Báo cáo khoa học: Site-directed mutagenesis, kinetic and inhibition studies of aspartate ammonia lyase from Bacillus sp. YM55-1

Site-directed mutagenesis, kinetic and inhibition studies of aspartate ammonia lyase from Bacillus sp. YM55-1 Vinod Puthan Veetil1, Hans Raj1, Wim J. Quax1, Dick B. Janssen2 and Gerrit J. Poelarends1

1 Department of Pharmaceutical Biology, Groningen Research Institute of Pharmacy, University of Groningen, The Netherlands 2 Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, The Netherlands

Keywords aspartase; aspartate ammonia lyase; Bacillus; deamination; enzyme mechanism

Correspondence G. J. Poelarends, Department of Pharmaceutical Biology, Groningen Research Institute of Pharmacy, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands Fax: +31 50 3633000 Tel: +31 50 3633354 E-mail: g.j.poelarends@rug.nl

(Received 23 January 2009, revised 6 March 2009, accepted 20 March 2009)

doi:10.1111/j.1742-4658.2009.07015.x

Aspartate ammonia lyases (also referred to as aspartases) catalyze the revers- ible deamination of l-aspartate to yield fumarate and ammonia. In the pro- posed mechanism for these enzymes, an active site base abstracts a proton from C3 of l-aspartate to form an enzyme-stabilized enediolate intermediate. Ketonization of this intermediate eliminates ammonia and yields the prod- uct, fumarate. Although two crystal structures of aspartases have been deter- mined, details of the catalytic mechanism have not yet been elucidated. In the present study, eight active site residues (Thr101, Ser140, Thr141, Asn142, Thr187, His188, Lys324 and Asn326) were mutated in the structurally char- acterized aspartase (AspB) from Bacillus sp. YM55-1. On the basis of a model of the complex in which l-aspartate was docked manually into the active site of AspB, the residues responsible for binding the amino group of l-aspartate were predicted to be Thr101, Asn142 and His188. This postulate is supported by the mutagenesis studies: mutations at these positions resulted in mutant enzymes with reduced activity and significant increases in the Km for l-aspartate. Studies of the pH dependence of the kinetic parameters of AspB revealed that a basic group with a pKa of approximately 7 and an acidic group with a pKa of approximately 10 are essential for catalysis. His188 does not play the typical role of active site base or acid because the H188A mutant retained significant activity and displayed an unchanged pH- rate profile compared to that of wild-type AspB. Mutation of Ser140 and Thr141 and kinetic analysis of the mutant enzymes revealed that these resi- dues are most likely involved in substrate binding and in stabilizing the enediolate intermediate. Mutagenesis studies corroborate the essential role of Lys324 because all mutations at this position resulted in mutant enzymes that were completely inactive. The substrate-binding model and kinetic anal- ysis of mutant enzymes suggest that Thr187 and Asn326 assist Lys324 in binding the C1 carboxylate group of the substrate. A catalytic mechanism for AspB is presented that accounts for the observed properties of the mutant enzymes. Several features of the mechanism that are also found in related enzymes are discussed in detail and may help to define a common substrate binding mode for the lyases in the aspartase ⁄ fumarase superfamily.

Abbreviations Ap, ampicillin; AspA, aspartase from E. coli; AspB, aspartase from Bacillus sp. YM55-1; FumC, fumarase C from E. coli; RCSB, Research Collaboratory for Structural Bioinformatics; PDB, Protein Data Bank.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2994

been purified and characterized from a number of Gram-positive and Gram-negative bacteria, including Escherichia coli, Hafnia alvei, Pseudomonas fluores- cens, Bacillus subtilis and Bacillus sp. YM55-1 [1–11]. Aspartate ammonia lyases (also referred to as asparta- ses) are microbial enzymes that catalyze the reversible deamination of l-aspartate (1) to yield fumarate (2) (Scheme 1). These enzymes have and ammonia (3)

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

+ NH3

–

+

– CO2

CO2

NH3

–O2C

–O2C

1

2

3

Scheme 1. Reversible deamination of L-aspartate catalyzed by aspartase.

this is far from complete and major issues remain unresolved. One issue concerns the identity of the gen- eral base catalyst that abstracts the C3 proton, and that of other essential catalytic and substrate-binding residues. Another issue concerns whether substrate binding induces a conformational change that moves other residues into the active site, as might be expected for an enzyme that is allosterically activated by its sub- strate. The crystal structure of AspA does not address these questions because it was solved in the absence of a bound ligand [12]. Attempts to obtain a crystal struc- ture of AspA (or any other aspartase) complexed with substrate, product or a competitive inhibitor have thus far proved unsuccessful.

thermophilic bacterium Bacillus The best studied example is the aspartase (AspA) from E. coli, for which the crystal structure has been eluci- dated [12]. AspA functions as a homotetramer, where each monomer consists of 478 amino acid residues, and is allosterically activated by its substrate (1) and Mg2+ ions, which are required for activity at alkaline pH [3,13]. The enzyme has a rather narrow substrate specificity; it is specific for 1 and 2, and only hydroxyl- amine can substitute for 3 as a substrate [14–16].

reaction (i.e. reverse the enediolate intermediate enzyme-stabilized

In our studies, we focus on aspartase (AspB) from the sp. YM55-1 [10,11]. This aspartase is of considerable biocatalytic interest because of its high activity and enantioselec- tivity, relative thermostability and lack of allosteric regulation by substrate or metal ions [10,11]. It also efficiently catalyzes the addition of ammonia to fumaric acid). Moreover, the broad nucleophile specificity of AspB enables the use of a range of alternative nucleophiles, such as methyl- amine, hydroxylamine, hydrazine and methoxylamine, in the conjugate addition reaction [22]. These proper- ties indicate that AspB is a promising biocatalyst for the synthesis of enantiopure N-substituted aspartic acids.

in its

Stereochemical, kinetic isotope and pH-rate studies indicate that AspA catalyzes an anti-elimination reac- tion, where an active site base (with a pKa of approxi- mately 6.5) abstracts the proton from C3 of 1 to form (4; an Scheme 2) [5,6,16–18]. The proposed enediolate inter- mediate (i.e. aci-carboxylate) can rearrange to elimi- nate ammonia (3) and form the product, fumarate (2). the The rate determining step is the cleavage of carbon-nitrogen bond, which may be facilitated by a general acid that protonates the leaving ammonia group. Additional support for the formation of an enzyme-stabilized enediolate as an intermediate during the deamination reaction is provided by inhibition studies demonstrating that 3-nitro-2-aminopropionate, when present resonance-stabilized nitronate (aci-form) state, binds very tightly to AspA as a transition state analog [19]. On the basis of the crystal structure of AspA and site-directed mutagenesis studies, two positively charged residues (Arg29 and Lys327) were proposed to bind the two carboxylate groups of 1 (C4 and C1, respectively) [12,20,21]. It was further postulated that Ser143 functions as the general acid catalyst [21].

