AU6261494A - Delivery system controlled through factors associated with hiv and cell - Google Patents

Description

DELIVERY SYSTEM CONTROLLED THROUGH FACTORS ASSOCIATED WITH HIV AND CELL.

The present invention relates to a method for the prevention or therapy of AIDS. In particular, the invention relates to methods for preventing the replication and spread of HIV and to agents for doing the same.

Human immunodeficiency virus (HIV) has been identified as the etiological agent in human acquired immunodeficiency syndrome (AIDS) (Barre-Sinoussi et al . , 1983; Gallo et al . , 1984) . Conventional therapeutic strategies have concentrated on antiviral drugs such as AZT and on the development of preventive vaccines. However, the intracellular immunisation approach (Baltimore, 1988) has lead to the development of molecular strategies for the inhibition of HIV replication (Malim et al . , 1989, Trono et al . , 1989, Sczakiel et al . , 1991, Sullenger et al . , 1990).

Molecular systems for in vivo cell specific therapy have been described whereby a gene encoding an anti-HIV product can be controlled in its expression by regulatory regions of genes active only in particular cells. This can be achieved in a number of ways, but two of the most attractive in AIDS therapy are tissue-specific expression of the antiviral gene product in cells susceptible to infection by HIV and restriction of this expression to cells actually infected by HIV.

Initial studies on cell-specific ablation therapy have utilised cytotoxic agents such as diphtheria toxin A or ricin A chain genes under the control of lens (Breitman et al . , 1987; Landel et al . , 1988) or pituitary (Behringer et al . , 1988) specific promoters. After icroinjection into mouse embryos and the production of transgenic animals, these constructs resulted in the destruction of either lens or pituitary cells. However, when associated with leaky promoter elements, these toxin genes are unsuitable for somatic therapy because of the constitutive cell lethality and the extreme sensitivity of mammalian cells to diphtheria and ricin toxins.

A more versatile toxin-encoding gene for potential use in human ablative therapy has been described (Borelli et al . , 1988) . The Herpes Simplex Virus type 1 thy idine kinase (tk) gene product is a conditional cell lethal and has been shown to be toxic to mammalian cells only in the presence of nucleoside analogues such as acyclovir (ACV) or ganciclovir (GCV) . These analogues kill actively cycling cells because they possess high affinity for the tk gene product with little or no affinity for endogenous mammalian tk. Model systems have demonstrated in vivo lymphocyte specific lethality by anti-herpetic drug treatment of tk transgenic mice (Borelli et al . , 1988, Heyman et al . , 1989). Specificity of conditional toxicity is due to lymphoid specific transcriptional control elements and quantitative flexibility is inherent within the levels of tk transgene expression and/or administered drug dose. Upon withdrawal of the drug in these studies, mature lymphocytes are restored to normal numbers. Thus, the in vivo ablative system is regenerative, reversible and does not affect stem cells.

Another potential system for use in anti-HIV therapy involves the expression in cells susceptible to HIV infection of a decoy gene.

Decoy genes encode proteins which act as antagonists to natural proteins involved in the replication of the HIV virus. For example, a decoy gene may encode a defective mutant of a transactivator protein which is capable of binding to the transactivator-responsive site on the host or viral genome, yet is incapable of activating transcription.

Transdominant mutations have been reported in a number of viral transactivators which abolish or attenuate the ability of the wild-type protein to transactivate the target gene. Examples include transdominant mutations of E1A (Glen et al . , 1987), tax ( achsman et al . , 1987) and VM65 (Friedman et al . , 1988). Similar mutations in HIV genes have been described for the Tat transactivator (Pearson et al . , 1990) and the Rev transactivator (Bevac et al , 1992).

Expression of such mutant proteins in a HIV-infected cell leads to competition with the natural transactivator and resultant loss of transactivating activity.

A potential disadvantage of the use of decoy gene approaches is that when a decoy is expressed in the absence of the infecting virus a host immune response may result from the production of the decoy gene product, leading to destruction of the host cell.

It is desirable, therefore, in HIV infection, to be able to control the specificity of gene expression so as to limit it to cells susceptible to HIV infection and preferably to cells which are infected by HIV. This end has been pursued in the prior art, not always related to HIV and AIDS, in a number of ways.

Firstly, as described above, promoter and enhancer elements which display at least some degree of tissue-specificity can be selected. However, use of such elements in isolation in gene transfer approaches leads to uncertainty in the levels of expression of the anti-HIV agent which will be achieved, since transgenes are invariably expressed in an integration- site dependent manner. Furthermore, especially in conditional cell ablation approaches, any leaking of the gene in unsuitable tissue-types will be unacceptable.

Locus control regions (LCRs) are elements which confer position-independent, copy number-dependent expression of genes in gene transfer approaches. They have also been shown to permit high levels of expression of cloned genes and to possess tissue-specific properties. First discovered in globin genes (Grosveld et al , 1987) these elements are believed to direct the creation of independent regulatory domains within the chromatin structure of cell genomes, thereby ensuring the activity of a co-transferred gene.

A number of LCRs other than those for globin genes have been described, for example in the CD2 gene in T-lymphocytes (Greaves et al . , 1989), the lysozyme gene in macrophages (Bonifer et al . , 1990) and Class II MHC genes in B cells (Carson and Wiles, 1993).

LCRs are known to be able to direct efficient tissue- specific expression of cloned genes. However, they are not known to be responsive to the infection status of a cell.

The HIV Long Terminal Repeat (LTR) promoter element is composed of cis-acting control sequences; enhancer, transcription start site and the Tat responsive region (TAR) (Rosen et al . , 1985, Jones et al . , 1986; Peterlin et al . , 1986; Garcia et al . , 1987; Muesing et al . , 1987; Siekevitz et al . , 1987; Feng and Holland, 1988; Harrich et al . , 1989). This basic structure is conserved for both HIV-1 and HIV-2 (Guyader et al . , 1987; Arya and Gallo, 1988; Markovitz et al . , 1990). The presence of Tat in transient transfection assays has been shown to dramatically increase the level of HIV LTR directed RNA transcription (Rosen et al.,1985; Muesing et al . , 1987; Laspia et al . , 1989). Tat trans- activation requires the presence of the TAR region which is situated 3' to the LTR and is strictly orientation and position dependent (Peterlin et al . , 1986; Sharp and Marcinial, 1989) . A common mechanism of Tat trans- activation is shared by HIV-1, HIV-2 and SIV (Emerman et al . , 1987; Guyader et al . , 1987; Fenrick et al . , 1989; Shibata et al . , 1990; Berkhout et al . , 1990b). However, the three Tat gene products are not completely interchangeable in their effects. The Tat gene product from both HIV strains and that of SIV effectively trans-activate the HIV- 2 promoter whereas the HIV-1 promoter and SIV promoter are only maximally trans-activated by their own specific Tat protein (Emerman et al . , 1987; Fenrick et al . , 1989; Shibata et al . , 1990). These functional studies are supported by structural studies of the Tat proteins and stem-loop structures of the TAR regions (Fenrick et al . ,1989; Berkhout et al . , 1990b). Thus, the HIV-2 promoter and TAR element function optimally in HIV-l or HIV-2 infected human cells or in SIV infected primate cells.

5* HIV promoter deletion constructs containing the TAR element show that Tat trans-activation of the HIV promoter is retained while basal levels of transcription are dramatically decreased (Arya and Gallo, 1988; Berkhout et al . , 1990; Zeichner et al . , 1991; Berkhout et al . , 1992). The 5* enhancer/promoter element of HIV-l, which contains two NFcB binding sites and three Spi sites, plays a dominant role is basal transcription. NFΛB is a cellular transcription factor found in many cell types including mitogen induced T cells. It has been demonstrated that mitogen induction can increase HIV-l directed basal transcription in CD4+ human T cells (Nabel and Baltimore, 1987) . This increase was localised to the NFcB sites by mutagenesis studies. Elimination of one or more Spi sites has been demonstrated to have an effect on trans-activation and to lower basal transcription from the HIV promoter (Jones et al . , 1986; Harrich et al . , 1989; Berkhout et al . , 1992) . Finally, the TATA sequence, is absolutely necessary for both basal and Tat-induced transcription (Garcia et al . , 1989; Li et al . , 1991; Berkhout et al . , 1991). Although HIV-l, HIV-2 and SIV have a similar arrangement of enhancer, promoter and TAR motifs, the effect of proximal upstream elements on the basal and Tat trans-activated transcription of the HIV-2 promoter has not been extensively analysed.

The targets of HIV infection are primarily CD4+ T- lymphocytes, but also include macrophages, dendritic cells and some cells located in the brain (reviewed in McCune, 1991) . These cells share very few common features except for being derived from common hematopoietic stem cells and susceptibility of HIV infection. Hematopoietic stem cells are not infected by HIV (Molina et al . , 1990; Davis et al . , 1991) . Specific molecular ablation of only HIV infected cells within the hematopoietic system requires differential regulation of the HIV promoter-directed gene expression between HIV infected (Tat positive) and uninfected (Tat negative) cells. While HIV infection is limited at the level of viral entry by CD4, non-CD4+ human cells can actively transcribe HIV sequences, as demonstrated by transfection studies (Barry et al . , 1991). Since therapy may entail the transduction of antiviral sequences into hematopoietic stem cells (thus creating a complete blood system with all precursor and mature cell types carrying the antiviral sequence) , the absence of basal transcription from the promoter in uninfected (Tat negative) cell types is required.

The attempts which have been made to date to use technology as described above for the purposes of antiviral therapy have, in some cases, recognised the existence of these problems. However, they have not been successful in demonstrating a workable solution thereto. For example, in WO90/11359, a Tar-transactivated cell ablation system is defined. However, the basal level of transcription from the system is so high that even HIV-negative cells are killed.

