Characterization of α-Ketoglutarate-dependent Taurine Dioxygenase from Escherichia coli *

The Escherichia coli tauD gene is required for the utilization of taurine (2-aminoethanesulfonic acid) as a sulfur source and is expressed only under conditions of sulfate starvation. The sequence relatedness of the TauD protein to the α-ketoglutarate-dependent 2,4-dichlorophenoxyacetate dioxygenase of Alcaligenes eutrophus suggested that TauD is an α-ketoglutarate-dependent dioxygenase catalyzing the oxygenolytic release of sulfite from taurine (van der Ploeg, J. R., Weiss, M. A., Saller, E., Nashimoto, H., Saito, N., Kertesz, M. A., and Leisinger, T. (1996) J. Bacteriol. 178, 5438–5446). TauD was overexpressed in E. coli to ∼70% of the total soluble protein and purified to apparent homogeneity by a simple two-step procedure. The apparent M r of 81,000 of the native protein and the subunit M rof 37,400 were consistent with a homodimeric structure. The pure enzyme converted taurine to sulfite and aminoacetaldehyde, which was identified by high pressure liquid chromatography after enzymatic conversion to ethanolamine. The reaction also consumed equimolar amounts of oxygen and α-ketoglutarate; ferrous iron was absolutely required for activity; and ascorbate stimulated the reaction. The properties and amino acid sequence of this enzyme thus define it as a new member of the α-ketoglutarate-dependent dioxygenase family. The pure enzyme showed maximal activity at pH 6.9 and retained activity on storage at −20 °C for several weeks. Taurine (K m = 55 μm) was the preferred substrate, but pentanesulfonic acid, 3-(N-morpholino)propanesulfonic acid, and 1,3-dioxo-2-isoindolineethanesulfonic acid were also desulfonated at significant rates. Among the cosubstrates tested, only α-ketoglutarate (K m = 11 μm) supported significant dioxygenase activity.

In the absence of sulfate, Escherichia coli can utilize aliphatic sulfonates as sulfur sources for growth. Sulfonates known to provide sulfur include ethanesulfonate, butanesulfonate, L-cysteate, isethionate (2-hydroxyethanesulfonate), and taurine (2-aminoethanesulfonate) (1,2). None of these sulfonates served as sulfur source under anaerobic conditions, nor could they be utilized as a source of carbon and energy or of carbon, energy, and sulfur under either aerobic or anaerobic conditions (1). The mechanisms of sulfur assimilation from aliphatic sulfonates are unknown, but it has been shown that sulfonate/sulfur enters the assimilatory sulfate reduction pathway at the stage of sulfite (3).
Recently, we have identified the tauABCD gene cluster, lo-cated at 8.5 min on the E. coli chromosome, which is specifically involved in the utilization of taurine as a sulfur source (2). Disruption of tauB, tauC, or tauD resulted in the loss of the ability to utilize taurine as a source of sulfur, but did not affect the utilization of a range of other aliphatic sulfonates as sulfur sources. The tau genes were only expressed during growth in the absence of sulfate or cysteine (2). The amino acid sequences of TauABC exhibit similarity to components of ABC-type transport systems (4). TauA has a putative signal sequence, indicating that it functions as a periplasmic binding protein, and the sequences of TauB and TauC are significantly similar to those of ATP-binding proteins and membrane components, respectively, of members of the ABC transporter superfamily. It thus appears that the proteins encoded by tauABC constitute an uptake system for taurine. The tauD gene product shows 30% sequence identity to a characterized ␣-ketoglutarate-dependent 2,4-dichlorophenoxyacetate dioxygenase of Alcaligenes eutrophus, encoded by the tfdA gene of plasmid pJP4 (5,6). This suggests that TauD is an ␣-KG 1 -dependent dioxygenase involved in the oxygenolytic release of sulfite from taurine. Here we report the purification of the TauD protein and describe its catalytic properties, which demonstrate that it is an ␣-KGdependent taurine dioxygenase.