Ser-143

O–

O

O

+ NH3

+ NH3

Arg-29

+

NH3

Lys-327

O–

O–

O–

–O2C

–O2C

–O2C

On the basis of these studies, a picture of the cata- lytic mechanism of AspA has emerged [16]. However, The crystal structure of AspB (also in the absence of a bound ligand) was recently solved (Fig. 1) and shows that the overall topology and active site architecture of AspB are similar to those observed in AspA and fuma- rase C (FumC) from E. coli, confirming its member- ship in the aspartase ⁄ fumarase superfamily of enzymes [23]. Similar to AspA, AspB functions as a homotetr- amer, but now each subunit is composed of 468 amino acid residues [11,23]. The functional tetramer contains four active sites, each harboring residues from three different subunits [23]. A structural model of the com- plex in which l-aspartate was docked manually into the active site of AspB suggested interactions that might be responsible for the binding and activation

4

2

3

H H 1 B:

Scheme 2. A schematic representation of the catalytic mechanism of AspA

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2995

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

Results

pH dependence of the kinetic parameters of AspB

The pH dependences of kcat and kcat ⁄ Km for the AspB- catalyzed deamination of l-aspartate (1; Scheme 1) were determined in 100 mm sodium phosphate buffer over the pH range 6.0–10.0 at 37 (cid:2)C. Both parameters show a bell-shaped dependence on pH, with limiting slopes of unity on either side of the pH maximum, indicating that both a basic group (ascending limb) and acidic group (descending limb) are important for catalysis. The pH dependences of kcat and kcat ⁄ Km are given by Eqns (1,2):

ð1Þ kcat ðpHÞ ¼ ðkcatÞmax=ð1 þ ½Hþ(cid:2)=K1 þ K2=½Hþ(cid:2)Þ

kcat=KmðpHÞ ¼ ðkcat=KmÞmax=ð1þ½Hþ(cid:2)=K1 þK2=½Hþ(cid:2)Þ ð2Þ

Fig. 1. A close-up of the active site of AspB [23]. The roles of the key active site residues (Thr101, Ser140, Thr141, Asn142, Thr187, His188, Lys324 and Asn326) and their interactions are discussed in the text. Suffixes A, B and C indicate that the residues originate from three different subunits. Prepared using PYMOL [39].

where K1 is the ionization constant of the basic group and K2 is the ionization constant of the acidic group, with both being important for catalysis [24].

log kcat ⁄ Km for

Fig. 2. pH-dependence of the deamination of ) and the T141A (d) and H188A L-aspartate (1) by wild-type AspB ( ( ) mutants. The curves were generated by a nonlinear least- squares fit of the data to the logarithmic form of Eqn (2). Errors given in the text are standard deviations.

A nonlinear least-squares fit of the pH-dependence of log (kcat ⁄ Km), which follows the ionizations in the free enzyme and the free substrate, to the logarithmic form of Eqn (2) gives pKa values of pK1 = 7.1 ± 0.1 and pK2 = 9.8 ± 0.2 (Fig. 2). A nonlinear least- squares fit of the pH-dependence of log (kcat), which follows the ionizations in the enzyme–substrate com- plex, to the logarithmic form of Eqn (1) yields pKa values of pK1 = 6.2 ± 0.4 and pK2 = 11.3 ± 2 (data not shown). Because data could not be collected at high enough pH values to clearly define the descending limb (i.e. the extension of these studies beyond pH 10 is precluded by enzyme denaturation), the pKa value

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2996

(i.e. polarization of the C4 carboxylate group) of the substrate [23]. It is noteworthy that this approach of manual docking of the substrate in the active site, which may provide valuable insight into how residues at the active site interact with the substrate, is compli- cated for AspA because this enzyme is allosterically activated by its substrate [13]. Hence, it is reasonable to expect some significant differences between the active site structure of the apoenzyme and that of the enzyme–substrate complex. In the docking model, the C1 carboxylate group of l-aspartate forms hydro- gen-bonding interactions with the hydroxyl group of Thr187, the amino group of Lys324 and the amide group of Asn326, whereas the C4 carboxylate group of the substrate forms hydrogen bonds with the hydroxyl groups of Ser140 and Thr141 [23]. According to the model, the amino group of l-aspartate forms hydro- gen-bonding interactions with the side chains of Thr101, Asn142 and His188 [23]. In the present study, we performed site-directed mutagenesis experiments on all of these residues to provide further insight into the mechanism for the AspB-catalyzed deamination reaction. The results obtained also have implications for our understanding of the catalytic mechanism of AspA.

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

for the descending limb may be somewhat less than the calculated value.

Table 1. Kinetic parameters for the deamination of L-aspartate by wild-type AspB and the Thr101, Asn142 and His188 mutants. Steady-state kinetic parameters were determined in 50 mM sodium phosphate buffer (pH 8.5) at 25 (cid:2)C. Errors are standard deviations. ND, not determined (a conservative estimate of the sensitivity of the assay indicates an at least 106-fold decrease in kcat ⁄ Km com- pared to that of wild-type AspB). In those cases where no activity was detected upon prolonged incubation with substrate, kcat values are given as < 0.001.

Enzyme

)1Æs)1)

kcat (s)1)

Km (mM)

kcat ⁄ Km (M

Wild-type T101A T101S N142A N142Q H188A H188Q H188R H188K

40 ± 7 > 0.38 > 34 < 0.001 > 0.88 0.7 ± 0.1 < 0.001 < 0.001 < 0.001

15 ± 2 > 1000 > 1000 ND > 1000 27 ± 2 ND ND ND

2.7 · 103 3.8 · 10)1 3.4 · 101 ND 8.8 · 10)1 2.6 · 101 ND ND ND

Construction and purification of the AspB mutants

The residues selected for mutagenesis in the present study were Thr101, Ser140, Thr141, Asn142, Thr187, His188, Lys324 and Asn326 (Fig. 1). Similar to wild- type AspB, all mutant proteins were constructed as His6-tagged fusion proteins, produced in E. coli TOP10, and purified to > 95% homogeneity (as assessed by SDS ⁄ PAGE) using a Ni-based immobilized metal affinity chromatography procedure [22]. This purification procedure is highly specific for His6-tagged proteins, eliminating the possibility that wild-type and mutant AspB enzymes isolated from E. coli TOP10 are contaminated with native aspartase from this host. A previous study has shown that the recombinant His6- tagged wild-type AspB has only slightly reduced activ- ity compared to the native protein (without fusion tag) [22]. The yield of mutant protein from cell culture varied in the range 8–20 mgÆL)1.

tions of the kinetic assays, no activity could be detected for these mutants. A conservative estimate of the sensitivity of the assay indicates an at least 106-fold decrease in kcat ⁄ Km compared to that of wild-type AspB. Substitution of His188 with an alanine, how- ever, resulted in an active enzyme with an approxi- mately 57-fold reduction in kcat and a 1.8-fold increase in Km, which results in a approximately 100-fold decrease in kcat ⁄ Km. Hence, the major effect of this mutation is on the value of kcat. Each AspB mutant was analyzed by nondenaturing PAGE (data not shown). The mutant enzymes were found to migrate comparably with the wild-type enzyme, which suggests that the oligomeric association of the mutants was still intact and that gross confor- mational changes are unlikely to be present. The struc- tural integrity of some mutants was also assessed by circular dichroism (CD) spectroscopy (see below).

