NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Coffin JM, Hughes SH, Varmus HE, editors. Retroviruses. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1997.

Cover of Retroviruses


Show details

The Retroviral Protease

Proteolytic processing at specific sites in the Gag and Gag-Pro-Pol (and sometimes Env) precursors by the viral PR is an essential step in the viral life cycle. Since PR has a central role in proteolytic processing, it provides an important target for the design of inhibitors of viral replication. This section describes the retroviral PR by way of introduction to its central role late in the viral life cycle.

Identification of the Viral Protease as an Aspartic Proteinase

Evidence for a role for a PR in the viral life cycle came from the observation that the Gag protein is initially synthesized as a larger precursor that is subsequently processed through a series of intermediates to the smaller protein products found in the mature virion (Vogt and Eisenman 1973; Vogt et al. 1975). Subsequently, a virion-associated PR was detected that had the specificity to produce these proteolytic cleavages (Von der Helm 1977; Yoshinaka and Luftig 1977a; Dittmar and Moelling 1978; Vogt et al. 1979; Lillehoj et al. 1988; for review, see Oroszlan and Luftig 1990). A demonstration of the critical role of PR in the viral life cycle came from genetic studies—initially with MLV and then with HIV-1—in which it was shown that mutations placed near the 5′end of the pol gene resulted in the production of noninfectious viral particles containing unprocessed Gag and Gag-Pro-Pol precursors (Crawford and Goff 1985; Katoh et al. 1985; Kohl et al. 1988). These experiments defined the pro gene downstream from gag and upstream of the RT-coding domain of the pol gene (see also Levin et al. 1984; Kramer et al. 1986; Farmerie et al. 1987). The pro gene appears in different reading frames relative to gag and pol, depending on the genus (see Chapter 2 and Fig. 5).

A major step was taken in our understanding of PR when sequence alignments revealed similarities with the aspartic proteinase family of enzymes (Toh et al. 1985). Proteinases are grouped into families based on the chemical entity that effects hydrolysis of the peptide bond. All aspartic proteinases share the feature that two aspartic acid residues, each placed in the highly conserved motif Asp-Thr/Ser-Gly, coordinate a water molecule used to hydrolyze the target peptide bond. Sequence alignments of retroviral genomes showed that this conserved motif is encoded in all known retroviruses.

Two features of aspartic proteinases are of special interest with respect to the viral PR. First, this class of enzymes appears to be restricted to eukaryotes, although it includes examples from highly diverse organisms such as pepsin and renin from mammals, rhizopus-pepsin from the fungus Rhizopus chinensis, and the ubiquitous lysosomal enzyme cathepsin D (Davies 1990). Second, the cellular versions of this enzyme are pseudodimers in which each of the two Asp-Thr-Gly motifs is part of a discrete domain that appears to have arisen by a gene duplication and fusion to encode the single-chain proteinase (Tang et al. 1978). The retroviral PR has only a single Asp-Thr-Gly motif and is too small (on the order of 100 amino acids long) to represent both domains of its cellular counterpart. This led to a model of the viral PR which predicted the existence of a dimer to make the active form of the enzyme (Pearl and Taylor 1987), a prediction that was borne out by subsequent structural studies (see below). Biochemical evidence also supports the assignment of retroviral PRs to the aspartic proteinase family of enzymes as shown by sensitivity to aspartic proteinase inhibitors (Katoh et al. 1987; Seelmeier et al. 1988; Richards et al. 1989), features of the catalytic mechanism (Meek et al. 1989; Hyland et al. 1991a,b), and sensitivity of the active site region to mutagenesis (Kohl et al. 1988; Le Grice et al. 1988; Nam et al. 1988; Loeb et al. 1989a).

Three-dimensional Structure of the Protease

The first structures of a retroviral protein were determined for the PR of ASLV (Miller et al. 1989a) and HIV-1 (Lapatto et al. 1989; Navia et al. 1989; Wlodawer et al. 1989). Several hundred structures of the HIV-1 PR have now been determined with various bound inhibitors; however, only a fraction of these structures has been published (for review, see Wlodawer and Erickson 1993). To date, the only other viral PR structures reported are those of HIV-2 (Mulichak et al. 1993; Tong et al. 1993), SIV (Rose et al. 1993; Zhao et al. 1993), FIV (Wlodawer et al. 1995), and EIAV (Gustchina et al. 1996), although others are certain to be described. The structures of these PRs reveal certain features that are likely to be conserved among all retroviral PRs, and these features can be seen in the ribbon diagram shown in Figure 13.

