Altered Plaque Formation by Recombinant Vaccinia Virus Expressing Simian Immunodeficiency Virus Nef

Abstract

The nef gene of primate lentiviruses encodes a myristoylated protein that is important for pathogenicity and the maintenance of high virus loads. A deletion in nef leads to a significant reduction of the pathogenicity of simian immunodeficiency virus (SIV) in macaques. At the cellular and biochemical levels, Nef has been shown to down-regulate CD4 and major histocompatibility complex class I molecules and to interact with cellular protein kinases. The importance of these activities for Nef function remains uncertain. We have prepared vaccinia virus recombinants expressing different alleles of SIV nef. When grown on TK⁻ 143 cells, recombinants constructed with the nef allele from SIVmac1A11 produced typical plaques while recombinants expressing the nef allele from SIVmac239-R1 gave rise to plaques with altered morphology. By using chimeric Nef proteins and site-directed mutagenesis, the amino acid responsible for altered plaque formation was mapped to a leucine at residue 211. In vitro phosphorylation of immunoprecipitates prepared from cells infected with the vaccinia virus recombinants resulted in labeled proteins of 62 and 90 kDa. The recombinants differed in the ability to stimulate phosphorylation, and the leucine at residue 211 was again found to be the determining amino acid. These results might help elucidate the role of nef in the pathogenesis of SIV.

The nef gene of the human and simian immunodeficiency viruses (HIV and SIV, respectively) encodes a small, myristoylated protein that associates with membranes and the cytoskeletal matrix (10, 12) and that appears to be a minor component of the virus particle (31). The exact role that Nef plays in the virus life cycle and how Nef functions at the cellular level are the subjects of much current research. What is clear from genetic studies is the importance of Nef for viral growth and pathogenesis.

Molecularly cloned SIVs containing deletions in the nef gene establish persistent infections with low viral loads and reduced pathogenicity in rhesus macaques (13). This contrasts with the high viral titers and debilitating disease typical of infection with the wild-type virus. Thus, Nef is required for SIV to reach its full pathogenic potential, and the same appears to be true of HIV. An analysis of the virus from a group of asymptomatic, long-term human survivors who were infected with HIV type 1 (HIV-1) by blood transfusions from a single donor found the donor virus to contain a deletion in the nef gene (6). One feature that could account for the behavior of nef-deficient viruses in vivo is the reduced infectivity of nef(−) HIV-1 mutants in vitro (1, 5, 21, 26).

Several lines of evidence suggest that Nef has the ability to alter cell activation and signal transduction pathways. Nef has been shown to downregulate CD4 and MHC class I molecules from the cell surface (8, 27). It has also been reported to interfere with the induction of interleukin 2 mRNA in T-cell lines (19). In biochemical studies, Nef has been observed to interact with Hck and Lyn (23), members of the Src family of nonreceptor tyrosine kinases, as well as to associate with a serine-threonine kinase (24, 25) whose characteristics are similar to those of p21-activated kinases (17, 22). Although the relevance of these activities for virus growth and disease progression remains uncertain, their study should provide insights into those properties of Nef that are crucial for biological function.

Here, we report the formation of atypical plaques by vaccinia virus expressing certain alleles of SIV nef, indicating the possibility of a new and useful system with which to study the biological functions of Nef.

TK⁻ 143 and BSC-1 cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 5% fetal bovine serum; 5′-bromodeoxyuridine (25 μg/ml) was added to the medium for TK⁻ 143 cells. Recombinant vaccinia viruses (WR strain) expressing Nef were generated by using standard procedures (20). The various nef genes were cloned in the vector pSC11 (3), which contains the β-galactosidase gene. Recombinants were identified by their blue phenotype in the presence of 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal), and Nef expression was confirmed by Western blotting.

The Nef1A11 and Nef157 genes were inserted at the SmaI site of pSC11 as blunted HindIII fragments. To prepare NcoI chimeras, pSC11-Nef1A11 and pSC11-Nef157 were digested with NcoI and PstI; both enzymes cut once, NcoI within Nef and PstI within the vector. The 3.7- and 5.0-kb fragments of each recombinant plasmid were gel purified, and then the 3.7-kb fragment of one Nef plasmid was ligated to the 5.0-kb fragment of the other Nef plasmid.

