WO1991005860A1 - Non-infectious hiv-1 particles and uses therefor - Google Patents

Abstract

HIV-1 mutant constructs which include a mutation in a nucleotide sequence which is present in wild-type HIV-1, in which it is critical in packaging HIV-1, as well as HIV-1 mutant expression products and HIV mutant viral particles, which are non-infectious. Vaccines which are HIV-1 mutant constructs or HIV-1 mutant expression products, as well as a method if immunizing an individual, comprising administering the vaccine in an appropriate amount.

Description

NON-INFECTIOUS HIV-1 PARTICLES AND USES THE REFOR Description

Background

Human immunodeficiency virus type 1 (HIV-1) is the causative agent of acquired immune deficiency syndrome (AIDS), which is characterized by immune suppression resulting from selective infection of helper cells and death of 0KT4+ T helper cells.

Sarin, P.S. Ann. Rev. Pharmacol., 28 :411-428 (1988). Clinical manifestations of the disease include severe immune deficiency, which is generally accompanied by malignancies and opportunistic infections.

There is a particular interest in finding ways to prevent infection by HIV and to counter the effects of the virus in an already infected individual because of the devastating effects of the virus and the fact that the mortality rate among such individuals is very high.

Despite the fact that much effort, time and money have been expended in developing a means of preventing HI" infection or of reducing or eliminating the effects of the virus once infection has occurred, only limited progress has been made to date in doing so. It has been suggested that various points in HIV infection of T cells and virus replication in infected cells should be targeted in developing chemotherapeutic agents useful in preventing or treating the disease. Such points in the HIV life cycle as attachment to T cells, reverse transcriptase activity, DNA transcription and/or translation and assembly and release of virus particles might be effectively targeted. Several drugs, including those, such as D -penicillamine (DPA), which have been used in the past for treatment of other diseases, and those, such as anti- sense oligonucleotide inhibitors, designed specifically to interfere with HIV, have been assessed for their effectiveness in preventing HIV infection of cells and/or inhibiting the effects of the virus in infected cells. Sarin, P., Ann. Rev. Pharmacol., 28:411-428 (1988); Norman, C., Science, 30:1355-1358 (1985).

Although these efforts are., in some instances, producing promising results, it is clear that at the present time, there is no effective means for interfering with HIV activity in infected cells. Summary of the Invention

The present invention relates to HIV-1 mutant constructs which include a mutation in a nucleotide sequence critical in packaging HIV-1 and which, when expressed in susceptible mammalian cells, produce non-infectious viral particles, as well as to HIV-1 mutant expression products. As described herein, HIV-1 particles which do not contain the viral genome have been made. Such mutant HIV-1 particles provide a means to produce vaccines which are based on intact, fully immunogenic, but non-infectious virus particles. Such vaccines can be used to induce an anti-HIV-1 response in an individual, either prior to or after infection with HIV-1, resulting in enhanced resistance by the individual to the virus. The invention further relates to vaccines which contain a HIV-1 mutant expression product and to methods of immunizing an individual by administration of the vaccine.

In particular, the present invention relates to HIV-1 packaging mutations which differ from wild- type HIV-1 in that they include at least one alteration (a deletion, insertion or substitution) in the nucleotide sequence of the corresponding region of the wild-type DNA critical for viral packaging or in the amino acid sequence of the corresponding region of the wild-type product critical for viral

packaging. Mutations in HIV-1 gag sequences, such as in p7^gag (alternatively referred to as p15^{ga g} ) , and in sequences between the first HIV-1 splice donor site and the gag initiation codon, which are sequences homologous to the defined ψ sites of other viruses, have been shown to result in blocking of packaging of viral RNA and in production of

non-infectious virus. An insertion mutation in p15^gag which resulted in a change in the intervening sequence between the two Cys-His boxes in p15^gag has been made and the mutated HIV-1 construct shown to specifically inhibit packaging of viral genomic RNA into infectious particles. Substitution mutations in one or both of the metal binding motifs of p7^{ga g} have also been made by substituting a tyrosine residue for each of the two cysteine residues present at the corresponding locations in wild-type HIV-1 p7^gag and have also been shown to produce non- infectious viral particles which have the same protein content as that of wild-type HIV-1. Deletion mutations have been made in sequences in the

HIV-1 genome now identified as the HIV-1 ψ site and, similarly, have been shown to produce non- infectious viral particles whose protein content is the same as that of the wild- type virus. Brief Description of the Deawings

Figure 1 is a partial HIV-1 nucleotide sequence (nucleotides 1351-1980) and the deduced amino acid sequence for that partial sequence, (Ratner, L. et al., Nature, 313:277-284 (1985)), on which the location of the mutations in HIV-1 gag described herein are indicated.

Figure 2 is a schematic representation of HIV-1 mutations. Figure 2A: location and size of deletions affecting the HIV-1 ψ site. Figure 2B: amino acid changes in the metal binding motifs of HIV-1 gag produced by various point mutations. The guanine nucleotide in the second position of codons coding for cysteine (C) residues was changed to an adenosine residue, producing codons coding for tyrosine (Y). Figure 3 shows results of assessment of transcription and translation of HIV-1 mutants in cos-1 cells. Figure 3A shows results of Northern Blot analysis showing a comparison of viral transcripts in RNA extracted from cos-1 cells transfected with wild-type or mutant HIV-1 genomes. Figure 3B shows the results of SDS-PAGE analyses of HIV-1 specific proteins in transfected cos-1 cells.

Figure 4 shows results of analysis of HIV-1 mutant particles by Western Blot.

Figure 5 shows results of assessment of the nucleic acid content of viral particles present in the supernatant from transfected Cos-1 cells.

Figure 6A is a schematic representation of the replication defective HIV-1 virus pΔ PAC-Hygro, in which there are the following alterations from wild type HIV-1: a deletion in the ψ site; alteration of the guanine nucleotide in the second position of codons coding for cysteine (C) residues to an adenosine residue, producing codons coding for tyrosine (Y); and replacement the nef gene with the hygromycine resistance gene.

Figure 7 is a schematic representation of the gag-coding region of HIV-1 with nucleotide numbers indicating the initiation codon, the cleavage sites between p17, p24 and p15, and the gag termination codon. The nucleotide differences between wild type and bCA20 are indicated. Below is shown the amino acid sequence of the two HIV-1 Cys-His boxes, and of the intervening sequence, where the bCA20 mutation was introduced. Figure 8 shows the result of immunoblot

analysis of cytoplasmic extracts from HT4(ΔE-dhfr) cells.

Figure 9 shows results of study of viral RNA hybridization by slot-blot analysis of the supernatant from HT4(ΔE-dhfr) cells.

