The Molecular and Biochernical Characterization of the MLRQ Subunit of NADH:Ubiquinone Oxidoreductase in the

Human Mitochondrial Respiratory Chain

Dhush y Kanagarajah

A Thesis submitted in codormity with the requirements for the degree of Master of Science

Graduate Department of Biochemistry University of Toronto

" Copyright by Dhushy Kanagarajah. 200 1

National Library I*l of Canada Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellington Street 395. rue Wellington Ottawa ON K1A O N 4 ûttawa ON K1 A O N 4 Canada Canada

The author has granted a non- L'auteur a accordé une licence non exclusive licence ailowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distriie or seil reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/fh, de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or othenirise de celle-ci ne doivent être imprimés reproduced without the author' s ou autrement reproduits sans son permission. autorisation.

The Molecular and Biochemical Characterization of the MLRQ Subunit of NADH:Ubiquinone Oxidoreductase in the

Humsn Mitochondrîal Respiratory Chain

Master of Science, 200 1 Dhushy Kanagarajah

Department of Biochemistry University of Toronto

Abstract

Isolated deficiency of NADH:ubiquinone oxidoreductase (Complex 1), the first enzyme

of the mitochondrial respiratory chah is the most comrnon cause of human

mitochondriocytopathies. In order to characterize the nuclear genes contributing to this

disease, the cDNA and genomic sequences encoding the MLRQ subunit of complex I

were determined. The NDUF.44 gene encoding MLRQ was localized to chromosome 7

p2 1-22 and a pseudogene was found on chromosome 1 p2 1. Tissue specific expression of

MLRQ at both mRNA and protein levels was examined. Overexpression of this subunit

in a patient exhibiting complex 1 deficiency is dso discussed. Extraction.

immunoprecipitation and cross-linking studies revealed that while the N-teminus of

MLRQ has a great afinity for phospholipids of the inner mitochondriai membrane. it

likely also associates with the MWFE subunit through other intemediary subunit(s).

forming part of the bulky staik region that bridges the two arms of complex 1.

Acknowledgements

First and foremost, 1 wish to express my heartfelt thanks and gratitude to Dr. Brian

Robinson for his guidance and support during the course of this degree. 1 tnily feel

privileged to have had the opportunity of working with such a great supervisor. 1 am also

gratefùl to my CO-supervisors, Dr. B. Sarkar and Dr. R. Baker for their t h e and

assistance with this project.

Behind every graduate student's thesis lies a support system so precious that failing to

acknowledge it would be unpardonable. 1 count myself lucky to have had the privilege of

working with al1 my colleagues @ast and present) at the Robinson lab, but I must

especially thank Agnieszka, Jessie, Maryanna, Nevi, So-Young and Tomoko. Their

kindness and fiiendship will never be forgotten. 1 am forever indebted to Dr. Sandy Raha

for al1 his expertise, encouragement and humour without which 1 would never have

survived. A very special thanks goes to Maureen Waite for her friendship and for always

making the time to lend me a hand.

This thesis would not have been completed without the extraordinary love and support

of each and every person whom 1 cal1 family. My most ardent supporters. they have been

with me throughout the trials and tribulations of graduate life. For this, 1 would like to

thank my uncle Milroy, my brother Dhilip and especially my husband Ramana for his

patience and understanding during these past few years. Finally but most importantly. 1

would like to express my gratitude to my parents for impressing upon me the principles

of hard work, perseverance and a firm belief in the merits of education. Their love and

prayers have been instrumental in the completion of this study.

I dedicate this thesis to the memory of my loved ones whose blessings and

encouragement stiil spur me on to greater accomplishments.

a-.

I I I

CONTRIBUTIONS TO THESIS

Screening of PACNAC libraries and FISH Mapping: The Toronto Centre For Applied Genomics

Tissue mitochondria: Dr. Sandeep Raha

Table of Contents

Abstract Acknowledgements Contributions to thesis Table of contents List of figures List of tables Abbreviations Glossary of Medical Terms

i v a..

Vtlt

Chapter 1 Introduction and Objectives The mitochondrion: structure, function, mode of inheritance and associated diseases Oventiew Part 1. Mitochondrial structure. function and inheritance

The mitoc hondrion Mitochondrial ultrastructure Mitochondrial DNA organization Mitochondrial replication. transcription and translation Mitochondrial protein import Energy metabolism

Part II. The mitochondrial respiratory chah complexes

Organization of the OXPHOS system The OXPHOS system: Role in electron transport and proton translocation Complex 1: The NADHxbiquinone oxidoreductase complex

Evolution of compler 1 Subunit composition of the NADH:ubiquinone oxidoreductase corn plex

(i) The flavoprotein fraction (FP) (ii) The iron-sulphur protein fraction (IP)

(iii) n i e hydrophobie protein fraction (HP) a) The rnitochondrially encoded subunits b) The nuclear encoded subunits

Structural mode1 of complex 1 (i) Assembly of Complex 1 (ii) Spatial Organization and Subunit Interaction in Human Complex 1 32

Energy conversion in cornplex 1 35 (i) Iron-sulphur clusters. flavin and semiquinones 35

(ii) Electron transfer in cornplex 1 37 (iii) Models for coupling electron flow with proton translocation 42

Complex 1 inhibitors 47 Complex II: The Succinate-ubiquinone oxidoreductase cornplex 48 Complex III: The Ubiquinol-femcytochrome c oxidoreductase complex 49 Complex IV: The cytochrome c oxidase complex 5 1 Cornplex V: The ATP synthase complex 52

Part M. Mitochondrial disorders 54

Typical symptoms of defects in energy metabolism Mitochondrial respiratory chain diseases

(i) MtDNA associated diseases (ii) Nuclear DNA associated diseases

(iii) Mitochondrial respiratory chain disorders associated with neurological diseases

Human Complex I deficiencies (i) MtDNA encoded defects in complex 1

(ii) Nuclear DNA encoded defects in complex 1 (iii) Free radical generation and complex 1 deficiency

Objectives and Rationale 62

Chapter 2 Cloning, molecular characterization and chromosomal localization of the MLRQ subunit of hurnan NADH:ubiquinone oridoreductase

Abstract Introduction Materiais and methods

Part 1. Molecular Characterization of MLRQ cDNA in various tissues. cells and patient cell lines

Tissue culture of cardiomyocytes and fibroblasts RNA isolation from tissues and cells cDNA synthesis and PCR cDNA cloning and sequencing of MLRQ Mutational screening of complex 1 deficient patients

Part II. Genomic characterization and Iocalization of the NDUE4-C (MLRQ) gene and pseudogene

Screening the PAC library Southem blot analysis of PAC clones Northem analysis of MLRQ expression Chromosomal localization Amplification. cloning and sequencing of MLRQ from genomic and PAC DNA Y AC library screening

Results and discussion Part 1. Isolation and characterization of MLRQ cDNA

MLRQ cDNA structure Mutational analysis of MLRQ cDNA in complex 1 deficient patients

Part II. Chromosomal localization and characterization of the iVDUE4-I gene and pseudogene

Library screening and FISH mapping MLRQ expression at the transcriptional level Amplification of the iVDUFA-I gene fiom genomic DNA Southem blot analysis Genomic organization of NDUFA-l The pseudogene on chromosome 1 Other MLRQ-like sequenccs in the genorne

Chapter 3 Biochemical characterization, protein expression and immunoprecipitation studies pertaining to 1MLRQ and related complex 1 subunits

Abstract Introduction Materials and methods

Part 1. MLRQ expression in hurnan tissues and cells Antibody generation Western blot analysis of MLRQ expression

Part II. Bacterial expression of MLRQ protein Design of MLRQ- fusion protein construc t Induction and puritkation of MLRQ-GST fusion protein Factor Xa cleavage of fusion protein

Part III. Anti-srnse expression of MLRQ in marnmalian celis Design of sense and anti-sense oriented pREP9 constructs Optimization of transfection conditions Transfection and selection with pREP9 Transfection and selection with the linearized vector pCDNA 3 . l + Western blot analysis of sense anti-sense expression in transfected cells Amplification of MLRQ From transfected cells

Part IV. Association of MLRQ with other complex 1 subunits Solubilization of beef heart mitochondria Immunoprecipitation of MLRQ, MWFE and 49 kD subunits Cross-linking with DST and EGS Imrnunoprecipitation of cross-linked bovine hem mitochondria SDS-PAGE and western blot analysis of MLRQ during extraction. immunoprecipitation and cross-linking

Results and discussion Part 1. Determining MLRQ fom. îûnction and expression

Tissue expression of MLRQ Bacterial expression of MLRQ: Attempts at defining subunit structure

vii

Antisense expression of MLRQ: Attempts to detemine subunit function 1 13 Part II. Subunit interactions of MLRQ within complex 1 117

Detergent solubilization of cornplex 1 subunits 117 Proximity and association of MLRQ with other complex I subunits 122 Cross-linking and immunoprecipitation studies 128

Chapier 1 Postulating the role of a supernumerary subunit such as MLRQ in complex 1 function and dysfunction Conclusions and Future directions

Part I. bfolecular structure of MLRQ Part II. Biochemical characterization of MLRQ Part III. Localization of MLRQ within complex 1 Part IV. Possible d e s for MLRQ in cornplex I function The final word

References 145

viii

List of figures

Chapter 1. 1.1 Electron micrograph of a mammalian mitochondrion 1.2 Mitochondrial structure and membrane oqanization 1.3 Organization of the rnitochondrial DNA genome 1.4 The protein irnport machinery of mitochondria 1.5 The glucose oxidation pathway 1.6 Mitochondrial respiratory chain and ATP-synthase 1.7 Three dimensional modrls of complex 1 from E. d i . !Y crussa

and B. taurus as detemined by rlectron cryo-microscopy 1.8 Mode1 of the ovrrlap in subunit composition between the different

subcompIexes of complex 1 1.9 Structural model of complex 1 1.10 A hypothetical model for direct energy conversion in Complex 1

Chapter 2. The nucleotide sequence of the human MLRQ subunit cDNA and its deduced arnino acid sequence Alignment of the predictrd hurnan MLRQ subunit protein sequence with that of bovine and mouse Schematic of !L'DC'F.-f-I structure and chromosomal allocation of the eene and its pseudogenes C

Sequence of the MLRQ pseudogene on PAC 2F23 Northem blot analysis of iVDUFrl4 transcripts in normal human tissues PCR amplification of NDC'Fcl4 from genomic DNA Southem analysis of PAC clones 2F23 and 96E2J Regulatory motifs and putative transcription factor binding sites in the 5' lower part of the NDD%il4 gene Comparison of the MLRQ cDNA sequence with the pseudogene sequences from PACs 2F23 and 69E11 in the region showing greatest homology 92

Chapter 3. 3.1 Tissue specific expression of the MLRQ subunit of complex 1 1 07 3.2 Westem blot analysis on mitochondna isolated fiom cultured fibroblasts of

a patient (5621-HT) and control(42 12) using various complex 1 antibodies 109 3.3 Bacterial expression of the MLRQ subunit as a GST-hsion protein 11 1 3.4 Purification and cleavage of GST-MLRQ fusion protein 112 3.5 Westem blot analysis of MLRQ expression in SV40 imrnortalized

fibroblasts transfected with sense and anti-sense pCNDA 3.1 + constructs 1 15 3.6 PCR amplification of sense and anti-sense MLRQ sequences to confirm

transfection of fibroblasts I I6

Extraction of the MLRQ subunit from bovine heart mitochondria 118 Solubilization of the MLRQ subunit compared to other cornplex 1 subunits 120 Hydropathy profile and membrane orientation of the MLRQ polypeptide 12 1 Immunoprecipitation studies using protein A agarose 133 Immunoblotting of the immunoprecipitated MLRQ. MWFE and 19 kDa subunits of complex 1 to determine subunit association 125 Immunoblots OF 1mM DST and 0.2mM EGS cross-linked bovine heart mitochondria with antibodies to the MLRQ and 49 kDa subunits 127 Immuno blotting of EGS cross-linked bovine heart mitochondria with complex 1 antibodies 129 Immunobiotting of EGS cross-linked bovine heart mitochondria immunoprecipitated with MLRQ and M WFE antibodies 133

Chapter 4. 4.1 Schematic representation of the proximity of MLRQ to other subunits of

complex 1

List of tables

Chapter 1. Nomenclature and properties of homologous complex I subunit genes of E. d i . B. triurzrs and H. sapiens C urrent molecular genetic kno wledge of human nuclear-encoded subunits of complex 1 of the mitochondrial electron transport chain Current hypotheses on the subunit location of FMN and iron-sulphur (Fe-S) clusters in the minimal nuclear-encoded functional unit of bovine hem Compiex 1 Composition and penetic origin of mitochondrial respiratory chah (OXPHOS) subunits The clinical presentation and incidence of isolated cornplex 1 deficiency attnbuted to nuclear encoded defects

Chapter 2. 2.1 Nuclear gene mutations in patients with isolated complex 1 deficiency 2.2 Summary of clinical information on patients screened for mutations in

the cDNA of IVDWA-I 2.3 Exon-intron splice junctions of the human iVDUE4-l genr

Chapter 3. 3.1 Subunit proxirnity in complex I as detemined by immunoprecipitation

Studies 3.2 Cross-linked products detected by immunodetection with various

cornplex 1 antibodies

Ab b revia tio ns

A aa ATP ATPase bp BCIP C CAMP CC CD CD cDNA CHAPS COX CPEO Da DCCD DDM DNA DST DTT EDTA EGS EPR EST FAD FeS FILA FISH FMN FP G GTP HQNO HT HTGS HP IgG IP IPTG kb kDa KSS

adenine amino acid adenosine triphosphate adenosine triphosphate sy nthase base pairs 5-bromo-4-chloro-3 -indoly lphosphate p-toluidine salt cytosine cyclic adenosine monophosphate cardiomyopathy and cataracts circular dic hroism cataracts and developmental delay DNA complementary to RNA 3-[(3-cholamidopropyI)dimethy1amrnonio]- 1 -propan-su1 fonat cytochrome c oxidase chronic progressive external ophthalmoplegia daltons N, hr -dicyclohexy lcarbodiimide n-dodecy l-P-D- maltoside deoxyribonucleic acid disuccinimidyl tartrate dithiothreitol ethylenediarninetetra-acetate ethylene glycolbis(succinimidylsuccinate) electron paramagnetic resonance expressed sequence tag flavin adenine dinucleotide iron sulfur center fatal infantile lactic acidosis fluorescence in situ hybridization flavin rnononucleotide tlavoprotein guanine guanosine triphosphate 2-n-hepty l-4-hydroxyquinoline N-oxide hepatopathy and tubuiopathy high throughput genome sequence hydrophobie protein immunoglobuiin G iron-sulfur protein isopropy lthiogalactoside kilo base pair kilodaltons Keams-Sayre syndrome

xii

LD LDAO LDH LHON LA' MELAS

MERRF MM MMC MNGIE mRNA MS mtDNA NaCl NAD NADH NADPH N ARP NBT nDNA NMR Oligo ORF OXPHOS PAC PAGE PCR PD Q QFR QH2 RNA T m rRNA S SDS SMP SQ SQR T tRNA TTFA UTR YAC

Leigh' s disease lauryldimethylamine oxide lactate dehydrogenase Leber's hereditary optic neuropathy lactate to pyruvate ratio mitochondnal encephalomyelopathy with lactic acidosis and stroke-like episodes myoclonus epilepsy with ragged red fibres mitochondrial myopathy myopathy and cardiornyopathy mitochondrial neurogastrointestinal encephalomyopathy messenger ribonucleic acid mild symptorns mitochondrial deoxyribonucleic acid sodium chloride nicotinamide adenine dinucleotide oxidized form nicotinamide-adenine dinucleotide reduced form nicotinamide adenine dinucleotide phosphate reduced form neurogenic muscular weakness, ataxia and retinitis pigmentosa para-nitro-blue tetrazo liurn ch10 ride nuclear deoxyribonucleic acid nuclear magnetic resonance oligodeoxyribonucleotide open reading frame oxidative phosphoiylation system P 1 -artificial chromosome polyacrylamide gel electrophoresis polymerase chain reaction Parkinson's disease ubiquinone menaquinol-fumarate oxidoreductase u b iquino 1 ribonucIeic acid rounds per minute nbosomal ribonucleic acid svedberg unit sodium dodecyl sulfate submitochondnal particles semiquinone succinate-ubiquinone oxidoreductase thymine transfer ribonucleic acid theony ltrifluoro-acetone untranslatecl region(s) yeast artificial chromosome

Glossary of medical terms

Alzheimer's disease - A progressive, neurodegenerative disease charactenzed by loss of function and death of nerve cells in several areas of the brain leading to loss of cognitive function such as memory and language.

Ataxia - Defective muscular coordination affecting baiance, gait, limb or eye movements.

Basal ganglia disease - Disease of the three large subcortical nuclei of the vertebrate brain that participate in the control of movement.

Cardiomyopathy - A general diagnostic term designating primary myocardial disease. often of obscure or unknown aetiology

Cataracts - An ocular opacity, partial or complete. of one or both eyes. on or in the lens or capsule. especiaily an opacity impairing vision or causing blindness.

Chronic progressive external ophthalmoplegia - Disorder where there is a progressive weakness of the extraocular muscles, eventually leading to a complete ophthalmoplegia.

Dystonia - Disordered tonicity of muscle

Encephalomyelopathy - Any disease involving the brain and spinal cord.

Encephalopathy - Any degenerative disease of the brain.

Epilepsy - Recurring disorder characterized by sudden seizure activity or temporary alterations of one or more brain functions arising from abnormal electrical brain activity.

Familial megalencephaly - An inherited disorder where the patient exhibits an enlargement of the head caused by blockage of outflow of cerbrospinal fluid.

Fatal infantile lactic acidosis - Acidosis caused by accumulation of lactic acid more rapidly than it c m be metabolized causing fatality in infants.

Friedreich's ataxia - An autosomal recessive inherited disorder that leads to the progressive dysfunction of the cerebellum, spinal cord and penpheral nerves. Symptoms consist of an unsteady gait (ataxia), slurred speech and jerks eye movemenrs.

Hepatopathy and renal tubulopathy - Enlargement of liver accompanied by disorders of the reabsorptive functions of the kidney with regard to specific nephron segments responsible for specific transport functions.

Hyperventricular cardiomyopathy - Myocardial disease where there is an enlargement of the ventricles in the heart.

xiv

Kearns-Sayre syndrome - Phenotype associated with single deletions of mtDNA. Core clinicai features are a progressive weakness of the muscle which moves the eyes (CPEO) and pigmentary retinopathy.

Lactic acidemia - Condition of high blood lactate resulting from an inborn error of metabolism.

Leigh's disease - A disease of pymvate metabolism manifesting in infancy with psychomotor retardation, dysphagia, hypotonia, atauia, weakness, extemal ophthalmoplegia, vision loss, hearing loss, and convulsions.

Leukodystrophy - An inherited metabolic disorder of the nervous sy stem, particularly the white matter.

Myoclonus - Twitching or spasm of a muscle or group of muscles.

Neurogenic - Arising from or caused by the nervous system.

Neuropathy - Disease involving inflammation or darnage to the peripheral nerves.

Ophthaimoplegia - Paralysis of the ocular muscles

Parkinson's disease - A progressive neurological disease where symptoms include shuffling gait, stooped posture, resting tremor, speech impediments, movement difficulties and an eventual slowing of mental processes and dementia.

Retinitis pigmentosa - Disease caused by overactivity of the pigrnented retinal epithelial cells. leading to damage and occlusion of photoreceptors and blindness.

Wilson's disease - An inherited (autosomal recessive) disorder where there is c~cessive quantities of copper in the tissues, particularly the liver and central nervous system.

Chapter 1

Introduction and objectives The mitochondrion: structure, function, mode of inheritance and

associated diseases

Overview

In recent years, mitochondrial defects have been implicated as playing a role in a wide

range of degenerative diseases, aging and cancer. Although studies on various hurnan

disorders resulting from mitochondrial dysfûnction have given some insight into the

complexities of mitochondrial genetics, the pathophysiology of mitochondrial diseases

remains a perplexing problem due to the interplay between the mitochondrial and nuclear

genomes. The essential role of mitochondrial oxidative phosphorylation in cellular energy

production, the generation of reactive oxygen species, and the initiation of apoptosis has

suggested a number of novel mechanisms for mitochondrial pathology. The importance

and interrelationship of these pathways are now being studied. In order to illustrate these

interrelationships, this section begins with the examination of mitochondrial structure,

DNA organization and protein import into the organelle. It then proceeds to look at the

function of the different complexes of the respiratory chain with particular emphasis on

the filst enzyme of the chah namely, NADH:ubiquinone oxidoreductase or complex 1. In

order to establish a solid foundation for the concepts and ideas that are presented in

succeeding chapters, a thorough discussion of the subunit composition, structural mode1

and energy conversion pathways of complex 1 is presented. This chapter concludes by

htroducing the clinicai spectnun of mitochondrial pathology arising fiom defects in both

the rnitochondrial and nuclear genomes, with emphasis being placed on human complex 1

deficiency.

Part 1. Mitochondrial structure, function and inhentance

The mitochondrion

Al1 reactions in cells involving growth and metabolism require energy. Mitochondria

were identified 5 1 yean ago as the organelles responsible for most celluiar energy-

production (reviewed by Gray et al, 1999). It is generally believed that mitochondria

represent the descendents of primitive bacterial cells (cyanobacteria) that became

symbiotically associated with primitive ancestors of the present eukaryotic organisms,

thereby increasing the host s energy-generating capacity (Gray et al, 1999). Each

mitochondrion performs a multitude of Functions which include the reactions of the Krebs

cycle, oxidative phosphorylation and P-oxidation (Darne11 et al, 1986). Mitochondria

display an amazing plasticity of form and distribution. The size, shape and quantity of

mitochondria Vary between tissues and even arnong different locations within the same

tissue (Munn, 1974). Although, their interna1 structural organization is highly conserved,

the extemal shape of mitochondria is variable. In addition to the classic kidney-bean

shaped organelles observed in electron micrographs (Fig. 1.1), mitochondria are also

fiequently found as extended reticular networks (Chen, 1988). These networks are

extremely dynamic in growing cells, with tubular sections dividing in half, branching and

fusing to create a fluid tubular web (Bereiter-Hahn and Voth, 1994). In differentiated

cells, such as those found in cardiac muscle or kidney hibules, mitochondria are often

localized to specific cytoplasmic regions rather than randomly distnbuted (Yaffe, 1999).

Typical mitochondria within a rat liver ce11 will exhibit an elliptical shape with an

approxirnate length of 1 -3pm and a width of O. 1 - 1 pm (Lehninger, 1964). The average

number of mitochondria within a rat liver ce11 is approxirnately 1 O00 (Munn, 1974), but a

range of 500-2500 has been reported. Therefore, rnitochondna can occupy nearly 20% of

the total cellular volume (Lehninger, 1964).

Figure 1.1. Electron micrograph of a mamrnaüan mitochondrion. The outer membrane as well as the highly folded, finger-like cristae which make up the inner membrane are clearly visible. Adapted fiom Fawcett, A., 1994.

Mitochondrial ultrastructure

The presence of a double membrane is a common feature of ail mitochondria (Fig. 1.2).

The outer and inner rnitochondrial membranes define the two submitochondrial spaces:

the intermembrane space between the two membranes, and the central matrix

cornpartment (Darne11 et ai, 1986). The outer mitochondrial membrane is freely

permeable to most of the small molecules ( 4 0 kDa) because it is covered with

hydrophilic pores or channels composed of the protein porin (Manella, 1982). The inner

mitochondrial membrane, because of its high protein and cardiolipùi content is only fieely

permeable to O?, CO2. H 2 0 and srnall metabolites (Darnell et al, 1986). It is highly

folded, with finger-like projections termed cristae which greatly increase its internai

surface area (Darnell et al, 1986). While the respiratory chah is situated in the imer

mitochondrial membrane, most of the reactions involving the oxidation of pynivate and

fatty acids take place in the mitochondrial matrix (Darnell et ai, 1986).

Mitochondrial DNA or~anization

The majority of mitochondrial proteins are nuclear encoded and imported into the

mitochondria from the cytoplasm. In addition, mitochondria possess their own unique

genome and this mitochondnal deoxyribonucleic acid (mtDNA) is inherited maternally in

humans, because sperm mitochondria do not survive afier fertilization (Giles et al, 1980).

The human mitochondrion typically contains 2 to IO copies of the mitochondrial genome

(Harding and Holt, 1993). MtDNA is a closed circular, double stranded (ch) molecule

consisting of 16,569 bp and has been completely sequenced (Anderson et al, 198 1 ;

reviewed by Wallace, 1993) (Fig. 1.3). It is a compact piece of genetic information with

little intervening, non-coding sequence with the exception of its short regulatory region

termed the D- or displacement loop (Anderson et al, 198 1 ; Tzagoloff and Myers, 1986).

I Cristae 1 Respiratory chah and ATP synthase

Inner membrane

Figure 1.2. Mitochondrial structure and membrane organization. A common feanire of al1 mitochondria is the presence of a double membrane. The outer membrane completely envelopes the inner membrane. The highly invaginated i ~ e r membrane houses the OXPHOS system. - Complex 1; - Complex II; - Complex III; 0- Complex IV; s -

MtDNA encodes genes for the 12s and 16s ribosomal ribonucleic acids (rRNAs), 22

transfer ribonucleic acids (tRNAs) and 13 messenger ribonucleic acids (rnRNAs) for

polypeptides which are components of the mitochondrial oxidative phosphory lation

system (Wallace, 1993).

Mitochondrial re~lication. transcri~tion and translation

Due to their different buoyant densities in alkaline cesium chloride gradients, the two

strands of mitochondrial DNA are referred to as the heavy (H) or guanine rich strand, and

the complementary light (L) strand which is cytosine rich (Larsson and Clayton, 1995).

DNA replication initiates within the D-loop region at the OH ongin of replication on the

H-strand (Larsson and Clayton, 1995). When the leading strand has elongated to two-

thirds of its total length, the OL ongin of replication on the L-strand (which is nested in a

cluster of five R N A genes) is then exposed and initiates lagging-strand replication

(Larsson and Clayton, 1995). Components that are crucial for the replication process

such as the mitochondrial specific y-DNA polyrnerase (Bolden et al, 1 977), mtDNA

helicase (Hehman and Hauswirth, 1992) and primase enzymes (Wong and Clayton, 1985)

are nuclear encoded factors which are imported into the mitochondria. The H-strand

encodes the 12s and 16s rRNAs, 14 tRNAs and 12 polypeptides of the respiratory

chain, while the L-strand encodes the ND6 subunit and eight tRNAs (Shofier and

Wallace, 1994). Mitochondrial transcripts begh at two promoter regions, PH and PL for

the H- and L-strand transcripts, respectively (Shoffner and Wallace, 1994)). With the

help of mitochondnal transcription factor (h-mtTFA) and possibly sorne other factor, a

HSP r

TRANSCRIPTION TERM SITE

Figure 1.3. Organization of the mitochondrial DNA genome. The circular, doubIe stranded 16.5 kb human mitochondriai DNA is iilustrated, indicating the locations of the encoded genes. MtDNA encodes two rRNAS ( 12S, 16s). 22 tRNAs ( for its own protein synthesis. and 13 rnRNAs for protein subunits of the respiratory chah complexes: ND 1-6 for complex 1, cyt b for cornplex III, CO 1-III for complex N and ATP 6 and 8 for complex V. (Adapted from Pitkanen. S.. Academic dissertation. 1997).

mitochondrial specific RNA polyrnerase produces long polycistronic transcripts which

are later processed (Larsson and Clayton, 1995). An additional transcript is also

produced fiom the PH promoter at a rate approximately 10-30 fold greater than the full

length PH transcript (Shoffner and Wallace, 1994). Following their release from the

primas, polycistronic transcript by a mitochondrial endonuclease (P), both R N A

molecules and coding transcripts are hrther modified to form mature functional tRNAs

and polyadenylated rnRNAs, respectively (Tzagoloff and Myers, 1986). The 13 mRNAs

are then translated in the mitochondria by rnitochondrial ribosomes which utilize the two

mitochondrial rRNAs and 22 mitochondrial tRNAs (Shoflher and Wallace. 1994). The

mitochondrial ribosomal complex is composed of small28S and large 39s subunits

(Tzagoloff and Myers, 1986). Mitochondrial ribosomes have been proposed to bind in a

non-specific manner to mitochondrial transcripts, whereby the small28S ribosome

subunit interacts with the mRNA molecule and scans for the initiation codon (Liao and

Spremulli, 1990). Because of this unique interaction, the translation of mitochondnal

transcnpts is believed to involve monoribosomes rather than polyribosomes as is seen in

cytoplasmic protein synthesis (Liao and Spremulli, 1990). This process may be aided by

mitochondriai translational initiation factors and subunit S5, a GTP binding component of

the small ribosomal complex (O Brien et al, 1990).

Mitochondrial rote in im~or t

Proteins destined for mitochondriai irnport have either an amino terminal targeting

sequence (rnost comrnon) or an intemal targeting sequence (Rassow et al, 1999) (Fig. 1.4).

The matrix targeting signals (MTSs) at the amino terminal are usually 20 to 60 amino

acid residues in Iength, rich in positively charged amino acids (arginine and lysine), rich in

hydroxylated arnino acids (serine and threonine), Iack acidic residues (aspartic acid and

glutamic acid), and have an arnino acid sequence necessary for forming an amphpathic

helical structure (von Heijne, 1986). The intemal targeting sequences however are poorly

characterized and are rnainly found in hydrophobie preproteins (Hermann and Neupen.

2000).

The proteins which participate in the translocation of proteins across the

mitochondrial outer membrane are systematically named TOM proteins; the

corresponding proteins of the imer membrane are named TfM proteins (Hurt et al, 1984).

The TOM complex is comprised of an array of import recepton and a protein conducting

channel (Komiya et al, 1988). Targeting sequences are recognized by the receptors

Tom20 and Tom70 and are transferred to a general insertion pore made up of TomrlO,

Torn22, Tom7 and Tom6 (Komiya et al, 1998). After passage across the outer

membrane, some preproteins first bind to proteins in the intermembrane space, others

immediately insert into import sites of the inner membrane (Komiya et ai, 1998).

Preproteins that translocate into the intermembrane space, interact with the TIM23

cornplex made up of proteins Tirn 17, T b 2 3 (which f o m the protein-conducting

channel), Tim44 (a hydrophilic ma& protein) and presumably an as yet unidentified 14

kDa subunit (Henman and Neupert, 2000). Translocation across the mitochondrial imer

membrane is strictly dependant on the mitochondrial membrane potential (Berthold et al,

1995). It is believed that the membrane potentiai exerts an electrophoretic effect on the

positively charged parts of the preproteins, irrespective of whether the charges are

localized at the amino terminus or in mature parts of the preprotein (Berthold et al, 1995).

In the matrix, the incoming preprotein is b o n d by the chaperone mtHSP7O which is

associated with T h 4 4 (Benhold et al, 1995). This process is ATP dependent and is

assisted by the CO-chaperone Mge 1, a mitochondrial homolog of the prokaryotic GrpE

protein which facilitates the release of Tim44 fiom mt-Hsp70 and tight binding to the

preprotein (Westermann et al, 1995). As soon as an amino-terminal presequence enters

the matrix cornpartment, it is usually cleaved off by the specific processing enzyme MPP

(mitochondrial processing peptidase) (Hurt et al, 1984).

Three different pathways are responsible for translocation into the imer membranes

depending on the targeting sequence of the preprotein (reviewed by Hermann and

Neupert, 2000). Proteins can be imported as described above, but become arrested at the

level of the TIM23 complex and laterally inserted into the lipid bilayer (Kaput et al,

1982). Secondly, proteins can be completely transportrd into the matrix from where they

reinsert into the inner membrane (Hart1 et al, 1986). To date, the only identified

component of this insertion complex is Oxal (Hel1 et ai, 1998). A third pathway into the

imer membrane is used by the hydrophobie preproteins that do not contain MTSs but

instead have intemal targeting signals (Palmisano et al, 1998). They bind to soluble

proteins Tim9 and Tim 1 O and are then hported via Tim 12 and a larger complex of

membrane proteins containhg T h 2 2 and TimS4 (Palmisano et ai, 1998). Recently, a

second soluble intermembrane space complex formed by T h 8 and Tim 13 was descnbed

which also seems to be involved in rnitochondrial protein import (Leuenberger et of,

1 999).

Preproteins carrying an amino terminal presequence

CYTOSOL

INTERMEMBRANE &h \

Figure 1.4. The protein import machinery of mitochondria. Transport of preproteins across the outer membrane is mediated by Tom proteins while transport across the inner membrane is mediated by Tirn proteins. Depending on the targeting sequence. protein impon into and across the inner membrane can follow four different pathways and requires the membrane potential Av. OM, outer membrane; IM, inner membrane; MPP, mitochondrial processing peptidase; Mgel, mitochondnal GrpE homologue; mtHSP70, mitochondrial heat shock protein of 7OkDa; Tom, translocase of the outer membrane; Tim, translocase of the inner membrane.

Enerw metabolism

Cellular energy requirements are satisfied mainly by the hydrolysis of phosphate

bonds in ATP (reviewed by Zubay, 1993). Therefore, in order to sustain all cellular

processes, ATP pools must be maintained at high enough levels. Mitochondna are the

organelles responsible for transforming energy fiom the oxidative breakdown of

carbohydrate, fatty acids and amino acids into ATP (Zubay, 1993). Glucose metabolism

begins in the cytosol with glycolysis, the conversion of one glucose molecule into two

molecules of pyruvate (Zubay, 1993). This results in the reduction of two

NAD-(nicotinamide adenine dinucleotide oxidized form) to two NADH (nicoiinamide

adenine dinucleotide reduced form), as well as a net energy yield of two ATP (Zubay,

1993 ). In mammalian systems, the mitochondrial malate-aspartate shut~le transfers the

two molecules of NADH formed in glycolysis into mitochondria (reviewed by Meijer and

van Dam, 1974). A less efficient glycerophosphate shuttle serves the sarne function in

tissue such as insect muscle (Voet and Voet, 1990). Oxidative decarboxylation of

pyruvate takes place in the mitochondrial matrix, followed by oxidation by the Krebs

cycle enzymes (Fig. 1.5) (Voet and Voet, 1990). The energy released by the oxidation of

one molecule of glucose is conserved in ten NADH, two FADH? (flavin adenine

dinucleotide reduced form) and two ATP molecules (Voet and Voet, 1990). NADH and

FADHz are M e r oxidized by the mitochondrial respiratory chah (Voet and Voet,

1990). In the absence of oxygen, the lactate dehydrogenase enzyme catalyzes a reaction

to produce lactate, the end-product of anaerobic glycolysis, and NAD', which can then be

recycled for the oxidative reaction (Zubay, 1993). Therefore, reasons for lactic acid

13

accumulation in the ce11 indude lack of sufficient oxygen or dysfunction of mitochondrial

enzymes required for oxidative phosphorylation (reviewed by Robinson, 1989).

GLUCOSE I

GLUCOSE C -@--a.

2 NAD+

MALATE-ASPART ' ATE

Figure 1.5. The glucose oxidation pathway in mamrnals. Complete oxidation of one molecule of glucose gives a net of 38 ATP molecules. Dashed lines indicate necessary reactions for continuation of glycol sis. LDH, lactate dehydrogenase; PDH-complex, pyruvate dehydmge 4 - Complex IP - Cornplex II; O - Cornplex III;

- Complex IV; - Complex V; - Quinone; a- Cytochrome c.

Part II. The Mitochondrial respiratory chain complexes

Organization of the OXPHOS svstem

The oxidative phoçphorylation system (OXPHOS) located in the inner mitochondnal

membrane consists of five distinct multimenc protein assemblies (Fig. 1.6):

1 > NADH-ubiquinone oxidoreductase (Cornplex 1) also called NADH-dehydrogenase or NADH-coenzyme Q reductase

2) Succinate-ubiquinone oxidoreductase (Complex II)

3) Ubiquinol-femcytochrome c oxidoreductase (Cornplex II 1)

4) Cytochrome c oxidase (Complex IV)

5 ) ATP synthase or FIFo-type ATPase (Complex V)

A new study by Schagger and Pfeiffer (2000) shows that these complexes are not

randomly distributed, but instead assemble into suprarnolecular structures. The work

involved solubilization of the membrane protein complexes through mild one-step

protocols using dodecylmaitoside (DDM), digitonin and Triton X- 100, followed by blue-

native PAGE to isolate the supramolecular stnictures. In marnrnalian mitochondria,

aimost al1 complex I is seen to assemble into supercomplexes comprising complexes I and

III and up to four copies of complex N (Schagger and Pfeiffer, 2000). In Saccharomyces

cerevisiae, complex IV is predominantly found associated with complex III and exists in

three forms, the fiee dimer, and two supercomplexes with another one or two complex N

monomen (Schagger and Pfeiffer, 2000). The amount of supercomplexes formed depends

on the ce11 s demand for energy. A respirasorne mode1 is proposed with two copies of

a Il iIIzNa building block and one copy of a IIIzW4 building block (Schagger and Pfeiffer,

2000) to fit the overall 1 :3:6 stoichiometries of complexes 1:HI:IV determined by Hatefi

(1 985). Association of complex II with any of the other respiratory chain complexes was

not identifid (Sc hagger and Pfeiffer, 2000).