Figure 13. Ribbon diagram of the dimeric retroviral protease.

Figure 13

Ribbon diagram of the dimeric retroviral protease. The structures of the HIV (brown), FIV (cyan), EIAV (yellow), and ASLV (purple) PR proteins are superimposed. The catalytic Asp residues (25 and 125) are also shown. Note the strong conservation of structure, (more...)

The protein is a dimer of two identical subunits. The aspartic acid residue in the active site lies in a loop that approaches the center of the dimer. This loop interdigitates with another loop that makes up the hydrophobic core of the enzyme. The two interlocking loops are characteristic of the ψ loop structure seen in the eukaryotic aspartic proteinases. The two subunits of the dimer are linked by a four-stranded antiparallel β-sheet involving both the amino and carboxyl termini of each subunit, in addition to other intersubunit interactions. There is a long cleft between the subunits where the substrate polypeptide binds; on the floor of this cleft are the catalytically important aspartic acids. Finally, each subunit has a segment (called the flap) of antiparallel β-sheet with a β-turn that extends over the substrate-binding cleft (at the top of the structure shown in Fig. 13).

A comparison of the structures of the HIV-1 PR with and without a bound inhibitor shows that PR undergoes significant structural changes with binding (Miller et al. 1989b). The most dramatic change is in the flap, which has been estimated to move by as much as 15 Å, allowing the substrate to enter the enzyme and the cleaved products to leave the enzyme (Gustchina and Weber 1990). Interior regions of the PR dimer also move, although these movements are small compared to those of the flaps.

PR hydrolyzes on the order of 10 of the 1500 peptide bonds present in the Gag and Gag-Pro-Pol precursors. The specificity in the choice of peptide bonds to be cleaved comes not from the catalytic mechanism of peptide bond hydrolysis but rather through the interaction between PR and its protein substrate. One of the major interests in the structure of PR is in using the structure to understand the basis of PR specificity. By convention, the peptide bond that is cleaved is referred to as the scissile bond, with the flanking amino acids going toward the amino terminus referred to as P1, P2, P3, etc. Flanking amino acids going toward the carboxyl terminus are named P1′, P2′, etc., with the scissile bond lying between the P1 and P1′amino acids. Biochemical characterization of PR shows that peptide substrates seven amino acids long (P4–P3′) are efficiently cleaved by PR, and shorter peptides are cleaved less efficiently or not at all (Darke et al. 1988; Kotler et al. 1988; Moore et al. 1989; Toth and Marshall 1990; Tozser et al. 1991). Thus, the evidence suggests that PR interacts with substrate over at least seven amino acids. The substrate side chains extend into pockets or subsites in the PR substrate-binding cleft, with the subsite named for the corresponding substrate side chain (i.e., the S1 subsite for the P1 side chain, etc). The available structural information about PR-substrate interaction comes from structures of peptide-mimetic inhibitors bound to PR (for review, see Fitzgerald and Springer 1991; Appelt 1993; Wlodawer and Erickson 1993; Ringe 1994).

Most of the structural studies on interaction of PR with its substrate have been done with the HIV-1 PR bound to a noncleavable inhibitor. The inhibitors bind in an extended β-sheet conformation in the substrate-binding cleft, and it is assumed that the bound substrate has a similar conformation. PR interacts both with the peptide backbone of the inhibitor and with the side chains that extend into the subsites. Most of the interactions with the backbone of the bound inhibitor occur with backbone carbonyls and amides of PR (Fig. 14). An additional main-chain interaction with the inhibitor involves the side chain carboxylate of the conserved PR residue Asp-29 with a main-chain nitrogen of the inhibitor, suggesting that this side-chain/main-chain interaction is conserved among retroviral PRs. Two other main-chain interactions are made flanking the cleavage site of the bound inhibitor through a water molecule coordinated by the main-chain nitrogens from residue Ile-50 present in the β-turn in each flap. The importance of this latter interaction has been demonstrated through the use of a synthetic enzyme made without the main-chain nitrogen of the flap residue (Baca and Kent 1993). The interactions with the main-chain atoms of the inhibitor/substrate correctly orient the peptide backbone in the substrate-binding cleft, but they provide no basis for specificity of site selection. The correct identification of cleavage sites must occur through the interactions with side chains of the inhibitor/substrate.

