Cleavage of p15 protein in vitro by human immunodeficiency virus type 1 protease is RNA dependent.

The human immunodeficiency virus (HIV) gag polyprotein is processed by the viral protease to yield the structural proteins of the virus. One of these structural proteins, p15, and its protease cleavage products, p7 and p6, are believed to be responsible for the viral RNA binding which is prerequisite for assembly of infectious virions. To better understand potential interactions between viral RNA, p15, and the HIV protease, we have synthesized p15 in an in vitro system and studied its processing by the viral protease. Using this system, we demonstrate that p15 synthesized in vitro is properly cleaved by the HIV protease in an RNA-dependent reaction. Mutation of cysteine residues in either zinc-binding domain of the p7 portion of p15 does not alter the RNA-dependent cleavage, but mutation of three basic residues located between the zinc-binding domains blocks HIV protease susceptibility. The results support a previously unrecognized role for the interaction of RNA and nucleocapsid-containing gag precursors that may have important consequences for virus assembly.     

Molecular characterization of gag proteins from simian immunodeficiency virus (SIVMne).

A simian immunodeficiency virus (SIV) designated SIVMne was isolated from a pig-tailed macaque with lymphoma housed at the University of Washington Regional Primate Research Center, Seattle. To better establish the relationship of SIVMne to other immunodeficiency viruses, we purified and determined the partial amino acid sequences of six structural proteins (p1, p2, p6, p8, p16, and p28) from SIVMne and compared these amino acid sequences to the translated nucleotide sequences of SIVMac and human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2). A total of 125 residues of SIVMne amino acid sequence were compared to the predicted amino acid sequences of the gag precursors of SIV and HIVs. In the compared regions 92% of the SIVMne amino acids were identical to predicted residues of SIVMac, 83% were identical to predicted residues of HIV-2, and 41% were identical to predicted residues of HIV-1. These data reveal that the six SIVMne proteins are proteolytic cleavage products of the gag precursor (Pr60gag) and that their order in the structure of Pr60gag is p16-p28-p2-p8-p1-p6. Rabbit antisera prepared against purified p28 and p16 were shown to cross-react with proteins of 60, 54, and 47 kilodaltons present in the viral preparation and believed to be SIVMne Pr60gag and intermediate cleavage products, respectively. SIVMne p16 was shown to contain covalently bound myristic acid, and p8 was identified as a nucleic acid-binding protein. The high degree of amino acid sequence homology between SIVs and HIV-2 around proven proteolytic cleavage sites in SIV Pr60gag suggests that proteolytic processing of the HIV-2 gag precursor is probably very similar to processing of the SIV gag precursor. Peptide bonds cleaved during proteolytic processing of the SIV gag precursor were similar to bonds cleaved during processing of HIV-1 gag precursors, suggesting that the SIV and HIV viral proteases have similar cleavage site specificities.     

Functional characterization of the cleavage specificity of the sapovirus chymotrypsin-like protease

Sapovirus is a positive-stranded RNA virus with a translational strategy based on processing of a polyprotein precursor by a chymotrypsin-like protease. So far, the molecular mechanisms regulating cleavage specificity of the viral protease are poorly understood. In this study, the catalytic activities and substrate specificities of the predicted forms of the viral protease, the 3C-like protease (NS6) and the 3CD-like protease-polymerase (NS6-7), were examined in vitro. The purified NS6 and NS6-7 were able to cleave synthetic peptides (15 to 17 residues) displaying the cleavage sites of the sapovirus polyprotein, both NS6 and NS6-7 proteins being active forms of the viral protease. High-performance liquid chromatography and subsequent mass spectrometry analysis of digested products showed a specific trans cleavage of peptides bearing Gln-Gly, Gln-Ala, Glu-Gly, Glu-Pro, or Glu-Lys at the scissile bond. In contrast, peptides bearing Glu-Ala or Gln-Asp at the scissile bond (NS4-NS5 and NS5-NS6, or NS6-NS7 junctions, respectively) were resistant to trans cleavage by NS6 or NS6-7 proteins, whereas cis cleavage of the Glu-Ala scissile bond of the NS5-NS6 junction was evidenced. Interestingly, the presence of a Phe at position P4 overruled the resistance to trans cleavage of the Glu-Ala junction (NS5-NS6), whereas substitutions at the P1 and P2′ positions altered the cleavage efficiency. The differential cleavage observed is supported by a model of the substrate-binding site of the sapovirus protease, indicating that the P4, P1, and P2′ positions in the substrate modulate the cleavage specificity and efficiency of the sapovirus chymotrypsin-like protease.     

