JOURNAL OF VIROLOGY, Nov. 2008, p. 10803–10810 0022-538X/08/$08.00⫹0 doi:10.1128/JVI.00812-08 Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 82, No. 21
The C Terminus of Foamy Retrovirus Gag Contains Determinants for Encapsidation of Pol Protein into Virions䌤 Eun-Gyung Lee and Maxine L. Linial* Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109 Received 15 April 2008/Accepted 12 August 2008
Foamy viruses (FV) differ from orthoretroviruses in many aspects of their replication cycle. A major difference is in the mode of Pol expression, regulation, and encapsidation into virions. Orthoretroviruses synthesize Pol as a Gag-Pol fusion protein so that Pol is encapsidated into virus particles through Gag assembly domains. However, as FV express Pol independently of Gag from a spliced mRNA, packaging occurs through a distinct mechanism. FV genomic RNA contains cis-acting sequences that are required for Pol packaging, suggesting that Pol binds to RNA for its encapsidation. However, it is not known whether Gag is directly involved in Pol packaging. Previously our laboratory showed that sequences flanking the three glycine-arginine-rich (GR) boxes at the C terminus of FV Gag contain domains important for RNA packaging and Pol expression, cleavage, and packaging. We have now shown that both deletion and substitution mutations in the first GR box (GR1) prevented neither the assembly of particles with wild-type density nor packaging of RNA genomes but led to a defect in Pol packaging. Site-directed mutagenesis of GR1 indicated that the clustered positively charged amino acids in GR1 play important roles in Pol packaging. Our results suggest that GR1 contains a Pol interaction domain and that a Gag-Pol complex is formed and binds to RNA for incorporation into virions. Foamy viruses (FV) are complex retroviruses that comprise one of two subfamilies. The FV replication strategy differs in many ways from that of orthoretroviruses such as human immunodeficiency virus (HIV-1) and is similar in some respects to that of hepadnaviruses such as human hepatitis B viruses (HBV). One of the major differences between FV and orthoretroviruses is the mode of Pol expression, regulation, and encapsidation into virions. Orthoretroviruses synthesize Pol as a Gag-Pol fusion protein. About one Gag-Pol is translated for every 20 Gag proteins, and such translational regulation balances the required levels of structural proteins and enzymatic proteins. The Gag-Pol fusion is packaged into virions by coassembly with self-assembling Gag proteins (reviewed in reference 9). However, FV Pol is synthesized from a separate spliced mRNA independently of Gag. This raises fundamental questions about how Pol expression is regulated and how Pol is selectively incorporated into assembling virus particles without Gag determinants. Recently, we have shown that Pol expression is regulated at the level of transcription by utilization of a suboptimal splicing site (14). The details of FV Pol packaging are not yet known. HBV also expresses its reverse transcriptase (RT) (P protein) independently of the structural proteins. However, HBV P protein is responsible for packaging of genomic RNA and initiating assembly of capsids (18), whereas FV Pol is not required for particle assembly or encapsdation of genomic RNA (1). Thus, the mechanism of FV Pol incorporation is unique, differing from those of orthoretroviruses and hepadnaviruses. The encapsidation of FV genomic RNA is essential for FV * Corresponding author. Mailing address: Division of Basic Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., Seattle, WA 98109-1024. Phone: (206) 667-4442. Fax: (206) 6675939. E-mail:
[email protected]. 䌤 Published ahead of print on 20 August 2008.
Pol packaging (11). Two cis-acting sequences called Pol encapsidation sequences (PES) are required for Pol packaging (19). One is located in the 5⬘ long terminal repeat (LTR), and the other is at the 3⬘ end of the pol gene. These sequences appear to partially overlap the cis-acting sequences required for RNA packaging (19). It has been suggested that Pol binds to a complex region in RNA for encapsidation (19). Since two RNAs are packaged, each is expected to bind only one or a few Pol dimers. Fewer Pol proteins are predicted to be in FV than in orthoretroviral particles. Not surprisingly, bacterially expressed FV protease (PR)-RT has shown to be more active and processive than HIV-1 RT in vitro (3). It has been proposed that genomic RNA is packaged through Gag binding and that Pol is incorporated by directly binding to RNA, so that RNA functions as a bridging molecule between Gag and Pol (19). However, these data do not exclude the possibility that direct interactions between Gag and Pol are required for RNA binding. The FV genome is organized like that of orthoretroviruses and contains open reading frames for the canonical proteins Gag, Pol, and Env that are derived from the LTR promoter, as well as additional 3⬘ open reading frames for nonstructural proteins Tas and Bet originating from an internal promoter (Fig. 1). There are fundamental differences in the sequences and proteolytic cleavages of Gag and Pol proteins between FV and orthoretroviruses. Unlike orthoretroviral Gag, FV Gag is not cleaved into the matrix (MA), capsid (CA), and nucleocapsid (NC) proteins found in mature virions. Instead, FV PR cleaves only once at the C terminus of Gag to release a p3 peptide, resulting in FV particles that are morphologically similar to the immature particles of orthoretroviruses. Another prominent feature of FV Gag is that it does not contain the orthoretroviral cysteine-histidine (Cys-His) motifs flanked with clustered positively charged residues in NC. Cys-His motifs are
10803
10804
LEE AND LINIAL
J. VIROL.
FIG. 1. Diagram of the FV genome showing the open reading frames for Gag, Pol, and Env as well as the accessory proteins Tas and Bet. The positions of the three glycine/arginine-rich (GR) boxes are in black at the C terminus of Gag. Arrows show the proteolytic cleavages by the FV PR encoded in Pol.