Figure 14. Schematic diagram of hydrogen bonding interactions between HIV-1 PR and a putative substrate.

Figure 14

Schematic diagram of hydrogen bonding interactions between HIV-1 PR and a putative substrate. Eight amino acids of the substrate are shown in the diagram. The main chain of the substrate, which as a β-sheet conformation is shown in the center (more...)

The side chains of the inhibitor are oriented into the PR subsites, where the interactions between substrate and PR side chains occur (Fig. 14). These interactions determine the specificity of PR in recognizing processing sites in Gag, Pro, and Pol (see below). Although the PR dimer is symmetric, it binds inhibitors asymmetrically (Miller et al. 1989b; Dreyer et al. 1993; Hosur et al. 1994), with subtle differences in the interactions between the amino half and the carboxyl half of the bound inhibitor. Subsites are evident in PR for substrate side chains P3 through P3′, with evidence of weaker interactions for the P4 and P4′side chains. Because of the β-sheet nature of the bound inhibitor, alternating side chains extend in the same direction along the substrate-binding cleft. In this orientation, there is opportunity for the P1 and P3 (as well as the P1′and P3′) side chains to interact.

Important residues critical for enzyme activity have been identified through mutagenesis. The importance of highly conserved residues for catalytic activity has been shown for both the HIV-1 PR (Loeb et al. 1989b; Louis et al. 1989) and the ASLV PR (Grinde et al. 1992a). In addition, mutations that alter enzyme activity as a function of pH have been documented (Ido et al. 1991; Konvalinka et al. 1992). A structural determinant has also been examined by fixing the orientation of a variably structured β-turn in the HIV-1 PR without affecting activity (Baca et al. 1993).

The diversity between the PR domains of HIV-1 and visna virus (both lentiviruses) is as great as that between human and fungal aspartic proteinases (Doolittle et al. 1989). Only about 20 residues of the PR are either invariant or have evolved with only highly conservative substitutions among the different genera of retroviruses (Rao et al. 1991). Thus, major features of structure, flap movement, and substrate recognition must be accommodated by different constellations of amino acid sequence. The structures of the HIV-1 and ASLV PRs are very similar, with most of the differences concentrated in variable lengths of surface loops (Rao et al. 1991) (see Fig. 13). As would be expected, the most conserved residues are those surrounding the catalytic aspartates and those lining the substrate-binding cleft.

Features of Protease Specificity

Information about PR specificity is derived from two sources. One approach to understanding specificity is to compare known cleavage sites in Gag, Pro, and Pol to infer the importance of specific amino acids at certain positions in defining a cleavage site. The second approach is to examine (and change) the interaction between PR and a substrate to establish the requirements of each entity in the cleavage reaction. Both of these approaches have yielded important, but as yet incomplete, information about the basis of PR specificity. Recent reviews of retroviral PR specificity are available (Oroszlan and Luftig 1990; Pettit et al. 1993; Dunn et al. 1994; Katz and Skalka 1994).

Consensus Sequence

An early comparison of PR cleavage sites revealed that they tend to be hydrophobic (Oroszlan and Copeland 1985). A simple model that can account in part for PR cleavage site selection is that a hydrophobic sequence is placed in an accessible (i.e., exposed) part of the precursor. This would distinguish the cleavage site from the other surface residues, which are typically hydrophilic, and from the other hydrophobic residues, which are typically buried in the interior of the protein and inaccessible. Since Gag and Gag-Pro-Pol can assemble in the absence of processing, it is possible that the individual protein domains fold into a functional form with the processing sites exposed as linkers between these domains. In addition, cleavage sites must assume an extended β-sheet conformation when bound to the PR, suggesting that the processing sites may be disordered stretches of protein between folded functional domains in the precursor. Different cleavage sites are cleaved at different rates. Thus, the sequence or context at some cleavage sites may be suboptimal as part of a mechanism to effect an ordered cleavage of the sites.

Alignments of available PR processing sites have been used to infer specific patterns of amino acids and to identify residues that define a cleavage site; these alignments have been used to develop algorithms to predict the presence of cleavage sites (Henderson et al. 1988b; Pettit et al. 1991; Poorman et al. 1991; Chou 1993). The cleavage sites of HIV-1, ASLV, and SIV have been studied in the greatest detail.