Figure 10 shows results of examination of

HT4(R7-dhfr), HT4 (WT-ΔE-dhfr) and HT4 (bCA20-ΔE-dhfr) by electron microscopy. Panel A: HT4(R7-dhfr); Panel B: HT4(WT-ΔE-dhfr); Panel C: HT4(bCA20- ΔE-dhfr). Detailed Description of the Invention

The present invention is based on results of an investigation of the RNA and the protein sequences involved in HIV-1 packaging. Assessment of the effect of various point deletion or insertion mutations has shown the critical nature of the HIV-1 gag region in packaging genomic RNA into the virus particle and has demonstrated the occurrence and location of a HIV-1 RNA packaging (ψ) site, which is also critical to packaging of the virus.

Specifically, point mutations that alter cysteine residues in one or both metal binding motifs of p7, a cleavage product of the gag precursor; insertion mutations in pl5 which change the length and the nature of the intervening sequences between the two

Cys-His p15^gag boxes; and deletion mutations between the first HIV-1 splice donor site and the gag initiation codon have been made and their effects on viral packaging have been demonstrated. Based on these results, mutations in HIV-1 RNA sequences now shown to be essential for packaging of the virus have been made and, for both types of HIV-1 mutants, HIV-1 RNA packaging has been shown to be inefficient. In particular, two types of HIV-1 mutants have been made and shown to be defective in their ability to package the viral RNA: 1) HIV-1 mutants in which there are alterations or mutations in gag coding sequences and 2) HIV-1 mutants in which there are alterations or mutations in

sequences in the HIV-1 genome which are homologous to the sequences of the defined ψ sites of other viruses. The two types of HIV-1 mutants produced are referred to, respectively, as HIV-1 gag mutants and HIV-1 ψ mutants. In one type of HIV-1 gag mutant, mutations in the form of substitutions of selected nucleotides which in wild-type HIV-1 encode cysteine residues, to result in a nucleotide sequence which encodes tyrosine in the corresponding location, have been introduced into the nucleotide sequences that encode the HIV-1 gag metal binding motifs. This type of HIV-1 gag mutant is referred to as a HIV-1 gag metal binding motif mutant. In a second type of HIV-1 gag mutant (referred to as a HIV-1 gag insertion mutant), a nucleotide insertion which results in alteration in the nature and the length of the intervening sequence between the two HIV-1 gag Cys-His boxes has been made. Three deposits have been made 10/13/89 and 10/16/89 at the American Type Culture Collection, Rockville, MD, in support of the subject application: a HIV-1 gag metal binding motif mutant, designated pA14-15HXB (ATCC Accession #68123); an HIV-1 gag insertion mutant designated Plasmid bCA20-dhfr (ATCC Accession #40682); and a HIV-1 ψ mutant, designated pA3HXB (ATCC Accession #68122). These deposits have been made under the terms of the Budapest Treaty and, upon grant of a U.S. patent, all restrictions on their availability will be irrevocably removed.

Both the gag and the ψ types of mutations were introduced into appropriate cells, in which they were expressed. The mutations introduced in gag coding sequences as described below, were shown to block packaging of the viral RNA. The viral

products resulting from expression of the HIV-1 gag metal binding motif mutants or the ψ mutants have been shown to be normal in protein .content (i.e., not substantially different in protein content from that of the wild type HIV-1), to be missing genomic RNA and to be noninfectious. The alteration in the HIV-1 gag insertion mutant has been shown to be lethal for viral replication and the viral particles generated have been shown to be noninfectious. Such viral products, which are essentially the same in protein content as wild type HIV-1, but are not infectious, can be used as an anti-HIV-1 vaccine.

The following is a description of: an assessment of HIV-1 packaging sequences, of the HIV-1 gag metal binding motif and of the HIV-1 Cys-His boxes and surrounding sequences; construction of HIV-1 mutants, referred to as HIV-1 mutant constructs, whose expression in host cells results in production of noninfectious HIV-1 mutants (referred to as mutant HIV-1 particles); and the use of the HIV-1 mutant particles or HIV-1 mutant expression

products, or a composition containing one or more mutant particles or mutant expression products, as a vaccine. Nucleotide locations referred to are those indicated by Ratner and co-workers; the relevant portion of the HIV-1 genome is represented in Figure 1. Ratner, L. et al., Nature 313:277-284 (1985). Assessment of HIV-1 Packaging Sequences and

Construction of HIV-1 Mutant Constructs with

Packaging Signal Deletions

Packaging, the process by which viral proteins and RNA get together to form an infectious particle, is an essential step in the retrovirus life cycle. Unspliced genomic RNA is an indispensable component of the viral particle and the retrovirus packaging process can discriminate between unspliced viral RNA and spliced viral or cellular mRNAs. This discrimination is very likely the result of interaction between specific viral RNA sequences and viral proteins. Varmus, H., Science 216: 812-820 (1982); and Weiss et al. , RNA Tumor Virus, Cold Spring

Harbor Laboratories, Cold Spring Harbor, New York (1984). For avian and murine retroviruses, RNA sequences essential for packaging have been mapped to a site near the 5' end of the viral genome.

Shank, P.R. and M. Linial, J. Virol. 36:450-456 (1980); Mann, R. and D. Baltimore, J.Viol. 54:401- 407 (1985); Linial, M. et al., Cell, 15:1371-1381 (1978); Koyamat, Harada F. and S. Kawai, J. Virol., 51:154-162 (1984); Watanabe, S. and H.M. Temin,

Proc. Natl. Acad. Sci. USA, 79:5980-5990 (1982); and Mann, R. and D. Baltimore, J.Virol., 54:401-407 (1985)). This site occurs within intervening sequences which are removed during splicing, producing an unpackageable RNA.

Using the defined psi sites of murine leukemia, spleen necrosis and avian sarcoma viruses as a guide, deletion mutations were constructed in homologous sequences in the HIV-1 genome to investigate whether this region acts as a packaging signal for the human virus. In particular, using oligome- diated mutagenesis, two deletion mutations were constructed: one in which nucleotides 293-331, inclusive, were deleted, to create the mutant pA3-HXB and one in which nucleotides 293-313, inclusive, were deleted, to produce pA4-HXB (Figure 2A). These deletions occur between the first splice donor site and the gag initiation codon, leaving the consensus sequence for the first splice donor site intact. Other deletions in the HIV-1 ψ site can also be made. Those shown (e.g., by the methods described herein) to be packaging defective mutants can be used, for examples as an anti-HIV vaccine, as is described for the pA3-HXB and pA4-HXB mutants. Construction of HIV-1 gag Mutants

The Metal Binding Motif of HIV-1 and Construction of

HIV-1 Mutant Constructs Containing Alterations in Viral RNA Recognition Sites

One or more proteins encoded by sequences at the 3' end of the gag gene are thought to be involved in the recognition of viral RNA, as deletion mutations at this locus appear to produce particles lacking viral RNA. Henderson, L.E. et al., J. Biol. Chem., 256:8400-8406 (1981); Oroszlan, S. and T.D.

