-- Syn. D. caesium Räs. 13. D. miniatum (L.) W. Mann var. cirsodes (Ach.) Vain. 66 Thalle lisse ou presque lisse à la face inférieure, à surface rarement granuleuse ou avec quelques papilles, 0.3-0.5 mm d’épaisseur. Thalle à un lobe (morpho, miniatum) ou quelques

175

lobes (morpho, intricatum), avec une face supérieure de brun clair à brun noir. De l’étage mésoméditerranéen à l’étage alpin. Sur roches calcaires ou non calcaires, en milieu sec, plus rarement légèrement humide— 13. D. miniatum (L.) W. Mann var. miniatum 6 Thalle réticulé à la face inférieure, 0.4-0.7 mm d’épaisseur, de couleur claire, gris ou brun à la face supérieure, Europe. De l’étage montagnard à l’étage alpin. En milieu humide, calcaire ou non—

14. D. arnoldianum Degel.

Traduction Martine Davoust

176

Appendix 2: The major mycosporine-like compounds and their physico-chemical properties

+ Mycosporine-like compounds R1 R2 R3 R4 λmax ε [M+H] ESI/MS Ref. (nm) (M-1cm-1) fragmentation pattern values Palythine NH H COOH H 320 36200 245 137, 150, 155, 162, [166] 184, 186, 199, 209, 230, 245 Palythine-serine NH CH2OH COOH H 320 10500 275 260 [166] Palythine-threonine NH CH(OH)CH3 COOH H 320 no 289 134, 151, 169, 245 [166] Shinorine N-CH(COOH)CH2OH H COOH H 332 44668 333 137, 168, 185, 186, [166] 197, 211, 230, 241, 255, 333 Asterina-330 N-CH2CH2OH H COOH H 330 4800 289 137, 150, 168, 186, [166] 197, 199, 213, 230, 243, 273, 289 Palythinol N-CH(CH3)CH2OH H COOH H 330 43500 303 137, 150, 168, 186, [166] 197, 199, 231, 243, 288, 303 Porphyra-334 N-CH(COOH)CH(OH)CH3 H COOH H 332 42300 346 137, 151, 168, 185, [166] 186, 197, 200, 243, 303, 347 Usujirene N-CH=CHCH3 (cis) H COOH H 356 50000 285 137, 149, 166, 179, [166] 193, 195, 197, 202, 223, 241, 285 Palythene N-CH=CHCH3 (trans) H COOH H 360 50000 285 137, 149, 166, 179, [166] 193, 195, 197, 205, 223, 241, 285

177

Palythenic acid N-C(COOH)=CHCH3 H COOH H 337 29200 329 138, 150, 175, 182, [166] 193, 197, 225, 237, 241, 251, 268, 283, 296, 314, 329 Mycosporine-2-glycine N-CH2COOH H COOH H 331 no 303 151, 164, 185, 200, [207] 244, 288 Mycosporine- methylamine- N-CH3 CH2OH COOH H 325 16600 289 274 [208] serine Euhalothece-362 N-CH=C(OH)CH2OH CH3 COOH H 362 no 331 213, 228, 242, 272, [41] 316 478-Da N-CH(COOH)CH(OH)CH3 H COOH Pen 335 33173 479 347, 420, 435 [47] 1050-Da H Pen-Glu 312, 58800 1051 649, 721, 883, 1015 [47] 340 508-Da MAA N-CH(COOH)CH(OH)CH3 H COOH Glu 334 no 509 285, 303, 347, 385, [48] 403, 407, 419, 421, 451, 465 612-Da MAA NH CH(CH3)O-Glu COOH Glu 322 no 613 151, 169, 170, 177, [48] 181, 187, 191, 199, 209, 212, 227, 230, 245, 289, 349, 375, 389, 392, 407, 451, 569 Mycosporine-serinol O CH2OH CH2OH H 310 27270 262 182, 184, 194, 212, [30] 216, 244 Mycosporine-glutaminol- O (CH2)2CONH2 CH2O-G H 310 no 465 235, 267, 285, 303, [46] glucoside 429, 447 Mycosporine-glutamicol- O (CH2)2COOH CH2O-G H 310 no 466 236, 268, 286, 304, [45], glucoside 430, 448 [46] Mycosporine-glutamine O (CH2)2CONH2 COOH H 310 no 317 280 [92] Mycosporine-glutamic acid O (CH2)2COOH COOH H 310 20893 317 255, 273, 299 [92] Mycosporine glutaminol O (CH2)2CONH2 CH2OH H 310 12542 303 235, 267, 285 [46], [90] Mycosporine glutamicol O (CH2)2COOH CH2OH H 310 17248 304 178, 210, 236, 258, [3], 268, 286 [46], [91]