Moreover, all the attempts which have been made to express gene products under the control of HIV promoters in vivo have failed to reproduce T-cell expression as is observed in viral infection. Transgenes controlled by HIV promoters invariably express ectopically in transgenic mice but fail to express in lymphoid tissues, including T-cells (Morrey et al . , 1991; Skowronski, 1991; Salomon et al . , 1994). Such expression patterns would be totally inappropriate for anti- HIV gene therapy as the antiviral agent would be expressed ectopically and most importantly not in T-lymphocytes, which it is desired to protect or specifically ablate in order to arrest the infective process. The methods described in, for example, WO 90/11359 would therefore not lend themselves to applications in gene therapy even if the basal level of transcription of the promoter used could be lowered to acceptable levels.

The use of this system in vivo would result in widespread death of cells other than to those susceptible to HIV infection.

According to a first aspect of the present invention, there is provided a method for the prevention or treatment of a viral infection comprising administering to a subject an effective amount of a delivery system, comprising a vector effective to express in a cell susceptible to infection by the virus a gene encoding an agent which is effective, directly or indirectly, to prevent, eliminate or attenuate the infection, wherein the gene is subject to control between an operative state and an inoperative state through factors specifically associated with the virus and the cell.

The gene encoding the agent according to the method of the invention is subject to regulation between an inoperative state and an operative state. By inoperative state, it is intended to signify that when the gene is in this state, the antiviral effects of the gene product are substantially absent. Where the antiviral gene product is a protein, the absence of antiviral effects will normally be due to a lack of production of the protein, which prevents the overburdening of the protein expression machinery of the cell. For example, if the antiviral agent is a cytotoxic agent in the inoperative state, not enough toxin will be produced to lead to cell death on a significant scale.

Conversely, in the operative state, sufficient antiviral agent is produced to prevent, eliminate or attenuate the infection of the subject by the virus. It is to be understood, however, that regulation between the two states may be achieved by means other than levels of transcription. For example, the splicing of the gene transcript may be regulated.

The antiviral agent used in the method of the invention may be any agent capable of preventing, eliminating or attenuating a viral infection, whether alone or in combination with a second agent. When a second agent is employed, this second agent may be a drug such as a prodrug which is administered to the subject in a non-targeted manner. This second agent may then be activated at the site of infection by the antiviral agent. Suitable antiviral agents include toxins, inhibitors of gene transcription and protein synthesis, immunoactive agents or other drugs capable of directly influencing the metabolism and viability of a cell, as well as alternative factors capable of activating a second agent, such as a drug. For example, the agent could be an enzyme or a ribozyme. Further included are antisense oligonucleotides or oligonucleotide analogues capable of disrupting the expression of a cellular or viral gene and decoy genes.

In general, as used herein the term "vector" signifies the nucleic acid which comprises the gene of the invention, while "delivery system" signifies the means used to introduce the vector to the target cell. These entities may be entirely separate, as in the case of a liposomally- delivered episomal vector, or closely related, as in the case of a viral vector.

In the method of the invention, a vector may be any nucleic acid capable of expressing a gene in a cell. Examples of nucleic acid vectors suitable for use in the present invention include circular and linear lengths of DNA which either integrate into the host genome or are maintained in episomal form, delivered by non-viral delivery systems such as naked DNA, liposomal or receptor-mediated delivery systems, or viral vectors such as retroviral vectors, adeno- associated or adnoviral vectors and other virus-based vectors known in the art.

Preferably, targeting of the vectors may be achieved by designing delivery systems so that they are selectively taken up by cells susceptible to infection by the virus. This may be effected by targeting the delivery systems to specific receptors found on cells infected by the virus or by targeting to progenitor cells such that the delivered gene is subsequently expressed in cells of a particular lineage susceptible to infection by the virus.

The delivery of vectors to pluripotent progenitor cells or the use of non-targeted delivery systems has an obvious drawback, namely that it does not allow selective delivery of the gene of the invention to virus-susceptible cells where more than one cell type derives from the progenitor and not all such cell types are susceptible to infection.

According to the present invention, the above-identified disadvantage is overcome by the use of genes which are under the control of tissue-specific factors. The advantages of the invention may be combined with the use of targeted delivery systems to provide a second level of control in the treatment of viral infections.

It has been found that regulated tissue-specific expression of antiviral agents may be effectively achieved in vivo in hematopoietic cells by the use of LCR sequences. In the treatment of an HIV infection, the use of CD2 and macrophage LCRs is particularly preferred. The use of tissue specific control elements is combined with the use of regulatory elements responsive to a virus associated factor. This will ensure that expression of the antiviral gene takes place only in cells actually infected by the virus. Also, the possibility of a host immune response to the product of the transfected gene, which may be expressed over a long period of time, will be significantly reduced.

This strategy is advantageously combined with a decoy gene as the antiviral agent. Suitable decoy genes for anti-HIV therapy include derivatives of the HIV-encoded proteins Rev, Tat, Vpu, Vpr and Nef. Production of mutated forms of Rev and Tat inhibits infection by the HIV virus. It has been found that by using a tissue-specific LCR, elevated levels of expression of transfected mutant or wild-type proteins may be achieved in lymphoid tissue. Alternatively, agents which function to ablate the cell or arrest viral growth may be used. Such agents include cytotoxic polypeptides, ribozy es, proteolytic enzymes and nucleases.

Furthermore, the use of regulatory elements responsive to virus associated factors allows the use of prodrug approaches for viral inhibition or selective cell ablation, because the antiviral gene will only be in an operative state in the presence of a viral infection.

By virus associated factor, it is intended to refer to a factor, capable of modulating the expression of a gene, which is dependent on the virus. For example, it may be expressed by the cell only in the specific presence (or absence) of a viral infection. Alternatively, it may be encoded by the virus. Preferably, this factor will modulate the transcription of the gene of the invention. However, it is envisaged that the factor may be active by other means, for example by specifically activating a product of the gene of the invention or by acting in unison therewith to potentiate a biological effect. For example, the gene of the invention may be subject to control by a viral transactivator.

Preferably, the gene of the invention is controlled by a viral promoter/enhancer which is transactivatable by the viral transactivator. Optionally, the promoter/enhancer elements responsible for the transactivation may be incorporated into other, non-viral or synthetic promoter/enhancers to confer the transactivatability thereon. Examples of transactivatable viral promoters will be apparent to those skilled in the art and include the adenovirus E1A promoter/enhancer, the hCMV-MIE promoter/enhancer, retroviral LTR elements such as HIV-l and HIV-2 LTRs as well as promoter/enhancers from herpesviruses and poxviruses. In the case of an HIV infection the Tat transactivator may be used. This factor acts via the TAR element which is comprised in a Tat-responsive promoter, such as the LTR. Expression from promoters regulated by the Tar element is greatly increased in the presence of the HIV Tat protein. Accordingly, transcription from the gene of the invention will be greatly increased in a cell infected by HIV, in which the Tat protein will be present.

In order to effect regulation through HIV-associated factors, the use of an HIV-2 LTR promoter is particularly preferred.

Use of the HIV-2 promoter for cell-specific gene expression has been described previously. However, this promoter displays levels of basal transcription which, in combination with certain agents, may be excessive. Accordingly, the HIV-2 promoter is preferably modified in order to reduce the basal levels of transcription.

Modification may be accomplished according to a number of strategies. For example, the use of negative regulatory elements which decrease the level of transcription is envisaged. Especially preferred, however, is the modification of the promoter sequences in order to reduce the basal levels of transcription. Promoter sequences may be modified in order to remove regulatory elements which respond to cell-specific regulatory factors. For example, the HIV-2 promoter contains elements responsive to SPI and NFKB factors. Preferably, therefore, elements responsible for activation by cell-specific factors may be mutated or deleted .

It will be appreciated that it must be ensured that deletion and mutation of elements responsible for activation by cell- specific factors, while reducing the basal level of transcription, must not adversely affect the level of response of the promoter to transactivation by virus- specific factors to an excessive extent. Preferably, therefore, not all the elements responsive to cell-specific factors are removed.

It will be appreciated that modification of the HIV-2 promoter may not be necessary in association with all antiviral agents. For example, a small amount of leakage from the promoter when used to express a decoy gene will not be excessively detrimental to the cells of the subject. However, even a small amount of leakage when the promoter is used to express a cytotoxic agent may prove unacceptable.

In order to put the method of the invention into effect, the invention further provides a polynucleic acid sequence comprising a gene as described above.

Further provided by the present invention are vectors comprising the polynucleic acid sequences according to the invention as well as the use of polynucleic acid sequences and vectors according to the invention in therapy.

Detailed description of the embodiments

The invention is described below, by way of example only, with reference to the following Figures:

Figure 1: Proposed method of HIV-2 tk retroviral mediated HIV specific intracellular molecular ablation.

Figure 2: A: HIV-2 LTR promoter constructs. NF/cB_d and NF/cB refer to the distal and proximal NFcB sites respectively. mSPl refers to a mutated SPI site. The HIV-2 promoter sequences were generated by PCR and then cloned upstream of the human growth hormone (hGH) reporter gene to make pH2ghl to pH2gh8. The same sequences were also cloned upstream of the Herpes Simplex Virus-1 thymidine kinase gene to make plas ids pH2tkl to pH2tk8.

B: Sequence of the HIV-2 promoter constructs.

