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

J. Biol. Chem., Vol. 277, Issue 22, 19938-19945, May 31, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/22/19938    most recent M112244200v1 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 Pohl, E. Articles by Hensel, R. Search for Related Content PubMed PubMed Citation Articles by Pohl, E. Articles by Hensel, R. Social Bookmarking                      What's this?

The Crystal Structure of the Allosteric Non-phosphorylating Glyceraldehyde-3-phosphate Dehydrogenase from the Hyperthermophilic Archaeum Thermoproteus tenax^*

Ehmke Pohl^§, Nina Brunner^¶, Matthias Wilmanns, and Reinhard Hensel^¶**

From the  European Molecular Biology Laboratory, Hamburg Outstation, Notkestraße 85, D-22603 Hamburg, Germany and the ^¶ Department of Microbiology, University of Essen, Universitätsstraße 5, D-45117 Essen, Germany

Received for publication, December 21, 2001, and in revised form, February 11, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

The NAD⁺-dependent non-phosphorylating glyceraldehyde-3-phosphate dehydrogenase (GAPN) from the hyperthermophilic archaeum Thermoproteus tenax represents an archaeal member of the diverse superfamily of aldehyde dehydrogenases (ALDHs). GAPN catalyzes the irreversible oxidation of D-glyceraldehyde 3-phosphate to 3-phosphoglycerate. In this study, we present the crystal structure of GAPN in complex with its natural inhibitor NADP⁺ determined by multiple anomalous diffraction methods. The structure was refined to a resolution of 2.4 Å with an R-factor of 0.21. The overall fold of GAPN is similar to the structures of ALDHs described previously, consisting of three domains: a nucleotide-binding domain, a catalytic domain, and an oligomerization domain. Local differences in the active site are responsible for substrate specificity. The inhibitor NADP⁺ binds at an equivalent site to the cosubstrate-binding site of other ALDHs and blocks the enzyme in its inactive state, possibly preventing the transition to the active conformation. Structural comparison between GAPN from the hyperthermophilic T. tenax and homologs of mesophilic organisms establishes several characteristics of thermostabilization. These include protection against heat-induced covalent modifications by reducing and stabilizing labile residues, a decrease in number and volume of empty cavities, an increase in -strand content, and a strengthening of subunit contacts by ionic and hydrophobic interactions.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

The non-phosphorylating glyceraldehyde-3-phosphate dehydrogenase (GAPN)¹ from the hyperthermophilic archaeum Thermoproteus tenax (Tt-GAPN) catalyzes the NAD⁺-mediated phosphate-independent irreversible oxidation of D-glyceraldehyde 3-phosphate (D-GAP) to 3-phosphoglycerate. The enzyme is regulated by a number of effectors that modulate the binding of its cosubstrate NAD⁺. NADH and ATP lower the affinity for NAD⁺, whereas AMP, ADP, glucose 1-phosphate, and fructose 6-phosphate increase the affinity. All effectors induce positive cooperativity, with NADH showing the highest Hill coefficient of all effectors of 1.9. Due to its allosteric properties, Tt-GAPN plays a crucial role in regulating carbohydrate catabolism, thus enabling the organism to respond immediately to physiological requirements (1). Generally, the activity of GAPN is counteracted by the NADP⁺-dependent phosphorylating glyceraldehyde-3-phosphate dehydrogenase (GAPDH), which preferentially exerts anabolic functions. In contrast to GAPN, which is expressed at similar levels under heterotrophic and autotrophic conditions in T. tenax, GAPDH is preferentially expressed under autotrophic conditions. Thus, the regulation of both enzymes occurs at different levels, allowing a short-term and a long-term adaptation to changing physiological requirements and growth conditions (2).

Although the phosphorylating T. tenax GAPDH belongs to the archaeal family of classical phosphate-dependent GAPDHs, sequence analyses suggested that Tt-GAPN is a member of the superfamily of aldehyde dehydrogenases (ALDHs) (3). These enzymes catalyze the irreversible oxidation of aldehydes to their respective carboxylic acids. The highest sequence identities are found for other D-GAP-specific ALDHs from bacteria as well as plants (1). In contrast to these NADP⁺-dependent enzymes, Tt-GAPN explicitly uses NAD⁺ as cosubstrate and moreover exhibits allosteric properties that have not been described for any other ALDH yet investigated (1).

