WO1989012461A1 - Antagonists of viral transactivating proteins - Google Patents

Abstract

A method for preparing antagonists of viral transactivating proteins is disclosed. In one aspect the method involves the preparation of antagonists from active domains of the viral transactivating protein. Pharmaceutical compositions comprising the antagonists which compositions are useful in treating an individual suffering from viral infection are also disclosed.

Description

I ANTAGONISTS OF VIRAL TRANSACTIVATING PROTEINS CROSS REFERENCE TO RELATED APPLICATION

This application is a continuation-in-part of application Serial No. 207,393 filed June 16, 1988.

BACKGROUND OF THE INVENTION

The present invention relates to the fields of virology and biochemistry. More particularly, the present invention relates to a method for preparing peptide inhibitors of viral gene expression and viral replication. Many prominent pathogenic and cancer-inducing viruses of animal and man encode genes whose protein pro¬ ducts function to transcriptionally activate the expression of specific viral and/or cellular genes. The evolution of viral genes which encode transactivation proteins may repre- sent a universal phenomenon that has developed to facili¬ tate the replication of pathogenic viruses in otherwise hostile cellular environments. Examples of viral trans¬ activating proteins whose function is important for viral gene expression and viral replication include those encoded by the ubiquitious human herpes simplex virus (O'Hare and Hayward, 1985), the cause of numerous pathologies including the common cold sore, shingles, chicken pox and genital herpes lesions; human T cell lymphotropic virus types I (HTLVI), the etiological agent of adult T cell leukemia and HTLV2 (Chen et al., 1985; Felber et al., 1985; Sodroski et al., 1985a; Seiki et al., 1986; Fujisawa et al., 1986); human papillomavirus type 16, which is etiologically asso¬ ciated with cervical cancer (Phelps and Howley, 1987; Phelps et al., 1988); other serotypes of human papilloma- virus which cause sexually transmitted condylomas (Hirochika et al., 1987); human hepatitis B virus which is strongly implicated in hepatocellular carcinoma (Twu and Schloemer, 1987; Spandau and Lee, 1988); bovine leukemia virus (Rosen et al., 1985; Derse, 1987); simian virus 40 (Brady et al., 1984; Keller and Alwine, 1984); pseudorabies virus (Ihara et al., 1983) and human adenovirus types 2 and 5.

An example of urgent medical importance is human immunodeficiency virus type 1 (HIV), the primary etiological agent of acquired immune deficiency syndrome (AIDS) . The HIV tat gene product is a transactivating protein that is essential for viral gene expression and for replication of the AIDS virus (for review, see Chen, 1986) .

Evidence suggests that similar mechanisms are involved in the transactivating functions of these viral gene products (Kingston et al., 1985 and Berk, 1986). The present invention provides a method for preparing peptide antagonists of viral transactivating proteins. The pep- tides are useful as therapeutic agents in the control of viral infection and viral pathogenicity.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates a flowsheet of the steps employed in the practice of the present invention.

Figure 2(a) illustrates that no transactivation is detected in HeLa cells microinjected with pHIV-LTRCAT alone.

Figure 2(b) shows that efficient transactivation is detected in HeLa cells microinjected with both plasmid pCV-1 (ptat) and pHIV-LTRCAT.

Figure 3 shows the amino acid sequence of the transactivating protein (tat-86) of HIV (BRU isolate).

Figure 4 shows a comparison of transactivation response which occurs upon coinjection of HIV-LTRCAT with ptat or with the tat-86 peptide. Figure 5 represents the microscopic scoring of tat-transactivated cells. Cells injected with pHIV-LTRCAT and treated with tat-86 yield mainly high grain density cells as shown by autoradiographic analysis (3+ and 4+) , whereas injected cells treated with deletion mutants (tat- dm37-72, tat-dm37-62) or with very low levels of tat-86 may show a significant proportion of 1+ and 2+ cells. Panel A shows phase contrast microphotographs of representative 1+, 2+, 3+.and 4+ cells at 200 X magnification. Panel B shows the same cells viewed at 200 X magnification, but out-of- phase to more clearly demonstrate the 1+ and 2+ cells.

Figure 6 shows the amino acid sequence of HIV-1 tat protein (BRU isolate) . Superimposed on the sequence are tentative functional regions I to IV. The absolute boundaries of these regions are not known; they are drawn for purposes of discussion and based on data of the inventors.

Figure 7 represents the inhibition of HIV tat- induced transactivation of the HIV LTR-driven CAT gene by chemically synthesized HIV tat deletion and amino acid substitution mutant peptides. HeLa cells on coverslips were microinjected with pHIV-LTRCAT. To different cover¬ slips were added: (upper panel tat-86 alone (2X10^_6M); (second panel from top) tat-86 (2X10^~6M) + tat-dm52-86 (10^_!5M); (second panel from bottom) tat-86 (2X10^"6M) + £a±-dm37-62,47Ala (10^_5M); (bottom panel) tat-86 (1X10^"6M) + tat-dm37-62,46Ala (10^_5M). Six hours after peptide addition, cells were fixed and analyzed by in situ hybridization using ³⁵S-dCTP labeled CAT plasmid DNA. The photographs show typical autoradiographic patterns.

Figure 8 illustrates the transcriptional activa¬ tion of the E2 gene by adenovirus 5 E1A region 3 (Lillie, Loewenstein, Green, and Green, 1987) . HeLa cells were microinjected with the E2 gene (in pE2) and E1A peptide or plasmid, as specified below, and then analyzed by immuno- fluorescence 18 to 24 hr. after injection with an E2- specific antibody. The photographs show typical immuno- fluorescence patterns but represent only a small area of the total quadrant and therefore are not accurate quanti¬ tative representations. Panel A; pE2 -^*- pm975 (encodes 289 amino acid ElA protein). Panel B: pE2 plus region 3 peptide. Panel C; pE2 plus region 1 peptide. Panel D; pE2 plus region 3 peptide preincubated with antibody specific to the N-terminus of region 3.

Figure 9 represents the sequence and localization of adenovirus 5 functional protein domains. The location of domains which encode cell DNA induction, transcriptional repression, transcriptional activation, and immortalization functions are schematically illustrated (see Lillie, Green and Green, 1986; Lillie, Loewenstein, Green and Green, 1987) .

Figure 10(a) shows conserved E5 regions among animal papillomavirus including reindeer papillomavirus (RPV), the deer papillomavirus (DPV), European elk papillo¬ mavirus (EEPV) and the bovine papillomavirus (BPVI) .

Figure 10(b) shows the conserved E5 region between human papillomavirus type 16 (HPV16), human papillomavirus type 18 (HPV18), and human papillomavirus type 6B (HPV6B) . Figure 11(a) shows the amino acid similarity between the adenovirus type 5 ElA 13S product and the human papillomavirus type 16 E7 oncoprotein (from Phelps et al., 1988) .

Figure 11(b) illustrates the cellular DNA induc- tion of quiescent NIH/3T3 cells by human papillomavirus type 16 E7 deletion mutant peptides (Rawls, Pusztai, and Green, unpublished data) .

Figure 11(c) illustrates transactivation of the adenovirus type 5 E2 promoter by human papillomavirus HPV16 E7 deletion mutant peptides (Rawls, Pusztai, and Green, unpublished data) .

STATEMENT OF THE INVENTION

The present invention provides a method for pre- paring antagonists of viral transactivating proteins. For purposes of the present invention, the terra "transactivating protein" includes oncogene products which have transactivat¬ ing function. In another aspect the present invention also provides compositions comprising the antagonists which com- positions are useful in treating an individual suffering from viral infection. Accordingly, the invention also con¬ templates a method for treating an individual suffering from a viral infection by administering an effective amount of an antagonist of the transactivating protein of said virus. Viral genes that encode transactivating proteins are defined by their ability to transcriptionally activate other viral genes. Putative transactivating genes are first identified by mutational analysis showing that the expres¬ sion of a specific viral gene is required for the subsequent expression of another viral gene or genes. Experimental proof that the viral gene encodes a transactivating protein is obtained by showing that cotransfection with a plasmid containing the gene and a plasmid containing the target promoter results in stimulation of transcription from the target. The amino acid sequence of the viral transactivat¬ ing protein is derived from the DNA sequence by the identi¬ fication of open translational reading frames. This method¬ ology is well proven and has been employed for the identifi¬ cation of the transactivating proteins of essentially all viruses listed above.

Briefly, the method of the present invention employs the following steps after the identification of a viral transactivating protein (see Figure 1). An active domain of the transactivating protein is determined. This determination can be accomplished by making N-terminal and C-terminal deletions of the transactivating protein and assaying the truncated proteins for transactivating activity. Although not critical, in most cases it is preferred that one obtain the minimal active domain thereby minimizing the size of the antagonist peptide. Amino acid substitution is then made in an active domain, preferably for conserved amino acid residues, to deactivate the trans¬ activating function of the peptide while not destroying its binding ability thereby producing an antagonist of the viral transactivating protein. Conserved residues may be deter¬ mined by comparing the sequence of virus from the same class. Viral sequences are contained in publicly available databases such as Genbank (Intelligenetics) and NBRF (National Biomedical Research Foundation) . It should be noted that fragments of transactivating proteins which do not have transactivating activity may possess the desired antagonistic activity. Hence, the mutation of active domains should only be considered a preferred method of obtaining peptide antagonists. The general methodology for practice of the present invention is outlined in Figure 1, and is believed to be applicable to all the viruses listed above.

