HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 25, 18939-18945, June 23, 2000

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ishijima, J. Articles by Kuramitsu, S. Search for Related Content PubMed PubMed Citation Articles by Ishijima, J. Articles by Kuramitsu, S. Social Bookmarking                      What's this?

Free Energy Requirement for Domain Movement of an Enzyme^*

Jun Ishijima, Tadashi Nakai^§, Shin-ichi Kawaguchi, Ken Hirotsu^§, and Seiki Kuramitsu^¶**

From the  Department of Biology, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan, the ^§ Department of Chemistry, Graduate School of Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan, ^¶ Harima Institute/SPring-8, The Institute of Physical and Chemical Research (RIKEN), Sayo-gun, Hyogo 679-5148, Japan, and  RIKEN Genomic Sciences Center, Wako, Saitama 351-0198, Japan

Received for publication, August 30, 1999, and in revised form, February 9, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Domain movement is sometimes essential for substrate recognition by an enzyme. X-ray crystallography of aminotransferase with a series of aliphatic substrates showed that the domain movement of aspartate aminotransferase was changed dramatically from an open to a closed form by the addition of only one CH₂ to the side chain of the C4 substrate CH₃(CH₂)C₍₎H(NH₃⁺)COO. These crystallographic results and reaction kinetics (Kawaguchi, S., Nobe, Y., Yasuoka, J., Wakamiya, T., Kusumoto, S., and Kuramitsu, S. (1997) J. Biochem. (Tokyo) 122, 55-63; Kawaguchi, S. and Kuramitsu, S. (1998) J. Biol. Chem. 273, 18353-18364) enabled us to estimate the free energy required for the domain movement.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The induced fit movement of an enzyme molecule upon the binding of substrate (1) has been characterized for many enzymes and is essential for catalysis. Domain closure of such enzymes can drastically change their active site environment from hydrophilic to hydrophobic, and such closure allows the enzymes to undergo reactions that are difficult in the aqueous phase. In order to elucidate the detailed mechanism of a given enzyme, it is necessary to estimate quantitatively the energy required for domain movement. For the last 2 decades, molecular dynamic calculations have been used to predict the domain movement of enzymes (2, 3). Recent studies indicate that single molecule measurements may also be useful in determining the energy required for domain fluctuation (4, 5). Despite these trials, it has generally proven difficult to confirm quantitatively the free energy required for domain movement of an enzyme. In this paper, we present a new method of estimating the energy required for domain movement by analyzing the reactions of two aminotransferases with a series of aliphatic -amino acid substrates.

Escherichia coli aspartate aminotransferase (EC 2.6.1.1) (AspAT)¹ with the bound coenzyme pyridoxal 5'-phosphate (PLP) reacts with an -amino acid to form the pyridoxamine 5'-phosphate (PMP) form of the enzyme and the -keto acid (6-11) as follows.

AspAT is a unique "dual substrate" enzyme that catalyzes this reaction for both acidic and hydrophobic amino acids (12, 13). In the catalytic mechanism, the catalytic group Lys²⁵⁸ is commonly used for both types of substrate. Arg³⁸⁶ is also frequently used for recognition of the -carboxyl group of both types of substrate.

With acidic substrates, AspAT undergoes a large domain movement. Arg²⁹²* ² moves from the outside to the inside of the active site with its accompanying water molecules and recognizes the -carboxyl group of the substrate (6-9). (This movement of Arg²⁹²* is similar to that for a hydrophobic substrate (see Fig. 3c).)

With hydrophobic substrates, it is known that the catalytic efficiency increases in proportion to the side chain length of a series of straight aliphatic substrates (Cn substrates) (12, 13). Surprisingly, consecutive additions of a single methylene group to the substrate (from C4 to C7) produce a constant effect on the stabilization energy of the transition state (ES) relative to the unbound state (E + S) (Refs. 12 and 13; see Fig. 5, inset). The energy contribution of one methylene group in the alkyl chain to AspAT is 0.65 kcal mol¹ for longer hydrophobic substrates with Cn, n  5. Similar phenomena have been reported for many enzymes (14-17). The energetic contribution of one methylene group in the substrate is 0.3 kcal mol¹ for horseradish peroxidase (14), -0.4 kcal mol¹ for Bacillus amyloliquefaciens subtilisin (15), 0.9 kcal mol¹ for Paracoccus denitrificans AroAT (16), and 1.5 kcal mol¹ for bovine -chymotrypsin (17). This apparent uniformity of the substrate-binding site will be achieved by fluctuation of the enzyme molecule. Therefore, the linear correlation of the free energy with the chain length (from C5 to C7) of the substrate for AspAT (Fig. 5) suggests an apparently uniform hydrophobic environment of the substrate-binding pocket of this enzyme.