Recently, ALDHs have been of considerable interest because human ALDHs are involved in physiological impairments such as alcoholism, Sjögren-Larsson syndrome, and carcinogenesis (3). Their broad substrate variety also lends them great biotechnological potential. Crystal structures of ALDHs have been solved for members of class 1 (4, 5), class 2 (6), class 3 (7), and class 9 (8) ALDHs; for the class 3-related enzyme from Vibrio harvei (9); and for GAPN from Streptococcus mutans (Sm-GAPN) (10). Although the sequence identities for the different classes can be as low as 20%, the structures determined so far adopt the same basic fold comprising three domains, one dinucleotide-binding domain, one catalytic domain, and one oligomerization domain. Unlike the classical Rossmann fold, the nucleotide-binding domain of ALDHs consists of only five parallel -strands. The interaction between the pyrophosphate moiety of the nucleotide and the protein occurs via the loop between strand _J and helix ₈. This arrangement results in a significantly different conformation of the cosubstrate compared with proteins with the Rossmann fold. The catalytic domain adopts an /-fold and contains the conserved active-site cysteine. The oligomerization domain is responsible for dimer and tetramer formation.

The catalytic mechanism of ALDHs can be described as a two-step mechanism consisting of acylation and deacylation of an active cysteinyl residue (11). The acylation involves the formation of a thiohemiacetal intermediate, followed by hydride transfer and reduction of the pyridine nucleotide. Subsequently, the resulting thioacyl intermediate is hydrolyzed, and the product and the reduced cosubstrate are released. Despite many biochemical and mutagenesis studies, the mechanism and the role of putatively essential residues of this reaction remain to be elucidated at a molecular level (12). For instance, the invariant Glu²⁶⁸ (numbering according to class 2 ALDH) has been implied to function as a general base, deprotonating the active-site cysteine in class 2 ALDH (6). In class 3 ALDH, however, this residue seems to be involved only in cosubstrate binding. In the case of Sm-GAPN, this residue has been suggested to play an essential role in deacylation by activating a water that hydrolyzes the intermediate (53).

In this work, the crystal structure of Tt-GAPN in complex with its natural inhibitor NADP⁺ is presented. The structure was solved by multiple anomalous diffraction and refined to a resolution of 2.4 Å with a crystallographic R-factor of 0.21. This represents the first crystal structure of a hyperthermophilic archaeal member of the ALDH superfamily and allows us to investigate the structural basis for substrate specificity and allosteric properties. Detailed comparisons of Tt-GAPN, which exhibits a half-life of thermoinactivation of ~30 min at 100 °C, with homologous enzymes from mesophilic organisms have been performed to unravel the characteristics of thermostabilization.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

Protein Expression, Purification, and Crystallization-- The protein was expressed in Escherichia coli and purified as described previously (1). Crystals were obtained by the hanging drop vapor diffusion method with 2.0 M sodium formate as a precipitating agent. The crystals belong to the hexagonal space group P6₂22, with unit cell dimensions of a = b = 185.2 Å and c = 132.1 Å (13). The selenium-methionine-substituted protein was produced using standard methods and purified by the same protocol developed for the wild-type protein (13). Full incorporation of selenium-methionine was verified by mass spectrometric analysis (data not shown), and the protein was crystallized under the same conditions as the wild-type protein.

X-ray Data Collection and Processing-- The crystals were frozen in a rayon loop using the reservoir solution plus 15-25% (v/v) glycerol as a cryoprotectant (14). All diffraction data were collected at cryogenic temperatures. The multiple anomalous diffraction data were collected on the European Molecular Biology Laboratory wiggler beam line BW7A equipped with a Mar300 imaging plate scanner. A native data set to a resolution of 2.4 Å was collected on the wiggler beam line BW7B with a Mar345 imaging plate scanner (15). All images were processed with DENZO and SCALEPACK (16). Integrated intensities were converted into structure factors using TRUNCATE (17), which is part of the CCP4 package (18). Further data collection statistics are summarized in Table I.

