Analysis of the Seven-Member AAD Gene Set Demonstrates That Genetic Redundancy in Yeast May Be More Apparent Than Real

Corresponding author: Stephen G. Oliver, School of Biological Sciences, University of Manchester, 2.205 Stopford Bldg., Oxford Rd., Manchester M13 9PT, United Kingdom., steve.oliver{at}man.ac.uk (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Saccharomyces cerevisiae has seven genes encoding proteins with a high degree (>85%) of amino-acid sequence identity to the aryl-alcohol dehydrogenase of the lignin-degrading, filamentous fungus, Phanerochaete chrysosporium. All but one member of this gene set are telomere associated. Moreover, all contain a sequence similar to the DNA-binding site of the Yap1p transcriptional activator either upstream of or within their coding sequences. The expression of the AAD genes was found to be induced by chemicals, such as diamide and diethyl maleic acid ester (DEME), that cause an oxidative shock by inactivating the glutathione (GSH) reservoir of the cells. In contrast, the oxidizing agent hydrogen peroxide has no effect on the expression of these genes. We found that the response to anti-GSH agents was Yap1p dependent. The very high level of nucleotide sequence similarity between the AAD genes makes it difficult to determine if they are all involved in the oxidative-stress response. The use of single and multiple aad deletants demonstrated that only AAD4 (YDL243c) and AAD6 (YFL056/57c) respond to the oxidative stress. Of these two genes, only AAD4 is likely to be functional since the YFL056/57c open reading frame is interrupted by a stop codon. Thus, in terms of the function in response to oxidative stress, the sevenfold redundancy of the AAD gene set is more apparent than real.

THE complete genome sequence of the brewers' and bakers' yeast Saccharomyces cerevisiae (GOFFEAU et al. 1996 ) is an important resource for the study of eukaryotic molecular genetics. This sequence has defined the extent of our knowledge of yeast genetics; less than 45% of the ~6000 protein-encoding genes uncovered on the organism's 16 chromosomes have had their function determined experimentally. Because the S. cerevisiae genome is relatively small (less than four times the size of that of the bacterium Escherichia coli, with only 50% more protein-encoding genes; BLATTNER et al. 1997 ) the extent of apparent genetic redundancy at the level of either their nucleotide sequences or of the amino-acid sequences of their predicted protein products (MEWES et al. 1997 ; WOLFE and SHIELDS 1997 ) seems remarkable. Most of the duplicated genes are members of families with just two or three members, but some gene families are significantly larger.

These duplicated open reading frames (ORFs) are arranged in blocks, called cluster homology regions (CHRs; GOFFEAU et al. 1996 ), that are found both at the chromosome ends and at internal sites within chromosome arms. It had long been known that yeast contains a number of multigene families whose members are predominantly associated with chromosome ends (e.g., the SUC, MAL, and MEL genes which are involved, respectively, in the fermentation of sucrose, maltose, and melibiose; CARLSON et al. 1985 ; CHARRON et al. 1989 ; NAUMOV et al. 1990 ). In the case of the smallest yeast chromosome (chromosome I), it has been suggested that redundant sequences have accumulated at the ends simply to increase its size and ensure correct disjunction in meiotic nuclear division (BUSSEY et al. 1995 ). Similar arguments may be made for some of the other small chromosomes (OLIVER 1995 ). In other cases (e.g., chromosomes III, OLIVER et al. 1992 , and XI, DUJON et al. 1994 ), the presence of repeat sequences, at internal sites on chromosome arms, that are normally associated with telomeres indicates that the chromosomes may have grown over evolutionary time by recombination events involving their ends. Several smaller CHRs are associated with recombination events between yeast transposons or their LTRs, either by the excision of a section of a chromosome and its reintegration at a new site (MELNICK and SHERMAN 1993 ) or by unequal crossing-over events caused by the misalignment of Ty elements at different positions within a chromosome arm (WICKSTEED et al. 1994 ).

These different mechanisms for gene duplication are probably insufficient to account for the full extent of redundancy in the yeast genome. WOLFE and SHIELDS 1997 have provided a more radical view of this phenomenon. They propose that the S. cerevisiae genome, at some stage in its evolutionary history, underwent a complete duplication and that the genome has been subsequently reduced to its present size via a series of deletion and reciprocal translocation events (KEOGH et al. 1998 ). Whatever the evolutionary history of the CHRs, it should be noted that they do not provide an explanation for the failure of classical (or "function first"; OLIVER 1996B ) genetics to detect >45% of yeast's genes. If these genes had not been revealed because they exist in more than one copy in the yeast genome, thus masking the effect of recessive mutations, then it would be expected that these "orphan" genes (DUJON 1996 ) would be overrepresented in the CHRs. In fact, if anything, it is the genes of experimentally determined function that are overrepresented in the CHRs (Table 1). It is possible to explain this, in terms of the Wolfe model (WOLFE and SHIELDS 1998; KEOGH et al. 1998 ), by supposing that homologous blocks competed for survival within the genome as it reduced in size following the ancient duplication event. Duplicated segments (CHRs) that contained genes that (through mutation or recombination) developed new, and selectively advantageous, functions would be more likely to survive. Mutations in genes that determine such functions are likely to produce some overt phenotype that would allow their detection by classical techniques.

