Hyperthermophilic dehydrogenase enzymes - Semantic Scholar

7 downloads 0 Views 176KB Size Report
Exeter Biocatalysis Centre, School of Biological Science and Chemistry, ... of factors including an increase in the number of salt bridges and hydrophobic.
Thermophiles 2003

Hyperthermophilic dehydrogenase enzymes J.A. Littlechild1 , J.E. Guy and M.N. Isupov Exeter Biocatalysis Centre, School of Biological Science and Chemistry, Stocker Road, Exeter EX4 4QD, U.K.

Abstract Archaeal dehydrogenases are often found to be of a specific class of dehydrogenase which has low sequence identity to the equivalent bacterial and eukaryotic counterparts. This paper focuses on two different types of hyperthermophilic dehydrogenase enzyme that have been cloned and over-expressed in Escherichia coli. The crystallographic structures of the apo form of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) from Sulfolobus solfataricus and the related holo form of GAPDH from Methanothermus fervidus have been solved to high resolution. The zinc-containing structure of ADH (alcohol dehydrogenase) from Aeropyrum pernix has also been solved as a quaternary complex with the cofactor NADH and the inhibitor octanoic acid. The results show that despite the low sequence identity to the related enzymes found in other organisms the fold of the protein chain is similar. The archaeal GAPDH enzymes show a relocation of the active site which is a feature of evolutionary interest. The high thermostability of these three archaeal dehydrogenases can be attributed to a combination of factors including an increase in the number of salt bridges and hydrophobic interactions, a higher percentage of secondary structure and the presence of disulphide bonds.

Introduction The Archaea have been identified as a new phylogenetic branch of life distinct from Bacteria and eukaryotes and are thought to be representative of a primordial organism due to their isolation from extremophilic environments. This article will discuss two different dehydrogenase enzymes that have been isolated and characterized from three different hyperthermophilic Archaea.

GAPDH (glyceraldehyde-3-phosphate dehydrogenase) GAPDH (EC 1.2.11.12) catalyses the oxidative phosphorylation of D-glyceraldehyde-3-phosphate to form 1,3diphosphoglycerate and is a key enzyme in glycolysis and gluconeogenesis. Most GAPDHs known to date are homotetramers with a subunit molecular mass of around 37 kDa. Bacterial and eukaryotic GAPDHs show high sequence similarity (over 40% identity) and usually utilize the cofactor NAD+ , with the exception of some plant chloroplast GAPDHs which are NADP+ -specific [1]. The GAPDH from the Archaea has been identified as a different form of the enzyme compared with the bacterial and eukaryotic counterparts, due to the low overall sequence identity (16– 20%) and to its dual cofactor specificity for both NADP+ and NAD+ . In addition to the low sequence similarity between archaeal GAPDHs and the enzymes from the two other kingdoms, there is difficulty in aligning residues implicated in the catalytic mechanism. The combination of these factors has Key words: Aeropyrum pernix, Archaea, dehydrogenase, hyperthermophilic, Sulfolobus solfataricus. Abbreviations used: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ADH, alcohol dehydrogenase. 1 To whom correspondence should be addressed (e-mail [email protected]).

led to the suggestion that archaeal GAPDHs are unrelated to their bacterial and eukaryotic counterparts, showing a convergent molecular evolution in the catalytic region of their structure. The crystal structures of two hyperthermophilic archaeal GAPDHs have been elucidated. The Sulfolobus solfataricus GAPDH has been solved in the apo form [2] and the Methanothermus fervidus GAPDH [3] structure has been determined with the bound cofactor NADP+ . These two GAPDH enzymes share 49% sequence identity. Both enzymes have been cloned and over-expressed in Escherichia coli [4,5]. Archaeal GAPDHs display the tetrameric structure and overall protein fold found in other GAPDH enzymes, although they have additional secondary structure elements. However, the residues of the catalytic domain implicated in the catalytic mechanism and in the substrate phosphate (Ps) and inorganic phosphate (Pi ) binding sites seem to be relocated between different structural elements, with only the active-site cysteine and preceding serine residue retained in a similar position. In bacterial and eukaryotic GAPDHs a conserved histidine residue is located on strand β1 of the catalytic domain, and is thought to act as a base extracting the proton from the activesite cysteine residue during catalysis. However, there is no histidine residue at this position in the archaeal GAPDH structure. Instead, another residue conserved in Archaea, His-219 (Sulfolobus GAPDH numbering) from the strand β4 of the catalytic domain, positions its imidazole group in approximately the same location. It is therefore proposed that His-219 plays the same role in the archaeal enzyme as its counterpart histidine residue in other GAPDH enzymes. The apo Sulfolobus GAPDH crystal structure reveals two sulphate-binding sites close to the inorganic phosphate (Pi ) and substrate phosphate (Ps) sites as seen in other GAPDH  C 2004

