D -Tyrosyl-tRNA Tyr Metabolism in Saccharomyces cerevisiae*

The Saccharomyces cerevisiae YDL219w ( DTD1 ) gene, which codes for an amino acid sequence sharing 34% identity with the Escherichia coli D -Tyr-tRNA Tyr deacylase, was cloned, and its product was functionally char-acterized. Overexpression in the yeast of the DTD1 gene from a multicopy plasmid increased D -Tyr-tRNA Tyr deacylase activity in crude extracts by two orders of mag-nitude. Upon disruption of the chromosomal gene, deacylase activity was decreased by more than 90%, and the sensitivity to D -tyrosine of the growth of S. cerevisiae was exacerbated. The toxicity of D -tyrosine was also enhanced under conditions of nitrogen starvation, which stimulate the uptake of D -amino acids. In relation with these behaviors, the capacity of purified S. cerevisiae tyrosyl-tRNA synthetase to produce D -Tyr-tRNA Tyr could be shown. Finally, the phylogenetic distribution of genes homologous to DTD1 was examined in connection with L -tyrosine prototrophy or auxotrophy. In the aux- otrophs, DTD1 -like genes are systematically absent. In the prototrophs, the putative occurrence of a deacylase is variable. It possibly depends on the L -tyrosine ana- bolic pathway adopted by the cell. Living cells -amino aminoacylation gene of S. cerevisiae was amplified by PCR using 0.1 m g of this DNA plus oligonucleotides CGCGGATCCGA-TTTACAATGAAGATTGTCTTACAAAAAGTC and GCTCTAGAGTCA- ATCTTATTGGTCACTGTCAAGAATGATTG as primers. The PCR fragment of expected size (549 base pairs) was purified using the Qiagen PCR purification kit-50, digested by Bam HI and Xba I, and inserted into the corresponding sites of plasmid pYES2 to give the plasmid pYES2-DTD1.Thenucleotide sequence of the cloned DTD1 gene differed from the sequence in the genomic data base by a C 3 T substitution located 293 bases downstream from the A of the ATG initiation codon. This change did not modify the amino acid sequence of the protein produced. The plasmid pYES2-DTD1 was used to transform the yeast strain DBY2057 by the lithium acetate method (24). Disruption of the DTD1 Gene— Disruption of the DTD1 gene was performed by the PCR-based method of Wach et al. (25) using the kanMX cassette as a selectable marker. This cassette contains the synthetase was determined using a M r of 2 3 47,712 and a light absorption coefficient of 0.439 A 280 units mg 2 1 ml, as deduced from the amino acid sequence. Enzymatic Assays— Measurements of initial rates of D -Tyr-tRNA Tyr deacylase activity were performed in crude extracts as described earlier, using E. coli D -[ 3 H]Tyr-tRNA Tyr as substrate (20). One unit of enzyme activity corresponds to 1 pmol of D -Tyr-tRNA Tyr hydrolyzed/min.Tyrosyl-tRNA synthetase activity was assayed during 10 min at 28 °C in 100 m l of a reaction mixture containing 20 m M Tris-HCl (pH 7.8), 7 m M MgCl 2 , 2 m M ATP, 0.1 m M EDTA, 150 m M KCl, 1 mg/ml bovine serum albumin, 2.5 m M 2-mercaptoethanol, 5.3 mg/ml crude brewer’s yeast tRNA, catalytic amounts of tyrosyl-tRNA synthetase, and indicated amounts of L -[ 14 C]tyrosine (500 Ci/mol) or D -[ 3 H]tyrosine (500 Ci/mol). The reaction was quenched by the addition of (i) 2.5 ml of ice-cold trichloroacetic acid (5%, w/w) containing 0.5% tyrosine and (ii) 10 m l of carrier RNA from yeast (4 mg/ml). The precipitate was recov- ered on Whatman GF-C filters, and the retained radioactivity was measured in a Beckman LS1801 scintillation counter. One unit of enzyme initial activity corresponds to 1 pmol of tyrosine transferred onto the tRNA/min.

