Cucumber mosaic virus as a presentation system for a double hepatitis C virus-derived epitope

Arch Virol (2007) 152: 915–928 DOI 10.1007/s00705-006-0916-7 Printed in the Netherlands

Cucumber mosaic virus as a presentation system for a double hepatitis C virus-derived epitope M. Nuzzaci1 , G. Piazzolla2 , A. Vitti1 , M. Lapelosa1 , C. Tortorella2 , I. Stella2 , A. Natilla1 , S. Antonaci2 , and P. Piazzolla1 1 2

Department of Biology, Plant Protection and Agrobiotechnology, University of Basilicata, Potenza, Italy Department of Internal Medicine, Immunology and Infectious Diseases, Section of Internal Medicine, University of Bari, Bari, Italy

Received September 13, 2006; accepted November 30, 2006; published online January 22, 2007 # Springer-Verlag 2007

Summary

Chimeric plant viruses are emerging as promising vectors for use in innovative vaccination strategies. In this context, cucumber mosaic virus (CMV) has proven to be a suitable carrier of the hepatitis C virus (HCV)-derived R9 mimotope. In the present work, a new chimeric CMV, expressing on its surface the HCV-derived R10 mimotope, was produced but lost the insert after the first passage on tobacco. A comparative analysis between R10- and R9-CMV properties indicated that R9-CMV stability was related to structural features typical of the foreign insert. Thus, in order to combine high virus viability with strong immuno-stimulating activity, we doubled R9 copies on each of the 180 coat protein (CP) subunits of CMV. One of the chimeras produced by this approach (2R9-CMV) was shown to systemically infect the host, stably maintaining both inserts. Notably, it was strongly recognized by sera of HCV-infected patients and, as compared with R9-CMV, displayed an enhanced ability to Author’s address: Prof. Pasquale Piazzolla, Department of Biology, Plant Protection and Agrobiotechnology, University of Basilicata, 85100 Potenza, Italy. e-mail: [email protected]

stimulate lymphocyte IFN-g production. The high immunogen levels achievable in plants or fruits infected with 2R9-CMV suggest that this chimeric form of CMV may be useful in the development of oral vaccines against HCV. Introduction Hepatitis C virus (HCV) is an emerging virus which is estimated to infect up to 200 million individuals worldwide, corresponding to more than 3% of the world population [44]. The propensity to establish a persistent infection makes this virus the major etiological agent of chronic hepatitis and hepatocellular carcinoma [23, 35]. Although all attempts to generate a vaccine against HCV proteins have failed, this is still the goal of all researchers engaged in this field. In this context, the production of plantderived vectors is a novel and particularly promising approach to the creation of vaccines. In fact, the expression of potentially immunogenic peptides, either in transgenic plants or on the outer surface of genetically engineered chimeric viruses, could offer remarkable advantages [39]. First of all, several plant components (fruits, leaves, roots) can be eaten, thus providing an easy and inexpensive route of

916

antigen (Ag) administration. Furthermore, Ag delivery by plant cells protects the Ag during passage through the acid environment of the stomach. Finally, plant-derived vaccines eliminate the risk of contamination by animal pathogens such as virus or prion proteins, thereby diminishing the safety concerns associated with the use of many currently available types of vaccines. In the last few years, an increasing number of studies have demonstrated the feasibility of this approach in animals and humans, as vaccinated hosts have been shown to develop mucosal and systemic immune responses to the designed Ag [40]. The successful oral immunization achieved in humans eating transgenic potatoes expressing the hepatitis B surface Ag (HBsAg) [42] confirms the extraordinary potential of this innovative vaccination strategy. The plant virus vector strategy, requiring engineering restricted to the virus genome, currently represents the ideal alternative to the vexing problem of plant genome manipulation, which is otherwise necessary for the creation of the abovementioned transgenic products. To date, several chimeric viruses, actively replicating in plants and expressing numerous copies of a foreign peptide, have been generated in a form that makes the epitope suitable for presentation to the immune system [6, 16, 19, 22]. In particular, both elongated and isodiametric plant viruses such as tobacco mosaic virus (TMV) [26, 43], cowpea mosaic virus (CPMV) [27, 31], alfalfa mosaic virus (AMV) [9, 46], tomato bushy stunt virus (TBSV) [15], potato virus X (PVX) [21], zucchini yellow mosaic virus (ZYMV) [2], plum pox virus (PPV) [10], and cucumber mosaic virus (CMV) [24, 25, 30, 47] have been successfully developed as epitope presentation systems. Our group has obtained interesting results using CMV, an isodiametric plant virus about 30 nm in diameter with a tripartite genome, as a vector for HCV epitopes. This plant virus is present in all tropical, subtropical and temperate regions of the world and is endowed with an extremely wide host range, including many edible plants (banana, cucumber, strawberry, carrot, pepper, tomato, lettuce, celery) [24]. CMV has been engineered to express on its surface the so-called R9 mimotope, a syn-