In this study, we determined the three-dimensional structure of AspAT complexed with a series of straight chain aliphatic amino acids. With hydrophobic substrates with three (alanine) or four (2-amino butyric acid) carbon atoms, Arg²⁹²* remained outside the active site, as expected from the previous results with other hydrophobic substrates (12, 18, 19). We were, however, surprised to observe that the positively charged -amino group of Arg²⁹²* moved to bind hydrophobic substrates with Cn, n  5. Domain closure was also observed, as with acidic substrates. A detailed analysis of these structural and kinetic phenomena enabled us to estimate the free energy required for domain movement. The energy thus obtained seems to explain the phenomena implied by single molecule measurements.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Chemicals-- A series of aliphatic -amino acids ((CH₃(CH₂)_n3C₍₎H(NH₃⁺)COO, n = 3-6) were covalently bound to PLP with NaBH₄ as a reducing agent. These PLP-amino acids (Cn-PLP, n = 3-6) were purified on a DOWEX 1-X8 column (Dow Chemical). The structure of each product was confirmed by atomic composition, electrospray ionization mass spectrometry, and NMR. X-ray crystallography of the enzymes complexed with these substrate analogs (Protein Data Bank (PDB) codes 1C9C, 1CQ6, 1CQ7, and 1CQ8) also confirmed the structure of these compounds.

Protein Purification and Crystallization-- AspAT was purified as described previously (11). It was then converted to the apoenzyme by adding 10 mM cysteine sulfinate and 0.5 M potassium phosphate (pH 8.0); excess PLP and cysteine sulfinate were removed using Sephadex G-50 gel filtration. The apoenzyme was converted to a holoenzyme by adding a 2-fold excess (with respect to enzyme concentration) of PLP-amino acid to the protein solution. Crystals of holoenzyme (Cn-PLP complex) were grown by the hanging drop/vapor diffusion method using ammonium sulfate as the precipitant. The 5-µl drops containing 40 mg ml¹ protein in 10 mM potassium phosphate, 10 µM PLP-amino acid, and 0.3 mM NaN₃ at pH 8.0 were mixed with an equal volume of reservoir solution that contained 35% saturated ammonium sulfate, 10 mM potassium phosphate, and 0.3 mM NaN₃ at pH 8.0 and then equilibrated against the reservoir solution at 20 °C. After 4 days, the drop was seeded with a small crystal obtained beforehand. The crystals reached their maximal size of 0.7 × 0.2 × 0.04 mm (C5-PLP complex) in 2-3 weeks.

Structure Determination by X-ray Crystallography-- The diffraction data sets were collected to 2.4-Å resolution (2.7 Å for the C4-PLP complex) with an R-AXIS IIc imaging plate detector (Rigaku), mounted on an RU-200 rotating anode generator (Rigaku), which was operated at 40 kV and 100 mA with monochromatized CuK radiation at room temperature. All of the data were processed and scaled using the programs DENZO and SCALEPACK (20). The conditions for data collection are summarized in Table I. The structures were determined by molecular replacement methods using the structure of the PLP form (PDB code 1ARS) or the PMP form (PDB code 1AMQ) as the starting model. Model refinement was performed by the CCP4 program suite (21) version 3.51, X-PLOR (22) version 3.851, and program O (23) version 6.2.2.

Kinetic Analysis-- The half-transamination reactions of the PLP form of the enzymes were followed spectrophotometrically at 360 nm as described previously (11). The reaction conditions used were 50 mM HEPES, 100 mM KCl, and 10 µM EDTA, pH 8.0, at 25 °C.

The kinetic parameters were determined using Reaction 2 and Equation 1 as follows,

(Eq. 1)

where E represents the enzyme, S the substrate, E·S the enzyme-substrate complex, ES the transition state, P the product, K_d the dissociation constant of E·S to E + S, k_max the maximum rate constant of the conversion of E·S into E + P, and k_app the apparent rate constant at a given substrate concentration. When the reaction condition was [S] K_d and the k_app value was directly proportional to the substrate concentration, Equation 2, instead of Equation 1, was used to determine the catalytic efficiency, k_max/K_d (11), as follows.

(Eq. 2)

The free energy difference between E + S and ES (G_T) was calculated from Equation 3 as follows,

(Eq. 3)

where R represents the gas constant, T the absolute temperature, k_B the Boltzmann constant, and h the Planck constant (11).

