IE83555B1 - Multivalent repressor of gene function - Google Patents

Description

The invention concerns the field of viral gene expression, more particularly the phenotypic expression of the rex (regulator of virion—protein expression) gene of HTLV—I and its equivalents in other retroviral species, such as rev of HIV—1. of HTLVVI Adult T—cell Leukemia (élg) as well as noncancerous conditions known as Tropical Spastic Parapesis. HTLV—II is etiologically related to some cases of variant T—cell hairy cell leukemia. Both virus groups are dividing their replication cycle, similarly to the DNA viruses, in an "early" and a "late" stage of gene expression. The "early" phase of gene expression is characterized by the expression of the regulatory proteins, while in the "late" phase the structural proteins are synthesizedI' The HTLV—I genome is coding for an activator of viral transcription termed Tax. The equivalent of Tax in HIV-1 is termed Tat.

Tax and Tat appear to act primarily on the retroviral LTR (long terminal repeat) for viral gene expression. In addition, HTLV~I encodes an activator of viral structural gene expression termed Rex. A functional Rex protein is responsible for the increased transport of unspliced viral IRNA out of the nucleus into the cytoplasm of the infected cell. There these IRNA species are constituting the viral genome and encoding the structural proteins.

Human Immunodeficiency Virus Type 1 (HIV-1) encodes a homologous protein termed Rev. The rev gene product is, as Rex in the HTLV—I system, absolutely required for the expression of the HIV-1 structural proteins.

In HIV-1 the selectivity of the induction noted above is due to an RNA target sequence required for Rev function termed Rev Response Element Taking Rex as an illustration, the complete function of the Rex protein in regulating expression of the HTLV—I Egg and en! genes requires at least three functionally distinct component activities: nuclear and nucleolar localization, i.e. the capacity to be transported from the cytoplasmic site of synthesis of all proteins to the nucleus and there to be concentrated in the nucleolar region; specific recognition (directly or indirectly) of the RexRE (RRX) sequence in viral RNAs; and Rex effector activity, the presently still unknown activity of this regulatory protein which actually mediates export from the nucleus to the cytoplasm of partially spliced viral mRNA species that include the RexRE sequence.

Regarding the structural locations in the Rex protein where these component activities of the complete Rex function reside (i.e. the functional domains), all that was known prior to the present invention is that a positively charged peptide domain in the first twenty amino acids at the amino terminus of Rex is required for nucleolar localization (H. Siomi As mentioned above both the reg gene product for HTLV~I and the rev gene product for HIV-1 are required for replication of the virus (see e.g. for HIV E. Terwilliger et al., J. Virol. QZ, [1988] 655). The crucial importance of Rex and Rev is underscored by the fact that in spite of their different primary structures, they are related functionally, and ETLV~I Rex is able to exert its function in the other viral species, i.e. in HIV-1 (L. Rimsky et a1., fiature 2Q; [1988] 738): thus even though — Rev and Rex do not share any significant homology on the nucleotide as well as on the amino acid level, — the nucleotide sequences and stem and loop structures of the RRE differ from those of the RexRE (RRX) in HTLV—I, — c0mputer~generated prediction of secondary structures of the Rex and Rev proteins reveal no significant similarities and — the Rex protein does not appear to bind to the same part of the RRE as the Rev protein does, of these proteins that are required for target RNA sequence recognition.

Mutations in regulatory proteins may yield a gene product with a dominant negative phenotype over the wild—type function (I. Herskowitz, Nature £22 [1987] 317). Dominant negative mutant proteins, known as trans—dominant repressors, a small group of which have been discovered recently in several unrelated viruses, represent a novel class of anti-viral agents. In genetic analyses, negative Iutations are those which cause a diminution or loss of a function of a gene. Dominant negative mutations are those that prevent other copies of the same gene, which have not been mutated (i.e. which have the wild type sequence), from functioning properly.

On the other hand recessive negative mutations do not so inhibit wild—type counterparts. Further, some dominant mutations inhibit wi1d—type genes only when the mutant and wild—type genes are located on the sane chromosome (DNA or RNA molecule). In this case the inhibiting mutation is said to be "cis—acting". Alternatively, a dominant mutation may inhibit the corresponding wi1d—type gene even when located on a separate chromosome.

This type is classified as a "trans—acting" dominant mutation or, more simply, as a transdominant mutation.

These differences in compositions and functions of these two regulatory proteins indicate that comparison of Rex structure with that of the Rev protein or its known mutants offers no guidance at all for selecting mutations that might produce trans-dominant repressors of the viral proteins.

A therapeutic application of the above concepts would involve the inhibition of production or overproduction of a deleterious gene product by manipulation of the gene to create dominant negative mutations whereby the resultant gene is encoding mutant regulatory proteins which when expressed disrupt the activity of the wi1d~type function (I. Herskowitz, Nature 829 [1987] 219). In the situation of viral, e.g. retroviral infections it thus appears highly desirable to provide corresponding transdominant repressors of virus function by the construction of similar inhibitors of essential regulatory genes, e.g. inhibitors of the rev or reg gene. This approach would provide the requisite tools for "intracellular immunization", an approach to the treatment of viral T H E Engineered transdominant versions of the HTLV-I Egg and, respectively, of the HIV-1 53! gene have now been made, the product of which blocks HTLV-I, HTLV-II or, respectively, HIV-1 replication. Furthermore, the product of some of these engineered transdominant versions of the reg or re! gene blocks goth HTLV-I (and in some instances HTLV-II) and HIV-1 (and in some instances HIV-2 and SIV) replication.

This appears to be the first reported occurrence of the preparation of viral repressors acting in more than one viral species, i.e. of transdominant gene products repressing the phenotypic expression of functionally equivalent genes of more than one viral species.

The invention thus concerns genes coding for proteins which transdominantly repress the phenotypic expression of functionally equivalent genes of more than one viral species and thus block replication of more than one viral species, particularly the mutant genes in pcRexM2, pcRexH7 and pcRexH8; H6, H7 and H13; and pHl0, disclosed hereunder.

It also concerns genes coding for proteins which transdouinantly repress the phenotypic expression of the reg gene of HTLV—I and/or HTLV—II and genes coding for proteins which transdoninantly repress the phenotypic expression of the 53! gene of HIV-1 and/or HIV-2 and/or SIV, particularly the mutant genes in pcRexH2, pcRexH7, pcRexH8, pcRexH17 and pcRex13A15; pH10, pA9/14, pA10/14, pH21, pM22, pHZ7, pH28, pH29 and pH32; and H6, H7 and H13, disclosed hereunder.

It also concerns a process for the preparation of these genes comprising isolating the corresponding vildtype gene from an appropriate expression system, putting this gene into an appropriate cloning system, introducing the desired mutation into the gene and recovering the resultant mutant gene from the clones having the desired mutation. It also concerns 3 process for the preparation of proteins as defined above which comprises expressing and amplifying a mutant gene as defined above in an appropriate expression and amplification system and recovering the expressed product therefrom.

It also concerns proteins which transdominantly repress the phenotypic expression of functionally equivalent genes of more than one viral species and thus block replication of more than one viral species, in particular the mutant proteins of pcRexH2, pcRexM7 and pcRexH8; H6, H7 and M13; and pH10, disclosed hereunder.

It also concerns proteins which transdominantly repress the phenotypic expression of the reg gene of HTLV—I and/or ETLV—II, and proteins which transdominantly repress the phenotypic expression of the rev gene of HIV 1 and/or HIV—2 and/or SIV, in particular the mutant proteins or pcReXH2, pcRexH7, pcRexM8, pcRexH17 and pcRex13A15; pM10, pA9/14, pA10/14, pH21, pHZ2, pH27, pH28, pH29 and pM32; and H6, H7 and H13, disclosed hereunder.

It also concerns a vector, e.g. a retroviral or plasmid vector containing a gene as defined above in a for: suitable for achieving delivery in a functional state into a target mammalian cell in vivo or in vitro.

It also concerns a pharmaceutical composition containing a gene or protein as defined above in a form suitable for achieving the desired prophylactic or therapeutic effect, together with a pharmaceutically acceptable carrier or diluent, e.g. in the form of cells taken from a patients’s body and treated in vitro prior to reinsertion.

It also concerns a method of treatment of viral infections comprising administering a gene or protein as defined above in a form suitable for achieving the desired prophylactic or therapeutic effect to a subject in need of such treatment, e.g. in the form of cells taken from a patient's body and treated in vitro prior to reinsertion.

Under "treatment" is to be understood the prophylactic as well as the curative treatment of viral infections, whereby "curative" includes the stabilization of a viral infection at a stage of latency.

It also concerns the genes, proteins and DNA segments defined herein for use as a pharmaceutical.

The invention also concerns inhibitors derived from the genes, proteins and DNA segments defined herein and able to nimic the transdouinant, i.e. primarily the RNA—binding domain in a mutant Rex or Rev protein as defined above, such as low molecular weight inhibitors or neutralizing monoclonal antibodies. Low molecular weight means herein a molecular weight below about 10 kD, especially below about 1 kD. u_,, Further aspects which the invention concerns are as listed hereunder: — A trans—dominant repressor of HIV-1 Rev function comprising a first and 3 second domain, the first domain having substantially the specific binding functions of vi1d—type HIV-1 Rev and the second domain not having the activation functions of wild-type HIV-1 Rev, the second domain being modified from wild-type HIV-1 Rev by one or more mutations; preferably the first domain comprises from about amino acid position 10 to about amino acid position 68 of vi1d—type Rev and the modified second domain is derived from about amino acid position 68 to about amino acid position 90 ;of wild-type Rev; especially, the above one or more mutations are missense or deletion mutations which occur between about amino acid position 68 and about amino acid position 90, preferably from about 78 to about 86, especially from about 78 to about 83 or 84 of wild—type Rev, the specific binding functions of the first domain of vild—type HIV-1 Rev remaining substantially functionally intact; particularly the repressors pM10, pM21, pMZ2, pMZ7, pM28, pM29 and pM32 disclosed thereunder; a trans-dominant repressor of HIV-1 Rev function comprising a first domain having substantially the specific binding functions of wild-type HIV-1 Rev, this transdominant repressor not having the activation functions of vild—type HIV-1 Rev; preferably the first donain comprises from about amino acid position 10 to about amino acid position 68 of vild—type Rev and the transdominant repressor lacks from about amino acid position 68 to at least about amino acid position 90 of wild-type Rev; particularly the repressors pA9/14 and pA10/14 disclosed hereunder: a DNA segment that encodes a trans-dominant repressor of the function of the HTLV-I Rex protein, the repressor being modified from a wild-type form of the Rex protein by at least one trans-dominant negative mutation in the ,1o_ peptide domain of the wild—type Rex protein that exhibits the effector activity of the Rex protein, this repressor having substantially the nucleolar localization activity of the wild—type form of the Rex protein; preferably such a DNA segment in which the peptide domain of the wild—type Rex protein comprises from about amino acid position 59 to about amino acid position 121, especially in any one of the following amino acid positions: 59, 60, 64, 65, 119, 120 and 121; particularly a DNA segment comprising any of the following mutant reg genes: H6, M7, M13 and variants and derivatives thereof which exhibit trans—dominant repression of BTLV—I Rex protein function; a corresponding trans—doninant repressor of the function of the HTLV-I Rex protein so modified from a vild—type form, preferably having the ability to repress either the function of the HIV-1 Rev protein or the function of the HTLV—II Rex protein; a method for identifying a specific inhibitor of the gene activation function of the Rex protein comprising the steps of: i) providing a genetic system comprising: — a DNA segment encoding an mRNA which comprises a regulatory response element that is recognized by the Rex protein, and at least one unused splice site (i.e. a region or intron that is bounded by splice recognition sequences but that has not been spliced out of the IRNA); ~ a DNA segment encoding a reg gene that is capable of being expressed to produce a protein product which induces export of the mRNA from the nucleus; ~ a host cell transformed by the DNA segment encoding the reg gene a and by the DNA segment encoding the mRNA and having the capability to express the protein product of the reg gene and to express the mRNA; ii) contacting a culture comprising the cells of this genetic system with an agent suspected of being a specific inhibitor of the Rex protein under conditions such that the agent enters the cells; "11- iii) determining the effect of this agent on export from the nucleus of the mRNA that comprises the unused splice site; and iv) determining the effect of the agent on export from the nucleus of a spliced form of the IRNA in which the splice site has been used; whereby a decrease in the export of the IRNA that comprises the unused splice site together with no decrease in the export of the spliced form of the mRNA indicates that the agent is a specific inhibitor of an activity of the HTLV—I rex gene or of an activity of a product of the {ex gene; the mRNA regulatory element that is recognized by the Rex protein preferably being derived from an mRNA of a virus selected from HTLV—I, HTLV—II and HIV-1; — an identification method as defined above wherein the decrease in the export of the mRNA that comprises the unused splice site is preferably detected by determining the level of production of a first protein, encoded by the mRNA that comprises the unused splice site, and the increase in export of the spliced form of the HRNA is preferably detected by determining the level of production of a second protein, encoded by the spliced form of the mRNA; ~ an identification method as defined above wherein the HRNA comprising the regulatory response element and the splice site is encoded by a plasmid couprising the 3' end of an HTLV—I provirus including the coding regions for the Rex and Tax proteins, the couplete en: gene, the Rex response element and the entire 3' LTR; preferably by plasmid pgTAX~LTR disclosed hereunder; and the £35 gene preferably is provided on plasmid pRex; ~ plasmid pgTAX—LTR; ,12_ ~ a reagent kit for screening agents to identify a specific inhibitor of the gene activation function of the Rex protein according to the above identification method, comprising: — a DNA segment encoding an mRNA which comprises a regulatory response element that is recognized by the Rex protein, and at least one unused splice site; — a DNA segment encoding a reg gene that is capable of being expressed to produce a protein product which induces export of the mRNA from the _nucleus; and — a container containing a host cell transformed by the DNA segment encoding the reg gene and by the DNA segment encoding the mRNA, the cell having the capability to express the protein product of the 535 gene and the mRNA; ~ a method of inhibiting replication of HIV-1, HTLV—I or HTLV—II comprising introducing a DNA segment as defined above into a cell having the ability to replicate one of these viruses and to express the DNA segment to produce a transdominant repressor of HTLV—I Rex function; and — a method of inhibiting HIV-1, HIV-2 and SIV, especially HIV-1 replication comprising introducing into a cell infected with HIV-1 a trans~dominant repressor of HIV-1 Rev function.

