Selection of High-Affinity Peptidic Serine Protease Inhibitors with Increased Binding Entropy from a Back-Flip Library of Peptide-Protease Fusions.

We have developed a new concept for designing peptidic protein modulators, by recombinantly fusing the peptidic modulator, with randomized residues, directly to the target protein via a linker and screening for internal modulation of the activity of the protein. We tested the feasibility of the concept by fusing a 10-residue-long, disulfide-bond-constrained inhibitory peptide, randomized in selected positions, to the catalytic domain of the serine protease murine urokinase-type plasminogen activator. High-affinity inhibitory peptide variants were identified as those that conferred to the fusion protease the lowest activity for substrate hydrolysis. The usefulness of the strategy was demonstrated by the selection of peptidic inhibitors of murine urokinase-type plasminogen activator with a low nanomolar affinity. The high affinity could not have been predicted by rational considerations, as the high affinity was associated with a loss of polar interactions and an increased binding entropy.

Distinctive binding modes and inhibitory mechanisms of two peptidic inhibitors of urokinase-type plasminogen activator with isomeric P1 residues.

Two isomeric piperidine derivatives (meta and para isomers) were used as arginine mimics in the P1 position of a cyclic peptidic inhibitor (CPAYSRYLDC) of urokinase-type plasminogen activator. The two resulting cyclic peptides showed vastly different affinities (∼70 fold) to the target enzyme. X-ray crystal structure analysis showed that the two P1 residues were inserted into the S1 specificity pocket in indistinguishable manners. However, the rest of the peptides bound in entirely different ways on the surface of the enzyme, and the two peptides have different conformations, despite the highly similar sequence. These results demonstrate how the subtle difference in P1 residue can dictate the exosite interactions and the potencies of peptidic inhibitors, and highlight the importance of the P1 residue for protease inhibition. This study provides important information for the development of peptidic agents for pharmacological intervention.

A cyclic peptidic serine protease inhibitor: increasing affinity by increasing peptide flexibility.

Peptides are attracting increasing interest as protease inhibitors. Here, we demonstrate a new inhibitory mechanism and a new type of exosite interactions for a phage-displayed peptide library-derived competitive inhibitor, mupain-1 (CPAYSRYLDC), of the serine protease murine urokinase-type plasminogen activator (uPA). We used X-ray crystal structure analysis, site-directed mutagenesis, liquid state NMR, surface plasmon resonance analysis, and isothermal titration calorimetry and wild type and engineered variants of murine and human uPA. We demonstrate that Arg6 inserts into the S1 specificity pocket, its carbonyl group aligning improperly relative to Ser195 and the oxyanion hole, explaining why the peptide is an inhibitor rather than a substrate. Substitution of the P1 Arg with novel unnatural Arg analogues with aliphatic or aromatic ring structures led to an increased affinity, depending on changes in both P1 - S1 and exosite interactions. Site-directed mutagenesis showed that exosite interactions, while still supporting high affinity binding, differed substantially between different uPA variants. Surprisingly, high affinity binding was facilitated by Ala-substitution of Asp9 of the peptide, in spite of a less favorable binding entropy and loss of a polar interaction. We conclude that increased flexibility of the peptide allows more favorable exosite interactions, which, in combination with the use of novel Arg analogues as P1 residues, can be used to manipulate the affinity and specificity of this peptidic inhibitor, a concept different from conventional attempts at improving inhibitor affinity by reducing the entropic burden.

Allosteric inactivation of a trypsin-like serine protease by an antibody binding to the 37- and 70-loops.

Serine protease catalytic activity is in many cases regulated by conformational changes initiated by binding of physiological modulators to exosites located distantly from the active site. Inhibitory monoclonal antibodies binding to such exosites are potential therapeutics and offer opportunities for elucidating fundamental allosteric mechanisms. The monoclonal antibody mU1 has previously been shown to be able to inhibit the function of murine urokinase-type plasminogen activator in vivo. We have now mapped the epitope of mU1 to the catalytic domain's 37- and 70-loops, situated about 20 Å from the S1 specificity pocket of the active site. Our data suggest that binding of mU1 destabilizes the catalytic domain and results in conformational transition into a state, in which the N-terminal amino group of Ile16 is less efficiently stabilizing the oxyanion hole and in which the active site has a reduced affinity for substrates and inhibitors. Furthermore, we found evidence for functional interactions between residues in uPA's C-terminal catalytic domain and its N-terminal A-chain, as deletion of the A-chain facilitates the mU1-induced conformational distortion. The inactive, distorted state is by several criteria similar to the E* conformation described for other serine proteases. Hence, agents targeting serine protease conformation through binding to exosites in the 37- and 70-loops represent a new class of potential therapeutics.