Copeland, Curr. Top. Microbiol. Immunol., 115:221- 233 ( 1985 ) ; and Covey , S . M . Nucleic Acid Res. ,

14:623-633 (1986). A metal binding motif of the form CysX₂CysX₃GlyHisX₄ Cys is a common feature of the carboxyl terminus of the gag precursor in all retroviruses. Berg, J., Science 232:485-486 (1986). This motif occurs once in the murine retroviruses, and twice in most other retroviruses studied thus far.

As nucleic acid binding properties have been attributed to proteins containing zinc finger motifs (Evans, R.M. and S.M. Hollenberg, Cell 52:1 (1986)), a direct role for the metal binding domain in RNA binding has been investigated for the Rous sarcoma (RSV) and the Moloney leukemia (MLV) viruses.

Karpel, R.L. et al., J. Biol. Chem. 262 :4961-4967 (1987). In RSV, deletion of the first of the two zinc finger motifs abolishes RNA packaging and infectivity. Deletion of the second motif reduces viral infectivity approximately 100-fold. Both zinc fingers are important for 70S RNA dimer formation (Meric, C. and P.-F. Spahr, J. Virol. 60:450-459

(1986); and Meric, C. et al., J. Virol. 62:3328-3333 (1988)). Similar results have been observed for MLV, where the lack of packaging of genomic RNA seems to correlate with nonspecific packaging of spliced viral RNA and cellular messengers. Meric, C. and S.P. Goff, J. Virol., 63:1558-1568 (1989); Prats, A C. et al., J. EMBO, 7:1777-1783 (1988); and Gorelick, R.J. et al., Proc. Natl. Acad . Sci. USA , 85:8420-8424 (1988)). One other function has been attributed to the gag zinc fingers; they are necessary for the correct positioning of the tRNA primer on the replication initiation site at the 5' end of the RNA genome. Prats, A.C. et al., J .EMBO 7:1777- 1783 (1988).

To investigate the role of the gag zinc fingers in virus assembly, oligo-mediated mutagenesis was employed to substitute tyrosines for the first two cysteines of the second metal binding motif, producing the mutant pA15-HXB (Figure 2B and Example 1). A second mutant, pAl4-15-HXB, in which both of the metal binding motifs were altered, was constructed by further mutating pA15-HXB to substitute tyrosines for the first two cysteines of the first metal binding motif. In a similar manner, substitution of an amino acid other than tyrosine for some or all of these cysteine residues can be made, using the procedures described herein.

The gag gene of retroviruses encodes proteins that are critical to viral assembly and release, stabilization of the virion, uncoating of the viral RNA, initiation of reverse transcription and perhaps integration. Bolognesi, D.P., et al., Science

199:183-186 (1978); Crawford, S. and S.P Goff, J. Virol, 49:909-917 (1984); Hsu, H.W., et al.,

Vϊrology 142:211-214 (1985); Meric, C. and P.-F Spahr, J. Virol. 60:450-459 (1986); Meric, C. and S. Goff, J. Virol. 63:1558-1568 (1989); Prats, A.C, et al., EMBO J. 7:1777-1783 (1988); Schultz, A.M. and A. Rein, J. Virol. 63:2370-2373 (1989); Schwartzenberg, P., et al., J. Virol. 49:918-924 (1984) Like other retroviruses, HIV-1 expresses its gag gene as a polypeptide precursor, Pr55^gag . This precursor is subsequently phosphorylated and

myristoylated, presumably self-assembles under the plasma membrane, and is then processed by the virally-encoded protease to yield the mature viral proteins, p17, p24 and p15; the latter is believed to be cleaved further into p9 and p7 polypeptide species. Mervis, R.J., et al., J. V i ro l.

62:3993-4002 (1988); Veronese, F.d.M . , e t al . , J. Virol. 62 : 795 - 801 ( 1988 ) p 17 app ears to be located at the inner leaflet of the lipid membrane, p24 probably constitutes the shell of the tubular core of HIV, and the exceedingly basic character of pl5 suggests that it might be associated with the viral RNA. Gelderblom, H.R., et al., Virology 156:171-176 (1987). As described herein, it has now been shown that a mutation in the p15^{ga g} coding sequence specifically blocks the packaging of the viral RNA.

HIV-1 p15^{ga g} is a 123 amino-acid long protein, encoded by the 3' end of the gag gene. It carries striking similarities with retroviral nucleocapsid proteins (NC). These similarities include, in the p9 region, two tandem copies, separated by 7 amino acids, of a conserved motif of three cysteine residues and one histidine residue, referred to as the Cys-His box (see Figures 1 and 7). Covy, S.N., Nucleic Acids Res.14:137-145) In this motif, if the N terminus-proximal cysteine is designated n, the two other cysteine residues are located at n+3 and n+13, whereas the histidine residue is at n+8. The Cys-His boxes of the NC of retroviruses have been hypothesized to form metal-binding domains (Berg, J., Science 232:485-487), by analogy with the " zinc-finger" protein sequences implicated in recognition of specific DNA sequences by a wide variety of eukaryotic transcription factors.

Recently, it was demonstrated that the avian

myeloblastosis virus (AMV) p12^gag binds ^{6 5}Zn(II) in a zinc blotting technique (Schiff, L.A., et al., Proc. Natl. Acad. Sci. USA 83:7246-7250 (1988).

However, no significant amount of metal has been detected in virions, and it is uncertain whether bound Zn ions are important for NC function.

Studies in various oncornaviruses have shown that NC is tightly associated with the genomic RNA and that purified NC binds nucleic acids in vitro, although in a non-specific manner (Davis, J.M., et al., J. Virol. 18:709-718 (1976); Nissen-Meyer, J. and A.K. Abraham, J. Mol. Biol. 142:19-28 (1980)). Also, in both Moloney leukemia virus (M-MuLV) and Rous sarcoma virus (RSV), mutants containing a lesion in the NC region of the gag gene have been shown to present a major defect in RNA packaging (Gorelick, R.J., et al., Proc. Natl. Acad. Sci. USA

85:8420-8424 (1988); Meric, C. and P.-F Spahr, J. Virol. 60:450-459 (1986); Meric, C. and S. Goff, J. Virol. 63:1558-1568 (1989)).

