|
|
||||||||
Department of Chemistry and Biochemistry, University of California, Santa Barbara, California 93106, USA
Reprint requests to: Thomas C. Bruice, Department of Chemistry and Biochemistry, University of California, Santa Barbara, CA 93106, USA; e-mail: tcbruice{at}chem.ucsb.edu; fax: (805) 893-2229.
| Abstract |
|---|
|
|
|---|
PQQH occurs with Glu 171CO2 and His 144Im as the base species in MDH and sGDH, respectively. The general-base-catalyzed hydroxyl proton abstraction from substrate concerted with hydride transfer to the C5 of PQQ is assisted by hydrogen-bonding to the C5 = O by Wat1 and Arg 324 in MDH and by Wat89 and Arg 228 in sGDH. Asp 297COOH would act as a proton donor for the reaction PQQH
PQQH, if formed by transfer of the proton from Glu 171COOH to Asp 297CO2 in MDH. For PQQH
PQQH2, migration of H5 to the C4 oxygen may be assisted by a weak base like water (either by crystal water Wat97 or bulk solvent, hydrogen-bonded to Glu 171CO2 in MDH and by Wat89 in sGDH). Keywords: molecular dynamics; PQQ; quinoproteins; methanol dehydrogenase; soluble glucose dehydrogenase; hydride transfer mechanism
Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.04673404.
| Introduction |
|---|
|
|
|---|
MD technique uses a force field (parameterized on a test set of molecules) to describe the atomic interactions of a system in space and time by evaluating the classical Newtonian equation of motion. The method involves direct incorporation of electrons and can be used for the study of large and complex systems, unlike quantum mechanical calculations. Enzyme cofactors and ligands or substrates often require additional parameterization. The high-resolution crystal structure of the enzymesubstrate or enzymeinhibitor complex is used as the starting coordinates for exploration of the structure and enzyme mechanism. Appropriate MD procedures allow the generation of either enzymesubstrate or enzymeintermediate complexes with hydrogens (ionizable functions may be set to be ionized or protonated) in explicit solvent. Explicit solvation mimics the experimental conditions (at an appropriate pH and ionic strength of biomolecules). A realistic treatment of the system boundaries is carried out either by periodic boundary conditions (an elegant procedure, but computationally expensive; Brooks et al. 1985) or by enclosing the system of interest in a sphere, with restraints and stochastic forces acting at the boundary. The latter procedure is called stochastic boundary molecular dynamics (SBMD; Brooks and Karplus 1989), which is useful in the study of large multimeric solvated systems while probing a specified region of interest. The potential of SBMD to elucidate otherwise inaccessible details about enzyme mechanisms is illustrated by recent studies on quinoprotein dehydrogenases from our group (Reddy and Bruice 2003, 2004; Reddy et al. 2003), which form the focus of the rest of this review.
Quinoproteins are a class of redox enzymes bearing orthoquinone cofactors that function as oxidants in respiration and energy transduction (Duine and Frank Jzn 1981; Duine 1991; Anthony 1996; Davidson 2001). Quinoproteins occur in the periplasm of Gram-negative bacteria and as amine oxidases in mammalian plasma. They relay electrons from the oxidation of substrates to 1e electron-accepting proteins such as cytochromes, cupredoxins (amicyanin or azurin), and ubiquinone. Quinoproteins use different cofactors either bound noncovalently (or freely dissociable species) like pyrroloquinoline quinone (PQQ; Salisbury et al. 1979) or derived from amino acids in the protein backbone of the enzyme, such as topaquinone (TPQ; Janes et al. 1990), lysine tyrosylquinone (LTQ; Wang et al. 1996), tryptophan tryptophylquinone (TTQ; McIntire et al. 1991), and cysteine tryptophylquinone (CTQ; Datta et al. 2000). The unique properties of quinoproteins have led to biotechnological applications such as biosensors, biotransformation of specific products, and bioremediation (Matsushita et al. 2002).
PQQ-containing dehydrogenases, the bacterial enzymes, constitute the largest quinoprotein subclass (Anthony and Ghosh 1998; Oubrie and Dijkstra 2000; Anthony 2001; Davidson 2001; Matsushita et al. 2002). The enzymes occur either in soluble or in membrane-associated forms. PQQ is synthesized exogenously and then associates with an apo-protein to form the active holoenzyme. In the active site of these enzymes, PQQ is complexed with a divalent metal ion. The PQQ enzymes are broadly classified as alcohol dehydrogenases (ADH) and glucose dehydrogenases (GDH). ADH, which catalyzes the oxidation of alcohols into the corresponding aldehydes, is classified into three types. Type I comprises methanol dehydrogenase (MDH) and ethanol dehydrogenase. Type II ADH is a soluble quinohemoprotein. Ethanol dehydrogenase and Type II ADH, isolated from different pseudomonad species, oxidize primary and secondary alcohols (Keitel et al. 2000; Chen et al. 2002; Oubrie et al. 2002). Type III ADH, a membrane-bound quinohemoprotein, is found only in acetic acid bacteria (Inoue et al. 1989; Matsushita et al. 1994). GDH enzymes (from Acinetobacter calcoaceticus) are of two types, either bound to membrane, mGDH, or in soluble form, sGDH. The two types of GDH do not have sequence homology (Cleton-Jansen et al. 1988), nor are they kinetically and immunologically similar (Matsushita et al. 1989). mGDH catalyzes the oxidation of aldose sugars (mainly monosaccharides) to the lactones, with the electron acceptor being ubiquinone in the membrane. sGDH has broader substrate specificity than mGDH.