MATRIX INTER MEMBRANE SPACE

Figure 1.6. Mitochondrial respiratory chain and ATP-synthase. Arrows indicate the direction of electron and proton (El? = hydrogen ion) flow. C 1 - V, complexes I - V: Q, coenzyme Q; cyt, cytochrome; FAD, flavin adenine dinucleotide; Fe-S, Von-sulphur clusters.

The OXPHOS svstem: Role in electron transoort and oroton translocation

The fmt four complexes, with ubiquinone and cytochrome c, make up the

rnitochondnal respiratory chain, also called the electron transport chah (Saraste, 1999).

The operation of the respiratory chain is characterïzed by two distinct processes that are

linked: electron transport and proton pumping (Saraste, 1999). NADH is oxidized by

complex 1 and succinate by cornplex II, followed by the electron transfer to ubiquinone

and then M e r to complex III, to cytochrorne c, to complex IV and to the final electron

acceptor, oxygen (Voet and Voet, 1990). The electron transport via complexes is coupled

to proton pumping fiom the rnatrix to the intermembrane space, producing an

electrochemical gradient (Voet and Voet, 1990). Because of this gradient, the protons

flow back from the intermembrane space to the mitochondnal matrix through complex V,

ATP synthase, and the released energy is captured in the form of ATP (Voet and Voet.

1990).

The mitochondrial electron transport chain houses at least 4 types of electron carriers:

flavins, iron sulfur clusters, quinone and cytochromes (Ohnishi, 1998). Complexes 1 and

II contain as prosthetic groups, flavin mononucleotide ( F m ) and flavin adenine

dinucleotide (FAD), respectively (Ohnishi, 1998). Protein subunits in complexes 1, II and

III bind iron sulfur clusters with two to four iron and sulphur atoms: [2Fe-ZS], [3Fe-3S]

or [4Fe-4S] (Saraste, 1999). The ubiquinone present in the mitochondnal inner membrane

is responsible for passing electrons on to the respiratory chain cytochrome system,

consisting of three types of cytochromes: a (a and a3), b, and c (c and cl) (Saraste, 1999).

The iron atoms in the iron sulphur clusters and in hemes of the cytochromes undergo

oxidation and reduction during respiration, cycling between the ferrous ( ~ e ' " and ferric

( ~ e ' ~ ) oxidation States (Saraste, 1999).

Comolex 1: The NADH-ubiauinone oxidoreductase complex

NADH-ubiquinone oxidoreductase or complex 1 is the largest of the membrane-bound

enzymes of the mitochondrial respiratory chain with a total molecular mass of - 1 o3

kilodaltons &Da) for monomenc complex 1 (Fig. 1.7) (Smeitink et 01, 1998a). The

mammalian complex 1 enzyme is composed of at least 43 subunits of which 7 are

mitochondnally encoded while the remaining subunits are encoded by the nuclear genorne

(Smeitink et al, 1998a). In cornparison, the fùngus Neurospora crassa which is widely

used as a simple eukaryotic mode1 organism in the shidy of complex 1, has 35 subunits

(Schulte et al, 1994), Pmcoccus denitrificans (designated NDH- 1 ) has 14 subunits

(Takano et al, 1996) and Escherichia coli complex 1 has 13 subunits (Blattner et ut, 1997).

However, Saccharomyces cerevisiae and other fermentative yeasts do not contain this

multi-enzyme complex of the respiratory chain but instead have a simple dimeric

diaphorase (discussed below) (Brody a al, 1997).

Evolution of Complex 1

Cornplex 1 is thought to have originated by fusion o f pre-existing protein assemblies

constituting modules for coupled electron transfer and proton transport. These Func tional

modules are defined by the homology of parts of complex 1 to other bacterial enzymes.

Complex 1 is found in purple bacteria and in the mitochondria of most eukaryotes. While

the eukaryotic NADH:ubiquinone oxidoreductase is referred to as complex 1, its bactenal

counterpart has traditionaliy been called NADH dehydrogenase type 1 (Friedrich and

Weiss, 1997). The known examples of bacterial complex 1 fiom purple bacteria consist of

14 different subunits while the mitochondrial complex contains at least 28 accessory

proteins which do not directly participate in electron and proton transport (Friedrich and

Weiss, 1997). There is no evidence that bacteria other than purple bacteria contain this

respiratory enzyme. In fact, a non proton pumping NADHxbiquinone oxidoreductase

with a single FAD redox group called NADH dehydrogenase type II appears to be more

widespread than complex 1 in bacteria (Matsushita et al, 1987; Yagi et al, 1992). Fungi

and plant mitochondria contain two of these non proton-purnping NADH:ubiquinone

oxidoreductases in addition to complex I (Friedrich and Weiss, 1997). These complexes

have a lower afinity for NADH as compared to complex 1 and most likely operate as

overflow outlets for an excess of reducing equivalents (Friedrich and Weiss, 1997).

Fermentative yeasts whch lack complex 1, use these complexes exclusively to oxidize

mitochondrial NADH (Friedrich and Weiss, 1997). A minimal bacterial complex

which is homologous to the respiratory complex 1, is found in cyanobacteria and

chloroplasts carrying only 1 1 subunits (Berger et al, 1993). However, this system is

thought to work as a NADPH:plastoquinone oxidoreductase in a cyclic photosynthetic

electron transfer (Friedrich et al, 1 995).

Most information about bactenal complex 1 comes fiom E. coli where complex i genes

are organized in the sotalled nuo-operon (Friedrich et al, 1995). Seven genes code for

peripheral proteins, including a11 proteins with binding motifs for NADH, FMN and al1

Fe-S clusters (Friedrich et al, 1995). The seven remaining genes code for the

hydrophobie, intrinsic membrane proteins. These 7 instrinsic membrane subunits, the

homologues of the E. coli NuoA, H and J-N are rnitochondrially encoded in animals and

fimgi (Attardi and Schatz, 1988), while al1 other subunits are nuclear-encoded in most

eukaryotes (Friedrich et (11, 1995) (Table 1.1). The evolution mode1 proposed by Finel

( 1998) suggests that the hydrophobie subunits of complex 1 evolved together with the

nuclear-encoded subunits until they were transferred from the mitochondnal chromosome

to the nucleus.

Table 1.1. Nomenclature and properties of homologous complex 1 subunit genes of E. d i , B. taurus and Ho sapiens

E. coli B. taurus H. sapiens Predicted Function

NuoA NuoB NuoC NuoD NuoE NuoF NuoG NuoH Nu01 NuoJ NuoK NuoL NuoM NuoN

ND3 PSST 30 (IP) 49 (IP) 24 (FP) 51 (FP) 75 (IP) ND t TYKY ND6 ND4L ND5 ND4 ND2

Q binding/e- transport NADH binding Q binding e' transport NAD H b indinde- transport e' transport H' pumping e- transport HI pumping

H' pumping H' purnping

The electron transfer moiety of complex I can be traced back to two different origins.

The fust is to the diaphorase part of soluble NAD'-reducing hydrogenase found in purple

bacteria such as Akaligenes eumphus, Desu[fovibriofnrctosovorans and cyanobacteria

Anabaena varibilis, Anacystk nidufam, while the second is to the formate hydrogenlyase

complex of E.coli (Friedrich and Weiss, 1997). Based on homology, the origin of the

proton transporting moiety of complex 1 can be related to bacterial Na'N and K'/H'

antiporten (Friednch and Weiss, 1997). It has also been proposed that NuoL and NuoH

and the sugar permeases of the bacterial phosphoenolpynivate-dependent

phosphotransferase system belong to a superfamily of pore-fonning proteins (Reizer et

al, 199 1). As the electron transfemng and proton transporthg modules were joined

together, they gave rise to the ancestor of complex 1 and the formate hydrogenlyase

(Friedrich and Weiss, 1997).

To promote electron transport in a specific direction and to prevent the capture of

electrons by other redox acceptors in the aqueous and membrane phases, there must be a

layer of insulating protein surroundhg the electron pathway that should be about 17 to

20 in thickness (Moser and Dutton, 1996). Many of the additional subunits of the

mitochondrial enzyme complex are thought to form this scaffold, keeping the redox

groups in the right position to prevent the electrons from escaping and forming reactive

oxygen species (Friedrich and Weiss, 1997). This results in safer energy conversion in

eukaryotes compared to bacteria. Because several of these additional subunits show no

sequence similaity between animal and fungi, they are thought to have emerged late in

evolution when these species had already diverged or they diverged so fast that a possible

homology cannot be seen (Friedrich and Weiss, 1997).

Subunit composition of the Human NADEubiquinone oxidoreductase complex

As previously elaborated, hurnan rnitochondrial complex 1 appears to be made up of

43 subunits of which 7 are mitochondrially encoded. The nuclear encoded subunits are

synthesized in the cytosol and transported into the mitochoncùia (Chomyn et al, 1988)

Treatment with chaotropic salts like sodium bromide resolves cornplex 1 into three

fractions. They are, narnely 1) the flavin protein (FP) fraction containing polypeptides

with a high fiavin, iron and sulphide content, 2) the iron protein Fraction (IP) which

contains a high iron and low flavin content and 3) the hydrophobic (HP) fraction

containing polypeptides with the lowest non-heme iron protein ratio (Galante et ai,

1979). The extrinsic membrane dornain has at least 20 subunits, most of which are

hydrophilic, including al1 the subunits of the FP and IP fractions (Galante et al, 1 979). It

also contains the FMN and most of the Fe-S clusters. The intrinsic membrane domain

which contains the HP fraction is an assembly of - 24 nuclear-encoded subunits and the 7

mitochondrial gene prducts (Belogmdov et al, 1994). However, this Fraction also

contains globular water-soluble subunits and being a part of the HP fraction does not

necessarily indicate that a subunit is hydrophobic or that it belongs to the membrane

domain (Walker et al, 1992). Fractionation using detergents (see later section) gives a

clearer picture of complex I structure.

(i) The Flavoprotein Fraction (FP)

The genes encoding the 3 subunits which make up the Flavoprotein (FP) fraction,

namely, NDUWI (5 1 kDa), NDUFV2 (24 kDa) and NDUFV3 (10 D a ) , have al1 been

characterized at both complementary DNA (cDNA) and nuclear DNA (nDNA) levels

(Table 1.2).

The 5 1 kDa subunit (encoded by N D U N I ) is known to hold the binding site for

NADH (Patel et al, 199 1) and also contains the FMN (Krishnamoorthy and Hinkle,

Table 1.2. Current rnolecular genetic knowledge of human nuclear-encoded subunits of cornplex 1 of the mitochondrial electron transport chain

Gene Subunit Mr

Flavoprotein fraction (FP) NDUFVI 5 l kDa 5 1 NDUFV2 24 kDa 24

iron-Sulphur protein fraction (IP) NDUFSI 75 kDa NDUFS:! 49 kDa

NDUFS3 30 kDa NDUFS-I AQDQ NDUFSS 15 kDa NDUFS6 DDGD NDUFAS BI3

Hydrophobic fraction (HP) NDUFA l NDUFA2 NDUFA3 NDUFAJ NDUFA6 NDUFA7 NDUFA8 NDUFA9 NDUFAIO NDUFAB 1 NDUFB l NDUFB2 NDUFB3 NDUFB4 NDUFBS NDUFB6 NDUFB7 NDUFB8 NDUFB9 NDUFB 10 NDUFC l NDUFC2 NDUFS7 NDUFS8

MWFE 8 8 B9 MLRQ B 1.l ASAT PGIV 39 kDa 42 kDa SDAP MNLL AGGG BI2 B 15 SGDH BI7 BI8 ASHi B22 PDSW KFYI MMTG PSST TYKY 17.2 kDa

Chromosome cDNA length ( O W

1331 bp 650 bp

22 1 bp

Z l l l bp 1388 bp

791 bp 398 bp 317 bp 371 bp 347 bp

210 bp 296 bp 25 1 bp 242 bp 383 bp 338 bp 515 bp

1 l j 0 bp 959 bp 263 bp 173 bp 215 bp 293 bp 586 bp 428 bp 383 bp 404 bp 473 bp 466 bp -il l bp 1.16 bp 356 bp 524 bp 527 bp 435 bp

References

Schuelke rr ul. 1998: de Coo er al, 1995 Hattori er al, 1995 de Coo er al. 1997

Chow er al. 199 1 Loeffen er ul. 199th Procaccio CJI ul, 1998 LoetTen et ul. 1998a van den Heuvel er ul. 1 9 ~ 3

Loe tEn er al. 1999 Loeffen rr al. 1998a Pata et ul. 1997 Russell er 01. 1997

Zhuchenko el al. 1996 Ton et ul. 1997 Loeffen er al. 1998b Kim et al. 1997 * Ton tir u1. 1997 Loetfén et ul. l998b Triepels et (11. 1998 Baens L'I d. 1994 Loeffen er ul. 1998b Triepels et al. 1999a

Loet'fen et ul. I998b Ton er al. 1997 Loeffen tir dl. I998b Ton et cd, 1997 Smeitink et al. 1998b Wong rr ul. 1990 Loeffen er ul. 1998b Gu er 01. 1996 Loeffen rr u1. 1998b Ton er al. 1997 Loeffen et ul. 1998b Hyslop er ul. 1996 Procaccio er al. 1997 Triepels rr al. 2000

(Chromosomal localizations: Ali et al. 1993; Duncan et al, 1992: Emahazion and Brookes. 1998: Emahazion et al. 1998)

* Kim et al, 1997 were another group that published the cDNA sequence of the human MLRQ homolog. Chromosomal localization of the XDLFA-I gene was not carried out by these authors.

1 988), a tetranuclear iron-su1fi.u cluster (Ohnishi et al, 1985) and a consensus motif for

the binding of the nuclear respiratory factor 11 (NRF-2) in its genomic structure, which is

thought to be involved in the transcriptional regdation of nuclear genes which code for

mitochondrial proteins (Schueke et al, 1998). 100% anti-sense hornology was found

between the 3' UTR of NDUFVI-mRNA and the 5' UTR of the mRNA for the y-

interferon inducible protein (IP-30) precursor (Schuelke et al, 1 998). It is hypothesized

that the NDUFVI-mRNA may act as an anti-sense suppressor, restraining translation of

IP-30 in tissues with high energy demand. This could therefore be the molecular link

between complex 1 deficiency and inflarnmatory myopathy which have been repeatedly

described to occur together (Schuelke et ol, 1998). The 24 kDa subunit contains four

strictly conserved cysteine residues for the binding of a binuclear iron-su1 fur c luster

(Ohnishi et al, 1985). This subunit has also been s h o w to bind GTP and possibly

exhibit GTPase activity when bound to the native complex 1 (Hegde, 1998). Mutational

studies on the 24 kDa subunit in Neurospora crassa has showed that this subunit is

absolutely essential for complex 1 activity and this may explain cases where the 24 kDa

subunit is reduced or absent in human mitochondrial diseases (Almeida et al, 1999). A

good example of this is show by Schapira et al (1 988) where the 24 kDa subunit appears

to be absent in a patient with rnitochondnal myopathy. Hattori et al ( 1998) have

reported a Ala29Val substitution in the mitochondrial targeting sequence of the 24 kDa

subunit in patients with Parkinson s disease (PD). This fkequency of homozygotes for

the mutation was significantly higher in PD patients than in control subjects and the

mutation may well be a cause of complex 1 deficiency in Parkinson s disease (Hattori et

al, 1998).

The 10 kDa subunit has no redox centers (de Coo et al? 1997) and is situated at close

proximity to the 24- and 5 1- kDa subunits (Yamaguchi and Hatefi. 1993). The

Iocalization of the NDUFV3 gene on chromosome 21q22.3 borders the location of the

gene for the mitochondnal ATP5O protein, which is thought to contribute to the Down

Syndrome phenotype (Chen and Antonarakis, 1995). Because Down syndrome has been

postulated to be a contiguous gene syndrome, de Coo et al (1 997) believe that the 1 O kDa

subunit might also be associated with this disease.

(ii) The Iron-Sulphur Protein Fraction (IP)

The Iron-sulfbr (IP) Fraction is made up of at les t 7 subunits, encoded by NDUFSI

(75 ma) , NDUFS2 (49 kDa), NDUFS.3 (30 kDa), NDUFS4 (AQDQ/18 kDa), NDLrFSS

(1 5 ma) , NDUFS6 (DDGD/ 13 kDa) and NDUFA.5 (B 13). Al1 of these subunits have

been charactenzed at the cDNA level (Table 1.2). The genomic DNA sequence has only

been determined for the B 13 subunit encoded by the NDUFAS gene (Tensing et al, 1 999)

and the 30 kDa subunit encoded by the NDUFS3 gene (Procaccio et al, 2000).

The 75 kDa subunit contains conserved cysteine motifs allowing for the existence of

one tetra-nuclear, one binuclear and possibly another tetranuclear iron-sulfur cluster

(Ohnishi, 1998). The 49 kDa subunit seems to be essential for complex 1 function as was

demonstrated by knockout mutants of the 49 kDa gene in N.crassa, which lacked NADH

dehydrogenase activity completely because the matrix arm of the cornplex failed to

assemble (Schulte and Weiss, 1995). Sequencing of the gas- 1 gene in Caenorhabdilb

elegum has revealed that it is a homologue of the 49 kDa subunit in the wom s

respiratory chain and that it plays a role in the determination of anesthetic sensitivity in

C. elegans (Kayser et al, 1 999). Danouzet et al ( 1 998) using bacterial genetics have

shown that the 49 kDa subunit is involved in the binding of piericidin and rotenone (both

quinone-related inhibitors) and thereby implicate this subunit in quinone binding. Both

the 49- and 30 kDa subunits contain highly conserved phosphorylation sites which are

thought to be involved in regulatory functions (Loeffen et al, 1998a). The 30 kDa subunit

shows similarity to a protein encoded by a gene on the chloroplast genome (Weiss er al,

199 1). An active CAMP-dependent protein kinase consensus phosphorylation site has

been determined in the AQDQ subunit (Papa et al, 1996) and phosphorylation of this

subunit in response to cholera-toxin treatment has been s h o w to be accompanied by a 2-

3 fold enhancement of rotenone-sensitive NADH oxidase respiration and

NADH:ubiquinone oxidoreductase activity of complex 1 (Scacco et al, 2000).

(iii) The Hydrophobic Protein Fraction (HP)

In humans, the iargest fraction of complex 1, the Hydrophobic fraction (HP), contains

the 7 mitochondrially encoded subunits as well as the 24 nuclear encoded subunits which

have been characterized on the cDNA level (Table 1.2).

a) The mitochondrially encoded subunits

The 7 subunits encoded by the mitochondrial genome are ND 1 (3 18 aa), ND2 (347 aa),

ND3 (1 15 aa), ND4 (459 aa), ND5 (603 aa), ND6 (1 74 aa) and ND4L (98 aa) (Chomyn et

al, 1985; Chomyn et al, 1986). Apolar stretches long enough to traverse the membrane

are characteristic of these subunits in a variety of organisms, although homologous

subunits Vary considerably in size (Weiss et al, 199 1). The great degree of sequence

identity between the ND2, ND4 and ND5 encoded subunits suggests that these genes

evolved from a single ancestral gene (Kikuno and Miyata, 1985). It has been proposed

that these mitochondrially encoded subunits are involved in proton translocation

(Guenebaut et al, 1998). The carboxyl group modifying reagent N,N -dicycIohexyl-

carbodiimide (DCCD) was found to act on complex 1 at the ND1 subunit site, thereby

blocking proton translocation and to the same extent electron transfer in complex 1 (Yagi

and Hatefi, 1988). The ND I subunit has historically been associated with the ubiquinone

reduction site from photolabeling studies done with rotenone analogues in isolated

complex 1 (Earley et ai, 1987). However, recent photoafinity labeling studies by Schuler

et al (1999) on mitochondrial electron transport particles has suggested that ND I is not

directly involved in the action of quinone binding. In human cells, the absence of the ND4

subunit has led to a failure to assemble other mtDNA-encoded subunits and a complete

loss of NADH:QI oxidoreductase activity (Hofhaus and Attardi, 1993). The ND5 subunit

has four cysteine residues that are conserved and thought to be able to ligate an iron-sulfur

cluster located within the membrane (Weiss et al, 199 1). The ND5 subunit contains the

same consensus sequence around its conserved myristoylated lysine that the proton

conducting COX subunit 1 does, although the exact purpose of the rnyristoylation is not

yet known (Plesofsky et al, 2000). Work done on a mouse A9 ce11 line has shown that a

fkimeshift mutation in the mitochondrial gene for the ND6 subunit, resulting in a

complete absence of the subunit caused a loss of assembly of the mtDNA-encoded

subunits and a serious impairment of oxidative phosphorylation function (Bai et ai,

1998). Similar loss of assembly was seen in the E35 stopper mutants of N.crassa which

were deleted of the ND2 and ND3 subunits (Alves and Videira, 1998).

b) The nuclear encoded subunits

NDUFAI ( M W F E ) , NDUFA2 (B8), NDUFA3 (B9), NDUFA4 (MLRQ) , NDUFA6

(8 M), NDUFA 7 (ASAT) , NDUFAB (PGIV), NDUFA9 (39 kDa), NDUFA 10 (42 kDa),

NDUFABZ (SDAP), NDUFB I (MNLL) , NDUFB2 (AGGG), NDUFB3 ( B 12), iVDUFB4

(B 1 5),lVDUFB5 (SGDH), NDUFB6 ( B 1 7), NDUFB7 ( B 18), NDUFBI (ASHI) ,

IVDUFB9 (B22), N D W I O (PDSW), NDUFCI (KFY 1), iVDUFC2 (MMTG), :VDUFS7

(PSST) and NDUFS8 (TYKY) (Table 1.2) are nuclear genes that encode for subunits in

the HP fraction of complex 1. So far, the genomic organization has only been determined

for NDUFAI (Zhuchenko et ai, 1996), NDUFA6 (Dunham et ai, 1999), NDUFB9 (Lin a

ai, 1999) and NDUFSS (de Sury et al, 1998).

Two consensus tetranuclear Fe-S clusters are present in the ïYKY subunit (Dupuis et

a/, 199 1 a), and a consensus binuclear Iron-sulfur pattern is found in the PSST (Arizmendi

et ai, 1992) and PGN (Dupuis et al, 199 1 b) subunits. The 10 kDa SDAP subunit is

known to have a phosphopantetheine attachment site unique to ac yl-carrying proteins

(Runswick et al, 199 l ) , as well as an EF-hand calcium binding domain (Triepels et al,

1999a).

ffiowledge of the functionality of other subunits in the HP fraction is scarce.

However, important information is emerging as more of these subunits are being

characterized. For example, the expression of the B 17 subunit was found to be highest in

kidney, indicating that the composition of complex I may be tissue specific and that

mutations in the NDUFB6 gene rnay cause distinctive phenotypes (Smeitink et al,

1998b). This clairn is also supported by the expression of subunits B8, B 12 and B 14, al1

of which were not found in the aorta, brain or kidney (Ton et al, 1997). Au et al ( 1999)

demonstrated that the MWFE subunit is essential for mammalian complex 1 activity by

showing that complex 1 activity was severely reduced (40%) in a MWFE mutant

Chinese hamster ce11 line. The phosphorylation of MWFE as well as the 42 kDa subunit

has been described (Sardanelli et al, 1995). The 822 subunit is considered a candidate for

BOR (branchie -oto-renal) syndrome because the NDUFB9 gene has been mapped to a 1 -

Mb deletion at chromosome 8q13 which also contains the gene for BOR syndrome (Gu n

al, 1996). Thus far however, mutational analysis of BOR families have yielded no

association between the lVDUFB9 gene and BOR syndrome (Lin et al, 1999). The 39 kDa

subunit related to hydroxysteroid reductase/isomerase was found to contain a NAD(P)-

binding motif and is proposed io participate in intramitochondrial fatty acid synthesis,

together with the acyl carrier protein (SDAP) of complex 1 (Friedrich et al, 1995;

Yamaguchi et al, 1998; Schulte et al, 1999). The cDNA sequence of the 42" complex 1

subunit has recently been deduced from both bovine and human heart mitochondria and

placed in the HP fraction (Skehel et al, 1998; Triepels et al, 2000). Electrospray mass

spectrornetry perfonned on complex 1 and two related subcomplexes in bovine heart,

identified a 17.2 kDa subunit (B 17.2) with an acetylated a-amino group (Skehel et al,

1998). S ince this protein shows 70% arnino acid identity to a 1 3 kDa human protein

associated with differentiation, it is hypothesized that this 1 3 kDa protein is part of the

B17.2 kDa homologue in human complex I (Skehel et al, 1998).

Electrospray ionization mass spectroscopy experiments carried out on complex 1 have

from time to time identified a 10566 Da protein (Fearnley et al, 1994). Sequencing of this

protein would most likely result in the identification of the 43' and final subunit of

complex 1 (Skehel et al, 1998).

Structural Mode1 of Complex 1

Vital to the understanding of the operation of this large, multi-functional enzyme of

the respiratory chah is the orientation and interaction of its subunits within the

mitochondrial imer membrane. Most of the information pertaining to the composition,

spatial and functional organization of cornplex 1 has been obtained from Bos taurus heart,

bacteria and fungi (GngorieR, 1999). Bovine heart complex 1, like that of Neurosporu

crarsa is an L-shaped enzyme with one arm (the extrinsic membrane domain) extending

into the mitochondrial matrix and another arm (the intrinsic membrane domain), which

remains in the membrane (Guenebaut et al, 1997; Gngoneff et al, 1998). The appearance

of the bovine complex diffen From that of N.crassa by possessing a significantly bigger

membrane-bound globular domain and also by having a thin staik region (30 diameter)

linking this globular ami with the intrinsic membrane domain like bacterial complex 1 (Fig.

1.7) (Grigorieff et al, 1998). The stak is thought to contain part of the electron transfer

Figure 1.7. Three dimensional models of complex 1 fkom E. coü, N. crassa and B. taunis as determined by electron cryo-rnicroscopy. A reconstruction of complex 1 from (a) E. coli (b) N. crassa and (c) B. taunrr is shown as determined by the electron cryo-microscopy studies of Guenebaut (1 998) and Grigoneff (1 998). Al1 three models share the L-shaped structure with an inûinsic membrane arm extending into the lipid bilayer and a peripheral arm promiding into the maaix. Additional protein mass is observed around the mitochondrial complexes when compared to complex 1 from E. coli, especially at the junction between the amis and around the membrane domain. The E. coli and bovine structures both show a narrowing between their membrane matrix domains (the s a ) .

pathway linking the NADH binding site in the globular am with the ubiquinone binding

site in the membrane domain (Gngorieff et al, 1998).

(i) Assembly of Complex 1

The assembly of this enzyme in N crassa occurs by the formation of a series of

intermediates (Tuschen et al, 1990). The matrix arm and the membrane portion of

complex 1 form independently and are joined in the course of assembly (Tuschen er al.

1990). The membrane arm is formed by the association of a 200 kDa and 350 kDa

assembly intermediates (Kumier et al, 1998). The larger 350 kDa assembly intermediate

is associated with two extra proteins of 80 and 30 kDa which are not constituent parts of

the mature complex 1 and are called complex 1 intermediate associated proteins. CIA84

and CIA30, respectively (Kuffner et al. 1998). These two proteins are considered single-

target chaperones specific for complrx I because they are integrated into the large

membrane arm during an early stage of complex 1 assembly, but are set fiee when the

complex has been assembled, only to be cycled once again to take part in the formation of

M e r intermediates (Kuffher et al, 1998). Deletion of the cia genes coding for these

proteins results in the severe disniption of the assembly process whereby the large

membrane arm intermediate is not formed, even though the matrix arm and the small

membrane arm intermediate are not affected (Kuffher et al, 1998). Homologues to the

CIA proteins have not been detected in prokaryotic genomes and no such chaperones are

yet known for other eukaryotic complex 1 enzymes.

(ii) Spatial organization and subunit interaction in human complex I

The use of strong but non-denaturing detergents such as LDAO (Iauryldimethylamine

oxide) has been used to resolve bovine mitochondrial complex I into two subcomplexes

called la and 1 P (Fig. 1.8) (Fine1 et al, 1992). Subcomplex Ia is composed of - 23

mostly hydrophilic subunits and contains al1 the redox centers. It is an active

NADH:dehydrogenase which can transfer electrons to a water soluble ubiquinone and is

likely to represent the peripheral ann and part of the membrane domain of the enzyme

(Fine1 et al, 1992). Subcomplex IP is made up of 17 mainly hydrophobic subunits

conesponding to the intnnsic membrane domain of complex 1 and has no known enzyme

activity (Finel et al, 1992).

Enzymatically active subcomplexes called IL. IS and IAS (Fig. 1.8) have also been

purified by sucrose gradient centrifugation in the presence of detergents (Fine1 et al.

1994). Al1 subcomplexes retained an NADH-oxidizing activity with ferricyanide or Q- 1

as acceptors, but lost activity with decylubiquinone as acceptor dong with a loss of

rotenone sensitivity during Q-1 reductase activity (Fine1 et al, 1994). Ih comprises 14 of

the subunits in subcomplex Ia and represents most or al1 of the peripheral arm of

complex I (Fine1 et al, 1994). The subcomplex ILS contains only approximately 13

subunits and like Ih, retains al1 the nuclear encoded subunits, whose homologues are

present in the bacterial genome and which are known to bind FMN and Fe-S clusters

(75-,51-, 49-, 30-, 24- Da, TYKY and PSST subunits) (Fine1 et al, 1994). n i e IS has

more subunits including the very hydrophobic ND4, but there are also subunits present in

MATRIX

Figure 1.8. Model of the overlap in subunit composition between the different subcomplexes of cornplex 1. Detergent ûeatrnent of complex 1 yields three major subcomplexes, the Ia Fraction which connibutes mainly to the penpheral am. the IP fraction and the smaller Iy fraction which mainly contribute to the membrane arm of the enzyme. The location of the smaller iS, 1A and IAS as well as the IPS and IPL subcomplexes are also shown. Subcomplexes that contain identical subunits are drawn above each other. IM refers to the inner mitochondrial membrane,

subcomplex 1 h that are missing fiom IS (Finel et al, 1994). None of the mitochondnally

encoded subunits have been found in subcomplexes Ia, Ih or IhS (Finel et al, 1994). This

suggests that most of the redox chemistry of complex I occurs in a hydrophilic domain on

the inside (mitochondrial matrix) of the membrane, with special sûucîural arrangements

required for proton translocation (Fine1 et al, 1994). The mitochondnally encoded ND

subunits may play a role in the formation of these structures (Fine1 et al, 1994).

Recently, a previously undetected fiagrnent referred to as Iy has been found by using

LDAO to disrupt complex 1 (Sazanov et al, 2000). This subcomplex has been ascertained

to contain the hydrophobie subunits ND 1, ND2, ND3 and ND4L, al1 of which are

mitochondrially encoded and the nuclear encoded KFYI subunit. Also, the Iy subcomplex

has been seen to release a fragment containing ND 1 and ND2 and the I P subcomplex has

been found to dissociate further to form IPL (containing ND4 and ND5) and IPS

(containing the rest of the Ip subunits) (Sazanov et al, 2000). Important associations

illustrated by this study also include the somewhat loose association of the 42 and 39

kDa subunits with Iy and the presence of the 39 kDa and 15 kDa (IP) subunits in both Iy

and la subcomplexes (Sazanov et al, 2000).

Other reports that give information on the special organization of complex 1 subunits

include findings by Han et al (1989) that the 75 kDa subunit exists on both sides of the

inner mitochondrial membrane and those by Patel et al (1988) which show that the 49

kDa and 30 kDa subunits are exposed to the intermembrane space. Determination of the

crystal structure of complex 1 in Neurospora crassa and imrnunolabeling the 49 kDa

subunit, has pinpointed it to the matrix-localized protruding arm of the complex

(Guenebaut et al, 1997). This is in agreement with its position in the E. coli cornplex 1.

where it has been localized to the so-called connecting hgment which links the peripheral

a m to the membrane arm (Friedrich et al, 1995; Leif et al, 1995).

Energy conversion in cornplex 1

The main fùnction of complex 1 is to transport electrons by oxidation of NADH

followed by reduction of ubiquinone. This process is accompanied by the translocation

of protons fiom the mitochondrial matrix to the intermembrane space.

(i) Iron-sulphur clusters, Flavin and Semiquinones

Several prosthetic groups catalyze these electron transport reactions, which are

hypothesized to be at least one Flavin Mononucleotide (FMN), six to eight Iron-Sulîùr

clusters (Albracht et al, 1997) and one or two species of semiquinone (de Jong er al, 1994:

Vinogradov et al, 1995).

The two types of iron-sulfur clusten encountered in complex 1 are [zF~-~s](''. '+ ' and

[JF~-JS](''. "' (Ohnishi et al, 1998). A binuclear cluster is made up of two iron atoms

which are bndged by two acid-labile inorganic sulfides and ligated to four cysteinyl sulfur

from the polypeptide chain of the apoprotein. Each iron is tetrahedrally coordinated to

two acid labile sulfur and two cysteinyl sulfur atoms (Ohinishi. 1998). A tetranuclear

cluster contains four iron atoms and four acid-labile sulfides arranged in a distorted cube

structure with the iron atoms bound to the polypeptide via four cysteine sulfur ligands

(Ohinishi, 1998). An important indication to the total predicted number of iron-sulhr

clusters in complex 1 and their possible subunit locations cornes from the hlly conserved

sequence motifs (Fearnley et al, 1992; Weidner et al, 1993). EPR analyses has led to the

spectral resolution of only six distinct bon-sulfur clusters, namely clusters N 1 a, N 1 b, N2.

N3, N4 and N5 associated with complex I so far (Ohinishi, 1998). While the 14 subunit

NADH:ubiquinone reductase of E.coZi contains one FMN and up to 9 Fe-S clusten as

prosthetic groups (Braun et al, 1998), mammalian compiex 1 contains at least four (4Fe-

4S] clusters (N-2, N-3, N-4, N-5) and two [2Fe-2S] clusters (N-la, N-l b). The 5 1 kDa

FP subunit contains an FMN and a tetranuclear iron-sulhir cluster (N-3). The 24 D a FP

subunit contains a binucIear iron-sulfûr cluster. The 75 kDa IP subunit contains a

tetranuclear (N-4) cluster, a binuclear cluster and probably another tetranuclear cluster.

The TYKY subunit contains eight conserved cysteine motifs for housing two tetranuclear

clusters, while the PSST subunit has a motif of 3 cysteine residues that could ligate

another tetranuclear cluster. However, the structural similarity of TYKY to bacterial low

potential feredoxins which cary two tetranuclear clusters in close vicinity, makes this

subunit a less likely candidate for carrying the N-2 cluster (Brandt, 1997). Although

other subunits with conserved cysteine residues such as the PGIV (Dupuis er ai, 199 1 ).

49- and 30-kDa (Preis et al, 1990; Feamiey and Walker. 1992) subunits have been

suggested to bind to Fe-S clusten, more recent studies show that it is unlikely (Fine1 et ai.

1 994). Although, the exact subunit location of the various iron-sulfùr cl usters is not

conclusively known, table 1.3 illustrating the two existing models, surnmarizes current

knowledge on this issue.

The non-covalently bound flavin of complex 1 can assume three different redox States.

namely, fully oxidized, semiflavîn and Mly reduced (Leif et al, 1995: Finel, 1993). The

oxidized form of flavin has four orders of magnitude higher affinity to its specific binding

site that its fully reduced form (Leif et al, 1995). A strong spin-spin interaction exists

between the semiflavin and iron sulfür cluster N3, both of which are located in the 5 1 kDa

subunit (Sled et al. 1994).

Table 1.3. Current hypotheses on the subunit location of FMN and iron-sulphur (Fe-S) clusters in the minimal nuclear-encoded functional unit of bovine heart Complex 1

Mode1 1 Mode1 2 (Albracht and de Jong, 1997) (Ohnishi, 1998)

Name of subunit

75 kDa 51 D a 24 kDa TYKY PSST

Prosthetic groups Prosthetic groups

Discovered in activated bovine heart submitochondnal particles (SMP). semiquinones

have also been a source of much debate (Suzuki and King, 1983). Allhough certain groups

have concluded that there is oniy a single species of semiquinone (SQ) (de Jong et al.

1994), rnost other researchers agree that two distinct species of semiquinone (SQNI and

S a r ) exist with fast and slow spin relaxation behaviour, respectively. SQNf is stronglp

spin-coupled with cluster N2 and Sas is located closer to the cytosolic side membrane

surface and is weakly spin-coupled with cluster N2. A thrd semi-quinone, undetectable

by EPR has aiso been hypothesized (Ohnishi, 1998; Friedrich et al, 1998).

