Mutational Bisection of the Mitochondrial DNA Stability and Amino Acid Biosynthetic Functions of Ilv5p of Budding Yeast

Corresponding author: Ronald A. Butow, Dallas, TX 75390-9148., butow{at}swmed.edu (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Ilv5p is a bifunctional yeast mitochondrial enzyme required for branched chain amino acid biosynthesis and for the stability of mitochondrial DNA (mtDNA) and its parsing into nucleoids. The latter occurs when the general amino acid control (GAC) pathway is activated. We have isolated ilv5 mutants that lack either the enzymatic (a^-D⁺) or the mtDNA stability function (a⁺D^-) of the protein. The affected residues in these two mutant classes cluster differently when mapped to the 3-D structure of the spinach ortholog of Ilv5p. a^-D⁺ mutations map to conserved internal domains known to be important for substrate and cofactor binding, whereas the a⁺D^- mutations map to a C-terminal region on the surface of the protein. The a⁺D^- mutants also have a temperature-sensitive phenotype when grown on a glycerol medium, which correlates with their degree of mtDNA instability. Analysis of an a⁺D^- mutant with a strong mtDNA instability phenotype shows that it is also unable to parse mtDNA into nucleoids when activated by the GAC pathway. Finally, the wild-type Escherichia coli ortholog of Ilv5p behaves like a⁺D^- mutants when expressed and targeted to mitochondria in ilv5 yeast cells, suggesting that yeast Ilv5p acquired its mtDNA function after the endosymbiotic event.

MtDNA is organized as protein/DNA complexes called nucleoids. Nucleoids have been isolated from yeast (MIYAKAWA et al. 1987 ; NEWMAN et al. 1996 ; KAUFMAN et al. 2000 ) and from several other organisms (VAN TUYLE and MCPHERSON 1979 ; SUZUKI et al. 1982 ; MIYAKAWA et al. 1996 ). Haploid yeast cells grown in rich-dextrose medium contain roughly 25 genome equivalents of mtDNA organized as 15 nucleoids (WILLIAMSON and FENNELL 1979 ). Among the characterized nucleoid proteins are those that would be anticipated to be constituents of an mtDNA/protein complex, such as Abf2p, an abundant HMG-box mtDNA-binding protein (DIFFLEY and STILLMAN 1991 ; MEGRAW et al. 1994 ; NEWMAN et al. 1996 ; ZELENAYA-TROITSKAYA et al. 1998 ; KAUFMAN et al. 2000 ); Rim1p, a single-stranded DNA-binding protein (VAN DYCK et al. 1992 ); and Mgm101p (MEEUSEN et al. 1999 ). Other proteins found associated with mtDNA were not anticipated to be components of mtDNA nucleoids or to function in mtDNA inheritance (KAUFMAN et al. 2000 ); these proteins include mitochondrial heat-shock and metabolic proteins. In addition to their "normal" functions in mitochondrial biogenesis and metabolism, some of these proteins also contribute to mtDNA stability and are thus bifunctional.

The notion of bifunctional proteins involved in mtDNA integrity was first revealed by the finding that Ilv5p, a mitochondrial NADPH-requiring enzyme [acetohydroxyacid (AHA) reductoisomerase] that catalyzes a step in the biosynthesis of branched chain amino acids (KAKAR and WAGNER 1964 ; PETERSEN and HOLMBERG 1986 ), can suppress the mtDNA instability phenotype of cells lacking Abf2p (ZELENAYA-TROITSKAYA et al. 1995 ). Suppression occurs when Ilv5p is expressed at levels only two- or threefold above that expressed from the single-copy ILV5 gene (ZELENAYA-TROITSKAYA et al. 1995 ). Moreover, wild-type mtDNA is also unstable in ilv5 cells, leading to the production of ^- petites (whose mtDNAs are amplified segments of the ⁺ mitochondrial genome). These effects are specific for ILV5, because blocking branched chain amino acid (BCA) biosynthesis by deletion of another gene in the pathway, ILV2, does not result in mtDNA instability (ZELENAYA-TROITSKAYA et al. 1995 ).

ILV5 is one of a number of genes in yeast whose expression is under general amino acid control (GAC; HINNEBUSCH 1988 ). Ilv5p, like many other amino acid biosynthetic enzymes, is induced when cells are starved for amino acids. This induction is under the control of Gcn4p, a positive regulator of the GAC response, which binds to sites upstream of amino acid responsive genes, including ILV5, to activate their transcription (HINNEBUSCH 1988 , HINNEBUSCH 1992 ). Additional insight into the role of Ilv5p in mtDNA maintenance came from the observation that activation of the GAC pathway, either by starving cells for isoleucine, leucine, and valine or by constitutive expression of Gcn4p, resulted in an increase in the number of mtDNA nucleoids without any change in the amount of mtDNA (MACALPINE et al. 2000 ). The increase resulted first from an elevation in mtDNA recombination, resulting in an increase in the number of mtDNA molecules, and second by the parsing of those molecules into additional nucleoid structures. Ilv5p was shown to be required for the latter but not the former step. The factor(s) under GAC that affects mtDNA recombination has not been identified.

