Binding of Sulfurated Molybdenum Cofactor to the C-terminal Domain of ABA3 from Arabidopsis thaliana Provides Insight into the Mechanism of Molybdenum Cofactor Sulfuration*

The molybdenum cofactor sulfurase ABA3 from Arabidopsis thaliana is needed for post-translational activation of aldehyde oxidase and xanthine dehydrogenase by transferring a sulfur atom to the desulfo-molybdenum cofactor of these enzymes. ABA3 is a two-domain protein consisting of an NH2-terminal NifS-like cysteine desulfurase domain and a C-terminal domain of yet undescribed function. The NH2-terminal domain of ABA3 decomposes l-cysteine to yield elemental sulfur, which subsequently is bound as persulfide to a conserved protein cysteinyl residue within this domain. In vivo, activation of aldehyde oxidase and xanthine dehydrogenase also depends on the function of the C-terminal domain, as can be concluded from the A. thaliana aba3/sir3-3 mutant. sir3-3 plants are strongly reduced in aldehyde oxidase and xanthine dehydrogenase activities due to a substitution of arginine 723 by a lysine within the C-terminal domain of the ABA3 protein. Here we present first evidence for the function of the C-terminal domain and show that molybdenum cofactor is bound to this domain with high affinity. Furthermore, cyanide-treated ABA3 C terminus was shown to release thiocyanate, indicating that the molybdenum cofactor bound to the C-terminal domain is present in the sulfurated form. Co-incubation of partially active aldehyde oxidase and xanthine dehydrogenase with ABA3 C terminus carrying sulfurated molybdenum cofactor resulted in stimulation of aldehyde oxidase and xanthine dehydrogenase activity. The data of this work suggest that the C-terminal domain of ABA3 might act as a scaffold protein where prebound desulfo-molybdenum cofactor is converted into sulfurated cofactor prior to activation of aldehyde oxidase and xanthine dehydrogenase.

otic molybdenum enzymes, the molybdenum atom is coordinated by the dithiolene group of molybdopterin, thus forming the molybdenum cofactor (Moco) 2 (2). According to the coordination chemistry of the molybdenum ligand, eukaryotic molybdenum enzymes can be divided into two groups; Moco with two additional oxo-ligands and a protein-derived cysteinyl sulfur is bound by enzymes of the sulfite oxidase family, whereas enzymes of the xanthine oxidase family have one oxygen, one inorganic sulfur, and one hydroxyl group ligated to the pterin-chelated molybdenum of the active enzyme. Among the four different molybdenum enzymes known in higher plants, sulfite oxidase and nitrate reductase belong to the sulfite oxidase family, whereas aldehyde oxidase (AO) and xanthine dehydrogenase (XDH) are members of the xanthine oxidase family (3). Although it is believed that all of these molybdenum enzymes basically incorporate the same type of Moco, only AO and XDH, but not enzymes of the sulfite oxidase family, require a final enzyme-dependent post-translational modification of the molybdenum center for activity (4). During this modification step, an oxo-ligand of the Moco in inactive AO and XDH enzymes is substituted by a sulfur atom in order to activate AO and XDH.