M. Nuzzaci et al.

thetic peptide derived from more than 200 hypervariable region 1 (HVR1) sequences of the HCV envelope protein E2 [33]. We have demonstrated that this chimeric R9-CMV, actively replicating in several hosts (tobacco, tomato, pepper), is able to elicit in rabbits a humoral response directed toward potentially neutralizing HCVepitopes. Furthermore, in patients with chronic hepatitis C, R9-CMV has been demonstrated to enhance HCV-specific cytotoxic T lymphocyte (CTL) responses playing a critical role in the clearance of intracellular pathogens [30]. The current study was undertaken with the aim of strengthening and improving the immunological properties of this particular Ag-presenting system. We evaluated the ability of CMV to support the expression of a new mimotope (which we herein call ‘‘R10’’), as well as the effect of doubling the number of R9 mimotopes expressed on each of the 180 protein subunits of CMV. Finally, a chimeric CMV (2R9-CMV), successfully expressing a second copy of R9 in a different position of the coat protein (CP) and actively replicating in the infected host, was produced. Interestingly, it was strongly recognized by serum samples isolated from HCVinfected patients and exhibited an enhanced capacity to stimulate interferon (IFN)-g production by peripheral blood mononuclear cells (PBMC) as compared with R9-CMV. These data strongly support the usefulness of chimeric forms of CMV for the development of innovative and promising oral immunization strategies. Materials and methods Prediction of cytotoxic T lymphocyte (CTL) epitopes Epitope prediction analysis was performed with the CTLPred program, a direct method for predicting CTL epitopes from an antigenic sequence, based on the consensus and combined use of two machine learning techniques, i.e., Artificial Neural Network (ANN) and Support Vector Machine (SVM) [5]. ANN scores were in the range 0=þ1; SVM scores in the range
1.5=þ1.5. Peptide synthesis, purification and preparation of polyclonal antisera The peptides (H2N-QTTVVGGSQSHTVRGLTSLFSPGA SQN-COOH), corresponding to the R9 mimotope sequence

Double expression of an HCV-derived epitope in CMV [33], and (H2N-TTHTTGGSASHQTSRLVSLFSPGAQQNCOOH), corresponding to the mimotope sequence herein defined as ‘‘R10’’ but named 877 by Frasca et al. [12], were synthesized and coupled with BSA according to Natilla et al. [24]. They were used to prepare the corresponding polyclonal antisera in rabbit, both giving a titre of 1:10000, as determined by indirect ELISA. Virus and RNA sources CMV-D, CMV-S, pseudorecombinant CMV-D=S, and chimeric CMVs were propagated in Nicotiana tabacum c.v. Xanthi plants and purified as described by Lot et al. [20]. The production of RNAs and pseudorecombinant CMV-D=S, derived from the RNA3 component of the CMV-S strain carrying the coat protein (CP) gene and the RNA1,2 components of CMV-D strain, were obtained as described by Natilla et al. [24]. Construction of chimeric clones The R10 mimotope nucleotide sequence was inserted in position 529 of the CP gene (nt 529), as described in Natilla et al. (2004), generating the modified plasmid pIVCPEcoRV. It was produced by creating a silent substitution (A414 ! T) in the third position of the codon for Ile138 of the pCMV3S movement protein (MP) to eliminate the EcoRV restriction

917 site in this gene, leaving EcoRV as the unique cloning site in the CP gene (nt 526–531). A pair of complementary oligonucleotides was synthesized, corresponding to the R10 mimotope sequence with a 50 and 30 EcoRV restriction site. The R10 insert was cloned into pIVCPEcoRV to obtain pIVCP-R10. At the same time, in the attempt to double the expression of the R9 mimotope on each chimeric virus particle, two R9 mimotope nucleotide sequences were inserted in positions 248 and 529 and in positions 392 and 529 of the CP gene, producing the modified plasmids pVCPaBamHI=EcoRV and pVICPBamHI=EcoRV, respectively. pVCPaBamHI=EcoRV was produced starting from pIVCPEcoRV, which was mutagenized via two site-directed substitutions (T249 ! A and A252 ! C) into codons for Gly83 and Ser84 of the CP gene (nt 247–252). Two pairs of complementary oligonucleotides, corresponding to the R9 mimotope sequence with a 50 and 30 BamHI or EcoRV restriction site, were synthesized. pVCPaBamHI=EcoRV was used to insert R9 in positions 248 and 529, giving pVCP-2aR9. pVICPBamHI=EcoRV was produced by generating BamHI as the unique restriction site in pIVCPEcoRV via site-directed mutagenesis (nt 391–396, with two substitutions T391 ! G and C392 ! G in the codon for Ser131 ! Gly131). pVICPBamHI=EcoRV was used to insert R9 in positions 392 and 529, giving pVICP-2R9 (Fig. 1). All these plasmids were obtained using the QuickChangeTM Site-Directed Mutagenesis Kit (Stratagene Europe,

