Engineered human carboxypeptidase B enzymes that hydrolyse hippuryl-L- glutamic acid: reversed-polarity mutants

Variants of human pancreatic carboxypeptidase B (HCPB), with specificity for hydrolysis of C-terminal glutamic acid and aspartic acid, were prepared by site-directed mutagenesis of the human gene and expressed in the periplasm of Escherichia coli. By changing residues in the lining of the S1' pocket of the enzyme, it was possible to reverse the substrate specificity to give variants able to hydrolyse prior to C- terminal acidic amino acid residues instead of the normal C-terminal basic residues. This was achieved by mutating Asp253 at the base of the S1' specificity pocket, which normally interacts with the basic side- chain of the substrate, to either Lys or Arg. The resulting enzymes had the desired reversed polarity and enzyme activity was improved significantly with further mutations at residue 251. The [G251T,D253K]HCPB double mutant was 100 times more active against hippuryl-L-glutamic acid (hipp-Glu) as substrate than was the single mutant, [D253K]HCPB. Triple mutants, containing additional changes at Ala248, had improved activity against hipp-Glu substrate when position 251 was Asn. These reversed-polarity mutants of a human enzyme have the potential to be used in antibody-directed enzyme prodrug therapy of cancer.      

In vivo characterization of tandem C-terminal thioesterase domains in arthrofactin synthetase

Macrocyclization of a peptide or a lipopeptide occurs at the last step of synthesis and is usually catalyzed by a single C-terminal thioesterase (Te) domain. Arthrofactin synthetase (Arf) from Pseudomonas sp. MIS38 represents a novel type of nonribosomal peptide synthetase that contains unique tandem C-terminal Te domains, ArfC_Te1 and ArfC_Te2. In order to analyze their function in vivo, site-directed mutagenesis was introduced at the putative active-site residues in ArfC_Te1 and ArfC_Te2. It was found that both Te domains were functional. Peaks corresponding to arthrofactin and its derivatives were absent in ArfC_Te1:S89A, ArfC_Te1:S89T, and ArfC_Te1:E26G/F27A mutants, and the production of arthrofactin by ArfC_Te2:S92A, ArfC_Te2:S92A/D118A, and ArfCDeltaTe2 was reduced by 95 % without an alteration of the cyclic lipoundecapeptide structure. These results suggest that Ser89 in ArfC_Te1 is essential for the completion of macrocyclization and the release of product. Glu26 and Phe27 residues are also part of the active site of ArfC_Te1. ArfC_Te2 might have been added during the evolution of Arf in order to improve macrocyclization efficiency.     

The substrate specificity of a recombinant cysteine protease from Leishmania mexicana: application of a combinatorial peptide library approach

The substrate specificity of CPB2.8DeltaCTE, a recombinant cysteine protease from Leishmania mexicana, was mapped by screening a fluorescence-quenched combinatorial peptide library. Results from library screening indicated a preference for Arg or Lys in the S(3) subsite and for hydrophobic residues, both aliphatic and aromatic, in S(2). The S(1) subsite exhibited a specificity for the basic residues Arg and Lys. Generally, the specificity of the primed subsites was less strict compared with the non-primed side which showed preference for Arg, Lys and Ala in S'(1), Arg, Pro and Gly in S'(2) and Lys, Arg and Ser in S'(4). By contrast, a strict preference for the basic residues Arg and Lys was found for S'(3). Overall, there was a trend for basic residues in alternating subsites and smaller residues in the primed sites compared with the non-primed sites. In addition, there were strict requirements for the amino acids in subsites S(3)--S(1). Fluorescence-quenched peptides from the library with the highest on-resin cleavage were resynthesised and their kinetics of hydrolysis by CPB2.8DeltaCTE assessed in solution phase assays. Several good substrates containing the quintessential dipeptide particular to cathepsin-L-like enzymes, -F-R/K-, in P(2) and P(1) were identified (e.g. Y(NO(2))-EKFR down arrow RGK-K(Abz)G, Abz=2-aminobenzoyl; k(cat)K(m)(-1)=4298 mM(-1)s(-1)). However, novel substrates containing the dipeptide -L/I-Q- in P(2) and P(1) were also well hydrolysed (e.g. Y(NO(2))-YLQ down arrow GIQK-K(Abz)G; k(cat)K(m)(-1)=2583 mM(-1)s(-1)). The effect of utilising different fluorescent donor--quencher pairs on the value of k(cat)K(m)(-1) was examined. Generally, the use of the Abz/Q-EDDnp donor--quencher pair (EDDnp=N-(2,4-dinitrophenyl)ethylenediamine) instead of K(Abz)/Y(NO(2)) resulted in higher k(cat)K(m)(-1) values for analogous substrates.     

