Expression of heterologous genes in oncolytic ... - CiteSeerX

Journal of General Virology (2008), 89, 389–396

DOI 10.1099/vir.0.83444-0

Expression of heterologous genes in oncolytic adenoviruses using picornaviral 2A sequences that trigger ribosome skipping Garth M. Funston,1 Susanna E. Kallioinen,1 Pablo de Felipe,1 Martin D. Ryan2 and Richard D. Iggo1 Correspondence

1

Richard D. Iggo [email protected]

2

Received 14 September 2007 Accepted 11 October 2007

School of Medicine, Biomolecular Sciences Building, University of St Andrews, St Andrews KY16 9ST, UK School of Biology, Biomolecular Sciences Building, University of St Andrews, St Andrews KY16 9ST, UK

Insertion of picornaviral 2A sequences into mRNAs causes ribosomes to skip formation of a peptide bond at the junction of the 2A and downstream sequences, leading to the production of two proteins from a single open reading frame. Adenoviral protein IX is a minor capsid protein that has been used to display foreign peptides on the surface of the capsid. We have used 2A sequences from the foot-and-mouth disease virus (FMDV) and porcine teschovirus 1 (PTV-1) to express protein IX (pIX) and green fluorescent protein (GFP) from pIX–2A–GFP fusion genes in an oncolytic virus derived from human adenovirus 5. GFP was efficiently expressed by constructs containing either 2A sequence. Peptide bond skipping was more efficient with the 58 aa FMDV sequence than with the 22 aa PTV-1 2A sequence, but the virus with the FMDV 2A sequence showed a reduction in plaque size, cytopathic effect, viral burst size and capsid stability. We conclude that ribosome skipping induced by 2A sequences is an effective strategy to express heterologous genes in adenoviruses; however, careful selection or optimization of the 2A sequence may be required if protein IX is used as the fusion partner.

INTRODUCTION Replication-competent oncolytic adenoviruses have been extensively tested in animals, and a virus lacking the E1B 55K gene was recently approved for cancer therapy in China (reviewed by Alemany, 2007). The main factor limiting widespread application of oncolytic adenoviruses is their low efficacy. Many groups have attempted to develop more active viruses, for example by inserting genes for toxic proteins or prodrug-activating enzymes (reviewed by Alemany, 2007). We and others have previously used internal ribosome entry sites (IRESs) to initiate the translation of prodrug-activating enzymes in adenoviruses but the results were disappointing (reviewed by de Felipe, 2002; Fuerer & Iggo, 2004; Lukashev et al., 2005). Ribosome skipping is a recently described mechanism that allows translation of multiple proteins from a single mRNA. It is based on the use of a picornaviral 2A sequence that causes the ribosome to continue translation after skipping the formation of one peptide bond. This process was first characterized in foot-and-mouth disease virus (FMDV) (Ryan & Drew, 1994). The process was originally termed ‘cleavage’ by analogy with the proteasemediated cleavages occurring at other sites in the FMDV polyprotein but this is misleading because the ‘cleavage’ results from failure to form a peptide bond during 0008-3444 G 2008 SGM

Printed in Great Britain

translation. Unlike reinitiation of translation by IRESs, skipping induced by 2A sequences gives approximately equal expression of the proteins upstream and downstream of the 2A site (de Felipe et al., 2006). 2A and 2A-like sequences have previously been used in biotechnology applications, but not in adenoviruses (reviewed by de Felipe et al., 2006). Protein IX is a small cement protein located between the hexons in the capsid of the adenovirus (Furcinitti et al., 1989). It is essential for the packaging of full-length viral genomes (Ghosh-Choudhury et al., 1987). Besides its role as a structural protein, it is a transcriptional activator (Rosa-Calatrava et al., 2001; Sargent et al., 2004a) and reorganizes promyelocytic leukaemia (PML) nuclear bodies (Rosa-Calatrava et al., 2003) (reviewed by Parks, 2005). Protein IX has been extensively studied as a platform to express foreign peptides on the surface of the capsid (Dmitriev et al., 2002; Le et al., 2004; Li et al., 2005; Meulenbroek et al., 2004; Vellinga et al., 2004, 2007). Since protein IX has been so well characterized, we selected it as a fusion partner to test whether 2A sequences could be used to express foreign genes in adenoviruses. We compared the 2A sequence from FMDV with the 2A sequence from porcine teschovirus-1 (PTV-1) that has previously been used in biotechnology applications (Holst et al., 2006). 389

