Catalytic importance of the substrate binding order for the FMNH2-dependent alkanesulfonate monooxygenase enzyme

The two-component alkanesulfonate monooxygenase system from Escherichia coli includes an FMN reductase (SsuE) and an FMNH2-dependent alkanesulfonate monooxygenase (SsuD) involved in the acquisition of sulfur from alkanesulfonates during sulfur starvation. The SsuD enzyme directly catalyzes the oxidation of alkanesulfonate to aldehyde and sulfite in the presence of O2 and FMNH2. The goal of these studies was to investigate the kinetic mechanism of SsuD through rapid reaction kinetics and substrate binding studies. The SsuD enzyme shows a clear preference for FMNH2 (Kd, 0.32 +/- 0.15 microM) compared to FMN (Kd, 10.2 +/- 0.4 microM) with a 1:1 binding stoichiometry for each form of the flavin. The kinetic trace of premixed SsuD and FMNH2 mixed with oxygenated buffer was best fit to a double exponential with no observed formation of the C4a-(hydro)peroxyflavin. However, when FMNH2 was mixed with SsuD and oxygenated buffer an initial fast phase (kobs, 12.9 s-1) was observed, suggesting that the mixing order is critical for the accumulation of the C4a-(hydro)peroxyflavin. Results from fluorimetric titrations with octanesulfonate imply that reduced flavin must bind first to promote octanesulfonate binding. When octanesulfonate was included in the kinetic studies the C4a-(hydro)peroxyflavin was observed at 370 nm when FMNH2 was not premixed with SsuD, which correlated with an increase in octanal product. There was a clear hyperbolic dependence on octanesulfonate binding, indicating that octanesulfonate binds in rapid equilibrium, and further results indicated there was a second isomerization step following binding. These results suggest that an ordered substrate binding mechanism is important in the desulfonation reaction by SsuD with reduced flavin binding first followed by either O2 or octanesulfonate.     

Characterization of a two-component alkanesulfonate monooxygenase from Escherichia coli.

The Escherichia coli ssuEADCB gene cluster is required for the utilization of alkanesulfonates as sulfur sources, and is expressed under conditions of sulfate or cysteine starvation. The SsuD and SsuE proteins were overexpressed and characterized. SsuE was purified to homogeneity as an N-terminal histidine-tagged fusion protein. Native SsuE was a homodimeric enzyme of M(r) 58,400, which catalyzed an NAD(P)H-dependent reduction of FMN, but it was also able to reduce FAD or riboflavin. The SsuD protein was purified to >98% purity using cation exchange, anion exchange, and hydrophobic interaction chromatography. The pure enzyme catalyzed the conversion of pentanesulfonic acid to sulfite and pentaldehyde and was able to desulfonate a wide range of sulfonated substrates including C-2 to C-10 unsubstituted linear alkanesulfonates, substituted ethanesulfonic acids and sulfonated buffers. SsuD catalysis was absolutely dependent on FMNH(2) and oxygen, and was maximal for SsuE/SsuD molar ratios of 2.1 to 4.2 in 10 mM Tris-HCl, pH 9.1. Native SsuD was a homotetrameric enzyme of M(r) 181,000. These results demonstrate that SsuD is a broad range FMNH(2)-dependent monooxygenase catalyzing the oxygenolytic conversion of alkanesulfonates to sulfite and the corresponding aldehydes. SsuE is the FMN reducing enzyme providing SsuD with FMNH(2).     

Crystal structure of Escherichia coli alkanesulfonate monooxygenase SsuD

The FMNH(2)-dependent alkanesulfonate monooxygenase SsuD catalyzes the conversion of alkanesulfonates to the corresponding aldehyde and sulfite. The enzyme allows Escherichia coli to use a wide range of alkanesulfonates as sulfur sources for growth when sulfate or cysteine are not available. The structure of SsuD was solved using the multiwavelength anomalous dispersion method from only four ordered selenium sites per asymmetric unit (one site per 20,800 Da). The final model includes 328 of 380 amino acid residues and was refined to an R-factor of 23.5% (R(free)=27.5%) at 2.3A resolution. The X-ray crystal structure of SsuD shows a homotetrameric state for the enzyme, each subunit being composed of a TIM-barrel fold enlarged by four insertion regions that contribute to intersubunit interactions. SsuD is structurally related to a bacterial luciferase and an archaeal coenzyme F(420)-dependent reductase in spite of a low level of sequence identity with these enzymes. The structural relationship is not limited to the beta-barrel region; it includes most but not all extension regions and shows distinct properties for the SsuD TIM-barrel. A likely substrate-binding site is postulated on the basis of the SsuD structure presented here, results from earlier biochemical studies, and structure relatedness to bacterial luciferase. SsuD is related to other FMNH(2)-dependent monooxygenases that show distant sequence relationship to luciferase. Thus, the structure reported here provides a model for enzymes belonging to this family and suggests that they might all fold as TIM-barrel proteins.     