Fig. 1. Map of site-directed mutagenesis and construction of the clones

918

M. Nuzzaci et al.

Amsterdam) and the following oligonucleotides: PM1 (50 GCTCCCTGCTTTGATTTCTTTCCAACCTACC-30 ); PM2 (5 0 - CCGCCTGAAATTGAGAAAGGATCCTATTTCGGTA GAAGGTTGTC-30 ); PM3 (50 -CAGTTCGGAAAGTACCTT CAGGATCCGATCTTTCCG-TCGCCA-30 ). Complementary oligonucleotides corresponding to the R9 and R10 mimotope sequences with a 50 and 30 EcoRV or BamHI restriction sites were synthesized. Annealing of these oligonucleotides generated compatible ends for cloning into the BamHI and EcoRV sites of the CP gene (pIVCPEcoRV, pVCPaBamHI=EcoRV and pVICPBamHI=EcoRV). Cloning and cell transformation were performed according to Sambrook and Russel [36]. Then, E. coli XLI-Blue supercompetent cells (Stratagene) were used for the transformation. The resultant plasmids, called pIVCP-R10, pVCP-2aR9 and pVICP-2R9, served as template for in vitro generation of the corresponding CMV-S chimeric RNA3s (IV-R10CMV, V-2aR9-CMV and VI-2R9-CMV). DNA sequencing confirmed the insertion of R10 or R9=R9 mimotope nucleotide sequences into the CP gene open reading frame. The R10 sequence was located in position 529 (R10-CMV), whereas the two R9 sequences were located in positions 529 and 248 (2aR9-CMV) or in 529 and 392 (2R9-CMV), respectively.

The cDNA was obtained by adding 20 pmol of the primer PA1 (50 -CTAAGTCGGGAGCATCCGTGAGATAG30 ), complementary to position 632–657 of the CMV-S CP gene, in each well. The plates were treated for 5 min at 95  C, quickly chilled and briefly vortexed. In the same wells, the volume was increased to 20 ml using a reaction mixture containing 400 units of M-MuLV reverse transcriptase (Life Technologies, Gaithersburg, MD, USA), followed by a 1-h incubation at 42  C. Finally, the entire 20 ml was transferred into centrifuge tubes. Six microlitres of this cDNA template were supplemented with 50 pmol of both the forward and reverse primers. R10-CMV was amplified using the primers PA1 and PA2 (50 -ACCACCCACACCACCGGAG-30 ), homologous to position 1–19 of the R10 mimotope sequence. 2aR9-CMV and 2R9-CMV were amplified with the primers PA1 and PA3 (50 - CAAACTACTGTTGTTGGAGGATCTCAAT-30 ), homologous to position 1–28 of the R9 mimotope sequence. CMV-D=S, as control, was amplified using the reverse primer PA1 and the forward primers PA2 and PA3. Each reaction was carried out in a 100-ml PCR reaction mixture with 2.5 units of DyNazyme TM II (Finnzymes, FIN). The PCR fragments were fractionated on a 1.5% agarose gel and stained with SYBR SafeTM DNA gel stain.

In vitro transcription, reconstitution of chimeric viruses, plant infection, and serial transmission to N. tabacum cv. Xanthi

Western blot analysis

In vitro transcripts of CMV-S genomic RNA3 were obtained using the T7 mMessage mMachine TM Kit (Ambion Europe Ltd., Cambridgeshire, UK), as described by Natilla et al. [24]. Each in vitro CMV-S RNA3 transcript was then supplemented with the other two CMV genomic RNAs (RNA1,2= RNA3 1:2) derived from CMV-D, to obtain the chimeric R10-CMV (nt 529), 2aR9-CMV (nt 248–529), and 2R9CMV (nt 392–529). A final RNA concentration of 1 mg=ml in 50 mM potassium phosphate, at pH 7.0, was used to inoculate N. tabacum cv. Xanthi plants at the four-leaf stage. The tissues were collected from systemically infected leaves seven days post-inoculation, and analyzed by IC-RT-PCR and Western blot analysis, then inoculated in new plants. This procedure was repeated for each passage.