The peptides of the present invention are used as therapeutic agents in the treatment of disease states asso¬ ciated with virus infection. The peptides may be adminis¬ tered as free peptides or pharmaceutically acceptable salts thereof. The term "pharmaceutically acceptable salts" refers to those acid addition salts or metal complexes of the peptides which do not significantly or adversely affect the therapeutic properties (e.g. efficacy, toxicity, etc.) of the peptides. The peptides should be administered to individuals as a pharmaceutical composition which, in most cases, will comprise the peptide and/or pharmaceutical salts thereof with a pharmaceutically acceptable carrier. The term "pharmaceutically acceptable carrier" refers to those solid and liquid carriers which do not significantly or adversely affect the therapeutic properties of the peptides. The pharmaceutical compositions containing peptides of the present invention may be administered to individuals, particularly humans, either intravenously, subcutaneously, intramuscularly, intranasally or even orally. The necessary dosage will vary with the particular condition being treated, method of administration and rate of clearance of the peptide from the body.

The above-described peptides may be prepared by any suitable synthesis method. Exemplary synthesis methods include solid-phase synthesis techniques as described in the textbook entitled "Solid-Phase Peptide Synthesis", Stewart & Young, Pierce Chemical Company, Rockford, Illinois (1984), solution synthesis and the fragment condensation synthesis methods. Those skilled in the art of biochemical synthesis will recognize that such synthesis methods require the use of a protecting group to stabilize a labile side chain to prevent the side chain from being chemically altered during the synthesis process. Protection of the alpha-amino group is most commonly required to insure proper peptide bond formation, followed by selective removal of the alpha-amino protecting group to permit subsequent peptide bond formation at that location. In selecting a particular side chain protecting group to be employed in the synthesis of such peptides, the protecting group should be stable to the reagents and conditions employed for removal of the alpha-amino protecting group at each step in the synthesis process and must be removable upon completion of the synthesis process under reaction conditions which will not detrimentally alter the peptide. For commercial- scale quantities of peptides, such peptides can be prepared using recombinant DNA or synthesis techniques.

Finally, it is recognized that by the use of recombinant DNA techniques, genes expressing peptide antago¬ nists discovered by the general methodology of Fig. 1 can be cloned into suitable viral vectors (e.g. retrovirus or adenovirus) for therapeutic delivery to humans. In par¬ ticular, this approach with HIV tat peptide antagonists is an attractive possibility for the somatic cell gene therapy of hemopoietic stem cells of individuals infected with the AIDS virus.

All peptide structures represented in the follow¬ ing description and claims are shown in conventional format wherein the amino group at the N-terminus appears to the left and the carboxyl group at the C-terminus at the right. Likewise, amino acid nomenclature for the naturally occur¬ ring amino acids found in protein and comprising the peptide inhibitors of the present invention is as follows: alanine (Ala;A), asparagine (Asn;N) , aspartic acid (Asp;D), arginine (Arg;R) , cysteine (Cys;C), glutamic acid (Glu;E), glutamine (Gln;Q) , glycine (GlysG), histidine (His;H), isoleucine (lie;I), leucine (Leu;L), lysine (Lys;K), methionine (Met;M) , phenylalanine (Phe;F), proline (Pro;P), serine (Ser;S), threonine (Thr;T), tryptophan (Trp;W), tyrosine (Tyr;Y), and valine (Val;V).

For sake of clarity and brevity in the explana¬ tion of the present invention, the description will focus on application of the invention to the following viral systems:

( i) the transactivating protein encoded by the HIV tat gene; ( ii) the transactivating protein encoded by the human adenovirus EIA gene, a model viral oncogene; and (iii) the cellular DNA synthesis-inducing protein encoded by the bovine papillomavirus type 1 (BPV1) E5 oncogene, and the cellular DNA synthesis-inducing and transactivating protein encoded by the human papillomavirus type 16 E7 oncogene. It will be understood that the description that follows is equally applicable to other viruses that encode a trans- activating gene, including the herpes simplex viruses and the human hepatitis B virus.

HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 (HIV)

The HIV tat transactivating protein. The AIDS virus, H_ψ , encodes a unique transactivating protein called tat. an

1987; Peterlin et al., 1986; Rosen et al., 1986; Sodroski et al., 1985b; Wright et al., 1986). The mechanism of tat transactivation is unclear. Evidence has been presented for (i) an increase in the steady state level of HIV- specific mRNA and/or (ii) an increase in the translational utilization of that mRNA (Cullen, 1986; Wright et al., 1986) . More recently, evidence was reported that tat increases HIV mRNA levels by enhancing the rate of trans¬ cription rather than stabilizing HIV specific mRNAs (Hauber et al., 1987). The presumed target of tat action, the TAR (transactivation response) region which extends from -17 to +80 within the LTR, is required for tat transcription and contains striking inverted and direct repeats (Rosen et al., /0

1986) . Direct repeats are found often within transcrip¬ tional regulatory sequences such as enhancers and promoters (Maniatis et al., 1987).

HIV tat specifically recognizes the HIV LTR and tat transactivation is essential for HIV replication. Thus, the tat system provides an attractive target for drug inter¬ vention in AIDS. The relatively small size of the tat gene product, 86 amino acids, places it within the limits of peptide synthesis. The open reading frame (ORF) has been defined by DNA sequence and mutational studies (for a review, see Chen, 1986) . Tat function is essential for viral gene expression and for- the replication of the AIDS virus. Hence, a therapy for AIDS can be developed based on peptide. antagonists of the tat protein that block its func- tion. The application of the invention to the identifica¬ tion and chemical synthesis of antagonistic tat peptides is described below. These and homologous peptides are useful as drugs to arrest the progress of AIDS.

Step 1 Development of Sensitive Assays for Biological Activity

Cultured human Hela cells (or other suitable cultured cells) are coinjected with (i) a recombinant plasmid that expresses a functional transactivating protein and (ii) a recombinant plasmid that contains the target promoter upstream of a suitable reporter gene. The expres¬ sion of the reporter gene is measured by immunofluorescence at the level of the protein gene product, or by in situ hybridization at the level of mRNA. In situ hybridization has been found to be an extremely sensitive and reliable method for measuring transactivation with low background, especially with ³⁵S-labeled DNA probes. / /

The initial task was to develop a specific assay to measure transactivation by transactivating peptides, in this case, HIV tat peptides. This was achieved by the development of a cell microinjection assay for HIV tat function. The first step is to show that the assay detects the activity of natural tat protein expressed constitutively from plasmid ρCV-1 (also referred to as ptat) (Arya et al., 1985) . The assay for transactivation function of tat peptides employs plasmid pHIV-LTRCAT (Okamoto and Wong- Staal, 1986) which contains HIV-LTR 5 ' of a chloramphenicol acetyltransferase gene (CAT) . CAT expression measured as mRNA is dependent on the efficient initiation of transacti¬ vation of HIV-LTR. As described below, this can be achieved by co-microinjection into the cell nucleus of pHIV-LTRCAT with plasmid ptat or by the addition of tat peptide to the medium of microinjected cells.

In the assay, HeLa cells are co-microinjected with pHIV-LTRCAT and pCV-1. The activation of the HIV LTR is measured 6 hr. later by in situ hybridization analysis that measures the expression of the CAT gene as mRNA. Referring to Figure 2(a), no detectable HIV-LTR transactivation is detected in cells microinjected with pHIV-LTRCAT alone. However, co-microinjection with pCV-1 dramatically induces HIV-LTR-driven expression of CAT, as shown by the massive cellular accumulation of CAT mRNA, visualized by the coalescence of autoradiographic grains above positive cells, see Figure 2(b). From 60 to 80% of microinjected cells are transactivated by co-microinjection. This assay for tat transactivation function is rapid, extremely sensi- tive and reliable. The assay system responds well to chemically synthesized tat peptides and deletion mutants that encode the tat transactivation domain. IZ

Step 2 Synthesis of a Large Peptide Representing the Putative Active Domain

The next step is to synthesize the largest peptide fragment that likely encodes a functional protein domain. For large proteins, a large minimal active peptide is mapped by genetic analysis in which essential and non- essential coding regions in the ORF of the transactivating gene are delineated. This is illustrated by the adenovirus ElA 289 amino acid protein as described later. The HIV tat protein of 86 amino acids is sufficiently small to permit initial chemical synthesis of the entire protein molecule as described below.

The full length tat 86 peptide was chemically synthesized by the solid-phase method of Merrifield (1963) using tBoc chemistry and a manual adaptation of the auto¬ mated procedure outlined in Clark-Lewis et al., (1986). Figure 3 shows the sequence of tat-86.