required for essential steps of virus replication including reverse transcription, integration, genomic RNA packaging, viral assembly, and infectivity (reviewed in reference 9). Instead, FV Gag contains three glycine- and arginine-rich regions (GR boxes) at the C terminus of Gag. GR box 1 has nucleic acid binding activity in vitro, and GR box 2 contains a nuclear localization signal (24, 28). The C terminus of Gag has been shown to contain domains important for RNA packaging as well as expression, cleavage, and packaging of the Pol protein (25). Mammalian orthoretroviral Pol is synthesized as a precursor that is cleaved by viral PR into individual pol-encoded enzymes, PR, RT, and integrase (IN). However, FV Pol is cleaved only once to release an 80-kDa Pol protein (Pol; PRRT) and a 40-kDa IN protein (Fig. 1). Only the Pol precursor protein is packaged into particles; neither PR-RT or IN is encapsidated (19, 23). Mutational analysis of the Pol cleavage site showed that precursor cleavage is required for IN, but not PR, activity and that RT activity is minimally affected by lack of cleavage (23). A hallmark of FV is that reverse transcription occurs concurrent with, or right after, viral assembly, leading to an infectious DNA genome (17, 22, 27, 30), but nothing is known about temporal regulation of reverse transcription. We previously showed that truncation mutations at the C terminus of Gag variously affected the packaging of both RNA and Pol into virions (25). However, that study utilized the premature termination codon mutations near the three GR boxes, and these mutations led to low levels of Pol protein expression (14), making the analysis of mutational effects on Pol packaging difficult. To overcome this problem, in the present study we engineered precise deletion or substitution mutations of the GR boxes. These mutations did not have negative effects on Pol levels, and they allowed us to further characterize the role of the C terminus of Gag in both RNA and Pol packaging. We have found that the first GR box contains determinants that are necessary for Pol packaging. Specifically, the clustered positively charged residues in GR box 1 are not required for RNA packaging but are required for Pol packaging, suggesting that a Gag-Pol complex is a precursor to Pol packaging. MATERIALS AND METHODS DNA mutagenesis and cloning. All of the mutations at the C terminus of Gag were constructed in the context of full-length prototype primate FV containing a cytomegalovirus immediate-early promoter (pcPFV) (25). Site-directed deletion or substitution mutations in Gag were obtained by two rounds of PCR using four oligonucleotides. Two outer oligonucleotides (forward and reverse) were
designed to anneal to the 5⬘ or 3⬘ end of the gag gene with the addition of two engineered unique restriction sites at each end. Two inner mutagenic oligonucleotides (in either the forward or reverse orientation) were designed to be complementary to the gag sequences except for the desired mutations. PCRs were as previously described (14). Each mutant construct was sequenced to confirm the presence of the correct mutational changes. Primer sequences will be supplied upon request. The deletion in a ⌬GR1 mutant starts at nucleotide position 2514 of the pcHFV (25) and ends at position 2594. A ⌬GR2 mutant has a deletion between positions 2661 and 2732 of the pcHFV vector. Cell cultures and transfections. 293T cells and FAB cells (29) were cultured in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin. Transient transfection was done using 1 mg/ml polyethyleneimine (Polysciences, Warrington, PA) as previously described (5). For the viral infectivity assay, the FAB assay was done as previously described (29). Western blot analysis. At between 45 and 48 h postransfection, cells were scraped off plates with 1⫻ sodium dodecyl sulfate (SDS) sample buffer (12.5% 4⫻ Tris-HCl–SDS [pH 6.8], 10% glycerol, 2% SDS, 1% 2-mercaptoethanol, 0.01% bromophenol blue), and lysates were prepared by using Qiashredder (Qiagen) in accordance with the manufacturer’s protocol. The supernatants were collected, and cell debris was cleared by low-speed centrifugation and filtration through a 0.45-m-pore-diameter syringe filter. Viral particles were pelleted through 20% sucrose cushions at 25,000 rpm for 2 h using an L7 ultracentrifuge (Beckman). The viral pellets were resuspended in 1⫻ SDS sample buffer prior to loading onto SDS-10% polyacrylamide gels. Western blot analyses were as previously described (14) using a 1:5,000 dilution of polyclonal rabbit anti-Gag antibody (1), a 1:800 dilution of monoclonal mouse anti-Pol antibody (25), or a 1:2,000 dilution of monoclonal mouse anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Proteins were visualized by the Odyssey detection system (Li-Cor, Lincoln, NE), according to the manufacturer’s protocol. The Odyssey