Vectors

Abstract

The present invention relates to a transfer vector for inserting a gene into a genetic locus of a baculovirus sequence. The transfer vector comprises an expression cassette comprising a eukaryotic promoter operably linked to the gene and a bipartite selection cassette. The present invention also relates to methods of using the transfer vector and derived bacmids and baculoviruses.

Claims

1. A suite of transfer vectors for inserting a plurality of genes that encode a plurality of proteins that are subunits of a protein complex into a baculovirus sequence comprising: i) a plurality of transfer vectors, wherein each transfer vector comprises an expression cassette comprising a eukaryotic promoter operably linked to at least one of the said genes; and ii) sequences flanking both sides of the expression cassette, wherein the sequences substantially correspond to sequences within a selected one of the genetic locus within the baculovirus selected from: ctx, egt, 39k, orf51, gp37, iap2 and odv-e56 and wherein different transfer vectors have different said flanking sequences.

2. A method for producing a recombinant bacmid, comprising: bringing a bacmid and the suite of transfer vectors of claim 1 together to allow homologous recombination; and selecting for a recombinant bacmid that comprises an expression cassette.

3. A method for producing a recombinant baculovirus comprising producing a recombinant bacmid by the method of claim 2 and culturing a eukaryotic cell containing the bacmid so that a baculovirus is produced.

4. The suite of transfer vectors of claim 1, wherein the plurality of transfer vectors are bacmids.

5. The suite of transfer vectors of claim 1, wherein the plurality of proteins interact to form a protein complex.

6. The suite of transfer vectors according to claim 5, wherein the protein complex is a virus like particle, or a chaperone complex.

7. The suite of transfer vectors of claim 1, wherein the plurality of transfer vectors comprise a recombinant baculovirus.

8. A cell in culture containing the suite of transfer vectors of claim 1.

9. A method for producing one or more proteins comprising culturing the suite of transfer vectors of claim 1.

Description

BRIEF SUMMARY OF THE DRAWINGS

(1) The present invention is now described by way of example only with reference to the following figures.

(2) FIG. 1A shows an improved selection of recombinants from ET cloning. Cartoon of DNA used for cat and bipartite recombinations. Both constructs had the same baculovirus flanking sequences (AcMNPV), Renilla luciferase expression cassette (Pp35-Rluc-Tph) and loxP sites (LoxP). The bipartite selectable marker also had zeocin resistance gene (ZeoR) and LacZα fragment. In the chloramphenicol resistant construct this was replaced with the cat gene.

(3) FIG. 1B is a bar graph showing the number of positive bacterial colonies following ET recombination. DNA used for recombination was produced either by PCR or restriction enzyme digest, as indicated.

(4) FIG. 1C is a digital image showing the correct insertion of Rluc expression cassette confirmed by PCR using one primer in the transferred DNA and one flanking the target locus in the baculovirus DNA. The correct PCR product (arrowed) would only be produced following correct integration. Results for 12 independent recombinants (1-12), no template PCR control (13), unmodified bacmid template (14) and transfer DNA template (15) are shown. Lane M is marker DNA.

(5) FIG. 1D is a bar graph showing renilla luciferase activity (relative light units) at 48 hours post infection in cell lysate from cells infected with passage 2 of recombinant bacmids 1-12 from C. Background activity in equivalent lysates from uninfected (cells) and unmodified bacmid (AcMNPV) infected cells were more than 106 fold lower.

(6) FIG. 2A shows the selective removal of marker genes. Cartoon showing strategy for ET recombination of an expression cassette (expression) into the AcMNPV Bacmid DNA and selective removal of only the marker genes (selection) by Cre mediated recombination.

(7) FIG. 2B shows the selective removal of marker genes. PCR using primers labeled a and b in A, lanes 1-2: two independent bacmid recombinants following ET recombination, lanes 3-6: four independent recombinants following Cre mediated recombination to remove the bacterial selectable markers, lane 7: no template, lane 8: unmodified bacmid DNA template, lane 9: plasmid DNA template containing the selectable marker cassette, lane 10: DNA marker. PCR products corresponding to the sizes predicted for the parental and recombinant products of the Cre mediated recombination are labeled P and R respectively.

(8) FIG. 2C shows a sequencing trace file of a representative recombinant from the PCR analysis in B confirming the presence of a defective loxP incorporating the loxP71 and loxP66 arms that render the recombinant incapable of undergoing further Cre-mediated recombination (sequences shown in SEQ ID NOS: 1 and 2).

(9) FIG. 2D is a bar graph showing renilla luciferase activity of the parental and recombinant bacmids from B when transfected into insect cells. Renilla luciferase activity was assayed at 48 hours post infection after passage of the recombinant virus. Background activity from uninfected cells is labeled Cells.