Our focus is on the PQQ enzymes, methanol dehydrogenase (MDH) and soluble glucose dehydrogenase (sGDH). In the course of reaction, the orthoquinone PQQ is reduced to hydroquinone (PQQH2). MDH catalyzes the oxidation of methanol to formaldehyde, a key step in biological metabolism in methylotrophic bacteria that have potential importance to biotechnology (Anthony 1986). The enzyme has a high affinity only for primary alcohols, with a Km value of ~520 µM (Keitel et al. 2000; Anthony 2001). MDH uses cytochrome cL as an electron acceptor to reoxidize the hydroquinone PQQH2 to orthoquinone PQQ in two separate single-electron transfers (Anthony 1992). MDH has also been isolated from autotrophic bacteria, whose electron acceptor is cytochrome c-551 in Paracoccus denitrificans (Anthony 1986; Harris and Davidson 1993a; Xia et al. 2003). The experimental studies of MDH are complicated by enzyme endogenous activity (i.e., the enzyme catalyzes the reduction of an electron acceptor in the absence of any added substrate, and further, if allowed to react in this manner, the enzyme becomes inactivated). This substrate-independent activity is suppressed by cyanide, which also protects the enzyme against inactivation (Duine and Frank 1980; Harris and Davidson 1993b).
A few available X-ray structures of MDH PQQ complexes are 1.94 Å (1H4I
[PDB]
) from Methylobacterium extorquens (Ghosh et al. 1995); 2.4 Å (4AAH
[PDB]
) from Methylophilus W3A1 (Xia et al. 1996); 1.9 Å (1G72
[PDB]
) from Methylophilus methylotrophus (Zheng et al. 2001); and 2.5 Å (1LRW
[PDB]
) from Paracoccus denitrificans (Xia et al. 2003). These enzyme structures from different bacterial sources have nearly the same active sites and are also similar at the sequence level. Attempts to crystallize the substrate methanol in the active site of MDH have not yet been successful. Although a crystal structure reported the presence of methanol (Xia et al. 1999), subsequent refinement of the structure showed no methanol in the active site (Zheng et al. 2001). MDH is a heterotetramer with two heavy and two light subunits with a total mass of 150 kDa. The catalytic chemistry is located in each heavy subunit of MDH with similar features of the active site. The light subunit of MDH has an extended structure and along with the heavy subunit forms a dimer. The subunits cannot be reversibly dissociated, and the function of the light subunit is unknown. The heavy subunit of MDH has a
-propeller fold, which is characteristic of eight four-stranded antiparallel
-sheets.
sGDH oxidizes a wide range of pentose and hexose sugars, mono- as well as disaccharides, to their corresponding lactones (Matsushita et al. 1989). The physiological electron acceptor for reoxidation of PQQH2 to PQQ in sGDH is unknown. Cytochrome b562 (Dokter et al. 1988), in addition to artificial electron acceptors such as N-methylphenazonium methyl sulfate (Olsthoorn and Duine 1996), serves well as an oxidant. sGDH is presently used in the accurate monitoring of blood glucose using diabetic control test strips (Ye et al. 1993; Yamazaki et al. 2000). sGDH is a homodimer (Dokter et al. 1986; Geiger and Gorisch 1986) of 100 kDa and basic in nature (pI = 9.5). The four sGDH X-ray structures isolated from A. calcoaceticus are 1.72 Å (1QBI
[PDB]
) apo enzyme (Oubrie et al. 1999b); 1.5 Å (1CRU
[PDB]
) sGDH bound to PQQ and methylhydrazine inhibitor (Oubrie et al. 1999a); and 2.2 Å (1C9U
[PDB]
) sGDH PQQ and 1.9 Å(1CQ1
[PDB]
) sGDH PQQH2 glucose structures (Oubrie et al. 1999c). Each subunit of sGDH binds PQQ, glucose, and three Ca2+ ions, with one of the Ca2+ ions located in the active site of the enzyme. The other four Ca2+ ions of sGDH have been implicated to have a role in the stabilization of the two subunits (Olsthoorn et al. 1997; Olsthoorn and Duine 1998b) and are absent in MDH. However, the Ca2+ ion in the active site is conserved in MDH and sGDH. Each subunit of sGDH is described by the
-propeller fold structure (Oubrie et al. 1999b), similar to MDH. But the fold in sGDH is made of six four-stranded antiparallel
-sheets. Although sGDH and MDH have certain similarities in their active sites, the amino acid sequence and the three-dimensional structures of the two enzymes are quite different.