It was OHNO 1970 who first pointed out the importance of gene duplication in evolution. One member of a duplicate pair of genes is freed, by the existence of the alternate copy, from the immediate demands of selection. It may thus evolve some novel function that may later be of selective advantage. It is probable that OHNO was thinking of novel functions that arose from changes in the biochemical activity or specificity of gene products. In the case of protein-encoding genes, such changes would arise due to mutation or recombination events involving their ORFs. Such changes might also permit the protein product of a gene to function in a novel cellular context, for instance, by adding or deleting a signal sequence that directs the protein to a particular cellular compartment. For instance, the CIT2 gene encodes the peroxisomal form of the citrate synthase enzyme, whereas CIT1 and CIT3 encode mitochondrially located forms of the same enzyme (LALO et al. 1993 ). Alternatively, alterations in the region of the gene upstream of its ORF may lead the two copies of a gene to be differently regulated (as is the case with CRY1 and CRY2, in which the second copy of the gene is expressed only when the first copy is deleted; LI et al. 1995 ). Finally, the increase in copy number can fulfill a demand for higher levels of gene expression, as is likely the case for a number of duplicated genes encoding ribosomal proteins (WARNER 1989 ).

The S. cerevisiae genome contains six telomere-associated ORFs whose predicted protein products are very similar (>85% amino-acid sequence identity) to aryl-alcohol dehydrogenase (AAD; MUHEIM et al. 1991 ; REISER et al. 1994 ) of the lignin-degrading fungus Phanerochaetae chrysosporium (DELNERI et al. 1999 ). A seventh ORF, on the left arm of chromosome XVI, is not related to the other six at the nucleotide-sequence level, but its predicted protein product shows an equivalent level of similarity to the fungal aryl-alcohol dehydrogenase. The discovery of this seven-member gene set in the S. cerevisiae genome stimulated us to determine whether the constituent genes were truly redundant, perhaps having only quantitative effects on phenotype, or whether there were differences in the control of their expression or in the substrate specificities of their gene products. We have constructed a series of single and multiple deletion mutants of the AAD gene set to perform phenotypic analyses (DELNERI et al. 1999 ). Here, we report an analysis, at both the informatic and experimental levels, of the relationships between the yeast AAD genes and their modes of expression. We demonstrate that the expression of two members of the gene set responds to the oxidative stress induced by glutathione antagonists in a Yap1p-dependent manner. It is likely that only one of these two genes encodes a functional protein product, suggesting that, in terms of the response to oxidative stress, the six- or sevenfold redundancy of the AAD gene set is more apparent than real.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and media:
The S. cerevisiae strains YPH499 (MATa ura3-52 lys2-801^amber ade2-101^ochre trp1-63 his3-200 leu2-1) and YPH500 (MAT ura3-52 lys2-801^amber ade2-101^ochre trp1-63 his3-200 leu2-1) were obtained from Phil Hieter (University of British Columbia; SIKORSKI and HIETER 1989 ). The collection of AAD strains containing single or multiple deletions in the seven-member AAD gene set was constructed via PCR-mediated gene replacement as described by DELNERI et al. 1999 . The yeast strain yap1 was obtained from Cristina Merlotti (University of Manchester). For routine culture, S. cerevisiae was grown on 2% yeast extract, 1% peptone, and 2% glucose (YPD). Glucose minimal medium (SD) contained 0.67% yeast nitrogen base (Difco Laboratories, Detroit), 0.5% ammonium sulfate, and 2% glucose. Amino acids or nucleic acid bases were added where required at the concentration of 20 mg/liter.

Informatic tools:
Homology searches were carried out on DNA sequences against the EMBL and GenBank databases using BLAST e-mail server facility (ALTSCHUL et al. 1990 ); amino-acid sequence homologies were identified using the nonredundant version of the SWISS-PROT database, the translated version of the EMBL nucleotide sequence database (TREMBL), and the FASTA search program (PEARSON and LIPMAN 1988 ). DNA and protein alignments were obtained using the GCG-bestfit package (DEVEREUX et al. 1984 ). To design evolutionary trees the TreeView program was used (PAGE 1996 ).

Preparation of total RNA:
Yeast cells were cultured under a number of different physiological conditions to assess AAD gene expression:

1. Noninduced: yeast were cultured in YPD or SD at 30° until OD₆₀₀ = 0.7.

2. Hydrogen peroxide shock: yeast cells were cultured in YPD or SD at 30° until OD₆₀₀ = 0.5, then hydrogen peroxide was added to a final concentration of either 0.4 or 1 mM. The cultures were allowed to grow for a further hour, following the shock.

3. General oxidative shock: yeast cells were cultured in YPD or SD at 30° until OD₆₀₀ = 0.5, then either diamide or DEME was added at a final concentration of 4.5 or 6 mM, respectively. The cultures were allowed to grow for a further hour following the addition of these agents.

Total RNA was prepared as described by DUTTWEILER and GROSS 1998 . Briefly, the harvested cells were resuspended in "multifunctional" buffer (phenol, formamide, EDTA, HEPES, aurintricarboxylic acid) containing glass beads (0.5 mM). The yeast cells were then broken by vigorous shaking in a vortex mixer for 3 min. After centrifugation, the supernatant containing the total RNA was recovered and heated at 65° for 5 min to denature the RNA.

Northern hybridization:
Plasmid G5 (PETES et al. 1978 ), containing a complete yeast ribosomal DNA repeat unit, was used as a probe to monitor variations in loading of total RNA extracts between lanes. All other DNA probes were synthesized by PCR amplification. A set of primers for each ORF was designed as listed in Table 2. PCR products were synthesized by 30-cycle reactions (94° for 30 sec; 55° for 1 min; 72° for 30 sec) in a Techne Progene thermocycler and their size and homogeneity checked by electrophoresis in a 1.2% agarose gel with TBE buffer. The average length of the PCR products was 200 bp and ~50 µg of PCR product was labeled with [³²P]dCTP by the random-priming method using the Rediprime random primer labeling kit (Amersham, Little Chalfont, U.K.) according to the manufacturer's instructions. Probes were purified using Stratagene NucTrap purification columns.