G. M. Funston and others

METHODS Cell lines. 293T and HT29 cells were supplied by ATCC. C7 cells (Amalfitano & Chamberlain, 1997) were provided by Dr J. Chamberlain (Department of Biochemistry, Box 357350, University of Washington Seattle, USA). HER911 cells (Fallaux et al., 1996) were supplied by Dr P. Beard (ISREC, Ch. des Boveresses 155, CH-1066 Epalinges, Switzerland). All cells were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen) with 10 % fetal bovine serum and 1 % penicillin/streptomycin. Antibodies. Rabbit antibody against the last amino acids of the PTV-

1 2A sequence (Holst et al., 2006) was provided by Dr D. Vignali (Department of Immunology, St. Jude Children’s Research Hospital, 332 N. Lauderdale, Memphis, USA). Rabbit anti-Ad5 hexon and antiprotein IX antibodies were provided by Professor W. Russell (BMS Building, North Haugh, University of St Andrews, St Andrews, Fife, UK). Rabbit anti-protein IX antibody (Caravokyri & Leppard, 1995) was provided by Dr K. Leppard (Department of Biological Sciences, University of Warwick, Coventry, UK) (this antibody was used in the anti-protein IX immunoblots shown in this report). Monoclonal mouse anti-a-tubulin antibody (clone B-5-1-1) was supplied by Sigma. Mouse anti-green fluorescent protein (GFP) antibody was supplied by Roche. Mouse anti-E1A antibody (M58) was supplied by BD Biosciences. Peroxidase-conjugated AffiniPure goat anti-mouse IgG (H+L; 115-035-003) and anti-rabbit IgG (H+L; 111-035-003) were supplied by Jackson ImmunoResearch. Viruses. The pIX–2A–GFP viruses are called vKM11 (FMDV 2A,

‘F2A’) and vKM31 (PTV-1 2A, ‘P2A’). The 2A sequences were inserted into the genome of vKH6 (Homicsko et al., 2005) by twostep gene replacement in yeast (Gagnebin et al., 1999). pPDF2 (unpublished) is a pcDNA3.1(+) (Invitrogen) derived vector with a hybrid cytomegalovirus (CMV)/T7 promoter and deletion of the neo cassette. pRS406 is a yeast integrating vector (Sikorski & Hieter, 1989). Overlapping protein IX-IVa2 fragments of Ad5 genomic DNA (ATCC VR5) were amplified by PCR and cloned into pPDF2 (protein IX region: primers oPF3 59-GCCGCCGCTAGCATGAGCACCAACTCGTTTGA-39 and oPF4 59-GGTACCCCATCATTATGGACGAATGCATGGAAA-39) and pRS406 (IVa2 region: primers oPF7 59-ATGCATGGATCCATAATGATGGCAATGGGCC-39 and oPF8 59-GCCACGGGTACCAGGGGCTGGACTATGACAC-39) to give pPF2 and pPF3, respectively. The PTV-1 2A site was inserted into pPF2 by inverse PCR with primers oPF5 (59-CCACGTCTCCTGCTTGCTTTAACAGAGAGAAGTTCGTGGCTCCGGA-39 and 59-CCCTCTAGAAACCGCATTGGGAGGGGAGGAAGCC-39) and oPF6 (59-TCCGTCGACGCGGCCGCGAATTCCAATGCGGTTTAAAACATAAATA-39). pPDF16 is an unpublished derivative of pL-P-D1D2AG from which a run of nine thymidines between FMDV 2A and GFP was deleted (de Felipe & Izquierdo, 2000, 2003). An FMDV 2A-GFP cassette was cloned from pPDF16 into pPF4 on an XbaI–NotI fragment to give pPF5. The remaining steps in the construction of the gene replacement vectors (pPF13 for F2A, pPF14 for P2A) are shown in Fig. 1b. pPF13 and pPF14 were linearized with SacII for insertion into vpKH6 (Homicsko et al., 2005). The resulting plasmids containing the modified, full-length viral genomes are called vpKM1 (F2A) and vpKM3 (P2A). They were cut with PacI to liberate the viral DNA and then transfected into C7 cells to produce virus. After plaque purification on SW480 cells the viruses were called vKM11 (F2A) and vKM31 (P2A). They were expanded on SW480 cells, purified by two rounds of CsCl2 equilibrium gradient centrifugation, buffer exchanged using HR400 columns (GE Healthcare) into 1 M NaCl, 100 mM Tris/HCl pH 8.0, 10 % glycerol and stored at 270 uC. The p.f.u. titre was determined on HER911 cells. Coupled transcription/translation in vitro. Coupled transcription/

