The 9.8 kDa Subunit of Complex I, Related to BacterialNa1-translocating NADH Dehydrogenases, is Requiredfor Enzyme Assembly and Function inNeurospora crassa

Isabel Marques1, Margarida Duarte1 and Arnaldo Videira1,2*

1Instituto de Biologia Moleculare Celular, Universidade doPorto, Rua do Campo Alegre823, 4150-180 Porto, Portugal

2Instituto de CienciasBiomedicas de Abel SalazarUniversidade do PortoRua do Campo Alegre 8234150-180 Porto, Portugal

A nuclear gene encoding a 9.8 kDa subunit of complex I, the homologueof mammalian MWFE protein, was identified in the genome of Neurosporacrassa. The gene was cloned and inactivated in vivo by the generation ofrepeat-induced point mutations. Fungal mutant strains lacking the9.8 kDa polypeptide were subsequently isolated. Analyses of mito-chondrial proteins from mutant nuo9.8 indicate that the membrane andperipheral arms of complex I fail to assemble. Respiration of mutant mito-chondria on matrix NADH is rotenone-insensitive, confirming that the9.8 kDa protein is required for the assembly and activity of complex I.We found a similarity between the MWFE homologues and the C-terminalpart of the nqrA subunit of bacterial Naþ-translocating NADH:quinoneoxidoreductases (Naþ-NQR), suggesting a link between proton-pumpingand sodium-pumping NADH dehydrogenases.

q 2003 Elsevier Science Ltd. All rights reserved

Keywords: mitochondria; NADH dehydrogenases; repeat-induced point-mutations; sodium translocation; Neurospora crassa*Corresponding author

Introduction

Mitochondrial NADH:ubiquinone oxidoreduc-tase (complex I, EC 1.6.5.3) transfers electronsfrom NADH to ubiquinone and links this processwith translocation of protons across the innermembrane, thereby contributing to the establish-ment of a transmembrane proton gradient neededfor the synthesis of ATP.1 Experiments with thefungus Neurospora crassa indicated that the enzymeis composed of two distinct subcomplexes,arranged perpendicularly to each other in anL-shaped structure, that undergo independentassembly. The membrane arm is embedded in themitochondrial membrane, while the peripheralarm is mainly protruding into the mitochondrialmatrix.2 Mammalian NADH:ubiquinone oxido-reductase might contain 46 subunits,3 seven ofwhich are encoded by mitochondrial DNA, andcontains non-covalently bound FMN and six toeight iron–sulphur clusters as prosthetic groups.4

Bacteria also possess enzymes equivalent toeukaryotic complex I, with a similar constitution

of prosthetic groups but with much fewer proteins,only 13–14 subunits. The prokaryotic enzymescontain homologues of the seven proteins encodedby mitochondria and homologues of seven pro-teins encoded by the nucleus in fungi or mammals.These core subunits are involved in the knowncomplex I activity, electron transport coupled toproton translocation across a membrane.5,6 Themany additional (“accessory”) subunits of mito-chondrial complex I are likely involved in otherfunctions of the enzyme (e.g. regulation or bio-synthetic activities) but their precise role remainslargely unknown. Besides complex I, two othertypes of NADH:quinone oxidoreductases can befound in respiratory chains, the alternative NADHdehydrogenases (NDH-2) and the bacterial Naþ-translocating NADH dehydrogenases (Naþ-NQR).One or more NDH-2 are present in differentorganisms, as single polypeptides or dimers,working without an energy coupling site.7 Naþ-NQR are complexes of six independent subunitsthat couple electron transfer to the pumping ofsodium ions across the bacterial membrane. Basedon comparison of protein sequences, no relation-ship between the three types of respiratory chainNADH dehydrogenases has been found so far.8,9

Mutations in both nuclear and mitochondrial

0022-2836/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved

E-mail address of the corresponding author:asvideir@icbas.up.pt

Abbreviations used: CI, complex I.

doi:10.1016/S0022-2836(03)00443-1 J. Mol. Biol. (2003) 329, 283–290

genes of complex I, including those encodingaccessory subunits, account for mitochondrialdisease.10,11 Recently, it was found that mutantforms of the MWFE subunit, including a conserva-tive R50K substitution, rendered complex I inactivein Chinese hamster cells interfering with complex Iassembly.12,13 The corresponding human NDUFA1gene, located on chromosome Xq24 and consistingof three exons, codes for a polypeptide of 70amino acid residues and is highly expressed incardiac and skeletal muscle.14 The protein isimported into mitochondria without proteolyticprocessing4 and is likely inserted in the mito-chondrial internal membrane.13 The MWFE poly-

peptide is highly conserved among animals,13 butno specific function was yet assigned to thisprotein.

