Redox-linked Gating of Nucleotide Binding by the N-terminal Domain of Adenosine 5′-Phosphosulfate Kinase*

Background: Adenosine 5′-phosphosulfate kinase (APSK) in plants contains a regulatory disulfide bond. Results: Analysis of oxidized APSK reveals altered nucleotide binding compared with the reduced enzyme. Conclusion: The N-terminal domain is responsible for redox-linked structural changes that regulate APSK activity. Significance: This work provides a molecular basis for understanding reciprocal regulation at a branch point in plant sulfur metabolism. Adenosine 5′-phosphosulfate kinase (APSK) catalyzes the phosphorylation of adenosine 5′-phosphosulfate (APS) to 3′-phosphoadenosine-5′-phosphosulfate (PAPS). Crystallographic studies of APSK from Arabidopsis thaliana revealed the presence of a regulatory intersubunit disulfide bond (Cys86–Cys119). The reduced enzyme displayed improved catalytic efficiency and decreased effectiveness of substrate inhibition by APS compared with the oxidized form. Here we examine the effect of disulfide formation and the role of the N-terminal domain on nucleotide binding using isothermal titration calorimetry (ITC) and steady-state kinetics. Formation of the disulfide bond in A. thaliana APSK (AtAPSK) inverts the binding affinities at the ATP/ADP and APS/PAPS sites from those observed in the reduced enzyme, consistent with initial binding of APS as inhibitory, and suggests a role for the N-terminal domain in guiding nucleotide binding order. To test this, an N-terminal truncation variant (AtAPSKΔ96) was generated. The resulting protein was completely insensitive to substrate inhibition by APS. ITC analysis of AtAPSKΔ96 showed decreased affinity for APS binding, although the N-terminal domain does not directly interact with this ligand. Moreover, AtAPSKΔ96 displayed reduced affinity for ADP, which corresponds to a loss of substrate inhibition by formation of an E·ADP·APS dead end complex. Examination of the AtAPSK crystal structure suggested Arg93 as important for positioning of the N-terminal domain. ITC and kinetic analysis of the R93A mutant also showed a complete loss of substrate inhibition and altered nucleotide binding affinities, which mimics the effect of the N-terminal deletion. These results show how thiol-linked changes in AtAPSK alter the energetics of binding equilibria to control its activity.

In plants, the N-terminal domain is also implicated in redox control of APSK to coordinate flux between the primary and secondary branches of the sulfur assimilation pathway (24). The primary route provides sulfur for the synthesis of cysteine and glutathione (4 -8), whereas PAPS generated in the secondary route is used for the synthesis of other metabolites (9 -11). The crystal structure of AtAPSK revealed a disulfide bond formed between Cys 86 in the N-terminal domain of one monomer and Cys 119 in ␣2 of the adjoining monomer ( Fig. 1A) (24). Kinetic analysis of the reduced (AtAPSK RED ) and oxidized (AtAPSK OX ) forms of the enzyme showed that formation of the disulfide reduced catalytic efficiency and increased effectiveness of substrate inhibition by APS. The functional differences suggest that cellular redox state may act as a novel regulatory feature in the plant APSK (23,24,29). For example, under oxidative stress conditions, decreased AtAPSK activity may enhance the flow of APS into primary sulfur metabolism to support glutathione synthesis for maintaining redox balance (24).