(ii) Electron transfer in complex 1

NADH, the electron donor of complex 1 binds to the FMN-containing 5 1 kDa subunit

(Deng et al, 1990). It is thought that FMN, because it can take up two electrons and

release them one at a tirne to one-electron acceptor such as an Fe-S cluster, is the

irnrnediate oxidant of NADH (Ragan, 1987; Weiss et al, 19%). A recent 3 2 ~

photolabeling study by Yamaguchi and CO-workers (2000) with purified berf heart

complex I showed that the 30 kDa and 42 kDa subunits, in addition to the 5 1 kDa subunit

were capable of binding NADH. The 39 kDa subunit as well as a 18-10 kDa subunit

were shown to bind NADP(H) (Yamaguchi et al, 3000). Hoviever. since there is no

evidence that complex 1 contains any flavin other than the FMN bound to the 51 kDa

subunit. it is not likely that these other subunits with nicotinarnide nucleotide binding

capabilities are acnially involved in binding reducing equivalents in vivo (Yamaguchi er al.

2000).

A strong homology between the 5 1- (Pilkington et al, 199 1 ) and 24-kDa (Tran-Betcke

er al. 1990) subunits of complex I with the a-subunit of the NAD+-reducing hydrogenase

of Alcaligenes eunophus and high homology between the N-terminal haif of the 75 kDa

subunit of complex I and the y-subunit of the same enzyme gives some insight into

electron flow in this enzyme. The homology of the PSST subunit to the 6 subunit of the

hydrogen-hydrogenase complex of Alcaligenes eurrophur, suggests the association of the

PSST subunit with the 75 kDa Fe-S protein (Robinson, 1993). Taken together. the

following scheme of electron flow between the iron-sulfur clusters of complex I c m be

envisioned: NADH -> 5 1 and 24 kDa of FP -> 75 kDa of IP -> 23 (TYKY) and 20 kDa

(PSST) of HP (Belognidov and Hatefi, 1996). The proximity of the FP and IP subuniü

of complex 1 was determined by Hatefi et al (1 993) through cross-linking experiments

(Fig. 1.9) and supports this model of electron transfer between the subunits.

It is known that electron transfer between FeS clusters c m easily occur over distances

of 1 .O to 1.5 nm (Onuchic et al, 1992). The pH dependent midpoint potential (E,) value

of cluster N2 is the highest among al1 iron-sulfur clusters in complex 1 and it's one

electron reduction or oxidation is coupled with the binding and release of one proton in

the physiological range (Ingledew and Ohnishi, 1980). Cluster N2 has also been found

to be only 8-1 1 A apart fkorn one of the EPR-detectable ubiserniquinones (Vinogradov er

al. 1995) and therefore has been assumed to serve as the electron donor to ubiquinone and

to be intimately linked to the proton translocation mechanism of the enzyme (de Jong and

Albracht, 1994).

The quinone binding site itself within complex 1 has been a matter of some debate.

While it is now widely accepted that there are at least two ubiquinone binding sites

(Vinogradov, 1993; Degli Esposti and Ghelli, 1994), their location within the complex has

not been clearly defined. Darrouzet et al ( 1 998) have shown through studies on

piericidin-resistant mutants of the bacterium Rhodobacter capstilatus that the 49 kDa

subunit is associated with the resistance and therefore involved in quinone binding. Their

model proposes that the ubiquinone binding sites are Iocated at the interface between the

membranous and peripheral domains of complex 1, whereby the polar cyclic head of the

quinone is bome by the 49 kDa subunit and the hydrophobic isoprenyl tail is bome by

the hydrophobic ND 1 subunit (Darrouzet et al, 1998). This model attempts to

accommodate the hdings that ND 1 and ND4 subunits may be involved in quinone

binding, as pathogenic mutations in these subunits causing complex 1 deficiency seem to

also slightly alter the interaction of the complex with quinones and its sensitivity to

rotenone (Majander et a/, 1991; Degli Esposti and Ghelli, 1994). Earlier photolabeling

studies on isolated complex I had actually pointed to the ND 1 subunit as a candidate

subunit for the isoprenyl tail binding site (Earley et ai, 1987). The latest studies however.

dispute this claim by showing through photolabeling studies on mitochondria1 electron

transport particles that it is the PSST subunit which couples electron transfer from its

iron-sulfur cluster N2 to quinone (Schuler et al, 1999). Schuler and colleagues showed

that the ND 1 subunit is probably not directly involved in quinone binding because

complex I inhibitors including rotenone were not able to prevent photoaffmity labeling of

this subunit at the same levels that totally blocked labeling of the high affinity PSST site.

This is substantiated by the fact that an isolated peripheral fragment of complex 1. devoid

of all the ND (mitochondrially encoded) subunits, is able to catalyze NADH-quinone

oxidoreduction (Friedrich et al, 1989; Finel et al, 1992). Recent work based on a

structural element containing a weak sequence motif that is common to the quinone sites

of bacteria, mitochondria1 bcl and photosystem I, points to the ND4 and ND5 subunits

as candidates for the locations of Q sites in complex I, either as a pair harbouring two

separate sites or as a pair forming a single site (Fisher and Rich, 2000). The two likely

candidates for quinone binding at this point in study seem to be the 49 kDa subunit of the

IP Fraction and PSST subunit of the HP fraction.

INTRINSIC ARiM

Figure 1.9. Structural mode1 of complex 1. The complex is portrayed as an elbow-shaped entity with a membrane bound intnnsic ann and an extrinsic ann which protmdes into the m a t r i The nurnbers inside the subunits represent their rnolecular rnass. NADH binds to the 51 kDa flavoprotein subunit containhg a 4Fe-4s center and electrons are passed through the 24 kDa, 75 kDa, TYKY and PSST iron sulphur proteins. Two quinone binding sites are postulated. The mtDNA encoded (ND) subunits are al1 part of the membrane arm of the complex and some of them are possibly invotved in proton pumping. rE, indicates the subunits that house the redox groups.

(iii) Models for coupling electron flow with proton translocation

The oxidation of one NADH by complex 1 is now considered to be linked to the

translocation of 4H ' per 2e' (Galkin et al, 1999). Electron transfer in proteins occurs

through either tunneling over large distances (up to 20 A) From one redox center to the

next or transfer dong covalent bonds (Moser et al, 1995).

Although many mechanistic models have been proposed over the years, none of them

have hlly taken into consideration factors such as the existence of cornplex 1 as a

multiprotein enzyme (Models by Mitchell, 1966 and Lawford and Garland, 1972), its

4W/2e' stoichiometry (Model by Suzuki and King, 1983), the Em7 potential of cluster N-

2 being -150 mV in bovine complex I and the pK, values of FMN and FMNH? (Model

by Knshnamoorthy and Hinkle, 1988), thennodynamic requirernents (Ragan. 1990),

reoxidation of the intemal ubihydroquinone by cluster N-2 (Model by Weiss et al. 199 1 ).

movement of 'translocating' semiquinones across the membrane and protonatiod

deprotonation occurring at opposite sides of the membrane (Model by Vinogradov.

1993), contributions to the proton translocation process by the electron input part of

complex 1 containhg FMN and the Fe-S clusters and other thermodynamic problems

associated with the dual Q-gated pump (Model by Degli Esposti and Ghelli, 1994). The

proposed hypotheses on the energy transducing mechanism of complex 1 can be roughly

divided into two types of either direct (Brandt, 1997; Dutton et al, 1998) or indirect

(Belogrudov and Hatefi, 1994; Yagi et al, 1998) coupling between electron transport and

proton translocation. Therefore, only the two most recent hypotheses representing each

mode1 will be discussed.

The direct coupling model proposed by Dutton et al (1998) is based on the Q-cycle

with a few variations to the aiready established rnechanism for complex III. Their mode1

proposes two quinone binding sites, Q, and Q,, that can exchange Q/QH2 with the

membrane pool (Qpool), with the Q, site having access to the protons on the matrix side

of the membrane and the Q, having access to protons on the cytosolic side of the

membrane. In addition. a non-pool exchangeable Q,, occupies a site that c m assume

either of two different conformations between the other quinone binding sites and ma.

even be covalently bound. One conformation provides access to the protons on the

rnatrix side of the membrane, presumably through a channel or a pore. The other

confomation provides access to protons on the cytosolic side of the membrane. This Q,,

site may very well be the novel redox group detected by Weiss's group (1997) within the

membrane ann of cornplex using ubiquinone- 10 depleted complex 1 purified fiom

iYcrassa. Only quinones are used in this model to manipulate proton motion (although

flavins are equally capable of binding and releasing protons, their localization in the

transmembrane domain seems to make it quite unlikely) (Dutton et al, 1998).

According to the model by Dutton et al (1998) (Fig. 1.10), electron transfer from

NADH in the matrix to the N-2 F e 4 cluster close to the matrix-membrane interface.

proceeds through the large nurnber of flavin and Fe-S redox cofacton forming a long

transfer chain. As N-2 is reduced, it injects a single electron into a Q drawn fiom the

membrane pool into the nearby Q, site, generating an unstable transition state

semiquinone (SQ) (Dutton et al, 1998). The SQ at the Q, site near the matnx acts as a

strong oxidant to pull two electronic charges across the membrane via a Q at the QnY site

Figure. 1.10. A hypothetical model for direct energy conversion in complex 1. Complex 1 is represented as a tranmembrane protein with a NADH, N N and iron-sulphur complex which deliven electrons to the [4Fe4S] cluster N2. Quinone binding sites are Q, and Qnx; Qpool is the membrane pool: Qny is a non-

pool exchangeable quinone; SQ represents semiquinone. The sequence proceeds as foilows: 1 ) NADH arrives at Complex 1, Qnz and Qnx sites can exchange with pool 2) Reduction of N2 occurs by the NADH subcompiex 3) N2 reduces a Q drawn from Qpool into Q,, forming an unstable. transition SQ 4) The SQ

in the Q,, site oxidizes the QH;! fixed in the Qny site. SQ of the Qnz site is reduced to QH7 a drawing two

protons fiom the mauix and QH2 of Qny site is oxidized to SQ releasing protons to the cytosol5) QH7 - in the Q,,, site adopt a geometry with access to cytosolic protons and the newly formed SQ assumes a geomeuy

with access to protons in the matrix 6) QH2 in the Q, site can reduce the SQ in the Qny site back to QH7 - and protons fmm the channe1 to matrix are bound and protons are released from the Qnx site. The top half of

the figure shows that as the QH2 is oxidized in the Qnx site two protons are released into the cytosol leaving a SQ anion. No further protons are released from Q, in the bottom half of the figure The reduced Qny site

assumes the original geometry and Qnz exchanges with the pool and steps 2-6 are repeated. Adapted from

Dutton et al, 1998.

frorn the Q,, site near the cytosolic side (Dutton et al, 1998). The QH2 in the Q, site is

therefore oxidized by the SQ in the Q, site. Thereby, as the SQ of the Q, site is reduced

to QHz, it binds 2 protons, ultimately drawn fiom the rnatx-ix (Dutton et al, 1998). As the

QH2 of the Qny site is oxidized to SQ, one (or two) protons are released to the cytosolic

channel (Dutton et al. 1998). While QH2 in the Q,, site is favoured to adopt a

conformation with access to cytosolic protons, the newly formed SQ rapidly assumes a

conformation with access to protons in the matrix (Dutton et al, 1998). In this geometry.

the QH2 in the Q,, site can reduce the SQ in the Qny site back to the QH2 and in the

process one or two protons from the channel to matrix are bound (Dunon r i al. 1998).

At the same time. one or two protons are released from the Q,, site (Dutton et al. 1998).

The intervening Qny site therefore acts as a proton pumping element, quite like the proton

pump of complex IV (Dutton er al, 1998). The overall reaction describes two elrctrons

camied by complex I frorn NADH to Q catalyzing the translocation of 4 or 6 proton

charges fiom the mitochondnal matrix to cytosol (Dunon et al, 1998). The

experirnentally observed SQ States QNf and ai, mentioned before (Vinogradov et al.

1995) are said to correspond to Q,, and Q,, respectively (Dutton et al. 1998). This

mode1 also supports the idea of two clear classes of inhibitors corresponding to the

difisible quinone binding sites Q, and Q, as has already been suggested by some

experimental studies (Friedrich et al, 1994).

The indirect coupling hypothesis put forth by Belognidov and Hatefi (1994) is based

on anaiogy to the ATP synthase complex in which there is no proton camer and energy

coupling between the catalytic and the proton-translocating sectors appears to take place

via conformational changes of the subunits. This mechanism is supported by their

findings that the proximities of the three subunits of FP to one another, the proximity of

the 51 kDa subunit of FP to the 75 kDa subunit of 1P and the proximities of al1 of the IP

subunits to one another and to some of the HP subunits are altered when the catalytic

sector of complex I is reduced by NADH or NADPH (Belogrudov and Hatefi, 1994).

That is, the proximity changes observed fiom NAD(P)H treatment of complex I involved

not only those subunits that contain electron carriers but also those that are devoid of

them (Belogrudov and Hatefi, 1994). Conformational changes beginning with the

reduction of FP, transmitted to subunits of IP and finally to some of the subunits of the

membrane sector of complex I (HP) serve as the device by which the energy derived frorn

electron transfer through the catalytic components of complex 1 is transduced and

conveyed to the subunits of the membrane sector (Belogmdov and Hatefi, 1994). The

resulting changes in the pK, of appropriate residues induced by these conformational

changes lead to proton uptake and release on opposite sides of the membrane (Belogrudov

and Hatefi, 1994).

Recent cryo-electron crystallization of two subcomplexes of bovine complex f lends

support to both mechanisms of energy transduction (Sazanov and Walker, 2000). The

large distances seen between subunits ND5 and ND4 (where proton pumping is thought

to take place) and the Ia subcomplex (where NADH oxidation and electron transport

reactions occur) argue in favour of the indirect mechanisrn of proton translocation

invo lving long-range conformational changes in the enzyme (Sazanov and Walker. 2000).

On the other hand, the ND2 (proton-translocating) and ND 1 (ubiquinone-binding)

subunits seen in the Iy subcomplex could participate in a direct coupling mechanism

(Sazanov and Walker, 2000).

Complex 1 Inhibitors

The sixty or so different families of compounds known to inhibit complex 1 include

those of both natural and commercial origin (reviewed by Degli Esposti. 1998). The two

most potent natural inhibiton of complex 1 are mtenone and piencidin. Rotenone. the

classical inhibitor of complex 1 belongs to a fmily of isoflavonoids extracted from

Leguminosae plants (Degli Esposti, 1998). Rotenone inhibition is time-dependent with a

K, as low as I nM being reported recentiy (Grivennikova et al, 1997). Piericidins. a group

of potent complex I inhibitors are 2,3-dimethoxy-4-hydroxy-5-methy-6-polyprenyI-

pyridine antibiotics produced by some Streptomyces strains (Miyoshi. 1998). Their close

sirnilarity to ubiquinones renders Piencidin A to be an effective inhibitor of complex 1

with a K, ranging between 0.6 and 1 nM (Degli Esposti, 1998). Cornplex 1 inhibitors are

thought to bind at or close to quinone binding sites and lunetic studies have suggested that

these inhibitors can be grouped into three classes, narnely rotenone, piericidin A and

capsaicin (Degli Esposti et al, 1994). Capsaicin which is part of the vanilloid farnily is a

pungent extract from red peppers and although it is a relatively weak inhibitor of compiex

1 compared to other natural products, it shows the rare property of competitive inhibition

venus quinone substrates in complex 1 (Degli Esposti, 1998). A study by Okun et al in

1999 showed through two independent approaches that al1 tested hydrophobic inhibitors

shared a common binding dornain with partial overlapping sites. While the rotenone site

overlaps with both the piericidin A and capsaicin site, the latter two sites were found not

to overlap (Okun el al, 1999). Other natural inhibitors of complex 1 are annonaceous

acetogenins such as rolliniastatin, myxobactenai antibiotics such as stigmatellin, and

antibiotics such as cochlioquinone B (Degli Esposti, 1998). Synthetic and commercial

inhibiton of complex I include various short-chah ubiquinones, pesticides. drugs.

neurotoxins and fluorescent dyes (Degli Esposti. 1998).

Complex II: The Succinate-ubiouinone oxidoreductase complex

Complex II refers generaily to succinate-ubiquinone oxidoreductase (SQR) or

menaquinol-fumarate oxidoreductase (QFR), membrane-bound enzyme complexes which

are alike in both structure and function (Maklashina et al. 1999). While SQR couples the

oxidation of succinate to fumarate in the Krebs cycle and the subsequent reduction of

ubiquinone in aerobic organîsms, QFR oxidizes menaquinol and reduces fumarate to

succinate as part of an anaerobic respiratory chah (Maklashina et al, 1999). The SQR

complex does not translocate protons, and therefore only feeds electrons to the electron

transport chah (Hagerhall, 1997; Ackrell et al, 2000). It consists of a penpherai domain.

exposed to the matrix in mitochondria and a membrane-integrai domain that vans the

membrane (Hagerhall, 1997). The penpheral part, which contains the dicarboxylate

binding site, is composed of a flavoprotein (FP; 64-79 kDa) subunit, with one covalently

bound FAD, and an iron-sullùr protein (IP; 27-3 1 D a ) subunit containing three iron-

sulfur clusters, a [2Fe-2S] cluster denoted S 1, a [4Fe-4S] cluster denoted S2 and a [3Fe-

4S] cluster denoted S3 (Ackrell et ai, 2000). The membrane-integral domain functions to

anchor the FP and IP subunits to the membrane and is required for quinone reduction and

oxidation (Ackrell et al, 2000). The anchor domain shows variability in composition and

primary sequence. It consists of one larger or two smaller hydrophobic peptides and

either one or two protoheme IX groups, with hexa-coordinated iron. or no heme at al1

(Ackrell et al. 2000). Inhibitors of this complex include Theonyltrifluoro- acetone

(TTFA), 3'-methyl carboxin and 2-n-heptyl-4-hydroxyquinoline N-oxide (HQNO)

(Hagerhall, 1997). Complex II is the only one of the respiratory chain enzymes which

has the same general composition in pro- and eukaryotes (Ackrell et al. 2000). This is in

contrast to complexes 1, III and IV of higher organisms which have various numben of

subunits present in addition to the "minimal functionai units'' found in their prokaryotic

counterparts.

Corn~ lex III: The Ubiauinol-ferricytochrome c oxidoreductase cornalex

The second proton translocation site within the electron transport chain is at complex

111 or the QH2-cytochrome c reductase complex. The reaction catalyzed by complex III is:

QH2 + 2H? + 2 cyt c, - - -, Q + 4H+ + 2 cyt crcd, where Q represents ubiquinone

(Voet and Voet, 1990). Recently, the complete crystal structure of the cytochrome b q

complex from bovine heart mitochondria was deterrnined (Iwata et ai, 1998). Mammalian

complex III is composed of eleven subunits with known amino acid sequences (Yu et al.

1998). Bacterial homologs are found of ody the three of the subunits and these are the

subunits that cary redox centres (Saraste, 1999). The key redox components are

subunits with two b cytochromes (b565 and b562), which are mitochondrially encoded

and have eight trammembrane helices with IWO heme b groups sandwiched between

helices B and D; a membrane-anchored cytochrome ci and a membrane-anchored FeS

protein canyinp a Rieske-type center (Fe&) and a ubiquinone (Saraste. 1999). Most of

the other eight subunits are small proteins that surround the metalloprotein nucleus. but

the two largest membrane spanning proteins in the complex, subunits 1 and II termed

"core proteins" face the mitochondrial matrix and are homologous to mitochondrial

processing peptidases which function in protein import (Iwata et al. 1998). Thus.

complex III, which exists as a rnacromolecular dimer within the membrane in bovine heart

mitochondria, may be multifùnctional (Iwata et al. 1 998).

Coupling the redox reaction to the generation of a proton gradient across the

membrane is performed by a mechanism called the Q cycle (Yu er al, 1998). According to

this model. one electron is transferred from the ubiquinol at the Q, site (near the

cytoplasmic side of the inner rnitochondrial membrane) to the Rieske iron-sulfur centre

and then to cytochrome c via cytochrome cl (Yu et al, 1998). The ubisemiquinone which

is generated then reduces cytochrome b566 heme which in tum transfers an electron to the

cytochrome b562 heme (Yu et al, 1998). A ubiquinone or ubisemiquinone bound at the Q,

site (near the matrix side of the membrane) then oxidizes the reduced b(562) heme (Yu et

al. 1998). Proton translocation is therefore the result of deprotonation of ubiquinol at

the Q, site and protonation of ubiquinol at the Qi site (Yu et al, 1998)..

Specific inhibitors of the complex include antimycin, myxothiazol and stigmatellin (Yu

et al, 1998). Patients with complex III deficiency ofken show lactic acidosis, muscle

weakness, atrwa, exercise intolerance, ocular myopathy or a multisystem disorder

(Kennaway, 1988; Slipetz et al, 199 1).

C o m ~ l e x IV: The cvtochrome c oxidase com~lex

Cytochrome c oxidase (COX) or complex IV is a Y-shaped complex which exists as a

large macromolecular dimer (Saraste, 1999). The enzyme catalyzes the following reaction:

4 cyt c,d + 4 W + O2 - --b 4 cyt cOX + 2H20 and it is the only non-equilibrium.

irreversible, step in oxidative phosphorylation (Voet and Voet. 1990). The crystal

structure of bovine heart COX at 2.8 A resolution was first obtained by Yoshikawa and

CO-workers (1998). This complex consists of 13 subunits. with the three largest subunits

which constitute the catalytic core of the complex, encoded by mtDNA and the ten

remaining subunits encoded by nuclear DNA (Robinson, 1998a). The core of this

complex is made up of the products of the COX 1, II and III genes which are

mitochondnally encoded (Robinson, l998a). Complex IV contains four redox centers. two

a-type hemes (a and a3) and two copper ions (CuA and Cus) (Saraste, 1999). The complex

IV substrate, cytochrome c. is a water-soluble hemoprotein that donates electrons on the

cytoplasmic side of the mitochondrial i ~ e r membrane (Saraste, 1999). Subunit II has a

dinuclear, mixed valence copper center CU^, which is the first site to receive electmns

fiom cytochrome c (reviewed by Capaldi. 1990). Subunit 1 contains the active site which

contains the heme iron and copper that are used to reduce O2 into two water molecuies

(reviewed by Capaldi, 1990). Electrons fiom subunit II are transferred to a low-spin

heme (cytochrome a) in subunit 1, and then to the bimetallic cytochrome a3/CuB active site

(reviewed by Capaldi, 1990). The protons needed for this reaction are taken from the

mitochondrial matrix side through two channels (reviewed by Calhoun et al, 1994). The

same channels are used to purnp one proton per electron across the membrane (reviewd

by Calhoun et al, 1994). Although subunit III has no prosthetic groups. it is vital to the

functioning of the other two core subunits (Robinson, 1998a). Inhibitors of complex IV

include cyanide. carbon monoxide, sulfide and azide which interact with the heme redox

centers (Tyler, 1992). Isolated cytochrome oxidase deficiency can be divided into five

basic categories, namely, Fatal infantile lacticacidemia. Classical Leigh's disease. Saguenay

Lac St Jean Leigh's disease, cardiomyopathy and mitochondrially encoded defects such as

Keams-Sayre. LHON and CPEO (Robinson, 1998a).

Cornplex V: The ATP svnthase

The proton motive force generated by the electron transport chain is utilized by the

K-ATP synthase to generate ATP from ADP and inorganic phosphate (Leslie er al.

1999). This 16 subunit mammalian complex appears as a mushroom shaped structure

with three parts designated F, (membrane associated), Fl(water soluble) and a stalk

(Leslie et al, 1999). This general structurai organization is in congruence with the crystal

structure of F1 -ATPase obtained at 2.8 A resolution from bovine heart mitochondria

(Abrahams et al, 1994). The F,-membrane associated component forms a channel through

which protons can traverse (Leslie et al, 1999). The electrochemical potential energy is

then utilized by the FI component, which faces the interior of the inner rnitochondrial

membrane to form ATP (Leslie et al, 1999). The FI headpiece is composed of five

subunits with the following stoichiometry:gB,y& (Leslie et al, 1999). The isolated F I

component which contains the catalytic nucleotide binding-site (P-subunits) is as active

as an ATPase and is therefore referred to as Fi-ATPase (Leslie el al. 1999). It is now

widely accepted that the three catalytic sites altemate between three different states -

open, loose and tight - which have differing affinities for nucleotides (Leslie et al. 1999).

Conformational changes required for inter-conversion behveen the three types of catalyric

sites is achieved by the rotation of the three catalytic P-subunits relative to the single

copy subunits y, 6 and E (Leslie et al, 1999). Efiapeptins and aurovertins are potent

inhibitors of the FI-ATPase catalytic sites (Leslie ei al. 1999). The F, and stalk domains

of the mammaiian structure are composed of eight subunits: a, b. cio, d, e. OSCP

(Oligomycin Sensitivity Conferring Protein), Factor A6 and A6L (Pedersen and Amzel.

1993). Subunits A and A6L are mitochondnally encoded while the remaining subunits are

nuclear encoded (Papa et al, 1992). Regulation is achieved by an inhibitory protein [FI.

which binds to the b-subunit in the absence of a proton gradient, thereby preventing the

ATPase complex From nuuiing in reverse and hydrolyzing ATP (reviewed by Pedersen

and Arnzel, 1993). Clinical consequences of defects in the ATP synthase complex are

best represented by NARP and Leigh's syndrome (reviewed by Schapira and Cock.

1 999).

Table 1.4. composition and genetic origin of mitochondrial respiratory chah (OXPHOS) subunits

Complex Subunits Nuclear Encoded mtDNA Encoded

7 WD1, ND2, ND3, ND4. ND4L. ND5, ND6) O 1 (cytochrome b) 3 (cytochrome oxidase 1.11 and III) 2 (ATPase 6 and 8)

Part III. Mitochondrial disorders

Tvoical symotoms of defects in enerw metabolism

A failure of mitochondria to produce ATP at normal rates because of genetic defects or

hypoxic situations leads to a rapid acceleration of glycogenolysis and glycolysis brought

about by increased ADP, AMP and inorganic phosphate concentrations (Robinson.

1998a). This increase is brought about through the initial defense of ATP levels through

the creatine phosphokinase and adeny late kinase equilibria (Robinson, 1 W8a). The

glycolytic response to a shortage of mitochondrial ATP production results in an increase

of lactic acid production (Robinson, 1998a). While the lactate production is increased.

inhibition of respiratory chain function alters the redox state of the ce11 to make it more

reduced in both the cytosoiic and mitochondrial compartments (Robinson, 1998). The

net resuit of this is a change in the ratio of lactate to pyruvate (LR ratio) both in the

intracellular and extracellular locations (Robinson, 1998a). Depending on the severity of

the defect, L/P ratios will be elevated from 10: 1 - 20: 1 in normal individuals to as much as

25: 1 - 40: 1 in affected individuals and to as much as 100: 1 in extreme cases (Robinson.

l998a). Collectively, patients with lacticacidemia represent about 1 in 5000 of the

population (Robinson, 1993).

One of the most common features seen in patients with defects in energy metabolism

is an increase in blood or CSF (cerebrospinal fluid) lactate (Robinson, 1998a). If lactic

acid levels are elevated there appears to be some correlation between the observed

elevation and the severity of symptoms in the patient (Robinson, 1998a). The three

major disorders of energy metabolism, namely, deficiency of the pyruvate dehydrogenase

complex. deficiency of NADH-ubiquinone oxidoreductase (complex 1) and deficiency of

cytochrome c oxidase (complex IV), seem to undertake the following progression of

seventy: Fatal Infantile Lactic Acidosis>Leigh disease>Psychomotor

Retardation>Ataxia>Retinal Degeneration (Robinson, 1993). These disorders are usually

nuclear in origin (Robinson et al, 1987; Robinson er 41. 1990; Tatuch er al. 1992). Defects

involving the oligomycin-sensitive ATPase (complex V) also follow this pattern. although

there are no patients with fatal neonatal lactic acidosis (Tatuch and Robinson. 1993). A

second set of symptoms that are cornmon and characteristic of maternally inherited

disorden (discussed below) include sensorineurd hearing loss, muscular weakness,

dementia and stroke-like episodes (Shofier et al, 1995; Robinson, 1 993). Other

symptoms include extemal opthalmoplegia, ptosis, exercise intolerance. myoclonus

epilepsy and hyperventricular cardiomyopathy (Robinson, 1998).

Mitochondr-ial Res~iratory Chain Diseases

Mitochondrial defects occur in a wide variety of degenerative diseases, errors in

metabolism, aging and cancer. Abnormalities in mitochondrial rnetabolism encompass

defects of fatty acid oxidation, tncarboxylic acid cycle enzymes as well as the respiratoty

chah and oxidative phosphorylation (OXPHOS) systems. Respiratory chain defects

represent the most cornmon biochemical deficiency of mitochondrial metabolism causing

diseases.

(i) MtDNA associated diseases

Various pathogenic mtDNA abnormalities have been identified. the most common

being (i) large-scale rearrangements (single deletions, multiple deletions, duplications). ( i i )

point mutations. insertions and deletions in transfer RNA (RNA), (iii) point mutations

in nbosomal rRNA (rRNA) and (iv) point mutations in protein encoding penes for

complex 1, cytochrome 6, Complex IV and Complex V (Hanna and Nelson. 1999).

Almost al1 of these disorders are characterized by mtDNA heteroplasmy. This is a state

characterized by mutated and wild-type genomes CO-existing in variable proportions and

with tissue-to-tissue and cell-to-cell differences within the sarne individual. The severity

of symptoms has been found to correlate to the percentage heteroplasmy in affecied

individuals with tissue heteroplasmy also being observed (Robinson. 1 998a). The

frequency of mtDNA point mutations is known to be approximately ten times that of

nuclear mutations. This has been attributed to the high level of oxygen free radicals

present in the mitochondria resulting fiom oxidative metabolism, the lack of adequate

DNA repair enzymes in the mitochondna and the absence of protective histone proteins

(Brown and Wallace, 1994). The various primary defects in mtDNA described above can

be found in association with many different clinical phenotypes such as Kearns-Sayre

syndrome (KSS), chronic progressive external ophthalmoplegia (CPEO), mitochondrial

encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS). my oclonus

epilepsy with ragged red fibres (MERRF), mitochondrial neurogastrointestinal

encephalomyopathy (MNGIE), neurogenic muscular weakness, ataxia and retinitis

pigmentosa (NARP), Lebers hereditary optic neuropathy (LHON) and Leigh's syndrome

(Brown and Wallace, 1994). While some mtDNA defects associate with multisystem

disorden such as KSS, MELAS or NARP, others are surprisingly tissue specific such as

that seen in LHON where the disease is confined to just the optic nerve (Brown and

Wallace, 1994).

(ii) Nuclear DNA associated diseases

Little is known about the nuclear gene defects that cause respiratory chain

dysfunction. The first nuclear gene defect reported to cause respiratory chain dysfunction

was in the flavoprotein subunit of complex 11, identified in two siblings with Leigh's

syndrome and severe deficiency of complex II activity (Bourgeron et al, 1995). Precise

p n e defects have also been identified in the AQDQ, TYKY, PSST and the 5 1 kDa

subunits of complex I (van den Heuvel et al, 1 998; Loeffen et al, 1 9 9 8 ~ ; Triepels et al.

1999b; Schueke et al, 1999). Linkage analysis in pedigrees with autosomal dominant

CPEO (mtDNA deletion disorders) has identified two loci, one on chromosome 1 Oq23.3-

24.3 and the other on chromosome 3p 14.1-2 1.2 (Suornalainen et al, 1995; Kaukonen ei al.

1996), but the gene has not yet been identified. Complementation studies have s h o w

that mtDNA depletion syndrome (clinical phenotypes include fatal infantile

hepatopathy, fatal infantile myopathy and a more benign infantile myopathy) and

isolated COX deficiency (Leigh's syndrome), are nuclear gene disorders (Carrozzo er al.

1998: Munaro et al, 1997).

Diseases are dso caused by mutations of nuclear genes encoding non-OXPHOS

mitochondrial proteins such as Frataxin in Friedreich's ataxia (Priller et al, 1997; reviewed

by Leonard and Schapira. 2000), Surf-1 (an assembly factor of COX) in COX-deficient

Leigh's syndrome (Zhu er al, 1998), ATP 7B in Wilson's disease (Lutsenko and Cooper.

1998) and others.

(iii) Mitochondrial respiratory chain disorders associated with neurological

diseases

Huntington's disease is inhented in an autosomal dominant fashion and brain tissue

from patients with this disease has s h o w a 5349% decrease in complexes IIAI1 activity.

a 32-38% reduction in complex IV activity as well as elevated L/P ratios in cerebrospinal

fluids (Gu et al, 1996). An anatomically specific complex I defect in the pars cornpacta

has also led to the association of Parkinson's disease (autosomal dorninantly inherited)

with mitochondrial complex 1 deficiency (Bindoff et al, 1989; Schapira et al, 1990). COX

deficiency has also been described in Alzheimer's disease (Sirnonian et al. 1994).

Human Com~lex 1 deficiencies

Human NADH:ubiquinone oxidoreductase deficiency can be present in a wide

spectnim of biochemical and symptomatic phenotypes. The observed range of severity

can range fiom fatal infantile lactic acidosis to adult onset exercise intolerance or optic

neuropathy (Robinson, 1998). Complex I deficiency can be caused by mutations in the

nuclear genome as evidenced by Mendelian inhentance or mutations in the mitochondrial

genome as demonstrated by matemal inheritance (Loeffen et al. 2000). Isolated complex I

deficiency is now known to be the most cornmon mitochondrial respiratory chain defect

(Loeffen er al. 2000).

(i) mtDNA encoded defects in complex 1

Mitochondrial DNA defects can cause complex 1 deficiency either by disturbing the

process of mitochondrial protein synthesis or by interfering at specific sites in the

process of oxidative phosphorylation (Robinson, IW8a). The majority of mtDNA

disease causing mutations are present in the heteroplasmic state, eventually leading to

depressed rates of electron transport and ATP synthesis (Robinson. 1998a).

Diseases of the fint kind have been associated with a characteristic red staining of

convoluted fibres around the mitochondria in muscle when viewed with the Gornori

trichrome stain (Robinson, 1998a). These fibres are called "ragged red fibreso' and

complex 1 diseases associated with them are: MELAS, MERRF, KSS, CPEO, rnyopathy

and cardiomyopathy (MMC) and mitochondnal myopathy (MM) (Wallace, 1 992).

ïhese diseases involve mutations in the mitochondrially encoded tRNA species or

mtDNA deletions and many aiso show a decrease in complex IV (COX) activity (Wallace.

1992).

Diseases of the second type are caused by a nurnber of point mutations in mtDNA as

well as in rntDNA encoding complex 1 polypeptides such as ND 1. ND4, ND5 and ND6

(Shoffner et al, 1995). These mutations are known to cause LHON (with no ragged red

fibres) and LHON with dystonia (basal ganglia disease) in most cases and MELAS in

some (Morgan-Hughes et al, 1999).

(ii) Nuclear encoded defects in complex 1

Isolated complex 1 deficiency in children seems in majority to be caused by mutations

in the nuclear DNA (Loeffen et al, 2000). Examination of the pattern of inheritance in

families with complex 1 deficiency has s h o w that that the disease appean to be

autosomal, or possibly in a few cases X-linked, in which the patient can experience a wide

variety of symptoms (Robinson, 1988a). The clinical presentation of isolated cornplex I

deficiency is surnmarized in table 1.5.

(iii) Free radical generation in complex 1 deficiency

One to two percent of al1 electrons passing down the electron transport chain are

diverted into the formation of the superoxide radicais 0i (Wong et ai. 1989). Superoxides

are formed by the interaction of molecular oxygen with a species of semiquinone (UQlo)

which is a natural intermediate in one electron reduction events in electron transport

through complexes I and III (reviewed by Robinson, 1998b). Superoxide is removed by

Table 1.5. The clinical presentation and incidence of isolated complex 1 deficiency attributed to nuclear encoded defects

Incidence Major Features

Most cornmon Leigh's disease (LD) +/- cardiomyopathy (WPW) or progressive detenoration - slow/fast

Fairly comrnon Fatal infantile lactic acidosis (FILA) cystic changes in white matter Uncornrnon Hepatopathy and tubulopathy (HT)

Cardiornyopathy and cataracts (CC) Cataracts and developmental delay (CD) Neonatal lactic academia followed by mild symptoms (MS) Normal development (some with exercise intolerance and congenital Catarac ts)

Severe cases of complex 1 deficiency have been s h o w not to assemble complex I (Robinson et al, IWO).

the enzyme ~ n " superoxide dismutase (MnSOD) (mitochondrial) and CU-?Z~-'

superoxide dismutase (CuZnSOD) (extrarnitochondrial but also exists in an extracellular

fom) (Pitkanen et al, 1996). These enzymes are able to convert superoxides to the stable

hydrogen peroxide which is di ffisible across biological membranes and removed by

catalase or glutathione peroxidase (Pitkanen el al, 1996). Free radicals were seen to be

produced in excess in patients with complex I deficiency and MnSOD production was

also seen to Uicrease to combat the production of superoxide (reviewed by Robinson.