detection system was shown to have an eightfold linear range (23). Quantitation of Pol packaging efficiency by Western blot analysis. The amount of Pol protein in virions was normalized to the level of Pol in transfected cells. This calculated level of Pol was further normalized to the number of virions determined by calculating the ratio of Gag proteins in the medium to Gag proteins expressed in the cells after the level of cellular Gag was normalized to that of an endogenous GAPDH protein. Iodixanol density gradient centrifugation and fractionation. Equilibrium centrifugation was carried out on a 20 to 55% iodixanol step gradient in phosphatebuffered saline. Particles pelleted as described above were loaded onto the gradient and centrifuged in a Ti55 rotor at 32,000 rpm and 4°C for 18 h. Aliquots of 320 l were collected from the top, and a total of 13 fractions were prepared to be analyzed for density. Proteins in each fraction were precipitated with 10% trichloroacetic acid and resuspended in 1⫻ SDS sample buffer for Western blot analysis. RPA. At between 45 and 48 h postransfection, RNA was extracted from cells and pelleted virions, and RNase protection assay (RPA) analysis was done as previously described (15) using an FV LTR-specific riboprobe as detailed previously (25). Reactions were run on 5% polyacrylamide gels, and the dried gels were directly scanned with a Molecular Dynamics PhosphorImager. Radioactively labeled protected bands were quantitated with ImageJ software. The RPA has shown to have a fivefold dynamic range (19). Calculation of viral RNA packaging efficiency. The packaging efficiencies of the FV RNA were determined by calculating the amount of FV RNA in virions
VOL. 82, 2008
FV Gag DETERMINANTS FOR Pol PACKAGING INTO VIRIONS
FIG. 2. Western blot analysis of gross mutations in the GR boxes. (A) Substitution and deletion mutations in the C-terminal GR boxes of Gag. In the GR1-HA mutant, the N-terminal 11 amino acids of GR box 1 (shown in the first line) were replaced by the HA epitope tag sequence. The first 27 amino acids were deleted in the ⌬GR1 mutant, and the N-terminal 24 amino acids were deleted in the ⌬GR2 mutant. Gag and Pol proteins in the GR mutants were examined by Western blot analysis using either anti-Gag antibody (B) or anti-Pol antibody (C). The Odyssey system was used for detection. Molecular mass markers (MWM) are shown in kilodaltons. Lanes 1 to 6 contain cell lysates, and lanes 7 to 12 contain pelletable virus particles obtained from the supernatants of 293T cells transiently transfected with each indicated proviral plasmid DNA.
normalized to the level of FV RNA in cells (measured by RPA). This calculated level of RNA was further normalized to the number of virions determined by calculating the ratio of Gag proteins in the medium to the Gag proteins expressed in the cells after the level of cellular Gag was normalized to that of an endogenous GAPDH protein (measured by the Odyssey detection system).
10805
proteins in cells (Fig. 2B, lanes 3 to 6) and in the pelletable supernatants (Fig. 2B, lanes 9 to 12). It is known that Env is required for FV egress from cells (1, 8). Therefore, as a control for nonspecific particle release, a mutant lacking the Env protein (⌬Env) was used and was found to synthesize intracellular Gag at wt levels but not to produce extracellular particles (Fig. 2B, lanes 2 and 8). All mutants synthesized at least wt levels of Pol proteins in cells (Fig. 2C, lanes 3 to 6). Cells transfected with the deletion of GR box 1 (⌬GR1) consistently expressed more Pol than wt (Fig. 2C, lane 4). The ⌬GR2 mutant showed Pol detected in pelletable supernatants, although less than wt (Fig. 2C, lanes 9 and 12). However, in spite of the higher expression of Pol in cells transfected with the ⌬GR1 mutant, Pol packaging was greatly reduced (Fig. 2C, lane 10). Further, we could not detect any extracellular Pol from cells containing the HA substitution mutant (GR1-HA) (Fig. 2C, lane 11). We quantitated the efficiency of Pol packaging normalized to wt, as detailed in Materials and Methods (Table 1). We found that the two mutations of GR box 1 reduced Pol packaging from 50-fold (⌬GR1) to 1,000-fold (GR1-HA), while ⌬GR2 retained about 40% of the wt Pol packaging level. Infectivity of the mutants was measured by the FAB assay as described previously (29) (Table 1). ⌬GR1 had 0.1% of wt infectivity, and GR1-HA is noninfectious. Despite only a 2.5-fold reduction in Pol packaging, ⌬GR2 showed a 100-fold reduction in infectivity, indicating the requirement for GR box 2 in other stages of viral replication. These results suggest that GR box 1 contains a sequence required for Pol packaging. GR box 1 is not required for RNA packaging. Since PES in genomic RNA are required for efficient incorporation of Pol into virions (11), one possible reason for the decreased or abolished Pol packaging seen in the GR box 1 mutants is that they lack viral RNA. To quantitate the RNA packaging efficiencies of the mutants, 293T cells were transfected with wt or mutant proviral plasmids and RPAs were performed (Fig. 3A and B). Each of the cell lysates except for the mock transfection showed a band of 260 nucleotides (nt), representing the 3⬘ end of the viral genomic RNA (Fig. 3B, cell). As previously