EXPERIMENTAL PROCEDURES
Materials-Restriction endonucleases and T4 DNA ligase were obtained from MBI Fermentas. DNase I and NADH came from Boehringer Mannheim, Pfu DNA polymerase from Stratagene, and horse liver alcohol dehydrogenase from Sigma. Lysozyme was from Fluka, as were taurine and all other sulfonated substrates except N-phenyltaurine, 4-phenyl-1-butanesulfonic acid, and 2-bromoethanesulfonic acid (Sigma) and isethionic acid (Aldrich). All other chemicals were from Aldrich, Fluka, or Merck and were of the highest purity grade available.
Taurine Desulfonation in Extracts of E. coli Cells Grown in Sulfate and Taurine-E. coli strain MC4100 (9) was grown under the conditions mentioned above in 100 ml of a modified sulfur-free M63 minimal medium (2) containing 250 M Na 2 SO 4 or taurine as sulfur sources. When the cultures had reached an A 650 of 0.9, cells were harvested by centrifugation at 5800 ϫ g for 20 min at 4°C, washed once with modified M63 minimal medium, resuspended in 1 ml of 20 mM Tris-HCl buffer (pH 7.0), and stored at Ϫ20°C until further use. Cell-free extracts were obtained after the addition of 5 g/ml lysozyme and 10 g/ml DNase I to the cells, incubation at 30°C for 30 min, and clarification by centrifugation at 35,000 ϫ g for 20 min.
Construction of a tauD Expression Plasmid-The tauD gene was placed under the control of the T7 RNA polymerase promoter of vector pET-24a(ϩ) (Novagen) in a three-step cloning procedure. First, an NdeI site at the translation start of tauD was introduced by polymerase chain reaction amplification of the tauD gene from plasmid pUC18ALA4 (2,10). The oligonucleotide primers used were EE1 (5Ј-CATGGAGAAGT-CATATGAGTGAAC-3Ј, with the change to introduce the NdeI site underlined) and EE2 (5Ј-CGGTGCTCGAAAGCTTAGGTTCGA-3Ј). In the second step, the 1041-base pair polymerase chain reaction product was digested with NdeI and SphI, and the resulting 226-base pair fragment encoding the N-terminal part of TauD was cloned in pUC19 (11), generating plasmid pME4140. In the third step, the final cloning of tauD in the expression vector pET-24a(ϩ) was done by a ligation involving the vector, the 226-base pair NdeI-SphI fragment of pME4140, and an 815-base pair SphI-HindIII fragment encoding the C-terminal part of TauD. This resulted in plasmid pME4141, in which tauD is under the control of the T7 RNA polymerase promoter.
Purification of TauD-For protein production, E. coli BL21(DE3) cells harboring the taurine dioxygenase expression plasmid pME4141 were grown at 30°C in a 5-liter Erlenmeyer flask containing 1000 ml of growth medium. When the culture had reached an A 650 of 0.6 -0.8, tauD expression was induced by the addition of isopropyl-␤-D-thiogalactopyranoside to a 1 mM final concentration and incubated for a further 3 h. Cells were collected by centrifugation at 5800 ϫ g for 20 min at 4°C, resuspended in 50 ml of 20 mM Tris-HCl buffer (pH 7.0), and stored at Ϫ20°C in 10-ml aliquots until further use. About 6 -9 g of cells (fresh weight) were collected from 1000 ml of culture.
For cell lysis and DNA digestion, 5 g/ml lysozyme and 10 g/ml DNase I were added to 10 ml of cell suspension, and the mixture was incubated at 30°C with constant shaking (180 rpm) for 30 min. Crude cell-free extracts were obtained by centrifugation of the lysate at 35,000 ϫ g for 20 min at 4°C. The extract was then fractionated with solid ammonium sulfate at 4°C. The TauD protein precipitated between 50 and 60% (w/v) (NH 4 ) 2 SO 4 and was collected by centrifugation at 48,000 ϫ g for 30 min at 4°C. The precipitate was dissolved in 3.5 ml of 20 mM sodium phosphate buffer (pH 7.0) and dialyzed for 15 h at 4°C against 1 liter of the same buffer, with gentle stirring.