Free radical production fiom the respiratory chain has been implicated in selective ce11

death or apoptosis (Korsmeyer et al, 1995). The bcl-2 family of proteins is known to

serve as a protective agent against apoptosis and bcl-2 levels are increased in cells with

respiratory chain defects involving complex I (Korsmeyer et al, 1995). Bcl-2 induction is

seen to proceed through a similar but different route as MnSOD induction (Korsmeyer er

of. 1995). Inbom errors of OXPHOS may therefore render a ce11 susceptible to death by

apoptosis and rnay be important in the pathogenesis and progression of respiratory chain

deficiencies (reviewed by Robinson, 1 998b).

Objectives and Rationale

Defects of mitochondrial metabolism result in a wide variety of human disorders which

c m present at any time from infancy to late adulthood and involve virtually any tissue

either alone or in combination. Abnormalities of the electron transport and oxidative

phosphorylation (OXPHOS) system. encoded by both mitochondrial and nuclear

genomes. are probably the most common cause of mitochondrial diseases. Of these.

isolated complex 1 deficiency is a major cause of abnormalities encountered in respiratory

chah enzymes. Only a limited number of enzymatic deficiencies of one or more of the

complexes of the rnitochondrial electron transport chain are associated with mutations in

the mtDNA. For complex 1 deficiency in particular, the several mtDNA mutations that

have been well charactenzed cannot explain al1 cases of isolated complex 1 deficiencies.

especially the most fiequently observed phenotype which presents in infancy and

expresses a highly progressive lethal course (LD). Theoretically, about equal nurnbers of

mutations are predicted for the mitochondrial and nuclear genomes. because even though

nuclear genes are subjected to a somewhat slower mutation rate, there are a greater number

of nuclear genes involved in complex 1. It is therefore of utmost importance that the

contribution of the numerous complex 1 subunits to isolated complex I deficiency be

established. In the ensuing chapters, the NADH:ubiquinone oxidoreductase complex

(Cornplex 1) will be the initial object of focus. In particular, the nuclear encoded MLRQ

subunit of this complex will be examined at both gene and protein levels. The strategy

used to clone, characterize and identify the chromosomal locus of NDUFA4 is presented

in chapter 2. Chapter 3 focuses on the biochemical characterization of this subunit by

looking at its expression in hurnan tissues and its association with other subunits in the

complex. This chapter also looks at strategies that were attempted in order to study the

structure and function of this subunit in complex 1 through bacterial protein expression

and marnmalian anti-sense expression studies, respectively. Finally, chapter 4 brings al1

these concepts together and outlines the possible role of a supemumerary subunit such as

MLRQ in the context of complex 1 f ic t ion and dysfunction in mitochondria. Thus. the

objectives of this study are to characterize the gene and gene product of a nuclear encoded

subunit of complex 1 and illustrate its role in the functioning of this first enzyme of the

OXPHOS system of mitochondria.

Chapter 2

Cloning, molecular characterization and chromosomal localization of the MLRQ subunit of human NADH:ubiquinone oxidoreductase

Abstract

The genornic DNA and cDNA sequeoces encoding the MLRQ subunit of human

NADH:ubiquinone oxidoreductase (cornplex 1 of the mitochondrial respiratory chain)

have been determined. The cDNA clones contain an open reading frame of 243 bp, 53 bp

of 5'-untranslated sequence (UTR) and a 3' UTR of 18 1 bp. MLRQ cDNAs obtained

from a number of different human tissue types such as brain, hem, liver and kidney as

well as fibroblasts and cardiomyocytes were sequenced and found to be identical.

Two clones narnely. 2F23 and 96E24 were isolated from a P 1 -artificiai chromosome

(PAC) library by using a MLRQ cDNA probe. Fiuorescence in situ hybridization (FiSH)

was then used to map 2F23 to chromosome 1 p21 and 96E24 to chromosome7 p2Lp22.

Searching the High Throughput Genome Sequence (HTGS) database with this

information reveaied that the NDUFA4 gene is situated on chromosome 7. spanning

approximately 6.5 kb and consisting of 4 exons. Partial sequencing of the MLRQ-like

gene product from PAC clone 2F23 reveaied that in contrast to NDUFA4, it is an

intronless gene with several basepair changes in its coding and non-coding regions and

therefore is a pseudogene.

Introduction

Arnong the inbom errors of metabolism, disturbances in mitochondrial energy

metabolism occur in 1110,000 live births (Bourgeron et al, 1995). Of these, isolated

complex I deficiency is the most cornrnon mitochondrial respiratory chain defect

(Loeffen et al, 2000). The clinical presentation and course of complex 1 deficiency can

Vary greatly due to the involvement of both the mitochondrial and nuclear genomes. as

well as factors related to oxygen free radical generation in the complex (Robinson et al.

1998). Complex 1 defects encoded in rnitochondnal DNA have been well characterized

at the molecular level (Morgan-Hughes et al, 1999). However, underlying genetic

defects cm be ascribed to mtDNA in only - 5% of complex 1 patients (Liang and Wong,

1998). In many isolated complex 1 deficiencies. no mutation in the mitochondrially

encoded subunits of cornplex I have been found. These cases have been attnbuted to the

nuclear genome and nuclear encoded complex 1 deficiency appears to be an autosomal. or

possibly in a few cases, an X-linked disease in which the patients can suffer from a wide

spectrum of symptoms (Smeitink et al, 1998a).

Although the cDNA charactenzation of human complex I subunits has recently been

completed (Loeffen et al. 1998b; Tnepels et al. 2000), the genomic structure and

chromosomal localization of many of these nuclear genes are yet to be determined.

Particularly in the hydrophobic (HP) fraction, the complete genornic DNA sequence has

only been elucidated for the NDUFAl (MWFE), NDUFA6 (B 14), NDUFS8 (TY KY) and

NDUFB9 (B22) genes (Zhuchenko et al. 1996: Dunham et al, 1999; de Sury er al. 1998:

Lin et al, 1999).

AQDQ, TYKY. PSST and the 51 kDa subunit are the only nuclear encoded subunits

in which mutations have been characterized thus far (van den Heuvel et al, 1998; Loeffen

et al. 1998~; Triepels et ai, 1999b; Schuelke et al, 1999). Table 2.1 characterizes the

mutations that have been found in patients with isolated complex 1 deficiency. A recent

study has also for the first time reported two mutations in the AQDQ subunit associated

with combined complex I and III deficiency (Budde et al. 2000).

Table 2.1. Nuclear gene mutation in patients with isolated complex 1 deficiency

AFFlECTED GENE SYMPTOMS AT LACTIC ACID MRI FINDINGS (MUTATION) PRESENTATION CONCENTRATION

Blood Urine CSF

(1) NDUFVI (5 1 kDa)Vomiting, 1' (R59X. T423M) strabismus, hypotonia

(2 ) NDUFVI (5 1 kDa)Vomiting, 1' (R59X. T423M) strabismus, hypotonia

(3) N D UFVI (5 l kDa) Infantile myoclonic (A34 1 V) epilepsy

(4) NDUFS4 (AQDQ)Vomiting, failure (5-bp duplication) thrive, hypotonia

(5) NDUFS7 (PSST) Feeding problems. (V 122M) dysanhria, ataxia

(6) NDUFS7 (PSST) Vomiting (V 122M)

(7) NDUFS8 (TY KY) Feeding difficulties. (W9L, R 102H) hypotonia, episodic

apnea and cyanosis

1' Atrophy

I' Atrophy

Atrophy. progressive I* macrocystic

leukodystrophy

N' Atrophy. basal ganglia abnormalities

SI.1' Syrnrnetrical hypodensi ties

N' Symmetricd hypodensities

1 ' Atrophy, syrnmetrical hypodensities

* N - normal; I - increased; S1.I - slightly increased Patients (1,2 & 3) (Schueike et al, 1999); Patient (4) (van den Heuvel et al, 1998); Patients (5 & 6) (Triepels et al, 1999b); (7) (Loeffen et al, 1998c)

The characterization of new human genes such as NDLIFAl (MLRQ) has greatly

benefited f'om bioinformatics' toois and databases such as dbEST, UniGene and HTGS,

which are divisions of GenBank. GenBank is a public sequence database that

incorporates dl known nucleotide and protein sequences, primarily through the direct

submission of sequence data from individual laboratories and from large-scale

sequencing projects. GenBank ensures comprehensive international coverage by data

exchange with the DNA DataBank of Japan (DDBJ), the European Molecular Biology

Laboratory (EMBL) and GenBank at the National Center for Biotechnology Information

(NCBI) (Benson et al, 1999). There are approximately 7 ,376 ,000 .0 bases in 6.2 15.000

sequence records as of April2000 in GenBank (www .ncbi.nlm.nih. eov/Genbank).

Expressed Sequence Tags (ESTs) are short (usually about 300-500 bp), single-pass

sequence reads from mRNA (cDNA) which are produced in large batches and represent a

snapshot of genes expressed in a given tissue or cDNA library by 'lagging" them

(Boguski et al. 1993). ESTs are a major source of new sequence records and genes in

GenBank and the discovery of new members of a gene farnily is Facilitated by retrieving

ESTs homologous to a known cDNA sequence of a different organism, from the EST

database (dbEST) (www.ncbi.nlm.nih.eov/dbEST). In order to organize the EST data in

a usehl fashion. NCBI created the UniGene collection of unique human genes and mouse

genes (www.ncbi.nlm.nih.~ov/schuler/uni~ene). UniGene starts with entnes in the

primate or rodent division of GenBank, combines these with ESTs of that organism and

creates clusten of sequences that share virtually identicai 3' untranslated regions (3'

UTRs). In this manner, the millions of human ESTs in GenBank are reduced - 20 fold to

sequence clusten, each of which may be considered as representing a single human p n e

(Benson et al, 1999). The High Throughput Genornic Sequences (HTGS) are unfinished

large-scaie genomic records that are in transition to a finished state, after which they will

be placed in the appropriate organism division. Search and retrieval of sequence data

over the world wide web (www) is perfomed rnost commonly by using the BLAST

family of search prograrns. Each BLAST search can be custornized according to the

required information such as blastn for aligning nucleotide query sequences against

nucleotide sequences and searching databases such as nr (non-redundant). HTGS. dbEST.

epd (eukaryotic promoter database) etc; blastx for aligning nucleotide query sequences

uanslated in al1 frames against protein sequences and blastp for aligning amino acid

query sequences against protein sequences and searching databases such as nr or

swissprot.

The rationale behind the characterization and molecular screening of the human

homolog of MLRQ was driven by a couple of factors. Linkage and complementation

studies performed by Scheffler and Day (Day et al. 1982) had long suggested the

existence of two X-linked subunits. The MWFE subunit has already been established as

one of these possibilities (Zhuchenko et al. 1996). MLRQ warranted investigation as the

other candidate subunit for X-linkage, when a search of the UniGene databaîe revealed

that researchers at the Whitehead institute had mapped a MLRQ-like gene to the human

X-chromosome through EST sequencing. These reasons. dong with the observation that

there is a preponderance of affected males in certain cases of complex 1 deficiency

(Orstavik et al. 1993), raised the possibility that the MLRQ gene may be situated on the

X-chromosome.

The nuclear encoded MLRQ subunit was found to coprecipitate with bovine complex

1 as a 9 kDa protein (Walker et al. 1992). It's hydropathy profile suggests that it can be

folded into a membrane spanning a-helk of the HP fraction and that it interacts with

other subunits of complex 1 through the more hydrophilic parts of its sequence (Walker et

al, 1992).

Charactenzation and mapping of the human genes for nuclear-encoded complex 1

subunits is essentiai to the understanding of the genetic mechanism of the disease as well

as to the understanding of the structure and Function of complex 1.

Materials and methods

Part 1. Molecular Characterization of MLRQ cDNA in various tissues, cells and patient ceil Lines

Tissue culture of cardiomyocytes and fibroblasts

Human fibroblasts were cultured in Eagle's a-minimal essential medium

supplemented with 10% (vlv) fetal calf serum and glucose to a final concentration of 4.5

mM. Cardiomyocytes were cultured in Iscoves medium supplemented with 10% (vlv)

fetal calf serum.

RNA isolation from cells and tissues

Total RNA was isolated from human tissues and cultured cell lines using TRlzol

Reagent (Gibco BRL) according to the manufacturer's protocol.

cDNA synthesis and PCR

First-suand cDNA was produced by reverse transcription of mRNA isolated frorn

human fibroblasts. cardiomyocytes and from heart, brain, kidney and liver autopsy tissue

using an oligo (dT,,.,,) primer and Superscript II reverse transcriptase (Gibco BRL) with

appropriate reverse transcription controls.

Oiigonucleotide pnmers forward (5'-ïTTAGCmAGGGCCTmTGGCT-3')

(corresponding to residues 8-30 of the bovine cDNA sequence) and reverse

(5' - ACAATACAGATCTCCAACATG-3' ) eorresponding to residues 405-426 of the

bovine cDNA sequence) designed from a human EST (Expressed Sequencr Tag)

sequence (accession number H65454) were used to ampli9 the cDNA sequence of the 5'

UTR and the coding region. PCR was carried out for 35 cycles ( 1 min denaturation at

94°C. 1 min annealing at 41°C. 1 min primer extension at 72°C. and a LO min primer

extension in the final cycle).

In order to ampliw the 3' region of the gene. RACE (Rapid Amplification of cDNA

Ends) PCR was employed using primer MLRQ 249 (5'-CCAATGATCAATACAAG

TTC-3') and Oligo dT,,,, with an Xbai site (5'-TTTITCTAGA-

'ITTTTTT-3'). PCR was carried out under the sarne conditions as before except with an

annealing temperature of 52°C.

The full-length MLRQ cDNA sequence was arnplified from the cultured ce11 lines.

tissues and a commercial human heart cDNA library by using fonvard primer MLRQ

(5'- CCTGGTGGCTAGGTCGG-3') and reverse primer MLRQ 478

(5'-AAGTTTCAGTTATTTATTGATTTAA-3'). PCR conditions were similar to those

outlined above except the annealing temperature was 52°C.

cDNA cloning and sequencing of MLRQ

The PCR products were gel purified and subcloned into TA" cloning vector

(Invitrogen) and competent DHSa E.coli cells were transformed with the constmct.

BIue/White screening was used to pick the colonies with the insert of interest and LB

media supplemented with ampicillin (100 pg/rnl) was inoculated with these colonies.

Cells were then harvested after ovemight growth at 37°C and plasmid rninipreps were

carried out on the cultures according to Maniatis et al (1989). The resulting DNA was

sequenced using the T7 DNA sequencing kit (Amersham Pharmacia) by the dideoxy

chain termination method with univenal forward and reverse prirners according to

manufacturer's instructions.

Mutational screening of compler 1 deficient patients

Patients with various complex 1 clinical presentations were screened for possible

mutations in the MLRQ cDNA sequence. These patients had abnonnally high L P ratios

and were diagnosed as complex 1 deficient. Sequencing was initially carried out with

cycle-sequencing kit (Arnersham Pharmacia), but due to a series of false positive

mutations. the MLRQ cDNA was amplified using PFU (Stratagene). subcloned and

sequenced as described above using T7 polymerase. The following patients were

screened for mutations from cultured fibroblasts.

Table 23. Summary of clnical information on patients screened for mutations in the cDNA of NDUFA4

CeU line Clinical Presentation L/P ratios

5634 Hepatomegaly with Rend Tubulopathy (HT) 93.132 I 2 1.78 1 (5) 7937 Fatal Infantile Lacticacidosis (FILA) and 7 1.776 I 13.449 (5)

Cardiomy opathy 8479 Leigh s Disease (LD) with cardiomyopathy 173.143 - 19.24 (6) 8693 Lacticacidemia; Myopathy; Optic atrophy 117.736 k 48.41 (8) 8768 Familial megalencephaly; Leukodystrophy; 130.462 I 27.54 (4)

Basal ganglia disease 8889 Fatal Infantile Lacticacidosis (FILA) 1 13.924 2 26.74 (5)

** L/P ratios for control ce11 lines used to compare each of these rneasurements range fiom 14.705 - 2.190 (6) to 19.355 - 2.55 (8)

Part II. Genomic characterization and localization of the NDUFA4 (MLRQ) gene and pseudogene

Screening the PAC library

Full-length MLRQ cDNA was used as a probe to search a Pl-artificial chromosome

(PAC) genornic DNA library, namely pCYPAC- 1 (RPCI-1). Screening of the library was

done following the protocols outlined by Osborne et al (1996) through the Toronto

Centre for Applied Genomics.

Southern blot analysis of PAC clones

PAC clones 2F23 and 96824 C U ~ N ~ ~ S were grown overnight at 37°C in LB media

supplemented with kanamycin (50 pg/ml). Plasmid DNA was isolated according io

Maniatis et al ( 1989).

Ovemight restriction was performed using enzymes EcoRI. MboI, BamHI and Nor1 on

5 pg of plasmid DNA from both 2F23 and 96824 clones. and using AvaalI on plasrnid

DNA from clone 2F23. Restricted DNA was loaded ont0 0.7% agarose gels and

electrophoresed ovemight. Alkaline transfer and southem hybridization was performed as

per Maniatis et al (1989) with 200 pg/rnl of Heparin as the blocking agent. 400 ng of

MLRQ (full-length) cDNA probe was spotted ont0 the nylon membrane (Amenham).

The DNA was UV cross-linked for 5 minutes. 50 ng of the MLRQ probe was labeled

with p' using the oligolabeling kit (Pharmacia). The probe was purified on a Sephadex

G-50 column washed with STE (O. 1 M sodium chloride, 10 rnM TrisHC1, pH 8.0 and 1

rnM EDTA, pH 8.0) and then mixed with salmon sperm DNA and denatured. The

denatured mix was added to the membrane and shaken overnight at 65°C. The

membrane was washed twice in 2X SSC (sodium chloride and sodium citrate) and 1%

SDS at room temperature for 5 mins. This was followed by two 15 minute washes in the

sane buffer at 6S°C and then a single wash with 0.2X SSC at 65°C. The membrane was

exposed to Kodax Biomax film to obtain an autoradiographic image.

Northem analysis of MLRQ expression

A nylon membrane blot (Human Normal Blot 1, Invitrogen) with 20 pg total RNA

from human tissues such as heart, brain, kidney, liver. lung, spleen and skeletal muscle

was probed with 0.5 pg of the 5 15 bp MLRQ cDNA. 60 ng of a DNA fragment fiom the

3 ' end of the p-actin gene was also used as a postive control probe. Probes were labeled

with "P using the nick translation system (Gibco) and purified on a QIAquick PCR

purification column (Qiagen). Hybridization and subsequent washes were c m i e d out in

accordance with the manufacturer s protocol. The autoradiograph was developed after

exposing the membrane to Kodax Biomax film.

C hromosomal localization

Biotinylated 96824 and 2F23 probes were prepared for FISH mapping to normal

human lymphocyte chromosomes according to methods employed by Ling et al (1998)

through the services of the Toronto Centre for Applied Genomics.

Amplification, cloning and sequencing of MLRQ from genomic and PAC DNA

Genornic DNA isolated fiom cardiomyocyte and fibroblast lines using alkaline lysis

methods, as well as DNA isolated fiom PAC clone 2F23 (Maniatis et al, 1989) was used

as template for PCR using prùners MLRQ 1 and MLRQ 478 under the same conditions

described above. Arnplified fragments were cloned into TA" vectors (Invitrogen) and

sequenced also as described above.

YAC library screening

A chromosome 7-specific (CEPH) YAC library was screened using full-length MLRQ

cDNA as probe, according to methods outlined by Kunz et al (1994) through the Toronto

centre for applied genomics.

Results and Discussion

Part 1. Isolation and characterization of MLRQ cDNA

MLRQ cDNA structure

The human MLRQ homologue was cloned from cultured fibroblast and

cardiomyocyte ce11 lines as well as from hem. brain, Iiver and kidney tissues and a hean

cDNA library. The full-length MLRQ cDNA sequence (accession no. AF301077) was

cornpiled frorn the overlapping sequences of the two PCR fragments (Fig. 2.1). Full-

length cDNA for MLRQ was found to be 5 15 bp in length with an open reading frarne of

243 bp encoding a protein 8 1 amino acids long (predicted molecular mass of 9370 Da).

This sequence was found to be identical to the human MLRQ cDNA clone frorn a Korean

fetal liver tissue library (accession number U94586) (Kim et al, 1997). with the exception

of an A to G transition at position 39 (with the start codon as pos. 1) and an additional 7

bp between the last polyadenylation site and the poly A tail. These polymorphisms cm

be attributed to the fact that these two clones are from the mRNA of individuals from

different ethnic populations. The compiled human MLRQ cDNA sequence shows 89%

(33 1/37 1) identity to its bovine counterpart (EMBL accession no. X64897) as well as

TCC TGG TGG CTA GGT CGG TTC TCT CCT TTC CAG TCG GAG ACC -13

1 TCT GCC GCA AAC ATG CTC CGC CAG ATC ATC GGT CAG GCC AAG 30

m e t l e u a r g g l n ile ile gly g l n a la l y s 10

AAG CAT CCG AGC TTG ATC CCC CTC TTT GTA TTT ATT GGA ACT 72 lys his p r o ser l e u i l e p ro l e u phe val phe i l e gly t h r 24

GGA GCT ACT GGA GCA ACA CTG TAT CTC T'TG CGT CTG GCA TTG 114 g ly ala t h r gly ala t h r l e u t y r l e u l e u arg l e u a l a l eu 38

TTC M T CCA GAT GTT TGT TGG GAC AGA m T AAC CCA GAG CCC 156 phe a s n pro asp val cys trp asp arg asn asn pro g l u pro 52

TGG AAC AAA CTG GGT CCC AAT GAT CAA TAC AAG TTC TAC TCA 198 t r p a s n lys leu gly pro asn asp g l n t y r l y s phe tyr ser 66

GTG M T GTG GAT TAC AGC AAG CTG AAG AAG GAA CGT CCA GAT 240 va l asn v a l asp t y r ser l y s l e u lys lys g l u a rg pro asp 80

TTC T M ATG AAA TGT TTC ACT ATA ACG CTG CTT TAG AAT G U 282 phe * 81

GGT CTT CCA GAA GCC ACA TCC GCA CAA TTT TCC ACT TAA CCA 324 GGA AAT ATT TCT CCT CTA AAT GCA TGA AAT CAT GTT GGA GAT 366 CTC TAT TGT AAT CTC TAT TGG AGA TTA CAA TGA TTA M T C a 4 0 8 TAA ATA ACT GAA ACT TGA AAA AAA AAA AAA AA?4 MW AAA M A 450 AM4 453

Figure 2.1. The nucleotide sequence of the human MLRQ subunit eDNA and its deduced amino acid sequence. The numben at the right of each line denote the position of the last nucleotide and amino acid. The stop codon is indicated by an asterisk. The polyadenylation signals are underlined.

85% (193/226) identity to the Mus musculris MLRQ-like protein mRNA (GenBank

accession no. U59509). At the deduced amino acid level. the human MLRQ subunit

protein exhibits sirnilarities of 93% (89% identity, 4% favoured substitutions) and 8 5 4

(80% identity. 5% favoured substitutions) with the bovine and mouse LMLRQ subunits,

respectively (Fig. 2.2).

hMLRQ: MLRQIIGQABXHPSLXPLFVFIGTGATGATLYLLRXrAR 47 bMLRQ : MLRQI IGQAKRHPSLIPLFIFIGAGGTGAALYVTRLALF'NPDVSWDR 4 7 mMLRQ: ------- Q ~ P s L I P L F V F I G A G G T G A A L ~ ~ D v ~ R 4 0

Figure 22. Alignment of the predicted human MLRQ subunit protein sequence with that of bovine and mouse. Perfectly conserved residues between the human (hMLRQ). bovine (bMLRQ) and mouse (rnMLRQ) sequences are highlighted.

The deduced polypeptide sequence of human MLRQ shows a relatively hydrophobic

N-terminal region (arnino acid positions 1-39) and a relatively hydrophilic carboxyl-

terminus (amino acid positions 40-8 1). As with the bovine polypeptide sequence. the

hydrophobic segment of about 25 amino acids (positions 15-39) is sufficiently long

enough to fom a trammembrane a-helix. It is interesting to note that the amino acid

substitutions seen between the human and bovine/mouse MLRQ polypeptide do not

affect the two domain structure of the protein. thereby suggesting that this feature is

cntical for the subunit's structure and function. Based on the predicted sequence of the

human MLRQ protein, it is safe to conclude that diis subunit has no consensus

phosphorylation sites and owing to a lack of cysteine-rich motifs (there is only one

cysteine at amino acid residue W), it is highly unlikely that the MLRQ subunit is an iron-

sulfur protein. It is also interesting to note that the bulky hydrophobic arnino acid

tryptophan, which has the lowest average occurrence in proteins (about 1.1 %) (Voet and

Voet. 1990) is seen twice (at amino acid residues 45 and 53) within the 8 1 arnino acid

MLRQ sequence.

Mutational analysis of MLRQ cDNA in complex I deficient patients

Since the contribution of mutations in the supemumerary subunits of complex 1, such

as MLRQ. to complex I deficiency had never been fully studied. mutational analysis of

this subunit was undenaken. A group of six young children in whom isolated complex I

deficiency had been confirmed in skin tissue (by vktue of the fact that other respiratory

chain enzymes such as COX and succinate cytochrome c reductase were found normal)

and in whom common mitochondrial DNA mutations had been excluded, were examined

for mutations in MLRQ cDNA. However. no disease causing mutations were found in

any of these patients. Recently. Triepels et al (2000) conducted an extensive study in

which 19 genes of the HP fraction. including that of MLRQ. were screened for molecular

aberrations in 14 young patients with isolated complex 1 deficiency. No diseasetausing

mutations or polyrnorphisms in the cDNA sequence of the NDUFA4 (MLRQ) gene were

detected by Triepels' group either. Therefore. mutations in the ORF of human NDUFA4

probably represent a minimal contribution to complex 1 deficiency.

Part II. Chromosomal localization and characterization of the NDUFAl gene and pseudogene

Library screening and FISH mapping

Screening of the pCYPAC-1 human genomic DNA library with the full-length MLRQ

cDNA as probe, generated two PAC clones, namely 2F23 and 96E24, with the latter

showing a much stronger hybridization signal. FISH mapping of the 2F23 and 96E24

clones to nomal human lymphocyte chromosomes assigned 2F23 to chromosome 1 p2 1

and 96824 to chromosome 7 p2 1 -p22 (Fig. 2.3a, b). Another PAC clone (69E 1 1 )

carrying a sequence with great similarity to the MLRQ cDNA sequence discovered

through a BLAST search of the non-redundant (nr) database, was also FISH mapped to

chromosome I q24 (Fig. 2Jb). nie fact that the two clones 2F23 and 96E24 obtained

from the PAC library were mapped to N O different chromosomes, suggested the

existence of either a pseudogene (non-transcnbed) or an isoform of MLRQ.

In order to validate this possibility, primers used to amplify the cDNA sequence of

MLRQ were used in a PCR with PAC 2F23 as the ternplate. The amplified sequence

(accession no. M206638) (Fig. 2.4) revealed that there were changes in its nucleotide

composition when cornpared to the sequence obtained from the cDNA of cardiomyocyte

and fibroblast ce11 lines (Fig. 2.1). 6 basepair changes in the coding region; 2 basepair

changes in the 5' region and 5 basepair changes in the 3' region were noted (Fig. 2.4).

The nucleotide changes seen in the PAC 2F23 sequence give rise io the substitutions

R3C, E L , T27S, R35H, Y65C and R78C. If this isoform of MLRQ is expressed, the

presence of the three cysteines would result in a protein that is considerably different in

structure from that of the known MLRQ subunit.

Confirmation that PAC 2F23 carries a pseudogene and not an isoform of MLRQ was

also done with the sequencing of the cDNA obtained from RNA isolated fiom heart,

brain, kidney and liver tissues. The MLRQ sequences amplified fiom these tissues were

identical to the sequences amplified fiom cardiomyocyte and tibroblast iines as well as a

heart cDNA library. The fact that the NDUFA4 gene maps to chromosome 7 was also

1 I

- 1 . 1 kb intron I

exon II

exon III

1 1

- 4.4 kb intron III

1 +TAA

PAC 2F23

Figure 2.3. Schematic of NDUFA4 structure and chromosoma1 allocation of the gene and its pseudogene, (a) The idiogram of human chromosomes 7 is shown iiiustrating the distribution of Iabeled sites determined with PAC 96E24 as probe. A schematic representation of the genomic organization of the NDCIFA4 gene is also provided. (b) The idiogram of human chromosome 1 is presented illustrating the distribution of labelled sites at p21 and q23-24 detennined with PACs 2F23 and 69E 1 1 as probes, respectively.

confmed by primers designed f?om the 3' region of the MLRQcDNA sequence which

were used to probe a YAC genomic library. The two YAC clones that were isolated,

namely C745fB and C932h3, were both FISH mapped to the p2Lp22 region of

shromosome 7 (Fig. 23a). Sequencing the 5' region of the two PAC clones, a complete

divergence in sequence between 2F23 and 96824 was discovered -75 bp upstream of the

start codon (Fig. 2.4).

Some diseases associated with chromosome 7 mutations are cystic fibrosis, Palister-

Hall syndrome, Pendred syndrome, GCK diabetes, Split handfoot malformation type 1

and Williams-Beuren syndrome. However, none of these disease genes have been

localized in proximity to the NDUFA4 gene loci 7 p2 1-22. Other genes that have been

mapped to the area 7p2 1-22 are those encoding carniosynostosis and interleukin 6 at p2.I

as well as those encoding ce11 division protein FtsJ, G protein-coupled receptor 30, islet

ce11 autoantigen, nucleotide diphosphate linked moiety X-type motif 1, v-maf avian

rnusculoaponeurotic fibrosarcoma oncogene family-protein K, platelet-derived growth

factor alpha polypeptide and replication protein A3 at p22. Other genes coding for

complex 1 subunits that have been localized to chromosome 7 are NDUFB2 (DDGD) and

the pseudogene possessing NDUFAS (8 13), both of which have been mapped to the q

a m of the chromosome.

MLRQ expression at the transcriptional level

To determine the size of MLRQ transcripts as well as to shed some light on the

question of an MLRQ pseudogene, northem analysis was perîormed on a blot containing

total RNA isolated fkom human heart, brain, kidney, liver, h g , spleen and skeletal

TGTTTAAAATTCTATGAAACGCAGACACTTTTTAGCTCAGGGCCTGGTGGCTAGG 1 6 5 - ***** -b

TCGGTTCTCTCCTTTCCAATCAGAGAGACCTCTGCCGCAAACATGCTCTGCCAGCTC

ATCGGTCAGGCCAAGAAGCATCCGAGCTTGATCCCCCTCTTTGTATTTATTGG 2 7 5

TGTTTGTTGGGACAGAAATAACCCAGAGCCCTGGAACAAACTGGGTCCCAATGAT 385

CAATACAAGTTCTGCTCAGTGAATGTGTGGATTACAGCGCTGGGGTGTC 440 4-

CAGATTTCTAAATGAAATGTTTCACTATAACACTGCTTTAGAATGAAGGTTTTCC 495

AGAAGCCACATCCGCACAATTTTTCACTTAACCAGGAAATATTTCTTCTCTAAAT 550 - - -

GCATGAAATCATGTTGGAGATCTCTATTGTAATCTCTATTGGAGATTACACTGAT 605

TAAATCAATAAATAACTGAAACTTG 6 3 0

Figure 2.4. Sequence of the MLRQ pseudogene on PAC 2F23. The structure of the MLRQ pseudogene on chromosome lp2 1 as determined by sequencing PAC 2F23 is presented. The 5' region in bold from pos. 1 to 130 represents sequence which differs fiom that of NDUFA4. The single bases in bold represent the nucieotide changes in the pseudogene compared to MLRQ cDNA. The asferiskî denote putative TATA and CAAT sequences while arrows point to the putative start and stop codons. The underlined sequences show the direct repeats at both ends of the coding region of MLRQ which are characteristic of Type II pseudogenes.

muscle. Probing with 3 2 ~ labeled Ml-length MLRQ cDNA, revealed a single hybndizing

band of about 500 bases in ail lanes (Fig. 2.5a). The DNA fragment from the p-actin

gene served as a positive control probe and indicated that RNA isolated from these

tissues was not degraded (although some degradation can be observed with skeletai

muscle RNA) (Fig. 2Sb, lane 7). Expression of the MLRQ ûanscript is seen highest in

heart tissue (Fig. 2.5a, lane l), which has the highest requirement for ATP, followed by

MLRQ 500 bases

Figure 2 5 . Northern blot analysis of NDUFA4 transeripts in normal human tissues. A blot

containing 20 pg of total RNA from normal human tissues (Invitmgen) was hybridized with 32~-labeled (a) Full length MLRQ cDNA probe and (b) DNA fragment from 3' end of p-actin gene at 42OC. Washes were performed as follows: once with 2X SSCf0.059bSDS at room temperature, three times with the same buffer at 42OC and twice with O. 1X SSUO. 1% SDS at 50°C. Lane 1, hem; lane 2, brain; lane 3, kidney; Iane 4, liver; Iane 5, h g ; lane 6, spleen; lane 7, skeletd muscle. Sizes of the hybridizing fragments are indicated on the right.

skeletal muscle and brain tissues (Fig. 2Sa, lanes 7 and 2). In comparison to the amount

of B-actin transcripts, MLRQ mRNA levels in liver tissue seem to be quite high. This is

not surprising since liver is the metabolic clearing house of the body, degrading fatty

acids and arnino acids whose byproducts ultimately enter the OXPHOS system. Northern

analysis performed by Kim et al (1 997) agrees with this pattern of expression seen with

the different tissue types. A more prudent positive control for this experiment might have

been the use of a DNA probe for banscripts of a mitochondrial protein such as citrate

synthase, which would have allowed comparison with protein expression as s h o w in the

next chapter (Fig. 3.1).

Amplification of the NDUFA4 gene from genomic DNA

Primers designed from the full-length MLRQ cDNA sequence were used to amplify

the NDUFA4 gene from genomic DNA isolated from fibroblasts and cardiomyocytes.

Products of varying sizes were amplified, including one that was of similar size to the

MLRQ cDNA. Sequencing of this amplified product showed that it was identical to the

MLRQ pseudogene sequence on PAC 2F23. The larger sized products including one thüt

was about 6.5 kb in size (corresponding to NDUFA4) (Fig. 2.6, lane 2) were not cloned

or sequenced.

Southem blot analysis

Another strategy that was approached simuItaneously in order to determine which

PAC clone held the true genomic sequence of MLRQ was Southem hybridization

(Fig. 2.7). Restriction products of the pCYPAC2 vectors 2F23 and 96E24 were probed

with the MLRQ cDNA sequence. A strong hybridinng signal with an insert of

Figure. 2.6. PCR arnpüfication of NDUFA4 from genomic DNA. Primers MLRQ I and MLRQ 478 were used to amplify the identified bands from genomic DNA isolated from a normal human fibroblast line. PCR conditions were as ~ O ~ ~ O W S : 35 cycles, TAnnealing= 52OC. Lane 1, negative conuol; Lane 2, products amplified from genornic DNA. PCR products were run on a 1.5% agarose gel. The largest product (about 6.5 kb) probably corresponds to the NDUFA4 genornic sequence) while the smaller fragments represent other MLRQ-like sequences in the human genome. The 478 bp product was sequenced and found to be identical to the MLRQ-like sequence from PAC 2F23.

Figure 2.7. Southeni anaiysis of PAC clones 2F23 and 96E24. Five rnicrograms of plasmid DNA isolated fiom PACs 2F23 and 96E24 were digested with the indicated restriction enzymes and electrophoresed through a 0.7% agarose gel. The DNA was transferred onto a Hybond-N+ (nylon) membrane and hybridized with a 32~-1abeled MLRQ cDNA probe. The blot was washed twice in 2X SSU1% SDS at room temperature, foilowed by two 15 minute washes in the same buffer at 65OC and a single wash with 0.2X SSC at 6S°C. Lane 1, h NindIII marker; lanes 2 and 8, EcoRI; lanes 3 and 9, MboI; lane 4, AvaII; h e s 5 and 10, BamHI; lanes 6 and 11, NotI; lane 7,50 bp marker.lanes 2-6 contain digested DNA fkom PAC 2F23 and lanes 8-11 contain digesteâ DNA from PAC 96E24. Arrows point to hybridizing hgments, however, band sizes are only provided for the strongly hybriduing signals.

approximately 4 kb was seen in Nit1 restncted 2F23 (lane 6), while no signal was

detected in the case of 96E24 restricted by the sarne enzyme (lane Il). This is surpnsing

especially because Nor1 is the enzyme that is supposed to release the genomic insert fiom

the pCYPAC2 vector. 2F23 cut with AvaII gives two Fragments approximately 1.5 kb and

2 kb in size (lane 4) and 2F23 restncted with BamHI gives a strong signal at 4kb (Iane

5). The only signals detected for PAC 96824 were in lanes with EcoRi restncted 96E24

(lane 8), which had a 1 kb fiagment and Mbol restricted 96E24 (lane 9) where a very

large fragment was detected. Cloning and sequencing of these hybridizing

DNA Fragments would have provided answers as to the sequence of genomic MLRQ.