RESULTS Mutations in GR box 1 decrease Pol packaging into virions. Previously we found that truncation mutations that removed the C terminus of FV Gag affected many processes of viral assembly (25). A Gag C-terminal truncation mutant lacking all three GR boxes and flanking sequences abolished genomic RNA packaging as well as Pol expression and packaging. However, a smaller deletion that removed only the regions downstream of GR box 2 had nearly wild-type (wt) levels of Pol expression and RNA packaging. These results suggested the possibility that GR box 1, GR box 2, and/or the flanking sequences contain determinants for RNA and Pol packaging. To investigate the roles of the C terminus of Gag in viral assembly, we generated precise deletions of either GR box 1 or 2 in the full-length pcPFV provirus (25) or used a previously described mutant containing a substitution of hemagglutinin (HA) sequences for GR box 1 (28) (Fig. 2A). Western blot analysis was used to examine intracellular and supernatant Gag and Pol protein expression and packaging (Fig. 2B and C). Cells transfected with the three Gag mutants expressed wt levels of Gag
TABLE 1. Pol packaging efficiency and infectivity of the GR box 1 mutants Infectivityb Construct
wt ⌬GR1 ⌬GR2 GR1-HA GR1Ala HAArg(⫹) Gly(⫺) Lys(⫹) GR2-subs GR3-subs miniGR1
Pol packaging efficiencya
IU/ml
Normalized to wt
1.00 0.023 ⫾ 0.010 0.420 ⫾ 0.050 0.001 ⫾ 0.002 0.007 ⫾ 0.008 0.350 ⫾ 0.080 0.600 ⫾ 0.210 0.640 ⫾ 0.520 0.180 ⫾ 0.006 0.200 ⫾ 0.090 0.450 ⫾ 0.100
(1.5 ⫾ 0.5) ⫻ 105 (1.2 ⫾ 0.3) ⫻ 102 (1.1 ⫾ 0.4) ⫻ 103 ⬍10 ⬍10 (4.9 ⫾ 1.5) ⫻ 104 (7.4 ⫾ 2.1) ⫻ 104 (6.1 ⫾ 1.3) ⫻ 104 (1.2 ⫾ 0.7) ⫻ 105 (1.0 ⫾ 0.3) ⫻ 105 (1.1 ⫾ 0.6) ⫻ 105
1.0 0.001 0.01 ⬍0.0001 ⬍0.0001 0.3 0.5 0.4 0.8 0.7 0.7
a The efficiencies of Pol packaging were normalized to that of the wt as measured by Odyssey Western blot analyses as shown in Fig. 2 and 4. The numbers are the averages and standard deviations from three or four independent assays. b Infectivity was measured by the FAB assay as described previously (29). The numbers are the averages and standard deviations from four independent assays.
10806
LEE AND LINIAL
FIG. 3. RNA packaging efficiencies of the GR mutants measured by RPA. (A) A 503-nt free riboprobe was in vitro transcribed. The dotted lines indicate non-FV LTR sequences in the expression vector. Three protected bands are predicted after hybridization with the riboprobe and subsequent RNase treatment. (B and D) RPA using cell lysates and pelletable particles. The size of the free probe is indicated. Control lanes for probes contain the reactions using the riboprobe in the absence of RNase (RNase⫺) or in the presence of RNase (RNase ⫹). Mock-transfected cells are included. ⌬Env is a deletion mutant of the env gene and is used as a control for nonspecific particle release. (C) Western blot analysis to measure the levels of Gag protein in the cell and pelletable supernatant with anti-Gag antibody, detected using the Odyssey system. Molecular mass markers (MWM) are indicated in kilodaltons.
described (25), 3⬘ genomic RNA was more readily detected with the LTR riboprobe than 5⬘ genomic RNA, although the reason is not known. RNA extracted from the pelletable supernatant from wt-transfected cells showed all three protected fragments of 427 nt (genomic or proviral DNA), 330 nt (5⬘end of genomic RNA), and 260 nt (Fig. 3B, virus). All three Gag mutant RNAs resulted in detection of 330- and 260-nt bands, although these were less intense than those of the wt. RNA from the negative control pellets (⌬Env or mock transfected) did not result in protected fragments. To quantitate the RNA packaging efficiencies of each mutant, portions of each cell lysate or viral pellet were quantitatively analyzed for Gag by Western blot analyses (Fig. 2B). All mutants expressed similar amounts of cellular Gag. However, all of the GR box mutants released less than one-seventh of the number of particles released by the wt in this experiment (Fig. 2B). The Western blot results, along with the RPA results, were quantitated using ImageJ software, and the RNA packaging efficiencies were calculated as described in Materials and Methods. All three mutants packaged RNA at 60 to 80% of the wt level (0.63 ⫾ 0.04, 0.81 ⫾ 0.07, and 0.68 ⫾ 0.21 [averages ⫾ standard deviations] for ⌬GR1, GR1-HA, and ⌬GR2, respectively). Thus, we can conclude that deletion or substitution of GR box 1 prevents Pol, but not RNA, packaging, while deletion of GR