(10) FIG. 3A shows identification of additional sites for expression of recombinant proteins in the AcMNPV genome. Cartoon of AcMNPV showing relative positions of loci used for protein expression.

(11) FIG. 3B is atable showing loci used for insertion and any additional changes made to locus.

(12) FIG. 3C is a bar graph showing relative Renilla luciferase activity at 48 hours post infection with virus modified to contain an additional firefly luciferase expression cassette at each locus indicated. All viruses had the same Renilla luciferase expression cassette at the p10 locus under control of the p35 promoter. Error bars indicted the standard deviation of five replicates for each locus.

(13) FIG. 3D is a bar graph of normalised firefly luciferase activity showing relative expression of a polyhedrin promoter-firefly luciferase polyhedrin terminator cassette inserted at each locus as indicated. Error bars indicate the standard deviation from 5 replicates for each locus. Firefly luciferase was normalised using the Renilla luciferase control expressed from the same genome. Firefly luciferase insertions at the orf11, v-fgf, pe and orf23 loci were excluded due to Renilla luciferase levels more than 2 logs lower than control virus. Virus Ph has firefly luciferase at the polyhedrin locus and Renilla luciferase at the p10 locus. Virus Ph* has the same p10 Renilla luciferase insertion but no firefly luciferase insertion.

(14) FIG. 4 shows expression of influenza M1 from egt and polyhedrin loci. Left panel shows Coomassie stained SDS PAGE gel of total protein from cells infected with control baculoviruses (lanes 1 and 4), egt-M1 (lane 2) and YM1-M1 (lane 3). The sizes of protein markers (lane M) are indicated on the left hand side of the gel. Right panel shows Western blot of duplicate gel probed with anti-H7N7 antiserum.

(15) FIG. 5A shows co-expression of influenza M1 and HA and formation of VLPs. Left panel shows Coomassie stained gel of total protein; right panel shows western blot of duplicate gel probed with anti-H7N7 influenza virus serum. Cells were infected with dual baculovirus expressing HA and M1 (lane 1), baculovirus expressing HA only (lane 2), cells control (lane 3), baculovirus expressing M1 only (lane 4).

(16) FIG. 5B shows negative stain EM images of influenza VLPs.

(17) FIG. 5C shows negative stain EM images of immunogold (HA) labelled influenza VLPs with gold particles arrowed.

(18) FIG. 6A shows production of baculovirus co-expressing 4 proteins. Cell lysate from cells infected with baculovirus expressing VP2 (lane 1), VP5 (lane 2), VP3 (lane 3) and VP7 (lane 4), uninfected cells (lane 5) or all 4 proteins (lanes 6-10), position of marker proteins (M) and size in kDa is as indicated.

(19) FIG. 6B shows production of baculovirus co-expressing 4 proteins. Cell lysate for cells expressing VP5, VP2, VP3, VP7 (lanes 1-4 and partially purified VLPs (lane 5).

(20) FIG. 6C is a table showing locus of insertion and viral promoter used for expression of each BTV protein.

(21) FIG. 6D shows immunogold negative stain EM (for VP5) showing purified BTV VLPs.

(22) FIG. 7 shows crystalline plates formed on overexpression of CCT5. Quadrilateral protein aggregates in culture media of very late infections with baculovirus expressing CCT5.

(23) FIG. 8A shows expression of CCT in insect cells. Coomassie stained cell lysate from cells infected with baculovirus expressing CCT1-CCT8 (lanes 1-8, respectively), cells only (lane 9) or baculovirus coexpressing 7 CCT subunits (lane 10).

(24) FIG. 8B shows a Western blot of duplicate gel using polyclonal anti-mouse CCT antiserum.

(25) FIG. 8C includes tables showing different genetic loci and promoters used to express each of the CCT subunits.

EXAMPLES

(26) The baculovirus expression system has well established potential for the production of large amounts of correctly folded eukaryotic proteins for enzymatic and structural studies. However, it is becoming increasingly clear that many, if not most, proteins are active within cells as complexes made up of the products of several distinct genes. Given the acceleration in the pace of discovery of protein structures, there is a real need to establish systems for the rapid and reliable production and purification of protein complexes. This is particularly true of larger complexes that require the simultaneous expression, and eukaryotic folding and processing, of many protein subunits. Previously the inventors have developed baculovirus transfer vectors capable of expressing 2, 3, 4 and 5 proteins from the same baculovirus genome. The inventors' main goal in the previous studies was to produce a system capable of synthesizing large amounts of viral protein complexes containing non-equimolar amounts of virus encoded proteins for structural and enzymatic studies. One of the observations from experience with these systems is that single baculoviruses expressing multiple genes are much more efficient at forming the desired protein complexes than co-infection of insect cells with multiple baculoviruses expressing single genes. In the current study the inventors have used new technologies to exploit this observation for the production of recombinant baculoviruses expressing multiple proteins for biologically relevant mammalian protein complexes.