translation reactions were performed in rabbit reticulocyte lysates 390

(TNT T7 system; Promega). Translation reactions performed in a total volume of 7 ml were programmed with 0.1 mg plasmid DNA and incubated in the presence of [35S]methionine [5 mCi (185 kBq); GE Healthcare] at 30 uC for 90 min. Reactions were analysed by autoradiography of SDS-PAGE gels. Immunoblotting. To test the protein IX expression cassette, 293T cells

were transfected with pPF2, pPF5 and pPF6, and harvested 24 h later. To test protein expression from viruses, SW480 cells were infected with vKH6, vKM11 and vKM31 at an m.o.i. of 0.5 p.f.u. per cell, the medium was changed after 4 h and the cells were harvested 12 h later. Nitrocellulose membranes were probed with primary antibodies against E1A, hexon, protein IX, 2A, GFP and tubulin, followed by secondary antibodies coupled to horseradish peroxidase and developed by using enhanced chemiluminescence (GE Healthcare). Measurement of plaque size. HER911 cells were infected at an

m.o.i. of 20 p.f.u. per well in six-well plates and covered with agar. The wells were stained with propidium iodide and pictures were taken 10 days after infection. The size of the plaques was measured using ImageJ (NIH, Bethesda, MD). A minimum of 22 plaques were measured per virus. Error bars are SEM. Photographs of individual plaques were taken with a Nikon Coolpix 990 camera, using mCherry filters (49008; Chroma) for propidium-stained cells and GFP filters (49002; Chroma) for viral GFP-expressing cells, on an Olympus CKX41 microscope. Burst assay. SW480 cells were infected at an m.o.i. of 0.1 p.f.u. per cell,

and harvested after 48 h. Virus was released from the cells by three rounds of freeze–thawing and titred by p.f.u. assay on HER911 cells. The burst size is expressed as output p.f.u. per input p.f.u. Cytopathic effect (CPE) assay. SW480 cells were infected at an

m.o.i. ranging from 0.0002 to 0.2 p.f.u. per cell. The medium was changed 4 h after infection, and cells were stained with crystal violet after 7 days. Heat stability assay. The viruses were heated to 45 uC for 0, 4, 8 or

12 min in storage buffer. The titre of infectious virus was then measured by plaque assay on HER911 cells. The wells were stained with propidium iodide and the plaques were counted.

RESULTS To test whether picornaviral 2A sequences can be used to express foreign genes in adenoviruses, 2A skipping sites were inserted after the protein IX gene in an oncolytic virus that targets colon cancer cells (Fig. 1a). GFP was cloned after the 2A sites to permit easy visualization of infected cells. Two different 2A sequences were tested: a 58 aa sequence that includes the FMDV 2A sequence itself (F2A) preceded by 39 aa from the FMDV 1D protein (Donnelly et al., 2001); and a 22 aa sequence that includes the PTV-1 2A-like sequence (P2A) preceded by a 3 aa spacer that improves cleavage (Holst et al., 2006). There was no spacer between protein IX and the picornaviral sequence in the FMDV construct. Expression of protein IX and GFP from pIX–2A–GFP fusion genes To test the pIX–2A–GFP cassettes in vitro before construction of recombinant viruses, they were cloned downstream Journal of General Virology 89

Adenoviral pIX–2A expression vectors

Fig. 1. (a) Schematic diagram showing the genome organization of the pIX–F2A–GFP and pIX–P2A–GFP viruses (vKM11 and vKM31). The viruses contain Tcf sites in the inverted terminal repeats and the E1B promoter. There is an RGD motif in the fiber HI loop. The protein IX region has been mutated to contain pIX–2A–GFP fusion genes with the indicated 2A sequences. The 2A skipping site is marked with an arrow. (b) Schematic diagram showing the cloning strategy. The final plasmids shown, pPF13 and pPF14, are yeast integrating vectors that were used to insert the pIX–2A–GFP cassettes into plasmids containing the entire viral genome (yeast artificial chromosome/bacterial artificial chromosomes, Gagnebin et al., 1999).

