Role of alphaOGG1 in the Maintenance of Mitochondrial Physiology Debora Lia

To cite this version:

Debora Lia. Role of alphaOGG1 in the Maintenance of Mitochondrial Physiology. Cellular Biology. Université Paris Saclay (COmUE), 2018. English. ￿NNT : 2018SACLS125￿. ￿tel-01820633￿

HAL Id: tel-01820633 https://tel.archives-ouvertes.fr/tel-01820633 Submitted on 22 Jun 2018

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

ROLE OF αOGG1 IN THE MAINTENANCE OF MITOCHONDRIAL PHYSIOLOGY

Thèse de doctorat de l’Université Paris-Saclay préparée à

NNT : 2018SACLS125 : NNT l’Université de Paris-Sud

Ecole doctorale n 577 Structure Et Dynamique Des Systemes Vivants (ED-SDSV) Spécialité de doctorat : BIOLOGIE

Thèse présentée et soutenue à Fontenay aux Roses, le 16/05/2018, par

DEBORA LIA Composition du Jury:

Filippo Rosselli President

Directeur de Recherche, Institut Gustave Roussy (FANC/BRCA signalling pathway and Cancers)

Guy Lenaers Rapporteur

Directeur de Recherche, Université d’Angers (PREMMi)

Miria Ricchetti Rapporteur

Group Leader, Institut Pasteur (Stem Cells and Development)

Anne Lombes Examinateur

Group Leader, Institut Cochin (Mitochondrie, Bioénergétique, Métabolisme Et Signalisation)

Anna Campalans Directeur de thèse

Chercheur, CEA (LRIG)

“Non è l’assenza di difetti che conta, ma la passione, la generosità, la comprensione e simpatia del prossimo, e l’accettazione di noi stessi con i nostri errori, le nostre debolezze, le nostre tare e virtù.” – Rita Levi-Montalcini

"Success consists of going from failure to failure without loss of enthusiasm." - Winston Churchill

A mia Madre, il mio sorriso, e a mio Padre, il mio eroe.

- 3 -

ACKNOWLEDGEMENTS

I would never have been able to finish my dissertation without the guidance of my thesis director, the help from friends, and the support from my family… so the acknowledgments will be a bit in English, a bit in French…and a bit in Italian of course  so that everyone could recognize himself in the best way.

First and foremost, I wish to thank my thesis director, Dr Anna Campalans. Annina, you have always been supportive since the first days I began working on our amazing project; Thank you for having helped me to become the person I am today, both professionally and personally. I know very well that I would never have arrived here without your help. You made me grow step by step, every day. Above all, you gave me the enthusiasm and the motivation to wake up and be happy with my work. And I think this is the greater gift a Teacher can give to a student. I feel privileged for the possibility I had to work with you. Thanks also for all the times you got angry or desperate with me, I needed that moments, especially that moments. Thank you for let me discovered the magic world of mitochondria. This thesis is the result of a beautiful journey we did together all these 3 years. You are a bit like family for me and I just want you to know that I own you a lot, and I will always do...

Thanks to Dr Pablo Radicella, for having allowed me to join your group in the first place. Thank you for all the good advices both professional and not, thank you for your kindness. For the Argentine barbecue. Thanks also to your wife, Maria.

My thesis committee guided me through all these 3 years. So, many thanks to Dr Anne Lombes, Dr Erika Brunet, Dr Karl Mann and especially to Dr Miria Ricchetti. Thanks for being part of my thesis committee, for carefully evaluating my work and for your helpful comments. All of you have been an inspiration in many ways. Many thanks also to all the members of my thesis committee.

Je remercie tous les membres et ex-membres de l’équipe LRIG pour le climat sympathique dans lequel ils m’ont permis de travailler. Les nombreuses discussions que j’ai pu avoir avec chacun des vous m’ont beaucoup apporté. Vous étes des gens spéciaux que j’ai eu l’honneur de connaître. Surtout merci à Dr Anne Bravard, Dr Jacquiline Bernardino-Sgherri, Dr Stefanie Marsin et Dr Anne Marie de Guilmi pour notre soirée entre filles…Merci Emilie pour ton bonheur.

I would like to thank Prashant who, as a good friend, was always willing to help and give his best suggestions…and also some hugs in my worst moments. It would have been a lonely lab without you.

Many thanks to Dr Jan Baijer for having taught me everything I know about cell sorting and FACs analysis. Thank you for having followed me in my crazy ideas.

Lamya, merci pour m’avoir si facilement suivie dans ma folie la plus totale, pour m’avoir accordé ta confiance et pour tous les beaux moments partagés ensemble.

- 5 -

Grazie Mary (la mia Abib). Insieme abbiamo imparato moltissimo sulla natura umana e su noi stesse in questi ultimi mesi. Mi hai raccolta che ero incasinata, mi hai dovuta cazziare, mi hai dovuto sopportare ma non sei scappata via. Grazie per esserci amica mia.

Grazie Eleonora (il mio Orgoglio). Non sono sicura che io sarei stata in grado di iniziare questo dottorato se non ti avessi incontrata. Grazie per avermi iniziato ai vini alsaziani ed ai cetrioli. Grazie per le sessioni di psicoterapia. Con te Parigi è diventata veramente casa mia… alla fine l’Universo è stato ed è davvero positivo.

Grazie Paola (il nostro Unicorno). Grazie per esserci anche se a distanza.

Grazie Elisa (la mia Topa). Le parole non mi bastano…Grazie per avermi fatto ritrovare il senno tutte le volte che lo perdevo. Grazie per il tuo calore, grazie per le nostre risate. Grazie per avermi accolta.

Grazie a Stefano, Gladys e Giacomo.

Thank you Haser, my dear friend, I know I’ve got a friend in you and I hope you know that you will never find my door close.

Grazie Parigi per tutti i tuoi Serbi, i tuoi Shannon, i tuoi Tournesol e le tue L’adada. Grazie ai nuovi amici (Salvatore (Babe), todos los amigos de el BoBo, Ghazi & Dasha, Elma, Ivana, Stafania (la mia Donnaccia)...) che me li hanno fatti scoprire insieme ad una città che mi ha accolta. Grazie Martina (la bella Pazza) sono felice di averti incontrata. Grazie Parigi, perché tra le tue rues non mi perdo.

Grazie agli amici di sempre, il mio porto sicuro: Emanuele, Nicola, Marco e Carola. Grazie per tutte le serate Scottish degli ultimi 12 anni. Grazie Ema (Cuore) per tutte le volte che mi hai detto: “Forza che me devi di’ mo?”. Grazie Nicola (Nata) per tutti i tuoi consigli saggi che io non seguo mai.

Ed in fine ma non per importanza, vorrei ringraziare la mia famiglia. La mia incredibile, unica famiglia. Durante i miei tre anni di dottorato il vostro appoggio è stato incondizionato e non avete idea di quanto mi sia stato necessario.

Grazie Nonna per la scheda telefonica fatta apposta per chiamare la nipote lontana. Grazie Zia Roberta e Zia Rossana per la vostra missione di salvataggio. Zia Rita per i vasetti di sottoli. Grazie alle mie sorelle/cugine per le pizze “all’andata e al ritorno”.

Grazie Mamma e grazie Papà. Voi siete la mia forza. Ho talmente tanti motivi per ringraziarvi che faccio fatica a dirne anche uno solo…grazie per tutte le volte che vi ho visti in prima fila ad aspettarmi all’aeroporto tornando a casa, per i pacchi stile tetris da giù, per avermi tollerata nei miei momenti critici, con i miei bronci e con i miei pianti isterici pieni di parolacce. Grazie per le chiamate su Skype ed i nostri saluti pieni di smorfie, risate, boccacce…grazie per le foto di Ginetta, per le lenzuola sempre pulite in camera mia. Grazie per le melanzane alla parmigiana. Grazie per tutte le piccole grandi attenzioni. Grazie per il Vostro Amore.

- 7 -

ACKNOWLEDGEMENTS 5 INDEX 9 RESUME 11 ABBREVIATIONS 21 Chapter I. INTRODUCTION 23

I Structure and functions of mitochondria 25 a Mitochondria, an overview 25 b The electron transport chain and mtROS production 26 c Structure of mtDNA nucleoids and their replication and transcription 29 d Impact of mtDNA alterations on medicine 36 II Quality controls of mitochondria 39 a ROS scavenging 39 b Mitochondrial Unfolded Proteins Response 40 c The maintenance of a healthy mitochondrial population: the fission/fusion 42 mechanisms i Mitochondrial inner membrane topology 45 d mtDNA maintenance: how to repair or tolerate DNA damage in 46 mammalian mitochondria i Mismatch repair pathway (MMR) 49 ii Homologous recombination and non-homologous end-joining 50 iii Translesion synthesis 50 iv mtDNA degradation 52 e Mitophagy programs 53 f The final response: apoptosis 54 Concluding remarks on mitochondrial dynamics, mitophagy and 57 apoptosis III Base excision repair in mitochondria 61 a Modified base excision and strand incision 64 b Gap filling by DNA polymerase and strand sealing by DNA ligase 66 IV The 8-oxoguanine lesion 67 a G:C to T:A transversion and the GO repair system 68 b The DNA glycosylase OGG1 protein 70 c The absence of OGG1 activity in the cell… what do we know so far? 74

Chapter II. PURPOSE OF THE WORK 79 Purpose of the work 81

Chapter III. RESULTS 83 Results – Section 1 85 a Characterization of the mitochondrial localization of αOGG1 85 b Role of the localization sequences at the N and C ter of αOGG1 in the sub- 89 cellular distribution of the enzyme Results – Section 2 95 a Evaluation of the mitochondrial physiology in OGG1 deficient cells 95 b Mitochondrial dysfunction of OGG1 deficient cells can be complemented 98 by expressing a functional αOGG1 in mitochondria c Influence of oxidative stress on mitochondrial localization of αOGG1 104 d Menadione induces the accumulation of 8oxoG in mtDNA 106

Preliminary results 110

- 9 -

a Quantification of mtDNA and mtRNA 110 b Evaluation of apoptosis in U2OS cells in the absence of OGG1 113

Chapter V. DISCUSSION AND CONCLUSIONS 117 Discussion and general conclusions of the work…many doors open 119

Chapter V. APPENDIX 133 Detailed protocols 135 TABLE 146 Submitted article Mitochondrial maintenance under oxidative stress depends on 149 mitochondrial but not nuclear α isoform of OGG1

BIBLIOGRAPHY 187 Bibliography 189

Resumé Le stress oxydant : Les mitochondries sont des organites présents dans la plupart des cellules eucaryotes où elles jouent un rôle majeur dans la production d'ATP cellulaire à travers la β-oxydation, le cycle de Krebs et la phosphorylation oxydative. Outre cette fonction essentielle dans le métabolisme énergétique, ces organites interviennent également dans la signalisation, la différenciation et la mort cellulaire, ainsi que dans le contrôle du cycle cellulaire.

Dans les cellules de mammifères, on suppose souvent que, parmi toutes les autres sources d’espèces réactives de l’oxygène (ERO) cellulaires, la chaîne respiratoire mitochondriale (ETC) est la source plus importante quantitativement et une des principales conséquences d’un dysfonctionnement mitochondrial est la surproduction des ERO. En fait, dans les complexes ETC, plusieurs centres redox sont présents qui peuvent transférer directement

− un électron à l'oxygène générant le radical anion superoxyde (O2 ). Même si à de faibles concentrations les ERO sont des composants essentiels de différentes voies de signalisation, un excès peut endommager plusieurs composants des mitochondries et peut également influencer l'assemblage correct des respirasomes, conduisant à un dysfonctionnement cellulaire. En fait, l'inhibition des activités ETC conduit au bloc de transfert d'électrons le long de la chaîne respiratoire, réduisant ainsi le potentiel de la membrane et les niveaux d'ATP cellulaire. Les électrons accumulés peuvent ensuite réagir avec de l'oxygène pour former des radicaux superoxyde, ce qui augmente la formation de EROmt.

Les ERO peuvent endommager les macromolécules cellulaires, et notamment l’ADN (mitochondrial, ADNmt et nucléaire, ADNn) et produire des douzaines de lésions chimiquement distinctes qui faussent la structure et les propriétés codantes de la molécule et peuvent provoquer des mutations et des aberrations chromosomiques. Les mutations dans l'ADNmt ou l'ADNn peuvent entraîner un déficit de l'ETC ou peuvent donner lieu à un dysfonctionnement mitochondrial global en affectant la signalisation, la transcription ou la traduction, provoquant ainsi un éventail varié et souvent dévastateur de maladies mitochondriales humaines. Les mitochondropathies regroupent un ensemble très vaste de maladies qui se caractérisent par un défaut de la chaine respiratoire. Elles touchent en général les tissus ayant un besoin énergétique important (muscles, cerveau, rein…). L’ADNmt, parce qu’il est au contact direct de la production des ERO, il est plus affecté que

- 11 - l’ADNn par le dommage oxydant (Alexeyev et al., 2008). Des études récentes montrent que la stabilité de l’ADN mitochondrial joue un rôle majeur dans le développement de plusieurs désordres chez l’homme. Les ERO favorisent l’induction de dommages à l’ADN ou la perturbation de la croissance ou de la mort cellulaire et donc contribuent à la carcinogenèse. En fait, des altérations de l’expression des gènes mitochondriaux (codant les complexes I, III, IV et V de l’ETC) ont d’ailleurs été trouvées dans de nombreux cancers chez l’homme (Chatterjee et al., 2006).

La plupart des maladies induites par le stress oxydant apparaissent avec l’âge car le vieillissement diminue les défenses anti-oxydantes et augmente la production mitochondriale de radicaux. Ça c’est le cas des maladies neurodégénératives comme, par exemple, Alzheimer ou Parkinson où des anomalies mitochondriaux sont impliquées (Wang et al., 2005).

La réparation par excision des bases : Les bases nucléiques sont soumises à l’attaque des radiations UV, d’agents alkylants et oxydants. Sur l’ADN les ERO sont une source d’altération des bases génomiques : on estime que 105 lésions oxydantes sont formées par cellule et par jour. La voie de réparation de l'excision de base (BER) est un mécanisme de réparation de l'ADN très conservé et est la voie principale pour l’enlèvement de ces dommages de bases d'ADN non volumineux (Wallace, 2014). Le mécanisme se déroule en quatre ou cinq étapes suivant la nature de la lésion. Tout d'abord une ADN glycosylase élimine la base endommagée et produit un site abasique (ou site AP). Une endonucléase spécifique clive ensuite le désoxyribose de ce site abasique. Une ADN polymérase remplit à nouveau l’espace libéré en utilisant la base intacte du brin opposé comme matrice. Enfin, une ADN ligase suture le brin réparé. Toutes les enzymes impliquées dans le BER sont codées par le génome nucléaire et doivent être importées dans les mitochondries.

Étant donné que la guanine a le potentiel redox le plus bas de toute autre base dans l'ADN, elle est facilement oxydée à la 8-oxoguanine (8-oxoG), qui est l'altération la plus fréquente induite par les ERO sur l'ADNn et l'ADNmt. Contrairement aux lésions UV qui induisent des distorsions de la structure de l’ADN, la 8-oxoG ne crée pas de déformation de l’ADN. Le fait que le carbone en C8 oxydé se trouve complètement enchâssé au sein de la double hélice rend cette lésion pratiquement « invisible ». Il ne s’agit pas d’une lésion bloquante pour les polymérases. Cependant, elle présente un potentiel mutagène très important. En effet, en

fonction de sa conformation (syn ou trans), la polymérase réplicative peut l’apparier à une cytosine (selon le mode classique de liaisons hydrogènes proposées par Watson-Crick) ou à une adénine (appariement de Hoogsteen). Si l’appariement 8-oxoG:A n’est pas réparé avant un 2ème cycle réplicatif, des transversions GC vers TA vont apparaître et ce type de mutations a fréquemment été recensé dans des oncogènes et des gènes suppresseurs de tumeurs. L’accumulation de 8-oxoG est limitée par l’activité de plusieurs enzymes : MTH1, une protéine qui régule l’incorporation de 8-oxoG dans le génome, MYH l’enzyme responsable de l'élimination des adenines mal incorporées face 8-oxoG et OGG1.

OGG1 est l’ADN glycosylase qui intervient dans la première étape de la réparation de la 8- oxoG par la voie de réparation par excision de bases (BER pour Base Excision Repair) dans le noyau ainsi que dans la mitochondrie. L’importance du rôle d’OGG1 dans la mitochondrie a été démontrée par des expériences de surexpression d’une forme recombinante d’OGG1 sur laquelle un signal d’adressage à la mitochondrie (MTS) supplémentaire a été rajouté ( Dobson et al., 2000). La surexpression de cette protéine permet de réparer plus efficacement la 8-oxoG mitochondriale, assure une meilleure maintenance du génome mitochondrial, augmente la viabilité cellulaire et favorise la neurogenèse. Cependant, les approches menées dans les études précédentes ne permettent pas de séparer les fonctions mitochondriales et nucléaires d’OGG1. De fait, la surexpression d’OGG1, même lorsque un MTS additionnel est ajouté, résulte en une augmentation de son niveau dans les deux compartiments cellulaires.

Mon projet vise à élucider les mécanismes qui mènent à un dysfonctionnement des mitochondries et/ou à la mort cellulaire lorsque la réparation de la 8-oxoG est spécifiquement affectée dans la mitochondrie. Différentes formes variantes d’OGG1 portant des mutations affectant l’activité enzymatique et pouvant ou pas être adressées à la mitochondrie ont été utilisées dans des approches de biologie moléculaire et cellulaire afin de mieux comprendre ce processus.

Dans la première partie de mon project je me suis intéressée à la caractérisation de la protéine OGG1. Mon étude présentée dans l’article joint, a montré que le signal d’adressage à la mitochondrie (MTS) localisé dans la région N-terminale d’OGG1 était fonctionnel et qu’il permettait l’import d’une partie du pool d’OGG1 dans la mitochondrie malgré la présence

- 13 - d’une séquence de localisation nucléaire (NLS) en C-terminal de la protéine. De plus, nous avons montré que deux sites d’initiation de la traduction sont présents en N-terminal d’OGG1 : une initiation de la traduction à la première méthionine (OGG1-AUG1 appelée dans la suite du texte OGG1-WT) résulte en une protéine fonctionnelle qui présente une double localisation nucléaire et mitochondriale, alors qu’une initiation de la traduction à la deuxième méthionine (OGG1-AUG2) conduit à une protéine fonctionnelle mais qui possède un MTS tronqué et donc une localisation exclusivement nucléaire. Nous avons par ailleurs montré par microscopie confocale et de super-résolution SIM-3 couleurs, que la protéine OGG1-WT mitochondriale colocalise avec l’ADNmt au sein des nucléoïdes.

Dans la seconde partie de mon travail nous avons aussi étudié le fonctionnement des mitochondries dans des cellules déficientes en OGG1, en situation basale ou après un stress oxydatif induit par un traitement des cellules avec de la ménadione. Après exposition à ces stress oxydatifs, les cellules déficientes en OGG1 présentent une diminution plus importante du potentiel de membrane mitochondrial et des niveaux plus élevés de ROS mitochondriaux. Ces phénotypes ont été observés dans des cellules d’osteosarcome humain (U2OS) transfectées avec des siRNA dirigés contre OGG1.

Les cellules déficientes en OGG1 ont ensuite été complémentées avec les protéines OGG1- WT (localisée dans le noyau et la mitochondrie) et OGG1-AUG2 (localisée exclusivement dans le noyau). Les séquences codant pour ces protéines ont été modifiées par des mutations silencieuses afin qu’elles ne soient pas affectées par les siRNA ciblant la protéine OGG1 endogène. La caractérisation de ces lignées, présentée dans le papier, a montré que la présence de la protéine OGG1-WT dans la mitochondrie est essentielle pour l’intégrité de la fonction mitochondriale après une exposition à un stress oxydant induit par la menadione. En effet, l’expression d’OGG1-WT dans les cellules déficientes en OGG1, contrairement à la forme exclusivement nucléaire, permet de préserver le potentiel de membrane des mitochondries et réduit la production de ROS mitochondriaux. Ces observations indiquent que la présence d’OGG1-WT dans la mitochondrie est indispensable à la fonction mitochondriale et qu’il ne s’agit pas d’un effet indirect dû à une meilleure réparation de l’ADN nucléaire.

Les mitochondries sont des organites extrêmement dynamiques qui subissent de façon continue des processus de fusion / fission qui sont essentiels pour le maintien de la fonction

mitochondriale (Ono et al., 2001; Youle & Narendra, 2011). Les cellules disposent d’un système de contrôle de qualité des mitochondries qui permet d’éliminer les mitochondries endommagées tout en préservant l’intégrité du réseau mitochondrial. La fission des mitochondries permet d’isoler les organites non fonctionnels qui sont éliminés par mitophagie. La fusion mitochondriale permet de restaurer un réseau mitochondrial fonctionnel. Les processus de fusion et de fission sont importants pour le maintien de l’intégrité des génomes mitochondriaux.

L’exposition des cellules à un stress oxydant fort résulte en une fragmentation massive des mitochondries. Afin d’évaluer la morphologie des mitochondries dans des cellules exprimant les différentes formes d’OGG1 après un traitement a la menadione, nous avons mis en évidence le réseau mitochondrial par marquage des cellules avec le MitoTracker rouge. Dans des cellules exprimant uniquement OGG1 endogène, le stress oxydant induit par la ménadione a comme conséquence une forte fragmentation du réseau mitochondrial. Par contre, dans les cellules qui surexpriment la forme OGG1-WT, la morphologie des mitochondries est préservée après le traitement et ces organites restent allongés et interconnectés comme dans des cellules non soumises à un stress oxydant. Cette observation suggère que le niveau endogène d’OGG1 est limitant lors d’une exposition des cellules à un stress oxydant. Par contre, les cellules exprimant les formes exclusivement nucléaires d’OGG1 (OGG1-AUG2) présentent après un stress oxydant un réseau mitochondrial fragmenté, indiquant que la présence d’OGG1 active dans la mitochondrie est essentielle pour la protection des mitochondries. Cette fonction protectrice nécessite l’activité ADN glycosylase d’OGG1 puisque l’expression dans la mitochondrie du mutant catalytique OGG1-K249Q conduit également à un réseau mitochondrial fragmenté après stress.

Lorsque les dommages mitochondriaux sont trop importants et qu’une fragmentation du réseau mitochondrial ne suffit plus à isoler puis à éliminer les mitochondries défectueuses, les mêmes mécanismes peuvent aboutir à la mort cellulaire (Furda et al., 2012; Shokolenko, et al., 2009). Mes résultats préliminaires montrent que les dommages induits par un stress oxydant fort conduisent la cellule vers l’apoptose en l’absence d’une forme fonctionnelle d’OGG1 dans la cellule.

- 15 -

Des études en microscopie de super-résolution suggèrent que chaque nucléoïde possèderait une seule copie d’ADNmt qui assure de façon localisée une transcription couplée à la traduction et à la biogenèse mitochondriale. Nos observations sur cellules fixées montrent que les protéines OGG1-WT et OGG1-K249Q sont toujours associées avec l’ADNmt, même en absence de traitement et indépendamment du cycle cellulaire et de l’état de réplication de l’ADN mitochondrial. Cette localisation pourrait assurer une réparation très rapide des 8-oxoG induites par le stress oxydant. Plusieurs études montrent une rapide diffusion des protéines localisées dans la matrice mitochondriale et la membrane externe au sein du réseau mitochondrial. Les protéines de la membrane interne et en particulier celles associées au nucléoïde présentent une diffusion beaucoup plus restreinte, ce qui donne une hétérogénéité intramitochondriale et qui pourrait permettre l’élimination sélective des mitochondries non fonctionnelles par mitophagie. Lors d’un traitement long avec la ménadione des cellules surexprimant le mutant OGG1-K249Q, nous avons remarqué que le pourcentage de cellules possédant un signal OGG1 mitochondrial diminuait, alors qu’il restait stable (voire augmentait) dans les cellules qui surexprimaient OGG1-WT. Ces expériences ont été réalisées dans des cellules qui expriment la forme endogène d’OGG1 en plus des formes surexprimées OGG1-WT et OGG1-K249Q. Une possibilité pour expliquer ces observations serait que l’accumulation de 8-oxoG dans le génome mitochondrial, qui ne peut pas être réparée par OGG1-K249Q, conduirait à un dysfonctionnement mitochondrial et à une dégradation ciblée de cet organite par mitophagie.

Enfin, une autre partie de ce projet est la compréhension des mécanismes qui, en l’absence de l’activité d’OGG1 dans la mitochondrie, induisent des dysfonctionnements mitochondriaux et conduisent à la mort cellulaire. L’ADNmt code seulement 13 protéines qui appartiennent toutes aux complexes de la chaine respiratoire mitochondriale, deux ARNr et 22 ARNt qui sont nécessaires pour la traduction des protéines mitochondriales. L’accumulation de 8-oxoG dans l’ADNmt pourrait contribuer à l’accumulation de délétions et même à la perte d’ADNmt qui ont été observées dans des cellules déficientes en OGG1 (Rachek et al., 2002). Ces évènements se traduisent notamment par une diminution de la respiration mitochondriale, par une augmentation de la production de ROS mitochondriaux et par une fragmentation du réseau mitochondrial. Après un stress oxydant, une perte d’activité plus importante des complexes respiratoires, contenant des sous-unités codées

par l’ADNmt mitochondrial, est en général observée. Cependant, en contradiction apparente avec ces observations, plusieurs études montrent qu’une perte partielle d’ADNmt n’a que très peu de conséquences sur les niveaux de transcription, sur la physiologie des mitochondries et sur la survie des cellules.

Au-delà de la perte de l’ADN mitochondrial, la présence de 8-oxoG dans le génome mitochondrial pourrait également avoir un impact sur la transcription mitochondriale. Des études in vitro ont montré que l’ARN polymérase mitochondriale est capable de transcrire à travers une 8-oxoG, mais avec un très fort taux d’erreur qui pourrait se traduire par la synthèse de protéines mutantes pouvant conduire à un dysfonctionnement de la chaine respiratoire (Nakanishi et al., 2013). Une autre possibilité est que la présence de 8-oxoG dans l’ADNmt conduise à une diminution globale du taux de transcription. Une particularité du génome mitochondrial est qu’il il est transcrit en un seul précurseur polycistronique subséquemment processé par différentes modifications post-transcriptionnelles. Donc, même si le niveau de transcription est identique pour tous les gènes mitochondriaux, les niveaux des transcrits spécifiques des gènes peuvent varier en fonction de leur stabilité. Afin de pouvoir mesurer l’impact de la présence ou de l’absence d’OGG1 sur la transcription en réponse à un stress oxydant, j’ai utilisé des oligonucléotides spécifiques des différents ARNmt (Taanman, 1999). Du fait que l’ADNmt ne présente pas d’introns, les mêmes oligonucléotides peuvent être utilisés pour quantifier les niveaux d’ADN (par qPCR) et d’ARN (par qRT-PCR) afin d’évaluer si le niveau de perte d’ADNmt peut à lui seul expliquer la diminution du niveau de transcription ou si la présence de 8-oxoG dans l’ADNmt a un effet direct sur la transcription. Mes resultates suggerent que, le traitement par menadione n'induit qu'une légère réduction de l’ADNmt total dans les cellules transfectées avec le siRNA contre OGG1 par rapport aux cellules transfectées par le siRNA control. Par contre, les niveaux de certains transcrits (tels que ND1) ont été considérablement réduits dans les cellules déficientes pour OGG1 alors que les niveaux d’expression de COX5, une sous-unité de la chaîne respiratoire codée par l'ADN nucléaire, n'ont pas été affectés par le traitement

oxydant. Afin d’analyser plus en détail les conséquences d’un défaut d’OGG1 dans la mitochondrie, des optimisations des sondes qui sont à cheval entre deux gènes sont maintenant en cours dans mon laboratoire.

- 17 -

Conclusion et perspectives : L’ensemble des observations obtenues avec mon travail soulève la question du rôle de la glycosylase OGG1 et donc de la réparation par excision de bases de l’ADN mitochondrial dans la physiologie de cette organelle et par extension de la cellule.

Il a été mis en évidence qu’un défaut de réparation de la 8-oxoG dans la mitochondrie pourrait conduire à la mort cellulaire par apoptose (notre résultats, Dobson et al., 2002; Hill & Evans, 2007; Rachek et al., 2002; W. Wang et al., 2011) . Plusieurs évidences suggèrent que les cellules cancéreuses sont plus sensibles au stress oxydant, et en particulier à l’accumulation de 8-oxoG dans leur génome que les cellules saines (Imaizumi, Kwang Lee, Zhang, & Boelsterli, 2015). Une inhibition de l’activité OGG1 pourrait donc spécifiquement provoquer la mort cellulaire des cellules tumorales après qu’elles aient été soumises à un stress oxydant (Gad et al., 2014; Huber et al., 2014). Des inhibiteurs d’OGG1 ont été récemment développés, et leur spécificité a été étudiée via des approches in vitro. Nous proposons d’évaluer l’effet de ces inhibiteurs in vivo, sur les cellules exposées ou non à un stress oxydant. Nous voulons tout d’abord analyser la capacité de ces inhibiteurs à pénétrer dans les cellules et à inhiber la réparation de la 8-oxoG dans les génomes nucléaires et mitochondriaux. Pour ce faire, des cinétiques de réparation après induction des dommages seront réalisées, en utilisant différents approches: par immunofluorescence en utilisant un anticorps spécifique contre la 8-oxoG, ou par des techniques de PCR optimisées pour évaluer le niveau de lésions dans la molécule d’ADNmt. Si les inhibiteurs sont fonctionnels in vivo, leur impact sur le fonctionnement des mitochondries et sur la survie cellulaire sera ensuite évalué.

Les résultats obtenus lors de mon stage de doctorat ont ouvert de nouvelles perspectives sur le mode d'action d’OGG1 dans les mitochondries. L'une d'entre elles est l'analyse de l'implication de cette glycosylase dans la modulation de la transcription mitochondriale comme un nouveau mécanisme par lequel cette protéine participe à la préservation de la physiologie mitochondriale. Nous proposons que cette fonction protectrice puisse être réalisée par l’enlèvement direct de la 8-oxoG. Grâce aux outils développés au laboratoire, nous proposons donc d’analyser les mécanismes moleculaire par lesquels OGG1 agit dans cet organite pour pallier les effets d’un stress oxydant. La compréhension de ces

mécanismes et des conséquences de la modulation de l’activité et/ou la localisation de l’ADN glycosylase sera utile afin d’explorer son potentiel comme cible thérapeutique.

- 19 -

ABBREVIATIONS  AIF, apoptosis inducing factor  ANT, adenine nucleotide translocase  BER: base excision repair  CPT, carnitine palmitoyl transferase  Cyp-D, cyclophilin D  Cyt C, cytochrome C  D-Loop, displacement loop  EndoG, endonuclease G  ETC, electron transport chain  FAD, flavin adenine dinucleotide  GSPx, glutathione peroxidase  H2O2, hydrogen peroxide  HR: homologous recombination  IMM, inner mitochondrial membrane  MMP: mitochondrial membrane potential  MMR: mismatch repair  MnSOD, manganese superoxide dismutase  MOMP: mitochondrial outer membrane permeabilization  MPT: mitochondrial permeability transition  mPTP, mitochondrial permeability transition pore  mtSSB, mitochondrial DNA single stranded binding protein  MYH (or MUTYH), MutY DNA glycosylase homologs  nDNA, nuclear DNA  NHEJ: non-homologous end-joining  NO, nitric oxide  O2−, superoxide radical

 OH, origin of heavy strand replication  OH , hydroxyl radical  ONOO−, peroxynitrite  OXPHOS, oxidative phosphorylation  PCD, Programmed cell death  POLγ, DNA polymerase gamma  ROS, reactive oxygen species  Smac, second mitochondria-derived activator of caspase  TFAM, mitochondrial transcription factor A  VDAC, voltage dependent anion channel

- 21 -

Chapter I. INTRODUCTION

- 23 -

I. Structure and functions of mitochondria

I.a. Mitochondria, an overview

Throughout the years, the interest in mitochondria has increased constantly. Human cells typically contain several hundreds of cytoplasmic mitochondria, their number depending on the tissue’s energy needs. These fascinating organelles are at the crossroad of many essential pathways in mammalian cells. In fact, not only they are the powerhouses of the cells by generating adenosine triphosphate (ATP) via oxidative phosphorylation but they also participate in the biosynthesis of steroids, pyrimidines and heme, in the homeostasis of calcium and iron and in the control of apoptosis. But a unique feature of these organelles is the fact that they possess their own genome, a small, circular molecule of 16kDa. The presence of this genetic system supports the hypothesis that around two billion of years ago these organelles were originally α- proteobacteria which at some point entered the ancestral eukaryotic cell and

fused with it becoming an endosymbiotic Figure 1. Schematic representation of a eukaryotic mitochondrion; the main compartments of the organelle entity (Sagan, 1967). are indicated.

During evolution, most Cell Biology by Gerald Karp mitochondrial genes were transferred to the nucleus, but the mitochondrial genome (mtDNA) in mammals still codes for 13 proteins, all involved in the electron transport chain, 2rRNA and 22 tRNA, essential for mitochondrial protein synthesis (Anderson S. et al., 1981). mtDNA is located in the matrix of mitochondria, an aqueous subcompartment enclosed by two membranes, the outer and the inner mitochondrial membranes (Outer Mitochondrial Membrane (OMM) and Inner Mitochondrial Membrane (IMM)) (Fig. 1). The smooth outer In the 1950s, seminal electron membrane has a similar composition to that of the cellular one and is microscopy studies led to the canonical view of mitochondria extremely rich in sphyngolipids (mainly phosphatidylcholine and as bean-shaped organelles and the discovery that phosphatidylethanolamine), which make this compartment highly mitochondria contained nucleic acids was made in the 1960s. permeable to ions and small molecules. The inner membrane is

- 25 - enriched in proteins and cardiolipin (diphosphatidyl glycerol), which contribute in determining its permeability properties and membrane potential. In fact, the passage of protons across the IMM is tightly regulated to generate a proton gradient between the 2 faces of the IMM, which is necessary for the central role played by mitochondria in the production of energy.

The vast majority of mitochondrial proteins are encoded by the nucleus, synthesized in the cytosol and imported into the organelle. In vertebrate cells it has been estimated that more than 3000 different proteins are actually imported in these organelles (Richly et al. 2003). During the transition from bacterial symbiont to proto-organelle, mitochondrial proteome evolution has been characterized by the loss of many original functions, the retargeting of others to different cellular structures and the wholesale addition of host- derived proteins—a veritable “hijacking of mitochondrial protein synthesis and metabolism” by the eukaryotic cell (Gabaldón & Huynen 2007).

In conclusion, thanks to their small but highly specialized DNA, these multifunctional organelles control the bioenergetics status of eukaryotic cells. In the following sections I will discuss the challenge of the maintenance of mtDNA and how a better understanding of mitochondrial genome preservation is fundamental for the comprehension of cellular physiology.

I.b. The electron transport chain and mtROS production

As has already been said, one of the main functions of mitochondria is to produce energy in the form of ATP from nutrients, mainly glucose and lipids, through the oxidative phosphorylation reaction.

This reaction consists in a multi-step sequence in which several redox centers are involved. These multiheteromeric structures (Fig. 2) comprise four protein complexes embedded in the inner mitochondrial membrane cristae. The first and second complexes, NADH-Coenzyme Q oxidoreductase (Complex I) and succinate-Coenzyme Q oxidoreductase (Complex II), respectively, pass electrons down the gradient of redox potential to a lipid- soluble carrier, ubiquinone (Coenzyme Q), to form ubiquinol. From this compound, electrons are transferred through Coenzyme Q-cytochrome c oxidoreductase (Complex III) to cytochrome c, another mobile carrier, which itself transfers the electrons to cytochrome c oxidase (Complex IV). Finally, Complex IV transfers the electrons to the final acceptor, oxygen, to produce water. Over 90% of the oxygen consumed in the cell is reduced to water during the reaction catalysed by Complex IV.

Figure 2 Creative drawing of the respiratory chain and human mitochondrial DNA.

Figure 2. Left: respiratory chain complexes. Mitochondrially encoded subunits, embedded in the midst of nuclear-encoded subunits, are shown in different colors: Complex I accomplishes the oxidation of NADH derived by the oxidation of fatty acids, pyruvate and amino acids, contains 7 subunits which are encoded by the mtDNA (subunits ND1–ND6 and ND4L) and at least 39 nuclear-encoded subunits; Complex II is entirely encoded from nDNA, accomplishes the oxidation of FADH2 derived from fatty acids and the Krebs' cycle and is composed of 4 subunits; Complex III holds 1 subunit, cytochrome b, encoded by the mitochondrial genome, and 10 subunits encoded by the nuclear genome; Complex IV is composed of 13 subunits, 3 of which are encoded by mtDNA (COX I–III) and the other 10 by nDNA; Complex V is composed of 2 mtDNA- encoded subunits (ATPase 6 and 8) and at least 13 nDNA-encoded subunits. Pi = inorganic phosphate; Cyt c = cytochrome c and CoQ = coenzyme Q are 2 highly hydrophobic, mobile, small electron carriers, both synthesized by nuclear genes.

Right: mtDNA. myt genes: complex I genes = blue; complex III cytb gene = green; complex IV genes = red; complex V genes= yellow. syn genes: tRNA genes = grey; rRNA genes = purple. Cyt b = cytochrome b; COI = complex I; COII = complex II; COIII = complex III.

Adapted from Zeviani et al., 2004

Transfer of electrons via the ETC from NADH or FADH2 to O2 creates an electrochemical proton gradient between the two versants of the inner mitochondrial membrane. This proton gradient couples ETC and oxidative phosphorylation (Walker et al.,

- 27 -

1995) to generate ATP from ADP when the protons diffuse down the gradient through the ATP synthase (Complex V).

The formation of the respiratory complexes is under the control of the nuclear and mitochondrial genomes (Fig. 2). Four of the five respiratory chain complexes (I, III, IV and V) may be considered as genetic mosaics because they contain both nuclear-encoded and mtDNA-encoded polypeptides. For this reason, a bidirectional flow of information between the nuclear genome and the mitochondrial genome is required for the production of energy in these organelles. Nowadays it is recognized that the mitochondrial respiratory chain complexes do not float passively in the lipid bilayer of the IMM but are arranged in supramolecular assemblies (respirasomes) (Schäfer et al., 2006; Perez et al., 2008). These structures support the proper and efficient channelling of electrons, together with the stabilization of the ETC.

It is often assumed that in mammalian cells the mitochondrial respiratory chain is quantitatively the most important cellular source of ROS (Lambert et al., 2009). In fact, in ETC complexes several redox centers are present and they may directly transfer an electron

− to oxygen generating the superoxide anion radical (O2 ). The spontaneous or enzymatic dismutation (which is a redox reaction in which a compound of intermediate oxidation state converts to two different compounds, one of higher and one of lower oxidation states) of such a radical yields hydrogen peroxide (H2O2), which, in turn, can produce the highly reactive hydroxyl radical ( OH). At low concentrations, ROS are essential components of different signalling pathways. In fact, gene expression responses are induced and finely regulated by cellular ROS. For this reason, their levels have to be accurately controlled (as will be discussed in next paragraphs). Most of the endogenous ROS are mainly produced by mitochondrial Complexes I and III (Murphy 2009). The amount of mitochondrial ROS production can vary depending on physio-pathological conditions (Vuda & Kamath 2016). Abundant evidence (Wojcik et al., 2010) shows that under normal conditions mitochondria

− are able to remove O2 through the mitochondrial manganese superoxide dismutase

(MnSOD) to produce H2O2. H2O2 is then detoxified in water by the mitochondrial detoxifying enzyme glutathione peroxidase (GSPx) (which will be discussed in detail in the following section) thus completely scavenging the ROS produced within the organelles. However, an excess of ROS can damage several components of mitochondria and can also influence the correct assembly of the respirasomes, hence affecting their metabolism and potentially

leading to cellular dysfunction. The inhibition of the ETC activities leads to the block of electron transfer along the respiratory chain, thereby reducing membrane potential and cellular ATP levels. The accumulated electrons may then react with oxygen to form superoxide radicals, thus increasing mtROS formation (Imaizumi et al., 2015). ROS may damage DNA (both mitochondrial and nuclear) and produce dozens of chemically distinct lesions that distort the structure and coding properties of the molecule and may cause mutations and chromosomal aberrations. Mutations in either mtDNA or nDNA can result in a deficit of the ETC and give rise to global mitochondrial dysfunction by affecting signalling, transcription or translation, therefore causing a diverse array of often devastating human mitochondrial diseases, impacting any organ in the body at any point during a person’s life (Friedman & Nunnari 2014).

I.c. Structure of mtDNA nucleoids and their replication and transcription

mtDNA is typically present in hundreds to thousands of copies in each single cell. This is an exclusively maternally inherited genome, since mitochondria in mammalian sperm are destroyed by proteolysis in the fertilized oocyte (Giles et al., 1980).

In accordance with the “endosymbiotic theory”, mtDNA organization is similar to that of bacterial nucleoids, consisting in discrete proteins-DNA complexes, and it is maintained from S. cerevisiae to human cells. Mitochondrial nucleoids are attached to the inner mitochondrial membrane (Meeusen & Nunnari, 2003). Super-resolution microscopy analysis found that their diameter is around 100nm and that each one contains between 1 and 2 copies of mtDNA (Kukat et al. 2011). Throughout the mitochondrial network, nucleoids show under the microscope a discrete punctuate distribution. Nucleoid packaging provides an efficient means of ensuring that mitochondrial genetic material is distributed throughout the cell’s mitochondria and is responsive to cellular metabolic needs: this facilitates the organization of transcription and translation in its immediate surroundings.

- 29 -

It has been proposed that mammalian mitochondrial nucleoids have a layered structure, and that transcription takes place in the nucleoid core while RNA processing and translation occur in its outer cortex (Fig. 3). The packaging of mitochondrial genome is mediated by nucleoid associated proteins, the most abundant being the transcription factor A, TFAM (Kukat et al., 2011), an HMG-box protein that constitutes the core of these structures. Figure 3. Spatial organization of nucleoid-associated proteins. The proteins that exhibit mtDNA-binding capacity are located in Some laboratories reported a very high TFAM the core region, whereas others might interact with proteins in the core region or other mitochondrial proteins. ATAD3, ATPase copy number of about 1000 proteins/mtDNA family AAA-domain-containing protein 3; HSP60, heat-shock protein 60; mtSSB, mitochondrial single-stranded DNA-binding genome or higher (Ekstrand et al., 2004, protein; mTERF, mitochondrial transcription termination factor; PHB; prohibitin; POLG, mitochondrial polymerase γ; mtRNAP, Takamatsu et al., 2002). The presence of TFAM mitochondrial RNA polymerase; TFBM, mitochondrial transcription factor B;TFAM mitochondrial transcription factor A. is extremely important for the maintenance of

From Bogenhagen et al., 2008. mtDNA and embryonic lethality with complete loss of mtDNA has been reported in knockout mice for this gene (Larsson et al., 1998). In vitro studies have demonstrated that TFAM alone is sufficient to fully package mtDNA in compact rosette-like structures containing loop extensions that can be observed by atomic force microscopy (Kaufman et al., 2007). Another important constituent of mitochondrial nucleoids is the mitochondrial single-stranded DNA- binding protein (mtSSB) which has a primary sequence and structure similar to bacterial SSB and has been shown to coat the extensive ssDNA regions in mtDNA replication intermediates (Maier et al., 2001), preventing them from renaturing. mtSSB, together with the replicative DNA helicase (TWINKLE), marks a sub-set of actively replicating mtDNA nucleoids.

Outside the animal kingdom, mtDNA varies enormously in size and form, however in mammalian species the mitochondrial genome is a double-stranded closed circular molecule between 16.2 and 16.7 kbp in size. It encodes 13 of the subunits of the ETC: ND1, ND2, ND3, ND4, ND4L, ND5 and ND6 (complex I); CYTB (complex III): COXI, COXII, COXIII (complex IV): and ATP6 and ATP8 (complex V) along with 22 tRNAs and 2rRNAs. It has one major non- coding region, the displacement loop (D-loop) (Fig. 4). The D-loop is formed by the stable incorporation of a third, short DNA strand known as 7S DNA and it comprises two hypervariable regions.

Figure 4. Map of human mtDNA. Light-strand promoter (LSP) transcription produces the ND6 mRNA molecule and primers for initiation of DNA synthesis at OH. The Heavy-strand transcription is initiated from two sites. The H1 site is located 16 bp upstream of the tRNAPhe gene and it produces a transcript, which terminates at the 3’ end of the 16S rRNA gene (TERM). The H2 site is close to the 5’ end of the 12S rRNA gene and it produces a polycistronic molecule, which corresponds to almost the entire H strand. The tRNA genes encoded on each of the two strands are indicated with the standard one-letter symbols for amino acids and their position is indicated with black squares. 2 mt- rRNA genes (fuchsia) and 13 protein-coding genes (olive, ND1-6 encoding members of NADH: ubiquinone oxido-reductase; blue, Cytb encoding apocytochrome b of ubiquinol: cytochrome c oxido-reductase; orange, COI-III encoding members of cytochrome c oxidase: aquamarine, ATPase 6, 8 encoding two members of Fo-F1 ATP synthase).

From Kyriakouli et al., 2008 Why mitochondria still contain part of their genetic material remains an open question. Several hypotheses have arisen. One explanation suggests that some hydrophobic proteins are difficult to import across the mitochondrial membranes. These molecules may therefore need to be produced within the organelle like, for example, cytochrome c oxidase subunit 1 and Cyt b, the most hydrophobic proteins found inside the mitochondrion. A second explanation for mtDNA retention could be the difference in codon usage between the nuclear and mitochondrial genomes, which may make further gene transfer from the mitochondrion difficult. Finally, an alternative explanation suggests that the molecular machines governing mitochondrial gene expression may be directly influenced by components of the respiratory chain and affected by the redox status of the organelles. So the transcription of genes important for metabolic control of the eukaryotic cells could be directly regulated by the mitochondria (Falkenberg et al., 2007).

- 31 -

The mammalian mitochondrial genome possesses two overlapping pairs of protein- coding regions and lacks introns. The noncoding region (NCR) contains the control sequences for transcription and replication. mtDNA transcription and replication are tightly coupled (St. John, 2016). The proteins necessary for mtDNA replication are encoded in the nucleus. The key nuclear-encoded mtDNA replication factors include TFAM, mtSSB, TWINKLE and the mitochondrial-specific polymerase gamma (POLG). The mechanism by which mtDNA replicates is still an unresolved issue, and three models have been proposed so far (Fig. 5):

 The strand-displacement model, proposed in the early 1970s (Robberson et al., 1972)  The RITOLS model, proposed in 2006 (Yasukawa et al., 2006)  The coupled leading- and lagging-strand synthesis (resembling bacterial genome replication) model (Holt et al., 2009)

Figure 5. A- The long-standing strand-displacement model of mammalian mtDNA replication in which single-stranded DNA intermediates are a hallmark. B- in the RITOLS model replication proceeds via long stretches of RNA intermediates laid on the mtDNA lagging-strand. C- Model resembling bacterial genome replication through coupled leading- and lagging-strand synthesis.

From McKinney & Oliveira 2013

The first model arose from the study of cesium chloride-purified mtDNA, The non-coding region (NCR) is the control region where nuclear-encoded transcription and replication visualized by electron microscopy; in this factors drive transcription and then replication of mitochondria. In human mtDNA, it spans model (Fig. 5-A) the synthesis of the approximately 1.1 kb between the mt-tRNA genes of leading strand (known as the heavy [H] phenylalanine and proline. A large part of the NCR often incorporates a linear third DNA strand of strand in vertebrate mtDNA because of its approximately 650 nucleotides, forming a stable D-loop structure. This additional strand is also called 7S DNA, a guanine-rich composition) initiates within name based on its sedimentation properties. It is generally accepted that the synthesis of 7S DNA is the non-coding D-loop region at a site primed using long RNA molecules of around 200 nt, called 7S RNA (Ojala et al., 1981). 7S RNA is synthesised designated as OH (origin of H-strand by transcription from LSP, and remains stably bound to synthesis). According to this mechanism, the template L-strand following synthesis (Xu and Clayton, 1996). the synthesis of the lagging strand (or light [L] strand) starts when two-thirds of the

leading sequence have been synthesized, when the OL (origin of the lagging strand synthesis) is exposed, after the displacement of the parental H strand due to the continuous replication.

More recently, the long-standing strand-displacement model has been challenged by two other models for mtDNA replication.

The first one is the RITOLS (Ribonucleotides Incorporated ThroughOut the Lagging Strand) model (Fig. 5B). Yasukawa and coauthors could not observe any evidence of the predicted extensive single-stranded replication intermediate region (which is a prerogative of the conventional model). By using psoralen/UV cross-linking, they provided compelling evidence for extensive tracts of hybrid RNA/DNA duplexes in replication intermediates. These hybrids are then replaced, or converted to DNA, via a maturation process. However, if RITOLS uses preformed transcripts, this mechanism could only be adopted for DNA sequences that are constitutively transcribed and in the absence of noncoding sequences, such as the mtDNA molecule (Holt & Reyes, 2012).