However. since the 96824 PAC clone was mapped to chromosome 7. this sequence was

first used in a High Throughput Genomic Sequence (HTGS) database search for a

genomic clone. Fomiitously, search results revealed that a clone DJ0855F 16 (accession

no. AC007029) contained the NDUFAI genornic sequence. Therefore. characterization

of the hybridizing fragments from PACs 2F23 and 96824 was not pursued.

Genomic organization of NDUFA4

The iVDUFA4 gene is composed of 4 exons that range From 58 bp (exon I I I ) to 237 bp

(exon IV). The three introns are estimated to range fiom 0.66 kb (intron 2) to nearly 4.4

kb (intron 3) in size with the entire NDUFA4 gene spanning approximately 6.7 kb

(Fig. 2.3a). The splice donor and acceptor sites in each of the four introns follow the

consensus GTIAG splicing sequence (Mount, 1982) (Table 2.3). Analysis of the

exonhntron structure of NDCIFA4 indicates that the predicted transmernbrane domain is

encoded primaily by exon II. while exons III and IV carry the hydrophilic domain.

Genomic and cDNA sequences of MLRQ reveal that it has three in-frame UAG stop

codons upstream of the start site. As with the bovine cDNA sequence, this suggests that

the import sequences for MLRQ lie within the mature protein and not at the N-terminus

(Walker et al, 1992). Many of the subunits in the hydrophobie Fragment possess this

feature and are brought into the mitochondnal IM by binding to the soluble Tim9 and

Tim 1 O proteins and imported through the Tim22TTim54 membrane protein cornplex via

Table 2.3. Exon-intron splice junctions of the human NDUFA4 gene

Exon Size (bp) 5' splice donor Intron Sue (bp) 3' Splice donor

I >95 CCGAGC gtaagt 1 Il19 Ser 1 4

II 89 TGTTTG gtaagt 2 660 CYS 44

III 58 TACAAG gtaaac 3 4397 LYS 63

IV 237

gtttag TTGATC Leu 1 5

ttttag TTGGGA C Y S 4 4

tccaag TTCTAC Phe 6 4

Exon and intron sequences are shown in upper-case and lower-case Ietters, respectively. The numbering of the amino acids corresponds to that of the mature MLRQ protein while the nucleotide sequence coding for these amino acids is italicized within the exon sequence.

Tim12 (see Fig. 1.4). In the PAC 2F23 sequence however, while the UAG in-frame stop

codon immediately preceeding the start codon is preserved, two additional UGA in-frarne

stop codons are also seen at different upstream sites. Three in-fiame AUG codons are

also present in the 5' UTR of the PAC 2F23 sequence. Two of these are between the two

UGA stop codons, while one is found M e r upstream (Fig. 2.4). Interestingly enough.

the 5' UTR of the presumed pseudogene on PAC 2F23 does show the classical TATA and

CCAAT boxes, but at positions -96 to -99 (1 07 to 1 10 on Fig. 2.4) and -2 1 to -25 (1 8 1 to

185 on Fige 2.4)' respectively, whereby they are rendered non-functional. Although AT-

rich boxes are found in the 5' regdatory region, no such promoter motifs are seen in the

130 bp sequence preceding the start codon of PAC 96E24 (Fig. 2.8).

cttagcggcaagaggcccgacctgccctccaggcgcqcc t ccca

Figure. 2.8. Regulatory motifs and putative transcription factor binding sites in the 5' lower part of theNDUFA4 gene. Only the downstream region of the gene that encompasses exon 1 is shown. Consensus motifs for the transcription binding sites are shown in italics with the narnes of the transcription factors given above. AT-rich boxes are underlined. The start codon is indicated in uppercase letters within the Kozak sequence context which is underlined and in bold. The three closest i n - h e stop codons upstream of the start start are shom in bold. Arrows Bank the CpG-island.

The absence of TATA and CAAT boxes is not however exclusive to the N'UFA4 gene.

The promoter regions of genes NDUFAI and NDUW2 do not contain these classical

promoter motifs either (Zhuchenko et al, 1996; de Coo et al, 1995; Hattori et al, 1995).

The translational initiation codon for the NDUFA4 gene occurs in the context

GCAPLACorgC, which deviates a little fiom the optimal GCCACCatgG as defmed by

Kozak (1996). However, one of the two positions said to exert the strongest effect for

translation within this consensus, an A at position -3, is preserved. The 5' end of

NDUFA4 1 is embedded in a CpG-island which is particularly apparent extending fiom nt

-597 to -338 ( W C content 76%). The G-C nch region could well extend into the first

intron to about nt +383 with the G+C content becoming 58%. This is a trend particularly

noticeable on housekeeping genes and other nuclear respiratory genes such as NDUFVZ,

NDWAland NDUFA.5 (Tensing et of, 1999). Nucleotide sequence analysis of this region

has revealed potential binding sites for transcription factors such as NRF2 (nuclear

respiratory factor 2) at positions -245 to -242; -96 to -93 (GGAA), NF1 at positions -337

to -354 (TGGCA), AP 1 at positions -586 to -583 and -333 to -330 (TGAC) as well as a

binding site for the ubiquitous transcription factor Sp 1 (-552 to -549). However, search

for the promoter region using an algorithm called PromoterInspector

~httr>:lleenomatix.pstde/c~ibinl ~romoterinsoector1 ~romoterinsoector.~I) did not

identiQ this region as a putative eukaryotic polymerase II promoter. Instead, this

algorithm which is capable of identifying highly specific localizations of promoter

regions in large genomic sequences (Scherf et al, 2000), identified a 192 bp region

between nt positions l9,88 1 and 20,072 of the 95,345 bp clone RP5-855F 16 (accession

number AC007029) which carries the genomic sequence of NDUFA4 as the only putative

promoter region in the clone. Binding sites for transcription factors IK1, GC, SP 1 and

AP2 as well as two sites for MZF 1 and IKZ have been identified in this 192 bp region. If

this is indeed the promoter modulating transcription of the NDUFA4 gene, it is acting

fiom a distance of approximately 34 kb.

The pseudogene on chromosome 1

While the human somatic cytochrome c gene (HCS) is represented by a single

expressed gene and 1 1 processed pseudogenes (Evans et al. 1988). pseudogenes have

rhus far only been found associated with complex I subunits such as the 24 D a subunit

(de Coo et al, 1995), the B 13 subunit (Pata et al, 1997). the 39 kDa subunit (www-

dsv.cea.fi/themalMitoPicW hnagesCxHumain/Xsome~Map~Human.html) and possibly

the DDGD (13 kDa) subunit (Loeffen et al, 1998a). It is now widely accepted that

pseudogenes can arise From two main mechanisms (Mighell et al. 2000). Type 1

pseudogenes aise through gene duplication of a functional, parental gene and may

remain transcriptionally active (Dover, 1989; Wilde, 1986). Retrotransposition on the

other hand, gives rise to type Il pseudogenes as a result of reverse transcription of the

processed parental transcnpt (Dover. 1989; Wilde, 1986). Type II pseudogenes are not

closely linked to their parental genes. but instead are found on different chromosomes

(Mighell et al. 2000). This is because the cDNA is reinserted into the genome at arbitrary

breaks in a chromosome and as a result, the pseudogene is normally transcriptionally

silent (Mighell et al, 200) . The fact that the two PAC sequences were localized to two

different chromosomes suggests that in the case of the NDUFA4 gene, retrotransposition

rnay have been the cause of pseudogene generation. The fact that none of the MLRQ

cDNA sequences determined from tissues such as hem, brain, liver and kidney

corresponded to the pseudogene sequence on PAC 2F23, further suggests that this MLRQ

pseudogene rnay be transcriptionaily silent. Processed pseudogenes are also known to

have a lack of introns, precise boundaries coinciding with the transcribed regions of the

gene and short direct repeats at both ends (Mighell et al. 2000). These features are

characteristic of the pseudogene in question. The pseudogene on PAC 2F23 is intronless.

is very similar to the transcribed gene except for the described basepair substitutions and

has two pentanucleotide repeats (Fig. 2.4). The fist direct repeat TGCAT starts at pos.

103 in the 5' region and at nucleotide pos. 550 in the 3' region. The second direct repeat

at ATGTT starts at nucleotide pos. 110 in the 5' UTR and at pos. 561 in the 3' region. In

fact, the entire nucleotide sequence starting from pos. 94 through to pos. 1 14

(TTCTTTAAGTGCATATATGTT) in the 5' UTR shows a great degree of sirnilarity to

the nucleotide sequence starting from pos. 541 through to pos. 565

(TTCTCTAAATGCATGAAATCATG'TT) in the 3' UTR. The origin of the MLRQ

pseudogene should become apparent afier a more thorough analysis of the upstrearn and

downstream flanking DNA sequences on PAC 2F23. This may in fact be possible with

recent results from the search of the HTGS database with the sequence from PAC 2F23.

An almost identical sequence (99%) was found in two separate clones. Homo sapiens

clone RP 1 1-24J 14 (accession number ACO16085) and Homo sapiens chromosome 1

clone RP 1 1 -270C 12 (accession number AL l 6 O 2 ) . These search results con firm the

fact that this pseudogene is indeed an intronless sequence.

Other MLRQ-like sequences in the genome

Also important to note is that other MLRQ-like sequences (in addition to the

previously mentioned one, mapped to the X-chromosome by researchen at the

Whitehead hstitute) have been reported. One sequence from PAC 69E 1 1 (accession

number AL021397) (Fig. 2.9) mapped to chromosome I q24 (Fig. 23b), is 88% identical

10 20 30 40 MLRQ cDNA 1 PAC 2F23 1 PAC 69Ell 1

MLRQ cDNA 51 P A C 2 F 2 3 51 PAC 69Ell 3

MLRQ cDNA 101 PAC 2F23 101 PAC 69Ell 53

MLRQ cDNA 151 200 PAC 2F23 151 200 PAC 69Ell 103 152

210 220 230 240 MLRQ cDNA 201 250 ?AC 2F23 201 250 PAC 69Ell lS3 202

268 270 280 290 MLRQ cDNA 251 300 PAC 2F23 251 300 PAC 69Ell 203 252

MLRQ cDNA 301 350 PAC 2F23 301 350 PAC 69Ell 253 302

360 370 380 390 MLRQ cDNA 351 400 PAC 2F23 351 400 PAC 69Ell 303 352

. .- MLRQ cONA 401 450 PAC 2F23 401 450 PAC 69Ell 353 402

MLRQ cDNA 451 PAC 2F23 451

Figure 29. Cornparison of the MLRQ cDNA sequence with the sequences from PACs 2F23 and 69Ell In the region showing greatest slmitarity. Identical DNA sequences arc shaded. The start and stop codons of the cDNA open rcading hune are shown in M d . The polyadenylation signais an underiined. The bold base 92 (usterisk) represents an adenine/guanine polymorphism within the h u m NDUFA4 coding sequence. The Iast 7 basepairs of the 3' UTR (with usrcriskr) denote another polymorphic site. Although num- bered consistentiy for convenience, the sequencc h m PAC 69E I 1 is actually reverse com- plementary to the MLRQ cDNA and PAC 2F23 sequences.

to the coding and 3' region of MLRQ cDNA. However. this sequence is intronless and

prematurely uuncated and reverse complementq to the actual MLRQ cDNA sequence.

Another sequence from clone 1 l89B24. mapped to chromosome Xq25-26.3 (accession

number AL030996) has been reported to show 84% identity to the coding and 3' region

of the MLRQ cDNA sequence. This intronless sequence too gives rise to a premature

stop codon resulting in a protein with only 18 amino acids. Recently. another sequence

from PAC clone 219d7 (accession nurnber AF225899) with 88% identity to the MLRQ

cDNA sequence. but iruncated after the fint three arnino acids. as well as a sequence

mapped to chromosome 12p 1 1-37.2-54.4 (accession number AC0 12 156) showing 86%

ideniity. but without the start codon at the expected position have also been reported. In

comparison to the aforementioned MLRQ-sequences however. clone 2F33 shows a

greater sirnilarity (97%) to the NDUFA4 cDNA sequence (Fig. 2.9). Even though the

NDUFA4 gene does not code for one of the 14 core complex 1 subunits found in bacteria

(Weidner et al, 1993). the existence of many MLRQ-like sequences seems to suggest that

this subunit may have appeared quite early on in the evolutionary course of complex 1. in

order to have undergone such a high degree of changes.

Chapter 3

Biochemical characterization, protein expression and immunoprecipitation studies pertaining to MLRQ and related

complex 1 subunits

Abstract

Expression of the MLRQ protein in human tissues such as heart. brain, liver. kidney.

placenta and muscle showed it to be of uniform size (9 D a ) with no evidence of post-

translational modifications that significantly affect its molecular mas. Overexpression

of this subunit in a patient with hepatomegaly with rend tubulopathy (HT) who had

decreased amounts of other complex 1 subunits seems to suggest a differential regulation

of MLRQ from that of other subunits in the enzyme. Bacterial expression and

purification of MLRQ as a GST-fusion protein was successful. Although transfection of

the MLRQ gene in the anti-sense orientation into transformed fibroblasts was successful.

downregulation of the subunit was not observed by immunodetection with the MLRQ

antibody. Solubilization studies demonstrated that MLRQ extracted differently from the

other complex 1 subunits exarnined, leading to the proposition that the region proximal to

its N-terminus is buried very deeply in the membrane m. The C-terminus being quite

hydrophilic is the exposed region that interacts with other subunits as evidenced by

immunoprecipitation studies with the MLRQ antibody directed towards epitopes at its C-

terminus. These studies also irnplicate MWFE as well as the 49 kDa and 8 18 subunits to

a lesser extent, as being subunits that are in close proximity to MLRQ. Cross-linking of

bovine heart rnitochondria with reagents DST and EGS support these findings and

suggest the association of MLRQ with MWFE through an as yet unidentified subunit.

Introduction

The biochemical characterization of MLRQ was necessitated by the fact that limited

information was available on its structure. spatial organization or function. MLRQ was

assigned as a subunit of complex 1 from studies by Walker et al ( 19%) who found that i t

was a peptide that eluted at 44% acetonitrile on a HPLC column after being isolated in a

2-step chromatographie procedure from complex I that had been dissociated in

guanidinium-hydrochloride. Even though Walker et a1 (1992) first identified it as a 9

kDa subunit in B. taurus, the position of this subunit which was isolated from urea-

solubilized complex 1 was not known with certainty. To date, the MLRQ subunit has

never been visualized on acrylamide or polyacrylamide gels from studies on the

resolution of complex I subunits (Walker et al, 1992: Fine1 et al, 1992; Fearniey et al.

1994; Sazanov et al, 2000). This inadequacy may be attnbuted to the fact that there are

other complex 1 subunits that migrate in the 8 to 10 kDa MW range such as the 10 kDa

(FP) subunit as weIl as the B8, B9. SDAP and AGGG subunits of the HP fraction.

Regardless, the first question that needed to be addressed was whether the MLRQ subunii

is indeed an integral part of complex 1, or whether it is simply a polypeptide that CO-

precipitates with complex I during extraction studies.

The molecular mass of MLRQ in B. taurus was determined from electrospray

ionization mass spectrometry (e.s.i.-m.s) by Feadey et al (1 994) to be 9324 (+/- 1 .O) Da

agreeing very well with the expected molecular mass of 9324.7 Da as calculated fiom the

sequence. However, no groups have thus far examined the expression of this subunit

through immunodetection, there by creating a need for suc h studies from whic h valuable

information on the expression of MLRQ in various human tissues as well as complex 1

deficient patients can be obtained.

MLRQ has also been detected by N-terminal sequencing or electrospray mass

spectromeûy, as a subunit in subcomplex Ia, which is an active NADH dehydrogenase

and represents the predominantly hydrophilic (globular) domain of intact complex I

(Fine1 et al, 1992). The fact that subcomplex IL, which is more water-soluble than la was

found not to contain MLRQ (Fine1 et al, 1992) is also in agreement with structural

information From the MLRQ arnino acid sequence which predicts it to have one

hydrophobie segment with the potential to be folded into a membrane-spanning U-helix.

Given this information, a few other questions regarding MLRQ such as the structure of

the subunit. its location in complex 1 and its interaction with other subunits of the enzyme

also arise.

Answers to these questions are essential in order to determine not only the role of the

MLRQ subunit but ultimately to augment existing knowledge on the structural

organization and fûnction of human complex 1 which is made up of many such

supemumerary subunits.

Materiais and methods

Part 1. MLRQ expression in human tissues and cells

Antibody Generation

High titer polyclonal MLRQ antibody was developed against a 14 arnino acid

sequence at the C-terminus (arnino acid residues 69-82: NVDYSKLKKEGPDF) of

bovine MLRQ. The human MLRQ protein sequence is identical in that region except for

arginine (R) in place of glycine (G) at amino acid position 79. The KLH-peptide was

emulsified by mixing with an equal volume of Freund's Adjuvant and injected into three

to four subcutaneous dorsal sites in 3-9 month old New Zealand white rabbits, for a total

volume of 1 .O mL (0.1 mg of peptide) per immunization (Research Genetics Inc.). The

animals were bled fiom the auricular artery. The blood was allowed to clot and serum

was collected by centrifugation. The anti-peptide antibody titer was determined with

enzyme linked immunosorbent assay (ELISA) with fiee peptide on the solid phase. The

antibody titer at third bleeding was determined to be 5 1,700.

Western blot analysis of MLRQ expression

25 pg of mitochondrial protein as quantified by the Lowry assay (Lowry et ni. 195 1 )

and prepared from human tissues such as brain, hem, liver, kidney, placenta and muscle

(Pitkanen et al. 1996) were electrophoresed on 16% SDS polyacrylamide gels. The

protein was transferred ont0 a nitrocellulose membrane (Xymotech) and subjected to

Western blot analysis. Primary antibodies targeted to the C-terminus of complex 1-

MLRQ and the citrate synthase enzyme (Research Genetics) were used to probe the blot.

The MLRQ antibody was used to probe the blot after the citrate synthase antibody was

stripped using a solution with a final concentration of 6M urea and 50mM Tris-HCI (pH

7.4). a rabbit IgG horseradish peroxidase was used as the secondary antibody.

Developrnent of the autoradiograph was carried out using the ECL kit (Arnersham-

Pharrnacia) according io manufacturer's instructions.

Expression of MLRQ in 50 pg of rnitochondrial protein from patient ce11 line 5624

and control ce11 line 4212 was also examined using antibodies against the MLRQ, ASHI.

B 18 and 49 kDa subunits. The blots were developed using goat anti-rabbit IgG alkaline

phosphatase conjugated secondary antibody and NBT/BCP (Biorad) (McEachern et al.

2000).

Part II. Bacterial expression of MLRQ protein

Design of MLRQ-fusion protein construct

m * * * b

Primers GST3 1 5' CAGTCGGGGATCCCTGCC 3' and GST 304 5' TGAATTCGA

AATCTGGACGTïC 3' (flanking the start and stop codons of MLRQ) were used to

amplify the MLRQ sequence from a human heart cDNA library. * indicates the bases

that were mutagenized in order to engineer the restriction sites EcoRI and BamHI into the

amplified sequence . The engineered MLRQ cDNA fragment was subcloned into the

TA" cloning vector (Invitropn) and then sequenced following the protocols descnbed in

chapter 2. The MLRQ insert of interest was obtained by restricting the recombinant

vector using enzymes BamHI and EcoRI. The pGEX-3X vector (Pharmacia) was also

restricted with the same enzymes and dephosphorylated according to the manufacturer's

protocol. T4 DNA Iigase (Pharmacia) was used to ligate the MLRQ insert into pGEX-

3X. Competent DHa cells were transformed with the pGEX-3X-MLRQ construct and

isolated clones were sequenced to ensure that the Factor Xa cleavage site ("Ile Glu Gly

Arg") in the fusion construct had been preserved. BL21 cells made competent using

RbCl (Maniatis et al, 1989) were then transformed with the sequenced pGEX-3X-MLRQ

constmct.

Induction and purification of MLRQ-GST fusion protein

After successful small-scale expression of the fusion protein. the protocol was

optimized to get a greater degree of solubility for the fusion protein during large scale

expressions. 100 rnL LB media supplemented w ith ampicillin was inoculated with B L2 I

cells canying the pGEX-3X-MLRQ construct and induced with O. 1 rnM IPTG overnight

at 30°C when the OD, reached between 0.6 to 1 .O. Pilot experiments exarnined

expression of the fusion protein at O, 1,2,3,4 hrs and ovemight after induction. Cells

were sedimented by centrifugation at 7,700g in an SS-34 rotor (SomaIl). The ensuing

pellet was resuspended in 5ml of Lysis buffer (20 mM Tris HCI, pH 8.0; 150 rnM NaCI:

1 mM EDTA; 1 mM Dm; 10% glycerol, 0.1 mg/ml Lysozyme and 10 pg/ml leupeptin).

Cells were lysed by freeze/thawing three times. 0.15g urea was added to the cells over 30

minutes at room temperature to give a final concentration of OSM. This was followed by

the addition of Triton x-100 over 1 hour CO give a final concentration of 1 %. Cells were

spun at 12,000 g for 30 minutes in an SS-34 rotor and the "soluble" supernatant was

applied to a prepared 1 ml Glutathione Sepharose 4B column (Pharmacia). The coiumn

was washed with 10 bed volumes of IX PBS. Fusion protein bound to the column was

eluted in lm1 batches by incubating the column in Glutathione elution buffer (10 mM

reduced glutathione in 50 mM Tris-HCl, pH 8.0) for 10 minutes at roorn temperature.

The pooled eluates were then analysed by SDS-PAGE.

Factor Xa cleavage of fusion protein

The pooled eluate from the Glutathione Sepharose 4B column was injected into a

dialysis slide (Pierce) and dialysed overnight in 50 m M Tris HCI, pH 7.5 and 150 rnM

NaCI buffer. Cleavage was performed in a Factor Xa to fusion protein ratio of 1: I O

(w/w) with the addition of 10 pg Factor Xa in a final concentration of 1mM CaC1, to the

dialysed solution and Ieft overnight at room temperature. After digestion, MLRQ protein

was retrieved in the Row-through by column purification on Glutathione Sepharose 4B as

described before and visualized on a SDS-PAGE gel by Coomassie staining.

Part III. Anti-sense expression of MLRQ in marnrnalian cells

Design of sense and anti-sense oriented pREP9 constructs

MLRQ was cloned into pREP9 vectors (a mammalian episomal vector confemng

neomycin resistance) in both the sense and anti-sense orientation. Ptimers Hind4 5'

** * ** TGGTGGCTAGGAAGCTTCTCT 3' and Xho339 5' ACCTTCATTCTCGAGCAGCGT

3' were used to ampli@ a 3 1 1 bp MLRQ fragment fiom human fibroblast cDNA through

PCR. * indicates the bases that were mutagenized to engineer the HindlII and B o 1 sites.

The resulting fragment encompassed 40 bp of the 5' region, the start and stop codons and

26 bp of the 3' region and was cloned in the sense orientation using the engineered

restriction sites. The anti-sense construct contained a 284 bp fragment that was also

** arnplified fiom cDNA using primers Barn30 5' AGTCGGATCCCTCTGCCG 3' and

** Hind339 5' ACCTTGATTCTAAACTTCGT 3'. * indicates the bases that mutagenized

to engineer the BmHI and HindIII sites. The resulting fragment was cloned into the

same vector in an anti-sense orientation to the CMV promoter using these restriction

sites. The anti-sense fragment covered 24 bp of the 5' region, the start and stop codons

and 26 bp of the 3' region. These fragments were f ~ s t subcloned into TA" vectors

(Invitrogen) and then sequenced using the T7 DNA sequencing kit (Arnersham

Pharmacia) before being cloned into the pREP9 vectors.

Optimization of transfection conditions

A vector with a coIourometric tag, pCMVB was used to detemine the DNA versus

ExGen 500 (MB t Fermentas) ratio required for optimal transfections Transfections were

camed out with DNA:ExGen 500 ratios of 2pg:4p1 : 2pg:gpI; 2pg: 1Opl; Zpg: 12pl;

4pg:4p1 ; 4pg:8pl; 4pg: 10p1; 4pg: 12p1: 4pg: 16~1. The desired volume of ExGen 500

diluted with 150 m M NaCI was added dropwise to an equal volume of the desired

quantity of pCMVP vector. also diluted wit!! 150 mM NaCI. The mixture was incubated

for 10 minutes at room temperature. Fibroblasts transformed with SV40 were collected

by centrifugation at 6000 rpm and washed twice in serum-free a-MEM + pyruvate +

uridine media. Cells were resuspended in 0.5 mL of the same media and the ExGen 500

+ plasmid mix was added to the cells. After a 5 minute spin ai 1500 rpm to increase

transfection efficiency, cells were plated in 6-well plates and transfection was allowed to

proceed for 4-5 hours. After washing twice with PBS, 2 rnL of the same media with

serurn was added and culture growth was monitored for two days. Cells were fixed with

2% formaldehydePBS and I mg/rnL X-gal (Gibco BRL) was added. Cells were

observed under the microscope for blue staining as an indicator of transfection efficiency.

Transfection and selection

Both lymphoblasts and transformed fibroblasts were transfected with sense and anti-

sense pREP9 constructs following the optimization protocol outlined above. Selection on

these ce11 lines transfected with the sense and anti-sense pREP9 constructs was begun

three days after transfection by treating the cells with increasing concentrations of the

neomycin analog G-418 sulfate (Gibco BRL), starting with 100 pg/rnL.

Transfection and selection with the linearized vector pCDNA 3.1+

Sense and anti-sense fragments were amplified as described above, cloned into

pCDNA 3.1+ vector and subsequently linearized using PvuI. 10 pl Superfect (QIAGEN)

was added to 2 pg of the DNA (sense, anti-sense or control pCDNA 3.1 +) in 100 pl of

serum-free a-MEM + pyruvate + uridine media and left for 5- IO minutes at room

temperature. SV40 transfomed ceils in 6-well plates were washed twice with PBS. 600

pl of growth media (a-MEM + pyruvate + uridine with sera) was added to the

SuperfecmNA mixture and applied to the cells. 2 rnL of the same media was added 6

hours after transfection. Two days after transfection. selection was started on the

transfected cells with 100 yg/mL G-418 sulfate (Neomycin analog). Three weeks after

the start of selection, viable clones were isolated using cloning rings and grown on

individual 100 x 20 mm plates. Selection was gradually increased upwards to 800 p$mL

of G4 18. Selection on untransfected SV40 transformed fibroblast cells with varying

concentrations of G 4 18 was also concurrently performed.

Western blot analysis of sense and anti-sense expression in transfected ceils

Mitochondria was isolated from pCDNA 3.1 + (control) as well as sense and anti-

sense constnict transfected SV40 cells lines (Pitkanen et al, 1996). 100 v g of

mitochondrid protein as assayed by the Lowry method (Lowry et al, 195 1). frorn various

clones transfected with sense and anti-sense constructs as well as the control fibroblast

line immortdized with SV40 were run on a 16% SDS-PAGE gel and western blotting

was perforrned as described by McEachem et al (2000). Expression levels of MLRQ and

the 49 kDa subunit was detected in these clones using the respective antibodies.

Amplification of MLRQ from transfected celh

Primers PCDNA-R 5' AGAAGGCACAGTCGAGGC 3' (designed from within the

pCDNA 3.1 + sequence) and either PCSense 5' GCCGCAAACATGCTCCGC 3' or

PCAnti 5' GAAATCTGGACGTT CCTTCTT 3' were used to amplify the MLRQ

sequence from sense and anti-sense pCDNA 3.1 + transfected fibroblast clones. PCR

was carried out at an annealing temperature of 50°C for 35 cycles.

Part IV. Association of MLRQ with other complex 1 subunits

Solubilization of beef heart mitochondria

Beef heart mitochondria were isolated as per Pitkanen et al, 1996 and initial studies on

solubilization involved extracting 1 mg/ml of mitochondria for 1 hour at 4°C and

centrifuging the sample at 10.000 rpm for 30 minutes in an Eppendorf centrifuge.

Detergents used included 0.2%, OS%, 1 %, 2% dodecyl maltoside (DDM), 0.5%. 1%. 7 9

Triton X- 100, 1 % octylglucoside, I % CHAPS, 1 % sodium deoxycholate, 1 C/o Nonidet

P40 with 0.5% sodium deoxycholate, 1% SDS, 1% DDM with 1.58 sodium

deoxycholate.

Further extraction studies were perforrned by resuspending beef hean mitochondria in

phosphate buffered saline (PBS) with 0.2% azide to a final concentration of 5 mg/rnL in

the presence of 1 mM aprotinin, 2mM leupeptin and 1 rnM benzarnide protease inhibitors.

Either 1% DDM or 2% Triton-X 100 or 2% Triton-X 1 14 were added to the mitochondria

and the sample was shaken for 1 hour at room temperanire. The sample was spun at

16,500 rpm using the ultracentrifuge rotor SW41 for 45 minutes. The pellet was

resuspended in 0.2% azide PBS. Two further extractions were perfonned on the resulting

supematants following the procedure outlined above. Aliquots were taken from

rnitochondria before DDM addition and also from the pellets and supematants resulting

frorn the three extractions, for western blot analysis. The supematants were concentrated

using Centricon tubes (Amicon) and the Lowry assay was used to quantify protein

(Lowry et al. 1951).

Immunoprecipitation of MLRQ, MWFE and 49 kDa subunits

1 ml of the concentrated supernatant ( I rng/rnl) frcm the first extraction with 18

DDM was treated with 10 pl of pre-bleed (non-immune) rabbit semm for 1 hour at 4 T

on a rocking platform. 50 11 of the homogeneous protein A-agarose suspension was

added to the sample and incubated for at least 3 hours at 4°C on a rocking platfom. The

beads were pelleted by centrifugation for 20 s at 12,000 g in a microfuge. Either 20 pl of

MLRQ antibody or 30 pl of 49 kDa antibody or 40 pl of MWFE antibody were added to

the supernatant and incubated for 1 hour at 4°C on a rocking platform. 50 pl of the

homogeneous protein A-agarose suspension was added to the mixture and incubated

again at 4°C on a rocking platfom for at least 3 hours. The complexes were collected by

centrifugation for 20 s at 12,000 g in a microfuge. The pellet was washed twice in lm1 of

washing buffer 1 (50 rnM Tris-HCl, pH 7.5, 150 m M sodium chloride (NaCl), 1%

Nonidet P40 and 0.5% sodium deoxycholate), twice in washing buffer 2 (50 m M Tris-

HCl, pH 7.5, 500 mM NaCl, O. 1 % Nonidet P40 and 0.05% sodium deoxycholate) and

Iastly in washing buffer 3 (50 mM Tris-HCl, pH 7.5,O. 1% Nonidet P40 and 0.05%

sodium deoxycholate). The washes were dl perfonned at 4°C for 20 minutes on a

rocking pladorm and the pellets were collected by centrifugation as described above.

After the lm traces of the final wash were removed from the agarose pellet, it was

resuspended in 40 pl of SDS gel loading buffer and heated to 100°C for 3 minutes before

being centrifuged for 20 s at 12,000 g in an Eppendorf centrifuge. The

irnmunoprecipitate was then andysed on a western blot with ECL (Amersham-

Pharmacia) development of the autoradiograph according to manufacturer's instructions.

The identity of the protein bands that appear during immunoprecipitation was confirmed

through M.4LDIToF sequencing (Keough et al. 2000) of these fragments excised from

the SDS-polyacrylarnide gels.

Cross-linking with DST and EGS

4.0 mg/mL of purified bovine heart mitochondria were dialyzed against 50 mM

Triethanolamine pH 8.0, containing 0.25 M sucrose at 4°C for 6 hours. 0.2 mM EGS

(Ethylene glycolbis(succinimidylsuccinate)) and 1 mM DST (Disuccinimidyl tartrate)

were used to cross-link samples of these mitochondria for 1 hour as described by

Yamaguchi and Hatefi (1993). However, Triton-X 100 was not added to the cross-linked

samples and the reaction was quenched by the addition of 5mM glycine for 30 minutes.

Subsequent experiments used 0.2M or 0.5M of the cross-linker EGS and the cross-linking

reactions were carried out for 2 hours at which point they were quenched with 20mM

glycine for 30 minutes.

Immunoprecipitation of cross-linked bovine heart mitochondria

The mitochondria cross-linked tvith 0.5 M EGS were solubilized with 1% DDM as

described in previous sections and the resulting supernatant diluted in 1mL PBS-aide ( 1

mg/mL) was used in immunoprecipitation reactions using both MLRQ and MWFE

antibodies, also as outlined in previous sections.

SDS-PAGE and western blot analysis of MLRQ during extraction,

immunoprecipitation and cross-linking

For visualization of protein, cross-linked samples as well as cross-linked and

immunoprecipitated samples were run on 520% Tris-glycine gradient gels (Novex) and

stained with Coomassie blue. Quantitation of protein sarnples was performed by the

Lowry assay (Lowry et al, 195 1 ). For western blot analysis, samples were

electrophoresed similarly but transferred ont0 nitrocellulose membranes (Xymotech) and

developed using ECL (Arnersham-Pharmacia) reagents as mentioned above.

Results and Discussion

Part 1. Determining MLRQ form, fùnction and expression

Tissue expression of MLRQ

Western blot analysis of mitochondria isolated fiom various human tissues (Fig. 3.1)

revealed that there is little variation in the apparent rnolecular mass of the MLRQ

subunit. The citrate synthase enzyme which is part of the TCA cycle was used as a

positive control and is reflective of the number of mitochondria is the tissue. Expression

of MLRQ was seen to be highest in heart (lane 2), an organ where cellular energy

production is most essential. The lowest expression of MLRQ was seen in mitochondria

from placenta (lane 5). Also, when compared to the expression of citrate synthase?

greater expression of MLRQ is seen in liver and kidney tissues (lanes 3 & 4). The high

level of MLRQ expression in liver is supported by results From Northem analysis (Fig.

MLPQ

Figure 3.1. Tissue specific expression of the MLRQ subunit of complex 1. 25 pg of total mitochondrial protein was separated by SDS-PAGE using a 16% polyacrylamide gel. The separated proteins were subjected to Western blot analysis. The blot was developed using horseradish peroxidase linked to anti rabbit IgG. Lane 1, brain; lane 2, heart; lane 3, livet; lane 4, kidney; lane 5, placenta; lane 6, muscle.

2.5) which also showed increased amounts of NDUFA-I transcnpts in this tissue.

However, the magnitude of MLRQ expression in tissues such as brain and skeletal

muscle (lanes 1 and 6) seems to be lower than that indicated by Northem analysis of

NDUFA 4 transcnpts in the same tissues. The level of MLRQ expression was also seen to

parallel that of the 49 kDa iron-sulfur subunit in other westem blots. Detection of the

MLRQ subunit with the expected molecular mass of 9 kDa suggests two possibilites.

One possibility is that this subunit undergoes no post-translational modifications that

significantly influence its molecular mass. The other is that the MLRQ subunit in al1

tissues is post-translationally modified.

Although expression of the MLRQ subunit was examined in many cases of complex 1

deficiency, noteworthy information about this subunit was obtained from the examination

of only one patient (5624), diagnosed with hepatomegaly with renal tubulopathy (HT).

No mutations were identified in the cDNA of this patient (see mutational screening in

previous chapter) and the size of the MLRQ protein as detected by westem blotting was

at the expected MW. However, the magnitude of MLRQ expression in 5624 (lane 1).

seemed to be two-fold that of the control4212 fibroblast ce11 line (iane 2) (Fig. 3.2d).