Structure Solution and Refinement-- As no suitable search model was available at the time of data collection, the structure was solved by multiple anomalous diffraction phasing. Eight of nine possible selenium positions were identified using the programs SHELXD (19) and SOLVE (20). SOLVE was used to refine the positions and to calculate an initial experimental map with a figure of merit of 0.56 at a resolution of 2.9 Å. The map was further improved by density modification using a solvent content of 0.75 with DM (21). This procedure increased the figure of merit to 0.79, and the resulting electron density map was readily interpretable. The complete protein model was built into the experimental map using the interactive graphics program O (22). The model was refined against the native data set using the maximum likelihood target function implemented in CNS (23). The refinement using data from 30 to 2.4 Å was completed by iterative cycles of model building and simulated annealing (24) and conjugate gradient least-squares coordinate and restrained B-factor refinement. The solvent model for low resolution data and an anisotropic overall B-factor correction of the program CNS were applied. During all steps of the refinement, 5% of the data were used to calculate the free R-factor (25). Further statistics describing the crystallographic refinement are summarized in Table II.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

Quality of the Model-- The final model contains 499 residues (positions 3-501), one NADP⁺ molecule, one sodium ion, and 385 solvent molecules. The first two residues are not defined in the density map and are presumably disordered. The side chains of residues 3, 319, 328, and 499-501 are not visible in the electron density map and were thus refined as alanines. The model possesses good geometry, with 90% of all residues in the most favored and 9.6% in the allowed regions of the Ramachandran plot as indicated by the program Procheck (26) (data not shown). The two residues (Lys⁴⁶⁶ and Phe⁴⁷¹) that are in generously allowed regions are located in a flexible loop. The r.m.s. deviations for bond lengths and angles compared with the standard values of Engh and Huber (27) are 0.009 Å and 1.48°, respectively. The refinement converged to a crystallographic R-factor of 0.21 (R_free = 0.24).

Overall Structure and Comparison with ALDH Structures-- The crystal structure of the Tt-GAPN monomer is shown as a ribbon diagram in Fig. 1a, with the annotation of secondary elements given in Fig. 2. The monomer is composed of three domains that are present in all ALDHs analyzed so far. The cosubstrate-binding domain consists of a central parallel five-stranded -sheet (_K, _J, _G, _H, and _I) surrounded by five helices (₁, ₅, ₆, ₇, and ₈) (shown in blue) and a small N-terminal antiparallel -sheet (_A, _B, _C, and _D) (shown in green). The catalytic domain (residues 266-441 shown in yellow) adopts an /-fold with the central parallel -sheet of _L to _X and helices ₉ to ₁₃. This domain contains the active-site cysteine (Cys²⁹⁷) that is highly conserved in the ALDH superfamily and the loop region of residues 455-460 that is implicated in substrate specificity (orange). The oligomerization domain is composed of three antiparallel -strands (red). Two of these strands (_E and _F) are part of the N-terminal half, whereas the third strand (_Z) originates from the C terminus. The remaining nine C-terminal residues are wrapped around the turn of the first two -strands.

View larger version (37K): [in this window] [in a new window]   Fig. 1.   a, ribbon diagram of the crystal structure of the Tt-GAPN monomer. The N-terminal -sheet of the cosubstrate-binding domain is shown in green; the remaining part of this domain is in blue. The catalytic domain is depicted in yellow, and the oligomerization domain is in red. The loop region of residues 455-460 that may be responsible for specificity is shown in orange. NADP+ is shown in a red ball-and-stick representation. b, ribbon diagram of the crystallographic tetramer. Each symmetry equivalent monomer is depicted in one color: monomer A, magenta; monomer B, blue; monomer C, orange; and monomer D, green. The three strands of the oligomerization domain in each subunit in the center of the figure are depicted in darker shades of the same color. The inhibitors bound in each monomer have been omitted for clarity. All figures were prepared using Molscript (47); its extended version, Bobscript (48); and/or raster3d (49).

View larger version (69K): [in this window] [in a new window]   Fig. 2.   Structure-based sequence alignment of Tt-GAPN, Sm-GAPN (10), and class 2 bovine ALDH (6) using DALI (29). The secondary structure elements of the Tt-GAPN structure are indicated by boxes (-helices) and arrows (-strands). The residues with a gray background are structurally equivalent. The active-site residues are highlighted in red; the residues involved in NADP+ binding in yellow; and the important loops of the active site in green. Tt, Tt-GAPN; Sm, Sm-GAPN; bov, class 2 bovine ALDH.