The second alternative model, referred to as the strand-coupled model (Fig. 5C), was proposed on the basis of the results of 2D-PAGE experiments (Holt et al., 2009). The hallmark of this mechanism of replication is the coupled leading and lagging DNA strand synthesis. In this case, replication proceeds bidirectionally until it finally comes to a halt at the D-loop

- 33 -

region, although Okazaki fragments have not been detected. However, this mode appears to occur only when cells are recovering from ethidium bromide-induced mtDNA depletion.

Regardless of the mode of replication, all these models require extension of an RNA primer and it is accepted that the transcription machinery is required to generate a nascent RNA that is subsequently cleaved, processed, and extended by DNA POLG. The minimal mitochondrial replisome machinery (Fig. 6) consists of POLG (the POLG complex comprises a catalytic subunit, POLGA, and it is supported by an accessory subunit, POLGB, which anchors POLGA to mtDNA during replication), the mtDNA helicase TWINKLE, mtSSB,

topoisomerases, RNase H1 and several transcription initiation and termination factors.

In the past 40 years, there has been an accumulation of controversial observations on whether the synthesis of mtDNA is coordinated with the cell cycle. For example, a higher activity of the mitochondrial DNA polymerase POLG has been reported during the cell cycle’s S and G2 phases (Radsak et al., 1978). It has also been found that the level of 7S DNA does vary according to cell cycle stage, with 7S DNA synthesis peaking in the S-phase (Antes et al., 2010). Finally, the coordination between cell cycle progression, mitochondrial biogenesis and mitochondrial energetics has been shown in relatively recent publications (Schieke et al., 2008; Garedew et al., 2012, Chatre & Ricchetti 2013).

The NCR contains not only the classical origin of heavy-strand replication, OH, but also the HSP (H1 and H2) and LSP promoters for transcription of the heavy- and light-strands, respectively. Both the heavy- and light-strands of mammalian mtDNA are transcribed as long, polycistronic molecules; the transcripts initiated at LSP may either be processed to generate (by cleavage or premature termination) Figure 6. mtDNA replication machinery: the helicase TWINKLE has 5’end to 3’ end directionality and unwinds the duplex DNA primers for the initiation of DNA synthesis template. The mtSSB stabilizes the unwound conformation, also stimulating POLG holoenzyme in its synthesis activity. The (RITOLS model) or arise from one long transcript. nascent DNA synthesized by POLG (green) is shown as a solid red line, while the RNA primer (jagged red line) created by the Of course, the requirement of transcription to mitochondrial RNA polymerase (orange) is degraded by RNase H1 (yellow). produce primers links the process of transcription

to mtDNA replication. Full length transcripts are then cut into functional tRNA, rRNA and mRNA molecules.

The H1 site is located 16 bp upstream of

the tRNAPhe gene and it produces a transcript From Kasiviswanathan et al., 2012

terminating at the 3’ end of the 16S rRNA gene, while the H2 site is close to the 5’ end of the 12S rRNA gene and it produces a polycistronic molecule, which corresponds to almost the entire H strand. In mitochondria, the existence of a single subunit RNA polymerase (RNAP) was first reported in yeast (Greenleaf et al., 1986) and later in human cells (Tiranti et al., 1997). These two proteins (Rpo41 in yeast and POLRMT, also known as h-mtRPOL, in human cells) show high sequence similarity at their C-terminal part to RNA polymerases encoded by

the T-odd lineage of bacteriophages (e.g. T7 and T3) (Masters et al., 1987). However, in contrast to the phage T7 RNA polymerase, the human POLRMT cannot interact with promoter DNA and initiate transcription on its own. The initiation of the transcription process involves 2 other protein families together with the RNAP: TFAM and the mitochondrial transcription factor B paralogues B1 and B2 (TFB1M and TFB2M, also called mtTFB1 and mtTFB2) (Falkenberg et al., 2007). The actual molecular events that are involved in transcription initiation are unknown, but these factors make up the basal transcription machinery and have been shown to function in vitro. Moreover, it is still not clear if there are subsets of specialized nucleoids for replication or transcription or if it is possible to transcribe both strands of mtDNA in the same nucleoid.

In recent years, the understanding of the mtDNA organization in nucleoids has greatly progressed. However, many additional questions remain to be answered. For example, it is still not clear if higher-order structures similar to chromatin in the nucleus are present in these organelles or how TFAM is removed to make mtDNA accessible for replication, transcription or repair. Furthermore, it’s still debated which is the relationship between mitochondrial fission and nucleoids distribution throughout the mitochondrial reticulum. The work of Lewis and colleagues revealed that for mtDNA synthesis and nucleoid distribution, ER tubules proximity is necessary (Lewis et al., 2016). In fact, in human cells the nucleoids actively engaged in mtDNA synthesis were found to be spatially and temporally linked to a small subset of ER-mitochondria contacts destined for mitochondrial division. This finding suggested that ER-mitochondria contacts At the ER-mitochondria contacts sites coordinate the licensing of mtDNA replication with destined for mitochondrial fission, mtDNA replication occurred upstream downstream mitochondrial division events to distribute of mitochondrial constriction and also before the assembly of the division newly replicated mtDNA to daughter mitochondria. machinery.

mtDNA organization has clinical implications and the mechanisms for nucleoid segregation are likely to play an important role in maternal - 35 - mtDNA inheritance (Kukat & Larsson 2013). In fact, ensuring the accurate distribution of nucleoids within cells is mandatory for human cellular homeostasis.

I.d. Impact of mtDNA alterations on medicine

In 1959, the first disorder linked to mitochondrial function was described by Dr. Luft (Ernster et al., 1959). Mitochondrial dysfunctions can arise from mutations in either mtDNA or nDNA. They can result in a deficit of a specific electron transport chain complex or give rise to global dysfunction of the organelle by affecting genes involved in mitochondrial processes such as signalling, DNA replication, DNA maintenance and protein synthesis. The term “mitochondrial disorders” was generally referred to diseases resulting from defects in the mitochondrial electron transport chain but, nowadays, it refers to a group of conditions that affect these organelles.

Point mutations in mitochondrial DNA are usually maternally inherited, while classic Mendelian inheritance (autosomal-dominant, autosomal-recessive or X-linked) is usually indicative of a nuclear mutation (DiMauro et al., 2004). In children, > 90% of mitochondrial disorders are the result of nuclear mutations (Bernier et al., 2002). On the other hand, for mitochondrial DNA mutations the age of onset of the disease is generally more advanced (Hancock, 2002). When searching for clues to suggest a mitochondrial DNA aetiology, three features are consistently associated with these mutations: progressive external ophthalmoplegia, myopathy and pigmentary retinopathy (Lamont et al., 1998; Khan et al., 2015) (Fig. 7).

Figure 7. Drawing showing the clinical features and the organs affected by mitochondrial disorders. From Khan et al., 2015

- 37 -

In 1989, Holt et al. described the first mutation (a deletion) in the mitochondrial genome associated with clinical pathology (Holt et al., 1989). Nowadays, >200 pathogenic point mutations, deletions, insertions and rearrangements of the mitochondrial genome are known (Gillis & Sokol, 2003; Khan et al., 2015; Alston et al., 2017). Polymerase chain reaction (PCR) can be used to screen for many among them. A defect in any of the mitochondrial- encoded genes can produce clinical disease. mtDNA disorders are known to affect ~1 in 4000 of the population (Gorman et al., 2015). Because each human cell contains multiple copies of independently proliferating mtDNA molecules, stochastic mutation events can either affect all of these molecules (homoplasmy) or only a proportion (termed heteroplasmy). The levels of heteroplasmy can vary greatly from cell to cell in the same tissue or organ, between organs within the same person and also between different individuals in the same family (Stewart & Chinnery, 2015). The levels of recombination between different nucleoids are negligible, therefore, we could consider the mtDNA lineages as clonal and thus imply that all mtDNA mutations originate within a single cell, within a single organelle, within a single molecule. A certain level of mtDNA heteroplasmy seems to be almost universally present in healthy humans. It has been shown that the proportion of mutated mtDNA must exceed a critical threshold level (from 50% to 60% for mtDNA deletions and up to >90% for some tRNA mutations) before the induction of respiratory chain dysfunctions. These differences in the levels of heteroplasmy and their specific threshold for “clinical manifestation” are both thought to contribute to the characteristic pattern of organ vulnerability observed for different mitochondrial diseases and to the heterogeneity observed in patients harbouring the same mtDNA mutation, making the treatment of such pathological conditions difficult.

II. Quality controls of mitochondria

Mitochondria are particularly vulnerable to the Eukaryotic cells have accumulation of damage during cell life, not only because of their evolved a wide arsenal of quality control mechanisms complex biogenesis but also because of their manifold functions in to preserve mitochondrial homeostasis and prevent eukaryotic cells. A hierarchical network of Quality Control (QC) cellular damage and eventual death. mechanisms is present in eukaryotic cells. It acts at the molecular and organelle levels and its components are highly interconnected and regulated.

In this section the different mechanisms used to minimize mitochondrial damage will be described. These mechanisms include processes for the maintenance of redox homeostasis and the prevention of oxidative stress (like for example ROS scavenging and Mitochondrial Unfolded Proteins Response), proper mitochondrial fusion and fission regulation, the removal and repair mtDNA alterations; and finally responses that eliminate excessively damaged mitochondria from the cell (i.e., mitochondrial autophagy/mitophagy). The intricate balance between these QC allows the rapid adaptation of mitochondria to stress and damage, and when one of these mechanisms is disrupted, the cell can easily trigger pathways that lead to apoptosis or to the progression of a disease state.

II.a. ROS scavenging

− Superoxide (O2 ) is the primary ROS generated from the respiratory chain and, in

mitochondria, it can be converted to hydrogen peroxide (H2O2) by the nuclear encoded superoxide dismutase mitochondrial isoform, which contains manganese in its active center

(MnSOD). Even if superoxide is not membrane permeable, H2O2 can efficiently pass through membranes and therefore its deleterious effect acts not only at its site of formation but also elsewhere. This compound can be detoxified to water by different enzymes. In mitochondria, once again, nuclear encoded enzymes such glutathione peroxidase (GSPX) or peroxiredoxin do the job. MnSOD and GSPX are small proteins that are key components of the mitochondrial ROS scavenging systems. These systems are regulated at different levels. For example, several regulatory proteins control the transcription of mammalian MnSOD through the interaction with different promoter elements (Li et al., 2010; Chung et al., 2011). The activity of this enzyme is also regulated at the post-translational level; for example, activation of the - 39 -

murine MnSod is promoted by the deacetylation of a conserved lysine residue mediated by the mitochondrial sirtuin Sirt3 (Tao et al., 2010).

In the presence of Fe(II) or Cu(I), H2O2 can also react non-enzymatically to form the extremely reactive hydroxyl radical for which no known scavengers exist (Fischer et al., 2012). It should be emphasized that ROS scavenging pathways, like all other QC mechanisms, are limited in their capacity and are therefore unable to completely prevent molecular damage. However, they are effective in controlling/decelerating the speed by which a critical threshold of damage is reached.

II.b. Mitochondrial Unfolded Proteins Response

By examining worms that were Figure 8. The existence of this broad transcriptional response demonstrates that a means of communication and coordination deficient for a nuclear-encoded cytochrome c exists between the mitochondria and the nucleus. oxidase subunit (cco-1), the impairment of From Sun et al., 2016 electron transport in these animals has been shown to trigger the activation of the mitochondrial unfolded protein response (UPRmt) (Lee et al., 2003; Durieux et al., 2011). The UPRmt is a form mitochondrial-to- nucleus communication, known as retrograde signalling, that contributes to ensure the maintenance of mitochondria quality. Initially characterized in mammalian cells, this stress response pathway starts either from a depletion of the mitochondrial genome or from an accumulation of misfolded proteins within the mitochondria (Zhao et al., 2002) causing proteolytic stress. This tightly controlled response consists in the triggering of nuclear transcription of mitochondrial chaperone proteins and involves crosstalk between mitochondria, endoplasmic reticulum and nucleus (Fig. 8).

In worms, it is now known that, when ROS increase, the UPRmt is regulated by a unique transcription factor termed Activating Transcription Factor associated with Stress-1 (ATFS-1). ATFS-1 has both a nuclear localization targeting sequence and a mitochondrial one (Nargund et al., 2012). Normally, ATFS-1 is efficiently imported into mitochondria, where it

is rapidly degraded by the Lon protease (Nargund et al., 2012). However its mitochondrial import is reduced during stress, resulting in a nuclear accumulation and therefore in a transcriptional response. In C. elegans, this transcription involves expression of nuclear genes such as the heat shock proteins (hsp) hsp-6 and hsp-60, which encode for chaperone proteins.

So far, in mammal cells, few players in the UPRmt signalling have been identified and the mammalian equivalent of ATFS-1 remains elusive. The accumulation of unfolded proteins in the mitochondrial matrix leads to activation of JNK2, which triggers c-Jun binding to AP-1 elements to up-regulate CHOP, the main transcription factor implicated in the mammalian UPRmt. CHOP heterodimerizes with C/EBPβ to activate the promoters of the UPRmt responsive genes (mitochondrial unfolded protein response elements-MURE). Among them we can find the homologues of the worm chaperone hsp-60, the HSP-60 (Aldridge et al., 2007).

ROS stress results in organelle perturbation by inducing both mutations and deletions in mtDNA which in turn can give rise to mutated proteins. On the other hand, ROS could also affect the protein folding, impeding the formation of stoichiometric complexes for OXPHOS in the organelles. Stoichiometric imbalance between mitochondrial and nuclear proteins seems to activate the UPRmt (Houtkooper et al., 2013). A recent study (Youle et al., 2010) demonstrated that the activation of the UPRmt induces the stabilization of PINK1 and PINK1/Parkin-dependent mitophagy: the authors confirmed the accumulation of PINK1 and Parkin in mitochondria without an apparent MMP loss. Mitophagy (discussed in the next section) appeared to be activated in an attempt to eliminate organelles with an overloaded matrix.

Mitochondrial-to-nuclear signalling and vice versa allows the restoration of mitochondrial function while, at the same time, re-wiring the cell to temporarily survive as best as possible without the benefit of full mitochondrial capacity.

- 41 -

II.c. The maintenance of a healthy mitochondrial population: the fission/fusion mechanisms

Mitochondria are dynamic organelles and their shape is determined by mechanisms of fusion and division, controlled by a growing set of “mitochondria-shaping” proteins, which influence crucial signalling cascades, including apoptosis. Due to the presence of two membranes in these organelles, fusion and fission events are extremely complex. Generally, we can say that the maintenance of a healthy mitochondrial population is dependent on balanced organelle fission and fusion that promote the remixing of mitochondrial material.

Already in the 90s, the molecular analysis of mitochondrial morphology began with the discovery of the D. melanogaster fusion factor fuzzy onions (FZO), which participates in the formation of the Nebenkern Structure. The Nebenkern structure originates from the fusion of mitochondria and is required for sperm mobility (Heles and Fuller 1997). FZO was the first identified component of the mitochondrial fusion machinery and it is considered as the founding member of a highly conserved family of dynamin-related proteins (DRPs) that, through their ability to self-assemble and hydrolyse GTP (guanosine triphosphatases- GTPases), mediate fission and fusion events by the remodelling of mitochondrial membranes. Subsequently, yeast genetic screens identified the orthologue, Fzo1, which has a conserved role in mitochondrial fusion (Hermann et al., 1998) and additional roles as a modulator of mitochondrial fusion and fission processes. The

Figure 9. OPA1 precursor molecules are matured and converted into L- The importance of fusion–fission in human health is OPA1 upon import into mitochondria. A fraction of L-OPA1 molecules is highlighted by a number of diseases caused by subsequently cleaved by OMA1. Under stress conditions (e.g., low ΔΨm, mutations in proteins involved in these processes. For low matrix ATP), OMA1 degrades L-OPA1 molecules completely into S- example, it has been well established how mutations in OPA1. OPA1 lead to optic atrophy, often in combination with Adapted from Rugarli & Langer, 2012 hearing loss and ophthalmoplegia. Loss of MFN2 leads to Charcot-Marie-Tooth type 2A, a disease GTPases of the dynamin family that characterized by neuropathy and associated muscle weakness as well as in some cases hearing loss, optic mediate these mechanisms are well atrophy, and pyramidal signs. It is thus clear that fusion–fission plays a critical role in maintaining a conserved between yeast, flies and healthy pool of mitochondria. mammals and with their combined actions

divide and fuse the two lipid bilayers that surround mitochondria.

In mammalian cells, fusion between mitochondrial outer membranes is mediated by membrane-anchored dynamin family members named mitofusins, MFN1 and MFN2, homologues of the yeast Fzo1, whereas the fusion of the inner mitochondrial membranes is mediated by a single dynamin family member called OPA1. MFN1 and MFN2 proteins form homo-oligomeric and hetero-oligomeric complexes during the fusion process, implying that they form complexes in trans between adjacent mitochondria (Detmer & Chan 2007, Koshiba T., et al., 2004). During the fusion process, the MMP is particularly important; its dissipation caused by the use of the ionophores (substances that permeabilize the inner mitochondrial membrane) generates an inhibition of the fusion mechanism and an increase in the fragmentation of the mitochondrial network (Legros et al., 2002). However, up until now, the mechanistic link between MMP and fusion mechanism has remained unknown.

Post-translational processing of OPA1 seems to be dependent on membrane potential (Fig. 9). Proteolytic OPA1 protein ensures cristae morphogenesis and the processing converts long OPA-1 (L-OPA1) into short isoforms maintenance of mtDNA, thus protecting cells against apoptosis (S-OPA1) as mitochondrial fusion requires the presence of (Landes et al., 2010). both OPA forms (Song et al, 2009), this process represents a central regulatory step determining mitochondrial morphology. In addition to that, the lipid

Figure 10. In yeast, mitochondrial fission consists in OM fission followed by IM division and is mediated by the dynamin- related protein Dnm1. During this process, cytoplasmic Dnm1 oligomerizes into a ring structure that constricts the organelle on the OM. Fis1 mediates the localization of Dnm1 on the mitochondrial outer membrane together with the adaptor proteins Mdv1 and Caf4. Fis1 is an integral outer membrane protein that interacts with the N-ter of Mdv1 and Caf4. Both Mdv1 and Caf4 bind Dnm1 at their C-ter. Fis1 and Dnm1 have mammalian homologues (FIS1 and DRP1, respectively), but no Mdv1 or Caf4 homologues have been identified so far.

From Detmer & Chan 2007 components of mitochondrial membranes are important factors for fusion processes; in - 43 - yeast, for example, members of the ergosterol synthesis pathway have been shown to be required for normal mitochondrial morphology (Altmann et al., 2005). Interestingly, ergosterol has been associated with yeast vacuole fusion, and phosphatidic acid is thought to play a part in generating the membrane curvature that is required for SNARE-mediated fusion (Choi et al., 2006; Vitale et al., 2001). Thus, specific lipids might have similar roles in distinct types of membrane fusion. In the last years, mitochondrial phospholipase D, an OMM enzyme that hydrolyses cardiolipin to generate phosphatidic acid, has also been identified as a protein important for mitochondrial fusion.

Mitochondrial division (Fig. 10) requires the recruitment of a dynamin-related protein (Dnm1 in yeast and DRP1 in mammals) from the cytosol. Dnm1 and DRP1 assemble into punctate spots on the mitochondrial reticulum, forming helical structures that wrap around the outer surface of mitochondria at constriction sites and lead to productive fission events (Smirnova et al., 2003; Shawn & Nunnari 2001). Consistently with this model, in vitro purified Dnm1 can indeed form helical rings and spirals whose dimensions are similar to those of constricted mitochondria (Ingerman et al., 2005). In yeast, recruitment and assembly of Dnm1 structures, required for fission activity, are mediated by three other proteins. One of these is Fis1, a mitochondrial integral outer membrane protein essential for fission (Tieu & Nunnari 2000; Mozdy et al., 2000). Fis1 binds indirectly to Dnm1 through one of two molecular adaptors, Mdv1 or Caf4 (Griffin et al., 2005). The mammalian homologue of Fis1, FIS1, is also essential for mitochondrial fission (Lee et al., 2004), but no homologues for Mdv1 or Caf4 are known so far.

Why do mitochondria continually fuse and divide? In recent years, several studies have shown that these processes have important consequences on the organelles’ morphology, function and distribution.

A major consequence of the fusion and fission process is that it allows mitochondria to exchange their content, including mtDNA and lipid membranes. These exchanges are crucial for maintaining the health of a mitochondrial population. These organelles should not be considered as autonomous compartments, instead, the hundreds of mitochondria in a typical cell exist as a population (Busch et al., 2014). When mitochondrial fusion is abolished, a large fraction of the mitochondrial population loses nucleoids (Chen et al., 2007). Fibroblasts lacking both MFN1 and MFN2 have reduced respiratory capacity, and individual

mitochondria show great heterogeneity in shape and membrane potential (Chen & Chan, 2005). Cells that lack OPA1 show similar defects, with an even greater reduction in respiratory capacity (Detmer & Chan 2007).

II.c.1. Mitochondrial inner membrane topology

While fusion and fission control the length, shape and number of mitochondria, the internal structures of these organelles are also extremely dynamic. Thanks to three- dimensional tomography of cryopreserved samples new views of inner membrane organization and plasticity have been provided (Mannella, 2006). The inner membrane can be divided into distinct structural and functional regions:

 The inner boundary membrane, comprising the regions in which the inner membrane is in close proximity to the outer one. These regions might be the sites for coupled outer and inner membrane fusion. This domain is enriched in proteins involved in protein translocation processes through the IMM, such as the TIM23 complex;  The cristae membrane, enriched in proteins that are involved in oxidative phosphorylation like Cyt c;  The cristae junctions, which are narrow ‘neck’ regions that separate the inner boundary membrane from the involuted cristae membrane; their regulation might be important for the relocalization of Cyt c during apoptosis (Scorrano et al., 2002).

There are several observations indicating how the inner membrane morphology is intimately related to bioenergetics (Mannella, 2006). For instance, purified mitochondria placed in low ADP conditions have a limited respiration and show an “orthodox” morphology, characterized by narrow cristae and few cristae junctions. Higher ADP concentrations induce a ‘condensed’ morphology, characterized by larger cristae and several cristae junctions per cristae compartment.

The shape of mitochondria affects also their distribution to specific subcellular locations. This function is especially important in highly polarized cells, such as neurons.

- 45 -

Mitochondrial fission also facilitates apoptosis by regulating the release of intermembrane- space proteins into the cytosolic compartment. As result of these cellular functions, mitochondrial dynamics have consequences in development, disease and apoptosis mechanisms (Detmer & Chan 2007).

II.d. mtDNA maintenance: how to repair or tolerate DNA damage in mammalian mitochondria In several species, mitochondria seem to possess different mechanisms for DNA maintenance, including base excision repair pathway (BER – which will be discussed in detail in the next chapter), mismatch repair (MMR), homologous recombination (HR), non- homologous end-joining (NHEJ), translesion synthesis and mtDNA degradation. However, so far, no direct evidence of the existence of a nucleotide excision repair pathway (NER) has been found. Nevertheless, some proteins considered as NER factors localize to the mitochondria and may cooperate with BER mechanisms.

Even though for many years it was believed that DNA repair mechanisms might not be required in mitochondria due to their heteroplasmic nature, the presence of many DNA repair proteins inside the organelle point to a different scenario. The mitochondrial genome is exposed to the same chemical assaults as nuclear DNA, leading to the formation of abasic sites, adducts, single and double strand breaks and alkylating and oxidative damage (Fig. 11).

Figure 11. Examples of lesions found in the mitochondria: Both spontaneous reactions that originate within the cell and exogenous agents from the environment may induce damage to DNA. Endogenous damaging agents includeFigure mismatches 11 Examples generated of lesions during foundDNA replication, in the mitochondria deamination of: bases,Both depurination spontaneous or depyrimidinat reactions thation andoriginate oxidative within damage the thatcell arisesand exogenous from the generation agents from of ROS the within environment the cell through may induce normal damage metabolism. to DNA.Exogenous Endogenous factors such damaging as ionizing agents radiation include cause toxic mismatches double-strand generated DNA breaks during (DSBs), DNA ultraviolet replication, (UV) deaminationradiation results of in bases,the formation depurination of cyclobutane or depyrimidination pyrimidine dimers andand alkylatingoxidative agents damage such thatas cisplatin arises lead from to theunwanted generation alkylation of and ROS DNA within crosslinks the cell through normal metabolism. Exogenous factors such as ionizing radiation cause toxic double-strand DNA breaks (DSBs), ultravioletFrom Saki (UV) & Prakash radiation 2017 results in the formation of cyclobutane pyrimidine dimers and alkylating agents such as cisplatin leadIn summary, to unwanted mammalian alkylation andmitochondria DNA crosslinks are capable of repairing a wide range of DNA lesions. The current knowledge of mtDNA repair systems together with other specific mitochondrial mechanisms for the maintenance of the organelle’s genome will be discussed below (Fig. 12).

- 47 -

Figure 12. DNA repair pathways in the nucleus and in the mitochondrion. All established pathways in mitochondria are shown with a squared shape, whereas documented pathways which need further confirmation are enclosed in an oval. It is considered that mitochondria have no NER pathway. The figure displays the DNA repair pathways which can be found in the organelles, but the presence or not of a given mechanism may depend on the organism considered.

Abbreviations: BER, base excision repair; MMR, mismatch; NHEJ, non-homologous end joining; HR, homologous recombination; TCR, transcription-coupled repair; GGR, global genome repair; TLS, translesion synthesis.

II.d.1. Mismatch repair pathway (MMR) The MMR pathway is a conserved repair mechanism that has been shown to be present in mammalian mitochondrial lysates (Mason et al., 2003). The specificity of MMR in the nuclear compartment is primarily for base-base mismatches and small insertions/deletions generated during DNA replication and recombination. Thereby, MMR contributes to the prevention of mutations in dividing cells. However, very surprisingly, in mitochondria it shows no bias in favour of the matrix strand and is thus prone to introduce mutations (Mason & Lightowlers 2003). So, it has been suggested that MMR activity is involved in the repair of small loops rather than mismatches. The eukaryotic MSH and MLH are the homologues of the E. coli mismatch repair proteins MutS and MutL respectively. MSH (Msh1 in S. cerevisiae, MSH2/6 in human cells) and MLH (MLH1 and 3 in mammalian cells, like human or mouse ones) are key components of this pathway in the nucleus. In the nuclear compartment MMR starts with the recognition of DNA mispairs by the MSH proteins followed by recruitment of the MLHs and activation of downstream molecules to repair the mismatch. MLH1 has been reported to be present in mammalian mitochondria, but further evidences are necessary to fully understand its role in these organelles (Martin et al., 2010). On the contrary, no evidence of mitochondrial presence of MSH2 has been found (Martin et al., 2010). In mitochondria, however, MMR appears to occur independently of the MSH and MLH proteins and relies instead on the Y-box binding protein (YB-1) (Lyabin et al., 2014). The YB-1 factor has previously been reported to play a role in nuclear BER and in the repair of cross-linked DNA. De Souza Pinto and colleagues demonstrated for the first time that this protein localizes to mitochondria of human cells and proposed that once bound to the DNA, YB-1 creates a local distorsion of the helix that could then function as signal to assemble a “mismatch repairosome” at the mismatch site through specific protein-protein interactions, which could or not involve some of the known BER proteins, to catalyse downstream steps in MMR (de Souza Pinto et al., 2009).

- 49 -

II.d.2. Homologous recombination and non-homologous end-joining

MtDNA molecules undergo It has been reported that mitochondria in plants and yeast predominantly through intra- molecular recombination, but possess the ability to repair double-strand breaks (DSBs) by either inter-molecular recombination is also observed after DSB homologous recombination (HR) or non-homologous end joining induction. (NHEJ) (Stein et al., 2015; Fritsch et al., 2014; Manchekar et al., 2006). Bacman et al., 2009 However, the presence of DSBs repair mechanisms in mammalian mitochondria are still under debate, even though there are some evidences for HR occurring in mitochondrial extracts (Kajander et al., 2001). For example the overexpression of mitochondria-targeted E. coli RecA in human cells enhances mtDNA repair after DSBs induction by bleomycin treatment (Paul et al., 2001), suggesting that an HR-like mechanism may occur in these cells. Moreover, Rad51, the central actor of HR in the nucleus of eukaryotic cells, also localizes to mitochondria in human cells and could potentially be involved in the repair of DSB in the organelle (Sage et al., 2010). It is also possible that HR in the mammalian mitochondrion occurs at low frequency, primarily under conditions where mtDNA is first depleted by treatment with either ethidium bromide or restriction endonucleases (Bacman et al., 2009). Some evidence suggests that NHEJ, relatively more error-prone when compared with the HR pathway, also takes place in mammalian mitochondria (Fukui & Moraes 2009). NHEJ proteins such as the MRX (Mre11p, Rad50p, Xrs2p) and Ku70/80 (Ku70p, Ku80p) complexes were observed in the organelle at the site of induced DSBs (Kalifa et al., 2012).

II.d.3. Translesion synthesis Lesions that potentially block replication, because they cannot be repaired or can escape repair, are tolerated thanks to mechanisms such as translesion synthesis (TLS) (Friedberg, 2005). This mechanism involves a group of specialized DNA polymerases that can bypass DNA lesions by being able to copy defective templates (Prakash et al., 2005). In eukaryotic nuclei, five polymerases perform TLS, including Rev1, polymerases kappa, eta, and iota (Polκ, Polη, and Polι), which belong to the Y-family of polymerases, and polymerase zeta (Polζ), which is a B-family polymerase. Concerning mitochondrial DNA POLγ, there is contradictory evidence on its capacity for mutagenic bypass of DNA lesions. While some studies have shown an arrest of the enzyme in presence of 8-oxoG in the template strand (Stojkovič et al., 2016), others reported the introduction of a dA opposite an AP site or an 8-

oxodG (Sale, 2013; Kasiviswanathan et al., 2013) or the insertion of any of the four deoxynucleotides opposite bulky DNA adducts derived from benzo[a]pyrene diol epoxide. So, the possibility of translesion synthesis in mitochondria is an open topic.

Polγ is not the only DNA polymerase in mammalian mitochondria. In fact, a particular primase, PrimPol, has been recently described in mammalian cells (García-Gómez et al., 2013). A certain amount of PrimPol is also present in the nucleus, but a larger fraction is located inside mitochondria. Human PrimPol has functional relationship to archeal PrimPol enzymes and it possesses DNA and RNA primase and DNA-dependent DNA and RNA polymerase activities. PrimPol is a unique enzyme capable of de novo DNA synthesis solely with dNTPs. Moreover, PrimPol has the capacity to tolerate lesions such as 8-oxoG and AP sites in template DNA therefore acting as a TLS DNA polymerase; for this reason, it is able to assist Polγ in “reading” template lesions (such 8-oxoG) or bypassing normally blocking lesions (like AP sites) by realigning the stalled primer terminus to an alternative template position ahead of the lesion. PRIMPOL gene silencing in human cultured cells has several deleterious effects on mtDNA. In fact, PRIMPOL small interfering RNA (siRNA) in human cells caused the decrease of mtDNA levels to half those of controls and prevented the recovery of mtDNA copy number after transient drug-induced mtDNA depletion (García-Gómez et al., 2013). Both results suggest that mtDNA replication is impaired when there is an abrupt loss of PrimPol. In agreement with these findings, mouse cells lacking PrimPol display inefficient replication of mitochondrial genome. However, given that a mouse lacking the gene is viable, PrimPol is not essential for mtDNA maintenance. Compatible with PrimPol being a nonessential gene, mitochondrial RNA polymerase (POLRMT), and not PrimPol, primes at one major initiation site on each strand, OriH and OriL, while primer elongation is performed by Polγ (Falkenberg et al., 2007). The leading-strand DNA synthesis is not always continuous; genotoxic damage has been linked to discontinuities in this strand synthesis that entail skipping the obstacle, resuming fork progression and leaving a gap behind (Langston & O’Donnell, 2006; Yeeles & Marians, 2011). In this regard, PrimPol might contribute to multipriming events at stalled replication forks given that replicative stress is constitutive during mtDNA replication, most likely due to a highly oxidative environment. The presence of PrimPol in human mitochondria suggests that it represents an ancestral cellular response to genotoxic insults contributing to tolerance of DNA damage inside mitochondria.

- 51 -

Polζ was also detected in the mitochondria of mammalian cells and elevated expression of the enzyme in breast tumors was associated with cell migration and invasion. The localization of the enzyme in mitochondria increases survival together with the invasive potential of cancer cells (Singh et al., 2015).

Finally, it has recenty been shown that in mouse mitochondria purified from brain Polβ was inside the mitochondrial membrane, however, liver samples, show no Polβ in the mitochondrial compartment (Sykora et al., 2017). To date, research has been unable to definitively answer whether this enzyme is in human mitochondria.

II.d.4. mtDNA degradation As it has been described, the mitochondrial genome is redundant, consisting of hundreds to thousands of copies per cell. Therefore, unlike the nuclear genome, a “repair or die” pressure is not imposed on mtDNA, and a substantial fraction of damaged molecules can be lost without detrimental effects for the cell, although the mechanisms by which this happens are not yet known. This loss can be compensated by replication of the remaining genomes.

In vitro, many cell lines are fairly tolerant to the loss of this molecule and can survive both a gradual loss of mtDNA through chronic treatment with ethidium bromide (King and Attardi 1989) and/or an acute destruction of a fraction (Alexeyev et al. 2008) or even all of their mtDNA (Kukat et al. 2008) by mitochondria-targeted restriction endonucleases.

Oxidative stress Induced by different substances (ethanol administration in animal models such mice or H2O2 addition to hamster fibroblasts in culture) causes a reversible mtDNA loss, underlining the strong link between oxidative stress (which may result in oxidative mtDNA damage) and mtDNA degradation. This evidence supports the hypothesis that the turnover of the mitochondrial genome is one of the protective mechanisms used by mitochondria to deal with excessive damage; however, the effects and consequences of persistent mtDNA damage remain poorly understood.

II.e. Mitophagy programs

As is clear from previous paragraphs, damage to the mitochondrial genome may affect and abolish the main source of cellular energy generation. However, because human cells possess hundreds of mitochondria, the general opinion was for a long time that the removal of damaged organelles through mitophagy was enough for the constant elimination of mutant mtDNA (Pellegrino & Haynes, 2015). Since damaged mitochondria generate a plethora of stress signals that could lead to cell dysfunction, this process is essential for cellular homeostasis maintenance.

Recognition and targeting of defective or superfluous organelles to the autophagosome involve distinct steps. While mitochondrial depolarization/fragmentation are two common prerequisites for mitophagy, mitochondrial fusion can block this process. The intimate link between the organelles’ shape and degradation suggests that those mechanisms are coupled at the molecular level (as will be described in the following section). The segregation of dysfunctional mitochondria from the fusing population results in the generation of small, depolarized organelles easier to eliminate. In mammals, several mitophagy mediators are involved: for example NIX participates in the developmental removal of mitochondria during erythropoiesis (Mortensen et al., 2010), and parkin and PINK1, whose genes are mutated in autosomal recessive forms of Parkinson’s disease, can regulate the selective elimination of damaged mitochondria in many cellular types (Valente et al., 2004; Kitada et al., 1998).

Figure 13. mtDNA damages which result in defective electron transport chain (ETC) protein complexes trigger the translocation of PINK1 and Parkin in proximity of the damaged protein complexes. Mitochondrial-derived vesicles (MDVs) carry selective mitochondrial cargo into peroxisomes and multivesicular bodies for lysosomal degradation. Cargo destined for lysosomes are generated by PINK and Parkin.

Adapted from Bingol & Sheng, 2016

- 53 -

PINK1/parkin proteins are the main regulators of this process in mammalian cells. In normal mitochondria, the multifunctional voltage-sensitive PINK1 kinase (putative kinase protein 1) that regulates this elimination process is imported into the mitochondrial matrix thanks to the presence of an MTS in its N-terminal part. Once inside the organelle, it is degraded by proteases (Voigt et al., 2016). However, under conditions of mitochondrial impairment or loss of mitochondrial potential, PINK1 cannot cross the inner membrane and is instead integrated and stabilized in the outer membrane of damaged mitochondria (Matsuda et al., 2010) where it recruits the E3 ubiquitin ligase Parkin. Parkin ubiquitinates proteins on the OMM, including the fusion proteins Mfn1/2 and the VDAC1 (Jin & Youle., 2013). The accumulation of ubiquitin-modifications is thought to facilitate the recruitment of the autophagy adaptor p62 (Narendra et al., 2008; Geisler et al., 2010). In this process double-membraned autophagosomes sequester and ultimately degrade mitochondria via lysosomal fusion (Youle & Narendra 2011). Because the materials degraded in this process can provide macromolecules for the synthesis of higher‐order structures (nucleic acids, proteins or organelles) in anabolic reactions, mitophagy can sustain cell metabolism, homeostasis, and survival (Gatica et al., 2015).

II.f. The final response: apoptosis

If all other protective mechanisms fail, stressed cells trigger the programmed cell death pathway (PCD or apoptosis). Apoptosis is a specific and evolutionarily conserved mechanism that if deregulated leads to cancer and immune diseases. There is a wide variety of signals and conditions, both physiological and pathological, that can trigger apoptosis and not all cells will necessarily die in response to the same stimuli.

There are two major pathways of apoptosis, the extrinsic one and the intrinsic one (Fig. 14), that respond to different signals in vertebrates. For both of them, the activation of a group of cysteine proteases, named caspases, is at the basis of a molecular cascade of events that leads to drastic morphological changes, including chromatin condensation, nuclear rupture and cell shrinkage (Hacker, 2000). Among all these morphological features, pyknotic nuclei, as a result of chromatin condensation, are the most characteristic trait of apoptosis.

Caspases can also alter the The extrinsic signaling pathways initiate apoptosis by integrity of the plasma membrane and, in the activation of transmembrane receptors. These death receptors are members of the tumor necrosis fact, another biochemical feature of factor (TNF) receptor gene superfamily (Locksley et al., 2001) that possess cysteine-rich extracellular domains apoptosis is the expression of cell surface and a cytoplasmic domain of about 80 amino acids called the “death domain” (Ashkenazi and Dixit, 1998). apoptotic markers (like normally inward- This cytosolic domain plays a critical role in transmitting facing phosphatidylserine) that allow the the death signal from the cellular surface to the intracellular signaling pathways. The sequence of recognition of apoptotic cells by the events that define the extrinsic phase of apoptosis go from ligand binding to the recruitment of cytplasmic adjacent ones, triggering quick adapter proteins. For example, the binding of Fas ligand to Fas receptor results in the binding of the adapter phagocytosis with minimal compromise protein FADD (Fig. 14) which then associates with to the surrounding tissue (Bratton et al., procaspase-8 via dimerization of the death effector domain. At this point, caspase-8 is activated and the 1997). During this process, however, execution phase of apoptosis is triggered. other proteins are also exposed on the cell surface, like the well known Annexin V. This is a phosphatidylserine-binding protein that interacts strongly and specifically with phosphatidylserine residues and can be used for the detection of apoptosis (Van Engeland et al., 1998; Arur et al., 2003).

Figure 14. Main characters of intrinsic and extrinsic apoptotic pathways

From abcam website Caspases are widely expressed and cleave many cellular substrates to dismantle cell contents, and for this reason these proteins exist as inactive proenzymes (McIlwain et al., 2013). The caspase family could be divided into initiator and executioner caspases. The initiator caspases-8 and -9 normally exist as inactive procaspase monomers that are

- 55 -

activated by dimerization. The dimerization facilitates autocatalytic cleavage of caspase monomers into one large and one small subunit, which results in stabilization of the dimer. Caspases-3, -6, and -7 are executioner caspases; their inappropriate activation is prevented by their production as inactive procaspase dimers that must be cleaved by initiator caspases. The cleavage allows conformational changes that bring the two active sites of the executioner caspase dimer together and create a functional mature protease (Riedl & Shi, 2004). Once activated, a single executioner caspase can cleave and activate other executioner caspases, leading to a domino effect able to amplify the initial event enormously. These proteases recognize and cleave proteins mostly at aspartic acid residues, although different caspases have different specificities. Once caspases are activated, there seems to be an irreversible commitment towards cell death.

For the intrinsic pathway, the involvement of mitochondria is essential and for this reason it is also known as mitochondrial apoptotic pathway; in fact, these organelles are not only the site where antiapoptotic and proapoptotic proteins interact, but also the origin of signals that initiate the activation of caspases.

The first clue suggesting that apoptosis might be regulated at the level of these organelles came when Bcl-2 was reported to be localized in the mitochondrial membrane (Hockenbery et al., 1990). Bcl-2 primary antiapoptotic function is to prevent the release of cytochrome c from the mitochondria. Bcl-2 is part of a larger family of evolutionarily conserved proteins that share one or more of the Bcl-2 homology (BH) domains that can either promote or inhibit apoptotic process (Luna-Vargas & Chipuk 2016). The stimuli that initiate the intrinsic pathway produce intracellular signals that activate direct proapoptotic

The Bcl-2 family (Garcia-Saez 2012) can be divided into: members of the family, such as Bax, which shuttles prosurvival (antiapoptotic) members. This from the cytosol to the mitochondrion to trigger the group includes Bcl-2 and closely related proteins like Bcl-XL, Bcl-W and Mcl-1, which formation of pores on the OMM. This leads to inhibit cell death by interacting physically with proapoptotic members and antagonizing their profound changes in the inner and outer function mitochondrial membranes structures which result in executioners Bax an Bak, which are believed to participate directly in MOM permeabilization an opening of the mitochondrial permeability the BH3-only proteins further classified as “sensitizers” (in the case of Noxa, Bfm or Bik) transition (MPT) pore (mitochondrial outer that antagonize the function of the antiapoptotic Bcl-2 family members, or “direct membrane permeabilization (MOMP)). The exact activators” like Bim and Bid that activate the mechanism that leads to the MOMP is still unclear, proapoptotic factors Bax or Bak however it has been proposed that once Bax

translocates in the mitochondrion together with Bak, which is constitutively bound to the OMM, it changes its own conformation, fits into the membrane, oligomerizes and induces the release of cytochrome c (Garcia-Saez, 2012).

Together with the Cyt C, the normally sequestered pro-apoptotic Smac/DIABLO and the serine protease HtrA2/Omi, involved in various mitochondrial functions essential for normal cell growth, are also simultaneously released (Wang & Youle 2009), Smac/DIABLO and HtrA2/Omi are reported to promote the PCD, once in the cytosol, by inhibiting anti- apoptotic proteins (Srinivasula et al., 2001; van Loo et al., 2002a; Schimmer, 2004). In the cytoplasm, Cyt C binds and activates APAF-1 (apoptotic protease activating factor 1), an adaptor protein. This event activates the formation of a heptameric caspase-9 complex, known as the apoptosome, that cleaves the executioner caspases-3, -6 and -7. After the cleavage, the caspases translocate to the nucleus as transcription factors and activate cell death genes.

Concluding remarks on mitochondrial dynamics, mitophagy and apoptosis

Both natural fission and extensive cellular stress can lead to mitochondrial depolarization, release of pro-apoptotic molecules, activation of caspases and, finally, apoptotic cell death. Under extensive damage, different mechanisms can induce mitochondrial outer membrane permeabilization (MOMP). MOMP constitutes one of the hallmarks of cell death, both apoptotic and necrotic (Kroemer et al., 2007). However, if MOMP is limited to a subset of the mitochondrial pool, it will result in the selective elimination of the depolarized organelles by mitophagy, suggesting that efficient recognition of dysfunctional mitochondria could set a higher threshold for MMP to rescue the cell and prevent cell death.

Mitochondrial dynamic processes and mitophagy mechanisms are highly integrated and it has been demonstrated that enhancing mitochondrial fusion or inhibiting fission result in decreasing mitophagy. Fission events were demonstrated to have a role in segregating dysfunctional mitochondria from the mitochondrial web (Gomes et al., 2008; Twig et al., 2008-a) and in sorting out mutant mtDNA copies (Malena et al., 2009). Fission events are associated with large changes in membrane potential. Interestingly, electron microscope tomography shows that fission events can generate asymmetric daughter mitochondria - 57 -

(Mannella et al., 2006) with an unequal nucleoid distribution and with a certain level of depolarization (Twig et al., 2008). These mitochondria can later be targeted by the mitophagy machinery. During fission elongated mitochondria are divided into pieces of manageable size for encapsulation and selective removal by mitophagy. However, reduction of organelle size independently of its energetic status does not trigger the elimination process, so the organelle size alone could be only one of the determining factors for the elimination (Twig et al., 2008), as the loss of Drp1 activity also decreases mitophagy under basal conditions. Fusion events are suggested as a complementary mechanism by which organelles quickly equilibrate mitochondrial matrix metabolites and membrane components (Twig et al., 2006; Wikstrom et al., 2009; Jakobs et al., 2006) together with intact mtDNA copies (Schon et al., 2010). The exact mechanism by which mitochondrial fusion prevents mitophagy is still unclear. Parkin has been suggested to regulate mitochondrial fusion in addition to its well-established function as a mitophagy adaptor (Lutz et al., 2009). Following fission, Parkin prevents or delays fusion of depolarized mitochondria by the elimination of mitofusins and in parallel triggers the accumulation of proteins that facilitate mitophagy. Knockout of Parkin in mice leads to accumulation of Mfn1 and to mitochondrial elongation (Poole et al., 2010; Yange et al., 2008). These results suggest that in the absence of Parkin, damaged mitochondria are “rescued” by fusion instead of being removed by mitophagy and compromising mitochondrial quality maintenance. In fact, mitochondrial fusion is a MMP- dependent process ensures that potentially dysfunctional organelles will avoid fusion, whereas intact mitochondria will benefit from sharing metabolites. In support of this view, knockdown of Mfn1 in HeLa cells was protective from the knockdown of Parkin (Poole et al., 2010; Yange et al., 2008). Also PINK1 promotes Parkin translocation to the organelle by acting on the mitofusine, in particular through the phosphorylation of Mfn2 (Chen & Dorn, 2013). In mouse cardiac myocytes, the removal of Mfn2 prevents depolarization‐induced recruitment of Parkin to damaged mitochondria, increasing cardiomyopathies in the animals (Chen & Dorn, 2013). These findings indicate that the proteins that are generally classified as “profusion proteins” have additional roles that affect mitophagy.

High cellular stress leads both to the excessive fission of mitochondria and to the induction of cell death response. Indeed, proteins involved in the regulation of mitochondrial morphology in human cells seem to play a crucial role in the intrinsic apoptosis pathway

(Hoppins et al., 2007). Some of the evidence that point to the intriguing and close connection between these two mechanisms include:

 The co-localization of Bax with Drp1 and Mfn2 mitochondrial fragmentation promoters at fission sites (Karbowski et al., 2002).  The delay of Cyt c release by knocking-down Drp1 (Suen et al., 2008; Parone et al., 2006).  In a variety of cell lines, simultaneous knock out of Bax and Bak results in mitochondrial fragmentation and decreased fusion rates. Little is known about how Bax and Bak mediate their effects on mitochondrial morphology, but Bax influences Mfn2 distribution on the mitochondrial outer membrane and Bak associates with Mfn1 and Mfn2 (Deliviani et al., 2006).  CytC is located into the cristae compartments; its mobilization and mitochondrial remodelling seem to be regulated by the levels of functional OPA1, independently of mitochondrial fusion (Frezza et al., 2006). Overexpression of OPA1 results in a block of Cyt C release and in the subsequent induction of apoptosis.  Drp1 has also been proposed to play a part in cristae remodelling during apoptosis (Germain et al., 2005).

So, it is clear that several structural changes occur in mitochondria during the early phase of PCD. The MOMP that is at the basis of Cyt c and SMAC/Diablo release seems to be facilitated by cristae remodelling. Those structural changes — fragmentation, MOMP and the opening of cristae junctions — occur at similar times, but their exact temporal sequence and causative links are controversial (Frank et al., 2001).

Finally, several findings indicate that the induction of mitophagy and apoptosis are partially regulated by the same proteins. In fact, Parkin undergoes extensive crosstalk with the apoptosis intrinsic pathway. It has been shown that Parkin’s mitochondrial translocation is blocked by pro-survival Bcl-2 proteins, and activated by BH3-only proteins under conditions of inhibited caspase activity (Hollville et al., 2014). Parkin induces apoptosis through degradation of anti-apoptotic Mcl-1 (Carroll et al., 2014). Consistently with this action, knockdown of USP30, which antagonizes Parkin-mediated mitophagy by deubiquitylating OMM proteins, promotes the activation of the mitochondrial apoptosis

- 59 -

pathway (Liang et al., 2015). This finding suggests that USP30 would drive mitochondria toward a mitophagy versus apoptosis decision event.

These findings indicate that there is a crosstalk between the different positive and negative feedback mechanisms regulating fission/fusion programs, mitophagy and both pro- survival and pro-death apoptosis signaling (Fig. 15).

.