Figure 1: Nucleotide and amino acid sequence of HTLV—I rex [the amino acid positions substituted by oligonucleotides (1) (amino acid positions 87, 88, _13_ :}- _1i_£:,,£_N.A_TF__ I 0 N for 5.1. and 6.1.: ) and (2) (position 94) are marked, as well as positions 82, 90, 91 and 97. The full sequence contains 567 nucleotides, coding for 189 amino acids.

Figure 1A: Location of the 29 mutations introduced into the HTLV—I rex gene. site—directed mutagenesis, missense substitutions were introduced at defined The HTLV~I rex gene encodes a 189 amino acid protein. Using amino acid residues (indicated by boxes) and are named according to their location within the {ex gene. The pA mutant is named for the extent of the deletion, i.e., J3A15 is deleted between the introduced mutation for M13 and M15.

Figure 2A: Rex immunoprecipitation: SDS/polyacrylanide gel electrophoretic analysis of 7 of the 30 rex mutants after Rex immunoprecipitation (shows that pcRexH2, pcRexH7, pcRexH8, pcRexH13, pcRexH14, pcRexH17 and pcRex13A15 are still producing Rex protein): A: pgtat (negative control for Rex antibody) B: pcRex C: pcRexH2 E: pcRexH8 F: pcRexH13 G: pcRexH14 H: pcRexH17 (the lower molecular weight in this lane is possibly due to a change in the modification of this protein by, e.g., phosphorylation) I: pcRexl3A15 -114, Biological phenotype of the mutants: As for Figure 2A, after Tat immunoprecipitation: Figure 2B: A: pgtat B: pgtat C: pgtat D: pgtat E: pgtat F: pgtat : pgtat H: pgtat I: pgtat K: pgtat pXF3 pcRev pcRex pcRexM2 pcRexH7 pcRexH8 pcRexH13 pcRexH14 pcRexH17 pcRex13A15 A Figure 2C: As for Figure 2B (shows that some of the Rex mutants are transdominant over vildtype Rex, namely pcRexH2, pcRexH7, pcRexH8, pcRexH14, pcRexH17 and pcRex13A15): A: pgtat B: pgtat C: pgtat D: pgtat E pgtat F pgtat G: pgtat H pgtat I pgtat pXF3 + pXF3 pXF3 pcRexH2 pcRexH7 pcRexH8 pcRexH13 pcRexH14 pcRexM17 pcRex13A15 pcRex + pcRex + pcllex + pckex + p<:Rex + pdmx+ pcRex + pcRex + -15_ Figure 2D: As for Figure 2B (shows that some of the Rex mutants are transdominant over vildtype Rev, namely pcRexM2, pcRexM7, pcRexH8 and pcRexH13 partially): A: pgtat + pXF3 + pXF3 B: pgtat + pcRev + pXF3 C: pgtat + pcRev + pcRexHZ D: pgtat + pcRev + pcRexH7 # E: pgtat + pcRev + pcRexH8 F: pgtat + pcRev + pcRexH13 G: pgtat + pcRev + pcRexH14 H: pgtat + pcRev + pcRexM17 I: pgtat + pcRev + pcRex13A15 Figure 3: Rex mutants which have been constructed Figure 4: Sequence of the 29 oligonucleotides synthesized to uutagenize the rex coding sequence. _16_ Figure 5: Location of mutations introduced into the HIV-1 rev gene. The HIV~1 rev gene encodes a 116 amino acid protein with the predicted sequence shown. The border between the two coding exons of 53! is indicated (SP).

Clustered point (pH) mutations were introduced by o1igonuc1eotide;directed mutagenesis, as indicated by the boxed residues. These mutations were named according to their location within Rev, with pfll the most N-terminal and pH14 the most C—ternina1. All introduced mutations affected from two to four adjacent amino acids and all (except pM7) introduced a unique BglII site into the 53! gene sequence. These introduced sites facilitated the subsequent construction of N— and C~terminal deletion (pa) mutants. The pa mutants are named for the extent of the deletion, e.g. pA11/14 is deleted between the introduced pM11 and pH14 mutations.

Figure 5A: Location of further missense and deletion mutations introduced into the HIV—1 reg gene (— = deleted amino acid).

Figure 6: DNA and corresponding amino acid sequence of pcREV.

Figure 7: DNA sequence corresponding to mutation sites H1—M14.

Figure 8: Inuunoprecipitation of the HIV-1 re! and tat trans—activators.

Figure 9: HIV-1 proviral rescue assay.

Figure 10: Subcellular localization of Rev and selected Rev mutants by indirect immunofluorescence.

Figures 11A and 11B: Analysis of Rev mutants for a dominant negative phenotype.

Figure 12: Figure 12A: _17_ Competitive inhibition of Rev function.

Domain structure of the HIV-1 re! trans—activator. The amino acid ful1—length protein is encoded by two exons separated by an intron largely corresponding to the viral en! gene.‘ The "binding" and "activation" domains are shown as hatched boxes encompassing residues 23-56 (approx.) and 78—83, respectively. S = splice junction; NL = highly basic region important for nuclear localization that shares considerable identity with the "Arg—rich" RNA binding motif. .3. _18_ E_9£_,5_iI3d 6- 3- = Figure 13: Amino acid sequence of the HTLV—I Rex protein and location of each of the 25 point mutations introduced. Nucleotides encoding each of the boxed amino acids were removed and replaced in—frane by an oligonucleotide encoding the aspartic acid—leucine dipeptide.

Structure of the Rex—responsive pgTAX—LTR expression vector. The end of the CR—1 HTLV—I provirus from the HindIII site at map position 5013 (H. Seiki et al., Proc. Nat. Acad. Sci. USA 89 [1983] 3618-3622) through the 3' LTR, was inserted into the pBC12/CHV This fragment contains the two coding exons for Tax expression vector. (white boxes), the complete en: gene and the entire 3' LTR of HTLV~I including the RexRE (RRX) (S.H. Hanly et al., genes Develop. 2 [1989] 1534‘1544).

Figure 14: Rex, but not Rev 01 IL~2, activates the expression of HTLV—I Env pCHV—IL~2 (B.R. Cullen, gel; fig [1986] 973-982). cells were transfected with pREX in the absence of pgTAX—LTR. Env In the final lane, protein production was analyzed by immunoprecipitation and gel electrophoresis. The Iigration of known molecular weight standards is indicated on the left.

Analysis of Rex function of ref mutants. After insertion into the bBC12/CHV expression vector, each of the 25 £35 mutants (designated Ml—M18 and H21-H27, see Fig. 13A; mutants designated H19 and M20 have not been constructed) was cotransfected with pgTAX—LTR, and the cultures were analyzed for Rex—dependent HTLV—I Env protein expression as described in Example 12. ,19l Figure 15: Simultaneous analyses by immunoprecipitation of HTLV—I Env, Tax and Rex proteins in COS cells cotransfected with pgTAX—LTR and vectors for the inactive (M1, M2, M6, M7, M13) or impaired (H15) Rex mutants or the wild—type pREX, pREV and pCMV—IL—2 vectors. (A) Env production; (B) Tax production; (C) Vild~type and mutant Rex protein production.

Figure 16: Subcellular localization of HTLV—I Rex mutants by immunofluorescence. The wild type, M6, M7 and H13 Rex proteins are localized in the nucleoli and nuclei of expressing cells. In contrast, the M1 Rex protein is detected only in the cytoplasn while the M2 protein is iistributed throughout the cell. The H15 mutant is localized in the nuclei Jf expressing cells but, in contrast to the wi1d—type Rex protein, appears to be excluded from the nucleoli.

Figure 1?: (A) Analysis of the ability of £35 mutants to inhibit function of the wild—type Rex protein. Each culture was cotransfected with pgTAX-LTR, pREX and either one Rex mutant (Lanes 1~6), pREX (Lane 7), pREV (Lane 8) or pCMV—TL~2 (Lane 9). HTLV«I Env production was analyzed as in Fig. 14.

(B) The HTLV—I reg transdoninant mutants inhibit the function of HIV-1 Rev protein. Cells were cotransfected with pgTAT, pREV and pBC/CNV—IL—Z or the H6, H7 and H13 transdoninant Rex mutants and assayed for Rev—induced production of the truncated 72 amino acid form of the Tat protein (Lane 2). H6, H7 and H13 (Lanes 3-5) completely inhibited the HIV-1 Rev protein as only the full—1ength 86 amino acid for: of the Tat protein was detected.

(C) The HTLV-I transdominant Rex mutants block Rex rescue of the replication of Rev—deficient HIV-1 provirus. Cells were cotransfected with a £ew—deficient HIV-1 proviral plasmid and pREX in the presence of the indicated fold excess of the H1, H6, H7 and H13 mutants.

Supernatant levels of HIV-1 p24 Gag protein were measured.

. D E T A I L E D D B S C R I P T I O N The procedures and techniques to be used in employing the present invention are known in the art.

Viral species is herewith to be understood as being a taxononically distinct species such as HTLV-I, HTLV—II, SIV, HIV-1 and HIV-2. The invention concerns in particular the field of retroviruses, especially human retroviruses.

The genes the expression of which it is a goal of the invention to repress are preferably genes coding for an undesirable property, such as a function resulting in activation of the provirus and maturation into infective particles., e.g. re! of HIV-1 and HIV-2.

IS reg and Rev as used herein mean HIV-1 reg and HIV~1 Rev, respectively, unless specified otherwise. Thus the equivalent rev and Rev of other viral species, such as HIV-2, are specified as "HIV-2 rel (revZ)" and "HIV-2 Rev (Rev2)", respectively, etc.