Figure 3: Differential Tat trans-activated expression from HIV-2 promoter constructs in transiently transfected COS cells. pH2ghl, pH2tk3, pH2tk4, pH2gh5, pH2gh7 and pH2gh8 were co-transfected into COS cells with a human β globin expression plasmid, p/3BSV328, in the absence (-) or presence (+) of Tat. RNA from the transfected cells was hybridised to ³P end labelled DNA probes specific for either the HIV-2tk or HIV-2hGH transcripts and to the 3¹ end of the human β globin gene. The arrows indicate the fragments protected by the correctly initiated HIV-2 hGH transcripts and HIV-2 tk transcripts and the 3 ' end of the β globin transfection control (212 bases) .

Figure 4: HIV-2 retroviral vector design and viral passage analysis of cells infected with HIV-2 tk containing retroviruses. A.) Schematic diagram of the pLA vectors showing the HIV-2 tk sequences cloned in the opposite orientation to the SV40 promoter and the 5' LTR. The Asp718 sites are as indicated. B.) Southern blot analysis on DNA from NIH 3T3 cells infected with retroviruses LAI, LA4 and LA5. Asp718 digested DNA was probed with a Bglll-BamHI fragment containing the tk gene. Plasmid control Asp718 digests are indicated by lanes marked pLAl, pLA4 and pLA5. DNA from populations and clones of infected 3T3 cells are indicated as pop 1, 31.1 and 31.2 for cells infected with LAI, pop 4, 34.1 and 34.2 for cells infected with LA4 and pop 5, 35.1 and 35.2 for cells infected with LA5. C.) Southern blot analysis on DNA from populations of CEM cells (pop 1, pop 4 and pop 5) infected with LAI, LA4 and LA5 probed with the tk gene. The first, third and fifth lanes are the respective plasmid controls. The populations were produced following co-cultivation with producers and 4 weeks of selection in 1-1.5 mg/ml G418. D.) Southern blot analysis on DNA from LAI infected HeLa clones, H1.2-H1.6. Asp718 digested DNA was probed with the tk gene.

Figure 5: Differential Tat trans-activated expression of tk in cells infected with LAI. A.) SI analysis of 3T3 cells infected with LAI and containing Tat. RNAs from an LAI infected 3T3 clone without

Tat (31.1) and three subclones containing Tat, (31.1A, B and C) were analysed by SI nuclease protection after hybridisation to a HIV-2 tk specific DNA probe and β actin probe. 30.ID is a Tat transfected, uninfected cell line. The β actin probe is a normalisation control for the amount of RNA in each lane. B.) SI analysis of CEM cell population (Cl) infected with LAI, before (-) and after (+) transient transfection with pSV2Tat. The RNA was hybridised to HIV-2 tk and β actin specific probes. The β actin probe is used to normalise the amount of RNA in each lane. C.) SI analysis of LAI infected HeLa clones (HI.4, HI.5 and HI.2), before (-) and after (+) transient transfection with pSV2Tat. Analysis for HIV-2 tk and β actin transcripts was carried out as above.

Figure 6: Differential Tat trans-activated ablation of HIV- 2 tk retrovirus-infected 3T3 and HeLa cells. GCV treatment was performed on LAl-infected Tat- negative 3T3 cells (31.1), LAl-infected Tat- positive (31.1A and 31.IB), uninfected Tat- positive (30.ID), and 3T3 cells (A) and LAl- infected (31.1), LA4-infected (34.1), and 3T3 cells (B) . GCV was also administered to LAl- infected HeLa cells without Tat (HI. ) and with Tat (H1.4A and H1.4B) as well as uninfected HeLa cells (C) and LAl-infected (HI.4), LA4-infected (H4), and LA5-infected ,(H5) HeLa cells (D) . Molar concentrations of GCV are as indicated.

Figure 7: A construct comprising the CD2 LCR and a 650 bp Nef fragment, which was used to generate transgenic mice.

Figure 8: A) Southern blot of CD2-nef Transgenic mice.

B) Tissue-specific regulation of nef transgene.

Figure 9: Distribution of CD4 and CD8 cell subsets in lymphoid tissues of CD2 Nef transgenic mice. Thymocytes, lymph node cells and spleen cells from transgenic, (A) line F, 1147 allele and (B) line B, 1191 allele, and non-transgenic littermates were stained for CD4 (PE ordinate) and CD8 (FITC abscissa) . 10^A cells were anlaysed in each sample using the Becton Dickinson FACScan. Relative fluorescence intensities are shown on a logarithmic scale, with percentages of double negative, double positive and CD4/CD8 single positive cells indicated.

Figure 10: Levels of expression of CD4 and other T cell surface markers on Nef transgenic thymocytes. Thymus cells from non-transgenic and transgenic littermates from (A) line F, 1147 allele and (B) line B, 1191 allele were stained with anti-CD4, CD8, CD3 and Thy-1 antibodies. Histogram plots show number of cells (ordinate) and intensity of fluorescence (logarithmic scale, abscissa) for non-transgenic cells (solid line) , transgenic cells (thick line) and no stain controls (dotted line) .

Figure 11: Thymocyte proliferation in Nef-transgenic mice. Varying numbers of thymocytes from non-transgenic (open symbols) and transgenic (solid symbols) littermates of Nef lines F, B, A and D were stimulated with anti-CD3e antibody and PMA and proliferation was measured via [³H]thymidine incorporation. PMA alone produced no activation (not shown) . Activation is plotted as [³H] c.p.m. incorporated versus cell number.

Figure 12: Construction of the HIV-2 TK CD2 transgene.

Figure 13: DNA slot blot showing HIV-2 TK CD2 transgenic mice generated by microinjection of embryos with the HIV-2 TK CD2 transgene.

Figure 14: S₁ analysis of HeLa cell RNA showing the expression of the HIV-2 TK CD2 transgene when transfected into HeLa cells. Figure 15: A: Southern blot analysis of DNA from tissues derived from a 20-copy HIV-2 YK CD2 founder animal;

B: SI protection RNA analysis of RNA derived from tissues of the founder animal of Figure 15A.

Table 1; Basal and Tat trans-activated levels of hGH expression after transient transfection in COS cells. COS cells were co-transfected with pH2ghl to pH2gh8, a β galactosidase expression vector, to normalise transfection efficiency and where indicated the pSV2Tat expression vector. Each transfection was carried out in duplicate or triplicate on at least three separate occasions.

Table 2: Basal and Tat trans-activated levels of hGH expression after transient transfection in CEM cells. CEM cells were co-transfected with pH2ghl to pH2gh8, a β galactosidase expression vector, to normalise transfection efficiency and where indicated the pSV2Tat expression vector. Each transfection was carried out in at least three independent experiments.

Table 3: CD4/CD8 T-cell subset changes in CD2-nef transgenic mice. Thymocytes from three transgenic and three non-transgenic mice from various aged litters were stained with anti-CD4 and anti-CD8 antibodies. 10⁴ stained cells per thymus were analysed on a Becton Dickinson FACScan. Double negative, double positive, CD4 single positive and CD8 single positive populations were gated and displayed as a percentage of total thymocyctes. The mean value and standard deviation for non-transgenic and transgenic mice are shown. Table : CD4 and CD8 down-regulation on T-cell surfaces in CD2-nef transgenic mice. Thymocytes from three transgenic and three non-transgenic mice from various aged litters were stained with anti- CD4 and anti-CD8 antibodies. 10* stained cells per thymus were analysed on a Becton Dickinson FACScan. DN, DP, CD4 SP and CD8 SP populations were gated and then displayed as a single parameter (CD4 or CD8) histogram with cell number (ordinate) against log of fluorescence intensity

(abscissa) . The mean fluorescence value for each population was obtained. These values were averaged for transgenic and non-transgenic littermates and the difference between the two is shown as a percentage of the average fluorescence for the non-transgenic mice. N/C, no change.

Materials and methods

Plasmid constructions

All the HIV-2 sequences used were based on the sequence of the HIV-2 LTR as previously described (Guyader et al . , 1987) . The regions of the HIV-2 LTR were generated by PCR amplification using the plasmid pHIV-2 LTR CAT (-556/+156) (Emerman et al . , 1987). Oligonucleotide primers were synthesised incorporating restriction enzyme sites for ease of cloning. The 3¹ end of each amplified fragment was from primer 9 (3¹) CGTGAACCGGCCACGACCCGTTCATGACCTAGGTG (5') which contains HIV-2 LTR sequences +96 to +116 and a Seal and BamHI restriction site on the end. The 5' ends also contained restriction enzymes for Hindlll, Bgl II and Xhol in primers 1-5 and Hindlll plus Xhol in primers 6-8. The HIV-2 LTR sequence in the primers was as follows (5'-3-):

Primer 2 CCTCATATTCTCTGTTATAAATATACC

Primer 3 AGGGACTTTCCAGAAGGGGCTGGTAACCACTCTGTATAAATATACC Primer 4 GAGGGACATGGGAGGAGC

Primer 5 TGGTGGGGAACGCCCTC

Primer 6 TTCGCCCACATATTCTCTG

Primer 7 AGAAGGGGCTGTAACCAACGTACGTATGCTTCGCCCACATATTCTCTG Primer 8 AGAAGCGGCTGTAACCAACGTACGTATGCTTCGCCCACATATTCTCTG

The constructs shown in Figure 2 were generated from the equivalent primer no. from 1-8 plus in each case primer 9. PCR amplification was carried out in lOmM Tris (pH 8.3), 50 mM KCL, 5mM MgCl₂ and 200 μM dNTPs with 200ng pHIV-2 LTR CAT template and 4ul of Taq poly erase (Boehringer Mannheim) with 94°C denaturation, 55°C annealing and 72°C extension. The amplified products were blunted using Klenow fragment then digested with Hindlll and BamHI before cloning into the Hindlll and BamHI sites of pøGH (Selden et al . , 1986) to make pH2ghl to pH2gh8. The same PCR amplified sequences were also fused to the Herpes Simplex Virus-1 thymidine kinase (tk) gene. pH2tkl, 3, 4 and 5 were made by ligating the HIV-2 LTR Bglll-Scal fragments from the respective pH2gh plasmids to BamHI linearlised pUC19 and a BamHI-Bglll (blunted with Klenow) fragment containing the tk gene (Wagner et al . , 1981). pH2tk7 and 8 were made by ligating the same tk fragment to Hindlll-BamHI linearised pUC19 and the HIV-2 LTR containing Hinglll-Scal fragment from pH2gh7 and pH2gh8 respectively. The constructions were checked by dideoxy sequencing (Sanger et al . , 1977).