This prompted us to look for and characterisethe MWFE homologue of N. crassa, since the fun-gus has been used as a powerful genetic system.15

Complex I of N. crassa contains at least 37 subunits,including an homologue of the mammalian MWFEpolypeptide. Overall, the fungal enzyme is quitesimilar to that of mammals. Up to now, only fourof the fungal subunits could not be related tomammalian subunits of complex I7,16,17 (unpub-lished data). Here, we report the identification andcharacterisation of the 9.8 kDa MWFE-homologue

Figure 1. Characteristics of the N. crassa MWFE homologue. (A) Genomic DNA structure. The boxes representsequences present in cDNA clones. (B) cDNA sequence and deduced primary structure of the 9.8 kDa polypeptide.Stop codons preceding in-frame the initiation codon as well as sequences used for primer design are underlined. Thedetermined cDNA sequence was combined with EST sequences (upstream of the initiation codon, EMBL accessionBG280238). (C) Alignment of the MWFE homologues from Varecia variegata (Vv), Gorilla gorilla (Gg), human (Hs),Mus musculus (Mm), hamster (Cg), Bos taurus (Bt), and N. crassa (Nc) with C-terminal residues of the nqrA subunitsof Naþ-translocating NADH dehydrogenase from H. influenzae (Hi), Neisseria meningitidis Z2491 (Ne), N. meningitidisMC58 (Nm), Pasteurella multocida (Pm), Haemophilus somnus (He) and Yersinia pestis KIM (Yp). Amino acid residuesidentical to N. crassa or present in at least seven sequences are shadowed. The asterisks below the sequences markconservative amino acid substitutions.

284 CI Assembly and Function in a Neurospora Mutant

of N. crassa. Our results indicate that, as in ham-ster cells, the protein is required for complex Iassembly and function. Apparently, lack of theprotein prevents assembly of the peripheral andmembrane domains of the enzyme. We also foundthat MWFE is similar to the C-terminal part ofthe nqrA subunit of bacterial Naþ-translocatingNADH dehydrogenases. Thus, for the first time,a link between proton-pumping and sodium-pumping NADH:quinone oxidoreductases ofrespiratory chains is recognized.

Results

Identification of the fungal MWFE homologueand its relationships

A computer search in the N. crassa genome,available from the Neurospora Sequencing Project,Whitehead Institute/MIT Center for GenomeResearch, led to the identification of a gene poten-tially encoding the fungal homologue of theMWFE subunit of mammalian complex I. Thegene structure is not easy to predict because thefungal protein is not highly homologous to theirmammalian counterparts (Figure 1). Therefore,we cloned and fully sequenced both strands of thecorresponding cDNA. Two in-frame stop codonspreceding the presumed ATG initiation codon con-firmed that the latter is correct. The single-copygene is located on linkage group II of the N. crassagenome and comprises four exons. The proteinhas 86 amino acid residues, a molecular mass of9750 Da, without a predicted cleavable signalsequence.

An interesting finding was the similaritybetween the MWFE protein and the C-terminalpart of the nqrA subunit of bacterial Naþ-trans-locating NADH:quinone oxidoreductases (Figure1(C)). In the region of similarity, 21 amino acidresidues are conserved between nqrA and theN. crassa 9.8 kDa subunit of complex I (24% iden-tity, or 55% similarity when conservative aminoacid substitutions are considered). This similarityis comparable to the homology between the fungal9.8 kDa polypeptide and mammalian MWFE. Alsoonly 21 amino acid residues are conserved betweenthe fungal and hamster subunits of complex I (30%identity). In addition, some very conserved resi-dues between the fungal and mammalian proteinsare also present in the Naþ-NQR A subunit.It should also be noticed that while some aminoacid residues of nqrA are conserved in the fungalsubunit of complex I and not in the mammalianproteins, some other residues are conserved inthe mammalian and not in the fungal protein.Combined with the fact that both MWFE andnqrA are subunits of (energy-coupling) NADH:quinone oxidoreductases, their relationship can beconsidered significant.

Since the peripheral and membrane arms ofcomplex I assemble independently of each other,

most null-mutants in subunits of one arm do notaffect assembly of the other arm of the enzyme.Therefore, we checked the presence of the 9.8 kDapolypeptide in different mutants of complexI. Mutant mitochondria were analysed by Westernblotting with antiserum against the 9.8 kDa protein(Figure 2). The polypeptide is clearly visible inwild-type mitochondria as well as in mitochondriafrom the peripheral arm mutants nuo21.3c18 andnuo24.19 In contrast, the 9.8 kDa polypeptide isonly faintly visible or not detected, respectively,in the membrane arm mutants nuo12.320 andnuo20.8.21 These results indicate that the 9.8 kDasubunit is part of the membrane arm of complexI. They are also in agreement with the findingsthat the protein is rather unstable when it is notassembled.13