The first insight into the mechanism of Moco sulfuration was obtained by cloning and biochemical characterization of the Moco sulfurase protein ABA3 from Arabidopsis thaliana (5,6). The NH 2 -terminal domain of ABA3 (ABA3-NifS) shares significant similarities to NifS-like cysteine desulfurases, whereas the C-terminal domain did not exhibit striking similarities to any other protein, except other Moco sulfurases and the recently identified mitochondrial amidoxime-reducing component (7). It was shown that, as typical for NifS-like enzymes, ABA3-NifS binds a pyridoxal phosphate cofactor that is essential for activity (6). Furthermore, L-cysteine and L-selenocysteine are decomposed by ABA3-NifS, with L-cysteine representing the preferred substrate with a K m value 4 times lower than that for the selenium substrate. During the decomposition of L-cysteine, L-alanine is released, and elemental sulfur is generated. The sulfur is immediately coupled to a conserved cysteine res-idue of ABA3-NifS, thus forming a protein-bound persulfide. Co-incubation of purified ABA3-NifS and cyanide-inactivated AO␣ from A. thaliana as target enzyme in the presence of L-cysteine resulted in activation of the AO␣ protein, indicating that the persulfide sulfur was transferred from the NifS-like domain of ABA3 to the Moco of AO␣. In vitro, the presence of the C-terminal domain is not required for sulfuration of xanthine oxidase family enzymes; however, there is strong evidence that it is needed in vivo. The tomato flacca mutant with a mutation in the Moco sulfurase C-terminal domain (8) is strongly reduced in root AO and XDH activities and does not reveal any activities in the shoots (9). Very recently, the A. thaliana mutant sir3-3 was isolated by a sirtinol resistance screen, which is based on reduced AO activities (10). sir3-3 was found to have a single point mutation in the C-terminal domain of the aba3 coding sequence, likewise indicating that the function of the C-terminal domain is required for proper activation of AO (and XDH). It was hypothesized that the function of the C-terminal domain of Moco sulfurases is related to the recognition of the respective target enzymes or one of the protein-bound cofactors (5,11). Other authors discussed the possibility of the C-terminal domain functioning as sulfur mediator during Moco sulfuration (8,12).
In this work, we report the biochemical and functional characterization of the C-terminal domain (ABA3-CT) of the Moco sulfurase ABA3 from A. thaliana. We assayed the protein for bound molecules that might be transferred to the ABA3 target enzymes AO and XDH. Moreover, the effect of the sir3-3 mutation was studied in plants as well as on the level of the recombinant protein. Finally, we discuss a possible role for ABA3-CT during Moco sulfuration.

Construction of Expression Vectors-A 921-bp open read-
ing frame encoding the C-terminal domain of ABA3 was subcloned into pQE80 (Qiagen, Hilden, Germany) as described before (6). The amino acid exchange R723K was introduced into ABA3-CT by PCR-based site-directed mutagenesis, and the respective cDNA was cloned into the BamHI site of pQE80, resulting in an NH 2 -terminal His 6 tag fusion of ABA3-CT/R723K.
Expression and Purification of ABA3-CT, ABA3-CT/R723K, AO␣, and AtXDH1 from A. thaliana-Expression of ABA3-CT and ABA3-CT/R723K was performed in freshly transformed Escherichia coli TP1000 cells (13). Cells were grown aerobically in LB medium in the presence of 100 g/ml ampicillin at 22°C to an A 600 ϭ 0.1 before induction with 15 M isopropyl-␤-Dthiogalactopyranoside and the addition of 1 mM sodium molybdate. After induction, cells were grown for a further 20 h at 22°C. Expression in E. coli RK5206 and RK5204 (14) was done likewise but in the absence of sodium molybdate. Overexpression of recombinant His 6 -tagged AO␣ (kindly provided by T. Koshiba, Tokyo, Japan) and AtXDH1 in the yeast Pichia pastoris was performed as described in Refs. 15 and 16, respectively. Cells were harvested by centrifugation and stored at Ϫ70°C until use. Cell lysis was achieved by several passages through a French pressure cell followed by sonication for 5 min. After centrifugation, His 6 -tagged protein was purified on a nickel-nitrilotriacetic acid superflow matrix (Qiagen, Hilden, Germany) under native conditions at 4°C according to the manufacturer's manual and eluted in elution buffer (50 mM sodium phosphate, pH 6.0, containing 300 mM sodium chloride, 250 mM imidazole, 10% glycerol). Eluted fractions were electrophoretically separated on 12% (for ABA3-CT) or 7.5% (for AO␣ and AtXDH1) SDS-polyacrylamide gels and stained by Coomassie Brilliant Blue.