The sequences of cleavage sites in the retroviral polyprotein precursors are diverse but some generalizations can be made. General features found in most precursor processing sites are a small amino acid at P4; glutamine or basic amino acids can appear along with aliphatic amino acids in the P3 position; a small amino acid is common at P2; hydrophobic amino acids are typical at P1 and P1′(flanking the scissile bond), although β-branched amino acids are never found at P1; in the primate lentiviral group, glutamine or glutamic acid can appear in the P2′position, but this position is otherwise aliphatic; and aromatic amino acids are excluded from P3′. These tabulations are limited by the fact that the cleavage sites recognized by any one retroviral PR are quite diverse in sequence and by the fact that different retroviral PRs may have subtle differences in specificity (for a recent example, see Konvalinka et al. 1995b). Cleavage sites in heterologous proteins have been used to expand the database of sites for the purposes of sequence comparison (Poorman et al. 1991). Several attempts have been made to group the diverse collection of sequences that can be cleaved by the PR into subgroups (Henderson et al. 1988b; Pettit et al. 1991). In one classification scheme, two families of sequences are identified that in one case have an aromatic/proline sequence at the site of cleavage and that in the other case have aliphatic residues at the scissile bond. Cleavage sites representative of these two sequences appear to be conserved at the amino terminus and the carboxyl terminus cleavage sites, respectively, of CA within the Gag precursor of most retroviruses (Pettit et al. 1991). It is not clear whether the functional differences of these sequence classes are manifest at the protein level in the Gag precursor or in the processed products, or serve to regulate the ordered cleavage of Gag in these different retroviruses.

Cleavage of Peptides

Peptide substrates with different amino acids substituted at the various positions have been used to examine the relative importance of the amino acid at each position. Altered peptide substrates have been used to probe the specificity of the ASLV PR (Kotler et al. 1988, 1989; Konvalinka et al. 1991; Strop et al. 1991; Grinde et al. 1992b) and the HIV-1 PR (Billich et al. 1988; Margolin et al. 1990; Phylip et al. 1990; Richards et al. 1990; Billich and Winkler 1991; Tozser et al. 1991, 1992; Griffiths et al. 1992). Although there are many important nuances to the requirements of individual cleavage site sequences, these studies in general have confirmed the hydrophobic nature of cleavage sites. A significant question that is not yet fully resolved is the extent to which the amino acid side chains of the substrate interact with each other (either directly or indirectly through their binding to PR). Some evidence indicates that the importance of a specific amino acid at one position can change depending on the flanking sequence (Henderson et al. 1988b; Partin et al. 1990; Pettit et al. 1991; Griffiths et al. 1992). In addition, comparison of singly and doubly substituted peptide substrates has revealed both cis and trans interactions between alternating and adjacent substrate positions (Ridky et al. 1996b). However, the extent to which these functional interactions define separate classes of processing sites is not yet clear. The potential for interaction appears to be greatest with alternating amino acids since, in the extended β-sheet conformation of the bound substrate, the alternating amino acids extend into adjacent subsite pockets in PR.


The other approach to the study of PR specificity is the use of mutant PRs to probe the effect of the mutation on specificity. The ASLV PR is noted for its poor catalytic efficiency (Grinde et al. 1992a). This may be to compensate for its abundance in the virion, since it is produced as part of the Gag precursor, rather than the less abundant Gag-Pro-Pol precursor. It has been possible to improve the catalytic efficiency of the ASLV PR for some substrates by changing amino acids in the enzyme subsites to be HIV-like in sequence (Grinde et al. 1992b; Konvalinka et al. 1992; Cameron et al. 1994). These studies have also demonstrated a direct interaction between substrate side chains and PR subsite residues. Further information about substrate specificity has also been obtained by direct comparison of HIV-1 PR with either HIV-2 PR or HIV-1 PR mutants (Loeb et al. 1989b; Tomasselli et al. 1990b; Tozser et al. 1992; Sardana et al. 1994; Moody et al. 1995) or by comparisons of the HIV-1 and the ASLV PRs (Tomasselli et al. 1990a; Konvalinka et al. 1992; Cameron et al. 1993) on a variety of substrates. In the extreme case, eight mutations have been introduced into the RSV PR to give it a more HIV-like specificity on peptide substrates (Ridky et al. 1996a). Taken together, these studies establish the importance of side chain interactions between PR and substrate in determining substrate selection.