Glu: glucose, Pen: pentose

178

Appendix 3: List of Dermatocarpon species in this study with their voucher barcode, collector, collection date, habitat and corresponding mycosporines contents (mean values ± SD, mg/g of dw. extract)

N° Lichens Voucher barcode Collector Date of collection Country Habitat notes Elevation Mycosporine Mycosporine (revised identification) glutaminol 1 glutamicol 2 1 D. luridum BM001085236 A.Willi Exsiccate series Austria, Tirol ad saxa inundata nd 18.06 ± 3.03 started in 1920 rivuli Tumpnerbach 2 D. luridum BM001085256 V. Räsänen 18/06/1936 Finland, Karelia ad saxa rivalia nd 17.77 ± 1.73 3 D. luridum BM001085255 Fagerström 16/06/1947 Finland, Karelia ad saxa humida prope nd 12.92 ± 1.20 cataractam 4 D. luridum BM001085235 F.W.Zopf 08/1897 Germany, Niedersachsen, auf Granitbloecken im nd 13.60 ± 1.43 Harz Mts Bett der Bode 5 D. luridum BM001085237 H.Ullrich 04/08/1963 Germany, Niedersachsen, an periodisch 450 m NN nd 22.64 ± 0.98 Harz Mts inundierten Silikatbloecken in der Radau 6 D. luridum BM001085252 Ahles 04/1861 Germany, Odenwald on rock of a mountain nd 15.73 ± 3.39 creek 7 D. luridum BM000731396 H. Thüs 2006 Germany, Odenwald 2.82 ± 0.19 21.00 ± 0.61 (under LOQ) 8 D. luridum BM001085254 Oakes 1849 N-America, USA, New nd 16.47 ± 2.78 Hamphire 9 D. luridum BM001085257 Macoun 01/06/1901 N-America, Canada, nd 17.15 ± 2.82 Ontario 10 D. luridum BM001085253 G.Merrill 29/10/1910 N-America, USA, Kansas on rocks in bed of a nd 11.30 ± 1.94 stream 11 D. luridum BM001085245 P.W.James 30/06/1959 UK, Scotland, West on inundated rocks nd nd Sutherland 12 D. luridum BM001085243 J.A.Crabbe 22/04/1905 UK, Scotland, Outer nd nd Hebrides 13 D. luridum BM001085241 P.J.Hunt 12/07/1961 UK, Scotland, Outer rocks at edge of nd 15.07 ± 1.20 Hebrides [illegible] in bog by inlet 14 D. luridum BM000974961 W.Johnson Received at BM in UK, North England, on stones in small nd 13.22 ± 2.77 1894 Cumberland stream

15 D. luridum BM001085246 F.J. Walker 28/07/1979 UK, North England, nd 6.75 ± 0.82 Cumberland (under LOQ) 16 D. luridum BM000974963 H.B. Holl Received at BM UK, SW England, Devon nd 19.51 ± 2.68 in 1887

179

17 D. luridum BM001085239 H.B. Holl Received at BM UK, SW England, Devon on wet rocks nd 10.07 ± 0.53 in 1886 18 D. luridum BM000974964 Ex Herb. Received at BM UK, SW England, Devon nd 10.52 ± 1.34 H.B.Holl in 1886