Figure 15. Model for mitochondrial dynamics, mitophagy and apoptosis crosstalk. The same stimuli that trigger mitochondrial‐dependent cell death can trigger mitophagy and will then determine whether a cell lives or dies. Elimination of damaged mitochondria is facilitated by mitochondrial fission and promotes cell survival. Mitophagic malfunction leads to the accumulation of dysfunctional mitochondria and makes the cell more susceptible to MMP and apoptosis. When cell death is induced, apoptotic executers inactivate pro-autophagic proteins, thus inhibiting mitophagy. Autophagic degradation of mitochondria is affected by mitochondrial shape/function, with heavily fused mitochondria being a poor substrate that evades autophagic degradation. The selectivity of the fusion machinery makes fusion and mitophagy complementary processes rather than competing ones.

Yellow: damaged mitochondria Red: mitochondria with Mitochondrial membrane permeabilization (MMP), ready to be digested; when cell death is induced, apoptotic executers inactivate pro- autophagic proteins, thus inhibiting autophagy.

From Rambold A S & Lippincott-Schwartz J, 2011

III. Base excision repair in mitochondria

The Base Excision Repair (BER) pathways Base modifications are the most common type of are primarily responsible for the removal of non-bulky base lesions that arise due to endogenous DNA damage, accounting for thousands of lesions oxidation, alkylation, deamination, and methylation. Purine nucleotides frequently per mammalian genome per day. Frequent DNA alterations undergo oxidation of the ring atoms. This chemical modification can lead to the include oxidative and alkylated base products, abasic sites, formation of guanine derivatives, namely 8- oxo-7,8-dihydroguanine (8oxoG), which misincorporated nucleotides. These lesions can alter the base- will be discussed in detail in the next paragraph. pairing properties of the genetic material, and thus can be mutagenic and potentially carcinogenic. Abasic sites, for example, are non-coding lesions; the hydrolytic deamination of C or m5C generates either a U:G or T:G mismatch, respectively, which has the potential to be mutagenic if not repaired before replication. Moreover, because some of these lesions (such as Fapy, or the alkylation of the adenine…) can block DNA polymerases, if not repaired they can induce cellular death (Lindahl & Wood, 1999; Barnes & Lindahl, 2004).

The DNA of all aerobic organisms is constantly exposed to reactive oxygen species (ROS) produced by normal cellular metabolism (Krokan & Bjørås, 2013). Oxidative lesions induced in nucleic acids by the exposure to ROS are preferentially eliminated by the Base Excision Repair (BER) pathway both in nuclear and mitochondrial compartments. So far this pathway is the best characterized in mitochondria (Fig. 15) (You et al., 1999). The enzymes involved in mitochondrial BER are all encoded by the nuclear genome and imported into those organelles.

BER is highly conserved from bacteria to humans, and it consists of five main steps:

 Base lesion recognition and removal by a specific DNA glycosylase  Cleavage of the abasic site  DNA end-processing  gap-filling  ligation

There are two types of BER, short-patch (SP or single nucleotide BER, SN) or long- patch (LP), and both are proposed to occur in mitochondria (Szczesny et al., 2008; Liu et al., 2008; Boesch et al., 2010; Akbari et al., 2008). SP-BER consists in the formation of a single nucleotide gap that is then filled by a DNA polymerase, while LP-BER generates a gap

- 61 - between 2-12 nucleotides that is subsequently filled and ligated restoring the DNA strand. The patch size of the LP-BER results in a replacement of the nucleotides surrounding the modified base. In the nuclear compartment, the main proteins required for the different steps of SP-BER repair are the initiating DNA glycosylase, AP-endonuclease APE1, DNA polymerase β (Pol β) and DNA ligase III (Lig3). In addition, poly(ADPribose) polymerase 1 (PARP1) and XRCC1 participate, even if apparently not in all types of short-patch BER (Campalans et al., 2005; Hanssen-Bauer et al., 2011). Long-patch repair mainly takes place in proliferating cells and uses replication proteins for processing subsequent steps after the glycosylase action and strand cleavage by APE1. These include the DNA polymerases δ/ε, PCNA, FEN1, and a role of Lig1 has also been proposed (Svilar et al., 2011). Which polymerase is used for the incorporation of the second nucleotide depends on the proliferation status of the cells (Akbari et al. 2009); usually Pol β is involved in non-proliferating cells and preferentially Pol δ/ε in proliferating ones. During this process, the original strand is displaced by the combined action of the PCNA (proliferating cell nuclear antigen) and the polymerase, leading to a “flap” structure which is subsequently cleaved FEN1 (flap endonuclease 1) to provide a ligatable 5′-phosphate (Dianova & Bohr 2001).

Very interestingly, Gredilla and colleagues have investigated the efficiency of the BER pathway in mitochondria throughout the murine lifespan. They focused their attention in central regions of mammalian cognition (brain cortex and hippocampus) which are severely affected during aging and in neurodegenerative diseases and they confirmed a regional specific regulation of mitochondrial DNA repair activities with aging. In cortical mitochondria, DNA glycosylase activities peaked at middle-age followed by a significant drop at old age. However, only minor changes were observed in hippocampal mitochondria during the whole lifespan of the animals (Gredilla et al., 2010).

Figure 16. Representation of the short-patch and long-patch base excision repair (BER) pathways in the nucleus and mitochondria of mammalian cells. The factors common to both compartments are presented on an orange background, the ones specific for the nucleus are on a blue background and factors specific for mitochondria are on a yellow background.

*DNA2 is present in the nucleus but its putative role in nuclear BER is not clear so far. APEX1, AP endonuclease; POLβ, δ, ε, γ, DNA polymerase β, δ, ε, γ; RFC, replication factor C; PCNA, proliferating cell nuclear antigen; DNA2, endonuclease/helicase; FEN1, flap endonuclease 1; LIG1, LIG3, DNA ligase 1, 3;

From Boesch et al., 2011

- 63 -

III.a. Modified base excision and strand incision

DNA glycosylases are generally relatively small, positively charged and most frequently single-domain proteins. Two mechanistic types of DNA glycosylases exist depending on whether they possess or not an associated intrinsic lyase activity in addition to their base excision activity: bifunctional or monofunctional glycosylases, respectively (Alexeyev et al., 2013; Jacobs & Schär, 2012). In mammalian cells eleven DNA glycosylases have been identified, among which six are monofunctional DNA glycosylases. These enzymes can be classified into four structurally (and functionally) distinct superfamilies: the uracil DNA glycosylases (UDGs), the helixhairpin-helix (HhH) glycosylases, the 3-methyl-purine glycosylase (MPG) and the endonuclease VIII-like (NEIL) glycosylases (Jacobs & Schar, 2012) (Table 1).

Table 1 List of DNA glycosylases in mammalian cells with their main substrates, mode of action and mutant phenotypes. They flip the damaged base out of the double helix and cleave the N-glycosidic bond of the damaged base, leaving an AP site. There are two categories of glycosylases: monofunctional and bifunctional. Monofunctional glycosylases have only glycosylase activity, whereas bifunctional glycosylases also possess AP lyase activity. Therefore, bifunctional glycosylases can convert a base lesion into a single-strand break Monofunctional DNA Glycosylases: monofunctional DNA glycosylases target damaged bases such as alkylated and deaminated DNA bases and hydrolyze the N-glycosidic bond leaving an apurinic/apyrimidinic (abasic or AP) site in the DNA strand. Because of the

lack of lyase activity, they must rely on AP endonuclease (APE1) to hydrolyze the phosphate backbone. The repair gap, containing a 3′-OH and a 5′-deoxyribose-phosphate (5′dRP) moiety, is tailored by the 5′-dRP lyase activity of Polβ (gap tailoring), and then in the nuclear compartment Polβ fills the single nucleotide gap, preparing the strand for ligation by either DNA ligase I (Lig1) or a complex of DNA ligase III (Lig3) and XRCC1 (DNA synthesis/ligation).

Bifunctional DNA Glycosylases: the glycosylases primarily involved in the removal of oxidized DNA bases are bifunctional and, unlike monofunctional DNA glycosylases, they possess an intrinsic AP-lyase activity so they are able not only to cleave the N-glycosidic bond, but also to incise the DNA backbone at the 3′ of the resulting AP site via either a -elimination or elimination reaction (Dodson & Lloyd 2002) and, in both cases, the ends have to be further processed by APE1. Only in the case of the Neil family, that leave a 3’ phosphate through a elimination, that latter step could be avoided by the action of XRCC1/PNK (Dodson & Lloyd 2002).

In mitochondria six different glycosylases have been described:

 two monofunctional o the uracil-DNA glycosylase 1 (UDG1; or uracil-Nglycosylase1 [UNG1]) (Anderson & Friedberg, 1980) o MUTYH (MYH) (Ohtsubo et al., 2000) a homolog of the E. coli MutY glycosylase  four bifunctional o 8-oxoguanine DNA glycosylase (functional OGG1-homolog of the Fpg E. Coli gene) (Takao et al., 1998; de Souza-Pinto et al., 2001; Audebert et al., 2002) which will be described in detail in the next section o nth (E. coli endonuclease III)-like 1 (NTHL1, or NTH1) (Karahalil et al. 2003) o Nei (E. coli endonuclease VIII)-like-1 and -2 (NEIL1 and NEIL2) (Hu et al., 2005; Mandal et al., 2012).

Little is known about how these enzymes find damaged bases in the genome. All known DNA glycosylases appear to use the same mechanism to excise the alteration on DNA: the base is “flipped out” of the stack into the active-site pocket of the enzyme, where

- 65 - hydrolysis of the glycosidic bond takes place (Friedman & Stivers, 2010). The active site of DNA glycosylases can only accommodate an extrahelical base.

Following the removal of a damaged DNA base by a monofunctional DNA glycosylase, the resulting abasic site is incised by AP endonuclease (APE1), mammalian orthologue of E. coli Xth. APE1 has a dual function. Its C-terminal domain harbours the endonuclease activity for the BER-pathway (Tell et al., 2009) while the N-terminal domain contains a nuclear localization signal and a redox-active domain essential for redox-based transcription regulation. For this reason, APE1 is also known as redox effector factor 1, Ref-1 (Tell et al., 2009). APE1/Ref-1 is essential for cellular survival of mammalian cells. As it has been introduced before, APE1 is the second enzyme in the BER-pathway and it hydrolyses the phosphodiester backbone immediately 5′ to an AP site to produce 3′ OH group and 5′deoxyribose-5-phosphate (Levin & Demple, 1990). The proteolytic truncation of 33 residues at the N-ter sequence gives rise to a mitochondrial isoform of this enzyme (Chattopadhyay et al. 2006). The knocking down of APE1 in neurons has been shown to induce a significant accumulation of oxidative DNA damage (AP sites and strand breaks) because of the absence of efficient repair (Yang et al., 2010). Moreover, the overexpression of this enzyme in cultured cells results in protection towards oxidative damage (Mantha et al., 2012).

III.b. Gap filling by DNA polymerase and strand sealing by DNA ligase

Once the AP endonuclease and the lyase activity of DNA polymerase generate a ligatable 5′-phosphate end, repair synthesis proceeds through the incorporation of one or more nucleotides. Seven DNA polymerase families are known so far (DNA pol families A–D, X, Y and RT). In the mitochondrial BER-pathway, this activity is performed by mtDNA Pol-γ (DNA pol family A member) (Yakubovskaya et al., 2007; Boesch et al., 2011). However, it was recently reported that Pol-β could play an important role in mitochondrial BER (Sykora et al., 2017). In the nucleus, gap filling is performed by the DNA Pol-β, although a smaller contribution of pol-λ (DNA pol family X member and the closest relative of pol-β) has also been reported (Braithwaite et al., 2010; Bjørås & Krokan, 2013). Both Pol-γ and Pol-β display lyase and polymerase activities. LP-BER, in mammalian mitochondria, is facilitated by the

DNA2 helicase/nuclease and a flap endonuclease-like activity most probably due to FEN1 (Kazak et al., 2013).

Mechanistically, the last step of BER is ligation of the nick after gap filling. DNA ligases use the energy of phosphoanhydride hydrolysis to make a phosphodiester bond. Of the three DNA ligases present in mammalian cells, Lig4 is mostly involved in the repair of double-strand breaks and it is not thought to have a role in BER. In the nuclear compartment, both Lig1 and Lig3 participate in BER; because knocking down Lig1 seriously affects nuclear DNA repair, Lig1 rather than Lig3 is thought to be the major ligase in nuclear LP-BER (Levin et al., 2000). Lig3, in contrast to the I and the IV, is the only one that is present both nuclei and mitochondria. Because the inactivation of murine Lig3 is embryonically lethal, functional studies on intact organisms are difficult. In addition, Lig3 is apparently restricted to vertebrates so simpler eukaryotic model systems have not been available. Several mouse models have been generated by conditional inactivation of Lig3 in either the nervous system (Lig3Nes-cre) or heart and skeletal muscle (Lig3Ckmm-cre). The first series of mice were born alive and were indistinguishable from littermates but, already after 2 weeks, they displayed disrupted neurogenesis together with ataxia and growth retardation, and died within 3 weeks. Similarly, Lig3Ckmm-cre mice displayed abnormal heart and muscle morphology and heart failure with precocious death within 3.5 and 4.5 weeks of age. Both Lig3Nes-cre and Lig3Ckmm-cre mice displayed massive mitochondrial dysfunction, likely due to defects in mtDNA replication and/or repair. Mitochondrial Lig3 activity is essential for cellular viability, most probably because of the role of this enzyme in mtDNA replication (Simsek et al., 2011; Ruhanen et al., 2011 Gao et al., 2011). Mitochondrial ligase function could also be replaced by Lig1 once the latter is tailored to enter mitochondrial network (Simsek et al., 2011).

Even though the enzymes discussed above are sufficient for the in vitro reconstitution of BER, interplay between BER enzymes and proteins involved in other aspects of DNA metabolism is necessary for the coordinated repair of DNA lesions [Hegde et al., 2010].

IV. The 8-oxoguanine lesion

Among the four DNA bases, Guanine has the lowest redox potential and for this reason it is the most susceptible to oxidation, yielding different products in DNA such as 8-oxo-7,8- dihydroguanine (8-oxoG) and 2–6- diamino-4-hydroxy-5-formamidopyrimidines (FapyG). 8-

- 67 -

oxoG is produced by ROS in a fairly large amount in both DNA (nuclear and mitochondrial) and nucleotide pools. Therefore, 8-OxoG in DNA can be the consequence of incorporation of 8-oxo- dGTP, formed in the nucleotide pool, or of direct oxidation of the DNA guanine base. The

MutY human homologue presence of 8-oxoG in DNA induces no significant local structural (MUTYH) alterations in its activity are associated with a hereditary changes (Lipscomb et al., 1995). It was shown to be highly mutagenic colorectal cancer syndrome termed MUTYH-associated by inducing transversions upon incorporation of dAMP opposite this polyposis (MAP) lesion by replicative DNA polymerases (Fig. 17) (Kuchino et al., 1987; for a review Boiteux et al., 2016). This alteration is quite stable and appears to accumulate during normal aging. Because of its massive production after oxidative stress, 8-oxoG can be used as biomarker .

Figure 17. Annealing of 8-oxoG characteristic: 8-oxoG can base pair with A or C. only the pair 8oxoG:C is substrate for the DNA glycosylase OGG1.

From Boiteux et al., 2016

IV.a. G:C to T:A transversion and the GO repair system 8-oxoG lesions at syn conformation are highly mutagenic due to their ability to mimic thymine properties; for this reason during replication an adenine is incorporated opposite the lesion and at a second round of replication a T will be paired with the A, leading to a G:C to T:A transversion. To counteract the mutagenic effect of 8-oxoG most species have what is called the GO repair system. Indeed, the GO system is found in organisms ranging from E. coli to humans. Cells deficient in the GO repair system show

high rate of G:C to T:A transversions. The GO system is constituted by three enzymes: MTH1, with 8-oxo-dGTP hydrolyzing activity, 8-oxoguanine DNA glycosylase (OGG1- discussed in detail in the next paragraph), and MutY homolog (MUTYH) with adenine DNA glycosylase activity. All together these enzymes minimize the mutagenic effect of 8- oxoG.

 Oxidative stress can oxidize the dNTP pool. MTH1 (mammalian homolog of the E. Coli MutT) hydrolyzes 8-oxo-dGTP to 8-oxo-dGMP, removing the triphosphate from the deoxynucleotide pool thus avoiding its incorporation in DNA.  Human MUTYH shares 41% sequence similarity to its E. coli counterpart and 79% homology to its mouse homologue (mMYH). MUTYH is responsible for removing adenines misincorporated opposite 8-oxoG; therefore, it reduces G:C to T:A transversions. Interestingly, whereas all the DNA glycosylases remove modified DNA base lesions, MYH (MUTYH) removes a normal DNA base. In fact, MYH (MUTYH) recognizes the 8- oxodG•A base pair and removes the normal DNA base adenosine. Deficiency in MUTYH has been definitively linked to cancer in human (Cheadle & Sampson 2005), emphasizing its critical role in removing premutagenic alterations. The gene for this glycosylase is located on the short arm of chromosome 1 between p32.1 and p34.3 and consists of 16 exons. In human cells, there are three major MUTYH transcripts (α, β and γ) and each transcript has a different 5' sequence or first exon and an alternative splicing, thus multiple forms of human MUTYH proteins are potentially present in nuclei and mitochondria (Ohtsubo et al., 2000).

- 69 -

Figure 18. 8-oxoguanine (GO) accumulates in DNA, via the incorporation of 8-oxo-dGTP from the nucleotide pool or because of direct oxidation of DNA. This increases the occurrence of A:T to C:G or G:C to T:A transversion mutations after two rounds of replication. MTH1 hydrolyzes 8-oxo-dGTP to 8-oxo-dGMP and pyrophosphate. 8-oxo-dGMP is further converted to nucleoside, 8-oxodeozyguanosine (8-oxo-dG), thus avoiding their incorporation into DNA. 8-oxoG DNA glycosylase (OGG1) removes 8-oxoG to initiate base excision repair (BER). OGG1 preferentially excises 8-oxoG opposite cytosine. MutY homolog (MUTYH) excises the adenine inserted opposite 8-oxoG in the template strand. OGG1 may enhance A to C transversion if it excises 8-oxoG opposite cytosine that had been inserted opposite 8-oxoG paired with the adenine in the template DNA (dashed blue line). Blue lines: BER pathways From Nakabeppu et al., 2014

IV.b. The DNA glycosylase OGG1 protein

In human cells, the enzyme responsible for the removal of 8-oxoG opposite to cytosine is the 8-oxoguanine DNA glycosylase 1, also known as OGG1 (Friedberg et al. 2006). The human homolog OGG1 sequence was first cloned in our laboratory by its homology with the yeast enzyme (Radicella et al. 1997).

This enzyme belongs to the helix-hairpin-helix glycosylases superfamily. Across the phylogeny, the OGG1 protein family is less represented and, while it is present in most eukaryotic genomes, its encoding genes The Helix-hairpin-Helix glycosylases: seem to be missing in bacteria and archaea Phylogenetic analysis in almost 100 genomes from bacteria, (Denver et al. 2003). Although they are archaea, and eukaryotes identified six distinct families of HhH DNA glycosylases: functional homologs, the Fpg protein from E.  Nth (homologs of the E. coli EndoIII protein)  OggI (8-oxoG DNA glycosylase I, discussed in the coli and the human OGG1 protein do not chapter)  MutY/Mig (A/G-mismatch-specific adenine glycosylase) present conserved regions. OGG1 shares, as  AlkA (alkyladenine-DNA glycosylase) a member of the HhH superfamily of DNA N-  MpgII (N-methylpurine-DNA glycosylase II)  OggII (8-oxoG DNA glycosylase II) glycosylases, a common ancestor gene with (Denver et al. 2003). Nth/Endo III of E. coli, not with Fpg. The gene

for OGG1 in humans is located on the shorter arm of chromosome 3 (3p25.3) and it consists of 12 exons. Sequence analysis of the promoter region revealed two CpG islands and the absence of TATA or CAAT boxes, suggesting a promoter corresponding to a housekeeping gene to be expressed in a constitutive way (Boiteux & Radicella, 2000). However, the presence of a NRF2 binding site within the promoter offers the idea that external stimuli, including oxidative stress, can modulate the expression of the enzyme (Boiteux & Radicella, 2000). hOGG1 is expressed at highest levels in the thymus, testis, intestine, brain and germinal centre of B cells (Sampath et al., 2012). Not only there is a tissue specific level of expression but aging and exercise can also modify the expression and the activities of this glycosylase. For example, an age-associated decrease in hOGG1 activity was demonstrated in human peripheral blood lymphocytes (Chen et al., 2003).

According to the NCBI databases, 8 different mRNAs could arise from the human OGG1 gene by alternative splicing. These isoforms are classified into two groups (type 1 and type 2) (Fig. 19) based on the last exon of their sequence. Type 1 alternative splice variants end with exon 7, while type 2 with exon 8. All variants share the N-terminal region, which contains a mitochondrial targeting signal.

- 71 -

A

B

Figure 19. Structures of multiple human OGG1 mRNAs and their products. (A) Alternatively spliced forms of OGG1 Structural features of different human OGG1 proteins. All forms of the OGG1 proteins contain a putative MTS at the common N-terminal region. There is a NLS only in the C-terminal end of OGG1–1a. The region exhibiting the highest homology to the corresponding region of yeast Ogg1 protein (aa 128–156). (B) Alternative splicing of the two main isoform of OGG1, the a and the b, respectively the OGG1-1a and OGG1-2a. These two proteins share the first 316aa of their sequences. At the N’ ter a MTS is present, while the NLS is present only in the aOGG1. The C’ ter is quite different between the two proteins and in fact the helix a0, necessary for the DNA glycosylase activity, is not present in the b isoform: for this reason the mitochondrial function of this protein remains a controversial issue.

Adapted from Nakabeppu et al., 1999; Hashiguchi et al., 2004; van der Kemp et al., 2004

There are two major OGG1 transcripts in human cells coding for αOGG1 (alternatively named OGG1 type 1a) and βOGG1 (type 2a). The α-isoform transcribed region spans less than 7 kbp (like its mouse counterpart) and encodes for a protein of 345 amino acids (39 kDa) that localizes mainly in nuclei, whereas βOGG1 is a 424 amino acids (47 kDa) protein that appears to be localized exclusively in mitochondria. The first 316 amino acids are common between the two proteins, while the C-terminus is completely different. In fact, human βOGG1 shares the first six exons with the α-form, but due to alternative splicing the seventh exon is replaced by a new one located 8–9 kbp downstream from the last exon of the α-isoform. The mouse Ogg1 gene codes for a 345 aa protein (mOGG1), corresponding to the human αOGG1, it is composed of 7 exons and it is about 6 kbp long. So far, βOGG1 appears to be absent in mice. While the α-isoform has been extensively characterized, no information regarding substrate specificity, binding activities or kinetic characteristics is available for βOGG1 and the function of its unique C-terminus (108 amino acid residues) is not known. Using a secondary structure prediction program, Hashiguchi and colleagues predicted the formation of an α-helix (αO) in the C-terminus of αOGG1, but no formation of this structure in βOgg1 (Fig. 19) (Hashiguchi et al., 2004). The presence of αO domain was confirmed by crystallography in a 313–323 amino acids region (Bruner et al., 2000; Nash et al., 1997; Norman et al., 2003). The absence of this region in the βOgg1 could explain why no enzymatic activity has been detected for this isoform. The 8-oxoG-DNA glycosylase/APlyase activity of human βOGG1 was reported in a first study (Roldan-Arjona et al., 1997), but not confirmed in another report (Hashiguchi et al., 2004). It should be noted that βOGG1 lacks residues, such as V317 and F319, that are essential components of the 8- oxoG binding site in αOGG1 (Hashiguchi et al., 2004).

At the N terminus of the αOGG1, amino acid residues 9–26 comprise a mitochondrial location signal (MLS), and a nuclear localization signal (NLS) is also present in amino acids 335–342 (Fig. 19). This protein has been extensively characterized both biochemically and structurally. The human αOGG1 requires the presence of a conserved lysine in position 249 in the active site for its DNA glycosylase activity; this Lys is proposed to be activated by a conserved aspartate in position 268. Both residues (Lys-249 and Asp-268) are located in the conserved helix–hairpin–helix structural element that is followed by a Gly/Pro-rich loop and a conserved aspartic acid motif (HhH-GPD), which is essential for DNA glycosylase as well as DNA binding activity. The Arg-154 residue is required for recognition of the base opposite

- 73 - the damage. The study of Bruner and colleagues also showed that a phenylalanine residue at position 319 interacts with 8-oxoG, and may be critical for 8-oxoG recognition by the OGG1 protein. Interestingly, to engage the enzyme catalysis, the binding of the protein to the lesion is not sufficient. For example, the K249Q mutant protein, previously shown to lack 8-oxoG DNA glycosylase activity and AP lyase activity, still maintains a significant binding activity to 8-oxoG and to abasic sites, suggesting that this mutation does not disrupt the recognition/binding mechanism (Fig. 19). The K249 residue appears to be critical for the good alignment of 8-oxoG during the enzymatic reaction and is involved in the nucleophile activation of a water molecule which allows D268 stabilization of the reaction intermediate. Thanks to atomic-force microscopy (AFM) data it has been suggested that the K249Q catalytically inactive isoform forms a DNA-protein co-crystal complex that is nearly identical to the catalytic active DNA-OGG1 complex (Bruner et al., 2000).

Human αOGG1 is a bifunctional DNA N-glycosylase/AP lyase using the conserved K249 residue as a nucleophile. Residue K249 plays the equivalent role of residue P1 of Fpg/Nei enzymes and thus is responsible for the nucleophile attack of C1′ of 8-oxoG nucleotide, which results in the formation of the imino-enzyme covalent intermediate. In the case of human αOGG1, the AP lyase actitity seems to be uncoupled from the excision step (Vidal et al., 2001).

Both phosphorylation and acetylation post-translational modifications have been described for the human αOGG1. The phosphorylation of Ser326 was suggested to mediate the relocalization of the protein to the nucleoli during the S phase (Luna et al., 2005) while the acetylation at Lys338 and Lys341 increases the glycosylase activity by increasing enzyme turnover (Bhakat et al., 2006) through the reduction of the protein’s affinity for AP sites. So far, any of these post-translational modifications have been described for the enzyme inside mitochondria.

IV.c. The absence of OGG1 activity in the cell… what do we know so far?

For the understanding of the roles of DNA glycosylases, the relevance of the oxidative lesions they act upon and the onset of several diseases, the generation of mouse models deficient or transgenic for those enzymes has been extremely useful (for a review, Jacobs &

Schar, 2012). However, the existence of multiple enzymes with redundant capacities and backup repair mechanisms make the task of defining these roles particularly challenging. Probably due to the presence of those backup activities Ogg1-knockout (or Ogg1−/−) mice are viable, fertile and do not show marked pathological defects in adulthood. An accumulation of 8-oxoG has been shown in the liver cells of these animals (in an age- dependent manner) in both nuclear and mitochondrial genomes (Klungland et al., 1999; Osterod et al., 2001; de Souza-Pinto et al., 2001). OGG1 deficiency is associated with a mutator phenotype in both yeast and E. coli models (Michaels & Miller, 1992; Thomas et al., 1997). Because a mutation in critical genes is known to be at the origin of the cancer process, we would expect an increase of tumor incidence in those animals. Surprisingly, investigations in multiple independent OGG1-deficient mouse models have failed to document a strong role for OGG1 deficiency on tumorigenesis in the absence of exogenous stress. The inactivation of OGG1, MTH1 or MUTYH individually has marginal effects, and only by inactivating two (double KO OGG1-MYH, Russo et al., 2004) or all of the critical compounds of the GO system Mth1, Ogg1 and Mutyh genes in the animal there is an enhanced spontaneous mutagenesis (Boiteux & Castaing, 2016 for a review).

Several studies using potassium bromate (KBrO3) to induce 8-oxoG have shown that

−/− Ogg1 mice treated with KBrO3 accumulate higher levels of 8-oxoG in the kidney, but do not develop kidney tumors (Arai et al., 2006). However, a role for OGG1 in oxidative stress- induced carcinogenesis has been reported. For example, studies by Sakumi and colleagues described an increase (5-fold higher) in spontaneous lung adenoma/carcinoma in aged (18 months after birth) Ogg1-/- mice, relative to WT controls [Sakumi et al., 2003]. This tumor formation was reported to be correlated with the accumulation of 8-oxoG in these animals. 8-oxoG content was shown to be significantly elevated in liver cells of MutY/Ogg1 double knockout mice (Russo et al., 2004). However, even if these results seem to link genomic 8- oxoG content with cancer proneness it is important to underline that MutY single mutant mice are primarily predisposed to intestinal tumors, thereby dissociating tissue 8-oxoG content from cancer susceptibility. Together these contradictory results indicate that 8-oxoG accumulation alone may be not a strong predictor of carcinogenesis in animal models.

Inactivating mutations, or mutations that alter the subcellular localization of OGG1, including the G12E point mutation that results in a loss of mitochondrial import, have been described in human tumors (Audebert et al., 2000; Audebert et al., 2002; Paz-Elizur et al., - 75 -

2006; Paz-Elizur et al., 2003). Furthermore, the chromosome region 3p25, containing the OGG1 gene, is frequently lost in lung and kidney cancer. The data concerning the correlations between human OGG1 alterations and cancer development showed infrequent somatic mutations in OGG1 in human gastric, head and neck, lung and kidney cancer (Weiss et al., 2005). The mutant αOGG1-S326C has been extensively investigated for its association with cancer risk. In fact, it shows a slightly reduced 8-oxoG-DNA N-glycosylase activity and it appears to be more sensitive than WT to inactivation by oxidizing agents (Bravard et al., 2009) and it might be a moderate risk factor for lung, digestive system and head and neck cancer (Zhou et al., 2015).

Mitochondria, the main cellular energy generator, are vital for proper neuronal function and survival and impaired mitochondrial functions are considered to be causative factor in neurodegenerative diseases. In a large group of disabling disorders of the nervous system (such as Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, and Alzheimer's disease) evidence for impaired mitochondrial dynamics (shape, size, fission- fusion, distribution, movement etc.) has emerged (Johri & Bea, 2012). For this reason, DNA repair defects that affect mtDNA could clearly be a significant cause of neurological disease. Several studies suggested a defect in human OGG1 activity in cells derived from the brain of patients affected by neurodegenerative diseases like for example Alzheimer disease (Wang et al., 2005; de Souza-Pinto et al. 2008; Liu et al., 2011). In these cells there are higher levels of 8-oxoG compared to those of a healthy population. Different mutants of the enzyme with reduced catalytic activities have been described in association with AD disease (Mao et al., 2007; Jacob et al., 2013). However, the consequences of OGG1 mutations in disease progression need to be confirmed in larger studies. Because in those studies the glycosylase is absent in both mitochondrial and nuclear compartment, it will therefore be important to determine the contribution of mitochondrial dysfunction to the neuropathology resulting from DNA repair defects.

To conclude, studies in human and mouse models have shown that, although impacting important biological processes and enhancing mutagenesis, inactivation of OGG1 alone does not unambiguously result in cancer predisposition or in severe pathologies. However, in mammals, 8-oxoG repair efficiency may be significantly different depending not only on the organ but also on the age of the organism. The complexity of 8-

oxoG repair in mammals makes even more challenging understanding the link between absence of OGG1, mitochondrial DNA stability and, by extension, organel functionality. Mitochondria are more than just the “energy powerhouses of the cell”: these organelles are involved in several facets of cellular metabolism and functions, from cell-cycle regulation to immune responses. Despite the importance of mitochondrial genome maintenance is rapidly gaining more recognition with the discovery of many DNA repair proteins the precise role of DNA repair pathways in the maintenance of mtDNA still remains unclear.

- 77 -

Chapter II. PURPOSE OF THE WORK

- 79 -

Purpose of the work

The purpose of my project was to understand the role of the glycosylase activity of OGG1 in the maintenance of mitochondrial DNA (mtDNA) and the impact of an alteration of its glycosylase activity on mitochondrial physiology. The recognition of the base lesion by the DNA glycosylase is the initiating step of Base Excision Repair (BER) pathway. BER is the main mechanism for the repair of oxidative DNA damage but little is known about its organization into these organelles that are considered to be the main ROS producers in the mammalian cells.

Based on these observations and context, I began my PhD work with the following main objectives:

 Characterization of the mitochondrial localization of αOGG1 by both biochemical and microscopy approaches. (Section 1 of Results)  Evaluation of mitochondrial physiology in cells deficient in OGG1. The main cellular model used was U2OS human cells where OGG1 has been silenced by transfection with siRNAs. (Section 2 of Results)  Characterization of different isoforms of OGG1 for their ability to complement the mitochondria deficiency observed in OGG1 deficient cells. The WT OGG1 present in both nucleus and mitochondria or different mutants impaired in either the mitochondrial import (OGG1-ΔMTS) or in glycosylase activity (OGG1-K249Q) have been used in order to understand if the presence of a functional glycosylase OGG1 in mitochondria is required to protect the cells exposed to an oxidative stress. (Section 2 of Results)  The final part of the project consists in the investigation of the molecular mechanisms by which the absence of the glycosylase activity in mitochondria induces the organelle dysfunction. For this part, the results obtained are still preliminary ones. (Section - Preliminary Results)

Most of the results obtained so far are in revision for publication in a submitted article presents as appendix of this manuscript.

- 81 -

Chapter III. RESULTS

- 83 -

Results – Section 1

a) Characterization of the mitochondrial localization of αOGG1

In human cells, 8 different isoforms have been described as arising by alternative splicing from the expression of the OGG1 gene. Among the several OGG1 isoforms present in human cells, the main two are the α (OGG1-1a) and the β (OGG1-2a). These two isoforms share the first 316 aa of their sequences and at the N- terminus of both of them it is possible to recognize the MTS (Mitochondrial Targeting Signal), while only the αOGG1 presents a Nuclear Localization Signal (NLS) at its C-terminus. For this reason, the αOGG1 was thought to be responsible for the 8-oxoG DNA glycosylase activity in the nuclear compartment, while the β one was proposed to be responsible of the repair activity in mitochondria (Nishioka et al., 1999). However, several observations cast a doubt on the latter. Indeed, in mouse cells only an isoform for this glycosylase which corresponds to the human αOGG1 is present. Furthermore, several studies have shown that βOGG1 has no glycosylase activity (Hashigchi et al., 2004; Van der Kempt et al., 2004). Mainly because of these two reasons, we hypothesized that the repair of 8-oxoG in mitochondria is dependent on the αOGG1. The first question we asked is whether this isoform is indeed present in mitochondria.

First, I verified the presence of αOGG1 in the mitochondrial compartment by Western blot (WB) (Fig. 20). For this purpose, I used a monoclonal antibody (ab124741) that recognizes an epitope at the C-terminal of the protein only present in this isoform.

Figure 20. Subcellular localization of αOGG1 in basal conditions. Western blot analysis of αOGG1 αOGG1 showing that the glycosylase is present in both nuclear and mitochondrial extracts. Antibodies Tubulin α against Tubulin α, Lamin B and MPPB were used to check the purity of the fractions. Lamin B W: Whole Cell Extract; N: Nuclear Fraction; M: mitochondrial Fraction MPPB

The different fractions, whole cell extract (WCE), nuclear and mitochondrial were obtained as described in detail in the Appendix 1. Tubulin, Lamin B and MPPB were used as markers for cytosol, nucleus and mitochondria respectively.

- 85 -

As it is possible to appreciate from Figure 20, αOGG1 is present both in the nuclear and the mitochondrial compartments, indicating that, consistently with a possible role in the repair of mitochondrial DNA, this isoform can be localized to the mitochondria. It is worth noticing that the αOGG1 bands show the same size in both fractions, suggesting that αOGG1 is not cleaved after the import in mitochondrial compartment.

In order to determine whether the mitochondrial-associated αOGG1 was inside the organelles, isolated mitochondria from U2OS cells were exposed or not to Triton and then treated with Protease K (Fig. 21). When mitochondrial membranes were disrupted by pre-treatment with Triton 1%, complete degradation of the proteins by proteinase K digestion was observed, indicating that all the proteins tested are indeed substrates for the protease. As expected, in the absence of a detergent pretreatment, a mitochondrial protein facing the cytoplasm are sensitive to protease digestion (as the cytoplasmic fragment of the outer membrane protein TOM22), whereas matrix proteins (TFAM and MPPB) are resistant to the treatment. αOGG1 behaved as TFAM and MPPB, resulting to be resistant to this digestion. Hence, we infer that the fraction of αOGG1 associated with mitochondria is inside mitochondria.

Figure 21. Mitochondria isolated from U2OS cells were treated with 5µg/mL of Proteinase K, after pre-treatment or not with 1% of Triton. As expected the portion of the outer membrane protein TOM22 exposed to the cytoplasm was degraded by the proteinase K digestion, while the part of the protein embedded in the outer mitochondrial membrane was protected. αOGG1 was protected from proteinase K digestion as were proteins located in the mitochondrial matrix such as TFAM and MPPB, indicating that αOGG1 was indeed imported inside the mitochondria.

The results I obtained are also in agreement with the analysis conducted by our collaborator Dr. Aurelio Reyes and presented in the figure 1 of the manuscript in the appendix of this thesis (Fig. 1 E and 1 F). As it is possible to appreciate from the presented blots of sucrose‐gradient purified mitochondria from HEK293 cells trypsin treatment did not cause the loss of OGG1 signal, underlining the intra-mitochondrial localization of this glycosylase (Fig. 1E of the manuscript in appendix). In order to evaluate if OGG1 is localized in mitochondrial matrix or in association with inner mitochondrial membranes, mitochondrial extracts were treated with high or low salt concentrations and the results obtained suggest a strong association of αOGG1 with the IMM (Fig. 1F of the manuscript in appendix).

Next, the results obtained were confirmed by microscopy. Unfortunately, our antibody against αOGG1 is not compatible with immunofluorescent approaches. For this reason, in my laboratory a plasmid expressing αOGG1(WT) tagged with a FLAG epitope was generated. I initially analyzed the localization of the resulting αOGG1(WT)-FLAG fused protein by WB (Fig. 22A) confirming that, like the endogenous protein, the ectopically expressed OGG1 is localized both in the nucleus and in mitochondria. Furthermore, the tagged protein showed the same susceptibility to proteinase K sequential digestion confirming its localization inside the mitochondria. Taken together, these experiments showed that αOGG1(WT)-FLAG behaves as the endogenous αOGG1 (Fig. 22B).

A B

αOGG1

Lamin B

Tubulin α

MPPB

Figure 22. Subcellular localization of αOGG1-FLAG in basal conditions. A) Western blot analysis of αOGG1-FLAG showing that the glycosylase is present in both nuclear and mitochondrial extracts. Antibodies against Tubulin α, Lamin B and MPPB were used to check the purity of the fractions. W: WCE ; N: Nuclear Fraction; M: mitochondrial Fraction. B) Mitochondria isolated from U2OS transfected cells were treated with 5µg/mL of Proteinase K as described in Figure 21, this analysis indicate that αOGG1-FLAG was indeed imported inside the mitochondria.

- 87 -

Therefore, by using the anti-FLAG antibody, I was able to characterize the subcellular localization of the αOGG1-FLAG fusion protein by confocal microscopy in U2OS cells transfected with the plasmid described above. 48 hours after the transfection, mitochondrial network was labelled by the use of MitoTracker Red or Deep Red. These are red or far red-fluorescent dyes that stain mitochondria in living cells and that are commonly used for mitochondrial localization. As it is possible to appreciate in Figure 23, the FLAG signal is detected in both nuclear and mitochondrial compartments, confirming the results obtained by biochemical fractionation.

Figure 23. U2OS cells were transfected with αOGG1- Flag plasmid, and 48h after transfection were incubated with 200nM of MitoTracker Red (purple). After fixation, cells were incubated with DAPI (blue) and an antibody against FLAG (green) and examined by confocal microscopy

Scale bar: 5µm.

b) Role of the localization sequences at the N and C ter of αOGG1 in the sub-cellular distribution of the enzyme

The sequence analysis conducted by our collaborator, Dr. Aurelio Reyes, revealed the presence of a TOM20 recognition motif (RKYF) in position 97-100 of αOGG1 suggesting the use of a canonical TOM mediated pathway to import protein into the mitochondria. The presence of a MTS in the N-terminal region of the protein is consistent with this hypothesis. We decided then to verify if mitochondrial targeting of the enzyme was dependent on this MTS sequence, by analyzing and quantifying by confocal microscopy the localization of different isoforms of OGG1 fused to the FLAG epitope.

To define the role of MTS we expressed full-length αOGG1 (WT) or a truncated version, lacking MTS, fused to the FLAG epitope and analyzed their subcellular localization. We also analyzed the impact of a mutation in the MTS that was identified in kidney cancer cells in which the glycine in position 12 is replaced by a glutamic acid (G12E) (Audebert et al, 2002). The presence of G12E substitution results in an alteration of the distribution of charges in the MTS sequence capable of disrupting the mitochondrial import of the protein. As shown above, the WT protein with both MTS and NLS targeting signals is present in both mitochondrial and nuclear compartments. The deletion of 10 amino acids from the N-terminus due to an initiation of the translation at the 2nd AUG in frame resulted in a protein mainly located in the nucleus in 94% of the cells. Similarly, the presence of the G12E mutation shown to inactivate the MTS when present in the αOGG1 also yielded an essentially nuclear protein (90% of the cells).

Using the same strategy, I evaluated the role of the NLS in the distribution of αOGG1 between the different compartments. The removal of the NLS gave rise to a protein mostly localized in mitochondria (Fig. 24).

- 89 -

A αOGG1 (WT)-Flag Merge αOGG1 (ΔMTS)-Flag Merge

Mito Mito

FLAG FLAG

DAPI DAPI

αOGG1 (G12E)-Flag Merge αOGG1 (ΔNLS)-Flag Merge

Mito Mito

FLAG FLAG

DAPI DAPI R o le o f th e L o c a liz a tio n S e q u e n c e s in th e a d d re s s in g o f th e p ro te in

B Mitochondrial Localization * * 1 0 0

* * * * s

l * * * *

l

e

c

f o

5 0 %

0

g g g g a a l l la la F F - - F -F ) ) )- ) T E S S 2 T L W 1 ( M N 1 G  ( (  ( G 1 1 G 1 G G G O G  G G O O O    Figure 24. Role of the localization targeting sequencesM ito c h MTSo n d r ianda l lo cNLSa liz ainti othen subcellular localization of αOGG1; A) U2OS cells were transfected with different isoforms of α-OGG1 (WT, G12E, MTS or NLS) and their subcellular localization was analyzed by confocal microscopy using an antibody against the FLAG epitpe (green). The nuclei have been detected using DAPI (Blue) and mitochondrial network is stained with MitoTracker RED (Red). B) Quantification of the percentages of transfected cells presenting the αOGG1 in mitochondrial network. For each constructs between 350-361 cells were evaluated and the experiment was repeated 3 times. Error bar: SD. Statistical significance between the different populations was evaluated with Welch’s t test, ****p<0.0001; **p=0.0004.

In order to determine if the import of the glycosylase in mitochondrial network was influenced by cell cycle phase I analyzed by confocal microscopy asynchronous U2OS cells transfected with the αOGG1(WT)-FLAG construct. Different markers were used to identify in which cell cycle phase were the transfected cells. The incorporation of EdU was used to visualize cells in S phase while an antibody against the kinetochore protein CENP-F was used to identify cells in G2 phase. Cells in G1 were negative for both EdU and CENP-F. Very interestingly, the proportion of cells with mitochondrial staining remained unchanged (Fig. 25) for the different cell cycle phases, demonstrating that the localization of the protein in these organelles is not influenced by cell cycle.

There are evidences that BER proteins are not freely soluble in the matrix of the organelles (Stuart et al., 2005). This is in agreement with our fractionation analysis presented in the Figure 1E of the manuscript in the appendix of this thesis, showing that OGG1 is associated to an insoluble mitochondrial fraction, in close proximity of the mtDNA in nucleoids. Because my confocal

A MERGED B

EdU CENPF OGG1-FLAG (MTS-Turquoise

OGG1-FLAG)

G1

M ito c h o n d ria l lo c a liz a tio n o f  O G G 1

1 0 0 N S N S N S

s 8 0

o p

S 1

6 0

G

G O

 4 0

f

o

% 2 0

0

1 S 2 G G G2 C e ll C y c le P h a s e

T h e K ru s k a l– W a llis te s t b y ra n k s , K ru s k a l– W a llis H te s t[1 ] (n a m e d a fte r W illia m K ru s k a l a n d W . A lle n W a llis ), o r O n e -w a y A N O V A o n ra n k s is a n o n -p a ra m e tric m e th o d fo r Figure 25. Subcellular localization of αOGG1 though the cell cycle. A) Confocalte s tin g microscopy w h e th e r s a m p leimagess o rig in a tofe f ro m th e s a m e d is trib u tio n .[2 ][3 ][4 ] It is u s e d fo r c o m p a rin g tw o o r m o re in d e p e n d e n t s a m p le s o f e q u a l o r d iffe re n t s a m p le s iz e s . It the subcellular localization of αOGG1 proteins (Green= Flag epitope). e Mitochondrialx te n d s th e M a n n – W networkh itn e y U te s ist w h e n th e re a re m o re th a n tw o g ro u p s . T h e p a ra m e tric stained with MTS-Turquise2 (blue). The different cell cycle phases were discriminatede q u iv a le n t o f th e Kbyru s usingk a l-W a lEdUlis te s t is th e o n e -w a y a n a ly s is o f v a ria n c e (A N O V A ). A s ig n ific a n t K ru s k a l-W a llis te s t in d ic a te s th a t a t le a s t o n e s a m p le s to c h a s tic a lly d o m in a te s (purple) and CENPF (red) which labelled the S and the G2 phase respectively.o n e Cellso th e r sina m G1p le . areT h e tne segativet d o e s n o t id e n tify w h e re th is s to c h a s tic d o m in a n c e o c c u rs o r for both EdU and CENPF. B) Quantification of the percentages of cells withf omitochondrialr h o w m a n y p a irs localizationo f g ro u p s s to c h a s tic d o m in a n c e o b ta in s . D u n n 's te s t,[5 ] o r th e m o re p o w e rfu l b u t le s s w e ll k n o w n C o n o v e r-Im a n te s t[6 ] w o u ld h e lp a n a ly z e th e s p e c ific of αOGG1 in the different cell cycle phases. Between 500 and 600 cellss a werem p le p a evaluatedirs fo r s to c h a s andtic d o m thein a n c e in p o s t h o c te s ts . experiment was repeated 2 times. Error bar: SD. Statistical significance between the different populations was evaluated with Kruskal-Wallis test

- 91 - microscopy images showed a punctate distribution of αOGG1 similar to the nucleoids one, I verified by microscopy approaches the co-localization of αOGG1 with markers of mitochondrial nucleoids in order to determine if the enzyme is a nucleoid associated protein. The two markers used were picogreen that stains mtDNA and TFAM, the main proteic component of mitochondrial nucleoids. The mitochondrial network was stained with MitoTracker Deep Red or alternatively with MTS- turquoise2. For the latter, U2OS cells were transfected with a plasmid expressing the cyan fluorescent protein harboring a MTS targeting signal. My results showed that αOGG1(WT)-Flag accumulated at punctate foci within mitochondria and that there is good co-localization between αOGG1(WT)-Flag and both picogreen and TFAM (Fig. 26). I evaluated Pearson's correlation coefficient by using Image J plugin JACOP (Bolte & Cordelières, 2006). The Pearson correlation coefficient has a value between +1 and −1, where 1 underlines a perfect correlation of the two elements, 0 no correlation and -1 a perfect anti-correlation (Fig. 26C). Thanks to the collaboration with Dr. Tristan Piolot at the Curie Institute in Paris, I had the possibility to confirm this last result with Structured Illumination Microscopy (SIM) (Fig. 26D) after the optimization of transfection conditions and staining protocol for this technique.

The results I obtained with those microscopy approaches have been also further confirmed with the biochemical analysis conducted by Dr. Reyes and presented in the figure 2 of the manuscript in the appendix of this thesis (Fig. 2 E). Briefly, mitochondria were isolated from HEK293 cells and fractionated on a top-down iodixanol density gradient. As expected, the established mitochondrial nucleoid protein TFAM was concentrated in the same fractions as the mtDNA. The analysis revealed that also αOGG1 co-purified with mtDNA.

A B Merged Merged

Mito Mito

Picogreen TFAM

FLAG FLAG

C D

Merged

Mito

Person correlation Picogreen/αOGG1-FLAG r=0.75 TFAM TFAM/αOGG!-FLAG r=0.87

FLAG

Figure 26. αOGG1-FLAG is associated to mitochondrial nucleoids. A) Confocal microscopy images of the subcellular localization of αOGG1-FLAG (Red). Mitochondrial network is stained with MitoTracker Deep Red (Purple), while nDNA and mtDNA are stained with picogreen (Green). B) U2OS double transfected with αOGG1-FLAG (Green) and MTS-Turquise2 (blue) were stained with antibodies to detect TFAM (Red) and FLAG (green). C) Pearson coefficients were calculated from 1000 nucleoids (8 cells) between picogreen/FLAG and TFAM/FLAG. D) 3D-SIM was used to confirm colocalization of FLAG (Green) and TFAM (Red) in mitochondrial compartment stained with MTS-Turquise2 (Blue). Scale bars= 5µM, in the inset 2µM.