Interdomain hydrolysis of a truncated Pseudomonas exotoxin by the human immunodeficiency virus-1 protease

The specificity of HIV-1 (human immunodeficiency virus-1) protease has been evaluated relative to its ability to cleave the three-domain Pseudomonas exotoxin (PE66) and related proteins in which the first domain has been deleted or replaced by a segment of CD4. Native PE66 is not hydrolyzed by the HIV-1 protease. However, removal of its first domain produces a molecule which is an excellent substrate for the enzyme. The major site of cleavage in this truncated exotoxin, called LysPE40, occurs in a segment that connects its two major domains, the translocation domain (II), and the ADP-ribosyltransferase (III). This interdomain region contains the sequence ...Asn-Tyr-Pro-Thr... which is similar to that surrounding the scissile Tyr-Pro bond in the gag precursor polyprotein, a natural substrate of the HIV-1 protease. Nevertheless, it is not this sequence that is recognized and cleaved by the enzyme, but one 6 residues away, ...Ala-Leu-Leu-Glu... in which the Leu-Leu peptide bond is hydrolyzed. A second, slower cleavage takes place at the Leu-Ala bond 3 residues in from the NH2 terminus of LysPE40. When domain I of PE66 is replaced by a segment comprising the first two domains of CD4, the resulting chimeric protein is hydrolyzed at the same Leu-Leu bond by HIV-1 protease. Enzyme activities toward synthetic peptides modeled after the sequences defined above in LysPE40 are in complete accord, relative to specificity, kinetics, and pH optimum, with results obtained in the hydrolysis of the parent protein. These findings demonstrate that ideas concerning the specificity of the HIV-1 protease that are based solely upon its processing of natural viral polyproteins can be expanded by evaluation of other multidomain proteins as substrates. Moreover, it would appear that it is not a particular conformation, but sequence and accessibility that play the dominant role in defining sites in a protein substrate that are susceptible to hydrolysis by the enzyme.     

Identification of HTLV-I gag protease and its sequential processing of the gag gene product

The full-length provirus of human T-cell leukemia virus type I (HTLV-I) was isolated from MT-2, a lymphoid cell line producing HTLV-I. In transfected cells, structural proteins of HTLV-I, the gag and env products, were formed and processed in the same manner as observed in MT-2 cells. The nucleotide sequence was determined for a region between the gag and pol genes of the proviral DNA clone containing an open-reading frame. The deduced amino acid sequences show that this open-reading frame encodes a putative HTLV-I protease. The protease gene (pro) of HTLV-I was investigated using a vaccinia virus expression vector. Processing of 53k gag precursor polyprotein into mature p19, p24, and p15 gag structural proteins was detectable with a recombinant plasmid harboring the entire gag- and protease-coding sequence. We demonstrated that the protease processed the gag precursor polyprotein in a trans-action. A change in the sequence Asp(64)-Thr-Gly, the catalytic core sequence among aspartyl proteases, to Gly-Thr-Gly was shown to abolish correct processing, suggesting that HTLV-I protease may belong to the aspartyl protease group. The 76k gag-pro precursor polyprotein was identified, implying that a cis-acting function of HTLV-I protease may be necessary to trigger the initial cleavage event for its own release from a precursor protein, followed by the release of p53 gag precursor protein. The p53 gag precursor protein is then processed by the trans-action of the released protease to form p19, p24, and p15.     

Chemical and immunological characterizations of equine infectious anemia virus gag-encoded proteins.

