WO2000052188A1 - Packaging cells for retroviral vectors - Google Patents

Abstract

A packaging cell and producer cell is described. The packaging cell comprises a first nucleotide sequence (NS) encoding a toxic viral envelope protein; and a second NS encoding a retrovirus nucleocapsid protein; wherein the expression of the first NS is regulatable in a temperature range from about 25 °C to about 40 °C. The producer cell comprises a packaging cell comprising a viral sequence sufficient to produce viral vector particles from the cell; wherein the viral vector particle titre obtained is dependent upon the temperature regulatable expression of the toxic viral envelope protein.

Description

PACKAGING CELLS FOR RETROVIRAL VECTORS

The present invention relates to packaging and producer cell lines for producing recombinant viral vectors. In particular, the present invention relates to methods for producing pseudotyped viral vectors with a broad host range which can be produced at in temperature regulated packaging and/or producer cells. Most specifically, the invention relates to the generation of pseudotyped retroviral vectors, from stable producer cell lines, having vesicular stomatitis virus-G protein (VSV-G) as the membrane-associated viral envelope protein

Retroviruses and vectors derived from them require an envelope protein in order to transduce efficiently a target cell. The envelope protein is expressed in the cell producing the virus or vector and becomes incorporated into the virus or vector particles. Retrovirus particles are composed of a proteinaceous core derived from the gag gene that encases the viral R A. The core is then encased in a portion of cell membrane that contains an envelope protein derived from the viral env gene.

The envelope protein is produced as a precursor, which is processed into two or three units. These are the surface protein (SU) which is completely external to the envelope, the transmembrane protein (TM) which interacts with the SU and contains a membrane spanning region and a cytoplasmic tail (Coffin 1992 In The Retroviridae, Pleum Press, ed Levy). In some retroviruses a small peptide is removed from the TM.

In order to act as an effective envelope protein, capable of binding to a target cell surface and mediating viral entry, the envelope protein has to interact in a precise manner with the appropriate receptor or receptors on the target cell. This must occur in such a way as to result in internalisation of the viral particle in an appropriate manner to deliver the genome to the correct compartment of the cell to allow a productive infection to occur. There have been many attempts to use the envelope protein derived from one virus to package a different virus, this is known as pseudotyping. The efficiency of pseudotyping is highly variable and appears to be strongly influenced by interactions between the cytoplasmic tail of the envelope and the core proteins of the viral particle. The process by which envelope proteins are recruited into budding virions is poorly understood, although it is known that the process is in some ordered as most cellular proteins are excluded from retroviral particles (Hunter 1994 Semin. Nirol. 5:71-83). In some retroviruses budding may occur in the absence of envelope proteins indicating that the env is not necessary for this process but conversely the core can influence the efficiency of envelope incorporation into the particle (Einfeld 1996 Curr. Top. Microbiol. Immunol. 214:133-176; Krausslich and Welker 1996 Curr. Top. Microbiol. Immunol. 214:25-63).

By way of example, there is evidence for a precise molecular interaction between a cytoplasmic domain of the envelope protein and the viral core in some retroviruses.

Januszeski et al (1997 J. Nirol. 71 : 3613-3619) have shown that minor deletions or substitutions in the cytoplasmic tail of the murine leukemia virus (MLN) envelope protein strongly inhibit incorporation of the envelope protein into viral particles. In the case of HIN-1, Cosson (1996 EMBO J. 15:5783-5788) has shown a direct interaction between the matrix protein of HIN- 1 and the cytoplasmic domain of its envelope protein. This interaction between the matrix and envelope protein plays a key role in the incorporation of the envelope protein into budding HIN-1 virions. This is shown by the fact that visna virus can only be efficiently pseudotyped with HIN-1 envelope protein if the amino terminus of the matrix domain of the visna virus gag polyprotein is replaced by the equivalent HIN-1 matrix domain (Dorfman et al, 1994 J.

Nirol. 68:1689-1696).

However the situation is complex, since truncation of the HIN-1 envelope protein is required for efficient pseudotyping of Molony murine leukemia virus (Mammano et al, 1997 J. Nirol. 71 :3341-3345), whilst truncation of the human foamy virus envelope protein reduced its ability to pseudotype murine leukemia virus (Lindemann et al, 1997 J. Nirol. 71 :4815-4820). There is also an environmental component to the interaction between the core of a retrovirus and the cytoplasmic tail of its envelope protein. Prolonged passage of EIAN in some cell lines results in a truncation of the glycoprotein, suggesting that host cell factors can select for a virus on the basis of the C-terminal domain of the envelope protein (Rice et al, 1990 J. Nirol. 1990 64: 3770- 3778).

These studies and those of many other workers indicate that it is not possible to predict that even closely related retroviruses may be able to pseudotype each other. Further more, if a given envelope protein fails to pseudotype a particular virus, it is not possible to predict the molecular changes that would confer the ability to pseudotype. Pseudotyping has met with some success, but is clearly constrained by the need for compatibility between the virus components and the heterologous envelope protein.

In the construction of retroviral vectors it is desirable to engineer vectors with different target cell specificity's to the native virus, to enable the delivery of genetic material to an expanded or altered range of cell types. One manner in which to achieve this is by engineering the virus envelope protein to alter its specificity. Another approach is to introduce a heterologous envelope protein into the vector to replace or add to the native envelope protein of the virus.

The MLN envelope protein is capable of pseudotyping a variety of different retroviruses. MLN envelope protein from an ampho tropic virus allows transduction of a broad range of cell types including human cells.

The envelope glycoprotein (G) of Nesicular stomatitis virus (NSN), a rhabdovirus, is another envelope protein that has been shown to be capable of pseudotyping certain retroviruses. Its ability to pseudotype MoMLN- based retroviral vectors in the absence of any retroviral envelope proteins was first shown by Emi et al (1991 Journal of Virology 65:1202-1207). WO94/294440 teaches that retroviral vectors may be successfully pseudotyped with NSN-G. These pseudotyped NSN-G vectors may be used to transduce a wide range of mammalian cells. Even more recently, Abe et al (J Nirol 1998 72(8) 6356-6361) teach that non-infectious retroviral particles can be made infectious by the addition of NSN-G.

Burns et al (1993 Proc. Νatl. Acad. Sci. USA 90: 8033-7) successfully pseudotyped the retrovirus MLN with NSN-G and this resulted in a vector having an altered host range compared to MLN in its native form. NSN-G pseudotyped vectors have been shown to infect not only mammalian cells, but also cell lines derived from fish, reptiles and insects (Burns et al 1993 ibid). They have also been shown to be more efficient than traditional ampho tropic envelopes for a variety of cell lines (Yee et al, 1994 Proc. Νatl. Acad. Sci. USA 91 : 9564-9568, Lin, Emi et al, 1991 Journal of Nirology 65:1202-1207). NSN-G protein can be used to pseudotype certain retroviruses because its cytoplasmic tail is capable of interacting with the retroviral cores.

The provision of a non-retro viral pseudotyping envelope such as NSN-G protein gives the advantage that vector particles can be concentrated to a high titre without loss of infectivity (Akkina et al, 1996 J. Nirol. 70: 2581-5). Retrovirus envelope proteins are apparently unable to withstand the shearing forces during ultracentrifugation, probably because they consist of two non-covalently linked subunits. The interaction between the subunits may be disrupted by the centrifugation. In comparison the NSN glycoprotein is composed of a single unit. NSN-G protein pseudotyping can therefore offer potential advantages.

However, there are certain disadvantages involved in using producer cell lines to manufacture retrovirus vectors pseudotyped with NSN-G. The first is the difficulty in producing stable cell lines that express NSN-G; the second is the limited life spans of such cell lines.

A number of workers have reported that constitutive high-level expression of NSN-G is toxic to most mammalian cells (eg Emi et al, 1991 Journal of Nirology 65:1202- 1207, Yee et al, 1994 Proc. Νatl. Acad. Sci. USA 91 : 9564-9568). A variety of approaches have been used to solve these problems. By way of example, Yee et al (1994 Proc. Νatl. Acad. Sci. USA 91 : 9564-9568) developed a scheme for producing NSN-G pseudotypes by first producing 293 cell lines that constitutively express gag- ol proteins and contain a retroviral genome. These cell lines were then transfected with a plasmid containing the VSN-G gene downstream of a human cytomegalovirus immediate early promoter followed by the splicing and polyadenylation signals derived from the rabbit β-globin gene. Maximal production of transducing particles was obtained between 48 and 72 hours after transfection.

WO96/35454 teaches that a tetracycline responsive promoter may be used in combination with a nucleotide sequence enocoding vesicular stomatitis virus (NSN-G) to derive a retroviral packaging cell line that inducibly expresses the NSG-G protein at levels sufficient to support high level virus production but without the toxic effects of constitutive expression of NSN-G. Ory et al (1996 Proc. Νatl. Acad. Sci. USA 93:11400-11406) used the tetR/NP 16 transactivator and tetracycline responsive operator (tetO) minimal promoter system for inducible, tetracycline-regulatable expression of NSN-G in the production of packaging 293 cell lines. Yang et al (1995 Human Gene Therapy 6:1203-1213) used a similar strategy linking seven copies of the tetO to a minimal HCMN promoter to construct packaging lines derived from ΝIH-3T3 cells. Chen et al (1996 Proc. Natl. Acad. Sci. USA 93: 10057-10062) modified the tetracycline-inducible system (Gossen & Bujard, 1992 Proc. Natl. Acad. Sci. USA 89: 5547-5551) by fusing the ligand binding domain of the estrogen receptor to the carboxy terminus of a tetracycline-regulated transactivator. Using this system, they constructed cell lines that expressed VSV-G in the absence of tetracycline. VSV-G expression could be induced by β-estradiol regardless of whether the cells were grown with or without tetracycline. However, induction of VSV-G expression was higher when tetracycline was not present. This allowed the construction of stable packaging cell lines that produced transducing viral particles.

Yoshida et al (1997) developed an adenovirus system to produce MoMLV vectors pseudotyped with VSV-G. First a cell line was produced containing a genome plasmid. Secondly this cell line was infected with three different adenoviruses, one encoding the gag-pol gene of MoMLV under the control of the tetracycline transactivator, the second encoding VSV-G under the control of the tetracycline transactivator and the third encoding a nuclear localising transactivator. Transducing particles could be harvested from the resultant cells for a limited time period. Other researchers developing systems to study the export and processing of VSV glycoprotein mutants have used vaccinia virus systems in which the glycoprotein gene was cloned downstream of a bacteriophage T7 promoter. Co-infection of cells with the glycoprotein encoding vaccinia and a vaccinia virus expressing T7 polymerase resulted in a high level of expression of the VSV-G protein (Lefkowitz et al, 1990, Virology 178;373-383).

Arai et al (1998 J. Virol. 72:1115-1121) commented on the fact that the cell lines, in which the expression of VSV-G was controlled by the tetracycline-inducible system, produced low titres of transducing particles in the presence of tetracycline when VSV- G expression should be repressed. This leaky virus production by these packaging cell lines before induction could cause both virus re-entry into the cell culture and accumulation of the vector DNA in the chromosomes during the process of selection and subsequent passages of the packaging cell lines harbouring the virus vector.

Arai et al (1998) reported the development of packaging cell lines in which a completely silent gene for the VSV glycoprotein was present to negate the above problem. This was achieved using a system in which a cassette was produced which encoded the CAG (the chicken β-actin gene promoter connected with the cytomegalo virus immediate-early promoter) followed by the 5' loxP sequence followed by the neo gene with an associated poly A signal followed by the 3' loxP sequence followed by the coding sequence for VSV-G and an associated poly A signal. When transfected into cells only the neo gene product is produced. If an adenovirus encoding the Cre recombinase is then introduced into the cell the neo sequence is removed by recombination and the VSV-G gene is expressed from the CAG promoter.