To elucidate the role of the N-terminal domain in AtAPSK and its contribution to substrate inhibition and redox regulation of activity, ITC was used to examine nucleotide binding to AtAPSK OX , which displayed differential sensitivity to substrate inhibition by APS (24). Extensive analysis of nucleotide binding to the reduced form of AtAPSK was recently described (23) and provides a basis of comparison with the oxidized disulfidelinked form. Here we show that disulfide bond formation in AtAPSK OX leads to enhanced APS binding in both the presence and absence of either the non-hydrolyzable ATP analog AMP-PNP or ADP. Thus, disulfide bond formation alters binding equilibria to disfavor formation of a catalytically productive ternary complex and suggests a role for the N-terminal domain in guiding the ligand binding sequence of AtAPSK. In contrast, truncation of the N-terminal domain (AtAPSK⌬96) decreased APS binding affinity without altering AMP-PNP binding. Examination of the three-dimensional structure of AtAPSK suggested that Arg 93 , which is conserved across the APSK from bacteria, fungi, plants, and animals, is critical for positioning of the N-terminal domain. Kinetic and calorimetric analysis of the R93A mutant is consistent with the proposed role of this residue as a critical anchor for positioning of the N-terminal domain. Based on the results of calorimetric and kinetic experiments, we suggest a model for how the N-terminal domain of AtAPSK modulates APS binding and how redox-linked gating helps determine the order of nucleotide addition.

EXPERIMENTAL PROCEDURES
Reagents-All chemicals and reagents were purchased from Sigma-Aldrich as analytical grade or better. The standard buffer condition for all experiments was 25 mM HEPES, pH 7.5, 200 mM KCl, 5% (v/v) glycerol, and either 1 mM tris(2-carboxyethyl)phosphine or 5 mM ␤-mercaptoethanol, unless stated otherwise. Reducing agents prevent disulfide bond formation in AtAPSK, as judged by non-reducing SDS-PAGE (24) of samples taken before and after ITC titrations.
Generation of Mutant Constructs, Protein Expression, and Purification-For bacterial expression of AtAPSK, the pET-28a-AtAPSK⌬77 bacterial expression construct, which encodes A. thaliana APSK isoform 1 without the plastid localization sequence (residues 1-77) and with an N-terminal hexahistidine tag, was used (30). Generation of the N-terminal truncation mutant AtAPSK⌬96 used pET-28a-AtAPSK⌬77 as template. Forward and reverse primers harboring NheI and EcoRI restriction sites, respectively, were used to PCR-amplify the new versions of the coding region, which were treated with restriction enzymes and subcloned into pET-28a. The R93A point mutant used was introduced into pET-28a-AtAPSK⌬77 by QuikChange mutagenesis (Agilent). All proteins were overexpressed in E. coli BL21(DE3), and purification by nickel affinity and gel filtration chromatographies was done as described previously (23,24). For purification of protein for ITC analysis, all buffers were supplemented with 5 mM ␤-mercaptoethanol. After nickel affinity purification, protein was dialyzed into 25 mM HEPES, pH 7.5, 200 mM KCl, 5% glycerol, and 5 mM dithiothreitol (DTT) and then loaded onto a Superdex-200 26/60 HiLoad FPLC size exclusion column equilibrated in the same buffer. Purified proteins were dialyzed against 25 mM HEPES, pH 7.5, 200 mM KCl, 5% glycerol, and either 1 mM tris(2-carboxyethyl)phosphine or 5 mM ␤-mercaptoethanol and flash frozen in liquid nitrogen and stored at Ϫ80°C. Protein concentrations of AtAPSK OX and AtAPSK⌬96 were determined spectrophotometrically (31) using calculated extinction coefficients of 15,930 and 21,540 M Ϫ1 cm Ϫ1 , respectively.
Enzyme Assays-Determination of steady-state kinetic parameters was performed as previously described using an enzyme-coupled spectrophotometric assay (23,24).
