Specialization of Function Among Aldehyde Dehydrogenases: The ALD2 and ALD3 Genes Are Required for ß-Alanine Biosynthesis in Saccharomyces cerevisiae

Corresponding author: Jeremy H. Toyn, Bristol-Myers Squibb, Experimental Station, E400/3227, Wilmington, DE 19880., jeremy.toyn{at}bms.com (E-mail)

Communicating editor: A. P. MITCHELL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The amino acid ß-alanine is an intermediate in pantothenic acid (vitamin B₅) and coenzyme A (CoA) biosynthesis. In contrast to bacteria, yeast derive the ß-alanine required for pantothenic acid production via polyamine metabolism, mediated by the four SPE genes and by the FAD-dependent amine oxidase encoded by FMS1. Because amine oxidases generally produce aldehyde derivatives of amine compounds, we propose that an additional aldehyde-dehydrogenase-mediated step is required to make ß-alanine from the precursor aldehyde, 3-aminopropanal. This study presents evidence that the closely related aldehyde dehydrogenase genes ALD2 and ALD3 are required for pantothenic acid biosynthesis via conversion of 3-aminopropanal to ß-alanine in vivo. While deletion of the nuclear gene encoding the unrelated mitochondrial Ald5p resulted in an enhanced requirement for pantothenic acid pathway metabolites, we found no evidence to indicate that the Ald5p functions directly in the conversion of 3-aminopropanal to ß-alanine. Thus, in Saccharomyces cerevisiae, ALD2 and ALD3 are specialized for ß-alanine biosynthesis and are consequently involved in the cellular biosynthesis of coenzyme A.

PANTOTHENIC acid (vitamin B₅) and ß-alanine are intermediates in coenzyme A (CoA) biosynthesis. In bacteria, pantothenic acid is synthesized by the condensation of pantoate, an intermediate in valine biosynthesis, with ß-alanine, produced by the decarboxylation of L-aspartate (WILLIAMSON and BROWN 1979 ; CRONAN 1980 ; JACKOWSKI 1996 ). In yeast, the derivation of pantoate involves the same enzymatic steps as in bacteria, while ß-alanine biosynthesis differs from that and is dependent upon polyamine degradation (mediated by the SPE genes) and upon the amine oxidase encoded by FMS1 (WHITE et al. 2001 ). Amine oxidases can degrade polyamines with the production of the aldehyde compound 3-aminopropanal (HOLTTA 1977 ; LARGE 1992 ), implying that further oxidation of 3-aminopropanal by an aldehyde dehydrogenase would also be required for ß-alanine biosynthesis in yeast.

The complete yeast genome encodes seven different members of the "nonspecific" aldehyde dehydrogenase family (WANG et al. 1998 ); see Table 1. ALD2 and ALD3 encode closely related cytosolic enzymes that are induced on ethanol media or in response to stress (NAVARRO-AVINO et al. 1999 ). ALD4 encodes the major K⁺-dependent mitochondrial enzyme (JACOBSON and BERNOFSKY 1974 ; TESSIER et al. 1998 ), and ALD5 encodes a minor K⁺-dependent mitochondrial enzyme that is induced on ethanol (KURITA and NISHIDA 1999 ). ALD6 encodes a Mg²⁺-activated cytosolic enzyme (DICKINSON 1996 ; MEADEN et al. 1997 ). MSC7/YHR039c encodes a protein with homology to aldehyde deydrogenases that affects meiotic sister-chromatid recombination (THOMPSON and STAHL 1999 ), and YMR110c is a hypothetical open reading frame that could code for an aldehyde-dehydrogenase-related protein.

Despite the multiple genes, only one physiological substrate, acetaldehyde, has been identified. Ald4p and Ald6p function in the production of acetate from acetaldehyde, a key intermediate during fermentation of sugars as well as during growth on ethanol, and are consequently important for acetyl-CoA production (DICKINSON 1996 ; MEADEN et al. 1997 ; WANG et al. 1998 ; TESSIER et al. 1999 ; REMIZE et al. 2000 ). In contrast, Ald2p, Ald3p, and Ald5p do not contribute to the oxidation of acetaldehyde in vivo (WANG et al. 1998 ; NAVARRO-AVINO et al. 1999 ; REMIZE et al. 2000 ). The double-deletion mutant (ald2 ald3) has been reported to grow slowly on ethanol, suggesting that Ald2/3p has a function under those conditions (NAVARRO-AVINO et al. 1999 ). An undefined role for Ald5p in heme biosynthesis has been prosposed, based on defective mitochondrial electron transport and the lack of cytochromes in the mitochondria of an ald5 deletion strain (KURITA and NISHIDA 1999 ). Consequently, it remains unclear what role these enzymes play in biosynthetic pathways or in the detoxification of exogenous aldehydes.