J. VIROL.
box 2 does not affect either RNA or Pol packaging. These results suggest that GR box 1 contains a domain required for Pol interaction. GR box 1 is required for Pol packaging but not VLP formation. In order to delineate which amino acid residues in GR box 1 are necessary for Pol packaging, we created a series of amino acid substitution mutants (Fig. 4A). First, we replaced the N-terminal 11 amino acids in GR1 with alanines (Fig. 4A, GR1Ala). Gag proteins of the GR1Ala mutant were expressed and detected in both cells and pelletable supernatants at the same level as for the wt (Fig. 4B, lane 4). Pol was expressed at wt levels in cells but was detected at very low levels in virus particles (Fig. 4C, lane 4). GR1Ala virus was not competent for Pol packaging and was replication defective (Table 1). One explanation for these results is that the extracellular Gag is in aggregates rather than bona fide virus-like particles (VLPs). In order to determine whether expression of GR1Ala or ⌬GR1 leads to production of VLPs, the density of extracellular pelletable Gag was measured. Concentrated supernatants from transiently transfected 293T cells with each mutant proviral plasmid DNA were layered onto iodixanol step density gradients. The density profiles of each fraction from the ⌬GR1, GR1Ala, and wt gradients were similar (Fig. 5A). For both mutants, Gag proteins were recovered in the same fractions as for the wt (fractions 8 and 9 in Fig. 5B to D) at the expected density for FV (27). Since VLPs would be expected to contain nucleic acid, we next examined the efficiency of genomic RNA packaged into the GR1Ala mutant (Fig. 3C and D). Although the GR1Ala-transfected cells expressed levels of Gag proteins equivalent to those for the wt, fewer mutant particles were released into the supernatant (Fig. 3C). We used the same FV LTR riboprobe for RPA as in Fig. 3A. Although the mutant showed somewhat reduced levels of viral genomic RNA in the pelletable supernatant (Fig. 3D), quantitation based on particle numbers showed that the GR1Ala mutant packaged genomic RNA as efficiently as the wt (efficiency of 1.10 ⫾ 0.34 relative to the wt). Taken together, these results show that elimination of GR box 1 by alanine substitution or deletion does not affect virus assembly or RNA packaging but prevents Pol incorporation into VLPs. Positively charged residues, but not glycine residues, in GR box 1 are required for Pol packaging. Since arginine/glycine residues are conserved in the GR boxes of all FV isolates, either or both of these two amino acids could be required for Pol packaging. We performed site-directed mutagenesis of the N-terminal 11 amino acids of GR box 1 to create the additional mutants depicted in Fig. 4A. To test whether introduction of arginine residues into the GR1-HA mutant can restore the ability to package Pol, we created an HAArg(⫹) mutant based on the sequence of GR1-HA that is defective for packaging, by replacing residues in the GR1-HA sequence with arginines at the same position as found in the wt. The importance of glycine residues was tested by replacing them in the wt sequence with amino acids found in the HA sequence to create Gly(⫺). To test the requirement for basic residues, lysines were substituted for arginines in the wt to create Lys(⫹). The levels of Gag and Pol expression were measured by Western blot analysis (Fig. 4B and C). In all three mutants, similar levels of intracellular Gag proteins were synthesized and released to the supernatant from the transfected cells (Fig. 4B, lanes 5, 8, and 11).
VOL. 82, 2008
FV Gag DETERMINANTS FOR Pol PACKAGING INTO VIRIONS
10807
FIG. 4. Substitution mutations in GR box 1. (A) The N-terminal 11 amino acids of GR1 were replaced with the corresponding sequences indicated for each mutant construct. The positively charged residues are underlined. (B and C) Gag and Pol protein expression and viral release were measured by Odyssey Western blot analysis with either anti-Gag antibody (B) or anti-Pol antibody (C). Blots from different gels have been aligned as indicated by lines between blots. MWM, molecular mass markers in kilodaltons.
Amounts of Pol equivalent to those for the wt were expressed intracellularly and packaged into pelletable particles (Fig. 4C, lanes 5, 8, and 11). Pol packaging efficiencies of the mutants were in a range of 35 to 64% of the wt level (Table 1). All three mutants had high infectious titers, at least 30% of wt (Table 1). These results indicate that GR box 1 Gly residues are not required for Pol packaging and that only basic residues in GR box 1 (either lysine or arginine) are required. GR box 1 can be replaced with other GR boxes for Pol packaging. Among the three GR boxes, the only conserved residues are Gly and Arg, although the number and relative position of Gly and Arg vary. Both GR boxes 2 and 3 are intact in the GR box 1 mutants that are deficient for Pol packaging, showing that GR boxes distal to GR box 1 cannot compensate for the lack of basic residues at the GR box 1 site. We next asked whether moving the other GR boxes to the position of GR box 1 would allow them to be used for Pol packaging. To test this, we substituted either GR box 2 or 3 for GR box 1 (Fig. 4A). Both the GR2-subs and GR3-
subs mutants expressed intracellular Gag proteins and produced extracellular pelletable Gag at levels equivalent to those for the wt (Fig. 4B, lanes 6 and 15). Both precursor and cleaved Pol were detected at levels at least as high as those for the wt in cells, and cleaved Pol was packaged as efficiently as wt Pol (Fig. 4C, lanes 6 and 15). We found that the mutants packaged high levels of Pol and were as infectious as the wt (Table 1). This suggests that for Pol interaction as few as two positively charged amino acids at the site of GR box 1 are sufficient. To examine whether any other amino acids near the basic residues in GR box 1 are required for Pol packaging, alanines were substituted for amino acids in GR box 1, leaving only the central five amino acids, Arg-Gly-Arg-Gly-Arg, intact (Fig. 4A, mini GR1 mutant). The mini GR1 mutant assembled Gag particles, packaged Pol 45% as efficiently as the wt, and was infectious (Fig. 4B and C, lane 7; Table 1). Thus, as few as two positively charged amino acids at the center of GR box 1 in the Gag protein are sufficient for the Gag-Pol interactions required for Pol incorporation into virus particles.