Structural comparison of Tt-GAPN was performed using the crystal structures of dimeric class 3 ALDH from rat (sequence identity of 21.2% (28)) (7), tetrameric class 2 bovine ALDH (28.5%) (6), and Sm-GAPN (33.9%) (10). The superposition of Tt-GAPN and ALDHs based on the structural alignment by DALI (29) results in r.m.s. deviations of 2.4 Å for class 2 ALDH and 2.5 Å for class 3 ALDH. The central -sheets of the cosubstrate-binding and catalytic domains superimpose well, with r.m.s. deviations of only ~1 Å for equivalent C atoms. The GAPN-specific sequences (1) that are not found in any ALDH superimpose generally well with the corresponding residues in the ALDH structure. The C atoms of the active-site loop (residues 294-297, shown in green in Fig. 2) are within 1 Å after the least-squares superposition of the C backbone. However, larger differences of >5 Å for equivalent C atoms can be found in the flanking helices and the connecting loops. The largest differences occur in the first 20 residues that form two short antiparallel -strands in Tt-GAPN. In class 2 bovine ALDH, the equivalent residues adopt a random coil conformation pointing into the solvent.

As expected from the higher sequence identity, the structures of Tt-GAPN and Sm-GAPN show the highest similarity. The superposition of structurally equivalent C atoms results in a r.m.s. deviation of 1.6 Å. The largest differences can be seen in the loop regions, whereas the cores of the catalytic and cosubstrate-binding domains superimpose very well, with deviations of <1 Å for the main chain atoms (Fig. 3). The Tt-GAPN structure possesses longer termini and seven deletions and insertions compared with Sm-GAPN. The first 14 N-terminal residues in Tt-GAPN are folded back over the N-terminal sheet and form an antiparallel -strand with residues 10-14 and one helical turn (residues 4-6). In addition, there are four one-amino acid insertions (Lys⁶⁵, Lys¹³², Thr³³³, and Leu³⁶⁹) and three deletions (following Ala¹²⁶, Asp¹⁴¹, and Lys²²⁵) that do not lead to large structural changes. The largest insertion of three amino acids (Ala³⁸³-Arg³⁸⁵) is located in helix ₁₂ of the catalytic domain and is simply accommodated by an additional N-terminal turn of the helix. The C terminus of Tt-GAPN is also extended by eight amino acids compared with Sm-GAPN. These residues are folded back over strands _E and _F of the oligomerization domain.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Least-squares superposition of structurally equivalent C atoms of Tt-GAPN (red) and Sm-GAPN (blue) (10).

Tetramer Formation-- In the crystal structure presented here, the functional Tt-GAPN tetramer is generated by the crystallographic 222 symmetry (Fig. 1b). This tetramer can be described as a dimer of two equivalent dimers (A-B, shown in magenta and blue; and C-D, depicted in orange and green) that form extensive interactions mainly via the oligomerization domain (shown in darker shades of the same color in Fig. 1b). Strand _X from the catalytic domain of subunit A interacts with strand _Z from the oligomerization domain of subunit B. These two strands are thus part of a 10-stranded parallel -sheet structure extending through both the catalytic and oligomerization domains. There are additional contacts between the helix _A and its symmetry mate. The A-B interface has a buried surface of ~3000 Ų and is mainly composed of hydrophobic residues. The tetramer is completed by weaker interaction between the two dimers A-B and C-D. The equivalent A-C and B-D interfaces provide the smallest subunit contact with a buried surface of 840 Å² for each interface. This interface is stabilized by hydrophobic interactions of residues Leu¹³⁰ and Lys¹³¹, which are located at the beginning of the tetramerization domain to their symmetry-related residues. In addition, helix ₁₄ of the catalytic domain interacts with C-terminal residues 492-498 via hydrophobic contacts between Val⁴³³ and the methylene groups of Arg⁴³⁶ on one side and between Tyr^492' and the methylene groups of Lys^495' and Val^497' on the other side. This contact is extended into the first domain by a salt bridge from Arg⁴³⁶ to Asp¹⁵⁰. The A-D and B-C dimer interfaces are composed of the following two sets of interactions: (i) strand _E of each oligomerization domain forms a -sheet with its symmetry-related strand, and (ii) the loop region of residues 430-432 interacts with its symmetry-related counterpart through four hydrogen bonds (Arg⁴³⁰ N-Asp^431' O² and Arg⁴³⁰ O-Val^432' N). The overall buried surface of the A-D (and the B-C) interface amounts to 1340 Å². In the center of the Tt-GAPN tetramer, a large 1080 Å³ cavity filled with well defined solvent molecules can be found. A similar feature is found in the Sm-GAPN structure; but, due to the smaller contact surface in Sm-GAPN, the space is solvent-accessible through open channels.

