Ó Springer-Verlag 2000

Mol Gen Genet (2000) 263: 527±534

ORIGINAL PAPER

T. Lisse á E. Schwarz

Functional speci®city of the mitochondrial DnaJ protein, Mdj1p, in Saccharomyces cerevisiae

Received: 8 October 1999 / Accepted: 21 January 2000

Abstract Inactivation of the gene for the mitochondrial DnaJ homolog, Mdj1p, in Saccharomyces cerevisiae results in temperature sensitivity and the loss of respiratory activity; the latter phenotype has been attributed to the loss of mitochondrial DNA. To investigate the functional speci®city of Mdj1p, non-mitochondrial DnaJ proteins were targeted to mitochondria and tested for their ability to substitute for Mdj1p. The tested DnaJ proteins were able to complement the two Mdj1p-linked phenotypes, i.e., respiratory activity and growth at 37 °C, to di€erent extents, ranging from full to very poor complementation. All DnaJ homologs ensured faithful propagation of the mitochondrial genome. N-terminal fragments of Mdj1p and Escherichia coli DnaJ comprising the well-characterized J domain partially substituted for Mdj1p. As the only hitherto known function of the N-terminal fragment is modulation of the substrate binding activity of the cognate Hsp70, we conclude that both Mdj1p-linked phenotypes ± maintenance of respiratory activity and the ability to grow at elevated temperature ± involve a mitochondrial Hsp70 partner protein. Key words Mitochondrial DnaJ á Mdj1p á Mitochondrial DNA á Complementation á Chaperone

Introduction Molecular chaperones of the Hsp70 family are ubiquitous and participate in many cellular processes, including synthesis, translocation, folding and degradation of Communicated by T. D. Fox T. Lisse á E. Schwarz (&) Institut fuÈr Biotechnologie, Martin-Luther-UniversitaÈt Halle-Wittenberg, Kurt-Mothes-Str. 3, 06120 Halle, Germany E-mail: [email protected], Tel.: + 49-345-5524856; Fax: + 49-345-5527013

proteins (for review see Hartl 1996; Rassow et al. 1997; Bukau and Horwich 1998). During these processes, Hsp70 performs a chaperone function that is controlled by a cycle of ATP binding and hydrolysis. DnaJ proteins act as co-chaperones by stimulating the ATPase activity and thus modulate the substrate binding capacity of Hsp70, but also function as Hsp70-independent chaperones (Langer et al. 1992; SchroÈder et al. 1993; Cyr 1995). DnaJ proteins, now also commonly referred to as Hsp40 proteins, share between 35±59% sequence identity. Typical DnaJ proteins have a characteristic modular structure (reviewed by Cyr 1997). The J domain, an approximately 70-amino acid segment, is the distinguishing feature of DnaJ-like proteins. In typical DnaJ proteins, the J domain lies close to the N-terminus. In many cases, the J domain is separated from the rest of the protein by a glycine-/phenylalanine-rich segment (G/F segment). This spacer region is followed by a central, cysteine-rich, domain with four repeats of the motif CxxCxGxG. This domain binds two Zn2+ atoms (Banecki et al. 1996; Szabo et al. 1996). Usually, the Cterminal region, with roughly 200 amino acid residues, is the least conserved part within this protein family. With the study of truncated derivatives of Escherichia coli DnaJ the structure-function analysis of this member of the Hsp40 family began. An N-terminal fragment consisting of the J domain and the G/F segment is suf®cient for the interaction with, and stimulation of the ATPase activity of DnaK (Wall et al. 1994, 1995; Szabo et al. 1996). Though not absolutely required for the interaction with DnaK, the adjacent G/F segment appears to activate the substrate binding properties of DnaK. Investigations of deletion constructs, analysis of point mutations in the J domain and NMR-based structural data for the N-terminal fragment (2±108) allowed the identi®cation of residues which very probably contact DnaK (Wall et al. 1994; Pellecchia et al. 1996). In contrast to the well characterized N-terminal portion, a detailed understanding of the function of the following segments is still lacking, though initial studies have been performed (Banecki et al. 1996; Szabo et al. 1996).

528

What determines the functional speci®city of a DnaJ protein? The genome of the yeast Saccharomyces cerevisiae encodes at least 18 DnaJ proteins (Zuber et al. 1998), a number which equals that of the known Hsp70 homologs in yeast. It is assumed that each member of the Hsp70 family has its cognate Hsp40 protein. Speci®c structures in DnaJ proteins may determine both their choice of a particular Hsp70 partner and their mode of action (Schlenstedt et al. 1995). To learn more about the functional speci®city of DnaJ proteins we asked whether the mitochondrial DnaJ, Mdj1p, could be functionally replaced by other DnaJ proteins and, if so, whether, as in the case of E. coli DnaJ, truncated versions of DnaJ proteins would suce to complement the lack of Mdj1p. Here, we present data which show that the two Mdj1p-related phenotypes, growth at 37 °C and respiratory activity, can be mediated by a number of DnaJ proteins. However, full rescue of temperature sensitivity is not in all cases linked to restoration of respiratory activity. Interestingly, the J domains and the adjacent G/F segment of both Mdj1p and E. coli DnaJ are sucient to ensure the maintenance of mitochondrial DNA.