of a hybrid CMV/T7 promoter in a plasmid expression vector (pPF5 and 6, Fig. 1b). Skipping at the arrows shown in Fig. 1(a) should lead to the formation of protein IX with the 2A site at the C terminus and GFP with a single additional proline at the N terminus. In vitro transcription/ translation in the presence of 35S-labelled methionine was http://vir.sgmjournals.org

used to study in vitro skipping at the 2A sites. Both 2A constructs showed efficient skipping to yield GFP and protein IX isoforms of the expected size (Fig. 2a; the wildtype protein IX, pIX–P2A and pIX–F2A proteins are predicted to be 14, 17.5 and 21 kDa, respectively). Fig. 2(a) also shows protein IX expressed from a construct containing 391


just pIX–P2A without a downstream open reading frame (pPF4, Fig. 1b). The pIX–P2A protein expressed from this construct is the same size as the pIX–P2A protein expressed from the pIX–P2A–GFP construct. Small amounts of pIX– 2A–GFP fusion proteins can also be seen (Fig. 2a). We conclude that ribosomal skipping occurs at the 2A site in the absence of exogenous viral proteins yielding proteins of the expected size, as described previously (Donnelly et al., 2001; Holst et al., 2006). To test whether skipping occurs in vivo, 293T cells were transfected with the fusion constructs, and cell lysates were immunoblotted for protein IX, 2A and GFP. There was a prominent GFP band after Coomassie blue staining of the gel or Ponceau S staining of the membrane (data not shown), indicating that this is an efficient expression strategy. Immunoblotting for 2A confirmed the presence of the 2A sequence in the pIX–2A proteins (Fig. 2b; since the antibody prefers the P2A epitope, the relative abundance of

the two forms cannot be assessed from this blot). Skipping at the P2A site was less efficient than at the F2A site, as shown by the presence of more pIX–2A–GFP fusion protein, but in both cases the amount of GFP was far greater than that of the fusion proteins (Fig. 2b). We conclude that the pIX–2A expression strategy gives good expression of GFP. To test skipping in the context of the virus, the pIX–2A– GFP cassettes were cloned into an oncolytic viral genome by two-step gene replacement in yeast (Gagnebin et al., 1999). The parental virus, vKH6, has an RGD peptide in the HI loop of the fiber protein and Wnt-responsive (Tcf/ LEF) transcription factor-binding sites in the E1A, E1B and E4 promoters (Homicsko et al., 2005). The viruses were produced in C7 cells, plaque purified on SW480 colon cancer cells to reduce the risk of unwanted recombination events, expanded on SW480 cells and purified by two rounds of equilibrium density-gradient centrifugation.

Fig. 2. (a) In vitro transcription/translation of protein IX, pIX–P2A, pIX–P2A–GFP and pIX– F2A–GFP proteins. Transcription/translation reactions were programmed with plasmids pPF5 (F2A) and pPF6 (P2A, see Fig. 1b). (b) Expression of protein IX, pIX–P2A–GFP and pIX–F2A–GFP genes after transfection of 293T cells with plasmids expressing protein IX (pPF2), pIX–P2A–GFP (pPF6) and plX– F2A–GFP (pPF5) from the CMV promoter. Western blots were probed with anti-protein IX, anti-2A and anti–GFP antibodies as indicated. Tubulin was used as a loading control. Note that the anti-2A antibody has a higher affinity for the P2A than the F2A epitope. (c) Western blots showing the presence of pIX–2A proteins in virions. Blots of purified virus were probed with anti-protein IX and anti-2A antibodies. Ad5 hexon was used as a loading control. pIX–F2A degradation products are marked with an asterisk. (d) Expression of protein IX and GFP after infection of SW480 cells with vKM11 (pIX– F2A–GFP), vKM31 (pIX–P2A–GFP) and vKH6 (parental virus). Cells were harvested 16 h after infection. Western blots were probed with anti-protein IX, anti-2A, anti-GFP and anti-E1A antibodies. Tubulin was used as a loading control. 392