Timing of Protease Action

Problem of Precursor Activation

Processing of the Gag and Gag-Pro-Pol precursors is intimately linked to assembly and budding as can be seen with several temperature-sensitive MLV mutants which assemble capsids at the nonpermissive temperature without budding or processing (Witte and Baltimore 1978). Premature activation and processing by PR would preclude proper assembly and would be fatal to the virus. Consistent with this expectation are the observations that expression of an ASLV or HIV-1 genome which encodes a duplication of the PR-coding domain (i.e., a linked dimer) expected to be active in the monomer form of the precursor results in the initiation of processing prematurely in the cytoplasm of the cell and results in a lower yield of released virus (Burstein et al. 1991; Kräusslich 1991). Premature activation of PR has been observed during acute HIV-1 infection, and this appears to result in a pool of processed products that are excluded from the assembly pathway (Kaplan and Swanstrom 1991). In addition, when the Gag-Pro-Pol precursor is overexpressed in cells, premature processing is induced (Weaver et al. 1990; Park and Morrow 1991; Karacostas et al. 1993). Under normal circumstances, however, retroviruses are able to delay the appearance of PR activity. How this activity is controlled is not known. At present, there is no evidence that a cellular PR is involved in activating the processing pathway; viruses with inactivating mutations in PR produce uncleaved precursors (Crawford and Goff 1985; Katoh et al. 1985; Kohl et al. 1988; Göttlinger et al. 1989; Peng et al. 1989; Stewart et al. 1990; Sommerfelt et al. 1992). Similarly, the observation of premature activation of a linked dimer PR argues against the existence of a cellular inhibitor that would delay PR activity (Burstein et al. 1991).

Possible Mechanisms of Regulation

One attractive model for regulating PR activity is based on the need for dimerization. Concentration of the Gag-Pro-Pol precursor during assembly may provide conditions that promote dimerization and thus activity. Another possible source of regulation may reside in the fact that the initial PR dimer is still embedded in the Gag-Pro-Pol precursor, an arrangement that might limit its activity. In this model, the fully active form of the enzyme appears only as mature PR is slowly released from the precursor. Although PR does have activity in the context of the precursor, if the cleavage at the amino terminus of PR is blocked, processing during viral assembly is severely inhibited (Oertle and Spahr 1990; Burstein et al. 1992; Zybarth et al. 1994). Similarly, a reduction in the rate of release of PR reduces the extent of Gag processing, again suggesting that most of the processing occurs using mature PR (Xiang et al. 1997).

A simple model for regulation of PR activity based on protein concentration does not account for the delay in PR activity seen with type-B and type-D retroviruses. In these cases, viral capsids comprising precursors of the Gag and Gag-Pro-Pol proteins form in the cytoplasm of the infected cell (Rhee and Hunter 1987). These capsids do not initiate processing until budding has begun or concluded. In this case, the concentration of viral PR should be fixed once the capsids are formed, yet processing is not initiated until a later point in the assembly process. It is not known what signal initiates processing, but whatever mechanism delays activation of PR in these cases could also delay activation of PR activity during the assembly of type-C-like retroviruses.

Other factors that may contribute to regulating the timing of PR activation have been proposed. PR has a pH optimum significantly below pH 7, although the enzyme is active at physiological pH. This has led to the suggestion that a drop in pH in the virion may activate PR (Skalka 1989). In addition, removal of the region upstream of the HIV-1 PR increases processing of a Gag-PR fusion in vitro, as does inclusion of the NC domain, suggesting a role for these regions in regulating PR activity in cis (Partin et al. 1991; Zybarth and Carter 1995).

The mechanism by which PR cleaves itself out of the Gag-Pro-Pol (or in some cases the Gag-Pro) precursor has not been resolved. Examination of the PR structure (Fig. 13) shows that the amino and carboxyl termini, which represent the sites of cleavage to release PR from the Gag-Pro-Pol precursor, are involved in a four-stranded β-sheet that represents a significant fraction of the interactions between the two subunits. If the maintenance of this structure is important for PR activity, then it is unlikely that a folded dimeric PR in the Gag-Pro-Pol precursor would unfold this structure to cleave its own amino and carboxyl termini in cis. If this structure cannot be unfolded once it is formed, then dimeric PR in Gag-Pro-Pol precursors would not be cleaved at all. This suggestion leads to a model where a fraction of the PR becomes active in the Gag-Pro-Pol precursor, cleaving the PR ends in monomeric Gag-Pro-Pol precursors in trans to generate mature PR subunits that can dimerize and efficiently complete processing.