Materials and methods

primers regf (5¢-CGAATATTCCCGGGCTACCTGCGTCAGACATTG-3¢) and regr (5¢-GCCATAAGCTTGTTATGTTTAGAGTATGATATAGG-3¢). The resulting PCR product was digested with SmaI and HindIII. The resident GAL1 promoter and part of the adjacent F1 origin were deleted from pYES2.0 by cleavage with SspI and HindIII, and replaced by the PCRgenerated MDJ1 promoter, from which the BamHI site had previously been deleted. Transfer of the genes for the DnaJ proteins from pGEMSu9 into pYESM was carried out in two ways. The genes MDJ1 and SCJ1 were recloned as fusions with the presequence using the restriction enzymes HindIII and SacI. The genes DNAJ, SIS1, XDJ1, YDJ1 were excised from pGEM3Su9 with BamHI and EcoRI, and inserted into pYESM which had been modi®ed by the insertion of the coding fragment for the Su9 targeting sequence by cloning with HindIII/BamHI. The resulting plasmids were designated pYESMDJ1, pYES-DNAJ, pYES-SCJ1, pYES-SIS1, pYES-XDJ1 and pYES-YDJ1 for Mdj1p, DnaJ, Scj1p, Sis1p, Xdj1p and Ydj1p, respectively. For the construction of deletion derivatives encompassing the N-terminal segments of Mdj1p (amino acids 56±189) and DnaJ (amino acids 1±108) two stop codons were introduced into the expression plasmids described above for Mdj1p and DnaJ by sitedirected mutagenesis using the QuikChange kit. The primers 189f (5¢-CGGTGGCGCAAGCAGATAGTAATCTATGTTCAGAC3¢) and 189r (5¢-GTCTGAACATAGATTACTATCTGCTTGCGCCACCG-3¢) for Mdj189p and 108f (5¢-GGCGGACGTGGTCGTTAATAGGCGGCGCGCGGTGCTG-3¢) and 108r (5¢-CAGCACCGCGCGCCGCCTATTAACGACCACGTCCGCC-3¢) for DnaJ108p were used for ampli®cation of the deletion constructs.

Recombinant DNA techniques and plasmid construction

Construction of yeast strains

Genomic DNA from S. cerevisiae strain D237-10B (ATCC 25657) and E. coli strain XL1-blue (Stratagene) served as a source for DNA ampli®cation, if not indicated otherwise. To facilitate subsequent cloning of the coding sequences for the di€erent DnaJ proteins, primers were designed in such a way that the resulting PCR fragment contained a BamHI or BglII site at the 5¢-end and a EcoRI or SacI site at the 3¢-end. In the case of the gene for Scj1p, the coding sequence of the 44-amino acid ER sorting signal was replaced by the mitochondrial targeting sequence. The following primer pairs were used in the PCRs: for DNAJ, dnajf (5¢-CGCGGATCCATGGCTAAGCAAGAT-3¢) and dnajr (5¢-CGAATTCCTATTAGCGGGTCAGGTC-3¢); for MDJ1, mdjf (5¢-GAAGATCTAACGAAGCATTCAAG-3¢) and mdjr (5¢-GCGAGCTCTATTAATTTTTTTTGTCACC-3¢); for SCJ1, scjf (5¢-CGCGGATCCCTAATTTTGGCGCAG-3¢) and scjr (5¢-GCGAGCTCTACTACAACTCATCTTTG-3¢); for SIS1, sisf (5¢-CGCGGATCCATGGTCAAGGAGACA-3¢) and sisr (5¢-GCGAGCTCTATTAAAAATTTTCAT C-3¢); for XDJ1, xdjf (5¢-CGCGGATCCATGAGTGGCAGTGAT-3¢) and xdjr (5¢-GCGAGCTCATCATTGGATACAGCAG-3¢); for YDJ1, ydjf (5¢-CGCGGATCCATGGTTAAAGAAAC-3¢) and ydjr (5¢-GCTCATCATTGAGATGCGCATTGAACACCTTCG-3¢). The PCR products were digested with BamHI/EcoRI, BglII/SacI or BamHI/SacI, and inserted into the vector pGEM3Su9 linearized with BamHI/EcoRI or BamHI/SacI. pGEMSu9 (a gift from B. Westermann, unpublished data) was derived from pGEM3 by insertion of the coding sequence for the mitochondrial matrix targeting signal of ATPase subunit 9 from Neurospora crassa between the HindIII and BamHI sites in the polylinker. Thus, by inserting the PCR products into pGEM3Su9 cleaved with BamHI and either EcoRI or SacI, the genes for the DnaJ proteins were fused in frame to a mitochondrial matrix targeting sequence. The authenticity of all constructs was con®rmed by DNA sequence analysis. pYESM, a derivative of pYES2.0 (Invitrogen), was constructed to express all DnaJ fusions from the endogenous MDJ1 promoter. The MDJ1 promoter was generated in a PCR using the plasmid pMDJ315 (Rowley et al. 1994) as the template, together with the