None of these approaches will actually solve the problem associated with genome reentry into cells once VSV-G expression has been initiated. Although the cell lines produced by Arai et al (1998) will be stable until infected with the Cre recombinase encoding adenoviruses, their results indicated that virus production dropped significantly 5 days after adenovirus infection allowing a limited number of harvests from each batch of producer cells.

Chen et al (1996 Proc. Natl. Acad. Sci. USA 93: 10057-10062) produced two tetracycline/β-estradiol - inducible cell lines which expressed VSV-G. The numbers of transducing particles produced every 48 hours increased over a sixteen days period after induction by either the removal of tetracycline from the medium or by the addition of β-estradiol in the absence of tetracycline in the cell line that produced the lowest level of VSV-G. However in the cell line that produced larger amounts of VSV-G, although an increase in titre was observed in the absence of tetracycline, a rapid fall in the number of transducing particles produced was found when high levels of VSV-G were produced upon β-estradiol induction in the absence of tetracycline. This fall was attributed to the toxic effects of high levels of VSV-G expression.

Another problem associated with some of the systems described above is that chemicals need to be added to the culture medium to induce transducing particle production. These chemicals could potentially have deleterious side effects on the patient and so would require removal prior to formulation. Moreover, in some methods the cultures will be contaminated with proteins from other viruses, such as adenoviruses, which could activate unwanted immune responses when used to treat any patients.

The present invention seeks to overcome some of the problems associated with the prior art by providing stable cell lines, capable of producing transducing viral particles, that are capable of expressing VSV-G and which are produced without the use of added chemicals.

Aspects of the present invention are presented in the accompanying claims and in the following description.

Thus, in one aspect, the present invention provides a packaging cell comprising: a first nucleotide sequence (NS) encoding a toxic viral envelope protein; and a second NS encoding a retrovirus nucleocapsid protein; wherein the expression of the first NS is regulatable in a temperature range from 25°C to 40°C.

Preferably the temperature range is from 28 °C to 40°C.

Preferably the temperature range is from 32°C to 40°C.

Preferably the temperature range is from 35°C to 40°C.

Preferably the temperature range is from 28 °C to 37°C.

Preferably the temperature range is from 32°C to 37°C.

Preferably the temperature range is from about 32°C to 37°C.

In a second aspect, the present invention provides a producer cell comprising: a first NS encoding a toxic viral envelope protein; a second NS encoding a retrovirus nucleocapsid protein; and a third NS comprising a retroviral sequence capable of being encapsidated in the nucleocapsid protein; wherein the retroviral vector particle titre obtained from same is dependent upon the temperature regulatable expression of the toxic viral envelope protein.

The present invention demonstrates that, despite the toxicity of VSV-G, it is possible to construct a stable retroviral producer line that is capable of producing transducing vector particles expressing VSV-G. Thus, the present invention demonstrates that the difficulty associated with producing stable cell lines capable of expressing VSV-G can be overcome.

In fact, the present invention is the first report of the propagation of stable cell lines that have the capacity to express high levels of the VSV-G protein. The present invention also demonstrates the surprising finding that a VSV-G cell line produced using a pHCMV-G plasmid has a temperature sensitive phenotype and a VSV-G cell line produced using a pRV67 plasmid has a temperature insensitive phenotype. This temperature sensitive VSV-G phenotype was characterised by the latent ability of a VSV-G cell line, transfected with the pHCMV-G plasmid, to produce significant quantities of a VSV-G protein at 32°C. Such a result has not been obtained before.

This finding is suprising as the published literature would suggest that if either of the two VSV-G proteins was to be temperature sensitive, it would be more likely to be that derived from pRV67 than the wild type (wt) protein derived from pHCMV-G. Moreover, the single amino acid coding sequence change found between the VSV-G protein encoded by pHCMV-G and the VSV-G protein encoded by pRV67, which results in the loss of a second glycosylation site, would not have been expected to account for the observed differences taking the current knowledge of VSV-G protein expression into account.

The surprising findings of the present invention are advantageous because they do not require the removal of chemicals or foreign viruses, such as adenovirus, from the viral particles produced.

The term "pseudotyping" refers to a a technique or strategy whereby an env gene is replaced with a heterologous env gene. Pseudotyping is not a new phenomenon and examples may be found in WO-A-98/05759, WO-A-98/05754, WO-A-97/17457, WO- A-96/09400, WO-A-91/00047 and Mebatsion et al 1997 Cell 90, 841-847.

The term "heterologous" refers to a nucleic acid sequence or protein sequence linked to a nucleic acid or protein sequence which it is not naturally linked.

The terms "variant" in relation to this aspect of the present invention include any substitution of, variation of, modification of, replacement of, deletion of or addition of one (or more) nucleic acid (s) from or to the nucleotide sequence providing the expression product of the resultant nucleotide sequence has a temperature sensitive phenotype, preferably having at least the same temperature sensitive phenotype as the expression product of the sequence shown as SEQ ID No. 1 or SEQ ID No. 2.

Sequence identity with respect to SEQ ID No. 1 or SEQ ID No. 2 can be determined by a simple "eyeball" comparison (i.e. a strict comparison) of any one or more of the sequences with another sequence to see if that other sequence has, for example, at least 75% sequence identity to the sequence(s).

Relative sequence identity can also be determined by commercially available computer programs that can calculate % identity between two or more sequences using any suitable algorithm for determining identity, using for example default parameters. A typical example of such a computer program is CLUSTAL. Advantageously, the BLAST algorithm is employed, with parameters set to default values. The BLAST algorithm is described in detail at http://www.ncbi.nih.gov/BLAST/blast_help.html, which is incorporated herein by reference. The search parameters are defined as follows, can be advantageously set to the defined default parameters.

Advantageously, "substantial identity" when assessed by BLAST equates to sequences which match with an EXPECT value of at least about 7, preferably at least about 9 and most preferably 10 or more. The default threshold for EXPECT in BLAST searching is usually 10.

BLAST (Basic Local Alignment Search Tool) is the heuristic search algorithm employed by the programs blastp, blastn, blastx, tblastn, and tblastx; these programs ascribe significance to their findings using the statistical methods of Karlin and Altschul (see http://www.ncbi.nih.gov/BLAST/blast_help.html) with a few enhancements. The BLAST programs were tailored for sequence similarity searching, for example to identify homologues to a query sequence. For a discussion of basic issues in similarity searching of sequence databases, see Altschul et al (1994) Nature Genetics 6:119-129. The five BLAST programs available at http://www.ncbi.nlm.nih.gov perform the following tasks:

blastp - compares an amino acid query sequence against a protein sequence database.

blastn - compares a nucleotide query sequence against a nucleotide sequence database.

blastx - compares the six-frame conceptual translation products of a nucleotide query sequence (both strands) against a protein sequence database.

tblastn - compares a protein query sequence against a nucleotide sequence database dynamically translated in all six reading frames (both strands).

tblastx - compares the six-frame translations of a nucleotide query sequence against the six-frame translations of a nucleotide sequence database.

BLAST uses the following search parameters:

HISTOGRAM - Display a histogram of scores for each search; default is yes. (See parameter H in the BLAST Manual).

DESCRIPTIONS - Restricts the number of short descriptions of matching sequences reported to the number specified; default limit is 100 descriptions. (See parameter V in the manual page).

EXPECT - The statistical significance threshold for reporting matches against database sequences; the default value is 10, such that 10 matches are expected to be found merely by chance, according to the stochastic model of Karlin and Altschul (1990). If the statistical significance ascribed to a match is greater than the EXPECT threshold, the match will not be reported. Lower EXPECT thresholds are more stringent, leading to fewer chance matches being reported. Fractional values are acceptable. (See parameter E in the BLAST Manual). CUTOFF - Cutoff score for reporting high-scoring segment pairs. The default value is calculated from the EXPECT value (see above). HSPs are reported for a database sequence only if the statistical significance ascribed to them is at least as high as would be ascribed to a lone HSP having a score equal to the CUTOFF value. Higher CUTOFF values are more stringent, leading to fewer chance matches being reported. (See parameter S in the BLAST Manual). Typically, significance thresholds can be more intuitively managed using EXPECT.

ALIGNMENTS - Restricts database sequences to the number specified for which high-scoring segment pairs (HSPs) are reported; the default limit is 50. If more database sequences than this happen to satisfy the statistical significance threshold for reporting (see EXPECT and CUTOFF below), only the matches ascribed the greatest statistical significance are reported. (See parameter B in the BLAST Manual).

MATRIX - Specify an alternate scoring matrix for BLASTP, BLASTX, TBLASTN and TBLASTX. The default matrix is BLOSUM62 (Henikoff & Henikoff, 1992). The valid alternative choices include: PAM40, PAM120, PAM250 and IDENTITY. No alternate scoring matrices are available for BLASTN; specifying the MATRIX directive in BLASTN requests returns an error response.

STRAND - Restrict a TBLASTN search to just the top or bottom strand of the database sequences; or restrict a BLASTN, BLASTX or TBLASTX search to just reading frames on the top or bottom strand of the query sequence.

FILTER - Mask off segments of the query sequence that have low compositional complexity, as determined by the SEG program of Wootton & Federhen (1993) Computers and Chemistry 17:149-163, or segments consisting of short-periodicity internal repeats, as determined by the XNU program of Claverie & States (1993) Computers and Chemistry 17:191-201, or, for BLASTN, by the DUST program of Tatusov and Lipman (see http://www.ncbi.nlm.nih.gov). Filtering can eliminate statistically significant but biologically uninteresting reports from the blast output (e.g., hits against common acidic-, basic- or proline-rich regions), leaving the more biologically interesting regions of the query sequence available for specific matching against database sequences.

Low complexity sequence found by a filter program is substituted using the letter "N" in nucleotide sequence (e.g.,

and the letter "X" in protein sequences (e.g., "XXXXXXXXX").

Filtering is only applied to the query sequence (or its translation products), not to database sequences. Default filtering is DUST for BLASTN, SEG for other programs.

It is not unusual for nothing at all to be masked by SEG, XNU, or both, when applied to sequences in SWISS-PROT, so filtering should not be expected to always yield an effect. Furthermore, in some cases, sequences are masked in their entirety, indicating that the statistical significance of any matches reported against the unfiltered query sequence should be suspect.

NCBI-gi - Causes NCBI gi identifiers to be shown in the output, in addition to the accession and/or locus name.

Most preferably, sequence comparisons are conducted using the simple BLAST search algorithm provided at http://www.ncbi.nlm.nih.gov/BLAST.

Other computer program methods to determine identify and similarity between the two sequences include but are not limited to the GCG program package (Devereux et al 1984 Nucleic Acids Research 12: 387) and FASTA (Atschul et al 1990 J Molec Biol 403-410).

In some aspects of the present invention, no gap penalties are used when determining sequence identity.

The present invention also encompasses nucleotide sequences that are complementary to the sequences presented herein, or any fragment or derivative thereof. If the sequence is complementary to a fragment thereof then that sequence can be used as a probe to identify similar promoter sequences in other organisms.

The present invention also encompasses nucleotide sequences that are capable of hybridising to the sequences presented herein, or any fragment or derivative thereof.

Hybridization means a "process by which a strand of nucleic acid joins with a complementary strand through base pairing" (Coombs J (1994) Dictionary of Biotechnology, Stockton Press, New York NY) as well as the process of amplification as carried out in polymerase chain reaction technologies as described in Dieffenbach CW and GS Dveksler (1995, PCR Primer, a Laboratory Manual, Cold Spring Harbor Press, Plainview NY).

Also included within the scope of the present invention are nucleotide sequences that are capable of hybridizing to the nucleotide sequences presented herein under conditions of intermediate to maximal stringency. Hybridization conditions are based on the melting temperature (Tm) of the nucleic acid binding complex, as taught in Berger and Kimmel (1987, Guide to Molecular Cloning Techniques, Methods in Enzymology, Vol 152, Academic Press, San Diego CA), and confer a defined "stringency" as explained below.