10808
LEE AND LINIAL
FIG. 5. Density gradient analysis of the GR box 1 mutants. (A) A 20 to 55% iodixanol step density gradient was used for centrifugation of pelletable Gag from each mutant; 13 fractions were collected, and the density of each fraction was measured. (B to D) Trichloroacetic acid-precipitated proteins in each fraction were run on SDS-polyacrylamide gel electrophoresis and probed with anti-Gag antibody. The Odyssey detection system was used to visualize proteins. Molecular mass markers (MWM) are indicated in kilodaltons. (B) wt; (C) ⌬GR1; (D) GR1Ala mutant. The input lane for each gradient contains 10% of the viral pellet used on the gradient.
DISCUSSION In this study, we examined the requirement for Gag sequences in packaging of Pol into FV. Distinct from orthoretroviruses and similar to HBV, FV require specific regions of genomic RNA for Pol packaging (11). However, such a requirement does not preclude the possibility that Pol must first bind to Gag prior to interaction with the RNA sequences to permit encapsidation. The PES lie near to, or overlap, the sequences required for genomic RNA packaging (19). Since FV Gag processing and Pol and RNA incorporation into virions appear to be coupled, it is difficult to analyze the complex interactions between Gag, Pol, and RNA in virus assembly by modifying only one variable without altering the other two variables. One strategy to overcome this hurdle is to use an FV four-vector system based on the cotransfection of cells with four plasmids independently expressing Gag, Pol, Env, and packageable RNA (11). These vectors were used by Peters et al. (19) to define two regions of PES essential for Pol packaging but dispensable for RNA packaging. However, a potential problem of the four-vector system is that each component is overexpressed so that the normal ratios of the viral compo-
J. VIROL.
nents are not retained. We first attempted to use the fourvector system to examine whether Gag-Pol interactions are required for Pol incorporation into virions. However, we found that a large excess of Pol expression relative to Gag from separate cytomegalovirus promoter-driven plasmids negated the normal regulation of Pol and interfered with the effects of mutations of Gag on Pol packaging (data not shown), so this approach was not pursued. We instead used Gag mutations in the context of the fulllength provirus. We searched for Gag mutations that uncouple RNA and Pol packaging, and in the present study, we describe several GR box 1 mutants which packaged genomic RNA similarly to the wt but were defective in Pol packaging. The existence of such mutants strongly suggests that interactions between Gag and Pol are required for RNA binding. We cannot rule out indirect roles for Gag in Pol packaging. One possibility is that Pol alone binds to RNA through the PES but that Gag binding is required to stabilize the Pol-RNA complex. A second is that Gag binds to Pol to bring it to the site of viral assembly at the pericentriolar region (31), where Pol is then packaged into assembling particles. However, we favor a model in which Gag interacts with Pol and then Gag in a complex binds to RNA, which brings Pol into the virions, as a “passenger.” Having RNA binding specificity in Gag rather than Pol makes sense, in that Pol must be able to travel along the entire length of the genome in order to synthesize cDNA. Having a high-affinity RNA binding site in Pol could be problematic in this regard. The model that the Gag-Pol complex is assembled into virions via binding to RNA, not through Gag-Gag interactions, predicts that Gag-Pol cannot coassemble with Gag through protein-protein interactions because the Gag structure in the complex is different from that of free Gag. Our model suggests that for FV, rather than Gag being physically linked to Pol as in orthoretroviruses, leading to Pol packaging through Gag-Gag interactions, a Gag-Pol complex forms and then Gag in the complex binds to RNA, leading to Pol incorporation. In orthoretroviruses such as HIV-1 and avian leukosis sarcoma virus, the NC domain, corresponding to the C terminus of FV Gag, has two conserved structural features involved in genomic RNA packaging (reviewed in reference 2). One is the Cys-His (CH) motifs themselves, and the other is the clustered basic residues surrounding the CH motifs. Mutational analysis show that it is the stretches of positively charged residues and not the CH motifs that are required for recognition and binding to packaging signal ()-containing RNA for packaging. However, the structure contributed mainly by the CH motifs also needs to be present to allow RNA packaging (10, 13, 21). In FV Gag, there appear to be redundant determinants for RNA packaging at the C terminus, since deletion of either GR box 1 or 2 does not abolish RNA packaging. It is possible that clustered positively charged residues in the GR boxes are involved in binding to sequences for genomic RNA packaging in FV as well, although this has not yet been examined. Basic residues in the GR boxes could play different roles in Gag-Pol and Gag- RNA interactions, which is reminiscent of situation in avian leukosis sarcoma virus NC, where the basic residues play different roles in Gag-Gag and Gag- RNA interactions (15). In the case of FV, Gag binding to RNA for RNA packaging may require only clustered positively charged residues, while Gag binding to Pol for incorporation into virions re-
