AU699168B2 - Selection and use of antiviral peptides - Google Patents

Description

WO 95/02071 PCT/US94/07781 SELECTION AND USE OF ANTIVIRAL PEPTIDES 1. INTRODUCTION The present invention relates to methods of screening and selecting antiviral compounds or peptides. A potential antiviral compound or peptide mimics an essential surface of a wild-type viral protein in an in vitro reconstitution assay. The compound or peptide is incubated with a wild-type viral protein, host transactivation factors which interact with the wild-type viral protein and an appropriate DNA promoter fragment.

Potential antiviral compounds and peptides are detected by formation of altered protein:protein and/or DNA:protein transcription complexes as compared to transcriptional complexes formed in the absence of such exogenous compounds or peptides.

2. BACKGROUND OF THE INVENTION Viruses that infect eukaryotic cells often encode transcriptional regulatory proteins to ensure the appropriate expression of genes during the viral life cycle. Disclosed examples include the Ela protein of adenovirus (Berk, 1986, Annu. Rev. Genet. 45-79; Flint and Shenk, 1989, Annu. Rev. Genet. 23: 141-161); large T antigen of simian virus 40 (Keller and Alwine, 1984, Cell 36: 381-389); and the E2 protein of papioma virus (Phelps and Howley, 1987, J. Virol. 61: 1630-1638).

Herpes simplex virus type-1 (HSV-1) contains a double-stranded, linear DNA genome comprised of approximately 152 kbp of nucleotide sequence, which encodes approximately 75 genes. The viral genes are transcribed by cellular RNA polymerase II and are temporally regulated, resulting in transcription and subsequent synthesis of gene products in roughly three discernable phases. These phases, or kinetic classes of genes, are referred to as the immediate-early, early and late genes.

Briefly, the program of HSV-1 gene expression can be summarized as follows.

Upon infection of susceptible cells, five immedia'e-early genes are transcriptionally activated through the agency of VP16, a 65kD protein present in the incoming virus particle (Patterson and Roizman, 1983, J. Virol. 46: 371-377; Campbell, et al., 1984, J.

Mol. Biol. 180: 1-9) to yield infected cell polypeptides 0, 4, 22, 27 and 47 (Periera, et al., 1977, Virology 27: 733-749). Transcription of these five genes does not require n f i; iig r .11

I)

WO 95/02071 PCT/US94/07781 prior viral protein synthesis (Clements, et al., 1977, Cell 12: 275-285). The products of immediate-early genes are required to activate transcription and regulate the remainder of the HSV genome. Expression of early proteins, many of which are involved in viral DNA synthesis, signals the onset of viral DNA synthesis, which in turn is required for maximum expression of late genes (Holland, et al., 1980, Virology 101: 10-24). Late genes specify virion structural proteins and can be further divided into two subclasses, ri and r2. Transcription of l1 genes 'om input genomes occurs at low levels in the absence of viral DNA synthesis but is maximal when progeny viral DNA is synthesized (Holland, et al., 1980, Virology 101: 10-24; O'Hare and Hayward, 1985, J. Virol. 53: 751-760). By contrast, the accumulation of r2 transcripts is stringently dependent upon viral DNA synthesis (Holland, et al., 1980, Virology 101: 10-24; O'Hare and Hayward, 1985, J. Virol. 53: 751-760). In addition to the requirement for viral DNA synthesis, maximum expression of both rl and r 2 genes requires functional immediate-early proteins (Holland, et al., 1980, Virology 101: 10-24; O'Hare and Hayward, 1985, J. Virol. 53: 751-760; Honess and Roizman, 1975, Proc. Natl. Acad. Sci. USA 72: 1276-1280).

The HSV-1 protein, infected cell polypeptide 4, herein ICP4 (also known as a4 or Vmw175), is an essential immediate-early protein localized in the nucleus of the host cell which can be purified as a homodimer (Metzler and Wilcox, 1985, J. Virol. 329-337). ICP4 is multifunctional, including being involved in the enhanced expression of early and late HSV-1 genes (Dixon and Schaffer, 1980, J. Virol. 36:189-203), as well as autoregulation at the transcriptional level. Transient expression assays performed in the presence and absence of wild type ICP4 with ICP4 promoter-reporter gent chloramphenicol acetyl transferase [CAT]) constructs implicate ICP4 as a negative regulator of its own expression (for example, see DeLuca and Schaffer, 1985, Mol. Cel.

Biol. 5: 1997-2008). Additionally, cell lines transfected with HSV-1 strains housing temperature sensitive or deletion mutants of ICP4 fail to induce early and late HSV-1 expression, yet overproduce ICP4 (DeLuca, et. al., 1985, J. Virol. 56: 558-570, Dixon and Schaffer, 1980, J. Virol. 36: 189-203).

The ICP4 gene has been dissected via mutation analysis so as to locate and define functional domains. DeLuca and Schaffer (1987, Nucleic Acids Res. 11: 4491-4511) generated nonsense mutations by inserting synthetic oligonucleotides in all three reading frames throughout the ICP4 coding region. Domains involved in both transactivation and ii 1 r! i! 1T C' 'f

I

_L i WO 95/02071 PCT/US94/07"781 -3autoregulation domain were found to reside within the amino terminal 60% of the ICP4 gene as shown by an ability to induce an early gene promoter CAT chimeric construction and an ability to negatively regulate an ICP4-CAT chimeric construction, respectively. A number of additional nonsense and deletion mutations were introduced into both ICP4 genomic copies of HSV-1 such that the resulting mutants expressed only defined subsets of the wild-type ICP4 amino acid sequence. Regions involved in transactivation and autoregulation domain were reconfirmed, while a DNA binding domain was localized within the amino terminal portion of the ICP4 gene (DeLuca and Schaffer, 1988, J. Virol. 62: 732-743).

The ICP4 coding region has been dissected further to assign functions to the multiple discrete domains of the protein. The DNA-binding capacity of ICP4 was further localized to amino acid residues between positions 445 and 487 (Shepard, et al., 1989, J.

Virol. 63: 3714-3728; Paterson and Everett, 1988, Virology 166: 186-191). Amino acid residues responsible for transactivation were further localized to an amino acid domain between positions 143 and 210 (Shepard, et al., 1989, J. Virol. 63: 3714-3728), or possibly from positions 275 to 490 and positions 840 to 1100 (Paterson and Everett, 1988, Virology 166: 186-191). Regions involved in dimerization, phosphorylation and nuclear localization have also been delineated (DeLuca and Schaffer, 1988, J. Virol. 62: 732-743; Shepard, et al., 1989, J. Virol. 63: 3714-3728; Wu and Wilcox, 1990, Nucleic Acids Res. 18: 531-538).

An ICP4 deletion mutant was constructed which retained primary structural domains for DNA binding and autoregulation, but lacks the domain required to confer transactivation (Shepard, et al., 1990, J. Virol. 64: 3916-3926). This ICP4 mutant, forms an ICP4 heterodimer containing one wild-type and one mutant ICP4 subunit. The ICP4 mutant subunit lacks both transactivation domains yet ret tins the ability to bind with DNA as well as retaining dimerization activity the ability to form a heterodimer with the wild type subunit). Heterodimer formation resulted in a dominant inhibitory phenotype with regard to HSV-1 growth.

ICP4 binds DNA non-specifically (Freeman and Powell, 1982, J. Virol. 44: 1084-1087) but it but also exhibits a preference for a consensus sequence. When this consensus sequence overlaps the start site of mRNA synthesis, ICP4 can inhibit transcription when provided in trans (Roberts, et al., 1988; J. Virol. 62: 4307-4320; U~ ~Is WO 95/02071 PCT/US94/0/781 DeLuca and Schaffer, 1985, Mol. Cell. Biol. 5: 1997-2008; O'Hare and Hayward, 1987, J. Virol. 56: 723-733). However, in cases where ICP4 serves to induce a gene, deletion of sites for which ICP4 shows relatively high affinity does not affect activation by ICP4 (Smiley, et al., 1992, J. Virl. 66: 623-631). In fact, exhaustive studies have failed to reveal evidence supporting the existence of ICP4-specific induction sequences (Coen, et al., 1986, Science 234: 53-59; Eisenberg, et al., 1985, Mol. Cell. Biol. 5: 1940-1947).

Additionally, mutant ICP4 mutant molecules have been isolated which exhibit reduced affinity for DNA in vitro, but which still retain the ability to stimulate gene expression in the context of the infected cell (Imbalzano, et al., 1990, J. Virol. 64: 565-574; Shepard and DeLuca, 1991, J. Virol. 65: 787-794). Despite these observations, a correlation exists between ICP4 mutants that can no longer bind DNA ?nd the ablation of its transactivation function (Shepard, et al., 1989, J. Virol. 63: 3714-3728; Paterson and Everett, 1988, Virology 166: 186-196).

Studies have also focused on the interactive role of cis-acting promoter elements and ICP4 in the transactivation of early and late viral genes. One early viral gene, thymidine kinase is transactivated by ICP4 (DeLuca, et al., 19'4, J. Virol. 52: 767-776; DeLuca and Schaffer, 1985, Mol. Cell. Biol. 5: 1997-2008; O'Hare and Hayward, 1985, J. Virol. 53: 751-760). Thymidine kinase is transcribed by RNA polymerase II and contains a CCAAT box and two Spl sites upstream of a TATA box (Jones, et al., 1985, Cell 42: 559-572; McKnight and Kingsbury, 1982, Science 217: 316-324). Studies of the tk promoter indicate that only cis sequences interacting with cellular transcription factors are required for expression during viral infection, no induction-specific sequences have been identified (Boni and Coen, 1989, J. Virol. 63: 4008-4092; Coen, et al., 1986, Science 234: 53-59; Eisenberg, et al., 1985, Mol. Cell.

