Press releases

Release issued 5th March 2003

To see a full, illustrated, PDF version of this article, go to:

http://www.biovex.com/downloads/EBR.pdf

Text:

The recent sequencing of the human genome has produced a plethora of potential novel drug targets and candidate genes for gene therapy. New drug discovery using large- scale functional genomics approaches, however, faces the major problem of target validation. Traditionally this has been achieved by pharmacology, using specific antagonists and agonists, or by using transgenic animals, particularly knock-out mice.

However, these approaches have considerable drawbacks, being extremely time-consuming, laborious and expensive. This is particularly the case if large numbers of potential target genes are to be tested.

Recently there has been much work developing other methods of validating genes as drug targets, such as the construction of gene vectors to overexpress or down- regulate the gene of interest. The use of vectors has many advantages over traditional approaches in that they are relatively easy, quick and cost-effective to manufacture, and they give the ability to spatially and temporally regulate the expression of the gene of interest, thus avoiding the complication of any developmental role of the gene product. The production of efficient vector systems is of particular relevance to the nervous system, as non-toxic gene delivery to neurons both in vitro and in vivo has proven to be difficult. This is because neurons are postmitotic terminally differentiated complex cells that are notoriously hard to transfect.

An ideal vector for gene delivery to the nervous system should have the following properties:

 It should infect neurons efficiently

 It should give high levels of gene expression

 It should be non-toxic

 It should give long-term gene expression

 Retrograde transport would facilitate delivery to otherwise inaccessible sites

 The vector should not elicit an immune response

 It should have the ability to deliver large or multiple genes

 There should be no replication in the target cell

 It should be easily produced and manipulated

To date no single delivery system has had all of these ideal attributes and most vector systems have characteristics that make them suitable only for specific applications.

NON-VIRAL VECTORS

There have been many different vectors developed for gene delivery to the nervous system based on both non-viral and viral systems (1). Non-viral vectors are usually based on either plasmid DNA complexed with liposomes that fuse with the host cell membrane and introduce the DNA intracellularly, or on naked plasmid DNA being fired into the host cell using a biolistic ('gene gun') approach. These systems have advantages over viral systems in that they are easy and cost-effective to use. However, to date, low efficiency of gene delivery, particularly to the nervous system in vivo, and the transient gene expression achieved has limited the use and development of these technologies.

RETROVIRAL VECTORS

Most vector systems currently being developed are based on viruses exploiting their ability to efficiently infect and deliver exogenous DNA to target cells. Retroviruses were the first viruses to be developed as vectors and most retroviral vectors are based on the Moloney murine leukemia virus (MMLV) (2). Retroviruses are relatively small (approximately 10kb) enveloped single stranded RNA viruses. These viruses infect a wide range of host cells where their genome is reverse transcribed to double stranded DNA and integrated into the host cell genome.

This integration provides a basis for long-term expression and means that the exogenous gene is replicated as the host cell divides and is therefore not diluted out. However this integration is also a potential drawback as it is not site specific and may cause the activation or inactivation of important genes (for example oncogenes or tumour suppressor genes). Another problem is that retroviruses do not efficiently infect non-dividing cells, as they require the onset of mitosis to enter the host cell nucleus for integration, which prevents their use for gene delivery to neurons of the nervous system.

LENTIVIRAL VECTORS

There is, however, a subgroup of retroviruses that are able to infect non-dividing cells, including neurons (3, 4). These are the lentiviruses, the best-known members being human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV). Vectors based on lentiviruses have the advantage of producing stable expression of exogenous genes in neurons for extended periods of time. Another advantage of these and most retroviral vectors is that they can be pseudotyped, which means that the envelope protein of the virus can be substituted for that of another virus allowing the target cell type to be altered. This could potentially allow any cell type to be targeted. The main problem with lentiviral vectors is safety, due to the possibility of insertional activation or inactivation and the potential for recombination to produce replication- competent virus in vivo.

Another limiting factor with retroviruses in general is their small size, allowing the insertion of only relatively small amounts of exogenous DNA before the efficiency of the vector's packaging becomes compromised.