In this study, we present evidence that ALD2 and ALD3 are specifically required for the conversion of 3-aminopropanal to ß-alanine in the metabolic pathway leading to pantothenic acid and coenzyme A. Despite a significant degree of amino acid conservation, none of the other aldehyde dehydrogenase genes play a role in ß-alanine production. These findings suggest that the "nonspecific" aldehyde dehydrogenases are functionally specialized to carry out different roles in cellular biosynthesis and therefore in fact have specific and differentiated biochemical functions.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains and gene deletions:
The parental yeast haploid, BY4742, and its gene deletion strain derivatives from the Saccharomyces Deletion Project (SGD; WINZELER et al. 1999 ) were obtained from Research Genetics (Huntsville, AL): BY4742 (MAT his3 leu lys2 ura3), BY4742-10753 (ald2), BY4742-10752 and BY4742-16071 (ald3), BY4742-11671 (ald4), BY4742-10213 (ald5), BY4742-12767 (ald6), BY4742-11002 (yhr039c), BY4742-16550 (ymr110c), BY4742-10595 (fms1), BY4742-13316 (ecm31), and BY4742-12304 (pan6 = yil145c). The corresponding homozygous diploid deletion strains were obtained from the same source. The identities of the ald strains were confirmed by PCR amplification of genomic DNA using oligonucleotides designed for this purpose by the SGD (http://www-sequence.stanford.edu/group/yeast_deletion_project/deletions3.html). The ald2ald3::HIS3 double deletion was constructed by microhomologous recombination (MANIVASAKAM et al. 1995 ) with a DNA construct made by PCR amplification. Oligonucleotide primers, DD5, 5'-tgctcaagaatgttcatataaagggagcgaatccttaaaaggggatccggtgattgattg-3', and DD3, 5'-taatatttcattctcttacgcttagcttacatggctgcgctggctgcaggtcgacggatc-3' (Sigma-Genosys, The Woodlands, TX), were used to prime PCR amplification of the HIS3 gene cassette from plasmid YDpH (BERBEN et al. 1991 ). This resulted in a DNA fragment containing the HIS3 gene flanked by 40 bp of DNA homologous to DNA sequences outside the tandem repeated sequence of the ALD2-ALD3 locus. His⁺ transformants were then selected in the haploid strain BY4742, and deletion of the ALD2-ALD3 locus was confirmed by PCR amplification of genomic DNA using two pairs of primers: 169seq, 5'-ttgtgatcacctgctctctg-3', with 170seq, 5'-cttgtcgacactcactgatc-3', which amplifies both wild-type and deleted loci, and HIS3, 5'-ggtggagggaacatcgttgg-3', with 170seq, which produces a PCR product only in the ald2ald3::HIS3 deletion strain. The ald2ald5::URA3 and ald3ald5::URA3 double-deletion strains were constructed in a similar fashion. Oligonucleotide primers A55, 5'-aacttcttcacaacattaacaaaaagccaaagaagaagaaggggatccggtgattgattg-3'. and A53, 5'-tctataatgtttatcatacataccttcaatgagcagtcaatggctgcaggtcgacggatc-3', were used to prime PCR amplification from plasmid YDpU (BERBEN et al. 1991 ), resulting in a DNA fragment containing the URA3 gene flanked by 40 bp of DNA homologous to the DNA sequence just outside of the ALD5 open reading frame. Ura⁺ transformants were selected in strains BY4742-10753 (ald2) and BY4742-16071 (ald3), and deletion of the ALD5 gene was confirmed by PCR amplification of genomic DNA using the primer pair A5validFOR, 5'-cgatgagaatggcttcaaag-3', with URA3, 5'-cctttgttacttcttccgcc-3', which produces a 1.3-kb PCR product only at an ald5::URA3 locus.

Yeast plasmids:
Genomic DNA fragments carrying the ALD2, ALD3, and ALD5 genes with flanking sequences were individually subcloned into the BamHI site of the CEN-LEU2 yeast shuttle vector YCplac111 (GIETZ and SUGINO 1988 ) or pRS315 (SIKORSKI and HIETER 1989 ). The genomic DNA fragments were made by PCR using total yeast genomic DNA as the reaction template and gene-specific oligonucleotides. For ALD2, a 2298-bp subclone was obtained using oligonucleotides 5'-ccctttggatccgctacctcttaatgtgtcac-3' and 5'-ccctttggatccaagatctacgtaatggtggg-3'. For ALD3, a 2227-bp subclone was obtained using the oligonucleotides 5'-ccctttggatcccatatgacgtctgttcttcc-3' and 5'-ccctttggatccacattcggagtcctgtcctc-3'. For ALD5, a 2471-bp subclone was obtained using the oligonucleotides 5'-ccctttggatccctcttgtggctatgtaagcc-3' and 5'-ccctttggatccctctggcacttgtatctacc-3'.

The yeast expression vector YEp195AC and the ADH1-FMS1 derivative for overexpression of the FMS1 gene are 2µ-based vectors containing the URA3 selectable marker gene, as previously described (WHITE et al. 2001 ). Plasmids were introduced into yeast cells using lithium acetate (GIETZ and SCHIESTL 1995 ).