The 9.8 kDa protein is required for complex Iassembly and function

We obtained a null-mutant of the 9.8 kDa sub-unit of complex I through the generation ofrepeat-induced point mutations (methylations andGC to AT transitions) in the respective gene. Wetook advantage of the fact that duplicatedsequences in the genome of N. crassa are frequentlyinactivated by repeat-induced point mutationswhen passed through a genetic cross.22 Thus, wecloned a 1.7 kb DNA fragment containing thenuo-9.8 gene and duplicated it in the genome ofN. crassa by transformation. The transformant wascrossed with a wild-type strain of the oppositemating type and individual ascospore progenywere isolated. Among them, the mutant strainnuo9.8, lacking the 9.8 kDa subunit of complex I,was identified by immunoblotting analysis ofmitochondrial proteins with antiserum against theprotein (Figure 2).

The state of assembly of complex I in mutantnuo9.8 was evaluated by gradient centrifugationanalysis of mitochondria solubilised with a non-denaturing detergent. Fractions of the gradientswere analysed for NADH:hexa-amminorutheniumoxidoreductase activity, a typical artificial activityof complex I or its peripheral arm alone.23 The gra-dient fractions were also checked for the presenceof complex I subunits by Western blotting analysisof aliquots with antiserum against specific proteins

Figure 2. Presence of the 9.8 kDa polypeptide in mito-chondria from complex I mutants. Total mitochondrialproteins (100 mg) from wild-type (1 and 7) and mutantsnuo9.8 (2), nuo20.8 (3), nuo12.3 (4) nuo21.3c (5) andnuo24 (6) were analysed by Western blotting with anantiserum against the 9.8 kDa protein.

CI Assembly and Function in a Neurospora Mutant 285

of the peripheral and membrane arms of theenzyme (Figure 3). In the wild-type strain, most ofthe oxidoreductase activity elutes in fractions 9–11(Figure 3(A)), in agreement with the elution profileof the proteins analysed by Western blotting(Figure 3(B)), reflecting the presence of complex Iin these fractions. In the mutant strain nuo9.8,both the oxidoreductase activity and the 30.4 kDasubunit of the peripheral arm of complex I24 elutewith a peak in fraction 8 of the gradient (Figure3(A) and (C), respectively). This material likelyrepresents the peripheral arm of complex I. Infact, the total NADH:ferricyanide reductase activi-ties of nuo9.8 mitochondria are about 40% ofthe wild-type values (not shown), a characteristicof mutants able to form the peripheral arm ofcomplex I.23 Most of the 12.3 kDa subunit of themembrane arm of complex I25 is detected in frac-tions 6–8 of the sucrose gradients, with a peak infraction 7. This material might represent an almostcomplete membrane arm of complex I or an inter-mediate of it. This phenotype resembles that ofthe nuo12.3 mutant itself.20 We conclude that theperipheral and membrane arms of complex I failto assemble in the nuo9.8 mutant.

Since the activities measured in the sucrose gra-dients do not reflect the physiological activity ofcomplex I, we wanted to know if electrons couldstill be accepted from NADH by nuo9.8 complexand passed on to the rest of the respiratory chain.

Therefore, we determined the effect of the complexI inhibitor rotenone upon the respiration of isolatedmutant mitochondria (Table 1). The oxygen uptakeof wild-type mitochondria respiring on pyruvate/malate, which generates NADH in the matrix, wasinhibited by about 50% with rotenone. While mito-chondria of the mutant nuo9.8 displayed similarlevels of respiration, they were mostly resistantto rotenone inhibition (about 15% inhibition). Thislevel of rotenone inhibition is also obtained withother non-functional mutants of complex I, likenuo5126 and nuo30.4,27 and can be considered back-ground inhibition due to some unspecific effect ofthe drug (not shown). These results suggest thatmutant complex I is not active and respirationis carried out by an internal alternative NADHdehydrogenase, in agreement with the resultsobtained with the sucrose gradients. Interestingly,we found that the respiratory activity of mutantmitochondria is about three times higher thanthat of wild-type when NADH is used directly assubstrate (Table 1). This was somehow unexpectedbecause this NADH oxidation reflects the activitiesof external alternative NADH dehydrogenases thatoxidise cytosolic NADH.28

Independent evidence that complex I is notassembled in the nuo9.8 strain came from analysisof proteins from mutant mitochondria in non-denaturing gels. In contrast to wild-type, no highmolecular weight complex I can be detected whenmutant mitochondria are subjected to Blue Nativeelectrophoresis (Figure 4). Complex I (CI) was

Figure 3. Complex I assembly in mutant nuo9.8. Mito-chondria were isolated, solubilised with Triton X-100and centrifuged in sucrose gradients. Fractions of thegradients (numbered 1–12 from top to bottom) were col-lected and analysed for NADH:hexa-amminorutheniumoxidoreductase activity, shown in (A) (filled circles,wild-type; open circles, mutant nuo9.8). The same frac-tions from the wild-type strain (B) and mutant nuo9.8(C) were analysed by Western blotting with a mixture ofindividual antisera against the subunits of complex Iindicated at the left.