The dissolved ammonium sulfate precipitate was chromatographed at room temperature on a 1-ml Resource Q anion-exchange column (Pharmacia Biotech Inc.) with a BioCAD SPRINT™ apparatus (PerSeptive Biosystems Inc.) at a flow rate of 3 ml/min. Proteins were eluted using a three-step gradient of NaCl in 20 mM sodium phosphate buffer (pH 8.0): 2 column volumes with NaCl increasing from 0 to 60 mM, 5 column volumes with 60 mM NaCl, and 2 column volumes with NaCl increasing from 60 to 1000 mM. Fractions containing active TauD were pooled, concentrated with a Centriprep 10 concentrator (Amicon, Inc.), adjusted to 16% (v/v) glycerol, and stored at Ϫ20°C until further use.
Enzyme Assay-Taurine dioxygenase activity was assayed at 30°C by using Ellman's reagent (5,5Ј-dithiobis(2-nitrobenzoic acid)), which produces a bright yellow color upon reaction with sulfite (12,13). The assay mixture (1-ml final volume) contained 500 M taurine, 1 mM ␣-KG, 100 M Fe(II)SO 4 (freshly made), 200 M sodium ascorbate, and appropriate amounts of TauD in 10 mM imidazole buffer (pH 6.9). The pH of all potential substrate and cosubstrate stock solutions was adjusted to 6.9. Reactions were started by the addition of enzyme and stopped by the addition of 800-l samples to a spectrophotometric cuvette containing 100 l of EDTA (0.5 M), 100 l of 5,5Ј-dithiobis(2nitrobenzoic acid) (1 mg/ml in 100 mM sodium phosphate buffer (pH 7.0)), and 100 l of deionized water. The colorimetric reaction was allowed to develop at room temperature for 3 min. To correct for unspecific reactions with enzyme thiol groups, formaldehyde was added (5 mM final concentration) to samples containing both sulfite and thiols. This led to complexation of the sulfite as formaldehyde sulfoxylate (12), after which it did not react with 5,5Ј-dithiobis(2-nitrobenzoic acid). Residual absorbance at 415 nm in these samples was due to thiols only and was used to make the appropriate correction. Sulfite calibration curves were prepared daily from fresh solutions of sodium sulfite in deionized water. Sulfite determination with Ellman's reagent was linear up to 110 M sulfite. One unit of activity is defined as the amount of enzyme forming 1 mol of sulfite/min at 30°C under standard assay conditions.
The K m values for taurine and ␣-KG were determined by keeping the concentration of one substrate constant at 600 M while varying the concentration of the other substrate (25-500 M taurine or 10 -100 M ␣-ketoglutarate). Sulfite production was followed over 3 min (steadystate conditions) by removal of aliquots every 15 or 30 s. Desulfonation rates were calculated by regression analysis.
Analytical Methods-Oxygen consumption during TauD catalysis was measured with an oxygen electrode (Rank Brothers, Bottisham, Cambridgeshire, United Kingdom) under standard assay conditions in a 500-l reaction volume. Protein concentrations were measured using the method of Bradford (14) with Bio-Rad reagent dye concentrate, following the manufacturer's instructions. Bovine serum albumin was used as a standard. The native M r of TauD was estimated by gel filtration on Superose 12 HR 10/30 and Superose 6 HR 10/30 columns (Pharmacia Biotech Inc.), calibrated with known M r protein standards: thyroglobulin (670,000), ferritin (440,000), catalase (232,000), aldolase (158,000), bovine serum albumin (67,000), ovalbumin (43,000), and RNase (13,700). The column was equilibrated (2 column volumes) and eluted with 50 mM sodium phosphate buffer (pH 7.2) containing 150 mM NaCl at a flow rate of 0.5 ml/min.