To investigate the role of HIV-1 p15^gag, a mutation was introduced into wild-type HIV proviral DNA, as described in Example 3. Briefly, the plasmid W13 , which contains an infectious copy of the HIV-HXB2-D proviral DNA, was modified by the insertion of an 8 nucleotide-long Cla I linker. The linker was inserted in a unique Apa I site present at position 2009 and the Cla I site was then blunted with Klenow in order to rectify the gag reading frame. As a result, the two residues which follow the first Cys-His box (arginine- alanine) in

wild-type HIV-1 are replaced by four amino acids: serine-isoleucine-alanine-methionine (Figure 7). The insertion thus results in alteration of the nature and the length of the sequence between the two Cys-His boxes, as shown in Figure 7.

Analysis of HIV-1 gag Metal Binding Motif Mutants and HIV-1 ψ Mutants in Cos-1 Cells

Expression of viral RNAs and proteins

Transfection of the African Green Monkey kidney cell line cos-1 with a plasmid carrying biologically active HIV-1 DNA and an SV40 origin of replication results in the recovery of infectious HIV-1 virus. Viral transcription, translation, assembly and budding occur normally in these cells and the recovered virus is capable of infecting susceptible cells. Feinberg, M.B., et al., Cell 46:807-817 (1986)). To produce virus from the wild-type and mutated molecular clones of HIV-1, as described above, cos-1 cells were transfected with the wild- type pHXB2gpt plasmid and one of the four mutated plasmids. These four mutated plasmids, as represented in Figure 2, are: pA3-HXB, in which nucleotides 293 to 331, as they occur in wild type HIV-1 are deleted; pA4-HXB, in which nucleotides 293-313, as they occur in wild type HIV-1, are deleted; pA15-HXB, in which the first two cysteines of the second metal binding motif of p7 are replaced by tyrosines; and pA14-15HXB, in which the first two cysteines of both of the metal binding motifs of p7 are replaced by tyrosines. An additional mutant, designated pΔPAC-Hygro (see Figure 6) has also been constructed. pΔPAC-Hygro includes the 39 bp

deletion present in pA3HXB and the alterations of cysteines in both zinc finger motif present in pA14-15HXB.

Forty-eight hours after transfection with one of the four mutated plasmids, the cos-1 cells were examined for evidence of viral gene expression. The SV40 - transformed African Green Monkey kidney cell line cos-1 (Gluzman, Y., Cell, 23:175-182 (1981)) was obtained from G. Khoury (NIH) and maintained in DME media supplemented with 10% fetal bovine serum. Northern Blot analysis comparing viral transcripts in RNA extracted from cos-1 cells transfected with wild-type or mutant HIV-1 genomes was carried out, using a HIV-1 specific probe which was a ³²P random primed full length viral DNA. 15/xg of total celluar

RNA was loaded in each lane. Results show (Figure

3A) that the patterns of RNA transcripts in cells transfected with the mutant plasmids are indistinguishable from that obtained from cells transfected with the wild-type plasmid. In each case, all three classes of HIV-1 mRNA are present: the 9.2 genomic mRNA, a 4.3 kd spliced mRNA encoding the env and the vif genes, and the heterogeneous

collection of messages in the range of 2 Kb which includes tat-III, rev, and nef mRNA. Thus, these HIV-1 mutations do not appear to affect the

expression of viral RNA.

The expression of HIV-1 structural proteins was investigated by metabolic labeling of transfected cos-1 cells with ³⁵ S methionine, followed by immuno- precipitation of viral proteins using human and monoclonal mouse sera with defined specificity of

HIV antigens. Veronese et al., Proc. Natl.

Acad. Sci. USA 82:5199-5202 (1985); Veronese, F.D., et al., AIDS Research and Human Retroviruges

3:253-264 (1987). SDS-PAGE analyses were performed using a 10% or a 15% polyacrylamide gel. The plasmid pHXB2BAMp3 was included as a negative control because it does not produce virus, due to a pos t-transcriptional defect. Results are shown in Figure 3B. This analysis revealed that all of the major structural proteins of the virus are present in cells transfected with either the HIV-1 wild-type or a mutant plasmid. The presence of gpl60, gp120, gp41, p24, P17, and P15 in all of these transfected cells indicates that the HIV-1 mutants do not produce major alterations in the synthesis and processing of gag and envelope precursors.

The relative expression of one of the regula- tory proteins, tat, was also investigated for these mutants. Cotransfection of wild- type or mutant HIV-1 plasmids with the plasmid PC15, in which a CAT gene is under the control of the HIV-1 LTR (Fein- berg, M.B., et al., 46:807-817 (1986)), indicated that each of the mutants produce normal levels of tat. Thus, results show that the four HIV-1 mutants do not exhibit substantial alterations in the expression of viral RNAs and proteins. Bological activity of viral mutants

To assess the biological activity of the HIV-1 particles produced by transfection of cos-1 cells with pHXB2gpt (wild type) and the four mutated plasmids, pA3HXB, pA4HXB, pA15HXB, and pA14-15HXB2, culture supernatants were harvested, filtered, and examined for the presence of reverse transcriptase activity and p24. All of the supernatants contained substantial, although varying, levels of reverse transcriptase (RT) and p24. Supernatants adjusted to contain comparable levels of p24 (15ng) were used to infect a H9 T leukemia cell line, which is susceptible to infection and permissive for HIV-1 replication. In H9 cells infected with supernatant from cos-1 cells transfected with pHXB2gpt (wild type HIV-1), both the percentage of HIV-1 antigen positive cells and the reverse transcriptase activity increased with time (Table 1). In contrast, H9 cultures exposed to supernatants derived from cos-1 cells transfected with mutant viruses were not positive in cellular p24 immunofluorescence, RT activity or particle-associated p24 assays. These assays were negative for at least 30 days after the H9 cells were exposed to mutant viral particles. These data indicate that the viral particles produced by all four HIV-1 mutants in cos-1 cells lack infectivity when transferred to a susceptible cell line.

Analys is of Viral Particles

The protein and the nucleic acid content of mutant viral particles from cos-1 cell transfections were studied to determine whether the lack of productive infection of H9 cells could be due to defective virions. Three different assays indicated that the amounts of HIV-1 protein present in supernatants of cos-1 cells transfected with the mutant HIV-1 plasmids were similar to that obtained with the wild-type plasmid. Viral particles in the cos-1 supernatants were pelleted by centrifugation and a quantitative analysis of the amounts of p24 present in the pellet was performed using a p24 ELISA. p24 values ranged from 15 to 20ng in each sample (Table 2).

