Identification and Characterization of Novel Genes Involved in Cytochrome

Identification and characterization of novel genes involved

in cytochrome c oxidase deficiencies

Woranontee Weraarpachai

Department of Human Genetics

McGill University, Montreal

October 2011

A thesis submitted to McGill University in partial fulfillment of the requirements of the

degree of Ph.D.

©Woranontee Weraarpachai, 2011

Table of Contents

Acknowledgements ...... 5

Abstract ...... 7

Abstract (French) ...... 9

Contributions of Authors ...... 12

Abbreviations ...... 14

General Introduction ...... 19

Introduction ...... 20

Part 1: Oxidative Phosphorylation ...... 20

Part 2: Mitochondrial Genome...... 22

Mitochondrial DNA structure and organization ...... 22

Mitochondrial DNA replication ...... 26

Mitochondrial DNA transcription ...... 27

Mitochondrial translation ...... 29

Part 3: Human diseases due to mitochondrial translation defect ...... 38

Part 4: Complex IV or cytochrome c oxidase (COX) ...... 43

COX structure and function ...... 43

Assembly of COX ...... 45

COX deficiency ...... 52

Thesis outline ...... 57

Chapter 1: Mutation in TACO1, a translational activator of COX I, results in cytochrome c oxidase deficiency and late-onset Leigh Syndrome ...... 60

Abstract ...... 61

Results and Discussion ...... 62

Materials and Methods ...... 69

Figure Legends...... 80

Figures...... 86

Chapter 2: Mutations in C12orf62, a factor that couples COX I synthesis to cytochrome c oxidase assembly, cause fatal neonatal lactic acidosis ...... 101

Abstract ...... 102

Introduction ...... 103

Results ...... 105

Discussion ...... 114

Materials and Methods ...... 118

Figure Legends...... 130

Figures...... 136

General Discussion ...... 147

References ...... 159

Appendix ...... 179

Acknowledgements

My Ph.D. thesis and my Ph.D. life would not be successful without these people: I would like to give a big thank you to Prof. Eric Shoubridge for his guidance and support throughout my graduate studies, his enthusiasm is very contagious and he has always been a source of new and interesting ideas. Having a good supervisor is very important and thanks to Eric, my love for doing research has but grown over the years.

Next I would like to express my gratitude to the other members of the lab past and present who have always been around if I needed help and made the lab a fun and enjoyable place that I never would like to leave. In no particular order I would like to thank Hana Antonicka for her teachings of lab techniques and for always being there with insightful advice about my project and my thesis; Florin Sasarman for being such a good teacher of the translation assay and for the time he took to discuss my project and my thesis with me; Olga Zurita, Tamiko Nishimura and Lissiene Silva Neiva for being the best companions both in and outside the lab; Timothy Johns for making my life in the lab easier, for the localization experiments, and generally having an answer for everything;

Neil Webb for his help with the size exclusion experiment; Isabelle Thiffault for her translation skills; Vincent Paupe, Alex Janer, Stephen Fung, Steven Salomon who make the Shoubridge lab the awesome place that it is; Scot Leary for teaching me the COX assay, and Timothy Wai and Brendan Battersby who were always there with more information than I could ask for.

Thanks to my advisory committee, Prof. David Rosenblatt and Prof. Greg Brown for their advice and suggestions during the committee meetings.

I would also like to thank the Thai government for the scholarship I received. My special thanks to my loving parents in Thailand, Manu and Jintana, and my brother,

Jirasak who despite the distance have always been there for me. I also thank the other members of my family and friends both in Thailand and Canada who in times of need have never let me down. Finally, my best friend and husband, Pira, with whom I shared both the stressful and happy moments of this journey.

Abstract

In mitochondria, ATP is generated by oxidative phosphorylation (OXPHOS), a process that requires five multimeric enzyme complexes. Electrons are passed along the first four enzyme complexes (complex I-IV) that make up the mitochondria respiratory chain, releasing energy that is stored in the form of a proton gradient across the mitochondrial inner membrane, and is subsequently used by the ATP synthase (complex

V) to produce ATP. Complex IV or cytochrome c oxidase (COX) is the terminal enzyme in the mitochondrial respiratory chain, catalyzing the oxidation of cytochrome c by molecular oxygen. It contains 13 structural subunits in mammals, 3 of which are encoded by mitochondrial DNA. Cytochrome c oxidase deficiencies can be caused by mutations in either mitochondrial or nuclear DNA. COX deficiency can result from mutations in the structural subunits or factors necessary for the assembly of the enzyme complex. In this thesis, two novel genes mutated in two subjects with COX deficiency have been identified. First, we identified a specific defect in the synthesis of the mtDNA-encoded

COX subunit 1 (COX I) in a pedigree segregating late-onset Leigh Syndrome and COX deficiency. We mapped the defect to chromosome 17q by microcell-mediated chromosome transfer and identified a homozygous single base pair insertion causing a premature stop in CCDC44, renamed TACO1 for translational activator of COX I.

TACO1 is a member of a large family of hypothetical proteins containing a conserved

DUF28 domain that localizes to the mitochondrial matrix. Expression of the wild-type cDNA restored TACO1 protein and rescued the translation defect. TACO1 is the first specific mitochondrial translational activator identified in mammals. Respiratory competence, mitochondrial translation and COX activity were normal in yeast strain

deleted for the orthologue YGR021w, suggesting that TACO1 has evolved a novel function in mammalian mitochondrial translation. Secondly, we studied a family in which the subject presented with severe congenital lactic acidosis and dysmorphic features associated with a COX assembly defect and a specific decrease in the synthesis of COX I.

Using a combination of microcell mediated chromosome transfer, homozygosity mapping, and transcript profiling we mapped the gene defect to chromosome 12, and identified a homozygous missense mutation causing an amino acid change from methionine to isoleucine in C12orf62, a gene apparently restricted to the vertebrate lineage. Expression of the wild-type cDNA restored C12orf62 protein levels, and rescued the COX I synthesis and COX assembly defect. C12orf62 is a very small (6 kDa), uncharacterized, single transmembrane protein that localizes to mitochondria. COX I, II and IV subunits co-immunoprecipitated with an epitope-tagged version of C12orf62, and

2D BN-PAGE analysis of newly synthesized mitochondrial COX subunits in subject fibroblasts showed that COX assembly was impaired, and the nascent enzyme complex unstable. We conclude that C12orf62 is required for coordinating the early steps of COX assembly with the synthesis of COX I.

Abstract (French)

Dans les mitochondries, l'ATP est généré par la phosphorylation oxydative

(PHOSOX), un processus qui nécessite cinq complexes enzymatiques multimériques. Le transport des électrons le long des quatre premiers complexes enzymatiques (complexes

I-IV, qui composent la chaîne respiratoire mitochondriale) libère l'énergie qui est stockée sous la forme d'un gradient de protons à travers la membrane interne mitochondriale et est ensuite utilisée par l'ATP synthétase (complexe V) pour produire de l'ATP. Le complexe IV ou cytochrome C oxydase (COX) est l'enzyme terminale de la chaîne respiratoire mitochondriale, catalysant l'oxydation du cytochrome c par l'oxygène moléculaire. Il contient 13 sous-unités structurelles chez les mammifères, dont 3 sont codées par l'ADN mitochondriale. Les déficiences en cytochrome C oxydase peuvent être causées par des mutations dans l'ADN mitochondriale ou l'ADN nucléaire. Les carences en COX peuvent être liées à des mutations dans les sous-unités structurelles ou à des mutations dans des facteurs nécessaires à l'assemblage du complexe enzymatique. Dans cette thèse, deux nouveaux gènes mutés ont été identifiés et caractérisés dans deux patients présentant un déficit en COX. Premièrement, nous avons identifié un défaut spécifique dans la synthèse de la sous-unité COX 1 de l’ADN mitochondriale (COX I) dans un pedigree présentant une apparition tardive du syndrome de Leigh et une carence en COX. Nous avons cartographié le défaut génétique au chromosome 17q par la technique de transfert de chromosomes à médiation microcellulaire. Nous avons, par la suite, identifié une mutation homozygote, une insertion d'une base causant l’apparition prématurée d'un codon stop qui entraîne l'arrêt de la synthèse de la protéine CCDC44, renommé TACO1 pour activateur de la traduction de la COX I. TACO1 est membre

d'une grande famille de protéines contenant un domaine conservé à fonction inconnue, nommé DUF28, qui se localise à la matrice mitochondriale. L'expression de l'ADN complémentaire de type sauvage de TACO1 compense le défaut de traduction de COX I.

TACO1 est le premier activateur spécifique de la traduction mitochondriale à être identifié chez les mammifères. Il a été observé que l’absence (knock-down) du gène codant l’orthologue de TACO1 chez la levure, le YGR021w, ne perturbait pas la compétence des voies respiratoires, la traduction mitochondriale, ni l'activité de COX.

Ceci suggère que TACO1 a évolué et a acquérit une nouvelle fonction dans la traduction mitochondriale chez les mammifères. Deuxièmement, nous avons étudié une famille dans laquelle le sujet présentait une acidose lactique congénitale grave et dysmorphie associée

à un défaut d’assemblage et une diminution de l’activité enzymatique de la COX due à un défaut spécifique dans la traduction de COX I. En utilisant une combinaison de techniques dont le transfert de chromosomes à médiation microcellulaire, la cartographie d'homozygotie et le profilage de transcription, nous avons cartographié le gène défectueux sur le chromosome 12. Nous avons identifié une mutation faux sens à l'état homozygote provoquant un changement d'acide aminé de méthionine en isoleucine dans le gène C12orf62, un gène qui semble restreint à la lignée des vertébrés. L'expression de l'ADN complémentaire de type sauvage de C12orf62 a restauré la synthèse de COX I et le défaut d'assemblage de la COX. C12orf62 est une très petite protéine transmembranaire (6 kDa), non caractérisée, qui se localise aux mitochondries. Les sous- unités COX I, II et IV co-immunoprécipitent avec un épitope marqué de la protéine

C12orf62. Les analyses de bleu d'électrophorèse sur gel de polyacrylamide natif (BN-

PAGE) en deux dimensions pour les sous-unités nouvellement synthétisées de la COX

mitochondriale ont démontré que la COX assemblée est altérée et que le complexe enzymatique naissant est instable dans les fibroblastes du patient atteint. Nous concluons que C12orf62 est nécessaire pour coordonner les étapes précoces de l’assemblage de la

COX et de la synthèse de COX I.

Contributions of Authors

Chapter 1 is a manuscript published in July 2009 in Nature Genetics, 41(7):833-

837, by Woranontee Weraarpachai, Hana Antonicka, Florin Sasarman, Jügen Seeger,

Bertold Schrank, Jill E. Kolesar, Hanns Lochmüller, Mario Chevrette, Brett A. Kaufman,

Rita Horvath and Eric A. Shoubridge. All the experiments in this chapter were designed by Woranontee Weraarpachai under the supervision of Prof. Eric A. Shoubridge.

Woranontee Weraarpachai did the chromosome transfers, mutation analyses, subcellular localization, RNA immunoblotting analyses, BN gel analyses, SDS gel analyses, enzyme measurements; Hana Antonicka did BN gel analyses, enzyme measurements, the molecular modeling, the copy number analyses and helped writing the manuscript; Florin

Sasarman performed translation analyses; Jürgen Seeger evaluated the index subject and affected children; Bertold Schrank evaluated the adult subjects; Jill E. Kolesar did the yeast studies; Hanns Lochmüller performed the linkage analysis; Mario Chevrette helped with the chromosome transfer studies; Brett A. Kaufman did the size exclusion experiments and helped with the yeast studies; Rita Horvath performed the histological, biochemical and genetic investigation of the index subject and family members.

Chapter 2 is a manuscript submitted for publication in the American Journal of

Human Genetics, by Woranontee Weraarpachai, Florin Sasarman, Tamiko Nishimura,

Hana Antonicka, Karine Auré, Agnès Rötig, Anne Lombès, Eric A. Shoubridge. All the experiments presented in this chapter were designed and executed by Woranontee

Weraarpachai under the supervision of Prof. Eric A. Shoubridge except for the following contributions: Florin Sasarman performed some of the translation analyses; Tamiko

Nishimura helped with the knockdown experiments; Hana Antonicka analyzed the data 12

from homozygosity mapping; Karine Auré and Agnès Rötig performed the mutation analyses in family members of the index subject; Anne Lombès performed the histological and biochemical investigation of the index subject.

Abbreviations

2-D: two dimensional

A. aeolicus: Aquifex aeolicus aa: amino acid

ADP: adenosine diphosphate

ATP: adenosine triphosphate

ATP6 & 8: subunits of complex V of the oxidative phosphorylation

BN: blue native bp: base pair cDNA: complementary DNA

COX: cytochrome c oxidase

CSB: conserved sequence blocks

CsCl: cesium chloride

CSF: cerebrospinal fluid

Cys: cysteine

Cyt b: subunits of complex III of the oxidative phosphorylation

D-loop: displacement loop of the mitochondrial DNA

DDM: n-dodecyl β-D-maltopyranoside

DMEM: Dulbecco’s modiﬁed Eagle's medium

DNA: deoxyribonucleic acid

E. coli: Escherichia coli

EDTA: ethylenediaminetetraacetic acid

FADH2: flavin adenine dinucleotide, reduced form

FBS: fetal bovine serum

FISH: fluorescent in situ hybridization fMet: N-formylmethionine

FMN: flavin mononucleotide

GDP: guanosine diphosphate

GTP: guanosine triphosphate

H-strand: heavy strand

HA: hemagglutinin

HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HMG: high mobility group

HSP: heavy-strand promoter

HyTK: hygromycin phosphotransferase-thymidine kinase

Ile: isoleucine

IF: initiation factor

IMS: intermembrane space

IP: immunoprecipitation kb: kilobase kDa: kilodalton

L-strand: light strand

Leu: leucine

LHON: lebers hereditary optic neuropathy

LSP: light-strand promoter

Lys: lysine

MELAS: mitochondrial encephalomyopathy, lactic acidosis, stroke-like episodes

MERRF: myoclonus epilepsy, ragged-red fibers

Met: methionine 15

mRNA: messenger ribonucleic acid

MRI: magnetic resonance imaging mt: mitochondrial mtDNA: mitochondrial DNA mtEFG: mitochondrial elongation factor G mtEFTs: mitochondrial elongation factor Ts mtEFTu: mitochondrial elongation factor Tu mtIF: mitochondrial initiation factor mtRF: mitochondrial releasing factor mtRRF: mitochondrial ribosomal recycling factor mtSSB: mitochondrial single stranded binding protein mtTFA: mitochondrial transcription factor A mtTFB: mitochondrial transcription factor B mTERF: mitochondrial transcription termination factor

NADH: nicotinamide adenine dinucleotide, reduced form

ND: subunits of complex I of the oxidative phosphorylation

OD: optical density

OH: heavy strand origin of replication

OL: light strand origin of replication

OXPHOS: oxidative phosphorylation

PAGE: polyacrylamide gel electrophoresis

PBS: phosphate buffered saline

PCR: polymerase chain reaction

PEO: progressive external ophthalmoplegia

Phe: phenylalanine 16

PMSF: phenylmethylsulfonyl fluoride

POLɣ: DNA polymerase gamma

POLRMT: mitochondrial RNA polymerase

PPR: pentatricopeptide repeat protein

RF: releasing factor

RFLP: restriction fragment length polymorphism

RNA: ribonucleic acid

RNAi: RNA interference

RRF: ribosomal recycling factor rRNA: ribosomal nucleic acid

RT-PCR: reverse transcription PCR

S. cerevisiae: Saccharomyces cerevisiae

SDS: sodium dodecyl sulfate siRNA: small interfering RNA

SNP: single-nucleotide polymorphism ssDNA: single-stranded DNA

TAS: termination-associated sequence

TCA: trichloroacetic acid

TFAM: mitochondrial transcription factor A tRNA: transfer ribonucleic acid

TTC: triphenyl tetrazolium chloride

UTR: untranslated region

Val: valine

YEPD: yeast extract peptone dextrose

YPGal: yeast peptone galactose 17

YPGly: yeast peptone glycerol

General Introduction

Mitochondria are the energy generating organelles of mammalian cells. They generate energy for cellular processes through oxidative phosphorylation (OXPHOS).

This process requires five enzyme complexes (Complex I to V), where the first four comprise the mitochondrial respiratory chain and complex V generates ATP from energy generated by electron transport along the respiratory chain. Defects in any of these complexes can result in disease. This thesis focuses on defects in complex IV (also known as cytochrome c oxidase, or COX). This complex is composed of 13 subunits, encoded by both nuclear and mitochondrial DNA. Mutations can occur in either the structural subunits or the assembly process. The purpose of this thesis was to identify novel genes that are responsible for COX defects in two different subjects with COX deficiencies.

I will first describe the oxidative phosphorylation (OXPHOS) system and the basic structure and function of mitochondria. I then describe the mitochondrial genome and its regulation, focusing on mitochondrial translation and the structure and assembly of the COX complex. I will also review what is known about the genetic basis for mitochondrial translation defects and COX deficiency.

Introduction

Part 1: Oxidative Phosphorylation

Oxidative phosphorylation is the process that generates energy in the form of ATP as a final result of a transfer of electrons from NADH or FADH2 to O2 by a series of electron carriers (Stryer 1995). The biochemical pathway is composed of 5 enzymatic complexes and takes place in the inner membrane of mitochondria: complex I (NADH: ubiquinone oxidoreductase), complex II (succinate: ubiquinone oxidoreductase), complex

III (ubiquinol: cytochrome c oxidoreductase), complex IV (cytochrome c oxidase) and complex V (ATP synthase). The first four complexes make up the respiratory chain or the electron transport chain and complex V is the ATP synthase. Electrons flow through acceptors of increasing electron affinity; the excess energy is used to pump protons against the concentration gradient across the inner membrane (two protons per one electron), and the energy released as protons flow back along their concentration gradient is used by complex V to synthesize ATP from ADP (reviewed in (Taanman 1997)). The subunits of the five enzyme complexes are encoded in both the nuclear and mitochondrial genomes except complex II whose subunits are entirely encoded by nuclear genes

Complex I or NADH: ubiquinone oxidoreductase is the first and the largest of the respiratory chain complexes. It is composed of 45 subunits, 7 of which are encoded by the mitochondrial genome. It contains one molecule of flavin mononucleotide (FMN) and up to eight non-heme iron-sulfur clusters (van den Heuvel and Smeitink 2001). It couples the oxidation of NADH to the reduction of ubiquinone. The drop in energy associated

with this process translocates four protons across the inner membrane, for each molecule of NADH oxidized.

Complex II or succinate:ubiquinone oxidoreductase is composed of two soluble proteins, the SDHA and SDHB subunits, which are anchored on the matrix side of the inner membrane by two membrane subunits which are SDHC and SDHD. The soluble proteins have succinate dehydrogenease activity and the membrane proteins have cytochrome b and ubiquinone binding sites. Complex II catalyses the oxidation of succinate to fumarate, during which electrons are transported from FADH2 to the ubiquinone pool. The energy released in this process is not enough for proton pumping.

The reduced form of ubiquinone functions as a shuttle molecule for electrons between complex I, II and III. Complex III or ubiquinol:cytochrome c oxidoreductase is composed of 11 subunits, 10 of which are encoded by nuclear genome, and cytochrome b which is encoded by the mitochondrial genome. It contains two b-cytochromes, one cytochrome c1 and one [2Fe-2S] cluster. It catalyzes the electron transfer from ubiquinone to cytochrome c with the coupled translocation of four protons across the inner membrane, per pair of electrons transferred.

Cytochrome c shuttles the electrons from complex III to complex IV. Complex IV or cytochrome c oxidase is the terminal enzyme in the mitochondrial respiratory chain and catalyzes the reduction of molecular oxygen by reduced cytochrome c, leading to a translocation of two protons across the membrane, for each pair of electrons transferred

The resulting electrochemical gradient is used to drive the synthesis of ATP by complex V or ATP synthase. Complex V consists of 14 subunits, of which 2 are encoded

by the mitochondrial genome. The translocation of three protons drives the synthesis of one molecule of ATP. This complex consists of two parts, F0 which spans the inner membrane of mitochondria and functions in proton translocation and F1 which is in the matrix and contains the catalytic center. The translocation of three protons through complex V drives the synthesis of one molecule of ATP (van den Heuvel and Smeitink

2001).

Part 2: Mitochondrial Genome

Mitochondrial DNA structure and organization

Mitochondria not only have an important role in oxidative phosphorylation

(OXPHOS) which is the process that generates ATP, but also perform other functions including ion homeostasis, intermediary metabolism and apoptosis (Newmeyer 2003;

Schatz 1995). Mitochondria have a smooth outer membrane and an invaginated inner membrane. The number of invaginations, known as cristae, vary depending on the respiratory needs and on cell type (Voet and Voet 2004). Since oxidative phosphorylation takes place on the inner membrane of the mitochondria, the respiratory rate varies with the membrane surface area. The inner compartment is known as the matrix. In the matrix, there are soluble enzymes and chemical intermediates of energy metabolism, as well as the mitochondrion’s own genetic machinery, including DNA, RNA and ribosomes. The mitochondrial ribosomes are used to translate mRNA transcribed from mitochondrial

DNA (mtDNA).

Mitochondria have their own genome set apart from the nuclear genome. The mitochondrial genome consists of a 16.6 kb circular double-stranded DNA with no introns (Figure 1). The heavy strand (H) is rich in guanines; the light strand (L) is rich in cytosines and this nomenclature is based on their buoyant densities in alkaline CsCl gradients (Clayton 1982). There is also a small section (1.1 kb) consisting of a triple-

DNA strand called 7SDNA or the D-loop (Clayton 1992) and this is the only noncoding region in mtDNA. The human mitochondrial genome contains 37 genes, 13 encoding for structural subunits of the respiratory chain, 2 for rRNAs and 22 for tRNAs (Grossman and Shoubridge 1996). Most of the genes are encoded by the H-strand except for 8 tRNAs and the ND6 gene (a structural subunit of complex I) which are encoded by the L- strand (Taanman 1999). The D-loop contains the origins of replication for the H-strand and the promoters for transcription of both strands, as seen in Figure 1(B) (Taanman

1999).

The human mitochondrial proteome consists of an estimated 1100–1400 distinct proteins (Calvo and Mootha 2010), most of which are encoded by nuclear genes. They are synthesized in the cytosol and then imported into mitochondria (Herrmann 2003).

Although many mitochondrial proteins have known functions, such as those involved in oxidative phosphorylation and the Krebs cycle, these account for only 70–75% of mammalian mitochondrial proteins, and the function of the other 25-30% remains elusive

(Pagliarini et al. 2008).

Normally, there are about 103-104 copies of mitochondrial DNA per cell and the mtDNA is packed in a DNA-protein structure called the nucleoid at approximately 2-10 copies per nucleoid (Legros et al. 2004; Satoh and Kuroiwa 1991). mtDNA is maternally 23

inherited (Salvatore Dimauro 2003) and in a healthy individual, only wild-type mtDNA molecules are present; this is called homoplasmy. Heteroplasmy refers to the co- existence, in the cells and tissues of an individual, of wild-type and mutated mtDNA molecules. Mutations in mtDNA can be present in either homoplasmic state (Smeitink et al. 2001) or in a heteroplasmic state. For a pathogenic mutation, the level of heteroplasmy must exceed a certain threshold for expression of a biochemical phenotype and the severity of the disease correlates with the amount of mutated mtDNA. This is called the threshold effect. The threshold can be different for different mutations and different tissues. Mitochondrial DNA mutations can arise sporadically in somatic tissues because mutations may undergo clonal expansion and reach the threshold level at a later age

(reviewed in (Ylikallio and Suomalainen 2011)).

(

Figure 1. (A) Map of human mtDNA encoding 13 proteins, 22 tRNAs and 2 rRNAs.

The origins of H-strand (OH) and L-strand (OL) replication and the direction of DNA synthesis are indicated by black arrows; the initiation of transcription sites for L-strand

(LSP), H-strand (H1, H2) and the direction of RNA synthesis are denoted by white arrows. The D-loop is shown in blue. The binding site for the mitochondrial transcription

terminator (TERM) is indicated. The 22 tRNA genes are depicted by the single letter code for the amino acid. The genes coding for the two rRNA (12S and 16S) and 13 protein coding genes are shown. CO I-III, subunits of complex IV; ND1-5, subunits of complex I; ATP6,8, subunits of complex V; Cyt b, subunit of complex III. (B) Schematic representation of the D-loop regulatory region. The three conserved sequence blocks

(CSB I, CSB II, and CSB III) are located just downstream of light strand promoter (LSP).

HSP, heavy strand promoter; LSP, light strand promoter; OH, origin of H-strand DNA replication (Falkenberg et al. 2007).

Mitochondrial DNA replication

There are two models for mtDNA replication: strand-displacement replication

(Clayton 1991a; Shadel and Clayton 1997) and strand-coupled replication (Bowmaker et al. 2003; Holt et al. 2000). In the strand-displacement model, the synthesis of the leading strand of the H-strand starts from the H-strand origin (OH), which is located within the D- loop and proceeds down about two-thirds of the mtDNA until a second origin (OL) is exposed and then the synthesis of the L-strand continues in the opposite direction. The replication from the H-strand origin (OH) is primed by transcription of the LSP (Chang and Clayton 1985) to initiate the replication. There are three conserved sequences downstream of LSP (CSB I, II and III) (Walberg and Clayton 1983). CSB II increases the stability of an RNA-DNA hybrid and transcription termination (Pham et al. 2006; Xu-B and Clayton-DA 1995) forming a close relationship between transcription and replication of mtDNA. Recently, it has been shown that there may be multiple origins on the L- strand in the strand-displacement model (Brown et al. 2005). 26

The strand-coupled replication model suggests that mtDNA replicates symmetrically, with leading and lagging strands syntheses progressing from multiple bidirectional replication forks within a broad zone (Bowmaker et al. 2003). Both models require a primase to prime replication of the L-strand (Wong and Clayton 1985). The termination event happens at the Termination-associated sequences (TAS) which are short (15bp) DNA elements conserved in vertebrates and located at the 3’ end of D-loop

H-strands that also promote D-loop formation (Doda et al. 1981; Madsen et al. 1993).

Both mechanisms of mtDNA replication require that the TWINKLE helicase unwinds the duplex DNA template. The mitochondrial ssDNA-binding protein (mtSSB) stabilizes the unwound conformation and stimulates DNA synthesis by the POLɣ holoenzyme

(POLɣAB2), which is composed of a catalytic subunit POLɣA that forms a heterotrimer with the mtDNA polymerase accessory subunit POLɣB (Fan et al. 2006; Yakubovskaya et al. 2006)(reviewed in (Falkenberg et al. 2007)). Replication of mtDNA is not tightly coupled to the cell cycle. Thus, during mitosis some mtDNA may replicate more than once and some might not replicate at all (Clayton 1991b) which results in differences in mtDNA copy number between cells.