Protein conformational change delayed by steric hindrance from an N-linked glycan.

Very few studies have attributed a direct, active, functional role to N-linked glycans. We describe here an N-linked glycan with a unique role for maintaining the active conformation of a protein of the serpin family. The distinguishing feature of serpins is the "stressed-to-relaxed" transition, in which the reactive center loop inserts as a ß-strand into the central ß-sheet A. This transition forms the basis for the conversion of serpins to the inactive latent state. We demonstrate that plasminogen activator inhibitor-1 (PAI-1) from zebrafish converts to the latent state about 5-fold slower than human PAI-1. In contrast to human PAI-1, fish PAI-1 carries a single N-linked glycan at Asn185 in the gate region through which the reactive center loop passes during latency transition. While the latency transition of human PAI-1 is unaffected by deglycosylation, deglycosylated zebrafish PAI-1 (zfPAI-1) goes latent about 50-fold faster than the glycosylated zfPAI-1 and about 25-fold faster than non-glycosylated human PAI-1. X-ray crystal structure analysis of glycosylated fish PAI-1 confirmed the presence of an N-linked glycan in the gate region and a lack of glycan-induced structural changes. Thus, latency transition of zfPAI-1 is delayed by steric hindrance from the glycan in the gate region. Our findings reveal a previously unknown mechanism for inhibition of protein conformational changes by steric hindrance from N-linked glycans.

Interconversion of active and inactive conformations of urokinase-type plasminogen activator.

The catalytic activity of serine proteases depends on a salt-bridge between the amino group of residue 16 and the side chain of Asp194. The salt-bridge stabilizes the oxyanion hole and the S1 specificity pocket of the protease. Some serine proteases exist in only partially active forms, in which the amino group of residue 16 is exposed to the solvent. Such a partially active state is assumed by a truncated form of the murine urokinase-type plasminogen activator (muPA), consisting of residues 16-243. Here we investigated the allosteric interconversion between partially active states and the fully active state. Both a monoclonal antibody (mU3) and a peptidic inhibitor (mupain-1--16) stabilize the active state. The epitope of mU3 is located in the 37- and 70-loops at a site homologous to exosite I of thrombin. The N-terminus((Ile16)) of muPA((16--243)) was less exposed upon binding of mU3 or mupain-1--16. In contrast, introduction of the mutations F40Y or E137A into muPA((16--243)) increased exposure of the N-terminus((Ile16)) and resulted in large changes in the thermodynamic parameters for mupain-1--16 binding. We conclude that the distorted state of muPA((16--243)) is conformationally ordered upon binding of ligands to the active site and upon binding of mU3 to the 37- and 70-loops. Our study establishes the 37- and 70-loops as a unique site for binding to compounds stabilizing the active state of serine proteases.

The binding mechanism of a peptidic cyclic serine protease inhibitor.

Serine proteases are classical objects for studies of catalytic and inhibitory mechanisms as well as interesting as therapeutic targets. Since small-molecule serine protease inhibitors generally suffer from specificity problems, peptidic inhibitors, isolated from phage-displayed peptide libraries, have attracted considerable attention. Here, we have investigated the mechanism of binding of peptidic inhibitors to serine protease targets. Our model is upain-1 (CSWRGLENHRMC), a disulfide-bond-constrained competitive inhibitor of human urokinase-type plasminogen activator with a noncanonical inhibitory mechanism and an unusually high specificity. Using a number of modified variants of upain-1, we characterised the upain-1-urokinase-type plasminogen activator complex using X-ray crystal structure analysis, determined a model of the peptide in solution by NMR spectroscopy, and analysed binding kinetics and thermodynamics by surface plasmon resonance and isothermal titration calorimetry. We found that upain-1 changes both main-chain conformation and side-chain orientations as it binds to the protease, in particular its Trp3 residue and the surrounding backbone. The properties of upain-1 are strongly influenced by the addition of three to four amino acids long N-terminal and C-terminal extensions to the core, disulfide-bond-constrained sequence: The C-terminal extension stabilises the solution structure compared to the core peptide alone, and the protease-bound structure of the peptide is stabilised by intrapeptide contacts between the N-terminal extension and the core peptide around Trp3. These results provide a uniquely detailed description of the binding of a peptidic protease inhibitor to its target and are of general importance in the development of peptidic inhibitors with high specificity and new inhibitory mechanisms.