A reverse transcriptase assay performed on the pelleted viral particles revealed a two-fold

reduction in mutant levels relative to the wild-type (Table 2). Finally, Western blot analysis showed that the wild-type and each of the mutants produced particles containing comparable amounts of gag and env proteins (Figure 4).

To determine whether the transfection frequency was similar among the wild-type and mutant transfections of the cos-1 cells, an immunofluorescense assay was performed on fixed cos-1 cells 48 hours after transfection, using an anti-p24 specific mouse monoclonal antibody. Veronese, F.D., et al., Proc. Natl. Acad. Sci. USA, 82:5199-5202 (1985). The fraction of cells producing positive signals was nearly identical among the different transfections, ranging between 17 and 20%. Together, these data indicated that the number and protein composition of viral particles produced by the HIV-1 mutant constructs were similar to that of particles produced by the wild-type.

The amount of viral RNA present in these virions was investigated by extracting RNA directly from virions present in the supernatant, immobilizing the nucleic acid on nitrocellulose slot blots and probing with a labelled Clal/EcoRI 3.8 gag-pol fragment. Results are shown in Figure 5. At least a 100-fold reduction in the RNA content was observed for both ψ and gag metal binding motif mutants, relative to the wild-type. Thus, for both ψ and gag metal binding motif mutants, HIV-1 RNA is not packaged efficiently in the viral particles. The lack of genomic RNA in mutant viral particles reflects a packaging defect, rather than the absence of intracellular genomic RNA, as evidenced by the fact that these transcripts are present in cells transfected with mutant genomes (Figure 3A).

Wild type and mutant HIV-1 particle morphology was examined by electron microscopy. Careful scoring of the sections indicated that the majority of the mutant particles are less electron dense than are wild- type viral particles. The appearance of empty mutant virions is consistent with the RNA and protein analysis, which indicate that the mutant virions lack genomic RNA. The Role of HIV-1 ψ and Metal Binding Motifs in Viral Packaging

These data indicate that the HIV-1 ψ site is located in the 5' leader region of the genomic viral RNA and that deletions located immediately 3' to the splice donor site as small as 21 base pairs can produce a defect in packaging. The results described here indicate that HIV-1 ψ site mutants exhibit normal patterns of gene expression in transfected cells and that the viral particles produced by these mutants, while lacking detectable RNA, appear normal in protein composition.

HIV-1 mutant viruses with lesions in the gag metal binding motifs were shown to produce non- infectious viral particles that are similar, if not identical, to those produced by ψ de le tion mutants . The fact that HIV - 1 p7 and ψ mutants both produce viral particles lacking genomic RNA is consistent with the idea that HIV-1 packaging involves an interaction between p7 and genomic RNA. If p7 interacts with genomic RNA, the metal binding motifs could play a direct role in binding. It is also possible that the gag metal binding motifs interact with other nucleic acids (e.g., the tRNA primer) or with other proteins tht are important for proper RNA packaging.

It is clear from the analysis of the packaging mutants that assembly of viral capsids occurs in the absence of RNA packaging. A particle whose protein composition is similar to that of wild type virions is assembled even when the genomic RNA cannot be packaged. Mutations in the metal binding motifs of p7, a protein that is probably directly involved in RNA packaging, do not affect capsid assembly, indicating that themotifs of p7 does not play a role in assembly.

The HIV-1 packaging mutations described can provide a means to produce noninfectious antigenic particles useful for induction of anti-HIV-1 immune responses. A cell line with a stably integrated mutant provirus has been constructed to provide a source of such noninfectious particles. Potential reversion of point mutations in the provirus should be minimized by the simultaneous presence of the ψ site deletion and the four gag cysteine mutations present in A14-15HXB. The possibility that a mutant viral genome might occassionally be packaged cannot be eliminated, but fortuitous packaging of an HIV-1 genome may not lead to a productive infection, both because of the packaging defect and because of the role of the gag metal binding motifs in correctly positioning the tRNA primer necessary for reverse transcription. If necessary, the HIV-1 genome containing packaging mutations can be divided into 2 independent constructs stably integrated in a cell line, and any rescued genome will be highly

defective. Analysis of HIV-1 p15^gag Mutants

Assessment of biological activity of viral mutants The phenotypic consequences of the p15^{ga g} mutation (Figure 7) were assessed, as described in

Example 3, by transfecting cos cells with bCA20 (the mutated construct) and scoring generation of viral particles by measuring p24 antigen produced and reverse transcriptase activity released in the supernatant. As shown in Table 3 (Example 3), bCA20- induced p24 and reverse transcriptase activity were approximately 40% and 30% of the corresponding wild-type activity, respectively. Thus, the

mutation in bCA20 only mildly interfered with release of viral particles.

The cos cell supernatant was used to infect H9 cells. An indirect immunofluorescence assay, using serum from an HIV-infected individual as detection antibody, was carried out to monitor the H9 cells. After three weeks, no positive cells were detected, demonstrating that the viral particles generated following transfection were noninfectious.

Further assessment of the effects of the p15^{ga g} mutation contained in bCΔ20 was carried out in a cell line (designated HT4 (bCA20-ΔE-dhfr)) which constitutively expresses an Env version of the mutation. Construction of the cell line is

described in Example 3. Cells selected on the basis of appropriate resistance were cloned and analyzed by polymerase chain reaction for the presence of the viral integrant. Analyses, described in Example 3, demonstrated that the mutation present in bCA20 did not affect the synthesis, the cleavage or the stability of the gag precursor.

The effect of the mutation contained in bCA20 on packaging of viral RNA was also assessed, by slot-blot analysis of the supernatant from

HT4(bCA20-ΔE-dhfr) cells and HT4(WT-ΔE-dhfr) cells, which express a wild-type gag sequence (the control cell line). Results showed that the amount of viral RNA in the supernatant from HT4 (bCA20-ΔE-dhfr) was dramatically reduced, in comparison with the amount present in control cell line supernatant.

