Synthesis and sequence optimization of GFP mutants containing aromatic non-natural amino acids at the Tyr66 position

In order to alter the fluorescence properties of green fluorescent protein (GFP), aromatic non-natural amino acids were introduced into the Tyr66 position of GFP in a cell-free translation system using a four-base codon method. Two non-natural mutants (O-methyltyrosine and p-aminophenylalanine mutants) out of 18 mutants showed blue-shifted but weak fluorescence compared with wild-type GFP. Then the aminophenylalanine mutant was sequence optimized by introducing random mutations around the Tyr66 site. For this purpose, a method for random mutation of non-natural proteins in a cell-free system was developed. Three aminophenylalanine mutants with Y145F, Y145L and Y145 M mutations were obtained, which exhibited increased fluorescence by 1.5-, 3- and 4-fold, respectively. These results indicate that random mutation around non-natural amino acids is useful strategy in order to improve protein functions that are reduced by non-natural amino acid incorporation. The method described here will be applicable to other non-natural mutant proteins in a high-throughput manner.     

Engineering cytochrome P-450cam to increase the stereospecificity and coupling of aliphatic hydroxylation

Site-directed mutants were constructed in cytochrome P-450_camto re-engineer the stereochemistry and coupling of ethylbenzenehydroxyiation. The reaction with wild-type (WT) enzyme producesone regioisomer 1-phenylethanol with 5% reduced nicotinamideadenine deoxyribonucleic acid to product conversion of and aratio of 73:27 for the R and S enantiomers respectively. Ethyibenzenewas modeled into the active site of WT P-450_cam in a rigid modeand oriented to optimize either pro-R or pro-S hydrogen abstraction.Residues T101, T185 and V247 make extensive contacts with thesubstrate in the static complexes and were therefore chosenfor site-directed mutagenesis. Single mutants T101M, V247A andV247M are more stereospedik producing 89,87 and 82% (R)-1-phenylethanolrespectively. The coupling of the reaction is doubled for thesingle mutants T185L, T185F and V247M. In an effort to engineerincreased stereospecificity and coupling into a single catalystthe T101M, T185F and V247M mutants were combined in a multiplemutant of P-450_cam.This protein hydroxylates ethyibenzene resultingin an R:S ratio of 87:13 for the 1-phenylethanols and 13% couplingof reducing equivalents to product. The catalytic stereospecificityand stoichiometry with T101M–T185F–V247M does notrepresent a summation of the changes observed for the singlemutants. A portion of the individual effects on substrate recognitionproduced by the single substitutions is either eliminated ordegenerate within the triple mutant.     

Concurrent mutations in six amino acids in beta-glucuronidase improve its thermostability

To achieve a thermostable beta-glucuronidase (GUS) and identify key mutation sites, we applied in vitro directed evolution strategy through DNA shuffling and obtained a highly thermostable mutant GUS gene, gus-tr, after four rounds of DNA shuffling and screening. This variant had mutations in 15 nucleic acid sites, resulting in changes in 12 amino acids (AAs). Using gus-tr as the template, we further performed site-directed mutagenesis to reverse the individual mutation to the wild-type protein. We found that six sites (Q493R, T509A, M532T, N550S, G559S and N566S) present in GUS-TR3337, were the key AAs needed to confer its high thermostability. Of these, Q493R and T509A were not reported previously as important residues for thermostability of GUS. Furthermore, all of these six mutations must be present concurrently to confer the high thermostability. We expressed the gus-tr3337 gene and purified the GUS-TR3337 protein that contained the six AA mutations. Compared with the wild-type protein which lost its activity completely after 10 min at 70 degrees C, the mutant GUS-TR3337 protein retained 75% of its activity when heated at 80 degrees C for 10 min. The GUS-TR3337 exhibited high activity even heated at 100 degrees C for 30 min on nitrocellulose filter. The comparison of molecular models of the mutated and wild-type enzyme revealed the relation of protein function and these structural modifications.     