ADENOVIRAL VECTORS

Adenoviruses are non-enveloped double stranded DNA viruses of approximately 35kb in length. They have many potential advantages for use as a vector. They can naturally infect a broad range of host cells (both dividing and non-dividing) where their genome remains as an episomal element within the nucleus of the host cell (5, 6, 7). They are also relatively easy to generate and manipulate and can be grown to high titre. The main limitations to the use of adenoviral vectors in the nervous system are their immunogenicity and cytotoxicity and the transient nature of transgene expression. This limited expression is usually due to the cytotoxicity of the vector and subsequent clearance of infected cells by the immune system. Adenoviral vectors are also not cell-type specific. Therefore after inoculation in the nervous system, all cell-types present (neurons, oligodendrocytes and astrocytes) are infected. More recently 'gutless' adenoviral vectors that contain no adenoviral genes have been developed with subsequent improvements in their characteristics (8). However the main problem with this system is that purifying the 'gutless' adenovirus away from the toxic helper virus required to produce the vector has proved difficult.

ADENO-ASSOCIATED VIRAL VECTORS

Adeno-associated virus (AAV) is a small single stranded nonpathogenic DNA virus that requires another virus (usually adenovirus or herpes simplex virus) to replicate. Upon infection, wild-type AAV integrates into the host genome in a site-specific manner, but when using a recombinant virus this site-specificity is lost, with the subsequent potential problems of insertional activation or inactivation. AAV vectors have been shown to efficiently infect neurons within the brain for extended periods of time with little cytotoxicity (9, 10). However there are several limitations to the use of recombinant AAV vectors. The size of the transgene is severely limited due to the small size of the vector, where only about 4kb of DNA can be inserted due to packaging constraints. Another potential concern is the contamination of pure AAV stocks with the helper virus that is required to grow the vector. Also, as AAV is a single stranded DNA virus, double strand synthesis is required for transgene expression, which does not occur in neurons due to the lack of DNA replication.

Properties of Disabled Genomic HSV-1 Vectors:

 Neurotropic

 Ability to be transported along axons

 Establishes life long latency in neurons

 Large genome allowing large or multiple inserts

 Does not integrate into the host genome avoiding the risk of insertional mutagenesis

 Replication incompetent and non-toxic to host cell

 Use of the promoters of the LAT region allows high levels of transgene expression for extended periods of time

 Able to grow to high titre

 Relatively easy, quick and cost-effective to produce

HERPES SIMPLEX VIRAL VECTORS

Herpes simplex virus type 1 (HSV-1) is a large double stranded DNA virus of 152kb in length that encodes approximately 80 genes. It has many of the properties required for an efficient viral vector for target validation in the nervous system - see Table 1 (reviewed in 11, 12). It is a neurotropic virus that has naturally evolved to infect neurons, where it establishes a latent phenotype for the life of the host cell. During latency, the virus is asymptomatic and remains extrachromosomal in the nucleus of the host cell, and therefore does not have the potential to cause insertional mutagenesis in the host cell. However during latency most of the HSV-1 genome (and any exogenous transgene) is transcriptionally silenced except for a limited region of the HSV-1 genome that remains active to generate the latency associated transcripts (LATs) (13). This activity during latency suggests that it is possible to utilise the promoters of the LAT region to produce long-term, or even lifelong, transgene expression from a latent vector. HSV-1 contains many genes that are not essential for viral growth, which can be replaced by large or multiple inserts (combinations of genes of interest or a gene of interest with a marker gene such as β- galactosidase or green fluorescent protein) without compromising vector function. Due to the large size of HSV-1 this means that approximately 35kb of genetic material can be accommodated without affecting the packaging ability of the virus. Also, in addition to its ability to infect neurons, HSV- 1 naturally evolved to be retrogradely and anterogradely transported along the length of axons. This conveys the ability of the vector to express the transgene of interest at sites that are anatomically connected to, but distant from, the site of inoculation, potentially allowing peripheral or remote administration. There are also practical advantages in using HSV-1 based vectors. The viral genome is well characterised and easily manipulated by conventional molecular biological techniques and the vector can be produced to high titre in vitro.

