INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The structural proteins and enzymes of retroviruses are synthesized as part of Gag or Gag-Pol polyproteins. Late in the viral life cycle these polyproteins, along with viral envelope proteins and genomic RNA, aggregate at the cell membrane to form and subsequently bud immature virus particles. These immature viral particles are characterized by electron-lucent centers when examined by electron microscopy. Proteolytic processing of the Gag and Gag-Pol polyproteins by the virus-encoded protease yields mature structural proteins and enzymes (for a review, see reference 19). For specific cleavage, the PR recognizes an eight-amino-acid sequence symmetrically placed around the cleavage junction (17). Proteins processed from the avian sarcoma/leukemia virus (ASLV) Gag polyprotein include MA, p10, CA, NC, and PR. In addition, a 22-amino-acid sequence between MA and p10 contains a PPPPYV sequence, termed the L assembly domain, required for a step late during the viral budding process (21, 24). There are also a small number of amino acids between CA and NC, and these are referred to as the spacer (SP) region. Deletion of the SP results in the loss of virus infectivity (6, 15). Similar deletions in human immunodeficiency virus type 1 (HIV-1) Gag produce noninfectious virus (13, 16).

When Rous sarcoma virus (RSV) Gag is expressed in various cell types, three CA-containing species are resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (6). The bands are named according to their relative migration position; the fastest-migrating band is termed CA1, the middle band is CA2, and the slowest band is CA3 (6). CA1 is the first CA-containing band that appears and results from cleavage at Met488 (see Fig. 1). It is replaced over time with the appearance of CA2 and CA3. The sequential appearance of the three CA-containing proteins results from PR cleavage at three sites at the boundary of CA and NC in Gag. In mature ASLV, more than 90% of the CA-containing protein is accounted for as CA2 and CA3. A mass spectroscopy analysis of these two forms of CA indicates that they share a common amino terminus (15). Their carboxyl termini were mapped to Met479 and Ala476. In this report, we show that preventing cleavage at any of the three carboxyl-terminal CA sites results in noninfectious virus. In contrast, a mutant that has CA2 and CA3 proteins with little CA1 was infectious. This suggests that infectious virus is dependent upon proper processing to form these two mature CA proteins.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. All reagents were as previously described (21). Oligodeoxynucleotides were purchased from Midland Certified Reagent Company (Midland, Tex.) and used directly for mutagenesis. The wild-type RSV gag gene is from pATV-8, an infectious molecular clone of the RSV Prague C strain. The plasmid pSV.Myr0 is a simian virus 40-based mammalian expression vector carrying a wild-type copy of the RSV gag allele, the product of which efficiently directs production of virus-like particles from COS-1 cells (22).

Oligodeoxynucleotide-directed mutagenesis. Site-directed mutagenesis of the RSV gag gene was carried out by overlap extension mutagenesis as described by Aiyar et al. (2). In most cases, the mutations create a new restriction enzyme site to facilitate identification of clones containing the desired mutation. The oligodeoxynucleotides used to introduce mutations are listed in Table 1.

Transfection of mammalian cells. COS-1 cells grown in Dulbecco's modified Eagle's medium supplemented with 3% fetal bovine serum and 7% calf serum (Hyclone Inc.) were transfected by the Lipofectamine Plus method (7) (Gibco-BRL). Plasmid DNAs, at a concentration of 25 µg/ml, were digested with XbaI and incubated with T4 DNA ligase before transfection. This removes the bacterial plasmid sequence and joins the 3' end of the gag gene with the simian virus 40 late polyadenylation signal (22).

Metabolic labeling and immunoprecipitations. In most experiments, cells in 35-mm dishes were labeled 48 h after transfection with L-[³⁵S]methionine (1,000 Ci/mmol, 75 µCi/ml of tissue culture medium) for 2.5 h at 37°C as described previously (22). The cells or growth medium from each labeled culture were mixed with lysis buffer containing protease inhibitors. Rabbit antiserum directed against whole RSV (reactive with MA, CA, NC, and PR Gag proteins) was added and immunoprecipitation carried out for 2.5 h at 4°C. Precipitates was collected using protein A-agarose (Gibco-BRL) and the proteins were subjected to electrophoresis (24). Pulse-labeling experiments were carried out with sets of identically transfected cells incubated in methionine-free medium for 30 min, labeled with [³⁵S]methionine for 15 min, and chased with a 1,000-fold excess cold methionine in serum-free medium for the indicated times. Immunoprecipitated proteins were separated by electrophoresis through a SDS-polyacrylamide gel (13%) and detected by fluorography as previously described (22). Overnight exposures were typically required.