The viral core proteins (p15, p26, p11, and p9) of equine infectious anemia virus (EIAV) (Wyoming strain) were purified by reverse-phase high-pressure liquid chromatography. Each purified protein was analyzed for amino acid content, N-terminal amino acid sequence, C-terminal amino acid sequence, and phosphoamino acid content. The results of N- and C-terminal amino acid sequence analysis of each gag protein, taken together with the nucleotide sequence of the EIAV gag gene (R. M. Stephens, J. W. Casey, and N. R. Rice, Science 231:589-594, 1986), show that the order of the proteins in the precursor is p15-p26-*-p11-p9, where a pentapeptide also found in the virus is represented by the asterisk. The data are in complete agreement with the predicted structure of the gag polyprotein and show the peptide bonds cleaved during proteolytic processing. The N-terminus of p15 is blocked to Edman degradation. The p11 protein is identical to the nucleic acid-binding protein of EIAV previously isolated (C. W. Long, L. E. Henderson, and S. Oroszlan, Virology 104:491-496, 1980). High-titer rabbit antiserum was prepared against each purified protein. These antisera were used to detect the putative gag precursor (Pr55gag) and intermediate cleavage products designated Pr49 (p15-p26-*-p11), Pr40 (p15-p26), and Pr35 (p26-*-p11) in the virus and in virus-infected cells. High-titer antisera to EIAV p15 and p26 showed cross-reactivity with the homologous protein of human T-cell lymphotropic virus type III/lymphadenopathy-associated virus.     

Processing of gag precursor polyprotein of human T-cell leukemia virus type I by virus-encoded protease.

The biological activity encoded in the putative protease gene (pro) of human T-cell leukemia virus type I was investigated by using a vaccinia virus expression vector. The 53-kilodalton gag precursor polyprotein was processed into the mature p19, p24, and p15 gag proteins when the gag and protease-coding sequence was expressed under the control of a vaccinia virus promoter, suggesting that the protease may be synthesized through the mechanism of ribosomal frame shifting. The processing defect of a protease mutant could be complemented by cointroduction of a wild-type construct into the cell, demonstrating that the pro gene encodes the biologically active protease molecules which are capable of processing the gag precursor polyprotein in vivo in trans. A study involving the use of a variety of mutants constructed in vitro revealed that the protease consists of a nonessential carboxy-terminal region and a part essential for its activity, including the putative catalytic residue, aspartic acid. Furthermore, a cluster of adenine residues positioned at the overlapping region between the gag and pro genes was shown to be involved in the ribosomal frameshifting event for the synthesis of protease. To mimic the formation of the 76-kilodalton gag-pro precursor polyprotein formed by ribosomal slipping, the coding frames of the gag and pro gene were adjusted. The processing of the gag-pro precursor polyprotein depended on an intact protease gene, implying that a cis-acting function of human T-cell leukemia virus type I protease may be necessary to trigger the initial cleavage event that leads to the release of protease from the precursor protein.     

Assessment of substrate specificity of hepatitis G virus NS3 protease by a genetic method

The RNA genome of hepatitis G virus (HGV) encodes a large polyprotein that is processed to mature proteins by viral-encoded proteases. The HGV NS3 protease is responsible for the cleavage of the HGV polyprotein at four different locations. No conserved sequence motif has been identified for the cleavage sites of the NS3 protease. To determine the substrate specificity of the NS3 protease, amino acid sequences cleaved by the NS3 protease were obtained from randomized sequence libraries by using a screening method referred to as GASP (Genetic Assay for Site-specific Proteolysis). Based on statistical analyses of the obtained cleavable sequences, a consensus substrate sequence was deduced: Gln-Glu-Thr-Leu-Val downward arrow Ser, with the scissile bond located between Val and Ser. The relevance of this peptide as a cleavable substrate was further supported by molecular modeling of the NS3 protease. Our result would provide an insight on the molecular activity of the NS3 protease and may be useful for the design of substrate-based inhibitors.     

Human immunodeficiency virus protease. Bacterial expression and characterization of the purified aspartic protease

The protease of human immunodeficiency virus has been expressed in Escherichia coli and purified to apparent homogeneity. Immunoreactivity toward anti-protease peptide sera copurified with an activity that cleaved the structural polyprotein gag p55 and the peptide corresponding to the sequence gag 128-135. The enzyme expressed as a nonfusion protein exhibits proteolytic activity with a pH optimum of 5.5 and is inhibited by the aspartic protease inhibitor pepstatin with a Ki of 1.1 microM. Replacement of the conserved residue Asp-25 with an Asn residue eliminates proteolytic activity. Analysis of the minimal peptide substrate size indicates that 7 amino acids are required for efficient peptide cleavage. Size exclusion chromatography is consistent with a dimeric enzyme and circular dichroism spectra of the purified enzyme are consistent with a proposed structure of the protease (Pearl, L.H., and Taylor, W.R. (1987) Nature 329, 351-354). These data support the classification of the human immunodeficiency virus protease as an aspartic protease, likely to be structurally homologous with the well characterized family that includes pepsin and renin.     

