Human rhinovirus 3C protease: a cysteine protease showing the trypsin(ogen)-like fold

Viral-encoded proteases cleave precursor polyprotein(s) leading to maturation of infectious virions. Strikingly, human rhinovirus 3C protease shows the trypsin(ogen)-like serine protease fold based on two topologically equivalent six-stranded β-barrels, but displays residue Cys147 as the active site nucleophile. By contrast, papain, which is representative of most cysteine proteases, does not display the trypsin(ogen)-like fold. Remarkably, in human rhinovirus 3C cysteine protease, the catalytic residues Cys147, His40 and Glu71 are positioned as Ser195, His57 and Asp102, respectively, building up the catalytic triad of serine proteases in the chymotrypsin–trypsin–elastase family. However, as compared to trypsin-like serine proteases and their zymogens, residue His40 and the oxyanion hole of human rhinovirus 3C cysteine protease, both key structural components of the active site, are located closer to the protein core. Human rhinovirus 3C cysteine protease cleaves preferentially GlnGly peptide bonds or, less commonly, the GlnSer, GlnAla, GluSer or GluGly pairs. Finally, human rhinovirus 3C cysteine protease and the 3CD cysteine protease–polymerase covalent complex bind the 5′ non-coding region of rhinovirus genomic RNA, an essential function for replication of the viral genome.     

Factors determining the relative stability of anionic tetrahedral complexes in serine protease catalysis and inhibition

Quantum mechanical ab initio (RHF/6-31+G*//RHF/3-21G) calculations were used to simulate the formation of the tetrahedral complex intermediate (TC) in serine protease active site by substrates and transition-state analog inhibitors. The enzyme active site was simulated by an assembly of the amino acids participating in catalysis, whereas the substrates and inhibitors were simulated by small ligands, acetamide (1) and trifluoroacetone (2), respectively. For the first time, the principal factors determining the relative stability of the TC in serine proteases are arranged according to their energy contributions. These include (a) formation of the new covalent bond between Ser195 O(gamma) and the electrophilic center of a ligand; (b) stabilization of the oxyanion in the oxyanion hole; (c) basic catalysis by His57; and (d) hydrogen bond between Asp102 carboxylate and N(delta) of the protonated His57. We have directly calculated the gas-phase relative free energy of formation of TC(AS)(2) and TC(AS)(1), the value of DeltaDeltaG(g)[TC(AS)(2,1)]. It is DeltaE(cov), the relative energy of the new covalent bond between the enzyme and the ligand formed in a TC that determines the experimentally observed large difference in the stability of TCs formed by substrates and TS-analog inhibitors of serine proteases. We demonstrated that the relative stability of TCs formed by a series of mono- and dipeptide amides and TFKs, derived from experimental kinetic data, can be rather well approximated by the sum of the theoretically calculated value of DeltaDeltaG(g)[TC(AS)(2, 1)] and the difference in hydration free energies of isolated ligands.     

Crystal structure studies and inhibition kinetics of tripeptide chloromethyl ketone inhibitors with Streptomyces griseus protease B

The bacterial serine protease, SGPB, was inhibited by two specific tripeptide chloromethyl ketones, N-t-butyloxycarbonyl-l-alanylglycyl-l-phenylalanine chloromethyl ketone (BocAGFCK) and N-t-butyloxycarbonyl-glycyl-l-leucyl-l-phenylalanine chloromethyl ketone (BocGLFCK). Crystals of the inhibited complexes were grown and examined by X-ray crystallographic methods. The peptide backbone of each inhibitor is bound by three hydrogen bonds to the main chain of residues Ser214 to Gly216. There are two well-characterized hydrophobic pockets, S₁ and S₂, on the surface of SGPB which accommodate the P₁ and P₂ side-chains of the BocGLFCK inhibitor. A conformational change of Tyr171 is induced by the binding of this inhibitor. Both inhibitors make two covalent bonds to the SGPB enzyme. The imidazole ring of His57 is alkylated at the N^?2 atom and O^γ of Ser195 forms a hemiketal bond with the carbonyl-carbon atom of the inhibitor. Comparison of the binding modes of the two tripeptides in conjunction with the differences in their inhibition constants (K_I) allows one to estimate the binding energy of the leucyl side-chain as ?2.6 kcal mol^?1. The importance of an electrophilic component in the serine protease mechanism, which involves the polarization of the susceptible carbonyl bond of a substrate or inhibitor by the peptide NH groups of Gly193 and Ser195 is discussed.     