Introduction of Cysteine Residue by Site-directed Mutagenesis-- The 5,5-dithiobis(2-nitrobenzoic acid) (DTNB)-titratable syncatalytic Cys³⁹⁰ (6, 24) was introduced into E. coli AspAT by three-primer polymerase chain reaction (PCR)-available site-directed mutagenesis. In the first PCR, the 192-bp fragment containing the mutation site was amplified as follows. The plasmid pKDHE19 (25), which carries the aspC gene, was denatured at 98 °C for 2 min in the PCR mixture (with primers 5'-CTTCTGGTCGCGTTAACGTGTGCGGGATGACACC-3' and 5'-GACGTTGTAAAACGACGGCCAG-3') without DNA polymerase. After the addition of KOD DNA polymerase (TOYOBO), 25 PCR cycles of 15 s at 98 °C, 5 s at 65 °C, and 30 s at 74 °C were performed. In the second PCR, the above amplified 192-bp DNA fragment and the oligonucleotide 5'-AATGAAACCACCAAACTTTACCTAGGCATTGACGGCATC-3' were used as PCR primers, and the 1129-bp DNA fragment was amplified by the same method as described above. This amplified fragment carrying the mutation site was cut by restriction endonucleases EcoRI and BstPI and was replaced by the corresponding fragment of the wild type aspC gene in the pKDHE19 plasmid (25). The DNA sequence of the resultant plasmid was analyzed using an ABI PRISM 377 DNA sequencer (Perkin-Elmer). No mutation was observed except for the designed mutation site.

Titration of SH Groups of the Enzymes-- We monitored the reaction of the SH groups of the enzymes with DTNB at 412 nm and 25 °C by a method similar to that described previously (26). The enzyme concentration used was about 0.5 mg ml¹. The buffer used contained 50 mM HEPES, 100 mM KCl, and 10 µM EDTA at pH 8.0.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Overall Structure of AspAT Complexes-- In order to reveal the binding mode of the series of aliphatic amino acids (Cn substrates; CH₃(CH₂)_n3C₍₎H(NH₃⁺)COO, n = 3-7), the substrate analogs covalently bound to PLP (Cn-PLP, n = 3-6) were synthesized. The enzyme-Cn-PLP complexes (Cn-PLP complexes), prepared as described under "Experimental Procedures" were crystallized, and their three-dimensional structures were determined (Table I and Figs. 1-3).

View larger version (44K): [in this window] [in a new window]   Fig. 1.   The domain movement of AspAT upon binding of hydrophobic substrates. a, AspAT is a dimer with identical subunits I (colored pink or green) and II (colored gray). The large domain of subunit I complexed with C3-PLP or C4-PLP (pink) was superimposed onto that complexed with C5-PLP or C6-PLP (green). Although the complexes were crystallized under the same conditions, the C3-PLP or C4-PLP complex adopted the open-like form, and the C5-PLP or C6-PLP complex adopted the closed-like form. The addition of only one CH2 group to the C4-PLP complex induced domain closure in the C5-PLP complex. This figure was produced using WebLab ViewerLite (MSI). b, the root mean square deviations of the C atoms in the small domain (residues 5-47 and 326-405) relative to the structure in the absence (PDB code 1ARS (8)) or presence (PDB code 1ART (8)) of substrate were normalized and are represented by a color gradient. Red indicates that the atomic coordinates of a C atom in the Cn-PLP complex are near those of the unliganded open form (PDB code 1ARS), and green indicates that the atomic coordinates are near those of the closed form complexed with an acidic substrate analog, 2-methyl-aspartate (PDB code 1ART). White indicates that an atom is more than 2 Å from either the open or closed form.

Fig. 1a shows the overall structure of the Cn-PLP complexes. AspAT is a dimer with two identical subunits. Each subunit is composed of a large domain (amino acid residues 48-325) and a small domain (residues 5-47 and 326-405) (Refs. 6, 8, and 9; see subunit I in Fig. 1a). In Fig. 1a, the large domains of all of the Cn-PLP complexes are superimposed. AspATs exhibit significant domain movement on substrate binding (6, 8, 9). The small domain of E. coli AspAT rotates by 5-6° to form the closed form of the enzyme (8, 9), and this domain closure changes the active site environment by expelling bulk water molecules from the active site. The positions of the C atoms in the small domains of the C3-PLP and C4-PLP complexes (shown in pink) were almost identical to those of the "open form" without a bound substrate (PDB code 1ARS) (8). In contrast to these analogs, the positions for the C5-PLP and C6-PLP complexes (green) were very close to those for the "closed form" complexed with 2-methylaspartate (PDB code 1ART) (8), which is an acidic substrate analog with high affinity. The root mean square deviations relative to C6-PLP were 0.980, 0.865, and 0.250 Å for the C3-PLP, C4-PLP, and C5-PLP complexes, respectively.