Calorimetric Measurements-ITC experiments were performed with a VP-ITC calorimeter (Microcal, Inc.). As described above, AtAPSK was isolated and maintained in the reduced form. The oxidized form (AtAPSK OX ) was generated by removal of reducing agent using buffer exchange in centrifugal spin filter units (M r 10,000; Amicon) into the standard reaction buffer containing 5 mM trans-4,5-dihydroxy-1,2-dithiane (i.e. oxidized DTT; Sigma). Samples were dialyzed overnight (4°C) against the same buffer. Stock solutions (100 mM) of ATP, AMP-PNP, ADP, and APS were dissolved in NaOH to attain a pH of 7.5 and stored at Ϫ20°C. AtAPSK complexed with various ligands were generated by incubating protein and 2 mM AMP-PNP, 500 M ADP, or 2 mM APS overnight (4°C), followed by a 4-h equilibration at 17°C before titrations. Prior to ITC experiments, appropriate dilutions were made with dialysis buffer. Protein and nucleotide solutions were degassed at room temperature before use. 20 -30 injections (10 l) of nucleotide were added into sample solutions containing either AtAPSK OX (15-30 M) or AtAPSK⌬96 (30 -50 M). Data were analyzed using either an identical binding sites model (Equation 1) or two-site binding model (Equation 2), as follows, where Q i tot represents the total heat after the ith injection, V 0 is the calorimetric cell volume, M i tot is the concentration of protein in the cell after the ith injection, ⌬H is the corresponding enthalpy change to APSK-nucleotide binding, n is the number of nucleotide binding sites on the APSK dimer, and K 1 is the equilibrium binding constant. In the latter, K 1 and K 2 are the observed binding constants for the first and second sites, and ⌬H 1 and ⌬H 2 are the corresponding enthalpy changes upon nucleotide binding to each site. Fitting of data was performed using Origin software.

Calorimetric Analysis of Nucleotide Binding to Oxidized
AtAPSK-Crystallographic and functional analysis of AtAPSK identified an intersubunit disulfide bond between Cys 86 and Cys 119 that may play a role in regulating enzymatic activity ( Fig.  1A) (24). Here we use ITC to examine nucleotide binding to AtAPSK OX for comparison with the reduced form of the enzyme (23).
ITC analysis of ATP and ADP binding to AtAPSK OX showed exothermic binding with a two-site model best fitting the data ( Fig. 2A and Table 1). Comparison of this ITC data with that obtained using AtAPSK RED (23) showed that the presence of the Cys 86 -Cys 119 linkage in AtAPSK decreased first site binding affinity for ATP and ADP by 45-and 217-fold, respectively. For the second site, a 9-fold decrease in the affinity for ATP was observed, but the affinity for ADP was comparable with that of the reduced enzyme. In contrast, AtAPSK OX displayed an 18-fold tighter binding affinity for APS with a binding isotherm that could be adequately fit to a single-site model (n ϭ 1.95). Previous ITC experiments indicated weaker asymmetric twosite binding to the reduced apoenzyme (23). Overall, formation of the disulfide bond in AtAPSK inverts the binding affinities at the ATP/ADP and APS/PAPS sites from those observed in the reduced enzyme.
Next, ITC experiments were used to determine if preformed complexes of AtAPSK OX and either phosphonucleotide (AMP-PNP and ADP) or APS altered the binding affinity for APS or phosphonucleotides, respectively ( Fig. 2B and Table 2). APS titration of the AtAPSK OX ⅐AMP-PNP and AtAPSK OX ⅐ADP complexes yielded binding affinities (K 1 ϭ 0.64 and 3.70 M, respectively) that were comparable with those determined for the reduced protein (K 1 ϭ 1.50 and 3.30 M, respectively) (23). Titrations of the AtAPSK OX ⅐APS complex with either AMP-PNP (K 1 ϭ 37.2 M) or ADP (K 1 ϭ 26.1 M) showed 12-and 40-fold higher K d values compared with those of the reduced protein (K 1 ϭ 3.10 and 0.65 M, respectively) (23). These results indicate that disulfide bond formation alters the nucleotide binding equilibria, leading to formation of a catalytically productive E⅐ATP⅐APS ternary complex and suggest a role for the N-terminal domain in determining the order of ligand addition.