Northern hybridization was carried out according to the method of ENGLER-BLUM et al. 1993 with some modifications. Normally, 20 µg of total RNA was loaded onto a 1% agarose gel containing 17% formaldehyde and separated by electrophoresis for 2–3 hr at 2 V/cm, using 3-[N-morpholino]propane-sulfonic acid (MOPS, pH 7.2) as the running buffer. The RNA was transferred to a positively charged nylon membrane (Boehringer Mannheim, Mannheim, Germany) by capillary transfer overnight with 20x SSC. The RNA was then fixed to the membrane by UV irradiation for 4 min. The membrane was prehybridized for 1 hr at 62° with a solution of Na₂HPO₄ (0.25 mM), EDTA (1 mM), SDS (20%), and blocking reagent (0.5%). This solution was replaced with the same solution containing the radiolabeled probe and was left to hybridize overnight at 62°. Membranes were washed three times for 20 min at 62° with 50 ml of prewarmed wash buffer containing 1% SDS, 20 mM Na₂HPO₄, and 1 mM EDTA.

The amount of radioactivity in a hybridization band was quantified using a molecular imager system (GS-363; Bio-Rad, Richmond, CA). The relative abundance of the AAD transcripts was determined by comparison with the amount of the rRNA in the same lane.

DNA sequencing:
Sequencing of genomic copies of AAD6 and AAD10 was conducted in the manner described by JAMES et al. 1995 using an ABI sequencing system. PCR products were amplified and sequenced from total yeast DNA. These products were generated using the following primers: for AAD6, Fp6 5'-taagaaatacgacgttggtg-3', Rp6 5'-ctctacgacttgtgtatgatttc-3'; for AAD10, Fp10 5'-acccaatatgattctcaccttc-3', Rp10 5'-ttgcctacatcatacccctta-3'. For each ORF, two sets of templates were prepared independently and the sequencing was done in duplicate.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Informatic analysis:
All seven ORFs (YNL331c, YDL243c, YCR107w, YJR155w, YFL056, YFL057c, YOL165c) were found to exhibit significant similarity to the oxidoreductase family of proteins, with the highest similarity (~70% sequence identity) to the AAD of the lignin-degrading fungus P. chrysosporium (MUHEIM et al. 1991 ; REISER et al. 1994 ) and the norsolinic acid reductase (NAR) involved in aflatoxin B1 biosynthesis in Aspergillus parasiticus (CARY et al. 1996 ; Table 3). For this reason, we have given these ORFs the gene name AAD and designated the gene number according to the chromosome on which the ORF is found; e.g., YNL331c is named AAD14. AAD6 corresponds to ORFs YFL056c and YFL057c, which are displaced from one another by -1 frameshift (see below). The AAD gene that is not telomere asscociated (YPL088) was found to have ~40% similarity at the protein level with AAD and NAR, but had no significant homology, at the nucleotide-sequence level, with the six members of the S. cerevisiae AAD gene family. Thus, we believe that AAD16 is not, sensu stricto, a member of the S. cerevisiae AAD gene family, even though there is a strong possibility that it encodes an oxidoreductase. Accordingly, we refer to the seven-member AAD gene set (including AAD16) and the six-member AAD gene family (excluding the nontelomeric gene).

Only the protein products of AAD14, AAD4, and AAD3 have a length comparable to that of the P. chrysosporium AAD enzyme. YFL056c and YFL057c are two contiguous small ORFs of about half the size of the P. chrysosporium AAD gene and correspond, respectively, to the N-terminal and C-terminal halves of the AAD polypeptide. These two small ORFs, are displaced by a -1 frameshift between nucleotides 14914 and 14915 in the chromosome VI sequence (MURAKAMI et al. 1995 ). The insertion of a G residue at this position creates a single ORF, YFL056-57c, which has the ATG start codon of ORF YFL056c and the TAA of ORF YFL057c as its stop codon. The ORF YJR155w (AAD10) is preceded by a stop codon at position 727057 in the complete sequence of chromosome X (GALIBERT et al. 1996 ). The presence of this stop codon produces an N-terminal truncation, reducing the size of the predicted gene product by ~23% as compared to that of the P. chrysosporium AAD polypeptide. Removal of this stop codon (to produce ORF YJR155+w) extends the region of amino-acid sequence similarity between Aad10p and the fungal enzyme by some 100 amino acids upstream of the putative ATG of ORF YJR155w and results in a predicted protein product of similar size to that of the P. chrysosporium AAD.

Both the frameshift between ORFs YFL056c/YFL057c and the stop codon within the YJR155+w may have been the result of sequencing errors, so both DNA regions were resequenced using, for each chromosome, two independently generated PCR products produced by amplifying genomic DNA from strain YPH499 (see MATERIALS AND METHODS). The new sequencing data confirmed the previous results, suggesting that AAD6 and AAD10 may not be functional genes. In addition to YJR155w, another ORF, YOL165c, has an N-terminal deletion corresponding to 61% of the P. chrysosporium AAD polypeptide. However, in this case, we can find no ORF with apparent similarity to the fungal AAD enzyme, either upstream of the start codon of YOL165c or downstream from its stop codon. It is, therefore, unlikely that AAD15 specifies a functional oxidoreductase.

The telomere-associated members (AAD3, AAD4, AAD6, AAD10, AAD14, and AAD15) likely form a true gene family, while AAD16 is more distantly related. Comparison of the DNA sequences of the promoter and the terminator regions of each member of the gene family reveals that the region upstream of AAD15 is not similar to the region upstream of other family members, which all display a significant degree of sequence homology to each other in these two DNA regions. This homology extends for ~500 bp upstream of the ATG and 2000 bp downstream of the stop codon. The downstream homology regions include a second ORF, which in the case of AAD6 (ORF YFL056/YFL057c) is a gene of known function, THI5 (YFL058c) (Figure 1). The ORFs downstream of AAD4, AAD10, and AAD14 are all transcribed from the opposite DNA strand and all appear to be homologues of THI5 (MURAKAMI et al. 1995 ). In fact, YNL332w has been identified as THI12 (VAN DYCK et al. 1995 ). No such homologue can be found downstream of AAD3 (YCR107w) since it is the last ORF on the right arm of chromosome III (OLIVER et al. 1992 ). AAD15 has a downstream ORF that is transcribed from the same DNA strand and is not a THI5 homologue. Other evidence leads us to propose that AAD3 and AAD15 are closely related (see below).