Here we describe the separation of the amino acid biosynthetic and DNA transaction functions of Ilv5p by the isolation of point mutants that are defective for one or the other of these two functions. Single amino acid changes in the protein lead to either loss of AHA reductoisomerase activity, but maintenance of mtDNA stability (a^-D⁺), or loss of wild-type mtDNA with no effect on AHA reductoisomerase activity (a⁺D^-). One strong a⁺D^- mutant we chose for further analysis was also unable to parse mtDNA into nucleoids in response to activation of the GAC pathway. The affected residues in the two classes of mutants mapped to distinct regions on the 3-D structure of the AHA reductoisomerase from spinach. We also noted that the a⁺D^- mutants have a strong temperature-sensitive (ts) phenotype when grown on a nonfermentable carbon source, indicative of a severe instability of wild-type mtDNA at the restrictive temperature. Expression of the ortholog of ILV5 from Escherichia coli, ILVc, when targeted to mitochondria, produced a phenotype similar to that of the a⁺D^- mutants. Our data suggest that the yeast mitochondrial AHA reductoisomerase may have evolved to function in mtDNA transactions.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and growth conditions:
Strains used in this study are listed in Table 1. Cells were grown at 30° in YP medium (1% yeast extract and 2% Bacto-peptone) that contained either 2% dextrose (YPD) or 3% glycerol (YPGly), in minimal YNB medium (0.67% yeast nitrogen base without amino acids) that contained either 2% dextrose (YNBD) or 3% glycerol (YNBGly), or in YNB medium that contained 1% casamino acids and either 2% dextrose (YNBD + cas) or 3% glycerol (YNBGly + cas). Additional nutritional supplements were added as required unless specified otherwise. Uracil auxotrophs were produced by plating on YPGly medium containing 5-fluoroorotic acid (US Biological; ROSE et al. 1990 ). Ilv5 mutant alleles were transplaced into the ILV5 locus by transformation of strain 14CWWilv5, using selection on YNBD medium.

Plasmid construction:
Plasmids used in this study are listed in Table 2. pRS-ILVc was constructed as follows. A 1339-bp fragment containing ILV5 upstream sequence and the first 50 amino acids was amplified using primers 5'-ACCTCTAGTGGATCCGTAGATGTAATC-3' and 5'-CGAAGTTGATATGCATCAAACCACG-3' and digested with BamHI and NsiI. The ILVc coding sequence was amplified from E. coli strain DH5 (GIBCO BRL, Gaithersburg, MD) using the primers 5'-GGAATCACCATGCATAACTACTTCAATAC-3' and 5'-GCATCAGCGCAAGCTTAACCCGCAAC-3' and digested with NsiI and HindIII. These PCR products were ligated into the unique BamHI/HindIII sites in pBluescriptII (Stratagene, La Jolla, CA) to produce pBS-ILVc. To produce pBS-ILVc+3', a 645-bp fragment containing the ILV5 3'-untranslated region was then amplified using primers 5'-AAAACCAATAAAGCTTAAAATAATATCAAG-3' and 5'-GAGAAAGAGGAACTCGAGAATTGAG-3', digested with HindIII and XhoI, and ligated into pBS-ILVc. To produce pRS-ILVc, the 3-kb BamHI/XhoI cassette, containing ILVc, was excised from pBS-ILVc+3' and ligated into pRS416.

Assays:
Western blot analysis was carried out as described (KAUFMAN et al. 2000 ). To measure petite formation in the various strains, mid-log phase cultures of strains grown in YNBGly + cas medium were transferred to YNBD + cas medium and grown for at least 30 generations. At various time points cells were plated onto YNBD + cas plates and grown for 3 days at 30°. ⁺ and petite colonies were distinguished by 2,3,5-triphenyltetrazolium chloride (TTC) overlay (OGUR et al. 1957 ), using 0.2% TTC (Sigma, St. Louis) with 0.8% agarose (GIBCO BRL). Between 50 and 300 colonies were counted for each time point.

Random mutagenesis of ILV5:
Low-fidelity PCR amplification of a 1703-bp fragment containing the ILV5 coding sequence was performed using the Diversify PCR random mutagenesis kit (CLONTECH, Palo Alto, CA) with the primers 5'-GGCTTTTACACCCAGTATTTTCCCTTTCC-3' and 5'-TTTCAGGCCACGAGGGTAGCTCATAAC-3'. The pool of mutagenized PCR products was then digested with BclI and BlpI, ligated into pRS-ILV5, which had been digested with the same enzymes, and transformed into competent E. coli DH10B cells (GIBCO BRL). Plasmid DNA was isolated from 10,000 bacterial colonies to produce a library of mutagenized pRS-ILV5, denoted pRS-ILV5^M. The pRS-ILV5^M library was used to transform strain 14WWilv5u^-, selecting on YNBD + cas plates. Transformant colonies were then overlaid with either TTC agar (OGUR et al. 1957 ) to identify a⁺D^- clones or replica plated to YNBD plates to identify a^-D⁺ clones. All mutant isolates were checked, first by Western analysis for the presence of full-length Ilv5p and, second, by isolating plasmid DNA and retransforming it into 14WWilv5u^- to test whether the phenotype was plasmid dependent. For all ilv5 mutants, the region between BclI and BlpI sites was sequenced and aligned against the wild-type sequence to identify mutations.

Fluorescence microscopy:
Nucleoids were visualized by incubation of mid-log phase cells with 1 µg/ml 4',6-diamidino-2-phenylindole (DAPI; Sigma) at 30° for 30 min, followed by four washes in dH₂O, or by fixing the cells with absolute ethanol, incubating in 1 µg/ml DAPI for a few seconds, and then washing four times in dH₂O. Cells were observed using a Leica microscope (model DMRXE) equipped with an HBO 100-W/2 mercury arc lamp and a x100 Plan-Apochromat objective lens. Differential interference contrast and DAPI were visualized using filter sets described previously (OKAMOTO et al. 1998 ).

Protein structure analysis:
Protein alignments were performed using the ClustalX multiple sequence alignment program v1.63b (THOMPSON et al. 1997 ), colored using the MacBoxshade program v2.01 (http://www.ch.embnet.org/software/BOX_doc.html), and then processed in Microsoft Word 98. The structure of AHA reductoisomerase from Spinacia oleracea (PDB ID: 1YVE) was modeled using Swiss-PdbViewer (http://www.expasy.ch/spdbv/; GUEX and PEITSCH 1997 ), rendered using POV-Ray v3.1 (http://www.povray.org), and then processed in Adobe Photoshop (Adobe Systems).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mutational bisection of Ilv5p functions:
To investigate further the bifunctional properties of yeast Ilv5p, we screened for mutations in ILV5 that would separate the BCA biosynthetic function from that of mtDNA stability. To isolate such unifunctional mutants, ilv5 cells were transformed with a library of ILV5 coding sequence mutations generated by low-fidelity PCR and screened for BCA biosynthetic function by growth on medium lacking leucine, isoleucine, and valine and for mtDNA stability by TTC overlay of colonies [which stains respiring colonies (⁺) red but does not stain nonrespiring (petite) colonies] that were grown on rich-dextrose plates (see MATERIALS AND METHODS). This screen yielded the desired mutants: those that retained BCA biosynthetic function but lost the ability to maintain ⁺ mtDNA (a⁺D^-) and those that were defective in BCA biosynthesis but that still maintain ⁺ mtDNA (a^-D⁺; Fig 1). Western blot analysis showed that all a⁺D^- and a^-D⁺ mutant strains expressed comparable amounts of full-length Ilv5p (our unpublished data). Some mutants that were both auxotrophic for branched chain amino acids and produced petites (a^-D^-) were also obtained but were not analyzed further.