(27) In particular, the inventors have adapted and improved new bacterial chromosome engineering technologies for the rapid and efficient production of recombinant baculoviruses expressing multiple proteins simultaneously. The new systems developed are 30 times more efficient than conventional processes and allow the routine insertion of genes at any locus within the baculovirus genome. These studies have allowed identification of 7 new genetic loci within the baculovirus genome that allow high level expression of recombinant protein. Furthermore, the inventors have demonstrated that multiple iterative rounds of recombination can be carried out allowing the generation of virus genomes expressing multiple proteins in a complex from separate single gene insertions. As examples, the inventors have expressed VLP protein complexes for influenza A (H7 subtype) and bluetongue virus (serotype 1), as well as the 8 subunit mammalian chaperone complex CCT (TCP), that is a focus for cancer studies.

(28) Materials & Methods

(29) Multi-locus single gene insertions are achieved in E. coli using lambda red recombination. The invention includes the use of baculovirus loci that have not previously been used for multi-protein expression, and the incorporation of selectable markers that can be removed from the baculovirus genome and subsequently reused. Unlike other systems, the recombinant protein genes are not flanked by repeat sequences, improving their genetic stability.

(30) The transfer vectors contain a region of AcMNPV to target homologous recombination, an expression cassette (AcMNPV promoter, polylinker, polyadenylation signal) and a bacterial selection cassette. The AcMNPV promoter used is from either viral late (e.g. p35) or very late (e.g. Polyhedrin, p10) genes. The bacterial selection cassette consists of mutant LoxP site—bacterial selectable marker—mutant Lox P site. The mutant Lox P sites are variants on the LoxP66 and LoxP71 variants (9). The LoxP sites for each selectable marker are designed such that on recombination the marker is deleted destroying the LoxP site that remains. Each selectable marker has different LoxP mutant arms so that recombination between different selectable marker genes inserted in the same baculovirus will not occur.

(31) Choice of System used for Multi-Locus Expression of Recombinant Proteins

(32) The original research plan involved the use of homologous recombination in insect cells, or the alternative method of ET recombination in E. coli, to allow efficient insertion of expression cassettes for recombinant proteins at different loci within the baculovirus genome. Our preliminary data had suggested that linearization of transfer vector resulted in a significant increase in the recombination frequency (up to ˜30%) at a locus (p10) that did not have positive selection other than the recombinant protein produced. At the start of the research project the two systems (ET recombination and insect cell recombination) were considered with regards to their reproducibility and time taken to complete insertion of a recombinant gene in a particular locus and check expression of the resulting protein. In terms of time taken for each round of expression the ET recombination system was faster, mainly because of reduced time needed to prepare the baculovirus genome for the insertion of genes at a second genetic locus in the baculovirus. In addition it was possible to design an approach that allowed confirmation of genetic insertions independently of transgene expression, and therefore the need to grow virus in insect cells to check expression was less critical than for the wholly insect cell based system. Secondly, although it is possible to co-introduce a marker with inserted genes in the baculovirus system, the number of such markers is limited and each could only be used once. Since one of the goals of the project was to express the mouse CCT (TCP1) complex (8 subunits) the ET recombination approach was adopted as the research priority at an early stage as this system was faster and more able to accommodate multiple genetic insertions.

Example 1

(33) Development of Reagents Allowing Efficient Selection of ET Recombinants

(34) Initial experiments using the ET approach were disappointing. Experiments were performed in which a reporter gene (luciferase) was inserted into a bacterial artificial chromosome carrying the full AcMNPV genome (Bacmid) and selected for by a chloramphenicol resistance gene that was co-integrated with the luciferase reporter. Although chloramphenicol resistant bacterial colonies were recovered, subsequent analysis revealed that they did not contain the correctly modified Bacmid with the luciferase reporter. Similar results were obtained when the reporter, which was designed to only be expressed in insect cells in the presence of replicating AcMNPV, was changed to GFP. However, when the GFP construct was used for recombination in E. coli, without chloramphenicol selection, and Bacmid DNA prepared for the whole transformed E. coli population, then this DNA resulted in a few fluorescent foci when transfected into insect cells (Sf21). These data suggested that the problem was not with the recombination itself, but with the post-recombination selection of bacteria containing the insertion. To overcome these problems the inventors designed a new selection cassette that incorporated a bipartite marker system based on LacZα fragment and Zeocin resistance gene flanked by modified LoxP recombination sites (FIG. 1A). Using this system enough Zeocin was added to the selection plates to reduce but not eliminate background colony growth and recombinants were selected based on blue colony phenotype in the presence of IPTG and X-gal. To assess the relative efficiency of recovering recombinants using the chloramphenicol (cat) and bipartite selection systems the inventors carried out an experiment where recombination competent E. coli containing unmodified Bacmid were electroporated with 30 ng (˜12.5 fmol) of either the chloramphenicol based or bipartite selection cassette (FIG. 1B). As an ideal system for incorporation of foreign protein coding sequences would not involve repeated PCR of the inserted gene the inventors also compared the efficiency of recombination between the PCR amplified DNA, that is the routine for ET recombination, and restriction enzyme released, gel purified, DNA. PCR primers were designed such that the ends of the PCR products corresponded to the ends of the restriction enzyme released fragments.