Living cells have developed various mechanisms to avoid the incorporation of D-amino acids into proteins. Thus, aminoacylation of tRNAs, binding of aminoacyl-tRNAs to elongation factor Tu, and peptide bond formation all favor the L-isomers of the amino acids (1). In addition, the intracellular concentrations of D-amino acids are reduced through the action of Damino acid dehydrogenases in bacteria (2,3) or of D-amino acid oxidases in various eucaryotes (4 -7). In the case of fungi, a D-amino acid acetylase is found (8). These enzymes often exhibit a broad specificity and can attack a number of D-amino acids, although with variable efficiency (2, 4, 8 -14). There also occur specific enzymes, such as the D-serine deaminase, which protects numerous organisms against the toxicity of D-serine (15,16).
In crude extracts of bacteria, yeast, and mammals, an enzyme activity capable of hydrolyzing the ester bond of D-Tyr-tRNA was observed very early (17). This activity, which is likely to recycle erroneously aminoacylated tRNA Tyr , was searched for because, at the same time, evidence had been gained that Escherichia coli and Bacillus subtilis tyrosyl-tRNA synthetases produced D-Tyr-tRNA Tyr in vitro (18,19) Recently, the gene encoding the above D-Tyr-tRNA Tyr deacylase could be isolated from E. coli (20). This gene, named dtd, corresponds to the yihZ open reading frame at 87.81 min on the E. coli genetic map. Disruption of the dtd gene established that its product accounts for more than 90% of the deacylase activity in crude extract. However, although the dtd gene is essential to afford protection against the toxicity of D-tyrosine added to the culture medium, it does not interfere with cell growth under standard conditions (20).
A comparative sequence analysis revealed homologs of the E. coli dtd gene in several other bacteria as well as in the yeast Saccharomyces cerevisiae, in the nematode Caenorhabditis elegans, in the higher plant Arabidopsis thaliana, in mice, and in man. Such a ubiquitous character suggests the hydrolysis of D-Tyr-tRNA Tyr to be a universal mechanism of defense against a harmful effect of D-tyrosine. However, before drawing such a conclusion, more cells have to be examined for the occurrence of a relationship between D-Tyr-tRNA Tyr hydrolysis and protection against D-tyrosine.
In the present study, we functionally characterize the S. cerevisiae YDL219w gene, which codes for a protein showing 34% identity with the E. coli D-Tyr-tRNA Tyr deacylase. The YDL219w gene was amplified by PCR 1 and cloned. Expression of this gene from a multicopy plasmid increased D-Tyr-tRNA Tyr deacylase activity in crude extracts, whereas disruption of the chromosomal YDL219w gene decreased deacylase activity and exacerbated the sensitivity of S. cerevisiae to D-tyrosine. We propose the name of DTD1 for the YDL219w gene.

MATERIALS AND METHODS
Brewer's yeast tRNA was from Roche Molecular Biochemicals. L-[ 14 C]tyrosine acceptance of this tRNA was 23 pmol/A 260 . Q-Sepharose was from Amersham Pharmacia Biotech. Nickel-nitrilotriacetic acidagarose was from Qiagen. D-Methylene[ 3 H]tyrosine (211 GBq/mmol) was custom prepared by Amersham Pharmacia Biotech. L-[ 14 C]tyrosine (18.4 GBq/mmol) was from NEN Life Science Products. Geneticin disulfate and unlabeled L-and D-tyrosine were from Sigma. Plasmid pYES2/GS-YGR185CY was from Invitrogen. YPD medium was made as described before (21). Yeast nitrogen base without amino acids and ammonium sulfate, from Difco, was used for the preparation of minimal media.
Cloning of the DTD1 Gene-Genomic DNA of S. cerevisiae strain DBY2057 (Table I) (22) was prepared by the procedure of Hoffman and Winston (23). Then, the DTD1 gene of S. cerevisiae was amplified by PCR using 0.1 g of this DNA plus oligonucleotides CGCGGATCCGA-TTTACAATGAAGATTGTCTTACAAAAAGTC and GCTCTAGAGTCA-ATCTTATTGGTCACTGTCAAGAATGATTG as primers. The PCR fragment of expected size (549 base pairs) was purified using the Qiagen PCR purification kit-50, digested by BamHI and XbaI, and inserted into the corresponding sites of plasmid pYES2 to give the plasmid pYES2-DTD1.