This is of significance, especially considering the fact that amounts of other complex 1

subunits such as 49 D a , ASHI and B 18 were underexpressed in this complex 1 deficient

ce11 line when cornpared to the control (Fig. 3.2a,b,c). These immunoblotting

experiments as well as the fact that no mutations were detected in any of the subunits

screened (49 kDa, ASHI, B 18, MLRQ, MWFE, PGIV, AQDQ, TYKY and PSST),

suggest a problem in the assembly of complex 1. In light of these fmdings, it seems likely

that the expression of the MLRQ subunit is regulated by a diEerent mechanism than that

Figure. 3.2. Western blot analysis on mitochondria isolated from cultured skin fibroblasts of a patient (5624 - HT) and control(4212) using various complex 1 antibodies. Mitochondna (LOO pg of protein) from cultured skin fibroblast cultures of (1) 5624 - a patient with Hepatomegaiy with renal tubulopathy (HT) and (2) 42 12 - a control subject, were run on a 16% polyacrylamide gel. electroblotted onto a nitrocellulose matnx and irnmunodetected with antibodies to the following cornplex 1 subunits (a) 49 kDa (b) ASHI (c) B 18 and (d) MLRQ. Immunoreactive proteins were visualized with aikaline phosphatase conj ugated anti-rabbit Ig G.

responsible for other complex I subunits. Ntematively, it is also a possibility that MLRQ

is one of the first subunits to be incorporated during the assembly of complex 1 and

therefore not affected by the titre of the other subunits that follow.

Bacterial expression of MLRQ: Attempts at defining subunit structure

In order to determine the secondary structure of the MLRQ protein through CD

spectroscopy and NMR, bacterial expression of the MLRQ protein was undenaken.

Initial efforts to express the MLRQ protein in bacteria included using pProEX Hta (Gibco

BRL) as well as PET-21 d(+) (Novagen) systems, both using a Histidine tag for

purification. Attempts to induce the protein with IPTG however proved unsuccessful

with both systems. Rationalizing that this was probably because proteins less than 1 O

kDa in size like MLRQ are generally difficult to express stably in E.coii because they

cannot fold correctly and are often subject to proteolytic degradation. expression of

MLRQ as a Glutathione-S-transferase (GST) fusion protein was attempted.

Preliminary trials revealed that bacterial expression of the MLRQ-GST fusion protein

becarne optimal at a temperature of 30°C and that greater expression of the protein

occurred during ovemight growth at that temperature. Expression was seen to increase

proportionally fiom 1 to 4 hours of induction and level off at ovemight induction (Fig.

3.3). Purification of the MLRQ-GST protein on GST Sepharose columns was also

successfùl, yielding pure samples of the fusion protein as determined on SDS-

polyacrylarnide gels (Fig. 3.4a,b,c). As seen in Fig. 3.4a (lanes 2,3 and 4), a

considerable amount of the fusion protein remains insoluble. Inducing the BL2 1 cells

carrying the pGEX-3X-MLRQ vector at a temperature below 30°C may have increased

- 35 kDa Fusion protein

Figure 3.3. Bacterial expression of the MLRQ subunit as a GST-fusion protein. Temporal induction of BL2L cens canying the pGEX-3X-MLRQ vector with

0.1 m M IPTG at 30°C. Cells were induced when OD6w reached a value of 0.7. Lane 1, uninduced sample at OD600= 0.7; lanes 2-6, induced samples after 1.2.3,4

houa and overnight, respectively. Arrow indicates the MLRQ-GST fusion protein with a combined rnolecular mass of 35 kDa.

OM Factor Xa Flow-through from column Elution of bound protein Cleavage following O/N Factor Xa cleavage

II n -1

Figure 3.4. mcation and cleavage of GST-MLRQ fusion protein. a) Purification of GST- MLRQ fusion protein on a Glutathione Sepharose 4B colurnn. Lane 1, E.coli BL2 l/pGEX-3X- MLRQ lysed (soluble fraction): Iane 2, lysed insoluble fraction; lanes 3 and 4, Bow-through from column upon applying soluble fraction; lanes 5 and 6, first and last washes of column. b) Lanes 1 and 2, fusion protein eluates upon addition of glurathione elution buffer c) Digestion products from overnight cleavage of GST-MLRQ fusion protein in a Factor Xa to fusion protein ratio of 1 : 10. d) Lanes 1 and 2, Flow-through from Glutathione Sepharose 4B column following application of Factor Xa cleavage products. The 35 kDa protein represents the GST-MLRQ fusion protein; the GST protein is 26 kDa and the MLRQ protein is 9kDa. Al1 protein samples were loaded onto 16% SDS-polyacrylamide gels and stained with Coomassie blue dye.

solubility of the fusion protein. Cleavage of the fusion protein with Factor Xa proved to

be quite difficult and successful cleavage was achieved oniy once (Fig. 3.4d).

Anti-sense expression of MLRQ: Attempts to determine subunit function

The anti-sense strategy was based on the principie that the anti-sense MLRQ mRNA

made by the expression vector introduced into a ce11 can undergo Watson-Crick

hybridization to MLRQ mRNA or pre-mRNA and may via a variety of mechanisms

inhibit translation of that mRNA into MLRQ protein. Overexpression of anti-sense

rnRNA c m modulate the transfer of information fiom the gene to the protein by

interfering with various levels of the process which includes transcription. RNA splicing.

polyadenylation. translation or termination by preventing the binding of necessary

regulatory proteins, interfering with the export of mRNA from nucleus to cytosol as well

as by inducing structural changes in RNA by binding to it that would result in its

degradation. It was postulated that by specifically blocking synthesis of the MLRQ

protein, its role in the assernbly and function of complex I could be determined.

The "sense" MLRQ consaict encompassed 40 bp of the 5' region. the start and stop

codons and 26 bp of the 3' region. The "anti-sense" MLRQ construct somewhat mirrored

the "sense" constmct covering 24 bp of the 5' region, the start and stop codons and 26 bp

of the 3' region but cloned in an anti-sense orientation to the CMV promoter on the

expression vector. Transfection efficiency was determined using different DNA to

transfection reagent (canier) ratios and was found to be optimal at 2 pg DNA with 10 pl

of the reagent. Transfection efficiency was determined using pCMVB which carries the

B-gdactosidase gene and changes colour upon addition of X-gai. After the optimization

described in the Materials and methods section, the pREP9 syaem was used to deliver

MLRQ in both "sense" and "anti-sense" orientations into both transformed fibroblast and

1 ympho blast contro 1 ce11 lines. Selection studies showed that non-tram fected control

lymphoblast and transfonned fibroblasts were both killed at a G4 1 8 concentration of 100

pg/mL. However, ce11 lines transfected with "sense" and "anti-sense" constnicts did not

do much better, failing to thrive past G418 concentrations of 200 pg/ml. Using

lymphoblasts as a recipient ce11 line also proved to be a problem, as it was hard to readily

distinguish the dead cells (non-transfected) fiom the live ones.

A new vector pCDNA 3.1+ was used as a shuttle to deliver these "sense" and "anti-

sense" MLRQ constructs into transformed fibroblasts, as this linearized vector was

capable of integrating into the cell's chromosomes, thereby making for stable

transfection. With the new vector and Superfect as the carrier, both "sense" and "anti-

sense" transfected cells were now able to survive G418 concentrations up to 600 pg/ml.

At a concentration of 800 pg/ml G4 18 however, clones transfected with the "sense"

vector survived, whereas those transfected with the "anti-sense" vector showed poor

growth. Mitochondna isolated fiom these clones at this point were subjected to

imrnunodetection with MLRQ antibody, but no difference in MLRQ protein expression

was detected between the "sense" and "anti-sense" clones or clones transfected with

pCDNA 3.1 + as control (Fig. 3.5).

Explanations for the inefficacy of the anti-sense strategy becarne somewhat apparent

when the MLRQ sequence in the respective orientation was successfully arnplified using

pnmers designed from within the pCDNA3.1 + vector fiom cells transfected with "anti-

sense" (Fig. 3.6, lane 4) and "sense" (Fig. 3.6, lane 6) constnicts but not fiom the

control clones (Fig. 3.6, lanes 1 and 2). Since it was known that the "sense" and "anti-

MLRQ - Figure 3.5. Western blot analysis of MLRQ expression in SV40 immortalized fibroblasts transfected with sense and anti-sense pCDNA 3.l+ constmcts. (a) 100 pg of rnitochondna isolated from fibroblasts lane 1, control untransfected; ln 2, transfected with pCDNA 3.1 + vector: In 3. transfected with anti-sense MLRQ pCDNA 3.1 + construct: In 4. transfected with sense MLRQ pCDNA 3.1+ constmct were subjected to western blot analysis with antibodies to the 49 kDa, B 18 and MLRQ subunits of complex 1. The transfected cells were harvested at 200 pg/pl of G418 resistance. (b) 50 pg of mitochondria isolated from fibroblasts lane 1. transfected with pCDNA 3.1+; ln 2, transfected with anti-sense MLRQ pCDNA 3. I+ construct; In 3, transfected with sense MLRQ pCDNA 3.1+ construct were irnmunodetected with antibodies to the 49 kDa and MLRQ subunits. Transfected cells were harvested at 800 pg/pl of G418 resistance.

Figure 3.6. PCR amplification of sense and anti-sense MLRQ sequences to confim transfection of fibroblasts. Lanes 1 and 2 are negative controls. Lanes 3 and 5 are positive controls. PCR amplification of Lane 1, anti-sense MLRQ sequence from pCDNA 3.1+ uansfected fibroblasts; ln 2, sense MLRQ sequence frorn pCDNA 3.1+ transfected fibmblasts; In 3, anti-sense MLRQ sequence frorn anti-sense MLRQ pCDNA 3.1+ vector; In 4. anti-sense MLRQ sequence from anti-sense MLRQ transfected fibroblasts harvested at 400 pg/pl G418 resistance; ln 5, sense MLRQ sequence from sense MLRQ pCDNA 3.1+ vector; ln 6, sense MLRQ sequence from sense MLRQ transfected fibroblasts harvested at 400pg/pl of G418 resistance. PCR reactions were

carried out 35 cycles at an annealing temperature of 50%. PCR products were run on a 1% agarose gel.

sense" MLRQ constnicts had been successfully delivered to the nucleus of the ce11 by the

carrier, this suggested a couple of other factors that were at play here. Firstly. due to

high-order mRNA structure, the construct designed to target the mRNA of MLRQ may

not have been effective in inhibiting its expression. Secondly. the presence of other

MLRQ-like mRNAs rnay have interfered with the anti-sense RNA. Lastiy, the

transcription rate of the MLRQ sequence in the "anti-sense" orientation rnay not have

been high enough to match the transcription of the native MLRQ mRNA in the cell. This

may have also been compounded by the slow rate of degradation or chemical half-life of

the native MLRQ mRNA in the cell compared to that of the anti-sense RNA. With

steady-state levels of MLRQ rnRNA not affected, an anti-sense effect at the protein level

would not be seen. The fact that clones with the expression vector containing the MLRQ

sequence in the "sense" orientation did not lead to increased levels of MLRQ protein ma.

also be attributed to the faster rate of degradation of any MLRQ mRNA produced by the

expression vector. Protection of mRNA by formation of ribonucleoproteins is also a well

documented phenornenon which has been reported to impair binding of anti-sense RNA

(Strickland et al. 1988). Conclusive statements about the expression of MLRQ mRNA.

in either sense or anti-sense orientations can only be made by examining levels of the

respective RNAs on a Northem biot.

Part II. Subunit interactions of MLRQ within complex 1

Detergent solubïiization of complex 1 subunits

Vatying concentrations of many detergents and detergent cocktails were used to

extract MLRQ fkom bovine heart mitochondria. Non-ionic detergents such as OctyI

glucoside, DDM, Triton X- 100, Triton X- 1 14 and Nonidet-P40, anionic detergents suc h

MLRQ -e

1 % Octyl 1°/6 DOM 1% CHAPS glucoside

1 2 3 4 5 6 S P S P S P

1°h DDM/l.S% sodium 1% NP40I0.S0h sodium deoxycholate deoxycholate

11

0.2% DDM 1 SDS 1 % Triton x-100

I n n 1 2 3 4 5 6 P S S P S P

49 kDa - *A-

MLRQ u

Figure 3.7. Extraction of the MLRQ subunit from bovine heart mitochondria. Mitochondria isolated from bovine heart were extracted with a variety of detergents and detergent cocktails and the solubilized protein was irnmunodetected with the MLRQ andfor 49 kDa antibodies (as indicated by arrows) to determine extraction efficiency. Al1 lanes contain 20 pg of extracted protein from single extractions (unless othenvise specif'ed) with the indicated detergent(s) S, supernatant fraction; P, pellet fraction; suffix of 1 or 2 indicate the number of extractions that resulted in the fraction (a) Lane 1 contains purified bovine heart rnitochondria. Lanes 2-8 represent corresponding pellets and supematants resultiog from extractions with: In 2, 1 Qo Octyl glucoside; In 3, 1 % DDM; In 4, 1 % CHAPS; ln 5, 1 % Deoxycholate: ln 6,0.5% Deoxycholate: ln 7, 0.5% Triton x-100 ; In 8, 1% Triton x-100 (b) Extraction with: lanes f -3, 1% DDM/1.5% sodium deoxycho1ate;lanes 4-6, 1 % Nonidet-P4010.546 sodium deoxycholate (c) Extractions with: lanes 1-2, 1 BDDM; lanes 3-4, 1% CHAPS; lanes 5-6, 1% Octylglucoside. (d) Extractions with: lanes 1-2,0.2% DDM, In 3-4, 1% SDS; lanes 5-6, 1% Triton x- 100

as SDS and sodium deoxycholate as well as zwitterionic detergents such as CHAPS were

used in order to determine the detergent solubility of MLRQ as compared to other

cornplex I subunits (Fig. 3.7). Preliminary experiments with detergents demonstrated

that DDM (Fig. 3.la, lane 3) and Triton X-100 (Fig. 3.7a, Ianes 7,s) were the most

effective at solubilizing complex 1 subunits. This can be noted by observing relative

amounts of the MLRQ subunit present in the pellet versus the supematant after extraction

with various detergents (Fig 3.7a). Although DDM has becorne the detergent of choice

in many expenments involving solubilization of other respiratory chah enzymes. only a

recent study corroborates o u . findings on the usefulness of this deterpnt in the

purification of complex I (Okun et al, 2000). The non-denaturing properties of DDM

make it a favourable detergent for selective solubilization. so that close subunit-subunit

interactions within the complex are preserved for examination (Okun et ul. 1000). n-

dodecyl-P-D-maltoside (DDM) is an analogue of octyl glucoside that possesses small.

unifonn micelles, a simple and chemically well-defined structure as well as a high critical

micelle concentration that permits easy removal by dialysis (Rosevear et al, 1980). The

corûormation of the sugar moiety on DDM has been implicated as the critical factor in

influencing the micelle forming abilities of this detergent (Rosevear et al, 1980).

However, even though subunits such as 49 kDa, B 18. B 17 and MWFE were successhilly

removed from the pellet and into the supematant with a single extraction with DDM.

solubilization of MLRQ proved to be more dificult, with a good portion of it still

remaining in the pellet fraction even after three extractions with the detergent (Fig. 3.8).

This result is quite interesting taking into consideration the fact that other subunits

belonging to the hydrophobie (HP) fraction such as B 18, B 17 and MWFE were easily

MLRQ

MWFE

Figure 3.8. Solubilization of the MLRQ subunit compared to other complex 1 subunits. Three extractions of bovine hem mitochondria were performed with 1% DDM and the supernatant and pellet fractions collected each time. 20 pg of each fraction was loaded ont0 a 16% SDS-polyacrylamide gel and western blot analysis was performed with antibodies to the complex 1 subunits MLRQ, 49 kDa, B 17, AQDQIl 8P, B 18 and MWFE. C, (control) pufied bovine heart mitochondria; SI. supematant from 1st extraction; Pl, pellet from fmt extraction; S2, supernatant from 2nd extraction; PZ, pellet from 2nd extraction; S3, supematant from 3rd extraction; P3, pellet from 3rd extraction.

Amino acids

(A)

N-terminus INNER MITOCHONDRIAL MEMBRANE

Figure 3.9. Hydropathy pronle and membrane orientation of the MLRQ polypeptide. (a) Hydropathy plot for the MLRQ subunit based on the hyrophobicity indices of Kyte and Doolittie (1982) averaged across six residues. (b) Redicted membrane topology of MLRQ based on its amino acid sequence.

extncted compared to MLRQ. While the 49 kDa subunit and MWFE have been assigned

to the a-subcomplex of the enzyme, the B 18 and B 17 subunits are achidly part of the

mainly hydrophobic P-subcomplex with 8 18 in the IPS fraction and 8 17 in the IPL

fraction associated with the ND4 and ND5 subunits. The Kyte-Doolittle hydropathy

profile indicates that the human MLRQ subunit has the potential to be folded into a

membrane spanning a-helix (Fig. 3.9). The fira 39 arnino acids are relatively

hydrophobic and it is thought that this protein is anchored in the inner membrane and

interacts via more hydrophilic parts of the sequence with other subunits of complex 1.

However. the extremely low detergent solubility of the MLRQ subunit compared to the

aforementioned hydrophobic subunits of the complex suggests that it is perhaps

embedded more deeply and heavily bound to inner membrane phospholipids than

originally hypothesized.

Proximity and association of MLRQ with other cornplex 1 subunits

Immunoprecipitation after extraction with detergents was carried out in order to

determine the association of MLRQ with other subunits of complex 1. Subunits that are

in close association with MLRQ should irnmunoprecipitate along with it. Antibodies to

either the 49 D a , MLRQ or MWFE subunits were added to 1 % DDM solubilized bovine

heart mitochondna followed by pelleting the immunoprecipitate with protein A agarose

and analyses of the immunoprecipitates through SDS-PAGE and western blotting. The

MLRQ antibody (produced in rabbit) proved to be quite effective in immunoprecipitating

the subunit frorn bovine heart mitochondria (Fig. 3.10a, lane 4). The MLRQ subunit at 9

D a and the IgG at 55 kDa and 30 kDa (doublet) were visible upon western blotting with

the MLRQ antibody. The identity of IgG was confirmed through MALDIToF protein

Figure 3.10. Immunoprecipitation studies using protein A agarose. Bovine heart mitochondria was extracted with 1% DDM and immunoprecipitated with MLRQ antibody as detailed in the materials and methods section (a) Lane 1. supernatant from immunoprecipitation after addition of protein A-agarose and MLRQ antibody; lane 2. fint wash with washing buffer 1 (50 m . Tris-HCl, pH 7.5. 150 rnM NaCl. 1 % Nonidet P-40 and 0.5% sodium deoxyc holatef: lane 3. last wash with washing buffer 3 (50 m M Tris-HCI, pH 7.5,O. 1% Nonidet P-40 and 0.05% sodium deoxycholate); lane 4. MLRQ immunoprecipitate. Immunodetection with the MLRQ antibody (ln 4) gives 4 bands: 9 kDa (MLRQ). 55 kDa (IgG) and a doublet at around 30 kDa (b) 25 pg of the imrnunoprecipitated sample was therefore run on a 520% Tris-glycine gradient gel, stained with coomassie blue and the bands excised and sequenced through MALDI-ToF peptide m a s fingerpnnting. The double band at 30 D a was found to be pan of the IgG.

sequencing of these bands (Fig. 3.10b). Ease of immunoprecipitation with the MLRQ

antibody which is directed at epitopes at the C-terminus of the subunit. suggests that the

hydrophilic C-terminus is a region of MLRQ that is exposed. As a positive control. the

supernatant from bovine heart rnitochondria after 3 extractions with 2% Triton X-IO0

which was previously shown to have solely the MLRQ subunit. also showed similar

results upon immunoprecipitation. In order to confirm that the band appearing at 9 kDa

was not just a ubiquitous fragment. the same sample immunoprecipitated with MLRQ.

treated with 1% SDS did not produce the 9 kDa MLRQ band upon immunoblotting with

the MLRQ antibody.

Immunoprecipitation of MLRQ and immunoblotting the resulting pellet with the

respective complex 1 antibodies revealed that neither the 39 D a , 15 kDa. 18 IP (AQDQ)

or ASHI subunits have an affinity for MLRQ in order to be immunprecipitated dong with

it (Fig. 3.11a, c, d, lane 2). MWFE and the B 18 subunit to sorne extent were the only

subunits detected in the MLRQ immunoprecipitate (Fïg. 3.11b, lane 2). In order to

confirm this. the MWFE and 49 kDa immunoprecipitates were probed with the MLRQ

antibody. The 9 kDa MLRQ subunit was indeed detected in the MWFE

immunoprecipitate (Fig. 3.11a, lane 3) and to a lesser extent in the 49 kDa

immunoprecipitate (Fig. 3.11a, lane 4). In addition, immunoblotting detected B 18 (Fig.

3.1 1 b, lane 3) and the 49 kDa subunit (Fig. 3.1 lc, lane 3) in the MWFE

immunoprecipitate and B 18 (Fig. 3.11b, lane 4) and MWFE (Fig. 3.11b, lane 4) in the

49 kDa immunoprecipitate, again in small amounts. Neither AQDQ (Fig. 3.11a, lanes 3

and 4) nor ASHI (Fig. 3.11c, laries 3 and 4) were detected in the MWFE or 49 D a

immunoprecipitates. These results indicate association between the 49 kDa, B 18, MLRQ

AQDW l8lP

MLRQ

Figure 3.11. Irnmuoblotting of the immunoprecipitated MLRQ, MWFE and 49 kDa subunits of complex 1 to determine subunit association. Bovine heart rnitochondna solubilized with 1 G/c DDM were immunoprecipitated with MLRQ, MWFE and 49 kDa antibodies (see materials and methods section), run on 520% Tris-Glycine gradient gels and transfemd to nitroceliulose membranes for western blot analysis. 20 pg of protein sarnple were loaded in each lane. Lane 1. control bovine heart mitochondria; lane 2, MLRQ immunoprecipitant; lane 3, MWFE immunoprecipitant; lane 4.49 kDa irnmunoprecipitant. Blots were irnmunodetected with antibodies to the following subunits (a) MLRQ and AQDQA8IP (b) MWFE and B 18 (c) 49 kDa and ASHI (d) 39 kDa and 15 kDa

and MWFE subunits. Of these, the association between MLRQ and MWFE seems to be

the most significant. The fact that MWFE was able to CO-immunoprecipitate MLRQ but

not vice-versa might be explained by the fact that the region of interaction between

MLRQ and MWFE may have been masked by the interaction of the MLRQ antibody

with MWFE. Another possibility is that the binding of the MLRQ antibody to the MLRQ

subunit may have in fact distorted the protein such that it no longer interacted with

MWFE. These findings support the emerging structura1 mode1 of complex 1 which

purports that MLRQ, MWFE and the 49 kDa subunit belong to the a-subcomplex of the

enzyme. Even though the 49 kDa subunit belongs to the more hydrophilic Ih fraction

within the ia subcontext. it is known to form part of the pocket for quinone binding and

is said to be in proximity to the ND1 subunit which is part of the membrane a m . Given

this information, it is likely that both MLRQ and MWFE which have a transmembrane

domain do interact with the 49 kDa subunit. The following table summarizes the

information obtained from immunoprecipitation studies.

Table 3.1. Subunit proximity in complex 1 as determined by immunoprecipitation studies

Subunits immunoprecipitated with

Subunits immunodetected MLRQ MWFX 49 kDa

MLRQ Yes MWFE S1 49 kDa No 18 IP (AQDQ) No 39 kDa No 15 kDa No ASHI No BI8 SI Not ck - Not checked; SI - slightly

Yes Yes SI No Not ck Not ck No SI

Sl S1 SI No Not ck Not ck No SI

Figure 3.12. hunob lo t s of 1mM DST and 0.2mM EGS cross-linked bovine heart mitochondria with antibodies to the MLRQ and 49 kDa subunits. Bovine hem mitochondria cross-linked with ImM DST and 0.2mM EGS for 1 hour and quenched with 5mM glycine, was analyzed by gel electrophoresis on 5-20% Tris-glycine gradient gels and immunoblotted as described under materials and methods. (a) Immunobloning was done with antibody to the 49 kDa subunit. Lanes 1 and 5,20 pg of non cross-linked bovine heart mitochondria; lanes 2 and 6.40 pg of cross-linked mitochondria; Ln 3 and 7, 20 pg of cross-linked mitochondria; lanes 4 and 8.8 pg of cross-linked mitochondria. (b) Immunoblotting was done with antibody to the MLRQ subunit Lane 1,20 pg non cross-linked bovine heart mitochondria; lanes 2 and 5,40 pg of cross-linked mitochondria; lanes 3 and 6,20 pg of cross-linked mitochondria; lanes 4 and 7,s pg of cross-linked mitochondria. Al1 lanes 2-4 contain IrnM DST cross-linked samples while al1 Ianes 6-8 contain 0.2mM EGS cross-linked samples.

Cross-linking and immunoprecipitation studies

Chernical cross-linkers EGS and DST are both water-insoluble, homobifunctional N-

hydroxysuccinimide esters (NHS-esters). P n m q amines such as the a-amine groups

present on the N-termini of peptides and proteins as well as in some cases, the side chains

of amino acids react with these NHS-esters to form covalent amide bonds. Prelirninary

expenments exarnined the cross-linking of the MLRQ, MWFE and 49 kDa subunits of

bovine heart mitochondria with 1m.M DST and 0.2 mM EGS by detecting with the

respective antibodies. At such low cross-Iinker concentrations, MWFE was not found to

cross-link with either DST or EGS. However, a slight band of 20 kDa is visible when the

cross-linked mitochondna is immunodetected with the MLRQ antibody (Fig. 3.12b. lane

2). In addition to the possibility of MLRQ associating with other subunits approximately

I 1 D a in size such as B8, B9 or AGGG (IO kDa), this band could also represent a dimer

of MLRQ (9 + 9). The 49 kDa subunit was however found to clearly cross-link, yielding

a band of approximately 60 kDa (Fig. 3.12a, lanes 2 ,3 ,6 and 7) and could in fact

represent the association between the 49 kDa subunit and a 13 kDa subunit of IP (49 +

13). This is in agreement with studies performed by Yamaguchi and Hatefi (1993) who

have shown the 49 kDa subunit to be linked to a 13 kDa subunit using an antibody made

to a mixture of the two 13 kDa proteins, B 13 and DDGD. The cross-linking pattems of

DST and EGS were essentially the same, despite the considerable differences in the

lengths of the reagents (molecular lengths of DST and EGS being 0.64 and 1.6 1 nm.

respectively).

When cross-Iinker concentrations of EGS were increased to 0.2M and OSM, more

obvious cross-linking pattems began to emerge. No differences between cross-linking

0.2M EGS 0.5M EGS Control 0.2M EGS 0.5M EGS Control 1 2 3 1 2 3

0.2M EGS O.5M EGS Control 0,2M EGS OSM EGS Control 1 2 3 1 2 3

Figure 3.13. Immunoblotthg of EGS cross-linked bovine heart mitochondria with complex 1 antibodies. 4 m g h l bovine heart mitochondria was cross-linked for 2 hours and quenched with 20mM glycine and analysed on western blots as described under materials and methods. Lane 1, 20pi of 0.2M EGS cross-Iinked rnitochondria (4 rngfml): ln 2,20@ of 0.5 M EGS cross-inked mitochondria (4 rnglrnl); in 3.20 pg of non cross-linked mitochondria. Immunoblotting was performed with antibodies to (a) AQDQ/18 IP subunit (b) ASHI subunit (c) MLRQ subunit (d) MWFE subunit (e) 49 kDa subunit (0 B 17 subunit (g) B 18 subunit. Molecular masses are indicated for prominent cross-linked products.

patterns were observed using 0.2M or 0.5M EGS. Cross-linking at these higher

concentrations and probing with 18 IP (AQDQ) revealed a double band approximately

36- and 40-kDa in size (Fig. 3.13a, lanes 1 and 2). The smaller of these unidentified

bands could represent dimerization of the AQDQ subunit ( 18 + 18) and was in fact a

possibility proposed by Yamaguchi and Hatefi (1993). who observed a similar sized band

when the EGS cross-linked LP fraction was probed with the 18IP antibody.

Probing the cross-linked mitochondria with the ASHI antibody, revealed a prominent

double band at 38- and 43-kDa as well as less prominent bands at approximately 60- and

1 10 kDa (Fig. 3.13b, lanes 1 and 2). Again. the smallest of the bands may represent a

dimer of the subunit ( 18 + 18).

Cross-linked samples immunodetected with B 17 revealed just one prominent band at

33 kDa (Fig. 3.13f, lanes 1 and t), while the same samples immunodetected with B 18

revealed two prominent bands at 30 kDa and 34 kDa as well as other bands of higher

mo1ecuIar weight and lesser intensity (Fig. 3.13g, lanes 1 and 2).

Cross-linking with 0.2M and 0.5M of EGS resulted in the appearance of many cross-

linked bands when probed with the 49 kDa antibody. the most prominent ones being at 60

kDa, 79 kDa, 103 kDa and 135 kDa (Fig. 3.13e, lanes 1 and 2). As mentioned above.

the 60 kDa band probably represents association between the 49 kDa subunit and one of

the 13 kDa subunits. The 79 kDa band most likely represents the binding of the 30 kDa

subunit (49 + 30), aiso detected by Yamaguchi and Hatefi (1993). The myriad of other

associations can only be speculated upon in a similar context. The fact that the 49 kDa

subunit is not buried in the membrane is evidenced by the number of cross-linking

associations formed by this subunit as compared to the other subunirs examined here.

Immunoblotting these cross-linked samples with the MLRQ antibody, detected a very

prominent band at approximately 20 kDa, a lighter band at approximately 3 1 kDa and a

slight doublet at 40 kDa and 45 kDa (Fig. 3.13c, lanes 1 and 2). While the prominent

band at 20 kDa may represent the aforementioned dimerization of MLRQ and the doublet

at 40- and 45- kDa, dimeric or trimeric cross-linking to other subuniü in the vicinity. the

band at 3 1 kDa may in fact represent cross-linking of MLRQ to the MWFE subunit. This

possibility arises from the presence of a single band of the same molecular mass detected

in the cross-linked samples using the MWFE antibody (Fig. 3.13d, lanes 1 and 2). A

molecular mass of 3 1 kDa however indicates that MLRQ and MWFE. in addition to their

own association are also closely linked to other complex 1 subunits. Even so. this

assumption leads to the question as to why no dimeric association between the two

subunits was detected. One possibility is that there is no direct binding between MLRQ

and MWFE and that any association between the two subunits, no matter how close. is

through an intermediate subunit or subunits. A similar scenerio is evident in the cross-

linking experiments of Yamaguchi and Hatefi (1993) who detected trimeric cross-linking

between the 75 kDa, 5 1 kDa and 24 kDa subunits, but no dimeric associations between

75- and 24- kDa or 75- and 9- kDa subunits. It has been established through other

expetiments as well, that the 9 kDa and 24 kDa subunits are linked to the 75 kDa subunit

as well as to the rest of complex 1, solely through their association with the 5 1 kDa

subunit.

Assuming trimeric cross-linking between these subunits (9 + 6 + X), subunit X which

is approximately 16 kDa in size, can be any one of B 18, SGDH (1 8 kDa) or PGN ( 19

kDa). ASHI (18 kDa) and B 17 can be eliminated as probable candidates for subunit X.

because a band of approximately 3 1 kDa was not visudized in the cross-linked samples

probed with either the ASHI (Fig. 3.13b, lanes 1 and 2) or B 17 (Fig. 3.13f, lanes 1 and

2) antibody. A cross-linked product approximately 3 1 kDa in size was detected with the

B 18 antibody (Fig. 3.13g, lanes 1 and 2) and this subunit could very well be the

intermediary subunit between MLRQ and MWFE, especially because it had been found

previously to immunoprecipitate with both subunits (Fig. 3.1 1b, lanes 2 and 3).

However. the validity of this association is questioned by the fact that cross-linked

products allowing for dimeric cross-linking between B 18 and MLRQ or MWFE was not

detected. While PGIV belongs to the Ia fraction of complex 1. as do MLRQ and M W .

the SGDH and B 18 subunits belong to the P subfraction. A four subunit association

including MLRQ and MWFE would involve only the smaller subunits such as B8. B9.

SDAP ( 10 kDa), MNLL (5.5 kDa), AGGG ( 1 O kDa) or KFYI ( 5 D a ) . However, without

antibodies to confim the identity of these other subunits, this is merely conjecture at this

point. Table 3.2 lists the cross-linked products detected by the complex 1 antibodies used.

As expected, a greater number of cross-linkings were seen with subunits that are not

embedded in the mitochondrial membrane.

Since complex 1 has 43 subunits and antibodies to most or al1 of them were not

available to this study, immunoprecipitation of the cross-linked proteins was attempted in

order to obtain enough of the cross-linked proteins on a SDS gel for protein sequencing.

Immunoprecipitation of cross-linked samples with MLRQ was not very successful

yielding only very minute quantities of MLRQ and none of any of the other subunits

cross-Iinked to it (Fig. 3.14% lane 3). Immunoprecipitation with MWFE on the other

hand was very successful, yielding relatively large amounts of the MWFE subunit

MLRQ MWFE

\

Figure 3.14. Immunoblotting of EGS cross-linked bovine heart mitochondria immunoprecipitated with MLRQ and MWFE antibodies. Mitochondna cross-linked for 2 hours with 0.5M EGS were solubilized with 1% DDM and subjected to immunoprecipitation with antibodies to MLRQ and MWFE as descnbed under materials and methods. Immunoprecipitates from the cross-linked samples were run on 5-208 Tris-glycine gradient gels and then analysed by western blotting with antibodies to (a) the MLRQ subunit and (b) the MWFE subunit. 20 pg of protein was loaded in each lane. Lane 1, non cross-linked bovine heart mitochondria; lane 2, cross-linked bovine heart mitochoadria; lane 3, cross-linked bovine heart rnitochondria immunoprecipitated with MLRQ antibody; lane 4, cross-linked bovine heart mitochondria imrnunoprecipitated with MWFE antibody. Arrows point to immunoprecipitated products.

Table 3.2. Cross-lînked products detected by immunodetection with various complex 1 antibodies

Antibody Used

Cross-linked products

+** many other less prominent cross-linked products detected

MLRQ MWF'E 49 kDa

(Fig. 3.14b, lane 4). However. none of the other subunits cross-linked to MWFE were

found to CO-immunoprecipitate with the subunit. This phenornenon may be linked to two

factors: first. an excess of cross-linking may have resulted in the loss of material or

reduced antigen availability in the MLRQ/MWFE subunits or both and secondly. the

relative sensitivity of the antigen epitopes on these subunits to the EGS cross-linker.

Although the identity of the other subunits that cross-link with MLRQ can be deduced to

some extent. positive identification needs evidence in the form of immunodetection with

antibodies to the respective subunits or protein sequencing of the immunoprecipitated

produc ts.

20 kDa 31 kDa 60 kDa

31 kDa

79 kDa

40 kDa

103 kDa

45 kDa

135 kDa + **

Chapter 4

Postulating the role of a supernumerary subunit such as MLRQ in complex 1 function and dysfunction

Conclusions and Future Directions

Part 1. Molecular stmcture of MLRQ

A 5 15 bp cDNA that codes for the human LMLRQ subunit belonging to the NADH:

ubiquinone oxidoreductase complex of the mitochondrial respiratory chain has been

cloned. The NDUFA4 pne encoding this MLRQ subunit has been mapped to

chromosome 7 p2 1-22 and a pseudogene has been localized to chromosome 1 p2 1. The

pseudogene sequence was found to have 97% similarity to the MLRQ cDNA sequence

and even though the start and stop codons are presewed, there are sufficient arnino acid

changes in the coding region of the gene itself, that if expressed would make it a very

different protein from MLRQ. The pseudogene was found to be intronless and possibly

not expressed in human tissues such as brain. heart, liver and kidney as well as fibroblasts

and cardiomyocytes. A BAC clone containing sequence from chromosome 1

documented in the HTGS database has now k e n found to match this intronless

pseudogene sequence. The NDUFA4 genomic sequence has been found to span

approximately 6.5 kb. with three introns estimated to be 1.1.0.66 and 4.4 kb in size. The

existence of many other MLRQ-like sequences seems to suggest that this supernumerary

subunit probably appeared quite early on in the evolutionary pathway of mamrnalian

complex 1, in order for its gene to have undergone the number of changes that it has.

Also, the high level of conservation seen in MLRQ nucleotide and amino acid sequences

between the mouse, bovine and human species suggests a definite Functional significance

for the subunit in complex 1.

Now that both the cDNA and genomic structure of MLRQ have been determined, an

extensive mutational detection study of NDUFA4 is possible. Although our screening of

MLRQ cDNA did not find any mutations in the ORF that are responsible for complex I

deficiency, a greater subset of patients need to be examined before a definite statement

can be made regarding the involvement of MLRQ in complex 1 deficiency. Future

studies should inciude examination of the promoter region of this gene and subsequent

screening for mutations. Also, further studies can answer whether mutations in the cis-

acting transcription elernents or in the trons-acting transcription factors for NDUFA-I are

responsible for cornplex 1 deficiency. Even though mutations have now been found in a

few core subunits of complex 1 encoded by the nuclear genome, nuclear-encoded genes

responsible for complex 1 deficiency in numerous other cases are yet to be detected. It is

panicularly important at this point that molecular studies look îùrther than abnomalities

in structural complex I proteins. Molecular studies conceming aberrant complex 1

function should focus on other genes involved in the regulation, rnitochondrial impon

and assembly of this large. mulu-protein enzyme complex. The recent identification of

the SURF1 gene in COX-deficient patients with Leigh syndrome (Tiranti et ai, 1998; Zhu

et al; 1998) is a good example of the emerging class of gene defects whch are

responsible for the abnormal hctioning of mitochondrial respiratory chah enzymes.