Altered mechanism of the alkanesulfonate FMN reductase with the monooxygenase enzyme

The two-component alkanesulfonate monooxygenase system from Escherichia coli is comprised of an FMN reductase (SsuE) and a monooxygenase enzyme (SsuD) that together catalyze the oxidation of alkanesulfonate to the corresponding aldehyde and sulfite products. To determine the effects of protein interactions on catalysis, the steady-state kinetic parameters for SsuE were determined in single-enzyme assays and in the presence of the monooxygenase enzyme and alkanesulfonate substrate. In single-enzyme kinetic assays, SsuE followed an ordered sequential mechanism, with NADPH as the first substrate to bind and NADP+ as the last product to dissociate. However, in the presence of SsuD and octanesulfonate the kinetic mechanism of SsuE is altered to a rapid equilibrium ordered mechanism, and the Km value for FMN is increased 10-fold. These results suggest that both the SsuD enzyme and alkanesulfonate substrate are required to ensure that the FMN reductase reaction proceeds to form the ternary complex with the subsequent generation of reduced flavin transfer.     

Detection of protein-protein interactions in the alkanesulfonate monooxygenase system from Escherichia coli

The two-component alkanesulfonate monooxygenase system utilizes reduced flavin as a substrate to catalyze a unique desulfonation reaction during times of sulfur starvation. The importance of protein-protein interactions in the mechanism of flavin transfer was analyzed in these studies. The results from affinity chromatography and cross-linking experiments support the formation of a stable complex between the flavin mononucleotide (FMN) reductase (SsuE) and monooxygenase (SsuD). Interactions between the two proteins do not lead to overall conformational changes in protein structure, as indicated by the results from circular dichroism spectroscopy in the far-UV region. However, subtle changes in the flavin environment of FMN-bound SsuE that occur in the presence of SsuD were identified by circular dichroism spectroscopy in the visible region. These data are supported by the results from fluorescent spectroscopy experiments, where a dissociation constant of 0.0022 +/- 0.0010 muM was obtained for the binding of SsuE to SsuD. Based on these studies, the stoichiometry for protein-protein interactions is proposed to involve a 1:1 monomeric association of SsuE with SsuD.     

Deletional studies to investigate the functional role of a dynamic loop region of alkanesulfonate monooxygenase

Several bacterial organisms rely on the two-component alkanesulfonate monooxygenase system for the acquisition of organosulfonate compounds when inorganic sulfur is limiting in the environment. This system is comprised of an FMN reductase (SsuE) that supplies reduced flavin to the alkanesulfonate monooxygenase (SsuD). Desulfonation of alkanesulfonates by SsuD is catalyzed through the activation of dioxygen by reduced flavin. The three-dimensional structure of SsuD exists as a TIM-barrel fold with several discrete insertion regions. An extensive insertion region near the putative active site was disordered in the SsuD structure, suggesting the importance of protein dynamics in the desulfonation mechanism. Three variants containing a partial deletion of the loop region were constructed to evaluate the functional properties of this region. There were no overall gross changes in secondary structure for the three SsuD deletion variants compared to wild-type SsuD, but each variant was found to be catalytically inactive. The deletion variants were unable to undergo the conformational changes necessary for catalysis even though they were able to bind reduced flavin. Rapid kinetic analyses monitoring the reductive and oxidative half-reactions indicated that the SsuD deletion variants failed to protect reduced flavin from unproductive oxidation. These studies define the importance of dynamic loop region for protection and stabilization of reduced flavin and reaction intermediates.     

