Multidisciplinary cycles for protein engineering: Site-directed mutagenesis and X-ray structural studies of aspartic proteinases

Roles of Tyr13 and Phe219 in the unique substrate specificity of pepsin B

Pepsin B is known to be distributed throughout mammalia, including carnivores. In this study, the proteolytic specificity of canine pepsin B was clarified with 2 protein substrates and 37 synthetic octapeptides and compared with that of human pepsin A. Pepsin B efficiently hydrolyzed gelatin but very poorly hydrolized hemoglobin. It was active against only a group of octapeptides with Gly at P2, such as KPAGF/LRL and KPEGF/LRL (arrows indicate cleavage sites). In contrast, pepsin A hydrolyzed hemoglobin but not gelatin and showed high activity against various types of octapeptides, such as KPAEF/FRL and KPAEF/LRL. The specificity of pepsin B is unique among pepsins, and thus, the enzyme provides a suitable model for analyzing the structure and function of pepsins and related aspartic proteinases. Because Tyr13 and Phe219 in/around the S2 subsites (Glu/Ala13 and Ser219 are common in most pepsins) appeared to be involved in the specificity of pepsin B, site-directed mutagenesis was undertaken to replace large aromatic residues with small residues and vice versa. The Tyr13Ala/Phe219Ser double mutant of pepsin B was found to demonstrate broad activity against hemoglobin and various octapeptides, whereas the reverse mutant of pepsin A had significantly decreased activity. According to molecular modeling of pepsin B, Tyr13 OH narrows the substrate-binding space and a peptide with Gly at P2 might be preferentially accommodated because of its high flexibility. The hydroxyl can also make a hydrogen bond with nitrogen of a P3 residue and fix the substrate main chain to the active site, thus restricting the flexibility of the main chain and strengthening preferential accommodation of Gly at P2. The phenyl moiety of Phe219 is bulky and narrows the S2 substrate space, which also leads to a preference for Gly at P2, while lowering the catalytic activity against other peptide types without making a hydrogen-bonding network in the active site.

Role of S'1 loop residues in the substrate specificities of pepsin A and chymosin

Proteolytic specificities of human pepsin A and monkey chymosin were investigated with a variety of oligopeptides as substrates. Human pepsin A had a strict preference for hydrophobic/aromatic residues at P'1, while monkey chymosin showed a diversified preferences accommodating charged residues as well as hydrophobic/aromatic ones. A comparison of residues forming the S'1 subsite between mammalian pepsins A and chymosins demonstrated the presence of conservative residues including Tyr(189), Ile(213), and Ile(300) and group-specific residues in the 289-299 loop region near the C terminus. The group-specific residues consisted of hydrophobic residues in pepsin A (Met(289), Leu/Ile/Val(291), and Leu(298)) and charged or polar residues in chymosins (Asp/Glu(289) and Gln/His/Lys(298)). Because the residues in the loop appeared to be involved in the unique specificities of respective types of enzymes, site-directed mutagenesis was undertaken to replace pepsin-A-specific residues by chymosin-specific ones and vice versa. A yeast expression vector for glutathione-S-transferase fusion protein was newly developed for expression of mutant proteins. The specificities of pepsin-A mutants could be successfully altered to the chymosin-like preference and those of chymosin mutants, to pepsin-like specificities, confirming residues in the S'1 loop to be essential for unique proteolytic properties of the enzymes. An increase in preference for charged residues at P'1 in pepsin-A mutants might have been due to an increase in the hydrogen-bonding interactions. In chymosin mutants, the reverse is possible. The changes in the catalytic efficiency for peptides having charged residues at P'1 were dominated by k(cat) rather than K(m) values.

Effects of pulsed electric fields on the activity and structure of pepsin

A continuous pulsed electric field (PEF) system integrated with six co-field flow PEF treatment chambers was used to study the inactivation of pepsin. The inactivation of pepsin activity was a function of applied electric field strength, electrical conductivity, and pH. The inactivation of pepsin by PEF followed a first-order model. The first-order inactivation kinetic constant of pepsin was 0.012 (1/mus) in 7.5 mM HCl (pH 2.0) at 34.2 kV/cm. Aggregation of pepsin was observed during PEF treatment; however, the inactivation took place before the formation of aggregates. Circular dichroism analysis showed that inactivation of pepsin by PEF was correlated to the loss of beta-sheet structure in a pepsin molecule. The relative residual activity of PEF-treated pepsin was correlated to the relative molar ellipticity at 215 nm. Both PEF- and heat-induced inactivation of pepsin were correlated with the alteration of the secondary structure (beta-sheet dominant structure) of pepsin.