Journal of General Virology 89

Adenoviral pIX–2A expression vectors Fig. 3. Measurement of plaque size. (a) Photographs of plaques 10 days after infection of HER911 cells. Left panels, mCherry filters were used to detect propidium iodide staining of dead cells. Right panels, GFP filters were used to detect GFP expression by vKM11 and vKM31. (b) Area of propidium-stained plaques. At least 22 plaques were counted for each virus. Error bars, SEM. P values: vKH6 versus vKM11, 0.002; vKH6 versus vKM31, 0.98; vKM11 versus vKM31, 0.05. (c) CPE assay. SW480 cells were infected at the indicated m.o.i. (p.f.u. per cell) and stained with crystal violet after 7 days. (d) Viral burst assay. SW480 cells were infected at an m.o.i. of 0.1 p.f.u. per cell and harvested 48 h later. The burst size is expressed as p.f.u. per input p.f.u. Error bars, SEM. (e) Heat stability assay. Pure virus was heated to 45 6C for the indicated times then titred on HER911 cells.

ratio of protein IX to hexon was similar for the parental virus, vKH6, and the two progeny viruses, vKM11 (pIX– F2A–GFP) and vKM31 (pIX–P2A–GFP). The pIX–F2A protein in the virion showed signs of degradation, with the appearance of lower bands (Fig. 2c, asterisk). Only trace amounts of the pIX–2A–GFP proteins were present, suggesting that the viruses may preferentially incorporate pIX–2A protein into the capsid. The ability of the viruses to express the pIX–2A and GFP proteins correctly was tested by infecting SW480 cells at an m.o.i. of 0.5 p.f.u. per cell. Cells were harvested 16 h after infection to avoid losing cells that detached from the plate. Immunoblotting for E1A showed that the early steps of infection were not affected by the presence of the pIX–2A proteins in the capsid (Fig. 2d). There was more pIX–2A– GFP protein present after infection with the P2A virus, again suggesting that skipping is less efficient with the P2A sequence. The level of the pIX–F2A protein was substantially higher than that of either the wild-type or P2A proteins (Fig. 2d). The same trend was visible after transfection of the CMV expression vectors (Fig. 2b). In both cases the ratio of GFP to pIX–P2A–GFP was much higher than the ratio of protein IX to pIX–P2A–GFP, which was close to one. The reason is unclear, since all three proteins initiate from the same ATG and have long half-lives. Despite these differences, both 2A viruses expressed GFP in equal amounts. We conclude that, consistent with findings in other biological systems, 2A sequences allow efficient expression of transgenes in adenoviruses. pIX–F2A capsids are unstable

Immunoblotting of the purified virions for 2A (Fig. 2c) confirmed the presence of the 2A epitope in the pIX–F2A and pIX–P2A proteins of the vKM11 and vKM31 viruses, respectively; the weaker F2A signal is expected given the higher affinity of the antibody for the P2A epitope. The http://vir.sgmjournals.org

The pIX–F2A virus was more difficult to produce than the pIX–P2A virus or the parental virus. This suggests that the pIX–F2A protein may be partially defective. Visual inspection showed that pIX–P2A plaques were generally larger than pIX–F2A plaques (Fig. 3a). To pursue this we measured plaque size by staining dead cells with propidium iodide and measuring the surface area of plaques photographed through mCherry filters (Fig. 3b). This showed that the pIX–F2A virus forms significantly smaller plaques than the other viruses (Fig. 3b). To test whether the 393


pIX–F2A virus is less cytopathic than the other viruses, SW480 cells were infected with log dilutions of virus. SW480 cells are highly permissive for the parental virus, vKH6, because they have high Tcf activity, leading to strong activation of Tcf-regulated promoters. The CPE of the pIX–P2A virus (vKM31) was similar to that of the parental virus, whereas that of the pIX–F2A virus (vKM11) was reduced 10-fold (Fig. 3c). To identify the reason for the reduction in activity of the pIX–F2A virus, a burst assay was performed. SW480 cells were infected at an m.o.i. of 0.1 p.f.u. per cell and virus was harvested after 48 h. The burst size of the pIX–P2A virus was twofold lower than that of the parental virus, whereas that of the pIX–F2A virus was reduced 90-fold (Fig. 3d). The CPE assay was performed in conditions where several cycles of infection were required, so the difference in burst size is a potential explanation for the reduction in CPE. Viruses defective in protein IX function rapidly lose activity at mildly elevated temperatures. To test whether this might have contributed to the reduction in CPE of the pIX–F2A virus, aliquots of virus were heated to 45 uC and plaque assays were performed to measure the decline in titre. This showed a marginal difference in stability of the pIX–P2A virus compared with the parental virus, and a large reduction in stability of the pIX–F2A virus (Fig. 3e). We conclude that addition of the FMDV 2A sequence to the C terminus of protein IX interferes with protein IX function and this leads to a reduction in CPE, burst size and stability of the capsid.