NADP⁺ Binding-- Tt-GAPN utilizes NAD⁺ exclusively as a cosubstrate, whereas NADP⁺ acts as strong inhibitor (K_D = 1 µM). Inhibitor binding reduces the affinity for the cosubstrate and induces positive cooperativity of NAD⁺ binding (1). Although no NADP⁺ was added during the enzyme purification and crystallization, the experimental electron density map unambiguously revealed NADP⁺ as a ligand (Fig. 4a). We assume that the inhibitor remained bound to the protein throughout the isolation, purification, and crystallization of the enzyme. The adenine dinucleotide phosphate is well defined in the electron density map, whereas the nicotinamide moiety shows weaker density (Fig. 4a), which is also reflected by higher B-factors. NADP⁺ is located in a small pocket and is stabilized by a number of salt bridges and hydrogen bonds to the protein. The oxygen atoms of the 2'-phosphate form hydrogen bonds to the main chain nitrogen atoms of Ile¹⁹⁴ and Gly²²⁴ and to the N atom of Lys¹⁹¹ and the O atom of Ser¹⁹³ (Fig. 4b). In addition, the 3'-hydroxyl group is hydrogen-bonded to the carbonyl oxygen of Thr¹⁶⁵. The diphosphate moiety forms one hydrogen bond to the main chain nitrogen of Ser²⁴⁴ as well as one to the side chain O. The 2'- and 3'-hydroxyl groups of the ribose ring at the nicotinamide are hydrogen-bonded to the Glu³⁹⁵ side chain oxygen atoms. Because this residue is strictly conserved (Fig. 2), it has been suggested that this interaction is responsible for directing the cosubstrate orientation (9). The nicotinamide group is connected to the O² atom of Glu²⁶³ and the main chain nitrogen atom of Gly²⁶⁵ by additional hydrogen bonds. These two interactions appear to stabilize the conformation observed in the crystal. As the binding site of NADP⁺ is similar to the cosubstrate-binding site identified previously in ALDH crystal structures (6, 10), the NAD⁺ cosubstrate of GAPN is most likely bound in the same pocket. This assumption is further confirmed by the observation of a similar position and conformation of NADP⁺ in the Tt-GAPN and Sm-GAPN structures (Fig. 4c) and comparable protein-ligand contacts. Remarkably, the Sm-GAPN structure shows an additional hydrogen bond from the O atom of the Thr¹⁸⁰ side chain to the phosphate, which is not present in the Tt-GAPN structure; and it has been speculated that this interaction is responsible for NADP⁺ selectivity over NAD⁺ (10). The loss of this hydrogen bond is, however, compensated by two additional hydrogen bonds to the 2'-phosphate of NADP⁺: (i) the side chain oxygen of Ser¹⁹³ and (ii) the main chain nitrogen of Gly²²⁴ serve as hydrogen donors for the negatively charged phosphate as described above.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   a, stereo view of a representative portion of the experimental electron density map after solvent flattening at a 1.5- level showing the inhibitor-binding site. The final model of NADP+ is superimposed as a ball-and-stick representation. b, close-up stereo view of the binding site showing the hydrogen bonds between the phosphate and protein atoms. The nicotinamide moiety has been omitted for clarity. c, position of NADP+ in the structures of Tt-GAPN (gray bonds) and Sm-GAPN (white bonds) after the least-squares superposition of the C trace of the two enzymes.