Chemical synthesis and expression of the HIV-1 protease gene in E. coli

The 297bp HIV-1 protease gene was constructed from five discrete synthetic fragments and expressed in E. coli. A soluble protein product of 11.5 Kd was detected by immunoblotting using protease specific antisera. A quantitative assay system, utilizing a synthetic nonapeptide spanning the cleavage site between p17-p24 in the gag polyprotein, was used to measure the specific protease activity in crude extracts. The protease hydrolyzed tyrosyl-proline bonds with an approximate specific activity of 43 pmoles/min/micrograms of total protein. The chemical synthesis of the protease gene and it's expression provides a feasible method for rapid mutant analysis, important for structure-function studies and rational design of potential inhibitors.     

Mutational analysis of human immunodeficiency virus type 1 protease suggests functional homology with aspartic proteinases.

Processing of the retroviral gag and pol gene products is mediated by a viral protease. Bacterial expression systems have been developed which permit genetic analysis of the human immunodeficiency virus type 1 protease as measured by cleavage of the pol protein precursor. Deletion analysis of the pol reading frame locates the sequences required to encode a protein with appropriate proteolytic activity near the left end of the pol reading frame but largely outside the gag-pol overlap region, which is at the extreme left end of pol. Most missense mutations within an 11-amino-acid domain highly conserved among retroviral proteases and with sequence similarity to the active site of aspartic proteinases abolish appropriate processing, suggesting that the retrovirus proteases share a catalytic mechanism with aspartic proteinases. Substitution of the amino acids flanking the scissile bond at three of the processing sites encoded by pol demonstrates distinct sequence requirements for cleavage at these different sites. The inclusion of a charged amino acid at the processing site blocks cleavage. A subset of these substitutions also inhibits processing at the nonmutated sites.     

Mutational analysis of tobacco etch virus polyprotein processing: cis and trans proteolytic activities of polyproteins containing the 49-kilodalton proteinase.

The genome of tobacco etch virus contains a single open reading frame with the potential to encode a 346-kilodalton (kDa) polyprotein. The large polyprotein is cleaved at several positions by a tobacco etch virus genome-encoded, 49-kDa proteinase. The locations of the 49-kDa proteinase-mediated cleavage sites flanking the 71-kDa cytoplasmic pinwheel inclusion protein, 6-kDa protein, 49-kDa proteinase, and 58-kDa putative polymerase have been determined by using cell-free expression, proteolytic processing, and site-directed mutagenesis systems. Each of these sites is characterized by the conserved sequence motif Glu-Xaa-Xaa-Tyr-Xaa-Gln-Ser or Gly (in which cleavage occurs after the Gln residue). The amino acid residue (Gln) predicted to occupy the -1 position relative to the scissile bond has been substituted, by mutagenesis of cloned cDNA, at each of four cleavage sites. The altered sites were not cleaved by the 49-kDa proteinase. A series of synthetic polyproteins that contained the 49-kDa proteinase linked to adjoining proteins via defective cleavage sites were expressed, and their proteolytic activities were analyzed. As part of a polyprotein, the proteinase was found to exhibit cis (intramolecular) and trans (intermolecular) activity.     

Actin, troponin C, Alzheimer amyloid precursor protein and pro-interleukin 1 beta as substrates of the protease from human immunodeficiency virus