Molecular dynamics simulations of lignin peroxidase in solution

The dynamical and structural properties of lignin peroxidase and its Trp171Ala mutant have been investigated in aqueous solution using molecular dynamics (MD) simulations. In both cases, the enzyme retained its overall backbone structure and all its noncovalent interactions in the course of the MD simulations. Very interestingly, the analysis of the MD trajectories showed the presence of large fluctuations in correspondence of the residues forming the heme access channel; these movements enlarge the opening and facilitate the access of substrates to the enzyme active site. Moreover, steered molecular dynamics docking simulations have shown that lignin peroxidase natural substrate (veratryl alcohol) can easily approach the heme edge through the access channel.     

In silico analyses of substrate interactions with human serum paraoxonase 1

Human paraoxonase (HuPON1) is a serum enzyme that exhibits a broad spectrum of hydrolytic activities, including the hydrolysis of various organophosphates, esters, and recently identified lactone substrates. Despite intensive site-directed mutagenesis and other biological studies, the structural basis for the specificity of substrate interactions of HuPON1 remains elusive. In this study, we apply homology modeling, docking, and molecular dynamic (MD) simulations to probe the binding interactions of HuPON1 with representative substrates. The results suggest that the active site of HuPON1 is characterized by two distinct binding regions: the hydrophobic binding site for arylesters/lactones, and the paraoxon binding site for phosphotriesters. The unique binding modes proposed for each type of substrate reveal a number of key residues governing substrate specificity. The polymorphic residue R/Q192 interacts with the leaving group of paraoxon, suggesting it plays an important role in the proper positioning of this substrate in the active site. MD simulations of the optimal binding complexes show that residue Y71 undergoes an "open-closed" conformational change upon ligand binding, and forms strong interactions with substrates. Further binding free energy calculations and residual decomposition give a more refined molecular view of the energetics and origin of HuPON1/substrate interactions. These studies provide a theoretical model of substrate binding and specificity associated with wild type and mutant forms of HuPON1, which can be applied in the rational design of HuPON1 variants as bioscavengers with enhanced catalytic activity.     

Regiospecificity of human cytochrome P450 1A1-mediated oxidations: the role of steric effects

Cytochrome P450 1A1 oxidizes a diverse range of substrates, including the procarcinogenic xenobiotic benzo[a]pyrene (B[a]P) and endogenous fatty acid precursors of prostaglandins, such as arachidonic acid (AA) and eicosapentaenoic acid (EA). We have investigated the extent to which enzyme-substrate interactions govern regio- and stereoselectivity of oxidation of these compounds by using docking and molecular dynamics (MD) simulations to examine the likelihood of substrate oxidation at various sites. Due to structural differences between the substrates analyzed, B[a]P and its diols (planar, rigid), and the fatty acids AA and EA (long, flexible), different docking strategies were required. B[a]P, B[a]P-7,8-diols, (+) 7S,8S- and (-) 7R,8R-diols, were docked into the active site of a homology model of P450 1A1 using an automated routine, Affinity (Accelrys, San Diego, CA). AA and EA, on the other hand, required a series of restrained MD simulations to obtain a variety of productive binding modes. All complexes were evaluated by MD-based in silico site scoring to predict product profiles based on certain geometric criteria, such as angle and distance of a given substrate atom from the ferryl oxygen. For all substrates studied, the in vitro profiles were generally reflected by the in silico scores, which suggests that steric factors play a key role in determining regiospecificity in P450 1A1-mediated oxidations. We have also shown that molecular dynamics simulations may be very useful in determination of product profiles for structurally diverse substrates of P450 enzymes.     

Functional role of a conserved arginine residue located on a mobile loop of alkanesulfonate monooxygenase