19 D. luridum BM001085250 C.Gueidan 25/08/2010 UK, SW England, Devon on slates in water 15.54 ± 0.98 16.83 ± 0.44 along the river 20 D. luridum BM00108562 H.B.Holl Received at BM UK, SW England, Devon rocks in the bed of the nd 16.57 ± 4.44 in 1886 river 21 D. luridum BM001085261 H.B.Holl Received at BM UK, SW England, Devon nd 27.78 ± 1.91 in 1886 22 D, luridum BM001085248 T.D.V.Swinscow 04/10/1963 UK, SW England, Devon among slate boulders nd 12.77 ± 0.12 in river 23 D. luridum BM001085263 G.Davies 05/1864 UK, SW England, Devon nd 27.04 ± 2.26

24 D. luridum BM000974962 H.B. Holl Received at BM UK, Wales nd 13.09 ± 1.22 in 1888 25 D. luridum BM000920143 V. Howden 06/09/2006 UK, Wales in river 3.65 ± 1.32 9.36 ± 1.79

26 D. luridum BM000974960 H.H.Knight 20/07/1908 UK, Wales, on submerged rocks nd 16.89 ± 1.38 Carmarthenshire 27 D. luridum BM001085244 P.W.James 04/04/1958 UK, Wales, semi-flooded stones nd 8.53 ± 0.91 Pembrokeshire in a fast flowing (under LOQ) stream 28 D. luridum BM001085242 P.W.James 04/04/1958 UK, Wales, semi-flooded stones nd 9.24 ± 1.06 Pembrokeshire in a fast flowing stream 29 D. luridum BM001085249 P.W.James 16/04/1965 UK, Wales, on semi-aq. rocks nd 10.96 ± 3.00 Carmarthenshire 30 D. luridum BM001085233 M. Steiner 25/07/1959 Austria, Tirol on emerged mica 2450 nd 10.36 ± 2.56 (D. cf. miniatum agg) rocks m NN. 31 D. luridum BM001085238 M. Steiner 25/07/1959 Austria, Tirol on emerged mica 2450 nd 6.55 ± 1.10 (D. cf. miniatum agg) rocks m NN. (under LOQ) 32 D. luridum BM001104171 H. Thüs & R. 10/2006 Austria, Salzburg amphibious in stream 1800-2000 2.03 ± 0.21 11.34 ± 2.74 (Dermatocarpon sp.) Tuerk m NN (under LOQ)

33 D. luridum BM001085234 H.A. Imshaug 28/05/1960 N-America, Canada, nd 11.29 ± 3.72 (D. miniatum agg.) Ontario 34 D. luridum BM000974959 A.Wilson 06/1910 UK, North England, rocks on shore of nd 9.15 ± 1.39 (D. miniatum agg.) Cumbria Derwent Water

180

35 D. luridum BM001085240 A.R.Vickery 29/06/1968 UK, Scotland, Inner on rocks beside Caol nd nd (D. miniatum agg.) Hebrides Lochan 36 D. luridum BM000734558 H.Thüs & 05/2011 UK, Scotland, Inner 13.50 ± 1.73 6.44 ± 1.46 (D. miniatum agg.) P.Cannon Hebrides (under LOQ)