VOL. 82, 2008
FV Gag DETERMINANTS FOR Pol PACKAGING INTO VIRIONS
quires more complex interactions inherent in the protein tertiary structure encompassing GR box 1. The FV PR proteolytic cleavages of Gag and precursor Pol (PrPol) are essential for virus replication (7, 23). Like other retroviral PRs, FV PR belongs to the same family of aspartic PRs as orthoretroviral PRs and functions as a homodimer with the active site formed by two PR monomers (20). Since only PrPol, and not the individual cleavage products PR-RT and IN, is incorporated into virus particles (19), it has been hypothesized that PrPol dimerizes during assembly into particles and that dimerization leads to activation of PR, which then cleaves both PrPol and Gag. For this reason, Gag processing has been considered to be indicative of Pol packaging into assembling particles (11, 19, 23). Inconsistently with this hypothesis, however, some of the GR1 mutants fail to package Pol into particles, although there is cleaved Gag in both cell lysates and pelletable supernatants. FV capsid assembly occurs in a pericentriolar region of the cytoplasm through a cytoplasmic targeting/retention signal (CTRS) in Gag (31). The Gag protein of a CTRS-negative mutant does not accumulates at the pericentriolar area, as visualized by immunofluorescence (31), and transmission electron microscopy did not reveal VLPs at the pericentriolar region (16), indicating that the CTRS-negative mutant is deficient in normal virus assembly. In the CTRS-negative mutant, cellular Gag proteins were cleaved about 50% as efficiently as in the wt (6). In addition, PR was shown to cleave its own precursor Pol protein in the absence of Gag when they were expressed in the cells transfected either with a mutant proviral DNA lacking gag (23) or with components of the four-vector system without the pGag plasmid (11). Taken together, the results imply that FV PR can be activated intracellularly, in the absence of Pol incorporation into assembling particles. It is possible that Pol interacts with Gag to cleave off the p3 peptide near the pericentriolar region or in other regions of the cytoplasm. It is unlikely that interactions between Gag and Pol take place in the nucleus, although both proteins were transiently localized there (12, 24, 28), since a Gag mutant lacking the nuclear localizing signal found in GR box 2 (24, 28) has normal proteolytic cleavages in both Gag and Pol. Immunofluorescence studies of FV-infected cells show that Pol proteins accumulate to appreciable levels in the cytoplasm outside of the pericentriolar region (S. F. Yu, unpublished data). Since defined RNA binding sites are required for Pol packaging, it is likely that only a small number of Gag-Pol complexes bind to the RNA and are packaged. This raises the question of what the biological role (if any) is for the intracellular FV Pol proteins that are not packaged but accumulate in both the nucleus and cytoplasm. In both HBV and duck HBV, high levels of the P protein (RT) are detected in the cytoplasm and have been shown to suppress the accumulation of mRNA and pregenomic RNA, resulting in the reduced production of viral proteins, including the P and capsid proteins (4, 26). Those authors speculate that high levels of intracellular P protein lead to the maintenance of chronic infection by the suppression of the immune response. However, it is not known whether intracellular FV Pol also has a role in negative feedback regulation of mRNA for viral proteins. In summary, we have identified Gag mutants which assemble particles of normal density that contain RNA but lack Pol.
10809
These results extend our understanding of the unique mechanism of Pol encapsidation and demonstrate that specific sequences in the C terminus of Gag (equivalent to the orthoretroviral NC region) are required, in addition to genomic RNA sequences. ACKNOWLEDGMENTS We thank Rachel Life for critical reading of the manuscript. This work was funded by NIH grant R01 CA 18282 to M.L.L. REFERENCES 1. Baldwin, D. N., and M. L. Linial. 1998. The roles of Pol and Env in the assembly pathway of human foamy virus. J. Virol. 72:3658–3665. 2. Berkowitz, R., J. Fisher, and S. P. Goff. 1996. RNA packaging. Curr. Top. Microbiol. Immunol. 214:177–218. 3. Boyer, P. L., C. R. Stenbak, P. K. Clark, M. L. Linial, and S. H. Hughes. 2004. Characterization of the polymerase and RNase H activities of human foamy virus reverse transcriptase. J. Virol. 78:6112–6121. 4. Cao, F., and J. E. Tavis. 2006. Suppression of mRNA accumulation by the duck hepatitis B virus reverse transcriptase. Virology 350:475–483. 5. Durocher, Y., S. Perret, and A. Kamen. 2002. High-level and high-throughput recombinant protein production by transient transfection of suspensiongrowing human 293-EBNA1 cells. Nucleic Acids Res. 30:e9. 