(35) The bipartite selection resulted in a 20 fold increase in the number of positive colonies compared to the chloramphenicol only selection when the DNA fragment recombining was generated by PCR and a 30 fold increase when released by restriction enzyme digestion of plasmid DNA (FIG. 1B). These differences were significant (t-test, p=0.03). For the chloramphenicol selection there was no difference between the number of colonies recovered by PCR and restriction enzyme generated DNA. However, for the bipartite selection the mean number of colonies was four times higher for restriction enzyme released DNA compared to PCR amplified DNA (t-test, p=0.05). To confirm that recombinants from the new bipartite selection were genuine, representing modified bacmid containing the inserted expression construct, PCR was carried out on bacmid DNA purified from positive colonies. One primer was designed to a sequence inside the Zeocin resistance marker and one primer targeted a sequence present only in the bacmid DNA flanking the correct insertion site and not in the transfer DNA. Thus PCR product would only be produced where recombination had occurred between the linear DNA used for recombination and the bacmid. 12 separate bacmid DNA samples were tested using this method (FIG. 1C, lanes 1-12), all were positive for the PCR product diagnostic for correctly targeted recombination. In contrast, neither bacmid DNA alone, nor plasmid DNA containing the DNA fragment used for recombination was able to act as template to produce the PCR product (FIG. 1C, lanes 14 and 15, respectively). To further confirm that the recombination had resulted in the recovery of infectious recombinant baculovirus the same 12 PCR positive bacmid clones were transfected into Sf21 cells and passaged twice, then Renilla luciferase activity in cells infected with each of the recombinants was assayed at 48 hours post infection. Cells infected with each of the 12 recombinant viruses had Renilla luciferase activities that were 106 fold above background (FIG. 1D). Based on these data the inventors proceeded to use the bipartite selection for further investigations.

Example 2

(36) Cre Mediated Removal of the Selectable Marker Allows Multiple Rounds of Recombination with the Same Bipartite Selection

(37) In order to express protein complexes with multiple different subunits using single locus insertions, the bipartite marker system was designed such that the selection cassette was flanked by modified LoxP sites. These sites incorporate both the lox66 and lox71 mutations that limit Cre mediated recombination to a single round (1) and a mutation in the spacer reducing homology to wildtype loxP sites. Thus, incubation of modified bacmid with Cre recombinase results in the removal of the bipartite selectable marker and inactivation of the lox recombination site but leaves behind the baculovirus expression cassette (FIG. 2A). To confirm that this strategy could be used successfully to engineer multiple insertions in the bacmid DNA, Cre recombination was used to remove the bipartite marker from bacmid in which the Renilla luciferase reporter had been inserted. Recombination was achieved in E. coli using the EL350 cell line (2), which has both an integrated lambda prophage expressing exo, bet, and gam under control of the temperature regulated λPL promoter and Cre recombinase under control of an arabinose inducible promoter. EL350 cells containing three of the bacmids modified to incorporate the Renilla luciferase-bipartite selection insert (FIG. 1) were induced with arabinose then spread on plates containing kanamycin, to select for the bacmid, and X-gal, to screen for colonies that had lost the selectable marker by Cre mediated recombination. Bacmid DNA was purified from four putative recombinants and Cre mediated recombination confirmed by PCR using primers flanking the selectable markers (FIG. 2B). All four recombinants had PCR products consistent with the size expected from successful Cre recombination. This was further confirmed by sequencing across the modified loxP site of the 4 recombinants (FIG. 2C). The recombinants contained a disabled loxP site that incorporates both the loxP71 and loxP66 mutations and is not competent for further rounds of recombination. To further confirm that Cre mediated bacmid recombinants remained viable in insect cells, bacmid DNA was transfected into insect cells and luciferase activity assayed after two passages as before. All recombinants had Renilla luciferase activity that was equivalent to the parental bacmids before Cre recombination (FIG. 2D).