Cell culture and transfection

All cells were grown at 37°C in 5% C0₂ and all media were supplemented with Penicillin lOu/ml and Streptomycin lOOμg/ml. COS-7 cells were maintained in DMEM with 10% fetal calf serum and transfected using DEAE-Dextran as previously described (Sambrook et al . , 1989). For transient transfection analysis the pH2gh and pH2tk plasmids (10-15μg DNA) were transfected at a ratio of 4:1 to pSV2Tat (Fenrick et al . , 1989). For tranεfections to be analysed by radio- immunoassay, 5μg actin LacZ was co-transfected as a normalisation control. For those to be assayed by Si analysis, 7.5-15μg of P0BSV328, a human β globin expression plasmid, was co-transfected. p/?BSV328 is a 4.9 kb Bglll fragment containing the human β globin gene cloned into the BamHI site of p/3BSV328 (Grosveld et al . , 1982). HeLa cells were maintained in DMEM with 10% fetal calf serum. They were transfected by the calcium phosphate co-precipitation technique with 48 hr exposure of the cells to the DNA- calcium phosphate co-precipitate and then harvested. CEM cells (human leukemic CD4+ cell line) were maintained in RPMI 1640 with 10% fetal calf serum. Reproducible CEM cell transfection was carried out using electroporation. CEM cells at lx 10⁷ cells/ml in 0.8 ml of serum-free RPMI 1640 were shocked at 500μF and 300V (BioRad Gene Pulser) . For transfection, lOOμg DNA of each pH2gh plasmid, 50μg pSV2Tat and 50μg pCMV LacZ expression vector (Tassios and LaThangue, 1990) as a normalisation control were used. The amphotropic retrovirus packaging cell lines PA317 (Miller and Buttimore, 1986) and AM12 (Markowicz et al . , 1988) were maintained in DMEM with 10% newborn calf serum. For production of infectious recombinant retroviruses the packaging cells were transfected with 12μg of each pLA vector using the calcium phosphate co-precipitation technique. NIH 3T3 cells were maintained in DMEM with 10% newborn calf serum. Retrovirally infected 3T3 cells were transfected with pSV2Tat and a hygromycin expression plasmid (from R. Zamoyska) in a 10:1 ratio using the calcium phosphate co- precipitation technique. 24 hr following transfection, the cells were split and stably transfected clones isolated by selection in 120μg/ml hygromycin B (Sigma) . Human growth hormone and LacZ assays

Human growth hormone reporter gene expression was measured 72 hr after transient transfection of COS and CEM cells. Supernatants were collected from the transfected cells and lOOμl used to assay for secreted human growth hormone in a radioimmunoassay using the Tandem-R HGH system (Hybritech) . From a standard curve produced in each assay the levels of hGH secreted were measured in ng/ml. The small background level in a mock transfected control was subtracted from the results of the transfected samples. The hGH levels were normalised by reference to a co-transfected LacZ expression vector. An actin promoter driving LacZ expression (Gunning et al . , 1987) was used in COS cells and a CMV promoter driving LacZ expression in the CEM cell transfection. Equal volumes of cell lysate for each sample in a transfection were assayed for LacZ activity with the substrate chlorophenol red β-D galactopyranoside (CPRG, Boehringer Mannheim) as previously described (Simon and Lis, 1987) .

RNA analysis

RNA was isolated from cells using 3M lithium chloride/ 6M Urea as previously described (Fraser et al . , 1990). SI analysis was carried out as described by Antoniou et al . , (1988) on 0.5-25μg of total cellular RNA using ³²P end labelled probes. The probe for the HIV-2 hGH transcript was a Pstl-Avall (dephosphorylated end) in pH2ghl and the probe for the HIV-2 tk transcript was a Xhol-Mlul (dephophorylated end) fragment from pH2tkl as shown in Figure 3. The probe (700 bases) for the 3' end of the human β globin gene was an EcoRI-Sall fragment where the EcoRI is within the final exon and labelled by filling with reverse transcriptase in the presence of α³² P dATP. The protected fragment was 212 bases. The same probe for β actin was used for mouse and human cell lines due to cross hybridisation. A Xhol-Aval fragment from pHF/?A-l (Gunning et al . , 1983) was dephosphorylated and end labelled to give a protected fragment of 145 cases with RNA from human cells and 112 bases with RNA from mouse cells. Fold induction of transcripts was quantified by a phosphorlmager (Molecular Dynamics) or Joyce-Loebl Chromoscan densitometer.

Retroviral vector construction, virus production and infections

Retroviral vectors were produced by cloning selected HIV-2 LTR sequences fused to the tk gene from the pH2tk plasmids into pLA. A double stranded oligonucleotide adapter containing 5'-3* Smal, BamHI and Xhol restriction enzyme sites was synthesised. The adapter contained a 5' overhang to anneal to a Xhol site and 3' overhand to anneal to a BamHI site, neither of which will reform the Xhol or BamHI site after ligation. pLA was constructed by ligating Pvul- BamHI linearised pLJ to the adapter and a Pvul-Xhol (residues 419-1560) containing part of the gag region of the Mo-MLV genome. The viruses were made by cloning the HIV-2 tk containing XhoI-BamHI fragment from pH2tkl, pH2tk4 and pH2tk5 and pLA5. Packaging cell lines, PA317 and AM12 were transfected as described above. One day after transfection the packaging cells were split and grown for 10-14 days in 0.4 mg/ml G418 (Gibco-BRL) . G418 resistant colonies were isolated and the higher titre clones identified by analysis of viral RNA in cell supernatants as previously described (Huszar et al . , 1989) . NIH 3T3 and HeLa cells were infected with supernatants of cloned packaging cells in 8μg/ml polybrene and were selected in 0.6 mg/ml G418. CEM cells were infected by co-cultivation with packaging cells as previously described by Stevenson et al . (1988). Infected populations were isolated after 4 weeks in G418 1-1.5 mg/ml. The cells infected by the HIV-2 tk vector were free of replication competent virus after extensive culture as determined by the method described by Markowicz et al . (1988) . DNA analysis

Restriction enzyme digested genomic DNA was transferred to Nytran nylon filters by the Southern procedure. Hybridisation and washing were as described in Sambrook et al . (1989). A 2.8 kb Bglll-BamHI fragment of the tk gene was used as an OLB labelled probe.

Ganciclovir treatment

3 x 10⁴ NIH 3T3 cells in DMEM with 10% fetal calf serum were plated per well in 24-well dishes (Costar) . After two days Ganciclovir (GCV) (Syntex) was added to the medium at various concentrations and this GCV containing medium was changed every two days. After eight days at 37°C, 5% C0₂ the cells were washed and stained with 70% ethanol and 5% methylene blue.

Example I

HIV-2 promoter constructs

Based on the results of HIV promoter studies which demonstrate that it is possible to obtain decreased levels of basal transcription from a modified HIV-l promoter while maintaining Tat trans-activation were constructed plasmids containing HIV-2 promoter sequences. The HIV-2 promoter was used because its TAR region is very responsive to HIV-l, HIV-2 and SIV derived Tat and can provide the widest range of application and testing in the future. Eight minimal promoter constructs (Fig. 2) were generated using PCR, incorporating base substitutions, deletions and consensus sequences in the factor binding sites of the wild type HIV- 2 LTR. These constructs all contain the HIV-2 TATA sequence and TAR region and include the following 5* elements: 1.) the wild type HIV-2 promoter up to -230, 2.) a minimal TATA promoter, 3.) a promoter with 2 NF/cB sites, 4.) one with 3 Spi sites, 5.) another with a single Spi site, 6.) the same Spi site mutated, 7.) the proximal NFcB site and the mutated Spi site and 8.) a mutated proximal NFcB site and a mutated Spi site. During the construction of the promoter sequences, where possible, the spacing between the factor binding elements and the TATA sequence was maintained.

Basal activity of HIV-2 promoter constructs is decreased

The ability of the HIV-2 promoters to direct basal transcription was tested by transient expression of the hGH reporter gene (Selden et al . , 1986) in transfected COS and CEM (CD4 positive human T) cells. As shown in Table 1, the basal levels of hGH expression were decreased in both cell lines with all of the modified promoter constructs. In COS cells the levels of hGH protein ranged from 0.52 to 1.2 ng/ml and are at least 4 fold lower than from the wild type promoter 1 (4.9 ng/ml). The basal levels of hGH protein produced in CEM cells from the modified constructs were 0.25 to 1.9 ng/ml, with the levels decreased 2 to 14 fold in comparison to the wild type promoter (3.5 ng/ml). Overall, the lowest level of basal activity was found with construct 4. Deletion of either NF«B or Spi sites effectively decreased basal activity. When basal activity was measured by analysis of mRNA in an SI protection assay, low levels of reporter gene transcripts were found to correspond to the reduced basal levels of hGH protein.