37 D. arnoldianum BM001104172 H. Thüs & R. 2006 Austria, Salzburg amphibious in stream 1800-2000 2.01 ± 1.06 13.44 ± 1.56 (D. arnoldianum sensu Tuerk m NN (under LOQ) Thüs & Schultz 2008) 38 D. arnoldianum BM001085258 A. & A. 01/08/1963 Austria, Salzburg on emerged gneiss 2250-2300 nd 7.80 ± 0.66 (D. arnoldianum sensu Schroeppel rocks of a mountain m NN (under LOQ) Thüs & Schultz 2009) stream 39 D. arnoldianum BM001085260 A.J.Huuskonen 19/07/1961 Finland, Laponia in rivulo prope 900 m NN nd 8.13 ± 1.04 (D. miniatum var. deversorium alpinum (under LOQ) miniatum) Societatis Fennicae. supra saxa 40 D. arnoldianum BM001085259 A.J.Huuskonen 28/07/1949 Finland, Laponia ad saxa rivulosa prope nd 11.69 ± 2.51 (D. miniatum var. viam "Haltian miniatum) turistipolku" cop 41 D. arnoldianum BM001085251 Arnold 04/09/1876 Austria, Salzburg on submerged rocks nd 35.02 ± 1.05 (D. rivulorum) 42 D. luridum BM 001085368 22/01/1905 UK, SW England, Devon among slate boulders nd nd in river 43 D. luridum BM 001085376 23/03/1905 Ireland, County Mayo, ad saxa inundata nd 14.08 ± 0.58 Achill Island rivuli Tumpnerbach 44 D. luridum BM 001085371 07/1841 Germany, Odenwald, nd 14.22 ± 0.16 Heidelberg 45 D. luridum BM 001085375 1920 Austria, Tirol nd 8.50 ± 0.10 (under LOQ) 46 D. arnoldianum BM 001085370 1886 Switzerland, St. Moritz nd 8.64 ± 1.08 (under LOQ) 47 D. rivulorum BM 001085358 1886 Switzerland, St. Moritz nd 16.00 ± 4.04 48 D. luridum BM 001085361 05/08/1897 Faroer Islands, Island of nd 4.56 ± 0.84 Skuvoy (under LOQ) 49 D. luridum BM 001085373 1886 UK, SW England, river on rocks in the bed of nd 10.65 ± 2.26 Erme above Ivybridge the river 50 D. luridum JB/001/07/2007 J.L.& B. Martin 10/07/2007 Norway, Bergen on suntan rocks at the 30 m 4.64 ± 0.48 32.85 ± 5.41 edge of a street 51 D. luridum JB/001/08/2005 J.L.& B. Martin 24/08/2005 France, Gard, Sénéchas on rocks close to river 280 m nd 20.07 ± 1.01 (30450) 52 D. luridum JB/001/04/2006 J.L.& B. Martin 18/04/2006 France, Ain, Brénod on stones in the 1230 m nd 33.50 ± 0.27 (01110) meadow of a coomb