The characteristic
-propellers have been reported in several proteins, with eightfold
-propeller of MDH observed in nitrite reductase (with heme as the prosthetic group; Fülöp et al. 1995), and PQQ enzymes such as ethanol dehydrogenase (Keitel et al. 2000) and quinohemoproteins type II ADH (Chen et al. 2002; Oubrie et al. 2002). The sixfold
-propeller of s-GDH is observed in neuraminidase from influenza virus (Varghese et al. 1983) and bacterial sialidase (Crennell et al. 1993). Besides eightfold and sixfold
-propellers, sevenfold
-propeller structure is characteristic of quinoprotein methylamine dehydrogenase (which contains TTQ cofactor; Vellieux et al. 1989), the
subunit of quinohemoprotein amine dehydrogenase (with CTQ cofactor; Datta et al. 2000), G-proteins (Wall et al. 1995; Lambright et al. 1996), phospholipase D (Stukey and Dixon 1999), copper-containing galactose oxidase (Ito et al. 1994), the human regulator of chromosome condensation (Renault et al. 1998), the integrin
-subunit (Springer 1997), and the N-terminal domain of the clathrin heavy chain (ter Haar et al. 1998). A fourfold
-propeller is described for hemopexin (Faber et al. 1995; Libson et al. 1995) and collagenase-3 structures (Li et al. 1995; Gomis et al. 1996).
On the basis of biochemical and kinetic data, hydride transfer and additionelimination mechanisms have been proposed for MDH (Frank et al. 1988; Anthony 1996; Itoh et al. 1997) and sGDH (Olsthoorn and Duine 1998a; Dewanti and Duine 2000). The hydride transfer mechanism involves a general-base-catalyzed proton abstraction and hydride equivalent transfer from substrate to the quinone carbonyl carbon C5 of PQQ, and subsequent tautomerization of the C5 reduced PQQ to the hydroquinone (Fig. 1A,C
). The additionelimination mechanism involves nucleophilic addition of substrate to the C5 carbon of PQQ assisted by general-base catalysis of proton abstraction from substrate, followed by formation of covalent hemiketal intermediate and subsequent product elimination (Fig. 1B,D
). The rate-limiting step for MDH and sGDH is conversion of the enzymesubstrate complex into the reduced coenzyme and product (Frank et al. 1988; Olsthoorn and Duine 1998a). The large deuterium isotope effect (~6) during the reductive phase of reaction indicates that both mechanisms are probable. The ease of nucleophilic addition to the C5 carbon of isolated PQQ by methanol, urea, ammonia, and amines has encouraged the assumption of formation of a covalent C5 PQQsubstrate complex in the initial step, characteristic of an additionelimination reaction (Dekker et al. 1982; Itoh et al. 1993). A crystal structure report on methylhydrazine bound to sGDH indicated formation of a covalent C5 PQQinhibitor complex (Oubrie et al. 1999a). However, the electron density obtained for the X-ray structure of C5-reduced PQQ bound to MDH (Zheng et al. 2001) and the
-anomer configuration of glucose in sGDH complex (Oubrie et al. 1999c) conforms to the hydride transfer mechanism (Fig. 1A,C
). Theoretical studies have reported that the hydride transfer occurs from methanol to the O4 oxygen of PQQ (Zheng and Bruice 1997) and in another, from methanol to the O5 of PQQ (Jongejan et al. 2001). In view of the above reports and to understand better the role of enzyme active-site residues and catalytic mechanisms, we discuss 3-nsec SBMD studies by comparing MDH PQQ methanol, MDH PQQH, and MDH PQQH with sGDH PQQ glucose, sGDH PQQH glucolactone, and sGDH PQQH structures.
|
| Molecular modeling procedures |
|---|
|
|
|---|
The 1.9 Å sGDH PQQH2 glucose complex with 600 crystal waters, PDB entry 1CQ1 [PDB] (Oubrie et al. 1999c) was chosen as the initial structure for MD studies on sGDH. The homodimer sGDH has 450 amino acid residues in each subunit. The cofactor in the X-ray structure was identified as hydroquinone, PQQH2 with planar C5 and C4 atoms. The 1CQ1 [PDB] coordinates were therefore used to simulate the reactant complex sGDH PQQ glucose. For MD simulations on the sGDH PQQH glucolactone, the generated PQQH and glucolactone were exchanged for PQQ and glucose, respectively, in the X-ray structure. For MD studies on sGDH PQQH, the generated PQQH was used.