Standard genetic techniques were used for growth and manipulation of yeast (Guthrie and Fink 1991). Transformation of yeast was carried out as described previously (Ausubel et al. 1995). As a recipient for the recombinant DnaJ proteins S. cerevisiae strain YTL1 was constructed from the haploid strain YBW16a (Westermann et al. 1996). YBW16a carries an inactivated chromosomal MDJ1 allele and possesses an extrachromosomal MDJ1 allele on plasmid pBWM11 [MDJ1, URA3, 2 l]. For selection marker exchange, YBW16a was transformed with pMDJ315 [MDJ1, LEU2, CEN]. From these transformants a clone, YTL1, was selected that had lost pBWM11. After transformation of YTL1 with the expression plasmids for the various DnaJ proteins, selection against pMDJ315 gave rise to yeast cells which either expressed one of the heterologous DnaJ proteins or Mdj1p. The isogenic null mutant, Dmdj1, was obtained by isolating an untransformed YTL1 clone which had lost pMDJ315. In vivo labeling of mitochondrial translation products Mitochondrial translation products were labeled according to Herrmann et al. (1994). In brief, 10 ml of YNB medium containing 3% galactose were inoculated from a preculture to an OD600 value of 0.1. Cells were grown overnight at 24 °C, harvested and washed with water. Before labeling, the cell suspension was adjusted to OD600 2.5. An aliquot (250 ll) containing 150 lg/ml cycloheximide was incubated in the presence of an 8-ll mixture containing all amino acids (2 mg/ml each), except methionine, and 20 lCi of [35S]methionine (10 mCi/ml, Amersham) for 15 min at 24 °C. Labeling was stopped by addition of 10 ll of stop mixture (0.1 M methionine, 13 mg/ml chloramphenicol) and the reaction was further incubated for 15 min. Then cells were lysed by adding 50 ll of lysis solution (2 M NaOH, 14% b-mercaptoethanol), and total cell protein was precipitated with TCA, washed with acetone and solubilized in SDS sample bu€er. Translation products were fractionated by SDS-PAGE, blotted onto nitrocellulose and visualized by autoradiography.

529 Tetrazolium overlay Mutants were grown to mid-log phase at 24 °C in liquid YNB medium containing 3 % glucose and 1% casamino acids. Aliquots containing 100±200 colony forming units were plated on YPD medium. After 3 days, cells were overlaid with 0.5% agar in 0.067 M phosphate bu€er pH 7.0, containing 0.1% 2,3,5-triphenyltetrazolium chloride (TTC). Respiring colonies were detected on the basis of their red color. Miscellaneous DNA techniques were carried out as described by Ausubel et al. (1995). SDS-PAGE, semi-dry immunoblotting and detection with the ECL detection kit (Amersham) was performed according to standard protocols and following the manufacturers' guidelines. Isolation of mitochondria was done as described previously (Daum et al. 1982), cultivation of cells was performed under selective growth conditions on YNB medium containing 3% glucose and 1% casamino acids.

Results and discussion Several heterologous DnaJ proteins complement the Dmdj1 phenotype, when imported into mitochondria Inactivation of the gene for Mdj1p leads to temperature sensitive growth and the absence of respiratory activity due to the loss of mitochondrial DNA. To determine whether growth at elevated temperature and maintenance of the organellar genome requires Mdj1p function speci®cally or can also be conferred by other DnaJ proteins, we tested several DnaJ proteins for their ability to substitute for Mdj1p. Besides DnaJ from E. coli, the following DnaJ proteins from S. cerevisiae were analyzed for complementation activity: Scj1p, a DnaJ protein which resides in the endoplasmic reticulum (ER) and functions together with Kar2p (Schlenstedt et al. 1995); Sis1p, a cytosolic DnaJ homolog which ful®lls an important role during translation (Zhong and Arndt 1993); Ydj1p, which in the cytosol of yeast mediates protein translocation across intracellular membranes and regulates protein turnover (Caplan et al. 1992; Lee et al. 1996); and Xdj1p, a DnaJ protein of hitherto unknown function and location (Schwarz et al. 1994). In order to direct the DnaJ proteins to the mitochondrial matrix space, they were expressed as preproteins with the mitochondrial targeting sequence of ATPase subunit 9 fused as a prepiece to their N-termini (Fig. 1). To control for the functionality of the subunit 9 presequence in the context of the DnaJ sequence, DNA encoding mature Mdj1p was fused to the heterologous mitochondrial sorting peptide (Fig. 1). As a recipient for the recombinant DnaJ proteins, S. cerevisiae strain YTL1 was constructed. YTL1 carries an inactivated chromosomal MDJ1 allele and an extrachromosomal MDJ1 allele on the plasmid pMDJ315 [MDJ1, LEU2, CEN] (Rowley et al. 1994). After transformation of YTL1 with the expression plasmids for the various DnaJ proteins, selection against pMDJ315 gave rise to yeast

Fig. 1 Schematic representation of the structural features of the various DnaJ proteins tested for complementation activity

cells which expressed one of the heterologous DnaJ proteins or Mdj1p. The isogenic mdj1 null mutant, Dmdj1, was obtained by isolating an untransformed YTL1 clone which had lost pMDJ315. Western analysis was performed to con®rm the mitochondrial location of the heterologous DnaJ proteins. To this end, protein extracts of isolated mitochondria were fractionated by SDS-PAGE, blotted and analyzed by immunolabeling with antisera directed against the speci®c DnaJ proteins. DnaJ proteins for which antisera were available (DnaJ, Mdj1p, Scj1p, Ydj1p) could be