The root mean square deviations of each C atom between the open form (PDB code 1ARS) and the closed form (PDB code 1ART) (8) were normalized and are shown by the color gradient in Fig. 1b.

These observations indicate that the enzyme conformations differed markedly between the C4-PLP complex (open form) and the C5-PLP complex (closed form).

Conformational Differences of -Helix 1 and Its Nearby Loop in the Cn-PLP Complexes-- Fig. 2 shows the conformational differences among the Cn-PLP complexes. The hydrophobic substrate analogs were bound near Ile¹⁷, Leu¹⁸, and Ile³⁷. Ile¹⁷ and Leu¹⁸ are located at the N-terminal side of -helix 1, and Ile³⁷ is in the loop region adjacent to -helix 1. With these interactions, -helix 1 will move toward the active site on binding of C5-PLP (or C6-PLP). Since Ile¹⁷ and Leu¹⁸ are in -helix 1, there was only a slight change of the 1 angle (the rotation around the C-C bond) of the side chain (PDB codes 1C9C, 1CQ6, 1CQ7, and 1CQ8) (data not shown). These residues in the substrate-free holoenzyme and in the C3-PLP and C4-PLP complexes had similar B-factors. These B-factors were higher, suggesting greater flexibility, than those in the C5-PLP and C6-PLP complexes and in the 2-methylaspartate complexes (PDB codes 1C9C, 1CQ6, 1CQ7, and 1CQ8) (data not shown).

View larger version (39K): [in this window] [in a new window]   Fig. 2.   The mechanisms of domain movement. The -helix 1 (Ile17-Arg25) and the loop region (Ala26-Ile37) of the C3-PLP or C4-PLP complex in the open form (these residues are colored pink and the others gray) change to the closed form on binding of C5-PLP or C6-PLP (colored green). Ile17 and Ile37 are anchored at the active site by hydrophobic interactions with the substrate. Leu18 is located at the N-terminal end of -helix 1 and pulls the helix into the active pocket to shield the active site from the solvent. This tugging of residues from Ile17 to Ile37 changes the enzyme from the open to the closed form.

In contrast, Ile³⁷ is situated in the flexible region of the consensus sequence of AspATs (Gly³⁶-X³⁷-Gly³⁸, where X is isoleucine in E. coli AspAT). When C5-PLP or C6-PLP was bound to E. coli AspAT, this flexible loop region moved into the active site with a marked change in main chain conformation. The side chain of Ile³⁷ largely moved from the solvent region into the active site (Fig. 2). The 1 angles of the side chain of Ile³⁷ for the C3-PLP, C4-PLP, C5-PLP, and C6-PLP complexes were 43, 28, 191, and 33°, respectively. These conformational changes are enabled by the presence of glycine on each side of Ile³⁷.

Movement of Arg²⁹² in the Active Site-- Fig. 3, a and b, shows the electron density maps of the active sites for the C4-PLP and C5-PLP complexes, respectively. The C3-PLP to C6-PLP complexes are superimposed in Fig. 3c. When C3-PLP or C4-PLP was bound to the enzyme (Fig. 3, a and c), the hydrophilic Arg²⁹²* was situated outside the cleft. This Arg²⁹²* is known to form salt bridges and hydrogen bonds with the -carboxyl group of bound aspartate or glutamate substrates (6, 8, 9, 27, 28)

View larger version (47K): [in this window] [in a new window]   Fig. 3.   Stereoviews of AspAT bound to substrate analogs. Only Arg292* and the substrate analogs are indicated for clarity. a, the substrate analog is C4-PLP (the moiety of coenzyme is labeled as PLP and that of substrate as C4). The active-site conformation of the C4-PLP complex is almost identical to that of substrate-free AspAT (8, 9). b, the substrate analog is C5-PLP. Each substrate analog and the side chain of Arg292* are superimposed onto the (Fo  Fc) omit map contoured at 2.5  (a) and 3  (b) with these atoms omitted. c, the structure of the C3-PLP complex (magenta), C4-PLP complex (yellow), C5-PLP complex (cyan), and C6-PLP complex (green). Note that the side chain orientation of Arg292* changes drastically depending on substrate. These figures were produced using the program O (23).

Despite the absence of a negatively charged side chain in the substrate, the positively charged Arg²⁹²* moved into the active site when hydrophobic C5-PLP or C6-PLP was bound (Fig. 3, b and c). At this time, the enzyme formed the closed form shown in Fig. 1. Molecular dynamic simulation was performed according to the method of Kasper et al. (29). The simulation could explain the domain movement of the molecule but not this side chain movement of Arg²⁹²* (data not shown).