Insofar as their preparation is not particularly described herein, the compounds, vectors, cell—lines, etc. used as starting materials or reagents are known and publicly available or may be obtained in conventional manner from known and publicly available materials, or equivalent materials may be prepared in conventional manner from known and publicly available materials. Thus e.g. the Rex gene may be recovered from any isolate of HTLV—I and the Rev gene fro: any isolate of HIV«I and pgTAX—LTR may be recovered e.g. from HUT102 or HT1. Alternatively, genes ’ may be created by chelical synthesis according to the genetic code to produce a protein having the required amino acid sequence. A transdouinant repressor of Rex or Rev function is made by standard recombinant DNA methods or by standard chemical methods for peptide synthesis, or by a combination of these methods, all of which are conventional. e21, Ln )-‘ The invention in one approach concerns transdominant repressors of the Rex function in HTLV—I, especially transdominant repressors of the Rex function in HTLV—I which are also active on the functionally equivalent but structurally unrelated Rev function in HIV-1.

Specifically, modified reg coding sequences were constructed and expressed and found to possess the above property.

The vildtype reg coding sequence (see Figure 1) was changed using a purchasable nutagenesis system in accordance with "O1igonucleotide—directed in vitro nutagenesis system Version 2", ‘ Amersham, England (1988), Code RPN.l523, hereinafter shortened as "Amersham protocol". The construction of the final expression vectors was carried out in stages entailing in succession: ) preparation of a bacteriophage M13 vector carrying the £35 coding sequence, 2) mutagenesis of the reg coding sequence and ) recloning of the mutated gene into mammalian expression plasmids. mutants were constructed, including one deletion nutant. The position and nature of the 29 site—directed mutations is indicated in Figure 3. The corresponding oligonucleotides used for nutagenesis are listed in Figure 4. They all carry a BglII restriction site. using e.g. the polymerase chain reaction. _ZZ_ (Catalog No. 70, AIDS Research and Reference Reagent Program, June 1989, NIH). amino acids Tat protein. This difference can readily be visualized upon immunoprecipitation analysis.

As shown in Figure 2C, 6 out of 30 mutants were found, namely pcRexH2, pcRexH7, pcRexH8,gpcRexH14,gpcRexH17 and pcRex13D15, which had a trans—dominant Rex repressor. The sane pattern also was found with pcRev (see Figure 2D), namely for pcRexH2, pcRexH7, pcRexH8 and partially pcRexM13, indicating that transdolinancy is not limited to the HTLV—I gene, and that some of the mutants are transdouinant for both genes, namely, in this particular instance, pcRexH2, pcRexH7 and pcRexH. l\J CI) While some of the results first obtained indicated that the most successful mutations were located between amino acid'position 87 and 94, and it thus appeared that a portion of the rex/rev gene lying between about amino acid position 82 and about position 97 was of particular significance in the engineering of trans—doninant Rex/Rev repressors, further testing has shown that the range for preferred positions of the mutations on the Rex protein is broader, i.e. that they may lie at least as near the Ngterminus as amino acid position 22 and at least as far toward the C-terminus as amino acid position 101.

In a further approach focussing on the Rev function in HIV-1, transdominant rev repressors have also been found. It is possible that some of these at least also inhibit the Rex function in HTLV-I or HTLV—II, but this has not been tested here.

On the other hand it has been found that some of these at least also inhibit the Rev function in HIV-2 and SIV,,c.

Extensive mutational analysis further has led to the delineation of at least two distinct functional domains within rev that appear to be essential for trans—activation. These domains are envisioned as comprising a "binding domain" which directs the Rev protein to its appropriate target substrate, and an "activation domain" which permits the functional consequences of the binding event, transcriptional activation, to be displayed.

Consequently, two functional domains within re: have been defined: a binding domain likely to direct the Rev protein to its cellular target, and an activation domain permitting the nuclear export of the incompletely spliced RNAs that encode the structural proteins Gag and Env.

Mutation of the activation domain of Rev results in the expression of defective Rev protein which acts as transdoninant inhibitor of Rev function.

Such mutants markedly inhibit HIV-1 replication when expressed in transfected cells in culture and are thus also transdoninant, as are the mutants obtained under 5.1.

Preferred under the above 9 transdoninant mutants are pH10, pH21 and pM32, especially pH10. l26_ .3.

In a yet further approach focussing on the Rex function in HTLV—I similarly to 5.1., transdominant rex repressors have also been found and a method developed that permits detection of Rex activities and is useful for the identification of specific inhibitors of Rex functioh which do not interfere generally with other viral or host cell functions. This Rex inhibitor detection method utilizes a genetic system comprising a Rex—responsive "reporter" gene that encodes an unspliced form of an mRNA that includes a regulatory element, a RexRE (RRX) for instance. In this system, a protein providing Rex function induces the export of this unspliced mRNA from the cell nucleus to the cytoplasm. In the absence of , Rex function, this mRNA is spliced before export to the cytoplasm, as indicated above. Upon contacting cells comprising this genetic system with an agent suspected of being an inhibitor of Rex function, specific inhibition of HTLV—I Rex function by that agent is indicated by a decrease in nuclear export of the unspliced form of this particular mRNA, together with no decrease in nuclear export of the spliced form of this same mRNA.

This method is useful for detecting, for example, chemical inhibitors of the HTLV~I Rex protein, e.g. inhibitors able to mimic the transdominant domain in a mutant Rex or Rev protein, e.g. low molecular weight chemical inhibitors, as well as transdominant mutant forms of Rex that act as repressors of Rex.

To identify transdominant negative mutations of the rgx gene, again a series of point mutations were produced that altered segments of two or three amino acids at various sites throughout the linear sequence of the Rex protein, and several transdominant repressors of HTLV—I Rex protein function identified among these mutants, several of which additionally transdominantly repressed the HTLV—II Rex and/or the HIV-1 Rev protein function and are thus, analogously to some of the mutants found under 5.1., also repressing the phenotypic expression of functionally equivalent genes of more than one viral species. ..27_ Accordingly, the invention also relates to a method for identifying a specific inhibitor of the gene activation function of the Rex protein, comprising the several steps defined under 3. above. As noted above the HTLV—I Rex protein is able to replace the function of the HIV-1 Rev protein. In addition it has now been found that Rex also can substitute for the analogous HTLV—II regulatory protein. Thus mRNAs from at least any of these three viruses, which have a response element that is recognized by Rex and at least one appropriate unused splice site, can be used in this method.

’ Preferably the mRNA comprising the response element and the splice site is encoded by a plasmid comprising the 3’ end of an HTLV-I provirus including the coding regions for the Rex and Tax proteins, the complete en! gene, the Rex response element RexRE (RRX) and the entire 3' LTR. An example of such a plasmid is pgTAX—LTR. For convenience in mutant analyses that require controlling the ratio of copies of a mutant reg gene to vild—type reg gene copies, the rex gene on pgTAX—LTR is inactivated by a recessive negative mutation and in this system, the active rex gene is provided on the separate plasmid designated pREX. However, for testing chemicals, for instance, the active rgx gene could be provided on the same plasmid or other vector DNA as the required mRNA of this system. Further, this active Eex gene might comprise a natural sequence variant isolated from a strain of HTLV—I other than that used in the present invention, or any other mutant form of Eex gene that is capable of being expressed to produce a protein product which provides the gene activation function of the Rex protein, including induction of export of the above IRNA from the nucleus.

The elements of the genetic system listed above could also be provided by using a DNA segment encoding the entire functional genome of a retrovirus as a part of this genetic system; it can be applied to infected cells. However, for safety reasons as well as convenience, this genetic system preferably is unable to produce any infectious virus. This is accomplished by design into the system of a genetic defect that prevents expression of at least one viral activity which is essential for production of any infectious virus from which some genetic element is used. For example, this may be done by omitting from the system at least a part of _28_ one viral gene or by some other mutation.

Preferably the host cell transformed by the reg gene and by the DNA segment encoding the mRNA is exemplified by COS cells which have been transfected by the plasmid vectors described above. Thus the term "transformation" as used herein encompasses the term "transfection" and indicates a genetic transformation involving a vector DNA that encodes an infectious agent, particularly a virus. After transfection such alvector can then spread from the minority of transformed cells in the culture to the majority of other cells by means of infectious virus particles, thereby providing a larger sample of host cells expressing the genes of interest.

In addition the transformation of the host cells in the present genetic system need not result in stable constructs; either stable or transient gene expression systems may be used to provide the required HRNA and {ex gene.

Thus a wide variety of known expression systems may be employed in identifying inhibitors according to the present method.

In this method a decrease in the export of the IKNA that comprises the unused splice site together with no decrease in the export of the spliced form of this mRNA indicates that the agent is a specific inhibitor of an activity of the HTLV—l reg gene or of an activity of a product of the reg gene. For the case of a chemical agent that is found to be a specific inhibitor by use of the present method, the node of action, in principle, could include specific inhibition of transcription or translation of the mRNA. More likely modes of action, however, include specific inhibition of one or more activities of the Rex protein, including nucleolar localization, recognition of the Rex response element, or the Rex effector function.

Advantageously, the decrease in the nuclear export of the mRNA that comprises the unused splice site is detected by determining the level of pfoduction of a first protein, this first protein being encoded by the mRNA that comprises the unused splice site (i.e., only the unspliced form of the mRNA encodes this first protein); and the increase in the nuclear export of the spliced form of the mRNA is detected by determining the level of production of a second protein, this second protein being encoded by the spliced form of this mRNA.

Preferably the mRNA, in the unspliced form, encodes the HTLV—I Env protein. Splicing of this mRNA results in a shorter mRNA that encodes another HTLV—I protein, Tax. export of the unspliced mRNA having an RexRE (RRX) element is detected by In e.g. Examples 12 and 13 below, the nuclear expression of the Env protein. Further, inhibition of such export of the unspliced mRNA, which results from inhibition of Rex function, is detected by a decrease in production of'the Env protein. However, since this decrease might also result from some general toxicity of an agent to the virus or the host cell, specific inhibition of £35 gene function is indicated by a decrease in Env expression together with no decrease in export of the spliced form of the nRNA, as reflected in no decrease in production of the ETLV-I Tax protein. Thus preferably simultaneous analysis of HTLV—I Env, Tax and Rex protein expression is effected.

In e.g. Examples 12 and 13 below the expression of the Env and Tax proteins is determined by immunoprecipitation with appropriate antibodies and electrophoretic analysis of the resulting precipitates.

Alternatives for, e.g., large scale screening of samples for specific inhibition of Rex function include for instance enzyme—linked immunoassay (ELISA) methods for Env and Tax, or alteration of the mRNA by genetic engineering to provide some other, more convenient products for indicating expression of the spliced and nonspliced forms. For example, the mRNA could be altered to encode an enzyme that can be detected by addition of a colorless substrate which produces a color upon hydrolysis, such as E. coli B~ga1actosidase. If the gene for this enzyme is inserted in place of the egg gene, the unspliced mRNA form would produce this enzyme while the spliced form would not. A second similarly convenient indicator gene could also be encoded in the mRNA so that it would be expressed in the spliced mRNA form, for example, by fusion to the Tax sequences. Further variations on the above approaches for rapid and efficient mass screening would be readily apparent to the man of the art.

In another aspect the invention relates to a reagent kit for screening agents to identify a specific inhibitor of the gene activation function of the Rex protein according to the method above, comprising the components listed under 3. above. This kit optionally further comprises any of the following: media that are used in the culturing of cells; reagents that are used in determining the level of nuclear export of either the spliced or the unspliced form of the reporter mRNA, either directly by nucleic acid hybridization, for example, or indirectly by immunological detection, for instance, of the protein products of the spliced and unspliced forms of this mRNA; and instructions for use of any of the above components of this kit for practising the method of the invention.

The above method has been used for screening various reg gene mutants for dominant negative mutations. When these Egg gene mutants were coexpressed with the plasmid pgTAX-LTR in the Rex inhibitor detection system of the invention, a class of mutations was found, similarly as under 5.1., comprising amino acid substitutions in the Rex protein at position 59-60, 64-65 and 119-121, which resulted in proteins that not only lacked Rex function but also acted as transdominant repressors of the function of the wild-type Rex protein and which also acted as transdominant repressors of the function of the wild-type Rev protein.