The nucleotide sequence of the cloned DTD1 gene differed from the sequence in the genomic data base by a C3 T substitution located 293 bases downstream from the A of the ATG initiation codon. This change did not modify the amino acid sequence of the protein produced. The plasmid pYES2-DTD1 was used to transform the yeast strain DBY2057 by the lithium acetate method (24).
Disruption of the DTD1 Gene-Disruption of the DTD1 gene was performed by the PCR-based method of Wach et al. (25) using the kanMX cassette as a selectable marker. This cassette contains the * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
kanamycin resistance gene of the E. coli transposon Tn903 fused to transcriptional and translational control sequences of the Ashbya gossypii TEF gene. Consequently, S. cerevisiae transformants are efficiently selected because of their acquired resistance to Geneticin (G418) (26). A DNA fragment containing the kanMX cassette from plasmid pFA6-kanMX4, flanked by 40 base pairs of the target locus, was amplified by PCR using oligonucleotides: AGCCAAGCATCTGTAGTCGT-CGATTCAAAAGTTATTTCAACGTACGCTGCAGGTCGAC and CC-CTTCATTAGTTAAAGAGCAACTCATCATTGCGCCGAATATCGATG-AATTCGAGCTCG. The S. cerevisiae strain DBY2057 was transformed by the resulting PCR fragment using the lithium acetate method. Before plating on YPD-agar medium supplemented with 300 g of Geneticin/ml, the cells were cultivated at 30°C during 4 h with shaking. Disruption of the DTD1 gene in Geneticin-resistant clones was verified by PCR amplification of genomic DNA using oligonucleotides TGAAGATTGTCTTA-CAAAAAGTCAGCCAAGC and TTGGTCACTGTCAAGAATGATTGTA-ACGGG. One positive clone was named DBY2057⌬DTD1 and was used for further studies.
Purification of Native S. cerevisiae Tyrosyl-tRNA Synthetase-S. cerevisiae tyrosyl-tRNA synthetase was partially purified from strain YP-ALS (27). Cells were grown at 30°C in 2 liters of YPD medium and harvested by centrifugation for 35 min at 3000 ϫ g. The pellet was suspended in a buffer (pH 7.8) containing 20 mM Tris-HCl, 0.1 mM EDTA, 10 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, and 1 g/ml of leupeptin, at an optical density of 100 at 650 nm. Cells were disrupted by sonication (10 min, 0°C). Cell debris was removed by centrifugation for 20 min at 8000 ϫ g. Streptomycin was added to the supernatant at 30 g/l to precipitate nucleic acids. After centrifugation for 30 min at 8000 ϫ g, the resulting supernatant was brought to 80% ammonium sulfate saturation, left to stand 1 h at 4°C, and centrifuged for 30 min at 8000 ϫ g. The pellet was dissolved in 20 ml of 10 mM potassium phosphate (pH 6.75), containing 0.1 mM EDTA, and 10 mM 2-mercaptoethanol (buffer A), and dialyzed against 2 liters of the same buffer. The resulting solution was diluted to reach 6 A 280 and was applied on a Q-Sepharose column (2.6 ϫ 16 cm) equilibrated in buffer A. The column was eluted with a 1.25-liter linear gradient from 0 to 500 mM NaCl in buffer A. Fractions showing tRNA Tyr aminoacylation activity were pooled, concentrated by ammonium sulfate precipitation (80% saturation), centrifuged for 30 min at 10,000 ϫ g, and dissolved in 20 mM Tris-HCl (pH 7.8), 0.1 mM EDTA, 10 mM 2-mercaptoethanol. The protein sample was then dialyzed against 1 liter of the same buffer and used for assaying tyrosyl-tRNA synthetase activity.