Comparison between Crystal and Solution Structures-- The reactivity of Cys³⁹⁰ against DTNB in pig cytoplasmic AspAT is increased during catalysis, when AspAT takes the closed form (6, 24). The Cys³⁹⁰ residue, called "syncatalytic cysteine," is in the small domain and is far from the catalytic center (~15 Å) but situated at the interdomain boundary. In order to confirm that the difference in conformation among the Cn-PLP complexes is not due to differences in crystal packing, Cys³⁹⁰ was introduced into E. coli AspAT, and the conformations of the enzyme complexes were monitored in solution by SH titration using DTNB (Fig. 4).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Time dependence of the titration of SH groups with DTNB. The titration curves for the C3-PLP and C4-PLP complexes, with the open form (Figs. 1-3), are distinct from the curves for the C5-PLP and C6-PLP complexes, with the closed form (Figs. 1-3). The reaction of SH groups of the enzyme with DTNB was monitored spectrophotometrically at 412 nm, pH 8.0, 25 °C.

The reaction rates for the C3-PLP and C4-PLP complexes were identical. The rates for the C5-PLP and C6-PLP complexes were also identical and were faster than those for the C3-PLP and C4-PLP complexes. These results coincided with the crystallographic results showing that the C3-PLP and C4-PLP complexes are in the open form and that the C5-PLP and C6-PLP complexes are in the closed form (Figs. 1-3).

Kinetic Background and Its Interpretation-- The reaction kinetics of AspAT (Fig. 5, open circles) were studied with Cn substrates (12, 13). When the free energy difference (G_T) between the transition state (ES) and the unbound enzyme plus the substrate (E + S) (Fig. 5, inset) was plotted against the total number of carbon atoms in the substrate, a linear correlation was observed for substrates from C5 to C7 (Fig. 5). This linear relationship suggests a uniform hydrophobic environment of the substrate-binding site. A similar uniform environment has been suggested for horseradish peroxidase (14), B. amyloliquefaciens subtilisin (15), P. denitrificans AroAT (16), and bovine -chymotrypsin (17). In all of the above cases, the substrate-binding pockets consist of different kinds of atoms. The linear kinetic relationship of these enzymes will be accomplished by fluctuation of the active site. Although the detailed reasons for this linear relationship remain to be elucidated, it was found to be useful for estimating the energy required for domain movement, as described under "Discussion."

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Correlation between GT and the number of carbon atoms in the substrate for aspartate aminotransferase () and aromatic amino acid aminotransferase (). The substrates (Cn substrates) are a series of aliphatic amino acids with linear side chains: CH3(CH2)n3C()H(NH3+)COO. The free energy difference (GT) between the unbound enzyme plus substrate (E + S) and the transition state (ES) (see Fig. 5, inset) was calculated using the equation GT = RT(ln(kBT/h)  ln(kmax/Kd)), where R is the gas constant, T the absolute temperature, kB the Boltzmann constant, and h the Planck constant. The buffer solution contained 50 mM HEPES and 100 mM KCl, pH 8.0, at 25 °C. For each enzyme, the solid line fitted to the data for C5, C6, and C7 was extrapolated to C2 (dotted line). This line represents the GT value for the closed form of the enzyme. The solid line fitted to the data for C3 and C4 represents the GT value for the open form of the enzyme (see "Discussion" for details).

For AspAT, the maximum value of 18.2 kcal mol¹ was observed with C4. The change of slope around C4 reflects the hydrophobicity of the active site. This conversion of the active site environment could be explained by the domain movement of the enzyme molecule, which is accompanied by expulsion of bulk water molecules from the active site (see below).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Many enzymes undergo a large conformational change that serves to expel bulk solvent from the active site and to properly position functional groups for catalysis. This movement, in response to substrate binding at the active site, was termed "induced fit" by Koshland and implies that multiconformational states of enzymes are in equilibrium with one another and are easily perturbed by ligands (1). Many experiments with a number of proteins have revealed domain movement in aqueous solution by using substrate analogs or have directly revealed movement by x-ray crystallography; examples are myosin (30), phosphoglycerate kinase (31), tyrosine kinase (32), and citrate synthase (33). Since the unbound enzyme molecule adopts multiple conformations, the induced-fit mechanism mediates against catalysis by increasing K_m without a corresponding increase in k_cat, compared with all of the molecular species of enzyme being in the active form in the absence of substrate. Although the K_m value of the enzyme is increased, the conformational fluctuation confers an advantage for the association and dissociation steps of enzymes with substrates. In this study, we determined the three-dimensional structure of AspAT complexed with a series of Cn-PLPs. Domain closure was observed for longer (C  5) hydrophobic substrate analogs (Figs. 1-3). These structural and kinetic phenomena enabled us to estimate the free energy required for domain movement of AspAT, as described below.