Analysis of Enzyme Reaction Products-Succinate was measured by an enzyme assay using a succinic acid test kit from Boehringer Mannheim, following the instructions of the manufacturer. The formation of aminoacetaldehyde, the unstable product of taurine desulfonation, was indirectly and qualitatively detected by its reaction with horse liver alcohol dehydrogenase (EC 1.1.1.1) in a coupled assay, which yielded substoichiometric amounts of ethanolamine. The coupled assay (1-ml final volume) contained 1 mM taurine, 1 mM ␣-KG, 100 M Fe(II)SO 4 (freshly prepared), 200 M sodium ascorbate, 175 M NADH, 0.5 units of taurine dioxygenase, and 3.2 units of horse liver alcohol dehydrogenase in 30 mM sodium phosphate buffer (pH 6.9). Ethanolamine production from taurine was followed as NADH oxidation by measuring the absorbance at 340 nm with time. An assay mixture with no alcohol dehydrogenase served as a reference. The presence of ethanolamine in the samples was detected, after phenyl isothiocyanate derivatization, by HPLC co-chromatography with an authentic ethanolamine derivative. To this end, the material in 600 -1600-l samples was lyophilized and redissolved in 30 l of 10 mM sodium phosphate buffer (pH 6.5). The dissolved material was then treated with phenyl isothiocyanate by the addition of 1 volume of absolute ethanol and 2.5 volumes of derivatization mixture (7:2:1 ethanol/triethylamine/phenyl isothiocyanate). After a 10-min reaction at room temperature, the derivatized sample was lyophilized, resuspended in 200 l of 10 mM sodium phosphate buffer (pH 6.5), filtered (0.2 m), and chromatographed on a C 18 -Nucleosil column (250 ϫ 4.6 mm, particle size of 7 m) at a flow rate of 1 ml/min (Pharmacia Biotech HPLC system). A two-component buffer system (buffer A: 10 mM sodium phosphate buffer (pH 6.5); buffer B: 80% methanol and 20% 50 mM sodium phosphate buffer (pH 6.5)) was used with the following gradient: column equilibration for 10 min with 10% buffer B, from 10 to 100% buffer B in 10 min, and 100% buffer B for 5 min. Elution of taurine and ethanolamine derivatives was monitored by measuring the absorbance at 254 nm.

Desulfonation of Taurine in Crude Cell
Extracts-Preliminary analysis of taurine desulfonation was performed with crude extract prepared from isopropyl-␤-D-thiogalactopyranoside-treated cells of E. coli BL21(DE3)(pME4141), which strongly overexpress the tauD gene product (Fig. 1). Desulfonation was assayed by determining the amount of sulfite formed during 15 min of incubation at 30°C in 10 mM imidazole (pH 6.75) containing 10 mM taurine, 10 mM ␣-KG, 50 M Fe(II)SO 4 , and 100 M L-ascorbate. Sulfite release and oxygen consumption were not detected in cell extracts prepared from uninduced cells or in extracts from induced cells incubated without taurine or ␣-KG. Activity in cell extracts was abolished by the addition of EDTA, and desulfonation of taurine thus required ␣-KG, ferrous iron, and oxygen. It therefore seemed likely that the TauD protein represents a new member of the ␣-KGdependent dioxygenase enzyme family (15).
The cell extract from E. coli MC4100 was examined for taurine desulfonation by following taurine-dependent sulfite release. Strain MC4100 carries a single chromosomal copy of the tau genes, which are fully expressed during growth in sulfate-free minimal medium containing 250 M taurine as a sulfur source (2). Extracts made from cells grown under these conditions showed a specific taurine desulfonating activity of 12.2 nmol/min/mg of protein, which corresponds to about twice the in vivo desulfonation rate required for growth at a doubling time of 60 min with taurine as a sulfur source. As expected, extracts from sulfate-grown cells showed no taurine desulfonating activity.
Enzyme Purification-The tauD gene product was purified to homogeneity from E. coli BL21(DE3)(pME4141) with a yield of 20% in a two-step procedure summarized in Table I. The specific taurine desulfonating activity obtained with crude extracts after overexpression was 60 times that measured in wild-type E. coli MC4100 cells grown in minimal medium containing taurine as a sulfur source. TauD eluted as a single symmetrical peak from the Resource Q anion-exchange column at a NaCl concentration of 15 mM. Densitometric scanning of SDS-polyacrylamide gels (Fig. 1) showed that TauD amounted to ϳ70% of the total soluble protein in crude extracts. Gel filtration chromatography on Superose 6 and Superose 12 HR columns was used to estimate a M r of 81,000 Ϯ 4400 for the native enzyme. The calculated subunit molecular mass from the tauD gene sequence of 32.41 kDa was estimated by SDSpolyacrylamide gel electrophoresis analysis as 37.4 Ϯ 1.4 kDa. Since mostly mono-and dimeric (but no trimeric) structures are reported among ␣-KG-dependent dioxygenases (16) and with regard to the sequence identity to 2,4-dichlorophenoxyacetate dioxygenase (6), our data suggested a homodimeric structure for the TauD enzyme.