Coupled transcription and translation of CA proteins. Vectors for expression of each of the three CA proteins were constructed by PCR amplification of the CA coding sequence using NTCAXhol (17) and one of each of the 6HXAHTCAP oligodeoxyribonucleotides: NTCAXhol, 5'TACTCGAGATGCCTGTAGTGAAATTAAGACAGAG3'; 6HXAHTCAP_a, 5'CTCTCTCTGTTAAGCTTCTACATAATTAAGGGCTGGAT3'; 6HXAHTCAP_b, 5'TAAGGACTGGTTAAGCTTCTACATGGCCGCGGCTATG3'; and 6HXAHTCAP_c, 5'CTGGATGACAGACACAAGCTTCTAGGCTATGC3' (italics indicate the HindIII sites, boldface indicates stop codons, and underlines indicate the start codon).

Virus infectivity. An evaluation of virus infectivity was made by testing the ability of wild-type and mutant viruses to establish a persistent infection after transfection of DNA bearing the mutant viral genomes into susceptible cells. The proviral DNAs were introduced into duplicate 65-mm dishes of quail (QT6) cells by calcium phosphate-mediated transfection. To confirm successful transfection, the first plate of each pair was radiolabeled with [³⁵S]methionine at 18 h posttransfection. Then the CA protein, Gag, and CA-related cleavage intermediates were immunoprecipitated from lysates and medium samples with anti-CA serum and analyzed as described previously (6, 22). The second plate of cells was serially passaged at approximately 3-day intervals. The persistence of the mutant viruses was tested at passages 3 through 6 either by immunoprecipitation of radiolabeled antigens from lysates and medium samples with anti-CA or by Western blotting of CA antigens pelleted from the culture medium.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Altering polyprotein cleavage at the CA C terminus. Previously, the analysis of deletion mutants of the SP region suggested that CA1, CA2, and CA3 resulted from cleavage at Met488, Ala476, and Met479, respectively (Fig. 1). We have confirmed this assignment by analyzing the migration patterns of in vitro-translated recombinant CA proteins that terminated at these three positions (data not shown). We therefore use CA1, CA2, and CA3 (as originally defined by Craven et al. [6]) to indicate capsid proteins terminating at Met488, Ala476, and Met479, respectively. We refer to the CA cleavage sites to form CA1, CA2, and CA3 as sites 1, 2, and 3, respectively (Fig. 1).

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Substitutions in the RSV Gag polyprotein. The RSV Gag polyproteins are represented by rectangular boxes. The vertical lines inside the boxes represent cleavage sites between the different proteins. A series of substitutions that change Gag are listed below the boxes and were expressed from mutant plasmids constructed by overlap mutagenesis (2) using the oligodeoxynucleotides listed in Table 1. Below the rectangle is the sequence of Gag, amino acids 469 to 492, containing the CA-SP-NC region presented in one-letter notation. Three PR cleavage sites are indicated at Ala476, Met479, and Met488 and are referred to as sites 2, 3, and 1, respectively. The full-length black line represents the wild type (WT). Amino acid substitutions are aligned with the expanded sequence. For infectivity, a plus sign indicates that the mutant is as infectious as the wild type while a minus sign indicates that the mutant is noninfectious.

View larger version (25K): [in this window] [in a new window]   FIG. 2.   (A) Effect of Gag substitutions on CA processing. All labeling and protein detection was done as described in Materials and Methods. The positions of migration of Gag and its CA cleavage products are indicated on the left. (B) Effect of Gag substitutions on site 2 or 3 processing.

The stable accumulation of CA1 is not required for infectious virus. The effects of altered CA processing on viral infectivity were examined by subcloning each of the above mutations into the RCAN virus vector and determining if infectious virus could be recovered after transfection of the DNA into quail cells. These results are summarized in Fig. 1. While persistent infection was detected for wild-type virus after three or six passages, no viral protein was detected either in cells or in the medium after the passage of mutants that were altered in the processing of site 1, indicating that these mutant viruses were noninfectious (Fig. 1). All mutations that were designed to decrease the cleavage rates of site 2, site 3, or both (Fig. 1) caused a similar block to infection. In contrast, persistent infection resulted after three passages of cells expressing a clone with the A477L substitution (data not shown). Since little CA1 is detected in the A477G,M479G mutant, these results suggest that the length of time that CA1 persists is not important for infectivity, although its transient presence maybe needed for formation of CA2 and CA3.