Biol. 5: 1940-947). Imbalzano, et al. (1991, J. Virol. 65: 565-574) determined that induction by ICP4 of the tk promoter was not dependent on the integrity of these cis sequences. Efficient induction by ICP4 was observed in the presence of only the tk TATA box.

The general transcription factor (GTF's): TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJ assemble on the DNA template in a very defined and ordered fashion along with RNA pol merase II during transcription initiation. The first factors to assemble are TFIID, TFIIA and TFIIB (Buratowski, et al., 1989, Cell 56: 549-561; ii 1 (r-e 1 hl:'~"iF WO 95/02071 PCT/US94/07781 Maldonado, et al., 1990, Mol. and Cell Biol. 10: 6335-6347). TFIID consists of many different proteins nucleated in a complex containing the TATA-binding protein (TBP), a 43 kD protein that specifically binds TATA boxes (Kao, et al., 1990, Science 248: 1646-1650; Peterson, et al., 1990, Science 248: 1625-1630; Hoffmann, et al., 1990, Nature 346: 387-390). TFIIB consists of a single protein (33kD) which binds to TFIID as well as TBP to form a DNA-protein complex termed "DB" (Ha, et al., 1991, Nature 352: 689-695; Peterson, et al., 1990, Science 248: 1625-1630). TFIIA is known to participate in the formation of "DA" and "DAB" complexes.

Imbalzano and DeLuca (1992, J. Virol. 66: 5453-5463) studied the effects of ICP4 mediated induction of two different TATA box sequences that possess different affinities for TBP a TATA sequence from an early (tk) and late (gC) HSV-1 promoter).

Chimeric promoters combining the identified cis elements (CCAAT and Spl elements) of the early tk promoter and late gC promoter were constructed and recombined in place of the wild-type tk promoter in an ICP deficient virus. Therefore, it was possible to measure tk expression under control of either the early gene tk TATA box or the late gene gC TATA box in the presence or absence of the upstream elements, CCAAT or Spl. The authors concluded that the level of inductic n of a promoter by ICP4 can be affected by the presence of specific binding sites for cellular transcription factors.

Adding the Spl and CCAAT sequences to the promoter region containing either a tk or gC TATA box results in a greater increase in uninduced than induced expression.

Therefore, a promoter containing only a TATA box is induced by ICP4 to a greater extent than a promoter containing a TATA box and upstream cis acting elements such as a CCAAT or Spl element. Additionally, the induction by ICP4 is influenced inversely by the apparent affinity of the TATA box lor TBP.

The literature has documented that ICP4 is an essential immediate-early HSV-1 viral protein that is involved, among other things, in the transactivation of early and late HSV-1 viral genes. Certain ICP4 mutants have shown an ability to form a heterodimer with wild-type ICP4 in vitro, which may or may not effect the ability of ICP4 to induce early or late gene expression. In addition, studies indicate a relation between induction by ICP4 and the apparent affinity of a specific TATA box for cellular transcription factors. A better understanding of the required interaction of essential viral proteins such as ICP4 with host cellular transcription factors to control HSV-1 gene expression might -uc-- 1 i.

WO 95/020 PCT/US94/07781 lead to the development of a useful assay to select antiviral peptides. Such antiviral compounds may provide favorable alternatives to presently available treatments.

3. SUMMARY OF THE INVENTION The present invention relates to methods of screening for and selecting compounds or peptides which possess antiviral activity. The invention is based on the interaction between viral and host cellular factors acting in trans to regulate expression of essential viral genes subsequent to host infection. A wild-type viral protein and at least one host transcription factor and a nucleic acid cis-acting sequence are incubated with a potential antiviral compound or peptide in conditions conducive to formation of a protein:protein and/or DNA:protein transcription complex. A potential antiviral agent forms unique complexes in comparison to protein:protein and DNA:protein interactions involving the wild type-viral protein.

In the present invention, the potential antiviral compound or peptide can be any inorganic or organic compound or peptide sequence able to interfere with formation of a wild-type viral-host transcription complex. For example, the peptides may be mutants of a wild-type protein or biologically active fragments thereof involved in transactivation of a viral gene. By way of example and not of limitation, mutant derivatives generated for screening may be chosen at random by site-directed mutagenesis of the gene encoding the target protein, by a rational peptide design mechanism, or by utilization of a microbial based selection scheme to generate a pool of mutants for entry into the antiviral screening method of the present invention. Antibodies to essential surface regions of a wild-type viral transacting factor may also be candidates for entry into the in vitro assay of the present invention. Additionally, potential antiviral compounds may be naturally occurring or synthetic compounds or a population of such compounds, any or all of which may be candidates for entry into the antiviral screening assay of the present invention. All of these, and additional methods of generating screening candidates, are available to the skilled artisan and have been described in the literature. For example; de;etion, addition, nonsense or point mutations as well as biologically active fragments of a wild-type viral protein with trans-acting capability, such as ICP4, is a candidate for antiviral screening.

The present invention provides in part for formation of a general viral-host transcription complex by mixing one or more known cellular transcription factors prior to t-I..

*I

WO 95/02071 PCT/US94/07781 -7or simultaneously upon incubation with a potential antiviral peptide and nucleic acid promoter fragment. The host transcription complex may comprise, but is not solely limited to, at least one the eukaryotic transcription facors selected from the group consisting of IFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIU.

The skilled artisan will be prompted by the teachings of this disclosure to utilize any of one or a combination thereof of host transcriptions factor(s) interacting to form a protein:protein complex with the wild type viral trans-acting factor(s); this protein:protein complex prompting formation of an active DNA:protein transcription complex in the presence of an appropriate cis-containing nucleic acid sequence. The gel mobility of such an in vitro reconstituted DNA:protein transcription complex will provide the standard for comparison in assays adding a potential antiviral compound or peptide.

The present invention also provides for a nucleic acid promoter sequence containing cis sequences binding a viral-host protein:protein complex. One skilled in the art is directed by this disclosure to choose a nucleic acid fragment known to interact with the protein:protein complex in vivo or in an in vitro reconstitution assay. An appropriate viral promoter would be a nucleic acid fragment containing cis sequences that bind the viral-host transcription complex. Alternatively, a synthetic nucleic acid fragment containing an appropriate cis sequence, such as a TATA sequence and/or an ICP4-binding site, can comprise a portion of the nucleic acid sequence in the assay. For example, ICP4 is essential for expression of early and late HSV-1 gene expression. Therefore, the nucleic acid fragment chosen for use in the assay contains cis sequences known to bind the ICP4-host transcription complex. It is against the protein:protein and DNA:protein binding characteristic in a control assay that potential antiviral peptides are compared and detected as inhibitors of wild-type viral gene expression, and hence, as antiviral S 25 compounds.

In one embodiment of the invention, methods are utilized to screen and select for compounds or peptides which possess antiviral activity against herpes simplex virus

(HSV).

4 In an embodiment of the invention regarding screening and selecting compounds or peptides which possess antiviral activity against HSV, the wild-type trans-acting factor is Si a trans-acting factor targeted for inhibition chosen from the group consisting of the immediate-early genes of HSV, including but not limited to ICP4, ICPO and ICP27. L nn~o~ WO 95/02071 PCT/US94/07781 In an embodiment of the invention regarding screening and selecting compounds or peptides which possess antiviral activity against HSV, the wild-type trans-acting factor targeted for inhibition is ICP4. Deletion, addition, nonsense and point mutations of ICP4, as well as generation of biologically active fragments of ICP4 are candidates for antiviral screening.

In an embodiment of the invention regarding targeting of ICP4 for inhibition, the nucleic acid sequence utilized in the in vitro reconstitution assay is any HSV gene regulated by ICP4, including but not limited to regulatory sequences involved in controlling self-expression of ICP4, expression of the HSV early gene, thymidine kinase, or the late HSV gene, gC.

In an embodiment of the present invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ICP4, the host transcription complex includes, but is not solely limited to, at least one of the eukaryotic general transcription factors selected from the group consisting of TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJ.

In a preferred embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ICP4, the host transcription complex includes TFIIB and TFIID, which form the eukaryotic host transcription complex for use in the in vitro reconstitution assay.

In an preferred embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ICP4, the host transcription complex comprises the TATA Binding Protein (TBP) and TFIIB to form the general transcription complex for use in the in vitro reconstitution assay. This preferred embodiment of the invention is based on the present disclosure that regulation of ICP4 gene expression is correlated with its ability to form a tripartite complex with the two general transcription factors, TFIID and TBP. The formation of this tripartite complex allows for the selection of antiviral compounds and peptides which interfere with the formation of this essential complex and its ability to act in trans with the ICP4 regulatory sequence.

In an additional embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling expression of ICP4, I.i-~y C3" $I~s -la~l: oi~ir~ WO 95/02071 PCT/US94/07781 the host transcription complex comprises use of either TBP and TFIIB alone as the host transcription factor for use in the in vitro reconstitution assay.

In another embodiment of the present invention which targets ICP4 for inhibition, the nucleic acid sequence utilized is a fragment from the regulatory region controlling expression of thymidine kinase. Any of the embodiments disclosed in this specification in regard to utilizing a DNA fragment from the ICP4 regulatory region may be utilized with the HSV thymidine kinase promoter.

In another embodiment of the present invention which targets ICP4 for inhibition, the nucleic acid sequence utilized is a fragment from the regulatory region controlling expression of the late HSV gC promoter. Any of the embodiments disclosed in this specification in regard to utilizing a DNA fragment from the ICP4 regulatory region may be utilized with the late HSV gC promoter.

It is an object of the invention to provide a method of selecting antiviral compounds or peptides based on utilization of an in vitro reconstitution assay of viral and host trans-acting factors with a cis-sequence. The reconstitution assay selects for potential antiviral peptides by identifying peptides which promote protein:protein and/or DNA:protein complexes distinct from the documented wild-type formation of these transcription complexes.