Media and growth conditions:
Media lacking pantothenic acid and halo assays were as previously described (WHITE et al. 2001 ). However, for experiments with ald5 strains, the addition of complete amino acid supplements was found to enhance the pantothenate auxotrophy. Pantothenic acid, ß-alanine, or spermine (Sigma, St. Louis) were added at the concentrations indicated in the text. 3-Aminopropanal (3-amino-proprionaldehyde) was made by hydrolysis of 3-aminoproprionaldehyde diacetal (Acros, Fairlawn, NJ, no. 269850250); a 10 mM solution of the diacetal was incubated in 1 M HCl at room temperature for 2 hr and then neutralized with 5 M KOH before being added to growth media. YPD media was 2% glucose, 2% bactopeptone, and 1% yeast extract (Difco). Anaerobic growth conditions were obtained using a GasPak 100 jar (Becton Dickinson no. 260626) with GasPak Plus envelopes and anaerobic indicators (Becton Dickinson nos. 271040 and 271051). Solid NaCl, KCl, or CaCl₂ (Sigma) was added directly to liquid medium to final concentrations indicated in the text and figure legends. The liquid growth curve was established by standard methods, using 96-well flat-bottom polystyrene assay plates and a Spectramax 384 Plus (Molecular Devices, Sunnyvale, CA).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALD2 and ALD3 mediate the conversion of 3-aminopropanal to ß-alanine in vivo:
While testing each of the ald gene deletion strains for pantothenic acid auxotrophy by the traditional replica-plating technique, we found that only the double deletion ald2ald3 exhibited complete pantothenic acid auxotrophy. This auxotrophy can be complemented by introduction of either the ALD2 or the ALD3 plasmids (see MATERIALS AND METHODS), confirming that the double deletion is the cause of the phenotype. Strain BY4742, its single gene deletion derivatives fms1, ecm31, and pan6, and the double-deletion strain ald2ald3 were replica plated onto media lacking pantothenic acid or supplemented with 3-aminopropanal, ß-alanine, or pantothenic acid (Fig 1A). All of these strains could grow when pantothenic acid was added to the media. However, the ald2ald3 strain could grow with ß-alanine instead of with pantothenic acid, and the fms1 strain could grow either on 3-aminopropanal or on ß-alanine instead of on pantothenic acid. This indicates that ALD2 and ALD3 function downstream of FMS1 in the ß-alanine and pantothenic acid biosynthetic pathway and are specifically required for the utilization of 3-aminopropanal. This is consistent with the model of 3-aminopropanal conversion via ß-alanine to pantothenic acid, requiring the genes FMS1, ALD2 or ALD3, ECM31, and PAN6, respectively (Fig 1B).

View larger version (32K): In this window In a new window Download PPT slide   Figure 1. Auxotrophic phenotypes caused by deletion of genes required for pantothenic acid biosynthesis. (A) Parental strain BY4742 and its deletion derivatives fms1, ald2ald3, ecm31, and pan6 were patched onto YPD agar media and incubated at 30° overnight to allow growth. The YPD plate was then replica plated onto media lacking pantothenic acid ("no additions") and supplemented with 20 µM 3-aminopropanal (3-AP), 100 µM ß-alanine (ß-ala.), or 1 µM pantothenic acid (panto.), as indicated. The replica plates were incubated for 2 days at 30°. (B) The order of the biochemical intermediates, and the order of function of the genes required, inferred from the replica-plating experiment in A.

Because the single deletions ald2 and ald3, as well as the single deletions of the remaining aldehyde dehydrogenase genes, did not appear to require exogenous ß-alanine or pantothenic acid for growth using the replica plate method (data not shown), we sought a more sensitive plating technique to assess pantothenic acid pathway metabolite requirements. We found that if an inoculum of fewer cells is used (by "spotting" a dilute liquid suspension of cells onto agar medium), a more sensitive assay is produced and partial auxotrophies can be detected. Therefore, five of the single deletion strains, ald2–ald6, the double deletion ald2ald3, and the parental strain BY4742 were spotted onto medium lacking pantothenic acid or onto medium supplemented with spermine, 3-aminopropanal, ß-alanine, or pantothenic acid (Fig 2). In contrast to the replica-plating result, the ald2 single mutant did not grow in the absence of ß-alanine or pantothenic acid after 3 days of incubation. This auxotrophy was completely reversed by introduction of a plasmid carrying the ALD2 locus (see MATERIALS AND METHODS, data not shown). Although this is the same phenotype as the ald2ald3 strain in the replica plate assay, it suggests that the single mutant ald2 is less deficient than the ald2ald3 mutant in pantothenic acid biosynthesis. In this sensitive spotting assay, increased inoculum size or an extended incubation time (e.g., to 7 days) allowed for detectable growth of the ald2 single mutant but not of the ald2ald3 double mutant (not shown).