Maximum stringency typically occurs at about Tm-5°C (5°C below the Tm of the probe); high stringency at about 5°C to 10°C below Tm; intermediate stringency at about 10°C to 20°C below Tm; and low stringency at about 20°C to 25 °C below Tm. As will be understood by those of skill in the art, a maximum stringency hybridization can be used to identify or detect identical nucleotide sequences while an intermediate (or low) stringency hybridization can be used to identify or detect similar or related nucleotide sequences.

In a preferred aspect, the present invention covers nucleotide sequences that can hybridise to the nucleotide sequences of the present invention under stringent conditions (e.g. 65°C and O.lxSSC). The present invention also encompasses nucleotide sequences that are capable of hybridising to the sequences that are complementary to the sequences presented herein, or any fragment or derivative thereof. Likewise, the present invention encompasses nucleotide sequences that are complementary to sequences that are capable of hybridising to the sequence of the present invention. These types of nucleotide sequences are examples of variant nucleotide sequences.

In this respect, the term "variant" encompasses sequences that are complementary to sequences that are capable of hydridising to the nucleotide sequences presented herein. Preferably, however, the term "variant" encompasses sequences that are complementary to sequences that are capable of hydridising under stringent conditions (eg. 65°C and O.lxSSC {lxSSC = 0.15 M NaCl, 0.015 Na₃ citrate pH 7.0}) to the nucleotide sequences presented herein.

As used herein, the term "pseudotype" refers to progeny virions bearing the genome of one virus encapsidated by the envelope protein of another.

The term "toxic" refers to an expression product, such as a VSV-G protein whose expression detrimentally affects the metabolism of the target cells, thus limiting the quantity of this expression product produced in such cells.

"Target cell" simply refers to a cell which a vector, whether native or targeted, is capable of infecting or transducing.

A "toxic-protein" refers to a protein which exhibits cytotoxic or cytostatic behaviour in cells. Preferred assays for the measurement of cytotoxicity include the MTT assay which uses a non-radioactive format to quantitative viable cell number in proliferation assays (e.g. Promega Corp., Madison, WI) or ³H incorporation into nascent DNA. A number of other assay formats however, may likewise be utilised to evaluate various parameters associated with cell toxicity, including for example, cell viability (dye exclusion, cell counting) and cell metabolism (dye reduction) (see Cell Biology, A laboratory Handbook, ed. J.E. Celis, Academic Press) for further details and methods. Other representative examples of toxic proteins include but are not limited to γ interferon, interleukin-2, TNF α and TNF β.

The term "viral envelope protein" refers to the protein embedded in the membrane which encapsulates the nucleocapsid and which protein is responsible for binding to and entry of the infectious virus into the target cell. The viral envelope protein may also be a fusogenic protein. A "fusogenic protein" refers to glycoproteins which cause cells within a culture to fuse in a multinucleate syncytia. Representative examples of fusogenic proteins include VSV-G and Rabies G protein.

The term "nucleocapsid" refers to at least the group specific viral core proteins (gag) and the viral polymerase (pol) of a retrrovirus genome. These proteins encapsidate the retrovirus-packagable sequences and themselves are further surrounded by a membrane containing an envelope glycoprotein.

With the present invention, the term NOI (i.e. nucleotide sequence of interest) includes any suitable nucleotide sequence, which need not necessarily be a complete naturally occuring DNA sequence. Thus, the DNA sequence can be, for example, a synthetic DNA sequence, a recombinant DNA sequence (i.e. prepared by use of recombinant DNA techniques), a cDNA sequence or a partial genomic DNA sequence, including combinations thereof. The DNA sequence need not be a coding region. If it is a coding region, it need not be an entire coding region. In addition, the DNA sequence can be in a sense orientation or in an anti-sense orientation. Preferably, it is in a sense orientation. Preferably, the DNA is or comprises cDNA.

The NOI or NOIs may be under the expression control of an expression regulatory element, usually a promoter or a promoter and enhancer. The enhancer and/or promoter may be preferentially active in a hypoxic or ischaemic or low glucose environment, such that the NOI is preferentially expressed in the particular tissues of interest, such as in the environment of a tumour, arthritic joint or other sites of ischaemia. Thus any significant biological effect or deleterious effect of the NOI on the individual being treated may be reduced or eliminated. The enhancer element or other elements conferring regulated expression may be present in multiple copies. Likewise, or in addition, the enhancer and/or promoter may be preferentially active in one or more specific cell types - such as any one or more of macrophages, endothelial cells or combinations thereof. Further examples include include respiratory airway epithelial cells, hepatocytes, muscle cells, cardiac myocytes, synoviocytes, primary mammary epithelial cess and post-mitotically terminally differentiated non-replicating cells such as macrophages neurons.

The term "promoter" is used in the normal sense of the art, e.g. an RNA polymerase binding site.

The term "enhancer" includes a DNA sequence which binds to other protein components of the transcription initiation complex and thus facilitates the initiation of transcription directed by its associated promoter.

The promoter and/or enhancer may be constitutively efficient, or may be tissue or temporally restricted in their activity. Examples of suitable tissue restricted promoters/enhancers are those which are highly active in tumour cells such as a promoter/enhancer from a MUCl gene, a CEA gene or a 5T4 antigen gene. Examples of temporally restricted promoters/enhancers are those which are responsive to ischaemia and/or hypoxia, such as hypoxia response elements or the promoter/enhancer of a grp7S or a grp94 gene. The alpha fetoprotein (AFP) promoter is also a tumour-specific promoter. One preferred promoter-enhancer combination is a human cytomegalovirus (hCMV) major immediate early (MIE) promoter/enhancer combination.

Preferably the promoters of the present invention are tissue specific. That is, they are capable of driving transcription of a NOI or NOI(s) in one tissue while remaining largely "silent" in other tissue types.

The term "tissue specific" means a promoter which is not restricted in activity to a single tissue type but which nevertheless shows selectivity in that they may be active in one group of tissues and less active or silent in another group. A desirable characteristic of the promoters of the present invention is that they posess a relatively low activity in the absence of activated hypoxia-regulated enhancer elements, even in the target tissue. One means of achieving this is to use "silencer" elements which suppress the activity of a selected promoter in the absence of hypoxia.

The level of expression of an NOI or NOIs under the control of a particular promoter may be modulated by manipulating the promoter region. For example, different domains within a promoter region may possess different gene regulatory activities. The roles of these different regions are typically assessed using vector constructs having different variants of the promoter with specific regions deleted (that is, deletion analysis). This approach may be used to identify, for example, the smallest region capable of conferring tissue specificity or the smallest region conferring hypoxia sensitivity.

A number of tissue specific promoters, described above, may be particularly advantageous in practising the present invention. In most instances, these promoters may be isolated as convenient restriction digestion fragments suitable for cloning in a selected vector. Alternatively, promoter fragments may be isolated using the polymerase chain reaction. Cloning of the amplified fragments may be facilitated by incorporating restriction sites at the 5' end of the primers.

The term "hypoxia" means a condition under which a particular organ or tissue receives an inadequate supply of oxygen

In accordance with the present invention, suitable NOI sequences include those that are of therapeutic and/or diagnostic application such as, but are not limited to: sequences encoding cytokines, chemokines, hormones, antibodies, engineered immunoglobulin-like molecules, a single chain antibody, fusion proteins, enzymes, immune co-stimulatory molecules, immunomodulatory molecules, anti-sense RNA, a transdominant negative mutant of a target protein, a toxin, a conditional toxin, an antigen, a tumour suppressor protein and growth factors, membrane proteins, vasoactive proteins and peptides, anti-viral proteins and ribozymes, and derivatives therof (such as with an associated reporter group). When included, such coding sequences may be typically operatively linked to a suitable promoter, which may be a promoter driving expression of a ribozyme(s), or a different promoter or promoters, such as in one or more specific cell types.

Suitable NOIs for use in the invention in the treatment or prophylaxis of cancer include NOIs encoding proteins which: destroy the target cell (for example a ribosomal toxin), act as: tumour suppressors (such as wild-type p53); activators of anti-tumour immune mechanisms (such as cytokines, co-stimulatory molecules and immunoglobulins); inhibitors of angiogenesis; or which provide enhanced drug sensitivity (such as pro-drug activation enzymes); indirectly stimulate destruction of target cell by natural effector cells (for example, strong antigen to stimulate the immune system or convert a precursor substance to a toxic substance which destroys the target cell (for example a prodrug activating enzyme). Encoded proteins could also destroy bystander tumour cells (for example with secreted antitumour antibody- ribosomal toxin fusion protein), indirectly stimulated destruction of bystander tumour cells (for example cytokines to stimulate the immune system or procoagulant proteins causing local vascular occlusion) or convert a precursor substance to a toxic substance which destroys bystander tumour cells (eg an enzyme which activates a prodrug to a diffusible drug).

Also, the delivery of NOI(s) encoding antisense transcripts or ribozymes which interfere with expression of cellular genes for tumour persistence (for example against aberrant myc transcripts in Burkitts lymphoma or against bcr-abl transcripts in chronic myeloid leukemia. The use of combinations of such NOIs is also envisaged.

Suitable NOIs for use in the treatment or prevention of ischaemic heart disease include

NOIs encoding plasminogen activators. Suitable NOIs for the treatment or prevention of rheumatoid arthritis or cerebral malaria include genes encoding anti-inflammatory proteins, antibodies directed against tumour necrosis factor (TNF) alpha, and anti- adhesion molecules (such as antibody molecules or receptors specific for adhesion molecules).

Examples of hypoxia regulatable therapeutic NOIs can be found in PCT/GB95/00322 (WO-A-9521927).

The expression products encoded by the NOIs may be proteins which are secreted from the cell. Alternatively the NOI expression products are not secreted and are active within the cell. In either event, it is preferred for the NOI expression product to demonstrate a bystander effector or a distant bystander effect; that is the production of the expression product in one cell leading to the killing of additional, related cells, either neighbouring or distant (e.g. metastatic), which possess a common phenotype.

The NOI or NOIs of the present invention may also comprise one or more cytokine- encoding NOIs. Suitable cytokines and growth factors include but are not limited to: ApoE, Apo-SAA, BDNF, Cardiotrophin-1, EGF, ENA-78, Eotaxin, Eotaxin-2, Exodus-2, FGF-acidic, FGF-basic, fibroblast growth factor- 10 (Marshall 1998 Nature Biotechnology 16: 129).FLT3 ligand (Kimura et al (1997), Fractalkine (CX3C), GDNF, G-CSF, GM-CSF, GF-βl, insulin, IFN-γ, IGF-I, IGF-II, IL-lα, IL-l β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8 (72 a.a.), IL-8 (77 a.a.), IL-9, IL-10, IL-11, IL-12, IL-13, IL-15, IL-16, IL-17, IL-18 (IGIF), Inhibin α, Inhibin β, IP- 10, keratinocyte growth factor-2 (KGF-2), KGF, Leptin, LIF, Lymphotactin, Mullerian inhibitory substance, monocyte colony inhibitory factor, monocyte attractant protein (Marshall 1998 ibid), M-CSF, MDC (67 a.a.), MDC (69 a.a.), MCP-1 (MCAF), MCP-2, MCP-3, MCP-4, MDC (67 a.a.), MDC (69 a.a.), MIG, MlP-lα, MlP-l β, MIP-3α, MIP-3β, MIP-4, myeloid progenitor inhibitor factor- 1 (MPIF-1), NAP-2, Neurturin, Nerve growth factor, β-NGF, NT-3, NT-4, Oncostatin M, PDGF-AA, PDGF-AB, PDGF-BB, PF-4, RANTES, SDFlα, SDFl β, SCF, SCGF, stem cell factor (SCF), TARC, TGF-α, TGF-β, TGF-β2, TGF-β3, tumour necrosis factor (TNF), TNF-α, TNF-β, TNIL-1, TPO, VEGF, GCP-2, GRO/MGSA, GRO-β, GRO-γ, HCC 1 , 1 -309. As used herein, a "vector" denotes a tool that allows or faciliates the transfer of an entity from one environment to another. In accordance with the present invention, and by way of example, some vectors used in recombinant DNA techniques allow entities, such as a segment of DNA (such as a heterologous DNA segment, such as a heterologous cDNA segment), to be transferred into a target cell. Optionally, once within the target cell, the vector may then serve to maintain the heterologous DNA within the cell or may act as a unit of DNA replication. Examples of vectors used in recombinant DNA techniques include plasmids, chromosomes, artificial chromosomes or viruses.