Effect of N-terminal Domain Deletion (AtAPSK⌬96) on Steady-state Kinetics-The location of the Cys 86 -Cys 119 disulfide bond in the N-terminal domain of AtAPSK (Fig. 1, A and B) and the effects of its formation on nucleotide binding ( Fig. 2 and Tables 1 and 2) and catalytic activity (23) suggest an important functional role for the N-terminal domain. For example, reduction of the disulfide decreases the K i of APS substrate inhibition by 15-fold compared with AtAPSK OX . To test if removal of the N-terminal domain affects AtAPSK activity, a truncated version of the enzyme lacking residues 77-96 was generated. This construct removes the N-terminal domain through the end of the ␣1-helix (Fig. 1A). The resulting AtAPSK⌬96 protein displayed only modest changes in kinetic parameters compared with AtAPSK RED (i.e. a 3-fold slower turnover rate and a 5-fold higher K m for APS) but completely eliminated the effect of APS substrate inhibition ( Table 3). The protein also showed no sensitivity to reducing agents, consistent with the loss of Cys 86 .
Calorimetric Analysis of the Effect of N-terminal Deletion (AtAPSK⌬96) on Nucleotide Binding-The effect of the N-terminal truncation on the steady-state kinetics led us to analyze nucleotide binding to the AtAPSK⌬96 apoenzyme and binary nucleotide complexes. ITC experiments showed that binding of ATP and ADP to AtAPSK⌬96 was exothermic and that binding to the first site influenced the second binding event because a two-site model best fit the observed data ( Fig. 3A and Table 4  Given the results of the ATP, ADP, and APS titrations with the apoenzyme form of AtAPSK⌬96, ITC analyses of ligand binding to nucleotide binary complexes were performed (Fig. 3,  B and C). ITC experiments demonstrate that formation of the AtAPSK⌬96⅐AMP-PNP and AtAPSK⌬96⅐ADP complexes increases affinity for APS ( Fig. 3B and Table 4) because binding can now be detected in contrast to the lack of signal for APS  Table 1). B, titration of AtAPSK OX ⅐AMP-PNP (f) and AtAPSK OX ⅐ADP (Ⅺ) complexes with APS at 17°C. The solid lines represent the fit to data using either two-site or one-site binding models (see Table 2).

TABLE 1 Thermodynamic parameters of nucleotide binding to AtAPSK OX
All titrations were performed at 17°C. ITC data were fit to either a one-site binding model (n ϭ number of sites shown below K 1 value) or a two-site binding model. For comparison, the previously determined binding constants (K 1 and K 2 ) determined for AtAPSK RED (23) are shown in the last column.  titration of the AtAPSK⌬96 apoenzyme (Fig. 3A). For each AtAPSK⌬96 binary complex, the affinity for APS was comparable with the previously reported values for the full-length reduced protein (24). These results suggest that the N-terminal region is not required for communication between the ATP/ ADP and APS/PAPS binding sites of AtAPSK because binding of either AMP-PNP or ADP still enhances APS binding, as observed in the full-length protein (23). Although ITC titrations of AtAPSK⌬96 apoenzyme with APS lack significant heat signature (Fig. 3A), the catalytic activity of the enzyme (Table 3) indicates that APS still binds to the enzyme. Preincubation of AtAPSK⌬96 with 2 mM APS followed by titration with either AMP-PNP or ADP showed ternary complex formation (Fig. 3C). As observed with the reduced protein (23), AMP-PNP titration of the APS complex showed an endothermic interaction. This variation in binding energetics indicates that structural and/or dynamic changes occur in the active sites of the reduced and oxidized forms of AtAPSK. Formation of an AtAPSK⌬96⅐APS⅐nucleotide complex yielded K d values for both AMP-PNP and ADP ( Table 4) that were comparable with those for the full-length reduced protein (23); however, the concentrations of APS needed to form the binary complex with AtAPSK⌬96 are probably beyond the physiological range. Overall, ITC analysis of ligand interactions in AtAPSK⌬96 supports an important role for the N-terminal region in nucleotide binding.   Table 4). B, titration of AtAPSK⌬96⅐AMP-PNP (f) and AtAPSK⌬96⅐ADP (Ⅺ) complexes with APS at 17°C. The solid lines represent the fit to data using either two-site or one-site binding models (see Table 4). C, titration of the AtAPSK⌬96⅐APS complex with AMP-PNP (Ⅺ) and ADP (f). The solid lines represent the fit to data using a two-site binding model (see Table 4).