DISCUSSION We have shown that picornaviral 2A sequences can be used to express transgenes in oncolytic adenoviruses. Both 2A sequences tested gave efficient GFP expression, but there were important differences in the activity of the viruses. The fusion partner, protein IX, has been extensively investigated as a site for expression of foreign peptides on the capsid of adenoviruses (Vellinga et al., 2005b). Indeed, entire scFv peptides have been successfully expressed on the capsid as protein IX fusions (Vellinga et al., 2007). It was therefore unexpected that the addition of a 57 aa sequence derived from FMDV to protein IX should attenuate the virus. Several results point to a loss of activity of the pIX–F2A protein. The viral burst size and CPE activity were both reduced at least 10-fold relative to the parental virus. Protein IX acts at multiple stages in the viral life cycle, including transactivation of early promoters and stabilization of the capsid (Parks, 2005). The defect could be a consequence of defective transactivation, but loss of transactivation by protein IX is thought to have only a small effect on the virus compared with the dramatic destabilization of capsids lacking protein IX. The pIX–F2A virus was temperature sensitive, so a capsid defect is a likely explanation for the other defects: the CPE assay and plaque size assay require prolonged incubation of infected cells at 37 uC, during which time an unstable virus could lose activity. The same could be true of the burst assay, which 394

was harvested after 48 h, but the magnitude of the reduction was so great that other effects of protein IX should be considered. The phenotype may be similar to that of a recently reported protein IX deficient virus with a severe defect in viral burst size that could not be solely explained by a defect in capsid stability (Sargent et al., 2004b). The FMDV 2A sequence used was longer than the PTV-1 sequence and included part of the FMDV 1D capsid protein. It is possible that the 1D sequence imposed a quaternary structure on protein IX incompatible with its normal functioning or that the 1D sequence binds to the protein IX N terminus, perturbing the interaction with hexon. Despite the effect of the FMDV peptide on protein IX function, peptide bond skipping at the F2A site was highly efficient, with only trace amounts of pIX–F2A–GFP protein produced. The PTV-1 2A-like sequence induced less efficient skipping, as shown by the presence of larger amounts of the pIX–P2A–GFP protein. This may not be important, given the apparently equal GFP expression from the two viruses and the fact that the P2A sequence had little, if any, effect on protein IX function. Careful examination of the relative expression levels of the proteins produced by the two protein IX expression cassettes reveals a conundrum. Since translation of both cassettes initiates at the protein IX ATG and the amount of the downstream protein (GFP) is similar, one would expect to see similar amounts of protein IX in both cases. This is clearly not the case. The pIX–F2A protein is present in larger amounts than the pIX–P2A protein, whether it is expressed from the virus or from the CMV promoter in a transfected plasmid. The ratio of pIX–P2A–GFP to pIX–F2A–GFP is the same with anti-protein IX antibody and anti-GFP antibody, so we can rule out differences in antibody affinity as an explanation for the greater amount of the pIX–F2A protein. Despite the lower skipping activity of the P2A sequence, the pIX–2A–GFP to GFP ratio strongly favours GFP in both cases, whereas the pIX–P2A–GFP to pIX–P2A ratio is almost one. In other words, there is much less pIX– P2A than expected. Compared to the parental virus, however, the pIX–P2A level is normal and it is the pIX– F2A protein that is too abundant. These differences cannot be explained by a difference in the half-life of the proteins (data not shown), or an idiosyncrasy of the anti-protein IX antibody, since they are also seen with a different antiprotein IX serum (data not shown). Partial explanations might include premature activation of the protein IX promoter in the pIX–F2A virus or, conceivably, sequestration of the pIX–P2A protein in a form that escapes Western blotting. Despite the strong evidence that protein IX tolerates the addition of many different heterologous peptides to its C terminus (Vellinga et al., 2006, 2007), addition of the F epitope tag has been shown to prevent accumulation of protein IX in nuclear inclusion bodies (Rosa-Calatrava et al., 2001). It is possible that the FMDV 2A site has a similar effect and that this contributes to the increase in pIX–F2A protein level. Journal of General Virology 89

