Marco conceptual

Work presented in this chapter has shown the use of a modified recombineering strategy developed by Zsolt Ruzsics to alter or mutate BACs containing Ad vector and virus genomes.

The transgene expression cassettes in both Ad19a and Ad5 E1 and E3 deleted GFP vectors were successfully replaced with a PCR generated galK/KnR selection cassette. The resultant BACs can now serve as intermediates for the insertion of any transgene into the expression cassette of Ad19a or Ad5 which has been cloned into a corresponding shuttle plasmid with homology to the CMV promoter and SV40 polyA tail of both BACs. Two vaccination-relevant transgenes, the HIVA polyprotein and the nucleocapsid protein of PVM, were first inserted into shuttle plasmids (if not already existent) and were then inserted into the expression cassette of the galK/KnR intermediates in second stage recombination reactions. The inserts

Figure 3.10: Reconstitution of recombinant Ad19a and Ad5 viruses and expression vectors

Agarose gel electrophoresis of restriction digests with PacI 1) BAd19awt, 2) BAd19aΔ19K, 3) BAd19aGFP, 4) BAd19aHIVA, 5) BAd19aPVM-N, 6) BAd5GFP and 7) BAd5HIVA alongside the 1kb DNA ladder (M). A 6415bp fragment (indicated by star) was generated containing the bacterial sequence from each BAC so that the remaining linearised virus or vector genomes could be isolated for reconstitution by gel isolation.

were generated by both restriction digest, providing > 90bp of homology, and PCR, providing < 50bp of homology, showing that both strategies were effective regardless of length of the homologous sequence. This strategy has created Ad19a and Ad5 E1 and E3 deleted vectors which express the HIVA and PVM-N transgenes when used to infect mammalian cells. The expression of these proteins, by the reconstituted viruses will be examined in Chapter 4.

To show that the recombineering system is capable of minimal modifications, a nonsense mutation, consisting of a 5bp sequence addition, was inserted into the E3- 19K gene of Ad19awt. Both the required galK/KnR selection cassette and the modified E3/19K gene were generated by PCR with the modified E3/19K gene only requiring 26bp of homology at its 3’ end (in combination with 50bp of homology at its 5’ end) for recombination to occur. This strategy has, upon reconstitution, created a virus that lost E3/19K production whilst having no effect on the expression of any surrounding E3 proteins tested (see Chapter 4). The effect of the nonsense mutation on E3/19K and surrounding E3 protein expression will also be examined in Chapter 4.

All BACs resulting from this study were checked by restriction digest and/or DNA sequencing and were found to be correct and were transformed into DH10B cells for long-term storage due to their lack of the SW102 recombination system.

This work has demonstrated that the modified recombineering strategy can be effectively used to make both large-scale and small-scale modifications to BACs containing Ad genomes. The use of the galK/KnR selection cassette allowed for easy positive and negative selection of BACs during the process, however, it was found that some BACs underwent an unknown mutational event during second stage recombination which resulted in the loss of the galK/KnR selection cassette without incorporating the transgene and therefore false-positive colony growth. However, the number of false-positive clones was consistently 30-100 fold lower than the number of correct clones. The false-positive clones could be easily identified by restriction digest and/or sequencing and discarded.

The second stage recombination process required for transgene insertion including downstream analysis for correct clones can be done in < 10days and therefore, once established, this technique provides a quick, simple and inexpensive methodology for vaccine or gene therapy vector creation.

More importantly, this technique can also be used to generate small mutations in an Ad genome or other large genomes allowing the targeted deletion or the expression of any gene or of any mutation. This allows the simple modification of existing viruses and vectors to examine the effect on virus phenotype, for example, the deletion of one or several E3 proteins in combination from an Ad genome for studies such as those published previously (Elsing and Burgert, 1998; Ruzsics et al., 2006) or the modification of conserved amino acids within Ad proteins, as published previously, (Sester and Burgert; 1994; Hilgendorf et al., 2003) which may lead to the design of better vectors for vaccination or gene therapy.