The cell microinjection/ situ hybridization assay was further developed to assay transactivation by tat peptides. T_a_t peptides were directly added to HeLa cells immediately after microinjection with pHIV-LTRCAT using the following procedure.

1. Cell microiniection and peptide treatment. HeLa cells are seeded in Dulbecco modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) on 22 mm² cover- slips in 35 mm plastic petri dishes. When the cells are 50-70% confluent, the nuclei of 100 cells in quadrants etched on the coverslip (each quadrant contains 400-600 cells) are microinjected with the reporter plasmid pHIV- LTRCAT (or control pSV2CAT) via glass capillaries under constant pressure of a plastic syringe. The CAT gene in pHIV-LTRCAT is under the control of the HIV LTR regulatory region and is not expressed significantly in HeLa cells unless cells are (i) coinjected with pCV-1 (a plasmid expressing the HIV tat 86 amino acid protein) or (ii) treated with biologically active tat-86 peptides or

5 deletion tat peptides containing the active regions for the tat transactivation function. The CAT gene in pSV2CAT is under the control of the simian virus 40 promoter, is expressed constitutively in HeLa cells, and thus serves as a control for nonspecific inhibition of transcription by

10 candidate tat peptide antagonists.

To assay a peptide for tat transactivation function, cells are microinjected with pHIV-LTRCAT (100 ug/ml), washed with DMEM, and peptide is added (10^-5-10^-7M) in 2 ml, of DMEM/0.1% bovine serum albumin (BSA). To assay y< 15 potential antagonists of tat transactivation function, a mixture of test peptide and chemically synthesized tat-86 peptide (2X10^~6M) (sufficient to transactivate 30-35% of cells, i.e. about 50% of maximum) is added to cells in 2 ml of DMEM/0.1% BSA. Cells are incubated at 37°C in a humidi- 20 fied incubator flushed with 5% Cθ and examined at 1 and 4

hr. for potential toxicity. At 2 hr. after peptide addi¬ tion, 1 ml of DMEM/30% FBS is added, and incubation con¬ tinued until 6 hr after peptide addition. Cells are washed twice with phosphate-buffered saline (PBS) , once with 4% 25 paraformaldehyde/PBS/5 mM MgCl₂/ fixed in paraformaldehyde for 15 min., washed with 70% ethanol, and stored in 70% ethanol at 4° until used for hybridization.

2. In situ hybridization and autoradiographv. Coverslips are incubated at room temperature for 10 min. in PBS/5 mM 30 MgCl₂ and for 10 min. in 0.1 M glycine/0.2 M Tris, pH 7.4, followed by 10 min. at 65° in 50% formamide/2X SSC (SSC = 0.15 M NaCl/0.015 M Na₃ citrate). Coverslips are placed cell side down on parafilm containing 20 ul amounts of hybridization mixture and incubated at 37° for 5 to 16 hr. The hybridization mixture consists of 2X SSC/10% dextran sulfate/2 mg/ml BSA/1 mg/ml sheared salmon sperm DNA/1 mg/ml E coli tRNA/50 mM dithiothreitol. Each 20 ul portion contains 4 X 10⁶ dpm of ³⁵S-dCTP labeled CAT plasmid DNA (1-2 X 10⁸ dpm/ug) labeled by the nick translation reaction. Post hybridization washings are performed by incubating coverslips at 37° for 30 min. in 50% formamide/2X SSC and for 30 min. in 50% formamide/lX SSC, followed by 3 to 4 washes at room temperature with shaking in IX SSC. Cells are dehydrated by incubating sequentially for 5 min. periods in 70%, 95%, and absolute ethanol, followed by drying in air and mounting cell side up on microscope slides. Slides are dipped in Kodak NTB-2 emulsion at 42 deg, dried for 1 hr., and exposed for 1-2 days at 4 deg in a light-tight slide box containing a dessicant. Slides are developed in D-19 developer in total darkness, fixed, washed, and dried for at least 1 hr. prior to examination.

3. Microscopic scoring of HIV tat Transactivated Cells in Autoradiographs. Experimental slides are routinely examined after autoradiography by phase microscopy at a magnification of 200X. Cells are scored from 4+ to 1+ as follows: 4+ response - the emulsion over cells is black, i.e. silver grains are packed, coalesce, and obscure the cell structure beneath the emulsion; 3+ response - silver grains are tightly packed and too numerous to count but cell structure is visible beneath the emulsion; 2+ and 1+ response - apparent at 400 X and higher magnificaion, grain counts are 100 to 200 for 2+ and about 50 to 100 per cell for 1+. Hereafter, 4+ and 3+ are referred to as "high grain density" positives whereas 2+ and 1+ are referred to as "low density." In most experiments, control HeLa cells not microinjected and cells microinjected with pHIV-LTRCAT but not exposed to ptat or tat peptides show no significant background grain counts with a ³^S-labeled CAT DNA probe. Thus, the assay is both specific and highly sensitive. Cells microinjected with pHIV-LTRCAT plus pCV-1 (see Figure 2) or microinjected with pHIV-LTRCAT and treated with high levels of active tat peptide (see Figure 4) reveal mainly 4+ and 3+ response. Cells microinjected with pHIV-LTRCAT and treated with lower level of tat peptides exhibit increasing proportions of 2+ and 1+ cells (see Figure 5) . This assay has the advantage that it mimics the clinical situation in which tat antagonist peptide would be administered to AIDS patients. The transactiva ion response by the chemically synthesized tat-86 peptide is remarkably efficient and rapid. A comparison with the transactivation response which occurs upon coinjection of HIV-LTRCAT with the tat-86 peptide is shown in Figure 4. These results demonstrate that the chemically synthesized full length tat protein effectively enters the cell nucleus and transactivates the HIV-LTRCAT. This finding provides the experimental basis for the delineation of a smaller functional peptide domain and for the development of peptide antagonists of tat.

Step 3 Identification of Smaller Peptides that Exhibit Transactivating Function bv Sequential Synthesis and Assay of N-terminal and C-Terminal Deletion Mutants

The next step is the synthesis of N-terminal deletion mutants to eliminate nonessential N-terminal sequences. Synthesis starting at the C-terminus (residue 86) was stopped after addition of amino acid residues 68, 52, 37 and 22, to yield deletion mutant peptides of 19, 35, 50 and 65 residues (see Table 1 below) . Tests for trans¬ activation function showed that the 65 and 50 residue t_a_t deletion mutants were quite active. These findings indi¬ cate that the N-terminal 36 amino acids are not essential for the tat transactivation function when added as peptide to cells. TABLE 1

Chemically Synthesized Tat-86 Peptides

1. N-TERMINAL DELETION MUTANTS

Tat-dm22-86_* CTTCY CKKCCFHCQV CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ VSLSKQPTSQ-

PRGDPTGPKE-COOH

5 Tat-dm37-86* CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ VSLSKQPTSQ PRGDPTGPKE-COOH Tat-dm52-86^a RRQRR RPPQGSQTHQ VSLSKQPTSQ PRGDPTGPKE-COOH Tat-dm68-86^g SLSKQPTSQ PRGDPTGPKE-COOH

2. N- AND C-TERMINAL DELETION MUTANTS

Tat-dm32-57_* NH₂-FHCQV CFTTKALGIS YGRKKRRQRR R-CONH₂ 10 Tat-dm32-62_* NH₂-FHCQV CFTTKALGIS YGRKKRRQRR RPPQGS-CONH₂ Tat-dm32-67* NH₂-FHCQV CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ V-CONH₂ Tat-dm32-72+ NH₂-FHCQV CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ VSLSKQ-CONH₂ Tat-dm37-57_* NH₂-CFTTKALGIS YGRKKRRQRR R-CONH₂ Tat-dra37-62_* NH₂-CFTTKALGIS YGRKKRRQRR RPPQGS-CONH₂ 15 Tat-dm37-67 NH₂-CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ V-CONH₂ Tat-dm37-72_* NH '2,-CFTTKALGIS YGRKKRRQRR RPPQGSQTHQ VSLSKQ-CONH,

Tat-dm49-57^a NH.-RKKRRQRR R-CONH,

TABLE 1 (continued)

3. AMINO ACID SUBSTITUTION. N-. C-TERMINAL DELETION MUTANTS

Tat-dm37-62.54Ala* NH₂-CFTTKALGIS YGRKKRRARR RPPQGS-CONH₂

Tat-dm37-62.47Ala^a NH₂-CFTTKALGIS AGRKKRRQRR RPPQGS-CONH₂

5 Tat-dm37-62.46Ala^a NH₂-CFTTKALGIA YGRKKRRQRR RPPQGS-CONH₂

Tat-dm37-62.41Ala^a NH₂-CFTTAALGIS YGRKKRRQRR RPPQGS-CONH₂

Tat-dm37-62,40Ala^c NH₂-CFTAKALGIS YGRKKRRQRR RPPQGS-CONH₂

Tat-dm37-62,39Ala^c NH₂-CFATKALGIS YGRKKRRQRR RPPQGS-CONH₂

Tat-dm37-62,38Ala* NH₂-CATTKALGIS YGRKKRRQRR RPPQGS-CONH₂

10 Tat-dm37-62.37Ala* NH₂-AFTTKALGIS YGRKKRRQRR RPPQGS-CONH₂ i

Tat-dm37-62.46Ala.47Ala^a NH₂-CFTTKALGIA AGRKKRRQRR RPPQGS-CONH₂

Tat-dm37-72.47Ala^a NH₂-CFTTKALGIS AGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