It is an object of the invention to provide antiviral compounds or peptides as a therapeutic treatment against an HSV infection. A pharmaceutically effective amount of the antiviral peptide will be targeted to HSV infected cells of the patient such that formation of the viral-host mediated transcription complex is blocked, thus inhibiting further HSV gene expression and viral growth within the patient.

It is an object of the invention to provide antiviral compounds or peptides as a therapeutic treatment against an HSV infection. A pharmaceutically effective amount of the antiviral peptide will be targeted to HSV infected cells of the patient such that formation of the ICP4 mediated tripartite complex is blocked, thus inhibiting further HSV gene expression and viral growth within the patient.

This and other objects of the invention will be more fully understood from the following description of the invention, the figures, and the claims appended hereto.

i i 9_1 4V WO 95/02071 PCT/US94/07781 4. BRIEF DESCRIPTION OF THE FIGURES Figure 1 illustrates activities of rTFIIB and rTBP purified from E. coli. Part A depicts the activity of transcription factors disclosed in Example Section 6. Fractionated HeLa cell factors and purified recombinant rTBP and rTFIIB were combined to S reconstitute transcription from the adenovirus major late promoter in a G-less cassette assay as described in Example Section 6. Part B depicts a gel shift assay using rTBP and rTFIIB proteins. The 130 bp BamHI-EcoRI fragment spanning the ICP4 transcriptional site was used as probe. Part C is a diagram of the 130 bp probe derived .rom the ICP4 promoter. It contains an ICP4-binding site overlapping the start site of transcription, a TATA box with the sequence TATATGA, and a Spl site with the sequence GGGCGGG.

Figure 2 illustrates that rTFIIB helps rTBP bind the TATA box. Part A shows that the 130 bp ICP4 probe labeled on the coding strand was incubated with the indicated mixture of proteins, treated with DNase I and then run on a sequencing gel. Varying amounts of rTBP (400, 200, 100 and 50 ng) were reacted with the probe alone (lanes 3-6) or with the probe in combination with a constant amount of rTFIIB (0.3 ug, lanes 7-10).

As controls, the probe was incubated with buffer alone (lane 2) or with rTFIIB alone (lane 11). For sizing, a G-ladder of the probe was included (lane The position of the TATA box is indicated; the numbers refer to positions relative to the start site of transcription. Part B shows the same experiment as Part A performed using probe labeled on the non-coding strand.

Figure 3 shows that ICP4 helps DNA-binding of the DB complex. Part A depicts a DNase 1 protection assay as performed and described in Figure 2 using the 130 bp ICP4 probe labeled on the coding strand. Varying amounts of rTBP (100, 50, 25, and 12.5 ng) and a constant amount of rTFIIB (0.3 zg) were reacted with the probe alone (lanes 3-6) or with the probe in combination with a constant amount of ICP4 (0.25 14g, lanes 7-10). As controls, the probe was incubated with buffer alone (lane 2) or with rTFIIB alone (lane 11). For sizing, a G-ladder of the probe was included (lane The positions of the TATA box and of the ICP4 binding site are indicated; the numbers refer to positions relative to the start site of transcription. Part B depicts results from the same experiment as shown in Part A except the assay was performed using probe labeled on the non-coding strand. Part C depicts a DNase I protection assay using HeLa TFIID in place lc k +A 11 WO 95/02071 PCT/US94/07781 -11of iTBP. The same assay in part A was performed substituting various amounts of Hela TFIID (DB fraction) for rTBP.

Figure 4 depicts formation of a novel complex upon the simultaneous additions of TBP, TFIIB, and ICP4. In Part A ICP4 (60 ng), rTFIIB (0.5 ug), and rTBP (50 ng) were used in the indicted combinations. In Part B ICP4, rTFIIB and rTBP were used with the indicated antibodies The 58S antibody reacts with ICP4 near the carboxy terminus (DeLuca and Schaffer, 1988, J. Virol. 62: 732-743). Part C shows that the tripartite complex and the DB complex respond similariy in response to decreasing concentrations of rTFIIB. Constant amounts of rTBP (60 ng) and ICP4 (12 ng) were titrated against varying amounts of rTFIIB 0.25, 0.125, 0.063, and 0.032 ig) in a standard gel shift assay. At low relative concentrations of ICP4, the inclusion of TFIIB in the binding reaction reproducibly enhanced the binding of ICP4; this is shown in the third lane from the right, in which 125 ng of TFIIB was added along with 10 ng of ICP4.

Figure 5 shows DNase I footprinting of the complexes isolated by gel shift.

DNase I footprinting of the complexes isolated by gel shift was performed as described in the Example section. Part A represents the preparative gels run to isolate the indicated protein-DNA complexes. DNA from these complexes were run on the 8% sequencing gel shown in Part B. On the extreme right of Part B is the G-ladder used to size the DNase I cleavage products; the accompanying numbers arc relative to the transcriptional start site.

Also indicated are the TATA homology and the sequence ATCGTC which has been shown to be important for specific DNA-binding by the ICP4 protein.

Figure 6 shows that ICP4 DNA binding alone is not sufficient to form the tripartite complex. Part A shows a gel shift assay in which about 12 ng of purified ICP4, n208, X25, nd3-8 or nd8-10 was used in conjunction with the indicated proteins in a standard gel shift assay using the 130 bp BamHI-EcoRI fragment spanning the ICP4 transcriptional start site. For rTBP, 60 ng was used; for rTFIIB, 250 ng was used. The intensity of this complex was used to obtain the value for the DB4 complex in Table 2.

Part B is a diagram of ICP4 and the mutant proteins along with their associated activities.

The domains of ICP4 are those described previously.

Figure 7 shows that ICP4 forms a tripartite complex with rTBP and rTFIIB on the thymidine kinase (TK) promoter, an inducible template. Constant amounts of ICP4 (12 ng) and rTBP (0.06 ig) were titrated against varying amounts of rTFIIB. On the i 6 i-.

i -t ::a -ij U~ r---a~cix~L.~

Y-

Yi

A

Nov 94 A.P.T. Patent and Trade Mark Attorneys IV WO 95/02071 PCT/US94/07781 -12left, the EcoRI-BglII fragment spanning the TK transcriptional start site was used as probe. On the right, the same fragment but bearing a mutated TATA box (LS -18/-29) was used as probe. "58S" refers to an anti-1CP4 monoclonal antibody.

Figure 8 is a model of how ICP4 interacts with its own promoter and with the TK promoter. The arrows indicate interactions as identified either by DNase I footprinting assays or by cooperativity in DNA-binding. The arrows beneath the templates indicate the start sites of transcription. The checkered boxes represent TATA boxes. The other boxes depict ICP4 binding sites. High affinity ICP4 binding sites positioning ICP4 over the start of transcription may result ii repression, whereas ICP4's potential interaction with many lower afftiity sites throughout the genome may result in transactivation by recruiting the GTFs without tightly binding to the start of transcription.

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to methods of screening for and selecting compounds or peptides which possess antiviral activity. The invention is based on the interaction between viral and host cellular factors acting in trans to regulate expression of essential viral genes subsequent to host infection. A wild-type viral protein and at least one host transcription factor and a nucleic acid cis-containing sequence are incubated with a potential antiviral compound or peptide in conditions conducive to formation of a protein:protein and/or DNA:protein transcription complex. A potential antiviral agent any such compound or peptide) forms unique complexes in comparison to protein:protein and DNA:protein interactions involving the wild type-viral protein in the absence of the exogenous agent. The in vitro reconstitution assay is based on a potential antiviral agent forming unique complexes in comparison to interactions involving a wild type viral protein. Pools of potential antiviral compounds or peptides can be screened using the in vitro reconstitution assay disclosed in the present invention. Potential antiviral compounds or peptides generated from this screening process may then be subjected to additional assays confirming antiviral activity and selecting for those with the greatest therapeutic potential.

In the present invention, the potential antiviral peptide can be any peptide sequence able to interact with a viral-host transcription complex. The peptide may be a mutant of a wild-type protein or biologically active fragments thereof. For example; deletion, 61 ::I i:F: e

I

5 i I i: i i;i;i:i ii-i ii WO 95/02071 PCT/US94/07781 -13addition, nonsense or point mutations as well as biologically active fragments of a wild-type viral protein with trans-acting capability is a candidate for antiviral screening.

For example, in a specific embodiment of the invention, addition, deletion, nonsense or point mutations a wA ll as biologically active peptide fragments of the wild type viral protein ICP4 are candidates for antiviral screening. One such series of potential antiviral peptides may comprise ICP4 transactivation domains. Potential antiviral peptides may be generated for presentation in the selection scheme of the present invention by any one of a number of methods described previously in the literature. For example, peptides presented for antiviral peptide screening may be chosen as a result of a rational design mechanism to generate one or several potential antiviral peptide; or by mutagenesis of cloned DNA, the cloned DNA sequence encoding the target wild-type gene or a portion thereof. Any number of DNA mutagenesis techniques which have been described in detail in available literature may be utilized (see Ausabel, et al, 1991, Current Protocols in Molecular Biology, "Mutagenesis of Cloned DNA":Chapter Additionally, potential antiviral compounds may be naturally occurring or synthetic compounds or a population of such compounds, any or all of which may be candidates for entry into the antiviral screening assay of the present invention. All of these, and additional methods of generating screening candidates are available to the skilled artisan.

The present invention provides in part for formation of a general viral-host transcription complex by mixing one or more known cellular transcription factors prior to or simultaneouly upon incubation with a potential antiviral compound or peptide -nd nucleic acid promoter fragment. The host transcription complex may comprise, but is not solely limited to, at least one of the eukaryotic general transcription factors selected from the group consisting of TFIID, TFIIA, TFIB, TFIIF, TFIIE, TFIIH and TFIIJ.