181

53 D. luridum JB/001/08/2006 J.L.& B. Martin 29/08/2006 France, Haut-Rhin, Sewen on rocks in border of 615 m nd 27.10 ± 0.52 (68290) the lake 54 D. luridum JB/001/03/2011 J.L.& B. Martin 29/03/2011 France, Ain on rocks in border of 1310 m 11.88 ± 1.28 34.89 ± 1.50 a forest path 55 D. luridum JB/001/09/2011 J.L.& B. Martin 22/09/2011 France, Rhône-Alpes, on vertical rocks in 607 m 15.62 ± 1.38 16.67 ± 2.55 Montromant (69610) border of path 56 D. luridum JB/001/07/2005 J. La Gabrielle 10/07/2005 France, Haut-Rhin, Sewen on flooded rocks 3.82 ± 0.76 22.08 ± 1.07 (68290) 57 D. luridum JB/001/00/2006 N. Lottin 2006 France, Corrèze, 7.85 ± 0.02 25.22 ± 1.19 Chavanac (19290) 58 D. luridum JB/001/08/2013 D.&O. Gonnet 02/08/2013 France, Savoie, saxicole, calcifuge 2250 m 15.35 ± 0.35 8.86 ± 1.17 Lanslebourg-Mont-Cenis (under LOQ) (73480) 59 D. luridum JB/001/07/2012 M. Bertrand 10/07/2012 France, Pyrénées on flooded rocks 1800 m 14.68 ± 0.25 11.74 ± 0.61 Orientales, Porta (66760) 60 D. luridum HL L10/11-5 M. Millot 2010 France, Corrèze on rocks close to river 11.28 ± 3.65 4.72 ± 2.64 (under LOQ) 61 D. rivulorum REN- L.J.-C. Massé 07/09/1973 Austria, Glocknergruppe, 2700 m nd 15.06 ± 3.70 MAS_73_03 Krefelder Hütte 5710) 62 D. fluviatile REN- L.J.-C. Massé 01/09/1963 France, Côtes d'Armor, on rocks of chaos nd 14.93 ± 1.58 (D. luridum) MAS_73_01 Lanrivain (22480) 63 D. luridum REN- L.J.-C. Massé 16/08/1966 Ireland, County Cork on aerohalines cracks nd 8.44 ± 1.20 MAS_73_02 of the slaty rocks (under LOQ) 64 D. miniatum var. REN- L.J.-C. Massé 25/05/1974 France, Mayenne, Saulges on calcareous rocks nd 15.73 ± 2.46 miniatum MAS_73_04 (53340) near Erve 65 D. miniatum var. REN- H. des Abbayes 04/05/1935 France, Ille et Vilaine, ? nd 8.70 ± 0.45 miniatum (morpho Abb_B002_C13_ Fougères (35300) (under LOQ) imbricatum) 06 66 D. miniatum var. REN- H. des Abbayes 15/08/1924 France, Vendée, Le on humid rocks at the nd 7.54 ± 1.05 miniatum Abb_B002_C13_ Tablier (85310) edge of Yon, but not (under LOQ) 01 bathed by the water 67 D. miniatum var. REN- H. des Abbayes 18/06/1958 N-America, USA, along shore on outside nd 8.46 ± 1.21 miniatum Abb_B002_C13_ Michigan of Davidson Island (under LOQ) 10 68 D. miniatum var. REN- H. des Abbayes 09/1932 France, Pyrénées on schist rocks where nd 7.78 ± 0.91 miniatum (morpho Abb_B002_C13_ Orientales, Banyuls water flow during rain (under LOQ) imbricatum) 05 (66650) 69 D. miniatum var. REN- H. des Abbayes 05/1927 France, H.Garonne, nd 10.55 ± 1.27 miniatum Abb_B002_C13_ Bagnères d. Luchon 02 (31110)

182

70 D. miniatum var. REN- H. des Abbayes 04/03/1932 France, Puy de Dôme, on rocks near nd 22.42 ± 3.37 miniatum (morpho Abb_B002_C13_ Royat (63130) Tiretaine imbricatum) 04 71 D. miniatum var. REN- H. des Abbayes 04/05/1935 France, Manche, Mortains on vertical walls of nd 5.09 ± 0.15 miniatum Abb_B002_C13_ (50140) shady Armorican (under LOQ) 08 sandstone 72 D. miniatum var. REN- H. des Abbayes 26/07/1935 France, Puy de Dôme, on shady rocks nd 5.99 ± 0.98 miniatum (morpho Abb_B002_C13_ Compains (63610) (under LOQ) imbricatum) 07 73 D. intestiniforme REN- H. des Abbayes 06/02/1937 France, Cantal 1300 m nd 7.61 ± 0.68 Abb_B002_C13_ (under LOQ) 09 74 D. miniatum var. REN- H. des Abbayes 10/1930 ? nd 16.56 ± 0.89 miniatum (morpho Abb_B002_C13_ imbricatum) 03 75 D. abbayesi REN- H. des Abbayes 09/1932 France, Pyrénées on rocks in gutter nd 7.63 ± 1.11 (D. luridum) ABB_B002_C07_ Orientales, Banyuls where water flow (under LOQ) 01 (66650) during rain 76 Placidium sp1 REN- H. des Abbayes 20/08/1961 N-America, USA, on east facing hillside 1417 m nd 10.95 ± 1.05 ABB_B002_C10_ Colorado, Custer County in open pines 05 77 Placidium sp2 REN- H. des Abbayes 08/07/1961 N-America, USA, in thick young 1386 m nd 8.10 ± 2.75 ABB_B002_C10_ Colorado, Custer County ponderosa pines with 04 open areas 78 Catapyrenium. REN- H. des Abbayes 30/06/1961 N-America, USA, in shady valley 1158 m nd 6.12 ± 1.70 cinereum ABB_B002_C09_ Kentucky, Meade County (under LOQ) 01 79 Placidium rufescens REN- H. des Abbayes 29/03/1935 France, Maine et Loire, on mossy soils and nd nd ABB_B002_C014 Liré (49530) rocks _02 80 D. meiophyllizum REN- H. des Abbayes 05/08/1950 Suède, Närke, Svennevad nd 12.95 ± 2.49 ABB_B002_C12_ 02 81 D. meiophyllizum REN- H. des Abbayes 21/07/1954 France, Côtes d'Armor, on flat stones flooded nd 11.35 ± 0.83 ABB_B002_C12_ Laniscat (22570) by high winter water 01 82 D. aquaticum REN- H. des Abbayes 14/07/1927 France, Côtes d'Armor, on submerged rocks nd 9.07 ± 0.86 (D. luridum) ABB_B002_C08_ Plaintel (22940) 01