Mechanism of Domain Movement-- AspATs from E. coli and many vertebrates have a flexible loop (in the case of E. coli AspAT, Gly³⁶-Ile³⁷-Gly³⁸) and an -helix 1 near the N terminus of the protein (Refs. 6-9; Figs. 1-3). The residues Ile³⁷ in the flexible loop and Ile¹⁷ and Ile¹⁸ in -helix 1 are very important for domain movement.

In the case of catalysis of a series of aliphatic substrates (Figs. 1-3), Ile³⁷ in the flexible loop and Ile¹⁷ and Leu¹⁸ in -helix 1 interact with longer substrates (Cn, n  5) and recognize the substrate as a hydrophobic plate. The interaction of these residues with the bound substrate will pull -helix 1 to cover the active-site entrance and will trigger domain movement.

This is also the case for acidic substrates (aspartate, glutamate, or their keto acids). The negatively charged carboxyl group of the substrate is neutralized by the positively charged Arg²⁹²*. These neutralized groups will also be recognized as a hydrophobic plate by Ile³⁷, Ile¹⁷, and Ile¹⁸, as described above.

To confirm the validity of this hypothesis, we introduced an I37G mutation into E. coli AspAT. This mutant could not undergo any domain movement on substrate binding with either acidic or hydrophobic substrates (data not shown).

Estimation of Energy Required for Domain Movement-- Previous crystallographic studies have indicated that AspAT takes either an open or a closed form but never takes an intermediate state (6-9, 27, 28). Upon binding of an acidic substrate, the enzyme molecule undergoes a large domain movement. Arg²⁹²* moves from the outside to the inside of the active site and recognizes the -carboxyl group of the substrate (6-9, 27, 28).

For hydrophobic substrates, Arg²⁹²* was expected to remain outside the active site (12, 18, 19). In this study, however, movement of the side chain of Arg²⁹²* and domain closure (Figs. 1-3) were observed for longer hydrophobic substrate analogs (C  5).

From the following reasonable assumptions, we could estimate the free energy required for domain movement (Fig. 6). These assumptions were supported by comparison with the homologous enzyme E. coli AroAT, as discussed under "Confirmation from a Homologous Enzyme."

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Thermodynamic cycle. a, summary of thermodynamic cycles (shown in b) for the reactions of AspAT with a series of substrates from C3 to C7. The number n represents the number of carbon atoms in the substrate. GopenE + S represents the sum of the free energy of an enzyme with unbound open form (GopenE) and that of a substrate (GS). GclosedE + S represents the sum of the free energy of an enzyme with unbound closed form (GopenE) and that of a substrate (GS). GopenES and GclosedES are the transition states of the open and closed forms, respectively, of the complex. b, the thermodynamic cycle for each substrate was calculated from the data points in Fig. 5. From these data, the energy required for domain closure of AspAT was estimated to be 1.9 kcal mol1 (see "Discussion" for details). c, summary of thermodynamic cycles for AroAT. This diagram was obtained by a method similar to that described for AspAT (see Figs. 5 and 6, a and b).

Assumption 1 is that the energy contribution of each constituent group of the substrate CH₃-, -CH₂-, and -C₍₎H (NH₃⁺) COO groups was additive. Assumption 2 is that the energy contribution of the group -C₍₎H(NH₃⁺)COO was identical for all of the substrates studied. This assumption is based on the following results: (a) the crystallographic analyses (Figs. 1-3); and (b) the linearity of G_T versus the total number of carbon atoms in the substrate (Fig. 5). Assumption 3 is that the hydrophobic environment around the substrate-binding pocket is uniform, and the energy contribution of the methylene and methyl groups is identical. This assumption is supported by the linearity of G_T from C5 to C7 substrates in Fig. 5.

From these assumptions, we can generate thermodynamic cycles of AspAT for Cn substrates, and we can estimate the free energy differences among molecular species. G_open^E + S and G_closed^E + S in Fig. 6 represent the Gibbs free energy for the open and closed forms, respectively, where G_open^E + S or G_closed^E + S is the sum of free energies of the unbound enzyme (G^E) and unbound substrate (G^S). G_open^ES and G_closed^ES represent the open and closed forms, respectively, complexed with Cn substrate in the transition state. Whether the enzyme will be closed depends on the energy difference between G_open^ES and G_closed^ES.