Adenoviral pIX–2A expression vectors

Based on these results, future development of this approach, if protein IX is used as the fusion partner, should be based on the PTV-1 sequence. The use of protein IX as a platform for modification of the adenoviral capsid has been extensively studied in the context of pseudotyped replication-defective vectors (Vellinga et al., 2004, 2005a, b, 2006, 2007). Interference with protein IX function did not emerge as a major issue in those studies, perhaps in part because replication-defective vectors are less sensitive to mild defects in viral functions. One important conclusion from that work was that the addition of spacers to lift peptides above the surrounding hexons greatly improves the exposure of epitopes added to the protein IX C terminus. It may be that addition of a long spacer between protein IX and F2A would prevent interference of F2A with protein IX function. We chose not to do this because the goal was to develop a transgene expression strategy that would minimally enlarge the viral genome. There are many other potential fusion partners in an adenovirus, so there should be no difficulty using the 2A sequences to express transgenes in adenoviruses if protein IX proves unsatisfactory.

mouth disease virus 2A site-directed mutants and naturally occurring ‘2A-like’ sequences. J Gen Virol 82, 1027–1041. Fallaux, F. J., Kranenburg, O., Cramer, S. J., Houweling, A., Van Ormondt, H., Hoeben, R. C. & Van Der Eb, A. J. (1996).

Characterization of 911: a new helper cell line for the titration and propagation of early region 1-deleted adenoviral vectors. Hum Gene Ther 7, 215–222. Fuerer, C. & Iggo, R. (2004). 5-Fluorocytosine increases the toxicity of

Wnt-targeting replicating adenoviruses that express cytosine deaminase as a late gene. Gene Ther 11, 142–151. Furcinitti, P. S., van Oostrum, J. & Burnett, R. M. (1989). Adenovirus

polypeptide IX revealed as capsid cement by difference images from electron microscopy and crystallography. EMBO J 8, 3563–3570. Gagnebin, J., Brunori, M., Otter, M., Juillerat-Jeanneret, L., Monnier, P. & Iggo, R. (1999). A photosensitising adenovirus for photodynamic

therapy. Gene Ther 6, 1742–1750. Ghosh-Choudhury, G., Haj-Ahmad, Y. & Graham, F. L. (1987).

Protein IX, a minor component of the human adenovirus capsid, is essential for the packaging of full length genomes. EMBO J 6, 1733–1739. Holst, J., Vignali, K. M., Burton, A. R. & Vignali, D. A. (2006). Rapid

analysis of T-cell selection in vivo using T cell-receptor retrogenic mice. Nat Methods 3, 191–197. Homicsko, K., Lukashev, A. & Iggo, R. D. (2005). RAD001

ACKNOWLEDGEMENTS

(everolimus) improves the efficacy of replicating adenoviruses that target colon cancer. Cancer Res 65, 6882–6890.

We thank Dr K. Leppard, Dr D. Vignali and Professor W. Russell for antibodies, K. MacKenzie and P. Zarsadias for technical assistance, Professor W. Russell for critical reading of the manuscript and the EU fp6 Theradpox STREP and Wellcome Trust for financial support. This publication reflects only the authors’ views. The European Commission is not liable for any use that may be made of the information herein.

Le, L. P., Everts, M., Dmitriev, I. P., Davydova, J. G., Yamamoto, M. & Curiel, D. T. (2004). Fluorescently labeled adenovirus with pIX-EGFP

REFERENCES Alemany, R. (2007). Cancer selective adenoviruses. Mol Aspects Med

28, 42–58. Amalfitano, A. & Chamberlain, J. S. (1997). Isolation and character-

ization of packaging cell lines that coexpress the adenovirus E1, DNA polymerase, and preterminal proteins: implications for gene therapy. Gene Ther 4, 258–263. Caravokyri, C. & Leppard, K. N. (1995). Constitutive episomal

expression of polypeptide IX (pIX) in a 293-based cell line complements the deficiency of pIX mutant adenovirus type 5. J Virol 69, 6627–6633. de Felipe, P. (2002). Polycistronic viral vectors. Curr Gene Ther 2,

355–378. de Felipe, P. & Izquierdo, M. (2000). Tricistronic and tetracistronic retroviral vectors for gene transfer. Hum Gene Ther 11, 1921–1931. de Felipe, P. & Izquierdo, M. (2003). Construction and characteriza-

tion of pentacistronic retrovirus vectors. J Gen Virol 84, 1281–1285.