This mutational analysis also produced a second class of negative mutants comprising substitutions at Rex amino acid positions 5-7 and 14-15 which lacked Rex function. appropriately targeted to the cell nucleus nor transdominant. that for a Rex protein mutant to serve as a transdominant repressor of Rex function, that mutant may need to have not only a nuclear targeting activity but also a distinct nucleolar localization activity of the Rex protein, These findings on the Rex mutants here also define approximate bounds of at least two functionally distinct peptide domains within the Rex protein, a first one involved in nuclear and nucleolar targeting and a second involved in effector activity. The Rex mutants deficient in nuclear targeting are located in the positively charged peptide domain at fhe amino terminus of Rex that has been previously shown to function as a nucleolar localization signal; when a peptide comprising the amino terminal twenty amino aceds was attached to another protein by recombinant DNA means, this domain induced both nuclear targeting and nucleolar localization of that protein in a pattern similar to that observed for Rex (H. Siomi et al. id="p-1988" id="p-1988" id="p-1988" id="p-1988" id="p-1988"

[1988] supra). Thus it is not likely that the mutant at Rex amino acid positions 141-143 lies in a region that is required for nucleolar localization, even though this mutant was targeted to the nucleus but failed to localize in the nucleolar region of the nucleus. Rather, the alteration in this mutant most likely affects the Rex nucleolar localization function in the amino terminal domain indirectly, for instance, through interference with proper protein folding.

The second major functionally distinct domain of Rex encompasses amino acids 59-60 (tyrosine-isoleucine), 64-65 (tyrosine—tryptophan) and 119-121 (threonine—phenylalanine—histidine). An alteration at each of these discrete sites (H6, H7 and M13 mutants) leads to the production of a Rex protein that both lacks biological activity and displays transdominant inhibitory properties. Five different mutations having no effect on Rex function separate the region of H6 and N7 from that of H13 in the linear sequence of the Rex protein, indicating that the interaction of these two discrete regions within this functional domain may require proper protein folding. Thus the entire linear portion of the Rex protein encompassing these two regions of amino acids that are most critical for Rex effector function appears to contribute to the effector function of the Rex protein and, therefore, represents the domain to be mutated to produce transdominant repressors of Rex.

The present findings do not address the domain of Rex in which is located the activity required for recognition of the RexRE (RRX) in an mRNA. repressor of Rex function a mutant rex protein must retain the ability to Accordingly, it is not known whether to serve as a transdominant bind (directly or indirectly) to the RexRE, or to the recognition element of some other virus that is recognized by Rex.

Another aspect of the invention relates to a DNA segment that encodes a transdominant repressor of the function of the HTLV—I Rex protein, as well as such a transdominant repressor. This repressor is a protein that is modified from a wild-type form of the Rex protein by at least one mutation that negatively affects the effector activity of the Rex protein.

This represssor also has substantially the nucleolar localization activity of the wi1d—type form of the Rex protein. In particular, the negative mutation of this repressor is one that affects an amino acid in the peptide domain of the wild~type Rex protein that comprises from about aminowacid position 59 to about amino acid position 121, more particularly in any of the following positions: 59, 60, 64, 65, 119, 120 and 121. The DNA segment encoding the Rex repressor is exemplified by any of the following mutant rex genes: H6, M7, M13 and variants and derivatives thereof which exhibit transdominant repression of HTLV—I Rev protein function.

The sequence of this DNA segment is derived from the Rex gene of any isolate of HTLV—I (L. Rimsky et al., Nature ggé [1988] 738-740; H. Seiki et al., Science 222 [1985] 1227-1229) or is created by chemical synthesis.

The transdominant repressor of Rex function is made by standard recombinant DNA methods or by standard chemical methods for peptide synthesis or by a combination of these methods, all of which are we11—knovn in the art of genetic engineering.

The mutations that negatively affect the effector activity of the Rex protein are exemplified as described herein. However other types of mutations designed to produce localized effects on the protein structure at or close to these same amino acid positions also are highly likely to produce variants and derivatives of Rex which exhibit transdominant repression of HTLV—I Rex protein function according to the present invention. Such localized defects include, for example, deletions or insertions of single amino acids or substitutions of chemically or structurally similar amino acids. On the other hand, more extensive deletions or insertions, or substitutions that disrupt secondary structure (e.g., a proline in a B—sheet region) are highly likely to have effects on distant parts of the protein through influence on protein folding; therefore, such mutations at the indicated positions within the domain required for Rex effector function are not likely to produce mutant Rex proteins that retain substantially the nucleolar localization activity of the wild—type form of the Rex protein.

The transdominant Rex mutants of 5.3. above were also tested for inhibition of Rev function and found to be repressors of HIV-1 Rev as well.

The anti—viral potential of this class of transdominant Rex mutants has been demonstrated using an assay for inhibition of HIV-1 replication.

In addition and as already mentioned above it has now been discovered that HTLV«I Rex can also functionally substitute for the analogous HTLV—II regulatory protein, even though the nucleotide sequence of the corresponding response element in HTLV—II has a somewhat different stem and loop structure fro: that of the RexRE (RRX) in HTLV-I.

The invention thus further relates to a method of inhibiting replication of HIV-1, HTLV—I and HTLV—II comprising introducing a DNA segment as defined above which encodes a transdoninant repressor of Rex function into a cell having the ability to replicate one of these viruses.

This cell also has the ability to express the DNA segment to produce the transdominant repressor. This cell may be one that was previously infected by one or more of these viruses or this cell may be an uninfected target cell for one or more of these viruses. ~34- Thus in the genetic system defined above HTLV—I Env expression by pgTAX—LTR is specifically induced in the presence of a protein having the gene activation function of the wild—type Rex protein. In the system including a gene that provides a protein having Rex function, if contact with an agent inhibits Rex function, Tax protein continues to be produced unless that agent affects some viral or cellular activity that is not related to the expression of the gene or the product of the gene that provides Rex function, i.e. unless the agent is not a specific inhibitor of that gene or its product. Thus, as indicated in Example 12 and Figure 15, advantageously the exemplary method for identifying specific inhibitors of -35l Rex function includes the simultaneous analysis of ETLV—I Env, Tax and Rex protein expression.

It may be noted that in the case of specific inhibition of Rex function by transdominant repressors described below and in Figure 15 the production of HTLV—I Tax protein evidently increases. This is probably a result of an increase in the nuclear export of the spliced form of the Env mRNA that is not exported prior to splicing due to a lack of Rex function.

However, in principle a specific inhibitor of Rex function may prevent splicing of the mRNA but not induce export of the unspliced RNA.

Accordingly, the present method for identifying specific inhibitors of Rex function requires only that there be no decrease in the nuclear export of the spliced mRNA that, in the present situation, produces the Tax protein.

The use of this system is exemplified in Example 12 below. In this mutational analysis again oligonucleotide—directed mutagenesis, in the M13 bacteriophage, was employed to alter the primary sequence of the rex gene, at 25 discrete sites (Fig. 13A). The boxed amino acids were replaced by the dipeptide aspartic acid—leucine by insertion of an in—frame oligonucleotide duplex which also contained the diagnostic Bglll restriction site. Each of these rgx mutants was then inserted into the pBC12/CHV eucaryotic expression vector (B.R. Cullen, Eell fig [1986] 973-982) and the mutations were verified by DNA sequencing.

Each of the rgx mutations was examined for biological activity by cotransfection with the pgTAX-LTR vector. While nineteen of these rex mutants displayed a wild—type phenotype, five mutants (H1, H2, H6, H7 and M13) lacked apparent en: gene activation activity and one mutant (H15) displayed only partial function (Fig. 14B). COS cells were next cotransfected with these six rex defective mutants and the pgTAX—LTR vector, followed by simultaneous analysis of HTLV—I Env, Tax and Rex protein expression as described in Example 12 (see also Fig. 15 A-C). While HTLV—I Env was only detected in the presence of the wild—type Rex protein (Fig. 15A, Lane 7) or the partially active H15 mutant (Fig. 15A, Lane 6), the 40 kD Tax protein was detected in all of the cultures (Fig. 15B). Thus the lack of en! gene expression observed with the H1, H2, H6, H7 and H13 mutants is due to the specific loss of Rex biological activity rather than -36_ non—specific, toxic effects of these proteins in the transfected COS cultures. Each of the mutant Rex proteins was also identified in these cultures indicating that all of the mutants were expressed in a stable manner (Fig. 15C). The M2, M6 and M13 mutants migrated in a manner that was indistinguishable from the wild—type Rex protein (Fig. 15C, Lanes 2, 3, 5 and 7), whereas the H7 and M15 proteins exhibited a smaller apparent molecular weight (Lanes 4 and 6} and the H1 mutant yielded an if electrophoretic doublet of proteins (Lane 1). Sequencing of the protein coding regions in the M1, M7 and H15 mutants failed to reveal any changes other than the specific mutations introduced. Thus the biochemical basis for these apparent differences in size likely reflects altered post—translational processing of these mutant Rex proteins.

Like the wi1d—type Rex protein, in situ immunofluorescent staining of cells transfected with the biologically inactive H6, H7 or M13 Rex mutants, as detailed in Example 12, revealed normal targeting to the nucleoli and nuclei of expressing cells (Fig. 16). In sharp contrast the M1 mutant protein was detectable only in the cytoplasmic compartment, while the M2 Rex mutant was distributed in an approximately homogeneous manner throughout the cell. Consistent with these findings is the fact that the M1 and M2 mutations altered basic amino acid residues located within the positively charged peptide domain that functions as a nucleolar localization signal. The partially active M15 mutation lead to a pattern of nuclear localization of mutant Rex protein, but unlike the wild—type Rex protein the M15 protein did not localize further within the nucleolar region of the nucleus and, in fact, H15 appeared to be excluded from the nucleoli (Fig. 16). acids at the N—terminus may be involved in or contribute to the nucleolar These results suggest that residues away from the basic amino localization of Rex.

The Egg mutants were also examined for their capacity to block the biological action of the wild-type HTLV—I Rex protein and the vild—type HIV-1 Rev protein (Fig. 17). When cotransfected with pgTAX-LTR and pREX in COS cells (Panel A), a 10fold molar excess of the H6, H7 and H13 nutants displayed a dominant negative phenotype in that the action of the wild—type _37_ Rex protein was markedly inhibited (Lanes 3-5). In contrast, the M1, M2 and M15 proteins acted as recessive negative mutants since the action of the wild-type protein was not altered (Lanes 1, 2, 6). Similarly, the Rev protein of HIV-1 did not interfere with the action of the Rex protein (Lane 8) nor did IL-2 (Lane 9).

The capacity of the trans-dominant Rex mutants to blocR the function of Rev in the HIV-1 system was examined next (Fig. 17B). Vhen cotransfected with pgTAT and pREV in COS cells, a 10-fold molar excess of H6, H7 or H13 Rex mutants inhibited the action of the Rev protein (Lanes 3, 4, 5) as evidenced by diminished expression of the 72 amino acid form of the Tat protein. The ability of these transdominant Rex mutants to block HIV-1 viral replication was also studied (Fig. 17C). Replication of a Rev—deficient HIV-1 provirus, pHXB2-Bam-p3 (L. Rimsky et al., Refuge 33; id="p-1988" id="p-1988" id="p-1988" id="p-1988" id="p-1988"

[1988] 738) in the presence of Rex and graded amounts of transdominant Rex mutants was studied by transfection of COS cells with these plasmids. As indicated by synthesized levels of the HIV-1 p24 Gag protein in culture supernatants, the transdominant Rex mutants (H6, H7 and H13) produced dose—related inhibition of HIV-1 replication. In contrast, the recessive negative M1 mutant of Rex was without significant effect on HIV-1 replication.

Together, these findings with the various £35 mutants of Examples 12 and 13 hereunder indicate the approximate boundaries within the Rex protein of at least two different structural domains having different activities. Thus one domain is defined by the H1 and H2 mutations, is located at the N—terminus, involves amino acids 5-7 and 14-15 and appears to play a role in targeting of the protein to the nucleus and thence to the nucleolus. This positively charged domain may also be involved in Rex binding either directly to the RexRE (RRX) or to other proteins that directly contact this RNA element. The second domain, which has been described above, is critical for the Rex effector function and, therefore, can be mutated to produce transdominant repressors.