Table 1. Measurements of the respiratory activities ofmitochondria. The activities are expressed in nmol ofO2/mg protein/min. Percent inhibition is indicatedbetween brackets

Mitochondria Wild-type nuo9.8

Malate 60.2 58.2þRotenone 31.7(47) 50.0(14)þAntimycin A 0(100) 0(100)

NADH 90.6 288.6þAntimycin A 0(100) 0(100)

Figure 4. Analysis of mitochondrial complexes by BlueNative electrophoresis. Mitochondrial proteins (200 mg)from wild-type (1) and mutant nuo9.8 (2) were solu-bilised with dodecyl maltoside and subjected to BlueNative electrophoresis. Complexes I (CI), III (CIII) andV (CV) are indicated.

286 CI Assembly and Function in a Neurospora Mutant

identified by Western blotting with antisera againstspecific subunits (not shown). Complex III (CIII)and complex V (CV) were tentatively assigned inthe Figure by comparison with the patternsobtained with bovine mitochondria.29 Resultsfrom crosses involving the nuo9.8 mutant furthersupport the conclusion that complex I is not activein this strain. Homozygous crosses with nuo9.8mutants resulted in the formation of barrenperithecia, failing to produce sexual spores, asshown with homozygous crosses between mutantscarrying a non-functional complex I.18,27 Inaddition, we recently identified and disrupted aninternal alternative NADH dehydrogenase ofN. crassa mitochondria, the NDI1 protein. Wefound that double mutants from crosses betweenndi1 mutants and mutants without a functionalcomplex I are not viable (unpublished results).Similarly, no double mutants were obtained froma cross between the nuo9.8 and the ndi1 mutant,while wild-type recombinants were readily iso-lated (Table 2). As a positive control, the crossbetween the ndi1 mutant and the complex Inuo21.3a mutant (which carries a functionalenzyme30) yielded double mutants.

Discussion

Here, we described the identification and charac-terisation of the 9.8 kDa subunit of complex I fromN. crassa, which is the fungal homologue of themammalian MWFE protein. This polypeptide ispresent in eukaryotic but not in prokaryotic com-plex I. MWFE is highly conserved among animalsand a small region of variability of about eightamino acid residues (residues 39–46) seems tobe involved in species-specific compatibility.13

The human and Xenopus proteins are also veryconserved (70% identity13). With cloning of theNeurospora homologue, we now show that thepolypeptide sequence can deviate considerably.Only two small regions (surrounding that regionof variability in animals) and a few scatteredamino acid residues are conserved between thefungal and animal proteins.

We have isolated an N. crassa mutant straindevoid of the 9.8 kDa polypeptide after the gener-ation of repeat-induced point mutations in therespective gene, in vivo. Analysis of complex Ifrom the null-mutant indicated that the 9.8 kDapolypeptide is required for the assembly andfunction of the enzyme. By the analysis of mito-chondrial proteins in sucrose gradients or in native

gels, we could not detect the formation of a fullcomplex in the mutant. In addition, respiration ofnuo9.8 mitochondria on matrix-generated NADHis insensitive to inhibition by the complex I inhibi-tor rotenone. Independent support of that con-clusion was obtained from mating experiments.As described for other non-functional complex Imutants of N. crassa,18 it was not possible to crossnuo9.8 with itself and obtain viable progeny(homozygous crosses) and it was not possible toobtain double mutants from crosses betweennuo9.8 and ndi1 mutants (which lack an internalalternative NADH dehydrogenase). These resultsare in agreement with previous findings in hamstercell mutants. It was shown that truncated formsor specific amino acid alterations in the MWFEprotein interfere with complex I formation andrender the enzyme inactive.12,13 We extended theseobservations showing that the peripheral andmembrane arms of complex I fail to assemble.