The discovery of two novel chaperones involved in the assembly of complex I in

N.crussa sets a precedent for such studies (Kuffner et al, 1998). Complementation of

complex 1 activity using techniques such as micro-cell mediated chromosome transfer

(Zhu et al, 1998) in combination with linkage analysis using polymorphic microsatellite

markers in patient groups could provide answers as to the underlying genetic cause of

their deficiency by identibing chromosome regions containing essential candidate genes.

Part II. Biochemical characterization of MLRQ

Tissue expression of the MLRQ protein revealed that it is ubiquitously expressed.

Higher expression of MLRQ in hurnan tissues with high metabolic energy rates such as

heart and muscle is indicative greater numbers of mitochondria in these tissues. Since

bacterial expression and purification of the MLRQ subunit as a fusion protein was

successful, future studies can look at fuie tuning cleavage conditions in order to examine

the structure of the purified protein using techniques such as CD spectroscopy. NMR and

X-ray crystailography. Expression of the MLRQ fusion protein in minimal media labeled

with 2~ (Deuterium) or doubly ewiched by "C and '%, followed by subsequent cleavage

will facilitate NMR study of the subunit. Useful information on the secondary and

tertiary structure of MLRQ can be obtained fiom CD spectra and X-ray crystal maps.

respectively. The tryptophans present in MLRQ make fluorescence spectroscopy a

technique that c m be employed to obtain an ernission spectra of the protein. However.

although these techniques would be useh1 in characterizhg the structure of the expressed

protein, any real information on the conformation of MLRQ can only be discemed by

examining this subunit in various lipid environments that h i c the i ~ e r mitochondrial

membrane. Because MLRQ is one subunit in a huge complex made up of 43 subunits,

even this approach cm only give a rough approximation of the true nature of the subunit

in compIex 1. In addition, the bacterially expressed MLRQ fusion protein itself c m be

used for production of antibodies against the full human MLRQ protein. Generation of a

good MLRQ monoclonal antibody might be invaluable not o d y for screening patients but

also in future immunoprecipitation studies. The discovery of a patient with hepatopathy

and rend tubulopathy, who showed underexpression of other complex I subunits but an

overexpression of MLRQ suggests a differential mechanism of regulation for this subunit

and a process of incorporation into the mitochondna that is not dependant on the titre of

other complex 1 subunits.

Part III. Localization of MLRQ within complex 1

A definite conclusion that can be drawn from our extraction studies is that the MLRQ

subunit has a great affinity for the inner membrane. Even though it has a considerable

hydrophilic domain which interacts with other subunits, the transmembrane domain of

the subunit anchors it very strongly to the membrane, thereby providing integrity and

stability to the entire complex. The fact that other subunits in the HP fraction were

solubilized easily (even subunits thought to be entirely in the membrane m) indicates

that MLRQ or at least its transmembrane domain is embedded quite deeply in the

membrane with perhaps a high a fh i ty for phospholipid membrane components.

Immunoprecipitation studies show that MLRQ is in close proximity to MWFE as well as

the 49 kDa and 8 18 subunits to a lesser extent. This reiterates the locaiization of the

subunit to the Ia subcomplex but aiso denotes an interaction with the hydrophobie Ib

subcomplex. It is likely then that MLRQ rnakes up part of the bulky s tak region of

complex 1 (Fig. 4.1). Definitive statements cannot be made about the association of

Figure 4.1. Schematic representation of the proximity of MLRQ to other subunits of complex 1. Circles represent individual subunits and the ournben inside the circles show their masses in kilodaltons. This mode1 represents a summary of the information on the putative associations between MLRQ and other subunits in the Ia and ID hctions as gathered fIom extraction, immunoprecipitation and cross-linking studies. 49,49 kDa subunit; 6, MWFE; 9, MLRQ; - 16, B 18, P G N or SGDH.

MLRQ to other complex 1 subunits without considerable proof of identity of the cross-

linked products. However, by immunodetection with complex 1 antibodies, it seems

likely that MLRQ is associated with MWFE through an intermediary subunit@).

Deducing the identity of this subunit using molecular mass considerations suggests that in

a trimeric cross-linking scenario, this subunit can be any one of B 18, SGDH (1 8 D a ) or

PGIV (1 9 kDa). The B 18 subunit stands out in particular because it is one of the subunits

immunoprecipitated by the MLRQ antibody. A four-subunit association would implicate

rnany of the smaller subunits of complex 1 such as B8, B9, SDAP (10 D a ) , MNLL (5 .5

D a ) , AGGG (1 0 kDa) or KFYI (5 kDa). Future studies need to focus on efficient

imrnunoprecipitation of these cross-linked products. Modification of cross-linking

conditions and times prior to immunoprecipitation would greatly enhance this endeavor.

Analytical ultracentrifugation of solubilized cornplex 1 is another approach that can be

taken to resolve the subunit-subunit interactions of MLRQ. Sedimentation velocity can

provide experimental evidence on the size and shape of MLRQ while sedimentation

equilibnurn can answer themodynamic questions about the molar mass of MLRQ, it's

association constants and stoichiometries of association with other subunits. Two-

dimensional blue native gel electrophoresis of complex 1 followed by western bloaing

can also yield beneficial information on the association of various subunits with MLRQ

in cornplex 1.

Part IV. Possible roles for MLRQ in complex 1 fûnction

It is difficult to envisage the fiuictionality of a subunit such as MLRQ from protein

hornology alone as database searches do not reveal significant sequence homology to any

other knoun protein. However, based on its biochernical and molecular characterization

cornbined with current knowledge on the structure and subunits of complex 1, certain

conclusions about the nature of this protein can be drawn. Even though MLRQ is not one

of the core proteins of complex 1 involved in electron transport and mutations have not

been found in any of the supemumerary subunits of complex 1, its role in the hctioning

of the enzyme may not necessarily be trivial. The only accessory proteins with known

functions were the 39 kDa subunit which together with the acyl carrying SDAP protein

was shown to be involved in the biogenesis of the complex (Friedrich et al. 1995; Walker

et al, 19%). More recently however, deletion mutation studies on Chinese hamster ceil

lines demonstrated how another supernumerary subunit, MWFE, is essential for cornplex

I activity (Au et al, 1999). This snidy gives relevant insight into the possible functioning

of MLRQ because of the similarities shared between the MLRQ and MWFE subunits.

They are both small MW integral membrane proteins with a hydrophobie N-terminus and

hydrophilic C-terminus and have been localized to the Ia subcomplex. Like MWFE.

MLRQ also lacks a signal sequence at the N-terminus of the protein also suggesting that

it is irnported into mitochondria and associated with complex 1 without requiring

proteolytic processing. More significantly, our studies with irnmunoprecipitation and

cross-linking have s h o w that the two subunits are in close proximity to one another and

may be associated with each other, even if it is through another subunit. Given this. the

MLRQ subunit may be functioning alongside with MWFE in the formation or integration

of cofactors in complex 1. The MWFE mutant ce11 line studied by Au et al (1999)

showed a complete loss of complex 1 activity as well as decreases in the expression of the

75-,5 1- and 24- kDa subunits which cany redox centres. Likewise, MLRQ may be

involved in the assembly or even in the import of these core subunits of complex 1.

similar to the subunits of complex III which serve as processing enzymes (Schulte et al.

1989). If the hctionality of the MLRQ subunit is to be pursued through anti-sense

strategies, certain changes need to be made to the attempted procedure. Firstly, since the 0

anti-sense strategy is somewhat of an erratic procedure (for the reasons discussed in the

previous chapter), constmcts containing different portions of the MLRQ cDNA used to

produce anti-sense RNA need to be tested. Previous studies have shown that the size of

the anti-sense sequence has a great effect on the extent of mRNA inhibition

(Scherczinger et al, 1992). However, a more useful approach might be to produce a

MLRQ knockout in mouse and examine complex 1 activity in mutant ce11 lines denved

From the organism if it proves viable. Expression of the MLRQ subunit in combination

with other human complex 1 subunits in yeast would also give answers as to whether

MLRQ is stable on its own or needs the other subunits to form part of the complex.

The MLRQ subunit is of particular importance because it possesses a trammembrane

domain as well as a hydrophilic C-terminus which connects the membrane arm of

complex 1 with the globular arm. It is likely then that this subunit forms part of the stalk

which has been found to be thick enough to form an insulating layer around the electron

pathway (Grigorieff, 1999). Ultimately, the MLRQ subunit together with the other

supernumerary subunits of mamrnalian complex I probably forms a scaffold, a theme

which is suggested by the position of the supemumerary subunits of COX, which

package the core electron transfemng subunits of the enzyme thereby reducing their

contact with the surrounding phospholipids. Future studies looking at complex 1

dysfûnction might also concentrate on oxygen fiee radical formation fiom supemumerary

subunits such as MLRQ and the role of these subunits in the assembly and function of

complex 1.

The finai word

This work has answered questions and dispelled some of the uncertainties about the

MLRQ subunit that were manifest at the beginning of this study. Contrary to what was

ociginally thought, the MLRQ subunit has been localized to chromosome 7. We now

know that it is not responsible for the X-linked mode of inheritance seen in certain cases

of complex 1 deficiency. As foretold by the studies of Day and Scheffler (1982). the

search is on for the other X-Iinked complex 1 subunit which can be confirmed through the

chromosomal localizations of the B9, B 12, B 15, SGDH, MMTG and 17.2 kDa subunits

or the discovery of a complex I assembly factor on the X chromosome. It can also be

said that mutations in the coding region of NDUFA4 are not a common cause of complex

1 dysfunction. The availability of the genomic structure of NDUFA4 now makes it

possible to screen for mutations in the promoter regions of the gene. The localization and

confirmation of a pseudogene of MLRQ as well as the discovery of other MLRQ-like

sequences has also helped explain the many incorrect localizations of the gene reported in

GenBank databases. Lastly, the fact that MLRQ is part of complex 1 as opposed to

merely coprecipitating with it has now been cemented by biochemical characterizations

of the subunit through cross-linking and immunoprecipitation studies. This is supported

by the ubiquitous expression of MLRQ in various human tissues, even though the

regulation of the subunit may differ fiom that of sorne of the other complex 1 subunits.

These investigations allow a better understanding of complex 1 fünction and dysfunction

through the study of one of its supemurnerary subunits and thereby fulfill the objectives

set forth in this study.

Abrahams. J.P.. Leslie. A.G.. Lutter. R. and Walker. J.E. 1994. Structure at 2.8 A resolution of F 1 -ATPase fiom bovine heart mitochondria. Nature. 37O:62 1-618.

Ackrell, B A C . 2000. Progress in understanding structure-function relationships in respiratory chain complex 11. FEBS Lea. 466: 1-5.

Ali. S.T.. Duncan. A.M.V.. Schappert. K.. Heng. H.H.Q.. Tsui. L.C.. Chow. W. and Robinson. B.H. 1993. Chromosornal localization of the human gene encoding the 5 1-kDa subunit of rnitochondrial complex 1 (NDUFVI) to 1 1 q 13. Genomics. 18:JX-439.

Albracht. S.P.J.. Mariette. A. and de Jong. P. 1997. Bovine-heart NADH:ubiquinone oxidoreductase is a monomer with 8 Fe-S clusters and 2 FMN groups. Biochirn. Biophys. Acta. 1318:92-106.

Almeida. T.. Duarte. M.. ~Melo. A.M.P. and Videira. R. 1999. The 24-kDa iron-sulphur subunit of complex 1 is required for enzyme activity. Eur. J. Biochem. 26586-92.

-4lves. P.C. and Videira. A. 1998. The membrane domain of complex 1 is not assembled in the stopper mutant E35 of Neurospora. Biochem. Crll. Biol. 76( 1 ): 139- 143.

Anderson. S.. de Bniijn. M.H.L.. Coulson. A.R. and Eperon. I.C.. Sanger. F. and Young. I.G. 1982. Complete sequence of bovine mitochondrial DNA: Conserved features of the rnammalian mitochondrial genorne. J. Mol. Biol. 156:683-717.

Arizmendi. LM., Runswick. M.J.. Skehel. J.M. and Walker. J.E. 1992. NADH: ubiquinone oxidoreductase from bovine heart mitochondria. A fourth nuclear encoded subunit with a homologue encoded in chloroplast genomes. FEBS Lrtt. 301(3):237-212.

Attardi. G. and Schatz. G. 1988. Biogenesis of mitochondria. Annu. Rev .Ce11 Biol. -l:389-333.

Au. H.C.. Seo. B.B.. Matsuno-Yagi. T. and Schemer. I.E. 1999. The NDUFA 1 gene product (MWFE protein) is essential for activity of complex I in mammalian mitochondna. Proc. Natl. Acad. Sci. USA. 9634354-4359.

Bai. Y.and Attardi. G. 1998. The mtDNA-encoded ND6 subunit of mitochondrial NADH dehydrogenase is essential for the assembly of the membrane a m and the respiratory function of the enzyme. EMBO. J. 17(16):48484858.

Belognidov. G. and Hatefi. Y. 1994. Catalytic sector of complex I (NADH: Ubiquinone oxidoreductase) subunit stoichiometry and substrate-induced conformation changes. Biochemistry. 33:457 14576.

Belogrudov. G.I. and Hatefi. Y. 1996. Intersubunit interactions in the bovine mitochondrial complex 1 as revealed by ligand blotting. Biochem. Biophys. Res. Commun. 227(1): 135- 139.

Benson. D.A.. Boguski. M.S.. Lipman, DL? Ostell. J.. Ouellette, B.F.. Rapp. B.A. and Wheeler, D.L. 1999. GenBank. Nucleic Acids Res. 27(1): 12- 17.

Bereiter-Hahn. J. and Voth. M. 1994. Dyarnics of mitochondria in living cells: shape changes. dislocations. fusion. and fission of mitochondria. Microscop. Res. Tech. 27(3): 198-2 19.

Berger. S.. Ellersiek. U.. Kinzelt. D. and Steinmuller. K. 1993. Immunopurification of a subcomplex of the NAD(P)H-plastoquinone-oxidoreductase from the cyanobacterium Synechocystis sp. PCC6803. FEBS Lett. 326(1-33246-250.

Benhold. J.. Bauer. M.F.. Schneider. H.C.. Klaus. C.. Dietmeier. K.. Neupert. W. and Brunner. M. 1995. The MIM complex mediates preprotein translocation across the mitochondrial inner membrane and couples it to the mt-Hsp7OIATP drivinç system. Cell 8l(7): 1085-1093.

Bindoff. L.A.. Birch-Machin. M.. Cartlidge, N.E.. Parker. W.D. and Turnbull. D.M. 1989. Mitochondrial fimction in Parkinson's disease. Lancet 2(8653):49.

Blattner. F R . Plunkett. G.. Bloch. C.A.. Pema. NT.. Burland. V.. Riley. M.. Collado- Vides. J.. Glasner. J.D.. Rode. C.K.. Mayhew. G.F.. Gregor. J.. Davis. N. W.. Kirkpatrick. H.A.. Goeden. MA.. Rose. D.J.. Mau. B. and Shao. Y. 1997. The complete genome sequence of Escherichia coli K- 12. Science 277(533 1 ): 1453- 1474.

Boguski, M.S.. Lowe. T.M. and Tolstoshev. C.M. 1993. dbEST-database for "expressed sequence tags" Nat. Genet. 4(4):332-333.

Bolden. A.. Noy. G.P. and Weissbach. A. 1 977. DNA polymerase of mitochondna is a gamma-polymerase. J. Biol. Chem. 252335 1-3356.

Bourgeron. T.. Rustin. P.. Chretien. D.. Birch Machin. M.. Bourgeois. M.. Viegas Pequignot. E.. Munnich. A. and Rotig, A. 1995. Mutation of a nuclear succinate dehydrogenase gene results in mitochondnal respiratory chain deficiency. Nat. Genet. 11:144-149.

Brandt. U. 1997. Proton-translocation by membrane-bound NADH:ubiquinone- oxidoreductase (complex 1) through redox -gated ligand conduction. Biochim. Biophys. Acta. 13 18:79-9 1.

Braun. M., Bungert. S. and Friedrich. T. 1998. Characterization of the overproduced NAûH dehydrogenase hgment of the NADH:ubiquinone oxidoreductase (complex 1) from Escherichia coli. Biochemistry. 37: 1 86 1 - 1867.

Brody. S.. Oh. C.. Hoja LT. and Schweizer. E. 1997. Mitochondrial acyl carrier protein is involved in lipoic acid synthesis in Saccharomyces cerevisiae. FEB S Lett. JO8(2): 2 1 7- 220.

Brown. M.D. and Wallace. D.C. 1990. Molecular bais of mitochondrial DNA disease. J. Bioenerg. Biomembr. 26(3):273-289.

Budde. S.M.S.. van den Heuvel. L.P.W.J., Janssen, A.J.. Smeets. R.J.P.. Buskens. C.A.F.. DeMeirleir. L.. Van Coster. R.. Baethmann. M.. Voit. T.. Trijbels. J.M.F. and Smeitink. J.A.M. 2000. Combined erizymatic complex 1 and III deficiency associated with mutations in the nuclear encoded 1VDLIFS.I gene. Biochem. Biophys. Res. Comm. 275:63-68.

Calhoun. M. W.. Thomas. J. W. and Gennis. R.B. 1994. The cytochrorne oxidase superfamily of redox-driven proton pumps. Trends Biochem. Sci. l9(8):3Z-j 30.

Capaldi. R A . 1990. Structure and assembly of cytochrome c oxidase. Arch. Biochem. Biophys. 280(2):252-262.

Carrozzo. R.. Hirano. M.. Fromentry, B.. Casali . C.. Santorelli . F.M.. Bonilla. E.. DiMauro. S.. Schon. E.A. and Miranda. A.F. 1998. Multiple mtDNA deletions features in autosomal dominant and recessive disease suggest distinct pathopeneses. Neurolog. SO( 1 ):99- 106.

C hen. H. and Antonarakis. S.E. 1995. Localization of a human homolog of the mouse Tiam- 1 gene to chromosome 3 1 q23. 1. Genomics 30( 1 ): 1 23- 127.

Chen. L.B. 1988. Mitochondrial membrane potential in living cells. Annu. Rev. Cell. Biol. 1:155-181.

Chomyn. A.. Cleeter. M.W.J.. Ragan. C.I.. Riley. M.. Doolittle. R.F. and Attardi. G. 1986. URF6. Last unidentified reading fiame of human mtDNA. Codes for an NADH Dehydrogenase subunit. Science. 2346 14-6 1 8.

Chomyn. A.. Marioaini. P.. Cleeter. M.W.J., Ragan. C.I.. Matsuno-Yagi. A.. Hatefi. Y.. Doolittle. R.F. and Attardi. G. 1985. Six unidentified reading frames of human mitochondrial DNA encode components of the respiratory-chain NADH dehydrogenase. Nature. 314593-597.

Chomyn. A.. Patel, S.D.. Cleeter. M. W.J.. Ragan. C.I. and Attardi. G. 1988. The site of synthesis of the iron-sulfur subunits of the flavoprotein and iron-protein fractions of human NADH dehydrogenase. J. Biol.Chem. 263(3 1 ): 163%- 16400.

Damell. J., Lodish, H. and Baltimore. D. 1986. Molecular ce11 biolow. Scientific American Books Inc. New York.

Darrouzet. E., Issartel, J.. Lunardi, J. and Dupuis, A. 1998. The 49-kDa subunit of NADH-ubiquinone oxidoreductase (Complex 1) is involved in the binding of piericidin and rotenone. two quinone-related inhibitors. FEBS Lett. 43154-38.

Day. C.E. and Scheffler. LE. 1982. Mapping of the genes of some components of the electron transport chain (complex 1) on the X chromosome of marnmals. Somatic Cell. Genet 8(6):69 1-707.

de Coo. R.F.M.. Buddiger. P.. Smeets. H.. van Kessel. A.G.. Morgan-Hughes. J.. Weghuis. D.O.. Overhauser. J. and van Oost. B. 1995. Molecular cloning and characterization of the active hurnan rnitochondrial NADH:ubiquinone oxidoreductase 24-Da gene (iVDUFV2) and its pseudogene. Genomics. 26:46 1 -466.

de Coo. R.F.M.. Buddiger. P.. Smeets. H. and van Oost. B.A. 1997. Molecular cloning and characterization of the human mitochondrial N.4DH:Oxidoreductase 10-kDa çene (NDUFC'3). Genomics 45A34-437.

de Coo. R.F.M.. Buddiger. P.. Smeets. H.J.M. and van Oost. B.A. 1999. The structure of the human NDUFV 1 gene encoding the 5 1 -kDa subunit of rnitochondrial complex 1. Mamrnalian Genome lO:49-53.

Degli Esposti. M. 1998. Inhibitors of NADH-ubiquinone reductase: an overview. Biochim. Biophys. Acta 1364:222-235.

Degli Esposti. M.. Crimi. M. and Ghelli. A. 1994. Natural variation in the potency and binding sites of rnitochondrial quinone-like inhibiton. Biochem. Soc. Trans. 22( 1 ):209- 2 13.

Degli Esposti. M. and Ghelli. A. 1994. The mechanism of proton and electron transport in mitochondrial complex 1. Biochim. Biophys. Acta 1 187(7): 1 1 6- 1 20.

De Jong. AM. and Albracht. S.P. 1994. Ubisemiquinones as obligatory intermediates in the electron transfer from NADH to ubiquinone. Eur. J. Biochem. 222(3):975-982.

Deng. P.S.. Hatefi. Y. and Chen S. 1990. N-arylazido-beta-alanyl-NAD+. a new NAD+ photoaffinity analogue. Synthesis and labeling of rnitochondrial NADH dehydrogenase. Biochemistry 29(4): 1094- 1098.

de Sury. R.. Martinez. P.. Procaccio, V.. Lunardi. J. and Issartel. J.P. 1998. Genomic structure of the human NDUFS8 gene coding for the iron-sulfur TYKY subunit of the mitochondrial NADH:ubiquinone oxidoreductase. Gene 2 15: 1 - 1 0.

Dover. G. 1989. Molecular drive: a cohesive mode of species evolution. Nature 299: 1 1 1 - 117.

Duncan. A.M.V.. Chow. W. and Robinson. B.H. 1992. Localization of the human 75-kDa Fe-S protein of NADH-coenzyme Q reductase gene (:VDUFSi) to 2q33 -+ q34. Cytogenet. Cell Genet. 6O:2 12-2 13.

Dunham, I.. Shimizu. N., Roe. B.A.. Chissoe. S.. Hunt. A.R.. Collins. J.E.. Bruskiewich. R.. Beare, D.M.. Clamp. M.. Smink. L.J.. Ainscough. R.. Almeida J.P.. Babbage. A.. Bagguley C. Bailey J. Barlow, K.. Bates. K.N.. Beasley. 0.. Bird. C.P.. Blakey. S.. Bridgeman. A.M.. Buck. D.. Burgess. I., Bumll. W.D. and OfBrien. K.P. 1999. The DNA sequence of human chromosome 22. Nature 402(676 1):489-495.

Dupuis. A.. Skehel. J.M. and Walker. J.E. 199 1 a. A homologue of a nuclear-coded iron- sulfur protein subunit of bovine mitochondria1 complex I is encoded in chloroplast genomes. Biochemistry 30( 1 1 ):2954-2960.

Dupuis, A.. Skehel. J.M. and Walker. I.E. 199 1 b. NADH:ubiquinone oxidoreductase from bovine mitochondria. cDNA sequence of a 19 kDa cysteine-rich subunit. Biochem J. 277(1):11-15.

Dutton. P.L.. Moser. C.C.. Sled. V.D.. Daldal. F. and Ohnishi. T. 1998. A reductant- induced oxidation mechanism for complex I. Biochim. Biophys. Acta. 1364245-157.

Earley. F.G.. Patel. S.D.. Ragan. I. and Attardi. G. 1987. Photolabeling of a mi tochondrially encoded subunit of NADH dehydrogenase with [3 Hldihydrorotenone. FEBS Lett. 219(1):108-112.

Emahazion. T.. Beskow. A.. Gyllensten. U. and Brookes. A.J. 1998. [ntron based radiation hybrid mapping of 15 complex I genes of the human electron transport chain. Cytogenet. Cell Genet. 82: 1 15- 1 19.

Emahazion T. and Brookes A.J. 1998. Mapping of the NDUFAI. NDUFrl6. NDUE4 7. iVDUF.4 BB. and NDUFS8 electron transport chain genes by intron based radiation hybrid mapping. Cytogenet. Cell Genet. 82: 1 12- I 14.

Evans. M.J. and Scarpulla R.C. 1988. The human somatic cytochrome c gene: two classes of processed pseudogenes demarcate a period of rapid molecular evolution. Proc. Natl. Acad. Sci. U S A. 85(24):9625-9629.

Fearnley. I.M. and Walker, J.E. 1992. Conservation of sequences of subunits of mitochondria1 complex 1 and their relationships with other proteins. Biochem. Biophys. Acta 1 140: 105- 134.

Feamley. I.M.. Skehel. J.M. and Walker. J.E. 1994. Electrospray ionization mass spectrometric analysis of subunits of NADHLJbiquinone oxidoreductase (complex I) From bovine heart mitochondria. Biochem. Soc. Trans. 2255 1 -555.

Finel. M. 1993. The proton-translocating NADH:Ubiquinone oxidoreductase: A discussion of selected topics. J. Bioenerget. Biornem. 25(4):357-365.

Finel. M. 1998. Organization and evolution of structural elements within cornplex 1. Biochim. Biophys. Acta. 1364: 1 12- 12 1. Fine]. M.. Majander. A.S.. Tynela. J.. de Jong, A.M.. .Albracht. S.P.J. and Wikstrom. M. 1994. Isolation and characterization of subcomplexes of the mitochondrial NADH:Ubiquinone oxidoreductase (complex 1). Eur. J. Biochem. 226237-241.

Finel. M.. Skehel, M.J.. Albncht. S.P.J.. Feamley, I.M. and Walker. J.E. 1992. Resolution of NADH:Ubiquinone oxidoreductase from bovine heart mitochondria into two subcomplexes. one of which contains the redov centers of the enzyme. Biochemistry 3 l(46): 1 1425- 1 1334.

Fisher. N. and Rich. P. 2000. A motif for quinone binding sites in respiratory and photosynthesis systems. J. Mol. Biol. 296. 1 153- 1 162.

Friedrich. T., Abelmann. A.. Brors. B.. Guenebaut, V.. Kintscher. L.. Leonard. K., Rasmussen. T., Scheide. D.. Schlitt. A.. Schulte. U. and Weiss. H. 1998. Redox components and structure of the respiratory NADH:ubiquinone oxidoreductase (complex 1). Biochim. Biophys. Acta. t 3 6 W 1 5-2 1 9.

Friedrich. T.. Hothaus. G.. Ise. W.. Nehls. U.. Sctunitz. B. and Weiss. H. 1989. A srnaIl iso form of NADH:ubiquinone oxidoreductase (complex 1) wi thout mitoc hondrially encoded subunits is made inchloramphenicol-treated Neirrospora crassu. Eur. J. Biochem. 180( 1 ): 1 73- 180.

Friedrich. T.. Ohnishi. T.. Forche, E.. Kunze. B.. Jansen. R.. Trowitzsch. W.. Hofle. G.. Reichenbach. H. and Weiss. H. 1994. Two binding sites for natunlly occumng inhibitors in mitochondrial and bacterial NADH:ubiquinone oxidoreductase (complex 1). Biochem. Soc. Tram. 22(1):220-230.

Friednch. T.. Steinrnuller. K. and Weiss, H. 1995. The proton-pumping respiratory complex 1 of bacteria and mitochondria and its homologue in chloroplasts. FEBS Lett. 367(2): 107- 1 1 1.

Friedrich. T. and Weiss. H. 1997. Modular evolution of the respiratory NADH: ubiquinone oxidoreductase and the origin of its modules. J. Theor. Biol. 187:529-540.

Galante. Y.M. and Hatefi. Y. 1979. Purification and molecular and enz)matic properties of mitochondrial NADH dehydrogenase. Arch. Biochem. Biophys. 192559468.

Galkin. A.S.. Grivemikova V.G. and Vinogradov. A.D. 1999. HL/2e' stoichiometry in NADH-quinone reductase reactions catalyzed by bovine heart submitochondrial particles. FEBS Lett. 451:157-161.

Giles, R.E.. Blanc. H., C m , KM. and Wallace. D.C. 1980. Maternai inheritance of human mitochondria1 DNA. Proc. Nat. Acad. Sci. USA 77167 1 5-67 19.

Gray. M.W.. Burger. G. and Lang, B.F. 1999. Mitochondrial evolution. Science. 283: 1476- 148 1.

Grigorieff. N. 1998. Three-dimensional structure of bovine NADH:ubiquinone oxidoreductase (complex 1) at 22 A in ice. I Mol Biol. 277: 1 O N - 1046.

GrigorietT. N. 1999. Structure of the respiratory NADH:ubiquinone oxidoreductase (complex 1). Curr. Opin. Struct. Biol. 9A76-483.

Grivemikova. V.G.. Malklashina. E.O.. Gavrikova. E.V. and Vinogradov. A.D. 1997. Interaction of the mitochondrial NADH-ubiquinone reductase with rotenone as related to the enzyme active/insctive transition. Biochim. Biophys. Acta 13 l 9 : 2 X - l X .

Gu. M.. Cooper. LM.. Gash. M.. Mann. V.M.. Javoy-Agid. F. and Schapira A.H.V. 1996. Mitochondrial defects in Huntington's disease caudate nucleus. Ann. Neurol. 39585-589.

Guenebaut. V.. Schlitt. A.. Weiss. H.. Leonard. K. and Friedrich. T. 1998 Consistent structure between bacterial and mitochondrial NADH:ubiquinone oxidoreductase (complex 1). J Mol Bi01 276: 105 1 12

Hagerhall. C. 1997. Succinate: quinone oxidoreductases - Variations on a consenred theme. Biochim. Biophys. Acta. 1320: 107- 14 1.

Han. A.L.. Yagi. T. and Hatefi. Y. 1989. Studies on the structure of NADH: ubiquinone osidoreductase corn plex: topo-mphy of the subuni ts of the iron-sulfur protein component. Arch. Biochem. Biophys. 275( 1 ): 166- 1 73.

Hanna. M.G. and Nelson. I.P. 1999. Genetics and molecular pathogenesis of mitochondrial respiratory chain diseases. Cell. Mol. Life Sci. 55: 69 1-706.

Harding. A.E. and Holt. I.J. 1 993. Mitochondnal Myopathies. In: Molecular basis of neurolo-W. Ed. Conneally. P.M. Blackwell Scientific Publications. Boston. p.245-26 1.

Hartl. F.U.. Schmidt, B.. Wachter. E.. Weiss. H. and Neupert. W. 1986. Transport into mitochondria and intrmitochondrial sorting of the Fe/S protein of ubiquinol-cytochrome c reductase. Ce11 47(6):939-95 2 .

Hattori. N.. Suzuki, H., Wang. Y.. Minoshima S.. Shimizu. N.. Yoshino. H.. Kurashima. R.. Tanaka M.. Ozawa. T. and Minino, Y. 1995. Structural organization and chromosomal localization of the human nuclear gene (NDUFV7) for the 24-Da iron- sulfur subunit of complex I in mitochondrial respiratory chain. Biochem. Biophys. Res. Commun. 216:77I -777.

Hanori, N.. Yoshino. H., Tanaka M.. Suzuki. H. and Mizuno. Y. 1998. Genotype in the 24-kDa subunit gene (NDUFV?) of mitochondrial complex 1 and susceptibility to Parkinson's disease. Genomics 3952-58.

Hegde. R. 1 998. The 24-kDa subunit of bovine heart mitochondrial NADH:ubiquinone oxidoreductase is a G protein. Biochem. Biophys. Res. Commun. 2.14620-629.

Hell. K.. Hemann. J.M.. Pratje. E.. Neupert. W. and Stuart. R.A. 1998. Oxal p. an essential cornponent of the N-tail protein export machinery in mitochondria. Proc Nat1 Acad Sci U S A 95(5):2250-2255.

Hemnann. J.M. and Neupert. W. 2000. Protein transport into mitochondria. Curr. Opin. Microbiology 3:î 10-2 14.

Hehman, G.L. and Hauswirth. W.W. 1992. DNA helicase from rnmmalian mitochondria. Proc. Natl. Acad. Sci. USA 89:8562-8566.

Hothaus. G. and Attardi. G. 1993. Lack of assembly of mitochondrial DNA-encoded subunits of respiratory NADH dehydrogenase and loss of enzyme activity in a hurnan cell mutant lacking the mitochondrial ND4 gene product. EMBO J. 12(8):3043-3048.

Hurt. E.C.. Pesold-Hurt. B. and Schatz . G. 1984. The amino-terminal region of an irnported mitochondrial precursor polypeptide c m direct cytoplasmic dihydrofolate reductase into the mitochondrial matrix. EMBO J. 3( 13):3 149-56.

Hyslop. S.. Duncan. A.M.V.. Pitkanen. and Robinson, B.H. 1996. Assignment of the PSST subunit gene of human mitochondrial complex I to chromosome 19p 13. Genomics. 37375-380.

Ingledew. W.J. and Ohnishi. T. 1980. An analysis of some thermodynamic properties of iron-sulphur centres in site 1 of mitochondria. Biochern J. 186(1): 1 1 1-1 17.

Iwata. S.. Lee. LW., Okada. K.. Lee. J.K., Iwata, M.. Rasmussen, B.. Link. T.A., Ramaswamy. S. and Jap. B.K. 1998. Complete structure of the 1 I -subunt bovine mitochondrial cytochrome bci complex. Science. 281:64-71.

Kaput, J.. Goltz. S. and Blobel. G. 1982. Nucleotide sequencr of the yeast nuclear gene for cytochrome c peroxidase precursor. Functional implications of the presequence for protein transport into mitochondna. J. Biol. Chem. 257(24): 15054-1 5058.

Kaukonen, J.A.. Amati. P.. Suornalainen, A., Rotig, A.. Piscaglia M.G.. Salvi. F.. Weissenbach. J.. Fratta. G.. Comi, G.. Peltonen. L. and Zeviani. M. 1996. An autosomal locus predisposing to multiple deletions of mtDNA on chromosome 3p. Am. J. Hum. Genet. 58(4):763-769.

Kayser. E.B.. Morgan. P.G. and Sedensky. M. 1999. GAS-1: A mitochondrial protein controls sensitivity to volatile anesthetics in the nematode Caenorhabditis elegans. Anesthesioiogy. 90(2): 545-554.

Kennaway. N.G. 1988. Defects in the cytochrome bel compled in mitochondnal diseases. J. Bioenerg. Biomemb. 20:325-35 1.

Keough. T.. Lacey. M.P.. Fieno. A.M., Grant. R.A.. Sun. Y.. Bauer. M.D. and Begley. K.B. 2000. Tandem rnass spectrometry methods for de finitive protein identification in proteomics research. Electrophoresis 2 1( 1 1 ):XE-Z?6j.

Kikuno. R. and Miyata. T. 1985. Sequence homologies among mitochondrial DNA- coded URFî,. URF4 and URFS. FEBS Lett. 189(1):85-88.

Kim. J.W.. Lee. Y.. Kang. H.B.. Choe. Y.K.. Chung. T.W.. Chang. S.Y.. Lee. K.S. and Choe. I.S. 1997. Cloning of the human cDNA sequence encoding the NADH:ubiquinone oxidoreductase MLRQ subunit. Biochem. Mol. Biol. Int. 43:669-675.

Komiya. T.. Rospen. S.. Koehler. C.. Looser. R.. Schatz. G. and Mihara. K. 1998. Interaction of mitochondrial targeting signals with acidic receptor dornains along the protein import pathwq: evidence for the 'acid chain' hypothesis. EMBO J. 17( l4):3886- 3398.

Korsmeyer. S.Y. Yin. X.M.. Oltvai. ZN.. Veis-Novack. D.J. and Linette. G.P. 1995. Reactive oxygen species and the regulation of ce11 death by the BcI-2 gene farnily. Biochim. Biophys. Acta 1271(1):63-66.

Kozak. M. 1996. lnterpreting cDNA sequences: Some insights from studies on translation. Mammalian Genome. 7563-574.

Krishnarnoorthy. G. and Hinkle. P.C. 1988. Studies on the electron transfer pathway. topography of iron-sulfur centers. and site of coupling in NADH-Q oxidoreductase. J. Biol. Chem. 263(33): 17566-1 7575.

Kuffner. R.. Rohr. A.. Schmiede. A.. KrulI. C. and Schulte. U. 1998. InvoIvement of ttvo novel chaperones in the assembly of mitochondrial NADH:ubiquinone oxidoreductase (cornplex 1). J Mol Biol. 283:4094 1 7.