To construct nef gene chimeras at the AflIII and BseRI sites, subclones of Nef1A11 and Nef157 were prepared in the vector pBluescript SK(+) (Stratagene) at the HindIII site. For AflIII, the orientation of the nef gene was T3 promoter-5′Nef3′-T7 promoter. Recombinant plasmids were digested with AflIII, which cuts once in nef and once in the vector. The 760-bp and 3.0-kb fragments were gel purified, and the 760-bp fragment of one Nef subclone was ligated to the 3.0-kb fragment of the other. For BseRI, recombinant plasmids with the orientation T7-5′Nef3′-T3 were digested with BseRI, which cuts once in Nef, and XhoI, which cuts once in pBluescript near the T7 promoter. The 620-bp and 3.2-kb fragments were purified and ligated. The various chimeric fragments were digested with HindIII, blunted, and then cloned in pSC11.

Nef157 genes containing single amino acid substitutions were prepared by using a combination of site-directed mutagenesis and common restriction sites. Nef1A11 cloned in pBluescript in the T7-5′Nef3′-T3 orientation was used as the template for a PCR-based, site-directed mutagenesis procedure (4). To change residue 196 from His to Gln, a primer with the sequence 5′-GCTGGCTGCATTAAATAATG was used (the underlined C indicates the mutagenized nucleotide); this oligonucleotide spans bases 574 to 593 of the 792-nucleotide SIV nef gene sequence. To change residue 211 from Pro to Leu, a primer with the sequence 5′-ATGCTAGAACCTCTCCCCAA was used, corresponding to bases 618 to 637. This mutagenesis procedure uses two flanking primers that recognize vector sequences; in this case, primers annealing to the T3 and T7 promoters were utilized. Amplified fragments were digested with EcoRI and XhoI and inserted into pBluescript SK(+). Cloned fragments were sequenced to confirm the presence of the mutations and to rule out other nucleotide changes. The recombinant plasmids were digested with BseRI and XhoI, and the 3.2-kb fragment corresponding to the 3′ portion of Nef1A11, which contains the mutated region, was ligated to the 620-bp BseRI-XhoI fragment corresponding to the 5′ portion of Nef157 (see the BseRI chimera construction described above). The resulting nef genes are chimeric fragments containing a mutated nucleotide in the Nef1A11 region of the gene and basically encode Nef157 with a single amino acid substitution at residue 196 or 211.

While preparing recombinant vaccinia viruses expressing SIV nef genes, we observed that one recombinant produced atypical plaques. Recombinants were constructed by using nef genes from SIVmac1A11 and SIVmac239-R1. SIVmac1A11 is an isolate that establishes low viral loads after intravenous injection into rhesus macaques (30). SIVmac239-R1 is a virus recovered from a macaque infected with molecularly cloned SIVmac239 (30); the premature stop codon in the SIVmac239 nef gene reverts when the virus is grown in monkeys (13), and therefore, the SIVmac239-R1 nef gene, renamed Nef157, consists of an open nef coding region. The vaccinia virus recombinant expressing Nef1A11 (vNef1A11) produced the usual vaccinia virus plaques when plated on monolayers of TK⁻ 143 cells, a human osteosarcoma cell line deficient in thymidine kinase (Fig. 1B). In contrast, the great majority of plaques produced by the vaccinia virus recombinant expressing Nef157 (vNef157) were atypical (Fig. 1C). Instead of the degeneration normally seen with vaccinia virus infection, the plaques produced consisted of an intact monolayer with a cluster of heaped-up cells in the center of the plaque, similar in appearance to foci of transformed cells. All cells within a plaque were infected with vaccinia virus, as determined by staining with X-Gal, and no cells were viable by the time the altered plaques had formed (1 to 2 days postinfection), as judged by their inability to take up the vital dye neutral red. The formation of altered plaques by vNef157 was dependent upon the cell line used for growth; infection of BSC-1 cells, a monkey kidney cell line, did not result in altered plaques.

FIG. 1.

Formation of altered plaques by vaccinia virus expressing Nef157. Monolayers of TK⁻ 143 cells were infected with vaccinia virus and photographed 2 days after infection. (A) Cells infected with the WR strain of vaccinia virus. (B) Cells infected with recombinant vaccinia virus vNef1A11. (C) Cells infected with recombinant vaccinia virus vNef157.