There are two main types of HSV-1 based vector: defective vectors (amplicons) and disabled genomic vectors. Both of these attempt to utilise the potential advantages of the virus with greater or lesser success.

DEFECTIVE HSV-1 VIRAL VECTORS (AMPLICONS)

Amplicons are defective HSV-1 virus particles that are produced from plasmids containing the gene of interest, an origin of replication and a packaging signal. These plasmids are incapable of replicating without an HSV-1 helper virus (14). The plasmid is then co-transfected with a helper virus where, during viral replication, the plasmid is also replicated and packaged into virus particles. These particles are then replication-incompetent. The advantages of this system are that the only manipulation that is required is to the amplicon plasmid and is therefore straightforward, and also that each virus particle contains multiple copies of the gene of interest thus allowing high (but varying) levels of gene expression. As the amplicon plasmid sequence is not surrounded by the rest of the HSV-1 genome, the silencing of promoters during latency there will not be as profound. Therefore the choice of promoter construct utilised in defective HSV-1 vectors may not be as crucial as in disabled HSV-1 vectors. The main disadvantage of these vectors is that the defective vector cannot be separated from the helper virus. Therefore the toxicity of the vector is dependent on the disablement of the helper virus used to grow the stock. More recently novel methods of producing stocks that are helper virus free have been developed (15), but despite these recent improvements several limitations remain. The plasmid sequences delivered by amplicons are not maintained in a true latent state within the nucleus of neurons and may be degraded. Consequently long-term expression has been difficult to achieve in the nervous system in vivo. Also the yields of vector obtained from the helper free systems are at present very low and therefore are not practical for widespread use.

DISABLED GENOMIC HSV-1 VIRAL VECTORS

The second method of generating HSV-1 based vectors is by inserting the transgene of interest directly into genomic HSV-1 that has been disabled at specific loci (11, 12). These disablements can be deletions of genes that are essential for viral replication and thus the vector can be rendered replication-incompetent. These essential genes have to be complemented for in trans by a cell line to allow the vector to be grown to high titre in culture. The transgene of interest is inserted in a shuttle plasmid where it is flanked by the sequence of HSV-1 DNA that is required to be deleted from the vector. Following co-transfection of this plasmid with infectious HSV-1, viral DNA homologous recombination occurs knocking out the HSV-1 gene and replacing it with the transgene.

Despite the properties of HSV-1 that make it attractive for use as a vector for target validation in the nervous system, two main features of the virus have to be overcome before the virus can be usefully exploited: viral toxicity and transience of transgene expression. Even a replicationincompetent HSV-1 vector is toxic to neurons without appropriate modification to prevent HSV-1 gene expression to remove this toxicity. Furthermore, as HSV-1 does not integrate into the host genome and HSV-1 gene expression is very efficiently silenced during latency, obtaining long-term transgene expression requires the use of particular latently active promoters. Upon infection, HSV-1 gene expression occurs in a cascade fashion (see Figure 1). That is, the powerful transactivator (VP16) induces the expression of the five immediate early HSV-1 genes, which in turn activate the early genes and then the late genes. The early genes are primarily involved in viral DNA synthesis. Following the expression of the early genes and subsequent to viral DNA synthesis, the immediate early genes activate the expression of the late genes. These late genes encode most of the structural proteins of HSV-1 and following their expression, viral DNA cleavage and packaging of progeny virions can occur. Due to the cascade nature of this gene expression it is hypothesised that the powerful transactivator VP16 and the immediate early genes would be good candidates for disablement to prevent HSV-1 gene expression from the vector and thus minimise toxicity. However, deletion of these genes is not straightforward to achieve without compromising the ability of the vector to be efficiently grown in culture. Firstly, VP16 cannot be deleted, as it is an essential structural protein of the virion. However the transactiving domain of VP16 can be removed whilst leaving the structural domain intact. Secondly, making a cell line that complements for deletions of the essential immediate early genes has proved difficult, as the proteins encoded by the immediate early genes are themselves toxic. This has been overcome by producing a cell line where the complementing immediate early genes, along with a complementing VP16 gene, are expressed from promoters that are only active upon infection by the virus, but not expressed during routine passage and growth of the cells.