Example 3

(38) Identification of Genetic Loci in the Baculovirus Genome Suitable for High Level Expression of Heterologous Proteins

(39) Despite the extensive protein expression work that has been undertaken in the baculovirus expression system most expression has focused on replacement of the polyhedrin or p10 genes to express recombinant proteins. Relatively little literature describes the use of alternative loci for the expression of recombinant protein. In order to test whether selection could be used efficiently at different baculovirus genetic loci, a second round of recombination was carried out on one of the bacmids already containing the Renilla luciferase gene. In these experiments, the same polyhedrin promoter-Firefly luciferase-polyhdrin terminator was inserted independently at a total of 13 different genetic loci (ctx, orf11, egt, orf23, v-fgf, 39k, orf51, gp37, iap2, chiA, pe, odv-e18 and odv-e56) generating dual expression baculoviruses for Renilla and firefly luciferase proteins (FIG. 3A). Loci were selected as sites for insertion in these experiments by determining which baculovirus genes are non-essential for virus growth in tissue culture, and whether the arrangement of baculovirus genes at particular loci favoured insertion of an additional expression cassette. Nevertheless, some additional changes were made at some loci, particularly changes that resulted in knock-out of particular baculovirus genes (FIG. 3B). Recombinant viruses were passaged twice in Sf21 insect cells and on the third passage cells were harvested at 48 hours post infection, lysed, and assayed for both firefly and Renilla luciferase activities. Renilla luciferase activity was used as a marker for virus replication and protein expression as all the recombinants had the same p35 promoter driven Renilla luciferase cassette in the p10 locus. Expression of firefly luciferase at each new locus was compared to a virus carrying firefly luciferase gene at the polyhedrin locus and the same Renilla luciferase reference gene. Of the 13 loci tested, 9 had Renilla luciferase activities which were at least 10.sup.5 fold above background and of these 8 (ctx, egt, orf51, gp37, iap2, chiA, odv-e18 and odv-e56) had Renilla luciferase activity which was at or above the activity measured for the parental Renilla luciferase only, and polyhedrin locus firefly luciferase controls (FIG. 3C). The four loci (orf11, v-fgf, pe and orf23) which resulted in virus which gave Renilla luciferase activity that was within 10 fold of the background activity were excluded from further analysis. For the remaining viruses, the Renilla luciferase activity was used as a reference to normalise firefly luciferase activity and obtain a measure of relative expression of the firefly luciferase from each locus (FIG. 3D). 7 loci (ctx, egt, 39k, orf51, gp37, iap2 and odv-e56) had firefly luciferase activity which was at least 10.sup.6 fold higher than background and similar to that when the same gene was expressed from the polyhedrin locus (FIG. 3D). Two viruses (chiA and odv-e18) had high levels of Renilla luciferase but relatively poor expression of firefly luciferase.

(40) This study reveals that high level expression of foreign proteins is possible from several genetic loci within the baculovirus genome and identifies seven loci (ctx, egt, 39k, orf51, gp37, iap2 and odv-e56), in addition to polyhedrin and p10 that give good expression. Of these sites 39k, orf51 and gp37 are insertions into the DNA flanking the coding region of the gene and do not directly interrupt protein expression. In contrast, insertions into the ctx, egt, iap2 and odv-e56 each would be expected to prevent expression of the corresponding proteins from these genes. The first three of these genes have previously been described as non-essential to the growth of the virus in cell culture (3-6). Truncation of the ODV-E56 protein has also been reported (7).

(41) Of the loci that did not give good expression of the firefly luciferase reporter, four (orf11, v-fgf, pe and orf23) also resulted in a reduced expression of Renilla luciferase marker protein which was present in all recombinants. As the focus of the study was to identify sites which were suitable for insertion of very late promoter expression constructs the precise reason for this reduced expression was not investigated. Possibilities include locus-specific effects on virus replication or transcription, and disruption of essential promoter or enhancer elements for flanking genes. The low level expression of firefly luciferase in the chiA and odv-e18 insertion viruses was unexpected for different reasons. Other reports have recorded insertion of recombinant protein expression cassettes into the chiA locus (8, 9). It is possible that the reduced level of expression seen with the firefly luciferase gene in this locus in our experiments is due to the effects on genes flanking the insertion. For odv-e18, recent reports using the same mutant virus have suggested that this protein is essential for budded virus production and cell to cell movement (10, 11). These studies were based on mutants in which there was a deletion within the coding sequence of odv-e18 and the upstream flanking gene. In the experiments, where odv-e18 was inactivated by mutation of the ATG of the coding sequence to GAT followed by insertion of the firefly luciferase cassette at this point in the gene it was possible to recover infectious virus. Expression of Renilla luciferase on the third passage equivalent to that in the parental virus suggests that there was no impairment of the ability of this mutant virus to replicate. However, given the ˜2 log reduction in firefly luciferase levels compared to virus without this mutation it is not possible to rule out the possibility that a small population of virus in which the mutation was repaired was complementing a second population expressing the reporter gene.