In most cases, vNef157 formed altered plaques in TK⁻ 143 cells. Occasionally, however, somewhat normal-appearing plaques were formed. These plaques were easily distinguished from the typical vaccinia virus plaque by the presence of large clumps of cells in the center and around the periphery of the plaques. The reason for this occasional difference in plaque morphology is not known.

Immunoblotting experiments demonstrated that both vaccinia virus recombinants express Nef. Therefore, the inability of vNef1A11 to form foci cannot be attributed to the lack of Nef expression. As shown in Fig. 2a, the Nef157 sample contained multiple Nef-specific bands in the 28- to 33-kDa region of the gel, perhaps as a result of degradation. A background band migrated near the Nef-specific immunoreactivity (Fig. 2a, sample VV); this band could not be reduced, despite preadsorption of antibody with extracts prepared from vaccinia virus-infected TK⁻ 143 cells. To confirm the specificity of the results, Nef antibody was blocked with baculovirus-expressed Nef157 prior to immunoblotting; the Nef-specific bands were eliminated by this treatment (Fig. 2b). The different migration of Nef1A11 and Nef157 has been seen previously (30).

FIG. 2.

Expression of SIV Nef by recombinant vaccinia virus. Samples of TK⁻ 143 cells infected with recombinant vaccinia virus were electrophoresed on 12.5% polyacrylamide gels and immunoblotted by being transferred to polyvinylidene difluoride membranes and probed with rabbit antiserum to SIV Nef. Sample VV is a vaccinia virus recombinant expressing SIV Gag-Pol. (a) Immunoblot probed with Nef antibody preadsorbed with extracts from vaccinia virus-infected TK⁻ 143 cells. (b) Immunoblot probed with Nef antibody blocked with baculovirus-expressed Nef157. The arrows point to the background band that comigrated with Nef. Marker protein molecular sizes are shown in kilodaltons.

To map the amino acid responsible for altered plaque formation, we constructed nef gene chimeras with the 1A11 and 157 alleles. Both nef alleles encode proteins of 263 residues, and there are 12 amino acid differences between the two (Fig. 3). By using common restriction sites that separated groups of divergent amino acids, the 3′ ends of the two nef alleles were switched in various constructs. Hybrid fragments were inserted into pSC11, and recombinant vaccinia viruses were generated.

FIG. 3.

Schematic representation of amino acid divergence between Nef1A11 and Nef157. The 792 nucleotides encoding the SIV Nef protein are represented along with the positions of the common restriction sites used to generate nef chimeras. The 12 amino acid residues that differ between Nef157 and Nef1A11 are indicated by asterisks. The identities of the four divergent residues downstream of the NcoI site are given with the Nef157 amino acid first and the Nef1A11 amino acid following the slash.

The results of chimera formation are presented in Table 1. By using the NcoI site, vaccinia virus recombinants containing the 5′ half of Nef1A11 and the 3′ half of Nef157 produced altered plaques. Thus, altered plaque formation mapped downstream of the NcoI site in a region of Nef containing only four divergent amino acids: residues 93, 165, 196, and 211.

TABLE 1.

Structure and properties of nef chimeras and mutants

Chimeric or mutant protein		Altered plaques	Phosphorylationb
Chi1c		−	W
Chi2		+	S
Chi3		−	W
Chi4		+	S
Chi5		−	W
Chi6		+	S
Mut1d		−	W
Mut2e		+	S

^aThe primary structure of the 263-amino-acid Nef protein is shown schematically. Residues 93, 165, 196, and 211 are the four divergent residues located downstream of the NcoI site. 

^bWeak phosphorylation (W) is characteristic of Nef1A11. Strong phosphorylation (S) is characteristic of Nef157. 

^cThe Nef157 portion of a chimera is represented by black bars; the Nef1A11 portion is represented by gray bars. 

^dAmino acid 211 of Nef157 was changed to proline, the Nef1A11 residue. 

^eAmino acid 196 of Nef157 was changed to histidine, the Nef1A11 residue. 