Purification of His 6 -tagged S. cerevisiae Tyrosyl-tRNA Synthetase-Plasmid pYES2/GS-YGR185CY from Invitrogen contains the TYS1 (YGR185c) gene of S. cerevisiae cytoplasmic tyrosyl-tRNA synthetase fused at its 3Ј-end with a DNA sequence coding for a C-terminal V5 epitope tag and a polyhistidine tag (His 6 ). The nucleotide sequence of the TYS1 gene cloned in this plasmid differed from the sequence in the genomic data base by two A3 G substitutions located 873 and 1110 bases downstream from the A of the ATG initiation codon. To correct the gene sequence, a 480-base pair SalI/StuI fragment of the pYES2/ GS-YGR185CY plasmid was deleted. For this purpose, the plasmid was digested by SalI, filled in, partially digested by StuI, and recircularized. Then, the resulting plasmid was digested by SalI and used to transform the yeast strain DBY2057 for recombination. A plasmid containing the wild-type insert sequence was named pYES2/GS-TYS1. The yeast clone harboring this plasmid was used for the preparation of His 6 -tagged tyrosyl-tRNA synthetase.
Cells were grown at 30°C in 2 liters of minimal medium supplemented with 0.5% ammonium sulfate, 2% galactose, and 1% raffinose. Crude extract preparation, nucleic acid precipitation with streptomycin sulfate, and ammonium sulfate precipitation of proteins were per-formed as described above in the case of purification of native tyrosyl-tRNA synthetase. The protein pellet was dissolved in 20 ml of 50 mM potassium phosphate buffer (pH 8.0) containing 20 mM imidazole, 0.3 M NaCl, and 10 mM 2-mercaptoethanol (buffer B) and dialyzed against 2 liters of the same buffer. The resulting sample was applied on a nickelnitrilotriacetic acid agarose column (1.3 ϫ 1.7 cm) equilibrated in buffer B. After a 10-ml wash with buffer B at a flow rate of 0.13 ml/min, the column was eluted with a 20-ml linear gradient from 20 to 300 mM imidazole in buffer B. Fractions showing tRNA Tyr aminoacylation activity were pooled and prepared for activity measurements as described for native protein.
Using the above procedure, the His 6 -tagged tyrosyl-tRNA synthetase was purified approximately 140-fold when compared with the crude extract. It was homogeneous according to SDS-PAGE analysis. Concentration of the His 6 -tagged tyrosyl-tRNA synthetase was determined using a M r of 2 ϫ 47,712 and a light absorption coefficient of 0.439 A 280 units mg Ϫ1 ml, as deduced from the amino acid sequence.
Enzymatic Assays-Measurements of initial rates of D-Tyr-tRNA Tyr deacylase activity were performed in crude extracts as described earlier, using E. coli D-[ 3 H]Tyr-tRNA Tyr as substrate (20). One unit of enzyme activity corresponds to 1 pmol of D-Tyr-tRNA Tyr hydrolyzed/min.
Tyrosyl-tRNA synthetase activity was assayed during 10 min at 28°C in 100 l of a reaction mixture containing 20 mM Tris-HCl (pH 7.8), 7 mM MgCl 2 , 2 mM ATP, 0.1 mM EDTA, 150 mM KCl, 1 mg/ml bovine serum albumin, 2.5 mM 2-mercaptoethanol, 5.3 mg/ml crude brewer's yeast tRNA, catalytic amounts of tyrosyl-tRNA synthetase, and indicated amounts of L-[ 14 C]tyrosine (500 Ci/mol) or D-[ 3 H]tyrosine (500 Ci/mol). The reaction was quenched by the addition of (i) 2.5 ml of ice-cold trichloroacetic acid (5%, w/w) containing 0.5% tyrosine and (ii) 10 l of carrier RNA from yeast (4 mg/ml). The precipitate was recovered on Whatman GF-C filters, and the retained radioactivity was measured in a Beckman LS1801 scintillation counter. One unit of enzyme initial activity corresponds to 1 pmol of tyrosine transferred onto the tRNA/min.

