DESCRIPCIÓN GPON

4.0. INTRODUCTION.

Previously there has been little reported work documenting the use of replication incompetent HSV vectors and gene delivery to the peripheral nervous system. The majority of reported vectors used have been those deleted for gC or tk and whilst these vectors are attenuated for growth compared to wild type virus, they are still cytotoxic in most cell types. The first reported replication incompetent vector used in the PNS was described by Dobson et al (Dobson et al. 1990), who demonstrated that a vector with an MMLVLTR//acZ cassette inserted into iCP4, was able to establish both acute and latent infections in sensory ganglia (L4 and L5). LacZ positive neurons were detected up to 24 weeks, the longest time point studied. They reported that p-galactosidase labelled neurons, as determined by histochemical staining, were more abundant after sciatic nerve inoculation rather than footpad inoculation, although only data representing sciatic nerve inoculation was shown. However, the number of lacZ

positive cells in these experiments at any time point was very low. As mentioned in chapter 3, it was speculated that virus replication would be required in order to achieve a latent virus infection in peripheral ganglia. Results such as those presented by Dobson et a/would support this hypothesis, as only very low gene transfer was seen. Here, the need to perform sciatic nerve injection in order to achieve reasonable gene delivery to DRG suggests that peripheral replication is required for a vector to reach nerve terminals following footpad inoculation.

More recently a conditionally replication incompetent vector capable of gene delivery to the PNS has been reported (Marshall et al. 2000). Here the virus was inactivated for VP 16, as described previously (Ace et al. 1989), contained a partial deletion in ICPO (which does not prevent replication) and a temperature sensitive (ts) mutation in ICP4. The mutation in ICP4 allowed replication at the permissive temperature of 31 °C, but reduced replication significantly (10^-fold) at the fully non-permissive temperature of 38°C. The vector contained an 1RES

C hapter 4________________________________________________Replication Inœ m petent Vectors

p-geo (a fusion between the p-galactosidase gene and the neomycin resistant gene) cassette inserted into the LAT region. The group found that following footpad inoculation of vector, histochemical detection of p-galactosidase in the DRG was evident at Sdays and increased through 25days, remaining consistent at 5 months. This study concluded that replication was probably not required for the establishment of a latent infection in DRG following footpad inoculation, since gene delivery was achieved even without detection of replicating vector. This therefore conflicted with the previous hypothesis that replication may be necessary for virus to reach DRG following peripheral administration. However, as the mouse footpad is likely to be below the fully non-permisive temperature of for the virus, 38°C, replication was possible even though it was not detected.

There have been reports of other non-replicating HSV vectors which have been deleted for various combinations of IE genes (Wu et al. 1996;Samaniego et al.

1997;Krisky et a/.1998;Samaniego et al. 1998). Collectively these vectors have demonstrated that deletion of multiple IE genes is necessary to minimise vector toxicity in vitro and thus it was reasoned that following further development such replication incompetent vectors would be of use in gene therapy protocols. However, as yet none of these vectors have been reported as have been tested

in vivo for gene delivery to the PNS and so their use in this capacity remains to be seen. As a preliminary investigation Krisky et al performed in vitro studies using primary DRG cultures. They found that following infection with a triple mutant for ICP4, ICP22 and ICP27, cells survived for 21 days and transgene expression persisted in some cells for at least 14days, indicating the potential of this type of vector in these circumstances.

In this chapter, studies are performed with replication incompetent vectors, testing their ability to reach spinal ganglia following peripheral administration. Since viruses impaired for IE gene expression are known to have reduced toxicity (Samaniego et al. 1998), it might be appropriate to use such vectors for peripheral gene delivery, if they were found to be effective, as they would probably have an increased safety profile. In addition, the multiple deletion of essential genes means that if such vectors are to be used in clinical situations then any recombination between replication incompetent vectors and wild type

virus should only result in the generation of further replication incompetent virus. Importantly, no wild type virus carrying the therapeutic gene should be produced.