From the crystallographic results in Fig. 1, we determined that the conformations of the C5-PLP and C6-PLP complexes were in the closed form and the conformations of the C3-PLP and C4-PLP complexes were in the open form. Therefore, the slope from C5 to C7 (0.65 kcal mol¹) in Fig. 5 represents the energy obtained from the hydrophobic interaction between the substrate and the substrate-binding pocket of the closed form enzyme. On the other hand, the slope from C3 to C4 (+0.3 kcal mol¹ CH₂¹) represents the energy loss due to contact between the hydrophobic substrate and the hydrophilic surface of the active pocket filled with water molecules. In Fig. 5, the extrapolated red line for AspAT (dotted line) coincides at C = 2 with G_T = 19.5 kcal mol¹. This G_T corresponds to the energy difference between G_open^E + S and G_closed^ES in Fig. 6a. The slope of this extrapolated line (0.65 kcal mol¹ CH₂¹ for AspAT) corresponds to the hydrophobicity of the substrate-binding pocket for the closed form of AspAT.

The C3-PLP or C4-PLP complex is in the open form (Figs. 1 and 3). If we extrapolate the G_T values for the open form to C2, we obtain a value for G_T of 17.6 kcal mol¹ (Fig. 5, red solid line). This G_T corresponds to the energy difference between G_open^E + S and G_open^ES in Fig. 6a.

As described above, the G_T value for the closed form (from C5 to C7) corresponds to G_closed^ES  G_open^E + S, and that for the open form (from C3 to C4) corresponds to G_open^ES  G_open^E + S. Both the x-ray crystallographic and kinetic results showed that the conformation around -C₍₎H(NH₃⁺)COO was identical between the open form, which corresponds to G_open^ES, and the closed form, which corresponds to G_closed^ES. Based on these results, we assumed that the energy changes from the open (G_open) to closed (G_closed) forms are similar between bound (G^ES) and unbound (G^E + S) enzymes. Since the free energy of unbound substrate (G^S) is included in both G_closed^E + S and G_open^E + S in Fig. 6, this energy difference corresponds to the energy for conformational change of an unbound enzyme from the open (G_open^E) to closed (G_closed^E) forms. The free energy change thus estimated for domain movement (G_closed^E  G_open^E) is 1.9 kcal mol¹ for aspartate aminotransferase.

Confirmation from a Homologous Enzyme-- E. coli AroAT (EC 2.6.1.57) is an isozyme of AspAT and catalyzes transamination reactions with substrate specificity different from that of AspAT (34-37). When we applied our analysis to the kinetic data of AroAT (Fig. 6c), we could confirm the correctness of our analysis, and we obtained the same energy of 1.9 kcal mol¹ for domain closure as that found for AspAT as follows.

The amino acid sequence of AroAT is 43% identical to that of AspAT, and both enzymes consist of two identical subunits of 44 kDa (34, 35). The pK_a values of the PLP -Lys²⁵⁸ Schiff base are also very similar between AspAT and AroAT (36), suggesting the similarity of the microenvironment of their active pockets. The similarity of the conformations of these enzymes was confirmed by x-ray crystallography (8, 37). The root mean square deviation between the unliganded form of AspAT (PDB code 1ASN) and AroAT (PDB code 3TAT) is 1.28 Å for the overall structure (37). When we applied the least-squares fit to the C atoms of the large domain residues of the unliganded forms of AspAT (PDB code 1ARS) and AroAT (PDB code 3TAT), the deviation was 1.85 Å for the overall structure. We also applied the least-squares fit to the 25 C atoms in the active pocket (the residue numbers are 36, 37, 70*, 107, 108, 109, 110, 140, 141, 142, 143, 189, 194, 222, 224, 225, 256, 257, 258, 263, 266, 292*, 296*, 297*, and 386, where an asterisk represents a residue from the other subunit). The root mean square deviation of these residues was 0.76 Å. The active site residues around the group -C₍₎H(NH₃⁺)COO of the substrate (Ile³⁷, Asn¹⁹⁴, Tyr²²⁵, Tyr²⁵⁸, Met³⁵⁹, Phe³⁶⁰, and Arg³⁸⁶) are completely conserved (34, 35), and the root mean square deviation was 0.56 Å for the C atoms. These findings suggest that these two enzymes have similar catalytic mechanisms with similar three-dimensional structures.