Active Site-- A number of residues that have been implied to be essential for the catalytic reaction in ALDHs are conserved in Tt-GAPN, suggesting that the molecular mechanism of catalysis is likely to be similar (see conserved active-site residues in Fig. 2). According to the reaction mechanism suggested by Marchal et al. (53), Cys²⁹⁷ (corresponding to Cys²⁸⁴ in Sm-GAPN) performs the nucleophilic attack on the aldehyde group of the substrate to form the thiohemiacetal. Asn¹⁶⁸ (Asn¹⁵⁴) then stabilizes the intermediate and facilitates the transfer of the hydride to the cosubstrate. Glu²⁶³ (Glu²⁵⁰) activates the subsequent deacylation step by directed hydrolysis of the thioester bond. However, in the conformation observed in the crystal structure presented here, the distance between the side chain atoms of Glu²⁶³ and Cys²⁹⁷ is ~6 Å (Fig. 5). In addition, the nicotinamide ring of NADP⁺ is located between these two residues, thus blocking any potential interaction. In the case of Sm-GAPN, the binding of D-GAP induces, however, a structural change to the active conformation as shown by the crystal structure of the active-site variant C284S Sm-GAPN in complex with NADP and D-GAP. The superposition of the binary complex of Tt-GAPN and the ternary complex of C284S Sm-GAPN shows that a similar conformational change would also allow the reaction in Tt-GAPN. In the case of Tt-GAPN, the tight binding of NADP⁺ may block this conformational change from the inactive binary complex to its active form.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Stereo view of the active site after the superposition of the crystal structures of Tt-GAPN and C284S Sm-GAPN in complex with its substrate determined at 3-Å resolution (12). NADP+, which is present in both structures, has been omitted for clarity. The C trace of the loop region of residues 454-460 in Tt-GAPN is shown in red; the corresponding loop in Sm-GAPN (residues 436-442) is shown in blue.

The substrate-binding pocket of C284S Sm-GAPN reveals that D-GAP is bound by a number of ionic interactions to the phosphate (e.g. by Arg¹⁰³, Tyr¹⁵⁵, and Arg²⁸³) and to the carbonyl oxygen at C-1 (Asn¹⁵⁴ and Ser²⁸⁴) (12). However, because Sm-GAPN uses both D-GAP and L-GAP as substrate, the interaction between Arg⁴³⁷ and the chiral C-2 does not confer any stereospecificity. The structure of Tt-GAPN, which uses only D-GAP as a substrate, possesses a similar anion-binding pocket, however, with significant differences. The equivalent residue of Arg⁴³⁷ is a histidine (His⁴⁵⁵) that is turned away from the putative substrate-binding site and forms a salt bridge to Asp⁴²⁴. The main chain residues 456-459 adopt a remarkably different orientation (Fig. 5), with equivalent atoms of this loop differing by up to 5 Å between the Tt-GAPN and Sm-GAPN structures. It is important to note that this loop possesses the same conformation in the binary wild-type Sm-GAPN and ternary C284S Sm-GAPN structures. These structural changes observed are therefore not due to substrate binding and may thus be responsible for the stereoselectivity of Tt-GAPN. In the structure presented here, the D-GAP-binding site is filled with three water molecules that mimic the position of the substrate. The exact reason for stereospecificity cannot be determined, but it is likely that Tt-GAPN-specific features such as His⁴⁵⁵ and the conformation of the loop region of residues 456-459 are involved. Clearly, further structural studies of Tt-GAPN in complex with its substrate are needed to answer this question.

Allosteric Regulation-- The distance between the cosubstrate-binding sites in the Tt-GAPN tetramer is ~30 Å for the A-B and C-D dimers. Thus, it is unlikely that there is a direct connection between these two sites to induce cooperativity. Presumably, an overall motion of domains or even monomers with respect to each other is responsible for cooperative binding. The effector molecules such as AMP, ADP, and glucose 1-phosphate bind most likely at a different site than the cosubstrate sites determined in the current structure. Binding of the effector might then induce a conformational change in the enzyme to the active conformation. Further structural studies of Tt-GAPN in complex with its allosteric regulators are needed to unravel their mode of action.