Tat-dm37-72.46Ala^a NH₂-CFTTKALGIA YGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

Tat-dm37-72.41Ala^a NH₂-CFTTAALGIS YGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

15 Tat-dm37-72.40Ala^c NH₂-CFTAKALGIS YGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

Tat-dm37-72,39Ala^c NH₂-CFATKALGIS YGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

Tat-dm37-72.46Ala. NH₂-CFTTKALGIA AGRKKRRQRR RPPQGSQTHQVSLSKQ-CONH₂

47Ala^a

4. FULL LENGTH TAT-86 PEPTIDES

20 Tat-86 NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTKALGIS Y

GRKKRRQRR RPPQGSQTHQ VSLSKQPTSQ PRGDPTGPKE-COOH

TABLE 1 (continued)

Tat-86.22Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTAATTCY CKKCCFHCQV CFTTKALGIS Y COOH^d

Tat-86.25Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTAY CKKCCFHCQV CFTTKALGIS Y COOH

Tat-86.27Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY AKKCCFHCQV CFTTKALGIS Y COOH

5 Tat-86.30Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTKALGIS Y COOH

Tat-86.31Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCAFHCQV CFTTKALGIS Y COOH

Tat-86.34Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTKALGIS Y COOH

Tat-86.37Ala^b NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV AFTTKALGIS Y COOH

Tat-86.41Ala^a NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTAALGIS Y COOH

10 Tat-86.47Ala^a NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTKALGIS A COOH

Tat-86.41Ala. NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTAALGIS A COOH ex.

47Ala^b

Tat-86,41Ala, NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTAALGIA Y COOH

46Ala^b

15 Ta t-86 . 46Ala . NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTKALGIA Y COOH

Tat-86 . 41Ala , NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTAALGIA Y COOH

46Ala^b

Tat-86 . 41Ala . NH₂-MEPVDP RLEPWKHPGS QPKTACTTCY CKKCCFHCQV CFTTAALGIA A COOH

46Ala , 47Ala^b

, ,

20 *Actιve in HIVLTR transactivation; ^aAntagonist of HIVLTR transaction; ^bPotential antagonist under test; ^cDiminished activity; ^ddesignates tat

To establish how much of the C-terminal domain is required for tat peptide transactivation, a series of peptid starting at residues 72, 67, 62 and 57 and terminating synthe after addition of amino acid residues 37 or 32 (see Table 1) were synthesized. DNA transfection studies have indicated th tat protein lacking the second exon (residues 73-86) can stil transactivate the HIV-LTR (Muesing et al., 1987). All of th N-terminal/C-terminal deletion mutants were tested for activi and found to be active (Table 1) . Deletion of 14 amino acids from the C-terrainus (and 36 amino acids from the N-terminus) yielded a mutant peptide (tat-dm 37-72) that exhibits activit close to that of the 50 amino acid peptide, tat-dm37-86, whi possesses the native C-terminus. Deletion of 10 additional amino acids produced a mutant peptide, tat-dra37-62, that still possesses clear but reduced transactivation activity. This particular mutant peptide has proven very useful because sing amino acid substitution mutations built into this peptide backbone by chemical synthesis are strong antagonists of tat transactivation (see step 6 below) . To summarize, the inventors' findings have led the to the conclusion that tat-86 may be divided tentatively int four functional domains (see Figure 6). These are referred t as regions I, II, III, and IV for the purpose of discussion a for use in the design of mutant peptides that will function as antagonists of HIV replication.

Step 4 Development of Biological Assays to Measure Peptide Antagonist Function

The inventors have shown that an efficient (and clinically relevant) method for introducing tat peptides into cells is by peptide addition to the culture medium following the protocol described above. To assay for antagonistic function, a level of tat-86 that induces 50% of maximal transactivation is first established. Candidate tat peptide are tested for antagonist activity by simultaneous addition with tat-86 to HeLa cells microinjected with pHIV-LTRCAT. As an internal control for non-specific inhibition of gene trans¬ cription by tat peptide derivatives, a quadrant of cells on the same coverslip is microinjected with pSV2CAT; constitutive expression of CAT under the control of the SV40 promoter is independent of tat expression.

Step 5 Identification of Peptide Antagonists Among Deletion Mutant Peptides that are Inactive or Possess Low Activity in the Transactivation Assay

In principle, tat deletion peptides containing only sequences that facilitate binding to putative target molecules could be antagonists of tat function. For example, they may bind to a critical site on the target molecule but may not possess the tat activation region. Thus, they may prevent fully active tat protein molecules from binding. Deletion mutant peptides that were inactive in the transactivation assa were tested for their ability to inhibit tat-86 function in th antagonist assay described in step 4. Peptide ;fca.t---dm52-86, which contains 5 of the 8 amino acids in basic region III, was found to be a good antagonist of tat-86 function, exhibiting 40% inhibition of total cells transactivated at 5 x 10^~6 M peptide (see Table 2 and Fig. 7). Tat-dm49-57, which contains only basic region III, exhibits antagonist activity at high peptide concentration (see Table 2) .

TABLE 2

Inhibition of Tat-Mediated HIVLTR Transactivation by Tat Deletion Peptides Added to Cells

Transactivated Cells % Inhibition

Total (High Density)

Ta_t Peptide 2 x 10_^"~5_M 1. x J ).^~5_M 5_ x 10_^"6-M

52-86 60 (80) • 52 (70) 40 (34)

49-57 31 (37) 20 (0) 0 (0)

Treatment with tat-86 alone gave an average of 54% total cel response and 30% high density response. The control values within each experiment were normalized to 100% for calculatio of the percent inhibition of total cell and high density cel response.

Step 6 Further Development of Peptide Antagonists by Chemical Synthesis of Amino Acid Substitution Mutant Peptides in Minimal Domains

Once active peptide domains and preferably minimal domains are established, amino acid substitution mutant peptides are designed and constructed as potential antagonist of tat-induced transactivation. The amino acid sequence of t minimal active domain is compared by computer search with th analogous sequence in all known virus strains and isolates. I this manner, conserved amino acid residues are identified. 2Λ

Logically, the mutation of conserved residues would likely yield a biologically inactive peptide that could serve as an antagonist. Ideally, the mutation does not reside in a critical binding site for the target molecule but instead at a site that is still vital for function, e.g., an activation region. The inventors have synthesized a series of such mutant peptides substituting alanine for individual amino acid specified in Table 1. Twenty five amino acid substitution mutant peptides built into tat minimal domain peptides or longer peptides were tested for their ability to transactivate HIVLTRCAT by the standard assay procedure. All tat peptides showing deficiency in transactivation, as detailed in Table 3, are potential antagonists. Using this approach, the inventors have identified 14 different HIV tat mutant peptides including several full length tat peptides that exhibit high levels of antagonist activity. These are described in the following sections.

TABLE 3

Transactivation of HIVLTR by Chemically Synthesized HIV Tat Mutant Peptides with Amino Acid Substitutions in Region I or II

Transactivated Cells (%) Total (High Density)

Tat Peptide Added l ϊ lfi^' -5 M i x 10 -6 M

Tat 37-62 Mutants 37-62 44 ( 15 ) 28 ( 9 ) 37-62 , 37Ala 40 ( 10 ) 20 ( 3 ) A3

Table 3 (continued)

37-62, 38Ala 27 (4) 10 (0)

37-62, 39Ala 4 (1) 0 (0)

37-62, 40Ala 15 (4) 7 (0)

37-62, 41Ala 5 (0) 0 (0)

37-62,46Ala 0 (0) 0 (0)

37-62, 47Ala 0 (0) 0 (0)

37-62, 46, 47Ala 1 (0) 0 (0)

Tat 37-72 Mutants

37-72 61 (34) 41 (21)

37-72, 39Ala 19 (2) 11 (1)

37-72,40Ala 10 (3) 3 (1)

37-72, 41Ala 2 (0) 0 (0)

37-72, 46Ala 0 (0) 0 (0)

37-72, 47Ala 0 (0) 0 (0)

37-72, 46Ala,47Ala 1 (0) 0 (0)

Tat 37-86 Mutants

37-86 68 (12) 39 (8)

37-86, 41Ala 4 (0) 0 (0)

37-86,47Ala 1 (0) 0 (0)

Tat 1-86 Mutants - chemically synthesized

1-86 72 (56) 54 (23) l-86,22Ala 8 (0) 2 (0) l-86,25Ala 7 (0) 1 (0) l-86,27Ala 2 (1) 0 (0) l-86,30Ala 15 (1) 8 (1) l-86,31Ala 21 (0) 5 (0) l-86,34Ala 55 (15) 49 (8) l-86,37Ala 35 (8) 18 (0) l-86,41Ala 2 (0) 0 (0) l-86,47Ala 7 (0) 1 (0) Table 3 (continued)