Fig. 2 Immunological detection of the heterologous DnaJ proteins in the mitochondrial protein fraction. Isolation of mitochondria was done as described. Mitochondrial proteins were fractionated by SDSPAGE. Mitochondrial extracts from cells containing Mdj1p (lane 1), DnaJ (lane 2), Scj1p (lane 3), Ydj1p (lane 4), Mdj189p (lane 5) and DnaJ108p (lane 6) targeted to mitochondria were reacted with the indicated antisera. As a control, the presence of the mitochondrial GrpE homolog Mge1p was assayed

530

detected in the mitochondrial protein fraction (Fig. 2). Mdj1p, DnaJ and Ydj1p migrated at the positions expected for the correctly processed mature proteins. The signal for Scj1p was observed at an apparent molecular weight of 45 kDa, indicating that maturation of this protein by the mitochondrial processing peptidase was somehow inhibited. With antiserum against Ydj1p we observed two signals, probably corresponding to both the farnesylated and non-farnesylated forms of Ydj1p. Non-farnesylated Ydj1p has a lower electrophoretic mobility (Caplan et al. 1992). Next, the clones were tested for both Mdj1p-linked phenotypes, respiratory activity and growth at elevated temperature. The cells were plated onto medium containing the fermentable carbon source, glucose (YPD), or the non-fermentable carbon source, glycerol (YPG), and subsequently incubated at 24 °C or 37 °C. As expected, the mdj1 null mutant, Dmdj1, was only able to grow at 24 °C on YPD medium. There was no growth at 37 °C or on non-fermentable carbon sources (Fig. 3). In contrast, cells with Mdj1p grew well under all conditions tested (Fig. 3). The heterologous DnaJ proteins possessed di€erent complementation capacities. Cells containing E. coli DnaJ targeted to mitochondria behaved like the wild-type. Cells harboring Xdj1p or Ydj1p in their mitochondria grew well both on fermentable and non-fermentable carbon sources at 24 °C. However, at 37 °C the ability of Xdj1p and Ydj1p to restore respiratory activity was strongly reduced, whilst growth on fermentable YPD medium was still maintained. Very poor growth on YPG medium was observed with Scj1p and Sis1p at 24 °C. In fact, after transfer to fresh YPG plates, these cells formed microcolonies which only became visible after more than 8 days of incubation at 24 °C. Obviously, respiratory activity is drastically impaired in these cells. At elevated temperature, growth on

Fig. 3 Phenotype of cells expressing heterologous DnaJ proteins targeted to the mitochondria. Cells were grown overnight at 24 °C in liquid YNB medium containing 3% glucose and 1% casamino acids. Ten-fold serial dilutions of a cell suspension (OD600 ˆ 0.1) were spotted onto yeast extract/peptone plates containing either 2% glucose (YPD) or 3% glycerol (YPG). YPD plates were incubated for 2 or 3 days at 37 or 24 °C, respectively; YPG plates were kept for 3 and 5 days at 37 and 24 °C, respectively

YPG was completely compromised. Furthermore, Scj1p was unable to confer growth on YPD medium at elevated temperature. It is important to note that Sis1p gave rise to a differential complementation phenotype. Cells containing mitochondrial Sis1p showed negligible growth on nonfermentable medium but grew well at 37 °C on YPD medium (Fig. 3). Two possibilities ± which are not mutually exclusive ± could explain di€erential complementation by Sis1p. The processes underlying the two phenotypes could require (i) di€erent levels of an active DnaJ protein, or (ii) di€erent DnaJ protein functions, e.g. as a co-chaperone or as a chaperone on its own right, for speci®c substrate proteins. In the context of the latter explanation it is noteworthy that Sis1p is the only DnaJ protein tested which lacks the cysteine motif. In E. coli DnaJ, this motif has been postulated to participate in the interaction with unfolded substrate proteins (Szabo et al. 1996). Taken together, the complementation experiments show that the two Mdj1p-dependent phenotypes, i.e., growth at 37 °C and respiratory activity, can be supplied by other DnaJ proteins. However, complementation of growth at 37 °C is not in all cases linked to full complementation of respiratory activity. Furthermore, depending on the heterologous DnaJ protein present, cells exhibited all gradations from nearly wild-type to the null mutant phenotype. Given the di€erent capacity of the heterologous DnaJ proteins to support respiratory growth we asked whether the rate of petite induction was a€ected by the foreign DnaJ proteins. Strains were cultivated on selective glucose-containing medium to ensure the presence of the plasmid encoding the respective DnaJ protein. To determine the fraction of respiration-de®cient colonies, cells were plated on YPD medium, and after 3 days colonies were overlaid with TTC-containing top agar (Ogur et al. 1957) (Fig. 4). In strains in which Mdj1p

Fig. 4 Petite induction. The fraction of petites (non-respiring cells) in the population was assessed after tetrazolium overlay as described in Materials and methods. The error bars represent standard deviations from the mean. n indicates the number of plates examined