HIV-2 promoter constructs can be trans-activated to varying levels

The HIV-2 promoter constructs were tested for their ability to be Tat trans-activated in transfected COS and CEM cells. Our results indicate that all the promoter sequences can be trans-activated by HIV-l Tat (but to varying degrees) in both cell lines (Table 1 and 2) . In COS cells the wild type promoter transfectant was optimally trans-activated for hGH expression by a factor of 20 while transfectants receiving constructs 3, 4, 5 and 7 were trans-activated 4 to 10 fold. The levels of hGH protein were decreased from 100 ng/ml as produced by construct 1, to 3, 5, 12 and 3.4 ng/ml from constructs 3, 4, 5 and 7 respectively. In CEM cells, Tat trans-activation greatly increased hGH production. Wild type promoter driven hGH production was increased to 426 ng/ml and the levels of hGH expression from constructs 3, 4 and 5 were raised to 32, 43 and 69 ng/ml respectively. Surprisingly, upon trans-activation, COS and CEM cells containing construct 5 (only one Spi site) produced higher levels of hGH than construct 4 (three Spi sites) . Also, the degree of trans-activation in CEM cells of construct 4 and 5 (153-fold and 138-fold) was higher than the Tat induced trans-activation produced by the wild type construct 1 (123-fold) . Construct 3 with two NFcB sites was trans-activated only by a factor of 43.

Table 1 - HIV-2 directed hGH expression in COS cell transfeotants

pH2gh Tat- fold decrease in Tat+ fold construct hGH ng/ml basal transcription hGH ng/ml trans-activati

4.90 ± 1.8 98.9 ± 49.7 20.2

0.97 ± 0.15 2.66 ± 1.10 2.7

0.76 ± 0.19 3.17 ± 0.66 4.2

0.52 ± 0.24 10 4.79 ± 1.95 9.2

1.20 ± 0.27 11.8 ± 7.5 9.8

0.76 ± 0.29 2.49 ± 1.05 3.3

0.68 ± 0.34 1.51 ± 0.22 2.2

Table 2 - HIV-2 directed hGH expression in CEM cell transfectants

pH2gh Tat- fold decrease in Tat+ fold construct hGH ng/ml basal transcription hGH ng/ml trans-activati

3.5 ± 0.2 426.0 ± 65 123.0

0.25 ± 0.08 14 1.1 ± 0.1 4.5

0.74 ± 0.25 31.9 ± 1.2 43.0

0.28 ± 0.07 12 42.9 ± 8.8 153.0

0.5 ± 0.2 68.9 ± 10.1 138.0

0.6 ± 0.2 7.1 ± 2.5 12.8

1.3 ± 0.6 4.1 ± 0.5 3.1

1.9 ± 0.7 0.8 ± 0.1

Transduction of HIV-2 tk sequences by retroviral vectors

Since the promoter constructs 1, 4 and 5 were the most highly trans-activated whilst having decreased basal activity, these promoters were cloned with the tk gene in the reverse orientation into the retroviral vector pLA to produce pLAl, pLA4 and pLA5 respectively (Fig. 4A) . Amphotropic producer clones LAI, LA4 and LA5 were found to have titers of 5 x 10³ , 1 x 10³ and 5 x 10³ CFU/ml respectively, as assayed by G418 resistance of infected 3T3 cells. When the proviral integrants of LAI, 4 and 5 infected 3T3, HeLa and CEM cells were tested, we found complete, intact viral (8.1 kb) sequences present in the genomic DNA (Fig. 4) . In addition, at a low frequency smaller tk-hybridizing bands (7.3 kb and 6.4 kb) were found, indicating some rearrangement of the retroviral genome. The 7.3 kb proviral integrant was found to delete a sequence (0.8 kb) immediately 3* to the polyadenylation signal of the tk gene.

Trans-activated tk expression occurs after retroviral mediated gene transduction.

In order to examine the effects of retroviral mediated gene transduction on expression of the HIV-2 tk sequences, we examined basal and trans-activated levels of HIV-2 tk mRNA in LAI infected 3T3, HeLa and CEM cells. The results of SI analysis on the 3T3 clone (31.1), CEM population (Cl) and three HeLa clones (HI.2, HI.4 and HI.5) are shown in Figure 5. We tested transcriptional trans-activation of the infected 31.1 clone by analysis of three stably Tat transfected subcloneε (31.1A, 31.IB and 31.1C) and found a 3 fold increase in tk expression (Fig. 5A) . Previous data suggests this to be the maximum level of Tat trans-activation expected in 3T3 cells (Tassios et al . , 1990) . Upon transient transfection of a Tat expression vector into HI.4 and HI.5 clones we observed a 5-6 fold increase in transcripts. The HI.2 clone can also be trans-activated but to a lesser extent, about 2 fold (Fig. 5C) . CEM cells are also trans-activated by a factor of 4 (Fig. 5B) . Using this transient assay only a fraction of the transfected HeLa and CEM cells actually take up the Tat expression vector. The efficiency of uptake as measured by transfection with a CMV-LacZ plasmid and ^-galactosidase staining was determined to be 2-6% (data not shown) . This indicates that the levels of trans-activation in these cells should be increased by an additional 16-50 fold.

Differential ablation of HIV-2 tk infected cells

To determine whether differential killing of retrovirally infected cells occurs in vitro, the cytotoxic effect of ganciclovir (GCV) was tested on the LAI infected 3T3 and HeLa cells after stable transfection of a Tat expression vector. The basal expression of the HIV-2 tk construct 1 resulted in complete ablation of the 3T3 cells (31.1) at GCV concentrations greater than 4 μM (Fig 6A and B) whereas less than 0.3 μM GCV was completely cytotoxic to Tat transfected subclones 31.1A and 31.IB (Fig. 6A) . HeLa cells (HI.4) infected with the same retrovirus were 40 fold more sensitive to GCV than the 3T3 clone and were killed at GCV concentrations greater than 0.1 μM (Fig 6C) .

Concentrations of GCV as low as 20 nM were sufficient to kill Tat expressing HeLa subclones H1.4A and H1.4B. These results demonstrate predictable Tat induced trans-activation of the HIV-2 promoter and differential GCV sensitivity in both mouse and human cells after gene transduction via retroviral vectors.

As a functional test of diminished basal activity from the modified HIV-2 promoter constructs, we compared the GCV sensitivity of LAI infected 3T3 cells with LA4 infected cells (Fig.6B). LAI infected 3T3 cells (31.1) were completely ablated with greater than 4 μM GCV while LA4 infected 3T3 cells (34.1) were less sensitive to the pro-drug and required greater than 16 μM GCV for complete ablation. As shown in Figure 6D, we also compared the GCV sensitivity of LAI infected HeLa cells with LA4 and LA5 infected cells. LAI infected HeLa cells (HI.4) were completely ablated with greater than 0.1 μM GCV but LA4 and LA5 infected cells required GCV concentrations of greater than 1 μM for cytotoxicity. Thus, the cytotoxic effect of GCV on the basal expressing cells is greatly reduced by the use of the modified HIV-2 promoters in the LA4 and LA5 retroviruses.

Example II

CD2 Nef transgenic mice express Nef in thymocytes and peripheral T cells

Four transgenic lines of mice were produced with a construct containing the human CD2 promoter and LCR element and a 621 bp Nef fragment (Figure 7) . Two different alleles of the HIV-l Nef gene were used to examine the effects of Nef in vivo, 1147 with threonine at amino acid position 15 and 1191 with an alanine at position 15. In vitro studies show CD4 downregulation with both alleles (Guy et al . , 1990) but only the 1147 allele is phosphorylated at position 15. DNA from the four lines; A (1147), F (1147), B (1191) and D (1191) was analyzed by Southern blot analysis and copy numbers were determined to be 6, 25, 48 and 26 respectively (Figure 8A) . Slot blot RNA analysis demonstrated expression of the Nef transgene in all four lines (Figure 8B) . Expression was tissue specific and observed only in the thymus and spleen (Figure 8B) . Quantitation of RNA by probing for glucose-6-phosphate dehydrogenase transcripts and phosphorimaging demonstrated that expression levels were consistent with copy number dependent expression in the four different lines. For A:F:B:D having transgene copy numbers 6:25:48:26 respectively, the ratio of RNA expression levels in thymus was 6:22:30:22. Western blot analysis confirmed expression of Nef protein in the spleen (Figure 8C) and the thymus (not shown) of all four transgenic mouse lines.

Thymocytes and peripheral T cells populations are altered in Nef transgenic mice

To examine whether Nef had any effects on the T cells of the transgenic mice, FACS analysis was performed on cells from the thymus and the peripheral lymphoid organs; spleen and lymph nodes. Antibodies specific for CD4 and CD8 were used for analysis of distinct T cell subpopulations. Figure 9 shows representative FACS analysis for transgenic and non-transgenic littermates. In all lines we observed a decrease in the percentage of CD4 single positive (SP) thymocytes and a concomitant increase in the percentage of CD4/CD8 double positive (DP) cells. While the percentages of double negative (DN) and SP CD8 cells remained similar in lines A and D and line B (Figure 9B) , the percentages of CD8 SP cells slightly decreased and the DN cells increased in line F (Figure 9A) . Similar statistically significant changes were found in all lines (Table 3) when comparisons were made of data from twelve litters of mice (three transgenic and three non-transgenic littermates per experiment) . Cell counts of the total number of thymocytes in lines A, B and D showed a decrease of 10% on average, but a large decrease (approximately 70%) in thymocyte cell number was observed in line F (Table 2) . The greatest reduction in absolute numbers for line F was observed in the CD4 SP subset and resulted in a 15 fold fewer cells. The total number of CD8 SP cells was decreased 12 fold while the number of DN cells were not significantly changed. In the D transgenic line although the reduction in total number of thymocytes was less than line F, again a significant decrease was observed in the CD4 SP subset.