TABLE 4 Thermodynamic parameters of nucleotide binding to AtAPSK⌬96⅐nucleotide complexes
All titrations were performed at 17°C with AtAPSK⌬96⅐nucleotide complexes preformed as described under "Experimental Procedures." ITC data were fit to either one-site (n ϭ number of sites shown below K 1 value) or two-site binding models.

Effect of Arg 93 Mutation in the N-terminal Domain of AtAPSK on Activity and Nucleotide
Binding-The role of the N-terminal domain as an important feature for redox regulation, nucleotide binding, and substrate inhibition by APS led us to reexamine the crystal structure of AtAPSK (24) for other residues involved in positioning of the N-terminal domain. In general, the N-terminal domain of one monomer binds along the surface of the adjacent monomer with interactions mediated primarily through van der Waals contacts (Fig. 1A); however, the region near the disulfide and ␣1 includes residues that make additional intersubunit (Tyr 126 -Asp 92 ) and intrasubunit (Arg 93 -Asp 171 ) interactions (Fig. 1B). Sequence alignment of the N-terminal region of APSK from a variety of plants, bacteria, fungi, and animals ( Fig. 1C) indicates that the arginine corresponding to Arg 93 of AtAPSK is highly conserved among these enzymes. In general, either an arginine or a lysine is found at this position in APSK from other species. Based on the crystal structure of AtAPSK (24), Arg 93 may play a role in positioning the flexible N-terminal domain along the adjacent monomer surface.
To test the functional role of Arg 93 , the R93A mutant was generated, expressed, and purified. Substitution of Arg 93 with an alanine modestly reduces k cat by 2-fold and increases the K m for APS by 3-fold (Table 3). Importantly, the R93A mutation eliminates the effect of substrate inhibition by APS, as observed with AtAPSK⌬96.
ITC experiments with ATP, ADP, and APS as titrants for the AtAPSK R93A mutant behaved similarly to AtAPSK⌬96 (Fig.  4A). Binding of ATP and ADP yielded K d values similar to those of the N-terminal deletion mutant and displayed differential first and second site binding (Table 5). Negligible heat signal was detected for APS binding to the free enzyme (Fig. 4A). This binding analysis suggests that the effect of the N-terminal domain on nucleotide binding in the reduced form of the protein is largely mediated by Arg 93 . Titrations of the AtAPSK R93A⅐AMP-PNP and AtAPSK R93A⅐ADP complexes with APS showed that formation of complex causes a significant increase in affinity for APS. Thus, the R93A mutant displays a ligand binding profile similar to that of AtAPSK⌬96 ( Fig. 4B and Table 5).

DISCUSSION
APSK controls the flux of sulfate through the secondary sulfur assimilation pathway, which produces a diverse array of metabolites required for normal growth and defense against environmental stress (2)(3)(4)(5)(6)(7)(8)(9)(10)(11). All forms of APSK from a variety of organisms studied to date exhibit substrate inhibition by APS (1, 15-16, 20 -23), which may provide a mechanism for controlling this enzyme in vivo. Recent identification of a regulatory disulfide bond in AtAPSK links cellular redox state to the degree of substrate inhibition by APS, as well as the rate of catalysis, in plants (23,24,27). Previous analysis of nucleotide binding to the reduced form of AtAPSK (23) provides a basis of comparison to better understand the molecular underpinnings of redox regulation and the role of the N-terminal domain in APSK.