Characterization of TauD-The purified enzyme had a specific activity of 1.64 units/mg of protein. Upon storage at Ϫ20°C at a protein concentration of 2.6 mg/ml in buffer without glycerol, Ͼ50% of the activity was lost within 3 weeks. When glycerol was added to a 16% (v/v) final concentration, the specific activity of the pure enzyme increased by ϳ4-fold during storage at Ϫ20°C for 10 weeks. We presume that the activation during storage reflects folding of unfolded polypeptide chains.
The effect of pH on enzyme activity was examined over a range of 4.5-10.8 using appropriate buffer systems (17). TauD exhibited a distinct activity optimum around pH 6.9.
In our initial studies with TauD, described above, we observed that the specific activity of the enzyme fell rapidly during incubation at 30°C. A similar effect has been seen before with the related ␣-KG-dependent 2,4-dichlorophenoxyacetate dioxygenase (6), so we investigated it further by measuring sulfite release from taurine over 30 min by enzyme that had been preincubated in imidazole buffer at 30°C for 15 min alone, with Fe(II)SO 4 , or with Fe(II)SO 4 and ascorbate. The data (not shown) indicate that temperature itself leads to rapid enzyme inactivation, which is enhanced by ascorbate, and that the observed inactivation is not due to oxidation of the enzymebound ferrous iron.
The dependence of TauD activity on the concentration of ferrous iron and ascorbate was investigated with pure enzyme under standard assay conditions. Enzyme activity was dependent on Fe(II) supplied as either a sulfate or chloride salt, and treatment of the enzyme with EDTA completely abolished activity. Maximal specific activity was obtained with Fe(II) concentrations between 5 and 150 M. Other divalent metal ions, including Mg(II), Ca(II), Mn(II), Ni(II), Co(II), Cu(II), and Zn(II), could not replace ferrous sulfate at a 100 M final concentration either as sulfate or chloride salts. Zn(II), Cu(II), and, to a lesser extent, Co(II), all added at 10 -50 M final concentrations, inhibited maximal taurine/␣-KG dioxygenase activity by 80 -95%. Ascorbate led to a 50% increase in activity when added at concentrations between 200 and 800 M in an assay mixture containing 100 M Fe(II).
Stoichiometry of the ␣-KG-dependent Taurine Desulfonation Reaction-Of the four products formed in the taurine/␣-KG dioxygenase reaction (Fig. 2), succinate and sulfite were determined quantitatively and shown to be produced in equimolar amounts. This is in accordance with the expected stoichiometry of the reaction and demonstrated that oxidative decarboxylation of ␣-KG and desulfonation of taurine are strictly coupled. The expected consumption of 1 mol of dioxygen/mol of taurine desulfonated was also experimentally verified.
The putative organic product of the TauD-catalyzed taurine oxygenation is 1-hydroxy-2-aminoethanesulfonic acid, which would decompose to aminoacetaldehyde and sulfite (Fig. 2). Attempts by gas chromatography-mass spectrometry to detect aminoacetaldehyde among the products resulting from the enzymatic desulfonation of taurine were unsuccessful. Since aminoacetaldehyde is highly reactive and readily undergoes polymerization reactions, we chose an indirect method to demonstrate qualitatively its formation from taurine. As described under "Experimental Procedures," liver alcohol dehydrogenase and NADH were added to the complete reaction mixture, and reduction of the desulfonation product by alcohol dehydrogenase was monitored by following NADH oxidation. NADH oxidation occurred in the complete incubation mixture, but was not observed when taurine dioxygenase, alcohol dehydrogenase, taurine, or ␣-KG was omitted (Fig. 3). The identity of ethanolamine as the product indirectly formed from taurine desulfonation was verified after phenyl isothiocyanate derivatization by HPLC co-chromatography with authentic ethanolamine derivative. The dehydrogenase reaction was found to be the rate-limiting step in the coupled overall reaction. Irrespective of the concentration of alcohol dehydrogenase, only substoichiometric amounts of ethanolamine were formed from tau- rine, i.e. maximally 40% of the aminoacetaldehyde produced was reduced to ethanolamine.