View larger version (41K): In this window In a new window Download PPT slide   Figure 1. Unifunctional ilv5 mutants. (A) An ilv5 strain transformed with wild-type ILV5 on a plasmid, empty vector (ilv5), or the indicated ilv5 a-D+ or a+D- mutant alleles on plasmids was pregrown in YNBGly + cas and then shifted to YNBD + cas medium. Aliquots were removed at the time points indicated and scored by TTC overlay (OGUR et al. 1957 ) for the fraction of petites in the population. (B) The strains in A were streaked on YNBD medium with or without supplementation with isoleucine, leucine, and valine (ILV) and incubated for 3 days at 30°.

Fig 1 shows that the four a⁺D^- mutants isolated, F331C, A273V, I267F, and W327R, were able to grow on minimal medium but had a ⁺ mtDNA instability phenotype. The a^-D⁺ mutants, two examples of which (G198D and D255E) are shown in Fig 1, could not grow on minimal medium but maintained mtDNA during outgrowth on rich-dextrose medium. Among the ilv5 a⁺D^- mutants, F331C and A273V produced petites at about the same rate as the ilv5 strain (Fig 1A). The rate of petite production for mutant I267F was much slower, whereas the rate for W327R was significantly greater than that of the ilv5 mutant alone (Fig 1A).

Both classes of unifunctional ilv5 mutants contain single missense mutations that cluster in different regions of the protein:
To determine the mutations within ILV5 responsible for the unifunctional phenotypes, plasmid DNA was isolated and the PCR-mutagenized region was sequenced from each of the mutants. We found that each mutant contained a single missense mutation within the coding region of ILV5 (Fig 2). By aligning the Saccharomyces cerevisiae Ilv5p sequence with orthologs from Neurospora crassa, S. oleracea (spinach), and E. coli, we observed that all four a⁺D^- mutations and six out of seven a^-D⁺ mutations were at conserved residues. DUMAS et al. 1995 analyzed the sequence of AHA reductoisomerase from a variety of species and found five regions of identity, which they designated domains I–V (Fig 2). Site-directed mutagenesis of domains II–V led to partial or total inactivation of enzyme activity. Interestingly, five out of seven a^-D⁺ mutations are at residues within domains I–V, while none of the a⁺D^- mutations is within those regions.

View larger version (74K): In this window In a new window Download PPT slide   Figure 2. Locations of a-D+ and a+D- Ilv5p mutations shown in an alignment with other AHA reductoisomerases. S. cerevisiae (Sc), N. crassa (Nc), S. oleracea (So; spinach), and E. coli (Ec) AHA reductoisomerase sequences are shown; respective GenBank accession numbers are P06168, P38674, X57073, and 65954. The boxed regions are the conserved amino acid residues that have been shown to be critical for substrate, NADPH, and Mg2+ binding and activity of AHA reductoisomerase (DUMAS et al. 1995 ) and define in part the conserved domains I–V of the enzyme. Above the alignment, amino acids are shown, corresponding to the a-D+ (blue) and a+D- (red) Ilv5p mutations. Gray indicates amino acid identity and yellow, conserved amino acids. The arrow indicates the location of a 140-amino-acid insert in the spinach and E. coli enzymes.

The crystal structure of spinach AHA reductoisomerase complexed with NADPH, two Mg²⁺ ions, and a competitive inhibitor of enzymatic activity (N-hydroxy-N-isopropyloxamate) has been solved to 1.65-Å resolution (BIOU et al. 1997 ). In agreement with mutagenesis and enzyme kinetic data (DUMAS et al. 1995 ), the conserved residues in domains I–V were found to bind NADPH and Mg²⁺ ions within the active site, which is deeply buried within the protein core. We mapped the mutated residues from both classes of mutants to the equivalent residues on the three-dimensional structure of the spinach enzyme (Fig 3). Although the S. cerevisiae and spinach proteins have significant sequence similarity (31% identity), the spinach protein contains a 140-amino-acid insertion that is absent from the fungal enzymes (see Fig 2, arrow). The spinach enzyme forms a homodimer, where the majority of the dimer interface is between two -helices (A17 and A18) and a loop (the "dimer loop") within that 140-amino-acid insert (BIOU et al. 1997 ). Deletion of the dimer loop causes the spinach enzyme to form an active monomer (WESSEL et al. 1998 ). Furthermore, two other -helices (A20, residues 462–482; A26, residues 529–538), which are present in yeast Ilv5p, were also found to be on the monomer surface and at the dimer interface (WESSEL et al. 1998 ). Interestingly, three of the four a⁺D^- equivalent residues (A273V, W327R, and F331C) are within or immediately adjacent to helices A20 or A26 (Fig 3A). The only a⁺D^- strain whose mutated residue is not within these two helices is I267F, which has by far the weakest petite-inducing phenotype (Fig 1A). By contrast, equivalent mutated residues in the a^-D⁺ mutants all lie buried within the core of the protein, and three of these mutations (Q88R, D255E, and E259G) are of residues known to bind either NADPH or Mg²⁺ (BIOU et al. 1997 ; Fig 3B). Taken together, these data suggest that the a⁺D^- mutations are located at or close to the surface of Ilv5p and thus may affect intermolecular interactions of the protein.