In AtAPSK, disulfide bond formation between Cys 86 and Cys 119 alters the steady-state kinetic behavior of the enzyme from a more active reduced form to a less active oxidized form, which is also more sensitive to substrate inhibition by APS (Table 3) (24). Here we show that oxidation of the disulfide in AtAPSK affects the nucleotide binding equilibria compared with the reduced enzyme (Fig. 5). Based on ITC analysis of binary and ternary complex formation in AtAPSK OX (Fig. 2 and The solid lines represent the fit to data using a two-site sequential binding model or a single set of non-interacting sites model (see Table 5). B, titration of AtAPSK R93A⅐AMP-PNP (f) or AtAPSK R93A⅐ADP (Ⅺ) complex with APS at 17°C. The solid lines represent the fit to data using a two-site binding model (see Table 5).
Tables 1 and 2), the binding affinities at the ATP/ADP and APS/PAPS sites are essentially inverted versus those observed in AtAPSK RED . The presence of Cys 86 -Cys 119 in AtAPSK reduces the K d for ATP and ADP by 45-and 217-fold compared with the reduced enzyme but enhances affinity for APS by 18-fold. Similar effects were also observed using preformed binary complexes and the addition of the second nucleotide.
These results suggest a mechanism for redox-linked gating of ligand binding at the ATP/ADP and APS/PAPS sites. In AtAPSK RED , the ligand binding equilibria favor the catalytically productive sequence of ATP binding first followed by the APS addition (Fig. 5). Disulfide formation shifts the binding preference in the apoenzyme to favor APS and reduces the affinity for ATP and ADP binding to either free enzyme or the E⅐APS complex (Fig. 5). Structurally, the Cys 86 -Cys 119 disulfide cross-links the two monomers in AtAPSK and probably reduces the conformational flexibility of the N-terminal domain (Fig. 1A). Acting like a molecular staple, the disulfide bond may restrict movement of the N-terminal domain to preorganize the APS binding site by mimicking the overall conformation of the ATP/ ADP-bound form of the enzyme, which leads to enhanced APS affinity. The comparable K d values for APS binding to AtAPSK OX and the AtAPSK RED ⅐ATP and AtAPSK RED ⅐ADP complexes suggest structurally similar binding sites for this ligand in different forms of the enzyme. The overall conformation of the oxidized protein may also structurally occlude the ATP/ADP site to decrease affinity for binding of these ligands. Thus, there is a fundamental difference between the oxidized form of AtAPSK and the ATP/ADP-bound complexes.
Multiple steady-state kinetic studies of bacterial, fungal, and mammalian APSK, as well as AtAPSK RED , suggest that forma-tion of an E⅐ADP⅐APS dead end complex is responsible for substrate inhibition (1, 15, 16, 20 -23); however, all of these forms lack the regulatory disulfide switch found in the plant APSK. Comparison of ligand binding in the reduced and oxidized forms of AtAPSK suggests another structural mechanism for substrate inhibition by APS in plants. The restriction in movement of the N-terminal domain in AtAPSK OX also strengthens the effectiveness of substrate inhibition by APS.
The effect of disulfide bond formation on activity and ligand binding and the position of this linkage in AtAPSK led us to further examine the contribution of the N-terminal domain (residues 77-96) to nucleotide binding. Removal of the N-terminal domain in AtAPSK modestly alters both k cat and K m but completely eliminates substrate inhibition by APS (Table 3), which is similar to the results reported for the human enzyme (20). Although substrate inhibition by APS was previously linked to the N-terminal domain of human APSK (20), here we provide the first experimental evidence for targeted changes in nucleotide binding mediated through the N-terminal domain of an APSK.