Nit-1 NADPH Nitrate Reductase Reconstitution-Extracts from the Neurospora crassa nit-1 mutant were prepared as described in Ref. 17 and stored in aliquots at Ϫ70°C. All reconstitutions were performed in nit-1 buffer (50 mM sodium phosphate, 200 mM NaCl, and 5 mM EDTA, pH 7.2) in the presence of 2 mM reduced glutathione and 5 mM sodium molybdate where appropriate. The reconstitution assay was performed in a 40-l reaction volume containing 20 l of gel-filtrated nit-1 extract. Complementation was carried out anaerobically for 2 h at room temperature. After the addition of 20 mM NADPH and incubating for 10 min, reconstituted NADPH-nitrate reductase activity was determined as described in Ref. 17.
Chemical Detection of Moco and Molybdopterin-Moco and its metal-free precursor molybdopterin (MPT) were detected and quantified by converting them to the stable oxidation product FormA-dephospho according to Ref. 18. Oxidation, dephosphorylation, QAE chromatography, and HPLC analysis were performed as described in detail in Ref. 19. FormA-dephospho was quantified by comparison with a standard isolated from xanthine oxidase for which the absorptivity was ⑀ 380 ϭ 13,200 M Ϫ1 cm Ϫ1 (18).
Enzyme Assays-Plant material was squeezed at 4°C in 2 volumes of extraction buffer (100 mM potassium phosphate, 2.5 mM EDTA, 5 mM dithiothreitol, pH 7.5), sonicated, and centrifuged. Enzyme activities of AO and XDH in plant extracts were detected by activity staining after native PAGE, as previously described in Ref. 20 for AO and in Ref. 16 for XDH. In vitro superactivation of recombinant AO␣ and AtXDH1 by ABA3-CT was performed aerobically in a total volume of 40 l of 20 mM Tris/HCl, pH 8.0. 0.5 g of AO␣ or 2 g of AtXDH1 were incubated with 10, 20, and 50 g of Moco-loaded ABA3-CT for 30 min at 22°C, followed by native PAGE and activity staining with indole-3-carboxaldehyde as substrate for AO␣ or hypoxanthine as substrate for AtXDH1 according to Refs. 15 and 16. The relative densities of the resulting activity bands were determined by using ImageJ software version 1.38 from the National Institutes of Health (available on the World Wide Web).
Immunoblot Analysis-For immunoblot analysis, AO␣ and ABA3-CT proteins were excised from 7.5% native polyacrylamide gels and subjected to 12% SDS-PAGE. After gel blotting, a primary monoclonal anti-His 6 antibody from mouse (1:1000 dilution) and a secondary horseradish peroxidase-conjugated anti-mouse Ig was used (Sigma; 1:10,000 dilution) to detect chemiluminescence by using the ECL system (Amersham Biosciences).
Determination of Protein Concentrations-Concentrations of total soluble protein were determined by use of Roti Quant solution (Roth, Karlsruhe, Germany) according to Ref. 21.
Wave Scan of ABA3-CT-Absorption spectroscopy was carried out in elution buffer using an Ultrospec 2100 pro (Amersham Biosciences).
Inductively Plasma-coupled Mass Spectrometry (ICP-MS)-For analyses of the molybdenum concentrations, a temperature-and pressure-controlled microwave digestion was performed (CEM Mars 5). The samples were prepared in closed vessels by HNO 3 with a concentration of 6% and heated up in three steps to 200°C (120, 160, and 200°C). The temperature of 200°C was maintained for 10 min. For measurement by ICP-MS, samples were diluted to 1% HNO 3 . A Micromass Platform ICP-MS was used. The hexapole ion optic cell was rinsed with hydrogen as a collision gas and with helium as a deceleration gas (both 4.1 ml/min). The hexapole bias was set to Ϫ2.0 V. The rate of plasma flow was 15.4 liters/min with an intermediate gas of 1.45 liters/min and a nebulizer gas flow rate of 0.94 liters/ min. The reference factor power was 1.3 kilowatts. The sample was introduced by using a Meinhard nebulizer (pumped) at a flow rate of 1 ml/min. The instrument was calibrated for the molybdenum masses 92, 95, 96, and 98 g/mol with plasma standard solutions in 1% HNO 3 by a five-point calibration up to 200 ppb. The detection limits are 0.17, 0.18, 0.23, and 0.25 ppb, respectively. As an internal reference 103-Rh was used (10 ppb).