The charges of PQQ, PQQH, PQQH, and glucolactone were evaluated using the ab initio procedures, which involve geometry optimization at the HF/631+G(d,p) level and evaluation of electrostatic potential at the MP2/631+G(d,p) level. In each structure the protonation sites of histidine residues were identified appropriately and the hydrogen atoms were generated. Based on the proposed hydride transfer mechanism (Fig. 1A,C
), the carboxyl moiety of Asp 297 in MDH PQQH and the imidazole of His 144 in sGDH PQQH glucolactone structures were protonated. The systems were solvated in an equilibrated TIP3P water sphere of 42 Å radii, using the center of mass of the cofactor as the origin. All the systems were minimized before, and after overlaying the explicit water molecules. Appropriate procedures of stochastic boundary molecular dynamics (Brooks and Karplus 1989) were followed. As per SBMD method, each system was partitioned into reaction (040.0 Å), buffer (41.0142.0 Å), and reservoir (beyond 42.0 Å) regions. Newtonian dynamics treats the reaction region or the localized region of interest, whereas Langevin dynamics was applied to simulate the unwanted degrees of freedom of the buffer region by stochastic and mean forces. Details of the methodologies were published elsewhere (Reddy and Bruice 2003, 2004).
| Comparison of MDH and sGDH structures |
|---|
|
|
|---|
-propeller fold of each heavy subunit. The PQQ is sandwiched between the indole ring of Trp 237 and the vicinal disulfide Cys 103Cys 104 (Fig. 2A
(gauche+ instead of gauche) of Cys 103,
(gauche+ instead of trans or gauche), and
(trans instead of gaucheor gauche+) of Asp 105. The active-site vicinal disulfide of MDH has also been reported in other PQQ enzymes, including ethanol dehydrogenase (Keitel et al. 2000), quinohemoproteins of type II ADH (Chen et al. 2002; Oubrie et al. 2002), and mercuric ion reductase (Schiering et al. 1991). The role of vicinal disulfide in enzyme catalysis is not clear and has been suggested to act as a shield from the bulk solvent at the entrance to the active site (Avezoux et al. 1995). MD investigations have established that the methyl group of methanol is encompassed by a hydrophobic cavity provided by the vicinal disulfide Cys 103Cys 104, Trp 259, Trp 531, and Leu 547 (Fig. 2A
|
Equatorial interactions of the active site
In the plane of the cofactor, various hydrogen-bond interactions occur between the cofactor and enzyme residues. Besides Ca2+ coordination, and other essential interactions that promote catalytic reaction in the active site (discussed in the following sections), there is a network of hydrogen-bonding interactions and salt bridges between the 2, 7, and 9 carboxylate oxygens of PQQH and enzyme residues (Fig. 2C,D
). Such interactions of MDH PQQH involve the C2 carboxyl group with side chains of Trp 467, Glu 55, and Arg 109 and the C9 carboxylate with side chains of Arg 109, Thr 153, and Ser 168 (Fig. 2C
). The O7B oxygen at the C7 of PQQH interacts with Thr 235 and the backbone amide of Gly 169 and Ala 170 of MDH. In sGDH PQQH structure, the C2, C7, and C9 carboxylates of PQQH make ion-pair interactions with Arg 408, Lys 377, and Arg 406, respectively (Fig. 2D
). Most of the nonbonded separations related to these interactions of the MD structures are consistent with corresponding crystal structures of sGDH and MDH (Oubrie et al. 1999c; Zheng et al. 2001).
Active-site Ca2+ coordination
The enzymes MDH and sGDH are inactive in the absence of the Ca2+ ion (Adachi et al. 1990; Richardson and Anthony 1992; Olsthoorn et al. 1997). The Ca2+ ion is not involved in the binding of methanol or glucose. It is likely to play a role in the polarization of the C5 = O5 bond of PQQ along with the active-site Arg residue in MDH and sGDH (Anthony et al. 1994; Zheng and Bruice 1997; Olsthoorn and Duine 1998a). In the X-ray structures of MDH and sGDH, Ca2+ binds to the same cofactor ligands, the C7 carboxylate oxygen, the N6 pyridine nitrogen, and the C5 carbonyl oxygen (Oubrie et al. 1999c; Zheng et al. 2001). The other ligands of Ca2+ are different, provided by the side chains of Glu 171 and Asn 255 of MDH (Fig. 3B
), whereas with sGDH, the ligands are main-chain carbonyl oxygen atoms of Gly 246 and Pro 247 and two water molecules (Wat55 and Wat89; Fig. 3D
).