When peripheral T cell populations were examined, variable decreases in CD4 SP cells in the lymph nodes and spleen were seen for lines A, B and D. However, line F exhibited large decreases in the percentage of CD4 SP cells in the lymph nodes. Normally non-transgenic lymph nodes contained 51.1% (+/- 3.4) CD4 SP while only 14.7% (+/- 6.7) were found in Nef transgenic lymph nodes, representing a 3.5 fold decrease. A decrease from 23.5% (+/- 1.9) to 8.5% (+/- 3.9) was observed in the CD8 SP subset. In addition, the total number of cells in the spleen was unchanged in the A, B and D lines but decreased in line F by a factor of 1.5 in the CD4 SP compartment. Thus, Nef gene expression consistently results in a decrease of the absolute number of CD4 SP cells in the thymus and in one line this effect extends to the peripheral lymphoid organs.

Nef expression results in a downregulation of CD4 on the surface of developing T cells

Since others have found downregulation of CD4 on T cell lines transfected with the 1147 and 1191 alleles of the Nef gene, we investigated whether T cells in Nef transgenic mice also downregulate CD4 in vivo. FACS histogram analysis revealed that surface levels of CD4 were decreased in thymocytes from Nef transgenic mice (Figure 10) . The downregulation was observed in all four lines (Table 4), for both Nef alleles. As controls, antibodies specific for CD8, CD3 and Thy-1 were used to examine the specificity of the downregulation effects. With both Nef transgenic alleles, CD8 levels were found to be slightly decreased and the normally high CD3 expressing population of thymocytes was found to be greatly reduced in line F and less so in the other three lines. This loss of CD3 high cells correlates well with the loss of CD4 SP cells in the thymus. Thy-1 levels did not change, thus demonstrating the specific effects of Nef.

To determine which subset of cells, was downregulated for CD4 levels, we performed histogram analysis on gated CD4 SP and DP thymocyte populations. In twelve litters of mice, consistent downregulation ranging from 17% to 72% of normal levels of surface CD4 was observed on all DP populations. Thymocytes from the F line showed the greatest levels of downregulation. Some downregulation (but not to the extent in the DP population) was also seen on SP cells ranging up to a 28% decrease from normal levels. When lymph node T cells were examined, some but not consistent downregulation of CD4 was observed only in line F (data not shown) , suggesting that CD4 low expressing cells rarely leave the thymus for the periphery.

Table 4 Cell surface marker downregulation in CD2-Nef transgenit thymocytes

T cell subset % Change in CD4 levels % Change in CD8 levels _ Age (weeks) SP DP DP

1147-A 2.0 -25.4 -41.6 -0.6 -21.

8.0 + 2.7 -16.9 + 4.4 -7.

13.5 -21.8 -39.7 -11.7 -16.

14.5 -21.2 -44.2 -8.4 -20.

1147-F 5.0 -46.0 -72.1 -22.0 -28.

7.0 -38.2 -73.4 -16.8 -31.

90 -30.9 -68.9 -28.7 -24.

1191 B 3.0 N/C -33.8 -8.0 + 11.

11.0 -1.4 -29.4 -10.0 -27.

12.0 -24.8 -42.8 + 5.7 -20.

17.5 -27.6 -47.2 -17.0 -25.

1191-D 5.0 N/C -36.0 +9.0 -16.

6.0 +5.0 -36.0 +9.4 -7.

7.0 -16.9 -40.6 + 3.4 -18.

T cell activation is decreased in Nef transgenic mice

Mitogen induced or anti-CD3e mediated activation assays were performed to examine whether downregulation of CD4 or loss of CD4+ cells had negative effects on thymocyte or peripheral T cell activation. Proliferation of thymocytes as measured by ³H thymidine incorporation after activation via the calcium ionophore (ionomycin) and phorbol ester (PMA) revealed small differences between Nef transgenic and non-transgenic cells of both alleles, demonstrating that the total response of transgenic thymocytes to mitogen is not impaired. However, when cells were activated via the T cell receptor-CD3 complex with anti-CD3e antibodies and PMA, a measureable difference between transgenic and non-transgenic thymocytes was observed (Figure 11B) . This decrease in activation was seen in transgenic lines carrying both the Nef 1147 allele and the 1191 allele. In these experiments the shift of the titration curves to the right clearly indicates that more transgenic thymocytes are required to achieve the equivalent activation levels observed in non-transgenic thymocytes. These changes correlate well with the quantitative loss of SP CD4 cells in the transgenic thymuses (Table 3) . In addition, decreases in CD3e mediated activation were observed in proliferation assays performed on peripheral T cells from Line F (data not shown) and correspond to the decrease in the number of CD4 SP cells in the lymph nodes.

It is apparent that nef is therefore responsible for downregulation of CD4 expression on thymocytes, which we have found to be due to intracellular sequestration of CD4.

Nef is accordingly not useful as a decoy protein when expressed indiscriminately of the infecion status of the cell. Example III

In order to demonstrate tissue-specific expression of a HIV- 2 promoter controlled gene product in T cells in the presence of an LCR, transgenic mice were created carrying a transgene expressing the HSV-1 TK gene under the control of the HIV-2 promoter in the presence of the CD2 LCR.

Construction of an HIV-2 LTR driven transcriptional unit with the inclusion of the CD2 LCR element for high level expression.

A 350 base pair fragment of the HIV-2 LTR containing the two consensus NFkB sites and three Spi sites, the transcriptional start site, the TATA sequence and the TAT binding region, TAR was ligated to a 2.8 kb HSV-1 thymidine kinase (TK) gene. The TK gene contains its own translational initiation codon, polyadenylation and termination sequences but no promoter. A 5.5 kb BamHl-Xbal fragment containing the human CD2 LCR was ligated downstream of the HIV-2 TK sequences. The complete construct (8.7 kb) can be isolated from plasmid sequences by digestion with Xhol.

Creation of HIV-2 TK CD2 transgenic mice

An 8.7 kb Xhol fragment containing the HIV-2 promoter, TK gene and the human CD2 LCR was microinjected into fertilized mouse eggs. Thirty-two offspring were born and tails biopsies were taken nine days after birth. DNA was prepared and Slot blot analysis was performed. As shown in the figure, when a human CD2 probe was hybridised to the Slot blot filter, signal was observed in 7 out of 32 tail DNA samples (columns A and B) . Transgene copy number controls (column C) show signal expected from 0, 1, 10 and 40 copies of the HIV-2 TK CD2 construct. Copy numbers of the transgenic founder mice range from 5 to over 100. Founder mouse 2A containing 20 copies of the HIV tk CD2 transgene (Figure 13) was sacrificed and DNA extracted from various tissues for use in Southern blot analysis. DNAs were digested with Pstl and Xbal to liberate an internal fragment within the transgene and run on a 1% agarose gel. The DNA was transferred onto a nylon filter and hybridized with a CD2 probe. All tissues analysed, namely; thymus (T) , spleen (S) , lymph node (LN) , brain (Br) , heart (H) , kidney (K) , lung (Lg) , liver (Li) and muscle (M) were positive for the transgene indicating that this founder was not a mosaic. The slower migrating band in lane LN indicates only a partial digest of this DNA. Copy number controls 0, 1, 10 and 20 copies per cell are indicated (Figure 15a) .

The HIV-2 TK CD2 transgene expresses highly in HeLa cells

HeLa cells were transiently transfected (24 hours) with the HIV-2 TK CD2 construct (HeLa 2) and with a HIV-2 TK construct (HeLa 1) as a control. At the same time, these cells were cotransfected with a CMV-globin quantiation control and pSV2-TAT for transactivated expression from the HIV-2 promoter. Total RNA was isolated from mock transfected HeLa cells and two transfected populations (HeLa 1 and 2) . When SI nuclease RNA protection was performed, both transfected populations showed a high level of Tat- transactivated TK transgene expression. Cotransfected genes, TAT and globin were also expressed. No expression was observed in the mock transfected HeLa cells. Thus, the HIV-2 TK CD2 construct is expressed in human cell line with no added transcriptional effect due to the presence of the CD2 LCR.

Lymphoid specific expression of the HIV tk CD2 transgene

SI analysis was performed on founder transgenic (tg) mouse 2A and a non-transgenic (non-tg) litter ate using a probe specific for tk transcripts (Figure 15B) . The probe is indicated in lane P and a positive cell line expressing the tk gene in lane +. The ?-actin probe was used to quantitate the amount of RNA in each tissue sample; kidney (K) , spleen (S) , thymus (T) , blood (Bl) , brain (Br) , liver (Li) , lung (Lg) , lymph node (LN) , muscle (M) and skin (Sk) . These results show expression in the lymphoid tissues (lymph nodes, spleen and thymus) and strongly suggest T cell specific expression. Although no expression is seen in the non-lymphoid tissues, some expression is seen in the brain. Basal expression is observed in the absence of tat. We would expect that transcription would increase 100-200 fold, particularly in the lymphoid tissues, with transactivation of the HIV promoter by tat.

References

Ancelle, R. , 0. Bletry, A.C. Baglin, V.F. Brun, M.A. Rey, and P. Godeau. 1987. Long incubation period for HIV-2 infection. Lancet 1: 688-689.

Antoniou, M. , E. deBoer, G. Habets, and F. Grosveld. 1988. The human 9-globin gene contains multiple regulatory regions: identification of one promoter and two downstream enhancers. EMBO J. 7: 377-384.

Arya, S.K., and R.C. Gallo. 1988. Human immunodeficiency virus type 2 long terminal repeat: Analysis of regulatory elements. Proc. Natl. Acad. Sci. USA 85: 9753-9757.

Baltimore, D. 1988. Intracellular Immunisation. Nature 335: 395-396.