Conversion of trypsin to a functional threonine protease

The hydroxyl group of a serine residue at position 195 acts as a nucleophile in the catalytic mechanism of the serine proteases. However, the chemically similar residue, threonine, is rarely used in similar functional context. Our structural modeling suggests that the Ser 195 --> Thr trypsin variant is inactive due to negative steric interaction between the methyl group on the beta-carbon of Thr 195 and the disulfide bridge formed by cysteines 42 and 58. By simultaneously truncating residues 42 and 58 and substituting Ser 195 with threonine, we have successfully converted the classic serine protease trypsin to a functional threonine protease. Substitution of residue 42 with alanine and residue 58 with alanine or valine in the presence of threonine 195 results in trypsin variants that are 10(2) -10(4) -fold less active than wild type in kcat/KM but >10(6)-fold more active than the Ser 195 --> Thr single variant. The substitutions do not alter the substrate specificity of the enzyme in the P1'- P4' positions. Removal of the disulfide bridge decreases the overall thermostability of the enzyme, but it is partially rescued by the presence of threonine at position 195.     

The structure of a Michaelis serpin-protease complex.

Serine protease inhibitors (serpins) regulate the activities of circulating proteases. Serpins inhibit proteases by acylating the serine hydroxyl at their active sites. Before deacylation and complete proteolysis of the serpin can occur, massive conformational changes are triggered in the serpin while maintaining the covalent linkage between the protease and serpin. Here we report the structure of a serpin-trypsin Michaelis complex, which we visualized by using the S195A trypsin mutant to prevent covalent complex formation. This encounter complex reveals a more extensive interaction surface than that present in small inhibitor-protease complexes and is a template for modeling other serpin-protease pairs. Mutations of several serpin residues at the interface reduced the inhibitory activity of the serpin. The serine residue C-terminal to the scissile peptide bond is found in a closer than usual interaction with His 57 at the active site of trypsin.     

Amino acid sequence of rat mast cell protease I (chymase)

The amino acid sequence has been determined for rat mast cell protease I (RMCP I), a product of peritoneal mast cells. The active enzyme contains 227 residues, including three corresponding to the catalytic triad characteristic of serine protease (His-57, Asp-102, and Ser-195 in chymotrypsin). A computer search for homology indicates 73% and 33% sequence identity of RMCP I with rat mast cell protease II from mucosal mast cells and bovine chymotrypsin A, respectively. When the structure of RMCP I is compared to those of cathepsin G from human neutrophils and two proteases expressed in activated lymphocytes, 48-49% of the sequences are identical in each case. RMCP I has six half-cystine residues at the same positions as in RMCP II, cathepsin G, and the two lymphocyte proteases, suggesting disulfide pairs identical with those reported for RMCP II. A disulfide bond near the active site seryl residue and substrate binding site, present in pancreatic and plasma serine proteases, is not found in RMCP I or in the other cellular proteases. These results indicate that RMCP I and other chymotrypsin-like proteases of granulocyte and lymphocyte origin are more closely related to each other than to the pancreatic or plasma serine proteases.     

Characterization of Rarobacter faecitabidus protease I, a yeast-lytic serine protease having mannose-binding activity.

Rarobacter faecitabidus protease I, a yeast-lytic serine protease, was characterized in order to elucidate the mechanism of lysis of yeast cells by this enzyme. The N-terminal amino acid sequence of the enzyme was found to be homologous to those of Lysobacter enzymogenes alpha-lytic protease and Streptomyces griseus proteases A and B around the catalytic His residue, showing that it is a mammalian type serine protease. In a study of its substrate specificity, it preferentially hydrolyzed the ester of alanine among amino acid p-nitrophenylesters. It also efficiently hydrolyzed succinyl Ala-Pro-Ala p-nitroanilide, the specific synthetic substrate for pancreatic elastase. With oxidized insulin B-chain, it hydrolyzed almost exclusively the peptide bond between valine 18 and cysteic acid 19 in the early step of the reaction, and thereafter it partially hydrolyzed Val12-Glu13, Ala14-Leu15, and Leu15-Tyr16. These results indicate that Rarobacter protease I is elastase-like in its substrate specificity, preferentially hydrolyzing the peptide bond of aliphatic amino acids. Its affinity for yeast cells was also investigated, and while Rarobacter protease I was adsorbed by yeast cells, pancreatic elastase was not. This difference was thought to account for the failure of pancreatic elastase to lyse yeast cells, even though its specificity is similar to that of the yeast-lytic enzyme. Rarobacter protease I was adsorbed by a mannose-agarose column and specifically eluted from the column with a buffer containing D-mannose or D-glucose. These monosaccharides also inhibited its yeast-lytic activity.(ABSTRACT TRUNCATED AT 250 WORDS)     