|
In the sGDH PQQ glucose MD complex, Ca2+ shows hepta-coordination similar to the X-ray structure (Fig. 3D
; Oubrie et al. 1999c). In the sGDH PQQH structure, the coordination to Pro 247 is absent, changing from hepta- to hexa-coordination of Ca2+ (Fig. 3E
). In the sGDH PQQH structure, Ca2+ coordination to the O5 protonated oxygen of PQQH is exchanged for a TIP3P (Jorgensen et al. 1983) water oxygen, which comprises the explicit solvent in which the enzyme is immersed during MD procedures (Fig. 3F
). Absence of Ca2+ coordination to the O5 oxygen of PQQH is observed in the MD structures of MDH and sGDH. The oxyanion O5 of PQQH is a strong base and, as expected, remains coordinated to Ca2+ until protonated (Fig. 3B,E
).
| Hydride transfer mechanism of MDH |
|---|
|
|
|---|
Glu 171CO2 is the only functional base in the active site positioned to be the general base catalyst. One of the carboxylate oxygens of Glu 171CO2 is at a distance 2.79 ± 0.16 Å from the methanol oxygen (Fig. 4A
). The proton abstraction from the hydroxyl of methanol by Glu 171CO2 is in concert with the hydride transfer, as the methanol carbon is at a distance of 3.77 ± 0.29 Å from the C5 carbon of PQQ. The hydrogen-bond interactions of the guanido nitrogens (NH1 and NH2) of Arg 324 with > C5 = O5 (2.81 ± 0.14 Å) and > C4 = O4 (2.93 ± 0.18 Å) oxygens are in position to aid PQQ reduction. Also, the amide side-chain nitrogen of Asn 387 is proximal (3.15 ± 0.25 Å) to the O4 oxygen of PQQ. The crystal water Wat1 forms a strong hydrogen bond with the O5 quinone carbonyl oxygen (2.71 ± 0.14 Å) and Asp 297CO2 (2.75 ± 0.14 Å).
|
Ammonia or methylamine has been used as an activator for MDH (from Hyphomicrobium X) with the phenazine ethosulphate dye-linked assay system (as artificial electron acceptor) at pH optimum 9.0 (Frank et al. 1988). Stopped-flow kinetics studies on MDH indicated that a high concentration of ammonia acts as an inhibitor (Harris and Davidson 1993b). Also, it has been demonstrated that ammonia has little kinetic effect in the presence of physiological cytochrome electron acceptor (Dijkstra et al. 1989). In view of these, MD simulations undertaken on the MDH PQQ methanol structure with NH3 have shown that NH3 diffuses away from the active site, interacting with the solvent at the surface. In any case, NH3 at pH 7.0 would exist as NH4+. In contrast, NH4+ remains in the active site during the simulation of the MDH PQQ methanol NH4+ structure, making hydrogen-bond interactions with the O5 oxygen of PQQ, Asp 297CO2, and Glu 171CO2. However, methanol moves to the entrance of the enzyme active site, near the indole nitrogen of Trp 531 and the Cys 103Cys 104 vicinal disulfide. Possibly, NH3 interacts with MDH at a different position, which does not disturb the environment of the active site (Harris and Davidson 1993b). Work is in progress to study the effect of imino and amino moieties substituted, instead of the quinone carbonyl oxygen O5 of PQQ in the MDH PQQ methanol structure.
The immediate product from methanol oxidation by MDH is formaldehyde. The means of departure of O = CH2 is unknown. It is possible that O = CH2 interacts with one of the many waters in the active-site vicinity, to form the stable hydrate (HO)2CH2.
Step 2: Hydride transfer to C5(= O) of PQQ leads to (H)C5(O) of PQQH. Thus, Glu 171CO2, a general base, was examined to see whether once protonated it acts as a general acid species to protonate the oxyanion O5 of PQQH. MD simulations establish that Glu 171CO2H formed during hydride transfer is not placed in a position to protonate the oxyanion (H)C5(O) moiety of PQQH (Reddy and Bruice 2003). It could be imagined that Glu 171CO2H transfers its proton to Asp 297CO2, which would allow the MDH PQQH complex to be a transient structure. During MD simulations, the average distance from the oxyanion O5 of PQQH and the protonated carboxylate oxygen of Asp 297CO2H is 2.57 ± 0.08 Å (Fig. 4B
), indicating that Asp 297CO2H could act as the required proton donor (Reddy et al. 2003). This feature is not obvious in the crystal structure because of a large separation (4.07 Å) between the carboxyl oxygen of Asp 297 and the O5 oxygen of the C5 reduced PQQ (Zheng et al. 2001). A hydrogen bond between the oxyanion O5 of PQQH and Wat1 (2.91 ± 0.27 Å) is observed. These interactions render the O5 oxygen to move farther from the guanido nitrogen NH1 of Arg 324, in contrast to the observed hydrogen bond (3.03 Å) in the X-ray structure. The guanido nitrogen NH1 of Arg 324 (3.27 ± 0.22 Å) is hydrogen-bonded to Asp 297CO2H. The O4 carbonyl oxygen at C4 of PQQH is involved in bidentate interactions with the guanido side chains of Arg 324 (2.73 ± 0.09 Å and 2.90 ± 0.16 Å), consistent with the crystal structure (Zheng et al. 2001). These interactions are expected to assist rearrangement of PQQH
PQQH.