Mitochondrial DNA transcription

Mitochondrial transcription is initiated by three promoters consisting of two H

(heavy) strand promoters: HSP1 (Heavy strand promoter1) and HSP2 (Heavy strand protmoer2) and one L (light) strand promoter: LSP (light strand promoter). Transcription originates from HSP2 and LSP to produce long polycistronic precursor RNAs. The transcription from HSP2 produces almost the entire H-strand while LSP transcribes 8 27

tRNAs and one subunit of complex I (ND6). Then tRNAs are excised from these polycistronic products to generate mature mRNAs and rRNAs. The transcription from

HSP1 creates a shorter message containing the two rRNAs, tRNAPhe and tRNAVal and terminates at a specific site in the tRNALeu gene downstream of the 16S rRNA. The termination site for HSP2 is upstream of the tRNAPhe coding region and no studies have determined the termination site for LSP.

The enzyme responsible for transcription is mitochondrial RNA polymerase

(POLRMT). To initiate transcription, POLRMT needs the assistance of mitochondrial transcription factor A (mtTFA, or TFAM)) and one of the two mitochondrial transcription factor B paralogs (mtTFB1 and mtTFB2, or TFB1M and TFB2M). It has been shown that mtTFB2 is a much more active transcription factor than mtTFB1 in vitro

(Falkenberg et al. 2002). mtTFB1 and mtTFB2 also have rRNA methyl transferase activity but mtTFB2 is a less efficient enzyme than mtTFB1 (Cotney and Shadel 2006;

Seidel-Rogol et al. 2003). Recently, it has been shown that only mtTFB2 is a transcription factor (Litonin et al. 2010; Metodiev et al. 2009; Scarpulla 2008; Sologub et al. 2009). On the other hand, conditional knockout has revealed that mtTFB1 has no major role in mtDNA transcription but rather is an essential methyltransferase that dimethylates two highly conserved adenines of the 12S rRNA (Metodiev et al., 2009).

The modification of 12S rRNA is crucial for the integrity of the small mitochondrial ribosomal subunit and mitochondrial translation is abolished in its absence (Metodiev et al., 2009).

The TFAM protein contains two tandem High Mobilty Group (HMG) box domains separated by a 27-amino acid residue linker and followed by 25 residue of C- 28

terminal tail. TFAM is important for specific DNA recognition and transcription activation (Dairaghi et al. 1995). TFAM binds specific sequences upstream of HSP and

LSP to allow structural alteration of mtDNA at a precise position in the promoter region and unwinds the promoter to permit mtTFB2 to bind to ssDNA and recruit POLRMT to the promoter. In addition, TFAM also directly interacts with mtTFB2 to possibly recruit the transcription machinery to the promoter (McCulloch and Shadel 2003; Ringel et al.

2011). Human mitochondrial transcription termination factors have also been identified: mTERF1 through to mTERF4 (Linder et al. 2005). The termination of HSP1 is mediated by mTERF1. Recently, using a knockout mouse model, mTERF3 was shown to be a negative regulator of mtDNA transcription (Park et al. 2007). Inactivation of mTERF2 in the mouse shows that mTERF2 is also involved in the regulation of mitochondrial transcription (Wenz et al. 2009). The conditional knockout of mTERF4 leads to defective ribosomal assembly and a drastic reduction in translation, demonstrating that it is an important regulator of translation in mammalian mitochondria (Camara et al. 2011

Mitochondrial translation

The mitochondrial translation system is vital because it synthesizes 13 highly hydrophobic structural proteins of the OXPHOS. This system localizes to the inner membrane, on its matrix side (Liu and Spremulli 2000) and needs many components originating from both the nuclear genome and the mitochondrial genome. The metazoan mitochondrial translation mechanism is very different from those of the bacterial and

eukaryotic cytoplasmic translation, and it is also much less understood. The polypeptides encoded by the mammalian mitochondrial DNA are synthesized from nine mono- cistronic and two bicistronic transcripts, both of which have overlapping reading frames

(Anderson et al. 1982). Some of the transcripts lack a complete stop codon, so the post- transcriptional addition of poly (A) tails generates a functional termination codon (Ojala et al. 1981). In mitochondrial DNA, there are no introns and the genetic code differs from the universal code. For example, the mammalian mitochondrial code for methionine is

AUA which in the universal code represents isoleucine. The mitochondrial stop codon is represented by AGA and AGG which represent nuclear arginine. Since there are only 22 tRNAs encoded by mitochondrial DNA and none is imported from the nuclear DNA, a single tRNA must be able to read all codons of a four-codon family.

There is only one mitochondrial tRNA for methionine, for both initiation and elongation, depending on the presence or absence of the formyl group (Mikelsaar 1983).

Mitochondrial ribosomes are different from cytoplasmic (80S) and prokaryotic ribosomes

(70S) in size. Mitochondrial ribosomes have sedimentation coefficients of 55S and are composed of a small subunit, 28S, and a large subunit, 39S. The mitochondrial ribosome is made up of 12S and 16S rRNAs and 77 ribosomal proteins, while prokaryotic ribosomal rRNAs have sizes of 16S and 23S. There is also a 5S rRNA in prokaryotic cells, in cytoplasmic ribosomes and in plant mitochondrial ribosomes. It has also been suggested to be present in highly purified mammalian mitochondria but the function in translation is unclear (Magalhaes et al. 1998). MicroRNAs were isolated from mitochondria (Kren et al. 2009) and the nuclear-encoded 5S rRNA was identified as one of the most abundant RNAs in human mitochondria (Entelis et al. 2001; Smirnov et al.

2008). Recently, PNPASE (polynucleotide phosphorylase) was shown to play a role in regulating the import of nuclear-encoded RNAs, including the 5S rRNA, into the mitochondrial matrix (Wang et al. 2010).

Mitochondrial ribosomes have a higher protein to RNA ratio than bacterial ribosomes (O'Brien 2002) while bacterial ribosomes have 65% RNA, mammalian mitochondrial ribosomes have only 33% RNA. This low amount of RNA reflects the reduction in size of the rRNA and is made up by an increase in the number of ribosomal proteins (Koc et al. 2001). There are 29 mitochondrial ribosomal proteins in the small subunit, with 14 proteins being homologs to the E. coli 30S ribosomal proteins (Cavdar

Koc et al. 2001), and 48 ribosomal proteins in the large ribosomal subunit, with 28 proteins being homologs to the E. coli 50S ribosomal proteins (Koc et al. 2001). The mechanism of the assembly of mammalian mitoribosomes is still not well understood. In yeast, the products of the MTG1 and MTG2 genes were shown to act as assembly factors of the large ribosomal subunit (Barrientos et al. 2003; Datta et al. 2005). The human homolog of the MTG1 gene was shown to partially rescue the mitochondrial translation defect and respiratory phenotype of MTG1 yeast mutant (Barrientos et al. 2003). There is also a yeast mitochondrial AAA protease which was shown to process the yeast mitochondrial ribosomal proteins MrpL32, which then joins the preassembled ribosomal particles close to the inner membrane and completes ribosomal assembly (Nolden et al.

2005).

Previously, a protein called ERAL1 (Era G-protein-like 1) was identified, which is the human ortholog of the bacterial Ras-like protein, and is localized in the mitochondrial matrix. ERAL1 was reported to have a function in the formation of the 28S 31

small mitoribosomal subunit by acting as an RNA chaperone for the 12S rRNA. ERAL1 is associated with mitoribosomal proteins and binds in vivo to 12S rRNA. Depletion of

ERAL1 resulted in 12S instability and loss of newly synthesized 28S subunits. Moreover,

ERAL1 depletion induces apoptosis prior to the mitochondrial translation defect

(Dennerlein et al. 2010; Uchiumi et al. 2010).

Mammalian mitochondrial mRNAs lack 5’ untranslated regions (UTRs) and are not capped. Conversely, yeast mitochondrial mRNAs have 5’UTRs that are recognized by membrane-bound mRNA-specific translational activator proteins to initiate translation

(Fox 1996; Green-Willms et al. 1998), and cytoplasmic mRNAs are capped and have 5’

UTRs. In plants, pentatricopeptide domain RNA-binding proteins regulate the stability, expression and translation of mitochondrial transcripts. In mammals, there is a pentatricopeptide repeat domain 1 protein (PTCD1) which is a matrix protein and it was shown to be associated with leucine tRNA and precursor RNAs that contain leucine tRNAs. Knockdown of PTCD1 in cells resulted in no change of mitochondrial mRNA levels but it increased precursor RNAs and leucine tRNA and overexpression of PTCD1 led to a reduction of these RNAs. Lowering PTCD1 in cells, causes increased levels of several mitochondria-encoded proteins and complex IV activity. This data suggests that

PTCD1 acts as a negative regulator of leucine tRNA levels and hence mitochondrial translation (Rackham et al. 2009). Another mammalian pentatricopeptide domain protein is pentatricopeptide repeat domain protein 3 (PTCD3). This protein was shown to associate with the small subunit of mitochondrial ribosomes and the 12S mitochondrial rRNA, and to have a function in translation. PTCD3 is not involved in RNA processing and stability because PTCD3 knockdown and overexpression did not affect mitochondrial

mRNA levels. However, lowering PTCD3 in cells decreased mitochondrial protein synthesis, mitochondrial respiration and the activity of complexes III and IV (Davies et al. 2009).

In recent literature, LRPPRC (leucine-rich pentatricopeptide repeat cassette), a weak human homolog of yeast Pet309p, which is the translational activator for COX1, was shown to bind to SLIRP (a stem-loop RNA-binding protein) and regulate the stability and handling of mature mRNAs for translation (Sasarman et al. 2010). The translation process takes place in three steps: initiation, elongation and termination, each requiring a number of specific factors. There are 2 initiation factors, three elongations factors, one release factor and two ribosomal recycling factors that have been cloned and sequenced in several mammalian species, including humans. As it was mentioned before, initiation of translation of yeast mitochondrial mRNAs requires message- specific activators, which bind 5’ UTRs and will mediate the interaction of mitochondrial mRNA, small subunits of mitochondrial ribosome and the inner mitochondrial membrane (Fox 1996). Some yeast mitochondrial mRNAs require specific activators for their translation for example,

Cox1p, Cox2p, Cox3p, Atp9p and cytochrome b have their own translational activators.

It is not known how selection of the initiation codon takes place in yeast mitochondria. It was suggested that in yeast, sites outside the initiation codon exist to correct positioning of the ribosome by mRNA-specific translational activators since site-directed mutagenesis of the AUG initiation codon of COX2 and COX3 mRNAs to AUA decreased translation several fold, but the residual translation is initiated at the AUA initiation codon (Fox 1996). Translation initiation and selection of the initiation codon in mammalian mitochondria cannot be explained with a similar mechanism to prokaryotic,

cytosolic or yeast translation since there is neither a cap, nor a Shine-Dalgarno sequence, nor a 5’UTR. In general, the mitochondrial mammalian start codon is located within three nucleotides from the 5’ end of the mammalian mitochondrial mRNAs (Montoya et al.

1981). In mammalian mitochondrial mRNAs other mechanisms must exist to initiate translation.

In the mitochondrial translation machinery, two mammalian mitochondrial initiation factors (mtIF2 and mtIF3) have been identified (Koc and Spremulli 2002; Liao and Spremulli 1991), while bacteria have three translation initiation factors (IF1, IF2 and

IF3), meaning that there is no ortholog of IF1 in mammalian mitochondria. mtIF2 promotes the binding of fMet-tRNA to mitochondrial ribosomes in the presence of GTP and an RNA template containing a 5’ terminal AUG. Previously, mtIF2 was shown to have both functions of bacterial IF1 and IF2, namely binding of the mitoribosome, and the formation of the initiation complex (Gaur et al. 2008). It has also been shown that mammalian mitochondrial 55S ribosomes preferentially form initiation complexes at a

5’-termial AUG codon over an internal AUG (Christian and Spremulli 2010). mtIF3 was shown to promote the dissociation of the 55S mitoribosomes and permitting the assembly of the initiation complex (Christian and Spremulli 2009; Koc and Spremulli 2002). mtIF3 also plays a role in positioning the start codon in the P site before the mtIF2 -dependent binding of fMet-tRNA. Figure 2 shows the translation process in detail.

Figure 2. Schematic of mammalian mitochondrial translation. Two mitochondrial initiation factors (IF2 and IF3), three elongation factors (EFTu, EFTs, EFG1) three termination factors (RF1a, RRF and EFG2) are involved in the synthesis of mtDNA- encoded polypeptides. Other factors implicated in mitochondrial translation regulation include C12orf65, Ict1 and LRPPRC.

After initiation, the process continues with the elongation step. There are three elongation factors (mtEFTs, mtEFTu and mtEFG,) that have been found in human mitochondria and all have bacterial orthologs (Hammarsund et al. 2001; Ling et al. 1997;

Xin et al. 1995). Elongation factor mtEFTu forms a ternary complex with GTP and aminoacylated tRNA, which brings the aminoacyl-tRNA to the accepter A site for decoding. After GTP hydrolysis, mtEFTu:GDP is converted back to mtEFTu:GTP with the concomitant release of GDP by the elongation factor mtEFTs. At this point, one more amino acid has been added to the growing peptide chain. mtEFTu has also been reported to have chaperone activity. It interacts with misfolded newly synthesized polypeptides from the mitochondrial ribosome, prevents thermal aggregation of proteins in vitro, and enhances protein refolding in vitro (Suzuki et al. 2007). Two isoforms of mtEFG named mtEFG1 and mtEFG2 have been identified in human mitochondria (Hammarsund et al.

2001). Elongation factor mtEFG1 catalyzes the translocation of peptidyl-tRNA from the ribosomal A site to the P site following peptide bond formation, and requires GTP. The tRNA then leaves via the E site and a new round of elongation cycle can continue. Even though mtEFG1 and mtEFG2 are 35% identical (Spremulli et al. 2004), recently it was shown that mtEFG2 has a function in ribosome recycling and lacks translocation activity

(Tsuboi et al. 2009).

Lastly, the translation process ends with the termination step. The proteins involved in mitochondrial translation termination are still elusive (Richter et al. 2010). In prokaryotes, there are two class I release factors, RF1 and RF2 which recognize three stop codons (UAA, UAG and UGA) and help hydrolysis of the ester bond between the tRNA and the nascent polypeptide. There is one class II release factor, RF3 that has

GTPase activity and stimulates the removal of RF1 and RF2 from ribosomes after the release of the polypeptide chain. Lastly, ribosomal recycling factor (RRF) releases the ribosomes from the mRNA at the stop codon (reviewed in (Petry et al. 2008)). In mammalian mitochondria, there is only one class I release factor mtRF1a/mtRF1L which recognizes UAA and UAG stop codons (Nozaki et al. 2008; Soleimanpour-Lichaei et al.

2007; Temperley et al. 2010), and two ribosome recycling factors, mtRRF (Rorbach et al.

2008; Zhang and Spremulli 1998) and mtRRF2 (mtEFG2) (Tsuboi et al. 2009) which dissociate mitoribosomal subunits, tRNA and mRNA from each other. More recently three other possible mitochondrial class I release factors named mtRF1, Ict1 and

C12orf65 (Antonicka et al. 2010) have been proposed. Ict1 has been reported to be a ribosome-dependent, codon-independent peptidyl-tRNA hydrolase and can release nascent polypeptides from ribosomes stalled on mRNAs lacking a termination codon

(Haque and Spremulli 2010; Richter et al. 2010). The functions of mtRF1 and C12orf65 need further studies. C12orf65 and Ict1 might have overlapping functions since Ict1 could partially rescue the defect found in mutated C12orf65 patient cells; however

C12orf65 lacks peptidyl-tRNA hydrolase activity in an in vitro assay in bacterial ribosomes. The data suggested that it could play a role in processing peptidyl-tRNAs that have been prematurely released during elongation (Antonicka et al. 2010).

Part 3: Human diseases due to mitochondrial translation defect

One of the largest categories of mitochondrial diseases is associated with mitochondrial translation defects. The mutation can be in mtDNA causing a mutation in tRNAs and rRNAs or the mutation can be in nuclear DNA, which could affect tRNA- modifying enzymes, ribosomal proteins, aminoacyl-tRNA synthetases, elongation factors, termination factors and proteins involved in stability and handling of mature protein mRNAs. Any mitochondrial translational defect will result in OXPHOS deficiency affecting all complexes except complex II where none of the subunits are encoded by mtDNA or it could affect only one complex such as the defect in LRPPRC resulting in only COX deficiency.

Most of the human mtDNA point mutations occur in mtDNA-encoded tRNA genes. The first mutation affecting mitochondrial translation that was described is the

A3243 mutation in tRNALeu(UUR). This mutation causes mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes syndrome (MELAS)

(Goto et al. 1990). Another common mutation is A8344G in tRNALys which results in myoclonus epilepsy with ragged red fibers (MERRF). The phenotype is encephalomyopathy with myoclonus, ataxia, hearing loss, muscle weakness and generalized seizures (Shoffner et al. 1990). The third common mutation is tRNAIle which is associated with cardiomyopathy and progressive external ophthalmoplegia (PEO)

(Scaglia and Wong 2008).

Mutations in mitochondria rRNA genes are associated with non-syndromic sensorineural deafness or aminoglycoside-induced deafness. A1555G in the 12S rRNA is

a common mutation (Prezant et al. 1993) which modifies the structure of rRNA and its interaction with aminoglycosides, resulting in a translation defect.

To date, several mutations in nuclear DNA involving different mitochondrial translation process have been identified. In general, the phenotypes are more severe and cause early death. Interesting, among patients with the same mutation different phenotypes and tissue specificity is seen even though all factors are involved in the same process. A summary of mutation in nuclear DNA encoded genes causing mitochondrial translational defects is shown in Table 1 (reviewed in (Rotig 2011)).

Table 1. Nuclear genes responsible for mitochondrial translation disorders in humans.

Number of Age at Gene Protein Clinical presentation Outcome Ref. patients onset tRNA-modifying enzymes 6 Sideroblastic anemia + (Bykhovskaya PUS1 Pseudouridine synthase 1 > 4 patients months– myopathy et al. 2004) 12 years tRNA 5- Death at 3–4 methylaminomethyl-2- 2–4 (Zeharia et al. TRMU 7 patients Liver failure months or thiouridylate months 2009) recovering methyltransferase Ribosomal proteins Mitochondrial ribosomal Agenesis of corpus Death at 2–22 (Miller et al. MRPS16 1 patient Neonat protein S16 callosum, hypotonia days D3 2004) Mitochondrial ribosomal Hypotonia, Death at 2–22 (Saada et al. MRPS22 1 family Neonat protein S22 cardiomyopathy days 2007) Brain anomalies, hypertrophic Alive at 6 (Smits et al. 1 patient cardiomyopathy, Neonat Cornelia de Lange-like years 2011b)

Aminoacyl-tRNA synthetases 3 Pontocerebellar Death before 2 (Edvardson et RARS2 Arginyl-tRNA synthetase 2 Neonat patients/families hypoplasia years al. 2007) (Scheper et al. DARS2 Aspartyl-tRNA synthetase 2 > 30 families Leukoencephalopathy 3–15 yr 2007) Hyperuricemia, pulm Death at 10 (Belostotsky SARS2 Seryl-tRNA synthetase 2 3 families hypertesion,renal 4 months months et al. 2011) failure Sideroblastic anemia + Death at 18– (Riley et al. YARS2 Lysyl-tRNA synthetase 2 2 families Infancy myopathy 26 years 2010) (Pierce et al. HARS2 Histidyl-tRNA synthetase 1 family Perrault syndrome 2–3 yr 2011) Elongation factors (Antonicka et al. 2006), (Coenen et al. 4 patients/ Encephalopathy/liver Death before 2 2004), (Smits GFM1 mtEFG1 Neonat families failure years et al. 2011a)and (Valente et al. 2007) 2 patients/ Encephalomyopathy/ Death at 7 (Smeitink et TSFM mtEFTs Neonat families cardiomyopathy weeks al. 2006) Death at 14 (Valente et al. TUFM mtEFTu 1 patient Leukodystrophy Neonat months 2007)

Termination factor Leigh syndrome, optic Alive at 7 and (Antonicka et C12orf65 2 families atrophy, 1 years 20 years al. 2010) ophthalmoplegia Translation activators Leigh Syndrome French Leucine-rich PPR-motif Death at 6–69 (Mootha et al. LRPPRC 55 patients Canadian variant Infancy containing protein months 2003) (LSFC)

Part 4: Complex IV or cytochrome c oxidase (COX)

COX structure and function

COX is the terminal enzyme of the respiratory chain that catalyzes the oxidation of cytochrome c by molecular oxygen. Human COX is approximately 200 kDa and is composed of 13 structural subunits. The three subunits that form the catalytic core (COX

I, II and III) are encoded by mitochondrial DNA and are similar to those from prokaryotes, which require only four subunits for a fully functional enzyme complex

(Saraste 1990). The other 10 subunits (COX IV, COX Va, COX Vb, COX VIa, COX

VIb, COX VIIa, COX VIIb, COX VIIc and COX VIII) are encoded by the nuclear DNA

(Grossman and Lomax 1997) and then imported into mitochondria (Kadenbach et al.

1983; Taanman 1997). The complex is embedded in the inner mitochondrial membrane with one part extending into the intermembrane space (IMS) and the opposite part in the matrix. It is active when dimerized. In mammalian mitochondria, COX is also found to associate with complex I and complex III within a macromolecular structure called supercomplex, thought to better facilitate the electron flux between complexes (Schagger and Pfeiffer 2000).

COX I is the largest and highly hydrophobic protein composed of 12 transmembrane domains. COX I contains two hemes, a and a3. Heme a3 interacts with a mononuclear copper site (CuB) and forms the heterobimetallic active site. COX II is the smallest and the least hydrophobic subunit of the enzyme core. COX II contains two Cu ions in a binuclear site called CuA. There are other cofactors necessary for function

including zinc, magnesium, calcium and sodium ions (Carr and Winge 2003). COX III is highly hydrophobic with seven transmembrane domains but it does not have any prosthetic groups (Tsukihara et al. 1996). It appears to have a role in the assembly of subunits I and II, and in facilitating proton translocation (Bratton et al. 1999; Hosler

2004). Cytochrome c binds to the subunit II of COX at the outer part of the mitochondrial inner membrane. Electrons are transferred to the bimetallic CuA site, following by heme a and heme a3, which then reacts with the CuB center to reduce molecular oxygen. One proton is pumped across the membrane for each electron transferred (Capaldi 1990;

Tsukihara et al. 1995).

There are ten nuclear encoded smaller subunits whose function remains largely obscure. Subunits COX Va, Vb and VIb are hydrophilic proteins while the rest are hydrophobic proteins that span the membrane once. COX Va is also the only subunit that does not interact with any of the core subunits (Tsukihara et al. 1996). Two of the nuclear subunits (VIa and VIIa) have tissue-specific isoforms (heart and liver isoforms) which display differences in protein properties (reviewed in (Shoubridge 2001)) such as tissue- specific regulation of COX activity by ADP via specific interaction with COX VIa subunit in bovine heart (Anthony et al. 1993). Heart isoforms are expressed in heart and skeletal muscle while liver isoforms are expressed in brain, liver, kidney, lymphoblasts and fibroblasts (Robinson 2000). A testis-specific COX VIb isoform and a lung-specific

COX IV isoform have also been identified (reviewed in (Fernandez-Vizarra et al. 2009)).

In S. cerevisiae, cytochrome c oxidase is composed of three mitochondrially encoded and eight nuclear encoded subunits which are similar to their mammalian counterparts (Taanman 1997). There are more studies of the function of nuclear COX subunits in yeast than in human. In yeast, the homolog of mammalian subunit VIa was shown to be involved in the dimerization of the enzyme while the homolog of mammalian subunit VIIc is necessary for maximal activity of the oxidase complex. In another yeast study, individual null-mutants for the homologs of mammalian subunits IV,

Va, Vb, VIc or VIIa did not show any oxidase activity, suggesting that these subunits might be important to the ordered sequence of assembly events (Taanman and Williams

2001)). So it seems that the nuclear encoded subunits are important for the regulation and stability of the fully assembled complex since the bacterial complex can function with only three catalytic subunits (Galati et al. 2009; Li et al. 2006).

Assembly of COX

The assembly of COX is a complicated process that involves the sequential incorporation of each subunit (Wielburski and Nelson 1983). There is no tight regulation between the synthesis of mitochondrial and nuclear encoded COX subunits (Nijtmans et al. 1995). Previous experiments have shown that inhibition of mitochondrial protein synthesis leads to increased proteolytic degradation of unassembled nuclear-encoded

COX subunits (Nijtmans et al. 1995). The half-life of the holoenzyme is thought to be about three days (Leary et al. 2002).

Three intermediates (S1-S3) have been identified in the assembly of the holoenzyme complex (S4) by blue-native PAGE and SDS-PAGE in the second dimension (Nijtmans et al. 1998). The first (S1) contains only COX I, the second (S2) includes COX I, COXIV and COX Va, and the third (S3) contains all subunits except VIa and either VIIa or VIIb. Addition of these subunits completes the holoenzyme complex

(S4) (Nijtmans et al. 1998; Shoubridge 2001).

Recently, it was shown that mammalian COX IV and COX Va subunits form a dimer before joining S1 and the addition of COX IV and Va subcomplex occurs after the incorporation of the prosthetic groups in COX I but before the addition of COX II

(Stiburek et al. 2005) and this is similar to Cox5p-Cox6p dimer in yeast where the presence of Cox6p is required for Cox5p stability (Glerum and Tzagoloff 1997).There is also an additional intermediate between S2 and S3 that contains COX I-COX II-COX IV-

COX Va (Fornuskova et al. 2010). With an anti-COX III antibody lacking, there is no evidence for the presence of COX III in this intermediate.

The assembly of COX requires a larger number of genes than those encoding the structural subunits of the complex. Of the 30 different genetic complementation groups for the assembly of COX in yeast about 20 have been assigned to accessory proteins

(Barrientos et al. 2002). They are required at different stages of assembly such as translational regulation, membrane insertion and processing, heme a synthesis, copper transport and insertion and assembly process, as shown in Figure 3. At least 12 of these have known human orthologs with known or unknown function. Most of the functions of

the assembly factors are inferred from studies in yeast. Before COX assembly, the core subunits need to be synthesized but unfortunately not much is known about the regulation of translation of mitochondrial encoded proteins

S1 S2 S4

Figure 3. Stages of COX assembly and COX assembly factors in mammals. Cox1 to

Cox8 are structural subunits. CuA, CuB, heme a and heme a3 are prosthetic groups and the rest are assembly factors. S2 and S3 are assembly intermediate and S4 is fully assembled

COX (Stiburek et al. 2006).

The first subunit to assemble is COX I which is inserted into the membrane by the help of Oxa1 protein. Yeast Oxa1 also helps in translocation of both N- and C-termini of yeast Cox2p (Hell et al. 2001). The yeast Oxa1 null mutant is respiratory deficient. The human homolog of Oxa1p is OXA1L which is also localized to mitochondria (Bonnefoy et al. 1994). Yeast Mba1p cooperates with Oxa1p in the cotranslational insertion process 47

(Ott et al. 2006). The lack of Mba1p and C-terminus (which is ribosomal-binding domain) in Oxa1p shows that there is an association of mitochondrial translation products with mtHSP70 which is a chaperone that interacts with unfolded polypeptides (Hartl and

Hayer-Hartl 2002). Yeast Mba1p has a similar sequence to mitochondrial ribosomal L45 protein of higher eukaryotes (Ott et al. 2006). Yeast Cox18p or Oxa2p are showed to be involved in topogenesis of the C-terminal domain of Cox2p, and yeast Cox18p null mutants exhibit isolated COX deficiency (Saracco and Fox 2002; Souza et al. 2000). A human COX18 ortholog has been identified and localized to mitochondria (Sacconi et al.