Substrate Range and Kinetic Constants-The enzyme showed a Michaelis-Menten-type saturation curve in response to increasing taurine concentrations, with a K m of 55 M and a V max of 4.1 units/mg of protein (Fig. 4A). Among 19 potential substrates tested, methanesulfonic acid, ethanesulfonic acid, isethionic acid, 2-bromoethanesulfonic acid, L-cysteic acid, sulfosuccinate, 4-aminobenzenesulfonic acid, 2-(4-pyridyl)ethanesulfonic acid, and N-phenyltaurine were not utilized by TauD. 2-5% of the activity observed with taurine was observed with 1-propanesulfonic acid, 1-dodecanesulfonic acid, 4-phenyl-1butanesulfonic acid, HEPES, and PIPES. Significant activity was supported by the compounds listed in Table II. The observed substrate range indicates that in addition to taurine, several 2-substituted ethanesulfonic acids are substrates for ␣-KG-dependent taurine dioxygenase. Among the unsubstituted 1-alkanesulfonates, pentanesulfonic acid was the best substrate.
The K m value for ␣-KG was 11 M (Fig. 4B). Using 30-min incubation times and 33 milliunits of enzyme, ␣-ketoadipate, pyruvate, ␣-ketobutyrate, ␣-ketovalerate, ␣-ketocaproate, ␣-ketoisovalerate, and oxalacetate were tested as cosubstrates for taurine desulfonation. When tested at 500 M to 10 mM, ␣-ketoadipate supported desulfonation at 4 -10% of the rate observed with ␣-KG, whereas the other ␣-keto acids were inactive. A second carboxyl group is thus required for recognition of the cosubstrate by the enzyme. DISCUSSION Although several reports on the degradation of primary aliphatic sulfonates by bacteria as carbon or sulfur sources have appeared (1,2,18,19), the enzymology of the desulfonation of these compounds is largely unexplored. The taurine desulfonation enzyme characterized in this study is, to our knowledge, the first pure enzyme reported that is capable of oxygenolytic cleavage of the C-S bond of 1-alkanesulfonates. In contrast to the partially purified bacterial monooxygenase systems for alkanesulfonate desulfonation described so far (18,19), it does not play a role in carbon metabolism, and its synthesis is regulated by the sulfur supply to the cell (2).
The biochemical properties of taurine dioxygenase allow it to be assigned to the ␣-KG-dependent dioxygenase group of enzymes (15,16). As a family, these enzymes catalyze a variety of reactions including hydroxylations, desaturations, and ring expansions. The family contains enzymes that are monomers, dimers, and tetramers with subunit sizes ranging between 26 and 85 kDa. Characterization of the reactions catalyzed by some of these enzymes revealed that not all show typical hydroxylation biochemistry in which ␣-KG acts as an oxygen acceptor and one atom of molecular oxygen is found in succinate and the other in the second organic product. The classifi-  cation defined by biochemistry does not satisfactorily correlate with sequence data since the group contains both enzymes with similar biochemical characteristics but very little sequence similarity and enzymes with a high degree of amino acid sequence similarity but more diverse biochemical features. All these enzymes could have a common mechanism of action (16). ␣-KG-dependent taurine dioxygenase exhibits a relatively relaxed substrate specificity, accepting not only taurine but also pentanesulfonic acid, hexanesulfonic acid, MOPS, and 1,3dioxo-2-isoindolineethanesulfonic acid (Table II). The first three of these alternative substrates have been shown to serve as sulfur sources for E. coli. Since E. coli mutants defective in taurine dioxygenase are still able to utilize these sulfur sources, there must be at least one additional system for their metabolism as sulfur sources (2). Desulfonation by ␣-KGdependent taurine dioxygenase of organosulfonates other than taurine might thus be of little or no importance for sulfur metabolism. To strengthen or refute this view, it will be interesting to explore the specificity of the TauA periplasmic binding protein, which is coexpressed with TauD, and to characterize the system or systems that catalyze desulfonation of other alkanesulfonates.