View larger version (116K): In this window In a new window Download PPT slide   Figure 2. Auxotrophic phenotypes caused by deletion of the aldehyde dehydrogenase genes. (A) Parental strain BY4742 and its deletion derivatives ald2, ald3, ald4, ald5, and ald6, and the double deletion ald2ald3 were grown overnight on medium lacking pantothenic acid (to exhaust internal pools of pantothenic acid metabolites). Cells were then harvested and washed in distilled water by centrifugation, and 5-µl droplets containing ca. 2000 cells were inoculated onto agar medium lacking pantothenic acid ("No additions") or supplemented with 1 µM pantothenate, 100 µM spermine, 100 µM 3-aminopropanal, or 200 µM ß-alanine, as indicated. Incubation was for 3 days at 30°. (B) Parental strain BY4742, deletion derivatives ald2, ald3, and ald5, and the double deletions ald2ald3, ald2ald5, and ald3ald5 were grown for 3 days on medium lacking pantothenic acid. Cells were harvested and washed in distilled water by centrifugation, and 2000 cells were inoculated into 50 ml of liquid media lacking pantothenic acid. Cultures were incubated at 30° with shaking at 200 rpm, and optical density measurements (575 nm) were taken at various intervals out to 150 hr.

ALD5 does not play a role in pantothenic acid metabolism:
In the spotting assay, a second deletion mutant, ald5, also did not grow on media lacking ß-alanine or pantothenic acid (Fig 2A). The ald5 phenotype could be rescued by the introduction of plasmids carrying the ALD5 locus (see MATERIALS AND METHODS), and, in addition, the homozygous diploid strain ald5/ald5 behaved in an identical manner (not shown), indicating that the ald5 deletion was the cause of the phenotype in this strain. In contrast to the ald2 strain, 3-aminopropanal supported strong growth of the ald5 strain (Fig 2A), indicating that the ald5 mutant was not defective for the conversion of 3-aminopropanal to ß-alanine. Because it grew on 3-aminopropanal but not on spermine, the requirements of the ald5 mutant for pantothenic acid pathway metabolites resembled those of fms1. However, we observed several phenotypic differences (not shown): first, the ald5 strain required more exogenous ß-alanine than the fms1 mutant did for strong growth in the spotting assay. Second, in the absence of pantothenic acid, increased inoculum size (e.g., replica plating) allowed for growth of the ald5 but not of the fms1 strain. Third, the addition of multiple amino acid supplements enhanced the ß-alanine auxotrophy of the ald5 mutant, consistent with an indirect role for the Ald5p enzyme in pantothenic acid biosynthesis.

Although these results indicate that Ald5p does not function directly in 3-aminopropanal conversion to ß-alanine, we sought a more sensitive and definitive measure of the phenotype of ald5 in pantothenic acid metabolism as it relates to growth and viability, particularly in the context of ald2 and ald3 mutations. Therefore, we quantified the growth capabilities of the parental strain BY4742, double mutants ald2ald5, ald3ald5, and ald2ald3, as well as the single mutants ald2, ald3, and ald5, in the absence of exogenous pantothenic acid. Both total growth and the apparent log-phase growth rate of the parental, ald3, ald5, and ald3ald5 strains were approximately equivalent, while growth of the ald2, ald2ald3, and ald2ald5 strains remained at or near zero (Fig 2B). The fact that the ald5 has no discernible effect on growth in this assay does not contradict the spotting assay result, because the terminal incubation time was much shorter in the spotting assay than in the liquid growth assay. In fact, the ald5 strain did yield visible growth in the spotting assay after 7 days of incubation (not shown). Taken together, these results clearly indicate that ALD5 does not encode an aldehyde dehydrogenase required for the conversion of 3-aminopropanal to ß-alanine and that Ald5p is unlikely to be involved either directly or indirectly in pantothenic acid metabolism.

ALD2 plays the predominant role in pantothenic acid production:
On glucose medium, endogenous FMS1 expression is rate limiting for both growth rate and pantothenic acid production. When FMS1 is overexpressed using the ADH-FMS1 allele on a plasmid, growth rate is accelerated and excess pantothenic acid is excreted into the medium. This excretion can be detected using a bioassay involving growth of a pantothenic acid auxotroph, such as the ecm31 mutant (WHITE et al. 2001 ). We therefore compared the phenotypes of ald2, ald3, and ald5 mutants further by testing the ability of these deletion strains to excrete pantothenic acid. Strains BY4742, ald2, ald3, and ald5 harboring the ADH-FMS1 plasmid were spotted onto a "lawn" of ecm31 cells, which require pantothenic acid for growth. After incubation, halos of growth formed around all spots except around the ald2 mutant. Again, this suggests that ALD2 is directly involved in pantothenic acid production and is responsible for the majority of the conversion from 3-aminopropanal to ß-alanine (Fig 3A). Halos of growth did not occur around any of the strains harboring an empty vector control (not shown).

View larger version (58K): In this window In a new window Download PPT slide   Figure 3. Pantothenic acid overproduction and excretion are dependent upon ALD2, but not upon ALD3 or ALD5. (A) Pantothenic acid excretion requires ALD2 activity. Parental strain BY4742 and its deletion derivatives ald2, ald3, and ald5, harboring the ADH1-FMS1 overexpression allele, were spotted (1 x 107 cells/spot) onto a lawn of the ecm31 deletion strain on media lacking pantothenic acid. Incubation was for 2 days at 30°, after which time halos of growing lawn cells could be seen surrounding yeast spots that were excreting pantothenic acid. (B) Rescue of ald strains by FMS1 overexpression. Deletion strains ald5, ald2, and ald2ald3, harboring the ADH1-FMS1 overexpression plasmid or the control vector YEp195AC, were spotted onto media containing pantothenic acid (+ panto.) or lacking pantothenic acid (- panto.), as indicated. Incubation was for 2 days at 30°.