To identify the critical residue, chimeras were constructed at the AflIII and BseRI sites. Altered plaque formation was found to map downstream of the BseRI site, eliminating residues 93 and 165. To determine which of the remaining residues influence plaque formation, a Nef1A11 amino acid was substituted for each of these positions in Nef157 by site-directed mutagenesis. As shown in Table 1, the vaccinia virus recombinant containing Nef157 with a substitution at residue 196 produced altered plaques, while the recombinant expressing Nef157 with a substitution at residue 211 gave typical vaccinia virus plaques. Therefore, a leucine at residue 211 (Nef157) was required for altered plaque formation, and any recombinant containing proline (Nef1A11) at this position was unable to produce altered plaques.

A serine/threonine kinase with properties similar to those of p21-activated kinases has been found to coprecipitate with HIV-1 Nef when cell extracts are treated with Nef antibody (22, 24). The kinase is detected by its ability to autophosphorylate in vitro. To test our chimeric Nef proteins for the ability to interact with kinase, in vitro phosphorylation was determined by using a previously published procedure (24). TK⁻ 143 cells were infected at various multiplicities of infection with recombinant vaccinia virus and left for 1 to 2 days. Harvested cells were washed and then processed as previously described (24). Supernatants from infected cells were incubated overnight with SIV Nef antiserum. Immune products were precipitated with protein A-Sepharose, washed, and then mixed with kinase buffer containing [γ-³²P]ATP. The reactions were stopped by washing, and then sodium dodecyl sulfate sample buffer was added. Controls included samples incubated with normal rabbit serum and Nef antiserum added to samples prepared from cells infected with a vaccinia virus recombinant expressing SIV gag-pol.

Proteins of 62 and 90 kDa were specifically phosphorylated in vNef157 and vNef1A11 samples, but the apparent phosphorylation was much greater in the vNef157 sample (Fig. 4). By using strong versus weak phosphorylation, the remaining Nef recombinants were characterized, and the results are shown in Fig. 5 and summarized in Table 1. Vaccinia virus recombinants characterized by strong phosphorylation are the same ones that form altered plaques. Thus, strong phosphorylation is associated with the leucine at residue 211.

FIG. 4.

Association of Nef and protein kinase in vaccinia virus-infected cells. Immunoprecipitates from TK⁻ 143 cells infected with recombinant vaccinia virus were processed for in vitro phosphorylation. Samples were electrophoresed on 10% polyacrylamide gels, and dried gels were autoradiographed for 1 to 2 days. Sample VV is a vaccinia virus recombinant expressing SIV Gag-Pol. The positions of the 62- and 90-kDa products whose phosphorylation is stimulated by SIV Nef are indicated on the right. Marker molecular sizes are shown in kilodaltons on the left.

FIG. 5.

Stimulation of phosphorylation by Nef recombinants. TK⁻ 143 cells were infected with vaccinia virus recombinants expressing chimeric and mutant forms of SIV Nef, and lysates were processed for in vitro phosphorylation. In the Phos. panel, the phosphorylation results are presented, and the locations of the 62-kDa (lower arrowhead) and 90-kDa (upper arrowhead) phosphorylated products are indicated. The expression of the various Nef proteins in the TK⁻ 143 cell extracts is shown by immunoblotting (Anti-Nef panel), and the specificity of the detection is demonstrated in the bottom panel by immunoblotting with antibody blocked with baculovirus-expressed SIV Nef157 (Blocked Antibody panel). The arrows to the right indicate the position of the background band that comigrated with Nef.

Altered plaque formation by vaccinia virus has been observed previously in a different experimental setting. Vaccinia virus encodes a secreted growth factor that is similar to epidermal growth factor (EGF), and vaccinia virus growing on the A431 cell line, which expresses high levels of the EGF receptor, forms foci. Presumably, focus formation results from the proliferation of uninfected cells in response to the secreted vaccinia virus growth factor (2).

It is theoretically possible that altered plaque formation by vNef157 involves cell proliferation. For example, Nef might increase the number of EGF receptors on the surface of TK⁻ 143 cells, making these cells more sensitive to the viral growth factor. However, Nef is an intracellular protein, unlike the secreted vaccinia virus growth factor, and although potentially released from lysed cells, it is more likely to be delivered to cells only through vaccinia virus infection (2). Because vaccinia virus quickly shuts down host DNA, RNA, and protein synthesis, stimulation of cell growth by Nef appears to be highly unlikely in this case.