Mutagenesis of Thr101, Asn142 and His188

reduced only approximately

On the basis of a previously reported structural model of the complex in which l-aspartate (1) was docked manually into the active site of AspB, the residues responsible for binding the amino group of 1 are predicted to be Thr101, Asn142 and His188 [23]. To investigate the importance of these residues to the mechanism of AspB, eight single site-directed mutants were constructed in which Thr101 was replaced with an alanine or serine (T101A and T101S), Asn142 with an alanine or glutamine (N142A and N142Q) and His188 with an alanine, glutamine, lysine or arginine (H188A, H188Q, H188K and H188R). These mutations are expected to completely remove the functional side chain (e.g. T101A) or to replace the side chain with another one that has a similar functional group (e.g. T101S). The mutation of Thr101 to an alanine has a large effect on the catalytic efficiency of AspB. For the T101A mutant, a plot of various concentrations of 1 versus the initial rates measured at each concentration remained linear up to 1 m. Hence, the T101A mutant could not be saturated. Accordingly, only the kcat ⁄ Km was determined, and this parameter is reduced approx- imately 7100-fold compared to that of wild-type AspB (Table 1). The mutation of Thr101 to another residue with an aliphatic hydroxyl group (serine), however, has a less drastic effect on kcat ⁄ Km. For the T101S mutant, which could also not be saturated with 1, the kcat ⁄ Km is 80-fold. Replacement of Asn142 with an alanine resulted in a mutant enzyme without detectable activity (Table 1), emphasizing the importance of this residue to the mechanism of AspB. However, substitution of Asn142 with a glutamine, which also contains a terminal amide group, resulted in an active enzyme with an approxi- mately 3000-fold reduction in kcat ⁄ Km.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2997

The T101A, N142A and H188A mutants were ana- lyzed by CD, and the spectra of these mutants were comparable to that of wild-type AspB, indicating that The activities of the mutants were assayed using 1 as the substrate. It was found that replacement of His188 with a glutamine, lysine or arginine essentially abol- ished enzymatic activity (Table 1). Under the condi-

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

A

Fig. 3. Far-UV CD spectra of wild-type (Wt) and mutant enzymes. (A) Superimposed spectra of wild-type AspB and the T101A, N142A and H188A mutants. (B) Superimposed spectra of wild-type AspB and the S140A and T141A mutants. (C) Superimposed spec- tra of wild-type AspB and the T187A, K324A and N326A mutants. Spectra were measured in 10 mM NaH2PO4 buffer (pH 8.5) at a protein concentration of approximately 5 lM.

the mutations did not result in any major conforma- tional changes (Fig. 3A).

Mutagenesis of Ser140 and Thr141

B

To investigate the importance of Ser140 and Thr141, which are the two residues implicated in binding the C4 carboxylate group of 1 [23], to the mechanism of AspB, seven single site-directed mutants were con- structed (S140A, S140R, S140K, T141A, T141V, T141R and T141K), as well as two double mutants (S140G ⁄ T141G and S140K ⁄ T141K). Mutation to Gly, is expected to remove the functional Ala and Val group, whereas mutation to Arg or Lys may introduce a new functional group. It was found that replacement of Ser140 and Thr141 with an arginine or lysine essen- tially abolished enzymatic activity (Table 2). The T141K mutant had a low amount of activity (approxi- mately 40 000-fold reduction in kcat ⁄ Km), whereas the S140R, S140K and T141R mutants had no detectable activity. The T141V mutant and the two double mutants (S140G ⁄ T141G and S140K ⁄ T141K) also were completely inactive. Substitution of Ser140 and Thr141 by an alanine, however, resulted in active enzymes. For the T141A mutant, there is a 133-fold decrease in kcat and an approximately nine-fold decrease in Km

C

Table 2. Kinetic parameters for the deamination of L-aspartate by wild-type AspB and the Ser140 and Thr141 mutants. Steady-state kinetic parameters were determined in 50 mM sodium phosphate buffer (pH 8.5) at 25 (cid:2)C. Errors are standard deviations. ND, not determined.

Enzyme

)1Æs)1)

kcat (s)1)

Km (mM)

kcat ⁄ Km (M

40 ± 7 > 40 < 0.001 < 0.001 0.3 ± 0.01 < 0.001 < 0.001 > 0.027 < 0.001 < 0.001

15 ± 2 > 400 ND ND 1.7 ± 0.1 ND ND > 400 ND ND

2.7 · 103 1.0 · 102 ND ND 1.8 · 102 ND ND 6.8 · 10)2 ND ND

Wild-type S140A S140R S140K T141A T141V T141R T141K S140G ⁄ T141G S140K ⁄ T141K

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2998

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

using 1. As a result, the kcat ⁄ Km is reduced 15-fold. The major effect of this mutation is observed in kcat. For the S140A mutant, only the kcat ⁄ Km could be determined, and this parameter is reduced 27-fold compared to that of wild-type AspB. Because the increase in Km for the S140A mutant is > 27-fold, the kcat must be > 40 s)1. Hence, the major effect of this mutation is likely observed in Km. The CD spectra of the S140A and T141A mutants revealed no significant differences compared to that of the wild-type protein (Fig. 3B). at positions Thr187 and Asn326 resulted in active enzymes. Substitution of Thr187 with a serine resulted in a mutant (T187S) with a surprisingly improved kcat (approximately five-fold) compared to that of wild- type AspB. The Km value, however, increased signifi- cantly (approximately 11-fold). As a result, the kcat ⁄ Km is reduced approximately 2.3-fold. For the T187A, N326A and N326Q mutants, only the kcat ⁄ Km values could be determined, and these values are reduced 6280-fold, 22 500-fold and 168 750-fold, respectively, compared to the kcat ⁄ Km of wild-type AspB.

Mutagenesis of Thr187, Lys324 and Asn326

The T187A, K324A, and N326A mutants were also analyzed by CD and showed no detectable differences from the wild-type protein (Fig. 3C).

pH dependence of the kinetic parameters of the T141A and H188A mutants

The substrate-binding model suggests that residues Thr187, Lys324 and Asn326 are responsible for bind- ing the C1 carboxylate group of 1 [23]. To investigate the importance of these residues for AspB activity, 10 single site-directed mutants were constructed in which Thr187 was replaced by an alanine or serine (T187A and T187S), Asn326 with an alanine or glutamine (N326A and N326Q) and Lys324 with an alanine, serine, valine, histidine, arginine or aspartic acid (K324A, K324S, K324V, K324H, K324R and K324D). It was found that replacement of Lys324 with a small (alanine), polar (serine), charged (histi- dine, arginine and aspartate) or hydrophobic (valine) residue resulted in mutant enzymes with no detectable activity (Table 3), emphasizing the essential role of Lys324 in the mechanism of AspB. Prolonged incuba- tion with 1 revealed a small amount of activity for the K324R mutant, whereas all other mutants had no detectable activity. The activity of the K324R mutant is too low to measure kinetic parameters. Mutations To determine whether Thr141 or His188 was responsi- ble for the ascending limb (i.e. the basic group on the enzyme important for catalysis) or descending limb (i.e. the acidic group on the enzyme important for catalysis) of the pH-rate profile of wild-type AspB, the pH dependence of kcat ⁄ Km for the T141A- and H188A-catalyzed deamination of 1 was determined over the pH range 6.0–10. Similar to wild-type AspB, the pH-rate profiles of the T141A and H188A mutants (comprising the only mutations at these positions that resulted in an active enzyme) are bell-shaped with slopes of unity (Fig. 2). For the T141A mutant, pKa values of pK1 = 6.8 ± 0.1 and pK2 = 11.1 ± 0.5 were found. For the H188A mutant, pKa values of pK1 = 7.4 ± 0.3 and pK2 = 9.4 ± 0.3 were found.

Table 3. Kinetic parameters for the deamination of L-aspartate by wild-type AspB and the Thr187, Lys324 and Asn326 mutants. Steady-state kinetic parameters were determined in 50 mM sodium phosphate buffer (pH 8.5) at 25 (cid:2)C. Errors are standard deviations. ND, not determined.