View larger version (38K): In this window In a new window Download PPT slide   Figure 3. Residues mutated in Ilv5p unifunctional mutants mapped on the spinach AHA reductoisomerase 3-D ribbon structure. One of the two monomers of the spinach protein (which is a dimer) is shown in two opposite orientations. Equivalent residues (with sidechains) to those affected in a+D- mutants are shown in red (A), and a-D+ mutants are shown in blue (B). The 140-amino-acid insert of the spinach protein, which is absent from yeast AHA reductoisomerases, is shown in green. The inhibitor, N-hydroxy-N-isopropyloxamate, is shown in yellow. A17, A18, A20, and A26 mark -helices referred to in the text. DL, dimer loop.

The bacterial ortholog of ILV5 complements BCA biosynthesis but not mtDNA instability in ilv5 yeast cells:
The gene encoding AHA reductoisomerase (ILV5 in S. cerevisiae) is highly conserved among prokaryotes and eukaryotes (PETERSEN and HOLMBERG 1986 ). Although Ilv5p is involved in mtDNA stability in S. cerevisiae, it is not known whether prokaryotic orthologs have any DNA-associated functions or whether that property evolved after the endosymbiotic events that gave rise to present-day mitochondria. To obtain additional insight into the bifunctional nature of Ilv5p, we expressed the ILV5 ortholog from E. coli (ILVc) in an ilv5 yeast strain and then assessed whether the bacterial gene could complement the ilv5 phenotypes. The coding region of ILVc was amplified by PCR and cloned into the centromeric vector, pRS-ILV5, in place of the ILV5 coding region downstream of the mitochondrial targeting sequence. In this way, expression of ILVc was under the control of the yeast ILV5 promoter and the protein (IlvCp) targeted to mitochondria. The plasmid, pRS-ILVc, was then transformed into an ilv5 strain (14WWilv5u^-), which is auxotrophic for isoleucine, leucine, and valine and defective in maintaining the integrity of ⁺ mtDNA when grown in rich-dextrose medium (ZELENAYA-TROITSKAYA et al. 1995 ). To determine whether ILVc could complement either of these ilv5 phenotypes, we examined BCA auxotrophy, as well as the stability of wild-type mtDNA among the transformants. ILVc transformants were able to grow at the same rate as wild type in medium lacking isoleucine, leucine, and valine, demonstrating that the bacterial AHA reductoisomerase complements the enzymatic activity of the yeast protein in the biosynthesis of BCAs (Fig 4A). By contrast, when ILVc transformants were grown in rich-dextrose medium, those cells not only produced petites, but also did so at a rate somewhat greater than that observed in control ilv5 cells carrying a plasmid without an insert (Fig 4B). Therefore, the E. coli AHA reductoisomerase, unlike its yeast ortholog, is not a bifunctional protein when expressed in yeast: it complements BCA biosynthesis, but does not contribute to the maintenance of wild-type mtDNA in cells lacking the yeast ILV5 gene.

View larger version (33K): In this window In a new window Download PPT slide   Figure 4. The E. coli AHA reductoisomerase (IlvCp) complements the defect in BCA biosynthesis but not the mtDNA instability phenotype of ilv5 yeast cells. (A) Ilv5 cells transformed with wild-type ILV5 on a centromeric plasmid (ILV5), empty vector (ilv5), or pRS-ILVc expressing IlvCp in mitochondria (ILVc) were scored for their requirement for isoleucine, leucine, and valine (ILV) as in Fig 1B. (B) Aliquots of cultures of the strains described in A were scored for petite production as in Fig 1A.

a⁺D^- mutants are impaired for growth at 37° on glycerol:
To characterize further the mtDNA instability phenotypes of the different a⁺D^- mutant ilv5 alleles, we examined the ability of the mutants to grow on glycerol medium at 37°. Increased temperature is known to exacerbate the phenotype of some mutants affected in mtDNA stability. abf2 cells, for example, which lose mtDNA on glucose medium, grow on glycerol medium at 30° but not at 37° (MEGRAW and CHAE 1993 ). To test the phenotypes of the two classes of ilv5 mutants (and cells expressing the bacterial ortholog, ILVc) at 37°, ilv5 cells were transformed with either the a^-D⁺ and the a⁺D^- mutant ilv5 alleles or the ILVc gene on centromeric plasmids, as well as with the empty vector and the wild-type ILV5 allele as controls. Cells were pregrown in YNBGly + cas medium at 30°, plated onto YNBGly + cas and YNBD + cas medium, and incubated at 30° or 37° (Fig 5A). As had been reported previously (ZELENAYA-TROITSKAYA et al. 1995 ), ilv5 cells have a slight ts phenotype on glycerol medium. In contrast, some of the a⁺D^- ilv5 mutants and ILVc are clearly ts for glycerol growth, the extent of which correlates well with the severity of the petite phenotype of each mutant (see Fig 1A). Consistent with their lack of an mtDNA instability phenotype, growth of the a^-D⁺ mutants on glycerol was unaffected at 37° (Fig 5A).

View larger version (67K): In this window In a new window Download PPT slide   Figure 5. ilv5 a+D- mutants are temperature-sensitive for growth on a nonfermentable carbon source and are recessive. (A) The ilv5 strains (described in Fig 1 and Fig 4) that were transformed with the centromeric plasmids containing the various mutant and wild-type ILV5 alleles as well as ILVc (indicated in the left side of the figure) were grown to mid-logarithmic phase in YNBGly + cas medium and then plated at decreasing dilutions onto YNBD + cas or YNBGly + cas medium and grown for 3 days at 30° or 37°. (B) The wild-type strain 14WW was transformed with plasmids containing either a+D- ilv5 mutants W327R and A273V or an empty vector (pRS416) and assessed for ts growth as in A. (C) Strain W327R-I containing an integrated copy of ilv5 a+D- mutant W327R at the ILV5 locus was transformed with a plasmid containing wild-type ILV5 or an empty vector (pRS416) and tested for ts growth as in A. The chromosomal ILV5 alleles of the strains are indicated along the right side of the figure.