Proteinaceous components of the general transcription complex may be isolated and substantially purified for use in the assay by any means known to one skilled in the art. Methods of isolation include, but are not limited to, isolation and purification of the protein directly from its in vivo source. Alternatively, the gene encoding the protein of interest may be cloned into an expression vehicle (such as an expression plasmid) via recombinant DNA techniques and transformed into a recombinant host cell. The recombinant host cell may be prokaryotic or eukaryotic, including but not limited to bacteria, yeast, mamm~iian cells including but not limited to cel lines of human, bovine, Sj1l Sii t

I

1 h r n W: r r-rr U r I^ WO 95/02071 PCT/US94/07781 -14porcine, monkey and rodent origin, and insect cells including but not limited to drosophila derived cell lines. Cell lines derived from mammalian species which may be suitable and which are commercially available, include but are not limited to, CV-1 (ATCC CCL70), COS-1 (ATCC CRL1650), COS-7 (ATCC CRL1651), CHO-KI (ATCC CCL61), 3T3 (ATCC CCL92), NIH/3T3 (ATCC CRL 1658), HeLa (ATCC CCL2), C1271 (ATCC CRL1616), BS-C-1 (ATCC CCL26) and MRC-5 (ATCC CCL171). The bacterial cell most used for expression of recombinant protein is Escherichia coli. There are various strains of E. coli available and are well known in the art.

The literature discloses a variety of expression vector host systems suitable for bacterial and fungal hosts, including plasmids, bacteriophages, cosmids and derivatives thereof. Examples of these systems in regard to bacterial expression are pBR322, pCR1, col Fl, phage lambda, M13 and filamentous viruses.

The expression vector may be introduced inro host cells via any one of a number of techniques including but not limited to transformation, transfection, protoplast fusion, and electroporation.

In a specific embodiment of the present invention, human TATA binding protein (hTBP) is obtained by the TBP expression plasmid, pETHIID, transformed into the E.

coli strain BL21; and recombinant TFIIB is obtained by the TFIIB expression plasmid, phIIB, also transformed into the E. coli strain BL21.

The skilled artisan will be prompted by the teachings of this disclosure to utilize any of one or a combination thereof of host transcriptions factor(s) interacting to form a protein:protein complex with the wild-type viral trans-acting factor(s); this protein:protein complex prompting formation of an active DNA:protein transcription complex in the presence of an appropriate nucleic acid sequence. The gel mobility of such an in vitro reconstituted DNA:protein transcription complex will provide the standard for comparison in assays substituting potential antiviral compounds for the wild type viral trans-acting factor(s).

Presented as an example and not a limitation of the invention, an in vitro reconstitution assay is devised so as to select for compounds or peptides which interfere with the ICP4-host nuclear transcription complex. Any compound or peptide which interferes with this wild type characteristic is by definition an antiviral peptide by way of inhibiting the role of ICP4 as a transactivator of viral gene expression. This is so because L. Mol. Biol. 180: 1-9) to yield infected cell polypeptides 0, 4, 22, 27 and 47 (Periera, ct al., 1977, Virology 27: 733-749). Transcription of these five genes does not require 1.

WO 95/02071 PCT/US94/07781 a loss of ICP4 induced transactivating activity for ICP4 equates to an inability for continued viral growth. By way of example, and not of limitation, a deletion mutant of ICP4 was tested as an antiviral compound. Incubation of ICP4 with general transcription factors TFIIB and TBP in the presence of an ICP4 promoter fragment spanning the TATA box results in a "super shift" as detected in a gel retardation assay in comparison to the mobility of the general transcription complex without ICP4. If a deletion mutant of ICP4, X25, is substituted for ICP4 in the gel retardation assay, no "super shift" is recorded. This indicates that the region deleted in X25 is important in formation of the supershift seen with wild-type ICP4. Therefore, any compound or peptide which mimics this region of ICP4 important in the wild-type supershift formation of the tripartite complex) will interfere with the wild type viral protein and hence, by definition, possess antiviral activity. Additional ICP4 mutants tested for ability to alter the wild-type super shift, as described in Example Section 6, are n208, nd3-8, nd8-10 and d8-10.

The present invention also provides for a nucleic acid promoter sequence containing cis sequences binding a viral-host protein:protein complex. The nucleic acid fragment of choice is one known to interact with the protein-protein complex in vivo or in an in vitro reconstitution assay. Alternatively, a synthetic nucleic acid fragment containing an appropriate cis-acting sequence, such as a TATA sequence, can be utilized as the nucleic acid fragment in the assay. Once the wild-type protein has been chosen (for example, as described thn aghout this specification for ICP4) it is possible to select the appropriate nucleic acid fragment for use in the assay. The appropriate nucleic acid fragment will contain cis-acting sequences (such as a TATA box element) known to interact with the wild-type general transcription complex either in vivo or in an in vitro reconstitution experiment.

Again, by way of example and not of limitation, an assay which targets ICP4 as the wild-type viral protein would utilize any HSV-1 viral promoter fragment that ICP4 is known to interact with in vivo. Any such regulatory sequence or portion thereof which results in a stable viral-host transcription complex would be a potential template for use in the assay. Such a promoter fragment might be, by way of example and not by limitation, the promoter region upstream of the gene encoding ICP4, since ICP4 is known to interact with this promoter in the autoregulation its own expression. Another promoter might be, by way of example and not by limitation, the promoter region upstream of an early or Sli: C W !Ti r frames throughout the ICP4 coding region. Domains involved in both transactivation and frames throughout the ICP4 coding region. Domains involved in both transactivation and t I l ii: WO 95/02071 PCT/US94/07781 -16late HSV-1 gene known to be transactivated by ICP4. A specific example of such a promoter is the promoter fragment upstream of the early HSV-1 gene encoding thymidine kinase. Another specific example of such a promoter is the promoter fragment upstream of the late HSV-1 gene encoding gC. The skilled artisan will be able to choose alternative promoter fragments to those known to interact with ICP4 in vivo on the basis of sequence homology in regards to cis-acting elements known to interact with ICP4 mediated transcription complexes. These alternative promoter fragments may be purified from existing cloned DNA sequences or may be produced by synthetic means known to one skilled in the art. Therefore, the choice of a wild type target (such as ICP4) determines the strategy of choosing an appropriate promoter fragment, as well as the general transcription factors to be utilized in the assay. The general patterns of the wild-type protein:protein and DNA:protein interactions are then elucidated. By way of example, and not of limitation, the interaction may be detected through a gel retardation experiment and/or DNA footprint analysis, both of which are discussed in detail in Example Section 6. A potential antiviral compound or peptide is added to for the experimental assay once the characteristics have been determined for the basic wild type viral-host factor/nucleic acid fragment interaction. If this addition causes a different result by way of, for example, mobility in a gel retardation assay or a difference in the ability to bind to the nucleic acid fragment (whether that difference be location on the template or the strength of binding), the peptide or compound is singled out as a potential antiviral agent.

Therefore, one skilled in the art is directed by this disclosure to choose a nucleic acid fragment known to interact an essential surface of a wild-type viral protein so as to alter formation of a protein:protein or DNA:protein complex in vitro. An appropriate viral promoter would be a nucleic acid fragment containing cis sequences that bind a viral-host transcription complex. Alternatively, a synthetic nucleic acid fragment containing an appropriate cis sequence, such as a TATA sequence, can be utilized as the nucleic acid sequence in the assay. For example, ICP4 is essential for expression of early and late HSV-1 gene expression. Therefore, the nucleic acid fragment chosen for use in the assay contains cis sequences known to bind the ICP4-host transcription complex. It is against the protein:protein and DNA:protein binding characteristic in a control assay that 7 pf 1 rlr WO 95/02071 PCT/US94/07781 -17potential antiviral compounds or peptides are compared and detected as inhibitors of viral gene expression, and hence, as antiviral compounds.

In one embodiment of the invention, methods are utilized to screen and select for compounds or peptides which possess antiviral activity against herpes simplex virus

(HSV).

In an embodiment of the invention regarding screening and selecting compounds or peptides which possess antiviral activity against HSV, the wild type trans-acting factor is a trans-acting factor targeted for inhibition selected from the group consisting of ICP4, ICPO, and ICP27.

In an embodiment of the invention regarding screening and selecting compounds or peptides which possess antiviral activity against HSV, the wild-type trans-acting factor targeted for inhibition is ICP4. Deletion, addition, nonsense and point mutations of ICP4, as well as generation of biologically active fragments of ICP4 are candidates for antiviral screening. Additionally, chemical compounds which inhibit essential surface interactions between ICP4 and host transcription complexes or the formation of a competent DNA:protein complex are candidates for antiviral screening.

In an embodiment of the invention regarding targeting of ICP4 for inhibition, the nucleic acid sequence utilized in the in vitro reconstitution assay is an HSV gene regulated by ICP4, including but not limited to regulatory sequences involved in controlling self-expression of ICP4, expression of the early HSV gene, thymidine kinase, or expression of the late HSV gene, gC.

In an embodiment of the present invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ICP4, the host transcription complex includes, but is not solely limited to, at least one of the eukaryotic transcription factors selected from the group consisting of TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJU.

In a preferred embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ICP4, the host transcription complex includes TFIIB and TFIID, which form the eukaryotic host transcription complex for use in the in vitro reconstitution assay.

In an preferred embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling self-expression of ;i ii a c ,T X Cy

J~

U |aiong With KNA poij merase iI auring uranscnpion iniuauu. 111i.n i 1IIuia-LVI %j uw...

ble are TFIID, TFIIA and TFIIB (Buratowski, et al., 1989, Cell 56: 549-561; WO 95/02071 PCT/US94/07781 -18- ICP4, the host transcription complex comprises the TATA Binding Protein (TBP) and STFIIB to form the general transcription complex for use in the in vitro reconstitution Sassay. This preferred embodiment of the invention is based on the present disclosure that regulation of ICP4 gene expression is correlated with its ability to form a tripartite complex with the two general transcription factors, TFIID and TBP. The formation of this tripartite complex allows for the selection of antiviral compounds or peptides which interfere with the formation of this essential complex and its ability to act in trans with the ICP4 regulatory sequence.