Determinants for Protein Stability-- Although no general strategy of thermostability has been established, it has become clear that all levels of protein structure from primary to quaternary structure may contribute to thermal stabilization (30-33). We will identify below individual determinants for thermostability in Tt-GAPN by comparing its crystal structure with those of mesophilic homologs, viz. Sm-GAPN (10) and class 2 bovine ALDH (6).

At elevated temperature, damage to the covalent structure is of particular importance, with several residues known to be prone to irreversible chemical reactions. Asparagine and glutamine residues are susceptible to deamidation of the side chain and cleavage of the peptide bond, and cysteine is vulnerable to oxidation and -elimination (34, 35). Consequently, there are fewer of these residues in Tt-GAPN compared with its mesophilic homologs (Table III). Nearly all of the remaining Asn and Gln side chains are hydrogen-bonded to other protein atoms, thus providing additional protection against irreversible damage. These hydrogen bonds stabilize the residues in a conformation that may prevent destruction. Certain sequence motifs that are known to promote asparagine-related damage (e.g. Asn-Gly, Asn-Ser, Asn-Thr, and Ser-Asn) (34) are also rare. Tt-GAPN contains only one of these motifs (at residues 417 and 418), and it is stabilized by two hydrogen bonds of the Asn⁴¹⁷ side chain to main chain atoms. In contrast to Tt-GAPN, class 2 bovine ALDH contains three and Sm-GAPN contains six of these "hot spots." The only cysteine that remains in Tt-GAPN is the active-site Cys²⁹⁷.

Electrostatic interactions have often been implied to be important for thermal stability (36). The higher number of charged residues in Tt-GAPN does not lead to an increase in the number of intramolecular salt bridges (Table III); but more charged residues are engaged in intermolecular interactions, and the total number is considerably higher in the hyperthermophilic enzyme than in the mesophilic ones. On the other hand, the number of hydrogen bonds per tetramer is not increased in Tt-GAPN (Table IV). These results correspond well with a comprehensive study by Szilagyi and Zavodszky (37), which revealed a clear preference for salt bridges, but no increase in hydrogen bonding in hyperthermophilic enzymes.

Structural comparison of Tt-GAPN with its mesophilic homologs reveals a considerable decrease in the number and volume of empty cavities. Hyperthermophilic Tt-GAPN possesses no internal cavities >2 ų, whereas the mesophilic homologs exhibit more and larger cavities (Table III). According to estimates by Eriksson et al. (38), the higher packing density would result in a gain of free energy of ~1-2 kcal/mol and thus would thermodynamically stabilize the protein. A similar tendency toward fewer cavities in hyperthermophilic enzymes has also been noted previously (37).

In hyperthermophilic proteins, the N and C termini are often more tightly bound, or even buried, to avoid initiation of thermal unfolding (39). This may lead to a kinetic stabilization of the protein. In Tt-GAPN, however, the N terminus is solvent-exposed, and the first three residues are invisible in the electron density map. The observed flexibility may represent a weak point in the current structure. The enzyme that is directly isolated from T. tenax cells is N-terminally blocked and shows higher thermal stability compared with the recombinant protein. Hence, the chemical modification may be responsible for the higher thermal stability by restricting the N-terminal flexibility. In contrast to the solvent-exposed N terminus, the prolonged C terminus is integrated in the hydrophobic subunit contact area and may thus confer stability to the tetramer.

Shortening and deletion of loops have been described as characteristic features of thermophilic proteins (40). The structure-based sequence alignment (Fig. 2) shows that the loops of the hyperthermophilic and mesophilic enzymes generally are of similar lengths; and consequently, loop shortening does not appear to contribute to the thermal stability of Tt-GAPN. At the secondary structural level, the Tt-GAPN structure follows the trend toward a higher percentage of -strands compared with its mesophilic counterparts (Table III). However, there is no notable increase in -helical content. In this respect, our results confirm previous comparative studies (37, 41). Additional indications of helix stabilization that have been described such as a decrease in -branched residues (42), an increase in intrahelical salt bridges (43), and a preference for N-capping residues such as Thr, Asp, Ser, and Gly (44) are not found in the Tt-GAPN structure. Only prolines seem to be slightly more numerous in hyperthermophilic GAPN, with five N-capping prolines compared with none in Sm-GAPN and two in class 2 bovine ALDH. Prolines at this particular position have been found to be only marginally favored in general and are slightly preferred in the so-called Pro box motif (hPXXhh) (45). None of the N-capping prolines are part of this motif; and hence, these residues are not expected to provide additional stabilization. Furthermore, the amino acid composition of loops indicates that conformationally restraining prolines are more numerous in the hyperthermophilic enzyme relative to the mesophilic enzymes (Table III). However, there are also more glycines, which provide additional flexibility when Tt-GAPN and Sm-GAPN are compared. The mesophilic bovine enzyme and Tt-GAPN have the same number of glycines in loops (Table III). In summary, there is no clear indication that the loops are stabilized by their amino acid composition in Tt-GAPN.