The term "expression vector" as used in the present invention refers to an assembly which is capable of directing the expression of a nucleotide sequence of interest (NOI). The NOI expression vector must include a promoter which, when transcribed, is operably linked to the NOI, as well as a polyadenylation sequence. Within certain embodiments of the invention, both the promoter and the polyadenylation sequence are from a source which is heterologous to the helper elements, gag-pol and env. Within other embodiments of the invention, the expression vectors described herein may be contained within a plasmid construct.

The term "expression cassette" refers to a recombinant DNA molecule containing a desired coding sequence and appropriate nucleic acid sequences necessary for the expression of the operably linked coding sequence in a particular host organism. Nucleic acid sequences necessary for expression in eucaryotic cells usually include promoters, enhancers, and termination and polyadenylation signals. The cassette can be removed and inserted into a vector or plasmid as a single unit.

The term "operably linked" denotes a relationship between a regulatory region (typically a promoter element, but may include an enhancer element) and the coding region of a gene, whereby the transcription of the coding region is under the control of the regulatory region. In the present invention, several terms are used interchangeably. Thus, "virion", "virus", "viral particle", "'retroviral particle", "retrovirus", and "vector particle" mean virus and virus-like particles that are capable of introducing a nucleic acid into a cell through a viral-like entry mechanism. Such vector particles can, under certain circumstances, mediate the transfer of NOIs into the cells they infect. A retrovirus is capable of reverse transcribing its genetic material into DNA and incoφorating this genetic material into a target cell's DNA upon transduction. Such cells are designated herein as "target cells".

A vector particle includes the following components: a retrovirus nucleic acid, which may contain one or more NOIs, a nucleocapsid encapsidating the nucleic acid, the nucleocapsid comprising nucleocapsid protein of a retrovirus, and a membrane surrounding the nucleocapsid. The heterologous NOI may be operably linked to a promoter and encode a protein that is expressible in a target cell.

For the purposes of this application, such a heterologous NOI is capable of being expressed from the retrovirus genome either from endogenous retroviral promoters such as long terminal repeat (LTR) , or from a heterologous promoter to which the heterologous gene or sequence is operably linked.

When the vector particles are used to transfer NOIs into cells which they transduce, such vector particles also designated "viral delivery systems" or "retroviral delivery systems". Viral vectors, including retroviral vectors, have been used to transfer NOIs efficiently by exploiting the viral transduction process. NOIs cloned into the retroviral genome can be delivered efficiently to cells susceptible to transduction by a retrovirus. Through other genetic manipulations, the replicative capacity of the retroviral genome can be destroyed. The vectors introduce new genetic material into a cell but are unable to replicate.

The vector of the present invention can be delivered by viral or non-viral techniques. Non-viral delivery systems include but are not limted to DNA transfection methods. Here, transfection includes a process using a non-viral vector to deliver a gene to a target mammalian cell.

Typical transfection methods include electroporation, DNA biolistics, lipid-mediated transfection, compacted DNA-mediated transfection, liposomes, immunoliposomes, lipofectin, cationic agent-mediated, cationic facial amphiphiles (CFAs) (Nature Biotechnology 1996 14; 556), multivalent cations such as spermine, cationic lipids or polylysine, 1, 2,-bis (oleoyloxy)-3-(trimethylammonio) propane (DOTAP)-cholesterol complexes (Wolff and Trubetskoy 1998 Nature Biotechnology 16: 421) and combinations thereof.

Viral delivery systems include but are not limited to adenovirus vector, an adeno- associated viral (AAV) vector, a herpes viral vector, a retroviral vector, a lentiviral vector, or a baculoviral vector.

Examples of retroviruses include but are not limited to: murine leukemia virus (MLV), human immunodeficiency virus (HIV), equine infectious anaemia virus (EIAV), mouse mammary tumour virus (MMTV), Rous sarcoma virus (RSV), Fujinami sarcoma virus (FuSV), Moloney murine leukemia virus (Mo-MLV), FBR murine osteosarcoma virus (FBR MSV), Moloney murine sarcoma virus (Mo-MSV), Abelson murine leukemia virus (A-MLV), Avian myelocytomatosis virus-29 (MC29), and Avian erythroblastosis virus (AEV).

A detailed list of retroviruses may be found in Coffin et al ("Retroviruses" 1997 Cold Spring Harbour Laboratory Press Eds: JM Coffin, SM Hughes, HE Varmus pp 758- 763).

Preferred vectors for use in accordance with the present invention are recombinant viral vectors, in particular recombinant retroviral vectors (RRV). The lentiviruses can be divided into primate and non-primate groups. Examples of primate lentiviruses include but are not limited to: the human immunodeficiency virus (HIV), the causative agent of human auto-immunodeficiency syndrome (AIDS), and the simian immunodeficiency virus (SIV). The non-primate lentiviral group includes the prototype "slow virus" visna/maedi virus (VMN), as well as the related caprine arthritis-encephalitis virus (CAEV), equine infectious anaemia virus (EIAV) and the more recently described feline immunodeficiency virus (FIV) and bovine immunodeficiency virus (BIV).

A distinction between the lenti virus family and other types of retroviruses is that lentiviruses have the capability to infect both dividing and non-dividing cells (Lewis et al 1992 EMBO. J 11 : 3053-3058; Lewis and Emerman 1994 J. Virol. 68: 510-516). In contrast, other retroviruses - such as MLV - are unable to infect non-dividing cells such as those that make up, for example, muscle, brain, lung and liver tissue.

Preferred vectors for use in accordance with the present invention are recombinant viral vectors, in particular recombinant lentiviral vectors.

Other examples of vectors include ex vivo delivery systems, which include but are not limited to DΝA transfection methods such as electroporation, DΝA biolistics, lipid- mediated transfection, compacted DΝA-mediated transfection.

The vector may be a plasmid DΝA vector. Alternatively, the vector may be a recombinant viral vector. Suitable recombinant viral vectors include adenovirus vectors, adeno-associated viral (AAV) vectors, Herpes-virus vectors, or retroviral vectors, lentiviral vectors or a combination of adenoviral and lentiviral vectors. In the case of viral vectors, gene delivery is mediated by viral infection of a target cell.

The vector of the present invention may be configured as a split-intron vector. A split intron vector is described in PCT patent application GB98/02885 and GB/98/02867. If the features of adenoviruses are combined with the genetic stability of retro/lentiviruses then essentially the adenovirus can be used to transduce target cells to become transient retroviral producer cells that could stably infect neighbouring cells.

A "packaging cell" refers to a cell which contains those elements necessary for production of infectious recombinant virus which are lacking in a recombinant viral vector. Typically, such packaging cells contain one or more expression cassettes which are capable of expressing viral structural proteins (such as gag, pol and env) but they do not contain a packaging signal.

The term "packaging signal" which is refered to interchangeably as "packaging sequence" or "psi" is used in reference to the non-coding sequence required for encapsidation of retroviral RNA strands during viral particle formation.

Packaging cell lines suitable for use with the above-described vector constructs may be readily prepared (see also WO 92/05266), and utilised to create producer cell lines for the production of recombinant vector particles.

A "producer cell" or "vector producing cell" refers to a cell which contains all the elements necessary for production of recombinant viral vector particles and retroviral delivery systems.

Preferably, the producer cell is obtainable from a stable producer cell line.

Preferably, the producer cell is obtainable from a stable producer cell line constructed using pHCMV-G and a second plasmid with a selectable marker.

In preferred packaging and producer cells, the toxic envelope protein sequences, and nucleocapsid sequences are all stably integrated in the cell. However, one or more of these sequences could also exist in episomal form and gene expression could occur from the episome. In such cell lines, retroviral sequences are capable of being packaged with the nucleocapsid proteins. Further, the retroviral sequences that are capable of being packaged may also contain one or more heterologous NOIs that are capable of being expressed in a target cell that is infected by the virions produced in the producer cell.

The packaging cell lines are useful for providing the gene products necessary to encapsidate and provide a membrane protein for a retrovirus and retrovirus nucleic gene delivery vehicle. As described below, when retrovirus sequences are introduced into the packaging cell lines, such sequences are encapsidated with the nucleocapsid proteins and these units then bud through the cell membrane to become surrounded in cell membrane and to contain the envelope protein produced in the packaging cell line. These infectious retroviruses are useful as infectious units per se or as gene delivery vectors.

The cells are useful for producing infectious pseudotyped retrovirus, and especially high titer virions which may also contain one or more NOIs sequences capable of being expressed in a target cell or tissue. The cells are thus useful for packaging a retrovirus genome which may also contain a heterologous NOI capable of being expressed in a target cell or tissue

Our invention is based on an unexpected observation made when constructing producer cell lines by introducing the plasmid pHCMV-G. This is the same plasmid as was used by Chen et al (1996) and Yee et al (1994) to express the VSV-G protein. Clonal cultures were selected and about 70% failed produce a detectable virus titre, the remaining 30% produced a variable amount of virus. It was observed that lowering the temperature of the incubator from 37°C to 32°C for a four day period increased the titre per cell by 30 fold. An increase in the surface expression of VSV-G at 32°C in comparison to 37°C was detected using monoclonal antibodies (specific to the cytoplasmic tail of VSV-G) and polyclonal antibodies (raised against the whole protein). It proved difficult to detect surface or cytoplasmic expression of VSV-G at 37°C, but detection was possible at 32°C, suggesting that total VSV-G expression is higher at the lower temperature. Northern blot analysis suggested that mRNA levels are lower in cells at 37°C than at 32°C.

Further work in which pHCMV-G was introduced into either TE671 or HT1080 cells found that in the majority if not all of the cell lines found to express VSV-G, expression was low at 37°C and increased significantly at 32°C. At present no cell lines have been obtained with this plasmid, which show high levels of VSV-G expression at 37°C. In contrast, cell lines which produced high levels of VSV-G at 37°C could be obtained when another plasmid, pRV67, was used. So far, the increase in expression at 32°C in comparison to 37°C has not been observed in the cell lines constructed using pRV67 that expresses a fully functional variant of VSV-G.

The reason for the temperature sensitive phenotype of the VSV-G cell lines produced using pHCMV-G is unknown. A single amino acid coding sequence change has been found between the VSV-G proteins encoded by pHCMV-G and pRV67 which results in the loss of the second glycosylation site, and this could account for the differences. Although both genes are downstream of the HCMV promoter the two plasmids possess different introns upstream of the coding sequences and different downstream sequences. However the differences between the two plasmids would not have been expected to account for the observed differences taking the current knowledge on VSV-G protein expression.

Virions of rhabdoviruses are enveloped; protruding through the envelope is a homotrimeric membrane glycoprotein, G. This protein is responsible for the attachment of the virus to cellular receptors and is required for virus infectivity. The glycoprotein of VSV serotype Indiana has 51 1 amino acids including a signal peptide, a hydrophobic membrane spanning domain of 20 amino acids, and a 29 amino acid hydrophilic carboxy-terminal cytoplasmic domain and usually has two asparagine- linked complex oligosaccharides attached (Rose and Gallione 1981 Journal of Virology 39: 519-528). A fatty acid chain (palmitate) is linked to the single cysteine residue in the cytoplasmic domain (Rose et al, 1984 Proceedings of the National Academy of Science USA 81 : 2050-2054).