In an additional embodiment of the invention which targets ICP4 for inhibition and utilizes a nucleic acid fragment from the regulatory region controlling expression of ICP4, the host transcription complex comprises use of either TBP and TFIIB alone as the host transcription factor for use in the in vitro reconstitution assay.

In another embodiment of the present invention which targets ICP4 for inhibition, the nucleic acid sequence utilized is a fragment from the regulatory region controlling expression of thymidine kinase. Any of the embodiments disclosed in this specification in regard to utilizing a DNA fragment from the ICP4 regulatory region may be utilized with the HSV thymidine kinase promoter.

In another embodiment of the present invention which targets ICP4 for inhibition, the nucleic ;i sequence utilized is a fragment from the regulatory region controlling expre -iut of the late HSV gC gene. Any of the embodiments disclosed in this specification in regard to utilizing a DNA fragment from the ICP4 regulatory region may bei utilized with the gC promoter.

a Therefore, potential antiviral agents can comprise any inorganic or organic compound, natural or synthetic, as well as any amino acid containing entity. The basis for inclusion as a potential antiviral agent is based in part on the ability to interact and effect an essential surface of the wild-type viral protein involved in transactivation of viral genes.

I gs I ji as ICP4. with host clluluiaru tansripi uion factoru iLo contrl HSV-1 gacneLU viiu LULIi SUgh as ICP4 with host cellular transcription factors to control HSV-1 gene expression might i n r i WO 95/02071 PCT/US94/07781 -19- 6. EXAMPLE: IN VITRO FORMATION OF A TRIPARTITE COMPLEX OF ICP4, TBP AND TFIIB 6.1. MATERIALS AND METHODS 6.1.1. PROTEINS Wild-type ICP4 was obtained from 108 Vero cells infected with HSV-1 (strain KOS). The truncated mutant, n208 (DeLuca and Schaffer, 1988, J. Virol. 62: 732-743) was obtained from n208-infected Vero cells. The mutant X25 was obtained following infection of X25 cells with KOS. X25 cells harbor the gene expressing the X25 mutant.

Since this gene retains the IE promoter/regulatory sequences, it can be induced to high levels by VP16 supplied by viral infection. X25 cells, their induction upon infection, and the subsequent purification of X25 homodimers were previously described (Shepard, et al.

1990, J. Virol. 64: 3916-3926). The viruses expressing the nd3-8 (n30-142, n774-1298) nd8-10 142-210, n774-1298), and d8-10 (C~142-210) ICP4 mutant proteins were constructed as described previously (DeLuca and Schaffer, 1988, J. Virol.

62: 732-743; Shepard, et al. 1990, J. Virol. 64: 3916-3926), using the previously characterized plasmids pndi3-i8, pndi8-il0, and pdi8-il0, respectively (Shepard, et al.

1990, J. Virol. 64: 3916-3926). General procedures for infection, harvesting of the cells, nuclear lysis, extraction, and subsequent chromatography to obtain purified ICP4 were as previously given (Shepard, et al. 1990, J. Virol. 64: 3916-3926; Imbalzano, et al. 1990, J. Virol. 64: 565-574; Shepard and DeLuca, 1991, J. Virol. 65: 299-307).

Human TATA binding protein (hTBP) was generated via the TBP expression plasmid, pETHIID, and the Escherichia coli strain BL21 (Kao et al., 1990, Science 248: 1646-1650). pETHIID-transformed BL21 cells were grown in 2 X YT medium 0.4% glucose in the presence of 50 /g/ml of ampicillin at 37 0 C until the A6 reached approximately 0.8. Isopropylthio-3-D-galactoside (IPTG) was added to a final concentration of 0.5 mM, and the cells were grown for 3 more hours at 30 0 C. The lysis of the cells and the fractionation of the extracts over DEAE-sephacel and heparin-sepharose was performed essentially as described (Kao et al.,1990, Science 248: 1646-1650; Imbalzano and DeLuca, 1992, J. Virol. 66:5453-5463). Fractions were assayed by gel shift with a 29 bp double stranded oligonucleotide encoding the adenovirus E1B TATA box, as described (Kao et al., 1990, Science 248: 1646-1650) and by immunoreactivity with a mouse anti-TFIID antibody provided by Dr. Robert Roeder,

I

s j! t nTe present mvenuon proviles in part ior iormauui ui a rcllulcu VIU-1ua transcription complex by mixing one or more known cellular transcription factors prior to r, *1 r II j, 1 ,y .f.i..ei~si^ WO 95/02071 PCT/US94/C7781 Rockefeller University. Relevant fractions were pooled and applied to a fast protein liquid chromatography (FPLC) mono Q column equilibrated with D 0.1 M KC1 at a rate of 1 ml/min. Buffer D was as previously described (Kao, et al., 1990, Science 248: 1646-1650). The TBP immunoreactive material present in the flow-through was applied to an FPLC Mono S column equilibrated with buffer D plus 0.1 M KCI at a rate of 1 ml/min. Bound proteins were eluted with a continuous 0.1 to 1.OM KCI gradient.

Immunoreactive fractions were frozen in liquid nitrogen and stored at -80 0

C.

Recombinant TFIIB was generated by the TFIIB expression plasmid, phIIB, obtained from Dr. Danny Reinberg and described previously (Ha, et al., 1991, Nature 353: 689-695). The procedures for bacterial expression, lysis and purification of TFIIB through the phosphocellulose column were as previously described (Ha, et al., 1991, Nature 353: 689-695 The TFIIB eluting from the phosphocellulose column at 0.5 M KCl was further fractionated on an FPLC superose 12 column.

General transcription factors were isolated via 80 liters of Hela cells grown in suspension which were used to make nuclear extracts by the method of Dignam, et al.

(1983, Nucleic Acids Res. 11: 1475-1489). The fractionation was similar to that of Reinberg and Roeder (1987, J. Biol. Chem. 262:3310-3321) as well as Samuels, et al., J. Biol. Chem. 257:14419-14427). The nuclear extracts were first fractionated through a phosphocellulose column to make 0.1 M KC1 flow-through 0.3 M K? I 0.5 M KCI and 1 M KC1 fractions. Subsequently, all fractions wer, lJjusted to 0.1 M KCI prior to their use in transcription or further fractionation. TFlA activity in fraction A was eluted from a DEAE-sephacel column with 0.3 M KCI b':ffer From fraction C, TFIIB, TFIIE/F, and RNA Pol II activities were cv aed from the DEAE-sephacel column with 0.1 M KCI (fraction CA), M KCI (fraction CB) and 1 M KCI (fraction CC), respectively. Fraction D, containing TFIID activity was loaded on a DEAE-sephacel column, and TFIID was eluted with 0.25 M KCI buffer (fraction DB).

The activity and relative requirements of the fractionated general transcription factors were assessed by performing transcription reactions using the adenovirus major later promoter hooked to the G-less cassette. In vitro transcription reactions contained 0.9 mg of template, 1 tl fraction AB (TFIIA), 4 ~l fraction CB (TFIIE/F), 2 1 fraction CC (Pol II), 4 Al fraction DB (TFIID), 0.017 /g rTFIIB, 15 units RNAse T1, 15 units RNasin, 7mM MgCl, 60 uM ATP and CTP, 25 AM UTP, 0.1 mM OMe GTP and 5 uCi

I

2 f a trans-acting factor targeted for inhibition chosen from the group consisting of the immediate-early genes of HSV, including but not limited to ICP4, ICPO and ICP27.

CK Bii WO 95/02071 PCT/US94/07781 -21and [a- 3 2 P] UTP. The transcription reactions were incubated at 30°C for one hour, and after phenol and chloroform extraction the RNA transcripts were separated from free nucleotides through a G-50 spin column. The RNA was then precipitated and resuspended in loading buffer and applied on a 4% denaturing rr ini gel.

6.1.2. GEL RETARDATION ASSAYS DNA-binding reactions were performed as follows. One nanogram of an end-labeled probe (3 x 104 to 6 x 104 cpm/ng) and the indicated mixture of proteins were incubated together at 30°C for 40 minutes in a buffer consisting of 10 mM N-2-hydroxyethylpiperzine-N'- 2-ethanesulfonic acid (HEPES, pH 5 mM ammonium sulfate, 8% (vol/vol) glycerol, 2% (wt/vol) polyethylene glycol 8000, 50 mM KC1, 5 mM B-mercaptoethanol, 0.2 mM ethylenediamine-tetraacetic acid (EDTA), and 25 jg of poly(dG)-poly(dC) per ml in a total volume of 30 The reactions were then electrophoretically separated at 200 V on a native 4% polyacrylamide gel containing TBE buffer (89 mM Tris, 89 mM boric acid, 0.1 mM EDTA, pH The gels were then dried and exposed to Kodak XAR-5 film with intensifying screens. Densitometric scans of the gel tracks were done by using a Hoefer scanning densitometer and compatible data processing software for the Macintosh.

6.1.3. STANDARD DNASE I FOOTPRINTING For footprinting, DNA-binding reactions were done as described in Example Section 6.1.2 except that the non-specific competitor was reduced to 0.06 xg/pl. After incubation, 30 !l of DNase I buffer (5 mM CaCI 2 and 10 mM MgC12) plus 2 J 1 of DNase I (5 ng) were added at room temperature. After 1 minute, 60 p 1 of stop buffer (0.2 M NaC1, 0.02 M EDTA, 1% (wh) SDS, 20 mg/ml tRNA, 1 mg/ml proteinase K) was added and allowed to incubate at 37 0 C for 10 minutes. The reactions were then extracted with phenol and chloroform, ethanol precipitated, resuspended in 95% formamide and run on a denaturing 8% polyacrylamide gel.