for vector detection. Mol Imaging 3, 105–116. Li, J., Le, L., Sibley, D. A., Mathis, J. M. & Curiel, D. T. (2005). Genetic

incorporation of HSV-1 thymidine kinase into the adenovirus protein IX for functional display on the virion. Virology 338, 247–258. Lukashev, A. N., Fuerer, C., Chen, M. J., Searle, P. & Iggo, R. (2005). Late

expression of nitroreductase in an oncolytic adenovirus sensitizes colon cancer cells to the prodrug CB1954. Hum Gene Ther 16, 1473–1483. Meulenbroek, R. A., Sargent, K. L., Lunde, J., Jasmin, B. J. & Parks, R. J. (2004). Use of adenovirus protein IX (pIX) to display large

polypeptides on the virion–generation of fluorescent virus through the incorporation of pIX-GFP. Mol Ther 9, 617–624. Parks, R. J. (2005). Adenovirus protein IX: a new look at an old

protein. Mol Ther 11, 19–25. Rosa-Calatrava, M., Grave, L., Puvion-Dutilleul, F., Chatton, B. & Kedinger, C. (2001). Functional analysis of adenovirus protein IX

identifies domains involved in capsid stability, transcriptional activity, and nuclear reorganization. J Virol 75, 7131–7141. Rosa-Calatrava, M., Puvion-Dutilleul, F., Lutz, P., Dreyer, D., de The, H., Chatton, B. & Kedinger, C. (2003). Adenovirus protein IX sequesters

host-cell promyelocytic leukaemia protein and contributes to efficient viral proliferation. EMBO Rep 4, 969–975. Ryan, M. D. & Drew, J. (1994). Foot-and-mouth disease virus 2A oligopeptide mediated cleavage of an artificial polyprotein. EMBO J 13, 928–933. Sargent, K. L., Meulenbroek, R. A. & Parks, R. J. (2004a). Activation

de Felipe, P., Luke, G. A., Hughes, L. E., Gani, D., Halpin, C. & Ryan, M. D. (2006). E unum pluribus: multiple proteins from a self-

of adenoviral gene expression by protein IX is not required for efficient virus replication. J Virol 78, 5032–5037.

processing polyprotein. Trends Biotechnol 24, 68–75.

Sargent, K. L., Ng, P., Evelegh, C., Graham, F. L. & Parks, R. J. (2004b).

Dmitriev, I. P., Kashentseva, E. A. & Curiel, D. T. (2002). Engineering

of adenovirus vectors containing heterologous peptide sequences in the C terminus of capsid protein IX. J Virol 76, 6893–6899. Donnelly, M. L., Hughes, L. E., Luke, G., Mendoza, H., ten Dam, E., Gani, D. & Ryan, M. D. (2001). The ‘cleavage’ activities of foot-and-

http://vir.sgmjournals.org

Development of a size-restricted pIX-deleted helper virus for amplification of helper-dependent adenovirus vectors. Gene Ther 11, 504–511. Sikorski, R. S. & Hieter, P. (1989). A system of shuttle vectors and

yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122, 19–27. 395


Vellinga, J., Rabelink, M. J., Cramer, S. J., van den Wollenberg, D. J., Van der Meulen, H., Leppard, K. N., Fallaux, F. J. & Hoeben, R. C. (2004). Spacers increase the accessibility of peptide ligands linked to

Vellinga, J., Van der Heijdt, S. & Hoeben, R. C. (2005b). The adenovirus

capsid: major progress in minor proteins. J Gen Virol 86, 1581–1588. Vellinga, J., Uil, T. G., de Vrij, J., Rabelink, M. J., Lindholm, L. & Hoeben, R. C. (2006). A system for efficient generation of adenovirus

the carboxyl terminus of adenovirus minor capsid protein IX. J Virol 78, 3470–3479.

protein IX-producing helper cell lines. J Gene Med 8, 147–154.

Vellinga, J., van den Wollenberg, D. J., van der Heijdt, S., Rabelink, M. J. & Hoeben, R. C. (2005a). The coiled-coil domain of the

Vellinga, J., de Vrij, J., Myhre, S., Uil, T., Martineau, P., Lindholm, L. & Hoeben, R. C. (2007). Efficient incorporation of a functional hyper-

adenovirus type 5 protein IX is dispensable for capsid incorporation and thermostability. J Virol 79, 3206–3210.

stable single-chain antibody fragment protein-IX fusion in the adenovirus capsid. Gene Ther. 14, 664–670.

396

Journal of General Virology 89