Tat 1-86 Mutants - E. coli produced

1-86 59 (20) 33 (10) l-86,22Ser 8 (0) 7 (1)

1-86, 41Ala 8 (2) 4 (0) l-86,46Ala 14 (4) 11 (3) l-86,47Ala 10 (0) 4 (1)

1-86,46, 47Ala 10 (0) 3 (0) l-86,20Thr,46, .47A1;a 10 (3) 5 (0)

The data are the mean of from two to six experiments

1. Strong Tat antagonists are produced by substituting alanin for region II amino acids in peptide backbone tat-dm37-62. Th inventors first substituted alanine for conserved amino acids within the minimal domain peptide tat-dm37-62. In particular, tat peptides with substitutions in amino acids 40 to 47 (regio II, putative activation domain, Figure 6) that are defective i transactivation function (Table 2) were investigated. Tat-dm37 62,41Ala, iai-dm37-62,46Ala, tat-dm37-62.47Ala. and the doubl mutant £&£-dm37-62,Ala46,Ala47 are strong antagonists (Table 4, see Fig. 7). Several of these antagonists reduced wild type tat transactivation activity by 80-90%. Of particular significance, cells expressing large amounts of HIVLTR-directe CAT RNA, i.e., "high density cells," were also reduced by 90%.

TABLE 4

Strong Inhibition by Tat Mutant Peptides of HIV-LTR Transactivation by Tat-86 Peptide Added to Cells

Transactivated Cells Inhibition

Total (High Density)

Tai Peptide 1 S 1QT⁵ M 1 £ 1__T⁵ M 5_ x 1Q_^" M 2 .1ST* M

Tat : 37-62 Mutants

37- •62, ,37Ala 2 (0) 0 (0) ND ND

10 37- •62, ,39Ala 12 (13) 0 (0) ND ND

37- •62, ,40Ala 47 (39) 4 (4) ND ND

37- ■62, ,41Ala 61 (32) 50 (35) 11 (13) 10 (0)

37- -62, , 46Ala 76 (91) 53 (63) 42 (52) 25 (20) b.

37- •62, , 47Ala 83 (97) 63 (95) 50 (69) 31 (46)

15 37- ■62, ,46,47Ala 85 (89) 67 (50) 41 (33) 15 (0)

The data are the mean of two to twelve experiments. HeLa cells on coverslips were microinjected with pCD12 and treated with Tat-86 (2 x 10^""^** M) or Tat-86 together with the indicated concentration of mutant Tat peptide. Treatment with Tat-86 alone gave an average of 54% total

20 cell response and 30% high density response. The control values within each experiment were normalized to 100% for calculation of the percent inhibition of total cell and high density cell response.

A

2. Effective tat antagonists are produced by substituting alanine for region II amino acids in peptide backbone tat- dm37-72. Next the same four tat amino acid substitutions wer generated within tat-dm37-72 backbone to determine whether antagonist activity could be increased in this manner. Since the native 37-72 backbone possesses stronger transactivation activity than the native 37-62 backbone, it was considered tha 37-72 may be more effectively taken up by cells or more tightl bound to its target. Consistent with this possibility, all four antagonists were effective, with several showing stronger activity in the 37-72 backbone compared to the same mutant peptides in the 37-62 backbone (see Table 5) .

TABLE 5

Strong Inhibition by Tat Mutant Peptides of HIV-LTR Transactivation by Tat-86 Peptide Added to Cells

Transactivated Cells

%. Inhibition

Total (High Density) τa_t Peptide l ϊ M - 5 M i ___ _ r⁵ i ___ in^-6 M a in - 6 M

Tat 37-72 Mutants

37-72,39Ala 15 (5) 0 (0) ND ND

10 37-72,40Ala 22 (10) 8 (0) ND ND 37-72,41Ala 70 (91) 68 (83) 41 (73) 22 (66) 37-72, 6Ala 76 (100) 62 (91) 47 (74) 37 (60) ? 37-72,47Ala 97 (100) 93 (100) 71 (83) 41 (70) ^**4 37-72,46,47Ala 82 (100) 54 (33) 21 (44) ND

15 The data are the mean of two to twelve experiments. HeLa cells on coverslips were microinjected with pCD12 and treated with Tat-86 (2 x 10^~δ M) or Tat-86 together with the indicated concentration of mutant Tat peptide. Treatment with Tat-86 alone gave an average of 54% total cell response and 30% high density response. The control values within each

20 experiment were normalized to 100% for calculation of the percent inhibition of total cell and high density cell response.

3. Tat antagonists in peptide backbone tat-dm37-86 containing alanine substituting for region II amino acids are also effective. Two of the substitutions that represent sites for strong antagonist activity, 41Ala and 47Ala, were built into the tat-dm37-86 backbone. Both of these peptides retain good antagonist activity (see Table 6) .

TABLE 6

Strong Inhibition bv Tat Mutant Peptides of HIV-LTR Transactivation by Tat-86 Peptide Added to Cells

. Transactivated Cells

% Inhibition Total (High Density)

Tat Peptide 2. x 10^~5 M 1 x lQT⁵ __M 5_ x 10^~β M

Tat 37-86 Mutants . 37-86, 1Ala 66 (98) 48 (50) 45 (33)

37-86,47Ala 75 (83) 54 (42) 40 (32)

The data are the mean of two experiments. HeLa cells on coverslips were microinjected with pCD12 and treated with Tat-86 (2 x 10^-6 M) or Tat-86 together with the indicated concentration of mutant Tat peptide. Treatment with Tat-86 alone gave an average of 54% total cell response and 30% high density response. The control values within each experiment were normalized to 100% for calculation of the percent inhibition of total cell and high density cell response. A °/

4. Tat antagonists in full length tat-86 backbone containing alanine substituting for region II amino acids. Finally, the same two amino acid substitutions were tested in the full-length tat-86 peptide backbone. Tat-86,41Ala is a strong antagonist (see Table 7). However, Tat-86,47Ala is leaky, i.e. it possesses weak transactivation activity, and accordingly its antagonist activity is diminished, but significant; presumably, the protein conformation imposed by the addition of N-terminal region I alters the activity of region II.

The inventors have tested full-length tat-86 antagonists in a very stringent assay for effectiveness in blocking HIV-LTR-driven gene expression from an integrated HIV-LTRCAT cell line. The results show that in particular tat-86,41Ala is a potent inhibitor of wild-type tat-86 and is apparently stable over the extended period of this assay (48 hrs). Remarkably, £iLt-86,4lAla blocks transactivation 99% at a 4-fold molar excess over that of tat-86. The inventors predict that this particular analog, those listed in Table 1, part 4, and their derivatives will be potent inhibitors of HIV-1 replication.

All of these mutant peptide antagonists are of potential value. Even a somewhat diminished antagonist in the full-length tat backbone may have therapeutic value because of important additional factors including peptide stability and uptake into cells. What is clear from these data is that the same substitutions that produce antagonists in the smallest peptide backbone will likely produce antagonists in larger peptide backbones. Finally, the inventors examined whether substi¬ tutions in other critical amino acids in the full length tat-86 molecule could result in antagonist activity. In principle amino acid substitution in conserved cysteine residues may alter "functional folding" but still permit binding of the mutant peptide to its target. Such molecules could prevent the binding of molecules that possess the active conformation. The full length tat mutant peptides shown in Table 1 are being tested presently for antagonist activity. Several of these exhibit antagonist activity in preliminary studies. Further studies in progress will determine their usefullness as antagonists of HIVLTR trans¬ activation.

TABLE 7

Strong Inhibition by Tat-86 Mutant Peptides of HIV-LTR. Transactivation by Tat-86 Peptide Added to Cells

Transactivated Cells % Inhibition Total (High Density)

Tat Peptide 2 x 1Q^~5 M 1 x 10^~5 M 5_ x 10^~6 M

Tat-86 Mutants - chemically synthesized

Ta_t-86,41Ala 78 (75) 49 (67) 30 (25)

Ta_t-86,47Ala 50 (62) 34 (33) 30 (8)

Tat-86 Mutants - E. coli produced

Ta£-86,41Ala 78 (79) 65 (54) 42 (58)

Tat-86,47Ala 54 (93) 38 (27) 36 (20)

The data are the mean of two experiments. HeLa cells on coverslips were microinjected with pCD12 and treated with Tat-86 (2 x 10^~6 M) or Tat-86 together with the indicated concentration of mutant Tat peptide. Treatment with Tat-86 alone gave an average of 54% total cell response and 30% high density response. The control values within each experiment were normalized to 100% for calculation of the percent inhibition of total cell and high density cell response.