With respect to cosubstrate specificity, the taurine desulfonation enzyme of E. coli occupies an intermediate position among the ␣-KG-dependent dioxygenases. Besides ␣-KG, only ␣-ketoadipate, but none of the other ␣-keto acids tested, supported a low level of desulfonation. The range of cosubstrates is thus more restricted than that of 2,4-dichlorophenoxyacetate dioxygenase (6), similar to the range of prolyl 4-hydroxylase (20), and less restricted than that of ␥-butyrobetaine hydroxylase, which utilizes only ␣-KG as a cosubstrate (21).
Other features of the E. coli ␣-KG-dependent taurine dioxy-genase, such as the requirement for ␣-KG and ferrous iron, the stimulation of activity by ascorbate, and its inhibition by divalent metal ions, are typical for ␣-KG-dependent dioxygenases (15,16). Unlike lysyl hydroxylase and prolyl 4-hydroxylase, taurine dioxygenase does not appear to catalyze an uncoupled decarboxylation of ␣-KG in the absence of taurine. Uncoupled ␣-KG decarboxylation also occurs in lysyl hydroxylase and prolyl 4-hydroxylase as a side reaction during substrate hydroxylation. It is thought to lead to the oxidation of enzymebound Fe(II), which subsequently must be reduced by ascorbate to reactivate the enzyme. The side reaction thus results in the absolute requirement of these enzymes for ascorbate (22,23), a property that was not observed with ␣-KG-dependent taurine dioxygenase. Taurine dioxygenase exhibited 30% amino acid identity to the 2,4-dichlorophenoxyacetate dioxygenase enzyme encoded by the tfdA gene of plasmid pJP4 (6,24) as well as 32 and 37% sequence identity to open reading frames of unknown function of Saccharomyces cerevisiae (GenBank™ accession number Z47973) and Mycobacterium tuberculosis (GenBank™ accession number Z77165), respectively (Fig. 5). It did not show significant similarity to other ␣-KG-dependent dioxygenases (16), emphasizing the genetic diversity of this group of enzymes and suggesting that they may have arisen by convergent evolution rather than from a common ancestor. While taurine dioxygenase and 2,4-dichlorophenoxyacetate dioxygenase did not react with each other's substrates, 2,3 , they are both dimers, with subunit molecular masses of 32 kDa, that exhibit maximal activity at pH 6.5-7.0. The alignment of the four proteins that 2 R. P. Hausinger, personal communication. 3 E. Eichhorn, unpublished data.
FIG. 5. Alignment of amino acid sequences from TauD, TfdA, and related proteins. TauD is taurine dioxygenase from E. coli (Swissprot accession number P37610) (2). TfdA is 2,4-dichlorophenoxyacetate dioxygenase from A. eutrophus (Swissprot accession number P10088) (6). Y is a hypothetical protein of unknown function from M. tuberculosis (GenBank™ accession number Z77165; open reading frame MTCY78.22c), and YЈ is a hypothetical protein of unknown function from S. cerevisiae (GenBank™ accession number Z47973; reverse complement translated from bases 13413 to 14650; only the sequence portion overlapping with the other three sequences is shown). Amino acid sequences were aligned using PILEUP (30). Conserved residues are boxed. Conserved His and Asp residues are marked with asterisks. do exhibit sequence homology (Fig. 5) revealed 4 histidine and 2 aspartate residues that are strictly conserved (marked with asterisks). Site-directed mutagenesis studies of human prolyl 4-hydroxylase (25) and lysyl hydroxylase (26) have indicated that 2 histidines and 1 aspartate constitute the endogenous ligands of the ferrous active site. In these ␣-KG-dependent dioxygenases and in the mechanistically related isopenicillin-N synthase, these residues are arranged in the His-X-Asp-X 53-57 -His motif (27), whose role in metal binding is supported by the crystal structure of the manganese form of a fungal isopenicillin-N synthase (28). Although the enzymes shown in Fig. 5 are not closely related to the above enzymes at a sequence level (Ͻ9% identity), a similar iron-binding motif might be provided by histidine 99, aspartate 101, and histidine 153 of the TauD sequence (Fig. 5). For the 2,4-dichlorophenoxyacetate dioxygenase enzyme, recent spectroscopic studies have indeed provided evidence that these residues may play a role in metal coordination (29). The function of the unknown open reading frames Y and YЈ (Fig. 5) cannot be speculated from these data, but it would be interesting to clarify if they also belong to the ␣-KG-dependent dioxygenase family.