2005).

The next steps consist of the addition of two hemes to COX I. Based on yeast studies, heme a is found in two chemical forms: heme a, which is low-spin heme and a3, which is a high-spin heme. The formation of heme a consists in the conversion of heme b into heme o by Cox10p (a mitochondrial membrane-associated farnesyltransferase). This process involves the conversion of C2 vinyl group on pyrrole ring a into a hydroxyethylfarnesyl group (Tzagoloff et al. 1993). In the next step, heme o is converted to heme a by the C8 methyl substituent on the pyrrole ring d of heme o which is oxidized into an aldehyde. The oxidation proceeds by two successive monooxygenase steps

(Brown et al. 2002) catalyzed by the Cox15p (heme a synthase) with combination with ferredoxin (Yah1p) and ferredoxin reductase (Arh1p) (Barros et al. 2002). The insertion of heme a into Cox1p has not been characterized. Since heme a moieties are located in the hydrophobic interior of Cox1p, the insertion is likely to occur concomitantly with the

translation of Cox1p rather than after assembly of the core complex (Carr and Winge

2003). The human homologs, COX10 and COX15 also perform the same function.

The function of SURF1 is still unclear but SURF1 is known to be involved in the early assembly steps of COX. Studies in the protobacteria Rhodobacter sphaeroides suggested that Surf1 assists in the insertion of heme a3 into Cox1 and in the maturation of the heme a3- CuB center (Smith et al. 2005). The yeast homolog of SURF1 is Shy1p which was shown to interact with the Cox1p translation machinery, including Mss51p and Cox14p (Fontanesi et al. 2008). Recently, analysis of tissues from patients with

SURF1 deficiency suggested that it could have a role in copper homeostasis (Stiburek et al. 2009).

Another important step in the COX assembly process is the delivery and insertion of the copper ions. Based on yeast studies, the function of human homolog of protein involved in copper deliver and insertion has been elucidated. COX17 is the copper chaperone that delivers copper to COX11 or SCO1 and SCO2 proteins. COX 11 is important for inserting CuB ion into COX I and SCO1 and SCO2 are important for formation of CuA site in COX II. Studies in SCO1 and SCO2 patient cell lines have shown that SCO1 and SCO2 have non-overlapping, cooperative functions in mitochondrial copper delivery and function as homodimers (Leary et al. 2004).

Additional roles for SCO1 and SCO2 have been recently proposed: SCO1 is involved in a posttranslational step in the maturation of COX II based on the result that SCO1 interacts with fully assembled COX in human muscle mitochondria, and that in SCO1 patients, the

mutant SCO1 was associated with subassembly intermediates containing COX II

(Stiburek et al. 2009). SCO2 alone is required for COX II synthesis, and also acts as a thiol-disulphide oxidoreductase to oxidize the copper-binding cysteins of SCO1 (Leary et al. 2009). Yeast Cox 19p was shown to be implicated in copper transfer to COX and Cox

19p null mutant has COX defect but not decrease in copper level. In addition, the Cox

19p mutant phenotype cannot be rescued by addition of exogenous copper salts (Nobrega et al. 2002). A human homolog of Cox 19p also has been identified (Sacconi et al. 2005).

The function of the human homologs of yeast assembly factors COX16 (Tay et al.

2004) and PET191 (C2orf64) (Huigsloot et al. 2011) is still unknown. A summary of the human COX assembly factors is shown in Table 2 (reviewed in (Mick et al. 2011)).

Table 2. Human COX assembly factors in different steps of the COX assembly process.

Protein Function Membrane insertion and processing OXA1L insertion of catalytic core COX18/OXA2 export of C-terminal domain of COX II

Heme synthesis and insertion COX10 Heme o synthesis; formation of heme a site in COX I COX15 Heme a synthesis FDX1 Heme a synthesis

SURF1 Involved in COX I biogenesis; heme a3 insertion

Copper transport and insertion

SCO1 Formation of CuA center (COX II)

SCO2 Formation of CuA center (COX II); COX II synthesis

COX11 Formation of CuB center (COX I) COX17 Copper delivery to COX 11, SCO 1 and SCO2 COX19 Copper trafficking

Unknown functions PET191 (C2orf64) unknown COX16 unknown

The detailed pathway of assembly of the COX complex is still incomplete and the function of some assembly factors of COX is still unknown. If the assembly pathway of

COX is resolved, it would provide vital information in finding the cause of many cytochrome c oxidase deficiency diseases.

COX deficiency

Deficiencies in oxidative phosphorylation have a wide range of clinical phenotypes (Chinnery and Turnbull 1999), including neurological, neuromuscular, cardiac, endocrine disorders, and even some cancers. In a Caucasian population in northern England, the disease frequency is 1:8000 (Chinnery et al. 2006). These disorders may follow Mendelian inheritance or maternal transmission, or occur as sporadic cases.

Mutations can arise in nuclear or mitochondrial DNA. Mutations that occur in mitochondrial DNA are usually, but not always, associated with milder phenotypes in the adult population. In the pediatric population, most disorders are inherited as autosomal recessive traits, and these are usually associated with more severe clinical manifestations and fatal outcome.

mtDNA mutations in the structural subunits of the catalytic core have been identified in a few patients, and all of the known mutations in subunits I, II and III genes are heteroplasmic, and a minimum threshold level of pathogenic mtDNA mutation must be surpassed to cause COX deficiency. Most isolated COX deficiencies are inherited as autosomal recessive disorders and generally have a very early age of onset and a fatal

outcome (reviewed in (Shoubridge 2001)). Observed phenotypes include classical Leigh syndrome, the Saguenay-Lac-St-Jean form of Leigh syndrome, fatal infantile COX deficiency, cardiomyopathy, myopathy and a reversible COX deficiency in muscle

(Robinson 2000). Interestingly, only two mutations in the structural subunits encoded in nuclear DNA has been found COX VIb1 (Massa et al. 2008) and COX IVi2 (Shteyer et al. 2009). Mutations have also been found in some of known human homologues of yeast assembly factors. Although genes are ubiquitously expressed, their mutations have been associated with a particular clinical tissue phenotype as shown in Table 3 (reviewed in

(Shoubridge 2001)).

Although several human COX assembly factors have been identified, the total number and identity of genes involved in COX assembly in mammals remains unknown

(Zhu et al. 1998). Moreover, the tissue-specificity of the observed clinical phenotypes remains an unexplained phenomenon, together with the extent of genetic heterogeneity in this group of patients. It is therefore believed that there exist other genes whose mutation can lead to defects in cytochrome c oxidase function, some of which may be specific to metazoans.

Table 3. Summary of reported gene defects associated with human COX deficiency.

Gene Function Type of Phenotype Reference mutation

MtDNA encoded genes

COX I Catalytic Frame shift Motor neuron (Comi et al. 1998) core disease

Nonsense Multisystem (Bruno et al. 1999)

Nonsense Myopathy, (Karadimas et al. 2000) myoglobinuria

Missense Sideroblastic (Gattermann et al. 1997) anemia

Missense MELAS-like (Tam et al. 2008) syndrome

COX II Catalytic Missense Encephalopathy (Clark et al. 1999) core (initiation codon)

Missense Myopathy (Rahman et al. 1999)

Nonsense Multisystem (Horvath et al. 2005)

COX III Catalytic Microdeletion Myopathy, (Keightley et al. 1996) core myoglobinuria

Frameshift Leigh-like (Tiranti et al. 2000)

Missense MELAS (Manfredi et al. 1995)

Nonsense Encephalopath, (Hanna et al. 1998) myopathy

Missense Exercise (Horvath et al. 2005) intolerance and rhabdomyolysis

Gene Function Type of Phenotype Reference mutation nDNA encoded genes

COX VIb1 Structural Missense Severe infantile (Massa et al. subunit encephalomyopathy 2008)

COX IVi2 Structural Missense Exocrine pancreatic (Shteyer et al. subunit insufficiency, 2009) dyserythropoietic anemia, calvarian hyperostosis.

SURF1 Early Frame shift, Leigh syndrome (Tiranti et al. assembly nonsense, 1998b; Zhu et al. splice site, 1998) rare missense

SCO2 Copper Nonsense, Hypertrophic (Jaksch et al. delivery missense cardiomyopathy 2000; encephalopathy Papadopoulou et al. 1999; Valnot et al. 2000a)

SCO1 Copper Frame shift, Hepatic failure (Valnot et al. delivery missense ketoacidosis 2000a)

COX10 Heme Missense Tubulopathy (Valnot et al. modification encephalopathy 2000b)

COX15 Heme Missense, Hypertrophic (Antonicka et al. modification splice site cardiomyopathy 2003b)

C2orf64 Early Missense Hypertrophic (Huigsloot et al. assembly cardiomyopathy 2011)

Gene Function Type of Phenotype Reference mutation

LRPPRC Post Missense Leigh syndrome French (Mootha et al. transcriptional Canadian 2003) expression of mt-encoded (Sasarman et al. OXPHOS 2010) subunits

FASTKD2 Mitochondrial Nonsense Infantile (Ghezzi et al. apoptosis enceplhalomyopathy 2008)

Thesis outline

Mutations in several factors and structural subunits involved in the biogenesis of

COX have been described. Nonetheless, the genetic defect responsible for many COX deficiencies remains unknown, and it is therefore presumed that there exist unidentified genes coding for assembly factors whose mutations lead to deficiencies in COX activity.

In this thesis, I investigated two different subjects with different clinical phenotypes associated with COX deficiency. The first subject presented with late-onset

Leigh syndrome (Chapter 1) and the second with fatal neonatal lactic acidosis and dysmorphology (Chapter 2). The aim of this thesis was first to identify the defective genes responsible for the COX deficiency (hereafter referred to as the disease genes), and second to characterize the function of the proteins in COX biogenesis. Using this combined information we hoped to shed light on the still mysterious processes of COX assembly and mitochondrial translation.

There are many ways to identify the genetic basis of mitochondrial defects, including direct sequencing of candidate genes or mapping by conventional linkage analysis, if an extended family is available. Lacking such material, I set out to identify the disease genes in these subjects by functional complementation using microcell-mediated chromosome transfer (Killary and Fournier 1995; Zhang et al. 1992; Zhu et al. 1998).

Using chromosome transfer one can introduce normal human chromosomes from a monochromosomal human:mouse hybrid cell line containing single human chromosomes tagged with HyTK (hygromycin B resistance) (Cuthbert et al. 1995) into subject cells. In 57

a recessive disorder, a single normal copy of the chromosome carrying a wild-type version of the disease gene would be predicted to rescue the biochemical phenotype in subject cells, allowing mapping of the defective gene to a specific chromosome (Zhu et al. 1998). Since deletions or rearrangements of the transferred chromosome sometimes occur during the microcell-mediated chromosome transfer procedure, one can find a correlation between the presence of a specific fragment of the complementing chromosome and the rescue of the disease phenotype, thus allowing one to finely map a critical chromosomal interval predicted to contain the disease gene. Sequencing of candidate genes in the region then leads to identification of the genetic mutation.

This thesis is presented in a manuscript format, separated in two chapters. In chapter 1, I mapped the disease gene in the first subject using the microcell-mediated chromosome transfer, and confirmed it by expression of the wild-type cDNA. I then show that the protein functions as a translational activator of COX I. In chapter 2, I identified the disease gene in the second subject using a combination of microcell- mediated chromosome transfer, homozygosity mapping and transcript profiling. This was confirmed by expression of the wild type cDNA. Characterization of the function of the protein suggests that it acts to couple COX I synthesis to assembly of the COX holoenzyme complex.

Chapter 1 is a manuscript published in July 2009 in Nature Genetics, 41(7):833-837, which identifies the genetic defect of the subject with late-onset Leigh syndrome associated with cytochrome c oxidase deficiency by using microcell-mediated chromosome transfer. Since the subject has a specific defect in the synthesis of COX subunit 1 (COX I), we investigated the function of the defective gene to explain the biochemical defect of the subject.

Chapter 1: Mutation in TACO1, a translational activator of COX I, results in

cytochrome c oxidase deficiency and late-onset Leigh Syndrome

Woranontee Weraarpachai1, 2,*, Hana Antonicka2,*, Florin Sasarman1,2, Jürgen Seeger3,

Bertold Schrank 4, Jill E. Kolesar1,2, Hanns Lochmüller5,6, Mario Chevrette7, Brett A.

Kaufman2, Rita Horvath5,8, Eric A. Shoubridge1, 2

1Department of Human Genetics, McGill University, Montreal, Quebec, H3A 2B4, Canada

2Montreal Neurological Institute, McGill University, Montreal, Quebec, H3A 2B4, Canada

3Department of Pediatrics, Deutsche Klinik für Diagnostik GmbH, Wiesbaden, Germany, Aukammallee 33, D-65191 Wiesbaden

4Department of Neurology, Deutsche Klinik für Diagnostik GmbH, Wiesbaden, Germany, Aukammallee 33, D-65191 Wiesbaden

5Friedrich-Baur Institute, Ludwig-Maximilians-University, Munich, Germany

Ziemssenstr. 1a, 80336 Munich, Germany

6Present address: Institute of Human Genetics, Newcastle University, Newcastle, NE1 4HH, UK

7Department of Surgery, Urology Division, McGill University and The Research Institute of the McGill University Health Centre, Montreal, Quebec, H3G 1A4, Canada

8Present address: Mitochondrial Research Group, Newcastle University, Newcastle, NE1 4HH, UK

* Co-first authors

Corresponding author: Eric A. Shoubridge, email: [email protected]; tel: 514- 398-1997; FAX: 514-398-1509

Abstract

Defects in mitochondrial translation are among the most common causes of mitochondrial disease (Taylor and Turnbull 2005), but the mechanisms that regulate mitochondrial translation remain largely unknown. In the yeast S. cerevisiae, all mitochondrial mRNAs require specific translational activators, which recognize sequences in 5’-UTRs, and mediate translation (Naithani et al. 2003). As mammalian mitochondrial mRNAs do not have significant 5’-UTRs (Montoya et al. 1981), alternate mechanisms must exist to promote translation. We identified a specific defect in the synthesis of the mtDNA-encoded COX I subunit in a pedigree segregating late-onset

Leigh Syndrome and cytochrome c oxidase (COX) deficiency. We mapped the defect to chromosome 17q by functional complementation and identified a homozygous single base pair insertion in CCDC44, a member of a large family of hypothetical proteins containing a conserved DUF28 domain. CCDC44, renamed TACO1 for translational activator of COX I, shares a remarkable degree of structural similarity with bacterial homologs (Shin et al. 2002), and our findings suggest that it is one of a family of specific mammalian mitochondrial translational activators.

Results and Discussion

The index subject (pedigree, Supplementary Figure 1), presented with childhood- onset and slowly progressive Leigh Syndrome due to an isolated COX deficiency. Blue-

Native polyacrylamide gel electrophoresis (BN-PAGE) analysis of subject fibroblasts showed greatly reduced steady-state levels of fully assembled COX (Figure 1a), but normal levels of the other respiratory chain complexes. No previously described COX subcomplexes were present in the fibroblasts from the subject, however we detected a very small amount of a ~135kDa subcomplex (S3*) (Figure 1b), containing at least subunits II, IV and VIc (Supplementary Figure 2).

To determine whether the rates of synthesis and turnover of the individual mtDNA-encoded COX subunits were altered, we pulse-labeled, then chased the mitochondrial translation products in subject fibroblasts. Remarkably, we found a specific defect in the rate of synthesis of the COX I subunit (Figure 1c), which was reduced by approximately 65%. The remaining radiolabeled material appeared as a smear of apparently truncated COX I polypeptides, with one major premature translation product (Figure 1d). The rates of synthesis of the other mitochondrially encoded polypeptides were similar to control in most cases (data not shown), though there was a consistent ~50-80% increase in some of the ND subunits, and in cyt b. Although the synthesis of COX III was decreased in the experiment shown (Figure 1c), this was not a consistent result. The compromised synthesis of COX I, and the consequent reduction in assembled COX, resulted in an increased degradation of all newly synthesized COX

subunits. COX II and III were practically undetectable at the end of the chase period (less than 15% of control), while 37% of control COX I could be detected, representing only the full-length polypeptide (Figure 1c).

A nuclear origin of the defect was confirmed by fusion of subject’s fibroblasts with human (143B) rho0 cells (data not shown), and we found no evidence for mtDNA depletion or large-scale mtDNA deletions (data not shown). Sequence analysis of the

COX I gene identified no mutations, and RNA blot analysis of mitochondrial transcripts showed normal transcript levels for all mtDNA-encoded COX subunits (Supplementary

Figure 3), as well as for two subunits of complex I and both ribosomal subunits, demonstrating that the COX I synthesis defect was not due to a decreased amount of

COX I mRNA, but rather to a mutation in a factor specifically responsible for COX I translation. Overexpression of 14 different known COX assembly factors (see Materials and Methods) did not rescue the defect (data not shown).

Genome-wide linkage analysis using microsatellite markers showed significant linkage to chromosome 17 with a maximum LOD score of 3.3 at D17S2193 (data not shown), but did not lead to identification of the gene defect. Using microcell-mediated chromosome transfer we found that transfer of a copy of chromosome 17q into subject fibroblasts rescued the biochemical defect in 40 out of 46 independent clones (Figure 2).

Copy number analysis of two rescuing and two nonrescuing clones using an SNP-Chip array identified two regions on chromosome 17q (49.1-53.1 Mb and 58.5-64.2 Mb) present exclusively in the rescuing clones (Supplementary Figure 4). These two regions

encompass 103 genes, five of which were predicted to code for mitochondrial proteins

(Supplementary Table 1) by independent mitochondrial targeting prediction programs

(Predotar, Target P, MitoPred, MitoProt and MitoPWolf).

Sequence analysis of the cDNA for CCDC44 (coil-coiled domain containing protein 44) from the subject fibroblasts detected a homozygous one-base-pair insertion at position 472 (472insC), resulting in a frameshift and the creation of a premature stop codon (Figure 3a, 3b). We confirmed the presence of this mutation in the genomic DNA from subject fibroblasts by restriction enzyme (Figure 3c). Steady-state levels of

CCDC44 mRNA were greatly reduced in subject fibroblasts, as predicted by the presence of a premature stop codon and consequent nonsense-mediated RNA decay (Figure 3d).

The mutation was absent in 100 unrelated controls, and segregated as expected in the pedigree; all affected individuals carried the homozygous mutation, whereas unaffected individuals were either heterozygous (both parents) or homozygous for the wild-type sequence (Supplementary Figure 1). We named this protein TACO1 for translational activator of mitochondrially encoded COX I.

Expression of wild-type TACO1 in subject fibroblasts rescued the COX assembly defect (Figure 4a), the COX I synthesis defect (Figure 4b), and COX activity (104% of control). TACO1 was undetectable by immunoblot analysis of mitochondria isolated from subject fibroblasts using polyclonal antibodies directed against both N- and C- terminal peptide sequences (Figure 4a). The cDNA for TACO1 encodes a protein of 297 amino acids, with a predicted mitochondrial presequence of 26 amino acids, which would

produce a 29.8-kDa mature protein. A C-terminal HA-tagged version of TACO1 expressed in HEK293 cells localized to mitochondria by immunocytochemistry (Figure

5a), and alkaline carbonate extraction of isolated mitochondria showed that it behaves as a mitochondrial matrix protein (Figure 5b).

To test whether TACO1 associates with other mitochondrial proteins, we separated native complexes present in mitochondrial from control fibroblasts cell by size exclusion. Immunoblot analysis of the individual fractions identified the majority of

TACO1 protein in a higher molecular weight complex of about 74 kDa (Figure 5C), indicating that TACO1 either functions as a multimer or is in a complex with another protein(s). A very similar elution pattern was found using an antibody against mitochondrial translation elongation factor Ts (EFTs), but not mitochondrial translation elongation factor Tu (EFTu), suggesting a potential interaction between mitochondrial translation factor(s) and TACO1. Overexpression of a C-terminal HA-tagged version of the protein in control and subject fibroblasts, which resulted in substantially higher steady-state levels of TACO1-HA compared to the endogenous protein, decreased translation of all mitochondrial polypeptides (Figure 6), and the assembly of all respiratory chain complexes (Supplementary Figure 5), suggesting that TACO1 interacts with a common element of the mitochondrial translation apparatus that can be titrated out. This effect was not simply due to the presence of the HA tag, as the tagged construct was as effective as the wild-type cDNA at rescuing the translation defect when expressed at appropriate levels. Although we were able to immunoprecipate TACO1-HA using an anti-HA antibody, it did not coimmunoprecipitate with EFTs (or EFTu, EFG1) (data not 65

shown), which indicates that if TACO1 does interact with translation elongation factors, the proteins do not form a stable complex.

The S. cerevisiae TACO1 ortholog, YGR021w, is 29% identical and 43% similar to the human protein; however, translation experiments on the ygr021wΔ deletion strain showed no translation defects (Supplementary Figure 6a). Furthermore, the COX activity of ygr021wΔ mutant cells grown on either glycerol or galactose was 70-100% of control, whereas that of the shy1Δ mutant, in which a known COX assembly factor is deleted, was less than 5% of controls when grown on galactose. The ygr021wΔ strain also grows on nonfermentable carbon sources, even acetate, with a similar doubling time as control strains (data not shown), and is respiratory-competent as indicated by TTC overlay

(Supplementary Figure 6B). These data indicate that the protein encoded by YGR021w is not essential for the synthesis of full-length Cox1p or respiratory competency in yeast.

Analysis of patterns of coexpression for YGR021w using the SPELL algorithm (Hibbs et al. 2007), however, show a strong enrichment for factors involved in mitochondrial translation in yeast, suggesting that YGR021w may play some nonessential role in mitochondrial translation.

To our knowledge, TACO1 is the first specific mitochondrial translational activator identified in mammals. As mammalian mitochondrial mRNAs lack significant

5’-UTRs (Montoya et al. 1981), most yeast genes involved in the translation of the mitochondrially encoded proteins lack mammalian homologs (Fontanesi et al. 2006).To date, one distant human homolog (LRPPRC) of the yeast translational activator Pet309p

has been reported (Mootha et al. 2003). Both proteins contain PPR motifs, consisting of degenerate 35 amino acid sequences that form antiparallel alpha helices, and which are thought to be involved in post-transcriptional mRNA metabolism, especially in mitochondria and chloroplasts (Delannoy et al. 2007; Lurin et al. 2004). Mutations in

LRPPRC lead to isolated COX deficiency in French-Canadian individuals with Leigh syndrome (Mootha et al. 2003). The protein is reported to be involved in the stabilization of the mRNAs for both COX I and COX III in mammals, without directly affecting their translation; however, the molecular mechanism remains unknown (Xu et al. 2004).

TACO1 is clearly necessary for the efficient translation of COX I, as only a small amount of full-length polypeptide is synthesized in a null background in which the steady-state levels of COX I mRNA are normal. We envision three possible mechanisms.

TACO1 might act by either securing an accurate start of COX I translation, by stabilizing the elongating polypeptide and ensuring completion of its translation, or alternatively, it could interact with the peptide release factor, ensuring that the polypeptide does not dissociate from the ribosome until synthesis is complete.

TACO1 is conserved through bacteria, but most of the homologous sequences, all of which contain the so-called DUF28 domain are annotated as hypothetical proteins.

Recently, a DUF28 family member PmpR (from Pseudomonas aeruginosa) has been shown to be involved in a negative regulation of the quorum-sensing response regulator gene by binding to an upstream promoter element (Liang et al. 1996). Another DUF28 family member, Aq1575, a hypothetical protein from the hyperthermophilic bacterium

Aquifex aeolicus (32% identical, 52% similar to TACO1), has been crystallized (Shin et al. 2002). The protein has a large cleft surrounded by three domains, one of which represents a novel protein fold, but no obvious active site or functional domain could be identified in the crystal structure. The predicted three-dimentsional structure of TACO1 is remarkably similar to that of the A. aeolicus homolog considering the evolutionary distance between them (Supplementary Figure 7). It is tempting to suggest that this ancient protein evolved as a specific translational activator in concert with the loss of the mitochondrial mRNA regulatory sequences that occurred with the extreme reduction in the size of the metazoan mitochondrial genome.

Materials and Methods

Subjects

The index subject is the fifth child of healthy consanguineous Turkish parents

(Supplementary Figure 1, V:5), three of the 8 siblings are also affected (V:3; V:7;V:8), five are healthy. One of the subject’s five cousins (V:10; both parents are siblings of the parents of the index subject) is also affected. The index subject developed normally till 5 years of age. From this time his gait became unstable and his active speech worsened; however his comprehension was relatively well preserved. The first detailed neurological examination at 10 years of age, indicated that the subject was a small, thin, and dystrophic child (weight and height < 3rd percentile). Examination of cranial nerves revealed bilateral optic atrophy with otherwise normal eye movements, no ptosis or nystagmus, decreased mimic, bilateral facial weakness, prominent dysarthria and mild dysphagia. He had a spastic tetraparesis with increased deep tendon reflexes,and Achilles clonus. He was able to sit without help, but was unable to stand or walk. Sensory symptoms and ataxia were not noted. Cognitive functions showed mild mental retardation.

Laboratory analysis showed increased serum lactate (4.1 mmol/l, normal < 2 mmol/l), CSF lactate (3.2 mmol/l, normal <2 mmol/l) and pyruvate (1.6 mg/dl, normal <

0.7 mg/dl). Other laboratory tests including CSF protein were normal.

Brain magnetic resonance imaging (MRI) showed bilateral, symmetric hyperintense lesions of the basal ganglia, typical for Leigh syndrome on T2/FLAIR 69

sequences. Multiple hyperintense lesions in both frontal subcortical areas and in the subcortical region of the hemispheres were also present. Brainstem and cerebellum were normal.

A muscle biopsy at 10 years of age showed a few hypotrophic fibers and a generalized COX deficiency in all muscle fibers. No ragged red fibers or succinate dehydrogenase hyper-reactive fibers were noted. Biochemical analysis of the respiratory chain enzyme complexes I-IV and pyruvate dehydrogenase in skeletal muscle showed a severe isolated COX deficiency with approximately 15% residual activity of the enzyme, the activities of the other enzymes were within normal range.

The subject’s siblings V:3, V:7, V:8 and cousin V:10 also presented with similar clinical symptoms, although the age of onset varied (16, 15, 4 and 14 years, respectively).

They are all of thin, small stature, with variable neurological presentation including bilateral optic atrophy, spastic tetraparesis, dystoniform movements, blurred speech and mild cognitive deficit. Interestingly, the affected girls showed a milder phenotype with preserved ambulation into the twenties, suggesting that sex-specific factors may influence the phenotype. In all four cases brain MRI showed bilateral, symmetric hyperintense lesions of the basal ganglia, typical for Leigh syndrome, with variable severity.