Active-site mobility in human immunodeficiency virus, type 1, protease as demonstrated by crystal structure of A28S mutant.

The mutation Ala28 to serine in human immunodeficiency virus, type 1, (HIV-1) protease introduces putative hydrogen bonds to each active-site carboxyl group. These hydrogen bonds are ubiquitous in pepsin-like eukaryotic aspartic proteases. In order to understand the significance of this difference between HIV-1 protease and homologous, eukaryotic aspartic proteases, we solved the three-dimensional structure of A28S mutant HIV-1 protease in complex with a peptidic inhibitor U-89360E. The structure has been determined to 2.0 A resolution with an R factor of 0.194. Comparison of the mutant enzyme structure with that of the wild-type HIV-1 protease bound to the same inhibitor (Hong L, Treharne A, Hartsuck JA, Foundling S, Tang J, 1996, Biochemistry 35:10627-10633) revealed double occupancy for the Ser28 hydroxyl group, which forms a hydrogen bond either to one of the oxygen atoms of the active-site carboxyl or to the carbonyl oxygen of Asp30. We also observed marked changes in orientation of the Asp25 catalytic carboxyl groups, presumably caused by the new hydrogen bonds. These observations suggest that catalytic aspartyl groups of HIV-1 protease have significant conformational flexibility unseen in eukaryotic aspartic proteases. This difference may provide an explanation for some unique catalytic properties of HIV-1 protease.     

An unusual helix-turn-helix protease inhibitory motif in a novel trypsin inhibitor from seeds of Veronica (Veronica hederifolia L.)

The storage tissues of many plants contain protease inhibitors that are believed to play an important role in defending the plant from invasion by pests and pathogens. These proteinaceous inhibitor molecules belong to a number of structurally distinct families. We describe here the isolation, purification, initial inhibitory properties, and three-dimensional structure of a novel trypsin inhibitor from seeds of Veronica hederifolia (VhTI). The VhTI peptide inhibits trypsin with a submicromolar apparent K(i) and is expected to be specific for trypsin-like serine proteases. VhTI differs dramatically in structure from all previously described families of trypsin inhibitors, consisting of a helix-turn-helix motif, with the two alpha helices tightly associated by two disulfide bonds. Unusually, the crystallized complex is in the form of a stabilized acyl-enzyme intermediate with the scissile bond of the VhTI inhibitor cleaved and the resulting N-terminal portion of the inhibitor remaining attached to the trypsin catalytic serine 195 by an ester bond. A synthetic, truncated version of the VhTI peptide has also been produced and co-crystallized with trypsin but, surprisingly, is seen to be uncleaved and consequently forms a noncovalent complex with trypsin. The VhTI peptide shows that effective enzyme inhibitors can be constructed from simple helical motifs and provides a new scaffold on which to base the design of novel serine protease inhibitors.     

Molecular structure of the alpha-lytic protease from Myxobacter 495 at 2.8 Angstroms resolution.

