The quest for new improved adenovirus gene therapy vectors against glioma tumours

Abstract: Gene therapy has received much attention the last decade as a method to correct a number of disorders arising from a defective gene. Gene therapy can be defined as the introduction of a functional genetic element into a cell for a therapeutic purpose. This is a very broad term and gene therapy can be applied to a wide range of diseases from genetic diseases such as cystic fibrosis to infectious diseases or even acquired genetic diseases such as cancers.Adenoviruses (Ad) are the second most common vector for gene therapy in clinical trials today, and these vectors are mostly based on serotype 2 or 5 (Ad2 and Ad5). It has been shown that Ad2 and Ad5 use a receptor that is often downregulated in malignant cells and they also suffer from shortcomings because of the high levels of pre-existing immunity against these serotypes in the society. Hence, new and improved vectors serving as alternatives to these serotypes need to be developed to make gene therapy a successful treatment option.The work presented herein is devoted to analyse what alternative adenovirus vectors could be used for treatment of glioma brain tumours. A number of different adenovirus serotypes were screened for their ability to infect human glioma tumour cells in vitro. Established cell lines as well as low-passage glioma cells from different donors were used. Adenovirus serotype 11p (Ad11p) proved to be a promising vector candidate because of its capacity to efficiently infect the glioma cells and its low prevalence in the society. The complete genome of this serotype was sequenced to further develop this as an alternative adenovirus vector. Furthermore, a number of cell lines were produced to generate E1 deleted Ad11p vectors. Other promising vector candidates were Ad16 and a chimpanzee adenovirus called CV23. Ad16 was the most efficient human serotype to infect human low-passage glioma cells and the prevalence for this serotype is also very low. The overall most efficient virus was surprisingly the non-human CV23 virus. This adenovirus has no prevalence in humans, but efficiently infects human cells in vitro. The first analysis was made on established glioma cell lines and was followed up by using low passage glioma cells from a number of different patients. The glioma cells were analysed when subjected to <20 passages (low passage) and then again at >40 passages (high passage). The cells at a higher passage number were significantly more permissive to Ad5 than the cells analysed at a low passage number. This could in part explain why some of the promising in vitro data for Ad5 have shown a limited success in vivo. In contrast, CV23 infected the low and high passage gliomas equally. This indicates that CV23 uses an internalisation mechanism subjected to less variation than the mechanism used by Ad5.We further characterised the receptor interaction of CV23 and found that none of the previously known primary receptors for adenoviruses were of importance for binding. We found that bovine serum albumin present in the growth medium was responsible for the high binding capacity to cells. Binding is a criterion for the first step of the infection, but not necessarily a good correlate to the infection capacity. CV23 infected human cells efficiently also in the absence of bovine serum albumin.