An interesting finding in the nuo9.8 mutant wasthat the activities of alternative NADH dehydro-genases were raised as compared with the wild-type strain, especially, the observation that therespiratory activity of mutant mitochondria ismuch higher when NADH is used directly assubstrate (Table 1). This reflects the activitiesof external alternative NADH dehydrogenases,which oxidise cytosolic NADH, and it is not clearhow a deficiency in matrix NADH oxidation dueto the complex I mutation would have an effect onthese external enzymes. An enhancement in exter-nal (and internal) NADH oxidase activities wasalso noticed in plant mitochondrial mutants withcomplex I deficiency.31,32 Preliminary results withspecific antibodies did not reveal higher levelsof external alternative NADH dehydrogenases innuo9.8 mitochondria as compared with the wild-type strain (not shown). It is possible that theexistence of pathways for exchanging matrix/cyto-solic NADH explain these observations. On theother hand, they might suggest a more directregulation of alternative NADH dehydrogenaseenzymes by complex I. Post-transcriptional regu-lation and direct enzyme regulation control theactivity of alternative oxidase, a protein thattransfers electrons from ubiquinol to oxygen andis thus a bypass of the cytochrome pathway.33 – 35

A major finding of this work was the similaritybetween the MWFE homologues and the C-termi-nal part of the nqrA subunit of bacterial Naþ-trans-locating NADH:quinone oxidoreductases.8 Despitethe fact that the nqrA protein is much bigger (447amino acid residues in Haemophilus influenzae)than MWFE, there are other examples of relation-ships between complex I subunits and biggerbacterial proteins. For instance, the a-subunit ofthe NADþ hydrogenase of Alkaligenes eutrophusrepresents a fusion between the 51 kDa and24 kDa subunits of complex I36 and the ORF5 sub-unit of formate hydrogenlyase of Escherichia colirepresents a fusion between the 49 kDa and30.4 kDa subunits of complex I.37 The conclusion

Table 2. Number of spores resulting from crossesbetween the ndi1 strain and complex I mutants

ProgenyCross nuo ndi1 Wild-type nuo, ndi1 Total

nuo9.8 £ ndi1 6 16 11 0 33nuo21.3a £ ndi1 3 6 6 3 18

CI Assembly and Function in a Neurospora Mutant 287

that MWFE and nqrA are related is based onsequence as well as on functional similarity, sinceboth polypeptides belong to (energy-coupling)NADH:quinone oxidoreductases. Contrary to whatis currently believed, this similarity suggests a rela-tionship between proton-pumping and sodium-pumping NADH dehydrogenases and might bereminiscent of a common origin of both complexes.On the other hand, MWFE homologues are notpresent in bacterial complex I.5 This might indicatethe existence of a specific function of mitochondrialcomplex I that is shared by bacterial Naþ-NQR.Clearly, a more precise definition of the roles ofMWFE and nqrA in their respective complexes isrequired.

Material and Methods

Molecular cloning and generation of antisera

Standard cloning techniques were employed.38,39 Theplasmid vectors pGEM-T easy (Promega) and pCSN4440

were used for bacterial subcloning and fungal trans-formation, respectively. A fragment of N. crassa genomicDNA of 1693 bp, containing the nuo-9.8 gene, was ampli-fied by PCR using the two specific primers 50-CGCTTTGTTCTGAGCTCCCGGC-30 and 50-GAAGCACCATTAGGCAGTGCTGT-30. The first (upstream) primer has arestriction site for Sac I (bolded) that was introduced bychanging two bases (underlined). The amplified 1.7 kbfragment was cloned in the pGEM-T easy vector inE. coli DH5a. Then, it was excised from the recombinantvector with the enzymes Sac I and Eco RV and cloned inthe pCSN44 vector, originating pCSN-15SE.

In order to express the 9.8 kDa protein, N. crassacDNA was amplified from the mycelial library M-1cloned in Uni-Zap XR (obtained from the FungalGenetics Stock Center) by PCR using the primer 50-ATGCCTGTCCCATTCGAGGCCCTC-30 that hybridises onthe initial ATG codon (bold). The second (downstream)primer was the same used in the amplification of geno-mic DNA. The PCR product was cloned in the pGEM-Teasy vector. Then, the segment with the gene was excisedfrom the recombinant plasmid with the enzymes Sph Iand Pst I (which cut in pGEM-T) and cloned in thepQE31 vector (Qiagen) in E. coli M15. The resultingfusion protein, consisting of the 9.8 kDa protein and 26extra amino acid residues at the N terminus, wasinduced with IPTG and purified in nickel columnsfollowed by SDS-PAGE. Rabbits were immunised asdescribed.41

Mutant isolation

N. crassa wild-type strains 74-OR23-1A and 74-OR8-1awere grown and handled according to standardprocedures.42 Spheroplasts were prepared from seven-day old conidia of the 74-OR8-1a strain and transformedwith the recombinant plasmid pCSN-15SE. The transfor-mants were selected in plates containing hygromycin B(100 mg/ml) and purified by successive asexual transfersin slants of Vogel’s minimal media containing 50 mg/mlof the same drug.20 In order to select single-copy trans-formants, genomic DNA of individual transformantswas isolated from 20 ml mycelial cultures25 and analysedby Southern blotting, using DNA encoding the 9.8 kDa

protein as a probe. A single-copy transformant was sepa-rately crossed with the 74-OR23-1A wild-type strain. Themutant strain nuo9.8 was identified among sporeprogeny by immunoblotting analysis of mitochondrialproteins with antiserum against the 9.8 kDa subunit ofcomplex I.