Selection of a thermostable variant of chloramphenicol acetyltransferase (Cat-86)

The moderate thermophile Bacillus stearothermophilus was usedas a host in which to detect more thermostable variants of theB.pumilus chloramphenicol acetyltransferase (Cat-86) protein.Seventeen mutants were isolated and detected by their abilityto grow in the presence of chloramphenicol at a previously restrictivetemperature (58°C). The genes encoding these proteins weresequenced; all 17 mutants carried the same C to T transitionthat conferred an amino acid substitution of alanine by valineat position 203 of the protein sequence. The wild-type and onemutant Cat-86 protein were purified to homogeneity using affinitychromatography, and kinetic and thermal stability studies wereundertaken. Both enzymes had similar sp. act. in the regionof 215 U/mg, with K_m values for chloramphenicol in the range13.8–15.4 µM and for acetyl CoA in the range 13.6–15.5µM. The A203V mutant shows greater stability than thewild-type Cat-86 protein at temperatures above 50°C andappears to pass through a transition state between 48 and 50°C.     

Protein engineering of the high-alkaline serine protease PB92 from Bacillus alcalophilus: functional and structural consequences of mutation at the S4 substrate binding pocket

Serine endoproteases such as trypsins and subtilisins are knownto have an extended substrate binding region that interactswith residues P6 to P3' of a substrate. In order to investigatethe structural and functional effects of replacing residuesat the S4 substrate binding pocket, the serine protease fromthe alkalophilic Bacillus strain PB92, which shows homologywith the subtilisins, was mutated at positions 102 and 126–128.Substitution of Val102 by Trp results in a 12–fold increasein activity towards succinyl-L-Ala-L-Ala-L-Pro-L-Phe-p-nitroanilide(sAAPFpNA). An X-ray structure analysis of the V102W mutantshows that the Trp side chain occupies a hydrophobic pocketat the surface of the molecule leaving a narrow crevice forthe P4 residue of a substrate. Better binding of sAAPFpNA bythe mutant compared with the wild type protein as indicatedby the kinetic data might be due to the hydrophobic interactionof Ala P4 of the substrate with the introduced Trp102 side chain.The observed difference in binding of sAAPFpNA by protease PB92and thermitase, both of which possess a Trp at position 102,is probably related to the amino acid substitutions at positions105 and 126 (in the protease PB92 numbering).Kinetic data forthe variants obtained by random mutation of residues Serl26,Prol27 and Serl28 reveal that the activity towards sAAPFpNAincreases when a hydrophobic residue is introduced at position126. An X-ray diffraction analysis was carried out for the threeprotease PB92 mutants which have residues Serl26-Prol27-Serl28replaced by Met-Ala-Gly(‘MAG’ mutant), Phe-Gln-Ser(‘FQS’ mutant) and Asn-Ser-Ala (‘NSA’mutant). Met 126 and Phel26 in the crystal structures of thecorresponding mutants are fixed in the same hydrophobic environmentas Trp102 in the V102W mutant.In contrast, Asnl26 in the ‘NSA’mutant is completely disordered in both crystal forms for whichthe structure has been determined. According to our kineticmeasurements none of the mutants with Met, Phe, Leu or Val atposition 126 binds sAAPFpNA better than the wild type enzyme.Resultsof the site-directed mutagenesis at position 127 imply thatpossible interaction of this residue with a substrate has almostno effect on activity towards sAAPFpNA and casein.     

The role of asparagine-32 in forming the receptor-binding epitope of human epidermal growth factor

The highly conserved asparagine residue at position 32 (Asn32)in the 'hinge' region of epidermal growth factor (EGF) separatesthe N- and C-terminal structural motifs of the EGF moleculeand is therefore an appropriate target for structure-functionstudies. Analogs of human EGF (hEGF) were generated in whichAsn32 was substituted with aspartate, glycine, isoleucine, lysine,proUne and tryptophan. The relative affinity of the EGF receptorfor mutant hEGF analogs was determined by radioreceptor competitionassay. A wide range of receptor affinities was observed dependingon the amino acid substitution. N32K and N32W hEGF analogs hadrelatively high receptor affinity, while the N32G and N32D analogsshowed decreased affinity, 35% and 25% respectively, relativeto wild type hEGF. However, no binding of the N32P analog wasdetected by radioreceptor competition assay. The N32P mutantdisplayed an NMR spectrum significantly different from thatof native wild type hEGF, indicating gross structural perturbation.In contrast, the N32K and N32D analogs exhibited spectra similarto that of native wild type hEGF. Genetically combining theN32D hEGF with an hEGF species having either the mutation L26Ghi the N-terminal region or L47A in the C-terminal region, generateddouble-mutant hEGF species whkh had relative affinities essentiallyequal to the product of the relative affinities of the parenthEGF mutants, indicating functionally independent changes inUgand-receptor interaction. These studies indicate the requirementfor H-bond donor functionality in the side chain of residuenumber 32 in forming a fully competent receptor-binding epitope.     