View larger version (33K): In this window In a new window Download PPT slide   Figure 1. Location of the AAD (solid) and THI5 (shaded) homologues in the yeast genome.

This analysis suggests that there has been a duplication of a telomeric region of ~4000 bp between chromosomes III, IV, VI, X, and XIV. It is not clear whether this duplication occurred before the postulated complete duplication of the yeast genome (WOLFE and SHIELDS 1997 ) or whether subsequent exchanges between chromosome ends have been involved (OLIVER et al. 1992 ; OLIVER 1995 ). When we compared the intergenic sequences between the AAD and THI5 homologues (~400 bp), the highest degree of nucleotide sequence identity was found for the intergenic regions on chromosomes IV and VI and for those on chromosomes X and XIV. The 500-bp region downstream of AAD3 (chromosome III) showed the highest level of sequence identity with the region downstream of AAD15 on chromosome XV. Thus, we conclude that the CHRs containing the six members of the yeast AAD gene family found on chromosomes IV, X, and XV derive, respectively, from the equivalent regions on chromosomes VI, XIV, and III (or vice versa). These pairwise relationships are also evident from the amino-acid sequence comparisons of the postulated protein products of these six genes (Figure 2).

View larger version (9K): In this window In a new window Download PPT slide   Figure 2. Evolutionary divergence among duplicated regions of ~1500 bp including the S. cerevisiae AAD genes (~1100 bp) and the intergenic sequences between the AAD and THI5 homologues (~400 bp).

AAD transcription responds to oxidative stress:
To determine whether the seven members of the AAD gene set were functionally redundant, we decided to examine whether or not they were coordinately regulated. As a first step, yeast grown under a number of different physiological conditions were examined for the presence of AAD gene transcripts. These conditions were designed to include situations (such as stationary phase, nitrogen starvation, and the presence of an aromatic aldehyde) that are relevant to the role of the AAD enzyme in P. chrysosporium (MUHEIM et al. 1991 ). In addition, we considered that the role of an AAD enzyme in S. cerevisiae (an organism that does not degrade lignocellulose) was likely to be associated with the reduction of potentially damaging oxidizing agents released into its environment by lignin degraders such as Phanerochaete. Accordingly, we examined the response of AAD gene expression to the oxidative stress produced by the glutathione antagonists, diamide and diethyl maleic acid (DEME). Finally, we examined an alternate stress condition (hyperosmotic shock) to discover whether AAD gene expression formed part of yeast's general stress response. Under none of these conditions, apart from addition of diamide and DEME (Figure 3A), could any AAD transcripts be detected. In particular, it should be noted that the oxidizing agent, H₂O₂, failed to stimulate AAD transcription either in rich or minimal medium (Figure 3B, see DISCUSSION).

View larger version (51K): In this window In a new window Download PPT slide   Figure 3. Northern blot probed with a 200-bp fragment of ORF YJR155w (top) and a fragment of the ribosomal array containing the 5S and the 35S genes as control (bottom). (A) RNA samples were taken 30, 60, 90, and 120 min after the addition of either diamide or DEME. (B) RNA samples were taken 60 min after the addition of either DEME (positive control), or 1 mM H2O2 or 0.4 mM H2O2.

Having detected the presence of AAD mRNA under oxidative stress conditions, it was possible to use our collection of aad deletion mutants (DELNERI et al. 1999 ) to validate this result and to examine the question of the differential regulation of gene expression between different members of the AAD gene set. AAD gene expression in response to oxidative stress was examined in the wild-type strain (YPH499), in a septuple mutant strain (7x) in which all the AAD ORFs had been deleted, and in single-gene deletants for each member of the AAD set. Northern hybridizations were carried out using two types of probe, a cloned copy of AAD10 and an AAD16 clone. Control experiments demonstrated that the AAD10 clone would detect transcripts from any other member of the gene family (Figure 4), but not transcripts from AAD16, whereas the AAD16 probe was specific for the detection of AAD16 mRNAs. The use of these two probes demonstrated that AAD16 was not transcribed in response to the addition of DEME and diamide, whereas a strong mRNA band of ~1.3 kb was detected using the AAD10 probe in both the wild-type strain and all single aad mutants.

View larger version (61K): In this window In a new window Download PPT slide   Figure 4. Northern blot probed with a 200-bp fragment of ORF YJR155w (top) and a fragment of the ribosomal array containing the 5S and the 35S genes as control (bottom). RNA samples from the wild-type strain, from the aad septuple mutant and from three different aad single mutants (ycr107c, yjr155w, and ynl331c) were taken 60 min after the introduction of DEME. Note that a strong RNA signal can be detected in the mutant lacking of the ORF YJR155w, suggesting that the 200-bp probe is not specific for the ORF YJR155w, but recognized also RNA from other AAD genes.

Stress-dependent AAD transcription requires a functional YAP1 gene:
The oxidative stress response in S. cerevisiae is mediated by the transcriptional activator, Yap1p (MOYE-ROWLEY et al. 1989 ), and so the role of this protein in the oxidative stress response of the AAD gene family was examined by repeating the Northern analysis in a yap1 deletant strain (YPH250, yap1). The results (Figure 5) demonstrated that AAD transcription in response to either DEME or diamide was absolutely dependent on YAP1. We searched the sequences of the AAD genes for close matches to the site to which Yap1p binds (TTACTAA; KUGE and JONES 1994 ). There is a Yap1-binding site upstream of ORFs YDL243c (AAD4) and YFL056-57c (AAD6). If the AAD gene expression upon oxidative stress is Yap1p dependent, then deletion of the two ORFs carrying the Yap1p-binding site in their promoter regions should lead to the disappearance of the signal corresponding to the AAD transcripts.