- 93 -

Finally, several proteins such as Pol, mtSSB and Twinkle have been found to associate with mtDNA only during replication in a highly regulated way (Rajala et al., 2014). To verify if the sub- mitochondrial localization of αOGG1 was dependent on the mtDNA replication status, I used EdU incorporation to visualize the mitochondrial nucleoids undergoing replication on transiently double transfected cells expressing αOGG1(WT)-FLAG and MTS-Turquoise2. An antibody against TFAM was used to label mitochondrial nucleoids. 500 nucleoids were analyzed and I evaluated their replication state together with the presence of the glycosylase. In the results presented in the histograms, it’s clear how the percentage of nucleoids positive for the presence of αOGG1 remains unchanged in both populations, replicating or not (Fig. 27B), indicating that the association of the glycosylase with the nucleoids is not dependent on DNA replication.

A B MERGE

EdU

TFAM

FLAG

Mito

C

MERGE EdU TFAM αOGG1

Figure 27. αOGG1 association with mtDNA is independent from mitochondrial DNA replication; A) immunostaining of double transfected U2OS cells with OGG1-FLAG and MTS-turquise2. EdU has been used to label replicating nucleoids (Purple) in mitochondrial compartment (blue). TFAM is in red while FLAG is in green. B) TFAM has been used as marker for the nucleoids and the percentage of positive signal for the glycosylase for replicating (EdU positive) or not replicating ones (EdU negative) have been evaluated by quantifing 500 nucleoids (4 transfected cells). Statistical significance between the different populations was evaluated with Mann-Whitney test. C) Magnification of the Fig. 27A in which is possible to appreciate that OGG1 is associated with the nucleoids independently of their replication status (EdU positive or not).

Results – Section 2

a) Evaluation of the mitochondrial physiology in OGG1 deficient cells

Mitochondrial dysfunction has been linked to a myriad of human diseases so identifying the functional status of the mitochondrial population is very important. But, how can we assess mitochondrial functionality? We can analyze several parameters among which membrane potential, ROS production and shape. In the second part of my project, I investigated whether the absence of OGG1 induced mitochondrial dysfunctions in U2OS cells, mainly after exposure to oxidative stress induced by menadione (MEN) treatment. MEN is reduced at the level of the complex I of the mitochondrial respiratory chain diverting the electron flow from the normal flux (Loor et al., 2010). Therefore, MEN is commonly used to induce a mitochondrial oxidative stress and to trigger apoptosis.

In order to downregulate OGG1 expression, U2OS cells were transfected with siRNA directed against a region of OGG1 present in all the 8 isoforms. 72 h after transfection, I measured a decrease of ≥ 90% in both the transcript and protein levels of αOGG1 (Fig. 28A). Then, to investigate whether, following MEN treatment, OGG1-depleted U2OS cells were more prone to mitochondrial ROS (mtROS) production or showed change in mitochondrial membrane potential (MMP) they were exposed to 50µM of MEN for 1h. I then stained the cells with alternatively MitoSOX Red, a mitochondrial superoxide-specific dye, or TMRE, an indicator of MMP. Mitosox is a fluorogenic dye selectively targeted to the mitochondria of live cells; in the organelles it is oxidized by superoxide and exhibits red fluorescence. TMRE is a cell permeant, positively-charged, red-orange dye that accumulates only in mitochondria with a normal membrane potential, so depolarized or inactive organelles fail to sequester TMRE. In combination with TMRE, MitoTracker Green was used to follow mitochondrial mass.

I first used flow cytometric analyses to determine the levels of mtROS as reflected by the fluorescence intensity of MitoSOX red. The MEN treatment markedly increased the levels of mitochondrial superoxide in cells transfected with siRNA against OGG1 when compared to the control population (Fig. 28B). Next, concerning the evaluation of MMP also using flow cytometry, while no difference was observed between both cell populations at basal conditions, menadione treatment induced a larger loss of mitochondrial membrane potential in OGG1 silenced cells (Fig. 28C). The total amount of mitochondria was very similar in the two groups analyzed, according to

- 95 - the staining with MitoTracker Green that remains unchanged. Together, these data show a higher menadione-induced mitochondrial dysfunction in OGG1 deficient cells.

A B

OGG1

Tubulin

C

Figure 28. Effects of menadione treatment on mitochondrial physiology of OGG1 silenced U2OS cells; A) U2OS cells were transfected alternatively with an siRNA control of with siRNA against OGG1; 72h after the silencing the WCE obtained from these two populations has been analyzed by Western blotting for the presence of OGG1; tubulin antibody was used as control. B) Mitochondrial superoxide production was quantified by using Mitosox in both populations, silenced or not (respectively U2OS siControl (top panel) and U2OS siOGG1 (bottom panel)) in the absence (grey) or presence (red) of 50 μM Menadione. Probability binning distribution difference analysis was used to quantify the effect of Menadione treatment. Chi-squared based T(X) metrics were calculated for Menadione treated samples as compared to their untreated counterparts, and are indicated in each panel. Results are representative of 4 independent experiments. C) Effect of Menadione-treatment on mitochondrial membrane potential and mitochondrial mass, as detected with TMRE and MitoTracker Green, respectively. TMRE and MitoTracker Green fluorescence was measured for U2OS siControl (top panels) and U2OS siOGG1 (bottom panels) cell lines, in the absence (left panels) or presence (right panels) of 50 μM Menadione. In the histograms were plotted MitoTracker Green and TMRE fluorescences in the absence (grey) or presence (red) of 50 μM Menadione. Probability binning distribution difference analysis was used to quantify the effect of Menadione treatment on mitochondrial membrane potential. Chi-squared based T(X) metrics were calculated for Menadione treated samples as compared to their untreated counterparts, and are indicated in each panel. Results are representative of 3 independent experiments.

Dr. Anna Campalans has also performed some real time microscopy analysis on living cells transfected with siControl or siOGG1 and stained with MitoTracker green (green) and TMRE (red). The purpose of this experiment was to analyze the recovery of the two different cell populations after exposure to 50 µM menadione for 1 hour. The results from these experiments are presented in the figure 5 of the manuscript in the appendix of this thesis (Fig. 5F). This approach confirmed the larger loss of mitochondrial membrane potential in cells transfected with siRNA against OGG1.

With the aim of evaluating if the effects of the absence of OGG1 on mitochondrial physiology were apparent in another cell model, primary mouse embryonic fibroblasts (MEF) from WT or Ogg1- /- mice were treated with menadione (50µM, 1h, 37 °C) and then stained with MitoSox Red or TMRE, followed by flow cytometry analysis.

As it is possible to appreciate from the histograms in Figure 29, the oxidative treatment induced in Ogg1-/- MEF an increase of mtROS production with a tendency similar to the one observed for U2OS silenced cells. However, under the experimental conditions used, no significant differences in MMP could be observed between WT and Ogg1-/- cells.

Taken together these results show that in OGG1-deficient cells mitochondria are more affected than in OGG1-proficient ones after the induction of a MEN-induced oxidative stress.

A B

Figure 29. Effects of menadione treatment on primary MEF cells isolated from WT or OGG1-/- KO mice; A) any substantial differences have been observed concerning mitochondrial membrane potential in response to oxidative treatment between WT or KO cells. B) in terms of mtROS production a significant increase is observed after stress in the absence of the enzyme (p=0.0196). The figure represent the mean of 3 independent experiments; Error Bars=SD. Statistical significance between the different populations was evaluated with Welch’s t test

- 97 -

b) Mitochondrial dysfunction of OGG1 deficient cells can be complemented by expressing a functional αOGG1 in mitochondria

In order to determine if the alpha isoform of OGG1 could complement the mitochondrial defects observed in OGG1 deficient cells, we established U2O2 cell lines stably expressing either the αOGG1(WT)-FLAG, or an isoform that is exclusively nuclear, the αOGG1(MTS)-FLAG. Silent mutations were introduced in the sequence of the ectopically expressed proteins in order to make them resistant to the siRNAs against OGG1. The ORFs have been cloned in a pIRES-AcGFP plasmid, containing an internal ribosome entry site (IRES). This permits both the gene of interest (cloned into the MCS) and the AcGFP gene to be translated from a single bicistronic mRNA. The presence of the AcGFP fluorescence allowed me to select the transfected cells by flow cytometry. Even if a special protocol for obtaining stable U2OS cells is not necessary I have optimized several parameters such as culture conditions, seeding densities, and G418 concentration used to optimize the selection of cells expressing the glycosylase. All the pIRES-AcGFP plasmids have been produced by the CEA facility CIGEX under the supervision of Dr. Campalans.

As it is possible to appreciate from the example in the following figure (Fig. 30), the plasmids carrying the silent mutations are able to escape the effect of the siRNA against OGG1. Moreover,

A B

siControl siOGG1 siControl

FLAG -

OGG1(WT) αOGG1(WT)-FLAG

α

FLAG

-

Mut) -

αOGG1(WT-Mut)-FLAG

OGG1(WT α Figure 30. Analysis of stable transfected U2OS cells; A) U2OS stably expressing αOGG1(WT)-FLAG or αOGG1(WT-Mut)- FLAG were transfected with siRNA control or against OGG1. 72h after, by confocal microscopy, I verified the effects of the introduction of the silent mutations in α-OGG1-FLAG sequence on the silencing. As it possible to observe the mutated plasmid has the capacity to escape the silencing. B) As expected, the introduction of silent mutations has no impact on the subcellular localization of the enzyme. In the figure DAPI was used for nuclear staining (blue) and OGG1-FLAG was visualized by the use of an antibody against FLAG and a secondary antibody coupled to Alexa 594 (red).

the introduction of those mutations did not influence the subcellular localization the DNA glycosylase.

So, wild-type U2OS cells or cells stably overexpressing the different isoforms of the protein were transfected with the siRNA against OGG1, removing like that the endogenous protein. 72 h after transfection, the cells were treated with menadione, stained with TMRE or MitoSOX, and analyzed in a microplate reader. These analyses confirmed my previous observations obtained by flow cytometry (Fig. 28). The results from these experiments clearly showed that the mitochondrial dysfunction that characterizes the OGG1-deficient cells is entirely complemented by the expression of a WT version of αOGG1 but not by the expression of a protein lacking the first 10 aminoacids and therefore a functional MTS (Fig. 31). These important findings indicate that the presence of the enzyme inside the mitochondria is essential to prevent organelle dysfunctions after oxidative treatment. A B C

Vinculine

FLAG

Figure 31. Effects of menadione treatment can be complemented with the WT form of αOGG1 but not with the αOGG1(ΔMTS) mutant; A) Western blots of WCE from U2OS stably expressing the WT or the DMTS isoform of the enzyme. The two populations have similar FLAG expression levels. B) Wild type U2OS or U2OS stably expressing αOGG1(WT) or αOGG1(ΔMTS) were transfected with siRNA against OGG1. Mitochondrial membrane potential was evaluated with TMRE probe in a plate reader from NT or treated cells with 50µM of menadione for 1h. The histograms correspond to the ratio TMRE/MitoTracker Green intensities. C) The same experimental design presented in B was used to evaluate mitochondrial ROS production by using the mitochondrial superoxide indicator, MitoSox. The histograms correspond to the ratio MitoSox/Hoecst 33342 fluorescences. The figure in B and C represent both the mean of 3 independent experiments; Error Bars=SD. Statistical significance between the different populations was evaluated with Mann-Whitney test

- 99 -

Mitochondria are extremely dynamic. These organelles fuse and divide constantly and their morphological changes can reflect different cell cycle phases, cellular metabolic demands but also cell damage. For these reasons, the classical ‘textbook’ view of mitochondria as single bean-shaped structures, results quite far from reality. To explore the effect of MEN on mitochondrial morphology, once again I used U2OS cells, which are well suited for this purpose because of their large cytoplasm and well developed mitochondrial network.

First of all, wild type U2OS were transfected with siRNA control or against OGG1. After 72h the cells were subjected to the oxidative stress and the state of the mitochondrial network was monitored. To be able to visualize it, I stained the cells with the MitoTracker Red probe before fixation. As is possible to appreciate from the Figure 32, MEN treatment had a strong impact on mitochondrial morphology, inducing a transition from a filamentous and branched conformation to a fragmented and round one. Interestingly, even though MEN treatment induced mitochondrial fragmentation in both cell populations, transfected either with the siRNA control or against OGG1, the latter ones were significantly more affected.

Normal Mitochondrial Network

Normal Figure 32. Mitochondrial morphology in non-treated cells or cells exposed to Menadione; U2OS cells were transfected with an siControl (siCN-Blue) or with siOGG1 (Red). The cells were exposed or Fragmented not to Menadione for one hour and mitochondrial network was stained with MitoTracker Red. Mitochondrial morphology was evaluated by confocal microscopy and the percentage of cells presenting a normal mitochondrial network was quantified. 80 cells were counted and the experiment repeated 3 times. Statistical significance between the different populations was evaluated with Mann-Whitney test ****p<0.0001; ** Not Treated 50µM MEN p=0.004

To assess whether αOGG1 could protect mitochondria from fragmentation, I transiently transfected U2OS cells with αOGG1(WT)-FLAG and exposed them to 50µM of MEN for 1h. Overexpressing the full-length protein clearly protected the mitochondrial network from the MEN- induced fragmentation. Interestingly, if instead of expressing the full-length DNA glycosylase, the transfection was performed with a construct coding for αOGG1(ΔMTS)-FLAG, the truncated glycosylase lacking its MTS and therefore not imported into mitochondria, failed to protect from the fragmentation induced by the MEN treatment.

In order to evaluate if the enzymatic activity of OGG1 was essential for the protective effect, cells were transfected with the catalytic mutant αOGG1(K249Q)-FLAG. All known DNA glycosylases/AP lyases of the HhH-GPD family use as the catalytic nucleophile a lysine residue on the enzyme. In the case of human OGG1, the ε-NH2 group of K249 is involved in the attack on C-1′of the 8-oxoG-containing nucleotide (Nash et al., 1997), resulting in the cleavage of the base. The OGG1 mutant K249Q exhibits substrate binding specificity similar to that of the WT protein but it is catalytically inactive (van der Kemp et al., 2004). Representative images and the quantification of the percentage of cells with a preserved mitochondrial network are presented in Figure 33. This quantification was conducted on cells showing the presence of OGG1 in nuclear and mitochondrial compartments for the constructs αOGG1(WT) and αOGG1(K249Q) and only in the nuclei for the αOGG1(ΔMTS), as determined by FLAG immunoreactivity. The results from these experiments show that the catalytic mutant αOGG1(K249Q)-FLAG also failed to protect mitochondria from fragmentation after exposure to menadione, strongly suggesting that the enzymatic activity of the DNA glycosylase is required for maintaining a normal mitochondrial network after exposure to an oxidative stress.

- 101 -

A

Not transfected WT MTS K249Q Figure 33. Mitochondrial network is protected by the overexpression

of WT isoform of αOGG1 in cells exposed to oxidative stress; A) U2OS cells were transiently transfected with different isoforms of αOGG1 and exposed or not to Treated Non Menadione for one hour. Mitochondrial network was stained with MitoTracker Red before fixation and then the cells were stained with FLAG antibody. Organelle morphology was

Menadione evaluated by confocal microscopy. Scale bars: 5µM. B) Quantification Normal Mitochondrial Network of the percentage of transfected cells presenting a normal mitochondrial network. 50 cells B n s * * * 8 0 were counted and the experiment N T repeated 3 times. Error Bars: SD.

W T

s l l  -M T S Statistical significance between the

e 6 0 c K 2 4 9 Q different populations was

f o

evaluated with Tukey’s multiple % 4 0 comparisons test p<0.0001

2 0

N T M e n

I have also conducted further analysis on mitochondrial morphology parameters by applying Kernel’s algorithm on the binarized images (as described in Koopman et al., 2005) for which examples are presented in Figure 34. My results show that both the length and the degree of branching were comparable after treatment between not transfected cells and the ones overexpressing αOGG1(ΔMTS)-FLAG or αOGG1(K249Q)-FLAG mutants, while, on the other hand, cells overexpressing αOGG1(WT)-FLAG had a better preserved mitochondrial population.

Vs

WT WT

Not Transfected Not

Vs

MTS

Δ

Not ransfected Not

vs vs

K249Q K249Q Not Transfected Not Not Transfected αOGG1(WT) αOGG1(DMTS) αOGG1(K249Q)

Figure 34. Measure of mitochondrial length and branching of U2OS cells transfected with different isoforms of αOGG1 exposed or not to Menadione; Binarized images of MitoTracker Red stained mitochondria were analyzed for Aspect Ratio (AR) and Form Factor (FF). The dots represent single mitochondria. For each conditions 8 cells were analyzed and the AR and FF were evaluated for more than 2000 mitochondria (between 2000 and 3000 mitochondria). The Kolmogorov–Smirnov test was used to evaluate the statistical significance of these morphological parameters between the different conditions analyzed and the p values are presented in the table.

Taken together these results clearly demonstrate that only the overexpression of the αOGG1(WT)-FLAG construct allows the maintenance of a normal mitochondrial network, underlining not only that the presence of the enzyme inside these organelles is mandatory, but also suggesting that the protection is due to its glycosylase activity more than to a simple structural role of the protein.

- 103 -

c) Influence of oxidative stress on mitochondrial localization of αOGG1

In order to evaluate if oxidative stress could have an impact on the subcellular localization of αOGG1, U2OS cells were transiently transfected with the WT or the K249Q mutated version of αOGG1 both tagged with a FLAG tag at their C-ter. After 48 hours, the transfected cells were treated with 50µM of menadione for 1, 2 or 3 hours, then washed and stained for 30 minutes with 200nM of MitoTracker Red. After fixation, the transfected cells were stained with an antibody against the FLAG epitope and the percentage of cells with mitochondrial localization of the glycoslase were evaluated.

If compared with the wild-type isoform, the αOGG1(K249Q) mutant presents a very different behavior after oxidative stress. Indeed, although before treatment the two populations showed the same percentage of cells with mitochondrial αOGG1, very few cells present detectable levels of αOGG1(K249Q)-FLAG in mitochondria after exposure to menadione while for the WT no differences were observed (Fig. 35 A-B). The quantification of cells with a FLAG mitochondrial staining in non- treated cells and at different times after exposure to menadione are shown in the graph presented in Figure 35B. While for the WT cells no difference were observed, for the mutant αOGG1(K249Q) we registered a progressive, significant loss of the mitochondrial staining at increasing times of menadione exposure. The possible meaning behind this observation will be addressed in the discussion.

A

B

αOGG1(WT)-FLAG αOGG1(K249Q)-FLAG

NT 1h 2h 3h

MEN Figure 35. Quantification of mitochondrial localization of the αOGG1(WT) and the mutant αOGG1(K249Q), in non treated cells (NT) and cells exposed for 1, 2 or 3 hours to Menadione A) The mitochondrial network was stained using MitoTracker Deep Red (red). The subcellular localization of both αOGG1 isoforms were analyzed by confocal microscopy using an antibody against FLAG (green), B) The percentage of cells with FLAG mitochondrial signal was quantified. For each conditions between 150 and 650 cells have been evaluated. The experiment has been repeated 3 times. Error Bars=SD. Statistical significance between the different populations was evaluated with Tukey’s multiple comparisons test

- 105 -

d) Menadione induces the accumulation of 8oxoG in mtDNA

The failure of the αOGG1 active site mutant to preserve the mitochondrial network suggests the possibility that the accumulation of 8-oxoG in mtDNA is, at least partially, responsible for the mitochondrial dysfunctions observed. The results presented above clearly indicate that overexpression of the glycosylase αOGG1 can prevent menadione induced mitochondrial dysfunction. It was then natural to investigate the effects if this treatment on mtDNA, in particular if menadione induces 8-oxoG on the molecule. An approach to evaluate the accumulation of 8-oxoG in the mitochondrial compartment is to detect the lesion by immunofluorescence. After 72h of transfection with either a siRNA against OGG1 or a control siRNA, cells were treated with 50µM of menadione for 1h at 37°C and then washed. Cells were stained for 30’ with MitoTracker Red, thus allowing a recovery period (Fig. 36). After fixation, cells were treated with RNase and the DNA was denaturated with NaOH, and the induction of 8oxoG was monitored using an antibody specific against the lesion. The signal of 8-oxoG in mitochondria was stronger after the exposure to menadione in both types of cells, and higher in the cells deficient for the glycosylase.

No Treatment 50 µM MEN

Figure 36. OGG1 deficiency increased accumulation of 8-oxoG in the mitochondrial

siControl DNA of U2OS cells in the presence of menadione. Fixed cells pretreated with RNase were subjected to a mild denaturation with 25mM NaOH before incubation with 8-oxoG monoclonal antibody (Green). Mitochondrial network was stained by using MitoTracker Red (mask indicated in the image), and the nucleus was visualized with DAPI

(Blue).

1

- siOGG1

Even if this protocol has been used by several authors (mostly from Nakabeppu’s group) to demonstrate the increase of 8-oxoG in mitochondrial compartment after oxidative stress I do not find it solid enough to perform quantitative analysis. Partially this is due to the fact that, following the proposed protocol, the antibody doesn’t seem to recognize only lesions in mtDNA, but the staining is also very diffuse in the cytoplasm even in regions without mitochondria.

Another more sensitive assay for quantifying mitochondrial DNA damage is the long- amplicon quantitative polymerase chain reaction (QPCR). The use of DNA polymerase amplification of long DNA products to study DNA damage has been described by Van Houten’s group already in 1997 and this approach has been used by the authors to demonstrate how mtDNA is more extensively damaged than nuclear DNA by oxidative stress (Yakes & Van Houten, 1997). The rationale of this technique is that the presence of DNA lesions in the long sequence probed will diminish the efficiency of amplification. Therefore, by using equal amounts of DNA from different samples (for example treated or not treated samples), DNA with fewer lesions will amplify to a greater extent than damaged DNA. The amplification of a short fragment of mtDNA of 200 bp, unlikely to contain lesions due to its small size, is used to normalize the amount of DNA used in the reaction. I decided to optimize this technique to evaluate 8-oxoG induction after menadione treatment. Many kinds of DNA lesions can slow down or block the progression of DNA polymerase. Because 8-oxoG does not significantly stall progression of DNA polymerase (Furda et al. 2012) the samples were pretreated with formamidopyrimidine DNA glycosylase (FPG), which excises 8-oxoG and generates a single strand break at the site of the lesion, therefore creating a block for the DNA polymerase. The same DNA templates have been treated or not with FPG and then amplified. As it is possible to appreciate from the gels presented in the example in Figure 37, menadione treatment induced a reduction in the yield of amplification of the long fragment. If without FPG this reduction is sometimes quite mild, after the pretreatment with the enzyme we observed a strong decrease in the amplification of the long fragment. This suggests that menadione treatment induces indeed the accumulation of 8-oxoG in mtDNA.

- 107 -

FPG - FPG + FPG - FPG + NC NT MEN NT MEN NC NT MEN NT MEN

8,9 kb

248 bp

Figure 37. Agarose gel for the evaluation of mtDNA damage after oxidative treatment. U2OS cells 6 were seeded in 10cm petri dishes at 0,75X10 cells/well and treated after 48h for 1h with 50µM of menadione at 37°C. The cells were then immediately harvested with trypsin and the DNA isolated, treated with FPG and amplified by QPCR. Equal amounts of DNA isolated from each cell line (treated (MEN) or not (NT)) were analyzed by QPCR for a small (0.248 kb) mitochondrial sequence (primer set 14,620/14,841) and a large (8.9 kb) mitochondrial sequence (primer set 5,999/14,841). QPCR samples not containing DNA served as negative controls (NC).

In order to verify if in the absence of OGG1 a greater amount of 8-oxoG was present in mtDNA, I performed the very same experiment on OGG1 deficient U2OS cells. After the expression of OGG1 was inhibited by siRNA, the cells were exposed to 50µM of MEN for one hour and immediately harvested. The presence of lesions in mtDNA was determined as above by amplification of the long and short fragments except that I performed the quantitation of PCR products fluorimetrically using the PicoGreen dsDNA quantitation reagent and fluorescence plate reader. DNA concentrations were calculated based on the slope of a standard curve. As above, the results obtained with the short sequence are used to normalize the data obtained with the large fragment.

As is possible to appreciate in Figure 38, loss of amplification of the long fragment of mitochondrial genome occurred at significantly higher levels in the absence of the DNA glycosylase when compared with cells in which the endogenous glycosylase is still present. After FPG treatment, almost 3-fold more DNA polymerase blocking lesions are present in siOGG1 cells if compared with siControl ones. These data demonstrate that a 60-min exposure to MEN 8-oxoG accumulates in mtDNA at significantly higher amounts in the absence of OGG1.

A siControl siOGG1 -FPG +FPG -FPG +FPG NT MEN NT MEN NT MEN NT MEN

Long Figure 38. Analysis menadione induced 8oxoG in U2OS cells silenced or not for OGG1; A) Example of Short agarose gel of QPCR reaction for small (0.248 kb) (primer set 14,620/14,841) and large (8.9 kb) (primer set 5,999/14,841) fragments of mtDNA. B B) The concentration of PCR products was estimated fluorimetrically using

the PicoGreen dsDNA quantitation reagent. The graph represents the standard curve performed with

c lambda phage DNA at known concentration. C) DNA pre-treated or not with FPG was amplified using the same primer pars as in A. The relative

amplification of the long fragment, PicoGreen Fluorescence PicoGreen DNA λ (ng/µL) normalized for the short one, is plotted. The results shown correspond to the amplification C obtained from Menadione treated 8oxoG induction after treatment cells, relative to NT cells for which amplification is set to 1.

The experiment was repeated 4 times. Error bars: SD

Statistical significance between the different populations was evaluated with Mann-Whitney test

- 109 -

Preliminary results

a) Quantification of mtDNA and mtRNA

Mitochondrial genome content and mtDNA transcription are critical for mitochondrial health. Our results obtained so far unveil a clear mitochondrial dysfunction in OGG1 deficient cells exposed to an oxidative stress and clearly suggest that the observed phenotypes are due to the absence of an active glycosylase in the mitochondrial compartment. To assess if there are any differences in transcription capacity between U2OS with or without OGG1 we developed a qPCR- based assay to determine the relative amount of mitochondrial transcribed ETC subunits before and after oxidative treatment.

In human cells, the mitochondrial genome is transcribed from the 2 DNA strands as single polycistronic precursors RNA and no intron sequences are present (Taanman, 1999) (Fig. 39A). This polycistronic RNA is further cleaved to give rise to the individual RNAs that are subjected to several post- transcriptional modifications that modulate their stability (Fig. 39B). As a consequence of this, even if the level of transcription is the same for all the mitochondrial genes, the relative amount of individual mRNAs could be greatly different (Chujo et al., 2012). In order to evaluate the levels of mitochondrial transcripts, we used TaqMan probes targeting mtDNA coding regions and inter-genic regions (Figure 39C). Because of the absence of introns in mitochondrial genome, the same TaqMan probes could also be used to control the level of mtDNA.

While Menadione treatment induces only a slight reduction in the levels of mtDNA in cells transfected with the siRNA against OGG1 compared to control cells, the levels of some transcripts (such as ND1) were dramatically reduced in the cells deficient for OGG1. The levels of COX5, a subunit of the respiratory chain encoded by nuclear DNA, were not affected by menadione treatment. Unfortunately, we registered huge variability between several replicates for some of the probes analyzed (it is the case for COX3), so the trend observed needs further experiments to be confirmed (Fig. 40).

A B

C

Figure 39. Variable steady-state levels and half-lives of human mitochondrial mRNAs. (A) Schematic representation of human mtDNA with its gene organization and transcriptional units. The outer and inner circles represent the H- and L-strands of mtDNA, respectively. Protein (blue) and ribosomal RNA (orange) genes are interspersed with 22 tRNA genes (yellow, with single-letter amino acid codes). (B) Copy number of mRNAs (means ± SD, n = 3) plotted against their half-lives in HeLa cells. The correlation factor (R2) of the plot is 0.601. (C) Position of the TaqMan probes used in our study on the heavy (in green) and the light (purple) strands of mtDNA.

Adapted from Chujo et al., 2012

- 111 -

A OGG1

**p=0,0022

Expression

RelativeGene Transcription

B ND1 ND4 tRNA-1

NS NS NS

siControl

siOGG1

Expression

RelativeGene mtDNAcontent

C ND1 COX3 tRNA-1 COX5

*p=0,0317 NS NS NS

Expression

RelativeGene Transcription

Figure 40. Analysis by real time PCR of DNA content and mRNA levels in cells exposed to menadione; A) U2OS cells were transfected alternatively with an siRNA control or with siRNA against OGG1; 72h after the silencing the RNA obtained from these two populations was extracted, retrotrascribed in cDNA and analyzed by Real time PCR to verify the efficiency of the silencing. B) Graphs represent the fold change of different regions of mtDNA (using probes described in Figure 39C, covering the regions encoding for ND1, ND4 and tRNA-1). Menadione treatment caused a mild, not significant loss of mtDNA in cells deficient for OGG1 when compared to control population. C) Oxidative damage caused an important reduction of ND1 transcrips while less obvious is the interpretation of the results obtained for COX3 and tRNA-1. The treatment does not affect the expression of the ETC subunit COX5 encoded by nuclear genome. The experiment was repeated 5 times. Error bars: SD. Statistical significance between the different populations was evaluated with Mann-Whitney test.

b) Evaluation of apoptosis in U2OS cells in the absence of OGG1

As it has been already discussed in the introduction, an increase in fission events results in the fragmentation of mitochondrial network under stress conditions. Fragmentation is essential to the removal of dysfunctional mitochondria through mitophagy.

In order to understand if the mitochondrial dysfunction observed in OGG1 deficient cells after exposure to menadione could trigger mitophagy, Dr. Campalans monitored the accumulation of LC3B aggregates, an autophagy/mitophagy indicator, and indeed observed a higher accumulation of LC3B in the cytoplasm of OGG1 deficient cells, mostly after exposure to the oxidative stress (Fig. 41). This result suggests that, in the absence of OGG1, an higher autophagy is induced as reflected by LC3B cytosolic accumulation. However, this observations need to be confirmed by further experiments.

A B

DAPI LC3B MT-DR Merged

LC3B aggregates 60

50

siControl 40 30

20 1

- 10 No Treatment No 0

NT Men siOGG1

sicontrol siOGG1

Figure 41. Menadione treatment induces

mitochondrial dysfunction in cells deficient in OGG1; A) Cells transfected with a siRNA against OGG1 or against a

siControl control sequence, were treated or not with menadione and incubated with an

antibody against LC3B (red), marker of 1 - autophagy. Mitochondrial network is visualized with MitoTracker Deep Red

and nuclear DNA is stained with DAPI. B) 1h 50µM Menadione 50µM 1h

siOGG1 Percentages of cells presenting LC3B aggregates was quantified. 4000 cells were evaluated in 2 independent experiments

- 113 -

Caspases are crucial enzymes of apoptosis. Among them, caspase-3 is a frequently activated death protease, which catalyzes the specific cleavage of many key cellular proteins. This enzyme is indispensable for triggering apoptotic chromatin condensation and DNA fragmentation in mammalian cells. Apoptotic pathways activated by caspase-3 are dependent on mitochondrial cytochrome c release and caspase-9 function and are essential for certain processes associated with the dismantling of the cell and the formation of apoptotic bodies.

With the aim of understanding if apoptosis is induced at higher levels in OGG1 deficient cells when compared with not silenced ones after exposure to menadione, U2OS cells, silenced or not, were grown on glass coverslips and treated with 50 μM MEN for 1 h and cells were then analyzed with a confocal microscope.

My preliminary analysis of confocal images revealed the expected fragmentation of the mitochondrial network in OGG1-deficient cells (Fig. 42) together with the accumulation of picnotic nuclei, indication of cell death, after exposure to 50 µM of menadione for one hour. Further preliminary experiments showed that the picnotic nucleus were positive for the staining with cleaved caspase-3 antibody, suggesting that cell death was triggered by apoptosis (Fig. 42). These preliminary data need to be confirmed and more thoroughly quantified.

MERGE DAPI MT-DR CAS-3

siControl

No Treatment No

1

-

siOGG1

siControl

1

1h 50µM Menadione 50µM 1h

- siOGG1

Figure 42. Cleaved caspase-3 staining in the U2OS cells with or without OGG1. After treatment of the cells with 50 μM Menadione for 1 h, the mitochondrial network was stained with 200 nM MitoTracker Deep Red for 30 minutes, then fixed in a 2% formaldehyde solution in warm, complete medium, washed, and permeabilized with Triton, as described in the appendix. Cells on cover glasses were then incubated with an anti-cleaved caspase-3 antibody followed by Alexa Fluor-594 secondary antibody (red).. Finally, cover glasses were washed in PBS and treated with 1 μg/ml DAPI solution in PBS (nuclei stain blue), mounted, and visualized with a confocal microscope. Enhanced activation of caspase-3 is observed only for siOGG1- U2OS cells after oxidant stress.

- 115 -

Chapter V.

DISCUSSION AND CONCLUSIONS

- 117 -

Discussion and general conclusions of the work…many doors open

Characterization of the mitochondrial localization of αOGG1: Although mitochondrial OGG1 was suggested to have a crucial role against mitochondrial DNA damage (de Souza-Pinto et al., 2001), which isoform of OGG1 is responsible for that activity in human cells is still unclear. It is very well known that through alternative processing of pre-mRNAs, individual mammalian genes often produce multiple mRNA and protein isoforms that may have related, distinct or even opposing functions but, why in human cells so many transcripts are codifying for different isoforms of OGG1 is still an open question. Nowadays, we still don’t know if there are tissue-specific splice variants of OGG1 or if different physiological conditions may have a role in the generation of one isoform instead of another. Indeed, it is not even clear if the reported mRNAs are translated into proteins in the cells (Furihata, 2015). In vitro several of these polypeptides are not even active as DNA glycosylases as they fail to excise 8oxoG from DNA; still, they could have other non-yet identified roles. The βOGG1 (also named OGG1-2a) localizes only in mitochondrial compartment and it was supposed to be the DNA glycosylase responsible for the repair of 8-oxoG in these organelles. However, no DNA glycosylase activity could be detected with the purified protein (Hashiguchi et al., 2004; Van der Kempt, 2004). The analyses I have conducted during the first part of my thesis project allowed us to conclude that αOGG1 (or OGG1-1a) is present in both mitochondrial and nuclear compartments (Fig. 20). In support to our findings on a possible role of this particular isoform in mitochondrial physiology maintenance, several groups observed that by targeting the h-αOGG1 with an additional MTS from MnSOD to mitochondria, cells showed an increased mitochondrial DNA repair and a better cellular survival after oxidative stress. This suggests that this enzyme may be indeed active in these organelles (Dobson et al., 2000; Rachek, 2002; Druzhyna et al., 2003 and 2005).

The experiments I have conducted show that αOGG1 is indeed imported inside the organelles (Fig. 21, Fig. 22, Fig. 23) and that the subcellular localization of this enzyme is dependent on the targeting signals at its N and C terminus (Fig. 24). These results indicate that αOGG1 is not exclusively localized in the nucleus, as it was previously proposed. For example, the work of Nishioka and colleagues of 1999, suggested that the presence of the NLS at the C terminus of the glycosylase was an impediment for the mitochondrial localization of the protein (Nishioka et al., 1999). More recently, also the review of Furihata makes no mention of the possibility of αOGG1 to be imported in mitochondrial compartment (Furhiata, 2015). On the contrary, my results showed how both the

- 119 - targeting signals are functional and necessary for the subcellular targeting of this enzyme to nuclear and mitochondrial compartments.

Mitochondrial localization of proteins: The vast majority of mitochondrial proteins are Dual import (in nuclear and encoded by nuclear genes, synthesized by ribosomes in the mitochondrial compartments) cytosol and subsequently imported into the organelle Alternative splicing Alternative transcription start site (Chainsaw et al., 2009). In particular, all the factors Alternative translation initiation dedicated to the maintenance and expression of mitochondrial DNA (mtDNA) are encoded from nuclear genome and so it is for BER proteins, among which OGG1. Even when the cells (human) are depleted of their mitochondrial DNA with ethidium bromate the enzyme that participate to BER are imported into these organelles (Stuart et al., 2004). The import is frequently dependent on an amino (N)-term amphipathic α helix, positively charged, which functions as a mitochondrial targeting signal (MTS) and that is recognized by the translocases of the outer mitochondrial membrane. Programs such as Mitoprot and Predator predict a very high probability of mitochondrial localization of αOGG1 (close to 90%), indicating that the residues from 9 to 26 at the common N-term of OGG1 proteins form a canonical MTS signal. My analysis show that this sequence is perfectly functional and sufficient to ensure the import of the enzyme into the organelle. The MTS may be processed at residue 23 (Trp; W) after being translocated into mitochondria. However, no evidence for this cleavage arose from my analysis (Fig. 20). Like for other enzymes involved in BER, such as LigIII or FEN1, OGG1 mRNAs possess potential alternative translation initiation (ATI) sites (Simsek et al., 2011; Kazak et al., 2013). It has been clearly established that alternative start AUG codons within a single transcript largely contribute to the diversity of the proteomes, especially to the mitochondrial one (Kazak et al., 2013). The presence of 2 AUGs in frame in OGG1 mRNA is a feature highly conserved among mammals (Fig. 23 and figure S1 of the manuscript at the appendix of this thesis). Interestingly, an initiation of translation at the second AUG would result in a protein with a truncated MTS, with the loss of two of the three arginine residues present in this sequence and necessary for the formation of the amphipathic helix. According to the prediction softwares the translation at this second AUG would result in a protein with reduced probability to be localized inside mitochondria, in agreement with the results I have obtained (Fig. 24). At this point, my analysis does not allow to determine if both of these initiation sites are used for the translation of two different proteins in the cell. In fact, we did not observe any significant shift in size between the nuclear and mitochondrial OGG1 (Fig. 20).

My microscopy analysis on αOGG1(WT)-FLAG showed that in almost the 30% of the transfected cells the protein is exclusively nuclear (Fig. 24B). To explain this observation our first hypothesis was to evaluate if there was a modulation of the cell cycle on the import of the enzyme. However, my experiments clearly indicated that the presence of the protein in mitochondrial compartment is independent from cell cycle. One possible explanation to this intriguing result may rely on a regulation in the choice of the translation initiation AUG. Based on its sequence context, the first AUG is likely to be a relatively weak start site thus facilitating ribosomal sliding (leaky mRNA scanning) (Kozak et al., 1996) that could produce two proteins from a single mRNA species. We cannot exclude that this second AUG could be used in particular conditions to favor nuclear import, providing like that the cells with an elegant mechanism to ensure proper relative levels of this DNA glycosylase in the two compartments.

The results I obtained by confocal microscopy showed for the first time that αOGG1-FLAG is not homogeneously distributed inside mitochondria but forms distinct foci, the same distribution that has been observed for the mitochondrial nucleoids where mtDNA is packed in close association with the IMM (Alam et al., 2003; Legros et al., 2004; Kukat et al., 2011). With the exception of the AP endonuclease, the enzymes catalysing mitochondrial BER are all associated with the particulate, insoluble fraction of the IMM (Stuart et al., 2005). My microscopy analysis (Fig. 26) showed that αOGG1 is actually a nucleoid-associated protein. By incubating purified mitochondria with increasing concentrations of NaCl, we could observe that the association of OGG1 with an insoluble fraction (probably mitochondrial membranes) resists an extraction with up to 500 mM of NaCl while TFAM is already solubilized with this conditions (Figure 2 of the article in the appendix). It is plausible that this distribution of BER proteins facilitates the efficiency of the mtDNA repair process (Boesch et al., 2009). The proteins that compose mitochondrial nucleoids have different functions, from DNA packaging and transcription to signalling functions. In the macromolecular structure of the nucleoid is possible to distinguish a core domain of key mtDNA-associated proteins, and a peripheral zone, consisting of factors which are less constitutively bound. Independently to the fact of being or not core proteins, some components such as Pol, mtSSB or Twinkle have been found to associate with mtDNA only during replication (Lewis et al., 2016), while my observations have also demonstrated that mtDNA replication state has no influence on the sub organelle localization of αOGG1 (Fig. 27), suggesting that its presence may be constitutive.

- 121 -

Evaluation of mitochondrial physiology in the absence of OGG1: In the present study, I analyzed the role of αOGG1 in the maintenance of mitochondrial physiology. mtDNA is proposed to be more susceptible to oxidative stress than the nuclear genome (nDNA) (Van Houten, 1997). Among all the types of mtDNA damage, oxidative damage is the most prevalent and, by far, the best studied. 8-oxoG is the most extensively studied DNA lesion resulting from oxidative damage. This alteration is highly susceptible to oxidation and can be further oxidized to yield several mutagenic base lesions, including spiroiminodihydantoin and guanidinohydantoin. When paired to cytosine, 8- oxoG is removed from DNA by OGG1 that initiates the evolutionarily conserved DNA repair mechanism Base excision repair (BER) process.

During my thesis, I have demonstrated that exposure of cells deficient for OGG1 to oxidizing agents results in a strong mitochondrial dysfunction. Indeed, in the absence of OGG1, cells show an increase in mitochondrial ROS production, a higher loss in mitochondrial membrane potential and a higher fragmentation of the network (Fig.28, Fig. 33, Fig. 34). It is also indeed clear that the overexpression of this enzyme in the organelles results in a protective effect (Fig. 31, Fig 33, Fig. 34), which is in line with findings of other authors (Torres-Gonzalez et al., 2014; Rachek et al., 2002; Chouteau et al., 2011). In fact, as previously reported, overexpression of a recombinant version of OGG1, specifically targeted to mitochondria by an additional MTS (MTS-OGG1) from mtSOD, protects the cells from an oxidative stress, likely due to an increased efficiency in the repair of 8- oxoG in mtDNA (Torres-Gonzalez et al., 2014; Rachek et al., 2002). Targeting human OGG1 to mitochondria potentially enhances mtDNA repair and cellular survival after oxidative stress in HeLa cells and attenuates or prevents mitochondrial dysfunction and apoptosis (Dobson et al., 2000). Moreover, Druzhyna and colleagues, have shown that the overexpression of OGG1 in mitochondria reduced the release of cyt c from the IMM and the activation of caspase 9 in oligodendrocytes after exposure to menadione (Druzhyna et al., 2003). My hypothesis was that accumulation of oxidative mtDNA damage contributes (or it is the cause) to the The 8-oxoG does not block polymerase progression. Formamidopyrimidine DNA-glycosylase (FPG) protein dysfunction of the organelles. Here, my analysis is specific for oxidized purines and possess N- showed that menadione acts as a mtDNA-damaging glycosylase and AP-lyase activities and thus generates agent (Fig. 37, Fig. 38). This quinone is metabolized a strand break after excision of 8oxoG residues from DNA. For this reason, DNA digestion with FPG (before by a 1-electron reducing enzyme, such for example performing the QPCR reaction) makes it more effective mitochondrial NADH-ubiquinone oxidoreductase in inhibiting the PCR and, like that, in evaluating the (complex I), and generates intracellular superoxide induction of DNA oxidative damage. and an unstable semiquinone radical that in turn

increases mitochondrial and intracellular ROS. The QPCR-based method I have optimized allowed us to measure a greater amount of oxidized residues within mtDNA of OGG1 silenced cells if compared with the control population, suggesting that the lack of 8-oxoG repair may be at the bases of the observed organelle dysfunctions.

How to preserve mitochondrial physiology after oxidative stress: As I have already said before, several groups suggest a possible role of αOGG1 in mitochondrial physiology maintenance after oxidative stress, (Dobson et al., 2000; Rachek, 2002; Druzhyna et al., 2003 and 2005). However, in all the studies presented so far, the approaches used for targeting the enzyme to the mitochondrial compartment do not allow a clear separation of the nuclear and mitochondrial functions of OGG1 because, albeit the presence of the additional MTS, an excess of the enzyme was also present in the nuclei of transfected cells. The main goal of my project was to elucidate if the specific loss of 8-oxoG repair activity in mitochondria (but not in nuclear compartment) has an impact on the organelles’ functions. In this regard, we generated several stable U2OS cell lines expressing different isoforms of αOGG1-FLAG (listed in the table 1) bearing silent mutations in order to escape to the silencing by siRNAs. The main characteristic of those isoforms are summarized in the table below.

Localization Glycosylase activity Nuclear Mitochondrial αOGG1 ( WT)- FLAG X X X αOGG1 ( ΔMTS)- FLAG X X αOGG1 (K249Q)- FLAG X X

Table 1 The canonical MTS recognized by TOM complex at the N’ ter and the NLS at the C’ ter of the WT are both functional so the resulting protein is located in nuclear and mitochondrial compartment (OGG1(WT)-FLAG) while the partial absence of the MTS due to the translation at the 2nd AUG in frame of the gene sequence result in a protein exclusively located in nuclei (OGG1(MTS)-FLAG). On the other hand the point mutation in the catalytic site (OGG1(K249Q)-FLAG of the enzyme doesn’t affect the localization of the enzyme but abolish the its glycosylase activity.

We demonstrated that expression of αOGG1(WT)-FLAG can preserve the morphology of mitochondrial network and complement the mitochondrial dysfunction observed after oxidative

- 123 - stress in the absence of the endogenous enzyme. While in the work of Chouteau and colleagues (Chouteau et al., 2011) the authors speculated that the protective effects of this enzyme may be associated with some alternative activities of OGG1, such as the stabilization of mtDNA-protein interactions, our observations with K249Q mutant suggest that the mechanism of action of OGG1 is associated with increased mtDNA repair capacities. While overexpression of the WT αOGG1 protects the mitochondrial network from fragmentation after exposure to oxidative stress, the mutant K249Q (like the ΔMTS) failed. These results indicate that the presence of an enzymatically active αOGG1 inside the organelles is necessary for the maintenance of mitochondrial functions under oxidative stress conditions. That said, we are still at the beginning of understanding by which molecular mechanism OGG1 absence impacts on mitochondrial physiology.

Oxidative damage of mtDNA and the consequences of the accumulation of 8-oxoG: It has been postulated that the accumulation of 8-oxoG in mtDNA above a threshold level could lead to the accumulation of deletions and, by the end, the loss of the entire mtDNA molecule (Rachek et al., 2002; Torres-Gonzalez et al., 2014). However it is still unclear how high levels of the 8-oxoG are linked to mitochondrial genome loss.

The 8-oxoG is considered as a pre-mutagenic lesion because of its capacity of pairing with adenine. However, Shokolenko and collegues have demonstrated that oxidative stress did not result in a significant increase in the rate of mtDNA mutagenesis (Shokolenko et al. 2009). Instead, DNA lesions such strand breaks (both single, SSB, or double strands, DSB) or abasic sites, prevailed over premutagenic base modifications by a factor of 10. Consistent with these observations is the work of Suter and Richter where they show that the 8-oxodG content in circular mtDNA is low and does not increase in response to oxidative insult. Removing instead of mutating: The high rate of polymerase However, they observed a high 8-oxoG blocking lesions induced in mtDNA by ROS suggests a content in mtDNA fragments corresponding to mechanism by which mitochondria may maintain the integrity of their genetic information.The hypothesis is consistent with degradation products (Suter & Richter, 1999). the observations of Yakes and Van Houten (Yakes & Van Houten, The accumulation of linear mtDNA preceeds 1997), who found that oxidative stress promoted a higher the elimination of damaged mitochondrial incidence of polymerase-blocking strand breaks and abasic sites in mtDNA than in nDNA. Therefore, degradation of severely genome (Alexeyev et al., 2013). damaged mtDNA emerges as a unique, mitochondria-specific mechanism for the maintenance of DNA integrity. These observations could be explained by the process of 8-oxoG elimination: in the

absence of OGG1, resulting in the accumulation of 8-oxoG in mitochondrial DNA, we could hypothesize a role of MYH in the generation of abasic sites after the elimination of the adenine paired with the 8-oxoG. In this case we would have an accumulation of fragmented mtDNA more susceptible to exonucleolytic degradation, and possessing high rates of 8-oxoG (Sheng et al., 2012).

On the other hand, if we assume that oxidative damage on mtDNA does not lead to an increase of 8-oxoG, then we may suppose that there should not be differences in presence or absence of the glycosylase activity of OGG1. However, my results clearly show that oxidative stress induces a more dramatic mitochondrial dysfunction in OGG1 deficient cells. Some of my preliminary results hint at a mechanism by which OGG1 absence (and so the increase in 8-oxoG levels) impacts on mitochondrial physiology through the regulation of mtDNA transcription. My preliminary results suggest that after exposure to menadione, OGG1 deficient cells present reduced level of some mitochondrial transcripts, of about the 50% for ND1 (Fig. 40). Recent studies have reported a loss of mtDNA in response to oxidative stress (Rothfuss et al. 2010; Furda et al. 2012). However, in my experiments I could not observe a significant loss of the molecule 30 minutes after menadione exposure (Fig. 40B). Thus, this loss of transcription previous to the loss of the mtDNA molecule, suggests a decreased rate of mitochondrial transcription in the presence of increasing amount of 8- oxoG in mtDNA. Because to start replication of mtDNA the very same transcription machinery is needed, we may suppose a consequent decrease of replication of the molecule and by the end a loss of mtDNA itself. The mechanism by which the accumulation in mtDNA of 8-oxoG, normally considered to be a non-blocking lesion, results in a block of transcription remains to be explored. Concerning the analysis on mtDNA transcription, my experiments showed a certain variability between the different transcripts analyzed (Fig. 40C). It is well established that the half-lives of mitochondrial mRNAs determine the levels of those molecules (Chujo et al., 2012). COX3 mRNAs are strongly stabilized by their interaction with complexes that suppress 3’deadenylation and/or 3’exonucleolytic degradation (Chujo et al., 2012). Maybe due to this reason, COX3 mRNA half-life is almost the double of the ND1 one (138’±10’ against 74’±2’ evaluated in HeLa cells, Chujo et al., 2012). This evidence could explain the weaker effect observed in my qRT-PCR experiments with the probe specific for COX3 (Fig. 40C) (which anyway presents a similar tendency to ND1 even if it is not significant). In order to rule out the bias induced by the half-lives of the mRNAs we used a probe directed against an intergenic region (tRNA-1, Fig 39C) and we could not observe any difference induced by menadione treatment on the levels of this RNA. However, this RNA is encoded by the light strand and we can not rule out that his may induce a difference because promoters HSP and

- 125 -

LSP are known to be functionally independent. In my opinion, this experiment should be repeated by using TaqMan-probes that recognize intergenic regions transcribed on the H-strand.