Examination of the two cell lines by electron microscopy was also carried out in order to

determine if the defect in viral RNA packaging correlated with morphological differences between them. Comparison of the morphology of

HT4(WT-ΔE-dhfr) with the morphology of HT4(R7-dhfr), a cell line infected with an Env⁺ , replication competent version of the same virus showed that the two were very similar (See Figure 10, Panel A and Figure 10, Panel B and Example 3). However, two dramatic differences were seen when particles released from HT4(bCA20-ΔE-dhfr) were compared morphologically with those from HT4(R7-dhfr):

particle diameter was approximately 50% larger for mutant-containing particles than for controls and the electron-dense region in the former was tightly apposed to the membrane, but the center of the particles was strikingly electron- luscent (Figure 10, Panel C). Uses of the Mutant Constructs and Mutant Expression Products of the Present Invention

Thus, HIV-1 mutant constructs which contain mutations in nucleotides which are present in wild-t/pe HIV-1 gag sequences have been constructed, as have mutant constructs which contain mutations in

HIV-1 RNA shown to be essential for viral packaging (HIV-1 ψ site). Further, HIV-1 mutant constructs which contain both a mutation in HIV-1 gag sequences and a mutation in HIV-1 RNA shown to be essential for viral packaging have been made. The viral particles produced by cells into which the mutant constructs are introduced can be used to provide protection to an individual against HIV-1 infection, as can mutant expression products (e.g., mutant gag products). For example, a mutant expression product, such as a gag mutant, can be introduced into an individual, resulting in an immune response and production of anti-mutant HIV-1 gag antibodies.

Alternatively, mutant HIV-1 viral particles produced in cells into which a mutant has been introduced (e.g., by transfection) can be used as a vaccine. In this case, viral particles which are produced using one (or more) of the mutations are obtained from the cells in which they are produced, combined as needed with additional components (e.g., an appropriate physiological buffer or carrier, an adjuvant) to produce a vaccine composition and administered (e.g., by injection) to an individual, either prior to or after HIV infection. In the immunized individual, the mutant HIV-1 particles, whose protein content is much like that of wild-type HIV-1, induce an immune response similar to that induced by the wild- type virus but, unlike the wild-type virus, do not infect the recipient because they are unable to do so.

The present invention will now be illustrated by the following examples, which are not intended to be limiting in any way. EXAMPLE 1 Site Directed Mutagenesis and Plasmid Constrution

The parent DNA used for these studies is the biologically active clone pHXB2gtp described by Fisher and co-workers. Fisher, A.G. et al., Nature, (London) 316:262-265 (1985). The Sacl-Sall (bp 39 to 5366) (Ratner, L. et al., Nature 313:277-284 (1985)) 5.3 kilobase pair (kb) fragment was cloned in M13mp18 and oligo-mediated site-directed mutagenesis was performed on the single stranded DNA according to the protocol of Kunkel et al . Kunkel T.A. et al., Methods in Enzymology, 154:367-382 (1987). The following oligomers were used for the mutagenesis:

A3: TGACGCTCTCGCACCCATCTCTCACCAGTCGCCGCCCCTC, deleting nucleotides 293 to 331 (Ratner, L. et al., Nature 313:277-284 (1985));

A4 : CTCTCTCCTTCTAGCCTCCGCTCACCAGTCGCCGCCCCTC , deleting nucleotides 293 to 313;

A14: TGCCCTTCTTTGCCATAATTGAAATACTTAACAATCTTTC, changing guanosine (G) 1508 and 1517 to adenines

(A);

A15 : TGTCCTTCCTTTCGATATTTCCAATAGCCCTTTTTCCTAG, changing G at position 1571 and 1580 to A. The introduction of the mutations was verified by sequencing different M13mpl8 recombinant DNAs using two specific oligomers A5 : CCATCGATCTAATTCTC and A16: GGCCAGATCTTCCCTAA. Ausubel, F.M. et al.,

Current Protocol in Molecular Biology, Green Publishing Associates and Wiley Interscience (1987).

A Bsshll/Ball 1.9 kb DNA fragment from the mutated M13mpl8 recombinant clones was used to transfer all the different mutations to the wild- type plasmid pHXB2gpt. The full length HIV1 mutant clones were designated pA3HXB , pA4HXB , pAl5HXB and pAl4-15HXB, according to the oligonucleotides used in the mutagenesis. The entire region that had been transferred from the M13mpl8 plasmids was sequenced using double stranded DNA sequencing methods

(Ausubel, F.M. e t al. , Current Protocols in Molecular Biology (Green Publishing and Wiley Interscience 1987)) to confirm the presence of the desired mutations and the absence of any other alterations. All DNA manipulations were according to standard procedures.

EXAMPLE 2 Transfection of Eukaryotic Cells with

Wild Type HIV-1 and HIV-1 Mutant

Constructs

Cos-1 cells were seeded at a density of

10⁶/100mm plate, 24 hours prior to transfection, in DME medium supplemented with 10% fetal calf serum and incubated at 37ºC in 5% CO₂. Transfection was carried out using 10 μg of plasmid DNA/plate as described by Chen and Okayama. Chen, C. and H.

Okoyama, Mol. Cell. Biol., 7:2745-2752 (1987). When viral particles were destined for RNA analysis, transfections in cos-1 cells were performed according to the method of Seldon et al. (Seldon, R.F. et al., Mol. Cell. Biol., 6:3173-3179 (1986)) to avoid the heavy DNA contamination of samples that was observed when calcium phosphate transformation was performed. RNA was extracted from supernatants containing equal amounts of p24 and was carried out in the presence of equal amounts of added tRNA , which was used to monitor the final recovery of RNA. RNA samples were resuspended in water at identical concentrations of tRNA (1 μg/ml) and 1, 0.3 and 0.1 equivalents of RNA were loaded on nitrocellulose, where 1 equivalent represents the amount of RNA obtained from cos-1 supernatants containing 18 ng of p24. RNA Slot Blot analysis was performed as previously described. Ausubel, F.M. et a l. , CurrentProtocols in Molecular Biology (Green Publishing and

Wiley Interscience, 1987). A 3.8 Kb- Clal/EcoRi gag-pol fragment from pHXB2 gpt was labeled with ³² P by random priming and used as a probe for the Slot

Blot. Nucleic Acid Analysis

RNA was extracted from cos-1 cells 48 hours after transfection, using the hot phenol method described by Queen and Baltimore. Queen, C and D. Baltimore, Cell, 33:741-748 (1983). For Northern blot analysis, 10 μg of DNase I-treated total cellular RNA was used per lane. RNA from mock transfected cos-1 cells was used as a negative control and 5 μg of RNA from H9 cells chronically infected with HXB2 virus was used as a positive control. The HIV-1 specific probe was a ³² P- labelled full length viral DNA. For analysis of proteins, cos-1 cells (4x10⁶) transfected 48 hours earlier with 10 μ g of each plasmid were labelled for 4 hours with 500 μCi of ³⁵ S methionine. As anegative control, cells were transfected with the plasmid pHXB2Bam p3, which does not produce virus due to a pos t-transcriptional defect. Feinberg, M.B. et al., Cell 46:807 (1986). Cell lysates were prepared and immunoprecipitations were performed as described. Veronese, F.D. et al., Proc. Natl. Acad. Sci. USA 82:5199 (1985). The HIV-1 positive human serum used in these experiments had demonstrated reactivity with all known viral structural proteins. Feinberg, M.B. et al., Cell 46:807 (1986). Immuno- precipitated proteins were resolved using a 10% SDS-polyacrylamide gel. Laemmli, U.K., Nature

227:680 (1970). Analysis of Viral Protein

Cos-1 cells (4x10⁶ ) transfected 48 hours earlier with pHXB2gpt; pA3-HXB; pA4-HXB; PA15-HXB; PA14-15-HXB and PHXB2Bam p3 were metabolically labeled with 500 μCi of ³⁵ S methionine for 4 hours.