531

was replaced by E. coli DnaJ, similar rates of petite induction to those seen with endogenous Mdj1p were observed. With Xdj1p and Ydj1p, the spontaneous rate of appearance of petites was slightly increased. Notably, up to three times more respiration-de®cient colonies relative to wild type arose with Scj1p and Sis1p. These observations con®rm the results of the growth test on non-fermentable carbon sources at 24 °C described above. Taken together, both sets of experiments indicate that DnaJ can best supply the Mdj1p-related functions. Xdj1p and Ydj1p exhibited limited complementation capacities whereas Sis1p and Scj1p can substitute for Mdj1p only very poorly. Why do certain DnaJ proteins complement Mdj1p so well whereas others only very poorly compensate for the lack of the endogenous DnaJ protein? One possible answer is that the concentrations of active protein di€er, due to variations in the eciency of import into the matrix, or in folding and/or in structural stability in the foreign cellular compartment. The possibility that the concentrations of active protein are too low appears unlikely, as extremely low concentrations of Mdj1p are sucient to confer the wild-type phenotype (B. Westermann unpublished results). Nevertheless, inecient folding to the native state could provide an explanation for the poor complementation activity of Scj1p. Apparently, when imported into mitochondria, Scj1p was not matured correctly by the mitochondrial processing enzymes. The mitochondrial import signal present in Scj1p might interfere with the folding and/or activity of the protein. Additionally, the possibility cannot be excluded that Scj1p, whose home compartment is the oxidizing ER, contains disul®de bonds in its native conformation, which are not formed in the reducing mitochondrial milieu. An alternative reason why certain DnaJ proteins ineciently substituted for Mdj1p, could be impaired interaction with mitochondrial Hsp70, since, for example, the speci®city of the Scj1p-Kar2p system localized in the ER is determined by a few essential, conserved residues in the J domain (Schlenstedt et al. 1995). A last formal possibility would be that certain target proteins required for growth at elevated temperature and for mitochondrial genome maintenance stringently depend on Mdj1p function. Recognition of these speci®c proteins by Mdj1p could rely on de®ned structures present in Mdj1p or E. coli DnaJ, but absent in other DnaJ proteins. As detailed below, this explanation is the least likely, in view of the complementing activity displayed by the N-terminal domains of Mdj1p and E. coli DnaJ. An N-terminal fragment of the DnaJ protein can phenotypically complement the Dmdj1 mutant The ®rst 108 amino acids of E. coli DnaJ, comprising the J domain and the G/F segment (subsequently referred to as N-terminal fragment), are sucient in vivo for the propagation of bacteriophage k and in vitro for the

stimulation of the ATPase activity of DnaK (Wall et al. 1994; Szabo et al. 1996). To analyze whether the same fragment of E. coli DnaJ or the corresponding domain of Mdj1p would suce to confer the wild-type phenotype, deletion derivatives of both proteins (DnaJ108p and Mdj189p) were generated. Strain constructions were carried out as described for the full-length DnaJ proteins (Fig. 1). Both N-terminal fragments conferred the ability to grow at elevated temperature in YPD medium, although DnaJ108p did not complement as well as Mdj189p (Fig. 3). Mdj189p conferred full respiratory activity at 24 °C and partially complemented at 37 °C. Interestingly, even DnaJ108p restored growth on YPG at 24 °C, albeit poorly. Also, the rate of spontaneous induction of the petite phenotype at 24 °C was comparable to that observed with Xdj1p and Ydj1p (Fig. 4). The only function of the N-terminal fragment from E. coli DnaJ known so far is interaction with DnaK (Hsp70). On its own, this domain cannot prevent aggregation of a denatured substrate protein, an observation which was taken as evidence that this part of DnaJ lacks chaperone activity (Szabo et al. 1996). Since our data clearly demonstrate complementation by the Nterminal fragment of Mdj1p we conclude that it is not the intrinsic chaperone activity of a particular DnaJ protein which is required for growth at elevated temperature and maintenance of respiratory activity. Rather, our studies indicate that both Mdj1p-linked phenotypes involve an Hsp70 partner protein. Whether Mdj1p interacts during these processes with Ssc1p, the major mitochondrial Hsp70 (Kang et al. 1990) and/or with Ssh1p, another mitochondrial Hsp70, which has been proposed to participate in mitochondrial genome maintenance (Schilke et al. 1996), is as yet unclear and will have to be clari®ed by further experiments. Interestingly, three recent publications demonstrate that the N-terminal fragment of the SV40 T/t antigen, which has homology to J domains, suces for bacteriophage k DNA replication and replication of SV40 viral DNA (Campbell et al. 1997; Kelley and Georgopoulos 1997; Srinivasan et al. 1997). These data support our ®ndings that the N-terminal domains of both DnaJ and Mdj1p can ensure the replication of the mitochondrial genome, a prerequisite for respiratory activity. Mitochondrial DNA is not depleted in strains with impaired respiratory activity Cells containing mitochondrial Scj1p and Sis1p grew extremely poorly on YPG. To test whether this de®ciency in respiratory activity might be due to instability of the mitochondrial genome, the presence of mitochondrial DNA was assayed by Southern analysis of digested total cellular DNA. Hybridization was carried out with ori5 DNA and MGE1 DNA, as mitochondrial and nuclear probes, respectively. Cells with mitochondrial Scj1p and Sis1p had levels of mitochondrial DNA comparable to those of wild-type cells (data not shown).