Another way to analyze the function of genes in the pantothenic acid pathway is to test the ability of the ADH1-FMS1 overexpression allele to rescue growth on medium lacking pantothenic acid. Strains lacking enzymes in the same pathway as FMS1 will be unable to grow, while strains lacking enzymes in unrelated pathways will grow. The deletion strains, ald2, ald5, and ald2ald3, were transformed with the ADH1-FMS1, or empty vector YEp195AC, and tested for growth on medium lacking pantothenic acid (Fig 3B). The ald2ald3 strain completely blocked the ADH1-FMS1- dependent stimulation of growth in the absence of pantothenic acid. Likewise, the ald2 strain was defective, although slow growth occurred with extended incubation time. Unlike the phenotype observed in the ald2 and ald2ald3 mutants, the ADH1-FMS1 allele rescued growth in the ald5 mutant in this assay, again indicating that ALD5 does not affect pantothenic acid biosynthesis. Taken together, these lines of evidence indicated that FMS1 and ALD2/ALD3 genes function together in the same pathway to mediate ß-alanine and pantothenic acid biosynthesis.

ALD3 compensates for the loss of ALD2 under conditions of osmotic stress:
The pantothenic acid excretion bioassay results, combined with the pantothenic acid auxotrophy profiles of the single mutants ald2 or ald3 and the double mutant ald2ald3, suggested that while both ALD2 and ALD3 function in the conversion of 3-aminopropanal to ß-alanine, the majority of this activity was dependent upon ALD2. Because previous evidence has indicated that the transcription of ALD2 and ALD3 is modulated by osmotic stress (MIRALLES and SERRANO 1995 ; NAVARRO-AVINO et al. 1999 ), we reasoned that the growth defect of the ald2 strain on medium lacking exogenous ß-alanine might be remediated under salt-induced osmostress, where ALD3 is upregulated. Single mutants ald2 and ald3, the double mutant ald2ald3, and the ß-alanine auxotroph fms1 were therefore streaked onto medium lacking ß-alanine, medium lacking ß-alanine but containing 0.5 M NaCl, or medium supplemented with ß-alanine (Fig 4). On medium lacking ß-alanine, sectors containing ald2 cells did not have any appreciable growth. However, growth of the ald2 strain was evident on medium lacking ß-alanine that contained 0.5 M NaCl. The ald3 strain grew under all conditions, while both ald2ald3 and fms1 strains grew only when exogenous ß-alanine was added to the medium. In the absence of exogenous ß-alanine, the effect of 0.3 M KCl on growth of the ald2 strain was less pronounced than that of 0.5 M NaCl, and 0.5 M CaCl₂ had no effect on the growth of any of the strains (not shown). The observation that the double mutant ald2ald3 remained auxotrophic for ß-alanine in the presence of 0.5 M NaCl or 0.3 M KCl suggested that ALD3, and not one of the other aldehyde dehydrogenase genes, was responsible for growth of the ald2 strain under conditions of Na⁺ or K⁺ stress.

View larger version (55K): In this window In a new window Download PPT slide   Figure 4. NaCl-induced osmotic stress causes ALD3-dependent remediation of the ß-alanine auxotrophy of ald2. Yeast deletion strains fms1, ald2ald3, ald2, and ald3 (as indicated) were grown overnight in liquid medium that lacked pantothenic acid or ß-alanine (to exhaust internal pools of the vitamin), then streaked onto (A) the same medium, (B) the same medium containing 0.5 M NaCl, and (C) the same medium containing 10 µM ß-alanine, and then incubated for 3 days at 30°.

Molecular oxygen is required for pantothenic acid biosynthesis:
Because amine oxidases utilize molecular oxygen (O₂) as a cosubstrate (in addition to amine compounds), O₂ should be required for pantothenic acid biosynthesis. To test this, strain BY4742 was streaked onto medium containing or lacking pantothenic acid and incubated under conditions of limited O₂ in an anaerobic jar. Control cultures were incubated at the same time on identical medium under aerobic conditions. After incubation for 3 days, growth was inhibited when pantothenic acid and O₂ were both absent, but growth took place when either pantothenic acid or O₂ was present (Fig 5). On a glucose medium, O₂ is required for a number of biosynthetic reactions, including sterol and unsaturated fatty acid synthesis. Presumably, these processes are sustained by the low level of O₂ remaining in the anaerobic jar, whereas the additional metabolic burden of pantothenic acid biosynthesis requires a higher concentration of O₂.