Biochemical Society

255

256

Biochemical Society Transactions (2004) Volume 32, part 2

Figure 1 Superposition of the active sites of S. solfataricus GAPDH and GAPDH from the bacterial source, Bacillus stearothermophilus [14] The common secondary structure elements are shown as a ribbon model. The active-site residues and Pi and Ps phosphates of the B. stearothermophilus enzyme are shown as blue bonds, those of the S.

Figure 2 The binding of the cofactor NADP+ in the GAPDH from M. fervidus The secondary structure is presented as a ribbon model with different subunits shown in different colours. NADP+ is shown as a ball-and-stick model, as is the sulphate ion which is bound at the inorganic phosphate site Pi . Figure produced using Bobscript [15].

solfataricus enzyme are shown as black bonds. Many active-site residues in S. solfataricus GAPDH are shifted by one β-strand towards the top of the β-sheet in relation to the B. stearothermophilus enzyme. Figure produced using Bobscript [15].

enzymes. The Pi site is co-ordinated by the hydroxy group of Ser-138, the main chain nitrogen and amide nitrogen of Asn-140, the guanidinium group of Arg-166 and imidazole groups of His-192 and His-193 (Figure 1). The serine residue preceding the active-site cysteine binds Pi in all known GAPDH enzymes. The Ps site is formed by the guanidinium groups of Arg-166 and Arg-167 which are also conserved in all known archaeal GAPDH enzymes, including the holo Methanothermus GAPDH enzyme which shows the NADP+ cofactor bound in the proposed binding pocket (Figure 2). Like other GAPDH enzymes the cofactor is bound in the syn conformation. However, it is stabilized by binding to an aspartic acid residue that is distinct from archaeal GAPDH enzymes. It is proposed that an ancestral GAPDH enzyme had a similar fold and a related quaternary structure but had a low turnover and broad specificity, using only a cysteine residue for catalysis. It seems that the active sites of archaeal and bacterial/eukaryotic GAPDH enzymes eventually converged to a similar three-dimensional location.  C 2004

Biochemical Society

Both of the archaeal enzymes belong to a different class from other GAPDH structures since they have additional secondary structural elements. The largest of these are found at the C-terminus where an additional α-helix of the nucleotide binding domain is followed by an extra strand of the catalytic domain. In this way these enzymes have an extra interdomain connection in comparison with their GAPDH counterparts. The Sulfolobus GAPDH has a disulphide bridge, unusual in cytoplasmic proteins, that links the catalytic and cofactor domains on the outside of the protein structure. It is speculated that this disulphide bond could act to stabilize the enzyme at high temperatures and potentially to regulate its enzymic activity. There are several salt-bridge clusters in the archaeal thermophilic GAPDH structures. One of these in the Sulfolobus GAPDH structure includes 14 charged residues, a sulphate molecule and His-299, and extends from the subunit interface into the active site. This ionic network is thought to be involved with the stability of the enzyme, and similar features have been found in other hyperthermophilic enzymes.