Elucidation of the contribution of active site and exosite interactions to affinity and specificity of peptidylic serine protease inhibitors using non-natural arginine analogs.

There is increasing interest in developing peptides for pharmacological intervention with pathophysiological functions of serine proteases. From phage-displayed peptide libraries, we previously isolated peptidylic inhibitors of urokinase-type plasminogen activator, a potential target for intervention with cancer invasion. The two peptides, upain-1 (CSWRGLENHRMC) and mupain-1 (CPAYSRYLDC), are competitive inhibitors of human and murine urokinase-type plasminogen activator, respectively. Both have an Arg as the P1 residue, inserting into the S1 pocket in the active site of the enzymes, but their specificity depends to a large extent on interactions outside the enzymes' active sites, so-called exosite interactions. Here we describe upain-2 (CSWRGLENHAAC) and the synthesis of a number of upain-2 and mupain-1 variants in which the P1 Arg was substituted with novel non-natural Arg analogs and achieved considerable improvement in the affinity of the peptides to their targets. Using chimeras of human and murine urokinase-type plasminogen activator as well as X-ray crystallography, we delineated the relative contribution of the P1 residue and exosite interactions to the affinity and specificity of the inhibitors for their target enzyme. The effect of inserting a particular non-natural amino acid into the P1 position is determined by the fact that changes in interactions of the P1 residue in the S1 pocket lead to changed exosite interactions and vice versa. These findings are of general interest when the affinities and specificities of serine protease inhibitors to be used for pharmacological intervention are considered and could pave the way for potential drug candidates for the treatment of cancer.

Towards universal systems for recombinant gene expression.

Recombinant gene expression is among the most important techniques used both in molecular and medical research and in industrial settings. Today, two recombinant expression systems are particularly well represented in the literature reporting on recombinant expression of specific genes. According to searches in the PubMed citation database, during the last 15 years 80% of all recombinant genes reported on in the literature were expressed in either the enterobacterium Escherichia coli or the methylotropic yeast Pichia pastoris. Nevertheless, some eukaryotic proteins are misfolded or inadequately posttranslationally modified in these expression systems. This situation demands identification of other recombinant expression systems that enable the proper expression of the remaining eukaryotic genes. As of now, a single universal system allowing expression of all target genes is still a distant goal. In this light, thorough experimental screening for systems that can yield satisfying quantity and quality of target protein is required. In recent years, a number of new expression systems have been described and used for protein production. Two systems, namely Drosophila melanogaster S2 insect cells and human embryonic kidney 293 (HEK293) cells stably expressing the EBNA-1 gene, show exceptional promise. The time has come to identify a few well-performing systems that will allow us to express, purify, and characterize entire eukaryotic genomes.

Oat (Avena sativa) seed extract as an antifungal food preservative through the catalytic activity of a highly abundant class I chitinase.

Extracts from different higher plants were screened for the ability to inhibit the growth of Penicillium roqueforti, a major contaminating species in industrial food processing. Oat (Avena sativa) seed extracts exhibited a high degree of antifungal activity and could be used directly on rye bread to prevent the formation of P. roqueforti colonies. Proteins in the oat seed extracts were fractionated by column chromatography and proteins in fractions containing antifungal activity were identified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) and database searches. Identified antifungal candidates included thaumatin-like proteins, 1,3-beta-glucanase, permatin precursor, pathogenesis-related protein type 1, and chitinases of class I and II. Class I chitinase could be specifically removed from the extracts and was found to be indispensable for 50% of the P. roqueforti inhibiting activity. The purified class I chitinase has a molecular weight of approximately 34 kDa, optimal chitinase activity at pH 7, and exists as at least two basic isoforms (pI values of 7.6 and 8.0). Partial sequencing of the class I chitinase isoforms by LC-MS/MS revealed a primary structure with high similarity to class I chitinases of wheat (Triticum aestivum), barley (Hordeum vulgare), and rye (Secale cereale). Oat, wheat, barley, and rye seed extracts were compared with respect to the abundance of the class I chitinase and decrease in antifungal activity when class I chitinase is removed. We found that the oat seed class I chitinase is at least ten times more abundant than the wheat, barley, and rye homologs and that oat seed extracts are highly active toward P. roqueforti as opposed to extracts of other cereal seeds.