Sunarizing the findings under 5.1., 5.2. and 5.3. above, it is concluded that: a) a generally applicable principle has been found for producing viral inhibitors by mutating a critical regulatory protein, such as Rex or Rev; b) this principle appears applicable to the production of mutant regulatory proteins acting on a plurality of viral species; and c) the specific transdoninant lutants which have been constructed and identified are: ~ HTLV—I Rex mutants effective in inhibiting HTLV~I Egg gene function: — pcRexH2, pcRexH7, pcRexH8, pcRexH17 and pcRex13A15 (see 5.1. and Examples 1-3); — H6, H7 and H13 (see 5.3. and Examples 12~13); — HIV-1 Rev mutants effective in inhibiting HIV-1 53! gene function: — pH10, p69/14 and pA10/14 - pH21, pH22, pH27, pH28, pH29 and pH32 (see 5.2. and Examples 11a and 11b); (see 5.2. and Examples 4~11); — HTLV—I Rex mutants effective in inhibiting HTLV—I Egg gene function and also effective in inhibiting HIV-1 Egg gene function: ~ pcRexH2, pcRexH7 and pcRexH8 (see 5.1. and Example 3); » H6, H7 and H13 (see 5.3. and Examples 12-13); Xx) C‘: -39‘ ~ HTLV—I Rex mutants effective in inhibiting HTLV—I and HTLV—II rex and HIV~I rev gene function: — H6, H7 and H13 (see 5.3. and Examples 12-13); — HIV~1 Rev mutant effective in inhibiting HIV-1 rev gene function and also effective in inhibiging HIV—2 rev and SIV__c rev gene function: — pH10 (see 5.2. and Example 11b). -40, . 1zx_LH_PLEs The following Examples illustrate the invention. They are not to be viewed as being limitative. .1.

Example 1: Constructiontof a transdolinant HTLV—I rex gene . Cloning of the reg coding sequence into the RP—DNA of bacteriophage H13 ug pCRex DNA were treated with 10 units each of restriction enzymes HindIII and EcoRI in 20 pl restriction enzyme incubation buffer (10 mM TrisAHCl ph 7.5; 10 mM MgCl2; 50 mH NaCl; 1 mM Dithiothreitol) at 37°C for 3 hours. The reaction mixture was directly loaded onto a 1 Z agarose gel (Seakem FMC Inc., Rockland, ME, USA) containing 1 pg/ml ethidium bromide and subjected to electrophoresis at 50 V; 25 nA for 4 hours in Tris—acetate buffer (T. Haniatis et al., Molecular Cloning, A Laboratory Manual [1982], Cold Spring Harbor Laboratory, New York, p. 156). The separated DNA was visualized on a 366 nu UV—1amp and an appropriate gel slice containing the This gel section was placed in a dialysis The DNA was pg of bacteriophage M13mp10 RF—DNA were treated with the restriction enzymes HindIII and EcoRI and the 7.2 kb vector H13 fragment was isolated from a 1 Z agarose gel in the same way. ng phage DNA and 1 ug cRex DNA were mixed in 20 ul ligation buffer (50 mM Tris-HC1 pH 7.4; 10 mM Hgclz; 10 mM Dithiothreitol; 1 mH ATP) with 1 unit of T4—DNA—ligase and incubated for 15 hours at 16°C. This reaction mixture was used directly to transform and plate out E. coli strain TG1 (Amersham protocol, p. 16-18).

The DNA of appropriate phage plaques was checked by endonuclease cleavage with the restriction enzymes HindIII and EcoRI, followed hy analytical gel electrophoresis through a 1 Z agarose gel containing 10 ug/ml ethidium bromide in Tris—acetate buffer. A bacteriophage carrying the rex coding sequence was identified and designated mplorex.

Single—stranded np10rex DNA was isolated using a large scale preparation protocol (Amersham protocol p.24-25). x.‘ 0 2. Hutagenesis of the rex coding sequence in Iplflrex As a prerequisite of the mutagenesis it was necessary to synthesize appropriate single—stranded DNA molecules. 29 oligonucleotides (see Figure 4) were made, which eventually led to 30 mutants, of which the following seven oligonucleotides led directly or (see Example 2) indirectly to the mutants which were found to be successful in terms of transdominant phenotype: (1) 5’— TG GAC AGA GTC TTA GAT CTG GAT ACC CAG TCT -3' Bg1II (2) 5'— AC TAT GTT CGG CCA GAT CTC ATC GTC ACG CCC -3’ BglII (3) 5’~ CC TAC ATC CTC ACA GAT CTC TGG CCA CCT GTC -3’ Bglll (A) 5'_ TCG GCT CAG CTC TTA CAT CTC TTA TCC CTC CA #3’ Bglll (5) 5'- AG crc TAC AGT TCA GAT CTC CTC GAC TCC ccr —3' BglII (6) 5'— GT TCC TTA TCC CTA GAT crc CCT CCT TCC CCA -3’ BglII (7) 5'— CT CCT TCC CCA CCA GAT CTA CCT CTA AGA ccc -3’ - BglII Oligonucleotide (1) is substituting the amino acid residues phenylalanine (position 30), phenylalanine (position 31) and serine (position 32) in the Rex protein by the amino acids leucine (position 30), aspartic acid (position 31) and leucine (position 32).

Oligonucleotide (2) is substituting the amino acid residues alanine (position 58) and tyrosine (position 59) in the Rex protein by the amino acids aspartic acid (position 58) and leucine (position 59).

Oligonucleotide (3) is substituting the amino acid residues proline (position 63) and tyrosine (position 64) in the Rex protein by the_amino acids aspartic acid (position 63) and leucine (position 64).

Oligopucleotide (4) is substituting the amino acid residues tyrosine (position 87), serine (position 88) and serine (position 89) in the Rex protein by the amino acids leucine (position 87), aspartic acid (position ) and leucine (position 89).

Oligonucleotide (5) is substituting the amino acid residues leucine (position 90) and serine (position 91) by the amino acids aspartic acid (position 90) and leucine (position 91).

Oligonucleotide (6) is substituting the amino acid residue serine (position ) by the amino acid leucine.

Oligonucleotide (7) is substituting the amino acid residues arginine (position 100) and glutamic acid (position 101) in the Rex protein by the amino acids aspartic acid (position 100) and leucine (position 101).

All oligonucleotides are introducing Bg1II restriction sites in frame of the rex coding sequence.

. The oligonucleotides have been synthesized on solid support on an Applied Biosystems 380A synthesizer using B—cyano—ethylphosphoamidite chemistry. Purification was done by 8 Z polyacrylamide gel electrophoresis, followed by elution of the main product and ethanol precipitation.

Phosphorylation of the oligonucleotides using ATP and polynucleotide kinase was carried out as described in the Amersham protocol, p. 13. The oligonuc1eotide—directed mutagenesis reaction was effected according to the x.) C’) This was followed by transformation and The DNA of different plaques was screened by restriction endonuclease analysis using Amersham protocol, p. 13-16. plating out of E. coli strain TG1 as described on pages 16-18. the enzymes BglII and EcoRI.

From all seven mutations one clone was identified carrying the introduced mutation in the rex coding sequence. These clones have been designated mp10rexH2, mp10rexMl, mp10rexM8, mp10rexH13, mp10rexH1l: mp10rexM16 and mp10rexH17. A further clone, np10rexH15, was used in the construction of the deletion mutant (see Example 2).

. Recloning of the mutated rex genes into iailalian expression plasmids The mutated rex genes were moved back from the bacteriophage H13 vectors into the original expression plasmid. 5 ug pcRex DNA were incubated with 10 units HindIII and 10 units EcoRI restriction endonuclease at 37°C for 3 hours. The reaction mixture was loaded directly onto a 1 Z agarose gel containing 10 pg/ml ethidium bromide and subjected to electrophoresis in Trisaacetate buffer at 50 V; 25 mA for 4 hours. The HindIII—EcoRI vector fragment was electroeluted out of the appropriate gel slice and precipitated with ethanol. ug of the mutants obtained under 2. were treated in the same way and fragments containing the rex coding sequence were isolated. ng of the isolated vector fragment and 1 ug of the isolated rex sequences were mixed together separately in 20 ul of ligation buffer in the presence of 1 unit of T4—DNA—ligase and incubated at 16°C for 15 hours.

The resultant reaction mixtures were directly used to transform E. coli strain HB101. The DNAs of different bacterial colonies were analysed using the restriction endonucleases HindIII, Asp718 and BglII, followed by analytical gel electrophoresis through a 1 Z agarose gel. Plasmids identified in this way carrying the reg genes were designated pcRexM2, pcRexH7, pcBexH8, pcRexM13, pgRexM14, pcRexH15, pcRexH16 and pcRexH17, respectively. ix) 0 _45_ Example 2: Construction of a deletion mutation in the rex coding sguence pg of pcRexH13 DNA were treated with 10 units each of the restriction enzymes BglII and EcoRI. The larger DNA fragment, containing the vector backbone and the 5' portion of the rex coding sequence was isolated as described above. the rex coding sequence was isolated. ng of the isolated pcRexH13 DNA fragment were mixed with 1 ug of the isolated pcRexH1S DNA fragment and incubated in 20 ul of ligation buffer in the presence of'1 unit of T4—DNA ligase at 16°C for hours. positions 87, 88 and 89 are identical with those in clone pcRexHl3 and position 90-94 are deleted.

The reaction mixture was then directly used to transform E. coli strain HB101. The DNA of different clones was screened by restriction endonuclease digestion employing the enzymes HindIII, Asp7l8 and Bglll. positive clone was identified and designated EcRex13A1 .

In parallel 5 ug of pcRexM15 DNA treated in analogous manner and the smaller DNA fragment, containing the 3' portion of This manipulation led to a rex coding sequence where amino acid _45_ Example 3: Biological activity 'a) Biological activity of the lutant genes in -annalian cells Meth. Enzymol. lég (1987) 684—703. At 60 hours post-transfection cultures were labeled for 3 hours with 300 uCi/I1 of 35S—cysteine and analyzed for expression of the HIV-1 Tat and the HTLV-I Rex protein by immunoprecipitation analysis. A rabbit anti-peptide antibody directed against amino acid residues 1-61 of Tat and a rabbit anti—peptide antibody directed against amino acid residues 173-189 of Rex were used in this L experiment as described in B.R. Cullen, J. Virol. Qg (1988) 2498-2501.

Precipitated proteins were resolved on SDS/polyacrylamide gels and visualized by autoradiography.

The rex mutant genes encoded in the vectors pcRexH2, pcRexH7, pcRexH8, pcRexM14, pcRexH17 and pcRex13A15 yielded a negative phenotype for £35 action in this assay system (see Figure 2B, Lanes D, E, F, H, I and K) whereas controls (Lanes B and C) and other mutants (Lane G) yielded a positive phenotype.

All of these phenotypically negative Rex mutant clones are able to produce a Rex—specific protein recognized by the polyclonal anti—Rex antibody described above (Figure 2A, Lanes C to E and G to I). In contrast, the mutation in clone pcRexH16 resulted in a protein undetectable by the Rex:specific antibody; this is believed to be due to a decreased protein ha1f—life (not shown). _47_ b) Transdoninant repression of vildtype HIV-1 rev and/or HTLV~I rex function by pcRexH2, pcRexH7, pcRexH8, pcRexH13, pcRexH14, pcRexH17 and pcRex13A15 The same experimental set—up was used to examine the ability of the rex mutants to inhibit in trans the function of the vildtype Rev and Rex protein. 0.25 ug of the genomic tat expression vector pgtat, 0.i5—ug of the vildtype rev (pcRev) or the vildtype rex (pcRex) expression plasmid and an excess of each rex mutant expression plasmid (5 ug) were mixed separately and transfected into Cos cells; The influence of the mutants on the wildtype Rev or Rex function was measured by Tat—specific immunoprecipitation as described above.

The result of this experiment shows that the six rex mutants pcRexHZ, pcRexH7, pcRexH8, pcRexH14, pcRexH17 and pcRexl3A15 have the ability to inhibit the vildtype Egg-mediated trans—activation (see Figure 2C, Lanes C, D, E, G, H and I), that the four rex mutants pcRexH2, pcRexH7, pcRexH8 and (partially) pcRexMl3 have the ability to inhibit the vildtype £eX—mediated trans—activation (see Figure 2D, Lanes C, D, E and F), while some of the mutants, namely, in this instance, pcRexH2, pcRexH7 and pcRexH8, are able to inhibit both the vildtype rex— and the vildtype £g!—Iediated trans—activation.