To determine whether the ts phenotype conferred by the a⁺D^- mutants was dominant or recessive, centromeric plasmids containing a⁺D^- mutants W327R or A273V or empty vector were transformed into the wild-type ILV5 strain, 14WW. Repetition of the previous experiment using these strains demonstrated that the glycerol ts phenotype of a⁺D^- mutants was recessive, as these transformants grew as well as wild-type ILV5 cells containing the empty vector (Fig 5B). To confirm this result, we constructed strain W327R-I, which contains an integrated copy of the ilv5 a⁺D^- mutant allele, W327R, and transformed this strain with either empty vector or a centromeric plasmid containing the wild-type ILV5 allele. The results of these growth experiments confirm that the ts glycerol growth phenotype for W327R is recessive (Fig 5C). Consistent with these data, the mtDNA instability phenotype in the a⁺D^- mutant strains is also recessive (our unpublished data).

a⁺D^- ilv5 mutant W327R produces ^- petites:
As expression of a⁺D^- ilv5 mutant W327R in ilv5 cells resulted in a hypermorphic phenotype compared to ilv5 cells (Fig 1 and Fig 5), we wondered whether the petites produced by this strain were ^- or ⁰. Previously we had shown that, unlike abf2 cells, which produce ⁰ petites, petites produced by ilv5 cells are ^- (ZELENAYA-TROITSKAYA et al. 1995 ). To determine this we pregrew strains 14WWilv5 and W327R-I in YPGly medium and then outgrew these strains in YPD medium until >90% of the cells were respiration incompetent as assessed by their inability to grow on a nonfermentable carbon source (data not shown). This culture was then fixed and DAPI stained, and the cells were observed by epifluorescence microscopy to determine whether the cells still contained mtDNA, which would indicate that ^- rather than ⁰ petites were produced. As shown in Fig 6, when compared to their ⁺ parents, the vast majority of cells in cultures of both 14WWilv5u^- and W327R-I, which are >90% respiration deficient, contain mtDNA. Moreover, the DAPI staining shows that most of the cells contain a few brightly staining punctate mtDNA nucleoids, characteristic of the nucleoid organization of ^- petites. During continued outgrowth of these cells, the ^- mtDNAs were stably maintained (data not shown). Therefore, even though a⁺D^- ilv5 mutant W327R is a hypermorph compared with ilv5 cells, both mutations result in the production of ^- petites.

View larger version (86K): In this window In a new window Download PPT slide   Figure 6. ilv5 and ilv5 a+D- mutant cells produce - petites. Strains 14WWilv5u- and W327R-I were grown in YPD medium to a point where >90% of the cells were unable to grow on a nonfermentable carbon source and were then DAPI stained and the nucleoids were observed by fluorescence microscopy (c and d). Representative cells are shown. The + precultures were also DAPI stained (a and b). Arrows point to nucleoid structures. Bar, 3 µm.

An a⁺D^- ilv5 mutant cannot reorganize nucleoids:
Recently we reported that ILV5 is essential for the parsing of mtDNA into nucleoids (MACALPINE et al. 2000 ). We defined parsing as an increase in nucleoid number resulting from an increase in the number of individual mtDNA molecules (due, for example, to increased mtDNA recombination) without a concomitant increase in the total amount of mtDNA. We found that nucleoid numbers, especially in ^- petite cells whose mtDNA consists of tandem repeats, increased dramatically (10-fold) as a result of activation of the GAC pathway. In ilv5 cells, this increase in nucleoid number failed to occur even though there was the same increase in the number of individual mtDNA molecules due to activation of the GAC pathway as there was in wild-type cells.

To determine whether the mtDNA parsing function of Ilv5p was also affected by an a⁺D^- ilv5 mutation, centromeric plasmids containing wild-type ILV5 or the W327R ilv5 mutant allele were transformed into an ilv5 derivative of the ^- petite strain HS40. These strains either were grown in YNBD medium with or without amino acid supplementation (casein) or were transformed with another centromeric expression plasmid, pRS-gcn4c, containing the gcn4c mutant allele and grown in YNBD + cas medium. The gcn4c allele has mutations in regulatory AUG codons upstream of the GCN4-coding sequence, so that Gcn4p is constitutively expressed and is insensitive to multivalent repression by isoleucine, leucine, and valine (MUELLER and HINNEBUSCH 1986 ). Cells were then stained with DAPI to determine the nucleoid distribution (Fig 7A), and the number of nucleoids per focal plane in 12 randomly chosen cells was counted for each sample (Fig 7B). ILV5 ^- cells had a few, brightly staining nucleoids when grown in YNBD + cas medium (Fig 7A, a), as is typical of many petites when grown in rich medium. In contrast, many more nucleoid structures were evident in cells grown in YNBD + cas medium that were constitutively expressing Gcn4p (Fig 7A, Fig B) or in cells grown in YNBD medium lacking amino acids (Fig 7A, Fig C). These results are in agreement with our previous observations (MACALPINE et al. 2000 ). The W327R a⁺D^- ilv5 mutant, however, failed to exhibit this pattern of increased nucleoid number in either condition that activates the GAC pathway (Fig 7A, d–f). Thus a single point mutation in Ilv5p results in both an instability of ⁺ mtDNA and a failure to parse mtDNA into nucleoids in response to activation of the GAC pathway. These data suggest that the functions of Ilv5p in mtDNA stability and nucleoid parsing may be mechanistically linked.

View larger version (49K): In this window In a new window Download PPT slide   Figure 7. ilv5 a+D- mutants are unable to parse mtDNA into nucleoids in - petites. (A) Strain 14WWilv5u- containing the petite mitochondrial genome HS40 was transformed with plasmids containing wild-type ILV5 (a–c), a+D- mutant W327R (d–f), empty plasmid (g and h), or a-D+ mutant D255E (i and j). These strains were grown on YNBD + cas with or without the gcn4c plasmid or on YNBD medium as indicated in the figure. Representative DAPI-stained cells are shown. Bar, 3 µm. (B) The number of nucleoids per focal plane in 12 randomly chosen cells was counted for each of the strains and conditions described in A.