Quantification of Cyanolyzable Sulfur-To ensure that ABA3-CT is completely free from sulfur-containing molecules that may be bound to the cysteinyl residues of ABA3-CT and give nonspecific background, the protein was preincubated with 10 mM Tris(2-carboxyethyl) phosphine hydrochloride overnight at 4°C. Tris(2-carboxyethyl) phosphine hydrochloride was removed by gel filtration on a Sephadex G-50 Nick gel filtration column (Amersham Biosciences) equilibrated in 0.1 M Tris acetate, pH 8.6. Subsequently, 450 l of ABA3-CT or ABA3-CT/R723K were incubated with 50 l of 0.5 M potassium cyanide overnight at 22°C. Formed SCN Ϫ was separated from the protein with a 5-kDa cut-off Vivaspin concentrator (Sartorius, Göttingen, Germany) and 450 l of ferric nitrate reagent (100 g of Fe(NO 3 ) 3 ϫ 9H 2 O and 200 ml of 65% HNO 3 per 1500 ml) were added to 450 l of the flow-through. The FeSCN complex was quantified at 460 nm using a KSCN standard curve.
Binding of Moco to ABA3-CT-As a Moco/MPT source, Moco carrier protein MCP of Chlamydomonas reinhardtii was chosen, since it was shown to specifically bind Moco and not MPT (22). Expression was performed as described in Ref. 22. The purified protein was heat-denatured at 90°C for 5 min, and after centrifugation, the supernatant was used as a Moco/MPT source for binding studies. 0 -12 M free Moco/MPT was incubated at 4°C for 15 min either with 6 M Moco/MPT-free ABA3-CT expressed in E. coli RK5204 (14) or in a control sample without protein. After incubation, unbound Moco/MPT was separated from ABA3-CT by centrifugation with a 10,000 molecular weight cut-off Vivaspin concentrator (Sartorius, Göttingen, Germany). The Moco/MPT-containing flowthrough was converted to the stable oxidation product FormAdephospho and quantified as described above. The amount of Moco/MPT bound to ABA3-CT and the concentrations of free ABA3-CT were determined according to the calculated concentrations of free and total Moco/MPT and the known concentration of total ABA3-CT. The k D value was calculated according to the assumption of one Moco-binding site per ABA3-CT monomer and based on the concept that at equilibrium, the population of ABA3-CT molecules ([A] Total ) will be split between free ABA3-CT ( Stabilization of Moco/MPT by ABA3-CT-A stabilizing effect on Moco/MPT by ABA3-CT was analyzed with 4.8 mg of protein in 600 l of elution buffer incubated at 4°C or 22°C, respectively. For each approach, two samples of equal volume (10 l each) were taken at the indicated time points, and FormA-dephospho was determined for each sample as described above.
Chemical Sulfuration of Plant Extracts-Reconstitution of AO and XDH activities in plant extracts was performed according to Ref. 23. 1 g of leaf material was squeezed at 4°C in 2 ml of extraction buffer (100 mM potassium phosphate, 2.5 mM EDTA, and 5 mM dithiothreitol, pH 7.5), sonicated, and centrifuged. Ammonium sulfate was added to 1.5 ml of the supernatant to a final concentration of 40%. After centrifugation, the ammonium sulfate-precipitated proteins were resolved in 400 l of extraction buffer and desalted on Sephadex G-50 Nick columns. 0.4 ml of the desalted extract as well as 0.5 M solutions of sodium sulfide and sodium dithionite were made anaerobic by degassing under vacuum and purging with nitrogen. For reconstitution, sulfide and dithionite were added through a septum to the anaerobic extracts to final concentrations of 20 mM. The reaction mixture was incubated for 30 min at 37°C and desalted on Sephadex G-50 Nick columns before protein concentrations were determined. Finally, 150 g of total protein of each extract were subjected to native PAGE and stained for AO and XDH activity as described above.