We show here for the first time that actin, troponin C, Alzheimer amyloid precursor protein (AAP), and pro-interleukin 1 beta (pro-IL-1 beta), are substrates of the protease encoded by the human immunodeficiency virus (HIV) type-1. As has been seen in other non-viral protein substrates of the HIV protease, the presence of Glu residues in the P2' position appears to play an important role in substrate recognition. Three of the four bonds cleaved in actin, two of the three in troponin C, and all of the bonds hydrolyzed in AAP and pro-IL-1 beta have a P2' Glu residue. In fact, Glu residues are accommodated in all positions from P4 to P4' surrounding the scissile bond in substrates of the HIV proteases, and as many as 4 adjacent Glu residues were seen in one of the bonds cleaved in AAP. This study of non-viral protein substrates has also revealed unexpected amino acids such as Gly, Arg, and Glu in the scissile bond itself rather than the more conventional hydrophobic amino acids. The HIV-2 protease hydrolyzed actin in a manner similar to that of the HIV-1 enzyme, but its cleavage of troponin C was distinct in that it split a bond adjacent to a triplet of Glu residues in P2, P3, and P4 that was refractory to the HIV-1 enzyme. Documentation of cleavage sites in the several important cellular proteins noted above has extended our understanding of the features in a substrate that are recognized by these multi sub-site proteases of retroviral maturation. Moreover, the present work adds to an accumulating body of evidence which demonstrates that these enzymes can damage crucial structural and regulatory cellular proteins if ever their activity is expressed outside the viral particle itself.     

A radiometric assay for HIV-1 protease

A rapid, high-throughput radiometric assay for HIV-1 protease has been developed using ion-exchange chromatography performed in 96-well filtration plates. The assay monitors the activity of the HIV-1 protease on the radiolabeled form of a heptapeptide substrate, [tyrosyl-3,5-3H]Ac-Ser-Gln-Asn-Tyr-Pro-Val-Val-NH2, which is based on the p17-p24 cleavage site found in the viral polyprotein substrate Pr55gag. Specific cleavage of this uncharged heptapeptide substrate by HIV-1 protease releases the anionic product [tyrosyl-3,5-3H]Ac-Ser-Gln-Asn-Tyr, which is retained upon minicolumns of the anion-exchange resin AG1-X8. Protease activity is determined from the recovery of this radiolabeled product following elution with formic acid. This facile and highly sensitive assay may be utilized for steady-state kinetic analysis of the protease, for measurements of enzyme activity during its purification, and as a routine assay for the evaluation of protease inhibitors from natural product or synthetic sources.     

Mechanism of inhibition of the retroviral protease by a Rous sarcoma virus peptide substrate representing the cleavage site between the gag p2 and p10 proteins.

The activity of the avian myeloblastosis virus (AMV) or the human immunodeficiency virus type 1 (HIV-1) protease on peptide substrates which represent cleavage sites found in the gag and gag-pol polyproteins of Rous sarcoma virus (RSV) and HIV-1 has been analyzed. Each protease efficiently processed cleavage site substrates found in their cognate polyprotein precursors. Additionally, in some instances heterologous activity was detected. The catalytic efficiency of the RSV protease on cognate substrates varied by as much as 30-fold. The least efficiently processed substrate, p2-p10, represents the cleavage site between the RSV p2 and p10 proteins. This peptide was inhibitory to the AMV as well as the HIV-1 and HIV-2 protease cleavage of other substrate peptides with Ki values in the 5-20 microM range. Molecular modeling of the RSV protease with the p2-p10 peptide docked in the substrate binding pocket and analysis of a series of single-amino acid-substituted p2-p10 peptide analogues suggested that this peptide is inhibitory because of the potential of a serine residue in the P1' position to interact with one of the catalytic aspartic acid residues. To open the binding pocket and allow rotational freedom for the serine in P1', there is a further requirement for either a glycine or a polar residue in P2' and/or a large amino acid residue in P3'. The amino acid residues in P1-P4 provide interactions for tight binding of the peptide in the substrate binding pocket.     

Quantitative separation of murine leukemia virus proteins by reversed-phase high-pressure liquid chromatography reveals newly described gag and env cleavage products.