Mutations affecting the activity of urokinase-type plasminogen activator

Mutagenesis throughout the single-chain urokinase-type plasminogenactivator (scu-PA) cDNA molecule, followed by expression ofthe mutant genes and secretion of the resulting mutant proteinsfrom yeast, has been used to determine the amino acid residuesimportant for activity of scu-PA molecules. Twelve out of 13colonies secreting variant scu-PA molecules with decreased abilityto form a zone of fibrinolysis had mutant genes with a singlecodon alteration in the serine protease encoding domain (B-chain).Many of these changes are of highly conserved residues in theserine proteases and are consequently of considerable interest.A model three-dimensional structure of the protease domain ofurokinase was used to explain the basis for the effects of thesedown mutations. The model showed that the strongest down mutationsresult from either interference of the mutated side chain withsubstrate binding at the active site or the introduction ofbulky or charged groups at structurally sensitive internal positionsin the molecule. Attempts to find second site revertants offive down mutants, altered either at the plasmin activationsite or near the serine at the active site, only resulted insame-site revertants, with the original or closely related aminoacids restored.     

The crystal structure of thermostable mutants of chimeric 3-isopropylmalate dehydrogenase, 2T2M6T

A chimeric 3-isopropylmalate dehydrogenase (IPMDH), 2T2M6T,was produced by replacing the amino acid sequences of the Thermusthermophilus enzyme with those of the Bacillus subtilis enzymefrom residues 75 to 113. Decreased thermostability of the chiaiericenzyme was recovered by either evolutionary engineering (I93L)or site-directed mutagenesis (S82R). The 3-D structures of themutants have been determined by X-ray diffraction at 2.1 Åresolution. Although S82R was refined routinely, (I93L) requiredthe preliminary rigid-body refinement of each domain. The X-factorswere reduced to 0.18 for both mutants. Removal of the unfavorabletorsion angle at isoleucine 93 may have made I93L more thermostablethan 2T2M6T. In the case of S82R, the replaced arginine residuecontributed to the extra hydrogen bond with water molecules.The large replaced residue decreased the entropy of the solvent,which may have caused the improvement in enzyme thermostability.Denatu ration by heating may be interpreted from these structuralresults.     

A new method for random mutagenesis of complete genes: enzymatic generation of mutant libraries in vitro

A new efficient in vitro mutagenesis method for the generationof complete random mutant libraries, containing all possiblesingle base substitution mutations in a cloned gene is described.The method is based on controlled use of polymerases. Four populationsof DNA molecules are first generated by primer elongation sothat they terminate randomly, but always just before a knowntype of base (before A, C, G or T respectively). Each of thefour populations is then mutagenized in a separate misincorporationreaction, where the correct base can now be omitted. The regenerationof wild-type sequences can thus be efficiently avoided. Also,the misincorporating nucleotide concentrations can be optimizedto give the three possible single mutations in close to equalratio. The mutagenesis can be precisely localized within a predeterminedtarget region of any size, and vector sequences remain intact.We have mutagenized the DNA coding for the -fragment of Escherichiacoli ß-galactosidase, and identified 176 differentbase substitution mutations by sequencing. The present methodgives mutant yields of 40–60%, when the mutants containabout one amino acid change per protein molecule. All typesof base substitution mutations can be generated and deletionsare rare. The efficiency of this method permits the use of relativelyelaborate screening systems to isolate mutants of either structuralgenes or regulatory regions.     

The role of the flap residue, threonine 77, in the activation and catalytic activity of pepsin A

Flexible loops, often referred to as flaps, have been shownto play a role in catalytic mechanisms of different enzymes.Flaps at the active site regions have been observed in the crystalstructures of aspartic proteinases and their residues implicatedin the catalytic processes. This research investigated the roleof the flap residue, threonine 77, in the activation of pepsinogenand the catalytic mechanism of pepsin. Three mutants, T77S,T77V and T77G, were constructed. Differences in amino acid polarityand hydrogen bonding potential were shown to have an influenceon the activation and catalytic processes. T77S activated atthe same rate and had similar catalytic parameters as the wild-typepepsin. The activation rates of T77V and T77G were slower andtheir catalytic efficiencies lower than the wild-type. The resultsdemonstrated that the threonine 77 polar side chain played arole in a proteolysis. The contribution of the side chain tozymogen activation was associated with the proteolytic cleavageof the prosegment. It was postulated that the hydroxyl groupat position 77 provided an essential hydrogen bond that contributedto proper substrate alignment and, indirectly, to a catalyticallyfavorable geometry of the transition state.     