Step 3: Asp 297CO2 has been suggested (Anthony 1996; Xia et al. 1999) to assist as a general base in the transfer of the H5 hydrogen at tetrahedral C5 of PQQH to provide the hydroquinone PQQH2 (Fig. 1A
). In the MD structure of MDH PQQH, Asp 297CO2 is at a distance of 3.43 ± 0.44 Å, and is not poised to assist in migration of the H5 hydrogen of PQQH. This is because Asp 297CO2 coordinated to Ca2+ is also preoccupied in hydrogen-bond interactions with the O5 protonated oxygen of PQQH (2.98 ± 0.26 Å) and Wat1 (2.64 ± 0.09 Å and 2.87 ± 0.12 Å; Fig. 4C
).
The average distance between the carboxylate oxygen of Glu 171CO2 and H5 at the tetrahedral C5 of PQQH is large (3.68 ± 0.26 Å). So Glu 171CO2 is unlikely to aid in proton isomerization of (H)C5(OH) C4( = O)
C5(OH) = C4(OH) . This reaction is very exergonic because of the gain in aromaticity. Thus, any catalysis of the rearrangement would amount to a solvation of the transferring proton. Interestingly, MD simulations suggest that a Glu 171CO2 ... H2O ... HC5 relay is involved in the solvation of the proton transfer. The water molecule involved can be either Wat97 or TIP3P, which exchanges during MD simulation. This is an interesting feature not noticed in the X-ray structures (Ghosh et al. 1995; Xia et al. 1999; Zheng et al. 2001). The reaction is assisted by hydrogen-bonding of the guanido nitrogens of Arg 324 with the O5 (2.86 ± 0.16 Å) and O4 oxygens (3.15 ± 0.33 Å) of PQQH (Fig. 4C
). In the MDH PQQH structure, Wat1 is hydrogen-bonded (2.88 ± 0.17 Å) to the O5 protonated oxygen of PQQH in good agreement with the crystal structure (Zheng et al. 2001). The hydrogen-bonding between NH1HH12 (Arg 324) and the carboxylate oxygen (OD1) of Asp 297CO2 is absent.
| Additionelimination mechanism of MDH |
|---|
|
|
|---|
|
| Hydride transfer mechanism of sGDH |
|---|
|
|
|---|
In the X-ray structure, the C1 glucose atom is directly above the C5 of PQQ, at a distance within 3.2 Å (Oubrie et al. 1999c). In
-D-glucose, the C1 hydrogen atom is axially positioned and points down, in the direction of the C5 of PQQ, at a separation of 1.2 Å from the latter. Therefore, the demonstrated absolute
-anomer specificity for glucose supports hydride transfer of the H1 at C1 of glucose to the C5 carbon of PQQ (Olsthoorn and Duine 1998b; Oubrie et al. 1999c). Such geometry of substrate and cofactor active-site groups has been observed in nicotinamide (Westheimer 1987; Karplus and Schulz 1989; Ramaswamy et al. 1994) and flavin-dependent oxidoreductases (Mattevi et al. 1996), which also use hydride equivalent transfer mechanisms.
Step 2: It has been suggested (Olsthoorn and Duine 1998b; Dewanti and Duine 2000) that the imidazolium of His 144 donates a proton to the oxyanion O5 of PQQH (Fig. 1B
). In the sGDH PQQH glucolactone structure, the average distance from the O5 of PQQH and His 144ImH+ is 4.79 Å. Thus, protonation of the oxyanion entity of PQQH by His 144 is not possible. The immediate product from glucose oxidation is glucolactone, which interacts with Arg 228 and His 144 of sGDH. The O5 of PQQH is engaged in hydrogen-bond interactions with the guanido nitrogens of Arg 228 (2.62 ± 0.08 Å and 2.86 ± 0.23 Å) and Wat89 (3.14 ± 0.18 Å; Fig. 4E
). The protonated NE2 of His 144 forms a hydrogen bond (3.05 ± 0.23 Å) with the O4 carbonyl oxygen of PQQH. It is therefore plausible that the rearrangement of (H)C5(O) C4( = O) of PQQH to C5(OH) = C4(OH) of PQQH2 is assisted by general acid protonation of the > C4 = O oxygen by His 144ImH+ and hydrogen bonds of Arg 228 to the oxyanion O5. Also, the imidazolium of His 144 is proximal (3.27 ± 0.55 Å) to the O1 oxygen of the glucolactone product.