Barre-Sinoussi, F. , J.C. Chermann, F. Rey, M.T. Nugeyre, S. Chamaret, J. Gruest, C. Daugret, C. Axler-Blin, F. Vezinet-

Brun, C. Rouzious, W. Rozenbaum, and L. Montagnier. 1983.

Isolation of T lymphotropic retrovirus from a patient at risk for acquired immunodeficiency syndromw (AIDS) . Science

220: 868-870.

Barry, P.A., E. Pratt-Lowe, R.E. Unger, and P.A. Luciw.

1991. Cellular factors regulate transactivation of human immunodeficiency virus type 1. J. Virol. 95: 1392-1399.

Behringer, R.R. , L.S. Mathews, R.D. Palmiter, and R.L. Brinster. 1988. Dwarf mice produced by genetic ablation of growth hormone-expressing cells. Genes and Dev. 2: 453-461.

Berkhout, B., A. Gatignol, A.B. Rabson,and K.T. Jeang. 1990a. Tar-Independent Activation of the HIV-l LTR: Evidence That Tat Requires Specific Regions of the Promoter. Cell 62: 757-767. Berkhout, B., A. Gatignol, J. Silver, and K.T. Jeang. 1990b. Efficient trans-activation by the HIV-2 Tat protein requires a duplicated TAR RNA structure. Nucl. Acids Res. 18: 1839- 1846.

Berkhout, B. , and K.T. Jeang. 1992. Functional roles for the TATA promoter and enhancers in basal and Tat-induced expression of the human immunodeficiency virus type 1 long terminal repeat. J. Virol. 66: 139-149.

Bevec, et al . , 1992, PNAS, 8_7, 9870-9874.

Bonfier, et al . , 1990, EMBO J. , 9,2843-2848.

Borelli, E. , R. Heyman, M. Hsi, and R.H. Evans.1988. Targeting of an inducible toxic phenotype in animal cells. Proc. Natl. Acad. Sci. USA 85: 7572-7576.

Breitman, M.L., S.Clapoff, J. Rossant, L.C. Glode, I.H. Maxwell, and A. Bernstein. 1987. Genetic ablation: targeted expression of a toxin gene causes microphthalmia in transgenic mice. Science 238: 1563-1565.

Carson, S. and Wiles, M. V. 1993. NAR 1, 9: 2065-2072.

Caruso, M. , and D. Klatzmann. 1992. Selective killing of CD4+ cells harboring a human immunodeficiency virus- inducible suicide gene prevents viral spread in an infected cell population. Proc. Natl. Acad. Sci. USA 89: 182-186.

Cheng, Y-C. , E.-S. Huang, J.-C. Lin, E.-C. Mar, J.S. Pagano, G.E. Dutschman, and S.P. Grill. 1983. Unique spectrum of activity of 9-[ (1,3-dihydroxy-2-propoxy)methyl]-guanine against herpesviruses in vitro and its mode of action against herpes simplex virus type 1. Proc. Natl. Acad. Sci., USA 80: 2767-1770.

Culver, K.W., Z. Ram, S. Wallbridge, H. Ishii, E.H. Oldfield, and R.M. Blaese. 1992. In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors. Science 256: 1550-1552.

Davis, B.R., D.H. Schwartz, J.C. Marx, C.E. Johnson, J.M. Berry, J. Lyding, T.C. Merigan, and A. Zander. 1991. Absent or rare human immunodeficiency virus infection of bone marrow stem/progenitor cells in vivo . J. Virol. 65: 1985- 1990.

Dufoort, G. , A.M. Courouce, P.R. Ancelle, and 0. Bletry. 1988. No clinical signs 14 years after HIV-2 transmission via blood transfusion. Lancet 2: 510.

Dzierzak, E.A. , T. Papayannopoulou, and R.C. Mulligan. 1988. Lineage-specific expression of a human _/3-globin gene in murine bone marrow transplant recipients reconstituted with retrovirus-transduced stem cells. Nature 331: 35-41.

Emerman, M. , M. Guyader, L. Montagnier, D. Baltimore, and M.A. Muesing. 1987. The specificity of the human immunodeficiency virus type 2 transactivator is different from that of human immunodeficiency virus type 1. EMBO J. 6: 3755-3760.

Feng, A., and E.C. Holland. 1988. HIV-l tat trans- activation requires the loop sequence within tar. Nature 334: 165-167.

Fenrick, R. , M.H. Malim, J. Hauber, S.Y. Le, J. Maizel, and B.R. Cullen. 1989. Functional analysis of the Tat trans¬ activator of human immunodeficiency virus type 2. J. Virol. 63: 5006-5012.

Fraser, P., J. Hurst, P. Collis, and F. Grosveld. 1990. DNasel hypersensitive sites 1, 2 and 3 of the human yø-globin gene cluster. Nucl. Acids Res. 18: 3503-3507. Friedman et al . , 1988, Nature, 335. 452-454.

Gallo, R.C., S.Z. Salahuddin, M. Popovic, G.M. Shearer, M. Kaplan, B.F. Haynes, T.J. Palker, R. Redfield, J. Oleske, B. Safai, G. White, P. Foster, and P.A. Markham. 1984. Human T-lymphotropic retrovirus, HTLV-III, isolated from AIDS patients and donors at risk for AIDS. Science 224: 500-503.

Garcia, J.A., D. Harrich, E. Soultanakis, F. Wu, R. Mitsuyasu, and R.B. Gaynor. 1989. Human immunodeficiency virus type 1 LTR TATA and TAR region sequences required for transcriptional regulation. EMBO J. 8: 765-778.

Garcia, J.A., F.K. Wu, R. Mitsuyasu, and R.B. Gaynor. 1987. Interactions of cellular proteins involved in the transcriptional regulation of the human immunodeficiency virus. EMBO J. 6: 3761-3770.

Glen et al . , 1987, Mol. Cell. Biol., 4_., 1004-1010,

Greaves et al . , 1989, Cell .56, 979-986.

Grosveld, G.C., E. deBoer, C.K. Shewmaker, and R.A. Flavell. 1982. DNA sequences necessary for transcription of the rabbit β globin gene in vivo . Nature 295: 120-126.

Grosveld et al . , 1987, Cell, 5_1, 975-985.

Gunning, P., J. Leavitt, G. Muscat, S.-Y. Ng, and L. Kedes. 1987. A human J-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA 84: 4831-4835.

Gunning, P., P. Ponte, H. Okayama, J. Engel, H. Blau, and L. Kedes. 1983. Isolation and characterisation of full-length cDNA clones for human alpha, beta, and gamma-actin mRNAs: Skeletal but not cytoplasmic actins have an a ino-terminal cysteine that is subsequently removed. Mol. Cell Biol. 3: 787-795.

Guyader, M. , M. Emerman, P. Sonigo, F. Clavel, L. Montagnier, and M. Alizon. 1987. Genome organisation and trans- activation of the human immunodeficiency virus type 2. Nature 326: 662-669.

Harrich, D. , J. Garcia, R. Mitsuyasu, and R. Gaynor. 1990. TAR independent activation of the human immunodeficiency virus in phorbol ester stimulated T lymphocytes. EMBO J. 9: 4417-4423.

Harrich, D. , J. Garcia, F. Wu, R. Mitsuyasu, J. Gonazalez, and R. Gaynor. 1989. Role of SPl-binding domains in in vivo transcriptional regulation of the human immunodeficiency virus type 1 long terminal repeat. J. Virol. 63: 2585-2591.

Heyman, R. , E. Borelli, J. Lesley, D. Anderson, D.D.Richman, S.M. Baird, R. Hyman, and R.H. Evans. 1989. Thymidine kinase obliteration: Creation of transgenic mice with controlled immune deficiency. Proc. Natl. Acad. Sci. USA 86: 2698-1702.

Huszar, D. , J. Ellis, and A. Bernstein. 1989. A rapid and accurate method for determining the titer of retrovirus vectors lacking a selectable marker. Technique 1: 28-32.

Jones, K.A. , J.T. Kadonga, P.A. Luciw, and R. Tijan. 1986. Activation of the AIDS retrovirus promoter by the cellular transcription factor, Spi. Science 232: 755-759.

Korman, A.J., J.D. Frantz, J.L. Strominger, and R.C. Mulligan. 1987. Expression of human class II major histocompatibility complex antigens using retrovirus vectors. Proc. Natl. Acad. Sci. USA 84: 2150-2154.

Landel, C.P., J. Zhao, D. Bok. and G.A. Evans. 1988. Lens- specific expression of reco binant ricin induces developmental defects in the eyes of transgenic mice. Genes and Dev 2: 1168-1178.

Laspia, M.F., A.P. Rice and M.B. Mathews. 1989. HIV-l Tat protein increases transcriptional initiation and stabilises elongation. Cell 59: 283-292.

Li, Y.C., J. Ross, J.A. Scheppler, and B.J. Franza. 1991. An in vitro transcription analysis of early responses of the human immunodeficiency virus type 1 long terminal repeat to different transcriptional activators. Mol. Cell. Biol. 11: 1883-1893.

Malim, M.H., S. Bohnlein, J. Hauber, and B.R. Cullen. 1989. Functional dissection of the HIV-l rev trans-activator - derivation of a trans-dominant repressor of rev function. Cell 58: 205-214.

Markovitz, D.M., M. Hannibal, V.L. Perez, C. Gauntt, T.M. Folks, and G.J. Nabel. 1990. Differential regulation of human immunodeficiency viruses (HIVs) : a specific regulatory element in HIV-2 responds to stimulation of the T-cell antigen receptor. Proc. Natl. Acad. Sci. USA 87: 9098-9102.