Example 4

(42) Expression of Protein Complexes Using Multi-Locus Baculovirus Expression

(43) Example 3 focused on the use of reporter genes to quantitatively assess the potential of different baculovirus loci to express recombinant protein. To establish whether it is possible to express and recover recombinant protein complexes, three specific complexes were targeted with different numbers of protein subunits. Virus-like particles were produced for Influenza (A/seal/Mass/1/80) H7 subtype by co-expression of the viral M1 and IIA proteins (2 protein complex), and for Bluetongue virus (BTV) serotype 1 by co-expression of VP2, VP3, VP5 and VP7 (4 protein complex). In addition, the inventors targeted expression of all 8 subunits of the mouse CCT chaperone complex to be used for further functional and structural studies.

(44) Influenza A VLPs

(45) For these experiments all influenza genes were from the SC35M mouse adapted strain derived from the H7N7 (A/seal/Mass/1/80) isolate, obtained from researchers at Philipps University Marburg, Germany

(46) To test whether the multi-locus expression system was able to produce protein complexes with two proteins, the coding sequence for the influenza A M1 protein was initially inserted into the egt locus under the control of the polyhedrin promoter. For comparison, the same gene was inserted into a conventional baculovirus expression vector (pAc-YM1) targeting the polyhedrin locus. Following recombination, both resulting baculoviruses expressed the M1 protein. Levels of expression of the protein from the egt and polyhedrin loci were both significantly higher than for any other viral or cellular protein in baculovirus infected cells when assessed by coomassie stained SDS-PAGE (FIG. 4).

(47) To generate a dual virus expressing both M1 and IIA, the bacterial selection cassette was removed from the M1 expressing Bacmid by Cre recombination and the HA gene from the same influenza strain (A/seal/Mass/1/80) inserted at the p10 locus by a second round of ET recombination. Co-expression of M1 and HA from the resulting virus were confirmed by SDS-PAGE and western blot analysis as before (FIG. 5A). In addition, influenza VLPs were isolated from the culture medium of infected cells, purified by density gradient ultracentrifugation and visualised by negative stain EM analysis (FIG. 5B). To ensure that VLPs visualised were indeed influenza, particles were immunogold labeled using antibodies specific for influenza HA (FIG. 5C).

(48) Taken together, these data were good evidence that the multilocus approach could be used successfully for recombinant protein expression and that it was practical to produce such a recombinant virus through two rounds of insertion of the bipartite selectable marker system. Flu VLP have been widely tested by others and shown to be immunogenic in mice and ferrets. This method of constructing the VLP has the advantage that it is possible to have the baculovirus genome with M1 already pre-integrated and ready for expression. Thus, to make VLP it should only be necessary to perform a single round of recombination to add the HA gene from whatever emerging flu subtype is required for vaccine production. For VLP based flu vaccines this would simplify the necessary cloning and potentially increase the speed with which new types of VLP could be made.

(49) BTV1 VLPs

(50) Unlike influenza, where VLP can be formed by expression of just two proteins, BTV VLP require the co-ordinated expression of four structural proteins (VP2, VP3, VP5 and VP7). In order to further demonstrate the usefulness of the new system for recombination, viruses were produced that expressed each protein singly and the combination of all four BTV structural proteins (FIG. 6). One of the advantages of the new system is that viruses expressing single proteins and multiple proteins can be made from the same set of transfer vectors. This facilitates production of control viruses expressing single proteins to act as markers for the position of proteins co-expressed in the full complex. For the BTV VLP a different set of loci were used from those in the influenza example (FIG. 6C). In addition, the new system was combined with the orf1629 knockout Bacmid described by Zhao et al., (12) to allow use of a conventional transfer vector at the polyhedrin locus to express one of the subunits of the complex. Formation of VLPs was confirmed by particle morphology under negative stain EM, combined with immunogold labelling for one of the outer capsid proteins (VP5) (FIG. 6D).

