Partner und Internationale Organisationen
(Englisch)
|
Michael A. Skinner, IAH, Compton (UK), Georges Beaud, IJM, Paris (F), Robert Drillien, ETS, Strasbourg (F), Marie-Paule Kieny, Transgène, Strasbourg (D), Claus-Peter Czerny, LMU, München (D), Gerd Sutter, GSF, Oberschleissheim (D)
|
Abstract
(Englisch)
|
The aim of our work was to develop recombinant vaccinia viruses (VV) expressing canine distemper virus (CDV) surface glycoproteins in the envelope. We expected that by presenting the CDV proteins to B cells and antigen presenting cells in this manner, a better immune response might be obtained. Successful expression of the HIV surface antigen in the envelope of recombinant vaccinia virus reported by Katz et al., as well as our own study demonstrating that a single chain antibody could be expressed on the surface of recombinant vaccinia virus, suggested to us that the approach was feasible.
To express the CDV proteins in the envelope of vaccinia virus we grafted the extracellular domain of the CDV H and F proteins onto the transmembrane and intraviral domains of various EEV envelope proteins.
The first approach that was chosen consisted of replacing a portion of, or the entire, VV envelope protein ectodomain coding sequences with the corresponding region of the CDV gene. This would obviously render the authentic VV protein non-functional. However, the fact that many knockout mutants of VV envelope proteins have been reported suggested that such recombinant viruses should be viable.
In a second approach we attempted to insert the chimeric protein coding sequences consisting of the VV transmembrane and intraviral domain and the CDV ectodomain into the VV thymidine kinas (TK) locus. In this case, the authentic VV gene was left intact at its natural location. Such recombinants should be easy to select on the basis of their TK-negative phenotype.
Finally, the third approach consisted of inserting the chimeric gene into the TK locus and subsequently deleting the authentic VV envelope gene. Alternatively, we planned to insert the chimeric gene into a knockout mutant in which the authentic gene had been deleted. In these recombinants we expected to obtain higher levels of expression of the chimeric gene than with co-expression as in the second group of recombinants.
We used three different promoters to drive the expression of the chimeric genes. The strong 11K promoter was expected to give high yields of recombinant protein. In case problems arose because of the high expression levels, we also planned to use the weak early/late 7.5K promoter. Finally, some constructs were made in which we used the natural promoter of the particular VV EEV gene, from which the transmembrane and intraviral domains of the recombinant protein were derived.
Recombinant vaccinia virus expressing the CDV H protein in the envelope
The first set of recombinants was made with the CDV H protein, which is a type II glycoprotein. The VV transmembrane and intraviral domains were derived from the A34R gene, which is also a type II glycoprotein. In these recombinants, the chimeric gene was designed such as to replace the authentic gene but to leave its promoter intact. The insertion plasmid also contained a marker gene allowing potent selection of recombinants.
In all recombinant viruses analyzed, the chimeric gene had been inserted, but no gene replacement was observed, indicating that a single crossing-over event had taken place leaving the authentic gene copy intact. More importantly, we were unable to detect expression of the recombinant protein.
We also generated recombinant viruses expressing chimeric CDV-VV proteins with transmembrane and intraviral domains derived from the A33R or A34R genes. In this case, the strong 11K promoter drove the expression of the recombinant proteins. The recombinant virus carrying the CDV H-A33R gene fusion expressed high levels of recombinant protein but we were unable to detect the protein in the EEV preparation. Similarly, the recombinants with the CDV H-A34R fusion gene expressed the protein well, albeit in a truncated form. Again, no expression in the envelope of EEV was observed.
The failure to detect recombinant proteins on the surface of EEV suggested to us that overexpression of these proteins might not be compatible with the their insertion into the EEV envelope. We therefore made recombinants in which the CDV H-A34R fusion gene was inserted into the TK locus and expressed from its natural promoter. With this recombinant we observed expression of the protein in infected cells but not in the envelope of recombinant viruses.
One possible explanation for the failure to detect the recombinant proteins in the envelope of EEV might be that the recombinant and authentic proteins compete for insertion into the EEV envelope. To address this possibility, a new series of recombinants were made in a parent virus in which the A34 R gene had been deleted. Chimeric genes containing the A34 A transmembrane and intraviral domains fused to the CDV H ectodomain. The chimeric genes were either placed under the control of the authentic A34 A gene promoter or under control of the 11K promoter. Downstream of these chimeric genes we also inserted the GUS gene to provide a means of selecting recombinant viruses.
Recombinant viruses expressing the fusion proteins were very difficult to purify since they formed no plaques. Despite this difficulty we managed to isolate a recombinant expressing the chimeric gene under the control of the 11K promoter. Unfortunately this virus produce very little EEV, in contrast to the parent A34 R deletion mutant which produces more EEV virus than wild-type virus.
The difficulties encountered in isolating recombinant viruses expressing chimeric CDV H-VV envelope proteins on the surface of virus particles could be explained by the large size of the recombinant protein. We therefore made a new series of recombinants in which only one half of the CDV ectodomain coding sequences were fused to the A34R intraviral and transmembrane domains. Also in these constructs we were unable to detect the recombinant fusion proteins in the envelope of EEV.
The difficulties we have encountered in isolating recombinant viruses expressing CDV H fusion proteins in the envelope can best be explained by assuming that expression of these proteins interferes at some level with the formation of the VV envelope.
Recombinant vaccinia virus expressing the CDV F protein in the envelope
Since we were unable to isolate recombinant vaccinia virus expressing CDV H proteins in the envelope, we made recombinants expressing the CDV F protein. Since this is a type I glycoprotein, we fused the F protein coding sequences to the B5 R gene intraviral and transmembrane domains, which is also a type I membrane protein. The chimeric proteins were placed under the control of the 7.5 K promoter and inserted into the VV TK locus. Three constructs were made either expressing the entire F protein, or the F1 or F2 portions of the F protein.
We readily obtained recombinant VV expressing the F1 and F2 portions of the F protein, however not the entire F protein. We have not yet been able to characterize the recombinant expressing the F2 subunit since our antibody does not detect this portion of the F protein.
Significantly, the recombinant VV virus carrying the F1 subunit gene fusion expressed that recombinant protein in the envelope as demonstrated by proteinase K shaving experiments.
|