The properties of the glycoprotein of VSV serotype Indiana strain San Juan and selected mutants of this protein have been extensively studied. The reason for this scientific interest is that this protein has been used as a model system to study the effects of protein modifications and sequences on processing and transport from the Endoplasmic Reticulum (ER) to the Golgi and on to the plasma membrane. In infected cells the protein is translated on membrane bound ribosomes and inserted into the rough endoplasmic reticulum (rER), where the correct folding of the protein is assisted by BiP (de Silva 1993 Journal Cell Biology 120: 647-655) and the formation of trimers occurs. During these processes the signal peptide is removed and the two asparagine- linked complex oligosaccharides are attached to the protein. The protein is then transported to the Golgi apparatus and the oligosaccharides are further processed. This involves the removal of some glucose and mannose residues and the addition of the peripheral sugars, N-acetlyglucosamine, galactose, sialic acid and fructose. The protein is then transported to the plasma membrane and is incorporated into virus particles.

The role of glycosylation in VSV-G is complex. Mutants in which one or both of the glycosylation sites have been mutated have been constructed for two closely related strains of VSV Indiana. In San Juan, the strain from which the plasmids pRV67 and pHCMV-G are based, at least one of the two glycosylation sites is required for efficient transport of protein to the cell surface. When both of the glycosylation sites were mutated very little of the G protein reached the cell surface at 37°C. However if only a single site was mutated surface expression levels were found to be similar to those found in the wt (Machamer et al, 1985, Molecular and Cellular Biology 5: 3074- 3083). The rate at which the proteins possessing a single mutation were processed was similar to the wt, when both glycosylation sites were mutated the protein tended to be retained in the ER in an aggregated form and trimers were not formed. Lowering the temperature to 30°C resulted in a low level of surface expression only in the double mutant (Machamer and Rose 1988, Journal of Biological Chemistry 263: 5955-5960). In a closely related strain, Orsay, significant levels of protein could be detected at the cell surface when protein containing the double mutation was expressed (Pitta et al, 1989, Journal of Virology 63: 3801-3809). Transport of the unglycosylated San Juan protein could be increased to similar levels to those observed in Orsay by mutating residue 172 to an aspartic acid (tyrosine in San Juan and aspartic acid in Orsay). A measurable increase in the amount of glycoprotein transport in San Juan was also observed if residue 231 was mutated from a glycine to an aspartic acid.

Of interest to the present invention, expression levels of G proteins, in which one or both of the two glycosylation site have been mutated, are lower than in the wt constructs for both the San Juan and Orsay strains (Pitta et al, 1989, Journal of Virology 63: 3801-3809; Machamer et al, 1985, Molecular and Cellular Biology 5: 3074-3083). This did not appear to be due to increased degradation. It is therefore possible that the expression level of VSV-G might be lower in pRV67 than in pHCMV-G.

In the San Juan strain 1 in 6 of the G molecules made are cleaved so as to release a fully glycosylated, soluble form, which is found in the cell medium. The remaining product, which contains the transmembrane and cytoplasmic domains, may be found in virions (Chen and Huang 1986 Journal of Virology 59: 210-215). As well as the observed differences in the amounts of protein transported to the cell surface, the percentage of the G protein which is truncated is higher in some of those mutants which are not glycosylated correctly. This truncation was affected by temperature, since more of the soluble form was observed at 39°C than at 31°C.

The published literature would suggest that if either of the two VSV-G proteins was to be temperature sensitive, it would be more likely to be that derived from pRV67 than the wt protein derived from pHCMV-G. Although virtually all of the studies quoted are not strictly comparable to the situation found in the production of a stable cell line, they have not reported that RNA levels decrease with increasing temperature in glycosylation mutants. The invention will now be further described only by way of examples in which reference is made to the following Figures:

Figure 1 which shows a diagrammatic representation of VSV-G expression cassettes - pRV67 and pHCMV-G;

Figure 2 which shows a photographic representation of western blots showing the expression of the VSV glycoprotein in cytoplasmic extracts of various cell lines;

Figure 3 which shows the sequence difference between VSV-G plasmids;

Figure 4 which shows a photographic representation of VSV-G expression in stable cloned cell lines derived from cell transfected with pHCMV-G;

Figure 5 which shows a photographic representation of the increase in VSV-G expression with the time of culture at 32°C;

Figure 6 which shows a diagrammatic representation of pONY2.1nlsZαcZ and pONY3;

Figure 7 which shows a graphical representation that pRV67 derived cell lines do not show temperature regulated VSV-G expression.

Figure 8 which shows a graphical representationn of transient vector production in a PI cell line after different times at 32°C;

Figure 9 which shows a graphical representation of transient vector production in PI cells transfected 12 hours prior to shift down to 32°C;

Figure 10 shows a graphical representation of the effect of temperature on vector production from stable cell lines; Figure 1 1 which shows a photographic representation of the effect of temperature on VSV-G expression in a VSV671 cell line;

Figure 12 which shows a pictorial representation of MLV plasmids;

Figure 13 which shows the production of MLV vector particles by the PI cell line;

Figure 14 which presents a Northern blot analysis showing the increase in VSV-G mRNA levels at 32°C;

Figure 15 which shows a photographic representation of a western blot performed to screen clones obtained from transfecting VSV-G expressing cells with CeB for MLV p30 expression.

Figure 16a which shows the effects of temperature on particle production in a packaging cell line, PI -CeB clone 12 transduced with MLV vector particles.

Figure 16b which shows the effects of temperature on particle production in a packaging cell line PI -CeB clone 7 transduced with MLV vector particles.

Figure 17 which shows the effect of temperature on particle production in an MLV producer cell line, derived from P 1.

Figure 18 which shows details of the EIAV gag/pol expression plasmid, pONYHYG, used to produce packaging cell lines.

Figure 19 which shows a photographic representation of a western blot performed to screen clones for EIAV gag/pol expression obtained from transfecting VSV-G expressing cells with pONYHYG. Figure 20 which shows a photographic representation of a western blot performed to screen clones for EIAV gag/pol expression obtained from limit diluting the ring clones AC4 and BD7.

Figure 21 which shows the effect of sodium butyrate on transducing particle production in an EIAV producer cell line.

Figure 22 which shows a photographic representation of a southern blot of chromosomal DNA samples isolated from a range of cell lines restricted with three different enzymes, and probed with a VSV-G probe.

Figure 23 which shows a photographic representation of a southern blot of chromosomal DNA samples isolated from a range of cell lines restricted with three different enzymes, and probed with an EIAV gag/pol probe.

Figure 24 which shows a photographic representation of a northern blot of total RNA samples isolated from a range of cell lines and probed with a VSV-G probe.

Figure 25 which shows the results on an analysis of the increase in VSV-G mRNA levels in cells grown at 32 ° C compared to the levels observed at 37° C.

Figure 26 which shows a photographic representation of a western blot analysis of the increase in the level of VSV-G protein in cells grown at 32° C compared to cells grown at 37° C.

Figure 27 which shows a photographic representation of a northern blot of total RNA samples isolated from a range of cell lines and probed with a probe to EIAV gag/pol.

Figure 28 which shows the results on an analysis of the increase in EIAV gag/pol mRNA levels in cells grown at 32°C compared to the levels observed at 37 °C. Figure 29 which presents SEQ ID No. 1; and

Figure 30 which presents SEQ ID No. 2.

In slightly more detail:

Figure 1 shows details of the two plasmids used to express the VSV glycoprotein.

Figure 2 shows western blots illustrating the expression of the VSV glycoprotein in cytoplasmic extracts of various cell lines. In all cases the protein was detected using the monoclonal antibody, P5D4, which binds to the cytoplasmic tail of the glycoprotein. The cells were transiently transfected with the plasmids and analysed for expression after 48hrs at 37°C.

Lanes numbered 1-6 contain the following samples: 1 = TE671 cells treated with fuGENE 6 with no DNA grown at 37°C

2 = TE671 cells transfected with pHCMV-G grown at 37°C

3 = HT1080 cells transfected with pRV67 grown at 37°C

4 - TE671 cells transfected with pRV67 grown at 37°C

5 = HT1080 cells grown at 37°C 6 = TE671 cells grown at 37°C

Figure 3 shows the one coding difference detected between the VSV-G genes in pHCMV-G and pRV67. The sequence shown is from amino acid 333 to 340 (the numbering assumes that amino acid 1 is the first residue in the signal peptide sequence).

Figure 4 shows an example of the results obtained when cytoplasmic extracts of cloned cell lines derived from cell transfected with pHCMV-G were analysed for VSV-G expression. Samples of the clones were grown at 37°C and 32°C for 3 days prior to harvesting and samples for both temperatures are shown. VSV-G expression was detected using Western blotting and the monoclonal antibody (P5D4) as described in Example 1. O samples are clones derived from HT1080 cells transfected with pHCMV-G and pCI-NEO and selected on lug/ml G418. P samples are clones derived from TE671 cells transfected with pHCMV-G and pCI-NEO and selected on lug/ml G418.

Figure 5 shows the time required for VSV-G expression to increase when a cloned cell line, PI, derived from TE671 cells transfected with pHCMV-G was shifted to 32°C. VSV-G expression was detected using Western blotting and the monoclonal antibody (P5D4) as described in Example 1. Panel A shows an SDS-PAGE gel with β- mercapto-ethanol; and Panel B shows a SDS-PAGE page without β-mercapto-ethanol. Lanes numbered 1-7 and M contain the following samples:

1 = PI cells grown at 37° C

2 = PI cells grown at 32°C for 32 hours.

3 = PI cells grown at 32°C for 56 hours.

4 = PI cells grown at 32°C for 80 hours. 5 = PI cells grown at 32°C for 104 hours.

6 = TE671 cells transfected with pHCMV-G, 48 hours post transfection, cells grown at 37°C.

7 = TE671 cells grown at 37°C. M = protein size markers.

Figure 6 shows details of the EIAV plasmids used to produce transducing particles from cell lines expressing VSV-G.

Figure 7 shows the results from experiments in which a cell line that expressed VSV- G, (A7) derived from pRV67, was transfected with pONY2.1nlsZαcZ and pONY3. The medium from such transfected cells was removed every day and the number of transducing particles present per ml was determined using D17 cells.

Figure 8 shows a study to determine the optimum conditions for analysing vector production after transient transfection of gag-pol and genome into VSV-G pseudotyping cell lines. A cell line (PI) derived from pHCMV-G, that had been shown to express VSV-G at 32°C, was transfected with pONY2.1nlsZ,αcZ and pONY3. The medium from such transfected cells was removed every day and the number of transducing particles present per ml was determined using D17 cells.

Figure 9 shows results from a similar study to that in Figure 9 using PI cells transfected with pONY2.1nlsZαcZ and pONY3. In this case the cells were shifted from 37°C to 32°C twelve hours after transfection. The medium from such transfected cells was removed every day and the number of transducing particles present per ml was determined on D17 cells.

Figure 10 shows the time course of production of MFG retroviral vector from TelCEBό lines containing HCMV-G. Three clones were analysed. They were grown at 37°C for 2 days and then shifted to 32°C. Culture supernatants were analysed by titration on D17 cell. Cell numbers were counted at each harvest point and data for VSV-7 cell line are shown.

Figure 11 shows HCMV-G expression from a typical TelCEBό cell line containing pHCMV-G (VSV671). The cell line VSV671 was grown at 32°C (lane3) and 37°C. Lanes numbered 1-2 show the following samples: 1 = VSV671 cells grown at 32°C, 2 = VSV671 cells grown at 37°C

Figure 12 shows plasmids used to produce MLV vectors.

Figure 13 shows the effect of reducing the temperature on the production of transducing particles from producer cell lines derived from TelCeBό cells transfected with pHCMV-G.

Figure 14 represents a northern blot of total RNA isolated from PI cells grown for different amounts of time at 32°C. The blot was probed with the complete coding region of the VSV-G gene. Lanes numbered 1-7 contain the following samples: 1 TelCeB cells grown at 37°C. 2 PI cells grown at 37° C.

3 PI cells grown at 32°C for 1 day.

4 PI cells grown at 32°C for 2 days.

5 P 1 cells grown at 32°C for 3 days.

6 P 1 cells grown at 32°C for 4 days.

7 VSV-G gene from pHCMV-G present on a 1692bp BamHl fragment.

Figure 29 presents SEQ ID No 1 which is the BamHl fragment containing the VSV-G gene present in pHCMV-G.