Homology modeling of Neurospora crassa geranylgeranyl pyrophosphate synthase: structural interpretation of mutant phenotypes

A model of the tertiary structure of the Neurospora crassa carotenogenic prenyltransferase, geranylgeranyl pyrophosphate synthase (GGPPS), is presented, based on structural homology with other prenyltransferases and on the crystal structure of recombinant avian farnesyl pyrophosphate synthase (FPPS). The conserved aspartate-rich motifs DDxx(xx)D and associated basic residues, considered to be the active sites for binding and catalysis in all prenyltransferases, are highly conserved in the N. crassa GGPPS protein, while other regions display a lower degree of sequence homology; thus the GGPPS model structure is predicted to be highly reliable in the active site region. A number of carotene-deficient mutants have been generated utilizing the repeat-induced point mutation (RIP) mechanism: mutant al-3RIP1 carries a Ser-to-Asn mutation in position 336 which falls within the predicted active site of the enzyme. Analysis of the model structure of this mutant indicates that Ser336 may be involved in substrate uptake. Two other mutants, al-3RIP3 and al-3RIP6, carry mutations in positions in the GGPPS protein, homologous to regions of the avian FPPS enzyme proposed to be involved in enzyme dimerization and substrate uptake, respectively, suggesting an explanation for the reduced carotene content of these mutants.      

Structure prediction of the EcoRV DNA methyltransferase based on mutant profiling, secondary structure analysis, comparison with known structures of methyltransferases and isolation of catalytically inactive single mutants

The EcoRV DNA methyltransferase (M·EcoRV) is an -adeninemethyltransferase. We have used two different programs to predictthe secondary structure of M·EcoRV. The resulting consensusprediction was tested by a mutant profiling analysis. 29 neutralmutations of M·EcoRV were generated by five cycles ofrandom mutagenesis and selection for active variants to increasethe reliability of the prediction and to get a secondary structureprediction for some ambiguously predicted regions. The predictedconsensus secondary structure elements could be aligned to thecommon topology of the structures of the catalytic domains ofM·HhaI and M·TaqI. In a complementary approachwe have isolated nine catalytically inactive single mutants.Five of these mutants contain an amino acid exchange withinthe catalytic domain of M·EcoRV (Val20-Ala, Lys81Arg,Cys192Arg, Asp193Gly, Trp231Arg). The Trp231Arg mutant bindsDNA similarly to wild-type M·EcoRV, but is catalyticallyinactive. Hence this mutant behaves like a bona fide activesite mutant. According to the structure prediction, Trp231 islocated in a loop at the putative active site of M·EcoRV.The other inactive mutants were insoluble. They contain aminoacid exchanges within the conserved amino acid motifs X, IIIor IV in M·EcoRV confirming the importance of these regions.     

Improvement in the alkaline stability of subtilisin using an efficient random mutagenesis and screening procedure

An efficient random mutagenesis procedure coupled to a replicaplate screen facilitated the isolation of mutant subtilisinsfrom Bacillus amyloliquefaciens that had altered autolytic stabilityunder alkaline conditions. Out of about 4000 clones screened,approximately 70 produced subtilisins with reduced stability(negatives). Two dones produced a more stable subtilisin (positives)and were identified as having a single mutation, either IIe107Valor Lys2l3Arg (the wild-type amino acid is followed by the codonposition and the mutant amino acid). One of the negative mutants,Met50Val, was at a site where other homologous subtilisins containeda Phe. When the Met50Phe mutation was introduced into the B.amyloliquefaciens gene, the mutant subtilisin was more alkalinestable. The double mutant IIe107Val/Lys2l3Arg) was more stablethan the isolated single mutant parents. The triple mutant (Met50Phe/IIel07Val/Lys2l3Arg)was even more stable than IIe107Val/Lys2l3Arg (up to two timesthe autolytic half-time of wild-type at pH 12). These studiesdemonstrate the feasibility for improving the alkaline stabilityof proteins by random mutagenesis and identifying potentialsites where substitutions from homologous proteins can improvealkaline stability.     