View larger version (91K): In this window In a new window Download PPT slide   Figure 5. Northern blot probed with a 200-bp fragment of ORF YJR155w (top) and a fragment of the ribosomal array containing the 5S and the 35S genes as control (bottom). RNA from the wild-type strain and from the yap1 strain were taken 60 min after the introduction of either DEME or 1 mM H2O2. Note that RNA samples from yap1 strain were overloaded by five times compared to W.T.

No AAD transcript was detected in the mRNA sample from the double mutant aad4 aad6, while it was present in the samples from other multiple mutants as long as either AAD4 or AAD6 was still present in the genome (Figure 6). This suggests that only two genes in the seven-member AAD gene set are expressed during oxidative stress. Because the AAD6 ORF is nonfunctional, only one of these two genes is likely to encode a functional product. Plate assays were performed to monitor sensitivity to DEME. The septuple mutant showed increased resistance to this agent. This result may suggest a role for the AAD gene set in "gating" the response to oxidative stress. However, no difference was seen between the wild-type response to DEME and that of the quadruple mutant, aad3 aad14 aad6 aad10 (in which only the marker KanMX was used for the gene deletion). Moreover, similar results were found when comparing the wild-type and the double mutant aad4 aad6. Given that nutritional markers were used to make the gene replacements for AAD16, AAD4, and AAD15 in the septuple mutant (DELNERI et al. 1999 ), these data mean that we cannot exclude the possibility that a marker effect (BAGANZ et al. 1997 ) is at least partly responsible for the observed change in resistance to DEME.

View larger version (45K): In this window In a new window Download PPT slide   Figure 6. Northern blot probed with a 200-bp fragment of ORF YJR155w (top) and a fragment of the ribosomal array containing the 5S and the 35S genes as control (bottom). (A) RNA samples from the wild-type and different mutants were analyzed 60 min after the introduction of DEME. (B) RNA samples from the wild-type and the double mutant aad4 aad6 (ydl243c and yfl056/57c) were taken 60 min after the introduction of either DEME or 1 mM H2O2.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The very high level of apparent genetic redundancy (MEWES et al. 1997 ; WOLFE and SHIELDS 1997 ) within a eukaryotic genome that is less than four times the size of that of the bacterium E. coli (BLATTNER et al. 1997 ) was probably one of the most unexpected results to emerge from the complete sequence determination of the S. cerevisiae genome (GOFFEAU et al. 1996 ). On general evolutionary grounds, much of this redundancy can be expected to be more apparent than real (MAYNARD SMITH 1989 ). Thus, many genes with identical, or nearly identical, protein products can be expected to play different biological roles by virtue of their being expressed in distinct physiological or developmental contexts or their protein products being directed to separate subcellular compartments (see Introduction). In some instances, members of a gene family can be expected to be truly redundant at the functional level, in which case each would make some fractional contribution to the organism's phenotype that only quantitative methods of analysis will reveal (OLIVER 1996A ). Finally, we must be aware that the complete yeast genome sequence gives us only a snapshot taken at a particular stage of this organism's evolutionary history and that some genes may be part way to evolving some novel function (during a time in which they are free from the constraints of selection; OHNO 1970 ), while still others (like, perhaps, AAD10 and AAD15) may be on their way to elimination from the genome. Whatever the explanation of genetic redundancy in any particular case, it represents a significant practical problem for the experimentalist and demands that we analyze all genes in a family or set together to reveal the functionality of individual members.

We decided to investigate the AAD gene set for a number of reasons. First, at seven members, it is one of the largest sets in the S. cerevisiae genome. Second, the set contains a family of six genes that exhibit a high degree of internal homology (>80% amino-acid sequence identity) and might thus be expected to exhibit genuine functional redundancy. Finally, although yeast is not a lignin degrader, our studies have confirmed that it does exhibit an aryl-alcohol dehydrogenase activity since stationary phase cultures can convert aromatic aldehydes into their corresponding alcohols (DELNERI et al. 1999 ).

YFL056c, YFL057c, and YOL165c are unlikely to be functional genes, since a frameshift splits the protein encoded by YFL056c and YFL057c, and YOL165c carries a deletion of ~60% of its 5' end. The YJR155w ORF is truncated by a stop codon located 228 bp downstream of the ATG. We do not know whether this deletion compromises AAD10 function. Moreover, it is possible that the effects of the stop codon are ameliorated by translational read-through; the UAG terminator could be decoded as glutamine by the normal Glu tRNA using a first-position wobble interaction (P. FARABAUGH, personal communication). Indeed, the UAG codon in ORF YJR155+w occurs at a site occupied by a glutamine codon in the AAD reading frames.

The six members of the AAD gene family comprise three pairs (AAD3 + AAD15, AAD6 + AAD4, AAD10 + AAD14, Figure 2) whose two genes are more related to one another than to the other four members of the family. The six members of the AAD gene family are contained within subtelomeric CHRs, and the IV-L, X-R, and XV-L segments derive, respectively, from the CHRs on chromosomes VI-L, XIV-L, and III-R (or vice versa). This could indicate that there were already three members of the AAD family prior to the duplication of the yeast genome postulated by WOLFE and SHIELDS 1997 . Alternatively, the six AAD-containing CHRs may be of more recent origin, being the products of a series of terminal translocations and rounds of mating. An answer to this evolutionary question may emerge from the genome sequencing of Saccharomyces species that originated immediately before or after the ancient duplication event. To understand the place of the AAD gene family in the contemporary S. cerevisiae genome, however, we need to discover the functions of the Aad proteins and determine whether they each have equivalent or distinctive roles in yeast physiology and development.