532

To investigate whether the de®ciency in respiratory activity displayed by strains with mitochondrial Scj1p and Sis1p could be due to mutation(s) in the mitochondrial genome, the cells were transformed with plasmid PHDJ315. Transformation of the MDJ1 allele restored full respiratory competence (data not shown). Thus, cells containing mitochondrial Scj1p and Sis1p maintained a functional organellar genome though they were severely inhibited in respiratory metabolism. Synthesis of mitochondrially encoded proteins is a€ected by foreign DnaJ proteins To address the question whether the di€erent phenotypes observed for yeast strains containing various foreign DnaJ proteins might be due to altered synthesis of mitochondrially encoded proteins, mitochondrial protein synthesis was monitored using an in vivo translation assay (Douglas and Butow 1976; Herrmann et al. 1994). Translation of mitochondrially encoded proteins was monitored by labeling with [35S]methionine. Labeled products, fractionated by SDS-PAGE, were blotted onto nitrocellulose membranes and visualized by autoradiography. The translation products were identi®ed by comparing our results to those of Herrmann et al. (1994) and con®rmed by immunodetection of CoxII (Fig. 5B). Immunodetection with antiserum against Ssc1p and quanti®cation of the signals demonstrated that comparable amounts of material were analyzed (Fig. 5A, lower panel). Translational activity was observed in all strains (Fig. 5A, upper panel). The ribosome-associated protein Var1 was synthesized at comparable levels in all strains. In contrast, di€erences were observed with regard to the translation products corresponding to subunits of the respiratory complex, CoxI-III, Cytb, ATPase 6, 8 and 9. Here, an overall reduction of translational activity was detected in extracts derived from cells expressing heterologous DnaJ proteins; this was most obvious in the case of mitochondrial Scj1p and Sis1p. Besides this general decrease in translation, the relative amounts of labeled polypeptides di€ered quantitatively between extracts derived from wild-type cells and those containing heterologous DnaJ proteins. Whereas the translation patterns of transformants with foreign DnaJ proteins looked similar, they di€ered from that of the Mdj1p-containing strain. In particular, CoxI and CoxII were underrepresented in extracts with foreign DnaJ proteins, while labeling of Cytb and ATPase 6 was increased relative to wild-type. It is known that multiple trans-acting nuclear genes are required for speci®c expression of COX genes encoded by the mitochondrial DNA (KloÈckner-Gruissem et al. 1987). Di€erent proteins are involved in ensuring mRNA maturation and stability of the intron-containing transcripts of COXI and CYTB (Arlt et al. 1998; van Dyck et al. 1998). Whether the altered levels of CoxI, CoxII, Cytb and ATPase 6 originate from aberrant transcriptional and/or (post-)translational processes will have to be elucidated by further studies. It

Fig. 5A, B Mitochondrial gene products. A Synthesis of mitochondrially encoded proteins in vivo. Upper panel Mitochondrial translation products were labeled with [35S]methionine as described in Materials and methods, fractionated by SDS-PAGE, blotted onto nitrocellulose and analyzed by autoradiography. CoxI, CoxII and CoxIII, correspond to subunits I, II and III of the cytochrome oxidase complex, respectively; Cytb, cytochrome b; Atp6, 8 and 9, subunits 6, 8 and 9 of the F0-ATPase. Lower panel Immunoblot analysis of the same membrane using antiserum against Ssc1p (Ssc1p is highlighted by the arrow). Using the program Phoretix 1D Quanti®er, the signal intensities for Ssc1p were quanti®ed, and are expressed as percentages relative to the signal obtained with extracts from cells containing Mdj1p, which was set to 100%. B Immunological identi®cation of CoxII. Nitrocellulose membrane from an in vivo translation experiment (left panel) was immunolabeled with monoclonal antibodies against CoxII (right panel). The positions of molecular mass markers (kDa) are indicated. The faint band above the CoxII signal probably results from the CoxII precursor

also remains to be clari®ed whether the requirement for a speci®c DnaJ function re¯ects a direct involvement in a process underlying the biogenesis of mitochondrially encoded proteins or is rather the consequence of the general ability of DnaJ proteins to assist protein folding. Next, we analyzed whether the steady-state concentrations of mitochondrial translation products might di€er depending on the DnaJ protein present. To this

533

Fig. 6 Steady-state levels of CoxII. Cells were lysed (as described by Ya€e and Schatz 1984), total protein was fractionated by SDSPAGE, blotted onto nitrocellulose membrane and probed with antiserum against CoxII. Labeling with antiserum against Mge1p proves that comparable amounts of cell extracts had been loaded in all the lanes. Protein extracts were loaded as follows: lane 1, Mdj1p wild-type (W303); lane 2, deletion mutant Dmdj1; lane 3-11, complementation mutants containing pmMdj1p, psMdj1p, psDnaJ, psScj1p, psSis1p, psXdj1p, psYdj1p, psMdj189p, psDnaJ108p, (for nomenclature compare Fig. 1)