Domain-selective ligand-binding modes and atomic level pharmacophore refinement in angiotensin I converting enzyme (ACE) inhibitors

Somatic ACE (EC 3.4.15.1), a Zn(II) metalloproteinase, is composed of functionally active N and C domains resulting from tandem gene duplication. Despite the high degree of sequence similarity between the two domains, they differ in substrate and inhibitor specificity and in their activation by chloride ions. Because of the critical role of ACE in cardiovascular and renal diseases, both domains are attractive targets for drug design. Putative structural models have been generated for the interactions of ACE inhibitors (lisinopril, captoril, enalaprilat, keto-ACE, ramiprilat, quinaprilat, peridoprilat, fosinoprilat, and RXP 407) with both the ACE_C and the ACE_N domains. Inhibitor-domain selectivity was interpreted in terms of residue alterations observed in the four subsites of the binding grooves of the ACE_C/ACE_N domains (S1: V516/N494, V518/T496, S2: F391/Y369, E403/R381, S1': D377/Q355, E162/D140, V379/S357, V380/T358, and S2': D463/E431, T282/S260). The interactions governing the ligand-receptor recognition process in the ACE_C domain are: a salt bridge between D377, E162, and the NH(2) group (P1' position), a hydrogen bond of the inhibitor with Q281, the presence of bulky hydrophobic groups in the P1 and P2' sites, and a stacking interaction of F391 with a benzyl group in the P2 position. In ACE_N these interactions are: hydrogen bonds of the inhibitor with E431, Y369, and R381, and a salt bridge between the carboxy group in the P2 position of the inhibitor and R500. The calculated complexes were evaluated for their consistency with structure-activity relationships and site-directed mutagenesis data. A comparison between the calculated interaction free energies and the experimentally observed biological activities was also made. Pharmacophore refinement was achieved at an atomic level, and might provide an improved basis for structure-based rational design of second-generation, domain-selective inhibitors.     

Alteration of substrate specificity of rat neurolysin from matrix metalloproteinase-2/9-type to -3-type specificity by comprehensive mutation

The substrate specificity of rat brain neurolysin was rapidly modified by semirational mutagenesis coupled with a yeast molecular display system. Neurolysin mainly recognizes substrates with sequential six residues close to the scissile bond in polypeptides, cleaving a peptide bond in the center position of the six residues. To alter the recognition of the P2' amino acid of substrates by neurolysin, six residues of neurolysin, Asp467, Arg470, Glu510, Tyr606, Tyr610 and Tyr611, which might be involved in the formation of the neurolysin S2' subsite, were individually and comprehensively substituted. The protein libraries of mutant neurolysins comprising 120 species were displayed on the yeast cell surface and screening was carried out using two fluorescence-quenching peptides, the matrix metalloproteinase-2/9- (MMPs-2/9-) and MMP-3-specific substrates, which consisted of similar amino acids, except for alanine (for MMPs-2/9) or glutamic acid (for MMP-3) at the P2' amino acid position. Among mutant neurolysins, the Y610L mutant neurolysin exhibited a marked change in substrate specificity. Steady-state kinetic analysis of the purified Y610L mutant neurolysin revealed that the binding efficiency toward the MMP-3-specific substrate was about 3-fold higher than that toward the MMP-2/9-specific substrate. These results indicate that Tyr610 of neurolysin is the important residue to recognize the P2' amino acid of substrates.     

Redesign of the substrate specificity of human cathepsin D: the dominant role of position 287 in the S2 subsite

Interest in the active site specificity of human cathepsin Dstems from the search for specific therapeutic agents againstmany of the sequentially and structurally homologous membersofthe aspartic proteinase family. The work presented here examinedone amino acid in the cathepsin D sequence, located in the S₂subsite, which contributes substantially to the specificityof enzyme-Ugand interactions at the enzyme active site. Previousstudies reported on the specificity of binding and catalysisby native and recombinant human cathepsin D explored throughkinetic studies using a systematic series of synthetic substrates.Utilizing a rulebased molecular model of human cathepsin D,Met287 was suggested as a candidate for mutagenesis to furtherexplore selectivity within the S₂ subsite of the cathepsin Dactive site. Met287 mutant derivatives of human cathepsin Dwere designed, expressed and characterized in kineticstudies.Native cathepsin D accommodates large hydrophobic residues inthe P₂ position of a substrate; positively charged residuesin P₂ are not favorable for catalysis.It was demonstrated thataltering Met287 of human cathepsin D to more polar amino acidsproduced active mutant enzymes with significantly altered substratespecificity.     