Figure 30 presents SEQ ID No 2 which is the BamHl fragment cloned into pSA91 to construct pRV67 which contains the VSV-G gene.

EXAMPLES

Example 1

COMPARISON OF TWO VSV-G PLASMIDS

Plasmids that express the VSV-G protein and that are widely used for pseudotyping both retroviral and lentiviral vectors are shown in Figure 1. They were constructed as follows: pHCMV-G was constructed by Yee et al (1994 Proc. Natl. Acad. Sci. USA 91 : 9564-9568). The BamHl fragment containing the VSV-G gene was isolated from pLGRNL (Emi et al, 1991 Journal of Virology 65:1202-1207) and was inserted into the unique BamHl site in pHCMV-Bam. pHCMV-Bam contains the human cytomegalovirus immediate early promoter followed by the splicing and polyadenylation signals derived from the rabbit β-globin gene.

pRV67 was constructed by cloning a BamHl fragment containing the VSV-G gene into pSA91. pSA91 is a derivative of pGWIHG (Soneoka et al 1995 Nucl. Acids Res.

23: 628-633) from which the gpt gene has been removed by digestion with BamHl. It contains the human cytomegalovirus immediate early promoter/intron A sequences followed by a multiple cloning site and a polyadenylation signals.

The expression of VSV-G from these plasmids is confirmed by western blot analysis of transiently transfected cells as shown in Figure 2. The two plasmids were transfected into HT1080 (ATCC CCL 121) and TE671 (ATCC 8805-CRL) using FuGENE 6™ (Boehringer Mannheim). 48 hours after transfection the cells were rinsed in phosphate buffered saline (PBS) and were lysed in with 1% v/v Nonidet P40 in PBS. The protein concentrations of the samples were measured using the BioRad DC protein assay as per the manufacturer's instructions and 5μg samples were analysed by SDS/PAGE followed by western blotting. The VSV-G protein was detected using a mouse monoclonal antibody raised against a peptide (YTDIEMNRLGK) in the cytoplasmic tail of VSV-G (P5D4), the antibody is conjugated to horse radish peroxidase (Boehringer Mannheim) and can be used directly which ECL detection reagents. Cells were cultured at the usual temperature of 37°C.

Unexpectedly, the protein products detected on the western blot from the two different transfections did not appear to be the same size. The VSV-G product derived from pHCMV-G ran slightly slower than the protein product from pRV67 (Figure 2, lane 2 compared with lane 4). However it was clear that both plasmids were capable of directing the expression of significant amounts of VSV-G. The difference in the absolute expression levels between the samples from HT1080 and TE671 cells (compare lanes 5 and 6) is almost certainly due to the significantly higher percentages of cells transfected when using TE671 cells.

To determine the basis of the difference in molecular weight the two plasmids were sequenced. When the coding sequences of the two proteins were compared, only two base changes were identified (see Figure 3). These resulted in a threonine (residue 338 in the protein if the first amino acid is the signal peptide is numbered as residue 1) in the protein derived from pHCMV-G being converted to an alanine in the protein derived from pRV67. This mutation results in the loss of the second glycosylation site in the pRV67 protein (the motif for N-linked glycosylation being Asn-X-Thr/Ser), and could explain the size difference in the two proteins observed in Figure 2.

Example 2

CONSTRUCTION OF CELL LINES EXPRESSING VSV-G

Two different approaches were used to produce cell lines expressing VSV-G. pHCMV-G and pCl-neo (Promega) were transfected into HT1080 and TE671 using FuGENE 6™ (Boehringer Mannheim) at a ratio of 10 to 1. pCI-oeo encodes for the aminoglycoside phosphotransferase gene down stream of an SV40 promoter, cells in which this gene is expressed should be resistant to the antibiotic Geneticin (Schering Corporation). The cells were then selected using lmg ml^"1 Geneticin, the assumption being that the majority of the cells that contained pCl-neo would also contain pHCMV-G.

An antibiotic resistance marker, zeocin resistance, was cloned downstream of the VSV-G gene in pRV67 to allow the selection of cells containing this plasmid using 50 μg ml^"1 Zeocin (Invitrogen). The gene was cloned downstream of a minimal SV40 promoter.

Cytoplasmic extracts from the clones were analysed for the presence of VSV-G protein using western blotting and the monoclonal antibody (P5D4) as described in Example 1. Samples of the clones were grown at 37°C and also at 32°C. An example of the results obtained with cells engineered to contain pHCMV-G is shown in Figure 4. They reveal the surprising result that the expression levels of VSV-G are dramatically reduced at 37°C as compared to 32°C. There was variability between the clones in terms of the maximum level of expression but even in this small sample clone 022 and clone PI were obtained that gave a strong temperature dependent control of VSV-G expression. It is assumed that the failure of all of these clones to make significant levels of VSV-G at 37°C contributed to their survival during propagation. The surprising result is that shifting the temperature down to 32°C revealed the latent ability of these cells to produce significant quantities of the VSV-G protein. Such a result has not been obtained previously and this is the first report of the propagation of stable cell lines that have the capacity to express high levels of the VSV-G protein. A few clones were obtained that expressed VSV-G using the pRV67 plasmid however in this case (see example 4; Figure 7). There was no increased expression after shift down to 32°C.

Example 3

CHARACTERISATION OF THE EFFECT OF TEMPERATURE ON VSV-G EXPRESSION IN CELL LINES DERIVED FROM PHCMV-G

The induction of VSV-G in cell lines derived from pHCMV-G when the culture temperature is shifted from 37°C to 32°C was characterised further. Monolayers of cells were grown in six well tissue culture dishes at 37°C until 50% confluent at which time the plates were shifted to 32°C. Cells were harvested every day as in Example 1 and the amount of VSV-G present was analysed using western blotting. Samples were detected using the monoclonal antibody P5D4 to VSV-G. The results are shown in Figure 5.

When the cell lysates were denatured with SDS and β-mercaptoethanol prior to analysis with SDS PAGE and Western blotting, the VSV-G protein can be seen migrating just in front of a cross reacting protein. This upper protein is present even in lane 7 that contains the parental TE671 cell line. We conclude that it is a cross reactive cellular protein. The lower protein band is clearly absent from this lane confirming that this is indeed the VSV-G protein. This is further supported by the data shown in panel B. Here the samples are run in their non-reduced form. That is, they are not treated with a reducing agent that breaks cysteine bonds in the protein. In this case there are no proteins detected in the TE671 negative control sample (see panel B) as presumably the cross reactive epitope on the cellular protein is not revealed without reduction.

A clearer picture is then obtained of how the amount of VSV-G present in the cells increases with time after the cells are shifted from 37°C to 32°C. Little VSV-G protein is detected up to 56 hours after the temperature shift. However significant amounts of the protein were present after 3 days at 32°C and appeared to reach a plateau. The doublet is due to the slightly faster mobility of the non-reduced VSV-G protein - this has been previously reported by Machamer and Rose (1988, Journal of Biological Chemistry 263; 5955-5960) - and the presence of denatured material in the preparation.

These data indicate that the induction of expression of VSV-G at reduced temperature is slow and takes at least three days to reach significant levels. This implies but does not prove that the mechanism is at the level of synthesis of VSV-G or some cellular factor rather than at the level of VSV-G protein conformational changes.

Example 4

PRODUCTION OF PSEUDOTYPED EIAV PARTICLES

We wished to determine whether the VSV-G expressing cell lines could be used to produce lentiviral vectors. In all of the following examples the assay system is based on the three-plasmid transfection method described previously (Soneoka et al, 1995 ibid). The plasmids used in these experiments were as follows; a genome plasmid, pONY2.1nlsZαcZ or a derivative of it, and a plasmid that expresses the EIAV gag-pol genes pONY3 or a derivative of it (GB patent application 9727135.7). The important features of these vectors are shown in Figure 6.

Construction of the EIAV vectors was as follows. An infectious proviral clone, pSPEIAV19, as described by Payne et al (1994, J.Gen. Virol. 75:425-9) was used as a starting point. A plasmid, pSPEIAVΔH, was constructed by the deletion of a Hindlll fragment, 5835-6571, from the region of the plasmid encoding for the envelope protein. A vector genome plasmid was constructed by inserting the EIAV LTR, amplified by PCR from pSPEIAV19, into pBluescript II KS+ (Stratagene). The MluVMlul (216/814) fragment of pSPEIAVΔH was then inserted into the LTR/Bluescript plasmid to generate pONY2. In addition, a BglU/Ncol fragment within pol (1901/4949) was deleted and a nuclear localising β-galactosidase gene driven by the HCMV IE enhancer/promoter was inserted in its place. This was designated pONY2.1nlsZαcZ.

An EIAV plasmid (pONY3) encoding the gag-pol genes was then made by inserting the MluVMlul fragment from pONY2 into the mammalian expression plasmid pCl-neo (Promega) such that the gag-pol protein is expressed from the HCMV IE enhancer/promoter.

Experiments were performed in which cells lines that had been shown by western blotting to produce a product recognised by the monoclonal antibody, P5D4, were transfected with pONY2.1nlsZαcZ and pONY3. The medium from such transfected cells was removed every day and the number of transducing particles present per ml was detected on D17 cells by using X-gal stain to detect the presence of β- galactosidase activity.

The results from A7, a cell line cloned from HT1080 cells transfected with pRV67, are shown in Figure 7. This cell line does not transfect efficiently (less than 5% of the cells being transfected) and no attempts have been made to optimise transfection levels. Also relatively few clones (less than 50) were screened to obtain a VSV-G expressing line from cells transfected with pRV67 and so the level of VSV-G had not been optimised. Despite this lack of optimisation VSV-G pseudotyped vector particles were released from the cultures. However in this case there was no increase in expression at 32°C, in fact levels were lower. This was also observed with the MLV vector and led us to conclude that this variant envelope is not subject to temperature regulation. It is possible that this protein is less toxic and that there is not such a strong selective pressure for a temperature sensitive phenotype. The situation was more complex in the cell lines derived from pHCMV-G where the best time to change the culture temperature after transfection had to be determined. The results of experiments using a cell line derived from TE671 are shown in Figures 8 and 9. Transfecting cells after they had stabilised at 32°C for 24hrs resulted in very low titres. Transfecting immediately and 18hrs before the shift down gave an improvement (Figure 8). The optimum period was about 12 hrs prior to shift down (Figure 9).

Although these data are encouraging in that pseudotyped vector particles are produced there are significant problems associated with such transient experiments. For example, the levels of VSV-G in the cells only become maximal three to four days after the temperature is shifted (see example 3). However as the transfection efficiencies at 32°C were found to be poor, the cells have to be shifted to 32°C after transfection. Three to four days after transfection the levels of expression from the transfected plasmids (encoding the gag/pol and genome components of the system) would be expected to be decreasing from maximal levels. These data therefore show that the VSV-G expressing stable cell lines have the potential to produce vector particles after transient transfection with the genome and gag-pol components. If pRV67 is used than some vector production is possible at 37°C. If pHCMV-G is used then a temperature shift to 32°C is critical for generating vector. However the timing of the temperature shift relative to the transfection was critical and titres were dependent upon transfection efficiency. To overcome these limitations, stable producer lines were developed in which all of the components were stably maintained in the cell.

Example 5

PRODUCTION OF STABLE CELL LINES THAT PRODUCE VSV-G PSEUDOTYPED MLV VECTOR PARTICLES

A cell line that produced transducing particles based on MLV was constructed. In the first instance we used a cell line, TELCeBό, that had already been produced to express the MLV gag-pol gene and a vector genome. This cell line is derived from TE671 cells and harbours the MFGnls/ cZ retroviral vector and a MoMLV gag-pol expression plasmid (CeB), (Cosset et al 1995 Journal of Virology 69:7430-7436).