Next, we tested whether ilv5 a^-D⁺ mutants retain the ability to reorganize nucleoids. Because those mutant cells are auxotrophic for isoleucine, leucine, and valine, the GAC pathway had to be induced by constitutive expression of Gcn4p. Accordingly, centromeric plasmids with no insert (pRS416; Fig 7A, Fig G and Fig H) or an a^-D⁺ ilv5 mutant allele (D255E; Fig 7A, Fig I and Fig J) were transformed into the ilv5 derivative of the HS40 ^- petite strain used above containing pRS-gcn4c and the strains grown in YNBD + cas medium. The results of these experiments show an increase in the number of nucleoids in the a^-D⁺ mutant strain D255E (Fig 7A, Fig J), similar to that observed in cells expressing wild-type ILV5 when cultured in minimal medium or constitutively expressing Gcn4p, compared with the same cells harboring the empty vector (Fig 7A, Fig G and Fig H).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have genetically bisected the mtDNA transaction and AHA reductoisomerase functions of the S. cerevisiae mitochondrial protein Ilv5p. Although Ilv5p was the first bifunctional protein to be shown to be involved in mtDNA inheritance in yeast (ZELENAYA-TROITSKAYA et al. 1995 ), recent evidence suggests that there may be many more such proteins in yeast mitochondria (KAUFMAN et al. 2000 ). This study is the first of its kind to isolate alleles that separate the two functions of a bifunctional protein involved in mtDNA stability. When the ilv5 mutation was originally shown to result in the destabilization of ⁺ mtDNA (ZELENAYA-TROITSKAYA et al. 1995 ), control experiments indicated that this destabilization was not the result of a block in branched chain amino acid biosynthesis, because deletion of another gene in the pathway, ILV2, had no effect on mtDNA stability. The isolation of a^-D⁺ Ilv5p mutants confirms and extends that result by showing that ⁺ mtDNA is stable in cells with a catalytically inactive Ilv5p.

The a^-D⁺ Ilv5p mutations we obtained clearly fall within or near domains of the protein that are conserved among AHA reductoisomerases; these correspond to regions of the protein that have been shown to be important for substrate and cofactor binding and for enzymatic activity (DUMAS et al. 1995 ; BIOU et al. 1997 ). To determine whether the positions of the mutated residues in the a⁺D^- mutants could give some insight into the DNA functions of Ilv5p, we mapped those mutations to the 3-D structure of the spinach ortholog, the only available structure for an AHA reductoisomerase (BIOU et al. 1997 ; HALGAND et al. 1999 ). Despite the uncertainties inherent in extrapolating this structure to yeast Ilv5p, it is nevertheless clear that the Ilv5p a⁺D^- mutations cluster in a different region of the protein than do the a^-D⁺ mutations. Specifically, three of the four Ilv5p a⁺D^- mutations lie within two adjacent -helices in the C terminus, whereas all the a^-D⁺ mutations are in residues in the pocket containing the active site and cofactor binding sites.

Given the mtDNA instability phenotype of the a⁺D^- mutants and the presence of Ilv5p among the proteins recovered with mtDNA following formaldehyde crosslinking of mitochondria in organello (KAUFMAN et al. 2000 ), it is reasonable to hypothesize that Ilv5p interacts directly with mtDNA. However, Ilv5p does not contain any known DNA-binding motifs, nor have we been able to detect any direct DNA-binding activity for wild-type Ilv5p or for any of the mutant Ilv5p's expressed and purified from E. coli (unpublished data). This raises the possibility that the defects in mtDNA transactions in the a⁺D^- mutants may be the result of aberrant interactions with other proteins, some of which may include nucleoid proteins. The inability of the bacterial ortholog of ILV5 (ILVc) to stabilize wild-type mtDNA when expressed and targeted to mitochondria of ilv5 cells but which is nevertheless able to complement the loss of branched chain amino acid biosynthesis suggests that the mtDNA stability function may have evolved after the endosymbiotic event. It is also possible that ILVc could have a DNA maintenance function in E. coli, which has yet to be identified, but still not be able to complement the mtDNA function of ILV5 in yeast mitochondria. It is intriguing that ILVc has a hypermorphic ⁺ mtDNA instability phenotype similar to that of the a⁺D^- mutant, W327R. Perhaps IlvCp, although unable to complement the ilv5 mtDNA instability phenotype, is recognized by the nucleoid as being similar to Ilv5p, and this recognition results in aberrant protein-protein interactions comparable to the W327R mutant. Null mutants of ILV5 have been shown previously to have a slight ts phenotype when grown at 37° on glycerol medium (ZELENAYA-TROITSKAYA et al. 1995 ), unlike abf2 cells, which are strongly ts (MEGRAW and CHAE 1993 ). In accordance with its petite phenotype, the a⁺D^- mutant, W327R, exhibited a hypermorphic ts phenotype at 37° on glycerol medium, indicating strong instability of ⁺ mtDNA. Surprisingly, mutants A273V and F331C, which have an mtDNA instability phenotype similar to that of an ilv5 strain, also exhibited a more severe ts phenotype than did ilv5 cells. Altogether, these observations indicate that the presence of a⁺D^- mutant Ilv5p's in mitochondria exacerbates the ⁺ mtDNA instability observed in ilv5 cells. Despite some a⁺D^- mutants being hypermorphs, all are recessive for both the ts and petite-inducing phenotypes. Preliminary data suggest that this may be related to the oligomerization state of Ilv5p, where the mutant protein is able to interact with wild-type Ilv5p in a heteroligomeric structure (data not shown), perhaps squelching the aberrant protein-protein interactions of the mutants.

