Autonome Parvoviren

Oncosuppressive properties and a low pathogenicity in adult animals
make the two rodent parvoviruses MVMp and H1 promising vectors for
cancer gene therapy. Recombinant vectors were developed from
genomes of these viruses in which a part of the
capsid-protein-coding sequence was replaced by a transgene. These
recombinant viruses can replicate their genome and express their
transgene, but they are unable to produce progeny viruses. Such
vectors are produced by cotransfection of the recombinant virus
genome and a helper plasmid which provides the capsid proteins.
During this procedure, RCVs (Replication Competent Viruses) can be
generated, probably by homologous recombination. As a consequence,
RCVs can express their capsid proteins and produce progeny viruses.
In the present work, second-generation H1- and MVMp- based
recombinant virus genomes were developed in order to minimise RCV
levels in the recombinant virus preparations. By constructing
chimeras and pseudotypes between MVMp and H1 viruses, virus
preparations with high titres were obtained which did not show any
detectable RCVs in plaque-assays. This corresponds to a
100-17000fold decrease of RCVs for the chimeras and pseudotypes. In
this work we also characterised the host-cell-tropism of the
chimeras and the pseudotypes. The chimeric vectors and Pseudotypes
allowed us to show that the restriction for viral replication of
the rat H1 genome in murine A9 cells is not due to the H1 capsids,
but can be ascribed to a 1800 bp long DNA region of the H1 genome,
located in the NS coding sequence. To our knowledge this is the
first report that convincingly shows that the tropism of a rodent
parvovirus is determined by the viral genome. Since a recombinant
MVMp genome packaged with either H1 or MVMp capsids led to viral
DNA replication in A9 cells, virus entry is not responsible for the
lack of recombinant H1 vector replication in these cells. After
transfection of A9 cells with H1 viral DNA, replication and viral
protein expression took place, whereas after H1 infection it did
not. Thus we can conclude that the cause for the restricted H1
virus replication has to be found after virus entry but before the
beginning of viral gene transcription. It is usually accepted that
viral gene transcription needs the conversion of single-stranded
Summary 134 DNA into double-stranded monomeric replicative form.
Most probably this conversion reaction is disordered in A9 cells
for the H1 genome. A defective double-stranded conversion might be
due to an interaction of VP proteins with the viral single stranded
DNA sequence. Another possibility is an interaction of a cellular
protein with the viral single stranded DNA which then leads to an
inhibition of the viral DNA conversion. It might be also possible
that a cellular factor which is needed for the conversion reaction
can not bind to the H1 genome in A9 cells. Interestingly, in most
of the tested murine cell lines the tropism was determined by the
MVMp genome, while for all human cell lines the tropism was
determined by the virus capsid. The second part of this work
focused on the effect that H1 virus infection on cellular gene
expression. For this purpose, the cDNA array technique was used. It
was previously shown that parvovirus H1 infection leads to the
activation of caspase 3 and apoptotic cell death of the human
monocytic U937 cell, in a way similar to the apoptotic cell death
induced by TNFα. Moreover, the c-myc gene, overexpressed in these
cells, is down-regulated during H1 infection. In agreement with
these data, we were able to show -by comparing the gene expression
profiles between buffer treated and H1 infected synchronised U937
cells- genes coding for caspases 2, 4 and 8 were upregulated after
H1 infection. In addition, different genes associated with the TNFα
complex were found to be up-regulated, as well as the mad4 gene,
which is able to inhibit the transcriptional activity of c-myc by
binding its main partner max. Moreover, 2 genes involved in
oestrogen-synthesis were found to be up-regulated after infection.
Oestrogen was shown to have some protective effect against TNFα
induced apoptosis in U937 cells. Induction of these genes might
reflect a defence mechanism of the cell. Also, a gene encoding for
a subunit of the 26s proteosome and different genes involved in
differentiation processes or in apoptosis were up-regulated in
infected cells. Few genes were shown to be down-regulated after
infection, such as the adenylate cyclase and cAMP dependent
transcription factor 1. Altogether, our data give a better
understanding of the molecular pathways which are used by the
parvovirus and enlighten the diversity of cellular responses to a
H1 virus infection in U937 cells.