The behavior of vNef157 is more reminiscent of that of Shope fibroma virus, a poxvirus that infects rabbits. Shope fibroma virus produces pocks in culture. Experiments have shown that cell division is not required, since pocks can form in the presence of mitotic inhibitors or when plating cells are irradiated prior to infection (11, 29). Cell aggregation has been suggested as a possible mechanism of pock formation.

For cells infected with vNef157, aggregation may also play a part. One widely held view of Nef function is that Nef influences cell activation and signal transduction pathways, perhaps through the stimulation of p21-activated kinases or the binding of Src family tyrosine kinases (7, 14, 17, 22, 23). Stimulation of these cellular pathways could lead to changes at the cell surface which, in turn, could alter the cell’s adhesion properties. In this scenario, Nef’s ability to downregulate CD4 and major histocompatibility complex class I molecules from the surface of T cells would be a specific example of a more general ability to influence the protein composition of the cell membrane.

Chimeric nef genes, consisting of sequences from Nef157 and Nef1A11, and site-directed mutagenesis were used to identify the amino acid responsible for altered plaque formation. We found that residue 211 was the critical position, with leucine (Nef157) instead of proline (Nef1A11) resulting in altered plaques. Two observations are relevant to residue 211. First, the leucine at residue 211 (the corresponding position in HIV-1 Nef is residue 181) is highly conserved in HIV and SIV Nef and is located in one of the four conserved regions of the nef sequence (28). Second, a conservative substitution of alanine for leucine at this position has been shown to reduce the HIV-1 Nef activity associated with virus replication (32). Thus, replacement of the leucine at residue 211 would be expected to have deleterious effects on Nef function, and our results confirm this expectation. Whether the proline substitution in Nef1A11 results in a protein with limited activity, an altered structure, or reduced stability is uncertain. The Western blotting results show that degradation of Nef occurs in vaccinia virus-infected cell samples, and the possibility exists that Nef1A11 is more unstable than Nef157. The comparative stability of these Nef proteins could be determined by pulse-chase analysis, and conformational differences could be determined by structure determination (9, 15).

Coprecipitation of Nef and a serine kinase was first reported for human T-cell lines expressing HIV-1 Nef (24). To detect the kinase, cell extracts were treated with Nef antibody and proteins present in the immunoprecipitates were phosphorylated in vitro. In those cells, labeled products of 62 and 72 kDa were observed. The 62-kDa product was also labeled in other cell types, including monkey and rat cells (25). Recent studies using various kinase inhibitors and immunological reagents have shown that the 62-kDa product represents a new member of the p21-activated kinase family, a group of kinases that are regulated by small, GTP-binding proteins and active in the cell’s response to environmental stimuli (16). The kinase is labeled by autophosphorylation (17, 22). In our experiments, labeled proteins of 62 and 90 kDa were observed following in vitro phosphorylation, and we presume that the 62-kDa product represents the same type of kinase.

Although vNef1A11 does not form altered plaques, Nef1A11 is still able to stimulate the phosphorylation of the 62- and 90-kDa proteins. However, the degree of phosphorylation is much less than that seen with Nef157. The residual phosphorylation could indicate that Nef1A11 retains enough activity to fulfill its role in the virus life cycle. Alternatively, as suggested in a recent report, the interaction with a p21-activated kinase may not be the crucial biological function of Nef (18).

Strong phosphorylation of the 62- and 90-kDa proteins required the leucine at residue 211. Thus, we could not genetically separate altered plaque formation from strong phosphorylation, at least with our two Nef alleles. The exact relationship between phosphorylation and altered plaque formation remains unclear, however, and future studies examining the effects on plaque formation of inhibition of phosphorylation may help to determine the connection between these two processes.

The expression of a simian virus protein by vaccinia virus leading to altered plaque formation on an osteosarcoma cell line is certainly an artificial system. Whether Nef acts directly on the host cell or through a vaccinia virus pathway remains to be determined. If altered plaque formation truly reflects Nef’s association with a p21-activated kinase, the infection of TK⁻ 143 cells with a vaccinia virus expressing Nef could be a useful system for studying the biological consequences of this interaction.