Deletion of the N-terminal domain of AtAPSK did not alter either the binding of ATP to free enzyme or the binding of APS to the E⅐AMP-PNP complex ( Fig. 3 and Table 4); however, a large effect on APS binding was observed because no heat signal change was detected in titrations with this ligand to free enzyme. The calorimetric results and the 5-fold increase in K m for APS (Table 3) suggest that removal of the N-terminal domain results in equilibria that strongly favor the ATP first binding sequence, which retains enhanced APS affinity to the binary complex (Fig. 5). In addition to disrupting APS interaction with the free enzyme, truncation of the N-terminal domain decreased affinity for ADP by 52-fold, although APS binding to the E⅐ADP complex was unaffected (Fig. 5). Overall, deletion of the N-terminal domain has two major effects on APSK that abolish the substrate inhibition effect of APS: 1) favoring of the catalytically productive ATP first binding sequence and 2) a higher K d for ADP that weakens formation of the E⅐ADP⅐APS dead end complex. Earlier work comparing the three-dimensional structures of full-length human APSK and an N-terminal truncated version (20) suggested a possible structural basis for differences in substrate inhibition between the two enzyme forms. The structure of human APSK lacking the N-terminal domain revealed an asymmetric dimer with ADP and APS present in one monomer and only ADP in the neighboring monomer, which led to a model in which the N-terminal domain is required for formation of a symmetric dimer, which represents the inhibited conformation of the enzyme. However, the binding isotherms observed in our ITC experiments ( Fig. 3 and Table 4) with AtAPSK⌬96 do not suggest that a greater degree of asymmetry exists in the truncated dimer compared with the full-length reduced protein. Currently, it is unclear how the AtAPSK⌬96 construct alters binding at the APS site because the N-terminal domain is not in direct contact with the site, although our results and the human APSK structural studies would imply that removal of the N-terminal domain alters movement of other structural features around the active site that are critical for substrate binding.
Multiple crystal structures of APSK from various organisms show that the N-terminal domain is flexible because it adopts different conformations in these models (17)(18)(19)(20)(23)(24)(25)(26). As shown here, the N-terminal domain in AtAPSK is a central feature for regulation of activity by guiding the sequence of nucleotide binding and substrate inhibition. Because positioning of the N-terminal domain of one AtAPSK monomer along the surface of the adjacent monomer (Fig. 1A) appears critical for ligand binding order, we reexamined the AtAPSK crystal structure for additional residues that orient this part of the protein. Structural and sequence comparisons (Fig. 1, B and C) suggested that Arg 93 may play a role in positioning the flexible N-terminal domain along the adjacent monomer surface by interaction with Asp 171 in ␣3. The R93A mutant recapitulated the steady-state kinetic (Table 3) and nucleotide binding parameters observed with AtAPSK⌬96 ( Fig. 4 and Table 5). As with removal of the entire N-terminal domain, the R93A mutant also abolished substrate inhibition by APS. Interaction between the enzymatic core of one monomer and the N-terminal domain of the other monomer in APSK appears essential for substrate inhibition and redox regulation.
Overall, our results suggest a two-tiered control of APSK activity in plants. The first mechanism is common across nearly all APSK and centers on positioning of the N-terminal domain over the catalytic site to control nucleotide binding order and to modulate substrate inhibition by APS. Conserved residues in ␣1 are critical for bending the N-terminal domain along the surface leading toward the active site. In plants, evolution of a second control switch, in the form of a redox-linked disulfide, further tunes the overall catalytic efficiency and substrate inhibition effects for additional regulation between the reduced and oxidized forms of APSK. This potentially provides reciprocal regulation at a metabolic branch point, in which oxidation of APS reductase activates flow into the primary sulfur assimilation pathway while attenuating APSK to decrease the levels of APS directed into the secondary pathway used for sulfonylation of multiple metabolites (24).
Although redox-regulation of proteins through disulfide bonds provides a critical response mechanism to changes in cellular environment (32,33), the structural and energetic basis for how this molecular switches work are often not well understood. AtAPSK provides a rare example for how thiol-linked changes directly alter the energetics of binding equilibria in a protein to control its activity and suggests a likely role for conformational dynamics in these processes.