To rule out the possibility that inhibition of transactivation by tat antagonists is due in part to non- specific inhibition of transcription, in several experi¬ ments cells on the same coverslip were injected with pSV2CAT or pH4WTCAT which express CAT from the SV40 or human histone 4 promoter, respectively. CAT expression from these pro¬ moters was not affected by tat peptide antagonists. In other experiments, the incorporation of ³⁵S-methionine into an acid-insoluble form was measured in the absence and presence of tat antagonist peptides. No reduction of incorporation was observed at 1 x 10^-5 M peptide and less than 10% reduction at 3 x 10^-5 M (data not shown) . We conclude that inhibition of HIV-LTR transactivation by tat peptide antagonists is specific and does not reflect a general block in the transcription of injected genes or an inhibition of cellular protein synthesis.

Development of additional tat peptide antagonists. Inasmuch as substitution in tat amino acid residues 41 or 47 generate antagonists of tat-86 in some degree in all peptide backbones tested, the inventors believe and predict that all substitutions in tat-dm37-62 and tat-dm37-72 back¬ bones that have been shown to be antagonists will also be antagonists in longer backbones or in full length tat-86. Furthermore, the inventors believe and predict that inasmuch as 41 and 47 substitutions are generally good antagonists in any backbone, a double substitution of 41 and 47 will be a good antagonist in any backbone. Lastly, the inventors recognize that substitutions with amino acids other than alanine for tat residues 41, 46, 47, 46 and 47, may produce superior antagonists. Gene therapy approach for the delivery of tat antagonist peptides to HIV infected cells. The inventors have synthesized full-length genes encoding some tat peptide antagonists. These mutant tat genes express tat peptides that have been shown to effectively block the function of tat-86. The inventors believe and predict that the expression of these tat mutant genes in cells will result in antagonism of HIV tat function. Such genes therefore are believed to be of value for the gene therapy of AIDS.. The inventors have cloned these tat mutant genes into several viral expression vectors. Transduction into hematopoietic stem cells of AIDS patients is a promising therapy and may avoid problems of drug delivery and drug stability.

ANTAGONISTS OF HUMAN ADENOVIRUS INFECTION Basic Approach. A number of peptides have been synthesized (Table 8 below) for use as potential mutants of th adenovirus ElA oncogene protein domain 3 (PD3). It has been previously demonstrated that the parent peptide, PD3, of 49 amino acids functions as an autonomous transcriptional activato of human adenovirus early genes (Lillie et al., 1987). Further more, it is now known that most of the chemically synthesized mutant peptides are defective in transactivating activity i.7i vivo . These mutant peptides are potential antagonists of PD3 function and may interact with human host cell transcription factors that regulate gene expression. Thus, in the long run the mutant peptides may be useful therapeutically for (1) the inhibition of viral infection in human beings, (2) the regul tion of human gene expression for the possible control of cancer and genetic diseases of man. Furthermore, in vivo and vitro assays have been developed to functionally assay PD3 mutant peptides as possible antagonists of transcriptional activation. Finally, preliminary data suggests that three o the mutant PD3 peptides are antagonists. The application of the method of the present invention to the identification an chemical synthesis of antagonistic peptides to PD3 is describ below.

TABLE 8 POTENTIAL PEPTIDE ANTAGONISTS OF VIRUS INFECTION AND HUMAN GENE EXPRESSION

1. PARENT HUMAN ADENOVIRUS 2 ElA ONCOGENE PROTEIN DOMAIN 3. PD3 (140-188)

NH₂-Glu-Glu-Phe-Val-Leu-Asp-Tyr-Val-Glu-His-Pro-Gly-His-Gly-Cys-Arg 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 Ser-Cys-His-Tyr-His-Arg-Arg-Asn-Thr-Gly-Asp-Pro-Asp-Ile-Met-Cys-Ser 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 Leu-Cys-Tyr-Met-Arg-Thr-Cys-Gly-Met-Phe-Val-Tyr-Ser-Pro-Val-Ser-COOH 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188

10 2. AD2 ElA PROTEIN DOMAIN 3 (PD3) SUBSTITUTION MUTANTS (SM)

PD3 parent H. -EEFVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH PD3-154Ala NH. -EEFVLDYVE HPGHGARSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH PD3-157Ala NH₂ -EEFVLDYVE HPGHGCRSAH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH PD3-171Ala^a NH₂ -EEFVLDYVE HYGHGCRSCH YHRRNTGDPD IMASLCYMRT CGMFVYSPVS-COOH

15 PD3-174Ala NH. -EEFVLDYVE HYGHGCRSCH YHRRNTGDPD IMCSLAYMRT CGMFVYSPVS-COOH PD3-174Ala^a NH. -EEFVLDYVE HYGHGCRSCH YHRRNTGDPD IMCSLCYMRT AGMFVYSPVS-COOH PD3-174Asn NH. -EEFVLDYVE HYGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYNPVS-COOH

3. AD2 ElA PROTEIN DOMAIN 3 (PD3) N-TERMINAL DELETION MUTANTS (dm)

dml(141-188) NH.-EFVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH

20 dm2(142-288) NH. -FVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH

TABLE 8 (Continued)

dm3(143-188) NH.-VLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH dm4(146-188) NH.-YVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH dm5(148-188) NH.-E HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH 5 dm6(150-188) NH.-PGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH dm7(152-188) NH₂-HGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH dm8(156-188) NH.-SCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH dm9(158-188)^a NH. -H YHRRNTGDPD IMCSLCYMRT CGMFVYSPVS-COOH

4. AD2 ElA PROTEIN DOMAIN 3 (PD3) C-TERMINAL DELETION MUTANTS (DM)

10 dmlO(140-187) NH₂-EEFVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSPV-CONH, dmll(140-186) NH.-EEFVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYSP-CONR, ^ dml2(140-185) NH.-EEFVLDYVE HPGHGCRSCH YHRRNTGDPD IMCSLCYMRT CGMFVYS-CONH,

5. PARENT AD2 ElA PROTEIN DOMAIN 2 (PD2) - TRANSCRIPTIONAL REPRESSION. CELLULAR DNA INDUCTION, AND IMMORTALIZATION DOMAIN

15 PD2-121-138 NH. -DLTCHEAGFPPSDDEDEE-COOH

6. PARENT AD2 ElA PROTEIN DOMAIN 1 (PD1) - TRANSCRIPTION REPRESSION.

CELLULAR DNA INDUCTION, AND IMMORTALIZATION DOMAIN

PD1-41-80 NH.PTLHELYDLDVTAPEDPNEEAVSQIFPDSVMLAVQEGIDL-COOH

Antagonistic activity on ElA transactivation.

30

Step 1 The Development of Sensitive Assays for Biological Activity

To measure the transactivation function of adeno¬ virus ElA PD3 peptides, the inventors have developed two methods of assay which are highly sensitive. First, is an in vivo assay in which a target ElA-inducible gene, early adenovirus gene E2 (present in plasmid pE2) , is co-microin- jected with PD3 peptide or PD3 peptide mutants into HeLa or other suitable cell lines. Transactivation is measured 8 to 18 hours later by immuno-fluorescence which detects the E2 gene product (Lillie et al., 1987) or by in situ hybrid¬ ization which detects viral mRNA. Figure 8 shows the transcriptional activation of the E2 gene by ElA region 3 (Lillie., Loewenstein, Green and Green, 1987) .

Step 2 The Chemical Synthesis of a Large Peptide

Representing the Putative Active Domain

As established by DNA mutational analysis, the ElA 289 ElA protein that functions as a transactivator encodes three functional domains, PD1, PD2, and PD3 (see Figure 9) (Lillie et al., 1986; 1987); only PD3 is essen¬ tial for the transactivation function of the ElA oncogene. Peptides encoding the entire PD3, PD2 and PD1 domains were prepared. It is now known that PD3 alone (49 amino acids, see Table 8) is an autonomous transactivator of early viral genes both in vivo and in vitro (Lillie et al., 1987) .

Step 3 Identification of Smaller Peptides that

Exhibit Transactivating Function by Seguential

Synthesis and Assay of N-Terminal and C-Terminal Deletion Mutants

A series of 9 N-terminal deletion mutants of PD3 were prepared (Table 8, dml to dm9) and tested for in vivo 7 and in vitro transactivation function. All N-terminal deletion mutants exhibited less than full activity by in vivo assay. All activity was lost after deletion of 13 N- terminal amino acids. These mutant peptides are candi- dates for ElA trans- activation antagonists. Three C- terminal deletion mutants have been synthesized (see Table 8).

Step 4 Development of Biological Assays to Measure Peptide Antagonist Function The inventors have shown that an efficient method of introducing PD3 peptides into cells is by co-microinjec¬ tion with the target plasmid encoding an ElA inducible gene. In vivo assays for antagonism involve coinjecting PD3 at a con- centration that exhibits half maximal activity with potential peptide antagonists and the reporter plasmid. Measurement of antagonist activity is performed by assaying the target gene product as described above. In vitro assays for antagonist function involve direct competition between PD3 and potential antagonists using the standard cell-free transcription assay. The level of antagonism is determined by comparison with control in vitro assays using PD3 alone.