Previous studies have shown ILV5 to be involved in two seemingly unrelated events involving mtDNA: the stability of ⁺ mtDNA (ZELENAYA-TROITSKAYA et al. 1995 ) and the parsing of mtDNA into nucleoids in response to activation of the GAC pathway (MACALPINE et al. 2000 ). Demonstration that a single point mutation in the a⁺D^- mutant W327R causes cells both to lose wild-type mtDNA and to fail to redistribute mtDNA into nucleoids under amino acid starvation conditions unites these two processes. When we initiated the screen to isolate unifunctional mutants of Ilv5p, we supposed that the a⁺D^- mutants would have lost the mtDNA stability function of the wild-type protein, similar to cells lacking Ilv5p, but nevertheless be catalytically active. However, although they all appear to be capable of sustaining branched chain amino acid biosynthesis to a similar degree, the mtDNA transaction phenotypes caused by the different a⁺D^- mutants are more complex than simple loss-of-function mutations. Given the number of proteins alone that are present in mtDNA nucleoids (25), significant synergism is likely among these factors in controlling mtDNA stability and inheritance. Consequently, it is plausible that the presence of a defective (a⁺D^-) Ilv5p could result in a more severe mtDNA instability phenotype than the simple absence of the protein. Further biochemical experiments will be required to analyze the composition and functionality of mtDNA nucleoids in a⁺D^- cells.

Regarding the mechanism of mtDNA instability in an ilv5 strain vs. an a⁺D^- mutant, it is interesting to note that although the severity of petite production in strain W327R is far greater than that in ilv5, both strains produce exclusively ^- petites. By contrast, mutations in nucleoid proteins, such as Mgm101p (MEEUSEN et al. 1999 ) and Abf2p, result in the production of ⁰ petites, suggesting that their defect is due to a failure to transmit mtDNA (CONTAMINE and PICARD 2000 ). Despite recognition of the structures of ^- mtDNAs for many years now, few mechanistic details are available on the recombination and amplification events that account for their formation. It is tempting to consider the possibility that the mtDNA parsing function of Ilv5p, that is, control of the number of individual mtDNA molecules per nucleoid, may be intimately related to mtDNA recombination events that could lead to the production of ^- genomes.

	ACKNOWLEDGMENTS

This work was supported by grant GM-33510 from the National Institutes of Health and by grants I-0642 and I-1211 from The Robert A. Welch Foundation.

Manuscript received March 25, 2002; Accepted for publication May 1, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BIOU, V., R. DUMAS, C. COHEN-ADDAD, R. DOUCE, and D. JOB et al., 1997  The crystal structure of plant acetohydroxy acid isomeroreductase complexed with NADPH, two magnesium ions and a herbicidal transition state analog determined at 1.65 Å resolution. EMBO J. 16:3405-3415.[Medline]

CONTAMINE, V. and M. PICARD, 2000  Maintenance and integrity of the mitochondrial genome: a plethora of nuclear genes in the budding yeast. Microbiol. Mol. Biol. Rev. 64:281-315.[Abstract/Free Full Text]

DIFFLEY, J. F. and B. STILLMAN, 1991  A close relative of the nuclear, chromosomal high-mobility group protein HMG1 in yeast mitochondria. Proc. Natl. Acad. Sci. USA 88:7864-7868.[Abstract/Free Full Text]

DUMAS, R., M. C. BUTIKOFER, D. JOB, and R. DOUCE, 1995  Evidence for two catalytically different magnesium-binding sites in acetohydroxy acid isomeroreductase by site-directed mutagenesis. Biochemistry 34:6026-6036.[Medline]

GUEX, N. and M. C. PEITSCH, 1997  SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714-2723.[Medline]

HALGAND, F., R. DUMAS, V. BIOU, J. P. ANDRIEU, and K. THOMAZEAU et al., 1999  Characterization of the conformational changes of acetohydroxy acid isomeroreductase induced by the binding of Mg2+ ions, NADPH, and a competitive inhibitor. Biochemistry 38:6025-6034.[Medline]

HINNEBUSCH, A. G., 1988  Mechanisms of gene regulation in the general control of amino acid biosynthesis in Saccharomyces cerevisiae.. Microbiol. Rev. 52:248-273.[Free Full Text]

HINNEBUSCH, A. G., 1992 General and pathway-specific regulatory mechanisms controlling the synthesis of amino acid biosynthetic enzymes in Saccharomyces cerevisiae, pp. 319–414 in The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression, edited by E. W. JONES, J. R. PRINGLE and J. R. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

KAKAR, S. N. and R. P. WAGNER, 1964  Genetic and biochemical analysis of isoleucine-valine mutants of yeast. Genetics 49:213-222.[Free Full Text]

KAUFMAN, B. A., S. M. NEWMAN, R. L. HALLBERG, C. A. SLAUGHTER, and P. S. PERLMAN et al., 2000  In organello formaldehyde crosslinking of proteins to mtDNA: identification of bifunctional proteins. Proc. Natl. Acad. Sci. USA 97:7772-7777.[Abstract/Free Full Text]

MACALPINE, D. M., P. S. PERLMAN, and R. A. BUTOW, 2000  The number of individual mitochondrial DNA molecules and mitochondrial DNA nucleoids in yeast are co-regulated by the general amino acid control pathway. EMBO J. 19:767-775.[Medline]

MEGRAW, T. L. and C. B. CHAE, 1993  Functional complementarity between the HMG1-like yeast mitochondrial histone HM and the bacterial histone-like protein HU. J. Biol. Chem. 268:12758-12763.[Abstract/Free Full Text]

MEGRAW, T. L., L. R. KAO, and C. B. CHAE, 1994  The mitochondrial histone HM: an evolutionary link between bacterial HU and nuclear HMG1 proteins. Biochimie 76:909-916.[Medline]

MEEUSEN, S. Q., Q. TIEU, E. WONG, E. WEISS, and D. SCHEILTZ et al., 1999  Mgm101p is a novel component of the mitochondrial nucleoid that binds DNA and is required for the repair of oxidatively damaged mitochondrial DNA. J. Cell Biol. 145:291-304.[Abstract/Free Full Text]

MIYAKAWA, I., N. SANDO, K. KAWANO, S. NAKAMURA, and T. KUROIWA, 1987  Isolation of morphologically intact mitochondrial nucleoids from the yeast, Saccharomyces cerevisiae.. J. Cell Sci. 88:431-439.[Abstract/Free Full Text]