Human Studies

We obtained informed consent from all investigated family members, and the research studies were approved by the Institutional review board of the Montreal

Neurological Institute and the Ludwig-Maximilians-University, Munich, Germany.

Cell lines

Primary cell lines were established from subject skin fibroblasts. The subject and control cell lines were immortalized by transduction with a retroviral vector expressing the HPV-16 E7 gene plus a retroviral vector expressing the catalytic component of human telomerase (htert) (Yao and Shoubridge 1999). The fibroblasts and HEK 293 line were grown at 37°C in an atmosphere of 5% CO2 in high glucose Dulbecco's modified

Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS).

Enzyme activity measurements

Spectrophotometric assays of whole cell fibroblast extracts, or crude mitochondrial pellets from yeast, were used to measure enzyme activities. COX activity was normalized to citrate synthase activity and specific activity was determined by protein content as described (Antonicka et al. 2003a; Capaldi et al. 1995). Protein concentration was measured by the Bradford assay.

Electrophoresis and immunobloting

Blue-Native PAGE was used to separate samples in the first dimension on 6%-

15% or 8%-12% polyacrylamide gradient gels, as previously described (Klement et al.

1995). Mitoplasts were prepared from fibroblasts by treatment with 0.8 mg of digitonin/mg of protein, solubilized with 1% lauryl maltoside, and 20 μg of the solubilized proteins were used for electrophoresis. For the two-dimensional analysis, the

BN-PAGE/SDS-PAGE was done as previously described (Antonicka et al. 2003a).

Individual structural subunits of complexes I, II, III, IV and V were detected by immunoblot analysis using commercially available monoclonal antibodies (Molecular

Probes), except for complex I, where a polyclonal antibody against subunit ND1 (a gift of

Anne Lombès, Paris) was used.

SDS-PAGE was used to separate denatured whole cell extracts or isolated mitochondria using 12% polyacrylamide gels followed by immunoblot analysis with indicated antibodies. The antibody against EFTu/Ts was a gift of Linda Spremulli

(University of North Carolina, Chapel Hill).

Pulse labeling of mitochondrial translation products

In vitro labeling of mitochondrial translation products was performed as previously described (Boulet et al. 1992). Briefly, cells were pulse-labeled for 60 min at

37o C in methionine/cysteine-free DMEM containing 200 µCi/ml of a

[35S]methionine/cysteine mix (Perkin Elmer) and 100 µg/ml of either emetine or anisomycin. The cells were chased for 10 min (PULSE) or 17.5 hours (CHASE) in regular DMEM. For chase studies, cells were incubated for 23 hours in 40 µg/ml chloramphenicol prior to labeling. Total cellular protein (50 µg) was resuspended in loading buffer containing 93 mM Tris-HCl, pH 6.7, 7.5% glycerol, 1% SDS, 0.25 mg bromophenol blue/ml and 3% mercaptoethanol, sonicated for 3–8 s, loaded and run on

15-20% polyacrylamide gradient gels. The labeled mitochondrial translation products were detected through direct autoradiography.

RNA blot analysis

We isolated RNA from subject and control fibroblasts using the RNeasy Kit

(Qiagen). Ten micrograms of total RNA were separated on a denaturing

MOPS/formaldehyde agarose gel and transferred to a nylon membrane. We labeled 300 to 500-bp-long PCR products of individual mitochondrial genes with [α-32P]-dCTP (GE

Healthcare) using the MegaPrime DNA labeling kit (GE Healthcare). Hybridization was performed according to the manufacturer’s manual using ExpressHyb Hybridization

Solution (Clontech) and the radioactive signal was detected using the Phosphoimager system.

Microcell-mediated chromosome transfer

Immortalized subject skin fibroblasts were fused with microcells carrying the q- arm of human chromosome 17 tagged with the hygromycin resistance gene, isolated from the B78MC57 mouse cell line (Cuthbert et al. 1995; Gagnon et al. 2006) by microcell- mediated chromosome transfer (Zhu et al. 1998).

Chromosome Copy number analysis

The chromosome copy number was determined in rescuing and nonrescuing clones using SNP Mapping GeneChip Nsp 250k Array (Affymetrix). This service was performed by The Centre for Applied Genomics, Hospital for Sick Children, Toronto.

The data were analyzed using the Copy Number Analysis Tool of GCOS Client software

(Affymetrix).

Mutation detection

We isolated total RNA from subject and control skin fibroblasts using RNeasy Kit

(Qiagen). TACO1 cDNA was amplified by using OneStep RT-PCR kit (Qiagen) and the gel purified PCR fragments were used for direct sequencing. Total genomic DNA from controls, subject fibroblasts and blood from family members was isolated using DNeasy

Kit (Qiagen). Primers specific for exon 3 of the TACO1 gene were used to amplify the

DNA, followed by either digestion with MwoI or direct sequencing. 74

cDNA constructs and virus production and infection

Retroviral vectors containing the cDNA sequence of COX assembly factors

(COX11, COX16, COX17, OXA1, SCO2, PET191, SURF1, OXA2, COX10, COX15.1,

COX19, COX23 and COX15.2) were created with the GatewayTM Cloning system

(Invitrogen) as previously described (Antonicka et al. 2003a). cDNAs from the individual genes were amplified by OneStep RT-PCRTM (Qiagen) using specific primers modified for cloning into Gateway vectors, except for TACO1, where the cDNA (Open

Biosystems, clone ID 4099917) in the Gateway modified vector pOTB7 was used. The

PCR constructs were cloned into Gateway-modified retroviral expression vectors, pLXSH or pBabe. For the C-terminal HA-tagged TACO1 (TACO1-HA), TACO1 cDNA was amplified using specific primers, cloned into pCR2.1®-TOPO® vector (Invitrogen), digested with EcoRI and cloned directly into EcoRI site of pLXSH or pBabe. The fidelity of cDNA clones was confirmed by DNA sequencing. Retroviral constructs were transiently transfected into the Phoenix packaging cell line using the HBS/Ca3(PO4)2 method (http://www.stanford.edu/group/nolan/protocols/pro_helper_dep.html). Subject and control fibroblasts were infected 48 hours later by exposure to virus-containing medium in the presence of 4 μg/ml of polybrene as described (Lochmuller et al. 1999).

TACO1 antibody production

A polyclonal antibody against two peptides (Ac-IKGPKDVERSRIFSKLC-amide and Ac-LEFIPNSKVQLAEPDLEQAAC-amide) from the human TACO1 protein was 75

prepared by 21st Century Biochemicals (Marlboro, MA). Crude serum and affinity purified antibodies were tested on cell lines overexpressing TACO1 protein and detected a band of approximately 28 kDa. The C-terminal affinity-purified antibody was used for further experiments.

Yeast strains

Yeast strains BY4741ρ +, Y1236ρ+, ygr021wΔ and shy1Δ were obtained from the yeast genome deletion collection (http://www- sequence.stanford.edu/group/yeast_deletion_project/deletions3.html). All strains were grown in standard media at 30C, with vigorous shaking for liquid cultures. For time course experiments, cells were grown in either YPGal or YPGly (1% yeast extract, 2% peptone, and either 2% galactose or 3% glycerol, respectively) and their growth monitored over the course of 46 hours by OD600 readings. For respiratory competency experiments, yeast strains were grown on YEPD plates (1% yeast extract, 2% tryptone/peptone, 2% glucose) into colonies and covered with a 2,3,5- triphenyltetrazolium (TTC) (Sigma) overlay (Barclay et al. 2001).

Yeast translation assay

Mitochondrial protein synthesis was assessed as in (Barrientos 2002), with certain modifications. Cells were grown to a maximum OD600 of 1.0 in 3ml of methionine drop 76

out media containing either 2% galactose or 3% glycerol. Cells were labeled with 35S methionine/cysteine mix for 20 min. After termination of the labeling reaction (including the addition of the NaOH/β-mercaptoethanol/PMSF solution), mitochondrial proteins were precipitated by the addition of 100% TCA to a final concentration of 15% and incubated on ice for 10 min. Samples were centrifuged and the pellet washed three times with cold acetone, followed by denaturation in 50µl Laemmli SDS-PAGE loading buffer.

Samples were separated on a 17.5% polyacrylamide gel.

Mitochondrial isolation and localization experiments

Fibroblasts were resuspended in ice-cold 250 mM sucrose/10 mM Tris-HCl/1 mM

EDTA (pH 7.4) and homogenized with 10 passes through a pre-chilled, zero clearance homogenizer (Kimble/Kontes, Vineland, NJ). Samples were centrifuged twice for 10 min at 600g to obtain a post-nuclear supernatant. Mitochondria were pelleted by centrifugation for 10 min at 10000g, and washed once in the same buffer. Mitochondria were either sonicated or further extracted with 100 mM alkaline carbonate as previously described

(Yao and Shoubridge 1999), and the relevant fractions were analyzed by SDS–PAGE.

Immunocytochemistry

HEK293 cells were plated on coverslips and transfected with TACO1-HA cDNA in pCR2.1®-TOPO® vector using Lipofectamine (Invitrogen). Twenty-four hours post-

transfection the cells were paraformaldehyde-fixed, solubilized by Triton X-100 and stained using anti-HA (Sigma) and anti-cytochrome c (BD Pharmigen) antibodies.

Secondary antibodies ALEXA488 (anti-mouse) and ALEXA 594 (anti-rabbit) secondary antibodies (Molecular Probes) were used for immunofluorescence detection.

Size exclusion chromatography

We separated soluble proteins from mitochondrial detergent extracts were separated on a Tricorn Superdex 200 10/30 HR column (GE Healthcare) as described

(Kaufman et al. 2007) and elution profiles determined by immunoblot analysis using antibodies against TACO1, EFTs and EFTu.

Molecular Modeling

The three-dimensional structure of TACO1 was modeled on the crystal structure of Aq1575 (Shin et al. 2002) homologue using the I-TASSER server (Wu et al. 2007;

Zhang 2007, 2008) and viewed with Swiss-MODEL using the Swiss-Pdb Viewer (Guex and Peitsch 1997) http://www.expasy.org/spdbv/.

Acknowledgements

We acknowledge the important contribution of the individuals who cared for

these subjects and the excellent technical assistance of Ira Kaus, Solvig Mueller-

Ziermann and Anja Zimmermann. We thank Timothy Johns for help with

immunocytochemistry and the cell culture. This work was supported in part by a grant

from the CIHR to EAS. EAS is an International Scholar of the HHMI. RH is supported

by the Deutsche Forschungsgemeinschaft HO 2505/2-1.

Web Resources

Phoenix-Helper dependent protocol,

http://www.stanford.edu/group/nolan/protocols/pro_helper_dep.html;

Saccharomyces Genome Deletion Project, http://www-

sequence.stanford.edu/group/yeast_deletion_project/deletions3.html;

Swiss PDB viewer, http://www.expasy.org/spdbv/.

Figure Legends

Figure 1. Compromised assembly of COX and impaired synthesis of COX subunit I in subject fibroblasts. (a) BN-PAGE analysis of fibroblasts from the subject and two controls. Antibodies against individual subunits of OXPHOS complexes I-V were used for immunoblotting as described in Materials and Methods. (b) BN-PAGE immunoblot analysis with an antibody against COX subunit IV; S3*, a novel COX subcomplex in subject fibroblasts. (c) Pulse-chase labeling of newly synthesized mitochondrial polypeptides (indicated on the left; ND, subunits of complex I; COX, subunits of complex IV; cyt b, subunit of complex III; ATP, subunits of complex V) in subject and control fibroblasts. (d) Magnification of the boxed region from (c) and a line graph of the intensities of individual bands. The arrows indicate a shorter version of the COX I polypeptide in the subject.

Figure 2. The biochemical defect in subject fibroblasts is rescued by microcell- mediated transfer of chromosome 17q. (a) BN-PAGE and (b) pulse labeling of mitochondrial polypeptides in control and subject fibroblasts and in one rescued and one nonrescued clone, obtained after the transfer of the q-arm of chromosome 17 into subject fibroblasts. Complex I-V, OXPHOS complexes; ND, subunits of complex I; COX, subunits of complex IV; cyt b, subunit of complex III; ATP, subunits of complex V.

Figure 3. Mutational analysis of TACO1 in the index subject. (a) Schematic diagram of TACO1 gene and protein. Position of the mutation in the subject DNA and the position of resulting premature STOP codon are indicated. Dark boxes in TACO1 gene denote the coding regions; grey boxes denote 5’ and 3’UTRs. White box in TACO1 protein indicates the predicted mitochondrial leader sequence. (b) Sequencing analysis of exon 3 of TACO1 indicating the position of the homozygous 472insC insertion in theaffected individual. (c) RFLP analysis of exon 3 of TACO1 amplified from genomic DNA of the subject and four controls and digested with MwoI. (d) RT-PCR of the full length TACO1 cDNA from two controls and subject fibroblasts.

Figure 4. Overexpression of TACO1 rescues the mitochondrial translation defect and the COX assembly defect in subject fibroblasts. (a) BN-PAGE analysis of the assembly of individual OXPHOS complexes in control and subject fibroblasts overexpressing TACO1 protein (upper panel). Immunoblot analysis of COX I and

TACO1 steady-state levels. Porin and the complex II 70-kDA subunit were used as loading controls (lower panel). (b) The mitochondrial translation products were pulse- labeled, then chased, in control and subject fibroblasts alone, and in cells overexpressing

TACO1. ND, subunits of complex I; COX, subunits of complex IV; cyt b, subunit of complex III; ATP, subunits of complex V.

Figure 5. TACO1 is a mitochondrial matrix protein. (a) Immunocytochemistry of

HEK293 cells transfected with TACO1-HA. Anti-HA and anti-cytochrome c antibodies were used as indicated. (b) SDS-PAGE immunoblot analysis of sonicated mitochondrial

(left panel) and alkaline carbonate extracts of mitochondria (right panel) from control fibroblasts using antibodies against TACO1, the 70kDa subunit of complex II and COX subunit 1 (COX I). (c) Mitochondria isolated from control fibroblasts were separated on a size exclusion column and individual fractions were run on SDS-PAGE and imunoblotted with antibodies against TACO1 and EFTu/Ts. The molecular weight of individual fractions was determined by calibration of the size exclusion column.

Figure 6. High levels of TACO1 or TACO1-HA expression depress mitochondrial translation in control fibroblasts. Pulse labeling of mitochondrial polypeptides (upper panel). ND, subunits of complex I; COX, subunits of complex IV; cyt b, subunit of complex III; ATP, subunits of complex V. The overexpression of TACO1 protein was confirmed by immunoblotting. (bottom panel). An asterisk (*) denotes the TACO1-HA protein. The 70 kDa subunit of complex II was used as a loading control.

Supplementary Table 1. Predicted mitochondrial proteins present in the rescuing regions on chromosome 17.

Supplementary Figure 1. Pedigree of the index subject (arrow) indicating affected and nonaffected individuals. The genotype for those individuals that could be studied is indicated in parenthesis as follows: (Hom WT) – homozygous wild-type; (Hom mut) – homozygous for 472insC mutation; (Het) – heterozygous for 472insC mutation.

Supplementary Figure 2. BN/SDS-PAGE 2D-gel analysis of subject and control fibroblasts with antibodies against COX subunits I, II, IV and VIc. COX, fully assembled complex IV. Previously identified COX subcomplexes S1, S2 and S3 are indicated. S3* is a novel COX subcomplex present in subject fibroblasts.

Supplementary Figure 3. RNA blot analysis of all three mitochondrial COX transcripts, two transcripts of complex I (ND1, ND2), and the two mitochondrial ribosomal RNAs (12S, 16S) in subject and two control fibroblast lines. Actin was used as a loading control.

Supplementary Figure 4. Chromosome copy number analysis of two rescued and two nonrescued clones. Three copies of a part of chromosome 17 were detected in these clones (as depicted by colored horizontal lines). The two regions present only in the rescued clones are indicated by orange boxes and the position of these regions on chromosome 17 is indicated on the top.

Supplementary Figure 5. Dominant-negative inhibition of mitochondrial translation and OXPHOS assembly in cells overexpressing high levels of TACO1. (A) BN-PAGE and SDS-PAGE analyses of control and subject cells overexpressing TACO1-HA.

Antibodies against subunits of individual complexes and anti-HA antibody were used.

(B) Correlation between the levels of TACO1-HA protein and the COX activity in control fibroblasts over-expressing TACO1-HA. Individual clones overexpressing

TACO1-HA were analyzed for COX activity (normalized to citrate synthase activity) and for TACO1-HA levels using an anti-TACO1 antibody (normalized to the levels of complex II – 70 kDa subunit). TACO1-HA levels were calculated relative to those present in clones transduced with the TACO1 cDNA, but with control levels of COX activity. The data were fitted with a sigmoidal curve using SigmaPlot.

Supplementary Figure 6. Deletion of the yeast ortholog of TACO1 (ygr021wΔ) does not produce a defect in the synthesis of mitochondrial COX polypeptides. (A) Pulse- labeled mitochondrial polypeptides in individual yeast strains were separated on SDS-

PAGE gel as described in Materials and Methods. Individual mitochondrial polypeptides are indicated on the left. Two different wild-type (WT) strains and the deletion mutant of

SHY1 (shy1Δ) were used as controls. (B) Wild-type (WT-Y1236ρ+) and ygr021wΔ strains were stained by TTC overlay. Red colonies indicate respiratory competent yeast.

Supplementary Figure 7. Molecular modeling of TACO1. Predicted 3D structure of

TACO1 (colour) using the crystal structure of the hypothetical protein Aq1575 from the thermophilic bacterium A. aeolicus (grey) as a template. The structure was modeled using the I-TASSER server (see Materials and Methods). The I-TASSER prediction gave a C- score of –1.99, a TM score of 0.48+ 0.15 and an RMSD of 10.8+ 4.6 Angstroms.

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Supplementary Table 1

Supplementary Figure 1

Supplementary Figure 2

Supplementary Figure 3

Supplementary Figure 4

Supplementary Figure 5

Supplementary Figure 7

Chapter 2 is a manuscript, under review for publication in the American Journal of

Human Genetics, in which we identify the gene responsible for a disease associated with

COX deficiency causing fatal neonatal lactic acidosis. A combination of microcell mediated chromosome transfer, homozygosity mapping and transcript profiling were used to find the disease gene. We also propose a function for the normal protein.

100

Chapter 2: Mutations in C12orf62, a factor that couples COX I synthesis to

cytochrome c oxidase assembly, cause fatal neonatal lactic acidosis

Woranontee Weraarpachai1, Florin Sasarman1, Tamiko Nishimura1, Hana Antonicka1,

Karine Auré2,3, Agnès Rötig4, Anne Lombès3,5, Eric A. Shoubridge1

1Department of Human Genetics and Montreal Neurological Institute, McGill University,

2 APHP, Hôpital Ambroise Paré, Boulogne-Billancourt, F-92100 France; UFR médicale

Paris-Ile-de-France-Ouest, F-78280 France,

3Institut Cochin ; Inserm UMRS 1016 ; CNRS UMR 8104 ; Université Paris Descartes,

Paris, F-75014 France;

4Université Paris Descartes and INSERM U781, 149 rue de Sèvres, 75015 Paris, France,

5APHP, Biochimie Métabolique, Unité de cardiogénétique et myogénétique moléculaire et cellulaire, GH Pitié-Salpêtrière, Paris, F-75651 France

Corresponding author: Eric A. Shoubridge

Montreal Neurological Institute, 3801 University Street

Montreal, Quebec, Canada H3A 2B4 [email protected]

Tel: 514-398-1997

FAX: 514-398-1509 101

Abstract

Defects in the assembly of cytochrome c oxidase (COX) are a common cause of a phenotypically diverse and genetically heterogeneous group of mitochondrial disorders.

We studied a family in which the index subject presented with severe congenital lactic acidosis and dysmorphic features associated with a COX assembly defect and a specific decrease in the synthesis of COX I, the subunit that nucleates the assembly of the holoenzyme. Using a combination of microcell-mediated chromosome transfer, homozygosity mapping, and transcript profiling we mapped the gene defect to chromosome 12, and identified a homozygous missense mutation in C12orf62, which segregated with the disease in the family. C12orf62 was not detectable by immunoblot analysis in fibroblasts from the subject, and expression of the wild-type C12orf62 cDNA from a retroviral vector completely rescued the biochemical phenotype. Further, siRNA- mediated knockdown of C12orf 62 recapitulated the biochemical defect in control cells, and exacerbated the COX deficiency in subject cells. C12orf62 is apparently restricted to the vertebrate lineage. It codes for a very small (6 kDa), uncharacterized, single transmembrane protein that localizes to mitochondria, and elutes in a high molecular weight complex (110 kDa) by gel filtration. COX I, II and IV subunits co- immunoprecipated with an epitope-tagged version of C12orf62, and 2D BN-PAGE analysis of newly synthesized mitochondrial COX subunits in subject fibroblasts showed that COX assembly was impaired, and the nascent enzyme complex unstable. We conclude that C12orf62 is required for coordinating the early steps of COX assembly with the synthesis of COX I. 102

Introduction

Complex IV or cytochrome c oxidase (COX) is the terminal enzyme of the respiratory chain that catalyses the oxidation of cytochrome c by molecular oxygen. It is composed of 13 subunits, three of which are encoded in mtDNA and form the catalytic core of the enzyme while the other 10 subunits are encoded in nuclear DNA and imported into mitochondria.

COX deficiencies (MIM 220110) present with a wide range of clinical phenotypes, usually with an early age of onset, and frequent fatal outcome. These include classical Leigh syndrome, the French Canadian form of Leigh syndrome, fatal infantile

COX deficiency, cardiomyopathy, myopathy, fatal infantile lactic acidosis and a reversible COX deficiency in muscles (Robinson 2000; Shoubridge 2001). Isolated COX deficiencies are generally inherited as autosomal recessive disorders, but the genetic basis for many COX deficiencies remains unknown. Mutations in several nuclear-encoded factors involved in the biogenesis of COX have been described (Papadopoulou et al.

1999; Valnot et al. 2000a; Zhu et al. 1998) and although mutations in the mtDNA- encoded structural subunits of the catalytic core have been identified in several patients

(Shoubridge 2001) only two mutations in nuclear-encoded structural subunits have been reported (Massa et al. 2008; Shteyer et al. 2009). Independent of the genetic cause, nearly all COX deficiencies result from a failure to assemble an adequate amount of functional holoenzyme.

103

The assembly of COX is a complex process involving the sequential incorporation of each subunit, and has been most extensively studied in the yeast S. cerevisae. While the broad outlines of assembly of the complex in humans and yeast appear similar (Barrientos et al. 2002; Mick et al. 2011; Nijtmans et al. 1998), many of the yeast proteins, especially those involved in the early assembly steps, have no obvious human orthologs. The exact function of many of these proteins is still unknown and the detailed pathway of assembly of the COX complex remains incomplete. Moreover, the tissue-specificity of the observed clinical phenotypes in humans remains an unexplained phenomenon. It therefore seems likely that as yet uncharacterized genes exist that code for essential COX assembly factors, and that some of these may be specific to metazoans.

Here we have investigated the genetic basis of a mitochondrial disorder in a subject with a COX assembly defect who presented with an unusually severe phenotype that included lactic acidosis, brain damage, and dysmorphy, leading to death within 24 hours of birth.

104

Results

Characterization of the COX defect in fibroblasts

COX activity in immortalized subject fibroblasts was reduced to 30-40% of control levels by spectrophotometric assays of whole cell extracts. This was associated with a specific reduction in the amount of fully assembled COX, normal levels of the other OXPHOS complexes, but no evidence of subassemblies of the COX complex by

BN-PAGE analysis (Figure 1A). The steady-state levels of COX I and COX II, two mitochondrially-encoded COX subunits, and of COX IV, a nuclear-encoded subunit, were decreased on immunoblot analysis, typical for a defect in COX assembly (Figure

1B). The specific reduction in the levels of fully assembled COX and COX subunits prompted us to assess the rate of synthesis and the stability of the mitochondria-encoded subunits of the OXPHOS complexes. To this end, mitochondrial translation products in fibroblasts from the subject and two controls were pulse-labeled with a mixture of [35S]- methionine and cysteine, and then chased either for 10 minutes to measure the rate of synthesis of the individual polypeptides (PULSE), or for 17 hours, to assess their stability

(CHASE). A specific decrease in the amount of pulse labeling of the COX I subunit was observed (approximately 50% compared to the lowest control), while the other mtDNA- encoded polypeptides were apparently synthesized at levels similar to those in control cells (Figure 2A). On the other hand, newly synthesized COX I appeared much more stable in subject fibroblasts during the CHASE experiment. To ensure that the decrease in the synthesis of the COX I polypeptide was not due to a reduction in the levels of its

105

mRNA, a Northern blot analysis was performed for all mtDNA-encoded COX subunits, as well as for a subunit of complex I (ND1), and both ribosomal subunits (Figure 2B).

The levels of the COX mRNAs in the subject were not reduced relative to controls when normalized to the rRNA levels, suggesting the presence of a mutation in a factor specifically involved with COX I translation.

Mapping of the gene defect in subject fibroblasts

To determine whether the defect was caused by a mutation in nuclear DNA or mitochondrial DNA, subject and control cells were fused with human (143) rho0 cells, which are devoid of mtDNA. The hybrids obtained with subject fibroblasts had a COX activity of 120% of that obtained with control fibroblasts. This strongly suggested that the subject biochemical defect is an autosomal recessive trait. To test whether known COX assembly factors could suppress the defect, fibroblasts from the subject were transduced with retroviral vectors expressing the cDNAs for several known COX assembly factors viz.: COX11, COX16, COX17, OXA1, SCO2, PET191, SURF1, OXA2, COX10, COX15.1,

COX19, COX15.2, SCO1 and COX23. None of these restored COX activity.

We next attempted to identify the defective gene by microcell-mediated monochromosomal transfer (Zhu et al. 1998). A panel of monochromosomal human:mouse hybrid cells containing single human chromosomes tagged with the chimeric selection marker HyTK was used as the donor cell line for human chromosomes

(Cuthbert et al. 1995). All 22 autosomes were transferred one at a time into subject 106

fibroblasts and the resulting clones were tested for complementation of the COX defect.

Rescued clones were considered to be those with specific COX activity higher than 60% of control (about double the COX activity in the subject fibroblasts). Instead of the expected rescue with a single chromosome carrying the wild-type version of the causal disease gene, we instead observed several chromosomes that could apparently restore

COX activity (Supplementary Table 1). Several of the rescued clones were analyzed by

BN-PAGE, and all displayed control levels of fully assembled COX (data not shown).

One possible explanation for this result was that a mouse chromosome, carrying the murine homolog of the gene mutated in the subject, was transferred along with the individual human chromosomes. FISH analysis confirmed the presence of mouse chromosomes in the rescued clones (data not shown), and genotyping with mouse microsatellite markers in the rescued and nonrescued clones also showed several mouse chromosome elements in the tested clones.