Co-transfection of the TELCeBό cells with pHCMV-G and pcDNA3 (Invitrogen) was carried out using calcium phosphate precipitation (Sambrook et al, 1989 Molecular Cloning: a laboratory manual, Cold Spring Harbor Laboratory Press). The plasmid pcDNA3 encodes the aminoglycoside phosphotransferase gene down stream of an SV40 promoter. Clones were selected using lmg ml^"1 Geneticin, 10 independent clones were isolated and the production of transducing particles from these cells was assayed. While seven out of ten clonal cultures produced no detectable titre, three clones produced detectable titres. These clones were called VSV-5, VSV-6.and VSV- 7.

Cultures were seeded with the same number of cells and were then incubated at either 37°C or 32°C for four days at which point the tissue culture medium was harvested. The number of transducing particles present per ml of medium was 5-fold higher at 32°C despite the lower number of producer cells at the lower temperature.

Flasks of cells were grown at 37°C until confluent and then were shifted to 32°C. A typical time course of virus production is shown in Figure 10. All three clones showed an increase in titre for seven days after the temperature shift. This increase was not simply due to an increase in the number of cells in culture which is shown for the VSV-7 cell line. After seven days at 32°C, the cells began to round up and detach and the yield of transducing particles decreased. An increase in the surface expression of VSV-G was detected using a monoclonal antibody (P5D4 specific to the cytoplasmic tail of VSV-G) and polyclonal antibodies (raised against the whole protein) at 32°C in comparison to 37°C. It proved difficult to detect surface or cytoplasmic expression of VSV-G at 37°C, but detection was possible at 32°C, suggesting that total VSV-G expression is higher at the lower temperature. Figure 11 shows the results of a Western blot performed on cell lysates from one of the stable producer cell lines developed. VSV-G expression was detected using the monoclonal antibody, P5D4. The presence of VSV-G can clearly be seen at 32°C, but not at 37°C.

In a second example the P 1 cell line described in example 2 which had been grown at 37°C for different periods of time (70 or 106 days after transfection with pHCMV-G) was transfected with the plasmids pMLV-GP (genome) and pCIE-GPSD (gag-pol). The plasmid pMLV-GP (see Figure 12) was constructed by M. Yap (University of Oxford). pHITl l l (Soneoka et al, 1995 ibid) was restricted with Kpnl to remove the region between the LTRs, this was replaced by a similar Kpnl fragment from pLNCX (Miller and Rosman 1989, Biotechniques 7: 980-990). The resultant plasmid was cut with Hpal and the enhanced green fluorescent protein gene (egfp) from pEGFP-Nl (Clontech) was inserted as a Smal-Hpal fragment.

pCIE-GPSD was constructed by inserting the gag/pol genes from MLV (derived from pHIT60; Soneoka et al, 1995 ibid) into a derivative of pCI-«eø (Promega) with an extended enhancer region. The dihydrofolate reductase gene (dhfr) from pSV2-DHFR (Subramani et al, 1981 Mol. Cell Biol.,1 : 854-64) was inserted after the gag-pol genes with a short spacer of 74 residues separating the genes, this should allow some of the ribosomes which translate the gag-pol genes to reinitiated and translate the dhfr gene.

8 hours after transfection a half of the cells were transferred to 32°C; supernatant samples were then taken every day for five days. The parental TE671 cells were used as a control against the possible ability of particles to transduce cells in the absence of VSV-G envelope protein; the results are shown in Figure 13. They demonstrate that a pool of PI derivatives that contain gag-pol and vector genome can produce VSV-G pseudotyped MLV vectors at least 10³/ml. This is a typical pool titre for a retroviral producer line and the selection of pure clones from this population will yield higher titre producer clones. Once again titres were significantly higher at 32°C than at 37°C. Transducing particles were not produced under the same conditions in control TE671 cells. The numbers of transducing particles produced was similar in PI cells that had been cultured for 106 or 70 days indicating that the expression levels of VSV-G in PI cells is stable. Example 6

Analysis of the molecular basis of the temperature regulation of VSV-G expression

Total RNA was prepared from the PI cell line and analysed by northern blotting using a probe homologous to VSV-G mRNA (Figure 14).

Figure 14 presents a Northern blot analysis showing an increase in VSV-G mRNA levels at 32°C. In example 3 it was shown that the amount of the VSV-G protein present in a stable cell line, PI, derived from TE671 cells transfected with pHCMV-G increased with time after the cells were shifted from 37 °C to 32°C. Only small amounts of the protein were detected at 37°C, whilst significant amounts of the protein were found to be present after 3 days at 32°C. Since this increase could be due an effect of temperature on transcription or translation an experiment was performed to investigate whether VSV-G RNA levels were affected by changes in culture temperature. PI cells were transferred from 37 °C to 32°C and cells were harvested every day for four days. The total RNA content of the cells was isolated using TRIZOL reagent (GIBCOBRL) and a northern blot was performed using standard methods. This blot was probed using a 1.6kb DNA fragment containing the whole of the VSV-G. The results shown in Figure 14 clearly demonstrate that messenger RNA levels for VSV-G increase in a similar manner to protein levels after the P 1 cells are shifted to 32°C.

Although the experimental work disclosed herein is directed to cell lines for producing infectious, recombinant retroviral vectors, the concepts and design body are broadly applicable to cell lines for the production of any viral vector where harmful or otherwise undesirable viral proteins must be produced by the cell in order for the viral vector to be produced. The constructs and methods of invention are used to prevent or minimise the production of these proteins until they are needed. At that time, expression is induced for a period of time necessary for the production of the proteins and the assembly of the viral vectors. Examples of such viral vectors include but are not limited to other RNA viral vectors besides retroviral vectors, and DNA viral vectors, such as adenoviral vectors, adeno-associated viral vectors, Herpesvirus vectors (preferably Herpes simplex I virus vectors), and vaccinia virus vectors. Examples of harmful or undesirable proteins include, for adenoviral vectors, products of the El, E2, E4, and major late genes; for adeno-associated viruses, the rep protein; and for Herpesvirus, the capsid protein. Methods for the construction of such cell lines will be readily apparent to those skilled in the art, given the teachings contained herein.

We have therefore shown that despite the toxicity of VSV-G it is possible to construct a retroviral producer line that releases vector particles carrying VSV-G. The method that we have developed involves passage of VSV-G expressing cell lines at 37°C and then screening cells for enhanced production at 32°C. We have also found that this effect may be dependent upon the integrity of the second glycosylation site in VSV-G as found in plasmid pHCMV-G. Mutation of this site in pRV67 although not influencing the ability of the protein to pseudotype vectors does not allow the selection of a temperature regulated producer clone although surprisingly there was sufficient expression at 37°C to allow some vector production.

This method is reproducible being obtained with two different parental cells, TE671 and HT1080 and three different retroviral MLV genomes, MFGnlsZαcZ and pMLV- GP and the EIAV vector genome pONY2.1nsZαcZ and two different gag-pol genes derived from MLV and EIAV. Example 7

PRODUCTION OF MLV BASED PACKAGING/PRODUCER CELL LINES IN WHICH THE VSV-G EXPRESSION IS TEMPERATURE REGULATED.

The construction and preliminary characterisation of P 1 , a temperature regulated VS V- G expressing TE671 based cell line, have been described in Examples 2 and 3. We have performed experiments to determine if this cell line can be used as a starting point to construct packaging and producer cell lines to produce MLV and lentiviral vector particles pseudotyped with VSV-G.

PI cells were transfected using FuGENE6™ (Roche) with CeB (Cosset et al. 1995) and 14 clones were selected with 3μg per ml blasticidin. The MLV gag/pol expression levels of these clones was determined using western blotting (Figure 15) using a monoclonal antibody R187 (Chesebro et al.127: 134-148) to MLV p30 gag, 15μg of total protein from cell lysates was loaded in each lane of the gel. These expression levels were compared with the expression levels found in two MLV based packaging cell lines, FLYA13 (Cosset et al. 1995 Journal of Virology 69:7430-7436) and TEFLY (based on TE671 cells), which produce MLV vector particles pseudotyped with the amphotropic MLV envelope. The MLV gag/pol expression in these packaging cell lines is encoded by the CeB plasmid. From Figure 15 it can be seen that five of the fourteen clones isolated express detectable amounts of gag/pol but that the expression levels are lower than the packaging cell lines.

The ability of two of the cell lines to function as packaging cell lines were studied in the following manner. MLV particles were produced using the three-plasmid transfection method described previously (Soneoka et al, 1995). The plasmids used in the transfection were as follows; a genome plasmid, pMLV-GP (Figure 12), pHIT60 (MLV gag/pol) and pHIT456 (MLV amphotropic envelope; the latter plasmids are as Soneoka et al. (1995). The resulting vector particles were used to transduce PI -CeB clones 7 and 12. The transduced cells were grown at 37°C for seven days, the cells were then transferred to six well plates and were cultured at either 37°C or 32°C. Supernatant samples were taken from the plates every day. These were filtered through a 0.45micron filter and were stored at -20°C until the end of the experiment when all of the samples were titered on HT1080 cells. The results, shown in Figures 16a and b. The number of transducing particles produced was significantly higher at 32°C than at 37°C for both of the cell lines. The pattern of particle production was similar to that observed for VSV671, the cell line constructed by introducing pHCMV- G into TelCeBό cells (see Figure 10, Example 5). The maximum titre obtained was lower in both cases than in VSV671 but this could be due to the lower gag/pol expression levels in the two PI -CeB clones tested.

A few clones were picked from the transduced PI -CeB clone 7 pool. The effect of temperature on transducing particle production is shown in Figure 17. The pattern is similar to that observed for the pooled samples.

Example 8

PRODUCTION OF EIAV BASED PACKAGING/PRODUCER CELL LINES IN WHICH THE VSV-G EXPRESSION IS TEMPERATURE REGULATED

To determine if PI can be used as a starting point to construct packaging and producer cell lines to produce lentiviral vector particles pseudotyped with VSV-G. PI cells were transfected using FuGENEό™ (Roche) with pONYHYG. This plasmid (see Figure 18) is a derivative of pONY3 (GB patent application 9727135.7) in which an antibiotic resistance marker is linked to the expression of EIAV gag/pol using an IRES (internal ribosome entry site). Potential EIAV gag/pol expressing cell lines were selected using 200 to 400μM hygromycin B (Roche). Seven clones were selected and the amount of gag/pol expressed by each clone was quantified using western blotting. EIAV protein was detected using a polyclonal serum from an EIAV infected horse and ECL. Four of the seven clones were shown to express detectable amounts of gag/pol (see Figure 19). However it is clear that in order to produce a commercially viable packaging cell line, typically more than seven clones would be assessed. In this study, stable cell lines were used in which the levels of gag/pol expression are far higher than those observed for the present clones. Two clones AC4 and BD7 were limit dilute cloned to ensure clonality and were reassessed for the levels of gag/pol that they expressed (Figure 20). Two clones were selected for further study, AC4 clone 4 (named EV2) and BD7 clone 2 (named EV1).

These clones were transfected with pONY4G, a derivative of pONY2.1LacZ (WO 99/32646) in which the β-galactosidase gene has been replaced with the gene for green fluorescent protein. A green clone from the EV2 transfected cells was selected for further analysis. Cells were grown for four days at 32°C, the medium was removed and was replaced with medium containing a range of sodium butyrate concentrations. Sodium butyrate has been shown to enhance gene expression from CMV promoters (. The results (Figure 21) indicate that titres of nearly 10⁴ transducing particles per ml may be achieved from an EIAV based producer cell line derived from PI.

Given that higher levels of gag/pol expression may be achieved in both EIAV and MLV producer cell lines and that genome expression levels are not optimal in the EIAV model, we believe that commercially useful producer cell lines can be developed using a system in which VSV-G expression is regulated by temperature.

Example 9

CHARACTERISATION OF MLV AND EIAV BASED PACKAGING/PRODUCER CELL LINES IN WHICH THE VSV-G EXPRESSION IS TEMPERATURE REGULATED.