The active site of pepsin is formed in the intermediate conformation dominant at mildly acidic pH

Pepsin is an aspartic protease that acts in food digestion in the mammal stomach. An optimal pH of around 2 allows pepsin to operate in its natural acidic environment, while at neutral pH the protein is denatured. Although the pH dependence of pepsin activity has been widely investigated since the 40s, a renewed interest in this protein has been fueled by its homology to the HIV and other aspartic proteases. Recently, an inactive pepsin conformation has been identified that accumulates at mildly acidic pH, whose structure and properties are largely unknown. In this paper, we analyse the conformation of pepsin at different pHs by a combination of spectroscopic techniques, and obtain a detailed characterisation of the intermediate. Our analysis indicates that it is the dominant conformation from pH 4 to 6.5. Interestingly, its near UV circular dichroism spectrum is identical to that of the native conformation that appears at lower pH values. In addition, we show that the intermediate binds the active site inhibitor pepstatin with a strength similar to that of the native conformation. Pepsin thus adopts, in the 6.5-4.0 pH interval, a native-like although catalytically inactive conformation. The possible role of this intermediate during pepsin transportation to the stomach lumen is discussed.

Functional implications of disulfide bond, Cys206-Cys210, in recombinant prochymosin (chymosin)

Prochymosin (chymosin) contains three disulfide bonds: Cys45-Cys50, Cys206-Cys210, and Cys250-Cys283. We have demonstrated that Cys250-Cys283 is indispensable for correct refolding of prochymosin, whereas Cys45-Cys50 is dispensable but has some contribution to the stability and substrate specificity of the enzyme. Here, we report the results about the functions of Cys206-Cys210 by site-directed mutagenesis studies. In a glutathione redox system C206A/C210A mutant exhibited oxidative refolding kinetics and efficiency ( approximately 40% reactivation) similar to those of the wild-type prochymosin, indicating that Cys206-Cys210 is also dispensable for refolding. However, C206S/C210S and single-site mutants (C210A, C210S, and C206A) showed only about 3 and 0-0.4% reactivation, respectively. This is quite different from the Cys45-Cys50 deficient mutants (C45A, C50A, C45A/C50A, C45D, C50S, C45D/C50S, C45A/C50S), which have comparable refolding efficiencies, implying that the substituents at position 206 and 210 play more important role in determining correct refolding than those at position 45 and 50. Urea-induced denaturation and fluorescence quenching studies indicated that the prochymosin mutants C206A/C210A and C206S/C210S were 2.1 and 4.8 kJ/mol less stable than prochymosin and some tryptophan residue in the mutated molecules was less exposed. However, the wild-type and mutant prochymosins shared similar far-UV CD and fluorescence emission spectra and similar specific potential activity, suggesting that the overall conformation was maintained after mutation. Activity assay and kinetic analysis revealed that mutation did not change the specific milk-clotting activity significantly but resulted in an increase in K(m) and k(cat) toward a hexapeptide substrate. On the basis of the above-mentioned perturbance of tryptophanyl microenvironment and the three-dimensional structure of chymosin, we proposed that deletion of Cys206-Cys210 may induce a propagated conformational change, resulting in a perturbance of the local conformation around active-site cleft and in turn, an alteration of the substrate specificity.

Structure of asparagine-linked oligosaccharides of an aspartic proteinase from the zygomycete fungus Rhizomucor pusillus

The zygomycete fungus Rhizomucor pusillus (previously called Mucor pusillus) secretes an aspartic proteinase containing two asparagine-linked, high-mannose type oligosaccharide chains at Asn79 and Asn188. For structural elucidation of the carbohydrate moieties, the protein was divided into two portions, an N-terminal portion containing Asn79 and a C-terminal portion containing Asn188, by a specific autocatalytic cleavage under alkaline conditions. Each of the asparagine-linked oligosaccharides was then released by peptide-N-glycosidase F digestion and pyridylaminated with a fluorescent reagent, 2-aminopyridine, at the reducing end. High-performance liquid chromatography analyses showed that the structure of the asparagine-linked oligosaccharide chain attached to residue Asn79 was Man5GlcNAc2, and that bound to residue Asn188 was Man5GlcNAc2 and Man6GlcNAc2. These observations suggest that the processing of mannose residues in asparagine-linked oligosaccharides in the Golgi apparatus of Rhizomucor resembles that in mammalian cells.