With this method, we could proceed to the analysis on cells depleted of the endogenous OGG1 and complemented with our FLAG-tagged constructs in order to verify the importance of the glycosylase activity on mtDNA transcription.

Mitochondria are extremely dynamic and fuse and divide constantly. Their morphological changes reflect different cell cycle phases, cellular metabolic demands but also cell damage. Various insults can induce mitochondrial fragmentation, which can be considered as a common stress response that allows to segregate and eliminate dysfunctional organelles from the network through mitophagy.

Oxidative treatments like exposure to menadione have a strong impact on mitochondrial morphology, with a transition from a filamentous and branched conformation to a fragmented and round one. Menadione treatment is also known to induce an accumulation of 8-oxoG in DNA of both nuclear and mitochondrial genomes (Oka et al., 2008). My analysis on mitochondrial morphology presented in this work show that overexpression of WT α-OGG1 reduces mitochondrial fragmentation induced by oxidative stress. Mitochondrial fragmentation is known to promote the opening of the mitochondrial transition pore with the subsequent collapse of the membrane potential, release of cyt c, degradation of the organelles and, by the end, apoptosis (Wu et al., 2011; Turner et al., 1998). Accumulation of oxidative damage in mtDNA may trigger those adverse effects (Jiang et al., 2013; Gredilla et al., 2010; Hiona et al., 2010).

We still did not decipher the molecular mechanisms that could explain why the overexpression of OGG1(K249Q) results in a loss of the protein from the mitochondria and the subsequent loss of the organelle. One fascinating hypothesis is that after the exposure to menadione, like what happens for the WT, also the OGG1(K249Q) mutant would be recruited to the DNA containing 8-oxoG but it will remain stably bound because of its capacity to recognize the lesion but its absence of enzymatic activity. On one hand, the impossibility of removing the 8-oxoG from mtDNA may cause the transcription block discussed above, the mtDNA loss and so, also the loss of the enzyme. However, on the other hand, the enzyme stuck on the molecule may block itself the transcription/replication processes. It has been proposed that the association of TFAM with mtDNA

can prevent BER proteins from freely accessing the DNA (Canugovi et al., 2011). This model could then be thought the other way round: the impossibility of the OGG1(K249Q) to dissociate from the double strand may block processes like transcription and/or replication that are necessary for mtDNA maintenance and the organelle physiology ending up with the removal of the organelle itself.

The domino effects – from mitophagy to apoptosis: As I have extensively discussed in the introduction, the loss of mitochondrial membrane potential is often used as an early marker of apoptosis and our studies suggest, in agreement with previous ones, that the absence of glycosylase activity in mitochondrial compartment, and therefore the eventual accumulation of oxidative damage to the mtDNA, could induce cell death. In the preliminary results, I have shown a strong increase of cleaved caspase-3 in cells depleted of OGG1 under oxidative treatment, when compared with the ones in which the enzyme is still present. The activation of the caspase-3 pathway is a hallmark of apoptosis and can be used an assay to quantify the cascade of events that leads to PCD. We have shown by real-time imaging, that the loss of the mitochondrial membrane potential of OGG1-deficient cells is followed by mitochondrial network loss and cell death. As future perspective, in order to determine whether the defect of OGG1 mitochondrial activity is the cause of cell death (and also understand by which mechanism the cell death is induced), the cells complemented with the different OGG1 variants, localized or not in the mitochondria, will be characterized.

Altogether, the data I have collected suggest that the higher mitochondrial dysfunction trigger cell death in OGG1 deficient human cells after exposure to menadione.

- 127 -

Conclusions and perspective: The bibliography concerning the role of OGG1 in mitochondria is still contradictory and, in my opinion, with our paper and with my thesis we unveil some very important points for the comprehension of BER in those organelles. The results obtained so far show that a catalytically active αOGG1 in mitochondria is essential to maintain mitochondrial functions and cell survival after induction of oxidative stress.

To be able to collect more information on the molecular mechanism/s at the basis of the organelles’ dysfunctions in the absence of the enzyme, we are planning to use two different approaches:

1. The use of other OGG1 mutants 2. The use of inhibitors against OGG1 glycosylase activity

1. Several mutants could be an alternative to the K249Q to understand the importance of the catalytic activity of the enzyme: An increase in 8-oxoG in DNA was reported for a model human leukemia cell line, KG-1, which was found to have a homozygous mutation in the OGG1 gene that results in the substitution of arginine 229 for glutamine (R229Q) (Hyun et al., 2000; Hill & Evans 2007). This amino acid substitution in OGG1 has been shown to alter the stability of the protein. While αOGG1(R229Q) has a glycosylase activity similar to that of the wild-type enzyme at low temperatures, it is inactivated at physiological temperature both in vivo and in vitro. Interestingly, the expression of αOGG1(R229Q) sensitizes KG-1 cells to killing by menadione. Already Chatterjee et al. demonstrated that mitochondrially targeted mutant αOGG1(R229Q) resulted in more cell death than nuclear-targeted mutant OGG1 upon exposure of cells to oxidative damage. Those results suggested that deficiencies of mitochondrial OGG1 (and therefore of the BER pathway) lead to reduced mitochondrial DNA integrity, resulting in decreased cell viability (Chatterjee et al., 2006). Very interestingly, other mutations of OGG1 have been identified in other species resulting in a thermosensitive enzyme; for example SAMP1 mice, which exhibit premature senescence, shortened lifespan, and elevated levels of 8-oxoG in DNA, have a homozygous R304W mutation in OGG1, which results in a thermolabile OGG1 protein that is devoid of enzymatic activity (Choi et al., 1999). However the capacity of this mutant to bind and excise the 8-oxoG are still lacking. For this reason, for our next experiments, we could use also other good characterized OGG1 mutants

such as H270L and F319A (Van der Kemp et al., 2004; Hashiguchi et al., 2004). Those two mutants present a reduced glycosylase activity (from 50 to 1000-fold) compared to the wild type. Band-shift analyses have shown that residues H270 and F319 are critical for the recognition of 8-oxoG. Mutation of those residues in OGG1 results in a protein without 8- oxoG binding or glycosylase activity (Van der Kemp et al., 2004). The use of one of those mutants would allow us to better understand if the deleterious effects we observed in cells expressing this mutant protein are due to the strong binding of α-OGG1(K249Q) to the mtDNA molecule containing 8-oxoG or to the loss of the catalytic activity and thus an effect of 8-oxoG itself.

2. In the last years, increasing evidence shows that The two classes of DNA glycosylase inhibitors are: DNA containing (i) AP site cancer cells are more sensitive to oxidative stress when analogues, which bind DNA N-glycosylases because these enzymes are end product compared to normal ones (Gad et al., 2014; Huber & Salah, inhibited and (ii) nucleotides with a stabilized glycosidic bond, which fail to be 2014). Because many authors have already suggested that processed by DNA N-glycosylases

a defect in the repair of 8-oxoG in mitochondria could lead Scharer & Jiricny, 2001 to cell death through apoptosis, we could then hypothesize that the specific inhibition of the glycosylase activity of OGG1 could offer a rationale for therapeutic strategies (Helleday et al., 2008; Gavande et al., 2016). Substrate analogues were the first synthetized inhibitors of human DNA glycosylases. These non-cleavable DNA substrates analogues have been initially used to elucidate the structural mechanisms for the lesion recognition and processing of the enzyme (Scharer & Jiricny, 2001). Morland and collegues showed that free 8-oxoG also inhibited the glycosylase activity most probably by being retained in the recognition site following N-glycosylic cleavage and inhibiting OGG1 turnover (Morland et al., 2005). Very recently some OGG1 inhibitors have been developed and they show promising specificity in in vitro studies (Srinivasan et al., 2012). Then, one fancy perspective of this work would be the evaluation of the effect of these compounds on cells exposed to oxidative stress or not. However, Srinivasan’s work remains incomplete because little or no information is available concerning the bioactivity of those molecules in vivo (at least in cellulo). In fact, it is still not clear not only if these inhibitors can enter in mitochondria but also if in general they can penetrate the cell so, it is clear that several studies will be necessary to use of this approach (Ramdzan et al., 2017). In this regard, ongoing studies in our lab are aimed at characterizing the activity of those compounds in U2OS cells.

- 129 -

In conclusion, delineating the multiple factors that underlie disorders that depend on mtDNA, mitophagy and apoptosis deregulation, including neurodegenerative (like Parkinson disease) disorders, aging process or cancer is essential.

Oxidative damage accumulates over time, especially in mtDNA of the human brain. It is proposed to play a critical role in aging and in the pathogenesis of several neurological disorders including Alzheimer's disease, amyotrophic lateral sclerosis and Parkinson's disease (Mecocci et al., 1993). A common pathological feature of these disorders is the loss of a subset of neurons in the central nervous system even if the causes of cell death are not well stablished. Of course, ROS may contribute to neuronal death by damaging lipids, proteins, and DNA, consequently BER that prevents DNA oxidative damage is critical to the maintenance of the genome and to the expression of functional, active proteins. The substantia nigra is located in the midbrain and it is critical in the development of many diseases and syndromes, including Parkinson's disease. Very interestingly, deletions in mtDNA have been found to accumulate with age and appear most commonly in the substantia nigra. Such deletions are found in patients with Parkinson’s disease more frequently than in age-matched controls (Bender et al., 2006; Schapira et al., 2008), thus, accumulation of mitochondrial DNA damage could contribute to mitochondrial dysfunction observed in Parkinson’s disease. There are at least two possible explanations for this: an excessive mitochondrial stress which may occur following exposure to environmental toxins, or defective mtDNA replication, or the failure to clear normally occurring damaged mitochondria by mitophagy (as is suggested to occur in patients with mutations in PINK1 and PARK2). For this reason, more precise and quantitative assays to evaluate mitophagy are needed. The mitochondrial dysfunction induced by 8-oxoG accumulation may cause an increase of mitophagy; we are currently setting up protocols to determine whether the absence of a functional glycosylase in the organelles mediates an increase in the mitochondrial turnover rate. The increase of risk for several cancers is associated with mutations and polymorphisms of OGG1; in fact, different variant forms of the glycosylase showing decreased 8-oxoG repair capacity have been associated with several cancer types. As I have already described in the introduction, mutations of the enzyme that result in a decreased glycosylase activity or that alter the subcellular localization, have been found in several cancer types (neck, lung, kidney…) (Weiss et al., 2005; Zhou et al., 2015). An invaluable strategy for devising new ways to treat and prevent these diseases will call for a consideration of the potential role of OGG1 in controlling mitochondrial turnover and a clarification of the tight link between 8-oxoG

accumulation in mtDNA, the resulting changes in mitochondrial metabolism and mitophagy during tumor formation (Ichim et al., 2015).

Finally, it is important to underline how limited and often contradictory is the knowledge about the role and the targeting of OGG1 to mitochondria. This project highlights the links between the base excision repair in mitochondria, the organelle physiological functions and, by extension, cellular survival. A better understanding of these mechanisms by which OGG1 acts in the mitochondrion to mitigate the effects of oxidative stress appears to have great importance for the generation of new potential therapeutic target.

- 131 -

Chapter V. APPENDIX

- 133 -

Detailed Protocols

Plating conditions for FACS analysis: Cell Model: U2OS (human osteosarcoma cells) Plate size: 6 well plate Number of cells per well: between 200.000 and 400.000 cells in 2000µL of complete medium (DMEM +/+) Reverse silencing For each well: 5µL Lipo RNAmix 60µL siRNA (2µM) 935µL Opti-MEM medium -/- Prepare the mix siRNA+lipofectamine and allow the formation of the complex for 15min at RT; after this time aliquot the 1000µL of the mix in the plate and add the cells (3000µl final volume) It isn’t necessary to change the medium After 72h the silencing is obtained

Plating conditions for Clariostar analysis: Cell Model: U2OS (human osteosarcoma cells) Plate size: 96 well plate Number of cells per well: between 7’500 and 10’000 cells in 200µL of complete medium (DMEM +/+) Reverse silencing For each well: 0.25µL Lipo RNAmix 3µL siRNA (2µM) 46.75µL Opti-MEM medium -/- Prepare the mix siRNA+lipofectamine and allow the formation of the complex for 15min RT; after this time aliquot the 50µL of the mix in the plate and add the cells (200µ final volume) It isn’t necessary to change the medium After 72h the silencing is obtained

Mitochondrial membrane potential TMRE-FACS analysis: Prepare PBS-BSA 0.4%, warm Treat the cells with 50 µM of Menadione in DMEM-/- for 1h at 37°C in 3000µL

- 135 -

Wash 2X DMEM+/+ Incubate 30’ with 50 nM MitoTracker-Green at 37°C in complete medium Remove medium Wash 1X warm PBS with Calcium and Magnesium Wash 1X with 1000µL Trypsin Incubate the cells with 250-500µL Trypsin for 5’ at 37°C Inactivate the trypsin with complete medium Centrifuge (2600 rpm) the cells and collect the pellet Prepare PBS-BSA 0.4%-TMRE 50nM and stain the cells for 15’-37°C

Using a flow cytometer, TMRE was detected with the filter commonly used for phycoerythrin (PE) (eg. 575/26 nm). MitoTracker Green was detected with the filter commonly used for the fluorescein isothiocyanate (FITC) (eg. 488 nm). Relative fluorescence intensity was used as measurement of mitochondrial membrane potential.

TMRE-Clariostar analysis:

Remove the medium and incubate with 100nM TMRE for 30’-37°C; Add together Hoechst 33342 1:1000 (working concentration 1µg/mL, stock concentration 1mg/mL) or 50 nM MitoTracker-Green Wash 1X DMEM+/+ Treat the cells with 50 µM of Menadione in DMEM-/- for 1h at 37°C in 200µL Wash once with Medium 199 -/- (withouf phenol red) Replace 199-/- medium and start the acquisition Measure Kynetics during at least 90’ with a CLARIOSTAR® microplate reader (BMG LABTECH)

Mitochondrial ROS production Mitosox-FACS analysis: Treat the cells with 50 µM of Menadione in DMEM-/- for 1h at 37°C in 3000µL Incubate 10’ with MitoSOX Red (5 μM) at 37°C for 10 minutes in PBS, protect from light. Wash 2X DMEM+/+ Wash 1X warm PBS with Calcium and Magnesium Wash 1X with 1000µL Trypsin Incubate the cells with 250-500µL Trypsin for 5’ at 37°C Inactivate the trypsin with complete medium

Centrifuge (2600 rpm) the cells and concentrate the pellets in 500µL of medium

Using a flow cytometer, MitoSOX Red was excited at 488 nm and fluorescence emission at 575 was measured. Relative fluorescence intensity was used as measurement of mitochondrial superoxide production.

Mitosox-Clariostar analysis: Remove the medium and incubate with Menadione 25 or 50µM in DMEM-/- 1h, 37°C in 200µL Wash 1X DMEM+/+ Incubate MAXIMUM 10 MINUTES 37°C dark with Mitosox: Dissolve the contents (50 μg) of one vial of MitoSOX™ mitochondrial superoxide indicator in 13 μL of dimethylsulfoxide (DMSO) to make a 5 mM MitoSOX™ reagent stock solution. Dilute the 5 mM MitoSOX™ reagent stock solution (prepared above) in HBSS/Ca/Mg or suitable buffer to make a 5 μM MitoSOX™ reagent working solution. Add together Hoechst 33342 1:1000 (working concentration 1µg/mL, stock concentration 1mg/mL) to use it the very same day; Wash 1X Medium 199 -/- Replace 199-/- medium and start the acquisition Measure Kynetics during at least 90’ with a CLARIOSTAR® microplate reader (BMG LABTECH)

Immunofluorescence Protocol: Cell Model: U2OS (human osteosarcoma cells) Plate size: 4 well plate Number of cells per well: between 20’000 and 40’000 cells in 500µL of complete medium (DMEM +/+) on EtOH washed cover slides Transfection 1 day after plating: Lipofectamine® 2000 Transfection Reagent –ThermoFischer Opti-MEM® I Reduced Serum Medium - ThermoFischer

• Mix A: Lipo2000 1µL +24 µL Opti-MEM medium -/- • Mix B 2.5 µL αOgg1-FLAG (100ng) + 22.5 µL Opti-MEM medium -/- Mix directly the two mixes and incubates RT for 5’ Add drop by drop the mix to the cells Change the transfection medium after 4h MAXIMUM with complete one After 24h from transfection, incubate cells with MEN 50µM for 60min in 250µL of DMEM without FBS and PS 37°C Wash 2X with complete medium

- 137 -

Stain with 50µM of MitoTrackers (Red or Deep Red) for 30’ at 37°C in complete medium Wash 2X with complete medium Fix FA 2% 20’ 37°C Wash 1X PBS Permeabilize 500µL TritonX-100 0,1% 5min RT 1h 30’ in blocking solution 1h30’ at 37°C in the 1st Antibody diluted in blocking solution Wash 3X 5/10min with DPBS + 0.1% Triton X-100 Incubate for 1h at at 37°C in the 2nd Antibody diluted in blocking solution Wash 3X 10min with DPBS + 0.1% Triton X-100 DAPI staining: Dilute 1:1000 DAPI in PBS (initial concentration 1mg/ml). Consider 250µL per well 5’ RT protected from light Wash 1X PBS Recover the coverslides and dry gently on paper Use 7µL of DAKO Mounting Medium-Sigma per coverslide Image acquisition was performed with a Leica confocal microscope SP8 with a 63x or 40x oil immersion objectives and processed with ImageJ. Staining EdU: Click-iT® EdU Alexa Fluor® 647 Imaging Kit- Invitrogen

Cell Model: U2OS (human osteosarcoma cells) Plate size: 4 well plate Number of cells per well: 40’000 cells in 500µL of complete medium (DMEM +/+) on EtOH washed cover slides Transfection: Lipofectamine® 2000 Transfection Reagent –ThermoFischer Opti-MEM® I Reduced Serum Medium - ThermoFischer

• Mix A: Lipo2000 1µL +24 µL Opti-MEM medium -/- • Mix B 2.5 µL αOgg1-FLAG (100ng) + 0.75 µL MTS turquoise (100ng) + 21.75 µL Opti-MEM medium -/- Mix directly the two mixes and incubate RT for 5’ Add drop by drop the mix to the cells

Change the transfection medium after 4h MAXIMUM with complete one Edu Staning: Incubate cells with EdU 50µM for 60min (this allows also the staining of replicating nucleoids) Wash 2X DPBS-Ca2+ Wash 2X PBS-3% BSA Fix FA 2% 20’ 37°C Wash 1X PBS Wash 1X PBS-3% BSA Permeabilize 400µL TritonX-100 0.5% 20min RT Make 1X Click-iT® EdU buffer additive by diluting the 10X solution 1:10 in deionized water. Prepare this solution fresh and use the solution on the same day. Prepare Click-iT® Plus reaction cocktail according to the table below (consider a final volume of 200µL per well)

Wash twice with 500µL of 3% BSA in PBS. Remove the wash solution Add 250 µl of Click-iT® Plus reaction cocktail to each well. Rock the plate briefly to ensure even distribution of reaction cocktail Incubate the plate for 30 minutes at RT, protected from light Wash each well once with 0.5 mL of 3% BSA in PBS 1h 30’ in blocking solution 1h30’ at 37°C in the 1st Antibody diluted in blocking solution Wash 3X 5/10min with DPBS + 0.1% Triton X-100 Incubate for 1h at at 37°C in the 2nd Antibody diluted in blocking solution Wash 3X 10min with DPBS + 0.1% Triton X-100 DAPI staining: Dilute 1:1000 DAPI in PBS (initial concentration 1mg/ml). Consider 250µL per well 5’ RT protected from light Wash 1X PBS Recover the coverslides and dry gently on paper

- 139 -

Use 7µL of DAKO Mounting Medium-Sigma per coverslides

Solutions: Blocking Solution For 10ml: - 100 µL Triton X100 10% - 100 µL Normal horse serum 100% - 300mg BSA (3% final) - PBS 1x to 10ml Primary antibodies ~250µL each well FLAG Mouse 1:2000 (ref F3165) TFAM Rabbit 1:1500 (ref ab155240) Secondary antibodies ~250µl each well Alexa M488 1:1000 Alexa R594 1:1000

Quantitative PCR Ricchetti’s protocol: TOTAL GENOMIC DNA EXTRACTION Resuspend the cells pellet in 200 ul of (Proteinase K 0.2mg/ml , SDS 0.2%, EDTA 5mM, PBS1X) Incubation: 3 hours at 50°C Isopropanol precipitation: Add 20ul of Sodium Acetate 3M pH5.2 and 300ul of isopropanol Incubation 20 minutes at -20°C Spin down the sample at 16000g for 30 minutes at 4°C Remove the supernatant and resuspend the pellet with Ethanol 70% Spin down the sample at 16000g for 5 minutes at RT Resuspend the DNA pellet in pure H2O (usually 100ul) Quantify the total genomic DNA (and storage at -20°C) Prepare 10ng/µL dilutions of the samples

FPG treatment

Upper Final µl/50µl Reagent MIX concentration

10X NEB 5 µl 1X buffer 1

Fpg 10 µg/mL 12.5 µl 2.5 µg/mL

BSA 10 0.5 µl 0.1 mg/mL mg/mL

H2O 7.0 µl

TOTAL 25 µl

DNA for assay µl/50µl Total amount

DNA 10 ng/µl 25 µl 250 ng

TOTAL 25 µl

- Mix both Upper Reagent mix (25 µl) and DNA solution (25 µl) to have a final volume of 50 µl - Incubate for 6 hours, at 37°C - Inactivation of FPG: 60°C, 10 minutes

- 141 -

Primers: Long fragment (8.9 kbp) Mitolong_for 5’-TCT AAG CCT CCT TAT TCG AGC CGA-3’ Mitolong_rev 5’-TTT CAT CAT GCG GAG ATG TTG GAT GG-3’

Short fragment (221 bp) Mitoshort_for 5’-CCC CAC AAA CCC CAT TAC TAA ACC CA-3’ Mitoshort_rev 5’-TTT CAT CAT GCG GAG ATG TTG GAT GG-3’ (same as Mitolong rev) Reaction Mix: ul H2O 29.5 MgCl 25mM 1 Buf5x 10 dNTPs 2,5mM 4 PF 1 PR 1 DMSO 2% 1 Phusion 0.5 total 48

DNA 2 (2.5ng/µL or 3ng/µL) Short-mito Step Temperature Time Number of cycles Hotstart 94°C 7 min 1 Initial dentaturation 94°C 1 min 1 Denaturation 94°C 15 s Annealing 60°C 45 s 22 Extension 72°C 45 s Final extension 72°C 10 min 1

Long-mito Step Temperature Time Number of cycles Hotstart 94°C 7 min 1 Initial dentaturation 94°C 1 min 1 Denaturation 94°C 15 s 25 Annealing/ 64°C 11 min extension Final extension 72°C 10 min 1

Picogreen quantification of PCR sample Dilute DNA ladder to 100ng/µL in 100µL -20µL DNA 500ng/µL -80µL TE 1X Dilute to 20ng/µL in 200µL -40µL DNA/samples 100ngµL -160µL TE 1X Standard curve 10ng/µL 5ng/µL 2.5ng/µL 1.25ng/µL 20ng/µL 50µL 25µL 12.5µL 6.25µL TE1X 50µL 75µL 87.5µL 93.75µL

Vtot 100µL 100µL 100µL 100µL Protocol -add 90µL of TE1X buffer to each well that will be used (samples+standard) -add 10µL of samples or standard to each well -prepare the solution containing the picogreen: 5µL/mL of TE1X -add 100µL of picogreen solution to each well Incubate for 10’ RT in dark Shake the plate gently for 20’’ Fluorescent Excitation: 485nm Emission: 530nm

- 143 -

Quantitative RT-PCR DNA and RNA were extracted from 1exp6 (maximum) U2OS cells, treated or not, by using QIAamp DNA Mini Kit and RNeasy plus MINI Kit (Qiagen) respectively and following manufactures’ instructions. Nucleic acid concentration was quantified by Nanodrop Spectrophotometer ND-1000 (Thermo Scientific, Waltham, MA) and adjusted to 100ng/µL. 300ng of RNA were retrotrascribed by using the kit SuperScript® VILO cDNA Synthesis Kit and Master Mix (Thermofisher).

Reaction Mix for RT-PCR SuperScript™ VILO™ Master Mix 2µL SuperScript™ Reverse Transcriptase (RT) 1µL Nuclease-free Water 4µL Template RNA (100ng/µL) 3µL Final Volume for each reaction 10µL 10min 25°C//1h 42°C//5min Cycle 85°C//∞16°C

The resulting cDNA were diluted 1/20e in a final volume of 200µL of RNAse free H2O. Quantitative PCRs were performed with an ABI PRISM 7300 sequence detection system (Applied Biosystems) by using the TaqMan Universal PCR Master Mix (Applied Biosystems). Each amplification was performed in duplicate.

Reaction Mix for qPCR

iTaq superMix 10µL

TaqMan Probe 1µL

Nuclease-free Water 5µL

Template cDNA (diluted 1/20e) 5µL

Final Volume for each reaction ~20µL

1X (10min 95°C) - 40X (15 sec Cycle 95°C//1min 60°C)

The relative amount of mtDNA and mtRNA was calculated by using the comparative CT method. Nuclear probes used for the normalization were RLPO (large ribosomal protein gene) for the cDNA and ACTR2 (actin- related protein 2 homolog) for the DNA samples. ND1-COX3 were amplified by PCR primers purchased from Applied Biosystems (Hs02596873_s1 and Hs02596866_g1 respectvely).

Protocol for measuring mitochondrial length and branching in digital images with ImageJ software- For color images, convert to black and white For black and white images: a. Select ‘free hand’ tool (4th tool from right) b. Draw Region of Interest (ROI) around the cell’s mitochondria. Omit any over-exposed regions c. ‘Process’ i. ‘Filter’ 1. ‘Convolve’ using filter from Koopman et al., 2005 (cut and paste), normalize kernel

0 0 -1 -1 -1 0 0

0 -1 -1 -1 -1 -1 0

-1 -1 3 3 3 -1 -1

-1 -1 3 4 3 -1 -1

-1 -1 3 3 3 -1 -1

0 -1 -1 -1 -1 -1 0

0 0 -1 -1 -1 0 0

d. ‘Image’ i. ‘Adjust’ 1. ‘Threshold’ a. ‘apply’ e. Re-draw ROI f. ‘Analyse’ i. ‘Set Measurements’ 1. check ‘Area,’ ‘Circularity,’ ‘Perimeter,’ ‘Fit Elipse’ (check shape descriptors) g. ‘Analyse’ i. ‘Analyse Particles’ 1. ‘size’ = 15 – Infinity 2. ‘Circularity’ = 0 – 1.00 3. From dropdown menu, ‘show outlines’ 4. check ‘display results’ and ‘clear results’ h. plot AR Aspect ratio (inverse of circularity) and Form factor (branching)

- 145 -

Table 1- Antibody

Protein Company Reference Application Dilution 16064-1- MPPB ptglab Western Blot 1/1000 AP Lamin B abcam ab20396 Western Blot 1/1000 FLAG SIGMA F1804 Western Blot and IF 1/2000 Tubulin α Abgent AJ1034a Western Blot 1/1000 hOGG1-1 abcam ab124741 Western Blot 1/10,000 TOMM22 abcam ab10436 Western Blot 1/1000 TFAM abcam ab155240 Western Blot and IF 1/2000 CENPF abcam Ab5 IF 1:500 Cleaved Cell Signaling Caspase- 5A1E IF 1:400 Technology 3 Cell Signaling LC3B 27755 IF 1:250 Technology

Table 2- Plasmids Primer name Constrcut Primer sequence α-OGG1F-WT α-OGG1-WT α- 5’-TCG AAT TCG GGCGGTGCTGCTGTGGAAATG CCT GCC CGC (Forward) OGG1-DNLS GCG CTT C -3’ α-OGG1F-DMTS (Forward) α-OGG1-DMTS 5’- T5’- TCG AAT TCG GGCATGGGGCATCGTACTCTAG-3’ α-OGG1R-FLAG α-OGG1-WT α- 5’- G5’- GC GGA TCC CTA CTT ATC GTC GTC ATC CTT GTA ATC (reverse) OGG1-DMTS GGC CCT GCC TTC CGG CCC TTT GGA AC-3’

α-OGG1R-DNLS- 5’- GC GGA TCC CTA CTT ATC GTC GTC ATC CTT GTA ATC GGC FLAG (reverse) α-OGG1-DNLS CCT TGC TGG TGG CTC CTG AGC ATG-3’ α-OGG1F-G12E For Site Directed 5’- CTT CTG CCC AGG CGC ATG GAG CAT CGT ACT CTA GCC TC- (Forward) Mutagenesis 3’ α-OGG1R-G12E For Site Directed 5’- GAG GCT AGA GTA CGA TGC TCC ATG CGC CTG GGC AGA (Reverse) Mutagenesis AG-3’ α-OGG1F-K249Q For Site Directed 5'-GGA GTG GGC ACC CAG GTG GCT GAC TGC-3' (Forward) Mutagenesis α-OGG1R-K249Q For Site Directed 5'-GCA GTC AGC CAC CTG GGT GCC CAC TCC -3' (Reverse) Mutagenesis HsOGG1a-Mut silent mutations 5’ – CCAGGTGTGGACGTTGACTCAGACTGAGGAGCAG – 3’ (Forward) siRNA HsOGG1a-Mut silent mutations 5’ – ACGTCCACACCTGGTCAGCTAGTACACCACTCCAGTGTG – 3’ (Reverse) siRNA siRNA-OGG1 GGAUCAAGUAUGGACACUG siRNA-control QIAGEN siAllStars Neg (ref. xxx)

- 147 -

Submitted Article

- 149 -

Mitochondrial maintenance under oxidative stress depends on mitochondrial but not nuclear α isoform of OGG1

Running title: Role of α-OGG1 in human mitochondria

Debora Lia1,2,3, Aurelio Reyes4, Julliane Tamara Araújo de Melo1,5, Tristan Piolot6, Jan Baijer1,2,3, Juan Pablo Radicella1,2,3, Anna Campalans1,2,3

1 Institut de biologie François Jacob (IBFJ). Institute of Cellular and Molecular Radiobiology, CEA, UMR967 INSERM, F-96265 Fontenay aux Roses, France

2 Université Paris Diderot, France

3 Université Paris-Sud, France

4 MRC Mitochondrial Biology Unit. University of Cambridge. Cambridge CB2 0XY. United Kingdom

5 Laboratório de Biologia Molecular e Genômica, Departamento de Biologia Celular e Genética, Centro de Biociências, Universidade Federal do Rio Grande do Norte, Natal, RN, Brazil

6 Institut Curie, CNRS UMR3215, INSERM U934, 75248, Paris, France / Present address: UMR7241 CNRS, INSERM, College-de-France, 75005, Paris, France

Correspondence to [email protected]. ORCID iD 0000-0001-7089-6891

Key words (3-6): mitochondria, DNA repair, 8-oxoG, OGG1

Summary statement

α-OGG1 is imported into mitochondria where its glycosylase activity is essential for mitochondrial maintenance in human cells exposed to oxidative stress.

Abstract

Accumulation of 8-oxoG in mtDNA and mitochondrial dysfunction have been observed in cells deficient for the DNA glycosylase OGG1 exposed to oxidative stress. In human cells up to eight mRNAs for OGG1 have been described as generated by alternative splicing and it is still unclear which of them ensures the repair of 8oxoG in human mitochondria. Here, we show that the α-OGG1 isoform, considered up to now to be exclusively localized in the nucleus, has a functional mitochondrial targeting signal and is imported into the mitochondria. We analyzed the submitochondrial localization of α-OGG1 with unprecedented resolution and we show that the glycosylase is associated with DNA in nucleoids. This association does not depend on the cell cycle or the replication of mtDNA. Using α-OGG1 deficient human cells, we show that the presence of α- OGG1 inside mitochondria and its enzymatic activity are required to preserve mitochondrial network in cells exposed to oxidative stress. Altogether, these results evidence a new role of α-OGG1 in the mitochondria and indicate that the same isoform ensures the repair of 8oxoG in both nuclear and mitochondrial genomes and that its activity in mitochondria is sufficient for the recovery of the organelle function after oxidative stress.

- 151 -

Introduction

The mammalian mitochondrial genome consists of a circular double-stranded DNA molecule of approximately 16 kb in length and accounts for 1% to 2% of the total DNA in the cell. It encodes for only 13 proteins, all of them involved in oxidative phosphorylation and ATP synthesis, as well as for two rRNAs (12S rRNA and 16S rRNA) and 22 tRNAs that are required for mitochondrial protein synth esis. However, up to 1500 proteins encoded by the nuclear genome are found inside mitochondria, including proteins required for replication, transcription and repair of mitochondrial DNA (mtDNA). mtDNA is highly packed in DNA-protein assemblies constituting the mitochondrial nucleoids, that can be clearly observed by fluorescence microscopy as discrete foci distributed throughout the mitochondrial network (Alam et al., 2003; Legros et al., 2004). Nucleoids are anchored to the mitochondrial inner membrane and are composed of one or two DNA molecules (Kukat et al, 2011) and several core proteins involved in DNA packaging, transcription and signaling. In particular, TFAM (Mitochondrial Transcription Factor A), a key activator of mitochondrial transcription as well as a participant in mitochondrial genome replication, is a major component of the nucleoid and is required for the mtDNA packaging. Besides the core proteins, many other proteins transiently associate with mitochondrial nucleoids (Gilkerson et al, 2013).

Even though many proteins known for their role in nuclear DNA repair (BER, mismatch repair, single and double strand break repair) are imported into mitochondria (Kazak et al, 2012), what mechanisms participate in the maintenance of mtDNA is still an open question. Mitochondrial biogenesis and dynamics, including degradation of damaged mitochondria by mitophagy, also play essential roles in preserving the stability of the mitochondrial genome. The scenario becomes even more complex if we take into consideration that each cell contains a large number of mtDNA molecules, up to several thousands, depending on the cell type. Thanks to cycles of fusion and fission that promote mixing and homogenization of mitochondrial contents, cells maintain a certain level of heteroplasmy in which mutant and wt mtDNA molecules coexist (Mishra & Chan, 2014). For most of the mtDNA mutations described so far, no phenotypic alteration is detectable unless mutant mtDNA molecules exceed 60% of the total mtDNA (Gilkerson et al., 2008, Schon and Gilkerson, 2010). Considering that accumulation of mutations or deletions and the loss of mtDNA are involved in many different human diseases, a better understanding of the cellular processes involved in the maintenance of this molecule is of major importance.

Oxidative mtDNA damage has been reported to be more extensive and persistent than nuclear DNA damage (Yakes and Van Houten, 1997; Richter et al., 1988). It has been proposed that this higher sensitivity of mtDNA could be due to the absence of histones in mitochondria, facilitating the access of ROS to mtDNA. However, this idea is not supported by the tight packaging of mtDNA around the TFAM protein in nucleoids resulting in a highly compacted structure. Accumulation of mtDNA damage could also be due to the proximity of the mtDNA to the continuous ROS production by the mitochondrial electron transport chain. Persistence of oxidative damage in mtDNA could lead to the accumulation of deletions or mutations at the origin of mitochondrial dysfunction (Shigenaga et al., 1994). Mitochondrial DNA damage has been linked to loss of mitochondrial membrane potential, increased reactive oxygen species (ROS) generation and cell death (Santos et al., 2003).

One of the most frequent DNA alterations induced by ROS is 8-oxoG, a product of oxidation of guanine. Several lines of evidence suggest that accumulation of 8-oxoG in mtDNA contributes to the mitochondrial dysfunction observed during aging and in neurodegenerative disorders (Druzhyna et al., 2008). Three enzymes are involved in the avoidance of the accumulation of 8-oxoG and its mutagenic consequences. MTH1 detoxifies the pool of nucleotides by hydrolyzing the 8-oxo-2′-deoxyguanosine triphosphate (8-oxo- dGTP) to 8-oxo-dGMP, the DNA glycosylase OGG1 excises 8-oxoG paired with cytosine in DNA, and MUTYH (MYH), another DNA glycosylase, removes the Adenine paired to the 8-oxoG after replication, thus giving another opportunity to OGG1 for removing the 8-oxoG (Oka et al., 2014). These three enzymes are mostly known for their role in the nucleus but they are also targeted to mitochondria and their deficiency causes mtDNA loss and results in mitochondrial dysfunction. Indeed, Ogg1 knock-out mice accumulate high levels of 8-oxoG in mitochondria (Souza-Pinto et al., 2001). It has also been shown that MTH1 and OGG1 play essential roles in the protection of mtDNA during neurogenesis, and that Mth1/ Ogg1 double knock-out mice accumulate 8-oxoG and display reduced mitochondrial membrane potential (Leon et al., 2016). Moreover, exposure of OGG1-deficient cells to oxidative stress results in mitochondrial dysfunction, with reduction of mtDNA levels, decrease of mitochondrial membrane potential, increase of mitochondrial fragmentation and reduced levels of proteins encoded by mtDNA, possibly linked to accumulation of oxidative mtDNA damage (Wang et al., 2011).

OGG1 initiates the Base Excision Repair (BER) pathway by specifically recognising and excising 8- oxoG. In mitochondria, OGG1 DNA glycosylase activity is thought to be followed by that of other BER proteins: AP endonuclease APE1, DNA polymerase Pol-gamma and DNA Ligase III. All these BER proteins are encoded by the nuclear genome and are imported into the mitochondria (Kazak et al, 2012). Pol-gamma and Ligase III are not only involved in BER but are essential for mtDNA replication, and defects in those proteins result in mtDNA instability leading to dramatic mitochondrial and cellular dysfunctions (Simsek et al., 2011; Trifunovic et al., 2004; Zhang et al., 2011). While for many years, Pol-gamma has been considered to be the only polymerase present inside mitochondria, recent reports have shown the presence of Pol-beta inside the organelle (Sykora et al., 2017). These findings indicate that from the recognition of the lesion to the ligation step, the very same BER enzymes act in both the nucleus and the mitochondria to ensure the repair of oxidative DNA damage. According to the information present in the National Center for Biotechnology Information (NCBI) many different OGG1 mRNAs (and potentially proteins) can be generated from the OGG1 gene by alternative splicing in human cells (Takao, 1998; Nishioka et al., 1999; Ogawa et al., 2014; Furihata et al., 2016). Experimental evidence is still lacking concerning which isoforms are indeed translated in the cell. All potential isoforms (OGG1-1a, −1b, −1c, −2a, −2b, −2c, −2d and −2e) share the same N-terminal domain containing a Mitochondrial Targeting Sequence (MTS). The OGG1-1a (also named α-OGG1) is the only isoform to have also a Nuclear Localization Signal (NLS) and has therefore been defined as the nuclear isoform of OGG1. Which of the OGG1 isoforms is responsible for the repair of the 8-oxoG in mtDNA is still unclear. Despite the very high conservation of the DNA glycosylases between mouse and human, to date, only the isoform corresponding to the human α-OGG1 has been identified in mouse and therefore proposed to ensure the 8-oxoG glycosylase activity in both nucleus and mitochondria. The artificial targeting of human α-OGG1 to mitochondria, by a fusion to the MTS of MnSOD allows the complementation of mitochondrial dysfunction observed in OGG1 deficient cells, suggesting that this isoform can also be active in mitochondria and efficient in the removal of 8-oxoG (Dobson et al., 2000; Rachek, 2002).

- 153 -

In this study, using a combination of biochemical and imaging techniques, we show that α-OGG1 is normally imported into mitochondria and that both the MTS and the NLS signals are functional and sufficient to determine the subcellular localization of the enzyme. In agreement with the function of the protein in DNA repair, we describe the association of α-OGG1 with mtDNA in the mitochondrial nucleoids independently of the cell cycle phase and the mitochondrial DNA replication status. Furthermore, our results show that the presence and glycosylase activity of α-OGG1 inside mitochondria are essential to preserve mitochondrial function in human cells exposed to oxidative stress.

Results

α-OGG1 is imported into mitochondria where it associates with the inner mitochondrial membrane

Using a monoclonal antibody that specifically recognizes an epitope only present in the α-OGG1 isoform, we showed that this isoform was present in both nuclear and mitochondrial protein fractions (Fig. 1A). Likewise, a α-OGG1 tagged with the FLAG epitope was produced and an anti-FLAG antibodies was used to study the subcellular localization of the fusion protein. As for the endogenous protein, α-OGG1-FLAG was present in both nuclear and mitochondrial fraction (Fig. 1B). U2OS cells expressing α-OGG1-FLAG were stained with MitoTracker Red prior to fixation and immunostained with an anti-FLAG antibody. As shown in Fig.1C, α-OGG1-FLAG protein was present in both the nucleus and the mitochondria (Fig. 1C).

To determine α-OGG1 sub-mitochondrial localization, we performed protease digestion assays on isolated mitochondria. Mitochondria from U2OS cells were treated with increasing amounts of Proteinase K to degrade mitochondrial proteins facing cytoplasm. α-OGG1 was protected from proteinase K digestion indicating that the protein was located inside the mitochondria (Fig1D). To better define the localization of α- OGG1, sucrose-gradient purified mitochondria from HEK-293 cells were exposed, prior to trypsin, to hypotonic buffer or to digitonin to disrupt the outer mitochondrial membrane. After washing and pelleting mitochondria, the organelles were lysed with 1% SDS in isotonic buffer and analyzed by Western Blot. These experiments indicated that α-OGG1 was either localized in the inner mitochondrial membrane or in the mitochondrial matrix (Fig. 1E). Finally, association of α-OGG1 with mitochondrial membranes was studied by treating purified mitochondria with low or high salt concentrations. α-OGG1 remained insoluble even at high salt concentrations (Fig. 1F). Altogether, these results show that α-OGG1 is located both in the nucleus and in the mitochondria where it associates with the inner mitochondrial membrane.

α-OGG1 is associated with mtDNA in nucleoids

Confocal microscopy showed that α-OGG1-Flag was not homogeneously distributed inside mitochondria but accumulated at punctate foci that resembled the mitochondrial nucleoids. To determine if α- OGG1 was actually a nucleoid-associated protein, co-localisation of α-OGG1-FLAG with nucleoid markers, was investigated. Co-localization between α-OGG1 and mtDNA and TFAM was observed (Fig. 2A, B) with Pearson correlation coefficients of 0.75 and 0.8, respectively. The spatial organization of α-OGG1 and TFAM in mitochondrial nucleoids was further explored by using Structured Illumination Microscopy (3D-SIM). This approach revealed that α-OGG1 was clearly surrounding TFAM, a marker for the nucleoid core (Fig. 2C, D). The association of α-OGG1 with mitochondrial nucleoids was further confirmed by fractionation on an iodixanol gradient of detergent-solubilized mitochondria from HEK 293T cells. The analysis of the gradient fractions

revealed that a fraction of α-OGG1 co-purified with mtDNA (Fig. 2E, fractions 9-11). Collectively, these results indicate that α-OGG1 associates with mtDNA in the nucleoids.

α-OGG1 contains functional nuclear and mitochondrial targeting signals

In agreement with the presence of α-OGG1 in both nuclear and mitochondrial compartments, several softwares for the prediction of targeting signals determine a very strong probability for α-OGG1 to be imported into mitochondria and identify a canonical Mitochondrial Targeting Sequence (MTS) at the N-terminus of the protein. Interestingly, a second in frame AUG codon was found at the N-terminal part of the OGG1 messenger and shown to be highly conserved among mammals (Fig. 3A and Fig. S1). Intriguingly, an initiation of translation at the second AUG would result in a protein with a truncated MTS and thus would have a reduced probability to be localized inside mitochondria (Fig. 3B). Finally, a TOM20 recognition motif (RKYF) in position 97-100 of α-OGG1, suggested that α-OGG1 can be imported into the mitochondria by the TIM/TOM pathway (Schulz et al., 2015; Model et al., 2008).

To define the role of the MTS and NLS targeting signals, we expressed full length α-OGG1 or truncated versions, lacking either the MTS or the NLS, fused to the FLAG epitope and analyzed their subcellular localization (Fig. 3C). While, as mentioned above, the full length α-OGG1 was present in both mitochondrial and nuclear compartments, disruption of the MTS (OGG1-MTS), due to the initiation of translation at the second in frame AUG, resulted in a protein localized exclusively in the nucleus. Interestingly, substitution of the glycine in position 12 by a glutamic acid, a mutation identified in kidney cancer cells (Audebert et al, 2002), also impaired the mitochondrial import of the protein (Fig 3C). Deletion of the NLS gave rise to a protein mostly localized in the mitochondria (Fig. 3D, E). Taken together, these results indicate that mitochondrial and nuclear targeting sequences co-exist in α-OGG1 and that both signals are functional and can determine the subcellular localization of the enzyme.

Import of α-OGG1 to mitochondria is independent of the cell cycle or mtDNA replication

As mitochondrial import has been shown to be modulated through the cell cycle (Harbauer et al., 2014), we decided to explore if the subcellular distribution of α-OGG1 was cell cycle-dependent. The incorporation of EdU was used to identify cells in S phase and an antibody against the kinetochore protein CENP-F was used for cells in G2 phase. Cells in G1 are negative for both EdU and CENP-F. As shown in Fig. 4A, OGG1(WT)-Flag was detected in mitochondria independently of the cell cycle phase and the proportion of cells with mitochondrial staining remained unchanged throughout the cell cycle. We further investigated whether the mtDNA replication status influenced the sub-organelle localization of α-OGG1, as it is the case for several proteins involved in mtDNA maintenance (Rajala et al., 2014). EdU incorporation was used to visualize the mitochondrial nucleoids undergoing replication on transiently double transfected cells expressing α- OGG1(WT)-Flag and MTS-Turquoise2. An antibody against TFAM was used to label mitochondrial nucleoids (Fig. 4C, D). More than 500 nucleoids were analyzed to determine their replication state and the presence or absence of the glycosylase. As can be appreciated in Fig. 4C, α-OGG1-FLAG could be detected in both EdU positive and negative nucleoids, clearly indicating that the association of the glycosylase with the nucleoids is not dependent on DNA replication. Altogether these results show that the mitochondrial import of α-OGG1 and its association to mtDNA is not dependent on the cell cycle phase or on mtDNA replication status.

- 155 -

OGG1-deficiency is associated with mitochondrial dysfunction after an oxidative stress

Considering that the only known activity for OGG1 is the repair of 8-oxoG, we next studied the impact of OGG1 deficiency in the response of mitochondria to menadione-induced oxidative stress that results in the induction of 8-oxoG in mtDNA (Oka et al., 2008). Transfection of U2OS cells with OGG1 directed siRNA resulted in more than 90% decrease of both transcript and protein levels of α-OGG1 72 h after transfection (Fig. 5A). Cells transfected with siRNA against OGG1 or control siRNA were exposed to menadione and mitochondrial parameters were evaluated. After menadione treatment, we found a higher increase in mitochondrial superoxide production in OGG1-deficient cells when compared to controls (Fig. 5B,C). In addition, the use of TMRE probe to monitor mitochondrial membrane potential revealed that menadione treatment induced a higher loss of mitochondrial membrane potential in OGG1 silenced cells, while no difference was observed between the two cell populations in basal conditions (Fig. 5D, E). The loss of mitochondrial activity observed was not due to a loss in mitochondrial mass as MitoTracker Green staining was not affected by the treatment.

To monitor single cell responses over time, we performed real time microscopy. Cells transfected with a siRNA against OGG1 or with a control siRNA, were treated with 50 µM of menadione for one hour, washed and imaged for 2 hours at a frame rate of one image every two minutes. Most of the mitochondria from cells transfected with a control siRNA showed a stable TMRE staining over time after menadione treatment, indicating the maintenance of a positive membrane potential. In contrast, approximately 40% of OGG1- deficient cells showed a progressive and fast loss of membrane potential leading to the inactivation of the entire mitochondrial network (Fig. 5F). These results revealed that human OGG1 deficient cells show a mitochondrial dysfunction when exposed to oxidative stress, with a major production of mitochondrial superoxide and loss of mitochondrial membrane potential.

α-OGG1 mitochondrial localization and enzymatic activity are required to prevent mitochondrial dysfunction in cells exposed to oxidative stress.