The structure of the flavin-dependent alkanesulfonate monooxygenase (SsuD) exists as a TIM-barrel structure with an insertion region located over the active site that contains a conserved arginine (Arg297) residue present in all SsuD homologues. Substitution of Arg297 with alanine (R297A SsuD) or lysine (R297K SsuD) was performed to determine the functional role of this conserved residue in SsuD catalysis. While the more conservative R297K SsuD possessed a lower k(cat)/K(m) value (0.04 ± 0.01 μM(-1) min(-1)) relative to wild-type (1.17 ± 0.22 μM(-1) min(-1)), there was no activity observed with the R297A SsuD variant. Each of the arginine variants had similar K(d) values for flavin binding as wild-type SsuD (0.32 ± 0.15 μM), but there was no measurable binding of octanesulfonate. The low levels of activity for the R297A and R297K SsuD variants correlated with the absence of any detectable C4a-(peroxy)flavin formation in stopped-flow kinetic studies. Single-turnover experiments were performed in the presence of SsuE to evaluate both the reductive and oxidative half-reaction. With wild-type SsuD a lag phase is observed following the reductive half-reaction by SsuE that represents flavin transfer or conformational changes associated with the binding of substrates. Evaluation of the Arg297 SsuD variants in the presence of SsuE showed no lag phase following reduction by SsuE, and the flavin was oxidized immediately following the reductive half-reaction. These results corresponded with a lack of detectable changes in the proteolytic susceptibility of R297A and R297K SsuD in the presence of reduced flavin and/or octanesulfonate, signifying the absence of a conformational change in these variants with the substitution of Arg297.     

Design and realization of a tailor-made enzyme to modify the molecular recognition of 2-arylpropionic esters by Candida rugosa lipase

Within a research project aimed at probing the substrate specificity and the enantioselectivity of Candida rugosa lipase (CRL), computer modeling studies of the interactions between CRL and methyl (+/-)-2-(3-benzoylphenyl)propionate (Ketoprofen methyl ester) have been carried out in order to identify which amino acids are essential to the enzyme/substrate interaction. Different binding models of the substrate enantiomers to the active site of CRL were investigated by applying a computational protocol based on molecular docking, conformational analysis, and energy minimization procedures. The structural models of the computer generated complexes between CRL and the substrates enabled us to propose that Phe344 and Phe345, in addition to the residues constituting the catalytic triad and the oxyanion hole, are the amino acids mainly involved in the enzyme-ligand interactions. To test the importance of these residues for the enzymatic activity, site-directed mutagenesis of the selected amino acids has been performed, and the mutated enzymes have been evaluated for their conversion and selectivity capabilities toward different substrates. The experimental results obtained in these biotransformation reactions indicate that Phe344 and especially Phe345 influence CRL activity, supporting the findings of our theoretical simulations.     

Molecular Modeling Study for Inhibition Mechanism of Human Chymase and Its Application in Inhibitor Design

Human chymase catalyzes the hydrolysis of peptide bonds. Three chymase inhibitors with very similar chemical structures but highly different inhibitory profiles towards the hydrolase function of chymase were selected with the aim of elucidating the origin of disparities in their biological activities. As a substrate (angiotensin-I) bound crystal structure is not available, molecular docking was performed to dock the substrate into the active site. Molecular dynamics simulations of chymase complexes with inhibitors and substrate were performed to calculate the binding orientation of inhibitors and substrate as well as to characterize conformational changes in the active site. The results elucidate details of the 3D chymase structure as well as the importance of K40 in hydrolase function. Binding mode analysis showed that substitution of a heavier Cl atom at the phenyl ring of most active inhibitor produced a great deal of variation in its orientation causing the phosphinate group to interact strongly with residue K40. Dynamics simulations revealed the conformational variation in region of V36-F41upon substrate and inhibitor binding induced a shift in the location of K40 thus changing its interactions with them. Chymase complexes with the most activecompound and substrate were used for development of a hybrid pharmacophore model which was applied in databases screening. Finally, hits which bound well at the active site, exhibited key interactions and favorable electronic properties were identified as possible inhibitors for chymase. This study not only elucidates inhibitory mechanism of chymase inhibitors but also provides key structural insights which will aid in the rational design of novel potent inhibitors of the enzyme. In general, the strategy applied in the current study could be a promising computational approach and may be generally applicable to drug design for other enzymes.     

Molecular docking and spatial coarse graining simulations as tools to investigate substrate recognition, enhancer binding and conformational transitions in indoleamine-2,3-dioxygenase (IDO)