183

83 D. aquaticum REN- H. des Abbayes 08/1931 France, Côtes d'Armor, on submerged rocks nd 8.80 ± 0.62 (D. luridum) ABB_B002_C08_ Lanrivain (22480) (under LOQ) 02 84 D. aquaticum REN- H. des Abbayes 09/1932 France, Vendée, Vouvant on submerged rocks nd 21.46 ± 1.39 (D. luridum) ABB_B002_C08_ (85200) (Verreries) 03 85 D. aquaticum REN- H. des Abbayes 30/03/1933 France, Finistère, Hanvec on rocks nd 14.48 ± 3.66 (D. luridum) ABB_B002_C08_ (29460) 04 86 D. aquaticum REN- H. des Abbayes 17/04/1933 France, Côtes d'Armor, on submerged rocks nd 26.40 ± 5.58 (D. luridum) ABB_B002_C08_ Laniscat (22570) 05 87 D. aquaticum REN- H. des Abbayes 17/06/1933 France, Loire Atlantique, on semi-flooded rock nd 14.27 ± 1.98 (D. luridum) ABB_B002_C08_ Massérac (44290) 06 88 D. aquaticum REN- H. des Abbayes 07/1933 France, Pyrénées Orient, on rocks flooded by a nd 9.95 ± 4.21 (D. luridum) ABB_B002_C08_ Argelès sur Mer (66700) torrent 07 89 D. fluviatile REN- H. des Abbayes 24/08/1960 N-America, USA, oak woods and nd 7.14 ± 0.82 (D. luridum) ABB_B002_C08_ Kansas, Greenwood sandstone outcrops (under LOQ) 08 County 90 Placidium lachneum REN- H. des Abbayes 25/06/1961 N-America, USA, in open pines on south 1036 m nd 8.79 ± 0.12 ABB_B002_C11_ Colorado, Custer County facing hill (under LOQ) 01 91 D. leptophyllodes JB/002/09/2013 J-Y. Monnat 09/2013 France, Brittany on submerged rocks 9.52 ± 0.80 9.50 ± 1.64 92 D. luridum JB/001/09/2013 J-Y. Monnat 09/2013 France, Brittany on submerged rocks 13.31 ± 1.32 9.28 ± 1.21 93 D. meiophyllizum JB/003/09/2013 J-Y. Monnat 09/2013 France, Brittany on submerged rocks 20.71 ± 1.00 9.64 ± 1.74 94 D. miniatum JB/004/09/2013 J-Y. Monnat 09/2013 France, Brittany on rocks close to the 23.58 ± 1.38 22.46 ± 0.93 sea

184

Appendix 4: Antioxidant activity of oxo-mycosporines and imino-mycosporines

Compound Method (mechanism) Concentration Radical inititor Standards Results Ref.