This set of experiments demonstrates on the protein level the transdominant repression of Rev and/or Rex function by the above rex mutants . -4g_ The results from Examples 1 to 3 also seem to indicate the existence of two functional domains in the HTLV—I Rex protein. Towards the amino terminus there are two mutants (pcRexM7, amino acid positions 58,59 and pcRexM8, amino acid positions 63,64) that are transdominant over both Rev and Rex proteins. A second cluster, containing mutants transdominant only over the Rex protein, is located in the middle of the coding sequence.

The additional rev/rex transdominant mutant (pcRexH2, amino acid " positions 30-32) located between the nuclear localization signal and the cluster comprised of mutants 7 and 8 could be part of a third functional domain or alternatively the introduced amino acid change might disturb the tertiary structure of the protein, resulting in the observed transdominant phenotype.

’.I\ fix.) _[,9, .2.

Example 4: Clustered point and deletion mutations in Rev The Rev protein is phosphorylated at serine inivivo and is localized predominantly to the cell nucleus where it is concentrated in the nucleoli. The mutational analysis described below addresses, among other factors, the relevance of these properties to the function of Rev as a trans—activator of HIV-1 structural gene expression.

Clustered point mutations (pH) were introduced into the HXB—3 strain of HIV-1 Rev by, again, oligonucleotide directed mutagenesis, as described by Taylor et a1., Nucleic Acids Res. 12 (1985) 8765-8785.

Specifically, a bacteriophage H13 nutagenesis system (Amersham Corp., Arlington Heights, IL, USA) was employed to introduce targeted nucleotide substitutions into full—length cDNA copies rev encoded by the expression corresponding amino acid sequences of pcREV appear in Figure 6.

A series of clustered point nutations were introduced into Rev (see M1—H14, Fig. 5), cloned, and their sequences confirmed using a dideoxynucleotide sequencing system (Stratagene, La Jolla, CA, USA). The mutations were generally spaced evenly throughout Rev and served to target each of the 11 serine residues therein. The DNA sequence at and surrounding each mutation (pH1—pH14) are provided in Figure 7, with the altered nucleotides being underlined.

Most of the noted mutations resulted in codons for aspartic acid (Asp) and leucine (Leu) replacing the residues "boxed" in Figure 5. For easeaof reference the mutations have been designated according to their location within Rev, e.g. pM1 being the most N—termina1 mutation and pM14 the most C—terminal mutation.

While the precise structure of the various pH mutants is shown in Figure 7, the majority of mutations affected only two codons. BxcePti0"5 to this generalization include M2, M4, M23, M24 and M25 which affected three codons and M6 _5o_ which affected five codons. In most cases the amino acid substitutions arose from a two—amino acid missense mutation, however, in pH6, Asp and Leu replaced four arginine residues, thereby resulting in a two amino acid PM4 Contains an additional adjacent single deletion, while amino acid substitution not observed in the parental REV, which involves the replacement of aspartic acid forftyrosine at position 23.

These additional changes arose from single base errors in the single-stranded DNA oligonucleotide primer used in the mutagenesis protoc Most point mutations resulted in the formation of unique BglII was constructed by the simple deletion of two sites: pfi7 adjacent serine residues (see Fig. 5). the same translational frame, thereby facilitating the subsequent construction of N—termina1 and C—termina1 deletion mutants (p0) of Rev. pA mutants are designated by the location of deletion, e.g. pA11/14 has a deletion between the introduced pM11 and pM14 mutations.

The BglII sites were all inserted in Example 5: Expression of Rev mutants The parental cytomegalovirus immediate early promoter based vector pBC12/CHV (B.R. Cullen, gel; fig [1986] 973-982) was used as a negative control. Also employed herein were the genomic tat gene expression vector pgTAT, and the secreted alkaline phosphatase gene expression vector pBC12/RSV/SEAP (Malim et al., supra, and Berger et al., gene 66 [1988] 1-10, respectively). ' Sixty hours post—transfection, the cultures were labelled with [35S]—cysteine and [32P]—inorganic phosphate ([32P]—Pi) in parallel as described by Halim et al., 1988, supra, and Hauber et al., J. Virol. gg [1988] 4801-04. The cells then were lysed with RIPA buffer and the relative level of re! and tat expression in the cultures assayed by immunoprecipitation analysis using rabbit polyclonal antipeptide antisera (Halim et al., [1988], supra, and Cullen et al., J. Virol. Q2 [1988] 2498-2501). More specifically, antisera to Rev amino acid residues 1-20 (REV1/20) was used for immunoprecipitation of the mutant proteins encoded by pM5 and pM6, while antisera to Rev amino acid residues 27-51 (REV27/51) was used for immunoprecipitation of all remaining mutant proteins.

Immunoprecipitation analysis of tat expression was performed using rabbit polyclonal antipeptide antisera to Tat amino acid residues 1-61 (TAT1/61). _52_ The immunoprecipitated proteins were resolved by electrophoresis on 14 Z discontinous SDS~acrylamide gels and visualized by autoradiography, The results of these experiments are depicted in Figure 8, wherein the relative migration of known protein molecular weight markers is depicted to the right of the figure.

Immunoprecipitation of the [35S]—cysteine and [32P]—Pi labelled cultures with anti—Rev antisera are depicted in Figures 8A and 8C;_ respectively. Immunoprecipitation of the [35S]—cysteine labelled cultures resulted in the majority of missense (pH) mutants yielding bands of an intensity and mobility comparable to the wild-type (Fig. 8A, lanes 1~14).

Exceptions to this generalization include mutant pH6, which yielded an intense band of slightly faster mobility (H, ~ 18 kD) and pfll, which yielded a faint band of significantly slower mobility.

Immunoprecipitation analysis of tat expression using anti—Tat antisera provided a qualitative assay for Eey function using pgTAT as a model indicator. As indicated above, absent rev, pgTAT expresses a fully spliced cytoplasmic tat mRNA which encodes the 86 amino acid (aa) two exon form of tat protein exclusively. In the presence of rev, however, pgTAT induces the cytoplasmic expression of an unspliced tat mRNA that encodes a truncated, one exon form of the protein 72 aa long.

Vild—type rev migrates at a relative molecular mass (M,) of 19 kilodaltons (kD) and is readily detected in Figure 8 (lane 0), while the 86 aa and the 72 aa forms of tat migrate at 15.5 kD and 14 kD, respectively.

Mock—transfected cultures yield no specific signal under these assay conditions (Malim et a1., 1988, supra), while inspection of Figure 8B reveals that comparable levels of total tat protein were synthesized in both the cultures transfected with the mutant expression vectors and those co—transfected with the indicator construction pgTat. This suggests that none of the mutant Rev proteins were toxic to the transfected cells.

This analysis also demonstrates that 5 of the missense mutants and 2 of the deletion mutants of Rev were inactive (Lanes 4-7, 10, 16 and 20), while all other Rev mutants appeared fully able to induce 72 aa Tat expression. Four of the inactive missense mutations are clustered between _53_ amino acid residues 23 and 56 (M4 to H7), while the fifth inactive mutant (pM10) is separated by two fully functional mutants and affects residues and 79. the 21 residues near the C—terminus of Rev (pA11/14) had little or no effect Deletion of either the 4 residues near the N—terminus (pA1/2) or on_rev function. In contrast, deletion of additional sequences between residues 9 and 17 (pal/3) resulted in loss of 533 function.

Example 6: Trans—activation capacity of rev mutants MA, USA), with standards supplied by the manufacturer.

COS cells cultures (35 mm) were co—transfected with 250 ng of pHIV—1Arev and 125 ng of a control vector or one of the modified mutant—containing vectors. Supernatant media were sampled 65 hours after transfection and assayed for p24 Gag protein expression levels. SEAP levels were measured in parallel. The resulting values are recited in Figure 9, with correction for the slight variability observed in supernatant SEAP levels (with mean SEAP activity set at 1.00 units, the observed standards deviation was t 0.14 and the range 1.30 to 0.73). r55- Little variation in the supernatant level of SEAP activity was seen in this experiment, thus demonstrating equivalent transfection efficiency and confirming the lack of mutant induced cellular toxicity.

Transfection of pHIV—1Arev alone resulted in no detectable p24 Gag protein in the culture supernatant (Lane NEG), while co-transfection with a wild—type rev gene expression vector effectively complemented the ability of pHIV—1ARev to induce the secretion of p24 Gag (Lane pcREV). V N Additionally, all Rev mutants testing positive in the pgTAT based assay visualized in Figure 8B achieved a level of activity between 50 and 100 Z of that noted for the wi1d—type rev construction in rescue assay, with the exception of H1, which achieved -30 2 activity, a reduction which may reflect the decreased in vivo stability of the H1 mutant. All mutants that were scored as negative in Figure 88 were fully negative in the rescue assay (i.e., 5 10 pg/ml of p24 Gag), with the exception of H7 which yielded a barely detectable level of supernatant p24 Gag protein.

Example 7: Rev phosphorylation not required for biological activity The HIV-1 Rev protein is phosphorylated at one or more serine residues when expressed in vivo. The mutations delineated in Figure 5 affect each of the 11 serine residues whithin the rev coding sequence.

Thus, the immunoprecipitated [3ZP]—Pi labelled Rev proteins transiently expressed in transfected COS cell cultures were monitored in order to identify the in vivo phosphorylation sites of the HIV-1 rev (Fig. 80). A comparison of the level of [32P]-Pi incorporation into rev with the level of [35S]—cysteine incorporated in a culture transfected in parallel (Fig. 8A) was used to assess the effect of individual mutations on the level of phosphate incorporation. The results of this comparison are depicted in Table 1: _56t Table 1 Phenotypic analysis of HIV-1 rev gene mutants Rev‘ Phosphorylationb Sub—cellularC Trans»dominantd Clone function localization repression wi1d—type ++ ++ N ‘ H1 + ++ ? HZ ++ + N M3 ++ ++ N>C H4 ~ ++ N>C — H5 - 1 C>N — M6 — C>N — M7 — ++ N>C — H8 ++ ++ N>C H9 ++ ++ N ,, H10 — ++ N ++ M11 ++ ++ N M12 ++ + N M13 ++ ++ N M14 ++ ++ N A]./?_ ++ _+_ N A1/3 ~ _ N>C — A9/14 — nd nd A10/14 ~ i N + All/14 ++ : N A12/14 ++ i N A13/14 ++ ++ N ‘ ++, 50-100 Z wild-type (wt); +, 5—SO Z Vt; —, < 5 Z wt; ++, comparable to wt; +, 30-60 2 wt; 1, 5-20 Z Vt —, no detectable phosphorylation; nd, not done; C Y, not detected by immunofluorescence; nd, not done; d ++, highly trans—dominant; +, moderately trans—dominant; —, not detectably trans—dominant Ix) C’) -57..

The analysis summarized in Table 1 identified four missense mutations which resulted in diminished phosphorylation. Of these, H2 and M12 had a moderate effect on phosphate incorporation (~30 Z and -60 Z inhibition, respectively) while H5 and H6 dramatically reduced the level of phosphate incorporation. The possible reasons for the dramatic effect of the H5 and M6 mutations (which do not affect any serine residues) on phosphorylation of re! will_be discussed in more detail below.

To further localize the phosphate receptor serine residues in Rev, a study of the level of phosphorylation of the deletion mutants (pA) (Fig. 8C, lanes 15-20) was performed. This analysis revealed that the pA13/14 mutation was normally phosphorylated while the pA12/14 deletion, and the larger C—terminal deletions, displayed only a low level of phosphorylation (-90 2 inhibition). Similarly, the pA1/2 Rev mutant, which " was effectively labelled with [35S]—cysteine, also displayed a low level of [3’P]~Pi incorporation (-90 Z inhibition). the HIV-1 regulatory protein in transfected cells. _53_ Example 8: Nuclear localization of Rev That the Rev protein is predominantly localized to the nuclei, particularly the nucleoli of expressing cells, was confirmed by analysis of the mutants using indirect immunofluorescence. The resulting phase contrast and corresponding immunofluorescence photographs of fixed, transfected COS cell cultures are shown in Figure 10; —~ The technique of indirect immunofluorescence used to localize the Rev protein within the transfected COS cell was that described in B.R. éullen, Meth. Enzymol. lgg [1987] 684 and B.R. Cullen, J. Virol.