Step 5 Identification of Peptide Antagonists among Deletion Mutant Peptides that are Inactive or Possess Low Activity in the Transactivation Assay As mentioned above, several deletion mutant pep¬ tides have been identified that are defective in transacti¬ vation function. These peptides are, therefore, candidate ElA transactivation antagonists. Step 6 Further Development of Peptide Antagonists by Chemical Synthesis of Amino Acid Substitution Mutant Peptides in Minimal Domains

Six single amino acid substitution mutants have been synthesized (see Table 8) . All five cysteines in PD3 have individually been substituted with alanine; four of the cysteines are stringently conserved among different adenovirus serotypes (see Figure 9). All cysteine substi¬ tution mutants were found to be defective in vivo . They fall into two categories, those with substantially reduced activity and two which exhibit no activity at all (PD3- 171Ala and PD3-179Ala, see Table 8). Not unexpectedly, these two cysteine mutants are completely defective in the in vitro assay for transcriptional activation. Preliminary evidence suggest that these two mutant peptides will func¬ tion as antagonists, as determined by the in vitro assay.

ANTAGONISTS OF ONCOGENIC PAPILLOMAVIRUSES

Basic Approach. The method of this invention has been applied to the bovine papillomavirus type 1 E5 gene (Green and Loewenstein, 1987) . This system is important because E5 is an oncogene and is involved in regulating cellular genes involved in cellular DNA replication. A large number of E5, peptides (Table 9 below), some of which represent potential antagonistic peptides of the bovine papillomavirus (BPV1) oncogene E5 have been prepared. It has been demonstrated that the chemically synthesized parent E5 peptide of 44 amino acids and several N-terminal deletion mutant peptides function as autonomous inducers of cellular DNA replication in growth-arrested cells (Green and Loewenstein, 1987) . These peptides may be useful therapeutically for the regulation of human DNA replication, and thus may be of importance for the control of some neo- plastic and genetic diseases of man. The method of the present invention is applied to the bovine papillomavirus E5 protein as described below. 31

Step 1 The Development of Sensitive Assays for Biological Activity

To routinely test the ability of E5 peptide and mutant peptides to induce cellular DNA synthesis, the inventors have used a microinjection assay. Briefly, mouse C127 cells are growth arrested by serum deprivation and then microinjected with E5 peptide preparations using glass capillaries. The induction of DNA synthesis is determined by the incorporation of [³H]thymidine, as measured by auto- radiography (Green and Loewenstein, 1987) .

Step 2 The Chemical Synthesis of a Large Peptide Representing the Putative Active Domain

The entire 44 amino acid E5 protein (see Table 9) was chemically synthesized and demonstrated to function as a potent inducer of cellular DNA synthesis (Green and

Loewenstein, 1987) . About 50% of cells incorporate radio¬ active thymidine after microinjection with as little as 10 ug/ml of the E5 44 amino acid peptide.

Step 3 Identification of Smaller Peptides that Exhibit DNA-Synthesis Inducing Function bv Seouential Synthesis and Assay of N-terminal and C-terminal Deletion Mutants

A series of 14 N-terminal deletion mutants of BPV1 E5 were chemically synthesized (see Table 9) and tested as described above for DNA-synthesis induction function. Remarkably, the C-terminal 13 amino acid peptide was found to be active. The C-terminal 10 and 12 amino acid peptides were devoid of activity. These represent potential E5 antagonists. The sequential addition of 7 amino acids to the C-terminal end substantially increased the DNA induction activity of the resulting peptides. Activity of the C- terminal 20 amino acid peptide (E5-dm 25-44), (see Table 9) was essentially that of the entire 44 amino acid E5 protein. These data demonstrate a minimal core domain of DNA induc¬ tion activity residing in the C-terminal 13 amino acid residues and a near full activity domain residing in the first 20 C-terminal amino acid residues.

Step 4 Development of Biological Assays to Measure Peptide Antagonist Function

The inventors have shown that an efficient method of introducing E5 peptides into growth arrested cells is by microinjection. Assays for antagonism involve co-injecting the E5 44 amino acid peptide at a concentration that exhibits half maximal activity together with potential E5 peptide antagonists. Measurement of antagonist activity is by comparison of DNA induction by the full length E5 44-mer with that of the mixture of E5 44-mer and candidate peptide antagonist.

Step 5 Identification of Peptide Antagonists Among Deletion Mutant Peptides that are Inactive or Possess Low Activity in the Transactivation Assay

As mentioned above, two deletion mutant peptides have been identified that are defective in DNA-synthesis induction function. These peptides are therefore candidate DNA-synthesis induction antagonists.

Step 6 Further Development of Peptide Antagonists by Chemical Synthesis of Amino Acid Substitution Mutant Peptides in E5 Minimal Domains

E5 is conserved among animal papillomavirus type 1 (BPV1) (see Figure 10(a)). Moreover, E5 - s—a-lso conserved among two strong candidates for causative agents of cervical carcinoma and other urogenital malignancies (human papillo- mavirus type 16 (HPV16) and HPV18)_r as well as one of the causative agents of venereal condylomas and juvenile laryn- geal papillomatosis (HPV6B) (see Figure 10(b)). ^<- /

Among the conserved sequences is the hydrophobic stretch which correspond to the essential middle domain of BPVl E5, and the conserved Cys-X-Cys sequence. A number of substitution mutants and N-terminal deletion/ substitution mutants in which Ala is substituted for one or both Cys at residues 37 and 39 have been prepared. Mutants with both substitutions are completely defective in their DNA synthesis-induction function, whereas those with single Cys substitutions are greatly reduced in activity. These substitution mutants may be peptide antagonists of E5 DNA- synthesis induction function.

Human Papillomavirus type 16 E7 oncoprotein. - peptide antagonists

Of considerable medical importance, the E7 oπco- gene of human papillomavirus type 16 (HPV16) encodes a transactivation function (Phelps et al, 1988). The protein encoded by E7 oncogene contains regions of remarkable similarity to those of human adenovirus ElA protein domains 1, 2, and 3 (see Figure 11(a)). The inventors have chemi- cally synthesized the full E7 oncoprotein of 98 amino acids as well as a series of deletion mutant peptides. They have demonstrated that regions 1 and 2 function as an oncogene to induce cellular DNA synthesis (Figure 11(b)). Further¬ more, they have shown that region 3 functions as a trans- activator (Figure 11(c)). Inasmuch as there is much evi¬ dence that HPV16 is involved in cervical cancer and other urogenital malignancies, the inventors predict that peptide antagonists of HPV16 E7 will prove useful for the preven¬ tion and therapy of some human urogenital malignancies. TABLE 9

POTENTIAL PEPTIDE ANTAGONISTS OF CELLULAR DNA REPLICATION

1. PARENT BPVl E5 44-mer

NH. - Met - Pro - Asn - Leu - Trp - Phe - Leu - Leu - Phe - Leu - Gly - 5 ^': 1 2 3 4 5 6 7 8 9 10 11

Leu - Val - Ala - Ala - Met - Gin - Leu - Leu - Leu - Leu - Leu - 12 13 14 15 16 17 18 19 20 21 22

Phe - Leu - Leu - Leu - Phe - Phe - Leu - Val - Tyr - Trp - Asp - 23 24 25 26 27 28 29 30 31 32 33

10 His - Phe - Glu - Cys - Ser - Cys - Thr - Gly - Leu - Pro - Phe - COOH. 34 35 36 37 38 39 40 41 42 43 44

2. BPVl E5 PEPTIDE SUBSTITUTION MUTANTS δ

E5(l-44) NH.-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCTGLPF-COOH

E5-37Ala,39Ala NH.-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FEASATGLPF-COOH

15 E5-43Ala NH.-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCTGLPF-COOH

E5-34Ala NH.-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDA FECSCTGLPF-COOH

E5-32Ala NH.-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYADH FECSCTGLPF-COOH

3. BPVl E5 PEPTIDE COMBINED N-TERMINAL DELETION (dm) AND SUBSTITUTION MUTANTS

TABLE 9 (Continued)

E5-dm(25-44),43Ala NH.-LLFFLVYWDH FECSCTGLAF-COOH

E5-dm(25-44),34Ala NH,-LLFFLVYWDA FECSCTGLPF-COOH

E5-dm(25-44),32Ala NH.-LLFFLVYADH FECSCTGLPF-COOH

E5-dm(25-44) ,37Ala,39Ala NH,-LLFFLVYWDH FEASATGLPF-COOH

E5-dm(25-44),33Ala NH.-LLFFLVAWDH FECSCTGLPF-COOH

E5-dm(27-44),37Ala NH₂-FFLVYWDH FEASCTGLPF-COOH

E5-dm(27-44),37Ala,39Ala NH₂-FFLVYWDH FEASATGLPF-COOH

E5-dm(27-44),39Ala, NH.-FFLVYWDH FECSATGLPF-COOH

10 E5-dm(32-44) ,37Ala,39Ala NH.-WDH FEASATGLPF-COOH

E5-dm(32-44),33Ala NH, -WAH FECSCTGLPF-COOH

E5-dm(32-44),34Ala NH.-WDA FECSCTGLPF-COOH δ

BPVl E5 PEPTIDE N-TERMINAL DELETION MUTANTS (dm)