We reasoned that identifying the mouse chromosomal region responsible for the rescue of COX activity could lead to the identification of the syntenic human chromosome region containing the defective gene. SNP genotyping of rescued and nonrescued clones revealed parts of mouse chromosomes 1, 11 and 15 exclusively in rescued clones. However, because the DNA samples contained both mouse and human

DNA, the genotype calling was not always unambiguous. As the subject was the child of consanguineous parents, we tried homozygosity mapping using an Affymetrix SNP array to identify regions of homozygosity in the subject that might be syntenic with the regions identified on the mouse chromosomes in the rescuing clones. The largest region of LOH 107

(loss of heterozygosity) was on chromosome 12 at position 43.5-54.2 Mbp. A region on mouse chromosome 15 (between 90-103.4 Mbp) is syntenic with a large region on human chromosome 12 (between 38.6-55.1 Mbp), and this focused our attention to this region.

Using an Illumina gene expression array we searched for transcripts in this region whose level of expression was low in fibroblast RNA from the subject compared to 11 normal and disease controls (data not shown). This analysis identified C12orf62, as the top candidate for the gene mutated in the subject.

Identification of a homozygous missense mutation in C12orf62

Sequencing of the C12orf62 cDNA obtained from subject fibroblasts revealed a homozygous missense mutation at position 88 (c.88G>A) (Figure 3A), predicting an amino acid change from methionine to isoleucine (p.Met19Ile) (Figure 3D). The mutation was confirmed by RFLP analysis in subject genomic DNA (Figure 3B). This mutation segregated with the disease phenotype in the subject’s family (Figure 3C). The two other affected siblings had the same homozygous mutation, whereas the unaffected sibling and the parents were heterozygous carriers of the mutation. C12orf62 contains two exons, only one of which is coding, and is predicted to produce a 57 amino acid protein of about 6 kDa. Orthologs of the protein appear to be largely confined to the vertebrates, and the mutated methionine, which is predicted to be in a transmembrane helix, is conserved in all C12orf62 orthologs.

108

Rescue of the COX deficiency by expression of the C12orf62 cDNA

Immunoblot analysis of mitochondrial extracts using polyclonal antibodies directed against both N- and C- terminal C12orf62 peptides showed that C12orf62 was undetectable in subject fibroblasts, but present in control cells, and also in an individual with TACO1 mutations (Figure 4 A, C), suggesting that the missense mutation destabilizes the protein. To investigate whether the lack of functional C12orf62 protein was responsible for the COX defect, wild-type C12orf62 was expressed in subject fibroblasts using a retroviral expression vector. Expression of C12orf62 in controls

(Figure 4A) had no measureable effect on the assembly of the OXPHOS complexes or mitochondrial translation, but it completely rescued the COX assembly defect (Figure

4A), the COX I translation in the PULSE assay (Figure 4B), and COX activity (92 % of control) in subject fibroblasts.

C12orf62 is a mitochondrial protein

None of the mitochondrial targeting prediction programs (Predotar, Target P,

MitoPred, MitoProt and MitoPWolf) predict C12orf62 to be targeted to mitochondria, although it is in the MitoCarta database (Pagliarini et al. 2008) and it was one of the highest scoring genes in a screen for OXPHOS gene expression (Baughman et al. 2009).

The mitochondrial localization of C12orf62 was demonstrated by immunocytochemistry in cells expressing a C-terminal myc-tagged version of C12orf62 (Figure 5A) that had been shown to completely rescue the biochemical defect in subject fibroblasts (data not 109

shown). C12orf62 was present exclusively in the pellet obtained after alkaline sodium carbonate extraction of mitochondria from HEK 293 cells, indicating that C12orf62, like

COX I is an integral membrane protein, and not membrane associated as is the 70kDa subunit of complex II (Figure 5B).

RNAi-mediated knockdown of C12orf62 recapitulates the biochemical phenotype of the subject

To further investigate the function of C12orf62, we used siRNA-mediated knockdown of the protein. Transient knockdown of C12orf62 in control fibroblasts, with two different siRNAs, resulted in a defect in COX assembly that was similar to that seen in the subject (Figure 6A), and reduced COX activity (38-53% of control), that correlated with the severity of the assembly defect. It also resulted in a decreased intensity of pulse labeling of COX I in the translation assay (Figure 6B). In the subject fibroblasts, the same knockdown exacerbated the biochemical phenotype, implying that the mutant protein, while not detectable by immunoblot analysis, retained some residual function. It also reduced the stability of COX I in the subject cells to levels which were indistinguishable from controls (Figure 6C). Interestingly, although the siRNA KD2 reduced COX activity to near background levels, it did not reduce the labeling intensity of COX I below that seen in KD1 (Figure 6B), implying that C12orf62 functions in some other aspect of COX biogenesis. We conclude that C12orf62 is specifically involved with the synthesis and

110

subsequent assembly of COX I, and that the presence of a small amount of the mutant protein in cells from the subject stabilizes newly synthesized COX I.

C12orf62 interaction with COX subunits

To investigate if the C12orf62 interacts or associates with COX I, a size exclusion experiment was performed in HEK 293 cells. The majority of C12orf62 appeared in the fraction corresponding to a molecular weight of about 110 kDa, containing COX I, and a small amount of COX IV (Figure 7), suggesting that C12orf62 interacts with newly synthesized COX I at an early step in holoenzyme assembly. (Fully assembled COX has a MW of about 220 kDa). A minority of C12orf62 also appeared in the same fraction as

COX II and COX IV, both of which are involved in early COX assembly steps. To confirm the interaction of C12orf62 and COX 1, immunoprecipitation with anti-FLAG antibody of control fibroblasts expressing C12orf62-FLAG was performed and showed that C12orf62 coimmunoprecipitated with a small fraction of COX I, COX II and COX

IV (Figure 8A). Reciprocal immunoprecipitation, using anti-COX I and COX II, verified the interactions (Figure 8B).

C12orf62 interaction with proteins involved in mitochondrial translation

To further investigate the function of C12orf62 in the synthesis of COX I, we searched for binding partners of C12orf62 using the FLAG-tagged construct, which we

111

had previously shown to be functional. Anti-FLAG-agarose beads were used for immunoprecipitation and the eluates were analyzed by mass spectrometry. C12orf62 coimmunoprecipitated with the translation elongation factor EFTu, LRPPRC and SLIRP, the latter two of which form an mRNP complex that regulates the stability and handling of mature mRNAs (Baughman et al. 2009; Sasarman et al. 2010). EFTu was not detected in the control and LRPPRC and SLIRP were >10-fold enriched. The results of the mass spectrometry analysis were validated by immunoblot analysis (Figure 8C).

Failure to assemble a stable COX complex in fibroblasts from the subject

The above data suggested that C12orf62 plays a role in the assembly or stability of the COX complex. To test this we pulse labeled the mitochondrial translation products in the presence of a reversible cytosolic translation inhibitor, and assessed the stability of the newly synthesized COX subunits after 8, 16 and 32 hours chase (Figure 9). The labeling intensity of the newly synthesized COX subunits was normalized to ATP6. In control cells the labeling intensity of all 3 COX subunits decreased by 3-5 fold even in the shortest chase (8 hours), indicating that they were apparently synthesized at levels in excess of that required for new assembly of the COX holoenzyme (Figure 9B). In the subject cells, although COX I was synthesized at ~25% of control levels, the newly synthesized protein was completely stable. This contrasted with COX II and COX III, whose turnover was much faster than in control cells (Figure 9B).

112

Incorporation into the holoenzyme complex of the newly synthesized mtDNA- encoded COX subunits was assessed by 2D BN-PAGE (Figure 10). In control cells COX

I is incorporated in two complexes, one corresponding to the fully assembled holoenzyme

(S4) and a lower molecular weight complex corresponding to an intermediate assembly complex (S3) (Nijtmans et al. 1998). A similar pattern is observed in the subject cells, where COX I stability appears normal in the fully assembled complex, but slightly decreased in the assembly intermediate. By contrast, newly synthesized COX II and III were less efficiently incorporated into the COX complex, and were rapidly turned over after a long chase. We conclude that C12orf62 is necessary for the assembly and stability of the nascent COX complex.

113

Discussion

We have identified a missense mutation in C12orf62 as the cause of a fatal, atypical, neonatal respiratory chain defect associated with a deficiency in COX assembly.

Several pieces of evidence support this conclusion. (1) Transfer of chromosome 12 or expression of wild-type C12orf62 cDNA rescues the biochemical defect in fibroblasts from the subject (2) The mutation (p.Met19Ile) occurs at an amino acid residue that is

100% conserved in all C12orf62 orthologs. (3) C12orf62 is not detectable by immunoblot analysis in fibroblasts from the subject, suggesting that the mutation destabilizes the protein (4) siRNA-mediated knockdown of C12orf62 in control fibroblasts recapitulates the phenotype caused by the Met19Ile mutation, and it exacerbates the biochemical defect in fibroblasts from the subject.

C12orf62 is a very small, uncharacterized protein of 6 kDa, predicted to be a single pass transmembrane protein. It appears largely confined to the vertebrate lineage.

Although it is not clear how it is imported into mitochondria, the data presented here demonstrate that it is an integral mitochondrial membrane protein, consistent with the structural predictions. It was identified in multiple tissues in the Mitocarta database, indicating that it is ubiquitously expressed. Knockdown or mutation of C12orf62 results in a marked, and specific, decrease in the labeling of COX I in a pulse translation experiment, and in a defect in the assembly/stability of the holoenzyme from the newly synthesized mtDNA-encoded COX subunits. These data suggest that the role of C12orf62 may be to couple the synthesis of COX I to subsequent assembly of the nascent subunits

114

into the holoenzyme complex. This function is reminiscent of Mss51p in S. cerevisiae, a protein that acts both as a translator activator for COX1 mRNA and, as part of a heteroligomeric complex, as an early assembly factor. Mss51p binds to the 5’ UTR of the

COX1 mRNA to control translation and also associates in a high molecular weight complex with several other proteins, including Cox14p, Surf1p, Coa1p, Coa3p to assist in early assembly (Mick et al. 2011). The current model proposes that Mss51p exists in two pools, one that is competent for activating COX1 translation and another that assists in chaperoning Cox1p in the assembly process. This then would serve to couple supply of newly synthesized Cox1p subunits with demand.

It is still not clear exactly how translation initiation of the mitochondrial mRNAs occurs in mammals, as most mammalian mitochondrial mRNAs lack significant 5’UTRs; however, it seems probable that a family of translational activator proteins exists, analogous to those in yeast. Only one such protein, TACO1, has been identified to date in mammals. It is essential for the efficient translation of full-length COX I, but the molecular mechanism remains unknown (Weraarpachai et al. 2009). Deletion mutants of the yeast ortholog of TACO1 (YGR021w) have normal Cox1p synthesis and are respiratory competent, indicating that the protein has evolved a different function in mammals. C12orf62 is essential for the normal level of COX I labeling in a pulse translation experiment, and the labeling intensity of newly synthesized COX I correlates with the severity of the COX assembly defect up to a point. Although we cannot be certain that this represents a true synthesis defect, rather than rapid proteolysis of the newly synthesized COX I peptide, a number of observations argue against the latter. 115

First, in fibroblasts that are null for TACO1, prematurely truncated forms of COX I accumulate in the pulse, and are only turned over in a longer chase (Weraarpachai et al.

2009), suggesting that the quality control system does not immediately recognize unassembled COX I subunits. Second, both LRPPRC and SLIRP, which form a ribonucleoprotein complex with COX I mRNA, and are essential for its stability

(Sasarman et al. 2010), are more than ten-fold enriched in the immunoprecipitate of

C12orf62, suggesting that C12orf62 plays a direct role in the synthesis of COX I. Finally, while newly synthesized COX I turns over rapidly in the context of other reported defects in COX assembly such as SCO1 (Leary et al. 2009) or TACO1 (Weraarpachai et al.

2009), it is completely stable in fibroblasts from the C12orf62 subject.

C12orf62 is present in a higher molecular weight complex of about 110 kDa, suggesting that it associates with other mitochondrial proteins. It coelutes on a size exclusion column with a fraction of COX I and COX IV, two subunits that nucleate the earliest step of COX assembly (Nijtmans et al. 1998). Immunoprecipitation experiments also showed that COX I, COX II and COX IV coimmunoprecipitate with C12orf62-

FLAG, and in reciprocal experiments, that immunoprecipitation of COX I and COX II brings down C12orf62-FLAG. These data all suggest that the majority of C12orf62 interacts with the newly synthesized core subunits of COX. It appears to play a crucial role in stabilizing the newly assembled enzyme based on the rapid turnover of COX II and COX III subunits in subject fibroblasts. Many small single transmembrane proteins are structural components of the OXPHOS complexes, and it has been proposed that they may function to organize their hydrophobic domains (Zickermann et al. 2010). One 116

possibility is that C12orf62 acts to stabilize COX I in the inner membrane, allowing it to assemble with COX IV and Va, and permitting the addition of heme a and the two copper atoms that are necessary to mature this subunit. This activity may make it unavailable for a role in translation.

The translation factor EFTu also coimmunoprecipitated with C12orf62-FLAG.

Although its major role is delivering aminoacyl-tRNAs to the acceptor (A) site on the ribosome during the elongation stage of translation, it also appears to have a chaperone function (Suzuki et al. 2007). It interacts with misfolded newly synthesized polypeptides from the mitochondria ribosome, prevents thermal aggregation of proteins in vitro, and enhances protein refolding in vitro. Thus EFTu could have a role in quality control of the nascent COX enzyme complexes through binding of C12orf62.

Surprisingly, despite a COX activity in the patient fibroblast at 30-40% of control, the phenotypic presentation and clinical outcome in the subject investigated here is much more severe than that of patients with SURF1 (Zhu et al. 1998) or TACO1 (Weraarpachai et al. 2009) mutations, whose fibroblasts have even less fully assembled COX. This likely reflects tissue-specific requirements for the individual COX assembly factors (Antonicka et al. 2006; Leary et al. 2007); however, as all of these factors are ubiquitously expressed, the molecular basis for this specificity remains unclear. A resolution of this question awaits the development of appropriate cellular or animal models of disease.

117

Materials and Methods

Subjects

The index subject was the first child of consanguineous healthy Portuguese parents (Figure 3C, II: 1). Two subsequent siblings later presented with an identical disease (II: 3, II: 4) while one is healthy. Oligohydramnios and septum lucidum cyst were noted at the 28th and 37th week of gestation. The subject was born at full term; her birth weight was 3250 grams (50th percentile) and her head circumference 37 centimeters (95th percentile). She presented with neurological and respiratory distress immediately after birth. She was dysmorphic with hypotelorism, micro-ophthalmia, ogival palate and a left single palmar crease. She had severe metabolic acidosis (pH=7.03, pCO2= 30, Base excess= -22) at 4 hours after birth, which was associated with ketonuria, thus confirming the abnormal intermediary metabolism. Lactate concentration was 23.4 mM in blood and

26.2 mM in CSF at 16 hours after birth. Death occurred 24 hours after birth. Autopsy disclosed brain hypertrophy with diffuse alteration of the white matter myelination and numerous cavities in the parieto-occipital region, brainstem and cerebellum, in association with hepatomegaly, hypertrophic myocardiopathy, renal hypoplasia and adrenal glands hyperplasia. Karyotype was normal 46, XX.

Informed consent was obtained from the subject’s parents, and the research study was approved by the Institutional Review Board of the Montreal Neurological Institute.

118

Cell lines

Primary cell lines were established from the index subject’s skin fibroblasts. The subject and control cell lines were immortalized by transduction with a retroviral vector expressing the HPV-16 E7 gene in addition to a retroviral vector expressing the catalytic component of human telomerase (htert) (Yao and Shoubridge 1999). The fibroblasts and

HEK 293 line were grown in high glucose Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum, at 37°C in an atmosphere of 5% CO2 .

Enzyme activity measurements

Enzyme activities were measured by spectrophotometric assays of whole cell extracts. COX activity was normalized to citrate synthase activity and specific activity was determined by protein content as described (Yao and Shoubridge 1999). Protein concentration was measured by the Bradford method. The normalized COX activity was reported as percentage of the average of several controls.

Electrophoresis and immunobloting

Mitoplasts were prepared from fibroblasts by treatment with 0.8 mg of digitonin per mg of protein, solubilized with 1% n-dodecyl maltoside (DDM). Twenty micrograms of the solubilized protein were used for electrophoresis as described in detail elsewhere

(Leary et al. 2009). Samples were run on 6%-15% non- denaturing polyacrylamide 119

gradient gels (Blue-Native PAGE), as previously described (Klement et al. 1995).

Individual structural subunits of complexes I, II, III, IV and V were detected by immunoblot analysis using commercially available monoclonal antibodies (Mitoscience), except for complex I, where a polyclonal antibody against subunit ND1 (a kind gift of

Anne Lombès, INSERM, Paris) was used. For two-dimensional analysis 30 g of 35S- labeled mitochondrial translation products were run in the first dimension on 8-12% non- denaturing polyacrylamide gradient gels, followed by a separation on 10-16% Tricine-

SDS gradient gels (Antonicka et al. 2003c). The gels were dried and exposed to

PhosphorImager (GE Healthcare) cassette for a minimum of 3 days.

SDS-PAGE was used to separate denatured whole cell extracts, mitochondrial extracts and fractions from both immunoprecipitation and size exclusion experiments.

Whole cells were extracted with 1.5% DDM/PBS after which 20 g of protein were run on 12% polyacrylamide gels. Isolated mitochondria were extracted with 50 mM HEPES buffer, pH7.6, 150 mM NaCl, 1% taurodeoxy cholate, containing complete protease inhibitors (Roche), and 80 g of protein were run on 12.5% polyacrylamide gels.

Fractions from the immunoprecipitation and size exclusion experiments were also run on

12.5% polyacrylamide gels, followed by wet transfer to a nitrocellulose membrane, and immunoblot analysis with the indicated antibodies. To immunoblot C12orf62 the transfer buffer contained 20% methanol and no SDS. Monoclonal antibodies against COX I, COX

II, COX IV, complex I-39 kDa and complex II-70 kDa were obtained from Mitoscience.

Monoclonal antibodies against actin and FLAG were obtained from Sigma and against

Porin from Calbiochem. LRPPRC and SLIRP antibodies were used as described 120

previously (Sasarman et al. 2010). The antibody against EFTu/Ts was a kind gift of Linda

Spremulli. C12orf62 antibodies were made as described later.

Pulse labeling of mitochondrial translation products

In vitro labeling of mitochondrial translation products was performed as previously described (Leary et al. 2009). Briefly, cells were pulse-labeled for 60 min at

37o C in methionine/cysteine-free DMEM containing 200 µCi/ml of a

[35S]methionine/cysteine mix (Perkin Elmer) and 100 g/ml of either emetine or anisomycin to inhibit cytosolic translation. To assess the mitochondrial translation rate the cells were chased for 10 min (PULSE). To assess the stability of the newly synthesized peptides the cells were chased for 17 hours (CHASE) in regular DMEM, or for 8, 16 and 32 hours in a time course experiment. For chase studies, cells were incubated for 23 hours in 40 g /ml chloramphenicol to inhibit mitochondrial translation prior to labeling. Total cellular protein (50 µg) harvested from cells after chasing was resuspended in loading buffer with a pH of 6.7 containing 93 mM Tris-HCl , 7.5% glycerol, 1% SDS, 0.25 mg bromophenol blue/ml and 3% mercaptoethanol, sonicated for

3–8 seconds, loaded and run on 15-20% polyacrylamide gradient gels. The labeled mitochondrial translation products were detected through direct autoradiography.

Preparations obtained after 17 and 32 hours of chase were used to run the second dimension BN-PAGE as described in the previous section.

121

Northern blot analysis

RNA was isolated from patient and control fibroblasts using the RNeasy Kit

(Qiagen). Ten micrograms of total RNA were separated on a denaturing

MOPS/formaldehyde agarose gel and transferred to a nylon membrane. PCR products

(300-500 bps) of individual mitochondrial genes (COX I, COX II, COX III, ND 1, 12S,

16S) were labeled with [-32P]-dCTP (GE Healthcare) using the MegaPrime DNA labeling kit (GE Healthcare). Hybridization was performed according to the manufacturer’s manual using the ExpressHyb Hybridization Solution (Clontech) and the radioactive signal was detected using the Phosphoimager system.

Microcell-mediated chromosome transfer

Immortalized subject skin fibroblasts were fused with microcells from a monochromosomal human:mouse hybrid cell line containing single human chromosomes tagged with the hygromycin resistance gene (Zhu et al. 1998).

Homozygosity mapping analysis

DNA extracted from subject fibroblasts was used for a genome–wide search for regions of homozygosity with the Affimetrix GeneChip 250K Nsp I (The Centre for

Applied Genomics, The Hospital for Sick Children, Toronto, Canada). The data were analyzed using Affymetrix Genotyping Console software (Affymetrix). 122

Mutation detection

Total RNA isolated using the RNeasy Kit (Qiagen) was used to amplify the

C12orf62 cDNA using the OneStep RT-PCR kit (Qiagen) and the gel-purified PCR fragments were used for direct sequencing. Total genomic DNA from controls, subject fibroblasts and blood from family members was isolated using the DNeasy Kit (Qiagen).

Primers specific for exon 2 of the C12orf62 gene were used to amplify the DNA, followed by either digestion with Nla III or direct sequencing.

cDNA constructs and virus production and infection

Retroviral vectors containing the cDNA sequence of COX assembly factors

(COX11, SCO1,COX16, COX17, OXA1, SCO2, PET191, SURF1, OXA2, COX10,

COX15.1, COX19, COX23 and COX15.2) or C12orf 62 were created with the GatewayTM

Cloning system (Invitrogen) as previously described (Antonicka et al. 2003a). cDNAs from the individual genes were amplified by the OneStep RT-PCRTM (Qiagen) using specific primers modified for cloning into Gateway vectors. The PCR constructs were cloned into Gateway-modified retroviral expression vectors, pLXSH or pBabe. For the C- terminal FLAG and Myc-tagged C12orf62 (C12orf62-FLAG, C12orf62-Myc), cDNA from C12orf62 was amplified by OneStep-RT-PCRTM (Qiagen) using specific primers with FLAG and Myc sequences inserted before the stop codon and subsequently cloned into pBabe. The fidelity of cDNA clones was confirmed by DNA sequencing. Retroviral constructs were transiently transfected into the Phoenix packaging cell line using the 123

HBS/Ca3(PO4)2 method

(http://www.stanford.edu/group/nolan/protocols/pro_helper_dep.html). Subject and control fibroblasts were infected 48 hours later by exposure to virus-containing medium in the presence of 4 g /ml of polybrene as described (Lochmuller et al. 1999).

C12orf62 antibody production

A polyclonal antibody against two human N and C terminal C12orf62 peptides

(MPTGKQLADIGYKT and CFQWRRAQRQAAEEQKT respectively) was prepared by

21st Century Biochemicals (Marlboro, MA). Crude serum and affinity purified antibodies were tested on cell lines overexpressing C12orf62 protein and detected a band of approximately 6 kDa. The mix of C-terminal and N-terminal affinity purified antibody was used for further experiments at a dilution of 1:150 for C-terminal and 1:175 for N- terminal.

Mitochondrial isolation and localization experiments

HEK 293 cells and fibroblasts were resuspended in ice-cold 250 mM sucrose/10 mM Tris-HCl/1 mM EDTA (pH 7.4) and homogenized with 10 passes through a pre- chilled, zero clearance homogenizer (Kimble/Kontes, Vineland, NJ). A postnuclear supernatant was obtained by two consecutive centrifugations for 10 min at 600g.

Mitochondria were pelleted by centrifugation for 10 min at 10000g, and washed once in 124

the same buffer. Mitochondria from HEK 293 cells (200 g) were further extracted with

100 mM sodium carbonate pH 11.5 as previously described (Yao and Shoubridge 1999) and the relevant fractions were analyzed by SDS–PAGE. COX I was used as a marker of an integral membrane protein, and complex II -70 kDa subunit as a marker for an inner membrane-associated protein.

Immunocytochemistry

Control fibroblasts overexpressing C12orf62-Myc were grown on coverslips for

24 hours, fixed in paraformaldehyde, solubilized by Triton X-100, and incubated with anti-c-Myc (Sigma) and anti-cytochrome c (Santa Cruz Biotechnology) antibodies. Anti- mouse ALEXA Fluor 594 and anti-rabbit ALEXA Fluor 488 secondary antibodies

(Invitrogen) were used for immunofluorescent detection.

Size exclusion chromatography

Mitochondria from HEK 293 cells (800 g) were extracted in 50 mM HEPES buffer, pH7.6, 150 mM NaCl, 1% taurodeoxy cholate and complete protease inhibitors

(Roche). The extracted samples were separated on a Tricorn Superdex 200 10/30 HR column (GE Healthcare) as described (Kaufman et al. 2007) and each fraction was TCA- precipitated. Elution profiles of C12orf62 were determined by immunoblot analysis using

125

antibodies against C12orf62. Antibodies against COX I, COX II and COX IV were used to detect the elution pattern of the COX complex.

Gene expression analysis

Total RNA was isolated from subject and control skin fibroblasts using the

RNeasy Kit (Qiagen). The RNA expression profile was obtained by Illumina gene expression profiling using HumanWG-6 Expression Bead Chip (Illumina) (The Centre for Applied Genomics, The Hospital for Sick Children, Toronto, Canada). The data were analyzed by screening for candidate genes showing the largest difference in gene expression between the subject and 11 controls standardized to the minimal variance in expression levels between controls to reduce false positives.

Mouse SNP genotyping analysis

DNA was isolated from several rescued and non rescued clones with different transferred human chromosomes, as well as from mouse control strains and subject fibroblasts by DNeasy kit (Qiagen). Illumina SNP genotyping using Mouse Bead Array technology was performed to identify the presence of mouse chromosomes (The Centre for Applied Genomics, The Hospital for Sick Children, Toronto, Canada). The data were analyzed using automated genotype clustering and calling to find mouse candidate

126

regions where the marker was present in the rescued clones but absent in the non-rescued clones.

siRNA transfection

RNA interference was used for transient knockdown of the expression of

C12orf62. Two Stealth RNA interference (RNAi) duplexes against human C12orf62

(Invitrogen) were designed using Block-iT RNAi Express. Stealth KD1

(C12ORF62HSS189308 with sequence of aacaggacuagagcguugaugguuu) , stealth KD2

(C12ORF62HSS131590 with sequence of gacauuggcuauaagaccuucucua) and the fluorescent oligo control Block-iT Alexa FluorRed (Invitrogen) were transiently transfected into subject and control fibroblasts at a final concentration of 12 nM using

Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s specifications.

The transfection was repeated on day 3 and 6 and the cells were harvested on day 8 and analyzed by COX assay, Blue-Native PAGE and in vitro labeling of mitochondrial translation products.

Protein Alignment

The protein sequences of C12orf62 from different species were used for multiple protein alignment with MultAlin. The human C12orf62 sequence was used to predict a transmembrane domain by using the TMHMM Server, v 2.0.