Oxygen consumption

Purified mitochondria were prepared as before.28 Toassess their quality, the activities of cytochrome c oxidase(EC 1.9.3.1) and malate dehydrogenase (EC 1.1.1.37)were measured in the presence and in the absence ofTriton X-10043 to calculate the latent activities of theenzymes. Oxygen uptake was measured polaro-graphically at 25 8C with a Clark-type oxygen electrode(Hansatech), in a total volume of 1 ml. The NADHassays were started with the addition of 1 mM NADHto the reaction medium (0.3–0.5 mg protein, 0.3 Msucrose, 10 mM potassium phosphate (pH 7.2), 5 mMMgCl2, 1 mM EGTA, 10 mM KCl, 4 mM carbonyl cyanidem-chlorophenylhydrazone and 0.02% (w/v) bovineserum albumin). The malate assays were started withthe addition of 10 mM malate to reaction medium con-taining 1 mM NADþ and 10 mM pyruvate. Rotenoneand antimycin A were added to final concentrations of40 mM and 0.2 mg/ml, respectively.

Protein analysis and enzymatic activities

The techniques used for protein determination,44 SDS-PAGE,45 blotting and incubation of blots withantisera,41,46 detection of alkaline phosphatase-conju-gated second antibodies,47 sucrose gradient centrifu-gation analysis of detergent solubilised mitochondrialproteins,30 NADH:hexa-amminoruthenium (III)48 andNADH:ferricyanide reductase activities49 and BlueNative gel electrophoresis of respiratory complexessolubilised with 1.7% dodecyl-b-D-maltoside29 havebeen published. Web-based computer programs such asBlast and ClustalW (and the Blosum 45 matrix to identifyconservative amino acid alterations) were used forbioinformatics.

Acknowledgements

We thank Monica Rodrigues for preparing rabbitantiserum, Dr Jorge Vieira for discussions and MrsLaura Pinto for excellent technical assistance.This research was supported by Fundacao para aCiencia e a Tecnologia from Portugal and thePOCTI program of QCA III through researchgrants to A.V. and fellowships to I.M. and M.D.

References

1. Hatefi, Y. (1985). The mitochondrial electron trans-port and oxidative phosphorylation system. Annu.Rev. Biochem. 54, 1015–1069.

2. Weiss, H., Friedrich, T., Hofhaus, G. & Preis, D.(1991). The respiratory-chain NADH dehydrogenase(complex I) of mitochondria. Eur. J. Biochem. 197,563–576.

3. Carroll, J., Shannon, R. J., Fearnley, I. M., Walker, J. E.

288 CI Assembly and Function in a Neurospora Mutant

& Hirst, J. (2002). Definition of the nuclear encodedprotein composition of bovine heart mitochondrialcomplex I: identification of two new subunits. J. Biol.Chem. 277, 50311–50317.

4. Walker, J. E. (1992). The NADH:ubiquinone oxido-reductase (complex I) of respiratory chains. Q. Rev.Biophys. 25, 253–324.

5. Weidner, U., Geier, S., Ptock, A., Friedrich, T., Leif, H.& Weiss, H. (1993). The gene locus of the proton-translocating NADH: ubiquinone oxidoreductase inEscherichia coli. Organization of the 14 genes andrelationship between the derived proteins and sub-units of mitochondrial complex I. J. Mol. Biol. 233,109–122.

6. Yagi, T. (1993). The bacterial energy-transducingNADH-quinone oxidoreductases. Biochim. Biophys.Acta, 1141, 1–17.

7. Videira, A. & Duarte, M. (2002). From NADH toubiquinone in Neurospora mitochondria. Biochim.Biophys. Acta, 1555, 187–191.

8. Hase, C. C. & Barquera, B. (2001). Role of sodiumbioenergetics in Vibrio cholerae. Biochim. Biophys.Acta, 1505, 169–178.

9. Barquera, B., Hellwig, P., Zhou, W., Morgan, J. E.,Hase, C. C., Gosink, K. K. et al. (2002). Purificationand characterization of the recombinant Na(þ )-translocating NADH:quinone oxidoreductase fromVibrio cholerae. Biochemistry, 41, 3781–3789.

10. Wallace, D. C. (1992). Diseases of the mitochondrialDNA. Annu. Rev. Biochem. 61, 1175–1212.

11. Triepels, R. H., Van Den Heuvel, L. P., Trijbels, J. M.& Smeitink, J. A. (2001). Respiratory chain complex Ideficiency. Am. J. Med. Genet. 106, 37–45.