end, total cell protein was subjected to SDS-PAGE, blotted and probed with antibodies against CoxII (Fig. 6). The steady-state concentrations of CoxII correlated well with the expression rates. The highest level of CoxII was observed in cell extracts from cells harboring Mdj1p. Extracts from all strains with heterologous DnaJ proteins showed reduced amounts of CoxII. Even with cells containing mitochondrial DnaJ or the J domain of Mdj1p which exhibited growth characteristics comparable to wild-type a decreased signal for CoxII was observed. In agreement with the poor ability of Xdj1p and Ydj1p to complement the respiration-negative phenotype of Dmdj1 mutant the, CoxII levels were severely reduced in those cell extracts. In extracts of cells with mitochondrial Scj1p and Sis1p the amount of CoxII was below the level required for immunodetection. In this report we have demonstrated that several nonmitochondrial DnaJ proteins are able to ensure mitochondrial genome maintenance, and presented evidence that the complete mitochondrial Hsp70 system may be required for DNA propagation. Though the molecular role of the Hsp70 system during propagation of the organellar genome remains to be analyzed in detail, Marszalek and co-workers have shown that Mdj1p plays a dual role in the inheritance of mitochondrial DNA: while at elevated temperatures Mdj1p is required for the activity of mitochondrial DNA polymerase, a second, as yet unidenti®ed, function in genome propagation has been assigned to Mdj1p at physiological temperature (Duchniewicz et al. 1999). The loss of respiratory activity in the Dmdj1 strain has hitherto been attributed to the absence of organellar DNA. The data presented here now re®ne our view about the contribution of DnaJ proteins to respiratory competence. While Scj1p and Sis1p ensured the maintenance of an intact mitochondrial genome, they were to a large extent incapable of supporting respiratory growth. From

this result we conclude that, apart from DNA maintenance, a speci®c DnaJ protein function is required for full respiratory activity. Whether this function is restricted to the role of the mitochondrial Hsp70 system in the biogenesis of mitochondrially encoded proteins (Herrmann et al. 1994; Westermann et al. 1996) is so far unclear. Apparently, the recently characterized second mitochondrial DnaJ protein, Mdj2p, is not sucient for these tasks (Westermann and Neupert 1997). The identi®cation and elucidation of the speci®c molecular mechanisms which stringently require Mdj1p during respiratory metabolism are the focus of our ongoing investigations. Acknowledgements We thank Drs. Roland Lill, Douglas Cyr and Jaroslaw Marszalek for providing antibodies against Scj1p, Ydj1p and DnaJ and Dr. Benedikt Westermann for providing ori5 DNA. We thank Drs. Walter Neupert and Benedikt Westermann for critically reading the manuscript and Dr. Elmar Wahle for stimulating discussions. We gratefully acknowledge the support of Dr. Rainer Rudolph, who provided optimal working conditions. This work was supported by a grant (SCHW375/2±1) to E.S. from the Deutsche Forschungsgemeinschaft (DFG).

References Arlt H, Steglich G, Perryman R, Guiard B, Neupert W, Langer T (1998) The formation of respiratory chain complex in mitochondria is under the proteolytic control of the m-AAA protease. EMBO J 17: 4837±4847 Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (1995) Current protocols in molecular biology. Wiley, New York Banecki B, Liberek K, Wall D, Wawrzynow A, Georgopoulos C, Bertoli E, Tanfani F, Zylicz M (1996) Structure-function analysis of the zinc ®nger region of the DnaJ molecular chaperone. J Biol Chem 25: 14840±14848 Bukau B, Horwich AL (1998) The Hsp70 and Hsp60 chaperone machines. Cell 92: 351±366 Campbell KS, Mullane KP, Aksoy IA, Stubdal H, Zalvide J, Pipas JM, Silver PA, Roberts TM, Scha€hausen BS, DeCaprio JA (1997) DnaJ/hsp40 chaperone domain of SV40 large T antigen promotes ecient viral DNA replication. Genes Dev 11: 1098±1110 Caplan AJ, Cyr DM, Douglas MG (1992) Ydj1p facilitates polypeptide translocation across di€erent intracellular membranes by a conserved mechanism. Cell 71: 1143±1155 Cyr DM (1995) Cooperation of the molecular chaperone Ydj1p with speci®c Hsp70 homologs to suppress protein aggregation. FEBS Lett 359: 129±132