127

Immunoprecipitation (IP)

For the FLAG IP, 400 g of mitochondria isolated from control fibroblasts overexpressing C12orf62 tagged with a FLAG-epitope at the C-terminus (C12orf62-

FLAG) were extracted in 100 l of 50 mM HEPES buffer, pH7.6, 150 mM NaCl, 1% taurodeoxycholate and complete protease inhibitors (Roche) on ice for 45 minutes, with occasional vortexing. The extract was centrifuged at 25000g at 4 C for 40 min and the supernatant was pre-cleared overnight with non-coated agarose beads (EZ viewTM Red

Protein G Affinity Gel, Sigma) to reduce non-specific protein binding to the beads. The precleared extract was used for immunoprecipitation with FLAG coated agarose bead

(EZ viewTM Red ANTI-FLAG M2 Affinity Gel from Sigma) according to the manufacturer’s instructions with some modifications. The incubation time was extended to overnight and the wash solution was 0.1 M Na-phosphate with 0.08% Tween 20. The proteins were eluted from the beads by using 0.1 M glycine pH2.5 with 0.08% Tween 20 and 1% DDM at room temperature. The IP fractions were analyzed by immunoblotting with antibody as described earlier and the eluates were TCA precipitated and analyzed by mass spectrometry (Orbitrap, Thermo Scientific, Watlham, MA) at the Institute de

Recherches Cliniques de Montreal.

For the COX I and COX II IP, 400 g of control fibroblasts overexpressing

C12orf62-FLAG were used. The extraction procedure is described above.

Immunoprecipitation was performed with Dynabeads protein A (Invitrogen) alone

(control IP) or coated with either COX I or COX II antibodies (MitoScience; experiment

128

IP) according to the manufacturer’s instructions (version no.004 with some exceptions).

Incubation of the antibody with the beads and the incubation of the extract with antibody

cross-linked to the beads were both carried out overnight. The beads were washed once

with 0.1 M glycine pH2.5, 0.08% Tween 20, and 0.05% DDM before the proteins were

eluted from the beads by using 0.1 M glycine pH2.5, 0.08% Tween 20, and 1% DDM at

room temperature. The IP fractions and eluates were analyzed as above.

Acknowledgements

This research was supported by a grant from the CIHR to EAS, by grants from

Agence Nationale pour la Recherche (ANR MNP 2008 Mitopark: ANR-08-MNPS-020-

01) and AFM (Association Française contre les Myopathies) to AL, and by a grant from

Agence Nationale de la Recherche (ANR-08-GENOPAT-043) to AR. EAS is an

International Scholar of the HHMI. We thank Timothy Johns and Neil Webb for

technical assistance.

Web Resources http://www.ncbi.nlm.nih.gov/omim http://rnaidesigner.invitrogen.com/rnaiexpress http://multalin.toulouse.inra.fr/multalin/ http://www.cbs.dtu.dk/services/TMHMM/

129

Figure Legends

Figure 1. Characterization of the COX defect in subject fibroblasts. (A) BN-PAGE analysis of the OXPHOS complexes in the subject and two control fibroblast lines.

Specific antibodies for subunits of the five OXPHOS complexes were used for immnoblotting. Complex II was used as a loading control. (B) SDS-PAGE analysis using antibodies against COX subunit1 (COX I), COX subunit2 (COX II) and COX subunit 4 (COX IV). The 70-kDa subunit of complex II, porin and actin were used as loading controls.

Figure 2. Defect in COX I translation and normal levels of mitochondrial mRNAs in subject fibroblasts. (A) Pulse labeled mitochondrial polypeptides were chased for 10 minutes (PULSE) and overnight (CHASE). The seven subunits of complex I (ND), one subunit of complex III (cyt b), three subunits of complex IV (COX), and two subunits of complex V (ATP) are indicated at the left of the figure. (B) Northern blot analysis of subject, TACO1 patient and control fibroblasts. Hybridization was performed with probes specific for the mitochondrial mRNAs encoding the three COX subunits and one subunit of complex I. The 12S and 16S mitochondrial rRNAs and actin were used as loading controls.

130

Figure 3. Mutation analysis of C12orf62 in the index subject. (A) Sequence analysis of exon2 of C12orf62 from genomic DNA showed a homozygous g.88G>A missense mutation in the subject fibroblasts. (B) RFLP analysis with Nla III of exon2 of C12orf62 amplified from genomic DNA from the subject, TACO1 patient and three controls fibroblasts. (C) Pedigree of the index subject (box) indicating affected and non-affected individuals. The genotype of individuals is shown in parenthesis as follow: (Hom mut) = homozygous for g.88G>A; (Het) = heterozygous for g.88G>A mutation. (D) Protein alignment of C12orf62 in different species. An arrow indicates the conserved methionine mutated in the subject. The position of the predicted transmembrane domain is indicated by a gray bar. The color bar reflects the degree of amino acid conservation between species.

Figure 4. Rescue of COX assembly and COX I translation defect in subject fibroblasts. (A) BN-PAGE analysis of the assembly of individual OXPHOS complexes in control and subject fibroblasts expressing C12orf62 from a retroviral vector (upper panel). SDS-PAGE analysis of C12orf62 indicates its level of expression using complex

II-70kDa subunit as a loading control (lower panel). (B) Pulse labeling with [35S] methionine/ cysteine of mitochondrial polypeptides and chase for 10 minutes (PULSE) in control and subject fibroblasts expressing C12orf62 from a retroviral vector. The seven subunits of complex I (ND), one subunit of complex III (cyt b), three subunits of complex

IV (COX), and two subunits of complex V (ATP) are indicated at the left of the figure.

131

(C) SDS-PAGE analysis of C12orf62 steady-state levels in two controls, TACO1 patient and subject fibroblasts. Complex II-70kDa subunit and complexI-39kDa subunit were used as loading controls.

Figure 5. C12orf62 is a mitochondrial membrane protein. (A) Immunocytochemistry of control fibroblasts transfected with C12orf62-Myc and detected with an anti-myc antibody using cytochrome c as mitochondrial protein marker. (B) SDS-PAGE analysis of alkaline sodium carbonate extracts of mitochondria from HEK 293 cells using antibodies against C12orf62, the complex II-70kDa as a marker of a membrane- associated protein and COX I as a marker of an integral inner membrane protein.

Figure 6. Knockdown of C12orf62 in control and subject fibroblasts results in a specific defect in COX assembly and COX I translation. (A) BN-PAGE analysis of the assembly of individual OXPHOS complexes in control and subject fibroblasts transiently transfected with two different siRNA constructs specific to C12orf62 (KD1 and KD2), with a fluorescent control siRNA (Alexa) and without siRNA (Mock).

Complex IV (OE) is a longer exposure of the same blot. (B) Pulse labeling with [35S] methionine/ cysteine of mitochondrial polypeptides of the samples in panel (A). The seven subunits of complex I (ND), one subunit of complex III (cyt b), three subunits of complex IV (COX), and two subunits of complex V (ATP) are indicated at the left of the figure. COX I synthesis was quantified by normalizing to the labeling of ND3. COX 132

activity was also measured in these samples. The values are shown under the translation gel expressed as percentage of the mock control. (C) The mitochondrial translation products were pulse-labeled and chased for 10 minutes (PULSE) or 17 hours (CHASE) in control and subject fibroblasts, and in cells which were transfected with one siRNA constructs specific to C12orf62 (KD2) or a fluorescent control siRNA (Alexa). COX I labeling was normalized to the labeling of ND3.

Figure 7. C12orf62 elutes in a high molecular weight complex of 110 kDa.

Mitochondria from HEK 293 cells were extracted and separated on a size exclusion column and individual fractions were run on SDS-PAGE gels and immunobloted with antibodies against C12orf62, COX subunit 1 (COX I), COX subunit 2 (COX II) and COX subunit 4 (COX IV). The molecular weight of individual fractions was determined by the elution profile of a set of standards. The fully assembled COX (S4) appears at ~220 kDa.

Figure 8. C12orf62-FLAG co-immunoprecipates with COX I, II and IV, and with

LRPPRC, EFTu, and SLIRP. (A) Mitochondria from fibroblasts expressing C12orf62-

FLAG or from control fibroblasts were extracted and incubated with anti-FLAG agarose beads. Each fraction was analyzed by immunoblotting with antibodies against FLAG,

COX I, COX II and COX IV. (B) Mitochondria from the control fibroblasts expressing

C12orf62-FLAG were extracted and incubated with magnetic beads either uncoated or coated with either COX I or COX II antibody. The input, unbound and eluate fractions 133

were analyzed by immunoblotting with antibodies against FLAG, COX I and COX II.

(C) The immunoprecipitation was performed with FLAG beads as described in (A) and each fraction was analyzed by immunoblotting with antibodies against FLAG, LRPPRC,

EFTu and SLIRP.

Figure 9. Synthesis and turnover of COX subunits in control and subject fibroblasts. (A) Pulse labeling with [35S]methionine/cysteine of mitochondrial polypeptides and a chase for 10 minutes (PULSE), and the time course of degradation by chasing for 8, 16 and 32 hours (CHASE) in control and subject fibroblasts. The seven subunits of complex I (ND), one subunit of complex III (cyt b), three subunits of complex

IV (COX), and two subunits of complex V (ATP) are indicated at the left of the figure.

(B) Histograms showing the quantification of the level of the three COX subunits normalized to the level of ATP6 in the pulse and chase experiment in (A).

Figure 10. COX subunits 2 and 3 (COX II and III) are not properly assembled into the nascent COX complex in subject fibroblasts. Second dimension BN-PAGE analysis of the sample pulse labeled with [35S] methionine/cysteine and chased for 17 and

32 hours. The gels were dried and the labeled mitochondrial translation products were detected through direct autoradiography. The arrows show different mtDNA-encoded polypeptides: one complex III subunit (cyt b), two subunits of complex V (ATP6 and

ATP8), COX subunit 1 in fully assembled complex IV (COX I-S4), COX subunit 1 in the 134

S3 intermediate of complex IV (COX I-S3), COX subunit 2 (COX II) and COX subunit 3

(COX III) in fully assembled COX. Quantification of the amount of label in each of the identified COX subunits, normalized to the label in ATP6 is shown beneath the figure for both the 17 and 32 hours chase.

Supplemental Table 1. Clonal analysis of microcell-mediated transfer of all 22 autosomes into subject fibroblasts. The total number of clones analyzed for each chromosome is indicated, as is the percentage of clones with greater than 60% of control

COX activity.

135

Figures

Figure 1

136

Figure 2

137

Figure 3

138

Figure 4

139

Figure 5

140

Figure 6

141

Figure 7

142

Figure 8

143

Figure 9

144

Figure 10

145

Supplementary Table 1

Chromosome Total Clones Rescued clones (% total) 1 37 12 (32%) 2 36 4 (11%) 3 42 6 (14%) 4 13 2 (15%) 5 47 12 (26%) 6 48 1 (2%) 7 43 3 (7%) 8 40 2 (5%) 9 3 0 10 30 6 (20%) 11 6 0 12 24 4 (17%) 13 18 1 (6%) 14 12 0 15 18 1 (6%) 16 17 1 (6%) 17 5 0 18 12 0 19 38 4 (14%) 20 33 6 (18%) 21 17 2 (12%) 22 27 0

146

General Discussion

The aim of this thesis was to identify the genes responsible for cytochrome c oxidase deficiency in two index subjects with different and unique clinical phenotypes. In these two subjects, none of the known COX assembly factors complemented the defects suggesting that novel genes were responsible for the defects. Once the genes were identified, the second aim of this thesis was to characterize their normal function. The two index subjects also had defects in mitochondrial translation, and up to the present the mechanisms that regulate mammalian mitochondrial translation remain largely unknown.

Understanding the function of the identified genes would thus contribute to a further understanding of the mechanism of mitochondrial translation.

In chapter 1, a mutation in TACO1 (Translational Activator of COX I) was identified in a child of consanguineous parents of Turkish origin who presented with encephalomyopathic disease severely affecting gait, vision, speech and growth, consistent with late onset Leigh syndrome (Seeger et al. 2010). Biochemically, the subject had impaired synthesis of COX I and a decrease in stability of COX I, II and III. TACO1 is a novel component of the mammalian mitochondrial translation apparatus that is necessary for the synthesis of full-length COX I. To date, this is the first specific mitochondrial translational activator identified in mammals. A frameshift mutation in TACO1, producing a premature stop codon, which completely eliminated expression of the

TACO1 protein, was also present in the subject’s siblings. The relatively late onset (5-16 years) and slow progression of the Leigh Syndrome phenotype in this family is rather

147

unusual, especially considering the magnitude of the COX deficiency, which was quite severe in fibroblasts and skeletal muscle (29% and 15% of control COX activity respectively in the index subject). The severity of the disease also showed a wide intra- familial variability. The two affected boys became symptomatic at age 5 (index subject) and 4 years respectively and they were unable to walk independently around age 10 years. In contrast, the three girls had the first symptoms in their teens (14, 15 and 16 years) and remained ambulatory into their twenties, suggesting that sex-specific factors may influence the phenotype.

Patients with all other previously described mutations in COX assembly factors

(SURF1, SCO1, SCO2, COX10 and COX15) present with clinical disease very early in life (first weeks or years), and have a rapidly deteriorating course, usually leading to early death (Antonicka et al. 2003b; Mootha et al. 2003; Papadopoulou et al. 1999; Tiranti et al. 1998a; Valnot et al. 2000a; Valnot et al. 2000b; Zhu et al. 1998). The differences in phenotypic presentation and clinical outcome of the COX deficiencies likely reflect tissue-specific requirements for the individual COX assembly factors (Antonicka et al.

2006; Leary et al. 2007); however, as all of these factors are ubiquitously expressed, the molecular basis for this specificity remains unclear.

We have previously shown that imbalances in the relative steady-state levels of the proteins involved in mitochondrial translation can substantially decrease the rate of mitochondrial translation, even in control cells (Antonicka et al. 2006). Consistent with this, overexpression of TACO1 resulted in a dominant negative defect in COX assembly

148

and activity. Very high levels of overexpression, achieved with a HA-tagged version of the protein, decreased translation of all mitochondrial polypeptides, suggesting that

TACO1 interacts with a common element of the mitochondrial translation apparatus that can be titrated out.

While several COX assembly factors with roles in the synthesis and assembly of prosthetic groups and in chaperoning COX subunits have been described (reviewed in

(Barrientos et al. 2009)), very little is known about the regulation of the synthesis and degradation of the individual mitochondrially encoded COX subunits in humans.

Mitochondrial translation has been studied extensively in the yeast S. cerevisiae and has resulted in the identification of one distant human homolog (LRPPRC) of the yeast translational activator Pet309p. This is probably not a true homolog as LRPPRC is involved in mitochondrial mRNA stability but does not act as a translational activator.

Whereas in yeast, mitochondrial mRNAs require mRNA-specific translational activators, which recognize sequences in their 5’-UTRs and mediate their subsequent translation, mammalian mitochondrial mRNAs lack 5’-UTRs, therefore most yeast genes involved in the translation of the mitochondrially encoded proteins lack mammalian homologs

(Fontanesi et al. 2006).

Both LRPPRC and Pet309p contain PPR motifs, which are present in a large family of proteins with roles in post-transcriptional RNA metabolism, especially in mitochondria and chloroplasts (Delannoy et al. 2007). Pet309p associates with the 5’UTR of the COX1 mRNA, and is necessary for both its stabilization and translation (Manthey

149

and McEwen 1995). Mutations in LRPPRC lead to isolated COX deficiency in French-

Canadian Leigh syndrome patients (Mootha et al. 2003). LRPPRC interacts with SLIRP, a stem-loop RNA-binding protein and functions in posttranscriptional mitochondrial gene expression as part of a ribonucleoprotein complex that regulates the stability and handling of mature mRNAs (Sasarman et al. 2010).

TACO1 is clearly necessary for the translation of full-length COX I as seen in the translation assay, however, the observation that some full-length polypeptide can be translated in its complete absence, suggests that it acts to optimize the efficiency of COX

I translation. Unlike some of the COX translational activators in yeast, which are integral inner membrane proteins, TACO1 behaves as a mitochondrial matrix protein; although we have not ruled out that it is peripherally associated with the inner mitochondrial membrane. It is present in a higher molecular weight complex of about 74kDa, suggesting either that it forms a trimer, or that it associates with other mitochondrial proteins. The detection of TACO1 and EFTs, a mitochondrial translation elongation factor, in the same size exclusion fractions suggests a possible interaction of TACO1 and the mitochondrial translation components involved in polypeptide elongation, but further studies are required to establish this relationship. This observation at least suggests that

TACO1 might act as a translational activator by either securing an accurate start of COX

I translation, or by stabilizing the elongating polypeptide and ensuring completion of its translation. Alternatively, it could interact with the peptide release factor, ensuring that the polypeptide does not dissociate from the ribosome until synthesis is complete. To date, it is not known how initiation of translation takes place in mammalian mitochondria 150

and since mammalian mitochondrial mRNAs lack 5’UTRs, TACO1, which acts as a mammalian translational activator of COX I, might operate through other parts of the

COX I transcript. Sequences downstream from the initiation codon could also be involved in ribosome binding to initiation translation, as is seen in E.coli for some leaderless mRNAs (Fox 1996) and the binding of TACO1 to COX I mRNA could occur through a similar mechanism downstream of initiation codon, but an interaction between TACO1 and COX I mRNA has not been directly confirmed.

While TACO1 is conserved through bacteria, most of the homolog sequences are annotated as hypothetical proteins. In E.coli, the hypothetical protein is yebC, the entire length of which forms the so-called DUF28 domain, which is found in all bacterial and yeast TACO1 homologs. The average length of the DUF28 domain is around 230 amino acids, but the functions of proteins consisting of the DUF28 homology domain remain largely unknown. Recently, a DUF28 family member PmpR (from Pseudomonas aeruginosa) has been shown to be involved in a negative regulation of the quorum- sensing response regulator by binding to an upstream promoter element in the gene

(Liang et al. 2008). Crystallization was achieved for another DUF28 family member,

Aq1575, a hypothetical protein from the thermophilic bacterium Aquifex aeolicus and a

TACO1 homolog (32% identical, 52% similar). Although no obvious active site or functional domain could be identified in the crystal structure (Shin et al. 2002), the protein has a large cleft surrounded by three domains, one of which contains a high proportion of negatively charged residues. Interestingly, a sequence similar to the DUF28 domain has been reported in LRPPRC from amino acid 676 to 841 (Liu and McKeehan 151

2002).The involvement of LRPPRC in RNA processing could suggest a similar function for TACO1 since its homologs also contain the DUF28 domain. The S. cerevisiae homolog of TACO1, YGR021w, was detected in highly purified mitochondria, but its function in yeast mitochondria is unknown. The deletion strain ygr021wΔ, which was respiratory competent, did not show any growth disadvantage on non-fermentable carbon sources, and synthesized all mitochondrial polypeptides at control levels, indicating that it does not have an obvious role in COX I synthesis. Therefore the function of TACO1 in the translation of COX I appears to be specific to mammals.

In chapter 2, a mutation in C12orf62 was identified in a female patient of

Portuguese ancestry with COX deficiency who was born at term suffering from lactic acidosis, cardiomyopathy, brain damage, renal hypoplasia and dysmorphy, and who died within 24 hours of birth. Laboratory investigations identified decreased synthesis but increased stability of COX I and decreased stability of COX II and III. C12orf62 is a novel component of the mammalian mitochondrial translation apparatus. The data suggested that it is necessary for the synthesis of COX I and is involved in the assembly process of COX. A missence mutation in C12orf62 was identified that produces an amino acid change from methionine to isoleucine and decreases the stability of the C12orf62 protein, causing a very early onset disease and an unusually severe phenotype compared to reported COX patients. Despite the higher residual COX activity (30-45%) measured in this subject’s fibroblasts, as compared to the TACO1 subject discussed in chapter 1, the severity of the phenotype is surprisingly more pronounced. This could be explained by the two functions of the C12orf62 protein, or by differences in tissue-specific activity, 152

where C12orf62 may be expressed at higher levels in some tissues. As an example of tissue specificity, in patients with a fatal hepatopathy due to mutations in the mitochondrial translation elongation factor EFG1 while severity was most pronounced in the liver, the heart was much less affected. The severity of the defect paralleled the steady-state of the mutant EFG1 protein. Our results demonstrate marked differences among tissues in the organization of the mitochondrial translation system and its response to dysfunction (Antonicka et al. 2006).

C12orf62 is essential for a normal level of COX I synthesis and indeed the level of newly synthesized COX I is proportional to the level of C12orf62, as is the severity of the COX assembly defect. In addition, C12orf62 must also play a role in the assembly of

COX by stabilizing the early steps of assembly, since we have shown that the C12orf62 protein does not appear to interact with the fully assembled COX. In comparison, patients with mutations in SURF1, COX10 and COX15, which influence a downstream step of assembly, do not have a defect in COX I synthesis (Leary et al. 2009). The function of

C12orf62 appears similar to that of Mss51p, which is a well studied translation activator for COX 1 in the yeast S. cerevisiae. Mss51p is important for COX 1 translational initiation through binding to the 5’ UTR of the COX1 mRNA (Perez-Martinez et al.

2003; Zambrano et al. 2007), and for elongation, by binding to the protein coding sequence of COX1 mRNA (Perez-Martinez et al. 2003). Mss51p and newly synthesized

Cox1p form a transient complex during Cox1p synthesis on the mitoribosome and need

Cox14p, Coa3p (Cox25p), Ssc1p and Mdj1p to be stabilized (Barrientos et al. 2004;

Fontanesi et al. 2011; Fontanesi et al. 2010; Mick et al. 2011; Mick et al. 2010). After 153

Cox1p synthesis, a complex containing Ssc1p-Mss51p-Cox1p-Cox14p-Coa3p-Coa1p remains stable until Cox1p continues to the downstream assembly steps (Mick et al.

2007; Pierrel et al. 2007). This complex is suggested to down-regulate Cox1p synthesis when COX assembly is impaired by trapping Mss51p and limiting its availability for

COX1 mRNA translation (Barrientos et al. 2004; Fontanesi et al. 2010). The mechanism for the release of Mss51p is unclear but data suggested that Shy1p and the addition of

Cox6p and Cox5p, which are the first two subunits to assemble with newly synthesized

Cox1p, are the factors responsible for ejecting Mss51p from the assembling complex. We believe that C12orf62 could similarly be involved in the initiation of synthesis of COX I and subsequent binding to COX I polypeptide, to finally dissociate at the COX assembly step and initiate a second round of COX I synthesis.

We demonstrated that the role of C12orf62 is specific to the synthesis of COX I.

Knockdown of C12orf62 in the subject’s fibroblasts results in a reduction of the residual

COX activity from 30% to 9% of control, however, the defect remains specific for COX I synthesis as well as assembly of COX, and does not become generalized. In contrast, knockdown of LRPPRC to a level similar to that of the Leigh Syndrome French Canadian patient, resulted in a COX-specific deficiency, while further knockdown of LRPPRC gave rise to a generalized OXPHOS deficiency (Sasarman et al. 2010).

Whether the reduction of COX I synthesis is due to low production or rapid proteolysis is still uncertain but three arguments are in favour of the former. First, mitochondrial mRNAs in mammals exist in a ribonucleoprotein complex with two

154

proteins, LRPPRC and SLIRP, which are necessary for their stability and translation

(Sasarman et al. 2010). In our experiments, this complex is coimmunoprecipitated with

C12orf62, suggesting that C12orf62 plays a direct role in the synthesis of COX I.

Secondly, in chapter 1, translation assays of TACO1 subject fibroblasts showed prematurely truncated forms of COX I accumulating in the pulse experiment, which only degraded in the 17.5 hours chase experiment, suggesting that there are no mechanisms to immediately degrade unassembled COX I subunits. Lastly, newly synthesized COX I is completely stable in fibroblasts from the C12orf62 subject while in comparison, rapid turnover is seen in the context of several other reported defects in COX assembly, such as those associated with mutations in SCO1 (Leary et al. 2009), COX 10, COX 15, SURF1

(Unpublished data), and TACO1 (chapter 1).(Weraarpachai et al. 2009) Moreover, the rapid turnover of COX I in cells carrying mutations in any of these other factors does not have a measurable effect on the synthesis of COX I.

C12orf62 is a 6 kDa integral membrane protein present in a higher molecular weight complex of about 110 kDa, suggesting that it associates with other mitochondrial proteins. The result from the size exclusion experiment showed that C12orf62 colocalizes with COX I and COX IV, which are early subunits that join in the COX assembly.

Immunoprecipitation experiments showed that C12orf62 interacts with COXI, COX II, and COX IV. Since there is no evidence that C12orf62 associates with the fully assembled COX in the 2D BN-PAGE or size exclusion experiments, we hypothesized that C12orf62 interacts with the newly synthesized core subunits of COX. It appears to play a crucial role in stabilizing the newly assembled enzyme based on the rapid 155

degradation of COX II and COX III subunits in the nascent COX complex in subject fibroblasts.

One possibility is that C12orf62 acts to stabilize COX I in the inner membrane, allowing it to assemble with COX IV and Va, and permitting the addition of heme a and the two copper atoms that are necessary to mature this subunit. This activity may make it unavailable for a role in translation. Alternatively, its presence in an early assembly complex could signal release of a factor that represses COX I synthesis. Another possibility is that C12orf62 indirectly affects levels of COX I through release of an intermediate protein involved in synthesis and thus low levels of C12orf62 are correlated with low levels of COX I. COX I subunits in the chase experiment were more stable in the C12orf62 subject fibroblasts compared to other patients with COX assembly factor mutations. The reason might be that the COX I subunits that are stored in the mitochondrial inner membrane complexed with heme and copper are abnormally stabilized by C12orf62, or one of its intermediates, and this impedes its release for protease degradation.

C12orf62 also coimmunoprecipitated with EFTu which plays a major role in protein synthesis by delivering aminoacyl-tRNAs to the acceptor (A) site on the ribosome during the elongation stage of translation. EFTu also has a chaperone function (Suzuki et al. 2007) which suggests that it could have a role in quality control of the nascent COX enzyme complexes.

156

Unlike TACO1, C12orf62 does not have yeast or bacteria homologs, is conserved only among vertebrates, and lacks known domains, suggesting that it might be an essential protein that is specific for mammals.

Initiation of translation of yeast S. cerevisiae requires a specific translational activator which will recognize the specific sequences in 5’UTRs of individual mRNAs.

The translational activators for COX1 mRNA are Pet309p and Mss51p, while Pet111p is the translational activator for COX2 mRNA and Pet54p, Pet122p and Pet494p are specific for COX3 mRNA (Mick et al. 2011), A balanced degree of the expression of translational activators is also important for the cooperative regulation of the individual subunits. While the mechanism of recruiting mRNAs to the inner membrane for translational machinery is similar in yeast and mammals, no specific translational activators which are similar to yeast have been found in mammals. In this thesis, these two proteins are the first two specific translational activators of one of the 13 mitochondrial proteins. Up to now only translational defects which are either generalized or affect a subset of subunits have been identified, such as the mutations in LRPPRC

(Sasarman et al. 2010) which affects COX I and COX II, and C12orf65 which affects all thirteen polypeptides (Antonicka et al. 2010). TACO1 and C12orf62 mutations are the first specific translation defects affecting the synthesis of COX I alone.