12. Au, H. C., Seo, B. B., Matsuno-Yagi, A., Yagi, T. &Scheffler, I. E. (1999). The NDUFA1 gene product(MWFE protein) is essential for activity of complex Iin mammalian mitochondria. Proc. Natl Acad. Sci.USA, 96, 4354–4359.

13. Yadava, N., Potluri, P., Smith, E. N., Bisevac, A. &Scheffler, I. E. (2002). Species-specific and mutantMWFE proteins: their effect on the assembly of afunctional mammalian mitochondrial complex I.J. Biol. Chem. 277, 21221–21230.

14. Zhuchenko, O., Wehnert, M., Bailey, J., Sun, Z. S. &Lee, C. C. (1996). Isolation, mapping, and genomicstructure of an X-linked gene for a subunit ofhuman mitochondrial complex I. Genomics, 37,281–288.

15. Davis, R. H. & Perkins, D. D. (2002). Timeline: Neu-rospora: a model of model microbes. Nature Rev.Genet. 3, 397–403.

16. Videira, A. (1998). Complex I from the fungus Neuro-spora crassa. Biochim. Biophys. Acta, 1364, 89–100.

17. Videira, A. & Duarte, M. (2001). On complex I andother NADH:ubiquinone reductases of Neurosporacrassa mitochondria. J. Bioenerg. Biomembr. 33,197–203.

18. Duarte, M. & Videira, A. (2000). Respiratory chaincomplex I is essential for sexual development inNeurospora and binding of iron sulfur clusters arerequired for enzyme assembly. Genetics, 156,607–615.

19. Almeida, T., Duarte, M., Melo, A. M. & Videira, A.(1999). The 24-kDa iron–sulphur subunit of complexI is required for enzyme activity. Eur. J. Biochem. 265,86–92.

20. Duarte, M., Sousa, R. & Videira, A. (1995). Inacti-vation of genes encoding subunits of the peripheraland membrane arms of Neurospora mitochondrial

complex I and effects on enzyme assembly. Genetics,139, 1211–1221.

21. da Silva, M. V., Alves, P. C., Duarte, M., Mota, N.,Lobo-da-Cunha, A., Harkness, T. A. et al. (1996).Disruption of the nuclear gene encoding the20.8-kDa subunit of NADH: ubiquinone reductaseof Neurospora mitochondria. Mol. Gen. Genet. 252,177–183.

22. Selker, E. U. (1990). Premeiotic instability of repeatedsequences in Neurospora crassa. Annu. Rev. Genet. 24,579–613.

23. Ferreirinha, F., Duarte, M., Melo, A. M. & Videira, A.(1999). Effects of disrupting the 21 kDa subunitof complex I from Neurospora crassa. Biochem. J. 342,551–554.

24. Videira, A., Tropschug, M. & Werner, S. (1990).Primary structure and expression of a nuclear-codedsubunit of complex I homologous to proteinsspecified by the chloroplast genome. Biochem. Bio-phys. Res. Commun. 171, 1168–1174.

25. Videira, A., Azevedo, J. E., Werner, S. & Cabral, P.(1993). The 12.3 kDa subunit of complex I fromNeurospora crassa: cDNA cloning and chromosomalmapping of the gene. Biochem. J. 291, 729–732.

26. Fecke, W., Sled, V. D., Ohnishi, T. & Weiss, H. (1994).Disruption of the gene encoding the NADH-bindingsubunit of NADH:ubiquinone oxidoreductase inNeurospora crassa. Formation of a partially assembledenzyme without FMN and the iron–sulphur clusterN-3. Eur. J. Biochem. 220, 551–558.

27. Duarte, M., Mota, N., Pinto, L. & Videira, A. (1998).Inactivation of the gene coding for the 30.4-kDasubunit of respiratory chain NADH dehydrogenase:is the enzyme essential for Neurospora? Mol. Gen.Genet. 257, 368–375.

28. Melo, A. M., Duarte, M., Moller, I. M., Prokisch, H.,Dolan, P. L., Pinto, L. et al. (2001). The externalcalcium-dependent NADPH dehydrogenase fromNeurospora crassa mitochondria. J. Biol. Chem. 276,3947–3951.

29. Schagger, H. (1995). Native electrophoresis for iso-lation of mitochondrial oxidative phosphorylationprotein complexes. Methods Enzymol. 260, 190–202.

30. Alves, P. C. & Videira, A. (1994). Disruption ofthe gene coding for the 21.3-kDa subunit of theperipheral arm of complex I from Neurospora crassa.J. Biol. Chem. 269, 7777–7784.

31. Gutierres, S., Sabar, M., Lelandais, C., Chetrit, P.,Diolez, P., Degand, H. et al. (1997). Lack of mito-chondrial and nuclear-encoded subunits of complexI and alteration of the respiratory chain in Nicotianasylvestris mitochondrial deletion mutants. Proc. NatlAcad. Sci. USA, 94, 3436–3441.