The structural proteins of murine type C retroviruses are proteolytic cleavage products of two different precursor polyproteins coded by the viral gag and env genes. To further investigate the nature and number of proteolytic cleavages involved in virus maturation, we quantitatively isolated the structural proteins of the Rauscher and Moloney strains of type C murine leukemia virus (R-MuLV and M-MuLV, respectively) by reversed-phase high-pressure liquid chromatography. Proteins and polypeptides isolated from R-MuLV included p10, p12, p15, p30, p15(E), gp69, and gp71 and three previously undescribed virus components designated here as p10', p2(E), and p2(E). Homologous proteins and polypeptides were isolated from M-MuLV. Complete or partial amino acid sequences of all the proteins listed above were either determined in this study or were available in previous reports from this laboratory. These data were compared with those from the translation of the M-MuLV proviral DNA sequence (Shinnick et al., Nature [London] 293:543-548, 1981) to determine the exact nature of proteolytic cleavages for all the structural proteins described above and to determine the origin of p10' and p2(E)s. The results showed that, during proteolytic processing of gp80env from M-MuLV (M-gp 80env), a single Arg residue was excised between gp70 and p15(E) and a single peptide bond was cleaved between p15(E) and p2(E). The structure of M-gPr80env is gp70-(Arg)-p15(E)-p2(E). The data suggest that proteolytic cleavage sites in R-gp85env are identical to corresponding cleavage sites in M-gp80env. The p2(E)s are shown to be different genetic variants of p2(E) present in the uncloned-virus preparations. The data for R- and M-p10's shows that they are cleavage products of the gag precursor with the structure p10-Thr-Leu-Asp-Asp-OH. The complete structure of Pr65gag is p15-p12-p30-p10'. Stoichiometries of the gag and env cleavage products in mature R- and M-MuLV were determined. In each virus, gag cleavage products (p15, p12, p30, and p10 plus p10') were found in equimolar amounts and p15(E)s were equimolar with p2(E)s. The stoichiometry of gag to env cleavage products was 4:1. These data are consistent with the proposal that proteolytic processing of precursor polyproteins occurs after virus assembly and that the C-terminal portion of Pr15(E) [i.e., p15(E)-p2(E)] is located on the inner side of the lipid bilayer of the virus.     

Monoclonal antibody to the amino-terminal L sequence of murine leukemia virus glycosylated gag polyproteins demonstrates their unusual orientation in the cell membrane.

To analyze cell surface murine leukemia virus gag protein expression, we have prepared monoclonal antibodies against the spontaneous AKR T lymphoma KKT-2. One of these antibodies, 43-13, detects an AKR-specific viral p12 determinant. A second monoclonal antibody, 43-17, detects a novel murine leukemia virus-related antigen found on glycosylated gag polyproteins (gp95gag, gp85gag, and gp55gag) on the surface of cells infected with and producing ecotropic endogenous viruses, but does not detect antigens within these virions. The 43-17 antibody immunoprecipitates the precursor of the cell surface gag protein whether in its glycosylated or unglycosylated state, but does not detect the cytoplasmic precursor of the virion gag proteins (Pr65gag). Based on these findings, we have localized the 43-17 determinant to the unique amino-terminal part of the glycosylated gag polyprotein (the L domain). We have determined that gp95gag contains L-p15-p12-p30-p10 determinants, whereas gp85gag lacks the carboxyterminal p10 determinant, and gp55gag lacks both p30 and p10 carboxy terminal determinants. Analysis of cell surface gag expression with the 43-17 antibody leads us to propose that the L domain plays a crucial role in (i) the insertion and orientation of murine leukemia virus gag polyproteins in the cell membrane and (ii) the relative abundance of expression of AKR leukemia virus versus Moloney murine leukemia virus glycosylated gag polyproteins in infected cells.     

Identification of protein intermediates in the processing of the p55 HIV-1 gag precursor in cells infected with recombinant vaccinia virus

We have used a recombinant vaccinia virus (VV) which expresses high levels of human immunodeficiency virus-1 (HIV-1) gag proteins to analyze the processing pathway of the gag p55 precursor. HIV-1 gag proteins were isolated from [3H]leucine-labeled VV:gag-infected H9 T lymphocytes by immunoprecipitation with either anti-p24, anti-p17, or anti-p6 antibodies. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis revealed that processing of the p55 precursor involves three major intermediates (p41a, p41b, and p39). The p41a and p39 proteins contain the p17 and p24 protein segments, and the p41b is comprised of p24 and p15 segments. On two-dimensional gels, each intermediate as well as the mature p24 and p17 proteins migrated as distinct species. [3H]Myristic acid labeling of the HIV-1 gag proteins revealed that in addition to p55 and p17, the p41a and p39 intermediates, but not p41b, are myristylated, confirming that myristylation occurs at the NH2 terminus before cleavage of the p55 precursor protein. We conclude that the myristylated HIV-1 gag p55 precursor is initially cleaved at random either at the p17/p24 junction or at two sites between p24 and p15 proteins, resulting in three intermediates (p41a, p41b, and p39) which are subsequently cleaved to yield mature gag proteins.