Recently developed genomic HSV-1 vectors have immediate early gene deletions in combination with an inactivation in VP16 (16, 17). These viruses have been shown to be non- toxic in a variety of neuronal cell types in vitro and in the nervous system in vivo. These viruses have numerous deletions within their genome, which is an important safety feature as it removes the possibility of the disabled vector reverting back to a wildtype HSV-1 phenotype.

The other main obstacle that needed to be overcome before using disabled HSV-1 vectors is the lack of long-term transgene expression. Most viral gene expression, and that of any exogenous promoters, is very tightly regulated and silenced as the virus enters latency. Therefore most gene expression, although high soon after inoculation in the nervous system, is very low after a few days. However, the virus does express certain RNA transcripts during latency (the LATs), which can be deleted without affecting the ability of the virus to go latent, and this can be exploited for the expression of exogenous genes. Two promoters, LAP1 and LAP2, control LAT gene expression. Extensive work has now shown that utilising the LAT promoters on their own does not confer long-term expression on an exogenous gene during latency. Recently, promoter constructs have been developed that utilise LAT promoter enhancer regions to confer long-term properties on strong exogenous promoters, such as the CMV promoter. These heterologous promoter constructs allow the strong short-term properties of the CMV promoter to be maintained for extended periods of time during latency.

This new approach has combined the low toxicity of a disabled HSV-1 vector (where the vector is replication-incompetent and there is no viral gene expression) with the heterologous LAT/CMV promoter systems, which allow high levels of transgene expression to occur in the long-term (16, 17). These vectors allow long-term non-toxic gene expression to various neurons in vitro, to the peripheral nervous system (dorsal root ganglia) and the central nervous system (brain and spinal cord) in vivo. These vectors also retain many of the useful characteristics of HSV-1, such as the ability to be transported to sites distal, but anatomically connected to, the inoculation site following a single injection. This not only aids delivery to otherwise anatomically inaccessible sites (for example dorsal root ganglia), but also greatly increases the area of transduction that occurs in the brain or spinal cord, for example. The ability of the vector to be axonally transported is an inherent property of HSV. Other viruses being developed as vectors do not naturally do this and so do not give as widespread gene expression as is possible with HSV. This advantage allows the vector to be administered either by a peripheral route to give gene delivery in the peripheral nervous system (for example delivery to the dorsal root ganglia following peripheral injection in the nerve or footpad), or to relatively inaccessible and discrete regions of the central nervous system (for example the substantia nigra following injection in the striatum).

With the recent sequencing of the human genome, lucidating the function of the novel genes that are rapidly being found is of major importance. These vectors provide a relatively easy approach to attributing a physiological function to these genes in the mature adult nervous system. Therefore disabled replication-incompetent HSV-1 vectors containing the appropriate promoter constructs with these novel transgenes could be used to quickly screen the relevance or not of particular genes in in vivo and in vitro models of disease states, or the function of the genes in the intact nervous system. These disabled HSV-1 vectors may be potentially powerful tools for assigning gene function to the excess of information currently being obtained. In the long- term this will be an extremely costeffective method of screening out genes of use for the potential development of novel drugs in a large-scale functional genomics-based approach to drug discovery.

By Dr James Palmer, Senior Scientist, and Dr Robert Coffin, Chief Scientific Officer, at BioVex Ltd

Dr James Palmer joined BioVex Ltd, a gene delivery and therapeutic vaccine company, in September 1999. He was previously a post-doctoral fellow at University College London in the academic laboratory of BioVex's Founding Scientist and current Research Director, Dr Robert Coffin. Dr Palmer is currently a Senior Scientist developing herpes simplex viral vector systems for gene delivery to the nervous system for target validation purposes. Prior to this, he obtained his PhD in Molecular Neuroscience from the University of Cambridge working in the Neurobiology Division of the MRC Laboratory of Molecular Biology, and his BSc in Pharmacology from the University of Bristol.