To determine if -OGG1 could rescue the mitochondrial defects observed in OGG1 deficient cells, we established U2OS cell lines stably expressing α-OGG1(WT)-Flag or the exclusively nuclear α-OGG1(MTS)- Flag. The deletion of 10 aminoacids at the N-terminus of OGG1 didn’t affect the enzymatic activity of the enzyme, as expression of α-OGG1(MTS)-FLAG resulted in the same level of 8-oxoG DNA glycosylase activity in total cell extracts than that of extracts from cells expressing the full-length protein (Fig. 6A). In order to specifically reduce the levels of the endogenous OGG1, silent mutations were introduced in the sequence of the ectopically expressed proteins to make them resistant to the siRNA against OGG1 (Fig. 6B).

Wild-type U2OS cells or cells stably overexpressing the different isoforms of the protein were transfected with the siRNA against OGG1. Three days after transfection, the cells were stained with TMRE or MitoSOX, treated with menadione, and analyzed. The oxidative stress induced a significantly higher mitochondrial ROS production along with higher loss of mitochondrial membrane potential in OGG1 deficient cells compared to cells transfected with the control siRNAs. Both parameters were rescued by expression of a WT version of α-OGG1 but not by the expression of the mutant isoform (Fig. 6C,D), indicating that the

presence of the α-OGG1 DNA glycosylase in the mitochondria is required for the recovery of the organelle from the oxidative stress.

Prominent fragmentation of the mitochondrial network was observed in U2OS cells after exposure to menadione. This fragmentation was characterized by a rapid transition from a filamentous and branched network to a fragmented one, with isolated, round-shaped mitochondria (Fig. 7A). To evaluate if overexpression of α-OGG1 in mitochondria could suppress the fragmentation of this organelle after menadione treatment, U2OS cells were transiently transfected with constructs expressing α-OGG1(WT)-FLAG, α- OGG1(MTS)-FLAG or the catalytically inactive mutant α-OGG1(K249Q)-FLAG (Fig. 7A). The latter mutant displayed the same subcellular distribution as the WT protein, and was localized both in the nucleus and in mitochondria (Fig. S2). The percentage of cells with a preserved mitochondrial network was quantified in basal conditions and after exposure to oxidative stress (Fig. 7B). Only the overexpression of the α-OGG1(WT)-FLAG protein resulted in a normal mitochondrial network in cells exposed to menadione. Further analysis of mitochondrial morphology parameters on binarized images corroborated that after menadione treatment mitochondrial length and degree of branching were comparable in non-transfected and α-OGG1(MTS)-FLAG or α-OGG1(K249Q)-FLAG overexpressing cells, while in α-OGG1(WT)-FLAG overexpressing cells the network was better preserved (Fig.7C,D).

Taken together, these results indicate that mitochondrial α-OGG1 is essential to preserve mitochondrial morphology and function after exogenous oxidative stress. In addition, because the catalytically inactive mutant α-OGG1(K249Q)-FLAG displays the same subcellular localization than the WT enzyme and co-localizes with mitochondrial nucleoids, our data strongly suggest that the DNA glycosylase activity and thus removal of 8-oxoG is required to protect mitochondrial physiology from oxidative stress.

Discussion

To be or not to be in the mitochondria

Despite clear evidence indicating the presence of a functional BER pathway inside human mitochondria, which isoform of the DNA glycosylase OGG1 is responsible for the recognition and excision of 8-oxoG in the mtDNA remained unclear. Although it was proposed that this activity could be performed by β- OGG1 (Oka et al., 2008), the lack of DNA glycosylase activity of this isoform renders its involvement in the repair of 8-oxoG unlikely (Hashiguchi et al., 2005). It was recently reported that in vitro purified OGG1 isoforms 1b and 1c present some enzymatic activity and could be responsible for the 8-oxoG removal from the mitochondrial genome (Furihata et al., 2015). However, there is no in vivo evidence for the existence of these OGG1 variants. It is not well known which transcripts are generated in different tissues and/or different physiological conditions and no experimental evidence determining which protein forms are indeed produced is available. In mouse, only a single transcript corresponding to the isoform α-OGG1 (OGG1-1a) has been identified and the protein has been shown to be functional in both nuclear and mitochondrial compartments. It is still unclear why so many OGG1 transcripts are found in human cells. Because several of the potential polypeptides do not show any DNA glycosylase in vitro, they could have other, still non identified roles, or they might simply not be translated. Indeed, several global analyses comparing transcriptome and proteome data - 157 - suggest that a large fraction of the reported mRNAs are never translated into proteins. Proteomic studies suggest that the vast majority of genes have a single dominant splice isoform that is the most highly conserved between different species (Tress et al., 2017). These considerations would be in agreement with α-OGG1, the only OGG1 isoform conserved between mouse and human, being the 8-oxoG DNA glycosylase acting both, in the nucleus and in mitochondria, in both species.

Artificial targeting of human α-OGG1 to the mitochondria by the addition of the MnSOD MTS showed that the human protein can be active in this organelle and improve mitochondrial DNA repair and cellular survival in cells exposed to oxidative stress (Dobson et al., 2000; Rachek, 2002; Druzhyna et al., 2005; Kim et al., 2014). However, these experiments did not address the question of whether the endogenous α-OGG1 protein, without the additional targeting sequence, is imported inside the mitochondrial compartment. It was suggested that the MTS in α-OGG1 was not sufficient to target the protein to the mitochondria because of the presence of the very strong NLS at the C-terminus of the DNA glycosylase (Nishioka et al., 1999), despite the fact that the mouse protein, with a similarity close to 94% with α-OGG1 was found in both, the nucleus and the mitochondria. We showed here that isoform α-OGG1 is efficiently imported in mitochondria and that the MTS at the N-terminal part of the protein is functional and sufficient to ensure the targeting of the enzyme to the organelle. Consistently, several algorithms predict a canonical MTS for α-OGG1 and a mitochondrial localization of the protein with a very high probability, close to 90% (Fig. 3). In fact, the strength of the α-OGG1 MTS is not very different from that of the MnSOD MTS used in the chimeric MTS-OGG1 construct (Rachek et al., 2002) and our study definitely showed that the addition of an ectopic MTS is not required to ensure the efficient import of α-OGG1 into the mitochondria. The presence of dual targeting signals found in α-OGG1 is clearly not an exception as many proteins with dual functional signals for nuclear and mitochondrial localization have been identified (Yogev and Pines, 2011; Kazak et al., 2013).

Whether and how the subcellular localization of α-OGG1 is regulated remains an open question. In our experiments we did not observe any significant change in the subcellular distribution of α-OGG1 through the cell cycle or as a consequence of the exposure of the cells to oxidative stress. While most of the BER proteins are present in both the nucleus and the mitochondria, very little is known concerning the molecular mechanisms involved in the regulation of their subcellular localization. In the case of Saccharomyces cerevisiae DNA glycosylase Ntg1 the regulation of its subcellular distribution is modulated by oxidative stress and involves sumoylation (Griffiths et al., 2009; Swartzlander et al., 2010). The Ntg1 sumoylated sites have been recently identified and two of them are conserved in the human orthologue hNTH1, suggesting a similar regulation in higher eukaryotes (Swartzlander et al., 2016), although this possibility has not been further explored.

It is interesting to note that two in frame AUG codons are present in the N-terminal part of OGG1 mRNA and that they are largely conserved among mammals (Fig. 3 and Fig. S1). As suggested by the prediction algorithms for mitochondrial targeting, our results clearly show that if translation is initiated at the second AUG, thus truncating the MTS, the resulting protein is exclusively localized into the nucleus. It has been clearly established that alternative start AUG codons within a single transcript largely contribute to the diversity of the proteome (Van Damme et al., 2014), and in particular to the mitochondrial proteome (Kazak et al., 2013). Although our fractionation experiments did not allow us to observe any significant shift in size between the nuclear and mitochondrial OGG1, we cannot exclude that this second AUG could be used in

particular conditions to favor nuclear import. OGG1 shares two characteristics with the proteins using several AUGs in frame: 1/ the first AUG has a suboptimal context for translation initiation, thus facilitating ribosomal sliding (leaky mRNA scanning) that could originate different polypeptides from the same transcript; and 2/ the 5’UTR is particularly long (350 nt compared to a normal length of around 200 nt) (Bazykin et al., 2011).

Interestingly, the three proteins involved in the GO system (David et al., 2007), dedicated to counteract the impact of the incorporation of 8-oxoG in DNA, OGG1, MTH1 and MUTYH share some characteristics. As for OGG1, several transcripts originated by alternative splicing have been identified for human MTH1 and MUTYH, (Oda et al., 1999; Ohtsubo et al., 2000) and some of them have two or three putative initiation codons, which are functional at least in vitro. Also as for OGG1, initiation of translation at the first AUG in their mRNAs gives rise to a protein containing a MTS, while initiation at the second AUG results in the loss of the MTS and yields a protein localized exclusively in the nucleus. Considering that all those proteins that are essential for the protection of mtDNA against the deleterious effects of 8-oxoG have evolved from very different families of proteins and do not have any sequence homology, it is striking to observe the very similar organization of the N-terminal sequences with the presence of several AUGs in frame. Although the molecular details have not been elucidated, alternative translation initiation has already been proposed to be the mechanism of choice for the generation of nuclear and mitochondrial isoforms of LIG3 (Lakshmipathy and Campbell, 1999). It is tempting to speculate that evolution has converged to this strategy in order to modulate the subcellular distribution of those enzymes. Further studies are required to better understand how the subcellular localization of the DNA glycosylase α-OGG1 and other BER proteins is modulated.

Our microscopy and biochemical observations show for the first time the association of α-OGG1 with the mtDNA in nucleoids. The use of 3-color structured illumination microscopy (3D-SIM) provided here a detailed view of this association and demonstrated the close proximity between the BER glycosylase and TFAM inside the mitochondrial network. While several proteins involved in DNA metabolism, such as Pol, mtSSB and Twinkle have been found to associate with mtDNA only during replication in a highly regulated way (Rajala et al., 2014; Lewis et al., 2016), our experiments showed that the association of the glycosylase with the nucleoid is not linked to the replication state of the mtDNA but rather constitutive. Considering the highly compacted structure of mtDNA in nucleoids and the fact that the binding of TFAM to 8-oxoG containing DNA inhibits OGG1 activity in vitro (Canugovi et al., 2010), other proteins may be required to facilitate access of the DNA repair proteins to the mtDNA. A role for CSB in remodeling nucleoids by removing TFAM from DNA and facilitating the access of DNA repair machineries has been proposed (Boesch et al., 2012). This could explain the increase in 8-oxoG levels in mtDNA detected in Csb -/- mice (Stevnsner et al., 2002; Osenbroch et al., 2009). α-OGG1 is strongly associated with a mitochondrial insoluble fraction, probably the inner mitochondrial membrane, as it resists extraction with up to 500 mM of NaCl and is protected from protease digestion coupled to an osmotic shock. This is in agreement with previous studies revealing the association of BER proteins and/or activities (DNA glycosylases, polymerase and ligase and to a minor extent the endonuclease) with an insoluble mitochondrial fraction (Stuart et al., 2005; Boesch et al., 2010). The mitochondrial nucleoids being themselves anchored to the inner mitochondrial membrane (Brown et al., 2011; Jakobs and Wurm, 2014; Gilkerson et al., 2013), this distribution of BER proteins could facilitate the efficiency of the DNA repair process.

- 159 -

Role of α-OGG1 in the maintenance of mitochondrial function

In our study, and in agreement with previous observations, a human cell line deficient in all OGG1 isoforms displays in response to a menadione treatment a higher loss of mitochondrial membrane potential and an accumulation of mitochondrial ROS, both indicators of mitochondrial dysfunction. The fact that the re- expression of the α-OGG1 isoform in those cells fully complemented those phenotypes (Fig. 6) suggests that this isoform is perfectly functional. The failure of the α-OGG1-MTS protein, showing an exclusively nuclear localization, to complement the mitochondrial dysfunction of OGG1 deficient cells indicates that the presence of the protein inside mitochondria is required to maintain mitochondrial functions.

Under stress conditions, an increase in the fission versus fusion results in a fragmentation of the mitochondrial network, thus isolating the dysfunctional mitochondria that will be selectively removed by mitophagy. Furthermore, a perfect balance between fusion and fission is essential for the maintenance of mitochondrial health. A defect in either mitochondrial fusion or fission processes results in mtDNA instability (Busch et al., 2014, Garcia et al., 2017). The overexpression of α-OGG1 in human cells helps in preserving the mitochondrial network under oxidative stress conditions, as has also been observed in previous studies concerning the mouse Ogg1 (Torres-Gonzalez et al., 2014) and are in agreement with the general idea suggesting that DNA glycosylase activity is the rate-limiting step in the BER pathway (Hollenbach et al., 1999). The inability of the mutant protein α-OGG1(deltaMTS) in protecting mitochondrial network clearly indicates that the presence of the protein inside mitochondria is essential and it is not an indirect consequence of a better repair of nuclear DNA. Considering the major effect of the overexpression of α-OGG1 in protecting mitochondrial morphology, we cannot rule out that α-OGG1 has another role, independent on DNA repair, in mitochondrial fusion or fission. Indeed, the levels of mitochondrial proteins involved in fission, DRP and FIS1, have been reported to be reduced in cells overexpressing the mouse Ogg1 (Torres-Gonzalez et al., 2014), suggesting that the decrease in fission could account for the reduction in mitochondrial fragmentation observed. However, as fragmentation occurs as a consequence of mitochondrial dysfunction and is dependent on the loss of membrane potential, we can also imagine that this simply reflects the better health of mitochondria in cells overexpressing OGG1. In order to explore whether the enzymatic activity of α-OGG1 is required for the maintenance of mitochondrial physiology after oxidative stress, we used a mutant in which Lys 249, essential for the cleavage of the N-glycosylic bond and the release 8-oxoG, is replaced by a Gln (K249Q) (Van der Kemp et al., 2004). This amino acid substitution does not induce any change in the structure of the protein (Bruner et al., 2000) nor in its subcellular localization (Fig. 2S). While overexpression of the wt α-OGG1 protects the mitochondrial network from fragmentation after exposure to menadione, the active site mutant K249Q failed to do so. This is in agreement with the results obtained with the OGG1 mutant R229Q. This mutation was initially identified in a leukemic cell line and results in a loss of enzymatic activity due to protein destabilization (Hyun et al., 2000; Hill and Evans, 2007). Targeting of the mutant protein MTS-hOGG1-R229Q to the mitochondria results in decreased mtDNA integrity and cellular survival after exposure to oxidative agents when compared to the wild type MTS-hOGG1 (Chatterjee et al, 2006). The fact that catalytically inactive α-OGG1 mutants could not preserve the mitochondrial morphology in cells exposed to oxidative stress make

us favor the hypothesis that the observed mitochondrial dysfunctions are the consequence of the accumulation of 8-oxoG in the mitochondrial genome.

The presence of 8-oxoG has been proposed to be an important cause of mtDNA mutations and deletions, which accumulate with aging and during disease progression (Druzhyna et al., 2008). It is well accepted that the accumulation of 8-oxoG in DNA results in an increase in mutagenesis due to the misincorporation during DNA replication of adenine opposite 8-oxoG. An increase in mtDNA point mutations has been shown in yeast strains deficient for OGG1 (Singh et al., 2001). However, Halsne et al., reported that in Ogg1-/- mutant or in Ogg1-/- Myh-/- double mutant mice mtDNA mutation rates were not different from that of WT animals (Halsne et al. 2012). The experiments presented here show that menadione treatment rapidly induces mitochondrial dysfunction in OGG1-deficient cells, with around 40% of the cells showing a massive loss of membrane potential in less than 30 minutes. Those rapid effects could hardly be explained by an increase in mutagenesis, specially taking into account the high number of copies of mtDNA and the high degree of heteroplasmy and complementation between different mtDNA molecules. So, how can the failure to remove 8-oxoG affect mitochondrial functions? It has been proposed that Polymerase γ is blocked or paused by 8- oxoG (Graziewicz et al., 2007; Stojkovič et al., 2016), leading to the loss of mtDNA instead of mutagenesis. Although other replisome-associated proteins such as Twinkle or Primpol have been suggested to help the polymerase bypassing 8-oxoG (Garcia-Gomez et al., 2013; Bianchi et al., 2013; Stojkovič et al., 2016), other studies are required to conclude on the role of those proteins in processing damaged mtDNA. Several reports have shown that oxidative stress induces the degradation of the mtDNA and mtRNA (Crawford et al., 1997 and 1998; Abramova et al., 2000; Shokolenko et al., 2009; Rothfuss et al. 2010; Furda et al. 2012). It is interesting to note that while low levels of 8-oxoG have been detected in the circular molecule, fragmented mtDNA had a very high 8-oxodG content, which was further increased after oxidative stress (Suter and Richter, 1999). Thus, whether the accumulation of 8-oxoG in mitochondrial DNA in the absence of the DNA glycosylase activity of α-OGG1 represents a direct block for transcription or replication or it induces degradation of the DNA molecules remains to be further explored.

Materials and Methods

Cell lines, Plasmids construction, transfections, RNA interference and treatments

U2OS, HeLa and HEK cells were cultured at 37°C in Dulbecco’s modified Eagle’s medium (DMEM, Sigma ref D5671), supplemented with 10% fetal bovine serum (FBS) (Sigma ref F7524) and 5% penicillin/streptomycin (Gibco ref 15140122) in 5% CO2.

pmTurquoise2-Mito was a gift from Dorus Gadella (Addgene plasmid # 36208)(Goedhart et al., 2012). α-OGG1 fusions to the FLAG epitope were generated by PCR, including the sequence coding for the FLAG in the reverse oligonucleotide, and cloned in the plasmid pCDNA3.1(-). The open reading frame of α-OGG1 was amplified by PCR to generate the fusion proteins OGG1(WT)-FLAG, OGG1(MTS)-FLAG and OGG1(NLS)- FLAG. The point mutations K249Q and G12E were introduced by site-directed mutagenesis from the hOGG1(WT)-FLAG construct, using the QuikChange Site-Directed Mutagenesis Kit (STRATAGENE). To facilitate the generation of stable cell lines, the same ORFs where cloned in the polycistronic pIRES2-AcGFP1 plasmid (CLONTECH) expressing the GFP in the second position of the cistron and thus allowing selecting

- 161 - the transfected cells expressing similar levels of the GFP by cell sorting. The oligonucleotides used to generate the different constructs are listed in Table S1

Cells were grown on coverslips, or on Ibidi µ-Slide 4 or 8 well plates for microscopy experiments and on T150 flasks for biochemical analysis. Transient and stable transfections were performed 24-48 hours after the seeding, using Lipofectamine 2000 (Invitrogen) and following a standard protocol with a ratio DNA/lipofectamine of 1µg/4µL. For stable cell lines expressing OGG1(WT)-FLAG or OGG1(MTS)-FLAG from pIRES plasmids, the culture medium was supplemented with 400 μg/ml of G418.

For the siRNA-mediated depletion of hOGG1, U2OS cells were transfected with siRNA against hOGG1 (GGAUCAAGUAUGGACACUG) or AllStars Negative Control siRNA (QIAGEN) using Lipofectamine RNAiMAX (Invitrogen).

For the induction of an oxidative stress we used menadione at 50µM for 1 hour in DMEM without FBS and antibiotics. All dilutions were done in pre-warmed DMEM immediately before use.

Immunofluorescence and Confocal microscopy

For microscopy cells were plated in 4 or 8 Well µ-Slide from ibidi. For the staining of the mitochondrial network, cells were rinsed twice with DMEM and incubated with 200nM of MitoTracker RED CMXRos and/or MitoTracker Deep RED (purchased from ThermoFisher Scientific) diluted in complete DMEM for 30 min at 37°C. When indicated, DNA was stained with picogreen at 1/500 (Quant-iT™ PicoGreen™ dsDNA Reagent, ThermoFisher Scientific). Cells were fixed with 2% of formaldehyde for 20min at 37°C, rinsed with PBS and permeabilized at room temperature in PBS–0.1% Triton for 5min. Cells were incubated in blocking solution (PBS containing 0.1% Triton, 3% BSA and 1% normal goat serum) at 37 °C for 1 hour or O.N. at 4°C. Cells were incubated for 1-2h at 37 °C with primary antibody diluted in blocking solution, washed three times for 5 min in PBS-0.1% Triton and further incubated with secondary Alexa Fluor fluorescent‐conjugated antibodies (Molecular probes) diluted in blocking solution for 45min at 37°C. When indicated, nuclear DNA was counterstained with 1 µg/ml 4′,6′-diamidino-2-phenylindole (DAPI). The primary antibodies and the dilutions used were as follows: FLAG (SIGMA, F3165) 1:2000, TFAM (abcam, ab155240), 1:2000, CENPF (abcam, Ab5) 1:500, LC3B (Cell Signaling Technology, 27755) 1:250. EdU, 5-éthynyl-2’-déoxyuridine, was added to the medium at a concentration of 1 mM to label replicating cells and stained following the manufacturer’s instructions (ThermoFischer Click-iT® EdU Alexa Fluor® 647).

Image acquisition was performed with a Leica confocal microscope SP8 with a 60x oil immersion objective or with a Nikon confocal A1 with a 63x oil immersion objective (NA 1,3 for both). Image analysis was performed with ImageJ (Schneider et al., 2012). Pearson correlation coefficients were calculated from 8 cells (around 1000 nucleoids) using the plugin JACOP (Bolte and Cordelières, 2006). Mitochondrial morphology alterations were characterized by applying the Kernel’s algorithm (Koopman et al., 2005) to binarized images. This approach allows to measure morphology parameters for individual mitochondria such as the aspect ratio (AR), a measure of mitochondrial length, and the form factor (F), which is a combined measure of mitochondrial length and degree of branching.

Super-resolution microscopy

Structured illumination (3D-SIM) was performed using a rotary-stage OMX v3 system (Applied Precision—GE Healthcare), equipped with 3 EMCCD, Evolve cameras (Photometrics). Signals from all channels were realigned using fluorescent beads before each session of image acquisition. All images were acquired with a PlanApo 100×/1.4 oil objective at 125 nm intervals along the z axis. Resolution checked on 100 nm beads show 125nm, 125, 260 nm in x,y and z axis respectively. Channels alignment was realized with ImageJ using UnwarpJ plugin (Sánchez Sorzano, et al., 2005).

WCE, nuclear and mitochondrial extractions

Whole cell extracts were obtained from 106 U2OS cells. Cell pellets were resuspended in TpL buffer (TP lysis buffer: 20mM Tris HCl pH 7.5, 250mM NaCl, 1mM EDTA, protease inhibitors) and sonicated in a Bioruptor® bath (30″ on/30″ off pulses for 10 min at maximum intensity). After sonication the samples were centrifuged at 20000xg for 30 minutes at 4°C, the pellets discarded and the supernatants kept on ice.

Nuclear protein extracts were obtained from 106 U2OS with the Nuclei Isolation Kit: Nuclei EZ Prep (SIGMA), according to manufacturer’s instructions.

Crude mitochondria extracts were prepared from 120x106 U2OS cells, as described by Wieckowski et al. 2009 with few modifications. Briefly, cells at 90-95% of confluence were detached by trypsination and collected by centrifugation at 600×g for 7 minutes at 4°C. After one wash with ice-cold PBS the cells were resuspended in ice-cold IB1 buffer (Isolation buffer 1: 225 mM mannitol, 75mM sucrose, 0.1 mM EGTA, 30mM Tris HCl pH 7.5) and disrupted with 15-30 strokes with a glass dounce homogenizer on ice. The integrity of the cells was checked under the microscope. The homogenate was centrifuged at 600xg for 7 minutes at 4°C. The supernatant was transferred into a pre-cooled ultracentrifuge vial and centrifuged at 7000xg for 12 minutes at 4°C. The supernatant was discarded and the pellet containing the mitochondria was gently resuspended in ice-cold IB2 buffer (Isolation buffer 2: 225 mM mannitol, 75mM sucrose, 30mM Tris HCl pH 7.5) and centrifuged again at 7000xg for 12 minutes at 4°C. The mitochondrial pellet was resuspended in IB2 buffer and re-pelleted at 10000xg for 12 minutes at 4°C. The crude mitochondrial pellet was resuspended in 120 µL of ice-cold MRB buffer (Mitochondrial Resuspension Buffer: 250mM mannitol, 0.5mM EGTA, and 5mM HEPES pH 7.4).

Protein contents of the different extracts were quantified by Bradford Assay.

Iodixanol gradient

Mitochondria isolated from HEK cells by differential and sucrose gradient centrifugation were treated with trypsin prior to lysis and fractionation on an iodixanol gradient as previously reported (Reyes et al., 2011). Protein and DNA were recovered from the different fractions (18 in total) and analysed by Southern and Western blotting. A mitochondrial DNA fragment amplified by PCR with specific oligos (Table S1) was incubated with DNA-labeling beads (GE Healthcare) in the presence of 50mCi of [a-32P] dCTP (3000 Ci/mmol, Perkin Elmer) for 30 min. Then, the Southern blot was hybridized to this mtDNA specific probe by overnight incubation at 65C in 7% SDS/ 0.25M sodium phosphate buffer (pH 7.4). Post hybridization washes were 1XSSC followed by 1XSSC/ 0.1% SDS twice for 30 min at 65C. Filters were exposed to phosphorscreens and scanned using a Typhoon phosporimager (GE Healthcare). The Western blots were incubated with antibodies against TFAM, the soluble glycyl-tRNA synthetase GARS and α-OGG1.

- 163 -

Proteinase K and trypsin partial digestion

25µg of mitochondrial extracts were incubated for 5 min at 4°C in 1 mL of MBR buffer with 1 mg/ml of BSA, with a final concentration of 5 or 25 µg/mL of Proteinase K (PK). When indicated a pretreatment with 1% of Triton- X100 for 5 min at 4°C was performed before the incubation with PK. PK was inactivated on ice by adding 10 mM PMSF for 5 min. The samples were then centrifuged at 14000rpm for 10’ at 4°C and the supernatant discarded. The resulting pellets were resuspended in 50 µL of MBR. Untreated mitochondrial extracts were used as a control.

Trypsin digestion

Sucrose-gradient purified mitochondria from HEK cells (2mg/ml) were resuspended in isotonic (20 mM HEPES pH 7.8, 2 mM EDTA, 210 mM mannitol, 70 mM sucrose) or hypotonic (10 mM HEPES pH 7.8) buffer and treated with or without 100 mg/ml of trypsin at room temperature for 30 min. Alternatively, mitochondria were treated with 200 or 500 mg/ml digitonin in isotonic buffer for 10 min at 4°C prior to trypsin treatment, as previously reported (Reyes et al., 2011). After washing and pelleting mitochondria three times, the organelles were lysed with 1% SDS in isotonic buffer. TOM20, TIM23, SDHA and EFTB were used as markers for the mitochondrial outer membrane, intermembrane space, mitochondrial inner membrane and mitochondrial matrix, respectively.

Extraction of mitochondrial membranes

Sucrose-gradient purified mitochondria from HEK cells (2 mg/ml) were sonicated for 1 min on ice followed by centrifugation at 10.000 g for 10 min at 4°C. The supernatant was treated with 150 or 500 mM NACl or 2% SDS on ice for 30 min followed by centrifugation at 122.000 g for 30 min at 4°C. Supernatants containing the soluble fractions were recovered and pellets with membrane-associated proteins were suspended in equal volume of isotonic buffer containing 0,2% SDS, as previously described (Martinez Lyons et al., 2016). TOM20 and CS were used as controls of integral membrane and soluble protein, respectively.

Western blotting analysis

Aliquots from each cell extracts were denatured by heating at 95 °C for 5 min. Loading of the extracts was normalized by protein content and 10-20 µg of protein extracts were separated in 10% SDS-acrylamide gel and transferred onto a nitrocellulose membrane. To confirm the amounts of protein in each lane, membranes were stained with Ponceau red and then blocked in 8% milk in PBS with 0.1% Tween 20 (TPBS) for 30 min at RT or O.N. at 4°C. After the blocking step the membranes were incubated for 90 min at RT with primary antibodies in 3% milk-PBS. All the antibodies used are listed in Table S2. After three 5 min washings with PBS, blots were incubated 45 min at RT with fluorescently labeled secondary antibodies: anti-Rabbit-800 (ref 05060- 250), anti-Mouse-800 (ref 05061-250), anti-Rabbit-700 (ref 05054-250) and/or anti-Mouse-700 (ref 050055- 250) (Diagomics) diluted 1:10000 in 2% milk-TPBS or HRP conjugated secondary antibodies (Promega anti- Rabbit W401B, anti-Mouse W402B) 1:5000 and then washed 3 times with TPBS for 7-10’ min. The analysis of the images was performed by scanning the membrane at 800 and 700nm simultaneously with an Odyssey® CLx or by exposure to X-ray films.

Determination of mitochondrial mass, membrane potential and ROS production

Analysis of mitochondrial mass, mitochondrial membrane potential (MMP) and intramitochondrial ROS (mtROS) production were performed by flow cytometry (BD LSRII) or microplate reader (CLARIOstar - BMG Labtech).

For Flow cytometry analysis 0.4exp6 U2OS were seeded in a 6 well plate and transfected with siOGG1 or with siAllStar as previously described. 72h after transfection the cells were treated 1h at 37°C with 50µM of menadione. To evaluate mtROS, U2OS cells were incubated with 5µM of MitoSOX® Red Mitochondrial Superoxide Indicator (ThermoFisher, ref. M36008) in PBS-BSA 0.4% for 10 min at 37°C protected from light, harvested with trypsin and directly analyzed. For MMP analysis cells were incubated with 100nM of MitoTracker GREEN FM (ThermoFisher, ref. M7514) for 30 min at 37°C, rinsed once and stained with 100 nM of TMRE (Tetramethylrhodamine, Ethyl Ester, Perchlorate, ThermoFischer, ref. T669) diluted in PBS-BSA 0.4% for 15’, 37°C protected from light. The cells were then harvested with trypsin and directly analyzed. For each sample, at least 50.000 cells were analyzed. Experiments were performed three times with similar results.

For plate reader measurements 10000 cells were reverse transfected with siOGG1 or siAllStars in a 96-well microplate for fluorescence assays, black-walled and with clear-bottom (Greiner Bio-One, ref. 655090). The cells were treated with 25 or 50 µM of menadione for 1h at 37°C and then rinsed once with complete DMEM. For the analysis of mtROS production after the chemical treatment the cells were stained for 10 min maximum with MitoSOX® Red, 5µM and Hoecst 33342 1:1000 diluted in pre-warmed PBS (+Ca2+ and Mg2+), the PBS was then replaced with 199 µl medium before the acquisition. To evaluate the MMP the cells were stained before the chemical treatment with 100 nM of TMRE and 100 nM of MitoTracker GREEN for 30 min at 37°C protected from light.

OGG1 glycosylase activity

A single-stranded 34-mer DNA containing an 8-oxoG at position 16 was 5 -end labelled with Cy5 and hybridized to the complementary oligonucleotide containing a cytosine opposite the lesion, yielding the 8- oxoG:C duplexes. Reactions were carried out at 37°C in a total volume of 14 µl, containing the specified amount of protein extract and 150 fmol of DNA probe. The reaction buffer was 22 mM Tris–HCl pH 7.4, 110 mM NaCl, 2.5 mM EDTA, 1 mg/ml BSA and 5% glycerol. Reaction mixtures containing 8-oxoG:C probe were incubated at 37 ◦C for 1 h, then NaOH (0.1 N final concentration) was added and the mixture was further incubated for 15 min at 37 ◦C and stopped by adding 6 µl of formamide dye (80% formamide, 10 mM EDTA, 0.02% bromophenol blue), followed by heating for 5 min at 95 ◦C. The products of the reactions were resolved by denaturing 7 M urea – 20% polyacrylamide gel electrophoresis. Gels were scanned using a Typhoon Multi- format Imager (GE Healthcare Life Sciences).

Acknowledgments

We thank the scientific and technical assistance of the Microscopy, Flow cytometry and Molecular Biology facilities of the IRCM Institute. We thank Samantha Lewis (Nunnari lab) for her help in the optimization of the protocol for visualization of replicating nucleoids. We would like to thank Paul-Henri Romeo, Christophe Carles and Chantal Desmaze for critically reading the manuscript. This research was supported by grants from Ligue contre le cancer, Electricité de France and the CEA (Impulsion).

Author contributions

- 165 -

D. Lia performed experiments and analysed the data. A. Reyes performed experiments for Fig 1 (B and C) and 2 (E) and edited the manuscript. JT de Melo performed preliminary experiments. T. Piolot performed SIM experiments presented in Fig 2 (C and D). J. Baijer analysed flow cytometry data (Fig. 5). J.P. Radicella edited the manuscript. A. Campalans conceived the study, designed and performed experiments, analysed the data and wrote the paper.

References

Abramova, N. E., Davies, K. J. A. and Crawford, D. R. (2000). Polynucleotide degradation during early stage response to oxidative stress is specific to mitochondria. Free Radic. Biol. Med. 28, 281–288.

Alam, T. I., Kanki, T., Muta, T., Ukaji, K., Abe, Y., Nakayama, H., Takio, K., Hamasaki, N. and Kang, D. (2003). Human mitochondrial DNA is packaged with TFAM. Nucleic Acids Res. 31, 1640–1645.

Audebert, M., Charbonnier, J. B., Boiteux, S. and Radicella, J. P. (2002). Mitochondrial targeting of human 8-oxoguanine DNA glycosylase hOGG1 is impaired by a somatic mutation found in kidney cancer. DNA Repair (Amst). 1, 497–505.

Bazykin, G. A. and Kochetov, A. V. (2011). Alternative translation start sites are conserved in eukaryotic genomes. Nucleic Acids Res. 39, 567–577.

Berquist, B. R., Canugovi, C., Sykora, P., Wilson, D. M. and Bohr, V. A. (2012). Human Cockayne syndrome B protein reciprocally communicates with mitochondrial proteins and promotes transcriptional elongation. Nucleic Acids Res. 40, 8392–8405.

Boesch, P., Ibrahim, N., Dietrich, A. and Lightowlers, R. N. (2010). Membrane association of mitochondrial DNA facilitates base excision repair in mammalian mitochondria. Nucleic Acids Res 38, 1478–1488.

Bolte, S. and Cordelieres, F. P. (2006). A guided tour into subcellular colocalisation analysis in light microscopy. J. Microsc. 224, 13–232.

Brown, T. A., Tkachuk, A. N., Shtengel, G., Kopek, B. G., Bogenhagen, D. F., Hess, H. F. and Clayton, D. A. (2011). Superresolution Fluorescence Imaging of Mitochondrial Nucleoids Reveals Their Spatial Range, Limits, and Membrane Interaction. Mol. Cell. Biol. 31, 4994–5010.

Bruner, S. D., Norman, D. P. and Verdine, G. L. (2000). Structural basis for recognition and repair of the endogenous mutagen 8-oxoguanine in DNA. Nature 403, 859–66.

Busch, K. B., Kowald, A. and Spelbrink, J. N. (2014). Quality matters: how does mitochondrial network dynamics and quality control impact on mtDNA integrity? Philos. Trans. R. Soc. Lond. B. Biol. Sci. 369, 20130442.

Canugovi, C., Maynard, S., Bayne, A.-C. V., Sykora, P., Tian, J., de Souza-Pinto, N. C., Croteau, D. L. and Bohr, V. A. (2010). The mitochondrial transcription factor A functions in mitochondrial base excision repair. DNA Repair (Amst). 9, 1080–1089.

Chatterjee, A., Mambo, E., Zhang, Y., Deweese, T. and Sidransky, D. (2006). Targeting of mutant hogg1 in mammalian mitochondria and nucleus: effect on cellular survival upon oxidative stress. BMC Cancer 6, 235.

Crawford, D. R., Wang, Y., Schools, G. P., Kochheiser, J. and Davies, K. J. A. (1997). Down-regulation of mammalian mitochondrial RNAs during oxidative stress. Free Radic. Biol. Med. 22, 551–559.

Crawford, D. R., Abramova, N. E. and Davies, K. J. A. (1998). Oxidative stress causes a general, calcium- dependent degradation of mitochondrial polynucleotides. Free Radic. Biol. Med. 25, 1106–1111.

David, S. S., O’Shea, V. L. and Kundu, S. (2007). Base-excision repair of oxidative DNA damage. Nature 447, 941–950. de Souza-Pinto, N. C., Hogue, B. A. and Bohr, V. A. (2001). DNA repair and aging in mouse liver: 8-oxodG glycosylase activity increase in mitochondrial but not in nuclear extracts. Free Radic Biol Med 30, 916– 923.

Dobson, A. W., Xu, Y., Kelley, M. R., LeDoux, S. P. and Wilson, G. L. (2000a). Enhanced mitochondrial DNA repair and cellular survival after oxidative stress by targeting the human 8-oxoguanine glycosylase repair enzyme to mitochondria. J Biol Chem 275, 37518–37523.

Dobson, A. W., Xu, Y., Kelley, M. R., LeDoux, S. P. and Wilson, G. L. (2000b). Enhanced mitochondrial DNA repair and cellular survival after oxidative stress by targeting the human 8-oxoguanine glycosylase repair enzyme to mitochondria. J. Biol. Chem. 275, 37518–37523.

Dobson, A. W., Grishko, V., LeDoux, S. P., Kelley, M. R., Wilson, G. L. and Gillespie, M. N. (2002). Enhanced mtDNA repair capacity protects pulmonary artery endothelial cells from oxidant-mediated death. Am. J. Physiol. Lung Cell. Mol. Physiol. 283, L205–10.

Druzhyna, N. M., Musiyenko, S. I., Wilson, G. L. and LeDoux, S. P. (2005). Cytokines induce nitric oxide- mediated mtDNA damage and apoptosis in oligodendrocytes: Protective role of targeting 8-oxoguanine glycosylase to mitochondria. J. Biol. Chem. 280, 21673–21679.

Druzhyna, N. M., Wilson, G. L. and Ledoux, S. P. (2009). Mitochondrial DNA repair in aging and disease. Mech Ageing Dev 129, 383–390.

Furda, A. M., Marrangoni, A. M., Lokshin, A. and Van Houten, B. (2012). Oxidants and not alkylating agents induce rapid mtDNA loss and mitochondrial dysfunction. DNA Repair (Amst). 11, 684–692.

Furihata, C. (2015). An active alternative splicing isoform of human mitochondrial 8-oxoguanine DNA glycosylase (OGG1). Genes Environ. 37, 21.

Garcia-Gomez S., Reyes A., Martinez-Jimenez M.I., Chocron E.S., Mouron S., Terrados G., Powell C., Salido E., Mendez J., Holt I.J. et al. (2013). PrimPol, an Archaic Primase/Polymerase Operating in Human Cells. Mol. Cell 52, 541–553.

Garcia, I., Jones, E., Ramos, M., Innis-Whitehouse, W. and Gilkerson, R. (2017). The little big genome: the organization of mitochondrial DNA. Front. Biosci. (Landmark Ed. 22, 710–721.

Garcia-Gomez, S., Reyes, A., Martinez-Jimenez, M. I., Chocron, E. S., Mouron, S., Terrados, G., Powell, C., Salido, E., Mendez, J., Holt, I. J., et al. (2013). PrimPol, an Archaic Primase/Polymerase Operating in Human Cells. Mol. Cell 52, 541–553.

Gilkerson, R. W., Schon, E. A., Hernandez, E. and Davidson, M. M. (2008). Mitochondrial nucleoids maintain genetic autonomy but allow for functional complementation. J. Cell Biol. 181, 1117–1128.

Gilkerson, R., Bravo, L., Garcia, I., Gaytan, N. and Herrera, A. (2013). The Mitochondrial Nucleoid : Integrating. Cold Spring Harb Perspect Biol 5, 1–10.

Goedhart, J., von Stetten, D., Noirclerc-Savoye, M., Lelimousin, M., Joosen, L., Hink, M. A., van Weeren, L., Gadella, T. W. J. and Royant, A. (2012). Structure-guided evolution of cyan fluorescent proteins towards a quantum yield of 93%. Nat. Commun. 3, 751.

Graziewicz, M. A., Bienstock, R. J. and Copeland, W. C. (2007). The DNA polymerase γ Y955C disease variant associated with PEO and parkinsonism mediates the incorporation and translesion synthesis opposite 7,8-dihydro-8-oxo-2′-deoxyguanosine. Hum. Mol. Genet. 16, 2729–2739.

- 167 -

Griffiths, L. M., Swartzlander, D., Meadows, K. L., Wilkinson, K. D., Corbett, A. H. and Doetsch, P. W. (2009). Dynamic compartmentalization of base excision repair proteins in response to nuclear and mitochondrial oxidative stress. Mol Cell Biol 29, 794–807.

Halsne, R., Esbensen, Y., Wang, W., Scheffler, K., Suganthan, R., Bjørås, M. and Eide, L. (2012). Lack of the DNA glycosylases MYH and OGG1 in the cancer prone double mutant mouse does not increase mitochondrial DNA mutagenesis. DNA Repair (Amst). 11, 278–285.

Harbauer, A. B., Opali ska, M., Gerbeth, C., Herman, J. S., Rao, S., Schonfisch, B., Guiard, B., Schmidt, O., Pfanner, N. and Meisinger, C. (2014). Cell cycle-dependent regulation of mitochondrial preprotein translocase. Science (80-. ). 346, 1109–1113.

Hill, J. W. and Evans, M. K. (2007). A novel R229Q OGG1 polymorphism results in a thermolabile enzyme that sensitizes KG-1 leukemia cells to DNA damaging agents. Cancer Detect. Prev. 31, 237–243.

Hollenbach, S., Dhénaut, A., Eckert, I., Radicella, J. P. and Epe, B. (1999). Overexpression of Ogg1 in mammalian cells: Effects on induced and spontaneous oxidative DNA damage and mutagenesis. Carcinogenesis 20, 1863–1868.

Hyun, J. W., Choi, J. Y., Zeng, H. H., Lee, Y. S., Kim, H. S., Yoon, S. H. and Chung, M. H. (2000). Leukemic cell line, KG-1 has a functional loss of hOGG1 enzyme due to a point mutation and 8- hydroxydeoxyguanosine can kill KG-1. Oncogene 19, 4476–9.

Jakobs, S. and Wurm, C. A. (2014). Super-resolution microscopy of mitochondria. Curr. Opin. Chem. Biol. 20, 9–15.

Kazak, L., Reyes, A. and Holt, I. J. (2012). Minimizing the damage: repair pathways keep mitochondrial DNA intact. Nat. Rev. Mol. Cell Biol. 13, 726–726.

Kazak, L., Reyes, A., Duncan, A. L., Rorbach, J., Wood, S. R., Brea-Calvo, G., Gammage, P. A., Robinson, A. J., Minczuk, M. and Holt, I. J. (2013). Alternative translation initiation augments the human mitochondrial proteome. Nucleic Acids Res. 41, 2354–2369.

Kim, S. J., Cheresh, P., Williams, D., Cheng, Y., Ridge, K., Schumacker, P. T., Weitzman, S., Bohr, V. A. and Kamp, D. W. (2014). Mitochondria-targeted Ogg1 and aconitase-2 prevent oxidant-induced mitochondrial DNA damage in alveolar epithelial cells. J. Biol. Chem. 289, 6165–6176.

Kukat, C., Wurm, C. a, Spåhr, H., Falkenberg, M., Larsson, N.-G. and Jakobs, S. (2011). Super- resolution microscopy reveals that mammalian mitochondrial nucleoids have a uniform size and frequently contain a single copy of mtDNA. Proc. Natl. Acad. Sci. U. S. A. 108, 13534–13539.

Lakshmipathy, U. and Campbell, C. (1999). The human DNA ligase III gene encodes nuclear and mitochondrial proteins. Mol Cell Biol 19, 3869–3876.

Legros, F., Malka, F., Frachon, P., Lombes, A. and Rojo, M. (2004). Organization and dynamics of human mitochondrial DNA. J Cell Sci 117, 2653–2662.

Leon, J., Sakumi, K., Castillo, E., Sheng, Z., Oka, S. and Nakabeppu, Y. (2016). 8-Oxoguanine accumulation in mitochondrial DNA causes mitochondrial dysfunction and impairs neuritogenesis in cultured adult mouse cortical neurons under oxidative conditions. Sci. Rep. 6, 22086.

Lewis, S. C., Uchiyama, L. F. and Nunnari, J. (2016). ER-mitochondria contacts couple mtDNA synthesis with mitochondrial division in human cells. Science (80-. ). 353, aaf5549.

Martinez Lyons, A., Ardissone, A., Reyes, A., Robinson, A. J., Moroni, I., Ghezzi, D., Fernandez- Vizarra, E. and Zeviani, M. (2016). COA7 (C1orf163/RESA1) mutations associated with mitochondrial leukoencephalopathy and cytochrome c oxidase deficiency. J. Med. Genet. 53, 846–849.

Mishra, P. and Chan, D. C. (2014). Mitochondrial dynamics and inheritance during cell division, development and disease. 15, 634–646.

Model, K., Meisinger, C. and Kühlbrandt, W. (2008). Cryo-Electron Microscopy Structure of a Yeast Mitochondrial Preprotein Translocase. J. Mol. Biol. 383, 1049–1057.

Nishioka, K., Ohtsubo, T., Oda, H., Fujiwara, T., Kang, D., Sugimachi, K. and Nakabeppu, Y. (1999). Expression and differential intracellular localization of two major forms of human 8-oxoguanine DNA glycosylase encoded by alternatively spliced OGG1 mRNAs. Mol Biol Cell 10, 1637–1652.

Oda, H., Taketomi, A., Maruyama, R., Itoh, R., Nishioka, K., Yakushiji, H., Suzuki, T., Sekiguchi, M. and Nakabeppu, Y. (1999). Multi-forms of human MTH1 polypeptides produced by alternative translation initiation and single nucleotide polymorphism. Nucleic Acids Res 27, 4335–4343.

Ogawa, A., Watanabe, T., Shoji, S. and Furihata, C. (2015). Enzyme kinetics of an alternative splicing isoform of mitochondrial 8-oxoguanine DNA glycosylase, OGG1-1b, and compared with the nuclear OGG1-1a. J. Biochem. Mol. Toxicol. 29, 49–56.

Ohtsubo, T. (2000). Identification of human MutY homolog (hMYH) as a repair enzyme for 2- hydroxyadenine in DNA and detection of multiple forms of hMYH located in nuclei and mitochondria. Nucleic Acids Res. 28, 1355–1364.

Oka, S., Ohno, M., Tsuchimoto, D., Sakumi, K., Furuichi, M. and Nakabeppu, Y. (2008). Two distinct pathways of cell death triggered by oxidative damage to nuclear and mitochondrial DNAs. Embo J 27, 421–432.

Oka, S., Leon, J., Tsuchimoto, D., Sakumi, K. and Nakabeppu, Y. (2014). MUTYH, an adenine DNA glycosylase, mediates p53 tumor suppression via PARP-dependent cell death. Oncogenesis 3, e121.

Osenbroch, P. O., Auk-Emblem, P., Halsne, R., Strand, J., Forstrøm, R. J., Van Der Pluijm, I. and Eide, L. (2009). Accumulation of mitochondrial DNA damage and bioenergetic dysfunction in CSB defective cells. FEBS J. 276, 2811–2821.

Rachek, L. I., Grishko, V. I., Musiyenko, S. I., Kelley, M. R., LeDoux, S. P. and Wilson, G. L. (2002). Conditional targeting of the DNA repair enzyme hOGG1 into mitochondria. J Biol Chem 277, 44932– 44937.

Rajala, N., Gerhold, J. M., Martinsson, P., Klymov, A. and Spelbrink, J. N. (2014). Replication factors transiently associate with mtDNA at the mitochondrial inner membrane to facilitate replication. Nucleic Acids Res. 42, 952–967.

Reyes, A., He, J., Mao, C. C., Bailey, L. J., Di Re, M., Sembongi, H., Kazak, L., Dzionek, K., Holmes, J. B., Cluett, T. J., et al. (2011). Actin and myosin contribute to mammalian mitochondrial DNA maintenance. Nucleic Acids Res. 39, 5098–5108.

Richter, C., Park, J. W. and Ames, B. N. (1988). Normal oxidative damage to mitochondrial and nuclear DNA is extensive. Proc. Natl. Acad. Sci. U. S. A. 85, 6465–7.

Rothfuss, O., Gasser, T. and Patenge, N. (2009). Analysis of differential DNA damage in the mitochondrial genome employing a semi-long run real-time PCR approach. Nucleic Acids Res. 38,.

Santos, J. H., Hunakova, L., Chen, Y., Bortner, C. and Van Houten, B. (2003). Cell sorting experiments link persistent mitochondrial DNA damage with loss of mitochondrial membrane potential and apoptotic cell death. J. Biol. Chem. 278, 1728–1734.

Schneider, C. A., Rasband, W. S. and Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675.

- 169 -

Schon, E. A. and Gilkerson, R. W. (2010). Functional complementation of mitochondrial DNAs: Mobilizing mitochondrial genetics against dysfunction. Biochim. Biophys. Acta - Gen. Subj. 1800, 245–249.

Schulz, C., Schendzielorz, A. and Rehling, P. (2015). Unlocking the presequence import pathway. Trends Cell Biol. 25, 265–275.