Heparan sulfate regulates ADAM12 through a molecular switch mechanism.

The disintegrin and metalloproteases (ADAMs) are emerging as therapeutic targets in human disease, but specific drug design is hampered by potential redundancy. Unlike other metzincins, ADAM prodomains remain bound to the mature enzyme to regulate activity. Here ADAM12, a protease that promotes tumor progression and chondrocyte proliferation in osteoarthritic cartilage, is shown to possess a prodomain/catalytic domain cationic molecular switch, regulated by exogenous heparan sulfate and heparin but also endogenous cell surface proteoglycans and the polyanion, calcium pentosan polysulfate. Sheddase functions of ADAM12 are regulated by the switch, as are proteolytic functions in placental tissue and sera of pregnant women. Moreover, human heparanase, an enzyme also linked to tumorigenesis, can promote ADAM12 sheddase activity at the cell surface through cleavage of the inhibitory heparan sulfate. These data present a novel concept that might allow targeting of ADAM12 and suggest that other ADAMs may have specific regulatory activity embedded in their prodomain and catalytic domain structures.

Expression, purification and insights into structure and folding of the ADAM22 pro domain.

The ADAMs (a disintegrin and metalloproteases) are an important class of enzymes in the regulation of human disease. The pro domains of ADAMs are responsible for the latency and secretion of mature enzymes. Unlike other metzincins, ADAM pro domains remain bound to the mature enzyme after secretion. To understand the functions of human ADAM pro domains and to determine three-dimensional structures, we have screened promising targets for expression and purification properties when using Escherichia coli as the host. The pro domain of ADAM22 (ADAM22-P) expressed in E. coli was folded, as determined by CD and NMR spectroscopy. An ADAM22-P fragment encoding residues 26-199 could be expressed in high amounts, remained soluble above 1 mM, and was suitable for structural studies by NMR spectroscopy. CD spectroscopy and predictions suggest that the secondary structure in ADAM22-P consists of beta-strands. Furthermore, our data indicate that the pro domains of ADAMs are expressed as two subdomains. The most N-terminal subdomain (ADAM22-P(N)) was found to be susceptible to proteolysis and was required for folding stability of the second subdomain (ADAM22-P(C)).

Catalytic properties of ADAM12 and its domain deletion mutants.

Human ADAM12 (a disintegrin and metalloproteinase) is a multidomain zinc metalloproteinase expressed at high levels during development and in human tumors. ADAM12 exists as two splice variants: a classical type 1 membrane-anchored form (ADAM12-L) and a secreted splice variant (ADAM12-S) consisting of pro, catalytic, disintegrin, cysteine-rich, and EGF domains. Here we present a novel activity of recombinant ADAM12-S and its domain deletion mutants on S-carboxymethylated transferrin (Cm-Tf). Cleavage of Cm-Tf occurred at multiple sites, and N-terminal sequencing showed that the enzyme exhibits restricted specificity but a consensus sequence could not be defined as its subsite requirements are promiscuous. Kinetic analysis revealed that the noncatalytic C-terminal domains are important regulators of Cm-Tf activity and that ADAM12-PC consisting of the pro domain and catalytic domain is the most active on this substrate. It was also observed that NaCl inhibits ADAM12. Among the tissue inhibitors of metalloproteinases (TIMP) examined, the N-terminal domain of TIMP-3 (N-TIMP-3) inhibits ADAM12-S and ADAM12-PC with low nanomolar Ki(app) values while TIMP-2 inhibits them with a slightly lower affinity (9-44 nM). However, TIMP-1 is a much weaker inhibitor. N-TIMP-3 variants that lack MMP inhibitory activity but retained the ability to inhibit ADAM17/TACE failed to inhibit ADAM12. These results indicate unique enzymatic properties of ADAM12 among the members of the ADAM family of metalloproteinases.

Generation of monoclonal antibodies for the assessment of protein purification by recombinant ribosomal coupling.