Southern blotting was used to begin a more detailed characterisation of the various VSV-G expressing cell lines produced (Figures 22 and 23). High molecular weight chromosomal DNA samples were isolated from TE671, VSV7, PI and the PI derivatives EV1 and EV2 using a kit (Promega cat: A1120). lOμg samples of DNA were digested either with EcoRI, Xhol or Hindlϊl restriction enzymes and were run on an 0.8% agarose gel. The gel was first probed with ³²P labelled DNA probes produced by random priming (Stratagen, Prime it random labelling kit) using a Xhol DNA fragment which contains the whole of the VSV-G coding region as a template. The restriction enzymes EcoRI and Xhol both cut either side of the coding region and hence complete copies of the VSV-G gene will be represented on the gel by a single band of 1668bp in the case of EcoRI and of 1683bp in the case of Xhol. Partial copies will be represented by bands of different sizes. Hindlll does not cut within the coding sequence and therefore may provide information on how many copies in the VSV-G gene have integrated into the genomes of the various cell lines selected. Although not 100% conclusive the resultant southern indicates that there is three copies of the VSV-G gene in PI, two of which are probably truncated, and a single complete copy in VSV7. There is no evidence from the Hindlll digest that the temperature regulation of VSV-G RNA expression is due to integration of the gene at a common site in PI and VSV7. Although derived from PI, ΕV1 looks to have under gone a deletion or insertion of a segment of DNA containing a Hindlll site in one of the Hindlll fragments containing integrated VSV-G sequences.

The gel was then striped and was re-probed with ³²P labelled DNA probes produced by random priming using a fragment of the ΕIAV coding region as a template. As would be expected ΕIAV sequences are only present in the ΕV1 and ΕV2 cell lines (Figure 23). Although the digests performed do not allow actuate quantification of the numbers of integrated copies of the gag/pol gene present, since all of the restriction enzymes used cut within the pONYHYG plasmid. The relative differences in the band intensities between EVl and EV2 suggests that there are more copies present in EVl .

Although we have shown that the levels of VSV-G mRNA present in the P 1 cells increases when cells are shifted from 37°C to 32° C (Example 6) an experiment was performed to ascertain if the level of induction is the same in the PI and VSV7 cell lines. The total RNA content of the cells was isolated using TRIZOL reagent (GIBCOBRL) and a northern blot was performed using standard methods. Total RNA was isolated from the cells grown at 37°C or at 32° C for four days. This was analysed by northern blotting and ³²P labelled VSV-G DNA probes prepared in the same manner as described for southern blotting (Figure 24). The amount of signal per lane was quantified using a phosphoimager. The RNA content of each lane ^" was standardised using probes to ribosomal RNAs and then the relative amount of VSV-G messenger RNA was calculated. The ratio of VSV-G mRNA in the cells at 32°C to 37° C was then calculated (Figure 25). The induction levels were found to be similar in the VSV7 and PI cell lines but a large amount of variation was found between the two gag/pol PI derivatives. The reason for this variation is unknown.

We considered that it was possible that temperature could also have an effect on the expression levels of MLV and EIAV gag/pol present in the cells. Promoter activity might be lower at 32°C than at 37° C. EVl and EV2 cells were grown at 37°C or at 32° C for six days and cell lysates were then prepared for western analysis as in Example 1. 15μg of total protein was loaded in each lane of a 10% tris /glycine gel, EIAV proteins were detected using polyclonal serum from an EIAV infected horse. Surprisingly the level of EIAV gag/pol present was significantly higher at 32°C than at 37° C (Figure 26). This experiment was repeated with the cell lines PI, TE671 and VSV7 included as control. In this case total RNA samples were isolated from the cells to determine by northern blotting if the increases observed at 32°C were due to effects on translation and/or protein stability or on mRNA levels as was the case for VSV-G. The northern blot was probed as previously described (Figure 27). In Figure 27 two different exposures are shown for each lane, the A lanes were exposed for a shorter time period than the B lanes. Gag/pol mRNA levels were found to increase in a similar manner to the levels of gag/pol protein present. The induction levels were calculated as for the VSV-G mRNA levels, samples being standardised relative to ribosomal RNA levels (Figure 28) The induction levels were of the same order of magnitude as was observed for VSV-G (Figure 25). Similar studies were performed on the VSV7 cell line, however in contrast to the EIAV gag/pol levels, MLV gag/pol levels in this cell line were not found to be higher at 32°C than at 37° C. This difference might give an insight into the mechanism behind the temperature regulation of VSV-G mRNA. Both the VSV-G and EIAV gag/pol genes are transcribed from CMV promoters, whilst the MLV gag/pol is transcribed from an LTR. It is possible that a temperature sensitive mutation in a factor that binds CMV promoters but not MLV LTRs is responsible for the observed phenotype.

All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are obvious to those skilled in molecular biology or related fields are intended to be within the scope of the following claims.

Claims

1. A packaging cell comprising:

(a) a first nucleotide sequence (NS) encoding a toxic viral envelope protein; and

(b) a second NS encoding a retrovirus nucleocapsid protein;

wherein the expression of the first NS is regulatable in a temperature range from 25°C to 40°C.

2. A packaging cell according to claim 1 wherein the first NS encodes a vesicular stomatitis virus (VSV)-G protein.

3. A packaging cell according to claim 1 or claim 2 wherein the expression of the first NS is upregulatable at about 32°C.

4. A packaging cell according to claim 3 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 1 or a variant thereof.

5. A packaging cell according to claim 4 wherein the expression vector is a plasmid.

6. A packaging cell according to claim 1 or claim 2 wherein the first NS is constitutively expressable at about 37°C.

7. A packaging cell according to claim 6 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 2 or a variant thereof.

8. A packaging cell according to claim 7 wherein the expression vector is a plasmid.

9. A producer cell comprising:

(a) a first NS encoding a toxic viral envelope protein;

(b) a second NS encoding a retrovirus nucleocapsid protein; and

(c) a third NS comprising a retroviral sequence capable of being encapsidated in the nucleocapsid protein;

wherein the retroviral vector particle titre obtained from same is dependent upon the temperature regulatable expression of the toxic viral envelope protein.

10. A producer cell according to claim 9 wherein the first NS encodes a VSV-G protein.

1 1. A producer cell according to claim 9 or claim 10 wherein the expression of the first NS is upregulatable at about 32°C.

12. A producer cell according to claim 1 1 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 1 or a variant thereof.

13. A producer cell according to claim 12 wherein the expression vector is a plasmid.

14. A producer cell according to claim 9 or claim 10 wherein the expression of the first NS is expressable at about 37°C.

15. A producer cell according to claim 1 1 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 2 or a variant thereof.

16. A producer cell according to claim 15 wherein the expression vector is a plasmid.

17. A producer cell according to claim 16 wherein the retroviral sequence comprises at least one nucleotide sequence of interest (NOI) that is capable of being expresssed in a target cell.

18. A virus producer cell comprising:

(a) a first NS encoding a toxic viral envelope protein; and

(b) a second NS comprising a viral sequence sufficient to produce viral vector particles from the cell;

wherein the viral vector particle titre obtained is dependent upon the temperature regulatable expression of the toxic viral envelope protein.

19. A producer cell according to claim 18 wherein the viral sequence comprises at least one NOI that is capable of expression in a target cell.

20. A method for producing a retroviral vector comprising

(a) incubating a producer cell with a second NS as defined in claim 9 and a third NS as defined in claim 9 under conditions and for a sufficient time to allow such a cell to enter a log phase of growth in a culture medium; and (b) introducing and maintaining an expression vector comprising a first NS encoding a toxic viral envelope protein into the culture medium at a temperature to induce optimal expression of the first NS;

such that sufficient retroviral vector transducing particles are produced from the cell.

21. A method according to claim 20 wherein the first NS encodes a VSV-G protein.

22. A method according to claim 20 or claim 21 wherein the expression of the first NS is upregulatable at about 32°C.

23. A method according to claim 22 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 1 or a variant thereof.

24. A method according to claim 23 wherein the expression vector is a plasmid.

25. A method according to claim 20 or claim 21 wherein the expression of the first NS is is constitutively expressable at about 37°C.

26. A method according to claim 25 wherein the first NS is obtainable from an expression vector comprising the sequence presented as SEQ ID No. 2 or a variant thereof.

27. A method according to claim 26 wherein the expression vector is a plasmid.

28. A method according to claim 27 wherein the retroviral sequence comprises at least one NOI that is capable of being expressed in a target cell.

29. A method according to claim 20 wherein the producer cell is obtainable from a group of cell lines selected from 022 and PI .

30. A method according to claim 20 wherein the second NS is derivable from an oncoretro virus and/or a lentivirus.

31. A method according to claim 20 wherein the third NS is derivable from an oncoretrovirus and/or a lentivirus.

32. A method of selecting a temperature regulatable retroviral producer cell line clone comprising:

(a) passaging VSV-G expressing producer cell lines at a first temperature between 25°C and 40°C;

(b) transfecting the producer cell lines at about 12 hours prior to temperature shift down to a second temperature between 25°C and 40°C; wherein the second temperature is lower than the first temperature;

(c) screening the producer cells for enhanced retroviral vector production at the second temperature;

(d) selecting producer cell lines showing an increase in retroviral titre for at least seven days after the temperature shift down is introduced.

33. A method according to claim 32 wherein the cell line is selected from a group consisting of 022 and PI.

34. A method for producing an retroviral vector comprising incubating the cell of claim 9 in a temperature regulatable culture medium.

35. A method for producing a retroviral gene delivery system comprising incubating the cell of claim 17 in a temperature regulated culture medium.

36. A method for producing a viral gene delivery system comprising incubating the cell of claim 19 in a temperature regulatable culture medium.

37. A retroviral vector as defined in claim 20 or a retroviral vector particle as defined in claim 9 or a retroviral delivery system as defined claim 35 wherein the retroviral vector or the retroviral vector particle or the retroviral delivery system is capable of transducing a target site.

38. A retroviral vector or a retroviral vector particle or a retroviral delivery system according to claim 37 wherein the retroviral vector or the retroviral vector particle or the retroviral delivery system is produced in sufficient amounts to effectively transduce a target site.

39. A retroviral vector or a retroviral vector particle or a retroviral delivery system according to claim 38 wherein the target site is a cell.

40. A cell transduced with a retroviral vector or a retroviral vector particle or a retroviral delivery system according to claim 39.

41. A retroviral delivery system according to any one of claims 35 to 40 for use in medicine.

42. Use of a retroviral delivery system according to any one of claims 35 to 41 in the manufacture of a pharmaceutical composition to deliver an NOI to a target site in need of same.

43. A method comprising contacting a cell with a retroviral delivery system according to any one of claims 35 to 42.

44. An expression vector for producing a retroviral vector as defined in claim 9 or a retroviral vector particle as defined in claim 20 or a retroviral delivery system as defined in claim 35 wherein the expression vector is a plasmid comprising the sequence presented as SEQ ID No. 1 or a variant thereof.

45. An expression vector for producing a producer cell defined in claim 9 or a retroviral vector particle as defined in claim 20 or a retroviral gene delivery system according defined in claim 35 wherein the expression vector is a plasmid comprising the sequence presented as SEQ ID No. 2 or a variant thereof.

46. Use of an expression vector according to claim 44 or claim 45 wherein the expression of the VSV-G protein is regulatable by temperature.

47. Use of an expression vector according to claim 44 wherein the expression of VSV-G is upregulatable at about 32°C.

48. Use of an expression vector according to claim 45 wherein the first NS is constitutive] y expressable at about 37°C.

49. Use of a cell comprising an expression vector according to claim 44 to enhance retroviral vector production wherein the enhanced retroviral production is dependent on the integrity of a second glycosylation site in the VSV-G envelope protein.

50. A stable temperature regulated producer cell line capable of expressing significant quantites of a VSV-G protein.

51. A stable producer cell line according to claim 50 wherein the cell line is selected from a group consisting of 022 and P 1.

52. A stable packaging cell or a stable producer cell with a temperature sensitive phenotype substantially as described herein and with reference to the accompanying drawings.