In considering the likely biological role of any novel gene revealed by systematic genome sequencing, the first recourse is to informatic analysis (e.g., MEWES et al. 1997 ). The yeast AAD gene set displays significant amino-acid sequence identity to a number of fungal and bacterial oxidoreductases whose sequences are in the public data libraries. The highest similarity was found with the P. chrysosporium aryl-alcohol dehydrogenase, an enzyme involved in the reduction of the aromatic aldehydes generated during the last step of lignin biodegradation. While S. cerevisiae does not degrade lignin itself, it may share an ecological niche (such as a rotting fig) with lignin-degrading fungi and thus be exposed to, and have to protect itself against, the products of lignin biodegradation. While our biochemical investigations have revealed an aryl-alcohol dehydrogenase activity in yeast, deletion analysis demonstrated that it was not due to the AAD gene set (DELNERI et al. 1999 ). However, the fungal aromatic alcohol oxidoreductases show extreme stereospecificity (B. HAHN-HAGERDAHL, personal communication) and thus we may not yet have found the appropriate substrate for the Aad enzymes of S. cerevisiae.

It is well established that exposure of yeast cells to agents such as H₂O₂, diamide, and DEME causes the expression of various oxidoreductase genes via the action of the Yap1 transcription factor (KUGE and JONES 1994 ; STEPHEN et al. 1995 ). Following an oxidative shock, the Yap1 protein (which is usually located in the cytosol) moves rapidly into the nucleus and binds to the promoter regions of its target genes (KUGE et al. 1997 ). The use of a yap1 strain enabled us to demonstrate that members of the AAD gene family are targets for the action of Yap1p. However, the Yap1-mediated expression of AAD genes only responded to the indirect oxidative stress mediated by anti-glutathione agents, such as DEME. Oxidizing agents, such as H₂O₂, failed to evoke AAD gene expression. Recent studies (FERNANDES et al. 1997 ; WEMMIE et al. 1997 ) have shown that the Yap1 protein contains two domains, one of which is required for the response to oxidizing agents and the other mediates the response to the glutathione antagonists. The fact that the AAD genes respond to only one type of oxidative stress in a Yap1-dependent manner may indicate that the bipartite Yap1 protein is able to activate transcription of some of its target genes in a way that differentiates between the two main classes of oxidative stress. However, it cannot be excluded that the depletion of glutathione is, itself, a necessary factor for the activation of AAD transcription.

It is not possible to distinguish any differences in the expression of the six members of the AAD gene family in response to DEME addition because the nucleotide sequences of these genes are too similar to allow any single-ORF probe to distinguish between them. Therefore, we turned to our collection of single and multiple aad deletion mutants (DELNERI et al. 1999 ) to elucidate the nature of the response of individual family members. This genetic approach enabled us to establish that the only two genes expressed in the presence of diamide or DEME were AAD4 and AAD6. Both genes contain a perfect match to the consensus Yap1-binding site in their promoter regions.

Hybridization array technology has been used to try to discover all the genes under Yap1p control (DERISI et al. 1997 ). The transcriptional activator was overexpressed and all genes showing a significant increase in their level of transcription were identified. These included AAD15, AAD14, AAD10, and AAD6, but not (curiously) AAD4. Given the degree of nucleotide sequence similarity between the members of the AAD gene family, it is not possible to distinguish between them using PCR products amplified from their respective ORFs. This result is presumably due to either variations in the amounts of PCR products arrayed or uneven hybridization across the glass slide. In contrast, a genetic approach using a collection of mutants, in which one or more members of the AAD gene set were deleted, enabled the unambiguous identification of which genes in the set are targets for the Yap1 factor since no AAD transcripts were detectable in aad4 aad6 double deletants.

Given the fact that AAD6 contains a frameshift in the middle of its ORF, the AAD4 gene would appear to be the only member of the gene set that expresses a functional protein in response to oxidative shock. It is possible that other members of the AAD family are transcribed under different physiological conditions that we have yet to investigate. What is clear from our study is that, as far as the oxidative stress response is concerned, the apparent sevenfold redundancy of the AAD gene set reduces to a single functional gene, AAD4. We suggest that this is unlikely to be a unique case and that the extensive genetic redundancy revealed by the complete genome sequence of S. cerevisiae is more apparent than real.

	ACKNOWLEDGMENTS

We thank Phil Farabaugh, Barbara Hahn-Hagerdahl, Derek Jamieson, and Linda Partridge for useful discussions; Andrew Hayes, Gregory Tomlin, Cristina Merlotti, and Simon Hubbard for their help and advice; and Carlo Bruschi for his encouragement. This work was supported by a Training and Mobility of Researchers Fellowship from the European Commission (EC) to D.D. and by a contract under the European Functional Analysis Network programme of the EC to S.G.O.

Manuscript received April 27, 1999; Accepted for publication August 13, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MYERS, and D. J. LIPMAN, 1990  Basic local alignment search tool. J. Mol. Biol. 215:403-410[Medline].

BAGANZ, F., A. HAYES, D. MARREN, D. C. J. GARDNER, and S. G. OLIVER, 1997  Suitability of replacement markers for functional analysis studies in Saccharomyces cerevisiae.. Yeast 13:1563-1573[Medline].

BLATTNER, F. R., G. PLUNKETT, III, C. A. BLOCH, N. T. PERNA, and V. BURLAND et al., 1997  The complete genome sequence of Escherichia coli K-12. Science 277:1453-1474[Abstract/Free Full Text].