Cell lysates were prepared and immunoprecipitations were performed using HIV-1 positive human

serum. which had demonstrated reactivity with all known viral structural proteins. Feinberg, M.B. et al., Cell 46:807 (1986). Immunopre-cipitated proteins were resolved using a 10% SDS-polyacrylamide gel electrophoresis. Laemmli, U.K., Nature. 227:680 (1970). Virus was pelleted by centrifugation for 3 hours at 27,000 rpm in a SW27 rotor. The pellet was resuspended in dissociation buffer (0.01 M Tris-HCL pH 7.3, 0.2% Triton X-100, 0.001M EDTA, 0.005M dithiothreitol (DTT), 0.006M KCL) if reverse transcriptase activity was to be measured, or in 0.2% Triton and Laemli buffer if protein analysis was to be performed. Western blot analysis and radio- immunoprecipitations (RIP) followed the procedure of Veronese et al. Veronese, F.D., et al., Proc. Natl. Acad. Sci. USA, 82:5199-5202 (1985). P24 analysis on tissue culture supernatants or on pelleted virus was performed.

Amounts of p24 gag protein (ng/ml) detected in the supernatant of each mutant 48 hours after transfection are shown in Figure 4. Each transfection was overlayed with 10 ml of medium. A

DuPont P24 ELISA kit was used and 3 different dilutions of each supernatant were analyzed. RT activity was measured after concentrating 3ml of Cos-1 supernatant from transfections of each mutant by centrifugation for 3 hours at 27,000 rpm.

Numbers refer to 1ml of supernatant and are the mean of three experiment. Analysis of protein content of wild-type and mutant particles by

radio-immunoprecipitation was also done.

EXAMPLE 3 Assessment of the role of HFV-1 p15-gag To address the role of HIV-1 p15^gag, a mutation was introduced into plasmid W13 (Kim, et al., J.

Virol., in press), which contains an infectious copy of the HIV-HXB2-D proviral DNA. (Shaw, et al.,

Science 226:1165-1171 (1984) W13 was modified by inserting an 8-nucleotide-long Cla I linker in a unique Apa I site present at position 2009, and then blunting this Cla I site with Klenow to rectify the gag reading frame. The mutated construct thpreby obtained is called bCA20-W13. The mutation results in the replacement of the two residues which

immediately follow the first Cys-His box, arginine-alanine, by a stretch of four amino acids, serine- isoleucine-alanine-methionine (Figure 7). Therefore, both the nature and the length of the intervening sequence between the two Cys-His boxes is altered.

To analyze the phenotypic consequences of the mutation, cos cells were transfected with bCA20 and generation of viral particles was scored by measuring the amount of p24 antigen as well as the reverse transcriptase activity released in the supernatant (Table 3).

a p24 antigen was measured using an ELISA assay system (DuPont-NEN, Inc., Billerica, MA)

b RT activity was determined as described (Kim, et al., J. Virol., in press)

bCA20-induced p24 and reverse transcriptase activities were approximately 40% and 30% of wild-type, respectively. This indicated that the mutation present in bCA20 only mildly interfered with the release of viral particles. The COS cell supernatant was then used to infect H9 cells, which were followed by an indirect immunofluorescence assay (Ho, D.D., et al., Science 226:451-453 (1984)), using serum from an HIV- infected individual as detector antibody. After three weeks, no positive cells were seen. This showed that the particles generated following transfection were non- infectious. Therefore, it could be concluded that the bCA20 mutation was lethal for viral replication.

To further study the consequences of the p15^{ga g} mutation contained in bCA20, a cell line which constitutively expresses an Env version of this mutant was generated. Expression of the HIV gag gene products is sufficient to generate viral particles in the absence of Env. Such cell lines are made as follows: Briefly, HT4-6C cells (a HeLa cell line expressing the CD4 molecule at its surface, a gift from B. Chesebro) Chesebro, B. and K. Wehrly, J. Virol. 62:3779-3788 (1988)), were trans- fected with construct bCA20-ΔE-dhfr, a modified version of bCA20-W13 which contains a translational frameshift in the env gene and the mutant dihydro- folate-reductase gene in place of the nef reading frame. Cells were selected for resistance to methotrexate, cloned, and analyzed by polymerase chain reaction to check for the presence of the viral integrant (not shown). Indirect immuno- fluorescence, using serum of an HIV-infected

individual as detector antibody, was also performed. The immunofluorescence seen in HT4(bCA20-ΔE-dhfr) was similar to that observed in HT4(WT-ΔE-dhfr), which expresses a wild-type gag sequence (not shown). p24 antigen and reverse transcriptase activity were also measured in the supernatants of these cell lines; the ratio of activity of bCA20 to wild-type was grossly similar to those observed with the transient transfection of the corresponding W13 viral constructs (Table 3). In addition, Western Blot analysis of cytoplasmic proteins was performed as described previously Trono and co-workers (Trono, D., et al., Cell, October 6, 1989), using an

anti-p24 monoclonal antibody (a gift from F.

Veronese) as detector antibody. HT4 (bCA20-ΔE-dhfr) and HT4(WT-ΔE-dhfr) gave similar patterns (Figure 8). Therefore, it could be concluded that the mutation present in bCA20 did not affect the

synthesis, the cleavage or the stability of the gag precursor.

To ask whether the mutation contained in bCΔ20 had deleterious consequences on the packaging of the viral RNA, a slot-blot analysis of the supernatant from HT4(WT-ΔE-dhfr) and HT4 (bCA20-ΔE-dhfr) cells was performed. For this, 700 μl of culture medium was mixed with 35 μl of 10 mg/ml proteinase K

(Boehringer Mannheim), 7 μl 10 mg/ml tRNA, 3.5 μl 0.5 M EDTA, 17.5 μl 20% SDS, incubated at 37°C for 45 min., phenol-extracted and ethanol precipitated in 0.4 M NaCl. The RNA was resuspended in 20 μl mM EDTA, denatured in 50% formamide - 17% formaldehyde, heated to 50ºC for 20 min., mixed with 120 μl

15xSSC, and bound to nitrocellulose by aspiration through a slot blot apparatus. Hybridization was then performed as previously described (Trono, D., et al., J. Virol. 62:2291-2299 (1988)), using a

[³²P] probe generated with T7 polymerase, complementary to nucleotides 8475 to 8900 of the HIV-1 genome. After hybridization, the filter was washed in 0.2xSSC three times at 68°C and exposed to X-ray film. Results showed that the amount of viral RNA present in the supernatant from HT4(bCA20-ΔE-dhfr) was dramatically reduced, compared to the control cell line, HT4 (WT-ΔE-dhfr) (Figure 9). Therefore, it was concluded that the bCA20 mutation specifically inhibited the packaging of the viral genomic RNA into particles.