View larger version (119K): In this window In a new window Download PPT slide   Figure 5. O2 is required for the growth of yeast in the absence of pantothenic acid. Yeast strain BY4742 was streaked onto medium containing pantothenic acid (+panto.) or lacking pantothenic acid (-panto.), as indicated, and incubated anaerobically (-O2) or in the presence of air (+O2), as indicated. Incubation was for 3 days at 37°.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Aldehyde dehydrogenases are required for pantothenic acid biosynthesis in yeast:
In yeast, the ß-alanine required for pantothenic acid biosynthesis is derived from the oxidation of the polyamine spermine, involving the amine oxidase encoded by FMS1 (WHITE et al. 2001 ). In other organisms, this oxidation of polyamines to carboxylic acids occurs in two steps: First, in the presence of O₂, amine oxidases convert the polyamines to aldehydes, and second, aldehyde dehydrogenases convert the aldehydes to carboxylic acids (LARGE 1992 ). On the basis of this hypothesis of a two-step chemical conversion, we tested three predictions: first, that yeast can utilize 3-aminopropanal for pantothenic acid biosynthesis; second, that aldehyde dehydrogenase(s) is required to convert 3-aminopropanal to ß-alanine; and third, that O₂ is required for pantothenic acid biosynthesis. We found that yeast was indeed able to utilize the aldehyde compound 3-aminopropanal for pantothenic acid production and that the aldehyde dehydrogenase genes ALD2 and ALD3 were required for the conversion of 3-aminopropanal to ß-alanine in vivo. Finally, wild-type yeast had an increased requirement for O₂ during growth in the absence of pantothenic acid. Confirmation that FMS1 functions in the same pathway as ALD2 or ALD3 came from pantothenic acid excretion experiments and complementation analysis, in which it was found that FMS1 is functionally dependent on ALD2 or ALD3. Thus, the amine oxidase Fms1p is required to make 3-aminopropanal from spermine, and the aldehyde dehydrogenases, Ald2p or Ald3p, are required to make ß-alanine from 3-aminopropanal (Fig 6), consistent with the biochemistry of polyamine degradation in other organisms (LARGE 1992 ).

View larger version (27K): In this window In a new window Download PPT slide   Figure 6. The pantothenic acid pathway of yeast. The ß-alanine branch involves degradation of polyamines, mediated by the SPE genes, the aldehyde dehydrogenase genes ALD2 or ALD3, and the amine oxidase encoded by FMS1. The pantoate branch, involving ECM31, PAN5, and PAN6, is substantially the same as that found in bacteria. AdoMet, S-adenosyl methionine; dcAdoMet, decarboxy-S-adenosyl methionine.

The role of aldehyde dehydrogenases in pantothenic acid biosynthesis is specific to ALD2 and ALD3:
Saccharomyces cerevisiae has seven known or putative "nonspecific" aldehyde dehydrogenases (WANG et al. 1998 ), and biochemical functions have been defined previously only for ALD4 and ALD6, both of which convert acetaldehyde to acetate (see Introduction). ALD6, like ALD2 and ALD3, encodes a cytosolic enzyme (MEADEN et al. 1997 ), but nevertheless we found it played no role in pantothenic acid biosynthesis. Thus, these three cytosolic aldehyde dehydrogenases are functionally specialized, most likely because they recognize different aldehyde compounds as substrates. On the other hand, the inability of the mitochondrial enzymes, Ald4p and Ald5p, to participate in pantothenic acid biosynthesis may be the result of not only substrate specificity (the Ald4p substrate is acetaldehyde), but also the subcellular location of the substrate in question; 3-aminopropanal is most likely produced and consumed in the cytosol, based on the localization of the enzymes involved.

The high degree of amino acid sequence identity between Ald2p and Ald3p, their genomic organization (encoded by 1518-bp tandem reading frames with 91% nucleotide homology, separated by a 690-bp intergenic region on chromosome XIII), and the stress response elements in their cognate promoters (NAVARRO-AVINO et al. 1999 ) suggest that the ALD2-ALD3 locus may have arisen through a fairly recent gene duplication event. However, despite their 91% amino acid sequence identity, we cannot rule out the possibility that a degree of functional specialization may also exist between ALD2 and ALD3, which is suggested by the finding that the ald2 strain was partially defective for pantothenic acid biosynthesis whereas ald3 was not defective in any of our assays. In other words, the participation of Ald3p in ß-alanine production became evident only in the context of the ald2 background. The most straightforward explanation for this is that Ald2p is responsible for the majority of ß-alanine production necessary for making pantothenic acid and that the phenotypic differences might simply be a reflection of different kinetic parameters with respect to conversion of 3-aminopropanal to ß-alanine or of different protein expression levels within the cell. This was supported by our observation that the growth defect of the single mutant ald2 (but not the double mutant ald2ald3) on medium lacking ß-alanine was remediated under Na⁺- or K⁺-induced osmotic stress conditions. Since transcription of ALD3 is known to be upregulated by osmostress (NAVARRO-AVINO et al. 1999 ), this suggested that increased ALD3 transcription (in the presence of high levels of NaCl or KCl) provided sufficient converting activity of 3-aminopropanal to ß-alanine to compensate for the absence of Ald2p, allowing the ald2 strain to grow in the absence of exogenous ß-alanine. Additionally, the fact that deletion of both of these genes was required for complete ß-alanine and pantothenic acid auxotrophy further indicated that Ald2p and Ald3p are both capable of converting 3-aminopropanal to ß-alanine in vivo. It is also possible that ALD2 and ALD3 might have additional functions, as suggested by their role during growth on ethanol and by their regulated gene expression (NAVARRO-AVINO et al. 1999 ).