Fine mapping and structural analysis of immunodominant IgE allergenic epitopes in chicken egg ovalbumin

Ovalbumin is a major allergen in hen egg white that causes IgE-mediatedfood allergic reactions in children. In this study, the immunodominantIgE-binding epitopes of ovalbumin were mapped using arrays ofoverlapping peptides synthesized on activated cellulose membranes.Pooled human sera from 18 patients with egg allergy were usedto probe the membrane. Five distinct regions were found to containdominant allergic IgE epitopes, these being L38T49, D95A102,E191V200, V243E248 and G251N260. The critical amino acids involvedin IgE antibody binding were also determined. These epitopeswere composed primarily of hydrophobic amino acids, followedby polar and charged residues and being comprised of ß-sheetand ß-turn structures. One epitope, D95A102, consistedof a single -helix. These results provide useful informationon the functional role of amino acid residues to evaluate thestructure–function relationships and structural propertiesof allergic epitopes in ovalbumin. They also provide a strategicapproach for engineering ovalbumin to reduce its allergenicity. Received January 9, 2003; revised April 23, 2003; accepted August 28, 2003.     

Theoretical and experimental evaluation of a CYP106A2 low homology model and production of mutants with changed activity and selectivity of hydroxylation

Steroids are important pharmaceutically active compounds. In contrast to the liver drug-metabolising cytochrome P450s, which metabolise a variety of substrates, steroid hydroxylases generally display a rather narrow substrate specificity. It is therefore a challenging goal to change their regio- and stereoselectivity. CYP106A2 is one of only a few bacterial steroid hydroxylases and hydroxylates 3-oxo-Delta4-steroids mainly in 15beta-position. In order to gain insights into the structure and function of this enzyme, whose crystal structure is unknown, a homology model has been created. The substrate progesterone was then docked into the active site to predict which residues might affect substrate binding. The model was substantiated by using a combination of theoretical and experimental investigations. First, numerous computational structure evaluation tools assessed the plausibility of its protein geometry and its quality. Second, the model explains many key properties of common cytochrome P450s. Third, two sets of mutants have been heterologously expressed, and the influence of the mutations on the catalytic activity towards deoxycorticosterone and progesterone has been studied experimentally: the first set comprises six mutations located in the structurally variable regions of this enzyme that are very difficult to predict by cytochrome P450 modelling (K27R, I86T, E90V, I71T, D185G and I215T). For these positions, no participation in the active-site formation was predicted, or could be experimentally demonstrated. The second set comprises five mutants in substrate recognition site 6 (S394I, A395L, T396R, G397P and Q398S). For these residues, participation in active-site formation and an influence on substrate binding was predicted by docking. These mutants are based on an alignment with human CYP11B1, and in fact most of these mutants altered the active-site structure and the hydroxylation activity of CYP106A2 dramatically.     

A low resolution model for the interaction of G proteins with G protein-coupled receptors

A model is presented for the interaction between G proteinsand G protein-coupled receptors. The model is based on the factthat this interaction shows little specificity and thus conservedparts of the G proteins have to interact with conserved partsof the receptors. These parts are a conserved negative residuein the G protein, a fully conserved arginine in the receptorand a series of residues that are not conserved but always hydrophobiclike the hydrophobic side of the C-terminal helix of the G proteinand the hydrophobic side of a helix in the C-terminal domainof the receptor. Other, mainly cytosolic, factors determinethe specificity and regulation of this interaction. The relationbetween binding and activation will be shown. A large body ofexperimental evidence supports this model. Despite the factthat the model does not provide atomic resolution, it can beused to explain some experimental data that would otherwiseseem inexplicable, and it suggests experiments for its falsificationor verification.     