RESULTS
Cloning of the DTD1 Gene-Comparative analysis using the BLAST program revealed the presence in S. cerevisiae of a gene homologous to the E. coli dtd gene encoding D-Tyr-tRNA Tyr deacylase. This gene, DTD1 (or YDL219w), is located on chromosome IV and contains one intron of 71 base pairs. To investigate its function and physiological importance, the DTD1 gene was amplified by PCR and cloned into the yeast expression vector pYES2. In the resulting plasmid (pYES2-DTD1), the DTD1 gene was placed under the control of the Gal1 portion of the divergent GAL1/GAL10 promoter. Cells DBY2057 transformed by plasmid pYES2-DTD1 were cultivated in minimal medium with galactose as the carbon source, i.e. under conditions inducing full expression of the cloned gene. D-Tyr-tRNA Tyr deacylase activity in a crude extract of the transformed strain was increased 190-fold as compared with strain DBY2057 (Table II). When cells were grown in the presence of glucose, the presence of the plasmid pYES2-DTD1 increased the deacylase activity by a factor of 1.7 only. These results showed that the D-Tyr-tRNA Tyr deacylase activity recovered in crude extracts followed the expected level of expression of the cloned DTD1 gene.
Inactivation of the S. cerevisiae DTD1 Gene-To determine whether the DTD1 protein was responsible for the previously reported D-Tyr-tRNA Tyr deacylase activity in yeast crude extracts (17), the DTD1 gene was disrupted. For this purpose, the chromosomal gene of the S. cerevisiae haploid strain DBY2057 was interrupted by the kanMX cassette using a PCR-based procedure.
The DTD1 gene inactivation led to a more than 10-fold decrease in the D-Tyr-tRNA Tyr deacylase activity in the crude extract (Table II). Remarkably, however, the inactivation did not affect the growth rate of yeast cells in rich YPD medium or minimal yeast nitrogen base medium.

D-Tyrosyl-tRNA
Tyr Metabolism in S. cerevisiae leucine, D-alanine, D-tryptophan, and D-tyrosine) inhibit the growth of wild-type S. cerevisiae cells (28). These D-amino acids are imported into yeast cells by the general amino acid permease corresponding to the GAP1 gene (28,29). The activity of this permease is controlled through a double mechanism that involves a derepression of the GAP1 gene expression under limiting nitrogen conditions, on the one hand, and an inactivation of the permease activity by ammonium ions, on the other hand. As a result, the toxicity of D-amino acids is enhanced when cells are grown in a medium devoid of ammonium salts (30 -32). Such observations led us to compare the toxicity of D-tyrosine in a minimal medium containing either L-proline or ammonium sulfate as the nitrogen source. Nitrogen starvation is known to induce the filamentous (pseudohyphal) growth of diploids of S. cerevisiae (33,34). Consequently, to avoid filamentation, we used haploid strains. On minimal medium agar plates supplemented with L-proline, the growths of the wild-type strain (DBY2057) and of the dtd1 mutant strain (DBY2057⌬DTD1) were identical. Upon addition of 30 M D-tyrosine, a significant inhibition of the growth of the mutant colonies only was observed (Fig. 1A). At 300 M D-tyrosine, the colony formation of either strain was strongly inhibited.
On minimal medium agar plates containing ammonium sulfate, the growth of all strains became less sensitive to the presence of D-tyrosine. This behavior reflects the inhibition of the GAP1 permease by ammonium ions (29). However, a relatively stronger effect of D-tyrosine on the growth of the mutant strain (DBY2057⌬DTD1) was visible again (Fig. 1B). Therefore, whatever the culture condition assayed, the dtd1 strain was more sensitive to D-tyrosine than the parental DTD1 strain.
To investigate further the relation between the toxicity of D-tyrosine and the expression of the DTD1 gene, strains DBY2057 and DBY2057⌬DTD1 were transformed by either plasmid pYES2-DTD1 or control plasmid pYES2. An overexpression of the DTD1 gene harbored by the pYES2-DTD1 plasmid was ensured by the presence of galactose. Under this condition, the growths of strains DBY2057(pYES2-DTD1) and DBY2057⌬DTD1(pYES2-DTD1) became indistinguishable whatever the concentration of D-tyrosine in the solid medium (Fig. 2). In agreement with this observation, levels of overproduced D-Tyr-tRNA Tyr deacylase activity in extracts from the two strains were very similar (Table II). Therefore, the addition in trans of a plasmid-borne functional DTD1 gene was enough to cure the specific sensitivity of strain DBY2057⌬DTD1 to D-tyrosine and to give this strain the phenotype of DBY2057(pYES2-DTD1). Moreover, because they each overexpress the DTD1 gene, the two pYES2-DTD1 carrying strains grew slightly better than DBY2057(pYES2) in the presence of D-tyrosine.