Specifically, the cells were treated with rabbit polyclonal anti—Rev peptide, antiserum followed by rhodamine conjugated goat anti-rabbit IgG.

The primary rabbit anti—Rev antibody was used at a 1:800 dilution. REVI/20 antibody was used for analysis of the majority of Rev mutants, with the exception of cultures transfected with the pM1, pM2, pM3, pal/2 and pal/3 vectors. REV27/51 antibody was used for these. The second antibody, rhodamine—conjugated goat anti-rabbit IgG (Boehringer Mannheim Biochemicals, Indianapolis, IN, USA), was used at a 1:50 dilution.

The Rev mutants displayed four categories of subcellular localization as indicated in the Figure. The categories were: N, nuclear/nucleolar with no detectable cytoplasmic expresion; N > C, slight cytoplasmic expression; N 3 C, clear cytoplasmic expression with some nuclear concentration; C > N, random distribution within the cell.

Representative examples of these distributions are shown in the Figure with the Rev proteins depicted being indicated in the upper right corner of the lower panel of pictures. The localization of each Rev mutant detected by this.assay is set-forth in Table 1.

As shown, the majority of mutants displayed a fully wild—type subcellular localization (N), including the trans-activation negative clone pH10 (Fig. 10D). However, several mutants yielded a very low but detectable level of cytoplasmic fluorescence as typified by pH4 (Fig. 10F). This phenotype did not clearly correlate with biological activity, as both active mutants (pM3, pH8) and inactive mutants (pM4, pM7) displayed this property, Additionally, one fairly extensive deletion mutant, pA1/3, displayed a subcellular localization intermediate between the wild—type pattern and the pattern induced by mutations in the basic domain of Rev (N 2 C, Fig. 10H).

This deletion is, however, not located close to the basic domain. While the reason for this aberrant localization is unclear, it does correlate with the lack of biological activity observed for pA1/3.

Mutation of the arginine rich domain of Rev (pH5, pH6) which displays homology to known nuclear localization signals, resulted in a high level of cytoplasmic Rev protein (C > N). These proteins, however, were not excluded from the cell nucleus (Fig. IOJ), suggesting that the basic domain of Rev is, in fact, a nuclear localization signal. k It is of interest to recall that pH5 and pH6 were not significantly phosphorylated in vivo despite retaining the sites proposed above as acceptors for phosphorylation. This may be because the kinase responsible for the phosphorylation of Rev is confined to the cell nucleus.

Thus, it appears possible that inappropriate subcellular localization of the pnb and pub Rev mutants is responsible for their low level of phosphorylation.

Trans—doninant repression of Rev function Example 9: Rev mutants which had lost the ability to trans—activate HIV-1 structural gene expression were examined for their ability to inhibit in trans the function of the wild~type Rev protein.

COS cells (35 mm) were co~transfected with 250 ng of the indicator construction pgTAT and: 290 ng pBC12/CMV (negative control) (Fig. 11A, lane 1); 40 ng pcREV (low level), 250 ng pBC12/CMV (lane 2); ng pcREV (high level) (lane 3); 40 ng pcREV, 250 ng pM4 (lane 6); ng pcREV, 250 ng pM7 (Lane 5); 40 ng pcREV, 250 ng pM10 (lane 6).

Sixty hours after transfection, the cultures were labelled with [35S]—cysteine and subjected to immunoprecipitation analysis using rabbit anti~Tat antisera. .

As shown in Figure 11A, pM4 (lane 4) and pH? (lane 5) had little effect on the activity of the wild-type Rev protein as measured by the induction of 72 aa Tat expression, whereas pH10 (lane 6) appeared to completely inhibit Rev function. This suggests that pM10 encodes a specific inhibitor of HIV-1 53! gene function.

In order to demonstrate that pM10 indeed acts by specifically preventing the cytoplasmic expression of the unspliced HIV-1 mRNA which encodes the 72 aa form of tat, the S1 nuclease protection assay of Halim et a1., (1988), supra was euployed. This assay (Fig. 118) quantitates the level of spliced (S) and unspliced (U) tat nRNA expressed in the cytoplasm of COS cells (100 mm) transfected with: 2.75 pg pBC12/CHV + 2.5 ug pcREV (Lane 1); 2.5 ug pgTAT + 2.75 vs DBC12/CHV (Lane 2); 2.5 ug pgTAT + 0.25 ug pcREV + 2.5 ug pBC12/CHV (Lane 3); 2.5 ugpgTAT + 2.75 U8 DCREV (Lane 4); 2.5 ug PETAT + 0.25 vs pcREv + 2.5 ug PH10 (Lane 5); 2.5 ug DZFAT + 0:25 pg pcREV + 2.5 ug pA10/14 (Lane 6). As shown, total input DNA was maintained at a total of 5.25 ug by inclusion of the parental expression vector pBC12/CMV as a negative control.

At 60 hours after transfection, cytoplasmic RNA was harvested for analysis and 5 ug aliquots were used in the S1 nuclease protection assay. The DNA probe used herein was a 798 basepair probe end—labelled at _51_ As expected spliced tat nRNA predominates in the cytoplasm of cells transfected with pgTAT alone (Fig. 11B, Lane 2) while unspliced tat mRNA is the dominant cytoplasmic species in the cytoplasm of cells that.

Z unspliced; coexpress the HIV~1 Rev protein (Figure 11B, Lanes 3 and 4). However, coexpression of pgTAT with both pcREV and pH10 restored the cytoplasmic predominance of the spliced form of tat mRNA (Figure 11B, Lane 5).

Although the total level of RNA loaded in Lanes 5 and 6 appear somewhat low, it is thus nevertheless apparent that pglg (Lane 5) was able to selectively inhibit the Rev induced cytoplasmic expression of unspliced tat gggé from the pgTAT vector. mRNA detected in the presence of both pcREV and pH10 is comparable to the Indeed, the relative level of unspliced tat level observed in the absence of Rev.

The same pattern is apparent for pA10/14 (Fig. 11B, Lane 6).

These results additionally demonstrate that a re! gene deletion extending 3' to the H10 mutation (pAl0/14) also encodes a trans—repressor of Rev function. A more extensive deletion which extends through the site of the M10 mutation, termed pA9/14, also displayed a dominant negative phenotype (Table 1). However, a deletion which extended to the site of the H8 mutation was no longer trandominant (data not shown). r62- Example 10: pH10 Rev mutant is a competitive inhibitor of rev function The experimental results presented in Figure 11A and 11B demonstrate that pM1O can repress wild~type Rev function when present in trans. However, these experiments were performed in the presence of a large excess of pM10. To more accurately quantitate the effectiveness of the trans—inhibition of Rev function, the ability of increasing levels of pH10 to inhibit the rescue of the pHIV—1Arev provirus mutant by a single level of pcREV was analyzed. This experiment also tested the effect of increasing levels of the pM4 and pAl0/14 Rev mutants, as well as the effect of simply increasing the level of expression of wi1d—type Rev itself.

COS cell cultures (35 mm) were co—transfected with 25 ng of pHIV-larev and 50 ng of pcREV, with an increasing fold molar excess of , either pcREV (V), pM4 (A), pA10/14 (0) or pM1O (C), as indicated in Figure 12, i.e., 10 fold means co—transfection of 500 ng of the indicated plasmid construction. Total input DNA was maintained at a total of 587.5 ng by inclusion of the parental expression vector pBCl2/CHV as a negative control. A SEAP gene expression vector was co—transfected as an internal control (12.5 ng/culture). The supernatant media were sampled at 65 hours and assayed by measurement of the level of supernatant p24 Gag expression.

The results of this assay are presented in Figure 12 relative to the level of p24 expression obtained in the absence of any competing rev vector, a level defined as 1.00. All values are expressed relative to the p24 Gag expression level observed in the culture transfected with 25 ng pHIV—lArev, 50 ng pcREV, 12.5 ng pBC12/RSV/SEAP and 500 ng pBC12/CHV. This control culture (0), which forms the basal value against which the competitive effects of the added Rev mutants were measured, was arbitrarily assigned a level of 1.00 unit of p24 Gag expression. The values presented herein are corrected for the slight variability observed in supernatant SEAP levels (mean SEAP activity was 1.00 1 0.27 with a range of 1.28 to 0.72).

As depicted, increasing the level of transfection of the wild—type rev expression vector pcREV was found to exert a mildly positive effect on viral replication, leading maximally to an -70 Z increase in the _53_ release of p24 Gag into the media. Co—transfection of the pM1O vector, in contrast, had a dramatically inhibitory effect on HIV-1 structural gene expression. The pattern of inhibition obtained is that expected for a competitive inhibitor of Rev function which displays the same affinity as wild~type Rev for its biological target. Thus, an equimolar amount of pH10 reduced p24 Gag expression -2 fold, a Zfold excess of pM10 reduced expression —3 fold, a Sfold excess -6 fold while a 10fold excess reduced p24 Gag expression by -93 Z.

‘ In addition to pH10, co—transfection of pA10/14 also reduced p24 Gag expression, however, this large deletion mutant of Rev was not as effective an inhibitor as pH10. It appears likely that the severely truncated protein expressed by pA10/14 is a less effective competitor because it lacks sequences which enhance the binding of Rev to its , biological target.

Finally, co~transfection of pH4 also had an inhibitory effect, although it was slight, less than two fold (32 Z), at the maximum dose used.

This effect is thought to arise from an activity (squelching) unrelated to RRE binding. _54_ Example 11: Domain structure of the HIV-1 Rev trans—activator The first of these domains (the RNA binding domain) defined by the four missense mutants pH4, pH5, pH6 and pH7, extends over about a 35 amino acid region between about amino acid position 10 and about amino acid position 68 of wild—type Rev and contains a highly basic sequence element which is essential for the nuclear localization of Rev. It also contains a nuclear localization (NL) signal (shaded). Mutants altered in this domain display a recessive negative phenotype.

In contrast, mutations within the second domain (the activation domain), which is centered approximately on amino acid residue 79, and defined i.a. by the missense mutant pHlO and the deletion mutants pA9/14 and pA10/14, also result in a loss of Rev function but, remarkably, the negative phenotype displayed by these mutants is trans—dominant. Two regions of Rev (hatched in Figure 12A) appear to be dispensable for protein function. As depicted herein, mutations in the Rev activation domain render Rev defective and result in the production of proteins which competitively inhibit wild—type Rev function. This trans—repression is sufficient to markedly reduce or suppress the replication of HIV-1 in transfected cells. The molecular basis for the dominant negative phenotypes displayed by pH10 and the related deletion mutants is not known yet. However, the observation thad both missense (M10) and deletion (A9/14, A10/14) mutants are able to inhibit Rev function in trans does suggest that the loss, rather than the acquisition, of an attribute is responsible. One possibility is that defective Rev protein molecules might be able to form mixed multimers with wild—type Rev protein subunits and hence inhibit the function of the wild—type protein in a transdominant manner. It is not, however, currently .65. known whether Rev functions as a monomer or as a multimer in vivo. An alternative hypothesis, based on earlier work involving the functional dissection of a number of prokaryotic and eukaryotic transcription factors, is that transcriptional trans—activators bear two distinct functional domains, a specific "binding domain" that directs the protein to its appropriate target substrate, and an "activation domain" that permits the functional consequence of the binding event, in this case transcriptional activation, to be displayed. In several systems, the binding domain has been shown to consist of a sequence—specific DNA binding element; however, in at least one case, that of the herpex simplex virus type 1 (HSV—1) trans-activator VP16, it appears that the binding domain instead mediates a specific interaction with a cellular transcription factor which in turn binds to target sequences in the HSV-1 genome. binding domain tends to result in a negative phenotype which is recessive at moderate levels of expression. In contrast, mutation or deletion of the activating domain of a transcription factor may result in mutants with a dominant negative phenotype. These mutants, which retain an intact binding domain, are believed to compete with the wild—type trans—activator for binding to the appropriate cellular target, yet are incapable of activating transcription once binding has occurred. In the case of the HSV—l VPI6 protein, overexpression of such a transdoninant mutant has been shown to inhibit wild—type VP16 function effectively, and hence to preclude replication of HSV—1 in normally permissive cells.