Ala226 to Gly and Ser189 to Asp mutations convert rat chymotrypsin B to a trypsin-like protease

In a previous successful attempt to convert trypsin to a chymotrypsin-like protease, 15 residues of trypsin were replaced with the corresponding ones in chymotrypsin. This suggests a complex mechanism of substrate recognition instead of a relatively simple one that only involves three sites, residues 189, 216 and 226. However, both trypsin-->elastase and chymotrypsin-->trypsin conversion experiments carried out according to the complex model resulted in non-specific proteases with low catalytic activity. Chymotrypsin used in the latter studies was of B-type, containing an Ala residue at position 226. Trypsins, however, contain a conserved Gly at this site. The substantially decreased trypsin-like activity of the G226A trypsin mutant also suggests a specific role for this site in substrate binding. Here we investigate the role of site 226 by introducing the A226G substitution into chymotrypsin-->trypsin mutants which were constructed according to both the simple (S189D mutant) and the complex model (S(1) mutant) of specificity determination. The kinetic parameters show that the A226G substitution in the S(1) mutant increased the chymotrypsin-like activity, while the trypsin-like activity did not change. In contrast, this substitution in the S189D chymotrypsin mutant resulted in a 100-fold increase in trypsin-like activity and a trypsin-like specificity profile as tested on a competing oligopeptide substrate library. Additionally, the S189D+A226G mutant is the first trypsin-like chymotrypsin that undergoes autoactivation, an exclusive property of trypsinogen among pancreatic serine proteases.     

Cis-Configured aziridines are new pseudo-irreversible dual-mode inhibitors of Candida albicans secreted aspartic protease 2

A series of cis-configured epoxides and aziridines containing hydrophobic moieties and amino acid esters were synthesized as new potential inhibitors of the secreted aspartic protease 2 (SAP2) of Candida albicans. Enzyme assays revealed the N-benzyl-3-phenyl-substituted aziridines 11 and 17 as the most potent inhibitors, with second-order inhibition rate constants (k(2)) between 56,000 and 121,000 M(-1) min(-1). The compounds were shown to be pseudo-irreversible dual-mode inhibitors: the intermediate esterified enzyme resulting from nucleophilic ring opening was hydrolyzed and yielded amino alcohols as transition-state-mimetic reversible inhibitors. The results of docking studies with the ring-closed aziridine forms of the inhibitors suggest binding modes mainly dominated by hydrophobic interactions with the S1, S1', S2, and S2' subsites of the protease, and docking studies with the processed amino alcohol forms predict additional hydrogen bonds of the new hydroxy group to the active site Asp residues. C. albicans growth assays showed the compounds to decrease SAP2-dependent growth while not affecting SAP2-independent growth.     

Interactions between immunoglobulin-like and catalytic modules in Clostridium thermocellum cellulosomal cellobiohydrolase CbhA

Cellobiohydrolase CbhA from Clostridium thermocellum cellulosome is a multi-modular protein composed starting from the N-terminus of a carbohydrate-binding module (CBM) of family 4, an immunoglobulin(Ig)-like module, a catalytic module of family 9 glycoside hydrolases (GH9), X1(1) and X1(2) modules, a CBM of family 3 and a dockerin module. Deletion of the Ig-like module from the Ig-GH9 construct results in complete inactivation of the GH9 module. The crystal structure of the Ig-GH9 module pair reveals the existence of an extensive module interface composed of over 40 amino acid residues of both modules and maintained through a large number of hydrophilic and hydrophobic interactions. To investigate the importance of these interactions between the two modules, we compared the secondary and tertiary structures and thermostabilities of the individual Ig-like and GH9 modules and the Ig-GH9 module pair using both circular dichroism (CD) spectroscopy and differential scanning calorimetry (DSC). Thr230, Asp262 and Asp264 of the Ig-like module are located in the module interface of the Ig-GH9 module pair and are suggested to be important in 'communication' between the modules. These residues were mutated to alanyl residues. The structure, stability and catalytic properties of the native Ig-GH9 and its D264A and T230A/D262A mutants were compared. The results indicate that despite being able to fold relatively independently, the Ig-like and GH9 modules interact and these interactions affect the final fold and stability of each module. Mutations of one or two amino acid residues lead to destabilization and change of the mechanism of thermal unfolding of the polypeptides. The enzymatic properties of native Ig-GH9, D264A and T230A/D262A mutants are similar. The results indicate that inactivation of the GH9 module occurs as a result of multiple structural disturbances finally affecting the topology of the catalytic center.     