C hapter 4________________________________________________Replication Incom petent Vectors

4.1. RESULTS.

4.1.1. Vectors with a Deletion In the Essential Gene ICP27.

ICP27 is an essential IE gene (Sacks at al. 1985) encoding a phosphoprotein. ICP27 mutants have been shown to display a variety of phenotypes as a result of the numerous regulatory functions the protein performs. Mutants do not replicate owing to a profound decrease in viral DNA synthesis and late gene expression (McMahan and Schaffer 1990;Sacks at al. 1985). In addition there is an over expression of some immediate early and early genes (McCarthy at al.

1989;Rice and Knipe 1990;McMahan and Schaffer 1990;Smith at al. 1992). ICP27 mutants also show an impairment of host protein shut-off since ICP27 normally acts to inhibit pre-mRNA splicing (Hardy and Sandri-Goldin 1994;Hardwicke and Sandri-Goldin 1994). The ICP27 protein is also toxic in vitro (Johnson at al. 1994). This information suggests that ICP27 deleted vectors are appropriate for testing in this chapter in that they require complementation in trans in order to replicate and thus will be replication incompetent in vivo. In addition, infected cells would not be subjected to any toxic effects of ICP27 expression, including the inhibition of pre-mRNA splicing.

Previous work in our laboratory has shown that a virus deleted for ICP34.5 and ICP27 can infect CNS neurons of rodents and primates in vivo with high efficiency (Howard at al. 1998). Here it was observed that a vector inactivated for ICP27 and ICP34.5 gave higher gene delivery efficiency than the ICP34.5 single mutant and in addition produced less damage around the site of injection

in vivo.

4.1.1.1. Vector Production.

Two vectors had previously been constructed; 1764 P2-/27-/vhs-pR20.9 and 1764 P2-/27-/vhs-pR20.5. Both vectors were based on the 1764 backbone virus deficient for ICP34.5/OrfP and VP 16, as described earlier. In addition both vectors were deleted for the endogenous LATP2 region (HSV nt 118768 to 120470 Dda\ to Hpal) (Lilley at al. 2001) and the ICP27 gene (HSV nt 113272 to 116869, Mlu\ to Mlu\ [this also deleted the nonessential genes UL55 and UL56]) (Howard at al. 1998). Each vector was insertionally inactivated for vhs by

insertion of an expression cassette, either pR20.9 or pR20.5 (see section 3.2.3.2.), at a unique Nru\ site in the vhs gene (HSV nt 91854). See figure 4.1.

4.1.1.2. In Vitro Vector Analysis on Permissive Cells.

Vectors were propagated on B130/2 cells (Howard at al. 1998). B130/2 cells are a BHK-derived cell line complementing the ICP27 deletion. VP16 was complemented by addition of 3mM HMBA to the media. During lytic replication on B130/2 cells the pR20.5 cassette expressed GFP and lacZ at high levels, as observed when the same cassette was inserted into UL43 flanking regions in a 1764 backbone (see figure 3.3.). However, the phenotype of the pR20.9 cassette was different to that previously described as both GFP and lacZ

expression were strong. When the pR20.9 cassette was inserted into UL43 or US5 flanking regions in a 1764 backbone, only low levels of p-galactosidase activity were detectable upon X-Gal staining (see section 3.2.3.3 ). Furthermore it has been found in our laboratory (Caroline Lilley, unpublished observations) that when the pR20.9 cassette is inserted into vhs flanking regions, in a vector also deleted for ICP4, ICP27 and LATP2, in vitro the cassette once again expresses lacZ and GFP at high levels. At present the reasons for the inconsistent phenotype of pR20.9 is unknown. It is possible that repression of LAP1 mediated by iCPO or ICP4 and sequences within LATP2, as suggested previously, does not occur when the cassette is inserted into the vhs gene. Or, when inserted into a replication incompetent vector, insufficient levels of ICP4 and/or ICPO are produced to mediate the repressive effects previously observed in 1764.

Chapter 4 Replication Incompetent Vectors