Oxo-mycosporine

• Phosphatidylcholine 30 µM AAPH. Ascorbic and uric acids moderate [58] peroxidation (HAT)

-. • Pyrogallol autooxidation 50 – 1000 µM O2 Superoxide dismutase na [171] (scavenging of ROS) at pH 8.2

• β-carotene bleaching (HAT) 10 – 200 µM α-Tocopherol 4–28% inhibition

• ABTS assay (ET) 2.5 – 10 µM ABTS.+ Ascorbic acid

- pH 6 12–36% inhibition (IC50 = 20 µM) Mycosporine glycine - pH 7.5 38–76% inhibition (IC50 = 4 µM)

- pH 8.5 46–82% inhibition (IC50 = 3 µM)

1 7 –1 –1 • Singlet oxygen scavenging 15 – 60 µM O2 α-Tocopherol kt = 5.6 × 10 M s vs α-tocopherol kt = 15.4 × [209] 7 –1 –1 in MeOH/CHCl3 medium 10 M s

• Enzymatic assay [210] -. - Superoxide dismutase O2 Bovine erythrocyte strong

- Catalase H2O2 strong • DPPH (ET) DPPH. α-Tocopherol na [178]

-. • NBT (scavenging of ROS) O2 Trolox IC50 = 350 µM vs trolox IC50 = 15 µM Mycosporine serinol

• Kit Bioxytech AOP-490 (ET) Quercetin and ascorbic weak acid 1 7 –1 –1 Singlet oxygen scavenging O2 α-Tocopherol kt = 5.9 × 10 M s vs α-tocopherol kt = 6.4 × [179] Mycosporine glutaminol glycoside 8 –1 –1 8 –1 –1 in aq. medium Ascorbic acid 10 M s , ascorbic acid kt = 1.6 × 10 M s Phosphatidylcholine 2, 10 – 30 µM AAPH. Ascorbic and uric acids strong [211] 4-Deoxygadusol peroxidation (HAT)

185

Compound/Mixture Method (mechanism) Concentration Radical inititor Standards Results Ref.

Imino-mycosporines • Phosphatidylcholine 50µM, 100 µM AAPH. Ascorbic and uric acids weak [212] peroxidation (HAT) 200 µM strong

Porphyra-334 1 • Lipid peroxidation 0-200 µM O2 Trolox moderate

. • DPPH (ET) DPPH weak [84]

Dehydrated compound of . • DPPH (ET) DPPH Trolox IC50 = 10 µg/mL vs standard IC50 = 5.8 µg/mL [84] porphyra-334 (NAC) • Phosphatidylcholine 10 – 30 µM AAPH. Ascorbic and uric acids weak [211] peroxidation (HAT)

-. • Pyrogallol autooxidation 100 – 1000 µM O2 Superoxide dismutase 13–46% inhibition [171] (scavenging of ROS) at pH 8.2

Shinorine • β-carotene bleaching (HAT) 10 – 200 µM α-Tocopherol 4–69% inhibition

• ABTS assay (ET) 2.5 – 10 µM ABTS.+ Ascorbic acids - pH 6 na - pH 7.5 na at 2.5 µM, 2 – 6% inhibition at 5, 10 µM - pH 8.5 2– 9% inhibition • Ferric thiocyanate (ET) 2 mg α-Tocopherol strong Usujilene [213] • Thiobarbituric acid (ET) 2 mg α-Tocopherol strong -. • ABTS assay (ET) O2 Trolox IC50 = 185 µM vs trolox IC50 = 182 µM 478-Da MAA (a pentose-bound [47] porphyra-334) • DPPH (ET) DPPH. Trolox na

-. 1050-Da MAA (3-aminocyclohexen- • ABTS assay (ET) O2 Trolox IC50 = 55 µM vs trolox IC50 = 182 µM 1-one and 1.3-diaminocyclohexen [47] . and two pentose and hexose sugars) • DPPH (ET) DPPH Trolox IC50 = 809 µM vs trolox IC50 = 128 µM

-. 508-Da MAA (a hexose-bound O2 weak ABTS assay (ET) Trolox and ascorbic acid [48] porphyra-334 derivative) IC50 = 29 mM vs standards IC50 = 0.16 mM

612-Da MAA (a two hexose-bound -. O2 comparable to standards palythine-threonine derivative) ABTS assay (ET) Trolox and ascorbic acid [48] IC50 = 0.25 mM vs standards IC50 = 0.16 mM