RNA cleavage without hydrolysis. Splitting the catalytic activities of binase with Asn101 and Thr101 mutations

Members of the microbial guanyl-specific ribonuclease family catalyse the endonucleolytic cleavage of single-stranded RNA in a two-step reaction involving transesterification to form a 2',3'-cyclic phosphate and its subsequent hydrolysis to yield the respective 3'-phosphate. The extracellular ribonuclease from Bacillus intermedius (binase, RNase Bi) shares a common mechanism for RNA hydrolysis with mammalian RNases. Two catalytic residues in the active site of binase, Glu72 and His101, are thought to be involved in general acid-general base catalysis of RNA cleavage. Using site-directed mutagenesis, binase mutants were produced containing amino acid substitutions H101N and H101T and their catalytic properties towards RNA, poly(I), poly(A), GpC and guanosine 2',3'- cyclic phosphate (cGMP) substrates were studied. The engineered mutant proteins are active in the transesterification step which produces the 2',3'-cyclic phosphate species but they have lost the ability to catalyse hydrolysis of the cyclic phosphate to give the 3' monophosphate product.      

Converting trypsin to elastase: substitution of the S1 site and adjacent loops reconstitutes esterase specificity but not amidase activity

The conversion of trypsin into a protease with chymotrypsin-like activity and specificity required substitution of fifteen residues in the S1 site and two surface loops with their chymotrypsin counterparts [Hedstrom,L., Szilagyi,L. and Rutter,W.J. (1992) Science, 255, 1249- 1253]. These residues may define a set of general structural determinants of specificity in the trypsin family. In order to test this hypothesis, we have attempted to convert trypsin into a protease with specificity for substrates containing small aliphatic residues by replacing the S1 site and these surface loops with the analogous residues of elastase. Five elastase-like mutant enzymes were constructed with various combinations of these substitutions. Four mutant enzymes catalyze the hydrolysis of MeOSuc-Ala-Ala-Pro-Ala-SBzl more efficiently than the hydrolysis of Suc-Ala-Ala-Pro-Phe-SBzl. This observation indicates that the mutant enzymes have elastase-like esterase specificity. The best mutant, Tr-->E1-2, is a more specific esterase than elastase: the ratio of the values of kcat/Km for MeOSuc- Ala-Ala-Pro-Ala-SBzl and Suc-Ala-Ala-Pro-Phe-SBzl is greater than 160 for Tr-->E1-2 and 50 for elastase. However, the esterase activity of Tr- ->E1-2 is 300-fold less than elastase; in addition, Tr-->E1-2 has no measurable amidase activity. Thus these substitutions do not construct a protease with elastase-like activity. These experiments indicate that a unique structural solution is required for each different specificity. Previous work suggested that instability of the S1 site is a major barrier to redesigning the specificity of trypsin. This view is corroborated by preliminary structural studies of Tr-->E1-2. One dimensional 1H NMR spectrum of Tr-->E1-2 suggests that the S1 site and the two surface loops of this mutant trypsin may be disordered.      

Identification of two new hydrophobic residues on basic fibroblast growth factor important for fibroblast growth factor receptor binding

Basic fibroblast growth factor (bFGF) is implicated in the pathogenesis of several types of vascular and connective diseases. A key step in the discovery of bFGF receptor antagonists to mitigate these actions is to define the functional epitopes required for receptor binding of the growth factor. Using structure-based site-directed mutagenesis, two critical areas on the bFGF surface for the high affinity receptor binding have already been identified [Springer, B.A., Pantoliano, M.W., Barberal, F.A., Gunyuzlu, P.L., Thompson, L.D., Herblin, W.F., Rosenfeld, S.A. and Book, G.W. (1994) J. Biol. Chem., 269, 26879-26884; Zhu, H.Y., Ramnarayan, K., Anchin, J., Miao, Y., Sereno, A., Millman, L., Zheng, J., Balaji, V.N. and Wolff, M.E. (1995) J. Biol. Chem., 270, 21869-21874; Zhu, H.Y., Anchin, J., Ramnarayan, K., Zheng, J., Kawai, T., Mong, S. and Wolff, M.E. (1997) Protein Engng, 10, 417-421]. According to these studies, one receptor binding site includes two polar residues Glu96 and Asn104 on bFGF whereas the other includes four hydrophobic residues Tyr24, Tyr103, Leu140 and Met142. Using a protein modelling technique, we report here the identification of a new hydrophobic patch on bFGF which includes residues Tyr73, Val88 and Phe93. The role of this area on receptor binding affinity was evaluated by mutating each of these residues individually and determining the mutated protein's (mutein's) receptor binding affinity. In addition, we examined the role of two other hydrophobic residues, Phe30 and Leu138, on bFGF for high-affinity receptor binding. These two residues are the neighbors of the hydrophobic residues Tyr24 and Tyr103, respectively. Replacement of Val88 and Phe93 with alanine reduced the receptor binding affinity about 10- and 80-fold, respectively, compared with wild-type bFGF. In contrast, substitution of Phe30 and Leu138 with alanine has no effect on the receptor binding affinities. We conclude that the newly identified hydrophobic residues, Val88 and Phe93, are crucial for the receptor binding. The present data, together with the previous identification of four hydrophobic residues (Tyr24, Tyr103, Leu140 and Met142), suggests that there are two hydrophobic receptor binding sites on the bFGF surface. Our findings can be employed in the discovery and design of potent bFGF antagonists using computational methods.      