Efficient cleavage to form CA2 and CA3 may require prior processing at downstream site 1. While infectious virus is obtained with an A477L mutant where little CA1 is evident, completely preventing the formation of CA1 in the L486G, M488G mutant results in noninfectious virus. This raises the possibility that cleavage at site 1 may facilitate the cleavage at upstream sites 2 and 3 similar to that of HIV-1 CA processing (20). In examining the results with the L486G,M488G mutant, we noticed that while the CA1 band was completely absent, the apparent amounts of CA2 and CA3 bands detected were reduced compared to those obtained with the wild type (Fig. 2A, compare media, lanes 1 and 2). Also, in the medium fraction, there were additional Gag intermediates migrating between Gag and the mature CA1, CA2, and CA3 bands not present in the wild type. Therefore we carried out a pulse-chase-type experiment (Fig. 3) to assess the relative cleavage rates at CA junctions. In the lysate fraction, the wild-type gag allele released CA1 during a 15-min labeling. With the A477G,M479G mutant, a band appeared as rapidly as CA1 in the wild type and in normal amounts compared to the wild type, even though the CA2 and CA3 bands were not present (Fig. 3). The L486G, M488G mutant did not release any CA proteins until after a 60-min incubation in the presence of excess unlabeled methionine. In addition, there were bands migrating between Gag and CA in the lysate during the chase. These results clearly demonstrate that processing at the upstream site 2 and 3 occurred much later than processing at site 1. They also suggest that efficient cleavage to form CA2 and CA3 may require prior cleavage to form CA1.

View larger version (34K): [in this window] [in a new window]   FIG. 3.   Kinetics of appearance of Gag and its cleavage products. Wild-type and mutant Gag alleles were separately expressed in COS cells and labeled with [35S]Met for 15 min. A 1,000-fold molar excess of unlabeled methionine was added to the tissue culture medium, and cells were further incubated for the indicated times. Cell lysates and media were immunoprecipitated and analyzed on an SDS-12% polyacrylamide gel as described in Materials and Methods.

L domain and CA processing. The RSV L assembly domain consists of a small proline-rich sequence that maps to the p2 region of Gag and is required for efficient budding of virus from cells (21, 24). Previously, when studying mutations in the L domain, we noticed that there was a correlation between the budding of viral particles from cells and the extent of processing at the CA-SP-NC cleavage sites (24). For instance, cleavage was observed almost exclusively at the CA1 site in cell lysates where the L domain was deleted from Gag (Fig. 4, Lysates, lane 4). To explore the possibility that the processing at the upstream sites 2 and 3 is dependent upon particle release, the A477L mutant was combined with an L domain deletion mutant. Two CA proteins whose migration corresponds to CA1 and CA3 were detected in the cell lysate expressed from the combined L domain A477L mutant (Fig. 4, Lysates, lane 2). This is in contrast to observing only CA1 in cell lysates from the L domain mutant (21, 24) and both CA2 and CA3 with the A477L mutant. This result shows that Gag particle release is not absolutely required for processing CA1 to CA3. However, the release of CA2 may still require the budding process to occur.

View larger version (53K): [in this window] [in a new window]   FIG. 4.   CA C-terminal processing defects caused by L domain deletions can be partially reversed by increasing the rate of cleavage of sites 2 and 3. A modified Gag containing an 11-amino-acid deletion of the p2b region was expressed in COS cells, and Gag and its cleavage products were analyzed as described in the legend to Fig. 2.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The presence of a spacer peptide between CA and NC in Gag is a conserved feature common to avian retroviruses and lentiviruses. Despite the lack of significant sequence identity between the respective HIV-1 and RSV CA and spacer sequences, proteolytic processing of this region is remarkably similar for both and occurs in a stepwise manner. Previously, it was shown that cleavage of the upstream HIV-1 CA site was reduced by the abolishment of the downstream CA site (20). In this report, we demonstrate that when cleavage of the RSV downstream CA site 1 was prevented, in the L486G,M488G mutant, the amounts of mature CA proteins resulting from cleavage at the upstream sites 2 and 3 were also reduced. Both the L486G,M488G and the A489L mutant, which has a single amino acid substitution in site 1, produce noninfectious virus. Additionally, these mutants show Gag processing intermediates that are absent in the wild type. We believe that these intermediates resulted from a delayed processing at CA junctions. However, it is also possible that amino acid substitutions at site 1 may cause a portion of Gag to improperly fold or somehow become refractory to further cleavage. It is conceivable that these intermediates might interfere with the correct assembly of virions. In addition, the L486G,M488G mutation would almost certainly alter the structure of the NC, and this change might contribute to the infectivity defect of this mutant.