186

Mycosporine glycine: palythine: Phosphatidylcholine 1.6 mM AAPH. Ascorbic and uric acids strong inhibition [58] palythinol: asterina-330 peroxidation (HAT) (27.9:9:4.3:1)

Shinorine: mycosporine glycine Phosphatidylcholine 1.2 mM AAPH. Ascorbic and uric acids moderate (2.4:1) peroxidation (HAT)

Asterina-330: palythine Phosphatidylcholine 0.9 mM AAPH. Ascorbic and uric acids moderate (8:1) peroxidation (HAT)

-. Asterina-330: palythine (86:14) • Pyrogallol autooxidation 50 – 1000 µM O2 Superoxide dismutase 15–85% inhibition [171] (scavenging of ROS) at pH 8.2 • β-carotene bleaching (HAT) 10 – 200 µM α-Tocopherol 48–81% inhibition

• ABTS assay (ET) 2.5 – 10 µM ABTS.+ Ascorbic acid - pH 6 na at 2.5 and 5 µM, 2.76% inhibition at 10 µM - pH 7.5 5– 16% inhibition (IC50 = 60 µM) - pH 8.5 12– 47% inhibition (IC50 = 10 µM)

Porphyra-334: shinorine Phosphatidylcholine AAPH. Ascorbic and uric acids moderate [58] (9.6: 1) peroxidation (HAT) 0.9 mM

Porphyra-334: α-tocopherol Phosphatidylcholine 50.02 µM AAPH. Ascorbic and uric acids increase 40% inhibition vs porphyra-334 50 µM [212] (2500: 1) peroxidation (HAT) (after 10 min)

-. Porphyra-334: shinorine (88:12) • Pyrogallol autooxidation 200 – 1000 µM O2 Superoxide dismutase 16– 46% inhibition [171] (scavenging of ROS) at pH 8.2

• β-carotene bleaching (HAT) 10 – 200 µM α-Tocopherol 11–71% inhibition

• ABTS assay (ET) 2.5 – 10 mM ABTS.+ Ascorbic acid - pH 6 na - pH 7.5 na at 2.5 and 5 µM

1.8% inhibition at 10 µM (IC50 = 400 µM)

- pH 8.5 3–15% inhibition (IC50 = 80 µM)

M-324 + M-322 DPPH (ET) 2. 4. 8 mg/mL DPPH. Ascorbic acid 27.6%, 45.8% and 72.2% inhibition, resp. [172]

187

Palythine + asterina -330+ M-312 DPPH (ET) 0.1, 0.2, 0.5 DPPH. Ascorbic acid 14.5%, 53.0% and 68.9% inhibition, resp. mg/mL

Shinorine + M-307 DPPH (ET) 0.2, 0.4, 0.8, 1.6 DPPH. Ascorbic acid 5.6%, 13%, 39% and 64.3% inhibition, resp. mg/mL

Mycosporine taurine + M-343 + In vivo assay 0.4, 4.0, 40 H2O2 Without MAAs but 9.3% inhibition at 0.4 µg/mL [214] dehydroxylusujirene 400 µg/mL treated with H2O2 21% inhibition at 4.0 µg/mL decrease with higher concentrations na: no activity detected; HAT: hydrogen atom transfer; ET: electron transfer; AAPH: 2'-azobis (2-amidinopropane) dihydrochloride; ABTS: 2, 2’-azinobis (3-ethylbenzothiazoline 6-sulfonate); DPPH: 1,1-diphenyl-2-picrylhydrazyl; NBT: nitro-blue tetrazolium; DCFH-DA: 2’, 7’-dichlorodihydrofluorescein diacetate; kt : rate constant

188

VU : VU :

Le Directeur de Thèse Le Responsable de l’Ecole Doctorale

(Nom et Prénom)

VU pour autorisation de soutenance

Rennes, le

Le Président de l’Université de Rennes 1

Guy CATHELINEAU

VU après soutenance pour autorisation de publication :

Le président de Jury,

(Nom et Prénom)