534 Cyr DM (1997) In: Gething MJ (ed) Molecular chaperones and protein-folding catalysts. Oxford University Press, Oxford, pp 89±95 Daum G, BoÈhni PC, Schatz G (1982) Import of proteins into mitochondria. J Biol Chem 257: 13028±13033 Douglas MG, Butow RA (1976) Variant forms of mitochondrial translation products in yeast: evidence for location of determinants on mitochondrial DNA. Proc Natl Acad Sci USA 73: 1083±1086 Duchniewicz M, Germaniuk A, Westermann B, Neupert W, Schwarz E, Marszalek J (1999) Dual role of the mitochondrial chaperone Mdj1p in inheritance of mitochondrial DNA in yeast. Mol Cell Biol 19: 8201±8210 Guthrie C, Fink GR (1991) Guide to yeast genetics and molecular biology. Academic Press, San Diego Hartl FU (1996) Molecular chaperones in cellular protein folding. Nature 381: 571±580 Herrmann JM, Stuart RA, Craig EA, Neupert W (1994) Mitochondrial heat shock protein 70, a molecular chaperone for proteins encoded by mitochondrial DNA. J Cell Biol 127: 893±902 Kang PJ, Ostermann J, Shilling J, Neupert W, Craig EA, Pfanner N (1990) Requirement for hsp70 in the mitochondrial matrix for translocation and folding of precursor proteins. Nature 348: 137±143 Kelley WL, Georgopoulos C (1997) The T/t common exon of simian virus 40, JC, and BK polyomavirus T antigens can functionally replace the J-domain of the Escherichia coli DnaJ molecular chaperone. Proc Natl Acad Sci USA 94: 3679±3684 KloÈckner-Gruissem B, McEwen JE, Poyton RO (1987) Nuclear functions required for cytochrome c oxidase biogenesis in Saccharomyces cerevisiae: multiple trans-acting nuclear genes exert speci®c e€ects on expression of each of the cytochrome c oxidase subunits encoded on mitochondrial DNA. Curr Genet 12: 311±322 Langer T, Lu C, Echols H, Flanagan J, Hayer M K, Hartl FU (1992) Successive action of DnaK, DnaJ and GroEL along the pathway of chaperone-mediated protein folding. Nature 356: 683±689 Lee DH, Sherman MY, Goldberg AL (1996) Involvement of the molecular chaperone Ydj1p in the ubiquitin-dependent degradation of short-lived and abnormal proteins in Saccharomyces cerevisiae. Mol Cell Biol 16: 4773±4781 Ogur M, St John R, Nagai S (1957) Tetrazolium overlay technique for population studies of respiration de®ciency in yeast. Science 125: 928±929 Pellecchia M, Szyperski T, Wall D, Georgopoulos C, WuÈthrich K (1996) NMR structure of the J-domain and the gly/phe-rich region of the Escherichia coli DnaJ protein. J Mol Biol 260: 236±250 Rassow J, van Ahsen O, BoÈmer U, Pfanner N (1997) Molecular chaperones: towards a characterization of the heat shock protein 70 family. Trends Cell Biol 7: 129±133

Rowley N, Prip-Buus C, Westermann B, Brown C, Schwarz E, Barrell B, Neupert W (1994) Mdj1p, a novel chaperone of the DnaJ family, is involved in mitochondrial biogenesis and protein folding. Cell 77: 249±259 Schilke B, Forster J, Davis J, James P, Walter W, Laloraya S, Johnson J, Miao B, Craig E (1996) The cold sensitivity of a mutant of Saccharomyces cerevisiae lacking a mitochondrial heat shock protein 70 is suppressed by loss of mitochondrial DNA. J Cell Biol 134: 603±613 Schlenstedt G, Harris S, Risse B, Lill R, Silver PA (1995) A yeast DnaJ homologue, Scj1p, can function in the endoplasmic reticulum with BiP/Kar2p via a conserved domain that speci®es interactions with Hsp70s. J Cell Biol 129: 979±988 SchroÈder H, Langer T, Hartl FU, Bukau B (1993) DnaK, DnaJ and GrpE form a cellular chaperone machinery capable of repairing heat-induced protein damage. EMBO J 12: 4137±4144 Schwarz E, Westermann B, Caplan AJ, Ludwig G, Neupert W (1994) XDJ1, a gene encoding a novel non-essential DnaJ homologue from Saccharomyces cerevisiae. Gene 145: 121±124 Srinivasan A, McClellan AJ, Vartikar J, Marks I, Cantalupo P, Li Y, Whyte P, Rundell K, Brodsky J, Pipas JM (1997) The amino-terminal transforming region of simian virus 40 large T and small t antigens functions as a J domain. Mol Cell Biol 17: 4761±4773 Szabo A, Korszun R, Hartl FU, Flanagan J (1996) A zinc ®ngerlike domain of the molecular chaperone DnaJ is involved in binding to denatured protein substrates. EMBO J 15: 408±417 Van Dyck L, Neupert W, Langer T (1998) The ATP-dependent PIM1 protease is required for the expression of introncontaining genes in mitochondria. Genes Dev 12: 1515±1524 Wall D, Zylicz M, Georgopoulos C (1994) The NH2-terminal 108 amino acids of the Escherichia coli DnaJ protein stimulate the ATPase activity of DnaK and are sucient for k replication. J Biol Chem 269: 5446±5451 Wall D, Zylicz M, Georgopoulos C (1995) The conserved G/F motif of the DnaJ chaperone is necessary for the activation of the substrate binding properties of the DnaK chaperone. J Biol Chem 270: 2139±2144 Westermann B, Neupert W (1997) Mdj2p, a novel DnaJ homolog in the mitochondrial inner membrane of the yeast Saccharomyces cerevisiae. J Mol Biol 272: 477±483 Westermann B, Gaume B, Herrmann JM, Neupert W, Schwarz E (1996) Role of the mitochondrial DnaJ homolog Mdj1p as a chaperone for mitochondrially synthesized and imported proteins. Mol Cell Biol 16: 7063±7071 Ya€e MP, Schatz G (1984) Two nuclear mutations that block mitochondrial protein import in yeast. Proc Natl Acad Sci USA 81: 4819±4823 Zhong T, Arndt KT (1993) The yeast SIS1 protein, a DnaJ homolog, is required for the inititiation of translation. Cell 73: 1175±1186 Zuber U, Buchberger A, Laufen T, Bukau B (1998) In: Fink AL, Goto Y (eds) Molecular chaperones in the life cycle of proteins. Marcel Dekker, New York, pp 241±274