Kunz. J.. Scherer. S.W., Klawitz. 1.. Soder. S.. Du. Y.Z.. Speich. N.. Kalff-Suske. M.. Heng. H.. Tsui. L.C. and Grreschik, K.H. 1994. Regional localization of 725 human chromosome 7-specific yeast artificial chromosome (YAC) clones. Genomics. 22:439- 448.

Larsson. N. and Clayton. D.A. 1995. Molecula. genetic aspects of human mitochondrial disorders. Annu. Rev. Genetics 29: 1 5 1 - 1 78.

Lawford. H.G. and Garland. P.B. 1972. Proton translocation by rat liver mitochondria with quinones as terminal acceptors. Biochem. J. 129(3):47-48.

Lehninger. A.L. 1964. The mitochondrion: Molecular basis of structure and function. W.A. Benjamin Inc. New York. p. 16-40.

Leit: H.. Sled. V.D.. Ohnishi. T.. Weiss. H. and Friedrich. T. 1995. Isolation and characterization of the proton-translocating NADH: ubiquinone oxidoreductase from Escherichia coli. Eur. J. Biochem. 230(2):538-548.

Leonard. J.V. and Schapira. A.H.V. 2000. Mitochondrial respiratory chain disorders II: neurodegenerative disorders and nuclear gene defects. The Lancet. 355339494.

Leslie. A.G.W.. Abrahams. J.P.. Braig. K.. Luner. R.. Mem. R.I.. Omss. G.L.. van Raaij. b1.J. and Walker. J.E. 1999. The structure of bovine mitochondrial FI-ATPase: an example of rotary catalysis. Biochem. Soc. Trans. 27:37-42.

Leuenberpr. D.. Bally. N.A.. Schatz. G. and Koehler. C.M. 1999. Diffèrent import pathways through the mitochondrial intemembrane space for inner membrane proteins. EMBO J. l8(17):48 l 6 4 8 Z .

Liao. H. and Spremulli. L.L. 1990. Effect of length and mRNA secondaq structure on the interaction of bovine mitochondrial ribosomes with messenger RNA. J. Biol. Chem. 265: 136 18- 13622.

Lin. X.. Wells. D.E.. Kimberling, W.J. and Kumar. S. 1999. Human NDUFB9 gene: eenomic organization and a possible candidate gene associated with deafness disorder L

mapped to chromosome 8q 13. Hum Hered. 49(2):75-80.

Ling. M.. McEachern. G.. Seyda. A.. MacKay. N.. Scherer. S.W.. Bntinova. S.. Beatty. B.. Giovannucci-Uzielli. M.L. and Robinson. B.H. 1998. Detection of a homozygous four base pair deletion in the protein X gene in a case of pqnivate dehydrogenase complex deficiency. Hum. Mol. Genet. i(5):jO 1-5

Loeffen J.. Smeets. R.. Srneitink. J.. Triepels, R.. Sengers. R.. Trijbels. F. and van den Hevel. L. 1999. The human NADH: ubiquinone oxidoreductase NDUFSj (1 5 ma): cDNA cloning. chromosomal localization. tissue distribution and the absence of mutations in isolated complex 1-deficient patients. J. Inherit. Metab. Dis. 22: 19-28.

Loeffen, J.L.C.M.. Smeitink, J.A.M.. Trijbels. J.M.F.. Janssen. A.J.M.. Triepels. R.H.. Sengers. R.C.A. and van den Heuvel. L.P. 2000. tsolated complex 1 deficiency in children: Clinical. biochemical and genetic aspects. Hum. Mut. 15: 123-1 3 4.

Loeffen. J.. van den Heuvel. L., Smeets. R., Triepels, R.. Sengers. R.. Trijbels. F. and Smeitink. J , 199th. cDNA sequence and chromosomal localization of the remaining three human nuclear encoded iron suiphur protein (IP) subunits of complex 1: The

human IP fraction is completed. Biochem. Biophys. Res.Comm. 1-758.

Loeffen J.L.. Triepels R.H.. van den Heuvel L.P.. Schuelke M.. Buskens C.A.. Smeets R.J.. Trijbels J.M. and Smeitink.. J.A.M. 1998b. cDNA of eight nuclear encoded subunits of NADH:ubiquinone oxidoreductase: human complex I cDNA characterization completed. Biochem Biophys Res Commun. ZJ3:4Ij-Q2.

Loeffen, J.. Smeitink, J.. Triepels. R.. Smeeets. R.. Schuelke. M.. Sengers. R.. Trijbels. F., Hamel. B., Mullaart. R. and van den Heuvel. L. 1998c. The first nuclear-encoded complex 1 mutation in a patient with Leigh syndrome. Am. J. Hum. Genet. 63: 1598- 1608.

Lowry . O.H.. Rosenbrough. N.J.. Fm. A.L. and Randall. R.J. 195 1. Protein measurement with the folin phenol reagent. J. Biol. Chern. 193:26j-?7j.

Lutsenko. S. and Cooper. M.J. 1998. Localization of the Wilson's disease protein product to rnitochondria. Proc. Natl. Acad Sci. USA 9560065-6009.

Majander. A.. Huoponen. K.. Savontaus. M.L.. Nikoskelainen. E. and Wikstrom. M. 1991. Electron tnnsfer properties of NADH:ubiquinone reductase in the NDlG460 and the ND41 1778 mutations of the Leber hereditary optic neuroretinopathy (LHON). FEBS Lett. 292( 1 -2):?89-X.

Maklashina. E. and Cecchini. G. 1999. Comparison of catalytic activity and inhibitors of quinone reactions of succinate dehydrogenase (Succinate-ubiquinone oxidoreductase) and fumarate reductase (Menaquinol-fumarate oxidoreductase) from Escherichicr coli. Arch. Biochem. Biophys. 369(2):223-232.

Maniatis, T.. Fritsch. E.R. and Sambrook. J. 1989. Molecular cloning: A laboratorv manual. Cold Spring Harbour Laboratory Press. New York. pp. 7.37-7.39: 9.3 1-9.57: 18.60- 18.75.

Mannella, C.A. 1982. Structure of the outer mitochondrial membrane: Ordered arrays of porelike subunits in the outer membrane fraction from ivezrrospora crassa mitochondria. J. Cell. Biol. 94:680-687.

Matsushita K., Ohnishi. T. and Kaback. H.R. 1987. NADH-ubiquinone oxidoreductases of the Escherichia coli aerobic respiratory chain. Biochemistry 26(24):7732-7737.

McEachem. G.. Kassovska-Bratinova S., Raha S.. Tamopolsky, M.A.. Turnbull . J.. Bourgeois, J. and Robinson. B. H. 2000. Manganese superoxide dismutase levels are elevated in a proportion of arnyotrophic lateral sclerosis patient ce11 lines. Biochem. Biophys. Res. Commun. 273(1):359-363.

Meijer, A.J. and van Dam. K. 1974. The metabolic significance of anion vansport in mitochondria. Biochim. Biophys. Acta 346:2 13-244.

Mighell, A.J.. Smith, N.R., Robinson. P.A. and Markham. A.F. 2000. Vertebrate pseudogenes. FEBS Lett. 468: 1 09- 1 14.

Mitchell. P. 1966. Chemiosmotic coupling in oxidative and photosynthetic phosphoiylation. Biol. Rev. Carnb. Philos. Soc. 41(3):445-502.

Miyoshi. H. 1 998. Structure-activity relationships of some complex 1 inhibitors. Biochim. Biophys. Acta 1364236-244.

Morgan-Hughes. J.A. and Hanna M.G. 1999. Mitochondrial encephalornyopathies: the enigma of genotype versus phenotype. Biochim. Biophys. Acta. I l IO(-): 1 25-45.

Moser. C.C.. Page. C.C.. Farid. R. and Dunon. P.L. 1995. Biological electron transfer. J. Bioenerg. Biornembr. 27(3):263-274.

Munaro. M.. Tiranti, V.. Sandona. D.. Lamantea. E.. Uziel. G.. Bisson. R. and Zeviani. M. 1997. A single ce11 con~plementation class is comrnon to several cases of cytochrome c oxidase-defective Leigh's syndrome. Hum. Mol. Genet. 6(2):32 1-228.

Munn. E.A. 1 974. The structure of mi tochondria. Academic Press. London. P. 5-32.

O'Brien. T.W.. Denslow. ND.. Anders. J.C. and Counney. B.C. 1990. The translation system of marnmalian mitochondria. Biochim. Biophys. Acta. 1050: 1 74- 178.

Ohnishi. T. 1998. Iron-sulfur clusten/semiquinones in cornplex 1. Biochim. Biophys. Acta, 1364: 186-206.

Ohnishi. T.. Ragan. C.I. and Haiefi. Y. 1985. EPR studies of iron-sulfur clusters in isolated subunits and subfractions of NADH-ubiquinone oxidoreductase. J. Biol. Chem. 260(5):2782-2788.

Ohnishi. T.. Sled. V.D.. Yano. T.. Yagi. T.. Burbaev. D. and Vinogradov. A. 1998. S tnicture- function studies of iron-su1 fur clusters and semiquinones in the NADH-Q oxidoreductase segment of the respiratory cliain. Biochim. Biophys. Acta l365:jO 1-308.

Okun. J.G.. Lummen. P. and Brandt, U. 1999. Three classes of inhibitors share a common binding domain in mitochondrial complex 1 (NADH:ubiquinone oxidoreductase). J. Biol. Chem. 274(5): 2625-2630.

Okun. J.G.. Zickermann. V.. Zwicker. K.. Schagger. H. and Brandt, U. 2000. Binding of detergents and inhibiton to bovine complex 1 - a novel purification procedure for bovine complex 1 retaining full inhibitor sensitivity. Biochim. Biophys. Acta 1459: 77-87.

Onuchic. LN.. Beratan. D.N., Winkler. J.R. and Gray. H.B. 1992. Pathway analysis of protein electron-transfer reactions. Annu. Rev. Biophys. Biomol. Struct. 2139-377.

Orstavik. K.H.. Skjorten. F.. Hellebostad. M.. Haga. P. ruid Langslet, A. 1993. Possible S linked congenital mitochondrial cardiomyopathy in three families. J. Med. Genet. 30(4):269-73

Osborne, LA.. Martindale. D.. Scherer. S. W.. Shi. X.M.. Huizenga J.. Heng. H.Q.. Costa. T.. Pober. B.. Lew. L.. Brinkman. J.. Rornmens. I.. Koop. B. and Tsui. L.C. 1996. Identification of genes from a 500-kb region at 7q11.23 that is comrnonly deleted in Williams syndrome patients. Genomics. 36(2):328-336.

Palmisano. A.. Zara. V.. Honlinger. A.. Voua. A.. Dekker. P.J.. Pfanner. N. and Palmieri. F. 1998. Targeting and assembly of the oxoglutarate carrier: genenl principles for biogenesis of carrier proteins of the rnitochondrial inner membrane. Biochem. J. 333 (1):151-158.

Papa. S.. Guemeri. F.. Zanotti. F.. Capoua. G.. Fiermonte, M.. Cocco. T.. AltendorE K.. and Deckers-Hebersteit. G. 1992. FO and FI subunits involved in the gate and coupling function of rnitochondrial H+ ATP synthase. Am. N. Y. Acad. Sci. 671 :3 45-358.

Papa. S.. Scardanelli. A.M.. Cocco. T.. Speranza. F.. Scacco. S.C. and Technikova- Dobrova. 2. 1996. The nuclear-encoded 18 kDa (IP) AQDQ subunit of bovine heart cornplex 1 is phosphorylated by the mitochondrial CAMP-dependent protein kinase. FEBS Lett. 379:299-30 1.

Pata. 1.. Tensing, K. and Metspalu. A. 1997. A human cDNA encoding the homologue of NADH:ubiquinone oxidoreductase subunit B 1 3. Biochem. Biophys. Acta. 1350: 1 1 5- 118.

Patel, S.D.. Aebersold. R. and Attardi. G. 1990. cDNA-derived amino acid sequence of the NADH-binding 5 1 -kDa subunit of the bovine respiratory NADH dehydrogenase reveals striking similarities to a bacterial NAD(+)-reducing hydrogenase. Proc. Natl. Acad. Sci. U S A 88(10):4225-4229.

Patel. S.D.. Cleeter. M. W. and Ragan, C.I. 1988. Trammembrane orgnnization of mitochondrial NADH dehydrogenase as revealed by radiochemical labeling and cross-linking. Biochem. J. 256(2):529-535.

Pedersen. P.L. Arnzel. L.M. 1993. ATP synthases. Structure. rcaction center. mechanism. and regulation of one of nature's most unique machines. J. Biol. Chem. 268(14):9937- 9940.

Pilkington. S.. Skehel, J.M.. Gemis. R.B. and Walker. J.E. 1991. Relationship between mitochondrial NADH-Ubiquinone Reductase and bacterial NAD-Reducing Hydrogenase. Biochemistry. 302 166-2 1 75.

Pitkanen, S. 1997. Biochemical and diagnostic studies on complex 1 detkiency. Academic dissertation. Helsinki. pp.8

Pitkanen. S.. Raha. S. and Robinson. B.H. 1996a. Diagnosis of complex 1 deficiency in patients with lactic acidemia using skin fibroblast cultures. Biochem. Mol. Med. 59(2): 134-1 37.

Pitkanen. S., Feigenbaum. A.. Laframboise. R. and Robinson. B.H. 1 W6b. NADH- coenzyme Q reductase (complex 1) deficiency: heterogeneity in phenotype and biochemical findings. J. Inhent. Metab. Dis. 19: 67-686.

Plesokky. N.. Gardner. N.. Videira A. and Bramble. R. 2000. NADH dehydrogenase in i \ r ~ ~ ~ r ~ ~ p ~ r ~ crussu contains rnyistic acid covalently linked to the ND5 subunit peptide. Biochim. Biophys. Acta. 1495223-230.

Preis. D.. van der Pas. J.C.. Nehl. U,. Rohlen, DA.. Sackrnann. U., Jahnke. U. and Weiss. H. 1990. The 49 kDa subunit of'NADH: ubiquinone reductase (complex I) from iVt.t~ro.sporu crassa mitochondria: primary structure of the gene and the protein. Curr. Genet. 18(1):59-64.

Priller. J.. Scherzrr. C.R.. Faber. P.W.. MacDonald. M E . and Young. .-\.B. 1997. Fratauin gene of Friedreich's ataxia is targeted to mitochondria. Ann. Neurol. 42:765- 369.

Procaccio. V.. Depetris. D.. Soularue. P.. Mattei. MG.. Lunardi. J. and Issartel. J.P. 1997. cDNA sequence and chromosomal localization of the iVDUFS8 human gene coding for the 23 kDa subunit of the mitochondrial cornplex I. Biochirn. Biophys. Acta. 135137-11.

Procaccio. V.. de Sury. R.. Martinez. P.. Depetns. D.. Rabilloud. T.. Soularue. P.. Lunardi. J. and Issartel. J.P. 1 998. Mapping to 1 q X of the human gene (NDUFS2) encoding the 49-kDa subunit of the mitochondrial respiratory Complex I and immunodetection of the mature protein in mitochondria. blarnmalian Genome. 9:483- 484.

Procaccio. V.. Lescuyer. P.. Bourges. 1.. Beugnot. R.. Dubo jal. H.. Depetris. D.. Mousson. B.. Montfort, M.. Smeets. H.. de Coo. R. and IssarteI. J.P. 2000. Human NDUFS3 gene coding for the 30-kDa subunit of mitochondrial Complrx 1: genomic organization and expression. iMammaIian Genome. 11 :808-8 10.

Ragan. I.C. 1987. Structure of NADH-Ubiquinone Reductase (Complex 1). Cur. Top. Bioenerget. 15: 1-35.

Ragan. C.I. 1990. Structure and function of an archetypal respiratory chain complex: NADH-ubiquinone reductase. Biochem. Soc. Tram. lS(4):j 15-5 16.

Rassow. J.. Dekker. P.J.T., van Wilpe. S.. Meijer. M. and Soll. J. 1999. The preprotein translocase of the mitochondrid inner membrane: Function and evolution. J. Mol.Bio1. 286: 105- 120.

Reizer. A.. Pao. G.M. and Saier. M.H. 1991. Evolutionary relationships arnong the pennease proteins of the bacterial phosphoenolppvate: sugar phosphotransferase system. Construction of phylogenetic trees and possible relatedness to proteins of eukaryotic mitochondria. J. Mol. Evol. 33(2): 1 79- 193.

Robinson, B.H. 1989. Lacticacidemia. In: The metabolic basis of inherited disease. Scriver. C.. Baunet. A.. Sly, W. and Valle. D. (Eds.): McGray-Hill Inc. New York. pp.869-880.

Robinson. B.H. 1993. Lacticacidemia. Biochim. Biophys. Acta. 1182:23 1-244.

Robinson. B.H. 1 998a. Nuclear defects affecting mitochondrial function. In: Mitochondrial DNA mutations in agine. disease and cancer. Singh. K. (Ed): Landes Bioscience. Texas. pp. 185-204.

Robinson. B.CI. 1998b. Human complex 1 deficiency: clinical spectrum and involvement of oxygen free radicals in the pathogenicity of the defect. Biochim. Biophys. Acta. f 363:27 1-286.

Robinson. B.H.. De Meirleir. L.. Glerum. M.. Sherwood. G. and Becker. L. 1987. Clinical presentation of mitochondrial respiratory chain defects in NADH-coenzyme Q reductase and cytochrome oxidase: clues to pathogenesis of Leigh disease. J. Pediatr. 1 10(2):2 16-12?.

Robinson. B.H.. MacKay. N.. Petrova-Benedict. R.. Ozalp. 1.. Coskun. T. and Stacpoole. P.W. 1990. Defects in the E2 lipoyl transacetylase and the X-lipoyl containing component of the pyruvate dehydrogenase complex in patients with lactic acidemia. J. Clin. Invest. 85(6): 182 1-1 824.

Rosevear. P.. VanAken. T.. Bauter. J. and Ferguson-Miller. S. 1980. Alkyl glycoside detergents: A simpler synthesis and their effects on kinetic and physical properties of cytochrome c oxidase. Biochemistry. l9:J 108-4 1 15.

Runswick. M.J.. Feamley. LM., Skehel. J.M. and Walker. J.E. 1991. Presence of an acyl carrier protein in NADH:ubiquinone oxidoreductase fiom bovine heart rnitochondria. FEBS Lett. 286( 1 -2): 12 1 - 124.

Riissell. M.W.. du Manoir. S.. Collins. F.S. and Brody. L.C. 1997. Cloning of the human NADH:ubiquinone osidoreductase subunit B 13:localization to chromosome 7q32 and identification of a pseudogene on 1 1 p 15. Marnmalian Genome 8:60-6 1.

Saraste. M. 1 999. Oxidative phosphorylation at t h e m de siecle. Science. 283: 1488- 1492.

Sazanov. L.A.. Peak-Chew. S.Y.. Fearnley. LM. and Walker. J.E. 2000. Resolution of the membrane domain of bovine complex 1 into subcomplexes implications for the structural organization of the enzyme. Biochemistry. 39:7229-7235.

Sazanov. L. A. and Walker. J. E. 2000. Cryo-electron crystallography of two sub- complexes of bovine complex I reveals the relationship between the membrane and peripheral arms. J. Mol. Biol. 302:W-464.

Scacco. S.. Vergari. R.. Scarpulla R.C.. Technikova-Dobrova. Z.. SardaneIli. A.. Larnbo. R.. Lonisso. V. and Papa. S. 2000. CAMP-dependent phosphorylation of the nuclear encoded 18-kDa (IF) subunit of respiratory complex I and activation of the complex in serum-starved mouse fibroblast cultures. J. Biol. Chem. 275(X): 17378-1 7582.

Schagger. H. and Pfeiff'er. K. 7000. Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J. 19(8): 1777-1 783.

Schapira. A.H.V. and Cock. H.R. 1999. Mitochondrial myopathies and encephalomyopathies. Eur. J. Clin. Invest. 29:886-898.

Schapira. AH.. Cooper. J.M.. Dexter. D.. Clark. J.B.. Jenner. P. and Marsden. C.D. 1990. Mitochondrial complex 1 deficiency in Parkinson's disease. J. Neurochem. 54(3):823-827.

Schapira. A.H.. Cooper. J.M.. Morgan-Hughes. J.A.. Patel. S.D.. Cleeter. M.J.. Ragan. C.1 and Clark. J.B. 1988. Molecular basis of mitochondriai myopathies: polypeptide analysis in complex-1 deficiency. Lancet 1(8584):500-503.

Scherf. M.. Klingenhoff. W. and Werner. T. 7000. Highly specific localization of promoter regions in large genomic sequences by PromotrrInspector: A novel context analysis approach. J. Mol. Biol. 297599-606.

Schuelke. M.. LoetTen. J.. Mariman. E.. Srneitink. J. and van den Heuvel. L. 1998. Cloning of the hurnan mitochondrial 5 1 kDa subunit (iVDUFI.'l) reveals a 100% antisense homology of its 3'UTR with the 5'UTR of the gamma-interferon inducible protein (IP-30) precursor: is this a link between mitochondrial myopathy and intlarnmation? Biochem Biophys Res Commun. 2.15599-606.

Schuelke. M.. Smeitink. J.. Mariman. E.. Plecko. B.. Trijbels. F.. Stockler-Ipsiroglu. S. and van den Heuvel. L. 1999. Mutant NDUFV 1 subunit of mitochondrial complex 1 causes leukodystrophy and myoclonic epilepsy. Nat. Genet. 2 1 260-26 1.

Schuler. F.. Yano. T.. Di Bernardo. S.. Yagi. T.. Yankovskaya V.. Singr. T.P. and Casida LE. 1999. NADHquinone oxidoreductase: PSST subunit couples electron transfer From iron-sulfur cluster N2 to quinone. Proc. Natl. Acad. Sci. USA 9 6 4 4 9 - 4153.

Schulte, U., Fecke, W.. h l l , C., Nehls, U., Schmiede, A.. Schneider, R.. Ohnishi. T. and Weiss, H. 1994. In vivo dissection of the mitochondrial respiratory NADKubiquinone oxidoreductase (complex 1). Biochim. Biophys. Acta. 1187: 1 2 1-1 24.

Schulte. U.. Haupt. V.. Abelmann, A., Fecke. W.. Bron. B.. Rasmussen. T.. Friedrich. T. and Weiss. H. 1999. A reductase/isomerase subunit of mitochondrial NADH:ubiquinone oxidoreductase (complex 1) carries an NADPH and is involved in the biogenesis of the complex. J. Mol. Biol. 292569-580.

Schulte. U. and Weiss. H. 1995. Generation and charactenzation of NADH: ubiquinone oxidoreductase mutants in Neurospora crama. Methods Enzymol. 26030 14.

Shoffner. J.M., Brown. M.D.. Stugard, C.. Jun, AS., Pollock. S.. Haas. R.H.. Kaufman. A.. Koontz. D., Kim. Y. and Graham. J.R. 1995. Leber's hereditary optic neuropathy plus dystonia is caused by a rnitochondrial DNA point mutation. AM. Neurol. 38(2): 163-1 69.

Shoffner. J.M. and Wallace. D.C. 1994. Oxidative phosphorylation diseases and mitochondrial DNA mutations: Diagnosis and treatment. Ann. Rev. Nutr. 14:535-568.

Sirnonian. N. and Hyman. B.T. 1994. Functional alterations in Alzheimer's disease: selective loss of mitochondrial-encoded cytochrome oxidase mRNA in the hippocarnpal formation. J. Neuropathol. Exp. Neurol. 53:508-5 12.

Skehel. J.M.. Fearnley. LM. and Walkrr. J.E. 1998. NADH:ubiquinone oxidoreductase from bovine heart mitochondria: sequence of a novel 17.2-kDa subunit. FEBS Lett. 438330 1-305.

Sled. V.D.. Rudnitzky. N.I.. Hatefi. Y. and Ohnishi. T. 1994. Thermodynamic analysis of flavin in mitochondrial NADH:ubiquinone oxidoreductase (complex 1). Biochemistry 33(33): 10069- 10075.

Slipetz. D.M., Aprille. J.R.. Goodyear. P.R. and Rozen. R. 1991. Deficiency of complex 111 of the rnitochondrial respiratory chain in a patient with Facioscapulohumeral disease. Am. J. Hum. Genet. 48502-5 10.

Srneitink. J.A.M.. Loeffen. J.L.C.M.. Triepels. R.H.. Smeets. R.J.P.. Trijbels. J.M.F. and van den Heuvel. L.P. 1998a. Nuclear genes of human complex 1 of the mitochondrial electron transport chain: state of the art. Hum. Mol. Genet. 7( 10): 1 573- 1 579.

Srneitink. J.. Loeffen. J.. Smeets. R.. Triepels. R.. Ruitenbeek, W.. Trijbels. F. and van den Heuvel. L. l998b. Molecuiar characterization and mutational analysis of the human B 17 subunit of the mitochondrial respiratory chain complex 1. Hum. Genet. 103:245-250.

Stein. C.A. 1999. Two problems in antisense biotechnology: in vitro delivery and the design of antisense exprîments. Biochim. Biophys. Acta. l189( 1 ):45-52.

Strickland, S.. Huarte. J., Belin. D.. Vassalli, A.. Rickles. R.J. and Vassalli, J.D. 1988. Antisense RNA directed against the 3' noncoding region prevents dormant mRNA activation in mouse oocytes. Science 241(4866):680-684.

Suomalainen, A., Kaukonen, J.. Amati, P., Timonen, R., Haltia. M., Weissenbach, J.. Zeviani, M.. Somer. H. and Peltonen. L. 1995. An autosomal locus predisposing to deletions of mitochondrial DNA. Nat. Genet. 9(2): 146- 1 5 1.

Suzuki. H. and King. T.E. 1983. Evidence of an ubisemiquinone ndical(s) From the NADH-ubiquinone reductase of the mitochondrial respiratory chain. J. Biol. Chem. 258(1):352-358.

Takano. S.. Yano. T. and Yagi. T. 1996. Structural studies of the proton-translocating NADH-quinone osidoreductase (NDH- 1 ) of P n r a c ~ ~ c z ~ s clenitrific*nns: identity. propeny. and stoichiometry of the periphenl subunits. Biochemistry 35(28):9 1 20-9 127.

Tatuch. Y.. Christodoulou. J.. Feigenbaum. A.. Clarke. J.T.. Wherret. J.. Smith. C.. Rudd. N.. Petrova-Benedict. R. and Robinson. B.H. 1992. Heteroplasmic rntDNA mutation (T-- --G) at 8993 can cause Leigh disease when the percentage of abnorrnal mtDNA is high. Am. J. Hum. Genet. 50(4):852-858.

Tatuch. Y. and Robinson, B.H. 1993. The mitochondrial DNA mutation at 8993 associated with NARP slows the rate of ATP synthesis in isolated lymphoblast mitochondria. Biochem. Biophys. Res. Commun. 192( 1 ): 123- 128.

Tensing. K.. Pata 1.. Wittig. 1.. Wehnert. M. and Metspalu. A.. 1999. Genomic organization of the human complex I 13-kDa subunit gene NDUFAj. Cytogenet. Cell Genet. 84( 1-2): 125- 127.

Tiranti. V.. Hoennagel. K., Carrouo. R.. Galimberti. C.. Munaro. M.. Granatiero. M.. Zelante. L.. Gasparini. P.. Marzella. R.. Rocchi. M.. Bayona-Bafaluy. M.P.. Emiquez. J.A.. Uziel. G.. Bertini. E.. Dionisi-Vici. C.. Franco. B.. Meitinger. T. and Zeviani. M. 1998. Mutations of SURF-1 in Leigh disease associated with cytochrome c osidase deticiency. Am J Hum Genet. 63: 1609- 162 1.

T'on. C.. Hwang. D.M.. Dempsey. A.A. and Liew. C.C. 1997. Identification and primary structure of five human NADH-ubiquinone oxidoreductase subunits. Biochem. Biophys. Res. Commun. 241(2):589-594.

Tran-Betcke, A.. Wamecke. U.. Bocker. C.. Zaborosch. C. and Friedrich. B. 1990. Cloning and nucleotide sequences of the genes for the subunits of NAD-reducing hydrogenase of Alcnligenes nltrophzîs H 16. J. Bacteriol. 172(6):2920-29X

Triepels. R.. Smeitink. J.. Loeffen. J., Smeets, C.. Buskens. F.. Trijbels. F. and van den Heuvel. L. 1999a. The human nuclear-encoded acyl carrier subunit (NDUFABI) of the mitochondnal complex I in hurnan pathology. 22: 163- 1 73.

Triepels. R., Smeitink. J., Loeffen. J.. Smeets. R.. Trijbels. F. and van den Heuvel. L. 2000. Characterization of the human complex 1 NDLIFB7 and 17.2-kDa cDNAs and mutational analysis of 19 genes of the HP fraction in complex 1-deficient patients. Hum. Genet. lO6(4):385-39 1.

Triepels. R.H.. van den Heuvel. L.P.. Loeffen. J.L.C.M.. Buskens. C.A.F.. Smeets. R.J.P.. Rubio Gozalbo. M.E.. Budde. S.M.S.. Mariman. E.C.. Wijburg. F.A.. Barth. P.G.. Trijbels. J.M.F. and Smeitink. J.A.M. 1999b. Leigh syndrome associated with a mutation in the iVDliFS7 (PSST) nuclear encoded subunit of complex 1. Ann. Neurol. -t6(6): 787-790.

Triepels, R.. van den Heuvel. L.P.. Loeffen. J.. Srneets. R.. Trijbels. F. and Smeitink. J. 1 998. The nuclear-encoded human NADH:ubiquinone oxidoreductase XDUFA8 subunit: cDNA cloning. chromosomal localization. tissue distribution and mutation detection in complex 1-deficient patients. Hum. Genet. 103:557-563.

Tuschen. G.. Sackrnann. U., Nehls. U.. Haiker, H.. Buse. G. and Weiss. H. 1990. Assembly of NADH: ubiquinone reductase (complex 1) in Neurospon mitochondria. Independent pathways of nuclear-encoded and mitochondrially encoded subunits. J. Mol. Biol, 213(4):845-857.

Tyler. D. 1992. The mitochondrion in health and disease. VCH Publishers Inc. New York.

Tzagoloff. A. and Myers. A.M. 1986. Genetics of mitochondrial biogenesis. AM. Rev. Biochem. 55249-285.

van den Heuvel. L.. Ruitenbeek. W.. Smeets. R.. Gelma-Kohan. 2.. Elpeleg. O.. Loeffen. J.. Trijbels. F.. Mariman. E.. de Bruijn, D. and Smeitink. J. 1998. Demonstration of a new pathogenic mutation in human complex 1 deficiency: a 5-bp duplication in the nuclear gene encoding the 18-kD (AQDQ) subunit. Am. J. Hum-Genet. 62262-268.

Vinogradov. A.D. 1993. Kinetics. control. and mechanism of ubiquinone reduction by the mammalian respiratory chain-linked NADH-ubiquinone reductase. I. Bioenerg. Biomembr. 25(4):367-375.

Vinogradov. A.D.. Sled. V.D.. Burbaev. D.S.. Grîvemikova. V.G.. Moroz. I.A. and Ohnishi. T. 1995. Energy-dependent Complex 1-associated ubisemiquinones in submitochondrial particles. FEBS Lea. 370(1-2):83-87.

Voet. D. and Voet. J.G. 1990. Biochemistv. John Wiley and Sons. New York. P.506- 525.

Von Heijne. G. 1986. Mitochondrial targeting sequences may form amphiphilic helices. EMBO J. 5(6):1335-1342.

Walker, J.E. 1992. The NADH:ubiquinone oxidoreductase (complex 1) of respiratory chains. Q Rev Biophys. 25253424.

Walker. J.E.. Arizmendi. J.M.. Depuis. A.. Feardey. LM.. Finel. M.. Medd. S.M.. Pilkington. S.J.. Runswick, M.J. and Skehel. J.M. 1992. Sequences of 20 subunits of NADH:ubiquinone oxidoreductase from bovine heart mitochondria. Application of a novel strategy for sequencing proteins using the polymerase chain reaction. J. Mol. Biol. 226(4): 105 1-72.

Walker. LE.. Skehel. J.M. and Buchanan. S.K. 1995. Structural analysis of NADH: ubiquinone oxidoreductase from bovine heart mitochondria. Methods Enzymol. 260: 14- 34.

Wallace. D.C. 1992. Diseases of the rnitochondrial DNA. Annu. Rev. Biochem. 61:1175- 1212.

Wallace. D.C. 1993. Mitochondrial diseases: Genotype versus phenotype. Trends in Genetics. 9: 138- 133.

Weidner. U.. Geier. S.. Ptock. A.. Friedrich. T.. Leif. H. and Leonard. K.R. 1993. The gene locus of the proton-translocating NADH: ubiquinone oxidoreductase in Eschcrkhirr coli. Oqanization of the 14 genes and relationship between the derived proteins and subunits of mitochondrial complex 1. J. Mol. Biol. 233( 1): 109- 112.

Weiss. H.. Friedrich. T.. Hofhaus. G. and Preis. D. 199 1 . The respiratoty-chain complex NADH dehydrogenase (complex 1) of rnitochondria. Eur. J. Biochem. 197563376.

Westermann. B.. Prip-Buus, C.. Neupert. W. and Schwarz. E. 1995. The role of the GrpE homologue. Mgel p. in mediating protein impon and protein folding in mitochondria. EMBO J. l4(14):3452-3460.

Wilde. C.D. 1986. Pseudogenes. CRC Cnt. Rev. Biochem. 19(4):323-352.

Wong. T.W. and Clayton. D.A. 1985. In vitro replication of human mitochondrial DNA: Accurate initiation at the origin of light strand synthesis. Ce11 42:95 1-958.

Wong. G.H.. Elwell. J.H.. Oberley. L.W. and Goeddel. D.V. 1989. Manganese superoxide dismutase is essential for cellular resistance to cytotoxicity of tumor necrosis factor. Cell58(5):923-93 1.

Yaffe. M.P. 1999. The machinery of mitochondrial inheritance and behaviour. Science 283: 1493-1497.

Yagi. T. and Hatefi, Y. 1988. Identification of the Dicyclohexylcarbodiimide-binding subunit of NADH-ubiquinone oxidoreductase (Complex 1). J. Biol. Chem. 263(3 l):l6ljO-l6I 55.

Yagi. T.? Xu. X. and Matsuno-Yagi. A. 1992. The energy-transducing NADH-quinone oxidoreductase (NDH- 1 ) of Paracocclrs denitrificans. Biochim. Biophys. Acta. 1101(2):181-183.

Yamaguchi. M.. Belogrudov. G.I. and Hatefi. Y. 1998. Mitochondrial NADH- Ubiquinone Oxidoreductase (Complex 1). Effect of substrates on the Fragmentation of subunits by trypsin. J. BioLChem. 273( l4):8094-8098.

Yamaguchi. M.. Belogrudov. G.. Matsuno-Yagi. A. and Hatefi. Y. 2000. The multiple nicotinamide nucleotide-binding subunits of bovine heart mitochondrial NADH:ubiquinone oxidoreductase (complex 0. Eur. J. Biochem. 2673329436.

Yamaguchi. M. and Hatefi. Y. 1993. Mitochondrial NADH:ubiquinone osidoreductase (complex 1): proximity of the subunits of the flavoprotein and the iron-sulfur protein subcomplexes. Biochemistry. 32: 1935-1 939.

Yoshikawa. S.. Shinzawa-Itoh. K. and Tsukihan. T. 1998. Crystal structure of bovine heart cytochrorne c oxidase at 2.8 A resolution. J. Bioenerg. Biomembr. 3O( 1 ): 7- 14.

Yu. C.A.. Xia. D.. Kim. H.. Deisenhofer. J.. Zhang. L.. Kachurin. A M . and Yu. L. 1998. Structural basis of functions of the mitochondrial cytochrome bc 1 complex. Biochim. Biophy. Acta. 1365(1-2): 15 1-1 58.

Zhu. 2.. Yao. J.. Johns. T,. Fu. K.. De Bie. 1.. Macmillan. C.. Cuthbert. A.P.. Newbold. R.F.. Wang. J.. Chevrette. M.. Brown. G.K.. Brown. R.M. and Shoubridge. E.A. 1998. SURFI. encoding a factor involved in the biogenesis of cytochrome c oxidase. is mutated in Leigh syndrome. Nat. Genet. 2O:X 7-343.

Zhuchenko. O.. Wehnert. M.. Bailey. S.. Sheng Sun. Z. and Chi Lee. C. 1996. Isolation. mapping, and genomic structure of an X-linked gene for a subunit of human rnitochondrial complex 1. Genomics 3728 1-288.

Zubay. G. 1993. Biochemistrv. W m C Brown Communication. Dubuque. pp.379-107.