Competitive inhibition of AspB and the T141A and H188A mutants

Enzyme

)1Æs)1)

kcat (s)1)

Km (mM)

kcat ⁄ Km (M

Wild-type T187A T187S K324A K324S K324V K324H K324D K324R N326A N326Q

40 ± 7 > 0.43 190 ± 7 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 > 0.12 > 0.016

15 ± 2 > 1000 163 ± 15 ND ND ND ND ND ND > 1000 > 1000

2.7 · 103 4.3 · 10)1 1.2 · 103 ND ND ND ND ND ND 1.2 · 10)1 1.6 · 10)2

These respectively

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

2999

It has previously been reported that 3-nitropropionate and d-malate are competitive inhibitors of the deami- nation activity of AspA with Ki values of 0.83 and 0.66 mm, observations [14]. prompted us to examine whether these compounds are also competitive inhibitors of AspB. Lineweaver–Burk reciprocal plots using three or four different inhibitor concentrations demonstrate that 3-nitropropionate and d-malate are competitive inhibitors of the aspartase activity of AspB with Ki values of 2.0 ± 0.5 mm and 68 ± 12 mm, respectively. These compounds are also competitive inhibitors of the aspartase activity of the T141A and H188A mutants. For the T141A mutant, Ki values of 0.5 ± 0.1 mm and 48 ± 12 mm were found for 3-nitropropionate and d-malate, respectively.

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

A

that replacing the histidine at position 188 with alanine results in a mutant with a surprisingly improved affin- ity for 3-nitropropionate (approximately 200-fold) and d-malate (approximately 170-fold).

Discussion

B

certain residues. The role of

On the basis of sequence analysis and crystallographic observations with AspA [Research Collaboratory for Structural Bioinformatics (RCSB) Protein Data Bank (PDB) entry 1JSW], AspB (RCSB PDB entry 1J3U) and FumC (in complex with a substrate analog, cit- rate; RCSB PDB entry 1FUO) [12,25], Fujii et al. [23] have made a structural model of the complex in which l-aspartate (1; Scheme 3) was docked manually into the active site of AspB. Guided by this substrate-bind- ing model, we have selected eight active site residues for mutagenesis to provide further insight into the mechanism of AspB. In addition, we have performed pH-rate and inhibition studies to further illuminate the mechanistic results obtained in these studies are interpreted and related to the mechanisms of AspA and other superfamily enzymes.

Fig. 4. Lineweaver–Burk reciprocal plots showing the competitive mode of inhibition of the H188A mutant by 3-nitropropionate (A) and D-malate (B). The curves were generated by fitting the data by nonlinear regression analysis using the equation for competitive inhibition, yielding the Ki values given in the text. For the H188A mutant, Ki values of 0.01 ± 0.001 mm and 0.4 ± 0.02 mm were found for 3-nitropropionate and d-malate, respectively (Fig. 4). The results show

Asn-142

T

A

h

O

s

Thr-101

H2N

r-

1

0

n - 1

H

4

1

2

His-188

OH

is - 1

Thr-141

8

8

Thr-141

O

N

Asn-326

O

O–

Asn-326

H

+ NH3

+ NH3

HN

O

Lys-324

O

NH2

Ser-140

O-

H

O–

–O2C

O

Thr-187

O

We first determined the probable chemical mecha- nism of the AspB-catalyzed reaction and pKa values of potential acid and base catalysts by pH-rate studies. From the ascending limb of the pH-rate profile of AspB, a residue with a pKa of 7.1 ± 0.1, which must be deprotonated for optimal activity, may be the gen- eral base catalyst involved in abstraction of the C3 proton of 1. From the descending limb, a residue with a pKa of 9.8 ± 0.2, which must be protonated for optimal activity, may be the general acid catalyst

4 B:H

+ NH3

Ser-140

H H 1 B:

H

O

Lys-324

Thr-187

2 + 3

Scheme 3. A schematic representation of the proposed catalytic mechanism of AspB.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3000

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

that His188 does not function as the typical general acid or general base catalyst in AspB, which is consis- tent with the absence of this residue in AspA.

involved in protonation of the leaving amino group of 1. However, it is important to emphasize that it is presently unknown whether the protonation state of the leaving group in the AspB-catalyzed reaction is the that of ammonia or ammonium ion [6,23]. If amino group is released as ammonia, a general acid catalyst may not be required for the reaction. The descending limbs of the kcat ⁄ Km versus pH and kcat versus pH profiles would then reflect the deprotonation of the amino group of the substrate (the pKa of the amino group of 1 = 9.8) and active-site bound inter- mediate (i.e. respectively. the putative enediolate), Hence, the observed pH dependence of the kinetic parameters for the AspB-catalyzed deamination of 1 is most simply explained in terms of the ionization of a single basic group at the active site and the ionization of the amino group of the substrate (1) or enzyme- bound intermediate (4) (Scheme 3).

One potential explanation for the loss in activity of the H188A mutant is that His188 could position and ‘lock’ the amino group of 1 in a favorable orientation for deamination. In this scenario, loss in activity of the H188A mutant is a result of the removal of optimal hydrogen-bonding interactions with the amino func- tionality, locating this group in an unfavorable posi- tion for the elimination reaction. Support for this view is provided by an analysis of the substrate specificity of the H188A mutant in the reverse amine addition reaction. In comparison to the wild-type AspB-cata- lyzed addition of methylamine to fumarate, the H188A mutant displays a two-fold increase in kcat and a higher Km for methylamine, suggesting that His188 is one of the residues that influences the nucleophile (i.e. amine) specificity of AspB (V. Puthan Veetil & G. J. Poelarends, unpublished results). Extrapolation of this observation to substrate binding supports the proposed role for His188 in binding the amino group of 1 (Scheme 3). The corresponding glutamine residue in AspA (Gln191) may participate in a similar hydro- gen-bonding interaction with the substrate. Another notable effect of the H188A mutant is the 170–200- fold increase in the binding affinity for the competitive inhibitors 3-nitropropionate and d-malate. This sug- gests that the decrease in catalysis for this mutant may be the result of a combination of effects, including the suboptimal positioning of the amino group and the slower release of product.

(Thr100 and Asn141) The substrate binding model of AspB suggests that His188 is one of the residues that interacts, via hydro- gen-bonding, with the amino group of 1 [23]. Intrigu- ingly, this histidine residue is conserved in FumC and other fumarase ⁄ aspartase superfamily members (e.g. argininosuccinate lyase, d-crystallin and adenylosucci- nate lyase) but is replaced by Gln191 in AspA [23]. Mutations of the conserved histidine residue severely impair catalysis in the former enzymes [26–28]. These results, taken together with crystallographic observa- tions, have led to several proposals for the catalytic role of the histidine [29–31]. These include roles for the histidine as the general base catalyst that abstracts the proton from C3 of the substrate, the general acid cata- lyst that protonates the leaving C2 group, or the criti- cal residue that activates an active site water molecule, which then functions as the C3 proton abstracting base [29–31].