Stability effects associated with the introduction of a partial and a complete Ca2+-binding site into human lysozyme

Two mutants of human lysozyme were synthesized. Mutant A92D,in which Ala92 was substituted by Asp, contains a partial Ca²⁺-bindingsite and mutant M4, in which Ala83, Gm86, Asn88 and Ala92 werereplaced by Lys, Asp, Asp and Asp respectively, contains thecomplete Ca -binding site of bovine a-lactalbumin. The Ca²⁺-bindingconstants of wild type human lysozyme and of mutants A92D andM4, measured at 25C and pH 7.5, were 2(1) x 102 M"1, 8(2)x l^M"1 and 9(0.5) x 10* M"1 respectively. Information gatheredfrom mkrocalorimetrk and CD spectro-scopic measurements indicatesthat the conformational changes of the M4 mutant lysozyme, inducedby Ca²⁺ binding, are smaller than those observed for bovinea-lactalbumin and for the Ca²⁺-binding equine lysozyme. At pH4.5, the thermostability of both the apo and Ca²⁺ forms of theA92D human was decreased in comparison with that of native humanlysozyme. In particular, within the apo form of this mutantan a-helix-containing sequence was destabilized. In contrast,at the same pH the thermostability of the apo and Ca²⁺ formsof the M4 mutant lysozyme was increased. The e-ammonium groupof the Lys83 side chain is assumed to be responsible for thestabilization of the apo form of this mutant.     

Introduction of internal cysteines as conformational probes in yeast phosphoglycerate kinase

Several mutants of yeast phosphoglycerate kinase, each containingonly one internal cysteine residue, were constructed from asingle mutant devoid of cysteine. These cysteines were introducedas local conformational probes in selected buried positions.The enzyme activity, conformational characteristics and stabilityindicated that the mutations introduced only small perturbationsin the molecule. The folding–unfolding process mediatedby guanidine hydrochloride under equilibrium conditions wasstudied by following the variations in ellipticity and the reactivityof the cysteine residue towards 5,5'-dithiobis(nitrobenzoate).The process was found to be reversible except for mutant C97A,V49C,suggesting that this region located in helix I might be crucialin determining an intermediate on the folding pathway. The transitionsobtained by the two signals did not coincide, indicating thatthe local structures, in several parts inside the molecule,are more sensitive to the denaturant than the overall conformation.     

Analysis of RNA phage fr coat protein assembly by insertion, deletion and substitution mutagenesis

A structure-function analysis of the icosahedral RNA bacteriophagefr coat protein (CP) assembly was undertaken using linker-insertion,deletion and substitution mutagenesis. Mutations were specificallyintroduced into either pre-existing or artificially createdrestriction enzyme sites within fr CP gene expressed in Escherichiacoli from a recombinant plasmid. This directs synthesis of wildtype protein that undergoes self-assembly and forms capsid-likeparticles indistinguishable morphologically and immunologicallyfrom native phage particles. A series of fr CP variants containingsequence alterations in the regions which are (i) exposed onthe external surface of capsid or (ii) located on the contactingareas between CP subunits were obtained and their assembly propertiesinvestigated. The majority of mutants demonstrated reductionof assembly ability and formed either CP dimers (mutations atresidues 2, 10, 63 or 129) or both dimer and capsid structures(residue 2 or 69). The exceptions were variants demonstratingnormal assembly and containing insertions at residues 2, 50or 129 of thefr CP. A third type of assembled structure wasformed by a variant with a single amino acid substitution I104T.The aA-helix region (residues 97-111) is particularly sensitiveto mutation and any alteration in this region decreases accumulationof mutant protein in E.coli. The relative contributions of particularfr CP domains in maintenance of capsid structural integrityas well as the possible capsid assembly mechanism are discussed.     

Engineering the xenobiotic substrate specificity of maize glutathione S-transferase I

Glutathione S-transferases (GSTs) are a heterogeneous family of enzymes that catalyse the conjugation of glutathione (GSH) to electrophilic sites on a variety of hydrophobic substrates. In the present study three amino acid residues (Trp12, Phe35 and Ile118) of the xenobiotic binding site (H-site) of maize GST I were altered in order to evaluate their contribution to substrate binding and catalysis. These residues are not conserved and hence may affect substrate specificity and/or product dissociation. The results demonstrate that these residues are important structural moieties that modulate an enzyme's catalytic efficiency and specificity. Phe35 and Ile118 also participate in k(cat) regulation by affecting the rate-limiting step of the catalytic reaction. The effect of temperature on the catalytic activity of the wild-type and mutant enzymes was also investigated. Biphasic Arrhenius and Eyring plots for the wild-type enzyme showed an apparent transition temperature at 35 degrees C, which seems to be the result of a change in the rate-limiting step of the catalytic reaction. Thermodynamic analysis of the activity data showed that the activation energy increases at low temperatures, whereas the entropy change seems to be the main determinant that contributes to the rate-limiting step at high temperatures.     