In conclusion, the work in this chapter has generated viruses and vectors which can be used to examine the potential of Ad19a as a vaccine vector in comparison to Ad5 in several settings and this work will be the subject of subsequent chapters.

Chapter 4: Examination of generated vector and virus phenotype

including transgene expression in human cells

4.1: Introduction

This chapter will investigate the phenotypes of the viruses and vectors generated in Chapter 3.

4

The adenoviruses, like most animal viruses, have ratios of particles to infectious units greater than one. Infectious units of Ads are recorded as plaque forming units (pfu) and are based on the number of plaques, marked areas of cell death, a known dilution of virus is able to generate on a monolayer culture of mammalian cells which support adenovirus infection and replication (Chapter 2.15.1). Ad particle number can be calculated by comparing the DNA concentration of a known dilution of disrupted Ad particles to known standards (Chapter 2.15.3). It has long been known that Ads differ dramatically in their particle/pfu ratios, ranging from 10:1 for the Ad5 and subgroup B Ads like Ad11 (Holterman et al., 2004) to values several orders of magnitude higher for Ads from subgroup F such as Ad40 and Ad41 (Brown et al., 1992). Published comparisons of Adenovirus particle/pfu ratios are rare. In 1967 it was shown that Ad19p had the second highest particle/pfu ratio of the entire catalogue of Ad serotypes discovered with a calculated ratio of 1600:1 compared to the Ad5 ratio of 20:1 (Green et al., 1967). In 1978 two Ad19 isolates were calculated to have particle/pfu ratios of 9,600:1 and 25,000:1, respectively (Newland and Cooney, 1978). It is unclear what the reason is for this vast inefficiency of particle assembly. Comparison of different serotypes is therefore complicated, and dependent on the aspect of the Ad life cycle studied, pfu’s or particles may be appropriate to quantitate Ads.

Currently, in a variety of vaccination and gene therapy settings both in vitro and in vivo Ad vector dose is given by particle number rather than by pfu and as such is not based on the number of virions capable of cell transduction but rather on the physical

number of virion particles. If particle numbers are used for comparative studies of different Ads to determine MOI or dose, transgene expression is adversely affected in an Ad with a high particle/pfu ratio, due to a lower number of pfu being used. Conversely, during in vivo studies, using identical pfu numbers rather than particle numbers may result in a larger immune response to the Ad with a higher particle/pfu ratio as a larger amount of immunogenic proteins is present. For a true comparison of different Ad serotypes the particle/pfu ratios must be taken into consideration. Therefore, we decided to base our assessment of the quality and quantity of the different Ads on their ability to infect/transduce cells, hence we measured their activity in a plaque forming assay.

4.1.2: Adenovirus vector transgene expression

Ad vectors based on differing serotypes, which contain expression cassettes with identical CMV promoters and SV40 polyA tails, should produce similar transgene expression. However, previous data suggested that sequences in cis and trans may influence transgene expression (Lusky et al.,, 1999). In 2006, despite having identical CMV promoter driven expression cassettes, the Ad19a and Ad5 eGFP expressing vectors showed marked differences in eGFP expression at identical transduction efficiencies (Ruzsics et al., 2006). The reasons remained unclear. One possibility was that an SV40 enhancer downstream of the expression cassette may have stimulated GFP expression. Therefore we have used an Ad5GFP vector with the same SV40 enhancer inserted downstream of GFP. As the expression level of the transgene is crucial for vector toxicity as well as for induction of immunity against the transgene, we tested transgenes other than eGFP to evaluate if we could observe the same enhanced expression profile. This examination of transgene expression from the created vectors is possible because expression of the inserted transgenes (HIVA and PVM-N) can be detected by western blot using antibodies SV5-Pk1 and R2052, respectively. In parallel, transduction efficiency can be monitored using FACS staining with the 2Hx-2 Ab, which recognises the hexon protein of any tested Ad (Cepko et al., 1981).