BUSSEY, H., D. B. KABACK, W. ZHONG, D. T. VO, and M. W. CLARK et al., 1995  The nucleotide sequence of chromosome I from Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 92:3809-3813[Abstract/Free Full Text].

CARLSON, M., J. L. CELENZA, and F. J. ENG, 1985  Evolution of the dispersed SUC gene family of Saccharomyces by rearrangements of chromosome telomeres. Mol. Cell. Biol. 5:2894-2902[Abstract/Free Full Text].

CARY, J. W., M. WRIGHT, D. BHATNAGAR, R. LEE, and F. S. CHU, 1996  Molecular characterization of an Aspergillus parasiticus dehydrogenase gene, norA, located on the aflatoxin biosynthesis gene cluster. Appl. Environ. Microbiol. 62:360-366[Abstract].

CHARRON, M. J., E. READ, S. R. HAUT, and C. A. MICHELS, 1989  Molecular evolution of the telomere-associated MAL loci of Saccharomyces.. Genetics 122:307-316[Abstract/Free Full Text].

DELNERI, D., D. C. J. GARDNER, C. V. BRUSCHI, and S. G. OLIVER, 1999  Disruption of seven hypothetical aryl-alcohol dehydrogenase genes from Saccharomyces cerevisiae and construction of a multiple knock-out strain. Yeast 15:1681-1689[Medline].

DERISI, J. L., V. R. IYER, and P. O. BROWN, 1997  Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680-686[Abstract/Free Full Text].

DEVEREUX, J., P. HAEBERLI, and O. SMITHIES, 1984  A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395.

DUJON, B., 1996  The yeast genome project: What did we learn? Trends Genet. 12:263-270[Medline].

DUJON, B., D. ALEXANDRAKI, B. ANDRE, W. ANSORGE, and V. BALADRON et al., 1994  Complete DNA sequence of yeast chromosome XI. Nature 369:371-378[Medline].

DUTTWEILER, H. M. and D. S. GROSS, 1998  Bacterial growth medium that significantly increases the yield of recombinant plasmid. Biotechniques 24:438-444[Medline].

ENGLER-BLUM, G., M. MEIER, J. FRANK, and G. A. MULLER, 1993  Reduction of background problems in nonradioactive Northern and Southern blot analyses enables higher sensitivity than ³²P-based hybridizations. Anal. Biochem. 210:235-244[Medline].

FERNANDES, L., C. RODRIGUES-POUSADA, and K. STRUHL, 1997  Yap, a novel family of eight bZIP proteins in Saccharomyces cerevisiae with distinct biological functions. Mol. Cell. Biol. 17:6982-6993[Abstract].

GALIBERT, F., D. ALEXANDRAKI, A. BAUR, E. BOLES, and N. CHALWATZIS et al., 1996  Complete nucleotide sequence of Saccharomyces cerevisiae chromosome X. EMBO J. 15:2031-2049[Medline].

GOFFEAU, A., B. G. BARRELL, H. BUSSEY, R. W. DAVIS, and B. DUJON et al., 1996  Life with 6000 genes. Science 274:546-567[Abstract/Free Full Text].

JAMES, C. M., K. J. INDGE, and S. G. OLIVER, 1995  DNA sequence analysis of a 35 kb segment from Saccharomyces cerevisiae chromosome VII reveals 19 open reading frames including RAD54, ACE1/CUP2, PMR1, RCK1, AMS1 and CAL1/CDC43. Yeast 11:1413-1419[Medline].

KEOGH, R. S., C. SEOIGHE, and K. H. WOLFE, 1998  Evolution of gene order and chromosome number in Saccharomyces, Kluyveromyces and related fungi. Yeast 14:443-457[Medline].

KUGE, S. and N. JONES, 1994  YAP1 dependent activation of TRX2 is essential for the response of Saccharomyces cerevisiae to oxidative stress by hydroperoxides. EMBO J. 13:655-664[Medline].

KUGE, S., N. JONES, and A. NOMOTO, 1997  Regulation of yAP-1 nuclear localization in response to oxidative stress. EMBO J. 16:1710-1720[Medline].

LALO, D., S. STETTLER, S. MANOTTE, P. P. SLONIMSKI, and P. THURIAUX, 1993  Two yeast chromosomes are related by a fossil duplication of their centromeric regions. C. R. Acad. Sci. 316:137-143.

LI, Z., A. G. PAULOVICH, and J. L. WOOLFORD, JR., 1995  Feedback inhibition of the yeast ribosomal protein gene CRY2 is mediated by the nucleotide sequence and secondary structure of CRY2 pre-mRNA. Mol. Cell. Biol. 15:6454-6464[Abstract].

MAYNARD SMITH, J., 1989 Evolutionary Genetics. Oxford University Press, New York.

MELNICK, L. and F. J. SHERMAN, 1993  The gene clusters ARC and COR on chromosomes 5 and 10, respectively, of Saccharomyces cerevisiae share a common ancestry. Mol. Biol. 233:372-388.

MEWES, H. W., K. ALBERMANN, M. BAHR, D. FRISHMAN, and A. GLEISSNER et al., 1997  Overview of the yeast genome. Nature 387:7-65[Medline].

MOYE-ROWLEY, W. S., K. D. HARSHMAN, and C. S. PARKER, 1989  Yeast YAP1 encodes a novel form of the jun family of transcriptional activator proteins. Genes Dev. 3:283-292[Abstract/Free Full Text].

MUHEIM, A., R. WALDNER, D. SANGLARD, J. REISER, and H. E. SCHOEMAKER et al., 1991  Purification and properties of an aryl-alcohol dehydrogenase from the white-rot fungus Phanerochaete chrysosporium.. Eur. J. Biochem. 195:369-375[Medline].