Heterologous Expression and Characterization of ABA3-CT-
ABA3-CT was expressed in E. coli TP1000, yielding a monomeric His 6 -tagged protein with a molecular mass of ϳ35 kDa. Since the E. coli strain TP1000 is characterized by high accumulation of the eukaryotic form of Moco, purified ABA3-CT was analyzed for its ability to bind Moco or its metal-free precursor MPT, respectively. In fact, FormA-dephospho, the stable oxidation product of both Moco and MPT, could be identified on ABA3-CT expressed in TP1000 (Table 1). FormA-dephospho was also detected when ABA3-CT was expressed under identical conditions in the MPT-accumulating and Moco-free E. coli strain RK5206. However, the amount was reduced to 10 -20% of that of the protein derived from TP1000 (data not shown).
Identification of Sulfurated Moco on ABA3-CT-ABA3-CT was analyzed for biologically active Moco by use of the nit-1 reconstitution assay. This assay is based on the transfer of Moco derived from an exogenous source to the nitrate reductase apoprotein of the Moco-deficient N. crassa nit-1 mutant, whereby reconstitution of NADPH-dependent nitrate reductase activity is achieved (15). When omitting supplementary molybdate from the reaction mixture, only active Moco and not molybdenum-free MPT can be detected, since only Moco is capable of reconstituting NADPH-nitrate reductase activity. In the presence of molybdate, however, MPT is nonenzymatically converted to Moco, thereby enabling reconstitution of NADPH-nitrate reductase activity as well. ABA3-CT as purified after heterologous expression in E. coli TP1000 yielded a specific NADPH-nitrate reductase activity of 533 Ϯ 46 nmol nitrite/(mg⅐min) in the absence of molybdate, whereas in the presence of molybdate, a specific activity of 1411 Ϯ 150 nmol nitrite/(mg⅐min) was observed (n ϭ 3). Although the majority of bound pterin obviously is represented by nit-1-inactive cofactor, these results indicate that more than 30% of the total cofactor bound to ABA3-CT is represented by nit-1-active Moco, which was able to reconstitute NADPH-nitrate reductase activity in the nit-1 extract.
Upon quantifying molybdenum contents of ABA3-CT by ICP-MS, 25% of the ABA3-CT monomers were found to contain molybdenum when the cells were expressed in the presence of 1 mM molybdate ( Table 1). The same protein fractions showed an average FormA saturation of about 35%, indicating that at least two-thirds of the MPT molecules bound to ABA3-CT contain molybdenum and thus resemble Moco. Since from these experiments it was not clear whether the Moco bound to ABA3-CT is present in the sulfo-form, which is required by the Moco sulfurase target enzymes AO and XDH, ABA3-CT was treated with potassium cyanide. By this procedure, molybdenum-bound terminal sulfur is released as thiocyanate, which after the addition of ferric nitrate reagent forms an FeSCN complex that can be quantified by its specific absorption at 460 nm. Interestingly, thiocyanate formation was found in a stoichiometric ratio of about 0.16 molecules of thiocyanate/ molecule ABA3-CT (Table 1), indicating that about 16% of the recombinant ABA3-CT molecules had cyanolyzable sulfur ligated to their molybdenum-active site. Interpretation of these data based on FormA contents indicates that more than 70% of the MPT is present as Moco, and about 64% of the Moco contains an additional terminal sulfur ligand, as essentially required by enzymes of the xanthine oxidase family. When ABA3-CT was expressed in the absence of supplementary molybdate, FormA as well as molybdenum contents were extremely reduced, and no thiocyanate formation was detected (Table 1), indicating that 1) molybdenum supply during expression in E. coli enhances Moco assembly on ABA3-CT, and 2) the sulfur detected as thiocyanate must derive from Moco. The latter is further supported by the finding that ABA3-CT expressed in the MPT-accumulating (i.e. Moco-free) E. coli strain RK5206 did not release thiocyanate upon cyanide treatment (data not shown).