The α-lytic protease was isolated from an extracellular filtrate of the soil microorganism Myxobacter 495. Trigonal crystals (space group, P3₂21) of this serine enzyme were grown from 1·3 m-Li₂SO₄ at pH 7·2. X-ray reflections from crystals of the native enzyme, comprising the 2·8 Å limiting sphere, were phased by the multiple isomorphous replacement technique. Five heavy-atom derivatives were used and the overall mean figure of merit 〈m?〉 is 0·83. The resulting native electron density map of α-lytic protease has been interpreted in conjunction with the published sequence (Olson et al., 1970) of 198 amino-acid residues.α-Lytic protease has a structural core similar to that of the pancreatic serine proteases (108 α-carbon atom positions are topologically equivalent (within 2·0 Å) to residues of porcine elastase) and its tertiary structure is even more closely related to the two other bacterial serine protease structures previously determined (James et al., 1978; Brayer et al., 1978b; Delbaere et al., 1979a). α-Lytic protease has the following distinctive features in common with the bacterial serine enzymes, Streptomyces griseus proteases A and B: an amino terminus that is exposed to solvent on the enzyme surface, a considerably shortened uranyl loop (residues 65 to 84), a major segment of polypeptide chain from the autolysis loop deleted (residues 144 to 155), a buried guanidinium group of Arg138 in an ion-pair bond with Asp194, and an altered conformation of the methionine loop (residues 168 to 182) relative to the pancreatic enzymes.At the present resolution, the members of the catalytic quartet (Ser214, Asp102, His57 and Ser195) adopt the conformation found in all members of the Gly-Asp-Ser-Gly-Gly serine protease family. The carboxylate of Asp102 is in a highly polar environment, as it is the recipient of four hydrogen bonds. The interaction between the N^ε2 atom of the imidazole ring in His57 and O^γ atom of Ser195 is very weak (3·3 Å) and supports the concept that there is little, if any, enhanced nucleophilicity of the side-chain of Ser195 in the native enzyme.The molecular basis for the observed substrate specificity of α-lytic protease is clear from the distribution of amino acid side-chains in the neighborhood of the active site. An insertion of five residues at position 217, and the conformation of the side-chain of Met192 account for the fact that the specificity pocket can bind only small residues, such as Ala, Ser or Val.     

Modeling of substrate and inhibitory complexes of histidine-aspartic protease

A three-dimensional structure of histo-aspartic protease (HAP), a pepsin-like enzyme from the causative agent of malaria Plasmodium falciparum, is suggested on the basis of homologous modeling followed by equilibration by the method of molecular dynamics. The presence of a His residue in the catalytic site instead of an Asp residue, which is characteristic of pepsin-like enzymes, and replacement of some other conserved residues in the active site make it possible for the enzyme to function by the covalent mechanism inherent in serine proteases. The detailed structures of HAP complexes with pepstatin, a noncovalent inhibitor of aspartic proteases, and phenylmethylsulfonyl fluoride, a covalent inhibitor of serine proteases, as well as with a pentapeptide substrate are discussed.     

The mechanism of replication of phi x174 DNA. XVI. Evidence that the phi x174 viral strand is synthesized discontinuously

Chymostatin is a naturally occurring inhibitor of serine proteases that have chymotryptic-like specificity. This tetrapeptide inhibitor is produced by various species of Streptomyces bacteria. Chymostatin reacts with the serine enzyme Streptomyces griseus protease A in the crystalline state to produce an adduct, the structure of which is in agreement with hemiacetal formation between the C-terminal l-phenylalaninal residue of the inhibitor and the O^γ atom of the active Ser195 residue of S. griseus protease A. The 2.8 Å difference electron density map of the complex is also consistent with the novel structural features previously deduced spectroscopically for chymostatin; i.e. an essential (for inhibition) aldehyde function in the C-terminal l-phenylalaninal residue, an unusual arnino acid, 2-(2-iminohexahydro-(4 S)-pyrimidyl)-(S)-glycine as the third residue from the C terminus and an N-terminal amino group blocked by a (1S)-carboxyphenylethyl-carbamoyl group. There is no significant movement of the active site residues of S. griseus protease A upon complexation with chymostatin.     

A collagenolytic serine protease with trypsin-like specificity from the fiddler crab Uca pugilator

A second collagenolytic serine protease has been isolated from the hepatopancreas of the fiddler crab, Uca pugilator. This enzyme cleaves the native triple helix of collagen under physiological conditions of pH, temperature, and ionic strength. In addition to its collagenolytic activity, the enzyme exhibits endopeptidase activity toward other polypeptides and small molecular weight synthetic substrates. The polypeptide bond specificity of this enzyme is similar to that of bovine trypsin as is its interaction with specific protease inhibitors. The amino-terminal sequence of this enzyme displays significant homology with other serine proteases, most notably with that of crayfish trypsin, and demonstrates that this enzyme is a member of the trypsin family of serine endopeptidases. The relatively unique action of this protease with regard to both collagenous and noncollagenous substrates has important implications concerning the specificity and mechanism of collagen degradation.     