same

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3001

To assess its role in catalysis, we have mutated His188 in AspB. Examination of the kinetic properties of the H188A mutant, which is the only mutation at this position that resulted in an active enzyme, shows that the major effect of this mutation is on the value of kcat. No significant change was observed in the CD spectrum of this mutant, demonstrating that the loss of activity resulting from replacement of His188 by alanine is not a consequence of the loss of structural integrity of the enzyme. Taken together, these observa- tions suggest an important role for His188 in catalysis. However, the pH-rate profile of the H188A mutant has the same overall shape as that of wild-type AspB, with pKa values of approximately 7.4 and 9.4, suggest- ing that the protonation or deprotonation of His188 is not responsible for the observed loss of activity on either side of the pH optimum. Therefore, we conclude In the substrate binding model of AspB, two other residues (Thr101 and Asn142) are implicated in bind- ing the amino group of 1 [23]. The corresponding res- idues in the FumC–citrate complex were found to assist His188 (corresponding to His188 in AspB) in binding an active site water molecule [23,25]. This geometry suggests that these three residues may be responsible for the positioning and activation of the nucleophilic water molecule in the FumC-catalyzed hydration reaction, as well as the positioning and protonation of the leaving hydroxyl group in the reverse dehydration reaction. To exam- ine the role of Thr101 and Asn142 in AspB, we have mutated these two residues. Complete removal of the functionality at position Thr101 by replacement with increase in Km an alanine leads to a significant (> 66-fold) and a large decrease in catalytic efficiency (> 7000-fold). The substitution at position Asn142 even results in a complete loss of activity. Both these alanine mutants retain their overall struc- tural integrity, as assessed by CD spectroscopy. These

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

that the most

observations, coupled with the enhancement in activ- ity observed with a serine (position Thr101) or gluta- mine (position Asn142) mutation at these positions, suggest likely role for Thr101 and Asn142 is to participate in a hydrogen-bonding inter- action with the amino group of 1 (Scheme 3). Simi- larly, the corresponding residues in AspA (Thr104 and Asn145), which are positionally conserved but have slighty different side chain orientations [12,23], may assist Gln191 in binding the amino group of the substrate.

In the substrate binding model of AspB, the hydro- xyl groups of Ser140 and Thr141 are hydrogen bonded to the C4 carboxylate group of 1 [23]. This suggests roles for the hydroxyl functional groups of Ser140 and Thr141 in binding the substrate, and possible interme- diates, in the reaction mechanism of AspB. Crystallo- graphic studies on FumC support this view [25]. In the structure of the FumC–citrate complex, Ser139 and Ser140 (corresponding to Ser140 and Thr141 in AspB) are hydrogen bonded to one of the carboxylate groups of citrate [23,25]. This geometry suggests the impor- tance of these residues in binding one of the two carboxylate groups of the substrate. In AspB, Ser140 the N-terminal end of and Thr141 are located at a-helix 6 in the positively charged environment created by a dipole moment of this helix [23].

stabilized carbanion) derived by abstraction of the C3 proton from 1, or an enol that could be obtained by protonation of an enolate anion, remains unknown. Presumably, Thr141, together with Ser140, polarizes the C4 carboxylate group and stabilizes the enediolate intermediate (4) formed upon abstraction of the C3 proton (Scheme 3). This proposed role for Thr141 in the mechanism of AspB shares similarity with the one proposed for Glu317 in mandelate racemase [32]. This the (R)- and enzyme catalyzes the equilibration of (S)-enantiomers of mandelate. On the basis of muta- genesis, crystallographic and kinetic isotope studies, Mitra et al. [32] proposed that Glu317 in mandelate racemase functions as a general acid catalyst in the concerted general acid–general base catalyzed forma- tion of a stabilized enolic tautomer of mandelic acid as a reaction intermediate. Glu317 can function as a general acid catalyst because it is protonated when substrate binds at the active site and it is properly positioned for partial proton transfer to the carboxyl- ate group of mandelic acid to form a strong hydro- it is clear gen-bonded enolic intermediate. However, from the pH-rate profile of the T141A mutant of AspB that Thr141 is not the acidic group responsible for the descending limb of the pH-rate profile of the wild-type AspB. Therefore, we conclude that protonation state of Thr141 does not show up in the pH versus kcat ⁄ Km profile of AspB and that the descending limb of this profile is caused by deproto- nation of either the substrate or an unknown acidic group on the enzyme important for catalysis. The observation that electrophilic catalysis by Thr141 in AspB (i.e. there is a 133-fold drop in the value of kcat for the T141A mutant) is not as important as that by Glu317 in mandelate racemase (i.e. there is a 4500- fold drop in the value of kcat for the E317Q mutant) [32] could be explained, at least in part, by the pres- ence of the dipole moment of a-helix 6, which likely assists Thr141 in stabilizing the negative charge that develops on one of the C4 carboxylate oxygens upon proton abstraction [23].

side-chain orientations are In the present study, Ser140 and Thr141 in AspB were mutated to assess their role in catalysis. The best characterized mutants are the S140A and T141A mutants, which have measurable activity and show no significant differences in their CD spectra compared to that of the wild-type AspB. Examination of the kinetic properties for the S140A mutant shows that there is a small effect on kcat ⁄ Km and a larger increase in Km, suggesting that Ser140 plays a major role in binding the C4 carboxylate group of 1 (Scheme 3). By contrast to these observations, examination of the kinetic prop- erties for the T141A mutant shows that the major effect of replacing Thr141 by alanine is on the value of kcat. This suggests an important role for Thr141 in catalysis. Another notable effect of the Thr141 muta- tion is the increase in the binding affinity for the sub- strate (assuming that the Km reflects substrate binding) and for the competitive inhibitors 3-nitropropionate and d-malate.

respectively) and a three-

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3002

A comparison of the crystal structures of AspA and AspB shows that Ser140 and Thr141 of AspB are posi- tionally conserved as Ser143 and Thr144 in AspA (although the slightly different) [23]. Replacement of Ser143 with glycine or threonine caused a significant decrease in kcat (10- and 100-fold, to four-fold increase in Km using 1 [21]. On the basis of these observations, Ser143 in AspA was proposed to func- tion as the general acid catalyst that protonates the leaving C2 amino group (Scheme 2) [21]. In view of the crystallographic observations with FumC [25] and A major part of the loss in activity of the T141A mutant is likely a result of the removal of optimal hydrogen-bonding interactions with the proposed eno- late anion intermediate formed at the C4 carbonyl position of 1, making the abstraction of a proton from the C3 position less favorable. Whether the intermediate is an enolate anion (i.e. a resonance-

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

[23]

it is doubtful

the substrate-binding model of AspB [23], and taken together with the mutagenesis results of the present study, that Ser143 functions as the general acid catalyst in AspA. However, in the absence structure of AspA in complex with of a crystal substrate (or a substrate analog), we can only speculate that the mechanistic roles of Ser143 and Thr144 in AspA share similarity with those proposed in the pres- ent study for Ser140 and Thr141 in AspB (Scheme 3).

[33]. Our

In the structure of the FumC–citrate complex, one of the other carboxylate groups of citrate interacts with the side chain of Lys324 (corresponding to Lys327 in AspA and Lys324 in AspB) [23,25]. Previ- ous studies have shown that mutation of Lys327 in AspA results in a six-fold increase in Km and a large (> 300-fold) decrease in kcat [20]. These observations suggest a role for the lysine residue in binding one of the two carboxylate groups of the substrate. Accord- ing to the substrate-binding model of AspB, Lys324 binds the C1 carboxylate group of 1 [23]. The stron- gest support for this orientation of substrate in the active site comes from the observation that substitu- tion of Lys327 in AspA with an asparagine changes the substrate specificity of the enzyme and allows it to process l-aspartate-a-amide experiments clearly show that Lys324 is an essential residue in the AspB-catalyzed reaction because mutations at this position result in a complete loss of activity. More- over, the mutant enzymes appear to retain an intact overall structural integrity (as shown for the K324A mutant by CD spectroscopy). It is therefore reason- able to conclude that Lys324 is crucial for substrate binding through an interaction with the C1 carboxyl- ate group (Scheme 3).