Indoleamine 2,3-dioxygenase (IDO) is an heme-containing enzyme involved in the regulation of important immunological responses and neurological processes. The enzyme catalyzes the oxidative cleavage of the pyrrole ring of the indole nucleus of tryptophan (Trp) to yield N-formylkynurenine, that is the initial and rate limiting step of the kynurenine pathway. Some indole derivatives have been reported to act as effectors of the enzyme by enhancing its catalytic activity. On the basis of the recent availability of the crystal structure of IDO, in this work we investigate substrate recognition and enhancer binding to IDO using molecular docking experiments. In addition, conformational transitions of IDO in response to substrate and enhancer binding are studied using coarse graining simulations with the program FIRST. The results enable us to identify (i) the binding site of enhancer modulators; (ii) the motion of an electrostatic gate that regulates the access of the substrate to the catalytic site of the enzyme; (iii) the movement of the anchoring region of a hairpin loop that may assist the shuttle of substrates/products to/from the catalytic site of IDO. These data, combined with available site-directed mutagenesis experiments, reveal that conformational transitions of IDO in response to substrate and enhancer binding are controlled by distinct combination of two conformational states (open and close) of the above structural motifs. On this basis, a molecular mechanism regarding substrate recognition and activity enhancement by indole derivatives is proposed.     

Structure-activity relationship of sphingomyelin analogs with sphingomyelinase from Bacillus cereus

The aim of this study was to examine how structural properties of different sphingomyelin (SM) analogs affected their substrate properties with sphingomyelinase (SMase) from Bacillus cereus. Using molecular docking and dynamics simulations (for SMase-SM complex), we then attempted to explain the relationship between SM structure and enzyme activity. With both micellar and monolayer substrates, 3O-methylated SM was found not to be degraded by the SMase. 2N-methylated SM was a substrate, but was degraded at about half the rate of its 2NH-SM control. PhytoPSM was readily hydrolyzed by the enzyme. PSM lacking one methyl in the phosphocholine head group was a good substrate, but PSM lacking two or three methyls failed to act as substrates for SMase. Based on literature data, and our docking and MD simulations, we conclude that the 3O-methylated PSM fails to interact with Mg(2+) and Glu53 in the active site, thus preventing hydrolysis. Methylation of 2NH was not crucial for binding to the active site, but appeared to interfere with an induced fit activation of the SMase via interaction with Asp156. An OH on carbon 4 in the long-chain base of phytoPSM appeared not to interfere with the 3OH interacting with Mg(2+) and Glu53 in the active site, and thus did not interfere with catalysis. Removing two or three methyls from the PSM head group apparently increased the positive charge on the terminal N significantly, which most likely led to ionic interactions with Glu250 and Glu155 adjacent to the active site. This likely interaction could have misaligned the SM substrate and hindered proper catalysis.     

Identification of critical steps governing the two-component alkanesulfonate monooxygenase catalytic mechanism

The alkanesulfonate monooxygenase enzyme (SsuD) catalyzes the oxygenolytic cleavage of a carbon-sulfur bond from sulfonated substrates. A mechanism involving acid-base catalysis has been proposed for the desulfonation mechanism by SsuD. In the proposed mechanism, base catalysis is involved in abstracting a proton from the alkane peroxyflavin intermediate, while acid catalysis is needed for the protonation of the FMNO(-) intermediate. The pH profiles of k(cat) indicate that catalysis by SsuD requires a group with a pK(a) of 6.6 ± 0.2 to be deprotonated and a second group with a pK(a) of 9.5 ± 0.1 to be protonated. The upper pK(a) value was not present in the pH profiles of k(cat)/K(m). Several conserved amino acid residues (His228, His11, His333, Cys54, and Arg226) have been identified as having potential catalytic importance due to the similar spatial arrangements with close structural and functional relatives of SsuD. Substitutions to these amino acid residues were generated, and the pH dependencies were evaluated and compared to wild-type SsuD. Although a histidine residue was previously proposed to be the active site base, the His variants possessed similar steady-state kinetic parameters as wild-type SsuD. Interestingly, R226A and R226K SsuD variants possessed undetectable activity, and there was no detectable formation of the C4a-(hydro)peroxyflavin intermediate for the Arg226 SsuD variants. Guanidinium rescue with the R226A SsuD variant resulted in the recovery of 1.5% of the wild-type SsuD k(cat) value. These results implicate Arg226 playing a critical role in catalysis and provide essential insights into the mechanistic steps that guide the SsuD desulfonation process.     