Step 3: His 144 has been suggested (Olsthoorn and Duine 1998b; Dewanti and Duine 2000) to play the role of general base in migration of the H5 hydrogen at the tetrahedral C5 of PQQH to provide the hydroquinone PQQH2 (Fig. 1B
). This role for His 144 in the sGDH PQQH structure is unlikely because the NE2 of His 144 is too far (4.82 ± 0.25 Å) from the H5 of PQQH (Fig. 4F
). According to the MD studies, any rearrangement of (H)C5(OH) C4( = O) of PQQH to C5(OH) = C4(OH) of PQQH2 would be assisted by two hydrogen-bond networks mediated by waters Wat55 and Wat89. These involve Asp 252 and Gln 246 through Wat55, and Arg 228 and His 144 through Wat89 (Reddy and Bruice 2004). Wat89, at a distance of 2.74 ± 0.19 Å from the H5 of PQQH, would be involved in the departure of the C5 hydrogen of PQQH, if this species should be intermediate to PQQH2. This is an interesting feature not suggested by X-ray studies (Oubrie et al. 1999c). The hydrogen-bond interactions of the guanido nitrogens of Arg 228 with the O5 (2.97 ± 0.18 Å) and O4 (2.94 ± 0.15 Å) oxygens of PQQH would have a positive influence on the hydroquinone formation (Fig. 4F
).
| Additionelimination mechanism of sGDH |
|---|
|
|
|---|
-anomer configuration, with the hydrogen at C1 pointing below, directly toward the C5 of PQQ (which facilitates hydride transfer; Oubrie et al. 1999c). The average distance between the C5 of PQQ and the O1 of glucose is large (3.46 ± 0.26 Å) for adequate juxtaposition (based on van der Waals separation), compared with the distance between the C5 of PQQ and the C1 of glucose (3.57 ± 0.17 Å). The results are consistent with the proposal that considerable rearrangements in the active site are required to form a covalent C5 PQQglucose complex (Oubrie et al. 1999c; Oubrie and Dijkstra 2000). Therefore, the additionelimination mechanism is not preferred for glucose oxidation by sGDH. | Conclusions |
|---|
|
|
|---|
|
|
PQQH2 in the MDH PQQH structure is assisted, not by Asp 297CO2, but by the water TIP3P/Wat97 mediated by Glu 171CO2 (Fig. 6The presence of Arg 324 and Asn 387 residues in MDH and Arg 228 and Asn 229 in sGDH close to the cofactor in the active sites is conserved. But the precise orientation of these residues with respect to the cofactor is slightly different in both enzymes. The position of the Arg residue is stabilized by proximal water (either Wat124 in MDH or Wat89 in sGDH). In both enzymes, hydrogen-bondings to Arg 324 of MDH and Arg 228 in sGDH polarizes the > C5 = O5 of PQQ and the tetrahedral C5O5 of PQQH, and PQQH to promote the reaction. The hydrogen-bonding of the amide nitrogen of Asn 387 of MDH or Asn 229 of sGDH to the O4 carbonyl oxygen of the cofactor may also aid the reaction. However, the interactions of Arg 324 and Asn 387 with the O5 and O4 of PQQH, respectively, are absent in the MDH PQQH structure. Bidentate hydrogen-bond interactions of Arg 324 with the O4 oxygen of PQQHstabilize MDH PQQH, but similar interactions of Arg 228 with the O5 oxyanion of PQQH are dominant in the sGDH PQQH glucolactone structure.
The hydrogen-bonding of the O5 oxygen of cofactor to Wat1 of MDH or Wat89 of sGDH is observed during the entire reaction in both enzymes. Distinctly, Ca2+ coordination to the O5 quinone carbonyl oxygen of PQQ is absent in the MDH PQQ methanol structure, whereas the same is observed in the sGDH PQQ glucose structure. In both enzymes, the ligands of Ca2+ to PQQH (in MDH PQQHand sGDH PQQH glucolactone) and PQQH (in MDH PQQH and sGDH PQQH) are the same. In sGDH, the hydrogen-bondings of Wat89 to the guanido of Arg 228 are observed throughout the entire reaction. In the MD structures of sGDH, the orientation of His 144 and Asn 229 residues are stabilized by hydrogen-bonding interactions with Asp 163CO2 and the hydroxyl oxygen of Ser 146, respectively, in agreement with the X-ray structure (Oubrie et al. 1999c).