Fujii et al. (Scheme 3), although other mecha- nisms cannot be ruled out. In this mechanism, an active site base (with a pKa of approximately 7.1) abstracts the proton from C3 of 1 to form an enedio- stabilized by both late intermediate (4), which is Ser140 and Thr141. The interaction of two hydroxyl functional groups with one carboxylate group is con- sistent with a dianionic aci-carboxylate intermediate. The identity of the residue with the pKa of 7.1 is not known. (It has been suggested previously [23] that Ser318 may function as the general base in AspB. However, this residue is not located in the presumed active site of the enzyme. To establish whether the loop containing Ser318 may undergo a large confor- mational change upon substrate binding, positioning Ser318 in the vicinity of C3 of the substrate, we have initiated studies that aim to solve the X-ray structure of the enzyme–substrate complex.) In the next step, the enediolate intermediate collapses and eliminates ammonia (3) to form the fumarate (2) product. Thr101, Asn142 and His188 could position and ‘lock’ the amino group in a favorable orientation for deami- nation, whereas Thr187, Lys324 and Asn326 bind the C1 carboxylate group of the substrate. Strong support for this arrangement in AspB (Scheme 3) also comes from the interactions observed in the recently solved crystal structure of the fumarase ⁄ aspartase superfam- ily enzyme adenylosuccinate lyase complexed with adenylosuccinate [31]. In the structure of this com- plex, Thr122 and Ser123 interact with the d-carboxyl- ate group (on which the negative charge accumulates in the aci-carboxylate intermediate) of the succinyl moiety of adenylosuccinate, whereas Thr170, Lys301 and Asn303 interact with the c-carboxylate group. The positional conservation of these carboxylate bind- ing residues (where Ser140, Thr141, Thr187, Lys324 and Asn326 in AspB are replaced by Thr122, Ser123, Thr170, Lys301 and Asn303 in adenylosuccinate lyase) suggests a similar mode of substrate binding for these two superfamily members.

In the region surrounding Lys324 in the structure of the FumC–citrate complex, there are two residues (Thr187 and Asn326) that are well positioned to assist the citrate carboxylate Lys324 in binding one of groups [25]. A comparison of this structure with those of AspA and AspB shows that these residues are posi- tionally conserved as Thr190 and Asn329 in AspA and Thr187 and Asn326 in AspB [23]. The results obtained in the present study support a role for Thr187 and Asn326 in assisting Lys324 to bind and position the substrate (through interactions with the C1 carboxylate group of 1) (Scheme 3) because mutants of these two residues display reduced activity with large effects on the Km for 1. Similar mechanistic roles may be proposed for the corresponding residues in AspA.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3003

In summary, the mutagenesis and pH-rate studies performed in the present study support the substrate- binding model and initial mechanism reported by The positional conservation of seven out of eight active site residues also suggests a mechanism for AspA that largely parallels that proposed for AspB (Scheme 3) where the residues in AspB (Thr101, Ser140, Thr141, Asn142, Thr187, Lys324 and Asn326) are replaced by the corresponding ones in AspA (Thr104, Ser143, Thr144, Asn145, Thr190, Lys327 and Asn329). This proposed mechanism needs to be corroborated by future crystallographic studies. The utility of 3-nitropro- prionate as a potent competitive inhibitor and potential crystallographic ligand for AspB could help to identify the additional features that are necessary for a fully active and specific aspartase.

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

Experimental procedures

The Netherlands). The cuvettes were mixed using a stirr ⁄ add cuvette mixer (Bel-Art Products, Pequannock, NJ, USA). The kinetic data were fitted by nonlinear regression analysis using grafit (Erithacus, Software Ltd, Horley, UK) obtained from Sigma-Aldrich Chemical Co. The CD spectra were recorded on a model 62A-DS spectropolarimeter from AVIV Biomedical, Inc (Lakewood, NJ, USA).

Materials

l-Aspartic acid, either the free acid or the sodium salt, and l-arabinose were purchased from Sigma-Aldrich Chemical Co. (St Louis, MO, USA). Ingredients for buffers and media were obtained from Duchefa Biochemie (Haarlem, The Netherlands) or Merck (Darmstadt, Germany). Molec- ular biology reagents, including restriction enzymes, PCR reagents, T4 DNA ligase, DNA ladders, agarose and pro- tein molecular weight standards, were obtained from F. Hoffman-LaRoche, Ltd. (Basel, Switzerland), Promega Corp. (Madison, WI, USA), Invitrogen Corp. (Carlsbad, CA, USA), Finnzymes (Espoo, Finland) or New England Biolabs (Ipswich, MA, USA). PCR purification, gel extrac- tion, and Miniprep kits were provided by Macherey-Nagel (Du¨ ren, Germany). Pre-packed PD-10 Sephadex G-25 col- umns were purchased from GE Healthcare Bio-Sciences AB (Uppsala, Sweden). Oligonucleotides for DNA amplifica- tion were synthesized by Operon Biotechnologies (Cologne, Germany).

Construction of AspB mutants

Two standard mutagenesis methods were used to introduce site-specific mutations into the aspB gene [36,37]. Most mutants were generated by the overlap extension PCR method using plasmid pUCBA as the template [36]. The final PCR products were gel purified, digested with NcoI and HindIII restriction enzymes, and ligated in frame with both the initiation ATG start codon and the sequence that codes for the polyhistidine region of the expression vector pBAD ⁄ Myc-His A. A few mutants were generated by the QuikChange mutagenesis method (Stratagene, La Jolla, CA, USA) using plasmid pBAD (AspB-His), which encodes His6-tagged wild-type AspB [22], as the template [37]. All mutant genes were completely sequenced (with overlapping reads) to verify that only the intended mutation had been introduced.

Bacterial strains, plasmids and growth conditions

Escherichia coli strains DH10B (Invitrogen Corp.) and XL1-Blue (Stratagene, La Jolla, CA, USA) were used for cloning and isolation of plasmids. Escherichia coli strain TOP10 (Invitrogen Corp.) was used in combination with the pBAD ⁄ Myc-His A vector (Invitrogen Corp.) for recom- binant protein production. Plasmid pUCBA [11], the DNA source for the aspB gene, was a kind gift from Y. Kawata (Department of Biotechnology, Tottori University, Japan). Escherichia coli cells were grown at 37 (cid:2)C in LB media. (15 gÆL)1), ampicillin (Ap; When required, Difco agar 100 lgÆmL)1) and ⁄ or arabinose (0.04% w ⁄ v) were added to the medium.

Expression and purification of AspB wild-type and mutants

The AspB enzyme, either wild-type or mutant, was pro- duced in E. coli TOP10 using the pBAD expression system. Fresh TOP10 cells containing the appropriate expression plasmid were collected from a LB ⁄ Ap plate using a sterile loop and used to inoculate 10 mL of LB ⁄ Ap medium. After overnight growth at 37 (cid:2)C, a sufficient quantity of the cul- ture was used to inoculate 1 L of LB ⁄ Ap medium in a 5 L Erlenmeyer flask to an initial A600 of approximately 0.02. Cultures were grown until A600 of 0.4–0.6 was reached at 37 (cid:2)C with vigorous shaking and then induced with arabi- nose [0.04% (w ⁄ v) final concentration]. Incubation was continued for 10–12 h at 37 (cid:2)C. Cells were harvested by centrifugation (6000 g for 15 min) and stored at )20 (cid:2)C until further use.