ADH (alcohol dehydrogenase) The ADHs (EC 1.1.1.1) belong to the oxidoreductase family, a class of enzymes responsible for the catalysis of all biological oxidation-reduction reactions. Alcohol dehydrogenases catalyse the cofactor-dependent interconversion of alcohols to the corresponding aldehyde or ketone. The enzymes are widely distributed throughout the three domains of life, Archaea, Bacteria and eukaryotes. The ADHs are generally separated into three distinct classes based on molecular size [6]. The

Thermophiles 2003

Figure 3 The active site of the A. pernix ADH with the amino acid chain depicted as a ribbon diagram The catalytic zinc ion is shown in pink with the zinc-coordinating residues in yellow. The bound inhibitor, octanoic acid, is shown in lilac, and the cofactor NADH is shown in pale blue. Figure produced using Bobscript [15] and rendered with Raster3D [16].

type I, or medium-chain, ADHs are the most common form of the enzyme, and the most extensively studied. Enzymes of this class may be dimeric or tetrameric, with a monomer size of approx. 370 residues, and all contain bound zinc ions. The type I ADHs include the mammalian ADH enzymes, with horse liver ADH being a well-studied example [7]. In general, bacterial type-I ADHs are tetramers, while those from higher eukaryotes are dimeric enzymes. Short-chain type II ADHs and long-chain type III ADH enzymes are also found in many organisms. The structures of two hyperthermophilic archaeal ADH enzymes have recently been determined in the apo and holo forms. Both enzymes are type I zinc-containing ADHs, both are tetrameric enzymes and specific for the cofactor NADH. The Sulfolobus solfataricus enzyme structure has been determined in the apo form [8], as has the Aeropyrum pernix ADH enzyme in the holo form with the inhibitor octanoic acid bound in the active site [9]. The sequence identity between the two enzymes is 39%. The Aeropyrum pernix ADH shares 24–25% identity with the eukaryotic horse liver and human ADH enzymes. The sequence identity is 28% to the ADH from the thermophilic bacterium Thermoanaerobium brockii and is lower when compared with other bacterial species. The Aeropyrum ADH is very thermostable, with a half-life for activity of over 2 h at 90◦ C. The structural studies show the archaeal enzymes to be tetramers with the four identical monomers each made up of the characteristic catalytic and cofactor binding domains.

There are two zinc ions found in the catalytic domain of the monomer, one of which plays a catalytic role and the other a structural role. The two archaeal enzyme structures can be superimposed and show that the inter-domain cleft is reduced in the Aeropyrum holo enzyme causing a domain displacement compared to the apo Sulfolobus enzyme. In the Aeropyrum enzyme the catalytic zinc is co-ordinated to the three amino acids Cys-54, His-79 and Asp-168. In addition the inhibitor octanoic acid forms the fourth member of the zinc’s tetrahedral geometry, as shown in Figure 3. The Sulfolobus apo enzyme shows an additional glutamic acid in place of the inhibitor. There is an equivalent amino acid Glu-80 in the Aeropyrum ADH which is at a distance of ˚ from the catalytic zinc, with the increased separation 5.7 A ˚ movement of the zinc between the two archaeal due to a 3 A enzyme structures. The second zinc ion is found in a loop of the catalytic domain and is thought to play a structural role. Bacterial ADH enzymes do not contain this second zinc ion. The structural zinc in the Aeropyrum ADH is co-ordinated by three cysteine residues and an aspartic acid residue. In the solved structure this second zinc was not found in all of the monomers and in its absence Cys-123 was found to move into a second position, allowing formation of a disulphide bond with Cys-115. Again, as in GADPH the formation of the disulphide bond would act to stabilize the enzyme. A recent study has predicted that up to 40% of intracellular cysteine in A. pernix may be in disulphide bonds [10]. The  C 2004

Biochemical Society

257

258

Biochemical Society Transactions (2004) Volume 32, part 2

important role of sulphur in biology is the subject of a recent review [11] and it is clear that its role in enzyme stability and catalytic regulation is only beginning to be understood. The archaeal hyperthermophilic ADH enzymes also show ion-pair networks at the interface of the A/B subunits. The enzyme tetramer is made up of a dimer of dimers, AB and CD. Extensive hydrophobic interactions also occur at the subunit interfaces and clearly contribute to the stability of the tetramer of these hyperthermostable ADH enzymes. This is reinforced when they are compared with the less stable mesophilic horse liver ADH which has no intersubunit ionic bonds and fewer hydrophobic interactions.