MURAKAMI, Y., M. NAITOU, H. HAGIWARA, T. SHIBATA, and M. OZAWA et al., 1995  Analysis of the nucleotide sequence of chromosome VI from Saccharomyces cerevisiae.. Nat. Genet. 10:261-268[Medline].

NAUMOV, G., H. TURAKAINEN, E. NAUMOVA, S. AHO, and M. KORHOLA, 1990  A new family of polymorphic genes in Saccharomyces cerevisiae: alpha-galactosidase genes MEL1-MEL7. Mol. Gen. Genet. 224:119-128[Medline].

OHNO, S., 1970 Evolution by Gene Duplication. Allen & Unwin, London.

OLIVER, S. G., 1995  Size is important, but. Nat. Genet. 10:253-254[Medline].

OLIVER, S. G., 1996a  From DNA sequence to biological function. Nature 379:597-600[Medline].

OLIVER, S. G., 1996b  From DNA sequence to biological function: the new "Voyage of the Beagle.". Biochem. Soc. Trans. 24:291-292[Medline].

OLIVER, S. G., Q. J. VAN DER AART, M. L. AGOSTONI-CARBONE, M. AIGLE, and L. ALBERGHINA et al., 1992  The complete DNA sequence of yeast chromosome III. Nature 357:38-46[Medline].

PAGE, R. D., 1996  TreeView: an application to display phylogenetic trees on personal computers. Comput. Appl. Biosci. 12:357-358[Free Full Text].

PEARSON, W. R. and D. J. LIPMAN, 1988  Improved tools for biological sequence comparison. Proc. Natl. Acad. Sci. USA 85:2444-2448[Abstract/Free Full Text].

PETES, T. D., L. M. HEREFORD, and K. G. SKRYABIN, 1978  Characterisation of two types of yeast ribosomal DNA genes. J. Bacteriol. 134:295-305[Abstract/Free Full Text].

REISER, J., A. MUHEIM, M. HARDEGGER, G. FRANK, and A. FIECHTER, 1994  Aryl-alcohol dehydrogenase from the white-rot fungus Phanerochaete chrysosporium: gene cloning, sequence analysis, expression, and purification of the recombinant enzyme. J. Biol. Chem. 269:28152-28159[Abstract/Free Full Text].

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].

STEPHEN, D. W. S., S. L. RIVERS, and D. J. JAMIESON, 1995  The role of the YAP1 and YAP2 genes in the regulation of the adaptive oxidative stress responses of Saccharomyces cerevisiae.. Mol. Microbiol. 16:415-423[Medline].

VAN DYCK, L., A. PASCUAL-AHUIR, B. PURNELLE, and A. GOFFEAU, 1995  An 8.2 kb DNA segment from chromosome XIV carries the RPD3 and PAS8 genes as well as the Saccharomyces cerevisiae homologue of the thiamine-repressed nMt1 gene and a chromosome III-duplicated gene for a putative aryl-alcohol dehydrogenase. Yeast 11:987-991[Medline].

WARNER, J. R., 1989  Synthesis of ribosomes in Saccharomyces cerevisiae.. Microbiol. Rev. 53:256-271[Abstract/Free Full Text].

WEMMIE, J. A., S. M. STEGGERDA, and W. S. MOYE-ROWLEY, 1997  The Saccharomyces cerevisiae AP-1 protein discriminates between oxidative stress elicited by the oxidants H₂O₂ and diamide. J. Biol. Chem. 272:7908-7914[Abstract/Free Full Text].

WICKSTEED, B. L., I. COLLINS, A. DERSHOWITZ, L. I. STATEVA, and R. P. GREEN et al., 1994  A physical comparison of chromosome III in six strains of Saccharomyces cerevisiae.. Yeast 10:39-57[Medline].

WOLFE, K. H. and D. C. SHIELDS, 1997  Molecular evidence for an ancient duplication of the entire yeast genome. Nature 387:708-713[Medline].

This article has been cited by other articles:

				R. Harrison, B. Papp, C. Pal, S. G. Oliver, and D. Delneri Plasticity of genetic interactions in metabolic networks of yeast PNAS, February 13, 2007; 104(7): 2307 - 2312. [Abstract] [Full Text] [PDF]

				M. Bekaert, H. Richard, B. Prum, and J.-P. Rousset Identification of programmed translational -1 frameshifting sites in the genome of Saccharomyces cerevisiae Genome Res., October 1, 2005; 15(10): 1411 - 1420. [Abstract] [Full Text] [PDF]

				A. Aranda and M.-l. del Olmo Exposure of Saccharomyces cerevisiae to Acetaldehyde Induces Sulfur Amino Acid Metabolism and Polyamine Transporter Genes, Which Depend on Met4p and Haa1p Transcription Factors, Respectively Appl. Envir. Microbiol., April 1, 2004; 70(4): 1913 - 1922. [Abstract] [Full Text] [PDF]

				R. Wightman and P. A. Meacock The THI5 gene family of Saccharomyces cerevisiae: distribution of homologues among the hemiascomycetes and functional redundancy in the aerobic biosynthesis of thiamin from pyridoxine Microbiology, June 1, 2003; 149(6): 1447 - 1460. [Abstract] [Full Text] [PDF]

				J. R. Dickinson, L. E. J. Salgado, and M. J. E. Hewlins The Catabolism of Amino Acids to Long Chain and Complex Alcohols in Saccharomyces cerevisiae J. Biol. Chem., February 28, 2003; 278(10): 8028 - 8034. [Abstract] [Full Text] [PDF]

				J. R. Dickinson, S. J. Harrison, J. A. Dickinson, and M. J. E. Hewlins An Investigation of the Metabolism of Isoleucine to Active Amyl Alcohol in Saccharomyces cerevisiae J. Biol. Chem., April 6, 2000; 275(15): 10937 - 10942. [Abstract] [Full Text] [PDF]