Intersubunit stabilization of oligomeric proteins has been considered more recently to represent one of the main factors in evolutionary strategies of thermoadaptation (46). Along with a series of recent reports that emphasize the importance of electrostatic and hydrophobic subunit interactions in hyperthermophilic oligomeric proteins (46), our comparison of subunit contacts of Tt-GAPN with its mesophilic homologs revealed remarkably increased hydrophobic interactions as well as a higher number of salt bridges (Table IV). Clearly, these two factors contribute significantly to thermostability.

	ACKNOWLEDGEMENTS

We are grateful to D. Lang for the initial contribution to this work. We also thank M. Wilm for the mass spectrometric analysis, R. van Silfhout and O. Mayans for support during data collection, and E. Lorentzen and P. Tucker for critically reading the manuscript and many valuable discussions.

	FOOTNOTES

^* This work was supported by Deutsche Forschungsgemeinschaft Grant He 1238/14-1.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 1KY8) 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 may be addressed. Tel.: 49-40-89902-192; Fax: 49-40-89902-149; E-mail: ehmke@embl-hamburg.de.

Present address: Bayer AG, Business Group Pharma, PH-R-AI 1, D-42096 Wuppertal, Germany.

^** To whom correspondence may be addressed. Tel.: 49-201-1833442; Fax: 49-201-1833990; E-mail: r.hensel@uni-essen.de.

Published, JBC Papers in Press, February 12, 2002, DOI 10.1074/jbc.M112244200

	ABBREVIATIONS

The abbreviations used are: GAPN, non-phosphorylating glyceraldehyde-3-phosphate dehydrogenase; Tt-GAPN, T. tenax GAPN; Sm-GAPN, S. mutans GAPN; D-GAP, D-glyceraldehyde 3-phosphate; GAPDH, phosphorylating glyceraldehyde-3-phosphate dehydrogenase; ALDH, aldehyde dehydrogenase; r.m.s., root mean square.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				J. S. Rodriguez-Zavala Enhancement of coenzyme binding by a single point mutation at the coenzyme binding domain of E. coli lactaldehyde dehydrogenase Protein Sci., March 1, 2008; 17(3): 563 - 570. [Abstract] [Full Text] [PDF]

				B. Siebers, B. Tjaden, K. Michalke, C. Dorr, H. Ahmed, M. Zaparty, P. Gordon, C. W. Sensen, A. Zibat, H.-P. Klenk, et al. Reconstruction of the Central Carbohydrate Metabolism of Thermoproteus tenax by Use of Genomic and Biochemical Data J. Bacteriol., April 1, 2004; 186(7): 2179 - 2194. [Abstract] [Full Text] [PDF]

				C. Dorr, M. Zaparty, B. Tjaden, H. Brinkmann, and B. Siebers The Hexokinase of the Hyperthermophile Thermoproteus tenax: ATP-DEPENDENT HEXOKINASES AND ADP-DEPENDENT GLUCOKINASES, TWO ALTERNATIVES FOR GLUCOSE PHOSPHORYLATION IN ARCHAEA J. Biol. Chem., May 23, 2003; 278(21): 18744 - 18753. [Abstract] [Full Text] [PDF]

				S. Marchal and G. Branlant Characterization of the Amino Acids Involved in Substrate Specificity of Nonphosphorylating Glyceraldehyde-3-Phosphate Dehydrogenase from Streptococcus mutans J. Biol. Chem., October 11, 2002; 277(42): 39235 - 39242. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/22/19938    most recent M112244200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services