(51) Mouse CCT

(52) The application of the present invention is not limited to the formation of virus-like particles. Indeed, as many cellular proteins are present in cells as complexes with more than one subunit, the efficient formation of these complexes has application to a range of fields. To demonstrate the usefulness of the system to other studies the chaperone complex CCT was chosen. This complex has implications in cancer research and with 8 subunits represents a significant challenge for protein expression. Indeed it would not have been possible to express this complex in the baculovirus system without the vectors and methods of the present invention. Eight transfer vectors, each expressing one of the subunits for the mouse CCT and each targeting a different locus, identified in Example 3 above, were constructed. These were used to generate recombinant virus expressing each subunit alone and combinations of subunits. For one of the subunits (CCT5) an unusual phenotype was noted when the protein was over expressed in cells. Crystalline needles were noted inside cells and, late in infection when cells had ruptured, quadrilateral plates of crystalline material were noted in the culture medium (FIG. 7). A purification scheme was generated for this protein and this was supplied with preliminary crystallisation conditions and infected cell material to the collaborator for the CCT part of the project.

(53) Although expression of some of the other CCT subunits resulted in visible aggregates in cells late in infection, none resulted in aggregates with such a regular appearance.

(54) All 8 CCT subunits had high level expression (FIG. 8A) and cross reacted with a polyclonal antiserum raised to mouse CCT. There was some cross reaction between one of the endogenous insect cell CCT subunits and the polyclonal antibody (FIG. 8B, lane 9). However, this endogenous signal only co-located with 3 subunits (CCT1, CCT5 and CCT6) which all accumulated to high level in infected cells and thus were clearly detectable. All other subunits migrated as slightly different molecular mass proteins and thus could be distinguished from the insect cell protein on the basis of migration.

CONCLUSIONS

(55) For all three protein complexes there were clear differences in the accumulation of the same recombinant protein when expressed alone or in presence of other proteins. In almost all cases there was a reduction in the level of protein expression when multiple proteins were expressed from the same virus. This was to a certain extent expected. Competition for proteins required for transcription, RNA processing, translation and protein folding would predict that two highly expressed baculovirus genes would have lower expression together than separately. However what was striking for both the BTV VLP and CCT examples was that reduction in protein expression was variable between different genetic loci. For example with BTV, VP2, VP5, VP3 and VP7 all resulted in significant and similar accumulation of protein when expressed on their own. However, when combined in a single baculovirus expressing all four proteins, VP2 and VP3 accumulated to lower levels than VP5 and VP7 (FIG. 6A). This effect cannot be explained by the position effect of genomic integration since in both the single and quadruple viruses the genes were expressed from the same genetic loci. One possible explanation would be that the effect was due to the different promoters that were used for the different genes. Both VP5 and VP7 were expressed under control of the p10 promoter and VP2 and VP3 were under control of the polyhedrin promoter. This would be consistent with a report in the literature where deletion of the p10 gene resulted in increased expression from polyhedrin locus 13. Thus duplication of the polyhedrin promoter in the presence of two p10 promoters would be predicted to selectively reduce expression from the polyhedrin promoter driven genes. However, this cannot be a complete explanation for the effect observed in the case of the CCT complex. Both CCT5 and CCT2 were expressed from the p10 promoter and gave high steady-state levels of recombinant protein when expressed alone (FIGS. 8A and B, lanes 2 and 5). However, when combined in the same virus, expression of CCT5 was substantially reduced compared to CCT2 (FIGS. 8A and B, lane 10). From these data, other cis acting sequences present at the p10 locus may be contributing to the enhanced relative expression of CCT2 and BTV VP5, which were both inserted at this locus. The relatively low apparent expression of influenza HA when inserted at the same site may be more related to turnover of this protein in the ER rather than transcriptional effects. Both CCT2 and VP5 accumulate in the cytoplasm.

(56) The inventors have improved the efficiency of the ET recombination system as applied to baculovirus to the point that it can be used for routine insertion of expression cassettes for recombinant proteins. Furthermore, they have identified seven genetic loci (ctx, egt, 39k, orf51, gp37, iap2 and odv-e56) that can be used for high level expression of protein and demonstrated that multi-protein complexes can be assembled using this system using three examples (influenza A and BTV VLPs, and CCT complex).

(57) All documents cited above are incorporated herein by reference.