In a typical purification experiment, cells of a 1 L culture lysis buffer were thawed and suspended in 10 mL of 10 mm imidazole, 300 mm NaCl, (50 mm NaH2PO4, pH 8.0). Cells were disrupted by sonication for 4 · 1 min (with 4–6 min of rest in between each cycle) at a 60 W out- put, after which unbroken cells and debris were removed by centrifugation (10 000 g for 30 min). The supernatant was filtered through a 0.45 lm pore diameter filter and incubated with Ni-nitrilotriacetic acid (1 mL slurry in a small column at 4 (cid:2)C for ‡ 18 h), which had previously been equilibrated with lysis buffer. The nonbound proteins

Techniques for restriction enzyme digestions, ligation, trans- formation and other standard molecular biology manipula- tions were based on previously described methods [34] or carried out in accordance with the manufacturer’s instruc- tions. The PCR was carried out in a DNA thermal cycler (model GS-1) obtained from Biolegio (Nijmegen, The Neth- erlands). DNA sequencing was performed by ServiceXS (Lei- den, The Netherlands) or Macrogen (Seoul, Korea). Protein was analyzed by PAGE under either denaturing conditions using SDS or native conditions on gels containing 7.5–10% polyacrylamide. The gels were stained with Coomassie bril- liant blue. Protein concentrations were determined by the method of Waddell [35]. Kinetic data were obtained on a V-650 spectrophotometer obtained from Jasco (IJsselstein,

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3004

General methods

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

all experiments. It was not possible to collect data at pH values above 10 as a result of enzyme denaturation. Conse- quently, data could not be collected to high enough pH values to clearly define the descending limb for wild-type AspB and the T141A mutant. In these cases, the pK2 value reported is an estimate obtained from grafit software.

were eluted from the column by gravity flow. The column was first washed with lysis buffer (10 mL) and then with buffer A (50 mm NaH2PO4, 300 mm NaCl, 20 mm imidaz- ole, pH 8.0; 10 mL). Retained proteins were eluted with buffer B (50 mm NaH2PO4, 300 mm NaCl, 250 mm imidaz- ole, pH 8.0; 3.0 mL). Fractions (approximately 0.5 mL) were analyzed by SDS ⁄ PAGE on gels containing 10% acrylamide, and those that contained purified aspartase were pooled and the buffer was exchanged against 50 mm NaH2PO4 (pH 8.0) using a pre-packed PD-10 Sephadex G-25 gelfiltration column. The purified enzyme was stored at 4 (cid:2)C or )80 (cid:2)C until further use.

Kinetics of reversible inhibition of AspB, T141A and H188A by 3-nitropropionate and D-malate

Circular dichroism spectra of the wild-type protein and the purified mutants were measured in 10 mm NaH2PO4 buffer (pH 8.5) at a concentration of approximately 5 lm in a CD cell with an optical path length of 1.0 mm.

CD spectroscopy

The reversible inhibition of AspB and the T141A and H188A mutants was examined using 3-nitropropionate and d-malate. For each experiment, a small amount of enzyme was diluted into 15 mL of assay buffer (50 mm NaH2PO4 buffer, pH 8.5). Subsequently, an aliquot of inhibitor from a stock solution (in 50 mm NaH2PO4 buffer, pH 8.5) was added to the diluted enzyme solution to yield the approxi- inhibitor concentration. After approximately mate final 30 min, aliquots (1 mL) of the resulting solution were removed and assayed using 10–15 concentrations (range 5–100 mm) of 1. The final concentrations of the inhibitors were in the range 0–10 mm. The mode of inhibition was determined from Lineweaver–Burk reciprocal plots [38]. The inhibition constants (Ki values) were obtained by fitting the data by nonlinear regression analysis using the equation for competitive inhibition provided in the grafit software.

Enzymatic assay

Acknowledgements

Kinetic assays were performed at 25 (cid:2)C in 50 mm NaH2PO4 buffer, pH 8.5, observing the increase in absor- bance at 240 nm corresponding to the formation of fuma- rate (2) (e = 2530 m)1Æcm)1) as described previously [22]. An aliquot of AspB, either wild-type or mutant, was diluted into buffer (20 mL) and incubated for ‡ 30 min at 25 (cid:2)C. Subsequently, a 1 mL portion was transferred to a 10 mm quartz cuvette and the enzyme activity was assayed by the addition of a small quantity (1–10 lL) of sodium l-aspar- tate (1) from a stock solution. The stock solutions were made up in 50 mm NaH2PO4 buffer (pH 8.5). The concen- trations of 1 used in the assay varied in the range 5–1000 mm.

pH dependence of the kinetic parameters of AspB, T141A and H188A

for (Department Japan)

The pH dependence of the steady-state kinetic parameters was determined in 100 mm sodium phosphate buffers with pH values in the range 5.5–10.0. The buffers were made up by combining appropriate quantities of 100 mm NaH2PO4, 100 mm Na2HPO4 and 100 mm Na3PO4 buffers to maintain constant ionic strength. For each pH value, a sufficient quantity of enzyme (from a stock solution in 50 mm NaH2PO4, pH 8.0) was equilibrated in buffer (20 mL) for ‡ 30 min at 37 (cid:2)C. The addition of enzyme did not signifi- cantly change the pH. Subsequently, aliquots (1 mL) were removed and assayed for activity using concentrations of l-aspartic acid (1) in the range 5–100 mm. Stock solutions of 1 were made in 100 mm NaH2PO4 buffer. The pH of the stock solutions was adjusted to each desired pH value (5.5– 10.0). The volume of substrate added was 20 lL or less in

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3005

This research was supported financially by The Neth- erlands Ministry of Economic Affairs and the B-Basic partner organizations (http://www.b-basic.nl) through B-Basic, a public–private NWO-ACTS programme. G. J. P. was supported by grants (VENI 700.54.401 and VIDI 700.56.421) from the Division of Chemical Sciences of The Netherlands Organisation of Scien- tific Research (NWO-CW). We thank Dr Yasushi of Biotechnology, Tottori Kawata University, the kind gift of plasmid pUCBA and Dr Yasuo Hata (Institute for Chemical Research, Kyoto University, Japan) for the kind gift of the PDB file of the homotetrameric AspB struc- ture. We gratefully acknowledge Gea K. Schuurman- Wolters (Department of Biochemistry, University of Groningen) for her assistance in acquiring the CD spectra. We are also grateful to Jeroen Bonet, Elisa Hoekstra and Sam Gijsberts for their assistance in the construction and purification of several AspB mutants. Finally, we thank Dr Oliver May, Dr Stefaan de Wildeman, Dr Friso van Assema and Dr Bernard Kaptein (DSM Geleen, The Netherlands) for insightful discussions.

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

14 Falzone CJ, Karsten WE, Conley JD & Viola RE

References

(1988) L-Aspartase from Escherichia coli: substrate specificity and the role of divalent metal ions. Biochem- istry 27, 9089–9093.

15 Emery TF (1963) Aspartase-catalyzed synthesis of

1 Rudolph FB & Fromm HJ (1971) The purification and properties of aspartase from Escherichia coli. Arch Bio- chem Biophys 147, 92–98.

2 Suzuki S, Yamaguchi J & Tokushige M (1973) Studies

N-hydroxyaspartic acid. Biochemistry 2, 1041–1045. 16 Viola RE (2000) L-Aspartase: new tricks from an old enzyme. Adv Enzymol Relat Areas Mol Biol 74, 295– 341.

on aspartase. I. Purification and molecular properties of aspartase from Escherichia coli. Biochim Biophys Acta 321, 369–381.