The sole lysine residue in porcine pepsin works as a key residue for catalysis and conformational flexibility

Pepsin contains a single lysine residue which protrudes from the enzyme's surface, behind the active site cleft, on the C-terminal domain. Mutations of pepsin by site-directed mutagenesis of the Lys-319 residue were generated to study the structure-function relationships. Kinetic parameters, pH activity profiles, along with conformational analysis using circular dichroism (CD), and molecular modelling were examined for the wild-type (non-mutant) and mutant enzymes. The pepsin mutations, Lys-319-->Met and Lys-319-->Glu, resulted in a progressive increase in the Km and similar decrease in kcat, respectively, as well as being denatured at a lower pH than the wild-type pepsin. CD analysis indicated that mutations at Lys-319 resulted in changes in secondary structure fractions which were reflected in changes in enzymatic activity as compared to the wild-type pepsin, i.e. kinetic data and pH denaturation study. Molecular modelling of mutant enzymes indicated differences in flexibility in the flap loop region of the active site, the region around the entrance of the active site cleft, sub-site regions for peptide binding, and in the subdomains of the C-terminal domain when compared to the wild-type enzyme. The results suggest that Lys-319, which is distal to the active site, is important to the flexibility/stability of the enzyme, as well as to its catalytic machinery.

Secretion and maturation study of endothiapepsin in Saccharomyces cerevisiae. A first step toward improving its substrate specificity

The gene encoding endothiapepsin (EAP), an extracellular aspartic proteinase from the filamentous ascomycete Cryphonectria parasitica, was expressed into Saccharomyces cerevisiae. Efficient secretion of an active and correctly processed enzyme was achieved when expressing the entire cDNA encoding prepro-EAP under the control of the galactose-inducible GRAP1 yeast promoter. Since three independent, site-directed mutations of EAP, including the substitution of an aspartyl catalytic residue, resulted in the intracellular accumulation of zymogen forms, we assumed that the EAP propeptide was autocatalytically processed. As a prerequisite to further improve the specificity of EAP, we therefore attempted to bypass this self-processing step in three different ways: 1) introduction of a Kex2-like recognition site between the pro and the mature part, 2) deletion of the prosequence (pre-EAP), and 3) co-expression in trans of the pre-EAP with its preprosequence. No improvement in the secretion of mutant enzymes was obtained in any of these experiments. As an alternative, we finally replaced the EAP processing site by the chymosin cleavage sequence of kappa-casein. Such a modification remained efficient in directing the secretion of active EAP only when a putative alpha-helix structural motif was conserved at the C terminus of the pro region.

Software for illustrative presentation of basic clinical characteristics of laboratory tests--GraphROC for Windows

GraphROC for Windows is a program for clinical test evaluation. It was designed for the handling of large datasets obtained from clinical laboratory databases. In the user interface, graphical and numerical presentations are combined. For simplicity, numerical data is not shown unless requested. Relevant numbers can be "picked up" from the graph by simple mouse operations. Reference distributions can be displayed by using automatically optimized bin widths. Any percentile of the distribution with corresponding confidence limits can be chosen for display. In sensitivity-specificity analysis, both illness- and health-related distributions are shown in the same graph. The following data for any cutoff limit can be shown in a separate click window: clinical sensitivity and specificity with corresponding confidence limits, positive and negative likelihood ratios, positive and negative predictive values and efficiency. Predictive values and clinical efficiency of the cutoff limit can be updated for any prior probability of disease. Receiver Operating Characteristics (ROC) curves can be generated and combined into the same graph for comparison of several different tests. The area under the curve with corresponding confidence interval is calculated for each ROC curve. Numerical results of analyses and graphs can be printed or exported to other Microsoft Windows programs. GraphROC for Windows also employs a new method, developed by us, for the indirect estimation of health-related limits and change limits from mixed distributions of clinical laboratory data.