Various new structural and mechanistic features of MDH and sGDH not amenable for experimentation have been explored using MD studies. Our studies provide evidence that MD procedures have the ability to clarify enzyme mechanisms, based on juxtaposition of the active-site residues. Is the hydride transfer mechanism a common feature of PQQ enzymes and other quinoproteins? It is early to answer conclusively, as more computational work is required on other PQQ dehydrogenases, and quinoproteins involving amino-acid-derived cofactors. Future MD studies on quinoproteins should be promising in providing valuable insights on structure and enzymatic mechanisms.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
Afolabi, P.R., Mohammed, F., Amaratunga, K., Majekodunmi, O., Dales, S.L., Gill, R., Thompson, D., Cooper, J.B., Wood, S.P., Goodwin, P.M., et al. 2001. Site-directed mutagenesis and X-ray crystallography of the PQQ-containing quinoprotein methanol dehydrogenase and its electron acceptor, cytochrome cL. Biochemistry 40: 97999809.[CrossRef][Medline]
Anthony, C. 1986. The bacterial oxidation of methane and methanol. Adv. Microbiol. Physiol. 27: 113210.[Medline]
. 1992. The c-type cytochromes of methylotrophic bacteria. Biochim. Biophys. Acta 1099: 115.[CrossRef][Medline]
. 1996. Quinoprotein-catalysed reactions. Biochem. J. 320: 697711.
. 2001. Pyrroloquinoline quinone (PQQ) and quinoprotein enzymes. Antioxid. Redox Sign. 3: 757774.
Anthony, C. and Ghosh, M. 1998. The structure and function of the PQQ-containing quinoprotein dehydrogenases. Prog. Biophys. Mol. Biol. 69: 121.[CrossRef][Medline]
Anthony, C., Ghosh, M., and Blake, C.C.F. 1994. The structure and function of methanol dehydrogenase and related quinoproteins containing pyrroloquinoline quinone. Biochem. J. 304: 665674.
Avezoux, A., Goodwin, M., and Anthony, C. 1995. The role of the novel disulphide ring in the active site of the quinoprotein methanol dehydrogenase from Methylobacterium extorquens. Biochem. J. 307: 735741.
Berendsen, H.J.C. and Hayward, S. 2000. Collective protein dynamics in relation to function. Curr. Opin. Struct. Biol. 10: 165169.[CrossRef][Medline]
Brooks III, C.L. and Karplus, M. 1989. Solvent effects on protein motion and protein effects on solvent motion. J. Mol. Biol. 208: 159181.[CrossRef][Medline]
Brooks III, C.L., Pettitt, M.B., and Karplus, M. 1985. Structural and energetic effects of truncating long range interactions in ionic and polar fluids. J. Chem. Phys. 83: 58975908.[CrossRef]
Chen, Z.-W., Matsushita, K., Yamashita, T., Adachi, O., Beilamy, H.D., and Mathews, F.S. 2002. Structure at 1.9 Å resolution of a quinohemoprotein alcohol dehydrogenase from Pseudomonas putida HK5. Structure 10: 837848.[Medline]
Cleton-Jansen, A.-M., Goosen, N., Wenzel, T.J., and van de Putte, P. 1988. Cloning of the gene encoding quinoprotein glucose dehydrogenase from Acinetobacter calcoaceticus: Evidence for the presence of a second enzyme. J. Bacteriol. 170: 21212125.
Crennell, S.J., Garman, E.F., Laver, W.G., Vimr, E.R., and Taylor, G.L. 1993. Crystal structure of a bacterial sialidase (from Salmonella typhimurium LT2) shows the same fold as an influenza virus neuraminidase. Proc. Natl. Acad. Sci. 90: 98529856.
Daggett, V. 2000. Long timescale simulations. Curr. Opin. Struct. Biol. 10: 160164.[CrossRef][Medline]
Datta, S., Mori, Y., Takagi, K., Kawaguchi, K., Chen, Z.-W., Okajima, T., Kuroda, S., Ikeda, T., Kano, K., Tanizawa, K., et al. 2000. Structure of a quinohemoprotein amine dehydrogenase with an uncommon redox cofactor and highly unusual crosslinking. Proc. Natl. Acad. Sci. 98: 1425814273.
Davidson, V.L. 2001. Pyrroloquinoline quinone (PQQ) from methanol dehydrogenase and tryptophan tryptophylquinone (TTQ) from methylamine dehydrogenase. Adv. Prot. Chem. 58: 95140.[Medline]
Dekker, R.H., Duine, J.A., Frank, J., Verwiel, J., and Westerling, J. 1982. Covalent addition of H2O, enzyme substrates and activators to pyrroloquinoline quinone, the coenzyme of quinoproteins. Eur. J. Biochem. 125: 6973.[Medline]
Dewanti, A.R. and Duine, J.A. 2000. Ca2+-assisted, direct hydride transfer, and rate-determining tautomerization of C5-reduced PQQ to PQQH2, in the oxidation of 3-D-glucose by soluble, quinoprotein glucose dehydrogenase. Biochemistry 39: 93849392.[CrossRef][Medline]
Dijkstra, M., Frank, J., and Duine, J.A. 1989. Studies on electron transfer from methanol dehydrogenase to cytochrome cL, both purified from Hyphomicrobium X. Biochem. J. 257: 8794.[