Markowicz, D. , S. Goff, and A. Bank. 1988. Construction and use of a safe and efficient amphotropic packaging cell line. Virology 167: 400-406.

McCune, J.M. 1991. HIV-l: the infective process in vivo . Cell 64: 351-363.

Miller, A.D. 1990. Progress toward human gene therapy. Blood 76: 271-278.

Miller, A.D. , and C. Buttimore. 1986. Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production. Mol. Cell. Biol. 6: 2895-2902. Molina, J.-M., D.T. Scadden, M. Sakaguchi, B. Fuller, A. Woon, and J.E. Groopman. 1990. Lack of evidence for infection or of effect on growth of hematopoietic progenitor cells after in vivo or in vitro exposure to human immunodeficiency virus. Blood: 76: 2476-2482.

Morrey, J. D. , et al . , 1991. J. Virol., 65, 5: 5045-5051

Muesing, M.A., D.H. Smith, and D.J. Capon. 1987. Regulation of mRNA accumulation by a human immunodeficiency virus trans-activator protein. Cell 48: 691-701.

Nabel, G. , and D. Baltimore. 1987. An inducible transcription factor activates expression of human immunodeficiency virus in T cells. Nature 326: 711-713.

Nabel, G.J., S.A. Rice, D.M. Knipe, and D. Baltimore. 1988. Alternative mechanisms for activation of human immunodeficiency virus enhancer in T cells. Science 239: 1299-1302.

Parrott, C. , T. Seidner, E. Duh, J. Leonard, T.S. Theodore, B.-W.A., M.A. Martin, and A.B. Rabson. 1991. Variable role of the long terminal repeat Spl-binding sites in human immunodeficiency virus replication in T lymphocytes. J. Virol. 65: 1414-1419.

Peterlin, B.M. , P.A. Luciw, P.J. Barr, and M.D. Walker. 1986. Elevated levels of RNA can account for the trans- activation of human immunodeficiency virus. Proc. Natl. Acad. Sci. USA 83: 9734-9738.

Platika, D. , E.A. Dzierzak, and R.C. Mulligan. 1989. Inducible selective ablation of HIV-tat expressing cells. J. Cell. Biochem. 16: 40.

Rosen, C.A. , J.G. Sodroski, and W.A. Haseltine. 1985. The location of cis-acting regulatory sequences in the human T cell lymphotropic virus type III (HTLV-lll/LAV) long terminal repeat. Cell 41: 813-823.

Ross, E.K., B.-W.A., A.B. Rabson, G. Englund, and M.A. Martin. 1991. Contribution of NF-kappa B and Spl-binding motifs to the replicative capacity of human immunodeficiency virus type l: distinct patterns of viral growth are determined by T-cell types. J. Virol. 65: 4350-4358.

Salomon, B. , et al . , 1994, J. Immunol. (In Press)

Sambrook, J. , E.F. Fritsch, and T. Maniatis. 1989. Molecular cloning: A handbook, 2nd edition (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press) .

Sczakiel, G. , and M. Pawlita. 1991. Inhibition of human immunodeficiency virus type 1 replication in human T cells stably expressing antisense RNA. J. Virol. 65: 468-472.

Selden, R.F., K.B. Howie, M.E. Rowe, H.M. Goodman, and D.D. Moore. 1986. Human growth hormone as a reporter gene in regulation studies employing transient gene expression. Mol. Cell. Biol. 6: 3173-3179.

Sharp, P.A., and R.A. Marciniak. 1989. HIV TAR: an RNA enhancer? Cell 59: 229-230.

Shibata, R. , T. Miura, M. Hayami, K. Ogawa, H. Sakai, T. Kiyomasu, A. Ishimoto, and A. Adachi. 1990. Mutational analysis of the human immunodeficiency virus type 2 (HIV-2) genome in relation to HIV-l and simian immunodeficiency virus SIV (AGM) . J. Virol. 64: 742-747.

Siekevitz, M. , S.F. Josephs, M. Dukovich, N. Peffer, W.-S. F., and W.C. Greene. 1987. Activation of the HIV-l LTR by T cell mitogens and the trans-activator protein of HTLV-I. Science 238: 1575-1578. Simon, J.A., and J.T. Lis. 1987. A germline transformation analysis reveals flexibility in the organisation of heat shock consensus elements. Nucl. Acids. Res. 15: 2971-2988.

Skowronski, J. 1991. J. Virol., 65, 2: 754-762.

Stevenson, M. , C. Meier, A.M. Mann, N. Chapman, and A. Wasiak. 1988. Envelope glycoprotein of HIV induces interference and cytolysis resistance in CD+4 cells: mechanism for persistence in AIDS. Cell 53: 483-496.

Sullenger, B.A., H.F. Gallardo, G.E. Ungers, and E. Gilboa. 1990. Overexpression of TAR sequences renders cells resistant to human immunodeficiency virus replication. Cell 63: 601-608.

Tassios, P.T., and N.B. LaThangue. 1990. A multiplicity of differentiation regulated ATF site-binding activities in embryonal carcinoma cells with distinct sequence and promoter specificities. New Biol. 2: 1123-1134.

Trono, E., M.B. Feinberg, and D. Baltimore. 1989. HIV-l gag mutants can dominantly interfere with the replication of the wild type virus. Cell 59: 113-120.

Ventakesh, L.K., M.Q. Arens, T. Subramanian, and G. Chinnadurai. 1990. Selective induction of toxicity to human cells expressing human immunodeficiency virus type 1 Tat by a conditionally cytotoxic adenovirus vector. Proc. Natl. Acad. Sci, USA 87: 8746-8750.

Wachsman et al . , 1987, Science, 235, 674-677.

Wagner, M.J., J.A. Sharp and W.C. Summers. 1981. Nucleotide sequence of the thymidine kinase gene of herpes simplex type 1. Proc. Natl. Acad, Sci. USA 78: 1441-1445. Wu, F.K., J. Garcia, R. Mitsuyasu, and R. Gaynor. 1988. Alterations in binding characteristics of the human immunodeficiency virus enhancer factor. J. Virol. 62: 218- 225.

Zeichner, S.L., J.Y. Kim, and J.C. Alwine. 1991. Linker- scanning mutational analysis of the transcriptional activity of the human immunodeficiency virus type 1 long terminal repeat. J. Virol. 65: 2436-2444.

Claims (27)

CLAIMS :

1. A delivery system comprising a vector effective to express in a cell susceptible to infection by the virus a gene encoding an agent which is effective, directly or indirectly, to prevent, eliminate or attenuate the infection, wherein the gene is subject to control between an operative state and an inoperative state in the cell through factors specifically associated with the virus and the cell for use in medicine.

2. A delivery system according to claim 1 wherein the gene comprises a first regulatory element which is responsive to tissue-specific factors and a second regulatory element which is responsive to virus-specific factors.

3. A delivery system according to claim 2 wherein the first regulatory element comprises one or more LCRs or functionally equivalent fractions thereof.

4. A delivery system according to claim 3 wherein the LCR is the CD2 LCR or a macrophage LCR.

5. A delivery system according to any one of claims 2 to 4 wherein the second regulatory element is responsive to a virus-encoded factor.

6. A delivery system according to claim 5 wherein the virus-encoded factor is a viral transactivator.

7. A delivery system according to claim 5 or claim 6 wherein the regulatory element is an optionally modified viral promoter.

8. A delivery system according to claim 7 wherein the viral promoter is the HIV-2 LTR promoter.

9. A delivery system according to of claim 8 wherein the HIV-2 LTR promoter is modified to reduce the basal level of transcription.

10. A delivery system according to claim 9 wherein the HIV-2 LTR promoter is the construct 4 or construct 5 identified in Figure 2.

11. A delivery system according to any preceding claim wherein the cell is a pluripotent progenitor cell.

12. A delivery system according to any preceding claim where the cell is a hematopoietic cell.

13. A delivery system according to any preceding claim wherein the delivery system is a non-viral delivery system.

14. A delivery system according to claim 13 wherein the delivery system is a liposomal, receptor-mediated or DNA transfection delivery system.

15. A delivery system according to any one of claims 1 to 12 wherein the delivery system is a viral delivery system.

16. A delivery system according to claim 15 wherein the delivery system is a retroviral, adenoviral or adenoviral- associated delivery system.

17. A delivery system according to any preceding claim wherein the agent effective to prevent, eliminate or attenuate the viral infection is a decoy protein.

18. A delivery system according to claim 14 wherein the decoy is an optionally modified form of Rev, Tat, Vpu, Vpr or Nef encoded by HIV.

19. A delivery system according to any one of claims 1 to 16 wherein the agent effective to prevent, eliminate or attenuate the viral infection is a cytotoxic agent.

20. A delivery system according to claim 19 wherein the cytotoxic agent acts in cooperation with a second agent.

21. A delivery system according to claim 20 wherein the second agent is a prodrug.

22. A polynucleic acid sequence comprising a gene encoding an agent which is effective, directly or indirectly, to prevent, eliminate or attenuate a viral infection in a cell, wherein the gene is subject to control between an operative state and an inoperative state through factors specifically associated with the virus and/or the cell, as described in any preceding claim.

23. A vector comprising the polynucleic acid sequence of claim 22.

24. A host cell transformed with a vector according to claim 23.

25. The polynucleic acid sequence of claim 22 or the vector of claim 23 for use in therapy.

26. Use of a delivery system according to any one of claims 1 to 21, a polynucleic acid sequence according to claim 22 or a vector according to claim 23 in the manufacture of a composition for use in antiviral therapy.

27. A method for the prevention or treatment of a viral infection comprising administering to a subject an effective amount of a delivery system according to any one of claims 1 to 21, a polynucleic acid sequence according to claim 22 or a vector according to claim 23.