3 Takagi JS, Ida N, Tokushige M, Sakamoto H &

17 Gawron O & Fondy TP (1959) Stereochemistry of the fumarase and aspartase catalyzed reactions and of the Krebs cycle from fumaric acid to d-isocitric acid. J Am Chem Soc 81, 6333–6334.

Shimura Y (1985) Cloning and nucleotide sequence of the aspartase gene of Escherichia coli W. Nucleic Acids Res 13, 2063–2074.

18 Hanson KR & Havir EA (1972) The enzymic elimina-

4 Karsten WE, Hunsley JR & Viola RE (1985) Purifica-

tion of aspartase and aspartokinase-homoserine dehydrogenase I from Escherichia coli by dye-ligand chromatography. Anal Biochem 147, 336–341.

tion of ammonia. In The Enzymes, Vol. 3 (Boyer PD ed.), pp. 75–166. Academic Press, New York, NY. 19 Porter DJT & Bright HJ (1980) 3-Carbanionic substrate analogues bind very tightly to fumarase and aspartase. J Biol Chem 255, 4772–4780.

5 Nuiry II, Hermes JD, Weiss PM, Chen C-Y & Cook PF (1984) Kinetic mechanism and location of rate- determening steps for aspartase from Hafnia alvei. Biochemistry 23, 5168–5175.

20 Saribas AS, Schindler JF & Viola RE (1994) Mutagenic investigation of conserved functional amino acids in Escherichia coli L-aspartase. J Biol Chem 269, 6313– 6319.

21 Jayasekera MMK, Shi W, Farber GK & Viola RE

6 Yoon M-Y, Thayer-Cook KA, Berdis AJ, Karsten WE, Schnackerz KD & Cook PF (1995) Acid-base chemical mechanism of aspartase from Hafnia alvei. Arch Biochem Biophys 320, 115–122.

(1997) Evaluation of functionally important amino acids in L-aspartate ammonia-lyase from Escherichia coli. Biochemistry 36, 9145–9150.

7 Takagi JS, Fukunaga R, Tokushige M & Katsuki H (1984) Purification, crystallization, and molecular properties of aspartase from Pseudomonas fluorescens. J Biochem (Tokyo) 96, 545–552.

22 Weiner B, Poelarends GJ, Janssen DB & Feringa BL (2008) Biocatalytic enantioselective synthesis of N-substituted aspartic acids by aspartate ammonia lyase. Chem Eur J 14, 10094–10100.

8 Takagi JS, Tokushige M & Shimura Y (1986) Clon- ing and nucleotide sequence of the aspartase gene of Pseudomonas fluorescens. J Biochem (Tokyo) 100, 697–705.

23 Fujii T, Sakai H, Kawata Y & Hata Y (2003) Crystal structure of thermostable aspartase from Bacillus sp. YM55-1: structure-based exploration of functional sites in the aspartase family. J Mol Biol 328, 635– 654.

24 Stivers JT, Abeygunawardana C, Mildvan AS,

9 Sun D & Setlow P (1991) Cloning, nucleotide sequence, and expression of the Bacillus subtilis ans operon, which codes for L-asparaginase and L-aspartase. J Bacteriol 173, 3831–3845.

Hajipour G & Whitman CP (1996) 4-Oxalocrotonate tautomerase: pH dependences of catalysis and pKa values of active site residues. Biochemistry 35, 814– 823.

25 Weaver T & Banaszak L (1996) Crystallographic

10 Kawata Y, Tamura K, Yano S, Mizobata T, Nagai J, Esaki N, Soda K, Tokushige M & Yumoto N (1999) Purification and characterization of thermostable aspar- tase from Bacillus sp. YM55-1. Arch Biochem Biophys 366, 40–46.

studies of the catalytic and a second site in fumarase C from Escherichia coli. Biochemistry 35, 13955– 13965.

11 Kawata Y, Tamura K, Kawamura M, Ikei K, Mizobata T, Nagai J, Fujita M, Yano S, Tokushige M & Yumoto N (2000) Cloning and over-expression of thermostable Bacillus sp. YM55-1 aspartase and site-directed muta- genesis for probing a catalytic residue. Eur J Biochem 267, 1847–1857.

26 Weaver T, Lees M & Banaszak L (1997) Mutations of fumarase that distinguish between the active site and a nearby dicarboxylic acid binding site. Protein Sci 6, 834–842.

12 Shi W, Dunbar J, Jayasekera MMK, Viola RE & Farber GK (1997) The structure of L-aspartate ammonia-lyase from Escherichia coli. Biochemistry 36, 9136–9144.

27 Lee TT, Worby C, Bao Z-Q, Dixon JE & Colman RF (1999) His68 and His141 are critical contributors to the intersubunit catalytic site of adenylosuccinate lyase of Bacillus subtilis. Biochemistry 38, 22–32.

13 Karsten WE, Gates RB & Viola RE (1986) Kinetic

28 Patejunas G, Barbosa P, Lacombe M & O¢Brien WE (1995) Exploring the role of histidines in the catalytic

studies of L-aspartase from Escherichia coli: substrate activation. Biochemistry 25, 1299–1303.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3006

V. Puthan Veetil et al.

Mechanism of aspartase from Bacillus sp. YM55-1

34 Sambrook J, Fritsch EF & Maniatis T (1989)

activity of duck d-crystallins using site-directed muta- genesis. Exp Eye Res 61, 151–154.

29 Blanchard JS & Cleland WW (1980) Use of isotope

Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

effects to deduce the chemical mechanism of fumarase. Biochemistry 19, 4506–4513.

35 Waddell WJ (1956) A simple ultraviolet spectrophoto- metric method for the determination of protein. J Lab Clin Med 48, 311–314.

36 Ho SN, Hunt HD, Horton RM, Pullen JK & Pease

30 Wu C-Y, Lee H-J, Wu S-H, Chen S-T, Chiou S-H & Chang G-G (1998) Chemical mechanism of the endo- genous argininosuccinate lyase activity of duck lens d2-crystallin. Biochem J 333, 327–334.

31 Tsai M, Koo J, Yip P, Colman RF, Segall ML &

LR (1989) Site-directed mutagenesis by overlap exten- sion using the polymerase chain reaction. Gene 77, 51–59.

37 Braman J, Papworth C & Greener A (1996) Site-direc-

Howell PL (2007) Substrate and product complexes of Escherichia coli adenylosuccinate lyase provide new insights into the enzymatic mechanism. J Mol Biol 370, 541–554.

ted mutagenesis using double-stranded plasmid DNA templates. Methods Mol Biol 57, 31–44.

38 Fersht A (1999) Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding, pp. 112–114. W.H. Freeman and Company, New York, NY.

32 Mitra B, Kallarakal AT, Kozarich JW, Gerlt JA, Clifton JG, Petsko GA & Kenyon GL (1995) Mechanism of the reaction catalyzed by mandelate racemase: importance of electrophilic catalysis by glutamic acid 317. Biochemistry 34, 2777–2787. 33 Asano Y, Kira I & Yokozeki K (2005) Alteration of

39 DeLano WL (2002) The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, CA (http:// www.pymol.org).

substrate specificity of aspartase by directed evolution. Biomol Eng 22, 95–101.

FEBS Journal 276 (2009) 2994–3007 ª 2009 The Authors Journal compilation ª 2009 FEBS

3007