Towards the identification of the allosteric Phe-binding site in phenylalanine hydroxylase

The enzyme phenylalanine hydroxylase (PAH) is defective in the inherited disorder phenylketonuria. PAH, a tetrameric enzyme, is highly regulated and displays positive cooperativity for its substrate, Phe. Whether Phe binds to an allosteric site is a matter of debate, despite several studies worldwide. To address this issue, we generated a dimeric model for Phe–PAH interactions, by performing molecular docking combined with molecular dynamics simulations on human and rat wild-type sequences and also on a human G46S mutant. Our results suggest that the allosteric Phe-binding site lies at the dimeric interface between the regulatory and the catalytic domains of two adjacent subunits. The structural and dynamical features of the site were characterized in depth and described. Interestingly, our findings provide evidence for lower allosteric Phe-binding ability of the G46S mutant than the human wild-type enzyme. This also explains the disease-causing nature of this mutant.     

Molecular modeling of formate dehydrogenase: the formation of the Michaelis complex

The formation of the reactive enzyme–substrate complex of formate dehydrogenase has been investigated by molecular dynamics techniques accounting for different conformational states of the enzyme. Simulations revealed that the transport of substrate to the active site through the substrate channel proceeds in the open conformation of enzyme due to the crucial role of the Arg284 residue acting as a vehicle. However, formate binding in the active site of the open conformation leads to the formation of a nonproductive enzyme–substrate complex. The productive Michaelis complex is formed only in the closed enzyme conformation after the substrate and coenzyme have bound, when required rigidity of the binding site and reactive formate orientation due to interactions with Arg284, Asn146, Ile122, and His332 residues is attained. Then, the high occupancy (up to 75%) of the reactive substrate–coenzyme conformation is reached, which was demonstrated by hybrid quantum mechanics/molecular mechanics simulations using various semiempirical Hamiltonians.     

Molecular docking and 3-D-QSAR studies on the possible antimalarial mechanism of artemisinin analogues

Artemisinin (Qinghaosu) is a natural constituent found in Artemisia annua L, which is an effective drug against chloroquine-resistant Plasmodium falciparum strains and cerebral malaria. The antimalarial activities of artemisinin and its analogues appear to be mediated by the interactions of the drugs with hemin. In order to understand the antimalarial mechanism and the relationship between the physicochemical properties and the antimalarial activities of artemisinin analogues, we performed molecular docking simulations to probe the interactions of these analogues with hemin, and then performed three-dimensional quantitative structure-activity relationship (3-D-QSAR) studies on the basis of the docking models employing comparative molecular force fields analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA). Molecular docking simulations generated probable 'bioactive' conformations of artemisinin analogues and provided a new insight into the antimalarial mechanism. The subsequent partial least squares (PLS) analysis indicates that the calculate binding energies correlate well with the experimental activity values. The CoMFA and CoMSIA models based on the bioactive conformations proved to have good predictive ability and in turn match well with the docking result, which further testified the reliability of the docking model. Combining these results, that is molecular docking and 3-D-QSAR, together, the binding model and activity of new synthesized artemisinin derivatives were well explained.     

Dynamics of Preferential Substrate Recognition in HIV-1 Protease: Redefining the Substrate Envelope

Human immunodeficiency virus type 1 (HIV-1) protease (PR) permits viral maturation by processing the gag and gag-pro-pol polyproteins. HIV-1 PR inhibitors (PIs) are used in combination antiviral therapy but the emergence of drug resistance has limited their efficacy. The rapid evolution of HIV-1 necessitates consideration of drug resistance in novel drug design. Drug-resistant HIV-1 PR variants no longer inhibited efficiently, continue to hydrolyze the natural viral substrates. Though highly diverse in sequence, the HIV-1 PR substrates bind in a conserved three-dimensional shape we termed the substrate envelope. Earlier, we showed that resistance mutations arise where PIs protrude beyond the substrate envelope, because these regions are crucial for drug binding but not for substrate recognition. We extend this model by considering the role of protein dynamics in the interaction of HIV-1 PR with its substrates. We simulated the molecular dynamics of seven PR-substrate complexes to estimate the conformational flexibility of the bound substrates. Interdependence of substrate-protease interactions might compensate for variations in cleavage-site sequences and explain how a diverse set of sequences are recognized as substrates by the same enzyme. This diversity might be essential for regulating sequential processing of substrates. We define a dynamic substrate envelope as a more accurate representation of PR-substrate interactions. This dynamic substrate envelope, described by a probability distribution function, is a powerful tool for drug design efforts targeting ensembles of resistant HIV-1 PR variants with the aim of developing drugs that are less susceptible to resistance.