Dr Robert Coffin is a Founding Scientist at Biovex Ltd, Lecturer in Virology and Gene Therapy at University College London (UCL) and Honorary Senior Lecturer at The Institute of Child Health. Following a PhD in Virology at Imperial College, Dr Coffin moved to UCL to develop herpes simplex virus (HSV) based vectors and to study the molecular basis of HSV latency. During the course of this work - which led to the foundation of Biovex Ltd where he is CSO - vectors were developed that are useful for studying, and potentially treating, neurological disease and for the immune-based treatment of various diseases. Dr Coffin is a world recognised expert in the field of HSV vectorology and latency, and has authored numerous scientific publications.

 The authors can be contacted at:

jpalmer@biovex.com and rcoffin@biovex.com

References

(1) Li and Huang, Nonviral gene therapy: promises and challenges, Gene Therapy 7: pp31-34, 2000

(2) Miller et al, Use of retroviral vectors for gene transfer and expression, Methods Enzymol. 217: pp581-599, 1993

(3) Naldini et al, In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector, Science 272: pp263-267, 1996

(4) Blömer et al, Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector, J. Virol. 71: pp6,641-6,649, 1997

(5) Akli et al, Transfer of a foreign gene into the brain using adenovirus vectors, Nat. Genetics 3: pp224-228, 1993

(6) Bajocchi et al, Direct in vivo gene transfer to ependymal cells in the central nervous system using recombinant adenovirus vectors, Nat. Genetics 3: pp229-234, 1993

(7) Le Gal La Salle et al, An adenovirus vector for gene transfer into neurons and glia in the brain, Science 259: pp988-990, 1993

(8) Kochanek et al, A new adenoviral vector: replacement of all viral coding sequences with 28 kb of DNA independently expressing both full-length dystrophin and beta-galactosidase, PNAS 93: pp5,731-5,736

(9) Kapplitt et al, Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain, Nat. Genetics 8: pp148-154, 1994

(10) Xiao et al, Gene transfer by adeno-associated virus vectors in to the central nervous system, Exp. Neurol. 144: pp113-124, 1997

(11) Fink et al, Gene transfer to neurons using herpes simplex virus-based vectors, Ann. Rev. Neurosci. 19: pp265-287, 1996

(12) Coffin and Latchman, Herpes simplex based vectors, Genetic manipulation of the nervous system, Academic Press, UK, 1996

(13) Stevens et al, RNA complementary to a herpes virus alpha gene mRNA is prominent in latently infected neurons, Science 235: pp1,056-1,059, 1987

(14) Sena-Esteves et al, HSV-1 amplicon vectors - simplicity and versatility, Mol. Ther. 2: pp9-15, 2000

(15) Fraefel et al, Helper virus-free transfer of herpes simplex virus type 1 plasmid vectors into neural cells, J. Virol. 70: pp7,190-7,197, 1996

(16) Palmer et al, Development and optimization of herpes simplex virus vectors for multiple long-term gene delivery to the peripheral nervous system, J. Virol. 74: pp5,604-5,618, 2000

(17) Lilley et al, Multiple immediate gene-deficient herpes simplex virus vectors allowing efficient gene delivery to neurons in culture and widespread gene delivery to the central nervous system in vivo, J. Virol. 75: pp4,343-4,356, 2001

Reprinted from European BioPharmaceutical Review Winter 2002 issue.

Copyright: Samedan Ltd. 2003

For more information on Target Validation see:

http://www.biovex.com/neurovex.html

(6th March 2003) BIOVEX ANNOUNCES PRELIMINARY PHASE I DATA ON ONCOVEX (GM-CSF) IN PATIENTS WITH BREAST CANCER AND MELANOMA

(18th December 2002) 11th International Conference on Cancer Gene Therapy