Both cell lines were also examined by electron microscopy, to see if the defect in viral RNA packaging correlated with morphological differences (Figure 10). The morphology of the virus particles observed in HT4(WT-ΔE-dhfr) was very similar to that observed in HT4(R7-dhfr), a cell line infected with an Env⁺ , replication competent, version of the same virus: cell-released, "mature", particles contained a condensed core surrounded by the viral lipid bilayer (Figure 10, Panel A and Panel B). By contrast, in particles released from

HT4(bCA20-ΔE-dhfr), two dramatic differences were noted: first, the diameter of the particles was approximately 50% larger than observed in the controls; second, the electron- dense region was tightly apposed to the membrane, but the center of the particles was strikingly electron- luscent

(Figure 10, Panel C). Still, approximately 1% of the bCA20 particles had a morphology close to that of wild-type, probably because of some leakage in the bCA20 phenotype, as already suggested by the RNA hybridization study on the cell supernatant. These electron microscopy findings indicate that the inability to package the viral RNA in bCA20 part- icles is accompanied by an increased diameter and an absence of "collapse" of the inner components of the virion, which normally reflects the final steps of maturation. It remains to be determined whether this block of maturation is primarily due to the absence of viral RNA in the particle, or is a direct consequence of the aberrant p15^{ga g} protein. Interestingly, the HIV-1 virions produced by cells transfected with the bCA20 p15^{ga g} variant provirus bear a notable resemblance to the virus-like part- icles released from Spodoptera trugiperda insect cells infected with a recombinant baculovirus expression vector encoding the p57^gag precursor of the simian immunodeficiency virus, SIV_mac

(Delchambre, M., et al., EMBO J. 8: 2653-2660

(1989)). Comparison of the morphologic features of the RNA-minus particles produced in these diverse settings suggests that the viral RNA itself may play an important role in the structural organization and maturation of the mature retroviral virion.

In conclusion, as a result of this assessment of the phenotype of an in vitro-engineered HIV-1 variant, bCA20, which contains a mutation between the two Cys-His boxes of the p15^gag protein, it has been demonstrated that this domain is critical to the packaging of the genomic RNA into the virus particle. In addition, it has been shown that the RNA-minus phenotype generated by the p15^gag lesion is associated with striking morphological anomalies, as shown by electron microscopy. Importantly, results indicate that it is possible to generate HIV particles that do not contain the viral genome.

This is of primary relevance for the development of a vaccine strategy based on intact, fully immunogenic, but non-infectious virus particles.

Claims

1. A HIV-1 mutant construct which differs from

wild-type HIV-1 in that the HIV-1 mutant construct has an alteration in a nucleotide sequence which is present in wild-type HIV-1 and critical for virus packaging.

2. The HIV-1 mutant construct of Claim 1 wherein the alteration in a nucleotide sequence is present in the HIV-1 RNA ψ site or is present in a nucleotide sequence encoding a gag

expression product.

3. A non-infectious mutant HIV-1 virus particle.

4. A non-infectious mutant HIV-1 virus particle which:

a) has an alteration in a nucleotide sequence which is present in wild-type HIV-1 and is critical for virus packaging, and b) has a protein content essentially the same as the protein content of wild-type HIV-1.

5. A HIV-1 mutant construct comprising an

alteration in a region of the nucleotide sequence of wild-type HIV-1 critical for virus packaging, the alteration selected from the group consisting of:

a) deletion of all or a portion of the HIV-1 ψ site; b) deletion of nucleotides 293 to 331, inclusive;

c) deletion of nucleotides 293 to 313,

inclusive;

d) deletion of a gag nucleotide sequence

critical for virus packaging;

e) insertion of nucleotide sequence

CATCGCGATC at a unique Apa I p15^gag site present at position 2009;

f) substitution of the G at position 1508 by an A and of the G at position 1517 by an

A; and

g) substitution of the G at position 1571 by an A and of the G at position 1580 by an A .

6. A HIV-1 mutant expression product encoded by a HIV-1 mutant construct of Claim 5.

7. A vaccine comprising a HIV-1 mutant expression product of Claim 6 and a physiologically acceptable carrier.

8. A vaccine comprising a HIV-1 virus particle that does not contain the viral genome.

9. A vaccine comprising a non- infectious HIV-1 virus particle.

10. A plasmid capable of expressing mutated HIV-1 DNA in mammalian cells, comprising mutated HIV-1 DNA, the mutated HIV-1 DNA consisting essentially of wild-type HIV-1 DNA from which nucleotides 293-313 as they occur in wild-type DNA are missing.

11. A plasmid capable of expressing mutated HIV-1 DNA in mammalian cells, comprising mutated HIV-1 DNA, the mutated HIV-1 DNA consisting essentially of wild-type HIV-1 DNA from which nucleotides 293-331 as they occur in wild-type DNA are missing.

12. A plasmid capable of expressing mutated HIV-1 DNA in mammalian cells, comprising mutated HIV-1 DNA, the mutated HIV-1 DNA consisting essentially of wild-type HIV-1 DNA in which an A is present in place of the G present at nucleotide 1508 of the wild-type HIV-1 DNA and an A is present in place of the G present at nucleotide 1517 of wild-type HIV-1 DNA.

13. The plasmid of Claim 11, further comprising wild-type HIV-1 DNa in which an A is present in place of the G present at nucleotide 1571 of the wild-type HIV-1 DNA and an A is present in place of the G present at nucleotide 1580 of the wild-type HIV-1 DNA.

14. A plasmid capable of expressing mutated HIV-1 DNA in mammalian cells, comprising HIV-1 DNA having inserted therein in a unique Apal site present at position 2009, an 8-nucleotide-long Clal linker.

15. A plasmid capable of expressing mutated HIV-1 DNA in mammalian cells, comprising HIV-1 DNA having inserted therein in a unique Apal site present at position 2009, the following nucleotide sequence: CATCGCGATG.