High Affinity Binding of MPT/Moco to ABA3-CT-A possible stabilizing effect of Moco by ABA3-CT was investigated under aerobic conditions at different temperatures. Under the conditions as described under "Experimental Procedures," the halflife of Moco/MPT at 22°C was about 67 h (Fig. 1A). At later time points, precipitation of ABA3-CT protein was observed, whereby further measurements were prohibited. When performing the experiment at 4°C, the half-life was enhanced to 125 h with more than 33% Moco/MPT still detectable after 334 h. The final fractions of both experiments were found to  contain intact Moco, since both fractions were able to reconstitute NADPH-nitrate reductase activity in extracts of the N. crassa nit-1 mutant in the absence and presence of additional molybdate (data not shown). These data suggest that ABA3-CT very efficiently protects Moco/MPT from rapid and irreversible degradation by oxidation.
In order to determine a dissociation constant (k D ) for the binding of Moco/MPT to ABA3-CT, 6 M cofactor-free purified ABA3-CT expressed in E. coli expression strain RK5204 was co-incubated with different concentrations (0 -12 M) of Moco/MPT extracted from recombinant C. reinhardtii MCP, which can be purified in large amounts and which specifically binds Moco rather than MPT. After coincubation, the samples were transferred to a 10,000 molecular weight cut-off concentrator, and unbound cofactor was separated from ABA3-CTbound cofactor by ultracentrifugation. Moco/MPT within the flow-through was immediately converted to the stable oxidation product FormA and quantified by HPLC. As a control, the same experiment was performed with free Moco/MPT in the absence of ABA3-CT. After quantification of free Moco/MPT in the flow-through of samples incubated in the presence and absence of ABA3-CT, determination of a k D value for cofactor binding to ABA3-CT was possible (Fig. 1B). The amount of Moco/MPT bound to ABA3-CT and the concentrations of free ABA3-CT were determined according to the calculated concentrations of free and total Moco/MPT and the known amount of total ABA3-CT. The maximum Moco/MPT saturation of ABA3-CT under the given experimental conditions was determined to be 0.89, suggesting a Moco/MPT to ABA3-CT ratio of 1:1. A k D value of 0.55 Ϯ 0.14 M was obtained, whereby high affinity binding of Moco/MPT to ABA3-CT is demonstrated.
Spectroscopic Properties of Moco-containing ABA3-CT-UVvisible spectra of Moco-loaded ABA3-CT showed absorption around 315 nm, between 350 and 400 nm, and also between 450 and 550 nm when the protein was analyzed directly after aerobic purification ( Fig. 2A). Very similar to this, also the anaerobically purified protein revealed absorption around 315 nm and between 350 and 400 nm directly after purification. However, in contrast to the aerobically purified protein, the third absorption band of anaerobically purified protein was shifted to the range between 500 and 580 nm. Upon oxidation in air for 20 h, the spectra of both aerobically and anaerobically purified protein developed the same distinct absorption around 315 nm and maxima at 395 and 465 nm, indicating that Moco bound to ABA3-CT is sensitive to oxidation, with the most sensitive absorption range between 350 and 580 nm. The addition of reducing agents like sodium dithionite (Fig. 2B) and oxidizing agents like potassium cyanide (Fig. 2C) to ABA3-CT for 2 or 20 h did not cause significant changes in the overall absorption. However, oxidation proceeded faster and was more pronounced in the presence of potassium cyanide in comparison with sodium dithionite. This effect has also been observed when using other reducing agents like dithiothreitol and ␤-mercaptoethanol or oxidizing agents like ferricyanide, respectively (supplemental Fig. 1). When fully oxidized ABA3-CT was titrated with sodium dithionite under anaerobic conditions, the absorption between 500 and 580 nm that has been observed earlier for anaerobically prepared protein was reconstituted (Fig. 3). These results not only confirm the redox sensitivity of this particular absorption band but also indicate that ABA3-CT is obtained in a (nearly) fully reduced state when purified under anaerobic conditions and in a partially reduced state when purified under aerobic conditions. It is noteworthy that the terminal sulfur ligand of the ABA3-CT-bound Moco obviously has no influence on the spectroscopic properties of the protein since the addition of cyanide, which is known to release the terminal sulfur ligand from the Moco of xanthine oxidase family enzymes (see "Experimental Procedures"), did not cause specific alterations of the spectrum.