REFERENCES CITED

(58) 1. Albert, H., Dale, E. C., Lee, E. & Ow, D. W. Plant J 7, 649-59. (1995). 2. Lee, E. C. et al. Genomics 73, 56-65 (2001). 3. Eldridge, R., Li, Y. & Miller, L. K. J Virol 66, 6563-71 (1992). 4. Flipsen, J. T., Mans, R. M., Kleefsman, A. W., Knebel-Morsdorf, D. & Vlak, J. M. J Virol 69, 4529-32 (1995). 5. O'Reilly, D. R. & Miller, L. K. Science 245, 1110-2 (1989). 6. Griffiths, C. M. et al. J Gen Virol 80 (Pt 4), 1055-66 (1999). 7. Braunagel, S. C., Elton, D. M., Ma, H. & Summers, M. D. Virology 217, 97-110 (1996). 8. Berger, I., Fitzgerald, D. J. & Richmond, T. J. Nat Biotechnol 22, 1583-7 (2004). 9. Fitzgerald, D. J. et al. Nat Methods 3, 1021-32 (2006). 10. McCarthy, C. B., Dai, X., Donly, C. & Theilmann, D. A. Virology 372, 325-39 (2008). 11. McCarthy, C. B. & Theilmann, D. A. Virology 375, 277-91 (2008). 12. Zhao, Y., Chapman, D. A. & Jones, I. M. Nucleic Acids Res 31, E6-6. (2003). 13. Chaabihi, H. et al. J Virol 67, 2664-71 (1993). 14. Summers, M. D. Adv Virus Res 68, 3-73 (2006). 15. Emery, V. C. & Bishop, D. H. Protein Eng 1, 359-66 (1987). 16. French, T. J., Marshall, J. J. & Roy, P. J Virol 64, 5695-700 (1990). 17. Latham, T. & Galarza, J. M. J Virol 75, 6154-65. (2001). 18. Pushko, P. et al. Vaccine 19, 142-53. (2000). 19. Ye, L. et al. Virology 351, 260-70 (2006). 20. Berger, I., Fitzgerald, D. J. & Richmond, T. J. Nat Biotechnol 22, 1583-7 (2004). 21. Fitzgerald, D. J. et al. Nat Methods 3, 1021-32 (2006). 22. French, T. J. & Roy, P. J Virol 64, 1530-6 (1990). 23. Devos, D. & Russell, R. B. Curr Opin Struct Biol 17, 370-7 (2007). 24. O'Neal, C. M., Clements, J. D., Estes, M. K. & Conner, M. E. J Virol 72, 3390-3 (1998). 25. Palomares, L. A., Lopez, S. & Ramirez, O. T. Biotechnol Bioeng 78, 635-44. (2002). 26. Mena, J. A., Ramirez, O. T. & Palomares, L. A. BMC Biotechnol 7, 39 (2007). 27. Weyer, U. & Possee, R. D. J Gen Virol 72 (Pt 12), 2967-74 (1991). 28. Bertolotti-Ciarlet, A., Ciarlet, M., Crawford, S. E., Conner, M. E. & Estes, M. K. Vaccine 21, 3885-900 (2003). 29. Kamita, S. G., Maeda, S. & Hammock, B. D. J Virol 77, 13053-61 (2003). 30. Crouch, E. A. & Passarelli, A. L. J Virol 76, 9323-34 (2002). 31. Mikhailov, V. S., Okano, K. & Rohrmann, G. F. J Virol 77, 2436-44 (2003). 32. Pijlman, G. P. et al. J Virol 76, 5605-11 (2002). 33. Pijlman, G. P., van den Born, E., Martens, D. E. & Vlak, J. M. Virology 283, 132-8 (2001). 34. Pijlman, G. P., van Schijndel, J. E. & Vlak, J. M. J Gen Virol 84, 2669-78 (2003). 35. Vanarsdall, A. L., Pearson, M. N. & Rohrmann, G. F. Virology 367, 187-95 (2007). 36. Vanarsdall, A. L., Mikhailov, V. S. & Rohrmann, G. F. Virology 364, 475-85 (2007). 37. Okano, K., Vanarsdall, A. L. & Rohrmann, G. F. Virology 359, 46-54 (2007). 38. Vanarsdall, A. L., Okano, K. & Rohrmann, G. F. J Virol 80, 1724-33 (2006). 39. Okano, K., Vanarsdall, A. L. & Rohrmann, G. F. J Virol 78, 10650-6 (2004). 40. Vanarsdall, A. L., Okano, K. & Rohrmann, G. F. Virology 326, 191-201 (2004). 41. Fang, M., Dai, X. & Theilmann, D. A. J Virol 81, 9859-69 (2007). 42. Wang, Y. et al. Virology 367, 71-81 (2007). 43. Yamagishi, J., Burnett, E. D., Harwood, S. H. & Blissard, G. W. Virology 365, 34-47 (2007). 44. Xi, Q., Wang, J., Deng, R. & Wang, X. Virus Genes 34, 223-32 (2007). 45. Li, Y. et al. Virus Genes 31, 275-84 (2005). 46. Stewart, T. M., Huijskens, I., Willis, L. G. & Theilmann, D. A. J Virol 79, 4619-29 (2005). 47. Kamita, S. G. et al. Proc Natl Acad Sci USA 102, 2584-9 (2005). 48. Milks, M. L., Washburn, J. O., Willis, L. G., Volkman, L. E. & Theilmann, D. A. Virology 310, 224-34 (2003). 49. Zhao, Y., Chapman, D. A. & Jones, I. M. Nucleic Acids Res 31, E6-6 (2003).