6.1.4. DNASE I FOOTPRINTING OF ISOLATED DNA:PROTEIN COMPLEXES In this case, the DNA-binding reactions were scaled-up 20 times. After the minute incubation, 300 /il of DNase I b-ffer plus 20 ,l of DNase I (50 ng) were added at room temperature. After 1 minute, 12 Al of 0.5 M EDTA was added. The reactions were then loaded and run o. a native TBE gel as described in Example Section 6.1.3. The wet gels were exposed overnight at 4 0 C and the resulting autoradiogram was used to cut out t4 I L 3:: 1* WO 95/02071 PCT/US94/07781 -22th- indicated bands. The bands were diced using a razor blade and the resulting pieces were put into an eppendorf tube containing 0.5 ,nls of 0.25 M ammonium acetate, 1 mM EDTA in order to elute the DNA. After an overnight incubation at 37 C, the eluted DNA was purified over a G-50 Sephadex spin column, ethanol precipitated and resuspended in 95 formamide. The samples were then appropriately normalized such that there were equivalent cpm per jl and then loaded on a 8% denaturing sequencing gel.

6.2. RESULTS 6.2.1. ACTIVITIES OF RECOMBINANT rTFIIB and rTBP As sources of TFIIB and TBP, human versions of these proteins were produced in bacteria from expression plasmids. SD -PAGE analysis of the preparations revealed that the proteins had been purified to near homogeneity and exhibited the appropriate molecular weights. To assess their functional integrity, two assays were performed. In the first (Figure 1A), the recombinant proteins were assayed for their ability to reconstitute an in vitro transcription system using the adenovirus major late promoter (AdMLP). It has been shown previously that recombinant TBP (rTBP) will substitute for bone-fide, mammalian TFIID when the AdMLP is used as substrate. Likewise, rTFIIB will substitute for mammalian TFIIB. The results in Figure 1A demonstrate that preparations of the recombinant proteins substitute for their natural analogs as well, thus demonstrating their functional integrity. In addition, it has also been show that TFIIB and rTBP interact to form a higher order complex with DNA substrates containing TATA boxes in gel retardation assays (Peterson et al., 1990, Science 248: 1625-1630). A 130 bp EcoRI-BamHI fragment spanning the ICP4 transcriptional start site was used as a probe. As diagrammed in Figure 1C, this fragment contains a TATA box and an iCP4 binding site. As shown in Figure 1B, rTFIIB failed to retard the probe to any significant extent when used alone. This was expected since it has been documented that rTFIIB lacks any specific DNA-binding capacity (Ha, et al. 1991, Nature 352: 689-695). In contrast to rTFIIB, rTBP did bind the probe when used alone, albeit at very low levels.

When rTBP and rTFIIB were assayed in combination, however, a striking amount of retarded product was obtained, presumably due to the formation of a DB complex as previously described (Peterson et al., 1990, Science 248: 1625-1630). To characterize this binding more fully, DNase I footprinting analysis was conducted (Figure In these i d

I

5~Y I I-.

t F-7 Ell i .r i M im i WO 95/02071 PCT/US94/07781 -23experiments, the 130 bp EcoRI-BamHI fragment was labeled on either the coding or noncoding strand, reacted with the indicated proteins, treated with enough DNase I to introduce on average a single nick per template, and then run on a denaturing polyacrylamide gel. While TFIIB alone did not result in any footprint, TBP produced a weak footprint on the TATA box at the intermediate TBP concentrations used. The further addition of more TBP resulted in widespread protection. By contrast, the inclusion of TFIIB in the titration series resulted in the protectinn of the TATA box at approximately 10-fold lower TBP concentratica and also the specific protection of sequences down toward the transcription initiation site. These effects are most clearly seen on the coding strand. These results indicate that rTFIIB significantly aids the DNA-binding potential of rTBP, presumably through the formation of a DB complex having enhanced affinity for the TATA homology. The results obtained in the footprinting experiment (Figure 2) indicates that the dissociation constant for TBP is approximately 10-fold lower in the presence of TFIIB.

6.2.2. ICP4 FACILITATES THE SPECIFIC DNA BINDING OF TBP CONTAINING COMPLEXES The probe used in the above experiments is derived from the ICP4 promoter and contains an ICP4 binding site located at the start of ICP4 transcription (Figure 1C). ICP4 was added to reactions similar to those that generated the footprints in Figure 2 to ascertain its effect on the formation of DB-TATA box complexes. In Figure 3A, rTFIIB and rTBP were reacted in the presence or absence of ICP4 with the 130 bp probe labeled on the coding strand. The amount of ICP4 used in this experiment is sufficient to give only a very weak footprint. Larger amounts of ICP4 results in a unique footprint extending form -10 to about +15 with respect to the start of transcription. This stretch contains the strong consensus site for specific binding by ICP4 (Faber and Wilcox, 1986, Nucleic Acids Res. 14: 6067-6083; Kristie and Roizman, 1986, Proc. Natl. Acad. Sci.

83: 3218-3222; Muller, 1987, J. Virol. 61: 858-865). More significantly, the inclusion of ICP4 enhanced the DB footprint as evidenced by the fact that at low concentrations of rTBP and rTFIIB where hardly any TATA-protection was detected in the absence of ICP4 a very strong footprint was obtained in the presence of ICP4. The enhancement was approximately 5-fold. The effect was reciprocal, the ICP4 footprint was enhanced by the DB complex, such that the region between -10 to +15 was clearly protected. To WO 95/02071 PCT/US94/07781 -24verify these results, DNase I footprinting was conducted using the non-coding strand (Figure 3B). Again, the assembly of the DB-TATA box was greatly facilitated by ICP4.

In order to assess whether these observations extend to a more physiologically relevant situation, human TFIID from HeLa cells was substituted for rTBP. Figure 3C shows the enhanced protection of the TATA box in the presence of ICP4 also occurred when hTFIID was used in place of rTBP. The demonstration of the same effect with the human TFIID fraction, which has many other molecular events described above may occur under conditions present in human cells, at least with respect to TFIID.

6.2.3. ICP4. TFIIB AND TBP FORM A TRIPARTITE COMPLEX ON DNA A gel retardation assay was used to confirm the existence of this tripartite complex. The assay involved reacting rTFIIB and rTBP, singly or in combination, in the presence of ICP4 and analyzing the products on native polyacrylamide gels to determine whether novel ICP4-containing complexes were formed. The EcoRI-BamHI fragment spanning the ICP4 transcriptional start site was used as a radiolabelled probe. The simultaneous use of iCP4. rTBP and rTFIIB resulted in a novel complex (DB4 in Figure 4) exhibiting a very low electrophoretic mobility, consistent with the supposition that it represents a tripartite complex containing ICP4, rTBP and rTFIIB. At the concentrations used, neither rTBP nor rTFIIB alone supershifted the ICP4-DNA complex. However, at very high protein concentrations, rTBP and rTFIIB alone were capable of altering the mobility of the ICP4-DNA complex. At relatively low concentrations used, the simultaneous addition of ICP4, rTBP and rTFIIB was necessary to generate a supershifted complex. Densitometric analysis of the data in Figure 4 confirmed the cooperativity in DNA binding seen in the previous DNase I protection assays. When used alone, ICP4 bound 9% of the available probe (Table When rTBP and rTFIIB were used together, 29% of the probe participated in DB complexes. When all three proteins were present, of the probe was taken up in tripartite complexes. If ICP4, rTBP and rTFIIB did not engage in cooperative interactions, then one would expect the latter value to be less than considering the fact that the probe was used in excess. Stated differently, if the binding of ICP4 and DB were independent events, then the proportion of the probe bound by both components would be the product of the proportions bound by the components used alone, The fact that the observed value was 15% (five times the value 'V expected of simple tripartite occupancy) reaffirms the conclusion that cooperative DNA 1 g (12 ng) and rTBP (0.06 fg) were titrated against varying amounts of rTFIIB. On the u 1 l I-

I

Fl WO 95/02071 PCT/US94/07781 binding is exhibited by ICP4, rTBP and rTFIIB when used in combination. Antibody supershifting experiments were performed to confirm the identities of the proteins participating in the DNA complexes identified in Figure 4A. In Figure 4B, a monoclonal antibody (58S) against ICP4 was used. As shown, the antibody supershifted only the two foremost retarded complexes, the two complexes identified as the DNA-ICP4 complex and the ICP4-rTBP-rTFIIB-DNA complex. However, when polyclonal antibodies against rTBP and rTFIIB were used, no supershifting was observed; not even the DB complex was affected. It is thus assumed that the antigenic epitopes of rTBP and rTFIIB are not available for antibody binding when the proteins are associated with DNA. This assumption is not groundless, given similar observations reported previously (Maldonado, et al., 1990, Mol. and Cell Biol. 10: 6335-6347).

TABLE 1. Complex-forming ability of TBP, TFIIB and ICP4 Protein(s) in reaction' TBP, TFIIB Wild-type ICP4 TFIIB, ICP4 TBP, ICP4 TBP,TFIIB, ICP4 Formation of probe in complexb DB ICP4 DB4 Unbound 29 9.0 82 10 84 10 84 15 3.0 15 58 I SPurified proteins added in each reaction to generate the data in Figure 4.

b The appropriate lane in Figure 4 was scanned with a densitometer, and the value shown represents the percentage of the total measured intensity present in each complex. DB, the complex labeled DB in Figure 4, having the mobility of that generated by the simultaneous addition of rTBP and rTFIIB; ICP4, the complex having the mobility of that generated by the sole addition of the purified ICP4 peptide; DB, the novel complex appearing only after the simultaneous addition of the ICP4 peptide, rTBP, and rTFIIB; unbound, the band having the same mobility as that generated by the probe run alone on the gel.