Similar to plant (24) and animal sulfite oxidase proteins (25) and to the recently described Moco carrier protein MCP from C. reinhardtii (22), the absorption around 395 nm is likely to derive from the ene-dithiolate group of bound Moco. Absorp- tion around 500 nm has been ascribed to a cysteine-to-molybdenum charge transfer band (25) for the Moco-binding domain of rat sulfite oxidase, and we therefore assume that the absorption of Moco-loaded ABA3-CT between 450 and 500 nm also is caused by a ligand-to-molybdenum charge transfer band. This assumption is supported by the observation that ABA3-CT obtained from E. coli RK5206, which accumulates metal-free MPT and is unable to generate Moco, lacks absorption in this particular range (supplemental Fig. 2). In contrast, no influence has been observed on the absorption at 315 nm, indicating that the origin of this absorption band is unrelated to the molybdenum-metal. Rather, it should derive from the MPT-moiety, since ABA3-CT, expressed in the MPT/Moco-free E. coli strain RK5204, is lacking any of the absorptions typically observed for Moco-loaded ABA3-CT (data not shown). Like for Mocobound ABA3-CT, absorption between 500 and 580 nm has been observed also for MCP from C. reinhardtii (22). Since this absorption is also absent in the ABA3-CT protein expressed in MPT-accumulating E. coli RK5206, it is probably related to the molybdenum center like the absorption between 450 and 500 nm.
The sir3-3/R723K Variant of ABA3-CT Is Affected in Moco Binding-Recently, a new aba3 mutant from A. thaliana, sir3-3, was isolated by a specific screening procedure based on the resistance of the mutant plant toward sirtinol treatment (10). The sir3-3 mutant was found to contain a G-to-A transversion at nucleotide 2168 within the coding region that caused an exchange from arginine to lysine at position 723 (R723K) within the C-terminal domain of the ABA3 protein. Since it was unclear whether and to what extent the mutation affects the function of ABA3, the activities of its target enzymes AO and XDH were analyzed in extracts of sir3-3 plants and compared with AO and XDH activities of other aba3 mutant plants, including aba3-1, aba3-2, and 13.5 (Fig. 4A). Although aba3-1 presents a single substitution of glycine 469 by glutamic acid within the NifS-like domain of ABA3, the mutants aba3-2 and 13.5 are characterized by deletions within the aba3 gene accompanied by frameshifts and early truncation of the aba3 open reading frame (5). In contrast to extracts of aba3-1, aba3-2, and 13.5, where no AO and XDH activities have been detected, extracts of sir3-3 plants showed a residual XDH activity, indicating that the sir3-3 mutation is leaky to a certain extent. In order to prove whether the lack of the terminal sulfur ligand is the primary cause of AO and XDH deficiency in the sir3-3 mutant, as was shown for the mutants aba3-1 and aba3-2 (23), leaf extracts of all mutants were treated anaerobically with sodium sulfide and sodium dithionite. By this treatment, the terminal sulfur of the Moco in AO and XDH proteins is inserted nonenzymatically. As Fig. 4B depicts, sulfide/dithionite treatment restored the activities of AO and XDH proteins in all mutant plant extracts, indicating that, like in aba3-1, aba3-2, and 13.5 plants, the sulfuration of the Moco of AO and XDH is affected also in the sir3-3 mutant. When introducing the sir3-  . Reconstitution of AO and XDH activities in extracts of sir3-3 mutants by in vitro sulfuration. 150 g of total extracts of the A. thaliana aba3 mutants aba3-1, aba3-2, 13.5, and sir3-3 were separated on native PAGE either without (A) or subsequent to anaerobic treatment with sodium sulfide and sodium dithionite (B). In gel activity staining for AO was performed by using a combination of indole-3-carboxaldehyde and 1-naphthaldehyde as substrate; XDH activity was developed in the presence of hypoxanthine as substrate.