The above genes were obtained mainly by using functional complementation techniques, and homozygosity mapping combined with microarray technology. In these cases, as is the case with many genes, linkage analysis was not possible due to lack of

157

family history. Identification of new genes in actual patients is clinically relevant, and in this case proved to be an excellent way to try to understand the mechanism of mammalian mitochondrial translation of COX subunits. These findings can be applied in prenatal diagnosis and for future mutation tests for the diagnosis of patients with COX deficiency. Recently, a new technology called exome sequencing, which was not available at the time these experiments were conducted, has been developed to identify mutations. This technology has proven both a fast and efficient way of mapping genes responsible for both rare and common diseases. With this new technology, future identification of other genes whose mutation leads to defects in COX should prove less labor and time-intensive and should lead to more insight into the mammalian mitochondrial translation process and the assembly of COX.

158

References

Anderson S, de Bruijn MH, Coulson AR, Eperon IC, Sanger F, Young IG (1982) Complete sequence of bovine mitochondrial DNA. Conserved features of the mammalian mitochondrial genome. J Mol Biol 156: 683-717

Anthony G, Reimann A, Kadenbach B (1993) Tissue-specific regulation of bovine heart cytochrome-c oxidase activity by ADP via interaction with subunit VIa. Proc.Natl.Acad.Sci.U.S.A. 90: 1652-1656

Antonicka H, Leary SC, Guercin GH, Agar JN, Horvath R, Kennaway NG, Harding CO, Jaksch M, Shoubridge EA (2003a) Mutations in COX10 result in a defect in mitochondrial heme A biosynthesis and account for multiple, early-onset clinical phenotypes associated with isolated COX deficiency. Hum Mol Genet 12: 2693-702

Antonicka H, Mattman A, Carlson CG, Glerum DM, Hoffbuhr KC, Leary SC, Kennaway NG, Shoubridge EA (2003b) Mutations in COX15 produce a defect in the mitochondrial heme biosynthetic pathway, causing early-onset fatal hypertrophic cardiomyopathy. Am J Hum Genet 72: 101-14.

Antonicka H, Ogilvie I, Taivassalo T, Anitori RP, Haller RG, Vissing J, Kennaway NG, Shoubridge EA (2003c) Identification and characterization of a common set of complex I assembly intermediates in mitochondria from patients with complex I deficiency. J Biol Chem 278: 43081-8

Antonicka H, Ostergaard E, Sasarman F, Weraarpachai W, Wibrand F, Pedersen AM, Rodenburg RJ, van der Knaap MS, Smeitink JA, Chrzanowska-Lightowlers ZM, Shoubridge EA (2010) Mutations in C12orf65 in patients with encephalomyopathy and a mitochondrial translation defect. Am J Hum Genet 87: 115-22

Antonicka H, Sasarman F, Kennaway NG, Shoubridge EA (2006) The molecular basis for tissue specificity of the oxidative phosphorylation deficiencies in patients with mutations in the mitochondrial translation factor EFG1. Hum Mol Genet 15: 1835-46

Barclay BJ, DeHaan CL, Hennig UG, Iavorovska O, von Borstel RW, von Borstel RC (2001) A rapid assay for mitochondrial DNA damage and respiratory chain inhibition in the yeast Saccharomyces cerevisiae. Environ Mol Mutagen 38: 153-8

Barrientos A, Barros MH, Valnot I, Rotig A, Rustin P, Tzagoloff A (2002) Cytochrome oxidase in health and disease. Gene 286: 53-63.

Barrientos A, Gouget K, Horn D, Soto IC, Fontanesi F (2009) Suppression mechanisms of COX assembly defects in yeast and human: insights into the COX assembly process. Biochim Biophys Acta 1793: 97-107 159

Barrientos A, Koor, D., Tzagoloff, A. (2002) Shy1p is necessary for full expression of mitochondrial COX1 in the yeast model of Leigh's syndrome. EMBO 21: 43-52

Barrientos A, Korr D, Barwell KJ, Sjulsen C, Gajewski CD, Manfredi G, Ackerman S, Tzagoloff A (2003) MTG1 codes for a conserved protein required for mitochondrial translation. Mol Biol Cell 14: 2292-302

Barrientos A, Zambrano A, Tzagoloff A (2004) Mss51p and Cox14p jointly regulate mitochondrial Cox1p expression in Saccharomyces cerevisiae. Embo J 23: 3472-82

Barros MH, Nobrega FG, Tzagoloff A (2002) Mitochondrial ferredoxin is required for heme A synthesis in Saccharomyces cerevisiae. J Biol Chem 277: 9997-10002.

Baughman JM, Nilsson R, Gohil VM, Arlow DH, Gauhar Z, Mootha VK (2009) A computational screen for regulators of oxidative phosphorylation implicates SLIRP in mitochondrial RNA homeostasis. PLoS Genet 5: e1000590

Belostotsky R, Ben-Shalom E, Rinat C, Becker-Cohen R, Feinstein S, Zeligson S, Segel R, Elpeleg O, Nassar S, Frishberg Y (2011) Mutations in the mitochondrial seryl-tRNA synthetase cause hyperuricemia, pulmonary hypertension, renal failure in infancy and alkalosis, HUPRA syndrome. Am J Hum Genet 88: 193-200

Bonnefoy N, Kermorgant M, Groudinsky O, Minet M, Slonimski PP, Dujardin G (1994) Cloning of a human gene involved in cytochrome oxidase assembly by functional complementation of an oxa1- mutation in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 91: 11978-82

Boulet L, Karpati G, Shoubridge EA (1992) Distribution and threshold expression of the tRNA(Lys) mutation in skeletal muscle of patients with myoclonic epilepsy and ragged- red fibers (MERRF). Am J Hum Genet 51: 1187-200.

Bowmaker M, Yang MY, Yasukawa T, Reyes A, Jacobs HT, Huberman JA, Holt IJ (2003) Mammalian mitochondrial DNA replicates bidirectionally from an initiation zone. J Biol Chem 278: 50961-9

Bratton MR, Pressler MA, Hosler JP (1999) Suicide inactivation of cytochrome c oxidase: catalytic turnover in the absence of subunit III alters the active site. Biochemistry 38: 16236-45.

Brown KR, Allan BM, Do P, Hegg EL (2002) Identification of novel hemes generated by heme A synthase: evidence for two successive monooxygenase reactions. Biochemistry 41: 10906-13.

160

Brown TA, Cecconi C, Tkachuk AN, Bustamante C, Clayton DA (2005) Replication of mitochondrial DNA occurs by strand displacement with alternative light-strand origins, not via a strand-coupled mechanism. Genes Dev 19: 2466-76

Bruno C, Kirby DM, Koga Y, Garavaglia B, Duran G, Santorelli FM, Shield LK, Xia W, Shanske S, Goldstein JD, Iwanaga R, Akita Y, Carrara F, Davis A, Zeviani M, Thorburn DR, DiMauro S (1999) The mitochondrial DNA C3303T mutation can cause cardiomyopathy and/or skeletal myopathy. J Pediatr 135: 197-202

Bykhovskaya Y, Casas K, Mengesha E, Inbal A, Fischel-Ghodsian N (2004) Missense mutation in pseudouridine synthase 1 (PUS1) causes mitochondrial myopathy and sideroblastic anemia (MLASA). Am J Hum Genet 74: 1303-8

Calvo SE, Mootha VK (2010) The Mitochondrial Proteome and Human Disease. Annu Rev Genomics Hum Genet

Camara Y, Asin-Cayuela J, Park CB, Metodiev MD, Shi Y, Ruzzenente B, Kukat C, Habermann B, Wibom R, Hultenby K, Franz T, Erdjument-Bromage H, Tempst P, Hallberg BM, Gustafsson CM, Larsson NG (2011 ) MTERF4 regulates translation by targeting the methyltransferase NSUN4 to the mammalian mitochondrial ribosome. Cell Metab 13: 527-39

Capaldi RA (1990) Structure and function of cytochrome c oxidase. Annu Rev Biochem 59: 569-96

Capaldi RA, Marusich MF, Taanman JW (1995) Mammalian cytochrome-c oxidase: characterization of enzyme and immunological detection of subunits in tissue extracts and whole cells. Methods Enzymol 260: 117-32

Carr HS, Winge DR (2003) Assembly of cytochrome c oxidase within the mitochondrion. Acc Chem Res 36: 309-16

Cavdar Koc E, Burkhart W, Blackburn K, Moseley A, Spremulli LL (2001) The small subunit of the mammalian mitochondrial ribosome. Identification of the full complement of ribosomal proteins present. J Biol Chem 276: 19363-74

Chang DD, Clayton DA (1985) Priming of human mitochondrial DNA replication occurs at the light-strand promoter. Proc Natl Acad Sci U S A 82: 351-5

Chinnery P, Majamaa K, Turnbull D, Thorburn D (2006) Treatment for mitochondrial disorders. Cochrane Database Syst Rev: CD004426

Chinnery PF, Turnbull DM (1999) Mitochondrial DNA and disease. Lancet 354 Suppl 1: SI17-21.

161

Christian BE, Spremulli LL (2009) Evidence for an active role of IF3mt in the initiation of translation in mammalian mitochondria. Biochemistry 48: 3269-78

Christian BE, Spremulli LL (2010) Preferential selection of the 5'-terminal start codon on leaderless mRNAs by mammalian mitochondrial ribosomes. J Biol Chem 285: 28379-86

Clark KM, Taylor RW, Johnson MA, Chinnery PF, Chrzanowska-Lightowlers ZM, Andrews RM, Nelson IP, Wood NW, Lamont PJ, Hanna MG, Lightowlers RN, Turnbull DM (1999) An mtDNA mutation in the initiation codon of the cytochrome C oxidase subunit II gene results in lower levels of the protein and a mitochondrial encephalomyopathy. Am J Hum Genet 64: 1330-9

Clayton DA (1982) Replication of animal mitochondrial DNA. Cell 28: 693-705

Clayton DA (1991a) Nuclear gadgets in mitochondrial DNA replication and transcription. Trends in Biochemical Sciences 16: 107-111

Clayton DA (1991b) Replication and transcription of vertebrate mitochondrial DNA. Annu Rev Cell Biol 7: 453-78

Clayton DA (1992) Transcription and replication of animal mitochondrial DNAs. Int.Rev.Cytol. 141: 217-232

Coenen MJ, Antonicka H, Ugalde C, Sasarman F, Rossi R, Heister JG, Newbold RF, Trijbels FJ, van den Heuvel LP, Shoubridge EA, Smeitink JA (2004) Mutant mitochondrial elongation factor G1 and combined oxidative phosphorylation deficiency. N Engl J Med 351: 2080-6

Comi GP, Bordoni A, Salani S, Franceschina L, Sciacco M, Prelle A, Fortunato F, Zeviani M, Napoli L, Bresolin N, Moggio M, Ausenda CD, Taanman JW, Scarlato G (1998) Cytochrome c oxidase subunit I microdeletion in a patient with motor neuron disease. Ann Neurol 43: 110-6

Cotney J, Shadel GS (2006) Evidence for an early gene duplication event in the evolution of the mitochondrial transcription factor B family and maintenance of rRNA methyltransferase activity in human mtTFB1 and mtTFB2. J Mol Evol 63: 707-17

Cuthbert AP, Trott DA, Ekong RM, Jezzard S, England NL, Themis M, Todd CM, Newbold RF (1995) Construction and characterization of a highly stable human: rodent monochromosomal hybrid panel for genetic complementation and genome mapping studies. Cytogenet Cell Genet 71: 68-76

Dairaghi DJ, Shadel GS, Clayton DA (1995) Human mitochondrial transcription factor A and promoter spacing integrity are required for transcription initiation. Biochim Biophys Acta 1271: 127-34 162

Datta K, Fuentes JL, Maddock JR (2005) The Yeast GTPase Mtg2p Is Required for Mitochondrial Translation and Partially Suppresses an rRNA Methyltransferase Mutant, mrm2. Mol Biol Cell 16: 954-63

Davies SM, Rackham O, Shearwood AM, Hamilton KL, Narsai R, Whelan J, Filipovska A (2009) Pentatricopeptide repeat domain protein 3 associates with the mitochondrial small ribosomal subunit and regulates translation. FEBS Lett 583: 1853-8

Delannoy E, Stanley WA, Bond CS, Small ID (2007) Pentatricopeptide repeat (PPR) proteins as sequence-specificity factors in post-transcriptional processes in organelles. Biochem Soc Trans 35: 1643-7

Dennerlein S, Rozanska A, Wydro M, Chrzanowska-Lightowlers ZM, Lightowlers RN (2010) Human ERAL1 is a mitochondrial RNA chaperone involved in the assembly of the 28S small mitochondrial ribosomal subunit. Biochem J 430: 551-8

Doda JN, Wright CT, Clayton DA (1981) Elongation of displacement-loop strands in human and mouse mitochondrial DNA is arrested near specific template sequences. Proc Natl Acad Sci U S A 78: 6116-20

Edvardson S, Shaag A, Kolesnikova O, Gomori JM, Tarassov I, Einbinder T, Saada A, Elpeleg O (2007) Deleterious mutation in the mitochondrial arginyl-transfer RNA synthetase gene is associated with pontocerebellar hypoplasia. Am J Hum Genet 81: 857- 62

Entelis NS, Kolesnikova OA, Martin RP, Tarassov IA (2001) RNA delivery into mitochondria. Adv Drug Deliv Rev 49: 199-215

Falkenberg M, Gaspari M, Rantanen A, Trifunovic A, Larsson NG, Gustafsson CM (2002) Mitochondrial transcription factors B1 and B2 activate transcription of human mtDNA. Nat Genet 31: 289-94

Falkenberg M, Larsson NG, Gustafsson CM (2007) DNA replication and transcription in mammalian mitochondria. Annu Rev Biochem 76: 679-99

Fan L, Kim S, Farr CL, Schaefer KT, Randolph KM, Tainer JA, Kaguni LS (2006) A novel processive mechanism for DNA synthesis revealed by structure, modeling and mutagenesis of the accessory subunit of human mitochondrial DNA polymerase. J Mol Biol 358: 1229-43

Fernandez-Vizarra E, Tiranti V, Zeviani M (2009) Assembly of the oxidative phosphorylation system in humans: what we have learned by studying its defects. Biochim Biophys Acta 1793: 200-11

163

Fontanesi F, Clemente P, Barrientos A (2011) Cox25 teams up with Mss51, Ssc1, and Cox14 to regulate mitochondrial cytochrome c oxidase subunit 1 expression and assembly in Saccharomyces cerevisiae. J Biol Chem 286: 555-66

Fontanesi F, Soto IC, Barrientos A (2008) Cytochrome c oxidase biogenesis: new levels of regulation. IUBMB Life 60: 557-68

Fontanesi F, Soto IC, Horn D, Barrientos A (2006) Assembly of mitochondrial cytochrome c-oxidase, a complicated and highly regulated cellular process. Am J Physiol Cell Physiol 291: C1129-47

Fontanesi F, Soto IC, Horn D, Barrientos A (2010) Mss51 and Ssc1 facilitate translational regulation of cytochrome c oxidase biogenesis. Mol Cell Biol 30: 245-59

Fornuskova D, Stiburek L, Wenchich L, Vinsova K, Hansikova H, Zeman J (2010) Novel insights into the assembly and function of human nuclear-encoded cytochrome c oxidase subunits 4, 5a, 6a, 7a and 7b. Biochem J 428: 363-74

Fox TD (1996) Genetics of Mitochondrial Translation, Translational Conrol. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.,p733-758

Gagnon A, Ripeau JS, Zvieriev V, Chevrette M (2006) Chromosome 18 suppresses tumorigenic properties of human prostate cancer cells. Genes Chromosomes Cancer 45: 220-30

Galati D, Srinivasan S, Raza H, Prabu SK, Hardy M, Chandran K, Lopez M, Kalyanaraman B, Avadhani NG (2009) Role of nuclear-encoded subunit Vb in the assembly and stability of cytochrome c oxidase complex: implications in mitochondrial dysfunction and ROS production. Biochem J 420: 439-49

Gattermann N, Retzlaff S, Wang YL, Hofhaus G, Heinisch J, Aul C, Schneider W (1997) Heteroplasmic point mutations of mitochondrial DNA affecting subunit I of cytochrome c oxidase in two patients with acquired idiopathic sideroblastic anemia. Blood 90: 4961- 72.

Gaur R, Grasso D, Datta PP, Krishna PD, Das G, Spencer A, Agrawal RK, Spremulli L, Varshney U (2008) A single mammalian mitochondrial translation initiation factor functionally replaces two bacterial factors. Mol Cell 29: 180-90

Ghezzi D, Saada A, D'Adamo P, Fernandez-Vizarra E, Gasparini P, Tiranti V, Elpeleg O, Zeviani M (2008) FASTKD2 nonsense mutation in an infantile mitochondrial encephalomyopathy associated with cytochrome c oxidase deficiency. Am J Hum Genet 83: 415-23

164

Glerum DM, Tzagoloff A (1997) Submitochondrial distributions and stabilities of subunits 4, 5, and 6 of yeast cytochrome oxidase in assembly defective mutants. FEBS Lett 412: 410-4.

Goto Y, Nonaka I, Horai S (1990) A mutation in the tRNALeu(UUR) gene associated with the MELAS subgroup of mitochondrial encephalomyopathies. Nature 348: 651-653

Green-Willms NS, Fox TD, Costanzo MC (1998) Functional interactions between yeast mitochondrial ribosomes and mRNA 5' untranslated leaders. Mol Cell Biol 18: 1826-34

Grossman LI, Lomax MI (1997) Nuclear genes for cytochrome c oxidase. Biochim Biophys Acta 1352: 174-92

Grossman LI, Shoubridge EA (1996) Mitochondrial genetics and human disease. Bioessays 18: 983-91.

Guex N, Peitsch MC (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18: 2714-23

Hammarsund M, Wilson W, Corcoran M, Merup M, Einhorn S, Grander D, Sangfelt O (2001) Identification and characterization of two novel human mitochondrial elongation factor genes, hEFG2 and hEFG1, phylogenetically conserved through evolution. Hum Genet 109: 542-50

Hanna MG, Nelson IP, Rahman S, Lane RJ, Land J, Heales S, Cooper MJ, Schapira AH, Morgan-Hughes JA, Wood NW (1998) Cytochrome c oxidase deficiency associated with the first stop-codon point mutation in human mtDNA. Am.J.Hum.Genet. 63: 29-36

Haque ME, Spremulli LL (2010) ICT1 comes to the rescue of mitochondrial ribosomes. Embo J 29: 1019-20

Hartl FU, Hayer-Hartl M (2002) Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 295: 1852-8

Hell K, Neupert W, Stuart RA (2001) Oxa1p acts as a general membrane insertion machinery for proteins encoded by mitochondrial DNA. Embo J 20: 1281-8

Herrmann JM (2003) Converting bacteria to organelles:evolution of mitochondrial protein sorting. Trends Microbiol 11: 74-79

Hibbs MA, Hess DC, Myers CL, Huttenhower C, Li K, Troyanskaya OG (2007) Exploring the functional landscape of gene expression: directed search of large microarray compendia. Bioinformatics 23: 2692-9

165

Holt IJ, Lorimer HE, Jacobs HT (2000) Coupled leading- and lagging-strand synthesis of mammalian mitochondrial DNA. Cell 100: 515-24

Horvath R, Schoser BG, Muller-Hocker J, Volpel M, Jaksch M, Lochmuller H (2005) Mutations in mtDNA-encoded cytochrome c oxidase subunit genes causing isolated myopathy or severe encephalomyopathy. Neuromuscul Disord 15: 851-7

Hosler JP (2004) The influence of subunit III of cytochrome c oxidase on the D pathway, the proton exit pathway and mechanism-based inactivation in subunit I. Biochim Biophys Acta 1655: 332-9

Huigsloot M, Nijtmans LG, Szklarczyk R, Baars MJ, van den Brand MA, Hendriksfranssen MG, van den Heuvel LP, Smeitink JA, Huynen MA, Rodenburg RJ (2011) A mutation in C2orf64 causes impaired cytochrome c oxidase assembly and mitochondrial cardiomyopathy. Am J Hum Genet 88: 488-93

Jaksch M, Ogilvie I, Yao J, Kortenhaus G, Bresser HG, Gerbitz KD, Shoubridge EA (2000) Mutations in SCO2 are associated with a distinct form of hypertrophic cardiomyopathy and cytochrome c oxidase deficiency. Hum Mol Genet 9: 795-801.

Kadenbach B, Jarausch J, Hartmann R, Merle P (1983) Separation of Mammalian Cytochrome c Oxidase into 13 Polypeptides by a Sodium Dodecyl Sulfate-Gel Electrophoretic Procedure. Anal.Biochem. 129: 517-521

Karadimas CL, Greenstein P, Sue CM, Joseph JT, Tanji K, Haller RG, Taivassalo T, Davidson MM, Shanske S, Bonilla E, DiMauro S (2000) Recurrent myoglobinuria due to a nonsense mutation in the COX I gene of mitochondrial DNA. Neurology 55: 644-9

Kaufman BA, Durisic N, Mativetsky JM, Costantino S, Hancock MA, Grutter P, Shoubridge EA (2007) The mitochondrial transcription factor TFAM coordinates the assembly of multiple DNA molecules into nucleoid-like structures. Mol Biol Cell 18: 3225-36

Keightley JA, Hoffbuhr KC, Burton MD, Salas VM, Johnston WS, Penn AM, Buist NR, Kennaway NG (1996) A microdeletion in cytochrome c oxidase (COX) subunit III associated with COX deficiency and recurrent myoglobinuria. Nat.Genet. 12: 410-6

Killary AM, Fournier RE (1995) Microcell fusion. Methods Enzymol 254: 133-52

Klement P, Nijtmans LG, Van den Bogert C, Houstek J (1995) Analysis of oxidative phosphorylation complexes in cultured human fibroblasts and amniocytes by blue-native- electrophoresis using mitoplasts isolated with the help of digitonin. Anal Biochem 231: 218-24.

166

Koc EC, Burkhart W, Blackburn K, Moyer MB, Schlatzer DM, Moseley A, Spremulli LL (2001) The large subunit of the mammalian mitochondrial ribosome. Analysis of the complement of ribosomal proteins present. J Biol Chem 276: 43958-69

Koc EC, Spremulli LL (2002) Identification of mammalian mitochondrial translational initiation factor 3 and examination of its role in initiation complex formation with natural mRNAs. J Biol Chem 277: 35541-9

Kren BT, Wong PY, Sarver A, Zhang X, Zeng Y, Steer CJ (2009) MicroRNAs identified in highly purified liver-derived mitochondria may play a role in apoptosis. RNA Biol 6: 65-72

Leary SC, Cobine PA, Kaufman BA, Guercin GH, Mattman A, Palaty J, Lockitch G, Winge DR, Rustin P, Horvath R, Shoubridge EA (2007) The human cytochrome c oxidase assembly factors SCO1 and SCO2 have regulatory roles in the maintenance of cellular copper homeostasis. Cell Metab 5: 9-20

Leary SC, Hill BC, Lyons CN, Carlson CG, Michaud D, Kraft CS, Ko K, Glerum DM, Moyes CD (2002) Chronic treatment with azide in situ leads to an irreversible loss of cytochrome c oxidase activity via holoenzyme dissociation. J Biol Chem 277: 11321-8

Leary SC, Kaufman BA, Pellecchia G, Guercin GH, Mattman A, Jaksch M, Shoubridge EA (2004) Human SCO1 and SCO2 have independent, cooperative functions in copper delivery to cytochrome c oxidase. Hum Mol Genet 13: 1839-48

Leary SC, Sasarman F, Nishimura T, Shoubridge EA (2009) Human SCO2 is required for the synthesis of CO II and as a thiol-disulphide oxidoreductase for SCO1. Hum Mol Genet 18: 2230-40

Legros F, Malka F, Frachon P, Lombes A, Rojo M (2004) Organization and dynamics of human mitochondrial DNA. J Cell Sci 117: 2653-62

Li Y, Park JS, Deng JH, Bai Y (2006) Cytochrome c oxidase subunit IV is essential for assembly and respiratory function of the enzyme complex. J Bioenerg Biomembr 38: 283-91

Liang H, Li L, Dong Z, Surette MG, Duan K (2008) The YebC family protein PA0964 negatively regulates the Pseudomonas aeruginosa quinolone signal system and pyocyanin production. J Bacteriol 190: 6217-27

Liang S, Hitomi M, Hu YH, Liu Y, Tartakoff AM (1996) A DEAD-box-family protein is required for nucleocytoplasmic transport of yeast mRNA. Mol Cell Biol 16: 5139-46

167

Liao HX, Spremulli LL (1991) Initiation of protein synthesis in animal mitochondria. Purification and characterization of translational initiation factor 2. J Biol Chem 266: 20714-9

Linder T, Park CB, Asin-Cayuela J, Pellegrini M, Larsson NG, Falkenberg M, Samuelsson T, Gustafsson CM (2005) A family of putative transcription termination factors shared amongst metazoans and plants. Curr Genet 48: 265-9

Ling M, Merante F, Chen HS, Duff C, Duncan AM, Robinson BH (1997) The human mitochondrial elongation factor tu (EF-Tu) gene: cDNA sequence, genomic localization, genomic structure, and identification of a pseudogene. Gene 197: 325-36

Litonin D, Sologub M, Shi Y, Savkina M, Anikin M, Falkenberg M, Gustafsson CM, Temiakov D (2010) Human mitochondrial transcription revisited: only TFAM and TFB2M are required for transcription of the mitochondrial genes in vitro. J Biol Chem 285: 18129-33

Liu L, McKeehan WL (2002) Sequence analysis of LRPPRC and its SEC1 domain interaction partners suggests roles in cytoskeletal organization, vesicular trafficking, nucleocytosolic shuttling, and chromosome activity. Genomics 79: 124-36

Liu M, Spremulli L (2000) Interaction of mammalian mitochondrial ribosomes with the inner membrane. J Biol Chem 275: 29400-6

Lochmuller H, Johns T, Shoubridge EA (1999) Expression of the E6 and E7 genes of human papillomavirus (HPV16) extends the life span of human myoblasts. Exp Cell Res 248: 186-93.