)i~'L ir"L Yaoe to interact with a viral-host transcription complex. The peptide may be a mutant of a wild-type protein or biologically active fragments thereof. For example; deletion,

'IT

A

I

1 l" a -~YLI~LPl~~iii-LL'i i' _I-~_IXLI WO 95/02071 PCT/US94/07781 The importance of TFIIB in the formation of the tripartite complex and most likely its participation was implied by the observation that in the absence of TFIIB, tripartite complexes did not form (Figure 4A). Figure 4C shows that the concentration dependence of TFIIB for tripartite complex formation is similar to that for DB complex formation.

This finding further highlights the importance of TFIIB for tripartite complex formation and also indicates that the tripartite complex can form at physiologically relevant concentrations of cofactors insofar as the in vitro assembly of the DB complex reflects a physiologically relevant phenomenon.

DNase I footprint analysis was conducted on the individual DNA-protein complexes to further identify the proteins and nucleic acid sequences involved (Figure Preparative gel retardation reactions were performed with the protein mixtures as indicated for Figure 5A and the 130 bp probe labeled on the coding strand. The reactions were treated with DNase I, run on native TBE gels, and exposed to film. The bands marked A through F were cut out, and their DNA was eluted. The eluted DNA was then run on a denaturing polyacrylamide gel to determine which sequences were protected from DNase I cleavage. DNA from band A represents nonretarded probe and was susceptible to DNase I cleavage throughout its entire length. DNA from band B represents probe complexed with ICP4 alone. In this case, sequences from about -10 to at least +10 with respect to the transcriptional start site were protected. This stretch includes the sequence ATCGTC known to be involved in ICP4-DNA interactions (Faber and Wilcox, 1986, Nucleic Acids Res. 14: 6067-6083). DNA from band D represents retarded probe resulting after reaction with ICP4 and rTBP. In this case, not only was the ICP4 footprint obtained, but also some very weak protection, covering the area extending from about -29 to -18, an area that contains the TATA box, was observed.

This result indicates that a low level of co-occupancy of the DNA template by rTBP and ICP4. DNA from band C represents retarded probe resulting after reaction with ICP4 and rTFIIB. In this case, the cleavage pattern was nearly identical to the pattern observed with ICP4 alone. DNA from band F represents the retarded probe resulting from adding rTBP and rTFIIB. The resulting DNase I profile demonstrated that sequences around -36 to -18 were protected in the complex. This stretch contains the TATA homology. DNA from band E represents the supershifted probe resulting from the reaction of ICP4, rTBP and rTFIIB altogether. In this case, both the TATA box and the _j 1 WO 95/02071 PCT/US94/07781 -27- ICP4 biding site were strongly protected, consistent with the supposition that the supershifted complex disclosed in Figure 4A represents a high-order complex containing ICP4, rTBP, and probably rTFIIB.

6.2.4. ICP4 DNA BINDING IS NOT SUFFICIENT TO FORM THE TRIPARTITE COMPLEX A set of ICP4 mutant proteins purified from mutant virus-infected cells were assayed for the ability to form tripartite complexes (Figure 6A). The primary structures of these proteins are shown relative to that of the wild type protein in Figure 6B. All five mutant proteins contain the DNA-binding domain, t .nd DNA and yield wild type footprints at least when assayed on the ICP4 promoter (Shepard and DeLuca, 1991, J.

Virol. 65: 299-307). Densitometry data from Figure 6A are shown in Table 2. For wild type ICP4, n208 and nd3-8, the proportion of the probe found in tripartite complexes was significantly greater that the product of the proportions bound by DB and the ICP4 proteins when assayed individually. As previously discussed, this finding is indicative of cooperative binding. In contrast, for X-25, nd8-10 and d8-10, the proportions of the probe bound in novel complexes were substantially reduced and approximately equal to the products of the proportions bound by DB and the mutant proteins alone. As hereinbefore discussed, this finding indicates simple tripartite occupancy of the probe resulting from independent binding events. With respect to mapping the region of ICP4 that is involved in the interaction with the DB complex, the following reasoning points to the region extending from amino acids 142 to 210. The failure of X25 to interact cooperatively implicates the region from 30 to 274. The wild type ability of nd3-8 eliminates amino acids 30 to 142 from consideration while implicating amino acids 142 to 274. The failure of nd8-10 and d8-10 discounts amino acids 210 to 274 as being necessary. Therefore, the residues residing between 142 to 210 contributes to the ability of ICP4 to cooperatively interact with the DB complex.

I i

*L

WO 95/02071 PCT/US94/07781 -28- 6.2.5. ICP4 FORMS A TRIPARTITE COMPLEX WITH rTBP AND rTFIIB ON THE THYMIDINE KINASE (TK) PROMOTER. AN INDUCIBLE TEMPLATE In gel retardation experiments of Section the BamHI-EcoRI fragment spanning the ICP4 transcriptional start site was used as probe. This fragment contains an ICP4 binding site overlapping the region where transcription initiates. The binding of ICP4 to this site is involved in repression of transcription. The EcoRI-BglII DNA fragment spanning the transcriptional start site of the inducible thymidine kinase promoter (TK) was chosen to determine if ICP4 could form a tripartite complex using a template whose transcription is induced by ICP4. This fragment, spanning -75 to +54 with respect to the start of transcription, contains a TATA box, an Spl site, and a low affinity ICP4 binding site located approximately 40 base pairs downstream of the transcriptional start site (Imbalzano et al, 1990, J. Virol. 64: 2620-2631). In Figure 7A, ICP4 alone retards the probe and the amount of complex 4 is increased by the addition of TFIIB.

The simultaneous addition of ICP4, TBP and TFIIB resulted in a complex of even lower mobility than complex 4 (marked with open circles). The anti-ICP4 monoclonal antibody 58S further retarded the novel complex formed in the presence of rTFIIB, rTBP and ICP4 (marked with a dash and open circle). The experiment in Figure 7B recapitulates the one in Figure 7A expect that one linker-scanning mutant of the probe was used. This mutant, designated LS-29/-18, harbors a defective TATA box (McNight and Kingsbury, 1982, Science 217: 316-324). This mutation reduced the ability of rTFIIB and rTBP to form the rTFIIB-rTBP complex. In addition, it abolished the ability of rTBP and rTFIIB to cause a supershifting of the ICP4-DNA complex. The result underscores the point that the observed supershifting in a phenomenon requiring the direct participation of rTBP.

Therefore, these experiments indicate that tripartite complexes can form on templates that are either induced or repressed by ICP4 and that this ability does not require the immediate juxtaposition of the TATA box and the ICP4 binding site.

I

.1 l^ with this promoter in the autoregulation its own expression. Another promoter might be, by way of example and not by limitation, the promoter region upstream of an early or "8k e;' a-

-I

WO 95/02071 PCT/US94/07781 -29- TABLE 2. Complex-forming ability of mutant ICP4 proteins Protein(s) in reaction" TBP, TFIIB Wild-type ICP4 TBP, TFIIB,ICP4 n208 TBP, TFIIB, n208 nd3-8 TBP, TFIIB, nd3-8 TBP, TFIIB, X25 nd8-10 TBP, TFIIB, nd8-10 d810 TBP, TFIIB, d8-10 Formation of probe in complexb ICP4 4.2 8.9 5.3 8.8 2.9 8.9 35 3.0 18 34 4.3 34 6.8 12 11 1.0 4.9 2.0 DB4 Unbound 76 18 67 82 19 69 61 47 34 1.8 53 87 1.7 74 93 0.3 86 Purified proteins added in each reaction to generate the data in Figure 6A. The ICP4 peptides are diagrammed in Figure 6B.

b Each lane in Figure 6A was scanned with a densitometer, and the value shown represents the percentage of the total measured intensity present in each complex. DB, the complex labeled DB in Figure 4 and BD in Figure 6A, having the mobility of that generated by the simultaneous addition of rTBP and rTFIIB; ICP4, the complex having the mobility of that generated by the sole addition of the purified ICP4 peptide; DB, the novel complex appearing only after the simultaneous addition of theICP4 peptide, rTBP, and rTFIIB; unbound, the band having the same mobility as that generated by the probe run alone on the gel.

ii -4

I

A

K3^.

nb1 r-M :i:u: B t- WO 95/02071 PCT/US94/07781 6.3. DISCUSSION The assembly of an active transcription complex on TATA-containing promoters begins with the binding of TFIID, TFIIB and TFIIA. The data presented throughout Example Section 6 show that ICP4 can participate in the formation of transcriptional complexes at this very early stage of assembly. The simultaneous addition of ICP4, rTFIIB and rTBP to a DNA template containing an ICP4 binding site and a TATA box resulted in a unique complex lower in electrophoretic mobility than the complexes obtained with any of the proteins used singularly or in dual combinations. ICP4 resides in this unique complex, as evidenced by the fact that the complex exhibited a characteristic ICP4 footprint (Figure 5) and by the fact that it was further shifted by an anti-ICP4 antibody (Figure 4B and rTBP resides in the complex, as evidenced by the fact that the complex protected the TATA box from DNase I cleavage (Figure The question of whether rTFIIB resides in the complex is implicated by the fact that in its absence the supershift complex does not form (Figure 4A). In addition, its presence facilitated the formation of the tripartite complex and the DB complex with the same concentration dependence. Consequently, it is concluded that ICP4, rTBP and rTFIIB form tripartite complexes on suitable templates.

Both protein-protein and protein-DNA interactions mediate the formation of the novel complex. Two separate lines of evidence underscore the importance of protein-protein interactions. First, ICP4, rTBP and rTFIIB help each other bind DNA.