On the basis of these structural similarities between AspAT and AroAT, we applied the same kinetic analysis to AroAT (Figs. 5 and 6c). The extrapolated blue dotted line for AroAT in Fig. 5 coincided with the extrapolated red line for AspAT at C = 2 with G_T = 19.5 kcal mol¹. The coincidence between the two enzymes suggests that the energy contribution of the -C₍₎H(NH₃⁺)COO group in the "closed form" of the enzyme is 19.5 kcal mol¹ and that the environment around the binding site of -C₍₎H(NH₃⁺)COO is identical in AspAT and AroAT. The slope of this extrapolated line, which corresponds to the hydrophobicity of the substrate-binding pocket for the closed form, was 1.7 kcal mol¹ CH₂¹ for AroAT. This value was smaller than that of 0.65 kcal mol¹ CH₂¹ for AspAT.

On the other hand, if we extrapolate the G_T values for AroAT between C3 and C4 to C2, we obtain a value for G_T of 17.6 kcal mol¹ (Fig. 5, solid blue line). This line for AroAT coincides with the line for AspAT at C2. This coincidence between the two "open form" aminotransferases suggests that the environment around the binding site of -C₍₎H(NH₃⁺)COO is identical in AspAT and AroAT. In summary, these observations indicate that the active pocket environment for the open and closed forms is almost identical in AspAT and AroAT except for the hydrophobicity of the substrate-binding pocket.

The free energy required for domain movement of AroAT (Fig. 6c), obtained by the same method used for AspAT, was estimated to be 1.9 kcal mol¹. The fact that we obtained the same energy for domain movement in these two aminotransferases supports the correctness of our analysis.

Comparison with Other Results Related to Domain Movement-- A number of previous studies have also attempted to analyze domain movement. Pfister et al. (38) measured the hydrogen-deuterium exchange of pig cytosolic AspAT and estimated the energy difference between open and closed forms to be 2-3 kcal mol¹. Since the hydrogen-deuterium exchange reflects not only domain movement but also the fluctuation of the domain itself, it is difficult to isolate the contribution of the domain movement from the data. Rhee et al. (39) estimated the free energy for burying the hydrophobic plug (Pro¹⁴-Phe¹⁸) of pig cytosolic aspartate aminotransferase into the active site to be 5.3 kcal mol¹ and concluded that the energy is more than enough to drive the conformational change. Radmacher et al. (4) measured height fluctuations on top of the lysozyme molecule with an atomic force microscope and suggested that the energy for this fluctuation is about 3 kcal mol¹, although they recommended caution when interpreting whether or not the height changes correspond directly to a change in the diameter of the enzyme. Karplus and McCammon (3) used molecular dynamic simulation to predict domain movement, but they did not quantitatively estimate the energy required for the process. The data obtained from single-molecule measurements of myosin movement along an actin filament (5) might also correspond to the energy of domain movement.

Conclusion-- In this paper, we tried to estimate from thermodynamic analysis the energy required for domain movement. The estimated energy of 2 kcal mol¹ is only 3 times as large as the energy of thermal fluctuations. This energy for domain closure of about 2 kcal mol¹ predicts that the enzyme fluctuates between the open and closed forms with a molar ratio of about 30:1 (RT ln(30/1) = 2 kcal mol¹). By using this domain fluctuation, an enzyme searches for its substrate; tight binding to the substrate then follows.

	FOOTNOTES

^* This work was supported in part by Grants-in-aid for Scientific Research 11878118 and 11169224 from the Ministry of Education, Science, Sports, and Culture of Japan and by Japan Society for the Promotion of Science ("Research for the Future" Program) Grant 96L00506.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and the structure factors (code 1C9C, 1CQ6, 1CQ7, and 1CQ8) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

^** To whom correspondence should be addressed: Dept. of Biology, Graduate School of Science, Osaka University, 1-1 Machikaneyama-cho, Toyonaka, Osaka 560-0043, Japan. Tel.: 81-6-6850-5433; Fax: 81-6-6850-5442; E-mail: kuramitu@bio.sci.osaka-u.ac.jp.

² An asterisk indicates that the residue is supplied by the other subunit of the dimer.

	ABBREVIATIONS

The abbreviations used are: AspAT, aspartate aminotransferase; AroAT, aromatic amino acid aminotransferase; PLP, pyridoxal 5'-phosphate; PMP, pyridoxamine 5'-phosphate; PDB, Protein Data Bank; Cn substrates, aliphatic amino acids with linear side chains [CH₃(CH₂)_n3 C₍₎H(NH₃⁺)COO]; Cn-PLP, aliphatic amino acid of n carbon atoms covalently bound to PLP by reduction with NaBH₄; Cn-PLP complex, E. coli AspAT complexed with Cn-PLP; PCR, polymerase chain reaction; DTNB, 5,5-dithiobis(2-nitrobenzoic acid).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg