Protein engineering最新文献

The virtue of the so-called 'proline concept' and the 'charge concept' for stabilizing protease-susceptible regions of a protein structure was compared on bovine pancreatic ribonuclease A. Alanine 20 and serine 21, both of which are located in a loop that is susceptible to the unspecific proteases subtilisin Carlsberg, subtilisin BPN', proteinase K and elastase, were replaced with proline or lysine by site-directed mutagenesis. The rate constant of proteolysis was decreased by up to three orders of magnitude for the proline mutants depending on the site of the mutation and the protease used. In contrast, substitution by lysine increased the proteolytic resistance by only one order of magnitude characterizing the 'proline concept' as superior to the 'charge concept'. Although the four applied proteases are considered to be unspecific, the degree of stabilization of the ribonuclease molecule varied considerably, indicating the impact of individual differences in their substrate specificity on the proteolytic resistance and degradation pathway of the target protein.

An intersubunit interactions study related to the active site has been performed on the wild-type cytidine deaminase (CDA) and on the mutant enzyme F137W/W113F. F137 is the homologous to the Bacillus subtilis CDA F125 involved in the subunit interactions. In the presence of SDS, wild-type human CDA dissociates into enzymatically inactive monomers without intermediate forms via a non-cooperative transition. Extensive dialysis or dilution of the inactivated monomers restores completely the activity. Circular dichroism measurements show that the secondary/tertiary structure organization of each subunit is unaffected by the SDS concentration, while the mutation Phe/Trp causes weakening in quaternary structure. The presence of the strong human CDA competitive inhibitor 5-fluorozebularine disfavours dissociation of the tetramer into subunits in the wild-type CDA, but not in mutant enzyme F137W/W113F. The absence of tyrosine fluorescence and the much higher quantum yield of the double mutant protein spectrum suggest the occurrence of an energy transfer effect between the protein subunits. This assumption is confirmed by the crystallographic studies on B.subtilis in which it is shown that three different subunits concur with the formation of each of the four active sites and that F125, homologous to the human CDA F137, is located at the interface between two different subunits contributing to the formation of active site.

The binding affinity and specificity of recombinant antibodies can be modified by site-directed mutagenesis. Here we have used molecular modelling of the variable domains of an enantiospecific antibody fragment to fine-tune its affinity so it is more suitable for the fractionation of the drug enantiomers. We have shown earlier that the Fab fragment of this antibody specifically recognizes one enantiomer from the racemic mixture of a medical drug and that it can be used for the fractionation of these enantiomers by affinity chromatography. However, the affinity was unnecessarily high, requiring harsh elution conditions to release the bound enantiomer. Thus, the continuous use of the antibody affinity columns was impossible. We made a homology model of the antibody and designed mutations to the antigen-binding site to decrease the affinity. Four out of five point mutations showed decreased affinity for the hapten. Two of the mutations were also combined to construct a double mutant. The affinity columns made using one of the single mutants with lowered affinity and the double mutant were capable of multiple rounds of enantioseparation.

Calcineurin (CN) is a heterodimer protein consisting of a 61 kDa catalytic subunit A and a 19 kDa regulatory subunit B. It plays a critical role in T-cell activation and is involved in many cellular processes. Regulation of CN is rather complex, including a number of factors such as divalent metal ions (primarily Ca(2+) and Mn(2+)), calmodulin (CaM) and autoinhibition (AI) segment. Previously, we reported that a loop 7 deletion mutant (V314) in subunit A exhibited high phosphatase activity, although the mechanism for the surprising activity enhancement and whether the activity change applies to other loop 7 residues were not known. In order to probe the role of loop 7, we have carried out extensive mutagenesis experiments, followed by systematic activity assays under a number of regulatory conditions. All mutants, including single deletion mutants Y315, N316 and double deletion mutant V314Y315, showed increased phosphatase activity. Significantly, activities of the mutants containing the V314 deletion, namely V314 and V314Y315, were no longer regulated by regulatory subunit B. These results, along with the structure analysis, suggest that loop 7 as a whole plays an important role in mediating CN's regulation through bridging the regulatory subunit and catalytic core and interaction with the AI segment of CN.

The tertiary structure of the central catalytic domain of insertion sequence ISLC3 isolated from Lactobacillus casei ATCC 393 was predicted using the homology modeling approach. The novel insertion sequence was isolated by us from the template bacteriophage phiA3 of L.casei ATCC 393. The number of amino acid residues of the ISLC3 central catalytic domain was 116 and was treated as the query sequence. There were five Web-available threading methods used to find some primary structure templates for the query sequence. These primary templates were further screened using the SWISS-MODEL Protein Modeling Server and the default parameter settings therein to give six final structure templates. All of these final structure templates were the integrase (IN) protein of retroviruses. Multiple sequence alignment using these IN sequences against the query one revealed the signature DDE motif. Based on the structures of these final templates, the structure of the query sequence was constructed using the InsightII/Discover/Homology programs. A metal ion, Mg(2+), was inserted into the center of the putative catalytic pocket formed by the DDE residues of the predicted structure in the final rounds of refinement by molecular dynamics (MD) simulations. The structure with a metal ion included was designated with Mg and that without a metal ion was designated free Mg. The average exposed surface area of some hydrophobic residues of both the predicted free Mg and with Mg structures were computed and compared with those computed for the six structure templates. Whereas the predicted with Mg structure was slightly more exposed than the predicted free Mg structure, the former appeared to be more stable than the latter, as revealed by the lower conformation energy recorded for the former during the structure refinement by MD simulations. To verify further the predicted structures, the coordinates of both predicted structures were fed into the ERRAT Protein Verification Server. It was found that the quality of the predicted with Mg structure was much better than that of the free Mg structure. The validation results also indicated that regions of the predicted with Mg structure that can be rejected at the 95% confidence level were approximately 20% whereas those which can be rejected at the same level for the six structure templates were approximately 10%. The predicted with Mg structure was also docked into a short oligonucleotide representing the substrate of the ISLC3 transposase using the DOCK_4.0.2 program. It was found that both Glu140 and Asp68 residues of the DDE motif of the predicted with Mg structure were able to form hydrogen bonds with the DNA substrate, which was similar to what was observed in a docking study using the retrovirus IN 1asu and its DNA substrate.

We present a protein fold recognition method, MANIFOLD, which uses the similarity between target and template proteins in predicted secondary structure, sequence and enzyme code to predict the fold of the target protein. We developed a non-linear ranking scheme in order to combine the scores of the three different similarity measures used. For a difficult test set of proteins with very little sequence similarity, the program predicts the fold class correctly in 34% of cases. This is an over twofold increase in accuracy compared with sequence-based methods such as PSI-BLAST or GenTHREADER, which score 13-14% correct first hits for the same test set. The functional similarity term increases the prediction accuracy by up to 3% compared with using the combination of secondary structure similarity and PSI-BLAST alone. We argue that using functional and secondary structure information can increase the fold recognition beyond sequence similarity.

The Structural Motifs of Superfamilies (SMoS) database provides information about the structural motifs of aligned protein domain superfamilies. Such motifs among structurally aligned multiple members of protein superfamilies are recognized by the conservation of amino acid preference and solvent inaccessibility and are examined for the conservation of other features like secondary structural content, hydrogen bonding, non-polar interaction and residue packing. These motifs, along with their sequence and spatial orientation, represent the conserved core structure of each superfamily and also provide the minimal requirement of sequence and structural information to retain each superfamily fold.

Hydrolysis of beta-lactam antibiotics by beta-lactamase enzymes is the most common mechanism of bacterial resistance to these agents. Several small-molecule, mechanism-based inhibitors of beta-lactamases such as clavulanic acid are clinically available although resistance to these inhibitors has been increasing in bacterial populations. In addition, these inhibitors act only on class A beta-lactamases. Here we utilized phage display to identify peptides that bind to the class A beta-lactamase, TEM-1. The binding affinity of one of these peptides was further optimized by the synthesis of peptide arrays using SPOT synthesis technology. After two rounds of optimization, a linear 6-mer peptide with the sequence RRGHYY was obtained. A soluble version of this peptide was synthesized and found to inhibit TEM-1 beta-lactamase with a K(i) of 136 micro M. Surprisingly, the peptide inhibits the class A Bacillus anthracis Bla1 beta-lactamase with a K(i) of 42 micro M and the class C beta-lactamase, P99, with a K(i) of 140 micro M, despite the fact that it was not optimized to bind these enzymes. This peptide may be a useful starting point for the design of non-beta-lactam, broad-spectrum peptidomimetic inhibitors of beta-lactamases.

This paper describes attempts to increase the kinetic stability of chitinase B from Serratia marcescens (ChiB) by the introduction of semi-automatically designed rigidifying mutations of the Gly-->Ala and Xxx-->Pro type. Of 15 single mutants, several displayed significant increases in thermal stability, whereas most mutants showed minor effects. All mutations with non-marginal effects on stability clustered in a limited, surface-exposed region of the enzyme, indicating that this region is involved in a partial unfolding process that triggers irreversible thermal inactivation (aggregation). A double mutant containing two stabilizing mutations in this region (G188A, A234P) displayed a 10-fold increase in half-life at 57 degrees C and a 4.2 degrees C increase in apparent T(m). These results show that entropic stabilization works well for ChiB and they pinpoint a region whose unfolding may be crucial for the kinetic stability of this enzyme.

A continuum electrostatics model is used to calculate the relative stabilities of 117 mutants of staphylococcal nuclease (SNase) involving the mutation of a charged residue to an uncharged residue. The calculations are based on the crystallographic structure of the wild-type protein and attempt to take implicitly into account the effect of the mutations in the denatured state by assuming a linear relationship between the free energy changes caused by the mutation in the native and denatured states. A good correlation (linear correlation coefficient of approximately 0.8) is found with published experimental relative stabilities of these mutants. The results suggest that in the case of SNase (i) charged residues contribute to the stability of the native state mainly through electrostatic interactions, and (ii) native-like electrostatic interactions may persist in the denatured state. The continuum electrostatics method is only moderately sensitive to model parameters and leads to quasi-predictive results for the relative mutant stabilities (error of 2-3 kJ mol(-1) or of the order of k(B)T), except for mutants in which a charged residue is mutated to glycine.