MIYAKAWA, I., C. OKAZAKIHIGASHI, T. HIGASHI, Y. FURUTANI, and N. SANDO, 1996  Isolation and characterization of mitochondrial nucleoids from the yeast Pichia jadinii.. Plant Cell Physiol. 37:816-824.[Abstract/Free Full Text]

MUELLER, P. P. and A. HINNEBUSCH, 1986  Multiple upstream AUG codons mediate translational control of GCN4.. Cell 45:201-207.[Medline]

NEWMAN, S. M., O. ZELENAYA-TROITSKAYA, P. S. PERLMAN, and R. A. BUTOW, 1996  Analysis of mitochondrial DNA nucleoids in wild-type and a mutant strain of Saccharomyces cerevisiae that lacks the mitochondrial HMG-box protein, Abf2p. Nucleic Acids Res. 24:386-393.[Abstract/Free Full Text]

OGUR, M., R. ST. JOHN, and S. NAGAI, 1957  Tetrazolium overlay technique for population studies of respiration deficiency in yeast genetics. Science 125:928-929.[Free Full Text]

OKAMOTO, K., P. S. PERLMAN, and R. A. BUTOW, 1998  The sorting of mitochondrial DNA and mitochondrial proteins in zygotes: preferential transmission of mitochondrial DNA to the medial bud. J. Cell Biol. 142:613-623.[Abstract/Free Full Text]

PETERSEN, J. G. L. and S. HOLMBERG, 1986  The ILV5 gene of Saccharomyces cerevisiae is highly expressed. Nucleic Acids Res. 14:9631-9651.[Abstract/Free Full Text]

ROSE, M. D., F. WINSTON and P. HEITER, 1990 Methods in Yeast Genetics: A Laboratory Course Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SIKORSKI, R. S. and P. HIETER, 1989  A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae.. Genetics 122:19-27.[Abstract/Free Full Text]

SUZUKI, T., S. KAWANO, and T. KUROIWA, 1982  Structure of three-dimensionally rod-shaped mitochondrial nucleoids isolated from the slime mould Physarum polycephalum.. J. Cell Sci. 58:241-261.[Abstract]

THOMPSON, J. D., T. J. GIBSON, F. PLEWNIAK, F. JEANMOUGIN, and D. G. HIGGINS, 1997  The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882.[Abstract/Free Full Text]

VAN DYCK, E., F. FOURY, B. STILLMAN, and S. J. BRILL, 1992  A single-stranded DNA binding protein required for mitochondrial DNA replication in S. cerevisiae is homologous to E. coli SSB.. EMBO J. 11:3421-3430.[Medline]

VAN TUYLE, G. C. and M. L. MCPHERSON, 1979  A compact form of rat liver mitochondrial DNA stabilized by bound proteins. J. Biol. Chem. 254:6044-6053.[Abstract/Free Full Text]

WESSEL, P. M., V. BIOU, R. DOUCE, and R. DUMAS, 1998  A loop deletion in the plant acetohydroxy acid isomeroreductase homodimer generates an active monomer with reduced stability and altered magnesium affinity. Biochemistry 37:12753-12760.[Medline]

WILLIAMSON, D. H. and D. J. FENNELL, 1979  Visualization of yeast mitochondrial DNA with the fluorescent stain "DAPI.". Methods Enzymol. 56:728-733.[Medline]

ZELENAYA-TROITSKAYA, O., P. S. PERLMAN, and R. A. BUTOW, 1995  ILV5 encodes a bifunctional mitochondrial protein involved in branched chain amino acid biosynthesis and maintenance of mitochondrial DNA. EMBO J. 14:3268-3276.[Medline]

ZELENAYA-TROITSKAYA, O., S. M. NEWMAN, K. OKAMOTO, P. S. PERLMAN, and R. A. BUTOW, 1998  Functions of the HMG box protein, Abf2p, in mitochondrial DNA segregation, recombination and copy number in Saccharomyces cerevisiae.. Genetics 148:1763-1776.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. Kucej, B. Kucejova, R. Subramanian, X. J. Chen, and R. A. Butow Mitochondrial nucleoids undergo remodeling in response to metabolic cues J. Cell Sci., June 1, 2008; 121(11): 1861 - 1868. [Abstract] [Full Text] [PDF]

				C. Gancedo and C.-L. Flores Moonlighting Proteins in Yeasts Microbiol. Mol. Biol. Rev., March 1, 2008; 72(1): 197 - 210. [Abstract] [Full Text] [PDF]

				B. K. Benson, G. Meades Jr, A. Grove, and G. L. Waldrop DNA inhibits catalysis by the carboxyltransferase subunit of acetyl-CoA carboxylase: Implications for active site communication Protein Sci., January 1, 2008; 17(1): 34 - 42. [Abstract] [Full Text] [PDF]

				X. J. Chen, X. Wang, and R. A. Butow Yeast aconitase binds and provides metabolically coupled protection to mitochondrial DNA PNAS, August 21, 2007; 104(34): 13738 - 13743. [Abstract] [Full Text] [PDF]

				D. A. Hall, H. Zhu, X. Zhu, T. Royce, M. Gerstein, and M. Snyder Regulation of Gene Expression by a Metabolic Enzyme Science, October 15, 2004; 306(5695): 482 - 484. [Abstract] [Full Text] [PDF]

				A. Elo, A. Lyznik, D. O. Gonzalez, S. D. Kachman, and S. A. Mackenzie Nuclear Genes That Encode Mitochondrial Proteins for DNA and RNA Metabolism Are Clustered in the Arabidopsis Genome PLANT CELL, July 1, 2003; 15(7): 1619 - 1631. [Abstract] [Full Text] [PDF]

				J. M. Bateman, M. Iacovino, P. S. Perlman, and R. A. Butow Mitochondrial DNA Instability Mutants of the Bifunctional Protein Ilv5p Have Altered Organization in Mitochondria and Are Targeted for Degradation by Hsp78 and the Pim1p Protease J. Biol. Chem., November 27, 2002; 277(49): 47946 - 47953. [Abstract] [Full Text] [PDF]