6. Eastman, S. W., and M. L. Linial. 2001. Identification of a conserved residue of foamy virus Gag required for intracellular capsid assembly. J. Virol. 75:6857–6864. 7. Enssle, J., N. Fischer, A. Moebes, B. Mauer, U. Smola, and A. Rethwilm. 1997. Carboxy-terminal cleavage of the human foamy virus Gag precursor molecule is an essential step in the viral life cycle. J. Virol. 71:7312–7317. 8. Fischer, N., M. Heinkelein, D. Lindemann, J. Enssle, C. Baum, E. Werder, H. Zentgraf, J. G. Muller, and A. Rethwilm. 1998. Foamy virus particle formation. J. Virol. 72:1610–1615. 9. Goff, S. P. 2007. Retroviridae: the retroviruses and their replication, p. 1999–2069. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Wolters Kluwer/Lippincott Williams and Wilkins, Philadelphia, PA. 10. Gorelick, R. J., T. D. Gagliardi, W. J. Bosche, T. A. Wiltrout, L. V. Coren, D. J. Chabot, J. D. Lifson, L. E. Henderson, and L. O. Arthur. 1999. Strict conservation of the retroviral nucleocapsid protein zinc finger is strongly influenced by its role in viral infection processes: characterization of HIV-1 particles containing mutant nucleocapsid zinc-coordinating sequences. Virology 256:92–104. 11. Heinkelein, M., C. Leurs, M. Rammling, K. Peters, H. Hanenberg, and A. Rethwilm. 2002. Pregenomic RNA is required for efficient incorporation of Pol polyprotein into foamy virus capsids. J. Virol. 76:10069–10073. 12. Imrich, H., M. Heinkelein, O. Herchenroder, and A. Rethwilm. 2000. Primate foamy virus Pol proteins are imported into the nucleus. J. Gen. Virol. 81:2941–2947. 13. Lee, E.-G., A. Alidina, C. May, and M. L. Linial. 2003. Importance of basic residues in binding of Rous sarcoma virus nucleocapside to the RNA packaging signal. J. Virol. 77:2010–2020. 14. Lee, E.-G., D. Kuppers, M. Horn, J. Roy, C. May, and M. L. Linial. 2008. A premature termination codon mutation at the C terminus of foamy virus Gag downregulates the levels of spliced pol mRNA. J. Virol. 82:1656–1664. 15. Lee, E.-G., and M. L. Linial. 2004. Basic residues of the retroviral nucleocapsid play different roles in Gag-Gag and Gag-psi RNA interactions. J. Virol. 78:8486–8495. 16. Life, R. B., E.-G. Lee, S. W. Eastman, and M. L. Linial. 2008. Mutations in the amino terminus of foamy virus Gag disrupt morphology and infectivity but not targeting of assembly. J. Virol. 82:6109–6119. 17. Moebes, A., J. Enssle, P. D. Bieniasz, M. Heinkelein, D. Lindemann, D. Bock, M. O. McClure, and A. Rethwilm. 1997. Human foamy virus reverse transcription that occurs late in the viral replication cycle. J. Virol. 71:7305– 7311. 18. Nassal, M., and H. Schaller. 1996. Hepatitis B virus replication—an update. J. Viral Hepatitis 3:217–226. 19. Peters, K., T. Wiktorowicz, M. Heinkelein, and A. Rethwilm. 2005. RNA and protein requirements for incorporation of the Pol protein into foamy virus particles. J. Virol. 79:7005–7013. 20. Pfrepper, K. I., M. Lo ¨chelt, M. Schnolzer, and R. M. Flu ¨gel. 1997. Expression and molecular characterization of an enzymatically active recombinant human spumaretrovirus protease. Biochem. Biophys. Res. Commun. 237: 548–553. 21. Poon, D. T., J. Wu, and A. Aldovini. 1996. Charged amino acid residues of human immunodeficiency virus type 1 nucleocapsid p7 protein involved in RNA packaging and infectivity. J. Virol. 70:6607–6616. 22. Roy, J., W. Rudolph, T. Juretzek, K. Gartner, M. Bock, O. Herchenroder, D. Lindemann, M. Heinkelein, and A. Rethwilm. 2003. Feline foamy virus genome and replication strategy. J. Virol. 77:11324–11331.
10810
LEE AND LINIAL
23. Roy, J., and M. L. Linial. 2007. Role of the foamy virus Pol cleavage site in viral replication. J. Virol. 81:4956–4962. 24. Schliephake, A. W., and A. Rethwilm. 1994. Nuclear localization of foamy virus Gag precursor protein. J. Virol. 68:4946–4954. 25. Stenbak, C. R., and M. L. Linial. 2004. Role of the C terminus of foamy virus Gag in RNA packaging and Pol expression. J. Virol. 78:9423–9430. 26. Yao, E., Y. Gong, N. Chen, and J. E. Tavis. 2000. The majority of duck hepatitis B virus reverse transcriptase in cells is nonecapsidated and is bound to a cytoplasmic structure. J. Virol. 74:8648–8657. 27. Yu, S. F., D. N. Baldwin, S. R. Gwynn, S. Yendapalli, and M. L. Linial. 1996. Human foamy virus replication—a pathway distinct from that of retroviruses and hepadnaviruses. Science 271:1579–1582.
J. VIROL. 28. Yu, S. F., K. Edelmann, R. K. Strong, A. Moebes, A. Rethwilm, and M. L. Linial. 1996. The carboxyl terminus of the human foamy virus Gag protein contains separable nucleic acid binding and nuclear transport domains. J. Virol. 70:8255–8262. 29. Yu, S. F., and M. L. Linial. 1993. Analysis of the role of the bel and bet open reading frames of human foamy virus by using a new quantitative assay. J. Virol. 67:6618–6624. 30. Yu, S. F., M. D. Sullivan, and M. L. Linial. 1999. Evidence that the human foamy virus genome is DNA. J. Virol. 73:1565–1572. 31. Yu, S. F., S. W. Eastman, and M. L. Linial. 2006. Foamy virus capsid assembly occurs at a pericentriolar region through a cytoplasmic targeting/ retention signal in Gag. Traffic 7:966–977.