32. Sabar, M., De Paepe, R. & de Kouchkovsky, Y. (2000).Complex I impairment, respiratory compensations,and photosynthetic decrease in nuclear and mito-chondrial male sterile mutants of Nicotiana sylvestris.Plant Physiol. 124, 1239–1250.

33. Affourtit, C., Albury, M. S., Crichton, P. G. & Moore,A. L. (2002). Exploring the molecular nature ofalternative oxidase regulation and catalysis. FEBSLetters, 510, 121–126.

34. Chaudhuri, M., Sharan, R. & Hill, G. C. (2002).Trypanosome alternative oxidase is regulated post-transcriptionally at the level of RNA stability.J. Eukaryot. Microbiol. 49, 263–369.

35. Umbach, A. L., Gonzalez-Meler, M. A., Sweet, C. R.& Siedow, J. N. (2002). Activation of the plantmitochondrial alternative oxidase: insights from

CI Assembly and Function in a Neurospora Mutant 289

site-directed mutagenesis. Biochim. Biophys. Acta,1554, 118–128.

36. Pilkington, S. J., Skehel, J. M., Gennis, R. B. & Walker,J. E. (1991). Relationship between mitochondrialNADH-ubiquinone reductase and a bacterial NAD-reducing hydrogenase. Biochemistry, 30, 2166–2175.

37. Videira, A. & Azevedo, J. E. (1994). Two nuclear-coded subunits of mitochondrial complex I aresimilar to different domains of a bacterial for-mate hydrogenlyase subunit. Int. J. Biochem. 26,1391–1393.

38. Videira, A., Tropschug, M., Wachter, E., Schneider, H.& Werner, S. (1990). Molecular cloning of subunits ofcomplex I from Neurospora crassa. Primary structureand in vitro expression of a 22-kDa polypeptide.J. Biol. Chem. 265, 13060–13065.

39. Sambrook, J. & Russell, D. W. (2001). MolecularCloning: a Laboratory Manual, 3rd edit., Cold SpringHarbor Laboratory Press, Cold Spring Harbor, NY.

40. Staben, C., Jensen, B., Singer, M., Pollock, J.,Schechtman, M., Kinsey, J. & Selker, E. U. (1989).Use of a bacterial hygromycin B resistance gene as adominant marker in Neurospora crassa transforma-tion. Fungal Genet. Newsletters, 36, 79–81.

41. Videira, A. & Werner, S. (1989). Assembly kineticsand identification of precursor proteins of complex Ifrom Neurospora crassa. Eur. J. Biochem. 181, 493–502.

42. Davis, R. H. & de Serres, F. J. (1970). Genetic andmicrobiological research techniques for Neurosporacrassa. Methods Enzymol. 17A, 79–143.

43. Melo, A. M. P., Roberts, T. H. & MØller, I. M. (1996).

Evidence for the presence of two rotenone-insensi-tive NAD(P)H dehydrogenases on the inner surfaceof the inner membrane of potato tuber mitochondria.Biochim. Biophys. Acta, 1276, 133–139.

44. Bradford, M. M. (1976). A rapid and sensitivemethod for the quantification of microgram quanti-ties of protein using the principle of protein–dyebinding. Anal. Biochem. 72, 248–254.

45. Zauner, R., Christner, J., Jung, G., Borchart, U.,Machleidt, W., Videira, A. & Werner, S. (1985). Identi-fication of the polypeptide encoded by the URF-1gene of Neurospora crassa mtDNA. Eur. J. Biochem.150, 447–454.

46. Towbin, H., Staechelin, T. & Gordon, J. (1979).Electrophoretic transfer of proteins from polyacryla-mide gels to nitrocellulose sheets: procedure andsome applications. Proc. Natl Acad. Sci. USA, 76,4350–4354.

47. Blake, M. S., Johnston, K. H., Russel-Jones, G. J. &Gotschlich, E. C. (1984). A rapid, sensitive methodfor detection of alkaline phosphatase-conjugatedanti-antibody on western blots. Anal. Biochem. 136,175–179.

48. Sled, V. D. & Vinogradov, A. D. (1993). Kinetics of themitochondrial NADH-ubiquinone oxidoreductaseinteraction with hexammineruthenium(III). Biochim.Biophys. Acta, 1141, 262–268.

49. Hatefi, Y. & Stiggall, D. L. (1978). Preparation andproperties of NADH:cytochrome c oxidoreductase(complex I–III). Methods Enzymol. 53, 5–10.

Edited by J. Karn

(Received 18 February 2003; received in revised form 28 March 2003; accepted 1 April 2003)

290 CI Assembly and Function in a Neurospora Mutant