Although the ald5 mutant exhibited an enhanced requirement for pantothenic acid pathway metabolites in the spotting assay, several lines of existing and new evidence indicate that the Ald5p protein does not play an essential role in pantothenic acid biosynthesis. First, the mitochondrial localization of Ald5p is inconsistent with a role in cytosolic metabolism. The observation that ald5 strains are defective for mitochondrial electron transport and cytochrome biogenesis (KURITA and NISHIDA 1999 ) supports that proposition. Second, the amino acid identity of Ald5p with either Ald2p or Ald3p is much lower than that between Ald2p and Ald3p, which is consistent with the hypothesis of aldehyde dehydrogenase functional specialization. Third, we have shown that the ald5 mutant does not directly affect pantothenic acid biosynthesis per se in either the ADH1-FMS1 allele complementation assay or the pantothenic acid excretion bioassay. Fourth, in liquid medium, the ald5 strain grew at a similar rate to the parental strain in the absence of exogenous pantothenic acid. Thus, the effect of ald5 on the pantothenic acid pathway is most likely of an indirect nature. Possible mechanisms include an increased cellular requirement for coenzyme A, a decreased Fms1p activity, and a decreased transport of spermine to the Fms1p enzyme. In all cases, overexpression of FMS1 using the ADH1-FMS1 allele would be expected to compensate for the partial pantothenate auxotrophy of ald5 either by increasing Fms1p activity or by putting Fms1p in parts of the cell it does not normally occupy. A full explanation will require a better understanding of the cellular role of ALD5.

In conclusion, the "nonspecific" aldehyde dehydrogenases do in fact have specialized functions in normal cellular metabolism. In the case of Ald2p and Ald3p, the function is in coenzyme A biosynthesis. This does not rule out possible "nonspecific" roles of these enzymes in protecting the cell from the occurrence of toxic aldehyde compounds, but does suggest that a number of different aldehyde dehydrogenase enzymes may be specialized for metabolic functions that occur during normal cellular growth and development.

Manuscript received June 20, 2002; Accepted for publication October 21, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BERBEN, G., J. DUMONT, V. GILLIQUET, P.-A. BOLLE, and F. HILGER, 1991  The YDp plasmids: a uniform set of vectors bearing versatile gene disruption cassettes for Saccharomyces cerevisiae.. Yeast 7:475-477.[Medline]

CRONAN, J. E., 1980  ß-alanine synthesis in Escherichia coli.. J. Bacteriol. 141:1291-1297.[Abstract/Free Full Text]

DICKINSON, F. M., 1996  The purification and some properties of the Mg²⁺-activated cytosolic aldehyde dehydrogenase of Saccharomyces cerevisiae.. Biochem. J. 315:393-399.[Medline]

GIETZ, R. D. and R. H. SCHIESTL, 1995  Transforming yeast with DNA. Methods Mol. Cell. Biol. 5:255-269.

GIETZ, R. D. and A. SUGINO, 1988  New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527-534.[Medline]

HODGES, P. E., A. H. Z. MCKEE, B. P. DAVIS, W. E. PAYNE, and J. I. GARRELS, 1999  Yeast Proteome Database (YPD): a model for the organization and presentation of genome-wide functional data. Nucleic Acids Res. 27:69-73.[Abstract/Free Full Text]

HÖLTTÄ, E., 1977  Oxidation of spermidine and spermine in rat liver: purification and properties of polyamine oxidase. Biochemistry 16:91-100.[Medline]

JACKOWSKI, S., 1996 Biosynthesis of pantothenic acid and coenzyme A, pp. 687–694 in Escherichia coli and Salmonella: Cellular and Molecular Biology, edited by F. C. NEIDHARDT, R. CURTISS III, J. INGRAHAM, E. LIN, K. LOW, B. MAGASANIK, W. REZNIKOFF, M. RILEY, M. SCHAECHTER and H. UMBARGER. American Society for Microbiology, Washington, DC.

JACOBSON, M. K. and C. BERNOFSKY, 1974  Mitochondrial acetaldehyde dehydrogenase from Saccharomyces cerevisiae.. Biochim. Biophys. Acta 305:277-291.

KURITA, O. and Y. NISHIDA, 1999  Involvement of mitochondrial aldehyde dehydrogenase ALD5 in maintenance of the mitochondrial electron transport chain in Saccharomyces cerevisiae.. FEMS Microbiol. Lett. 181:281-287.[Medline]

LARGE, P. J., 1992  Enzymes and pathways of polyamine breakdown in microorganisms. FEMS Microbiol. Rev. 88:249-262.