Shigenaga, M. K., Hagen, T. M. and Ames, B. N. (1994). Oxidative damage and mitochondrial decay in aging. Proc. Natl. Acad. Sci. U. S. A. 91, 10771–8.

Shokolenko, I., Venediktova, N., Bochkareva, A., Wilson, G. L. and Alexeyev, M. F. (2009). Oxidative stress induces degradation of mitochondrial DNA. Nucleic Acids Res. 37, 2539–48.

Simsek, D., Furda, A., Gao, Y., Artus, J., Brunet, E., Hadjantonakis, A. K., Van Houten, B., Shuman, S., McKinnon, P. J. and Jasin, M. (2011). Crucial role for DNA ligase III in mitochondria but not in Xrcc1- dependent repair. Nature 471, 245–248.

Singh, K. K., Sigala, B., Sikder, H. A. and Schwimmer, C. (2001). Inactivation of Saccharomyces cerevisiae OGG1 DNA repair gene leads to an increased frequency of mitochondrial mutants. Nucleic Acids Res. 29, 1381–8.

Sorzano, C. Ó. S., Thévenaz, P. and Unser, M. (2005). Elastic registration of biological images using vector-spline regularization. IEEE Trans. Biomed. Eng. 52, 652–663.

Stevnsner, T., Thorslund, T., de Souza-Pinto, N. C. and Bohr, V. A. (2002a). Mitochondrial repair of 8- oxoguanine and changes with aging. Exp Gerontol 37, 1189–1196.

Stevnsner, T., Nyaga, S., de Souza-Pinto, N. C., van der Horst, G. T. J., Gorgels, T. G. M. F., Hogue, B. a, Thorslund, T. and Bohr, V. a (2002b). Mitochondrial repair of 8-oxoguanine is deficient in Cockayne syndrome group B. Oncogene 21, 8675–82.

Stojkovič, G., Makarova, A. V., Wanrooij, P. H., Forslund, J., Burgers, P. M. and Wanrooij, S. (2016). Oxidative DNA damage stalls the human mitochondrial replisome. Sci. Rep. 6, 28942.

Stuart, J. A., Mayard, S., Hashiguchi, K., Souza-Pinto, N. C. and Bohr, V. A. (2005). Localization of mitochondrial DNA base excision repair to an inner membrane-associated particulate fraction. Nucleic Acids Res 33, 3722–3732.

Suter, M. and Richter, C. (1999). Fragmented Mitochondrial DNA Is the Predominant Carrier of Oxidized DNA Bases †. Biochemistry 38, 459–464.

Swartzlander, D. B., Griffiths, L. M., Lee, J., Degtyareva, N. P., Doetsch, P. W. and Corbett, A. H. (2010). Regulation of base excision repair: Ntg1 nuclear and mitochondrial dynamic localization in response to genotoxic stress. Nucleic Acids Res 38, 3963–3974.

Swartzlander, D. B., McPherson, A. J., Powers, H. R., Limpose, K. L., Kuiper, E. G., Degtyareva, N. P., Corbett, A. H. and Doetsch, P. W. (2016). Identification of SUMO modification sites in the base excision repair protein, Ntg1. DNA Repair (Amst). 48, 51–62.

Sykora, P., Kanno, S., Akbari, M., Kulikowicz, T., Baptiste, B. A., Leandro, G. S., Lu, H., Tian, J., May, A., Becker, K. A., et al. (2017a). DNA polymerase beta participates in mitochondrial DNA repair. Mol. Cell. Biol. MCB.00237–17.

Sykora, P., Kanno, S., Akbari, M., Kulikowicz, T., Baptiste, B. A., Leandro, G. S., Lu, H., Tian, J., May, A., Becker, K. A., et al. (2017b). DNA polymerase beta participates in mitochondrial DNA repair. Mol. Cell. Biol. 37, MCB.00237–17.

Takao, M., Aburatani, H., Kobayashi, K. and Yasui, A. (1998). Mitochondrial targeting of human DNA glycosylases for repair of oxidative DNA damage. Nucleic Acids Res 26, 2917–2922.

Torres-Gonzalez, M., Gawlowski, T., Kocalis, H., Scott, B. T. and Dillmann, W. H. (2014). Mitochondrial 8-oxoguanine glycosylase decreases mitochondrial fragmentation and improves mitochondrial function in H9C2 cells under oxidative stress conditions. AJP Cell Physiol. 306, C221–C229.

Tress, M. L., Abascal, F. and Valencia, A. (2017). Alternative Splicing May Not Be the Key to Proteome Complexity. Trends Biochem. Sci. 42, 98–110.

Trifunovic, Aleksandra, Wredenberg, Anna, Maria Falkenberg1, Johannes N. Spelbrink3, Anja T. Rovio3, Carl E. Bruder4, Mohammad Bohlooly-Y4, Sebastian Gidlo¨ f1, 2, Anders Oldfors5, Rolf Wibom6, Jan To¨ rnell4, Howard T. Jacobs3 & Nils-Go¨ran Larsson, 2 (2004). Premature ageing in mice expressing defective mitochondrial DNA polymerase. Nature 429, 417–423.

Van Damme, P., Gawron, D., Van Criekinge, W. and Menschaert, G. (2014). N-terminal Proteomics and Ribosome Profiling Provide a Comprehensive View of the Alternative Translation Initiation Landscape in Mice and Men. Mol. Cell. Proteomics 13, 1245–1261. van der Kemp, P. A., Charbonnier, J. B., Audebert, M. and Boiteux, S. (2004). Catalytic and DNA-binding properties of the human Ogg1 DNA N-glycosylase/AP lyase: Biochemical exploration of H270, Q315 and F319, three amino acids of the 8-oxoguanine-binding pocket. Nucleic Acids Res. 32, 570–578.

Wang, W., Esbensen, Y., Kunke, D., Suganthan, R., Rachek, L., Bjoras, M. and Eide, L. (2011). Mitochondrial DNA damage level determines neural stem cell differentiation fate. J Neurosci 31, 9746– 9751.

Wieckowski, M. R., Giorgi, C., Lebiedzinska, M., Duszynski, J. and Pinton, P. (2009). Isolation of mitochondria-associated membranes and mitochondria from animal tissues and cells. Nat. Protoc. 4, 1582–1590.

Yakes, F. M. and Van Houten, B. (1997). Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress. Proc Natl Acad Sci U S A 94, 514– 519.

Yogev, O. and Pines, O. (2011). Dual targeting of mitochondrial proteins: Mechanism, regulation and function. Biochim. Biophys. Acta - Biomembr. 1808, 1012–1020.

Zhang, L., Chan, S. S. L. and Wolff, D. J. (2011). Mitochondrial disorders of DNA Polymerase 7 dysfunction: From anatomic to molecular pathology diagnosis. Arch. Pathol. Lab. Med. 135, 925–934.

- 171 -

Table S1. List of oligonucleotides used in this study

Primer name Constrcut Primer sequence α-OGG1F-WT α-OGG1-WT α- 5’-TCG AAT TCG GGCGGTGCTGCTGTGGAAATG CCT GCC (Forward) OGG1-NLS CGC GCG CTT C -3’ α-OGG1F-MTS (Forward) α-OGG1-MTS 5’- T5’- TCG AAT TCG GGCATGGGGCATCGTACTCTAG-3’ α-OGG1R-FLAG α-OGG1-WT α- 5’- G5’- GC GGA TCC CTA CTT ATC GTC GTC ATC CTT (reverse) OGG1-MTS GTA ATC GGC CCT GCC TTC CGG CCC TTT GGA AC-3’ α-OGG1R-NLS- 5’- GC GGA TCC CTA CTT ATC GTC GTC ATC CTT GTA FLAG (reverse) α-OGG1-NLS ATC GGC CCT TGC TGG TGG CTC CTG AGC ATG-3’ α-OGG1F-G12E For Site Directed 5’- CTT CTG CCC AGG CGC ATG GAG CAT CGT ACT CTA (Forward) Mutagenesis GCC TC-3’ α-OGG1R-G12E For Site Directed 5’- GAG GCT AGA GTA CGA TGC TCC ATG CGC CTG GGC (Reverse) Mutagenesis AGA AG-3’ α-OGG1F-K249Q For Site Directed 5'-GGA GTG GGC ACC CAG GTG GCT GAC TGC-3' (Forward) Mutagenesis α-OGG1R-K249Q For Site Directed 5'-GCA GTC AGC CAC CTG GGT GCC CAC TCC -3' (Reverse) Mutagenesis HsOGG1a-Mut silent mutations 5’ – CCAGGTGTGGACGTTGACTCAGACTGAGGAGCAG – 3’ (Forward) siRNA 5’ – HsOGG1a-Mut silent mutations ACGTCCACACCTGGTCAGCTAGTACACCACTCCAGTGTG – (Reverse) siRNA 3’ mtDNA probe F For southern blot 5'-TTACAGTCAAATCCCTTCTCGTCC-3' mtDNA probe R For southern blot 5'-GGATGAGGCAGGAATCAAAGACAG-3'

Table S2. List of antibodies used in Western blotting and Immunofluorescence (IF) experiments.

Protein Company Reference Application Dilution MPPB ptglab 16064-1-AP Western Blot 1/1000 Lamin B abcam ab20396 Western Blot 1/1000 FLAG SIGMA F1804 Western Blot and IF 1/2000 Tubulin α Abgent AJ1034a Western Blot 1/1000 hOGG1-1 abcam ab124741 Western Blot 1/10,000 TOMM22 abcam ab10436 Western Blot 1/1000 TFAM abcam ab155240 Western Blot and IF 1/2000 TOM20 Santa Cruz sc-11415 Western Blot 1/2000 TIM23 SIGMA SAB1100941 Western Blot 1/2000 SDHA abcam ab14715 Western Blot 1/2000 ETFB Proteintech 17925-1-AP Western Blot 1/1000 GARS abcam ab42905 Western Blot 1/1000 CS Proteintech 16131-1-AP Western Blot 1/1000 CENPF abcam Ab5 IF 1:500

- 173 -

Legend to the Figures

Figure 1. αOGG1 is present in both nuclear and mitochondrial compartments. A) Nuclear and mitochondrial fractions were prepared from U2OS cells and analyzed by Western blot using an antibody against alpha-OGG1 Antibodies directed against Lamin B and MPPB were used as markers for the nuclear and mitochondrial extracts respectively. B) Nuclear and mitochondrial fractions were prepared from U2OS cells expressing OGG1-Flag and analyzed by Western blot as in panel A. C) U2OS cells transiently transfected with the construct expressing α-OGG1-Flag were analyzed by Immunofluorescence using an antibody against the FLAG epitope (green). MitoTracker Red was used to stain the mitochondrial network (magenta). DAPI was used to stain nuclear DNA (blue). Scale bars, 5µm. D) Mitochondria purified from U2OS cells were incubated with different concentrations of Proteinase K. When indicated, mitochondria were exposed to 1% of Triton prior to Proteinase K treatment. Extracts were analyzed by Western blot. E) Trypsin sequential digestion assays were performed on purified mitochondria from HEK cells. Where indicated, mitochondria were pre-incubated with digitonin (200 or 500 mg/ml) or exposed to hypotonic shock as described in material and methods. F) Crude mitochondrial extracts were extracted with increasing concentration of NaCl and insoluble (P) and soluble (S) fraction were isolated and analyzed by Western blot.

Figure 2. α-OGG1 is associated with mtDNA in mitochondrial nucleoids. A) Confocal microscopy images of U2OS cells transfected with α-OGG1-FLAG and stained with an antibody against FLAG (red). Mitochondrial DNA is stained with picogreen (green) and the mitochondrial network with MitoTracker Deep Red (magenta). Pearson coefficient (picogreen / OGG1-FLAG) = 0,75, calculated from 8 cells (approximately 1000 nucleoids). Scale bars, 5 µm (2 µm for the insets). B) Confocal microscopy images of U2OS cells transfected with α- OGG1-FLAG and MTS-Turquoise2 and stained with antibodies against FLAG (green) and TFAM (red). MTS- Turquoise2 (cyan) was used to label mitochondrial network. Pearson coefficient (TFAM and OGG1-FLAG) = 0,8, calculated as in A. Scale bars, 5 µm (2 µm for the insets) C) 3D-SIM was used to visualize α-OGG1-FLAG (green), and TFAM (blue) in the mitochondrial network stained with MitoTracker red (red). Scale bars, 2 µm. D) Details of single nucleoids imaged with 3D-SIM. Scale bar, 0,5 µm. E) Iodixanol gradient fractions of Mitochondria isolated from HEK cells were analyzed by southern blot using a probe against mitochondrial DNA (upper panel) and Western blots using antibodies against TFAM, GARS and α-OGG1. Fractions 1-4 were blank in all cases and are not presented.

Figure 3. α-OGG1 mitochondria-targeting signal (MTS) and nuclear localization signal (NLS) are functional and determine the subcellular localization of the protein. A) Alignment of the N-terminal sequences of OGG1 from human (α-OGG1 isoform), mouse and yeast showing the position of two AUG in frame both human and mouse. The sequence of the MTS is indicated. B) Prediction for mitochondrial import calculated with different softwares considering initiation of translation at the first or the second AUG for human and mouse OGG1 proteins. C) schematic representation of the WT α-OGG1 or truncations of the MTS (MTS; initiation of translation at AUG2) or the NLS (NLS) that were fused to the FLAG epitope. D) The subcellular localization of different proteins was analyzed by confocal microscopy using antibodies against the FLAG (green). Nuclear DNA is stained with DAPI (blue) and the mitochondrial network with MitoTracker Red (red). E) Percentage of cells with a detectable α-OGG1 staining in the mitochondria were quantified. 350 cells were

counted for each construct and the experiment was repeated three times. Scale bars, 5 µm and 2 µm in the insets.

Figure 4. Mitochondrial localization of OGG1 is not dependent on the cell cycle phase or mtDNA replication. A) α-OGG1-FLAG was visualized using an anti-FLAG (green), and mitochondrial network was stained with MTS-Turquoise2 (cyan). Cell cycle phases were discriminated by using incorporation of EdU (labels S phase; purple) and an antibody against CENPF (labels G2 phase; Red). The cells negative for both EdU and CENPF are the cells in G1. Scale bar: 5 µm. Higher magnification images of the boxed regions, are presented for the merge MTS-Turquoise and OGG1-FLAG, scale bar: 2 µm. B) The percentage of cells with detectable α-OGG1 in mitochondria were determined in the different cell cycle phases. Approximately 500 transfected cells were analyzed from two independent experiments. C) Replicating mtDNA was labelled by incorporation of EdU (purple), nucleoids visualised with an antibody against TFAM (red) and α-OGG1-FLAG stained using an anti-FLAG (green). Mitochondrial network was stained with MTS-Turquoise2. Filled and broken arrows indicate replicating and non-replicating nucleoids, respectively. The association of α-OGG1 with nucleoids could be detected in both conditions. Scale bar: 2 µm. D) TFAM was used as a marker for nucleoids and the % of nucleoids with a positive signal for α-OGG1-FLAG was determined in both replicating (EdU positive) and non replicating (EdU negative) populations. More than 500 nucleoids were analyzed from 4 cells.

Figure 5. Menadione treatment induces mitochondrial dysfunction in in OGG1-deficient cells. A) U2OS cells were transfected with a siRNA against OGG1 or against a control sequence. 72 hours after transfection, protein extracts were prepared and analyzed by Western Blot using antibodies against α-OGG1. B) Effect of Menadione-treatment on the level of mitochondrial superoxide, as detected with the fluorigenic substrate MitoSOX. MitoSOX fluorescence distributions were measured for U2OS siControl (top panel) and U2OS siOGG1 (bottom panel) cell lines, in the absence (grey) or presence (red) of 50 μM Menadione. Probability binning distribution difference analysis was used to quantify the effect of Menadione treatment. Chi-squared based T(X) metrics were calculated for Menadione treated samples as compared to their untreated counterparts, and are indicated in each panel. Results are representative of several independent experiments. D) Effect of Menadione-treatment on mitochondrial membrane potential and mitochondrial mass, as detected with TMRE and MitoTracker Green, respectively. TMRE and MitoTracker Green fluorescence was measured for U2OS siControl (top panels) and U2OS siOGG1 (bottom panels) cell lines, in the absence (left panels) or presence (right panels) of 50 μM Menadione. E) TMRE and MitoTracker Green fluorescence distributions were plotted as univariate histograms for U2OS siControl (top panels) and U2OS siOGG1 (bottom panels) cell lines, in the absence (grey) or presence (red) of 50 μM Menadione. Probability binning distribution difference analysis was used to quantify the effect of Menadione treatment on mitochondrial membrane potential. Chi-squared based T(X) metrics were calculated for Menadione treated samples as compared to their untreated counterparts, and are indicated in each panel. Results are representative of several independent experiments. F) Living cells transfected with siControl or siOGG1 and stained with MitoTracker green (green) and TMRE (red) were imaged on confocal microscopy. Representative images obtained at different times after exposure to 50 µM menadione for 1 hour are shown. Scale bar, 5 µm.

- 175 -

Figure 6. Mitochondrial dysfunction of OGG1-deficient cells can be complemented with the WT form of α-OGG1 but not with the α-OGG1(MTS) mutant. A) 8-oxoG glycosylase activity was measured in whole cell extracts purified from HeLa cells transiently transfected with plasmids encoding for the WT version of α- OGG1-FLAG or the deletion mutant α-OGG1(MTS)-FLAG. B) Western blots on extracts from U2OS cell lines expressing either α-OGG1-FLAG or α-OGG1(MTS)-FLAG in which silent mutations have been introduced into the sequence targeted by the siRNA. Protein levels are not affected by the transfection with siRNA against OGG1 resulting in the silencing of the endogenous protein. C) wild type U2OS or cells stably expressing α- OGG1(WT)-FLAG or α-OGG1(MTS)-FLAG were transfected with siRNA against OGG1 72 hours before being exposed to menadione. Mitochondrial ROS production from NT and treated cells was measured in a plate reader using the superoxide indicator MitoSOX Red. Values correspond to the ratio MitoSOX / Hoechst signals. D) The same experimental design as in C was used to evaluate mitochondrial membrane potential using TMRE as a probe. MitoTracker Green was used as a marker of mitochondrial mass and the results displayed as a ratio TMRE/MitoTracker Green. For both C and D measurements were performed continuously over a period of 90 minutes after Menadione exposure and the values obtained at 60 minutes are displayed in the graph.

Figure 7. Overexpression of the WT form of α-OGG1 but not the α-OGG1(MTS) or α-OGG1(K249Q) mutants protects mitochondria from Menadione-induced fragmentation. A) U2OS cells transiently expressing different variants of α-OGG1-FLAG, were exposed or not (NT) to Menadione. Transfected cells were stained using antibodies against FLAG (red). The mitochondrial network was stained with MitoTracker Red (purple). Mitochondrial morphology was visualized by confocal microscopy. Only images corresponding to cells transfected with the WT form of α-OGG1 are presented. B) Percentage of cells expressing the WT or the different mutants of α-OGG1 (MTS and K249Q) presenting a fragmented mitochondrial network. At least 50 cells were counted. Bars indicate the Mean with SD from three independent experiments. Statistical significance between the different populations was evaluated with Tukey HSD test, p<0,0001. C) Binarized images were obtained from the mitochondrial red staining. Representative cells expressing each of the α- OGG1-FLAG variants, exposed or not to Menadione, are presented. D) After image segmentation, mitochondrial morphology parameters Aspect ratio (AR) and form factor (FF) were determined for each mitochondria. Dots represents the values for non-transfected cells (black) and for cells overexpressing the WT α-OGG1-FLAG (grey). More than 2000 mitochondria from 8 independent cells treated with menadione are displayed for each population. E) Frequency distribution analysis of single mitochondria according to FF (E) and AR (F) morphological parameters. The analysis was performed for at least 2000 single mitochondria from 8 cells expressing each of the variants. Images were acquired from two independent experiments. Bars indicate Mean +/- SD. Statistical significance between the different populations for the different bins was evaluated with Tukey HSD test, p<0,0001. Scale bars, 5 µm.

Supplementary Figure 1. Conservation of uAUG in the 5’ UTR region of α-OGG1. A) First exons containing full or nearly full 5’UTR sequences from vertebrate OGG1 sequences available in Ensembl were multialigned with ClustalW2 (www.ebi.ac.uk/Tools/msa/clustalw2). The two regions containing the starting codons AUG1 and AUG2 are shown. B) The degree of conservation and strength of Kozak sequence was investigated using WebLogo (http://weblogo.berkeley.edu/logo.cgi). Sequence logo was calculated for the

heptanucleotide surrounding the AUG triplets from position -3 to +4. AUG1 has a G in position -3 in some species but is still fairly leaky. A stronger Kozak sequence was find for AUG2, due to an A in -3 position and a G in +1 position.

Supplementary Figure 2. Subcellular localization of the α-OGG1(WT) and of the mutant α- OGG1(K249Q). MitoTracker RED is used to stain mitochondrial network (red) and OGG1-FLAG is visualized with an antibody anti-FLAG (green). Scale bar, 5 µm and 2 µm in the insets.

- 177 -

- 179 -

H. sapiens M. musculus S. cervisiae

- 181 -

- 183 -

Supplementary Figure 1.

- 185 -

Supplementary Figure 2.

Bibliography

- 187 -

Bibliography 1. Abramova, N. E., Davies, K. J. & Crawford, D. R. Polynucleotide degradation during early stage response to oxidative stress is specific to mitochondria. Free Radic. Biol. Med. 28, 281–288 (2000). 2. Acin-Perez, R., Fernandez-Silva, P., Peleato, M. L., Perez-Martos, A. & Enriquez, J. A. Respiratory active mitochondrial supercomplexes. Mol. Cell 32, 529–539 (2008). 3. Akbari, M., Visnes, T., Krokan, H. E. & Otterlei, M. Mitochondrial base excision repair of uracil and AP sites takes place by single-nucleotide insertion and long-patch DNA synthesis. DNA Repair (Amst). 7, 605–616 (2008). 4. Akepati, V. R. et al. Characterization of OPA1 isoforms isolated from mouse tissues. J. Neurochem. 106, 372–383 (2008). 5. Alam, T. I. et al. Human mitochondrial DNA is packaged with TFAM. Nucleic Acids Res. 31, 1640–1645 (2003). 6. Aldridge, J. E., Horibe, T. & Hoogenraad, N. J. Discovery of genes activated by the mitochondrial unfolded protein response (mtUPR) and cognate promoter elements. PLoS One 2, e874 (2007). 7. Alexeyev, M. F. et al. Selective elimination of mutant mitochondrial genomes as therapeutic strategy for the treatment of NARP and MILS syndromes. Gene Ther. 15, 516–523 (2008). 8. Alexeyev, M., Shokolenko, I., Wilson, G. & LeDoux, S. The maintenance of mitochondrial DNA integrity--critical analysis and update. Cold Spring Harb. Perspect. Biol. 5, a012641 (2013). 9. Altmann, K. & Westermann, B. Role of essential genes in mitochondrial morphogenesis in Saccharomyces cerevisiae. Mol. Biol. Cell 16, 5410–5417 (2005). 10. Alston, C. L., Rocha, M. C., Lax, N. Z., Turnbull, D. M., Taylor, R. W., The genetics and pathology of mitochondrial disease J Pathol. 241(2): 236–250 (2017). 11. Ambro, L., Pevala, V., Bauer, J. & Kutejova, E. The influence of ATP-dependent proteases on a variety of nucleoid-associated processes. J. Struct. Biol. 179, 181–192 (2012). 12. Anderson, C. T. & Friedberg, E. C. The presence of nuclear and mitochondrial uracil-DNA glycosylase in extracts of human KB cells. Nucleic Acids Res. 8, 875–888 (1980). 13. Anderson, S. et al. Sequence and organization of the human mitochondrial genome. Nature 290, 457–465 (1981). 14. Antes, A. et al. Differential regulation of full-length genome and a single-stranded 7S DNA along the cell cycle in human mitochondria. Nucleic Acids Res. 38, 6466–6476 (2010). 15. Arai, T., Kelly, V. P., Minowa, O., Noda, T. & Nishimura, S. The study using wild-type and Ogg1 knockout mice exposed to potassium bromate shows no tumor induction despite an extensive accumulation of 8-hydroxyguanine in kidney DNA. Toxicology 221, 179–186 (2006). 16. Araujo, P. R. et al. Before It Gets Started: Regulating Translation at the 5’ UTR. Comp. Funct. Genomics 2012, 475731 (2012). 17. Arur, S. et al. Annexin I is an endogenous ligand that mediates apoptotic cell engulfment. Dev. Cell 4, 587–598 (2003).

- 189 -

18. Audebert, M., Radicella, J. P. & Dizdaroglu, M. Effect of single mutations in the OGG1 gene found in human tumors on the substrate specificity of the Ogg1 protein. Nucleic Acids Res. 28, 2672– 2678 (2000). 19. Audebert, M., Charbonnier, J. B., Boiteux, S. & Radicella, J. P. Mitochondrial targeting of human 8-oxoguanine DNA glycosylase hOGG1 is impaired by a somatic mutation found in kidney cancer. DNA Repair (Amst). 1, 497–505 (2002). 20. Bacman, S. R., Williams, S. L. & Moraes, C. T. Intra- and inter-molecular recombination of mitochondrial DNA after in vivo induction of multiple double-strand breaks. Nucleic Acids Res. 37, 4218–4226 (2009). 21. Bacsi, A., Chodaczek, G. & Hazra, T. Increased ROS generation in subsets of OGG1 knockout fibroblast cells. Mech. ageing … 128, 637–649 (2007). 22. Banerjee, A., Yang, W., Karplus, M. & Verdine, G. L. Structure of a repair enzyme interrogating undamaged DNA elucidates recognition of damaged DNA. Nature 434, 612–618 (2005). 23. Barnes, D. E. & Lindahl, T. Repair and genetic consequences of endogenous DNA base damage in mammalian cells. Annu. Rev. Genet. 38, 445–476 (2004). 24. Bazykin, G. A. & Kochetov, A. V. Alternative translation start sites are conserved in eukaryotic genomes. Nucleic Acids Res. 39, 567–577 (2011). 25. Bernardini, J. P., Lazarou, M. & Dewson, G. Parkin and mitophagy in cancer. Oncogene 36, 1315–1327 (2017). 26. Bernier, F. P. et al. Diagnostic criteria for respiratory chain disorders in adults and children. Neurology 59, 1406–1411 (2002). 27. Berquist, B. R., Canugovi, C., Sykora, P., Wilson, D. M. 3rd & Bohr, V. A. Human Cockayne syndrome B protein reciprocally communicates with mitochondrial proteins and promotes transcriptional elongation. Nucleic Acids Res. 40, 8392–8405 (2012). 28. Blainey, P. C., van Oijen, A. M., Banerjee, A., Verdine, G. L. & Xie, X. S. A base-excision DNA- repair protein finds intrahelical lesion bases by fast sliding in contact with DNA. Proc. Natl. Acad. Sci. U. S. A. 103, 5752–5757 (2006). 29. Boesch, P., Ibrahim, N., Dietrich, A. & Lightowlers, R. N. Membrane association of mitochondrial DNA facilitates base excision repair in mammalian mitochondria. Nucleic Acids Res. 38, 1478–1488 (2010). 30. Boesch, P. et al. DNA repair in organelles: Pathways, organization, regulation, relevance in disease and aging. Biochim. Biophys. Acta 1813, 186–200 (2011). 31. Bogenhagen, D. F., Rousseau, D., Burke, S. The Layered Structure of Human Mitochondrial DNA Nucleoids. The Journal of Biological Chemistry 283, 3665-3675 (2008). 32. Boiteux, S., Coste, F. & Castaing, B. Repair of 8-oxo-7,8-dihydroguanine in prokaryotic and eukaryotic cells: Properties and biological roles of the Fpg and OGG1 DNA N-glycosylases. Free Radic. Biol. Med. 107, 179–201 (2017). 33. Bosshard, M., Markkanen, E. & van Loon, B. Base excision repair in physiology and pathology of the central nervous system. Int. J. Mol. Sci. 13, 16172–16222 (2012).

34. Bratton, D. L. et al. Appearance of phosphatidylserine on apoptotic cells requires calcium- mediated nonspecific flip-flop and is enhanced by loss of the aminophospholipid translocase. J. Biol. Chem. 272, 26159–26165 (1997). 35. Bravard, A. et al. Oxidation status of human OGG1-S326C polymorphic variant determines cellular DNA repair capacity. Cancer Res. 69, 3642–3649 (2009). 36. Brown, T. a et al. Superresolution fluorescence imaging of mitochondrial nucleoids reveals their spatial range, limits, and membrane interaction. Mol. Cell. Biol. 31, 4994–5010 (2011). 37. Bruner, S. D., Norman, D. P. & Verdine, G. L. Structural basis for recognition and repair of the endogenous mutagen 8-oxoguanine in DNA. Nature 403, 859–866 (2000). 38. Busch, K. B., Kowald, A. & Spelbrink, J. N. Quality matters: how does mitochondrial network dynamics and quality control impact on mtDNA integrity? Philos. Trans. R. Soc. Lond. B. Biol. Sci. 369, 20130442 (2014). 39. Campalans, A., Marsin, S., Nakabeppu, Y., O’Connor, T. R., Boiteux, S., Radicella, J. P., XRCC1 interactions with multiple DNA glycosylases: A model for its recruitment to base excision repair DNA Repair (Amst). 12;4(7):826-35 (2005) 40. Canugovi, C. et al. The mitochondrial transcription factor A functions in mitochondrial base excision repair. DNA Repair (Amst). 9, 1080–1089 (2010). 41. Carroll, R. G., Hollville, E. & Martin, S. J. Parkin sensitizes toward apoptosis induced by mitochondrial depolarization through promoting degradation of Mcl-1. Cell Rep. 9, 1538–1553 (2014). 42. Chainsaw A, Koehler CM, Milenkovic D, Lithgow T, Pfanner N Importing mitochondrial proteins: machineries and mechanisms. Cell 138: 628– 644 (2009). 43. Chatre, L. & Ricchetti, M. Prevalent coordination of mitochondrial DNA transcription and initiation of replication with the cell cycle. Nucleic Acids Res. 41, 3068–3078 (2013). 44. Chatterjee, A., Mambo, E. & Sidransky, D. Mitochondrial DNA mutations in human cancer. Oncogene 25, 4663–4674 (2006). 45. Chatterjee, A., Dasgupta, S. & Sidransky, D. Mitochondrial subversion in cancer. Cancer Prev. Res. (Phila). 4, 638–654 (2011). 46. Chatterjee, A., Mambo, E., Zhang, Y., Deweese, T. & Sidransky, D. Targeting of mutant hogg1 in mammalian mitochondria and nucleus: effect on cellular survival upon oxidative stress. BMC Cancer 6, 235 (2006). 47. Chattopadhyay, R. et al. Identification and characterization of mitochondrial abasic (AP)- endonuclease in mammalian cells. Nucleic Acids Res. 34, 2067–2076 (2006). 48. Cheadle, J. P. & Sampson, J. R. Exposing the MYtH about base excision repair and human inherited disease. Hum. Mol. Genet. 12 Spec No, R159-65 (2003). 49. Chen, H., Chomyn, A. & Chan, D. C. Disruption of fusion results in mitochondrial heterogeneity and dysfunction. J. Biol. Chem. 280, 26185–26192 (2005). 50. Chen, H., McCaffery, J. M. & Chan, D. C. Mitochondrial fusion protects against neurodegeneration in the cerebellum. Cell 130, 548–562 (2007).

- 191 -

51. Chen, S.-K. et al. Age-associated decrease of oxidative repair enzymes, human 8-oxoguanine DNA glycosylases (hOgg1), in human aging. J. Radiat. Res. 44, 31–35 (2003). 52. Chen, Y. & Dorn, G. W. 2nd. PINK1-phosphorylated mitofusin 2 is a Parkin receptor for culling damaged mitochondria. Science 340, 471–475 (2013). 53. Choi, S.-Y. et al. A common lipid links Mfn-mediated mitochondrial fusion and SNARE- regulated exocytosis. Nat. Cell Biol. 8, 1255–1262 (2006). 54. Chourasia, A. H., Boland, M. L. & Macleod, K. F. Mitophagy and cancer. Cancer Metab. 3, 4 (2015). 55. Chourasia, A. H. et al. Mitophagy defects arising from BNip3 loss promote mammary tumor progression to metastasis. EMBO Rep. 16, 1145–1163 (2015). 56. Chung, Y. W., Kim, H. K., Kim, I. Y., Yim, M. B. & Chock, P. B. Dual function of protein kinase C (PKC) in 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced manganese superoxide dismutase (MnSOD) expression: activation of CREB and FOXO3a by PKC-alpha phosphorylation and by PKC- mediated inactivation of Akt, respective. J. Biol. Chem. 286, 29681–29690 (2011). 57. Crawford, D. R., Abramova, N. E. & Davies, K. J. Oxidative stress causes a general, calcium- dependent degradation of mitochondrial polynucleotides. Free Radic. Biol. Med. 25, 1106–1111 (1998). 58. Crawford, D. R., Wang, Y., Schools, G. P., Kochheiser, J. & Davies, K. J. Down-regulation of mammalian mitochondrial RNAs during oxidative stress. Free Radic. Biol. Med. 22, 551–559 (1997). 59. Dammak, M., Shirazi-Adl, A. & Zukor, D. J. Analysis of cementless implants using interface nonlinear friction--experimental and finite element studies. J. Biomech. 30, 121–129 (1997). 60. Davuluri, R. V, Suzuki, Y., Sugano, S. & Zhang, M. Q. CART classification of human 5’ UTR sequences. Genome Res. 10, 1807–1816 (2000). 61. de Souza-Pinto, N. C. et al. Repair of 8-oxodeoxyguanosine lesions in mitochondrial dna depends on the oxoguanine dna glycosylase (OGG1) gene and 8-oxoguanine accumulates in the mitochondrial dna of OGG1-defective mice. Cancer Res. 61, 5378–5381 (2001). 62. de Souza-Pinto, N. C., Wilson, D. M. 3rd, Stevnsner, T. V & Bohr, V. A. Mitochondrial DNA, base excision repair and neurodegeneration. DNA Repair (Amst). 7, 1098–1109 (2008). 63. de Souza-Pinto, N. C., et al. Novel DNA mismatch repair activity involving YB-1 in human mitochondria DNA Repair (Amst). 8(6): 704–719 (2009) 64. Delivani, P., Adrain, C., Taylor, R. C., Duriez, P. J. & Martin, S. J. Role for CED-9 and Egl-1 as regulators of mitochondrial fission and fusion dynamics. Mol. Cell 21, 761–773 (2006). 65. Denver, D. R., Swenson, S. L. & Lynch, M. An evolutionary analysis of the helix-hairpin-helix superfamily of DNA repair glycosylases. Mol. Biol. Evol. 20, 1603–1611 (2003). 66. Detmer, S. A. & Chan, D. C. Complementation between mouse Mfn1 and Mfn2 protects mitochondrial fusion defects caused by CMT2A disease mutations. J. Cell Biol. 176, 405–414 (2007). 67. Detmer, S. A. & Chan, D. C. Functions and dysfunctions of mitochondrial dynamics. Nat. Rev. Mol. Cell Biol. 8, 870–879 (2007).

68. Dezor, M. et al. Expression of 8-oxoguanine DNA glycosylase 1 (OGG1) and the level of p53 and TNF-αlpha proteins in peripheral lymphocytes of patients with Alzheimer’s disease. Folia Neuropathol. 49, 123–131 (2011). 69. Dhenaut, A., Boiteux, S. & Radicella, J. P. Characterization of the hOGG1 promoter and its expression during the cell cycle. Mutat. Res. 461, 109–118 (2000). 70. Dianova, I. I., Bohr, V. A. & Dianov, G. L. Interaction of human AP endonuclease 1 with flap endonuclease 1 and proliferating cell nuclear antigen involved in long-patch base excision repair. Biochemistry 40, 12639–12644 (2001). 71. DiMauro, S., Bonilla, E. & De Vivo, D. C. Does the patient have a mitochondrial encephalomyopathy? J. Child Neurol. 14 Suppl 1, S23-35 (1999). 72. Dimauro, S., Tay, S. & Mancuso, M. Mitochondrial encephalomyopathies: diagnostic approach. Ann. N. Y. Acad. Sci. 1011, 217–231 (2004). 73. Dobson, A. W., Xu, Y., Kelley, M. R., LeDoux, S. P. & Wilson, G. L. Enhanced mitochondrial DNA repair and cellular survival after oxidative stress by targeting the human 8-oxoguanine glycosylase repair enzyme to mitochondria. J. Biol. Chem. 275, 37518–37523 (2000). 74. Dobson, A. W. et al. Enhanced mtDNA repair capacity protects pulmonary artery endothelial cells from oxidant-mediated death. Am. J. Physiol. Lung Cell. Mol. Physiol. 283, L205-10 (2002). 75. Dodson, M. L. & Lloyd, R. S. Mechanistic comparisons among base excision repair glycosylases. Free Radic. Biol. Med. 32, 678–682 (2002). 76. Druzhyna, N. M., Hollensworth, S. B., Kelley, M. R., Wilson, G. L. & Ledoux, S. P. Targeting human 8-oxoguanine glycosylase to mitochondria of oligodendrocytes protects against menadione- induced oxidative stress. Glia 42, 370–378 (2003). 77. Druzhyna, N. M., Wilson, G. L. & LeDoux, S. P. Mitochondrial DNA repair in aging and disease. Mech. Ageing Dev. 129, 383–390 (2008). 78. Durieux, J., Wolff, S. & Dillin, A. The cell-non-autonomous nature of electron transport chain- mediated longevity. Cell 144, 79–91 (2011). 79. Ekstrand, M. I. et al. Mitochondrial transcription factor A regulates mtDNA copy number in mammals. Hum. Mol. Genet. 13, 935–944 (2004). 80. ERNSTER, L., IKKOS, D. & LUFT, R. Enzymic activities of human skeletal muscle mitochondria: a tool in clinical metabolic research. Nature 184, 1851–1854 (1959). 81. Falkenberg, M., Larsson, N.-G. & Gustafsson, C. M. DNA replication and transcription in mammalian mitochondria. Annu. Rev. Biochem. 76, 679–699 (2007). 82. Fischer, F., Hamann, A. & Osiewacz, H. D. Mitochondrial quality control: an integrated network of pathways. Trends Biochem. Sci. 37, 284–292 (2012). 83. Frank, S. et al. The role of dynamin-related protein 1, a mediator of mitochondrial fission, in apoptosis. Dev. Cell 1, 515–525 (2001). 84. Frezza, C. et al. OPA1 controls apoptotic cristae remodeling independently from mitochondrial fusion. Cell 126, 177–189 (2006).

- 193 -

85. Friedberg, E. C. Suffering in silence: the tolerance of DNA damage. Nat. Rev. Mol. Cell Biol. 6, 943–953 (2005). 86. Friedman, J. R. & Nunnari, J. Mitochondrial form and function. Nature 505, 335–343 (2014). 87. Friedman, J. I. & Stivers, J. T. Detection of damaged DNA bases by DNA glycosylase enzymes. Biochemistry 49, 4957–4967 (2010). 88. Fritsch, E. S., Chabbert, C. D., Klaus, B. & Steinmetz, L. M. A genome-wide map of mitochondrial DNA recombination in yeast. Genetics 198, 755–771 (2014). 89. Fukae, J. et al. Expression of 8-oxoguanine DNA glycosylase (OGG1) in Parkinson’s disease and related neurodegenerative disorders. Acta Neuropathol. 109, 256–262 (2005). 90. Fukui, H. & Moraes, C. T. Mechanisms of formation and accumulation of mitochondrial DNA deletions in aging neurons. Hum. Mol. Genet. 18, 1028–1036 (2009). 91. Furda, A. M., Marrangoni, A. M., Lokshin, A. & Van Houten, B. Oxidants and not alkylating agents induce rapid mtDNA loss and mitochondrial dysfunction. DNA Repair (Amst). 11, 684–692 (2012). 92. Furda, A., Santos, J. H., Meyer, J. N. & Van Houten, B. Quantitative PCR-based measurement of nuclear and mitochondrial DNA damage and repair in mammalian cells. Methods Mol. Biol. 1105, 419–437 (2014). 93. Furihata, C. An active alternative splicing isoform of human mitochondrial 8-oxoguanine DNA glycosylase (OGG1). Genes Environ. Off. J. Japanese Environ. Mutagen Soc. 37, 21 (2015). 94. Gabaldon, T. & Huynen, M. A. From endosymbiont to host-controlled organelle: the hijacking of mitochondrial protein synthesis and metabolism. PLoS Comput. Biol. 3, e219 (2007). 95. Gad, H. et al. MTH1 inhibition eradicates cancer by preventing sanitation of the dNTP pool. Nature 508, 215–221 (2014). 96. Gao, Y. et al. DNA ligase III is critical for mtDNA integrity but not Xrcc1-mediated nuclear DNA repair. Nature 471, 240–244 (2011). 97. Garcia, I., Jones, E., Ramos, M., Innis-Whitehouse, W. & Gilkerson, R. The little big genome: the organization of mitochondrial DNA. Front. Biosci. (Landmark Ed. 22, 710–721 (2017). 98. Garcia-Gomez, S. et al. PrimPol, an archaic primase/polymerase operating in human cells. Mol. Cell 52, 541–553 (2013). 99. Garcia-Saez, A. J. The secrets of the Bcl-2 family. Cell Death Differ. 19, 1733–1740 (2012). 100. Garedew, A., Andreassi, C. & Moncada, S. Mitochondrial dynamics, biogenesis, and function are coordinated with the cell cycle by APC/C CDH1. Cell Metab. 15, 466–479 (2012). 101. Gatica, R. et al. Over-expression of muscle glycogen synthase in human diabetic nephropathy. Histochem. Cell Biol. 143, 313–324 (2015). 102. Gavande, N. S. et al. DNA repair targeted therapy: The past or future of cancer treatment? Pharmacol. Ther. 160, 65–83 (2016). 103. Geisler, S. et al. PINK1/Parkin-mediated mitophagy is dependent on VDAC1 and p62/SQSTM1. Nat. Cell Biol. 12, 119–131 (2010).

104. Gerhold, J. M. et al. Human Mitochondrial DNA-Protein Complexes Attach to a Cholesterol- Rich Membrane Structure. Sci. Rep. 5, 15292 (2015). 105. Germain, M., Mathai, J. P., McBride, H. M. & Shore, G. C. Endoplasmic reticulum BIK initiates DRP1-regulated remodelling of mitochondrial cristae during apoptosis. EMBO J. 24, 1546–1556 (2005). 106. Giles, R. E., Blanc, H., Cann, H. M. & Wallace, D. C. Maternal inheritance of human mitochondrial DNA. Proc. Natl. Acad. Sci. U. S. A. 77, 6715–6719 (1980). 107. Gilkerson, R. W., Schon, E. A., Hernandez, E. & Davidson, M. M. Mitochondrial nucleoids maintain genetic autonomy but allow for functional complementation. J. Cell Biol. 181, 1117–1128 (2008). 108. Gilkerson, R. et al. The mitochondrial nucleoid: integrating mitochondrial DNA into cellular homeostasis. Cold Spring Harb. Perspect. Biol. 5, a011080 (2013). 109. Gillis, L. A. & Sokol, R. J. Gastrointestinal manifestations of mitochondrial disease. Gastroenterol. Clin. North Am. 32, 789–817, v (2003). 110. Gorman, G. S. et al. Prevalence of nuclear and mitochondrial DNA mutations related to adult mitochondrial disease. Ann. Neurol. 77, 753–759 (2015). 111. Graziewicz, M. A., Bienstock, R. J. & Copeland, W. C. The DNA polymerase gamma Y955C disease variant associated with PEO and parkinsonism mediates the incorporation and translesion synthesis opposite 7,8-dihydro-8-oxo-2’-deoxyguanosine. Hum. Mol. Genet. 16, 2729–2739 (2007). 112. Gredilla, R., Bohr, V. A. & Stevnsner, T. Mitochondrial DNA repair and association with aging- -an update. Exp. Gerontol. 45, 478–488 (2010). 113. Gredilla, R., Garm, C., Holm, R., Bohr, V. A. & Stevnsner, T. Differential age-related changes in mitochondrial DNA repair activities in mouse brain regions. Neurobiol. Aging 31, 993–1002 (2010). 114. Greenleaf, A. L., Kelly, J. L. & Lehman, I. R. Yeast RPO41 gene product is required for transcription and maintenance of the mitochondrial genome. Proc. Natl. Acad. Sci. U. S. A. 83, 3391– 3394 (1986). 115. Griffin, E. E., Graumann, J. & Chan, D. C. The WD40 protein Caf4p is a component of the mitochondrial fission machinery and recruits Dnm1p to mitochondria. J. Cell Biol. 170, 237–248 (2005). 116. Griffiths, L. M. et al. Dynamic compartmentalization of base excision repair proteins in response to nuclear and mitochondrial oxidative stress. Mol. Cell. Biol. 29, 794–807 (2009). 117. Grishko, V. I., Rachek, L. I., Spitz, D. R., Wilson, G. L. & LeDoux, S. P. Contribution of mitochondrial DNA repair to cell resistance from oxidative stress. J. Biol. Chem. 280, 8901–8905 (2005). 118. Gu, A., Ji, G., Yan, L. & Zhou, Y. The 8-oxoguanine DNA glycosylase 1 (ogg1) decreases the vulnerability of the developing brain to DNA damage. DNA Repair (Amst). 12, 1094–1104 (2013). 119. Hacker, G. The morphology of apoptosis. Cell Tissue Res. 301, 5–17 (2000). 120. Hales, K. G. & Fuller, M. T. Developmentally regulated mitochondrial fusion mediated by a conserved, novel, predicted GTPase. Cell 90, 121–129 (1997). - 195 -

121. Halsne, R. et al. Lack of the DNA glycosylases MYH and OGG1 in the cancer prone double mutant mouse does not increase mitochondrial DNA mutagenesis. DNA Repair (Amst). 11, 278–285 (2012). 122. Hancock, M. R. Mitochondrial dysfunction and the role of the non-specialist laboratory. Ann. Clin. Biochem. 39, 456–463 (2002). 123. Hanssen-Bauer, A. et al. XRCC1 coordinates disparate responses and multiprotein repair complexes depending on the nature and context of the DNA damage. Environ. Mol. Mutagen. 52, 623–635 (2011). 124. Hashiguchi, K., Stuart, J. a, de Souza-Pinto, N. C. & Bohr, V. a. The C-terminal alphaO helix of human Ogg1 is essential for 8-oxoguanine DNA glycosylase activity: the mitochondrial beta-Ogg1 lacks this domain and does not have glycosylase activity. Nucleic Acids Res. 32, 5596–608 (2004). 125. Hashiguchi, K., Bohr, V. a & de Souza-Pinto, N. C. Oxidative stress and mitochondrial DNA repair: implications for NRTIs induced DNA damage. Mitochondrion 4, 215–22 (2004). 126. Hegde, M. L. et al. Oxidative genome damage and its repair: implications in aging and neurodegenerative diseases. Mech. Ageing Dev. 133, 157–168 (2012). 127. Helleday, T., Petermann, E., Lundin, C., Hodgson, B. & Sharma, R. A. DNA repair pathways as targets for cancer therapy. Nat. Rev. Cancer 8, 193–204 (2008). 128. Hermann, G. J. et al. Mitochondrial fusion in yeast requires the transmembrane GTPase Fzo1p. J. Cell Biol. 143, 359–373 (1998). 129. Hill, J. W. & Evans, M. K. A novel R229Q OGG1 polymorphism results in a thermolabile enzyme that sensitizes KG-1 leukemia cells to DNA damaging agents. Cancer Detect. Prev. 31, 237– 243 (2007). 130. Hiona, A. et al. Mitochondrial DNA mutations induce mitochondrial dysfunction, apoptosis and sarcopenia in skeletal muscle of mitochondrial DNA mutator mice. PLoS One 5, e11468 (2010). 131. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D. & Korsmeyer, S. J. Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336 (1990). 132. Hollensworth S.B., Shen C., Sim J.E., Spitz D.R., Wilson G.L., LeDoux S.P. Glial cell type-specific responses to menadione-induced oxidative stress. Free Radic Biol Med.; 28(8):1161-74 (2000) 133. Hollville, E., Carroll, R. G., Cullen, S. P. & Martin, S. J. Bcl-2 family proteins participate in mitochondrial quality control by regulating Parkin/PINK1-dependent mitophagy. Mol. Cell 55, 451– 466 (2014). 134. Holt, I. J. et al. Mitochondrial myopathies: clinical and biochemical features of 30 patients with major deletions of muscle mitochondrial DNA. Ann. Neurol. 26, 699–708 (1989). 135. Holt, I. J. Mitochondrial DNA replication and repair: all a flap. Trends Biochem. Sci. 34, 358– 365 (2009). 136. Hoppins, S., Lackner, L. & Nunnari, J. The machines that divide and fuse mitochondria. Annu. Rev. Biochem. 76, 751–780 (2007). 137. Houtkooper, R. H. et al. Mitonuclear protein imbalance as a conserved longevity mechanism. Nature 497, 451–457 (2013).