Lurin C, Andres C, Aubourg S, Bellaoui M, Bitton F, Bruyere C, Caboche M, Debast C, Gualberto J, Hoffmann B, Lecharny A, Le Ret M, Martin-Magniette ML, Mireau H, Peeters N, Renou JP, Szurek B, Taconnat L, Small I (2004) Genome-wide analysis of Arabidopsis pentatricopeptide repeat proteins reveals their essential role in organelle biogenesis. Plant Cell 16: 2089-103

Madsen CS, Ghivizzani SC, Hauswirth WW (1993) Protein binding to a single termination-associated sequence in the mitochondrial DNA D-loop region. Mol Cell Biol 13: 2162-71

Magalhaes PJ, Andreu AL, Schon EA (1998) Evidence for the presence of 5S rRNA in mammalian mitochondria. Mol Biol Cell 9: 2375-82

Manfredi G, Schon EA, Moraes CT, Bonilla E, Berry GT, Sladky JT, DiMauro S (1995) A new mutation associated with MELAS is located in a mitochondrial DNA polypeptide- coding gene. .Neuromuscul.Disord. 5: 391-398

168

Manthey GM, McEwen JE (1995) The product of the nuclear gene PET309 is required for translation of mature mRNA and stability or production of intron- containing RNAs derived from the mitochondrial COX1 locus of Saccharomyces cerevisiae. EMBO J. 14: 4031-4043

Massa V, Fernandez-Vizarra E, Alshahwan S, Bakhsh E, Goffrini P, Ferrero I, Mereghetti P, D'Adamo P, Gasparini P, Zeviani M (2008) Severe infantile encephalomyopathy caused by a mutation in COX6B1, a nucleus-encoded subunit of cytochrome c oxidase. Am J Hum Genet 82: 1281-9

McCulloch V, Shadel GS (2003) Human mitochondrial transcription factor B1 interacts with the C-terminal activation region of h-mtTFA and stimulates transcription independently of its RNA methyltransferase activity. Mol Cell Biol 23: 5816-24

Metodiev MD, Lesko N, Park CB, Camara Y, Shi Y, Wibom R, Hultenby K, Gustafsson CM, Larsson NG (2009) Methylation of 12S rRNA is necessary for in vivo stability of the small subunit of the mammalian mitochondrial ribosome. Cell Metab 9: 386-97

Mick DU, Fox TD, Rehling P (2011) Inventory control: cytochrome c oxidase assembly regulates mitochondrial translation. Nat Rev Mol Cell Biol 12: 14-20

Mick DU, Vukotic M, Piechura H, Meyer HE, Warscheid B, Deckers M, Rehling P (2010) Coa3 and Cox14 are essential for negative feedback regulation of COX1 translation in mitochondria. J Cell Biol 191: 141-54

Mick DU, Wagner K, van der Laan M, Frazier AE, Perschil I, Pawlas M, Meyer HE, Warscheid B, Rehling P (2007) Shy1 couples Cox1 translational regulation to cytochrome c oxidase assembly. Embo J 26: 4347-58

Mikelsaar R (1983) Human mitochondrial genome and the evolution of methionine transfer ribonucleic acids. J Theor Biol 105: 221-32

Miller C, Saada A, Shaul N, Shabtai N, Ben-Shalom E, Shaag A, Hershkovitz E, Elpeleg O (2004) Defective mitochondrial translation caused by a ribosomal protein (MRPS16) mutation. Ann Neurol 56: 734-8

Montoya J, Ojala D, Attardi G (1981) Distinctive features of the 5'-terminal sequences of the human mitochondrial mRNAs. Nature 290: 465-70

Mootha VK, Lepage P, Miller K, Bunkenborg J, Reich M, Hjerrild M, Delmonte T, Villeneuve A, Sladek R, Xu F, Mitchell GA, Morin C, Mann M, Hudson TJ, Robinson B, Rioux JD, Lander ES (2003) Identification of a gene causing human cytochrome c oxidase deficiency by integrative genomics. Proc Natl Acad Sci U S A 100: 605-10.

169

Naithani S, Saracco SA, Butler CA, Fox TD (2003) Interactions among COX1, COX2, and COX3 mRNA-specific translational activator proteins on the inner surface of the mitochondrial inner membrane of Saccharomyces cerevisiae. Mol Biol Cell 14: 324-33

Newmeyer DDaF-M, S. (2003) Mitochondria:releasing power for life and unleashing the machineries of death. Cell 112: 481-490

Nijtmans LG, Spelbrink JN, Van Galen MJ, Zwaan M, Klement P, Van den Bogert C (1995) Expression and fate of the nuclearly encoded subunits of cytochrome-c oxidase in cultured human cells depleted of mitochondrial gene products. Biochim.Biopohys.Acta 1265: 117-26

Nijtmans LG, Taanman JW, Muijsers AO, Speijer D, Van den Bogert C (1998) Assembly of cytochrome-c oxidase in cultured human cells. Eur J Biochem 254: 389-94.

Nobrega MP, Bandeira SC, Beers J, Tzagoloff A (2002) Characterization of COX19, a widely distributed gene required for expression of mitochondrial cytochrome oxidase. J Biol Chem 277: 40206-11.

Nolden M, Ehses S, Koppen M, Bernacchia A, Rugarli EI, Langer T (2005) The m-AAA protease defective in hereditary spastic paraplegia controls ribosome assembly in mitochondria. Cell 123: 277-89

Nozaki Y, Matsunaga N, Ishizawa T, Ueda T, Takeuchi N (2008) HMRF1L is a human mitochondrial translation release factor involved in the decoding of the termination codons UAA and UAG. Genes Cells 13: 429-38

O'Brien TW (2002) Evolution of a protein-rich mitochondrial ribosome: implications for human genetic disease. Gene 286: 73-9

Ojala D, Montoya J, Attardi G (1981) tRna punctuation model of Rna processing in human mitochondria. Nature 290: 470-4

Ott M, Prestele M, Bauerschmitt H, Funes S, Bonnefoy N, Herrmann JM (2006) Mba1, a membrane-associated ribosome receptor in mitochondria. Embo J 25: 1603-10

Pagliarini DJ, Calvo SE, Chang B, Sheth SA, Vafai SB, Ong SE, Walford GA, Sugiana C, Boneh A, Chen WK, Hill DE, Vidal M, Evans JG, Thorburn DR, Carr SA, Mootha VK (2008) A mitochondrial protein compendium elucidates complex I disease biology. Cell 134: 112-23

Papadopoulou LC, Sue CM, Davidson MM, Tanji K, Nishino I, Sadlock JE, Krishna S, Walker W, Selby J, Glerum DM, Coster RV, Lyon G, Scalais E, Lebel R, Kaplan P, Shanske S, De Vivo DC, Bonilla E, Hirano M, DiMauro S, Schon EA (1999) Fatal

170

infantile cardioencephalomyopathy with COX deficiency and mutations in SCO2, a COX assembly gene. Nat Genet 23: 333-7.

Park CB, Asin-Cayuela J, Camara Y, Shi Y, Pellegrini M, Gaspari M, Wibom R, Hultenby K, Erdjument-Bromage H, Tempst P, Falkenberg M, Gustafsson CM, Larsson NG (2007) MTERF3 is a negative regulator of mammalian mtDNA transcription. Cell 130: 273-85

Perez-Martinez X, Broadley SA, Fox TD (2003) Mss51p promotes mitochondrial Cox1p synthesis and interacts with newly synthesized Cox1p. Embo J 22: 5951-61

Petry S, Weixlbaumer A, Ramakrishnan V (2008) The termination of translation. Curr Opin Struct Biol 18: 70-7

Pham XH, Farge G, Shi Y, Gaspari M, Gustafsson CM, Falkenberg M (2006) Conserved sequence box II directs transcription termination and primer formation in mitochondria. J Biol Chem 281: 24647-52

Pierce SB, Chisholm KM, Lynch ED, Lee MK, Walsh T, Opitz JM, Li W, Klevit RE, King MC (2011) Mutations in mitochondrial histidyl tRNA synthetase HARS2 cause ovarian dysgenesis and sensorineural hearing loss of Perrault syndrome. Proc Natl Acad Sci U S A 108: 6543-8

Pierrel F, Bestwick ML, Cobine PA, Khalimonchuk O, Cricco JA, Winge DR (2007) Coa1 links the Mss51 post-translational function to Cox1 cofactor insertion in cytochrome c oxidase assembly. Embo J 26: 4335-46

Prezant TR, Agapian JV, Bohlman MC, Bu X, Oztas S, Qiu WQ, Arnos KS, Cortopassi GA, Jaber L, Rotter JI, al e (1993) Mitochondrial ribosomal RNA mutation associated with both antibiotic-induced and non-syndromic deafness. Nat.Genet. 4: 289-294

Rackham O, Davies SM, Shearwood AM, Hamilton KL, Whelan J, Filipovska A (2009) Pentatricopeptide repeat domain protein 1 lowers the levels of mitochondrial leucine tRNAs in cells. Nucleic Acids Res 37: 5859-67

Rahman S, Taanman JW, Cooper JM, Nelson I, Hargreaves I, Meunier B, Hanna MG, Garcia JJ, Capaldi RA, Lake BD, Leonard JV, Schapira AH (1999) A missense mutation of cytochrome oxidase subunit II causes defective assembly and myopathy. Am J Hum Genet 65: 1030-9

Richter R, Rorbach J, Pajak A, Smith PM, Wessels HJ, Huynen MA, Smeitink JA, Lightowlers RN, Chrzanowska-Lightowlers ZM (2010) A functional peptidyl-tRNA hydrolase, ICT1, has been recruited into the human mitochondrial ribosome. Embo J 29: 1116-25

171

Riley LG, Cooper S, Hickey P, Rudinger-Thirion J, McKenzie M, Compton A, Lim SC, Thorburn D, Ryan MT, Giege R, Bahlo M, Christodoulou J (2010) Mutation of the mitochondrial tyrosyl-tRNA synthetase gene, YARS2, causes myopathy, lactic acidosis, and sideroblastic anemia--MLASA syndrome. Am J Hum Genet 87: 52-9

Ringel R, Sologub M, Morozov YI, Litonin D, Cramer P, Temiakov D (2011) Structure of human mitochondrial RNA polymerase. Nature

Robinson BH (2000) Human cytochrome oxidase deficiency. Pediatr Res 48: 581-5.

Rorbach J, Richter R, Wessels HJ, Wydro M, Pekalski M, Farhoud M, Kuhl I, Gaisne M, Bonnefoy N, Smeitink JA, Lightowlers RN, Chrzanowska-Lightowlers ZM (2008) The human mitochondrial ribosome recycling factor is essential for cell viability. Nucleic Acids Res 36: 5787-99

Rotig A (2011) Human diseases with impaired mitochondrial protein synthesis. Biochim Biophys Acta 1807: 1198-205

Saada A, Shaag A, Arnon S, Dolfin T, Miller C, Fuchs-Telem D, Lombes A, Elpeleg O (2007) Antenatal mitochondrial disease caused by mitochondrial ribosomal protein (MRPS22) mutation. J Med Genet 44: 784-6

Sacconi S, Trevisson E, Pistollato F, Baldoin MC, Rezzonico R, Bourget I, Desnuelle C, Tenconi R, Basso G, DiMauro S, Salviati L (2005) hCOX18 and hCOX19: two human genes involved in cytochrome c oxidase assembly. Biochem Biophys Res Commun 337: 832-9

Salvatore Dimauro MD, and Eric A. Schon,Ph.D. (2003) Mitochondrial Respiratory- Chain Diseases. The New England Journal of Medicine 348: 2656-2668

Saracco SA, Fox TD (2002) Cox18p is required for export of the mitochondrially encoded Saccharomyces cerevisiae Cox2p C-tail and interacts with Pnt1p and Mss2p in the inner membrane. Mol Biol Cell 13: 1122-31

Saraste M (1990) Structural features of cytochrome oxidase. . Q Rev Biophys 23: 331- 366

Sasarman F, Brunel-Guitton C, Antonicka H, Wai T, Shoubridge EA (2010) LRPPRC and SLIRP interact in a ribonucleoprotein complex that regulates posttranscriptional gene expression in mitochondria. Mol Biol Cell 21: 1315-23

Satoh M, Kuroiwa T (1991) Organization of multiple nucleoids and DNA molecules in mitochondria of a human cell. Exp Cell Res 196: 137-40

172

Scaglia F, Wong LJ (2008) Human mitochondrial transfer RNAs: role of pathogenic mutation in disease. Muscle Nerve 37: 150-71

Scarpulla RC (2008) Transcriptional paradigms in mammalian mitochondrial biogenesis and function. Physiol Rev 88: 611-38

Schagger H, Pfeiffer K (2000) Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. Embo J 19: 1777-83.

Schatz G (1995) Mitochondria: beyond oxidative phosphorylation. Biochim.Biophys.Acta 1271: 123-126

Scheper GC, van der Klok T, van Andel RJ, van Berkel CG, Sissler M, Smet J, Muravina TI, Serkov SV, Uziel G, Bugiani M, Schiffmann R, Krageloh-Mann I, Smeitink JA, Florentz C, Van Coster R, Pronk JC, van der Knaap MS (2007) Mitochondrial aspartyl- tRNA synthetase deficiency causes leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation. Nat Genet 39: 534-9

Seeger J, Schrank B, Pyle A, Stucka R, Lorcher U, Muller-Ziermann S, Abicht A, Czermin B, Holinski-Feder E, Lochmuller H, Horvath R (2010) Clinical and neuropathological findings in patients with TACO1 mutations. Neuromuscul Disord 20: 720-4

Seidel-Rogol BL, McCulloch V, Shadel GS (2003) Human mitochondrial transcription factor B1 methylates ribosomal RNA at a conserved stem-loop. Nat Genet 33: 23-4

Shadel GS, Clayton DA (1997) Mitochondrial DNA maintenance in vertebrates. Annu Rev Biochem 66: 409-35

Shin DH, Yokota H, Kim R, Kim SH (2002) Crystal structure of conserved hypothetical protein Aq1575 from Aquifex aeolicus. Proc Natl Acad Sci U S A 99: 7980-5

Shoffner JM, Lott MT, Lezza AM, Seibel P, Ballinger SW, Wallace DC (1990) Myoclonic epilepsy and ragged-red fiber disease (MERRF) is associated with a mitochondrial DNA tRNA(Lys) mutation. Cell 61: 931-7

Shoubridge EA (2001) Cytochrome c oxidase deficiency. Am J Med Genet 106: 46-52.

Shteyer E, Saada A, Shaag A, Al-Hijawi FA, Kidess R, Revel-Vilk S, Elpeleg O (2009) Exocrine pancreatic insufficiency, dyserythropoeitic anemia, and calvarial hyperostosis are caused by a mutation in the COX4I2 gene. Am J Hum Genet 84: 412-7

Smeitink J, van den Heuvel L, DiMauro S (2001) The genetics and pathology of oxidative phosphorylation. Nat Rev Genet 2: 342-52

173

Smeitink JA, Elpeleg O, Antonicka H, Diepstra H, Saada A, Smits P, Sasarman F, Vriend G, Jacob-Hirsch J, Shaag A, Rechavi G, Welling B, Horst J, Rodenburg RJ, van den Heuvel B, Shoubridge EA (2006) Distinct clinical phenotypes associated with a mutation in the mitochondrial translation elongation factor EFTs. Am J Hum Genet 79: 869-77

Smirnov AV, Entelis NS, Krasheninnikov IA, Martin R, Tarassov IA (2008) Specific features of 5S rRNA structure - its interactions with macromolecules and possible functions. Biochemistry (Mosc) 73: 1418-37

Smith D, Gray J, Mitchell L, Antholine WE, Hosler JP (2005) Assembly of cytochrome-c oxidase in the absence of assembly protein Surf1p leads to loss of the active site heme. J Biol Chem 280: 17652-6

Smits P, Antonicka H, van Hasselt PM, Weraarpachai W, Haller W, Schreurs M, Venselaar H, Rodenburg RJ, Smeitink JA, van den Heuvel LP (2011a) Mutation in subdomain G' of mitochondrial elongation factor G1 is associated with combined OXPHOS deficiency in fibroblasts but not in muscle. Eur J Hum Genet 19: 275-9

Smits P, Saada A, Wortmann SB, Heister AJ, Brink M, Pfundt R, Miller C, Haas D, Hantschmann R, Rodenburg RJ, Smeitink JA, van den Heuvel LP (2011b) Mutation in mitochondrial ribosomal protein MRPS22 leads to Cornelia de Lange-like phenotype, brain abnormalities and hypertrophic cardiomyopathy. Eur J Hum Genet 19: 394-9

Soleimanpour-Lichaei HR, Kuhl I, Gaisne M, Passos JF, Wydro M, Rorbach J, Temperley R, Bonnefoy N, Tate W, Lightowlers R, Chrzanowska-Lightowlers Z (2007) mtRF1a is a human mitochondrial translation release factor decoding the major termination codons UAA and UAG. Mol Cell 27: 745-57

Sologub M, Litonin D, Anikin M, Mustaev A, Temiakov D (2009) TFB2 is a transient component of the catalytic site of the human mitochondrial RNA polymerase. Cell 139: 934-44

Souza RL, Green-Willms NS, Fox TD, Tzagoloff A, Nobrega FG (2000) Cloning and characterization of COX18, a Saccharomyces cerevisiae PET gene required for the assembly of cytochrome oxidase. J Biol Chem 275: 14898-902.

Spremulli LL, Coursey A, Navratil T, Hunter SE (2004) Initiation and elongation factors in mammalian mitochondrial protein biosynthesis. Prog Nucleic Acid Res Mol Biol 77: 211-61

Stiburek L, Hansikova H, Tesarova M, Cerna L, Zeman J (2006) Biogenesis of eukaryotic cytochrome c oxidase. Physiol Res 55 Suppl 2: S27-41

174

Stiburek L, Vesela K, Hansikova H, Hulkova H, Zeman J (2009) Loss of function of Sco1 and its interaction with cytochrome c oxidase. Am J Physiol Cell Physiol 296: C1218-26

Stiburek L, Vesela K, Hansikova H, Pecina P, Tesarova M, Cerna L, Houstek J, Zeman J (2005) Tissue-specific cytochrome c oxidase assembly defects due to mutations in SCO2 and SURF1. Biochem J 392: 625-32

Stryer L (1995) Biochemistry, 4th edn. W.H. Freeman, New York

Suzuki H, Ueda T, Taguchi H, Takeuchi N (2007) Chaperone properties of mammalian mitochondrial translation elongation factor Tu. J Biol Chem 282: 4076-84

Taanman JW (1997) Human cytochrome c oxidase: structure, function, and deficiency. J.Bioenerg.Biomembr. 29: 151-63

Taanman JW (1999) The mitochondrial genome: structure, transcription, translation and replication. Biochim Biophys Acta 1410: 103-23

Taanman JW, Williams SL (2001) Assembly of cytochrome c oxidase: what can we learn from patients with cytochrome c oxidase deficiency? Biochem Soc Trans 29: 446-51

Tam EW, Feigenbaum A, Addis JB, Blaser S, Mackay N, Al-Dosary M, Taylor RW, Ackerley C, Cameron JM, Robinson BH (2008) A novel mitochondrial DNA mutation in COX1 leads to strokes, seizures, and lactic acidosis. Neuropediatrics 39: 328-34

Tay SK, Nesti C, Mancuso M, Schon EA, Shanske S, Bonilla E, Davidson MM, Dimauro S (2004) Studies of COX16, COX19, and PET191 in human cytochrome-c oxidase deficiency. Arch Neurol 61: 1935-7

Taylor RW, Turnbull DM (2005) Mitochondrial DNA mutations in human disease. Nat Rev Genet 6: 389-402

Temperley R, Richter R, Dennerlein S, Lightowlers RN, Chrzanowska-Lightowlers ZM (2010) Hungry codons promote frameshifting in human mitochondrial ribosomes. Science 327: 301

Tiranti V, Corona P, Greco M, Taanman JW, Carrara F, Lamantea E, Nijtmans L, Uziel G, Zeviani M (2000) A novel frameshift mutation of the mtDNA COIII gene leads to impaired assembly of cytochrome c oxidase in a patient affected by Leigh-like syndrome. Hum Mol Genet 9: 2733-42.

Tiranti V, Hoertnagel K, Carrozzo R, Galimberti C, Munaro M, Granatiero M, Zelante L, Gasparini P, Marzella R, Rocchi M, Bayona-Bafaluy MP, Enriquez JA, Uziel G, Bertini E, Dionisi-Vici C, Franco B, Meitinger T, Zeviani M (1998a) Mutations of SURF-1 in 175

Leigh disease associated with cytochrome c oxidase deficiency. Am J Hum Genet 63: 1609-21

Tiranti V, L DA, Pareyson D, Mora M, Carrara F, Zelante L, Gasparini P, Zeviani M (1998b) A novel mutation in the mitochondrial tRNA(Val) gene associated with a complex neurological presentation. Ann Neurol 43: 98-101

Tsuboi M, Morita H, Nozaki Y, Akama K, Ueda T, Ito K, Nierhaus KH, Takeuchi N (2009) EF-G2mt is an exclusive recycling factor in mammalian mitochondrial protein synthesis. Mol Cell 35: 502-10

Tsukihara T, Aoyama H, Yamashita E, Tomizaki T, Yamaguchi H, Shinzawa-Itoh K, Nakashima R, Yaono R, Yoshikawa S (1995) Structures of metal sites of oxidized bovine heart cytochrome c oxidase at 2.8 A [see comments]. Science 269: 1069-74

Tsukihara T, Aoyama H, Yamashita E, Tomizaki T, Yamaguchi H, Shinzawa-Itoh K, Nakashima R, Yaono R, Yoshikawa S (1996) The whole structure of the 13-subunit oxidized cytochrome c oxidase at 2.8 A. Science 272: 1136-44.

Tzagoloff A, Nobrega M, Gorman N, Sinclair P (1993) On the functions of the yeast COX10 and COX11 gene products. Biochem Mol Biol Int 31: 593-8.

Uchiumi T, Ohgaki K, Yagi M, Aoki Y, Sakai A, Matsumoto S, Kang D (2010) ERAL1 is associated with mitochondrial ribosome and elimination of ERAL1 leads to mitochondrial dysfunction and growth retardation. Nucleic Acids Res 38: 5554-68

Valente L, Tiranti V, Marsano RM, Malfatti E, Fernandez-Vizarra E, Donnini C, Mereghetti P, De Gioia L, Burlina A, Castellan C, Comi GP, Savasta S, Ferrero I, Zeviani M (2007) Infantile encephalopathy and defective mitochondrial DNA translation in patients with mutations of mitochondrial elongation factors EFG1 and EFTu. Am J Hum Genet 80: 44-58

Valnot I, Osmond S, Gigarel N, Mehaye B, Amiel J, Cormier-Daire V, Munnich A, Bonnefont JP, Rustin P, Rotig A (2000a) Mutations of the SCO1 gene in mitochondrial cytochrome c oxidase deficiency with neonatal-onset hepatic failure and encephalopathy. Am J Hum Genet 67: 1104-9.

Valnot I, von Kleist-Retzow JC, Barrientos A, Gorbatyuk M, Taanman JW, Mehaye B, Rustin P, Tzagoloff A, Munnich A, Rotig A (2000b) A mutation in the human heme A:farnesyltransferase gene (COX10 ) causes cytochrome c oxidase deficiency. Hum Mol Genet 9: 1245-9. van den Heuvel L, Smeitink J (2001) The oxidative phosphorylation (OXPHOS) system: nuclear genes and human genetic diseases. Bioessays 23: 518-25.

176

Voet D, Voet JG (2004) Biochemistry, 3rd edn. Wiley, Hoboken, N.J.

Walberg MW, Clayton DA (1983) In vitro transcription of human mitochondrial DNA. Identification of specific light strand transcripts from the displacement loop region. J Biol Chem 258: 1268-75

Wang G, Chen HW, Oktay Y, Zhang J, Allen EL, Smith GM, Fan KC, Hong JS, French SW, McCaffery JM, Lightowlers RN, Morse HC, 3rd, Koehler CM, Teitell MA (2010) PNPASE regulates RNA import into mitochondria. Cell 142: 456-67

Wenz T, Luca C, Torraco A, Moraes CT (2009) mTERF2 regulates oxidative phosphorylation by modulating mtDNA transcription. Cell Metab 9: 499-511

Weraarpachai W, Antonicka H, Sasarman F, Seeger J, Schrank B, Kolesar JE, Lochmuller H, Chevrette M, Kaufman BA, Horvath R, Shoubridge EA (2009) Mutation in TACO1, encoding a translational activator of COX I, results in cytochrome c oxidase deficiency and late-onset Leigh syndrome. Nat Genet 41: 833-7

Wielburski A, Nelson BD (1983) Evidence for the sequential assembly of cytochrome oxidase subunits in rat liver mitochondria. Biochem J 212: 829-34

Wong TW, Clayton DA (1985) In vitro replication of human mitochondrial DNA: accurate initiation at the origin of light-strand synthesis. Cell 42: 951-8

Wu S, Skolnick J, Zhang Y (2007) Ab initio modeling of small proteins by iterative TASSER simulations. BMC Biol 5: 17

Xin H, Woriax V, Burkhart W, Spremulli LL (1995) Cloning and expression of mitochondrial translational elongation factor Ts from bovine and human liver. J Biol Chem 270: 17243-9

Xu-B, Clayton-DA (1995) A persistent RNA-DNA hybrid is formed during transcription at a phylogenetically conserved mitochondrial DNA sequence. Mol.Cell Biol. 15: 580- 589

Xu F, Morin C, Mitchell G, Ackerley C, Robinson BH (2004) The role of the LRPPRC (leucine-rich pentatricopeptide repeat cassette) gene in cytochrome oxidase assembly: mutation causes lowered levels of COX (cytochrome c oxidase) I and COX III mRNA. Biochem J 382: 331-6

Yakubovskaya E, Chen Z, Carrodeguas JA, Kisker C, Bogenhagen DF (2006) Functional human mitochondrial DNA polymerase gamma forms a heterotrimer. J Biol Chem 281: 374-82

177

Yao J, Shoubridge EA (1999) Expression and functional analysis of SURF1 in Leigh syndrome patients with cytochrome c oxidase deficiency. Hum Mol Genet 8: 2541-9.

Ylikallio E, Suomalainen A (2011) Mechanisms of mitochondrial diseases. Ann Med

Zambrano A, Fontanesi F, Solans A, de Oliveira RL, Fox TD, Tzagoloff A, Barrientos A (2007) Aberrant translation of cytochrome c oxidase subunit 1 mRNA species in the absence of Mss51p in the yeast Saccharomyces cerevisiae. Mol Biol Cell 18: 523-35

Zeharia A, Shaag A, Pappo O, Mager-Heckel AM, Saada A, Beinat M, Karicheva O, Mandel H, Ofek N, Segel R, Marom D, Rotig A, Tarassov I, Elpeleg O (2009) Acute infantile liver failure due to mutations in the TRMU gene. Am J Hum Genet 85: 401-7

Zhang C, Baumer A, Maxwell RJ, Linnane AW, Nagley P (1992) Multiple mitochondrial DNA deletions in an elderly human individual. FEBS Lett.. 297: 34-38

Zhang Y (2007) Template-based modeling and free modeling by I-TASSER in CASP7. Proteins 69 Suppl 8: 108-17

Zhang Y (2008) I-TASSER server for protein 3D structure prediction. BMC Bioinformatics 9: 40

Zhang Y, Spremulli LL (1998) Identification and cloning of human mitochondrial translational release factor 1 and the ribosome recycling factor. Biochim Biophys Acta 1443: 245-50

Zhu Z, Yao J, Johns T, Fu K, De Bie I, Macmillan C, Cuthbert AP, Newbold RF, Wang J, Chevrette M, Brown GK, Brown RM, Shoubridge EA (1998) SURF1, encoding a factor involved in the biogenesis of cytochrome c oxidase, is mutated in Leigh syndrome. Nat Genet 20: 337-43.

Zickermann V, Angerer H, Ding MG, Nubel E, Brandt U (2010) Small single transmembrane domain (STMD) proteins organize the hydrophobic subunits of large membrane protein complexes. FEBS Lett 584: 2516-25

178

Appendix

179

180

181

182