Modeling of substrate and inhibitory complexes of histidine-aspartic protease

A three-dimensional structure of histo-aspartic protease (HAP), a pepsin-like enzyme from the causative agent of malaria Plasmodium falciparum, is suggested on the basis of homologous modeling followed by equilibration by the method of molecular dynamics. The presence of a His residue in the catalytic site instead of an Asp residue, which is characteristic of pepsin-like enzymes, and replacement of some other conserved residues in the active site make it possible for the enzyme to function by the covalent mechanism inherent in serine proteases. The detailed structures of HAP complexes with pepstatin, a noncovalent inhibitor of aspartic proteases, and phenylmethylsulfonyl fluoride, a covalent inhibitor of serine proteases, as well as with a pentapeptide substrate are discussed.     

Engineering subtilisin E for enhanced stability and activity in polar organic solvents

We examined the effect of a novel disulfide bond engineered in subtilisin E from Bacillus subtilis based on the structure of a thermophilic subtilisin-type serine protease aqualysin I. Four sites (Ser163/Ser194, Lys170/Ser194, Lys170/Glu195, and Pro172/Glu195) in subtilisin E were chosen as candidates for Cys substitutions by site-directed mutagenesis. The Cys170/Cys195 mutant subtilisin formed a disulfide bond in B. subtilis, and showed a 5-10-fold increase in specific activity for an authentic peptide substrate for subtilisin, N-succinyl-L-Ala-L-Ala-L-Pro-L-Phe-p-nitroanilide, compared with the single-Cys mutants. However, the disulfide mutant had a 50% decrease in catalytic efficiency due to a smaller k(cat) and was thermolabile relative to the wild-type enzyme, whereas it was greatly stabilized relative to its reduced form. These results suggest that an electrostatic interaction between Lys170 and Glu195 is important for catalysis and stability in subtilisin E. Interestingly, the disulfide mutant was found to be more stable in polar organic solvents, such as dimethylformamide and ethanol, than the wild-type enzyme, even under reducing conditions; this is probably due to the substitution of uncharged Cys by charged surface residues (Lys170 and Glu195). Further, the amino-terminal engineered disulfide bond (Gly61Cys/Ser98Cys) and the mutation Ile31Leu were introduced to enhance the stability and catalytic activity. A prominent 3-4-fold increase in the catalytic efficiency occurred in the quintet mutant enzyme over the range of dimethylformamide concentration (up to 40%).     

The nucleotide sequence of gene 21 of bacteriophage T4 coding for the prohead protease

We have sequenced gene 21 coding for the bacteriophage T4 prohead protease. The sequence codes for a protein of 212 amino acids (aa) with an Mr of 23,251. A second possible in-frame initiation site was also found which would code for an Mr 18,440 protein. Evidence is presented that this second site is used in vivo. The only striking homology of gp21 to other proteins is with the serine proteases. The protein is homologous to a short aa sequence around the active site, but has a His where the active site Ser is normally found. However, mutation of this His to Ser gave a functional protein that could not be inhibited by serine protease inhibitors. We have located three sites in the gene that give rise to temperature-sensitive mutations. One of these is towards the N-terminus of the gene, the other two flank the region that shows homology with serine proteases. Attempts to overproduce the protein in Escherichia coli failed due to the extreme lability of the enzyme. A frame-shift mutation in the gene was therefore constructed which allowed the synthesis of large amounts of a stable N-terminal fragment of the protein.     

A novel trypsin-like serine protease (hepsin) with a putative transmembrane domain expressed by human liver and hepatoma cells

Recombinant clones with cDNA inserts coding for a new serine protease (hepsin) have been isolated from cDNA libraries prepared from human liver and hepatoma cell line mRNA. The total length of the cDNA is approximately 1.8 kilobases and includes a 5' untranslated region, 1251 nucleotides coding for a protein of 417 amino acids, a 3' untranslated region, and a poly(A) tail. The amino acid sequence coded by the cDNA for hepsin shows a high degree of identity to pancreatic trypsin and other serine proteases present in plasma. It also exhibits features characteristic of zymogens to serine proteases in that it contains a cleavage site for protease activation and the highly conserved regions surrounding the His, Asp, and Ser residues that participate in enzyme catalysis. In addition, hepsin lacks a typical amino-terminal signal peptide. Hydropathy analysis of the protein sequence, however, revealed a very hydrophobic region of 27 amino acids starting 18 residues downstream from the apparent initiator Met. This region may serve as an internal signal sequence and a transmembrane domain. This putative transmembrane domain could be involved in anchoring hepsin to the cell membrane and orienting it in such a manner that its carboxyl terminus, containing the catalytic domain, is extracellular.     