Computer-based de novo designs of tripeptides as novel neuraminidase inhibitors

The latest influenza A (H1N1) pandemic attracted worldwide attention and called for the urgent development of novel antiviral drugs. Here, seven tripeptides are designed and explored as neuraminidase (NA) inhibitors on the structural basis of known inhibitors. Their interactions with NA are studied and compared with each other, using flexible docking and molecular dynamics simulations. The various composed tripeptides have respective binding specificities and their interaction energies with NA decrease in the order of FRI > FRV > FRT > FHV > FRS > FRG > YRV (letters corresponding to amino acid code). The Arg and Phe portions of the tripeptides play important roles during the binding process: Arg has strong electrostatic interactions with the key residues Asp151, Glu119, Glu227 and Glu277, whereas Phe fits well in the hydrophobic cave within the NA active site. Owing to the introduction of hydrophobic property, the interaction energies of FRV and FRI are larger; in particular, FRI demonstrates the best binding quality and shows potential as a lead compound. In addition, the influence of the chemical states of the terminal amino acids are clarified: it is revealed that the charged states of the N-terminus (NH(3) (+)) and C-terminus (COO(-)) are crucial for the tripeptide inhibitory activities and longer peptides may not be appropriate. In addition, the medium inhibiting activity by acetylation of the N-terminus indicates the possible chemical modifications of FRI. Experimental efforts are expected in order to actualize the tripeptides as potent NA inhibitors in the near future.     

Modified substrate specificity of L-hydroxyisocaproate dehydrogenase derived from structure-based protein engineering

L-2-Hydroxyisocaproate dehydrogenase (L-HicDH) is characterized by a broad substrate specificity and utilizes a wide range of 2-oxo acids branched at the C4 atom. Modifications have been made to the sequence of the NAD(H)-dependent L-HicDH from Lactobacillus confusus in order to define and alter the region of substrate specificity towards various 2- oxocarbonic acids. All variations were based on a 3D-structure model of the enzyme using the X-ray coordinates of the functionally related L- lactate dehydrogenase (L-LDH) from dogfish as a template. This protein displays only 23% sequence identity to L-HicDH. The active site of L- HicDH was modelled by homology to the L-LDH based on the conservation of catalytically essential residues. Substitutions of the active site residues Gly234, Gly235, Phe236, Leu239 and Thr245 were made in order to identify their unique participation in substrate recognition and orientation. The kinetic properties of the L239A, L239M, L236V and T245A enzyme variants confirmed the structural model of the active site of L-HicDH. The substrates 2-oxocaproate, 2-oxoisocaproate, phenylpyruvate, phenylglyoxylate, keto-tert-leucine and pyruvate were fitted into the active site of the subsequently refined model. In order to design dehydrogenases with an improved substrate specificity towards keto acids branched at C3 or C4, amino acid substitutions at positions Leu239, Phe236 and Thr245 were introduced and resulted in mutant enzymes with completely different substrate specificities. The substitution T245A resulted in a relative shift of substrate specificity for keto-tert-leucine of more than 17000 compared with the 2-oxocaproate (kcat/KM). For the substrates branched at C4 a relative shift of up to 500 was obtained for several enzyme variants. A total of nine mutations were introduced and the kinetic data for the set of six substrates were determined for each of the resulting mutant enzymes. These were compared with those of the wild-type enzyme and rationalized by the active site model of L-HicDH. An analysis of the enzyme variants provided new insight into the residues involved in substrate binding and residues of importance for the differences between LDHs and HicDH. After the protein design project was complete the X-ray structure of the enzyme was solved in our group. A comparison between the model and the experimental 3D structure proved the quality of the model. All the variants were designed, expressed and tested before the 3D structure became available.      