138. Hu, J. et al. Repair of formamidopyrimidines in DNA involves different glycosylases: role of the OGG1, NTH1, and NEIL1 enzymes. J. Biol. Chem. 280, 40544–40551 (2005). 139. Huber, K. V. M. et al. Stereospecific targeting of MTH1 by (S)-crizotinib as an anticancer strategy. Nature 508, 222–227 (2014). 140. Hyun, J. W. et al. Leukemic cell line, KG-1 has a functional loss of hOGG1 enzyme due to a point mutation and 8-hydroxydeoxyguanosine can kill KG-1. Oncogene 19, 4476–4479 (2000). 141. Imaizumi, N., Kwang Lee, K., Zhang, C. & Boelsterli, U. A. Mechanisms of cell death pathway activation following drug-induced inhibition of mitochondrial complex I. Redox Biol. 4, 279–288 (2015). 142. Ingerman, E. et al. Dnm1 forms spirals that are structurally tailored to fit mitochondria. J. Cell Biol. 170, 1021–1027 (2005). 143. Jackson, E. B., Theriot, C. A., Chattopadhyay, R., Mitra, S. & Izumi, T. Analysis of nuclear transport signals in the human apurinic/apyrimidinic endonuclease (APE1/Ref1). Nucleic Acids Res. 33, 3303–3312 (2005). 144. Jacob, K. D. et al. Alzheimer’s disease-associated polymorphisms in human OGG1 alter catalytic activity and sensitize cells to DNA damage. Free Radic. Biol. Med. 63, 115–125 (2013). 145. Jacobs, A. L. & Schär, P. DNA glycosylases: in DNA repair and beyond. Chromosoma 121, 1– 20 (2012). 146. Jakobs, S. High resolution imaging of live mitochondria. Biochim. Biophys. Acta 1763, 561– 575 (2006). 147. Jakobs, S. & Wurm, C. A. Super-resolution microscopy of mitochondria. Curr. Opin. Chem. Biol. 20, 9–15 (2014). 148. Jiang, Z. et al. Possible role of mtDNA depletion and respiratory chain defects in aristolochic acid I-induced acute nephrotoxicity. Toxicol. Appl. Pharmacol. 266, 198–203 (2013). 149. Jin, S. M. et al. Mitochondrial membrane potential regulates PINK1 import and proteolytic destabilization by PARL. J. Cell Biol. 191, 933–942 (2010). 150. Jin, S. M. & Youle, R. J. The accumulation of misfolded proteins in the mitochondrial matrix is sensed by PINK1 to induce PARK2/Parkin-mediated mitophagy of polarized mitochondria. Autophagy 9, 1750–1757 (2013). 151. Johri, A. & Beal, M. F. Mitochondrial dysfunction in neurodegenerative diseases. J. Pharmacol. Exp. Ther. 342, 619–630 (2012). 152. Kajander, O. A., Karhunen, P. J., Holt, I. J. & Jacobs, H. T. Prominent mitochondrial DNA recombination intermediates in human heart muscle. EMBO Rep. 2, 1007–1012 (2001). 153. Kalifa, L. et al. Mitochondrial genome maintenance: roles for nuclear nonhomologous end- joining proteins in Saccharomyces cerevisiae. Genetics 190, 951–964 (2012). 154. Karahalil, B., de Souza-Pinto, N. C., Parsons, J. L., Elder, R. H. & Bohr, V. A. Compromised incision of oxidized pyrimidines in liver mitochondria of mice deficient in NTH1 and OGG1 glycosylases. J. Biol. Chem. 278, 33701–33707 (2003).

- 197 -

155. Karbowski, M. et al. Spatial and temporal association of Bax with mitochondrial fission sites, Drp1, and Mfn2 during apoptosis. J. Cell Biol. 159, 931–938 (2002). 156. Kasiviswanathan, R., Minko, I. G., Lloyd, R. S. & Copeland, W. C. Translesion synthesis past acrolein-derived DNA adducts by human mitochondrial DNA polymerase gamma. J. Biol. Chem. 288, 14247–14255 (2013). 157. Kaufman, B. A. et al. The mitochondrial transcription factor TFAM coordinates the assembly of multiple DNA molecules into nucleoid-like structures. Mol. Biol. Cell 18, 3225–3236 (2007). 158. Kazak L., Reyes A., He J., Wood S.R., Brea-Calvo G., Holen T.T., Holt I.J. A cryptic targeting signal creates a mitochondrial FEN1 isoform with tailed R-Loop binding properties. PLoS One. 13;8(5):e62340 (2013). 159. Kazak, L. et al. Alternative translation initiation augments the human mitochondrial proteome. Nucleic Acids Res. 41, 2354–2369 (2013). 160. Kazak, L., Reyes, A. & Holt, I. J. Minimizing the damage: repair pathways keep mitochondrial DNA intact. Nat. Rev. Mol. Cell Biol. 13, 659–671 (2012). 161. Khan, N. A., Govindaraj, P., Meena, A. K., Thangaraj K., Mitochondrial disorders: Challenges in diagnosis & treatment Indian J Med Res. 2015 Jan; 141(1): 13–26 (2015) 162. Khan, A. U. H. et al. Human Leukemic Cells performing Oxidative Phosphorylation (OXPHOS) Generate an Antioxidant Response Independently of Reactive Oxygen species (ROS) Production. EBioMedicine 3, 43–53 (2016). 163. Kim, S.-J. et al. Mitochondria-targeted Ogg1 and aconitase-2 prevent oxidant-induced mitochondrial DNA damage in alveolar epithelial cells. J. Biol. Chem. 289, 6165–6176 (2014). 164. King, M. P. & Attardi, G. Human cells lacking mtDNA: repopulation with exogenous mitochondria by complementation. Science 246, 500–503 (1989). 165. Kitada, T. et al. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature 392, 605–608 (1998). 166. Klionsky, D. J. et al. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy 12, 1–222 (2016). 167. Klungland, A. et al. Accumulation of premutagenic DNA lesions in mice defective in removal of oxidative base damage. Proc. Natl. Acad. Sci. U. S. A. 96, 13300–13305 (1999). 168. Koshiba, T. et al. Structural basis of mitochondrial tethering by mitofusin complexes. Science 305, 858–862 (2004). 169. Kozak M. Interpreting cDNA sequences: some insights from studies on translation. Mamm Genome. 1996;7:563–574 (1997) 170. Kroemer, G., Galluzzi, L. & Brenner, C. Mitochondrial membrane permeabilization in cell death. Physiol. Rev. 87, 99–163 (2007). 171. Krokan, H. E. & Bjoras, M. Base excision repair. Cold Spring Harb. Perspect. Biol. 5, a012583 (2013). 172. Kuchino, Y. et al. Misreading of DNA templates containing 8-hydroxydeoxyguanosine at the modified base and at adjacent residues. Nature 327, 77–79 (1987).

173. Kukat, A. et al. Generation of rho0 cells utilizing a mitochondrially targeted restriction endonuclease and comparative analyses. Nucleic Acids Res. 36, e44 (2008). 174. Kukat, C. & Larsson, N.-G. mtDNA makes a U-turn for the mitochondrial nucleoid. Trends Cell Biol. 23, 457–463 (2013). 175. Kukat, C. et al. Super-resolution microscopy reveals that mammalian mitochondrial nucleoids have a uniform size and frequently contain a single copy of mtDNA. Proc. Natl. Acad. Sci. U. S. A. 108, 13534–13539 (2011). 176. Kuznetsov, N. A., Koval, V. V & Fedorova, O. S. Mechanism of recognition and repair of damaged DNA by human 8-oxoguanine DNA glycosylase hOGG1. Biochemistry. (Mosc). 76, 118–130 (2011). 177. Kyriakouli, D. S., Boesch, P., Taylor, R. W. & Lightowlers, R. N. Progress and prospects: gene therapy for mitochondrial DNA disease. Gene Ther. 15, 1017–1023 (2008). 178. Lakshmipathy, U. & Campbell, C. The human DNA ligase III gene encodes nuclear and mitochondrial proteins. Mol. Cell. Biol. 19, 3869–3876 (1999). 179. Lambert, A. J. & Brand, M. D. Reactive oxygen species production by mitochondria. Methods Mol. Biol. 554, 165–181 (2009). 180. Lambert, A. J., Buckingham, J. A., Boysen, H. M. & Brand, M. D. Low complex I content explains the low hydrogen peroxide production rate of heart mitochondria from the long-lived pigeon, Columba livia. Aging Cell 9, 78–91 (2010). 181. Lamont, P. J. et al. Clinical and laboratory findings in referrals for mitochondrial DNA analysis. Arch. Dis. Child. 79, 22–27 (1998). 182. Landes, T. et al. OPA1 (dys)functions. Semin. Cell Dev. Biol. 21, 593–598 (2010). 183. Langston, L. D. & O’Donnell, M. DNA replication: keep moving and don’t mind the gap. Mol. Cell 23, 155–160 (2006). 184. Larsson, N. G. et al. Mitochondrial transcription factor A is necessary for mtDNA maintenance and embryogenesis in mice. Nat. Genet. 18, 231–236 (1998). 185. Lee, S. S. et al. A systematic RNAi screen identifies a critical role for mitochondria in C. elegans longevity. Nat. Genet. 33, 40–48 (2003). 186. Lee, Y., Jeong, S.-Y., Karbowski, M., Smith, C. L. & Youle, R. J. Roles of the mammalian mitochondrial fission and fusion mediators Fis1, Drp1, and Opa1 in apoptosis. Mol. Biol. Cell 15, 5001–5011 (2004). 187. Lee, Y.-K., Youn, H.-G., Wang, H.-J. & Yoon, G. Decreased mitochondrial OGG1 expression is linked to mitochondrial defects and delayed hepatoma cell growth. Mol. Cells 35, 489–497 (2013). 188. Legros, F., Lombes, A., Frachon, P. & Rojo, M. Mitochondrial fusion in human cells is efficient, requires the inner membrane potential, and is mediated by mitofusins. Mol. Biol. Cell 13, 4343– 4354 (2002). 189. Legros, F., Malka, F., Frachon, P., Lombes, A. & Rojo, M. Organization and dynamics of human mitochondrial DNA. J. Cell Sci. 117, 2653–2662 (2004).

- 199 -

190. Leon, J. et al. 8-Oxoguanine accumulation in mitochondrial DNA causes mitochondrial dysfunction and impairs neuritogenesis in cultured adult mouse cortical neurons under oxidative conditions. Sci. Rep. 6, 22086 (2016). 191. Levin, D. S., McKenna, A. E., Motycka, T. A., Matsumoto, Y. & Tomkinson, A. E. Interaction between PCNA and DNA ligase I is critical for joining of Okazaki fragments and long-patch base- excision repair. Curr. Biol. 10, 919–922 (2000). 192. Levin, J. D. & Demple, B. Analysis of class II (hydrolytic) and class I (beta-lyase) apurinic/apyrimidinic endonucleases with a synthetic DNA substrate. Nucleic Acids Res. 18, 5069– 5075 (1990). 193. Levinger, L., Morl, M. & Florentz, C. Mitochondrial tRNA 3’ end metabolism and human disease. Nucleic Acids Res. 32, 5430–5441 (2004). 194. Lewis, S. C., Uchiyama, L. F. & Nunnari, J. ER-mitochondria contacts couple mtDNA synthesis with mitochondrial division in human cells. Science 353, aaf5549 (2016). 195. Li, M. et al. Human AP endonuclease/redox factor APE1/ref-1modulates mitochondrial function after oxidative stress by regulating the transcriptional activity of NRF1. Free Radic. Biol. Med. 53, 237–248 (2012). 196. Li, Z. et al. ROS leads to MnSOD upregulation through ERK2 translocation and p53 activation in selenite-induced apoptosis of NB4 cells. FEBS Lett. 584, 2291–2297 (2010). 197. Lightowlers, R. N., Taylor, R. W. & Turnbull, D. M. Mutations causing mitochondrial disease: What is new and what challenges remain? Science 349, 1494–1499 (2015). 198. Lindahl, T. & Wood, R. D. Quality control by DNA repair. Science 286, 1897–1905 (1999). 199. Lipscomb, L. A. et al. X-ray structure of a DNA decamer containing 7,8-dihydro-8-oxoguanine. Proc. Natl. Acad. Sci. U. S. A. 92, 719–723 (1995). 200. Liu, P. et al. Removal of oxidative DNA damage via FEN1-dependent long-patch base excision repair in human cell mitochondria. Mol. Cell. Biol. 28, 4975–4987 (2008). 201. Liu, S. et al. Parkinson’s disease-associated kinase PINK1 regulates Miro protein level and axonal transport of mitochondria. PLoS Genet. 8, e1002537 (2012). 202. Luft, R. The development of mitochondrial medicine. Proc. Natl. Acad. Sci. U. S. A. 91, 8731– 8738 (1994). 203. Luna-Vargas, M. P. A. & Chipuk, J. E. The deadly landscape of pro-apoptotic BCL-2 proteins in the outer mitochondrial membrane. FEBS J. 283, 2676–2689 (2016). 204. Lutz, A. K. et al. Loss of parkin or PINK1 function increases Drp1-dependent mitochondrial fragmentation. J. Biol. Chem. 284, 22938–22951 (2009). 205. Lyabin, D. N., Eliseeva, I. A. & Ovchinnikov, L. P. YB-1 protein: functions and regulation. Wiley Interdiscip. Rev. RNA 5, 95–110 (2014). 206. Maier, D. et al. Mitochondrial single-stranded DNA-binding protein is required for mitochondrial DNA replication and development in Drosophila melanogaster. Mol. Biol. Cell 12, 821–830 (2001).

207. Manchekar, M. et al. DNA recombination activity in soybean mitochondria. J. Mol. Biol. 356, 288–299 (2006). 208. Mandal, S. M. et al. Role of human DNA glycosylase Nei-like 2 (NEIL2) and single strand break repair protein polynucleotide kinase 3’-phosphatase in maintenance of mitochondrial genome. J. Biol. Chem. 287, 2819–2829 (2012). 209. Mannella, C. A. Structure and dynamics of the mitochondrial inner membrane cristae. Biochim. Biophys. Acta 1763, 542–548 (2006). 210. Mantha, A. K., Dhiman, M., Taglialatela, G., Perez-Polo, R. J. & Mitra, S. Proteomic study of amyloid beta (25-35) peptide exposure to neuronal cells: Impact on APE1/Ref-1’s protein-protein interaction. J. Neurosci. Res. 90, 1230–1239 (2012). 211. Mao, G. et al. Identification and characterization of OGG1 mutations in patients with Alzheimer’s disease. Nucleic Acids Res. 35, 2759–2766 (2007). 212. Martin, S. A. et al. DNA polymerases as potential therapeutic targets for cancers deficient in the DNA mismatch repair proteins MSH2 or MLH1. Cancer Cell 17, 235–248 (2010). 213. Mason, P. A. & Lightowlers, R. N. Why do mammalian mitochondria possess a mismatch repair activity? FEBS Lett. 554, 6–9 (2003). 214. Mason, P. A., Matheson, E. C., Hall, A. G. & Lightowlers, R. N. Mismatch repair activity in mammalian mitochondria. Nucleic Acids Res. 31, 1052–1058 (2003). 215. Masters, B. S., Stohl, L. L. & Clayton, D. A. Yeast mitochondrial RNA polymerase is homologous to those encoded by bacteriophages T3 and T7. Cell 51, 89–99 (1987). 216. Matsuda, N. et al. PINK1 stabilized by mitochondrial depolarization recruits Parkin to damaged mitochondria and activates latent Parkin for mitophagy. J. Cell Biol. 189, 211–221 (2010). 217. McIlwain, D. R., Berger, T. & Mak, T. W. Caspase functions in cell death and disease. Cold Spring Harb. Perspect. Biol. 5, a008656 (2013). 218. Mecocci, P. et al. Oxidative damage to mitochondrial DNA shows marked age-dependent increases in human brain. Ann. Neurol. 34, 609–616 (1993). 219. Meeusen, S. & Nunnari, J. Evidence for a two membrane-spanning autonomous mitochondrial DNA replisome. J. Cell Biol. 163, 503–510 (2003). 220. Michaels, M. L. & Miller, J. H. The GO system protects organisms from the mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-dihydro-8-oxoguanine). J. Bacteriol. 174, 6321– 6325 (1992). 221. Mishra, P. & Chan, D. C. Mitochondrial dynamics and inheritance during cell division, development and disease. Nat. Rev. Mol. Cell Biol. 15, 634–646 (2014). 222. Morland, I., Luna, L., Gustad, E., Seeberg, E. & Bjoras, M. Product inhibition and magnesium modulate the dual reaction mode of hOgg1. DNA Repair (Amst). 4, 381–387 (2005). 223. Mortensen, K. I., Churchman, L. S., Spudich, J. A. & Flyvbjerg, H. Optimized localization analysis for single-molecule tracking and super-resolution microscopy. Nat. Methods 7, 377–381 (2010).

- 201 -

224. Mozdy, A. D., McCaffery, J. M. & Shaw, J. M. Dnm1p GTPase-mediated mitochondrial fission is a multi-step process requiring the novel integral membrane component Fis1p. J. Cell Biol. 151, 367–380 (2000). 225. Murphy, M. P. How mitochondria produce reactive oxygen species. Biochem. J. 417, 1–13 (2009). 226. Nakabeppu, Y. Cellular levels of 8-oxoguanine in either DNA or the nucleotide pool play pivotal roles in carcinogenesis and survival of cancer cells. Int. J. Mol. Sci. 15, 12543–12557 (2014). 227. Nakanishi, N., Fukuoh, A., Kang, D., Iwai, S. & Kuraoka, I. Effects of DNA lesions on the transcription reaction of mitochondrial RNA polymerase: implications for bypass RNA synthesis on oxidative DNA lesions. Mutagenesis 28, 117–123 (2013). 228. Narendra, D., Tanaka, A., Suen, D.-F. & Youle, R. J. Parkin is recruited selectively to impaired mitochondria and promotes their autophagy. J. Cell Biol. 183, 795–803 (2008). 229. Nargund, A. M., Pellegrino, M. W., Fiorese, C. J., Baker, B. M. & Haynes, C. M. Mitochondrial import efficiency of ATFS-1 regulates mitochondrial UPR activation. Science 337, 587–590 (2012). 230. Nash, H. M., Lu, R., Lane, W. S. & Verdine, G. L. The critical active-site amine of the human 8- oxoguanine DNA glycosylase, hOgg1: direct identification, ablation and chemical reconstitution. Chem. Biol. 4, 693–702 (1997). 231. Nakanishi, N., Fukuoh, A., Kang, D., Iwai, S., Kuraoka, I. Effects of DNA lesions on the transcription reaction of mitochondrial RNA polymerase: implications for bypass RNA synthesis on oxidative DNA lesions. Mutagenesis. 28(1):117-23 (2013). 232. Nemec, A. A., Wallace, S. S. & Sweasy, J. B. Variant base excision repair proteins: contributors to genomic instability. Semin. Cancer Biol. 20, 320–328 (2010). 233. Nishioka, K. et al. Expression and differential intracellular localization of two major forms of human 8-oxoguanine DNA glycosylase encoded by alternatively spliced OGG1 mRNAs. Mol. Biol. Cell 10, 1637–1652 (1999). 234. Norman, D. P. G., Chung, S. J. & Verdine, G. L. Structural and biochemical exploration of a critical amino acid in human 8-oxoguanine glycosylase. Biochemistry 42, 1564–1572 (2003). 235. Oda, H. et al. Multi-forms of human MTH1 polypeptides produced by alternative translation initiation and single nucleotide polymorphism. Nucleic Acids Res. 27, 4335–4343 (1999). 236. Ogawa, A., Watanabe, T., Shoji, S. & Furihata, C. Enzyme kinetics of an alternative splicing isoform of mitochondrial 8-oxoguanine DNA glycosylase, ogg1-1b, and compared with the nuclear ogg1-1a. J. Biochem. Mol. Toxicol. 29, 49–56 (2015). 237. Ohtsubo, T. et al. Identification of human MutY homolog (hMYH) as a repair enzyme for 2- hydroxyadenine in DNA and detection of multiple forms of hMYH located in nuclei and mitochondria. Nucleic Acids Res. 28, 1355–1364 (2000). 238. Oka, S., Leon, J., Tsuchimoto, D., Sakumi, K. & Nakabeppu, Y. MUTYH, an adenine DNA glycosylase, mediates p53 tumor suppression via PARP-dependent cell death. Oncogenesis 3, e121 (2014). 239. Oka, S. et al. Two distinct pathways of cell death triggered by oxidative damage to nuclear and mitochondrial DNAs. EMBO J. 27, 421–432 (2008).

240. Ono, T., Isobe, K., Nakada, K. & Hayashi, J. I. Human cells are protected from mitochondrial dysfunction by complementation of DNA products in fused mitochondria. Nat. Genet. 28, 272–275 (2001). 241. Osenbroch, P. O. et al. Accumulation of mitochondrial DNA damage and bioenergetic dysfunction in CSB defective cells. FEBS J. 276, 2811–2821 (2009). 242. Osterod, M. et al. Age-related and tissue-specific accumulation of oxidative DNA base damage in 7,8-dihydro-8-oxoguanine-DNA glycosylase (Ogg1) deficient mice. Carcinogenesis 22, 1459–1463 (2001). 243. Panduri, V. et al. Role of mitochondrial hOGG1 and aconitase in oxidant-induced lung epithelial cell apoptosis. Free Radic. Biol. Med. 47, 750–759 (2009). 244. Park, S. Y. et al. Resistance of mitochondrial DNA-depleted cells against cell death: role of mitochondrial superoxide dismutase. J. Biol. Chem. 279, 7512–7520 (2004). 245. Parone, P. A. et al. Inhibiting the mitochondrial fission machinery does not prevent Bax/Bak- dependent apoptosis. Mol. Cell. Biol. 26, 7397–7408 (2006). 246. Paul, R., Dalibart, R., Lemoine, S. & Lestienne, P. Expression of E. coli RecA targeted to mitochondria of human cells. Mutat. Res. 486, 11–19 (2001). 247. Paz-Elizur, T. et al. Reduced repair of the oxidative 8-oxoguanine DNA damage and risk of head and neck cancer. Cancer Res. 66, 11683–11689 (2006). 248. Paz-Elizur, T. et al. DNA repair activity for oxidative damage and risk of lung cancer. J. Natl. Cancer Inst. 95, 1312–1319 (2003). 249. Pellegrino, M. W. & Haynes, C. M. Mitophagy and the mitochondrial unfolded protein response in neurodegeneration and bacterial infection. BMC Biol. 13, 22 (2015). 250. Poole, A. C., Thomas, R. E., Yu, S., Vincow, E. S. & Pallanck, L. The mitochondrial fusion- promoting factor mitofusin is a substrate of the PINK1/parkin pathway. PLoS One 5, e10054 (2010). 251. Prakash, S., Johnson, R. E. & Prakash, L. Eukaryotic translesion synthesis DNA polymerases: specificity of structure and function. Annu. Rev. Biochem. 74, 317–353 (2005). 252. Rachek, L. I. et al. Conditional targeting of the DNA repair enzyme hOGG1 into mitochondria. J. Biol. Chem. 277, 44932–44937 (2002). 253. Radicella, J. P. & Boiteux, S. [Repair of oxidized guanine in mammals: OGG1 genes]. C. R. Seances Soc. Biol. Fil. 191, 755–763 (1997). 254. Radicella, J. P., Dherin, C., Desmaze, C., Fox, M. S. & Boiteux, S. Cloning and characterization of hOGG1, a human homolog of the OGG1 gene of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U. S. A. 94, 8010–8015 (1997). 255. Radsak, K. & Schutz, E. Changes of mitochondrial DNA polymerase-gamma activity in synchronized mouse cell cultures. Eur. J. Biochem. 89, 3–9 (1978). 256. Rajala, N., Gerhold, J. M., Martinsson, P., Klymov, A. & Spelbrink, J. N. Replication factors transiently associate with mtDNA at the mitochondrial inner membrane to facilitate replication. Nucleic Acids Res. 42, 952–67 (2014).

- 203 -

257. Ramdzan, Z. M. et al. The DNA repair function of CUX1 contributes to radioresistance. Oncotarget 8, 19021–19038 (2017). 258. Richly, E., Chinnery, P. F. & Leister, D. Evolutionary diversification of mitochondrial proteomes: implications for human disease. Trends Genet. 19, 356–362 (2003). 259. Richter, C., Park, J. W. & Ames, B. N. Normal oxidative damage to mitochondrial and nuclear DNA is extensive. Proc. Natl. Acad. Sci. U. S. A. 85, 6465–6467 (1988). 260. Riedl, S. J. & Shi, Y. Molecular mechanisms of caspase regulation during apoptosis. Nat. Rev. Mol. Cell Biol. 5, 897–907 (2004). 261. Robberson, D. L., Kasamatsu, H. & Vinograd, J. Replication of mitochondrial DNA. Circular replicative intermediates in mouse L cells. Proc. Natl. Acad. Sci. U. S. A. 69, 737–741 (1972). 262. Roldan-Arjona, T. et al. Molecular cloning and functional expression of a human cDNA encoding the antimutator enzyme 8-hydroxyguanine-DNA glycosylase. Proc. Natl. Acad. Sci. U. S. A. 94, 8016–8020 (1997). 263. Rothfuss, O., Gasser, T. & Patenge, N. Analysis of differential DNA damage in the mitochondrial genome employing a semi-long run real-time PCR approach. Nucleic Acids Res. 38, e24 (2010). 264. Russo, G. J. et al. Drosophila Miro is required for both anterograde and retrograde axonal mitochondrial transport. J. Neurosci. 29, 5443–5455 (2009). 265. Sagan, L. On the origin of mitosing cells. J. Theor. Biol. 14, 255–274 (1967). 266. Sage, J. M., Gildemeister, O. S. & Knight, K. L. Discovery of a novel function for human Rad51: maintenance of the mitochondrial genome. J. Biol. Chem. 285, 18984–18990 (2010). 267. Saki, M., Prakash, A., DNA damage related crosstalk between the nucleus and mitochondria Free Radical Biology and Medicine 107, 216-227 (2017) 268. Sale, J. E. Translesion DNA synthesis and mutagenesis in eukaryotes. Cold Spring Harb. Perspect. Biol. 5, a012708 (2013). 269. Sampath, H., McCullough, A. K. & Lloyd, R. S. Regulation of DNA glycosylases and their role in limiting disease. Free Radic. Res. 46, 460–478 (2012). 270. Santos, J. H., Hunakova, L., Chen, Y., Bortner, C. & Van Houten, B. Cell sorting experiments link persistent mitochondrial DNA damage with loss of mitochondrial membrane potential and apoptotic cell death. J. Biol. Chem. 278, 1728–1734 (2003). 271. Schafer, E. et al. Architecture of active mammalian respiratory chain supercomplexes. J. Biol. Chem. 281, 15370–15375 (2006). 272. Scharer, O. D. & Jiricny, J. Recent progress in the biology, chemistry and structural biology of DNA glycosylases. Bioessays 23, 270–281 (2001). 273. Schieke, S. M., McCoy, J. P. J. & Finkel, T. Coordination of mitochondrial bioenergetics with G1 phase cell cycle progression. Cell Cycle 7, 1782–1787 (2008). 274. Schimmer, A. D. Inhibitor of apoptosis proteins: translating basic knowledge into clinical practice. Cancer Res. 64, 7183–7190 (2004).

275. Schon, E. A. & Gilkerson, R. W. Functional complementation of mitochondrial DNAs: mobilizing mitochondrial genetics against dysfunction. Biochim. Biophys. Acta 1800, 245–249 (2010). 276. Scorrano, L. et al. A distinct pathway remodels mitochondrial cristae and mobilizes cytochrome c during apoptosis. Dev. Cell 2, 55–67 (2002). 277. Shadel G S, Clayton D A. Mitochondrial DNA maintenance in vertebrates. Annu Rev Biochem. 1997;66:409–435 (1997) 278. Shadel, G. S. & Horvath, T. L. Mitochondrial ROS signaling in organismal homeostasis. Cell 163, 560–569 (2015). 279. Shaw, J. M. & Nunnari, J. Mitochondrial dynamics and division in budding yeast. Trends Cell Biol. 12, 178–184 (2002). 280. Shigenaga, M. K., Hagen, T. M. & Ames, B. N. Oxidative damage and mitochondrial decay in aging. Proc. Natl. Acad. Sci. U. S. A. 91, 10771–10778 (1994). 281. Shokolenko, I., Venediktova, N., Bochkareva, A., Wilson, G. L. & Alexeyev, M. F. Oxidative stress induces degradation of mitochondrial DNA. Nucleic Acids Res. 37, 2539–2548 (2009). 282. Simsek, D. et al. Crucial role for DNA ligase III in mitochondria but not in Xrcc1-dependent repair. Nature 471, 245–248 (2011). 283. Singh, B. et al. Human REV3 DNA Polymerase Zeta Localizes to Mitochondria and Protects the Mitochondrial Genome. PLoS One 10, e0140409 (2015). 284. Smirnova, E., Griparic, L., Shurland, D. L. & van der Bliek, A. M. Dynamin-related protein Drp1 is required for mitochondrial division in mammalian cells. Mol. Biol. Cell 12, 2245–2256 (2001). 285. Song, Z., Ghochani, M., McCaffery, J. M., Frey, T. G. & Chan, D. C. Mitofusins and OPA1 mediate sequential steps in mitochondrial membrane fusion. Mol. Biol. Cell 20, 3525–3532 (2009). 286. Srinivasan, A. & Gold, B. Small-molecule inhibitors of DNA damage-repair pathways: an approach to overcome tumor resistance to alkylating anticancer drugs. Future Med. Chem. 4, 1093– 1111 (2012). 287. Srinivasula, S. M. et al. A conserved XIAP-interaction motif in caspase-9 and Smac/DIABLO regulates caspase activity and apoptosis. Nature 410, 112–116 (2001). 288. St John, J. C. Mitochondrial DNA copy number and replication in reprogramming and differentiation. Semin. Cell Dev. Biol. 52, 93–101 (2016). 289. Stein, A., Kalifa, L. & Sia, E. A. Members of the RAD52 Epistasis Group Contribute to Mitochondrial Homologous Recombination and Double-Strand Break Repair in Saccharomyces cerevisiae. PLoS Genet. 11, e1005664 (2015). 290. Stevnsner, T. et al. Mitochondrial repair of 8-oxoguanine is deficient in Cockayne syndrome group B. Oncogene 21, 8675–8682 (2002). 291. Stewart, J. B. & Chinnery, P. F. The dynamics of mitochondrial DNA heteroplasmy: implications for human health and disease. Nature reviews. Genetics 16, 530–542 (2015). 292. Stojkovic, G. et al. Oxidative DNA damage stalls the human mitochondrial replisome. Sci. Rep. 6, 28942 (2016).

- 205 -

293. Stuart, J. A., Bourque, B. M., de Souza-Pinto, N. C. & Bohr, V. A. No evidence of mitochondrial respiratory dysfunction in OGG1-null mice deficient in removal of 8-oxodeoxyguanine from mitochondrial DNA. Free Radic. Biol. Med. 38, 737–745 (2005). 294. Stuart, J. A. et al. DNA base excision repair activities and pathway function in mitochondrial and cellular lysates from cells lacking mitochondrial DNA. Nucleic Acids Res. 32, 2181–2192 (2004). 295. Stuart, J. A., Mayard, S., Hashiguchi, K., Souza-Pinto, N. C. & Bohr, V. A. Localization of mitochondrial DNA base excision repair to an inner membrane-associated particulate fraction. Nucleic Acids Res. 33, 3722–3732 (2005). 296. Suen, D.-F., Norris, K. L. & Youle, R. J. Mitochondrial dynamics and apoptosis. Genes Dev. 22, 1577–1590 (2008). 297. Sun, N., Youle, R. J. & Finkel, T. The Mitochondrial Basis of Aging. Mol. Cell 61, 654–666 (2016). 298. Suter, M. & Richter, C. Fragmented mitochondrial DNA is the predominant carrier of oxidized DNA bases. Biochemistry 38, 459–464 (1999). 299. Sutovsky, P. et al. Ubiquitin tag for sperm mitochondria. Nature 402, 371–372 (1999). 300. Svilar, D., Goellner, E. M., Almeida, K. H. & Sobol, R. W. Base excision repair and lesion- dependent subpathways for repair of oxidative DNA damage. Antioxid. Redox Signal. 14, 2491–2507 (2011). 301. Swartzlander, D. B. et al. Regulation of base excision repair: Ntg1 nuclear and mitochondrial dynamic localization in response to genotoxic stress. Nucleic Acids Res. 38, 3963–3974 (2010). 302. Swartzlander, D. B. et al. Identification of SUMO modification sites in the base excision repair protein, Ntg1. DNA Repair (Amst). 48, 51–62 (2016). 303. Sykora P., Kanno S., Akbari M., Kulikowicz T., Baptiste B.A., Leandro G.S., Lu H., Tian J., May A., Becker K.A., Croteau D.L., Wilson D.M., Sobol R.W., Yasui A., Bohr V.A. DNA polymerase beta participates in mitochondrial DNA repair. Mol Cell Biol. pii: MCB.00237-17 (2017) 304. Szczesny, B., Tann, A. W., Longley, M. J., Copeland, W. C. & Mitra, S. Long patch base excision repair in mammalian mitochondrial genomes. J. Biol. Chem. 283, 26349–26356 (2008). 305. Taanman, J. W. The mitochondrial genome: structure, transcription, translation and replication. Biochim. Biophys. Acta 1410, 103–123 (1999). 306. Tahbaz, N., Subedi, S. & Weinfeld, M. Role of polynucleotide kinase/phosphatase in mitochondrial DNA repair. Nucleic Acids Res. 40, 3484–95 (2012). 307. Takamatsu, C. et al. Regulation of mitochondrial D-loops by transcription factor A and single- stranded DNA-binding protein. EMBO Rep. 3, 451–456 (2002). 308. Takao, M., Aburatani, H., Kobayashi, K. & Yasui, A. Mitochondrial targeting of human DNA glycosylases for repair of oxidative DNA damage. Nucleic Acids Res. 26, 2917–2922 (1998). 309. Tao, R. et al. Sirt3-mediated deacetylation of evolutionarily conserved lysine 122 regulates MnSOD activity in response to stress. Mol. Cell 40, 893–904 (2010). 310. Tatsuta, T. & Langer, T. Quality control of mitochondria: protection against neurodegeneration and ageing. EMBO J. 27, 306–314 (2008).

311. Tell, G., Quadrifoglio, F., Tiribelli, C. & Kelley, M. R. The many functions of APE1/Ref-1: not only a DNA repair enzyme. Antioxid. Redox Signal. 11, 601–620 (2009). 312. Thomas, D., Scot, A. D., Barbey, R., Padula, M. & Boiteux, S. Inactivation of OGG1 increases the incidence of G . C-->T . A transversions in Saccharomyces cerevisiae: evidence for endogenous oxidative damage to DNA in eukaryotic cells. Mol. Gen. Genet. 254, 171–178 (1997). 313. Tieu, Q. & Nunnari, J. Mdv1p is a WD repeat protein that interacts with the dynamin-related GTPase, Dnm1p, to trigger mitochondrial division. J. Cell Biol. 151, 353–366 (2000). 314. Tiranti, V. et al. Identification of the gene encoding the human mitochondrial RNA polymerase (h-mtRPOL) by cyberscreening of the Expressed Sequence Tags database. Hum. Mol. Genet. 6, 615–625 (1997). 315. Torres-Gonzalez, M., Gawlowski, T., Kocalis, H., Scott, B. T. & Dillmann, W. H. Mitochondrial 8-oxoguanine glycosylase decreases mitochondrial fragmentation and improves mitochondrial function in H9C2 cells under oxidative stress conditions. Am. J. Physiol. Cell Physiol. 306, C221-9 (2014). 316. Tress, M. L., Abascal, F. & Valencia, A. Alternative Splicing May Not Be the Key to Proteome Complexity. Trends Biochem. Sci. 42, 98–110 (2017). 317. Trifunovic, A. et al. Premature ageing in mice expressing defective mitochondrial DNA polymerase. Nature 429, 417–423 (2004). 318. Turner, N. A. et al. Oxidative stress induces DNA fragmentation and caspase activation via the c-Jun NH2-terminal kinase pathway in H9c2 cardiac muscle cells. J. Mol. Cell. Cardiol. 30, 1789– 1801 (1998). 319. Twig, G. et al. Fission and selective fusion govern mitochondrial segregation and elimination by autophagy. EMBO J. 27, 433–446 (2008). 320. Twig, G. et al. Tagging and tracking individual networks within a complex mitochondrial web with photoactivatable GFP. Am. J. Physiol. Cell Physiol. 291, C176-84 (2006). 321. Valente, E. M. et al. Hereditary early-onset Parkinson’s disease caused by mutations in PINK1. Science 304, 1158–1160 (2004). 322. Van Damme, P., Gawron, D., Van Criekinge, W. & Menschaert, G. N-terminal proteomics and ribosome profiling provide a comprehensive view of the alternative translation initiation landscape in mice and men. Mol. Cell. Proteomics 13, 1245–1261 (2014). 323. van der Kemp, P. A., Charbonnier, J.-B., Audebert, M. & Boiteux, S. Catalytic and DNA-binding properties of the human Ogg1 DNA N-glycosylase/AP lyase: biochemical exploration of H270, Q315 and F319, three amino acids of the 8-oxoguanine-binding pocket. Nucleic Acids Res. 32, 570–8 (2004). 324. van Engeland, M., Nieland, L. J., Ramaekers, F. C., Schutte, B. & Reutelingsperger, C. P. Annexin V-affinity assay: a review on an apoptosis detection system based on phosphatidylserine exposure. Cytometry 31, 1–9 (1998). 325. van Loo, G. et al. Caspases are not localized in mitochondria during life or death. Cell Death Differ. 9, 1207–1211 (2002).

- 207 -

326. Vidal, A. E., Hickson, I. D., Boiteux, S. & Radicella, J. P. Mechanism of stimulation of the DNA glycosylase activity of hOGG1 by the major human AP endonuclease: bypass of the AP lyase activity step. Nucleic Acids Res. 29, 1285–1292 (2001). 327. Vitale, N. et al. Phospholipase D1: a key factor for the exocytotic machinery in neuroendocrine cells. EMBO J. 20, 2424–2434 (2001). 328. Voigt, A., Berlemann, L. A. & Winklhofer, K. F. The mitochondrial kinase PINK1: functions beyond mitophagy. J. Neurochem. 139 Suppl, 232–239 (2016). 329. Vuda, M. & Kamath, A. Drug induced mitochondrial dysfunction: Mechanisms and adverse clinical consequences. Mitochondrion 31, 63–74 (2016). 330. Walker, J. E., Collinson, I. R., Van Raaij, M. J. & Runswick, M. J. Structural analysis of ATP synthase from bovine heart mitochondria. Methods Enzymol. 260, 163–190 (1995). 331. Wallace, D. C. Mitochondrial defects in cardiomyopathy and neuromuscular disease. Am. Heart J. 139, S70-85 (2000). 332. Wallace, D. C. Mitochondrial diseases in man and mouse. Science 283, 1482–1488 (1999). 333. Wallace, D. C. Mitochondria and cancer. Nat. Rev. Cancer 12, 685–698 (2012). 334. Wallace, S. S. Base excision repair: a critical player in many games. DNA Repair (Amst). 19, 14–26 (2014). 335. Wang, C. & Youle, R. J. The role of mitochondria in apoptosis*. Annu. Rev. Genet. 43, 95–118 (2009). 336. Wang, J., Xiong, S., Xie, C., Markesbery, W. R. & Lovell, M. A. Increased oxidative damage in nuclear and mitochondrial DNA in Alzheimer’s disease. J. Neurochem. 93, 953–962 (2005). 337. Wang, W. et al. Mitochondrial DNA damage level determines neural stem cell differentiation fate. J. Neurosci. 31, 9746–9751 (2011). 338. WARBURG, O. On the origin of cancer cells. Science 123, 309–314 (1956). 339. Weiss, J. M., Goode, E. L., Ladiges, W. C. & Ulrich, C. M. Polymorphic variation in hOGG1 and risk of cancer: a review of the functional and epidemiologic literature. Mol. Carcinog. 42, 127–141 (2005). 340. Wiederhold, L. et al. AP Endonuclease-Independent DNA Base Excision Repair in Human Cells National Institutes of Health. 15, 209–220 (2004). 341. Wikstrom, J. D., Twig, G. & Shirihai, O. S. What can mitochondrial heterogeneity tell us about mitochondrial dynamics and autophagy? Int. J. Biochem. Cell Biol. 41, 1914–1927 (2009). 342. Wojcik, M., Burzynska-Pedziwiatr, I. & Wozniak, L. A. A review of natural and synthetic antioxidants important for health and longevity. Curr. Med. Chem. 17, 3262–3288 (2010). 343. Wu, S., Zhou, F., Zhang, Z. & Xing, D. Mitochondrial oxidative stress causes mitochondrial fragmentation via differential modulation of mitochondrial fission-fusion proteins. FEBS J. 278, 941– 954 (2011). 344. Xie, M., Doetsch, P. W. & Deng, X. Bcl2 inhibition of mitochondrial DNA repair. BMC Cancer 15, 586 (2015).

345. Yakes, F. M. & Van Houten, B. Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress. Proc. Natl. Acad. Sci. U. S. A. 94, 514–519 (1997).

346. Yakubovskaya, E., Chen, Z., Carrodeguas, J. A., Kisker, C. & Bogenhagen, D. F. Functional human mitochondrial DNA polymerase gamma forms a heterotrimer. J. Biol. Chem. 281, 374–382 (2006). 347. Yang, J.-L., Tadokoro, T., Keijzers, G., Mattson, M. P. & Bohr, V. A. Neurons efficiently repair glutamate-induced oxidative DNA damage by a process involving CREB-mediated up-regulation of apurinic endonuclease 1. J. Biol. Chem. 285, 28191–28199 (2010). 348. Yang, Y. et al. Pink1 regulates mitochondrial dynamics through interaction with the fission/fusion machinery. Proc. Natl. Acad. Sci. U. S. A. 105, 7070–7075 (2008). 349. Yasukawa, T. et al. Replication of vertebrate mitochondrial DNA entails transient ribonucleotide incorporation throughout the lagging strand. EMBO J. 25, 5358–5371 (2006). 350. Yeeles, J. T. P. & Marians, K. J. The Escherichia coli replisome is inherently DNA damage tolerant. Science 334, 235–238 (2011). 351. Yogev, O. & Pines, O. Dual targeting of mitochondrial proteins: mechanism, regulation and function. Biochim. Biophys. Acta 1808, 1012–1020 (2011). 352. You, H. J. et al. Saccharomyces cerevisiae Ntg1p and Ntg2p: broad specificity N-glycosylases for the repair of oxidative DNA damage in the nucleus and mitochondria. Biochemistry 38, 11298– 11306 (1999). 353. Youle, R. J. & Karbowski, M. Mitochondrial fission in apoptosis. Nat. Rev. Mol. Cell Biol. 6, 657–663 (2005). 354. Youle, R. J. & Narendra, D. P. Mechanisms of mitophagy. Nat. Rev. Mol. Cell Biol. 12, 9–14 (2011). 355. Youle, R. J. & van der Bliek, A. M. Mitochondrial fission, fusion, and stress. Science 337, 1062– 1065 (2012). 356. Yuan, J., Shaham, S., Ledoux, S., Ellis, H. M. & Horvitz, H. R. The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta-converting enzyme. Cell 75, 641– 652 (1993). 357. Yuzefovych, L. V et al. Mitochondrial DNA Repair through OGG1 Activity Attenuates Breast Cancer Progression and Metastasis. Cancer Res. 76, 30–34 (2016). 358. Zhang, L., Chan, S. S. L. & Wolff, D. J. Mitochondrial disorders of DNA polymerase gamma dysfunction: from anatomic to molecular pathology diagnosis. Arch. Pathol. Lab. Med. 135, 925–934 (2011). 359. Zhao, Q. et al. A mitochondrial specific stress response in mammalian cells. EMBO J. 21, 4411–4419 (2002).

360. Zhou, P.-T., Li, B., Ji, J., Wang, M.-M. & Gao, C.-F. A systematic review and meta-analysis of the association between OGG1 Ser326Cys polymorphism and cancers. Med. Oncol. 32, 472 (2015).

- 209 -

- 211 -

Titre : Fonction De L'αOGG1 Sur La Maintenance De l'altération la plus fréquente induite par les ROS sur les La Physiologie Mitochondriale deux, l'ADNc et l'ADNmt. Si la fourche de réplication contourne le 8-oxoG avant son élimination, un A est Mots clés : ADN repair, Stabilité ADNmt, Réparation souvent inséré sur le brin d'ADN opposé et les par Excision de Bases, Stress Oxydant réplications subséquentes corrigent la transversion de G à T. Lorsqu'il est associé à la cytosine, le 8-oxoG est Résumé : Les mitochondries sont des structures éliminé de l'ADN par l'ADN glycosylase de 8-oxoguanine uniques dans la cellule mammifère. Ces organites (OGG1) qui, de cette manière, initie le procédé BER. portent leur propre génome (ADN mitochondrial, ADNmt) OGG1 est une glycosylase de ménage bifonctionnelle qui se compose d'une petite molécule qui codifie pour 13 qui, conjointement avec d'autres enzymes BER polypeptides de la chaîne de transport d'électrons (ETC), différentes, est présente dans les compartiments 22 ARNt et 2 gènes d'ARNr pour sa propre synthèse nucléaires et mitochondriaux, soulignant l'importance de protéique. Le MTDNA est proposé pour être plus maintenir l'intégrité de l'ADNmt pour le fonctionnement susceptible au stress oxydatif que le génome nucléaire cellulaire normal. Il a été démontré que la surexpression (ADNn) parce que non seulement il manque d'histones d'une version recombinante d'OGG1, spécifiquement protectrices, mais aussi en raison de sa proximité avec destinée aux mitochondries par un signal de ciblage les complexes ETC qui sont les principaux producteurs mitochondrial supplémentaire (MTS) (OGG1-MTS), de ROS dans les cellules de mammifères. Parmi tous les protège les cellules d'un stress oxydatif, probablement types de dommages à l'ADNmt, les dommages oxydatifs en raison d'une efficacité accrue dans la réparation De sont les plus répandus et, de loin, les mieux étudiés. La 8-oxoG dans l'ADNmt. L'objectif principal de notre projet voie de réparation de l'excision de base (BER) est un est d'élucider si la perte spécifique de l'activité de mécanisme de réparation d'ADN conservé de façon réparation 8-oxoG dans les mitochondries (mais pas évolutive qui répare les dommages de base d'ADN non dans le compartiment nucléaire) a un impact sur les volumineux. Puisque la guanine a le potentiel redox le fonctions organelles et / ou sur la viabilité cellulaire et plus bas de toute autre base dans l'ADN, elle est aussi pour dévoiler le mécanisme / s Derrière les effets facilement oxydée à la 8-oxoguanine (8-oxoG) qui est protecteurs d'OGG1 sur la physiologie mitochondriale et la maintenance d'ADNmt

Title : Role Of αOGG1 In The Maintenance Of oxoG) that is the most frequent alteration induced by Mitochondrial Physiology ROS on both, nDNA and mtDNA. If the replication fork bypasses the 8-oxoG before its removal, an A is often Keywords : DNA Repair, mtDNA stability, Base inserted on the opposite DNA strand and subsequent Excision Repair, Oxidative Stress replications fix the G to T transversion. When paired with cytosine, 8-oxoG is removed from DNA by the 8- Abstract : Mitochondria are unique structures within the oxoguanine DNA glycosylase (OGG1) that in such a way mammalian cell. These organelles carry their own initiates the BER process. OGG1 is a bifunctional genome (mitochondrial DNA, mtDNA) which consists of housekeeping glycosylase that, together with other a small molecule that codifies for 13 polypeptides of the various BER enzymes is present in both nuclear and electron transport chain (ETC), 22 tRNA and 2 rRNA mitochondrial compartments, highlighting the importance genes for its own protein synthesis. MtDNA is proposed of maintaining mtDNA integrity for normal cellular to be more susceptible to oxidative stress than the functioning. It has been demonstrated that the nuclear genome (nDNA) because not only it lacks overexpression of a recombinant version of OGG1, protective histones but also because of its proximity to specifically targeted to mitochondria by an additional ETC complexes which are the main ROS producers in Mitochondrial Targeting Signal (MTS) (OGG1-MTS), mammalian cells. Among all the types of mtDNA protects the cells from an oxidative stress, likely due to damage, oxidative damage is the most prevalent and, by an increased efficiency in the repair of 8-oxoG in mtDNA. far, the best studied. Base excision repair (BER) pathway The main goal of our project is to elucidate if the specific is an evolutionarily conserved DNA repair mechanism loss of 8-oxoG repair activity in mitochondria (but not in that repairs non-bulky DNA base damages. Since nuclear compartment) has an impact on the organelles’ guanine has the lowest redox potential of any other functions and/or on cell viability and also to unveil the bases in DNA, it is readily oxidized to 8-oxoguanine (8- mechanism/s behind the protective effects of OGG1 on mitochondrial physiology and mtDNA maintenance.

Universite´ Paris-Saclay Espace Technologique / Immeuble Discovery Route de l’Orme aux Merisiers RD 128 / 91190 Saint-Aubin, France