Peptidyl boronates inhibit Salmonella enterica serovar Typhimurium Lon protease by a competitive ATP-dependent mechanism

Lon is a homo-oligomeric ATP-dependent serine protease that functions in the degradation of damaged and certain regulatory proteins. This enzyme has emerged as a novel target in the development of antibiotics because of its importance in conferring bacterial virulence. In this study, we explored the mechanism by which the proteasome inhibitor MG262, a peptidyl boronate, inhibits the peptide hydrolysis activity of Salmonella enterica serovar Typhimurium Lon. In addition, we synthesized a fluorescent peptidyl boronate inhibitor based upon the amino acid sequence of a product of peptide hydrolysis by the enzyme. Using steady-state kinetic techniques, we have shown that two peptidyl boronate variants are competitive inhibitors of the peptide hydrolysis activity of Lon and follow the same two-step, time-dependent inhibition mechanism. The first step is rapid and involves binding of the inhibitor and formation of a covalent adduct with the active site serine. This is followed by a second slow step in which Lon undergoes a conformational change or isomerization to increase the interaction of the inhibitor with the proteolytic active site to yield an overall inhibition constant of 5-20 nM. Although inhibition of serine and threonine proteases by peptidyl boronates has been detected previously, Lon is the first protease that has required the binding of ATP in order to observe inhibition.     

Functional consequences of the Kaposi's sarcoma-associated herpesvirus protease structure: regulation of activity and dimerization by conserved structural elements

The structure of Kaposi's sarcoma-associated herpesvirus protease (KSHV Pr), at 2.2 A resolution, reveals the active-site geometry and defines multiple possible target sites for drug design against a human cancer-producing virus. The catalytic triad of KSHV Pr, (Ser114, His46, and His157) and transition-state stabilization site are arranged as in other structurally characterized herpesviral proteases. The distal histidine-histidine hydrogen bond is solvent accessible, unlike the situation in other classes of serine proteases. As in all herpesviral proteases, the enzyme is active only as a weakly associated dimer (K(d) approximately 2 microM), and inactive as a monomer. Therefore, both the active site and dimer interface are potential targets for antiviral drug design. The dimer interface in KSHV Pr is compared with the interface of other herpesviral proteases. Two conserved arginines (Arg209), one from each monomer, are buried within the same region of the dimer interface. We propose that this conserved arginine may provide a destabilizing element contributing to the tuned micromolar dissociation of herpesviral protease dimers.     

Kinetic studies of human immunodeficiency virus type 1 protease and its active-site hydrogen bond mutant A28S.

Human immunodeficiency virus type 1 (HIV-1) protease optimally catalyzes in the pH range of 4-6 in contrast to nearly all of the other eukaryotic aspartic proteases, which catalyze best in the pH range of 2-4. A possible structural reason for the higher optimal pH of HIV-1 protease is the absence of a hydrogen bond to the carboxyl group of active-site Asp25, which is nearly universally present in others. To investigate this hypothesis, we have mutated residue 28 in HIV-1 protease from alanine to serine. Both the wild-type and the mutant A28S enzymes have been overexpressed in Escherichia coli using a chemically synthesized gene and purified for a comparative study in enzyme kinetics. The kcat and Km values were determined by a radiometric assay for the wild-type enzyme from pH 3.2 to 7.0, and for the mutant enzyme from pH 3.2 to 6.0. The low pK values of the active site of the free enzyme, pKe1, are 3.3 and 3.4 for the wild-type and mutant enzymes, respectively. The low pK values of the active site of the enzyme bound to substrate, pKes1, are 5.1 and 4.3 for the wild-type and mutant enzymes, respectively. The high pK values of the free enzyme, pKe2, are 6.8 and 5.6, and the corresponding ones for the substrate-bound enzyme, pKes2, are 6.9 and 6.0 for the wild-type and mutant enzymes, respectively. The lowering of pK values in mutant HIV-1 protease indicates that the hydroxyl group of Ser28 forms a new hydrogen bond to active-site Asp25 to increase its acidity.