Roles of catalytic residues in {alpha}-amylases as evidenced by the structures of the product-complexed mutants of a maltotetraose-forming amylase

The crystal structures of the four product-complexed singlemutants of the catalytic residues of Pseudomonas stutzeri maltotetraose-forming-amylase, E219G, D193N, D193G and D294N, have been determined.Possible roles of the catalytic residues Glu219, Asp193 andAsp294 have been discussed by comparing the structures amongthe previously determined complexed mutant E219Q and the presentmutant enzymes. The results suggested that Asp193 predominantlyworks as the base catalyst (nucleophile), whose side chain atomlies in close proximity to the C1-atom of Glc4, being involvedin the intermediate formation in the hydrolysis reaction. WhileAsp294 works for tightly binding the substrate to give a twistedand a deformed conformation of the glucose ring at position–1 (Glc4). The hydrogen bond between the side chain atomof Glu219 and the O1-atom of Glc4, that implies the possibilityof interaction via hydrogen, consistently present throughoutthese analyses, supports the generally accepted role of thisresidue as the acid catalyst (proton donor).     

Engineering of the substrate-binding region of the sublilisin-like, cell-envelop proteinase of Lactococcus lactis

The substrate-binding region of the cell-envelope proteinaseof Lactococcus lactis strain SK11 was modelled, based on sequencebomology of the catalytic domain with the serine proteinasessubtilisin and thermitase. Substitutions, deletions and insertionswere introduced, by site-directed and cassette mutagenesfe ofthe prtP gene encoding this enzyme, based on sequence comparisonboth with subtilisin and with the homologous L.lactis strainWg2 proteinase, which has different proteolytic properties.The engineered enzymes were investigated for thermal stability,proteolytic activity and cleavage specificity towards smallchromogenk peptide substrates and the peptide _g1-casein(l–23).Mutations in the subtilisin-like substrate-binding region showedthat Ser433 is the active site residue, and that residues 138and 166 at either side of the binding cleft play an importantrole in substrate specificity, particularly when these residuesand the substrate are oppositely charged. The K748T mutationin a different domain also affected specificity and stability,suggesting that this residue is in close proximity to the subtilisin-likedomain and may form part of the substratebinding site. Severalmutant SK11 proteinases have novel properties not previouslyencountered in natural variants. Replacements of residues 137–139AKTalong one side of the binding cleft produced the 137–139GPPmutant proteinase with reduced activity and narrowed specificity,and the 137–139GLA mutant with increased activity andbroader specificity. Furthermore, the 137–139GDT mutanthad a specificity towards _g1,-casein(l–23) closely resemblingthat of L.lactis Wg2 proteinase. Mutants with an additionalnegative charge in the binding region were more stable towardsautoproteolysis.     

The potential of P1 site alterations in peptidomimetic protease inhibitors as suggested by virtual screening and explored by the use of C-C-coupling reagents

A synthetic concept is presented that allows the construction of peptide isostere libraries through polymer-supported C-acylation reactions. A phosphorane linker reagent is used as a carbanion equivalent; by employing MSNT as a coupling reagent, the C-acylation can be conducted without racemization. Diastereoselective reduction was effected with L-selectride. The reagent linker allows the preparation of a norstatine library with full variation of the isosteric positions including the P1 side chain that addresses the protease S1 pocket. Therefore, the concept was employed to investigate the P1 site specificity of peptide isostere inhibitors systematically. The S1 pocket of several aspartic proteases including plasmepsin II and cathepsin D was modeled and docked with approximately 500 amino acid side chains. Inspired by this virtual screen, a P1 site mutation library was designed, synthesized, and screened against three aspartic proteases (plasmepsin II, HIV protease, and cathepsin D). The potency of norstatine inhibitors was found to depend strongly on the P1 substituent. Large, hydrophobic residues such as biphenyl, 4-bromophenyl, and 4-nitrophenyl enhanced the inhibitory activity (IC50) by up to 70-fold against plasmepsin II. In addition, P1 variation introduced significant selectivity, as up to 9-fold greater activity was found against plasmepsin II relative to human cathepsin D. The active P1 site residues did not fit into the crystal structure; however, molecular dynamics simulation suggested a possible alternative binding mode.