Use of a minimum perturbation approach to predict TIM mutant structures

A minimum perturbation conformational search approach is usedto model the structures of the yeast triosephosphate isomerase(TIM) single mutant in which the catalytic base Glul65 is changedto Asp, and the double mutant in which Glul65 is changed toAsp and Ser96 to Pro. In chicken TIM this double mutant is referredto as a pseudo–revertant because some of the catalyticactivity lost due to the first mutation is regained when thesecond mutation occurs. Three minimum energy structures werecalculated for the Asp 165 conformation in the yeast TEM singlemutant and another three for the double mutant One of the calculatedminimum energy conformations for Aspl65 in the E165D structureagrees well with the X–ray structure. However, this conformationis not that of the lowest energy and is not one of the threemost common conformers for Asp found by Ponder and Richards.This suggests that when an amino acid is introduced it may notbe able to conform to the more general rules that apply to proteinstructures of evolutionary origin. While the van der Waals energylargely determines the allowed minima, the relative rankingof the final minima is determined by electrostatic effects andcan therefore be affected by the inclusion of crystal watersin the calculation. When the E165D calculation is repeated withan active–site water molecule fixed in its E165D X–raystructure position, the relative ranking of the minima shiftsand the X–ray conformation for Asp 165 is the lowest interactionenergy conformer. Two of the E165D calculated minimum energystructures are essentially identical to two of the S96P/E165Dminima. All of the calculated minima for both the E165D andS96P/E165D mutants position the Asp side chain such that theanti–orbital, and not the more basic syn–orbital,of the carboxylate would be utilized for proton abstraction.This observation may explain why the chicken TIM S96P/E165Dmutant, for which the X-ray structure indicates that the syn–orbitalis used, is a pseudo–revertant while the yeast TIM doublemutant is not; no X–ray structure is available for thelatter. The multiplicity of minima found in the present analysismakes clear that predicting the exact orientation of a singleside chain is not as simple as might be expected.     

Mapping of residues involved in the interaction between the Bacillus subtilis xylanase A and proteinaceous wheat xylanase inhibitors

The Bacillus subtilis xylanase A was subjected to site-directed mutagenesis, aimed at changing the interaction with Triticum aestivum xylanase inhibitor, the only wheat endogenous proteinaceous xylanase inhibitor interacting with this xylanase. The published structure of Bacillus circulans XynA was used to target amino acids surrounding the active site cleft of B.subtilis XynA for mutation. Twenty-two residues were mutated, resulting in 62 different variants. The catalytic activity of active mutants ranged from 563 to 5635 XU/mg and the interaction with T.aestivum xylanase inhibitor showed a similar variation. The results indicate that T.aestivum xylanase inhibitor interacts with several amino acid residues surrounding the active site of the enzyme. Three different amino acid substitutions in one particular residue (D11) completely abolished the interaction between T.aestivum xylanase inhibitor and B.subtilis xylanase A.     

Molecular mechanics analysis of drug-resistant mutants of HIV protease

Drug-resistant mutants of HIV-1 protease limit the long-termeffectiveness of current anti-viral therapy. In order to studydrug resistance, the wild-type HIV-1 protease and the mutantsR8Q, V32I, M46I, V82A, V82I, V82F, I84V, V32I/I84V and M46I/I84Vwere modeled with the inhibitors saquinavir and indinavir usingthe program AMMP. A new screen term was introduced to reproducemore correctly the electron distribution of atoms. The atomicpartial charge was represented as a delocalized charge distributioninstead of a point charge. The calculated protease–saquinavirinteraction energies showed the highly significant correlationof 0.79 with free energy differences derived from the measuredinhibition constants for all 10 models. Three different protonationstates of indinavir were evaluated. The best indinavir modelincluded a sulfate and gave a correlation coefficient of 0.68between the calculated interaction energies and free energiesfrom inhibition constants for nine models. The exception wasR8Q with indinavir, probably due to differences in the solvationenergy. No significant correlation was found using the standardmolecular mechanics terms. The incorporation of the new screencorrection resulted in better prediction of the effects of inhibitorson resistant protease variants and has potential for selectingmore effective inhibitors for resistant virus.