We recently described a conceptually novel method for the purification of recombinant proteins with a propensity to form inclusion bodies in the cytoplasm of Escherichia coli. Recombinant proteins were covalently coupled to the E. coli ribosome by fusing them to ribosomal protein 23 (rpL23) followed by expression in an rpL23 deficient strain of E. coli. This allowed for the isolation of ribsomes with covalently coupled target proteins which could be efficiently purified by centrifugation after in vitro proteolysis at a specific site incorporated between rpL23 and the target protein. rpL23-GFP-His is among the fusion proteins used in our previous study for ribosomal coupling of C-terminally His-tagged green fluorescent protein. To assess the efficiency of separation of target protein from ribosomes, by site-specific proteolysis, we required monoclonal antibodies directed against rpL23 and GFP. We therefore purified rpL23-GFP-His, rpL23-His and GFP from E. coli recombinants using affinity, ion exchange and hydrophobic interaction chromatography. These proteins could be purified with yields of 150, 150 and 1500 microg per gram cellular wet weight, respectively. However, rpL23-GFP-His could only be expressed in a soluble form and subsequently purified, when cells were cultivated at reduced temperatures. The purified rpL23-GFP-His fusion protein was used to immunize balb/c mice and the hybridoma cell lines resulting from in vitro cell fusion were screened by ELISA using rpL23-His and GFP to select for monoclonal antibodies specific for each protein. This resulted in 20 antibodies directed against rpL23 and 3 antibodies directed against GFP. Antibodies were screened for isotypes and their efficiency in western immunoblots. The most efficient antibody against rpL23 and GFP were purified by Protein G Sepharose affinity chromatography. The purified antibodies were used to evaluate the separation of ribosomes from GFP, streptavidin, murine interleukin-6, a phagedisplay antibody and yeast elongation factor 1A by centrifugation, when ribosomes with covalently coupled target protein were cleaved at specific proteolytic cleavage sites. We conclude that the generated antibodies can be used to evaluate ribosomal coupling of recombinant target proteins as well as the efficiency of their separation from the ribosome.

Initiation of protein synthesis in bacteria.

Valuable information on translation initiation is available from biochemical data and recently solved structures. We present a detailed description of current knowledge about the structure, function, and interactions of the individual components involved in bacterial translation initiation. The first section describes the ribosomal features relevant to the initiation process. Subsequent sections describe the structure, function, and interactions of the mRNA, the initiator tRNA, and the initiation factors IF1, IF2, and IF3. Finally, we provide an overview of mechanisms of regulation of the translation initiation event. Translation occurs on ribonucleoprotein complexes called ribosomes. The ribosome is composed of a large subunit and a small subunit that hold the activities of peptidyltransfer and decode the triplet code of the mRNA, respectively. Translation initiation is promoted by IF1, IF2, and IF3, which mediate base pairing of the initiator tRNA anticodon to the mRNA initiation codon located in the ribosomal P-site. The mechanism of translation initiation differs for canonical and leaderless mRNAs, since the latter is dependent on the relative level of the initiation factors. Regulation of translation occurs primarily in the initiation phase. Secondary structures at the mRNA ribosomal binding site (RBS) inhibit translation initiation. The accessibility of the RBS is regulated by temperature and binding of small metabolites, proteins, or antisense RNAs. The future challenge is to obtain atomic-resolution structures of complete initiation complexes in order to understand the mechanism of translation initiation in molecular detail.

Soluble expression of recombinant proteins in the cytoplasm of Escherichia coli.

Pure, soluble and functional proteins are of high demand in modern biotechnology. Natural protein sources rarely meet the requirements for quantity, ease of isolation or price and hence recombinant technology is often the method of choice. Recombinant cell factories are constantly employed for the production of protein preparations bound for downstream purification and processing. Eschericia coli is a frequently used host, since it facilitates protein expression by its relative simplicity, its inexpensive and fast high density cultivation, the well known genetics and the large number of compatible molecular tools available. In spite of all these qualities, expression of recombinant proteins with E. coli as the host often results in insoluble and/or nonfunctional proteins. Here we review new approaches to overcome these obstacles by strategies that focus on either controlled expression of target protein in an unmodified form or by applying modifications using expressivity and solubility tags.

Advanced genetic strategies for recombinant protein expression in Escherichia coli.