An alternative model is that PR in the Gag-Pro-Pol precursor is able to cleave its amino and/or carboxyl termini in cis to release preformed PR dimers. The observation of first-order kinetics in the generation of the amino terminus of PR from either a fusion protein (Louis et al. 1994) or a truncated miniprecursor (Co et al. 1994) suggests that this cleavage can occur by an intramolecular mechanism, and molecular models have been developed to account for an active PR with a partially disrupted structure. Finally, although there is little evidence for the role of a cellular protease, the possibility of a rare “initiating” cleavage by a cellular protease cannot be completely excluded. However, such an activity would have to be widely distributed in nature given the fact that retroviral PRs can be activated in a variety of prokaryotic and eukaryotic environments. Arguments for an initial cleavage of Gag-Pro-Pol by a distinct protease activity have been made for HIV-1 (Lindhofer et al. 1995).

Other Roles for the Protease

The presence of processed viral proteins in infected cells demonstrates the presence of intracellular viral PR activity and raises the possibility that PR may also cleave cellular proteins. Cleavage of the NF-κB precursor and cytoskeleton proteins has been detected in infected cells. This has led to speculation that cleavage of such proteins may have a role in viral replication and cytopathogenicity (Shoeman et al. 1990; Riviere et al. 1991; see also Shoeman et al. 1993 and references therein).

Upon purification of EIAV cores, the NC protein undergoes a slow cleavage event mediated by PR (Roberts and Oroszlan 1989). This observation led to the suggestion that PR may play a part early in infection, perhaps to cleave the NC after viral entry into the cell (Roberts et al. 1991). PR inhibitors have been used to block infection either by treating the host cell or by pretreating viral particles, and the block in replication was inferred to be after the bulk of viral DNA synthesis had occurred (Baboonian et al. 1991; Venaud et al. 1992; Nagy et al. 1994). However, other workers have failed to document an effect by a PR inhibitor early after infection (Jacobsen et al. 1992; Kaplan et al. 1996). A cleavage site in the HIV-1 NC protein has been detected in vitro within the first Cys-His box, but this cleavage occurs only after removal of zinc (Wondrak et al. 1994). The HIV-1 PR also cleaves the viral Nef protein (Freund et al. 1994), although the significance of this cleavage is unknown.

Protease Inhibitors

One of the early confirmations that the viral PR was in fact a member of the aspartic proteinase family of enzymes was the demonstration that the HIV-1 PR is somewhat sensitive to some inhibitors of other aspartic proteinases (Katoh et al. 1987; Seelmeir et al. 1988; Richards et al. 1989). The ability to inhibit PR, along with the central role that PR has in the viral life cycle, makes it an attractive target for the development of small molecule inhibitors. The development of PR inhibitors represents a major ongoing effort in drug design and discovery (Chapter 12.

The tightest binding inhibitors of any enzyme are transition state analogs. The transition state of the aspartic proteinase-catalyzed reaction occurs with the addition of a water molecule, coordinated by the active site aspartates, to the peptide bond. Presumed mimics of the transition state are represented by a hydroxyl group extending from the inhibitor at the site of the target peptide bond down toward the aspartates. This hydroxyl group displaces the enzyme-bound water molecule and interacts directly with the active site aspartates. The addition of flanking hydrophobic side chains to mimic the P1/P1′and P2/P2′side chains of a peptide substrate has formed the basic structure of a number of HIV-1 protease inhibitors with inhibitory constants (K i) in the nanomolar and even subnanomolar range. These substrate-based inhibitors have many chemical forms, but they assume similar conformations in the substrate-binding cleft of the PR (Fig. 14). Among the initial set of inhibitors that were discovered, a subset was able to penetrate cells and viral particles to inhibit the PR in the context of viral replication (Ashorn et al. 1990; Erickson et al. 1990; McQuade et al. 1990; Meek et al. 1990; Roberts et al. 1990). Further development of HIV-1 PR inhibitors has produced compounds with high potency, acceptable pharmacokinetic properties, and bioavailability (for some examples, see Roberts et al. 1990; Vacca et al. 1994; Kempf et al. 1995; Patick et al. 1996). Several PR inhibitors have been approved for use in the treatment of HIV-1 infection (see Chapter 12.

Copyright © 1997, Cold Spring Harbor Laboratory Press.
Bookshelf ID: NBK19438


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...