As shown in Figure 2, rTFIIB enhanced the DNA binding of rTBP by 10-fold. Figure 3 shows that ICP4 enhanced the DNA-binding potential of the rTFIIB-rTBP complex by the effect being reciprocal. Consequently, the combination of ICP4, rTBP and TFIIB results in a 50-fold stimulation of specific DNA binding. This cooperativity can best be interpreted as being a consequence of protein-protein interactions that translate into lower rates of dissociation from the DNA template. Similar phenomenon have been reported for the repressor protein of phage lambda and the large T antigen of SV40 (Deb and Tegtmeyer, 1987, J. Virol. 61: 3649-3654; Hochschild and Ptnashe, 1986, Cell 44: 681-687). The other line of evidence implicating the critical importance of protein-protein interactions is that fact that specific ICP4 mutants failed to cooperatively form the tripartite complex despite their ability to bind DNA. This result indicates that efficient assembly of the complex is not a consequence of simple tripartite occupancy of 7i 7 WO 95/02071 PCT/US94/07781 the DNA template but requires that specific protein-protein contacts be made between the participating components. As demonstrated in Figure 6, the ability of ICP4 to cooperatively form the tripartite complex maps between amino acids 142 and 274.

Evidence that specific protein-DNA interactions are involved in the assembly of the tripartite complex includes not only the footprint data but also the fact that a test DNA bearing a mutated TATA box fails to form the structure (Figure When the results of Figure 4 and Figure 7 are compared, it is evident that tripartite complexes form more efficiently on the fragment derived from the ICP4 promoter than on the fragment derived from the TK promoter. Two fa: t'is probably contribute to this difference. First, the binding site on the ICP4 promoter fragment exhibits a much higher affinity for ICP4 than does the site on the TK promoter fragment (Imbalzano, et al., 1990, J. Virol. 64: 2620-2631). Second, the distance between the ICP4 and TBP binding sites is different for the two promoters. Therefore, tripartite complexes may form with varying degrees of efficiency, depending on the relative affinities of the TATA box and ICP4 binding sites for their respective proteins and on the degree of separation between the participating cis-acting sites. Additionally, if the ICP4 binding site in the ICP4 promoter is mutated such that ICP4 no longer binds to the residual sequence, tripartite complexes do not form.

Collectively, these data demonstrate the importance of the nucleating effect of DNA on the cooperative assembly of the tripartite complex. This is further underscored by the observations that in the absence of DNA, these proteins have a relatively low affinity for each other.

within amino acids 142-274 of ICP4 is a serine rich tract that is highly conserved among the ICP4 analogs of a variety of herpes viruses (McGoech, et al., 1986, Nuc.

Acids Res. 14:1727-1744), which has been genetically implicated as a site of phosphorylation (DeLuca and Schaffer, 1988, J. Virol. 62: 732-743), and is a target for cellular kinases A and C. It is conceivable that the phosphorylation state of this tract modulates ICP4's interaction with the DB complex. This area has also been previously identified as being involved in tranactivation (Shepard and DeLuca, 1991, J. Virol. 787-794). n208 and nd3-8 are capable of activating transactivation and are able to form the complex, while X25 and nd8-10 are deficient for both transactivation and complex formation. Therefore, the data presented in Example Section 6 indicates that the tripartite complex has implications for the events involved in ICP4 function.

ii I INTERNATIONAL SEARCH REPORT International application No.

PCT/US94/07781 A r m; a ,r ,i ,I ;I WO 95/02071 PCT/US94/07781 -32- However, interaction with the DB complex cannot be the sole mechanism by which ICP4 stimulates gene expression. As shown in Figure 6, the d8-10 and d6- 10 like mutations (Paterson and Everett, 1990, Nucleic Acids Res. 16:11005-11025) are ene-half to one-third as active as wild type protein in transient expression assays, and viruses expressing these mutant proteins are able to generate high enough levels of early and late gene products to grow a, 2% of the level of wild type virus. Therefore, multiple mechanisms may be involved in ICP4 stimulation of gene expression, one involving the tripartite complex region of ICP4 between amino acids 142 and 210 and the others possibly involving some yet to be identified activities or interactions mediated by the carboxyl-terminal 524 amino acids of ICP4.

Figure 8 is a pictograph summarizing the interactions disclosed in the present invention. With respect to the question of transactivation, it is important to note that the results reported herein repeatedly underscore the fact that ICP4 increase the DNA-affinity of the DB complex. This is most dramatically seen in Figure 3. As depicted in Figure 3, ICP4 increases the DNA affinity of the DB complex by fivefold. This is also true in the case of human TFIID (Figure 3C). Consequently, the data indicates that ICP4 enhances gene expression by recruiting DB to the DNA template, an acknowledged rate limiting step in the assembly of the transcriptional preinitiation complex (Lin and Green, 1991, Cell 64: 971-981). Alternatively, it is conceivable that ICP4 transactivates by substituting, or by serving as a bridging agent, for one or more of the other cofactors among the myriad participating in the preinitiation complex. These two hypotheses are not mutually exclusive and may together explain how ICP4 operates on the molecular level to stimulate gene expression.

If indeed ICP4 acts as a nucleating agent by binding DNA and stabilizing the DNA association of the proteins which it contacts, then DNA binding is predicted to play a critical role in ICP4 functioning by this mechanism. However, the specific binding site mediating the formation of the tripartite complexes on the TK promoter shown in Figure 7 is not necessary for ICP4 induction in the context of the viral chromosome (Halpern, et al., 1984, J. Virol. 50: 733-738). No single site or collection of sites that uniquely specify induction by ICP4 has been identified (Smiley, et al, 1992, J. Virol.

66:623-631). However, the sequence requirement for specific ICP4 binding is known to be highly degenerate (DiDinato, et al., 1991, J. Mol. Biol. 219:451-470; Michael, et al.,

''I

ti f -U j d 2 41 111 1 ttd f WO 95/02071 PCT/US94/07781 -33- 1988, Science 239:1531-1534), resulting in numerous binding sights of variable affinity for ICP4 scattered throughout the HSV genome. Therefore, the abundance of ICP4 binding sites may make the absence of any single ICP4 site of little consequence in the context of the entire viral chromosome. Alternatively, interactions with other proteins i 5 bound to the DNA may substitute for the effect of specific DNA binding seen in this study. In addition, viral genes may be transactivated by ICP4 by alternative mechanism, such as implied by the activity and properties of d8-10.

An exception to this hypothesis is when a high-affinity site overlaps the start of transcription as in the case of the ICP4 promoter. Configured in this way, an ICP4 binding site results in repression when ICP4 is supplied in trans. The data presented in Example Section 6 is based on use of this promoter and show that ICP4 enhances the formation of the DB complex on DNA. This seeming paradox is reconciled upon considering the RNA Polymerase II and the preinitiation complex often assemble over the transcriptional start site of a promoter (Buratowski, et al., 1989, Cell 56: 549-561).

Consequently, ICP4's occupancy of this site may preclude the further assembly of the transcription complex beyond DB formation or interfere with the activity of transcription activation factors for upstream activating factors such as Spl despite facilitating the formation of the preincubation complex.

Therefore, ICP4 interacts with the transcriptional preinitiation complex and this interaction may have physiological significance for viral replication. In this context, similar scenarios for screening and selecting for antiviral compounds will exist for other viral trans-acting factors, including but not limited to the 80kDa IE protein of human cytomegalovirus, which interacts directly with the evolutionarily conserved carboxy-terminal domain of rTBP (Hagemeier, et al, 1992, J. Virol. 66: 4454-4462); S 25 VP16, a transactivator of HSV which functions at a different stage of in the viral transcription program than does ICP4, also interacts with rTBP (Smiley, et al, 1992, J.

Virol. 66:623-631); the Zta of Epstein-Barr virus and the Ela protein of human adenovirus (Horikoshi, et ai, 1991, Proc. Natl. Acad. Sci. USA 88: 5124-5128; Lee, et al., Cell 67: 365-376; Leiberman and Berk, Genes Dev. 5: 2241-2454). It has been shown that the ICP4 analog of pseudorabies virus complements Ela mutants of adenovirus for early gene expression, implying a similar path-way of action (Feldman, et al., Proc. Natl. Acad. Sci. USA 79: 4952-4956).

h^ fl J I _A II

Claims (11)

1. A method of selecting a potential antiviral compound or peptide which comprises mixing said potential antiviral compound or peptide with a wild-type HSV immediate-early viral protein, at least one general transcription factor which interacts with said wild-type viral protein and a cis-acting nucleic acid sequence, and detecting an altered protein:protein or DNA:protein interaction in comparison to the interaction detected in the absence of said potential antiviral compound or peptide.

2. The method of claim 1 wherein said wild-type viral protein is ICP4.

3. The method of claim 2 wherein said cis-acting nucleic acid sequence comprises the HSV promoter fragment upstream of the gene encoding ICP4. 15 4. The method of claim 2 wherein said cis-acting nucleic acid sequence comprises the HSV promoter fragment upstream of the gene encoding thymidine kinase. The method of claim 2 wherein said cis-acting nucleic acid sequence comprises the HSV promoter fragment upstream of the gene encoding gC.

6. The method of claim 1 wherein said general transcription factor is selected from the group consisting of TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJ.

7. The method of claim 2 wherein said general transcription factor is selected from the group consisting of TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJ.

8. The method of claim 3 wherein said general transcription factor is selected from the group consisting of TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIJ.

9. The method of claim 6 wherein said geneial transcription factors comprise TFIID and TFIIB. The method of claim 7 wherein said general transcription factors comprise TFIID and TFIIB.

11. The method of claim 8 wherein said general transcription factors comprise TFIID and TFIIB. RA411 T 0 gels were exposed overnight at 4°C and the resulting autoradiogram was used to cut out t

12. The method of claim 9 wherein a TATA Binding Protein replaces TFIID.

13. The method of claim 10 wherein a TATA Binding Protein replaces TFIID.

14. The method of claim 11 wherein a TATA Binding Protein replaces TFIID. 16th day of September 1998 Dated University of Pittsburgh of the Commonwealth System of Higher Education By their Patent Attorneys A.P.T. Patent and Trade Mark Attorneys o a~~ o o o or o or c or r or oo oo o o r o I D I o r r o I 'i B s t -t