On minimal medium agar plates supplemented with glucose, the sensitivity to D-tyrosine of strain DBY2057⌬DTD1 still responded to the transformation with plasmid pYES2-DTD1. However, as clearly shown on the plates containing 30 or 100 M D-tyrosine (Fig. 3), the strain DBY2057⌬DTD1 transformed by pYES2-DTD1 remained slightly more sensitive to D-tyrosine than the control DTD1 strain transformed by either pYES2-DTD1 or pYES2 plasmid. This behavior reflects the specific repression of the cloned DTD1 gene on the plasmid under glucose conditions. Indeed, the intracellular level of D-Tyr-tRNA Tyr deacylase activity derived from plasmid pYES2-DTD1 in the context of strain DBY2057⌬DTD1 was 2-fold smaller than that arising from the chromosome in the wild-type strain DBY2057 carrying control plasmid pYES2 (Table II).
Aminoacylation of tRNA with D-Tyrosine in S. cerevisiae-Altogether, the above results suggested that the DTD1 gene product can protect S. cerevisiae against the toxicity of externally added D-tyrosine through an intracellular hydrolysis of D-Tyr-tRNA Tyr . Such a conclusion implies that misacylation of b Yeast cells were grown in minimal medium (yeast nitrogen base without amino acids and ammonium sulfate) supplemented with Lproline (0.5 mg/ml) as nitrogen source and 2% glucose or galactose as carbon and energy source. With nontransformed strains DBY2057 and DBY2057⌬DTD1, uracil was also added to the medium (50 g/ml). Specific activity of the deacylase was measured in crude extracts obtained by sonication of steady-state cultures (optical density of 7-10 at 650 nm) resuspended at an optical density of 100 at 650 nm. tRNA Tyr by D-tyrosine can occur in vivo. Because E. coli and B. subtilis tyrosyl-tRNA synthetases accept D-tyrosine as a substrate (18 -20), we asked for whether S. cerevisiae tyrosyl-tRNA synthetase also can produce D-Tyr-tRNA Tyr . In a first set of experiments, native tyrosyl-tRNA synthetase was partially purified from a S. cerevisiae crude extract by chromatography on a Q-Sepharose column. Using crude brewer's yeast tRNA and radioactive D-tyrosine (3.5 M) as substrates, the obtained tyrosyl-tRNA synthetase sample could be shown to produce D-tyrosylated tRNA at an initial rate of 4 units/mg of total protein in the assay. With 3.5 M L-tyrosine, under the same experimental conditions, the rate of tRNA Tyr aminoacylation was 680 units/mg.
In the second set of experiments, homogeneous His 6 -tagged tyrosyl-tRNA synthetase of S. cerevisiae was used for aminoacylation assays. With 1 M D-tyrosine, the initial rate of tRNA aminoacylation was equal to 0.9 ϫ 10 Ϫ3 s Ϫ1 . Under the same reaction conditions, the enzyme aminoacylated tRNA with 1 M L-tyrosine at the rate of 0.13 s Ϫ1 . Therefore, the ratio of initial rates with L-tyrosine or D-tyrosine measured with tagged protein was comparable to that determined with partially purified native tyrosyl-tRNA synthetase (ratio values of 145 and 170, respectively). Finally, the initial rate of D-Tyr-tRNA Tyr formation by tagged tyrosyl-tRNA synthetase in the presence of 1 M D-[ 3 H]tyrosine was reduced by at least 98% upon addition of 5 M nonradioactive L-tyrosine to the incubation mixture.