Although the rev gene product is a posttranscriptional trans—regulator of gene expression it appears reasonable that the concept of two distinct functional domains should also be applicable in this case. In RNAS that encode the HIV-1 structural proteins. Mutations in this latter domain might thus result in competitive inhibitors of wild—type Rev function. This is the phenotype observed for the pM10 and pA10/14 mutants Importantly, mutation of the, './1 of Rev and these mutants may indicate the existence of a discrete activation domain. Conversely, the second, more N—termina1 essential region of the Rev protein defined by this mutational analysis may serve the same function as the "binding domains" defined in several transcription factors. Mutants altered in this domain (e.g., pH4, pH7) do in fact display a generally recessive negative phenotype, although a low but significant inhibition is observed at high expression levels. Further transdominant mutants have also been found (see Figure SA and Examples 11a and 11b) which allow the localization of the Rev activation domain to be refined as extending from about amino acid position 68 to about amino acid position 90, particularly from about position 78 to about position 86 and especially from about position 78 to about position 83 or 84 of wild—type Rev.

Example 11a: Further transdominant HIV-1 Rev mutants assays were internally controlled for transfection efficiencies).

In addition to those found as described under Examples 4 to 11 above, the following further mutants have been found to transdoninantly inhibit the vi1d—type HIV>1 Rev function: pH21, pH22, pH27, pH28, pH29 and EH3 .

Phenotypic analysis of further HIV41 53! gene Iutants Table 2 Clone Phenotype‘ pBC12/CMV (vector alone) pM15 _ pM16 pM17 pH18 pH19 pH20 pH21 pH22 pM23 pH24 pM25 pA9/19 p018/19 p018/23 pA22/14 pA23/14 pH27 pH28 pH29 pH32 pH33 pM34 pM35 pM36 ++ ++ ++ ++ ++ ++ ++ ++ ‘ ++ 40-1 + 5-3 < 5 b as Z inhibition of wi1d—type Rev function, with a 10fo1d excess of mutant Rev over vi1d—type Rev Transdominant repression" 91 97 Z (relative to wi1d—type Rev activity) 9 tat expression in the cytoplasm.

Ax) C) _70_ .3.

Example 12: Analyses of Rex function For testing reg mutants for Rex function (Fig. 14), each mutant DNA was cotransfected with pgTAX-LTR into COS cells using DEAE—dextran (B.P. Cullen, Heth. Enzymol. lég [1987] 692-693). All plasmids were added at a concentration of 1.25 ug/ml. Forty-eight hours after transfection, the cells were metabolically labelled with 35S—cysteine for 2 hours, cellular extracts prepared, and the samples were immunoprecipitated with the .5 alpha human monoclonal antibody that specifically reacts with the HTLV—I envelope protein. Immunoprecipitates were analyzed on SDS—1O Z polyacrylamide gels.

For simultaneous analysis of HTLV—I Env, Tax and Rex protein production (Fig. 15), COS cells were cotransfected with pgTAX—LTR case, the radiolabeled cellular extract was used for the three In each immunoprecipitations and electrophoretic analyses; only the relevant region of each of the resultant autoradiograns is presented in each panel of Fig. 15.

For subcellular localizationof HTLV—I Rex mutants by immunofluorescence (Fig. 16), COS cells were transfected with the indicated expression plasmids and fixed with paraformaldehyde 43 hours later. The cells were then sequentially stained with rabbit ant—Rex peptide antiserum (1:100 dilution) and goat anti—rabbit IgG conjugated to rhodamine as previously described (B.R. Cullen [1987] supra).

Example 13: Inhibition of Rex and Rev function For analysis of the ability of £55 mutants to inhibit function of the wild—type Rex protein (Fig. 17A), COS cultures were cotransfected with three plasmids including pgTAX—LTR (1.25 ug/ml), pREX (0.1 ug/ml) and 1 ug/ml of the Rex mutants (Lanes 1~6), pREX (Lane 7), pREV (Lane 8) or pCHV—IL—2 (Lane 9). Env production was analyzed by immunoprecipitation with the 0.5 alpha monoclonal antibody and electrophoresis through SDS—10 Z poly- acrylamide gels. For analysis of inhibition of the function of HIV~1 Rev protein, COS cells were cotransfected with pgTAT, pREV and a 10-fold molar excess of pBC/CMV-IL-2 or the H6, H7 and H13 transdominant Rex mutants.

After 48 hours of culture and biosynthetic labeling with 35S-cysteine, cellular extracts were assayed by immunoprecipitation for Rev—induced , production of the truncated 72 amino acid form of the Tat protein (Lane 2).

At a 10:1 molar ratio, the H6, H7 and H13 (Lanes 3-5) mutants completely inhibited the action of the HIV-1 Rev protein as only the full—length amino acid form of the Tat protein was detected. For analysis of inhibition of replication of HIV-1, COS cells were cotransfected with the reX—deficient HIV-1 proviral plasmid pHXB2—Bam-p3 (H.R. Feinberg et al., Qell fig [1986] 807-817) and pREX in the presence of the indicated fold excess of the H1, H6, H7, and H13 mutants. Total DNA concentration in the transfection cocktail was maintained at a constant level by the addition of varying amounts of the pBC/CHV—IL—2 parental vector. Three days after transfection, supernatant levels of the HIV-1 p24 Gag protein were measured by ELISA (Coulter Immunology kit). The H6, H7 and H13 mutants produced dose—related inhibition of HIV—1 p24 production while the recessive negative H1 mutant did not. _72r PHARHACOLOGICAL ASPECTS HIV~1 is the predominant etiologic agent of AIDS; HTLV—I is causing i.a. ATL; HTLV—II is etiologically related to some cases_of variant T—cell hairy cell leukemia. The HIV-1 Rev and the HTLV—I Rex trans—activators have been shown to be essential for viral replication in culture and Rev and Rex are therefore potential targets for chemotherapeutic intervention in afflicted patients.

Mutant rex and re! genes encoding transdoninant repressors of Further, corresponding transgenic mice appear to be immune to Rex or Rev are thus indicated for use as "intracellular innunogens" for the treatment of diseases caused by HTLV—I or by HTLV—II or, respectively, of HIV—1~induced diseases including AIDS and ARS (ARC). Further, since at least some of the transdominant Rex mutant proteins of the invention have the particular attribute of being able to inhibit both HTLV—I Rex and HIV-1 Rev protein action, they are thus indicated for use in the treatment of infections by both virus types. This property may be of particular value in patients coinfected with more than one of these vital pathogens or in those whose infection has not been distinguished between these two agents. ,73_ The existence of therapeutic agents effective on more than one viral species appears to have been nowhere disclosed prior to the present invention. In view of its broad applicability the above concept appears to be indicated not only in the therapy of the diseases caused by viral species encoding Rev and Rex but also in further viral diseases caused by organisms having genes similarly regulated.

The invention is thus indicated for use in the prophylaxy and therapy of viral, particularly retroviral diseases such as ATL (adult T—cell leukemia), AIDS (acquired immunodeficiency syndrome), ARS or ARC (AIDS—re1ated syndrome or complex), SIV (simian immunodeficiency virus) such as SIV..c, FIV (feline immunodeficiency virus), EIAV (equine infectious anemia virus), visna virus and bovine immunodeficiency virus infections, especially human retroviral diseases, more especially human retroviral diseases caused by pathogens regulated by the reg or re! gene or equivalents thereof, such as ATL, AIDS and ARS (ARC).

Of particular benefit is thereby the multivalent aspect of the repressor effect since it is of advantage in the treatment of multiple, especially double infection by virus, such as is often seen in i.v. drug users coinfected by HIV-1 and HTLV—I, or in treatment in situations of single infection with increased risk of further infection, such as in HIV infection, or in prophylaxy in situations where it is desired to protect against infection by a spectrum of different viral species. above principles of multivalent transdominancy could be applied. _7[,l The therapeutic potential of the invention is immediately apparent, since the repression of e.g. the Rex function of HTLV—I and the Rev function of HIV—1 blocks viral replication, thus preventing the formation of infective virus particles, and is thus expected to perpetuate the latent stage of infection. Thus the cells of subjects already infected with the HIV-1 virus but also having, integrated into their genome, the gene for a transdominant repressor would remain functional and the subjects indefinitely free of symptoms of disease, without the need for long—term therapy.

Viewed in this light the genes according to the invention are pharmaceuticals in themselves, for single or multiple administration either directly in vivo or indirectly in vitro, preferably as part of a vector, e.g. a retroviral or plasmid vector, in a form suitable for achieving delivery in a functional form into target mammalian cells; for example insertion of genes that encode such transdominant inhibitors of viral replication may be effected in vitro into cells of patients by direct implantation into the genome of lymphoid cells derived from infected individuals and these cells may be administered to the donor patient after insertion has been effected. Since ETLV—I and HTLV—II as well as HIV-1 replicate in various types of T—cel1s, the diseases they cause would appear to be particularly suited.

One application of this would thus parallel the gene therapy concept disclosed in e.g. T. Friedmann, Science Eff (16 June 1989) 1275 or P.M. Lehn, Bone Marrow Transplant l (1987) 243: hematopoietic stem cells are extracted from e.g. AIDS/ATL patients and cultivated in vitro, the mutated gene according to the invention, coding for a transdominant repressor for the function to be repressed, such as the Rev/Rex function, is implanted into these cells using retroviral vectors; the now viral—resistant progeny—producing stem cells are returned to the immune system of the original patient, where they are expected to proliferate in view of their acquired selective advantage over non—treated stem cells; in due time the population of hematopoietic cells will consist entirely of cells producing the transdominant factor and be virus—resistant. l75_ Methods on how to effect this are already known in the art, see e.g. USP 4868116. delivering the mutated genes according to the invention into target Vectors, e.g. retroviral or plasmid vectors for mammalian cells such as bone marrow cells are disclosed or referred to in, e.g., Science E33 (16 June 1989) 1275. and/or Rex transdominant genes are cloned into retroviral vector systems.

Thus, for example, various Rev After retroviral—mediated gene delivery into e.g. HIV—infected human cell lines the inhibitory effect of the transdominant mutants is readily ascertained by inhibition of viral production. secdnd column, "Infectious diseases"). kw CD A further mode of using the invention includes insertion not of a gene but of a repressor protein according to the invention into target cells. Administration e.g. orally or parenterally is effected in conventional manner in a form allowing intracellular penetration, such as by liposome—mediated delivery.

For these uses the exact dosage will of course vary depending upon the compound employed, mode of administration and treatment desired; ascertaining the most suitable dosage in a particular situation is within the skill of the man of the art. l\J C? The invention is further indicated for use in the design and engineering of anti—viral drugs based on transdominancy. A means has now been found for manipulating viral gene function which appears to be of general applicability for several viral species, although the structural basis therefor varies widely in the various viruses. This unexpected finding opens the way for studies aimed at designing further specific, possibly low molecular weight, possibly non-peptidic transdominantv inhibitors of viral replication, in particular the design of inhibitors able to mimic the transdominant, i.e. primarily the RNA—binding domain in the mutant Rev or Rex proteins, such as low molecular weight inhibitors or neutralizing monoclonal antibodies.

It is to be understood that various combinations or changes in form and detail can be made to the invention as described above without departing from the scope of the present Claims.

It is also to be understood that further mutants as described above, including mutants among the specific mutants already constructed and disclosed herein but which have not been characterized but may be characterized upon more detailed investigation as being transdominantly inhibitory and/or multivalent, and further mutants in accordance with the principles described above but not specifically disclosed herein, also fall within the scope of the present claims.

Claims (1)

1.Claims A DNA molecule coding for a mutant viral protein which is a modified form of a product of a gene essential for virus function and transdominantly represses the phenotypic expression of one or more of the following genes: i) the wild—type Ex gene of HTLV—l or llTLV—lI, the mutant viral protein being modified from a wild-type form of the Rex protein; and ii) the wild-type rev gene of HIV-1, HIV-2 or SIV, the mutant viral protein being modified a wild—type form of the Rev protein. A DNA according to claim 1, whereby the phenotypic expression of the reg gene of HTLV-I is repressed. A DNA molecule according to claim 1, coding for an HTLV—I Rex protein having a mutation between amino acid positions 30 and 101 of the wild—type Rex protein of