MANIVASAKAM, P., S. C. WEBER, J. MCELVER, and R. H. SCHIESTL, 1995  Micro homology-mediated PCR targeting in Saccharomyces cerevisiae.. Nucleic Acids Res. 23:2799-2800.[Free Full Text]

MEADEN, P. G., F. M. DICKINSON, A. MIFSUD, W. TESSIER, and J. WESTWATER et al., 1997  The ALD6 gene of Saccharomyces cerevisiae encodes a cytosolic, Mg²⁺-activated acetaldehyde dehydrogenase. Yeast 13:1319-1327.[Medline]

MIRALLES, V. J. and R. SERRANO, 1995  A genomic locus in Saccharomyces cerevisiae with four genes up-regulated by osmotic stress. Mol. Microbiol. 17:653-662.[Medline]

NAVARRO-AVIÑO, J. P., R. PRASAD, V. J. MIRALLES, R. M. BENITO, and R. SERRANO, 1999  A proposal for nomenclature of aldehyde dehydrogenases in Saccharomyces cerevisiae and characterization of the stress-inducible ALD2 and ALD3 genes. Yeast 15:829-842.[Medline]

REMIZE, F., E. ANDRIEU, and S. DEQUIN, 2000  Engineering of the pyruvate dehydrogenase bypass in Saccharomyces cerevisiae: role of the cytosolic Mg²⁺ and mitochondrial K⁺ acetaldehyde dehydrogenases Ald6p and Ald4p in acetate formation during alcoholic fermentation. Appl. Environ. Microbiol. 66:3151-3159.[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]

TESSIER, W. D., P. G. MEADEN, F. M. DICKINSON, and M. MIDGLEY, 1998  Identification and disruption of the gene encoding the K⁺-activated acetaldehyde dehydrogenase of Saccharomyces cerevisiae.. FEMS Microbiol. Lett. 164:29-34.[Medline]

TESSIER, W., M. DICKINSON, and M. MIDGLEY, 1999  The roles of acetaldehyde dehydrogenases in Saccharomyces cerevisiae.. Enzymol. Mol. Biol. Carbonyl Metab. 7:243-247.

THOMPSON, D. A. and F. W. STAHL, 1999  Genetic control of recombination partner preference in yeast meiosis: isolation and characterization of mutants elevated for meiotic unequal sister-chromatid recombination. Genetics 153:621-641.[Abstract/Free Full Text]

WANG, X., C. J. MANN, Y. BAI, L. NI, and H. WEINER, 1998  Molecular cloning, characterization, and potential roles of cytosolic and mitochondrial aldehyde dehydrogenases in ethanol metabolism in Saccharomyces cerevisiae.. J. Bacteriol. 180:822-830.[Abstract/Free Full Text]

WHITE, W. H., P. L. GUNYUZLU, and J. H. TOYN, 2001  Saccharomyces cerevisiae is capable of de novo pantothenic acid biosynthesis involving a novel pathway of ß-alanine production from spermine. J. Biol. Chem. 276:10794-10800.[Abstract/Free Full Text]

WILLIAMSON, J. M. and G. M. BROWN, 1979  Purification and properties of l-aspartate--decarboxylase, an enzyme that catalyses the formation of ß-alanine in Escherichia coli.. J. Biol. Chem. 254:8074-8082.[Abstract/Free Full Text]

WINZELER, E. A., D. D. SHOEMAKER, A. ASTROMOFF, H. LIANG, and K. ANDERSON et al., 1999  Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis. Science 285:901-906.[Abstract/Free Full Text]

This article has been cited by other articles:

				L. Kuepfer, U. Sauer, and L. M. Blank Metabolic functions of duplicate genes in Saccharomyces cerevisiae Genome Res., October 1, 2005; 15(10): 1421 - 1430. [Abstract] [Full Text] [PDF]

				F. Saint-Prix, L. Bonquist, and S. Dequin Functional analysis of the ALD gene family of Saccharomyces cerevisiae during anaerobic growth on glucose: the NADP+-dependent Ald6p and Ald5p isoforms play a major role in acetate formation Microbiology, July 1, 2004; 150(7): 2209 - 2220. [Abstract] [Full Text] [PDF]

				U. Genschel Coenzyme A Biosynthesis: Reconstruction of the Pathway in Archaea and an Evolutionary Scenario Based on Comparative Genomics Mol. Biol. Evol., July 1, 2004; 21(7): 1242 - 1251. [Abstract] [Full Text] [PDF]

				J. Onodera and Y. Ohsumi Ald6p Is a Preferred Target for Autophagy in Yeast, Saccharomyces cerevisiae J. Biol. Chem., April 16, 2004; 279(16): 16071 - 16076. [Abstract] [Full Text] [PDF]

				M. K. Chattopadhyay, C. W. Tabor, and H. Tabor Spermidine but not spermine is essential for hypusine biosynthesis and growth in Saccharomyces cerevisiae: Spermine is converted to spermidine in vivo by the FMS1-amine oxidase PNAS, November 25, 2003; 100(24): 13869 - 13874. [Abstract] [Full Text] [PDF]