Synthesis of D-Tyr-tRNA Tyr by tagged tyrosyl-tRNA synthetase could be further established by aminoacylation assays conducted in the presence of pure E. coli D-Tyr-tRNA Tyr deacylase (20). Tyrosyl-tRNA synthetase concentration in the assay (270 nM) was adjusted so that within 10 min, 150 nM of Dtyrosylated tRNA was produced in the absence of deacylase. When 150 nM deacylase was present, the production of D-Tyr-tRNA was reduced by more than 96%. In parallel experiments, we verified that the presence of D-Tyr-tRNA deacylase did not affect the formation of L-Tyr-tRNA Tyr . All these results reinforce our initial views that mischarging of D-tyrosine onto tRNA Tyr by S. cerevisiae tyrosyl-tRNA synthetase is at the origin of at least a part of the toxicity of this D-amino acid.

DISCUSSION
The present study shows that, like E. coli, S. cerevisiae harbors a gene encoding a protein with D-Tyr-tRNA Tyr deacylase activity. This gene confers protection to the cell against one harmful effect of D-tyrosine. Consequently, the phylogenetic distribution of the dtd/DTD1 homologs may be considered again.
When the available complete genome sequences are examined, it is striking to note that, systematically, the organisms that are auxotrophic for L-tyrosine (Mycoplasma pneumoniae, Mycoplasma genitalium, Rickettsia prowazekii, Borrelia burgdorferi, Treponema pallidum, Chlamydia trachomatis, and Chlamydia pneumoniae) lack a dtd/DTD1-like gene. Consequently, it is tempting to conclude that those cells, which do not synthesize L-tyrosine, do not produce D-tyrosine and therefore do not need a deacylase activity. This idea implies that Dtyrosine can be made as a side product of the anabolic pathways for L-tyrosine synthesis.
Whatever the considered organism, the biosynthesis of Ltyrosine from prephenate always involves two steps: a decarboxylation-dehydrogenation step and a transamination step (35)(36)(37). However, the time sequence of these two steps depends on the cell. In organisms like E. coli or B. subtilis, the decarboxylation-dehydrogenation of prephenate takes place first. It produces 4-hydroxyphenylpyruvate. The last step of L-tyrosine biosynthesis is the addition of an amino group to the future C␣ atom of L-tyrosine (37). One may assume that, upon this transformation, D-tyrosine can appear as a side product and requires D-Tyr-tRNA Tyr deacylase to circumvent the ensuing toxicity.
In other organisms such as most cyanobacteria, the transamination of prephenate is made first leading to L-arogenate. In a second step, L-arogenate is transformed into L-tyrosine by aro-  genate dehydrogenase (35,38). In this case, the production of D-tyrosine would be avoided because the arogenate dehydrogenase only works with L-arogenate. Interestingly, Synechocystis sp. PCC6803, which possesses a L-tyrosine pathway via L-arogenate, does not show any dtd/DTD1-like gene.
From the genome data only, it is difficult, however, to predict which pathway is used for the biosynthesis of L-tyrosine. The reason is that the enzymes of the two pathways can share sequence homologies. For instance, the Synechocystis arogenate dehydrogenase shows 28% amino acid sequence identity with the prephenate dehydrogenase of B. subtilis. In addition, in organisms like Serratia, Erwinia, Aeromonas, and Pseudomonas (37,39), the two pathways can co-exist. Moreover, some biochemical data do not support the occurrence of a strict correlation between the presence of a dtd/DTD1 homolog and the production of L-tyrosine via the 4-hydroxyphenylpyruvate pathway. For instance, prephenate dehydrogenase activity could not be detected in extracts of Streptomyces griseus (40) or Corynebacterium glutamicum (41), whereas the two bacteria exhibit a dtd/DTD1-like gene. At least in such cases, sources of cellular D-tyrosine distinct from the L-tyrosine synthesis pathways have to be searched for.
At this stage, it is worth mentioning the finding of DL-dityrosine in the S. cerevisiae ascospore wall (42)(43)(44). The catabolic turnover of this dipeptide is likely to generate free D-tyrosine. The biosynthetic routes leading to DL-dityrosine-containing macromolecules have not yet been established. Epimerization of specific positions in the macromolecule or the direct incorporation of D-tyrosine have been envisaged (43). Such an incorporation of D-tyrosine would definitely support the idea of an active metabolic pool of this amino acid in the yeast cell.