Discussion Several questions arise as part of these studies. Clearly the GAPDH enzymes have evolved independently in the Archaea and despite being low in sequence identity to other GAPDH enzymes they have a similar quaternary structure and protein scaffold. However with the exception of the catalytic cysteine and its neighbouring serine the active site is relocated on different secondary structural elements. It is thought that the preference for the cofactor NADP+ over NAD+ is due to its role in the archaeal cell being mainly involved in gluconeogenesis. The presence of disulphide bridges in both the GAPDH and ADH discussed in this paper is an interesting observation. Despite previous assumptions that they would not be found in cytoplasmic proteins, this finding is in agreement with other results from other hyperthermophilic archaeal proteins whose structures also reveal the presence of disulphide bridges that have been implicated in stabilization and control of enzymic activity [12,13]. The study of the hyperthermophilic archaeal dehydrogenase enzymes described in this paper supports the idea that the Archaea are representative of primordial organisms. It also provides an understanding of the mechanisms involved

 C 2004

Biochemical Society

in the ability of these dehydrogenase enzymes to withstand and operate at the high temperatures required of them within organisms thriving in conditions ranging from 80 to 90◦ C. This work was supported by the Biotechnology and Biological Sciences Research Council as a CASE studentship to J.E.G. (with Chirotech Technology, Cambridge, U.K.) and a postdoctoral fellowship to M.N.I.

References 1 Ferri, G., Stoppini, M., Meloni, M.L., Zapponi, M.C. and Iadarola, P. (1990) Biochim. Biophys. Acta 1041, 36–42 2 Isupov, M.N., Fleming, T.M., Dalby, A.R., Crowhurst, G.S., Bourne, P.C. and Littlechild, J.A. (1999) J. Mol. Biol. 291, 651–660 3 Charron, C., Talfournier, F., Isupov, M.N., Littlechild, J.A., Branlant, G., Vitoux, B. and Aubry, A. (2000) J. Mol. Biol. 297, 481–500 4 Jones, C.E., Fleming, T.M., Cowan, D.A., Littlechild, J.A. and Piper, P.W. (1995) Eur. J. Biochem. 233, 800–808 5 Talfournier, F., Colloc’h, N., Mornon, J.-P. and Branlant, G. (1998) Eur. J. Biochem. 252, 447–457 6 Reid, M.F. and Fewson, C.A. (1994) Crit. Rev. Microbiol. 20, 13–56 7 Eklund, H., Nordstrom, B., Zeppezauer, E., Soderlund, G., Ohlsson, I., Boiwe, T., Soderberg, B., Tapia, O., Branden, C.I. and Akeson, A. (1976) J. Mol. Biol. 102, 27–59 8 Esposito, L., Sica, F., Raia, C.A., Giordano, A., Rossi, M., Mazzarella, L. and Zagari, A. (2002) J. Mol. Biol. 318, 463–477 9 Guy, J.E., Isupov, M.N. and Littlechild, J.A. (2003) J. Mol. Biol. 331, 1041–1051 10 Mallick, P., Boutz, D.R., Eisenberg, D. and Yeates, T.O. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 9679–9684 11 Giles, N.M., Watts, A.B., Giles, G.I., Fry, F.H., Littlechild, J.A. and Jacob, C. (2003) Chem. Biol. 10, 677–693 12 Singleton, M.R., Isupov, M.N. and Littlechild, J.A. (1999) Structure 7, 237–244 13 Toogood, H.S., Smith, C.A., Baker, E.N. and Daniel, R.M. (2000) Biochem. J. 350, 321–328 14 Skarzynski, T., Moody, P.C.E. and Wonacott, J.A. (1987) J. Mol. Biol. 193, 171–187 15 Esnouf, R.M. (1997) J. Mol. Graph. Model. 15, 132–134 16 Merritt, E.A. and Bacon, D.J. (1997) Methods Enzymol. 277, 505–524 Received 19 September 2003