3/R723K mutation into heterologously expressed ABA-CT, it was found that the altered protein bound less Moco/MPT relative to the control protein ( Table 2). Besides significant reduction of Moco/MPT, a strong reduction of thiocyanate formation has also been observed for the R723K variant, suggesting that the amount of terminal sulfur is dramatically reduced and that most, if not all, of the Moco bound to this protein is present in the desulfo-form.
Activation of AO␣ and AtXDH1 by Moco-containing ABA3-CT-The separately expressed NifS-like domain of ABA3 was shown previously to be capable of activating recombinant AO␣ in vitro, whereas ABA3-CT did not show a stimulating effect on the activity of AO␣ (6). However, the ABA3-CT protein used for those assays was expressed under different conditions and did not contain Moco. In the present work, AO␣ and AtXDH1 as purified after heterologous expression in P. pastoris were co-incubated with Moco-loaded ABA3-CT for 30 min at 22°C prior to determination of AO and XDH activities. In fact, Moco-loaded ABA3-CT superactivated the activity of both AO␣ and AtXDH1, indicating a stimulating function of ABA3-CT, which led to the activation of so far inactive forms of AO␣ and AtXDH1 (Fig. 5, A and B). The appearance of a new activity band indicates the formation of a complex generated in the presence of Moco-loaded ABA3-CT and probably consisting of ABA3-CT and its target enzymes AO␣ and AtXDH1. In order to prove this assumption, the newly appearing activity band was excised from the native polyacrylamide gel and subjected to an additional SDS-PAGE. Subsequent immunoblotting using anti-His 6 antibodies showed that this band in fact contained both proteins, ABA3-CT and AO␣ (Fig. 5C). Such complex formation has been observed earlier in co-incubation experiments with full-length ABA3 and AtXDH1 (16). In contrast, co-incubation of ABA3-NifS with AO␣ or AtXDH1, respectively, did not result in formation of such a complex (6), indicating that the C-terminal domain of ABA3 rather than the NifS-like domain mediates interaction with the target enzymes of ABA3. To our surprise, ABA3-CT was found to accumulate also in the lower activity band after co-incubation with AO␣ (Fig. 5C), which may indicate the general necessity of a proteinprotein interaction for activation of AO␣.

DISCUSSION
The process of Moco sulfuration in eukaryotes, as required by molybdenum enzymes of the xanthine oxidase family, depends on the specific function of Moco sulfurase enzymes. All eukaryotic Moco sulfurases known to date generally feature two domains (5, 8, 11, 26 -28), of which only the NH 2 -terminal NifS-like domain is basically understood, since it was shown to be responsible for the mobilization of sulfur from L-cysteine (6). Furthermore, in vitro ABA3-NifS alone was shown to be able to sulfurate the Moco of recombinant AO␣ (6). However, it was known that in vivo besides the NifS-like domain also the C-terminal domain of Moco sulfurases is required for activation of AO and XDH. For the Moco sulfurase mutant flacca of tomato, a 6-bp deletion within the C-terminal domain was reported to be the primary cause of combined AO and XDH deficiency (8). Recently, the aba3 mutant sir3-3 from A. thaliana has been predicted to be deficient in AO-activity, since it presents an impaired capacity to oxidize the sirtinol derivative 2-hydroxy-1-naphthaldehyde to 2-hydroxy-1-naphthoic acid, which activates auxin signaling (10). This mutant harbors a point mutation within the C terminus coding region of aba3, leading to a substitution of the strictly conserved arginine 723 of the ABA3 protein by a lysine (R723K). Interestingly, a human xanthinuria  In gel activity staining for AO was performed by using indole-3carboxaldehyde as substrate, and XDH activity staining was developed in the presence of hypoxanthine as substrate. B, densitometric determination of relative activities of AO␣ and AtXDH1 after co-incubation with ABA3-CT based on A. C, immunoanalysis of the upper and lower activity bands of AO␣ co-incubated with ABA3-CT after native PAGE, separation of the respective bands on SDS-PAGE, and gel blotting. AO␣ and ABA3-CT proteins were detected using an anti-His 6 antibody, which also cross-reacts with the 116 kDa marker band. The upper bands (upper bands 1 and 2) of two independent co-incubations are shown.