Preparations enriched by a specific protein are rarely easily obtained from natural host cells. Hence, recombinant protein production is frequently the sole applicable procedure. The ribosomal machinery, located in the cytoplasm is an outstanding catalyst of recombinant protein biosynthesis. Escherichia coli facilitates protein expression by its relative simplicity, its inexpensive and fast high-density cultivation, the well-known genetics and the large number of compatible tools available for biotechnology. Especially the variety of available plasmids, recombinant fusion partners and mutant strains have advanced the possibilities with E. coli. Although often simple for soluble proteins, major obstacles are encountered in the expression of many heterologous proteins and proteins lacking relevant interaction partners in the E. coli cytoplasm. Here we review the current most important strategies for recombinant expression in E. coli. Issues addressed include expression systems in general, selection of host strain, mRNA stability, codon bias, inclusion body formation and prevention, fusion protein technology and site-specific proteolysis, compartment directed secretion and finally co-overexpression technology. The macromolecular background for a variety of obstacles and genetic state-of-the-art solutions are presented.

Soluble expression of aggregating proteins by covalent coupling to the ribosome.

Ribosomes are extremely soluble ribonucleoprotein complexes. Heterologous target proteins were fused to ribosomal protein L23 (rpL23) and expressed in an rpL23 deficient Escherichia coli strain. This enabled the isolation of 70S ribosomes with covalently bound target protein. Isolation of recombinant proteins from 70S ribosomes was achieved by specific proteolytic cleavage followed by efficient removal of ribosomes by centrifugation. By this procedure we isolated active green fluorescent protein, streptavidin (SA), and murine interleukin-6 (mIL-6). Approximately 500microg of each protein was isolated per gram cellular wet weight. By pull-down assays we demonstrate that SA covalently bound to the ribosome binds d-biotin. Ribosomal coupling is therefore suggested as a method for the investigation of protein interactions. The presented strategy is in particular efficient for the expression, purification, and investigation of proteins forming inclusion bodies in the E. coli cytoplasm.

Dialysis strategies for protein refolding: preparative streptavidin production.

We have investigated different dialysis strategies for the refolding of recombinant streptavidin, and present a novel dialysis setup featuring gradual dilution dialysis and continuous protein feeding into a dialysis sack. A denaturing dialysis buffer is exchanged gradually by dilution with refolding buffer and it is demonstrated that the refolding yield can be increased from 45 to 75% by lowering the dilution rate. In addition, continuous feeding of protein to the dialysis sack increases the yield by 5 to 10%. The principle of gradual dilution dialysis is amenable to stringent regulation and we suggest it to be applied for other insoluble protein targets.

Production of recombinant thermostable proteins expressed in Escherichia coli: completion of protein synthesis is the bottleneck.

Heterologous expression and high yield purification of proteins are frequently required for structural and functional investigations. Purification of recombinant thermostable proteins is essentially trivial since unwanted mesophilic host protein can efficiently be removed by heat denaturation. However, heterologous expression in E. coli often results in truncated protein forms. In many cases, this is a consequence of abundant codons in heterologous genes, which are decoded by rare tRNAs in E. coli-a combination that can be responsible for translational stalling and termination during protein biosynthesis. Other complications may originate from potential initiation codons and ribosomal binding sites present inside the open reading frame of the target gene or from other less well defined phenomena such as mRNA instability. Separation of full-length protein from truncated forms is a serious chromatographic problem that can be solved in the expression step. We have investigated the heterologous expression and purification of two translation initiation factors from the hyperthermophilic sulphate-reducing archaeon, Archaeoglobus fulgidus. Expression in E. coli was optimised to avoid truncated forms completely by complementation with the plasmids pSJS1244, pRIG, pCODON+ and pLysSR.A.R.E harbouring and expressing genes encoding rare tRNAs corresponding to the codons AGA, AGG, AUA, CUA, GGA, AAG and CCC. Two expression strains, C41(DE3) and C43(DE3) were found highly advantageous when combined with rare tRNA encoding plasmids as compared to BL21(DE3). We have also investigated the effects of site directed mutagenesis on rare lysine encoding AAG doublets as well as two methionine residues preceded by potential ribosomal binding sites. The expression approach presented here has enabled us to purify gram quantities of full-length protein by one step of ion-exchange chromatography and is generally applicable to many other heterologously expressed thermostable proteins.