E5-dm(16-44) (29-mer) NH, -MQLLLLLFL LLFFLVYWDH FECSCTGLPF-COOH 15 E5-dm(17-44)(28-mer) NH.-QLLLLLFL LLFFLVYWDH FECSCTGLPF-COOH E5-dm(19-44)(26-mer) NH. -LLLLFL LLFFLVYWDH FECSCTGLPF-COOH E5-dm(21-44) (24-mer) NH,-LLFL LLFFLVYWDH FECSCTGLPF-COOH E5-dm(25-44) (20-mer) NH, -LLFFLVYWDH FECSCTGLPF-COOH E5-dm(27-44) (18-mer) NH. -FFLVYWDH FECSCTGLPF-COOH 20 E5-dm(28-44) (17-mer) NH. -FLVYWDH FECSCTGLPF-COOH E5-dm(29-44) (16-mer) NH. -LVYWDH FECSCTGLPF-COOH

E5-dm(30-44)(15-mer)

E5-dm(31-44)(14-mer) NH,-YWDH FECSCTGLPF-COOH E5-dm(32-44)(13-mer) NH₂-WDH FECSCTGLPF-COOH E5-dm(33-44)(12-mer) NH₂-DH FECSCTGLPF-COOH E5-dm(35-44)(10-mer) NH, -FECSCTGLPF-COOH

BPVl E5 PEPTIDE C-TERMINAL DELETION MUTANTS (dm)

E5-dm(l-43) NH₂-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCTGLP-CONH, E5-dm(l-42) NH,-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCTGL-CONH₂ 10 E5-dm(l-41) NH₂-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCTG-CONH₂ E5-dm(l-40) NH₂-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSCT-CONH₂ E5-dm(1-3,9) NH,-MPNL WFLLFLGLVA AMQLLLLLFL LLFFLVYWDH FECSC-CONH,

BPVl E5 PEPTIDE COMBINED C-TERMINAL AND N-TERMINAL DELETION MUTANTS

E5-dm(25-43) NH. -LLFFLVYWDH FECSCTGLP-COOH 15 E5-dm(25-42) NH₂ -LLFFLVYWDH FECSCTGL-COOH E5-dm(25-41) NH. -LLFFLVYWDH FECSCTG-COOH E5-dm(25-40) NH. -LLFFLVYWDH FECSCT-COOH E5-dm(25-39) NH. -LLFFLVYWDH FECSC-COOH

REFERENCES

Arya, S. K., Guo, C, Josephs, S. F. and Wong-Staal, F., 1985. Science 229, 69-73.

Berk, A. J., 1986. Ann. Rev. Genet. 20, 45-79.

Brady, J. N., Bolen, J. B., Radonovich, M. , Salzman, N. and Khoury, G., 1984. Proc. Natl. Acad. Sci. U.S.A. 81, 2040-2044.

Chen, I. S. Y., 1986. Cell 47, 1-2

Chen, I. S. Y., Slamon, D. J., Shah, N. P., Quan, S. G. and Wachsman, W. , 1985. Science 229:54-58.

Clark-Lewis, I., Aebersold, R., Ziltener, H., Schrader, J. W., Hood, L. E. and Kent, S. B. H., 1986. Science 231, 134-139.

Cullen, B. R., 1986, Cell 46, 973.

Dayton, A. I., Sodroski, J. G., Rosen, C. A., Goh, W. C. and Haseltine, W. A., 1986. Cell 44, 941-947.

Derse, D. , 1987. J. Virol. 62, 1115-1119.

Feinberg, M. B., Jarrett, R. F., Aldovini, A., Gallo, R. C. and Wong-Staal, F., 1986. Cell 46, 807.

Felber, B. K. , Paskalis, H., Kleinrnan-Ewing, C, Wong-Staal, F. and Pavlakis, G., 1985. Science 229, 675-679. Fujisawa, J. , Seiki, M. , Sato, M. and Yoshida, M. , 1986. EMBO J. 5, 713-718.

Green, M. and Loewenstein, P. M. , 1987. Cell 51, 795-802.

Hauber, J. , Perkins, A., Heimer, E. P. and Cullen, B. R. , 1987. Proc. Natl. Acad. Sci. U.S.A. 84, 6364.

Hirochika, H., Broker, T. R. and Chow, L. T. , 1987. J. Virol. 61, 2599-2606.

Ihara, S., Feldman, L., Watanabe, S. and Ben-Porat, T.,- 1983. Virology 131, 437-453.

Keller, J. M. and Alwine, J. C, 1984. Cell 36, 381-389.

Kingston, R. E. , Baldwin, A. S. and Sharp, P. A., 1985. Cell 41, 3-5.

Lillie, J. W. _r Green, M. and Green, M. R., 1986. Cell 46, 1043-1051.

Lillie, J. W. , Loewenstein, P. M. , Green, M. R. and Green, M. , 1987. Cell 50. 1091-1100.

Maniatis, T. , Goodbourn, S. and Fischer, J. A., 1987. Science 236. 1237-1245.

Merrifield, R. B., 1963. J. Am. Chem. Soc. 85, 2149-2154

Muesing, M. A., Smith, D. H. and Capon, D. J. , 1987. Cell 48, 691-701.

O'Hare, P. and Hayward, G. S., 1985. J. Virol. 61, 190-199. -/I Okamoto, T. and Wong-Staal, F., 1986. Cell 47, 29-35.

Peterlin, B. M., Luciw, P. A., Barr, P. J. and Walker, M. D., 1986. Proc. Natl. Acad. Sci. U.S.A. 83, 9734.

Phelps, W. C. and Howley, P. M. , 1987. J. Virol. 61, 1630-1638.

Phelps, W. C, Yee, C. L. and Howley, P. M. , 1988. Cell 53, 539-547.

Rosen, C, Sodroski, J., Kettmann, R., Burny, A. and Haseltine, W.A. , 1985. Science 227, 320-322.

Rosen, C, Sodroski, J. G., Goh, W. , Dayton, A., Lippke, J. and Haseltine, W. A., 1986. Nature (London) 555-559.

Seiki, M. , Inoue, J., Takeda, T. and Yoshida, M. , 1986. EMBO J. 5, 561-565.

Sodroski, J., Rosen, C, Goh, W. C. and Haseltine, W. , 1985a. Science 228, 1430-1434.

Sodroski, J. G., Rosen, C, Wong-Staal, F., Slahuddin, S. Z., Popovic, M. , Arya, S., Gallo, R. C. and Haseltine, W. A., 1985b. Science 227, 171-173.

Spandau, D. F. and Lee, C. H., 1988. J. Virol. 62, 427-434.

Twu, J. S. and Schloemer, R. H., 1987. J. Virol. 61, 3448-3453.

Wright, C. M. , Felber, B. K. , Paskalis, H. and Pavlakis, G. N., 1986. Science 234, 988.

Claims

WHAT IS CLAIMED IS:

1. A method for preparing antagonists of viral replication which comprises

(a) producing peptide fragments of the transactivating protein of a virus; and (b) selecting said fragment which exhibits antagonist activity toward replication of said virus.

2. A method for preparing antagonists of viral replication which comprises

(a) producing a peptide fragment containing an active domain of the transactivating protein of a virus or a full length peptide of the transactivating protein of a virus; and

(b) making at least one amino acid substitu¬ tion in said peptide fragment containing an active domain or in said full length peptide to deactivate the transactivating activity of the peptide while not destroy¬ ing its binding ability thereby producing an antagonist to replication of said virus.

3. A method of claim 1 in which the virus is selected from the group consisting of. human immuno¬ deficiency virus type 1, herpes simplex virus, human adenovirus, human papillomavirus, and human hepatitis B virus.

4. A method of claim 2 in which the virus is selected from the group consisting of human immuno¬ deficiency virus type 1, herpes simplex virus, human adenovirus, human papillomavirus, and human hepatitis B virus.

5. A method of claim 1 in which the virus is human immunodeficiency virus type 1.

6. A method of claim 2 in which the virus is human immunodeficiency virus type 1.

7. An antagonist produced by the method of claim 5

8. An antagonist produced by the method of claim 6

9. A peptide containing an active domain of the transactivating protein of HIV useful in producing antagonists to HIV selected from the group consisting of tat-86. tat-dm22-86, tat-dm37-86. ±_a_t-dm32-57, tat-37-57. tat-dm37-62 and tat-dm37-72.

10. An antagonist of replication of HIV selected from the group consisting of tat-dm52-86, tat- dm49-57, £a±-dm37-62,41Ala, £a£-dm37-62, 6Ala, tat-dm37- 62,47Ala, £aJt-dm37-62,46Ala,47Ala, tat_-dm37-72,41Ala, Ja£-dm37-72,46Ala, tat-dm37-72, 7Ala, £a±-dm37-72,46Ala, 47Ala, t_a£-dm37-86,41Ala, £a_fc-dm37-86,47Ala, tat-86,41Ala. and tat-86,47Ala.

11. A pharmaceutical composition comprising an antagonist of claim 10 and a pharmaceutical carrier.