The removal of the spacer peptide from immature CA was previously shown to be important for HIV-1 assembly (20). The RSV A477G,M479G mutant, which produces only immature CA, is noninfectious. However, since the A477G,M479G mutant introduces changes in the CA coding sequence as well as changes the cleavage sites, we cannot exclude the possibility that the replication defect is caused by an altered CA protein. The L domain deletion mutant also shows CA2 and CA3 processing defects that can be partially rescued by increasing the cleavage rate of these two sites with the A477L substitution. Nevertheless, these viruses are noninfectious. This is not surprising, since the A477L substitution would not correct the budding defect caused by a loss in L-domain function.

CA1 exists transiently for periods of time after the release of particles from cells (Fig. 4). Similarly, in HIV-1, a CA protein containing the spacer peptide, called CA-p2 or CA-SP1, exists transiently as a result of the low rate of cleavage of the CA-SP1 site. RSV CA1 and HIV-1 CA-SP1 may play a role in early events of particle assembly, since deletion of the spacer sequences result in the production of noninfectious viruses (6, 15). In the A477L mutant, little CA1 was detected, consistent with a shorter-than-normal half-life for this cleavage intermediate. It was therefore interesting to find that the A477L mutant was infectious. Bowzard et al. (2a) have confirmed this result with the finding that a substitution of Leu for Ser480 in a gag allele also produces infectious virus. This mutation is predicted to increase the rate of cleavage to form CA2 and CA3 similarly to the A477L mutant. Like the A477L substitution, this mutation caused the rapid appearance of the mature CA species, presumably due to enhanced cleavage at sites 2 and 3, and allowed full infectivity. Although we cannot eliminate the importance of CA1 for viral replication, these results suggest that the length of time CA1 persists is not essential for the viral life cycle.

In our SDS-PAGE analysis, we noticed that the immature CA1 protein migrates unusually rapidly, consistent with the observations of Pepinsky et al. (15). Gel electrophoresis analysis of the CA bands obtained from mutant proteins that bear single or double amino acid substitutions has located the region responsible for the aberrant migration to that around cleavage sites 2 and 3. The mutants with Gly substitutions at positions 475, 477, and 479 showed a loss of the three normal bands and the appearance of a novel band. Pepinsky et al. (15) also observed that a deletion of the amino acids Ala, Ala, and Met from positions 477 to 479 resulted in a change in migration of the detected CA-containing bands. We speculate that the Gag region between 475 and 481 may be associated with changes in the residual structure of CA in SDS. Alternatively, the substitutions themselves may be responsible for the change in migration of the bands. We believe that this is unlikely because of the very small size of the region where substitutions result in changes in migration.

The structures of the entire RSV CA protein (5) and the C-terminal domain of RSV CA with and without the 12-amino-acid spacer peptide attached (12) have been determined by nuclear magnetic resonance spectroscopy. In each case, the very end of the CA protein from the Pro at Gag position 469 through the spacer peptide was found to be disordered or to exist in multiple conformations. The addition of the spacer peptide to the C-terminal domain caused no obvious change in either the C terminus of CA or the remainder of the C-terminal domain. A similar situation has been described for the HIV-1 CA protein (12, 23). However, Accola et al. (1) have reported the prediction of an -helix in the CA-p2 boundary in HIV-1 Gag. Moreover, HeLa cells transfected with proviral DNA containing amino acid substitutions that should disrupt this predicted -helix produced heterogeneous particles. Conversely, amino acid substitutions at the same site predicted to maintain the -helix structure produced homogeneous particles with a cone-shaped core similar to mature HIV-1 particles (1). This suggests that the peptides, p2 for HIV and presumably SP for RSV, form a helical structure in the CA-NC cleavage intermediate. Deletions in the CA or SP regions of RSV Gag lead to the production of heterogeneously sized particles (14). The changes in migration of RSV Gag peptides with mutations in and around sites 2 and 3 described in this report would be consistent with the different CA species having different conformations, although this has not been found in existing structural studies (5, 12). In another study of the in vitro assembly of HIV-1 CA, the spacer peptide was also found to alter the conformation and packing of CA subunits in assembled particles (11). Thus, it is possible that the three-dimensional structures of purified CA proteins as now reported may not represent the full range of conformations that these proteins adopt in particles. It is possible, for instance, that the spacer peptide in CA1 becomes structured when it is assembled into higher-order arrays.

A model of retrovirus core assembly has suggested that the cores of all retroviruses could be composed of closed hexagonal lattices made from CA proteins (8). It is conceivable that the two forms of mature RSV CA fit into different structural environments in an assembled core. Therefore, the loss of one form or the dramatic alteration of the ratio of the mature forms would be expected to compromise the core structure and possibly viral replication. In the A477L mutant, the ratio of the two CA proteins was different from that of the wild type. This result does not invalidate this model, since one of the two forms of CA may be needed only in small amounts. A better understanding must await determination of high-resolution structures of the assembled cores.