Leaf discs (15 mg) of plants infected with CMV-D=S and chimeric viruses were homogenized in 50 mM Tris-HCl pH 6.8, 100 mM DTT, 30% glycerol, 2% sodium dodecyl sulphate, and 0.1% bromophenol blue (1:20 w=v). Samples were fractionated by SDS-polyacrylamide gel electrophoresis and then electroblotted onto nitrocellulose membranes for 1 h at 100 V. The membranes were blocked with powdered milk and incubated, in different experiments, with R9 or R10 mimotope polyclonal antiserum (diluted 1:2000) and with CMV polyclonal antiserum (Phyto-Diagnostics Bio-Rad, Marnes, France) (diluted 1:2000). Finally, they were treated with goat anti-rabbit alkaline phosphatase-conjugated antibodies (Abs) (Sigma Chemical Co., Milan, Italy) (diluted 1:2000). The reactivity was detected using the Sigma Fast TM kit (Sigma).

Immuno-capture RT-PCR (IC-RT-PCR) analysis

Computer-assisted analysis

ELISA plates were coated with rabbit antiserum to the R9 [24] or R10 mimotope, which was added to each well (200 ml diluted 1:10 in 50 mM carbonate buffer, pH 9.6), followed by 2 h incubation at 37  C. Crude plant extracts, in extraction buffer (PBS, 2% PVP, 0.05% Tween 20) (1:5 w=v), were obtained from Xanthi tobacco leaves infected with CMV D=S or with chimeric CMVs. After incubation, the plates were washed with PBS containing 0.05% Tween 20 (washing buffer), then 200 ml of plant extract was added to each well and kept at 4  C overnight.

Mimotope secondary structure predictions were carried out by using the NNPREDICT program. This program is able to predict the secondary structure type for each residue in an amino acid sequence on the basis of a two-layer, feed-forward neural network [17]. The ribbon models of wild-type and chimeric CPs were obtained by using the Swiss-Pdb Viewer program. The isoelectric points (pIs) of the mimotopes (R9 and R10) and of chimeric CPs were calculated using the Lasergene Software Package, DNASTAR, Inc.

Double expression of an HCV-derived epitope in CMV The electrostatic properties of the CMV protein subunit were calculated with the Swiss-Pdb Viewer, utilizing the Poisson-Boltzmann method [14]. Patients Thirty untreated patients with chronic HCV infection (17 males and 13 females; mean age 51 years, range 20–63 years), admitted to the Department of Internal Medicine of Bari Medical School, were enrolled in the study after giving informed consent. Each patient had had abnormal alanine aminotransferase (ALT) serum levels for at least 6 months before inclusion in the study. Exclusion criteria included alcoholism, use of hepatotoxic drugs, clinical and=or histological evidence of autoimmune hepatitis, inherited metabolic disorders and co-infection with other hepatotropic viruses (i.e., HBV and HDV). HCV infection was assessed in all patients using the Ortho Third-generation HCV Elisa Test System (Ortho Diagnostic System, Raritan, NJ, USA), followed by detection of serum HCV-RNA by nested reverse transcription-polymerase chain reaction (RT-PCR), using primers of the 50 noncoding region of HCV. The HCV genotype was determined by Inno-Lipa HCV II (Innogenetics N.V., Ghent, Belgium), which allows genotyping of the 6 major HCV types and their most common subtypes. By this approach, we found genotype 1b in 19 patients, 2a=2c in 9 patients, and 3a in 2 patients. Percutaneous needle liver biopsy specimens were obtained from all patients and used for histological examination. In 25 patients, histological features of minimal or mild portal=periportal necro-inflammation and absent or minimal fibrosis (grading 1–2=staging 1–2) [37] were found, whereas five specimens showed moderate or severe periportal inflammation (grading 2–3) associated with bridging fibrosis or cirrhosis (staging 3–4). Twenty HCV-negative healthy donors were included in the study as controls.

Reactivity of HCV patient sera to chimeric R9-CMV and 2R9-CMV To compare the antibody reactivity to the wild-type CMVD=S and to CMV chimeric forms showing the ability to actively replicate in the host plant, serum samples were collected from all HCV-infected patients (n ¼ 30) and healthy donors (n ¼ 20) and stored at
80  C until assayed by ELISA. Purified virus preparations of CMV-D=S, R9-CMV or 2R9-CMV, at a final concentration of 10 mg=ml in carbonate buffer (50 mM NaHCO3, pH 9.6), were used to coat ELISA plates overnight at 4  C. The plates were washed with PBS containing 0.05% Tween 20 (washing buffer) and incubated for 2 h at 37  C in 300 ml of blocking buffer (PBS 5% BSA= 0.1% Tween 20) per well. The plates were then emptied and air-dried before being incubated with increasing dilutions of serum samples from patients and controls (prepared by using dilution buffer: PBS, 1% BSA, and 0.1% Tween 20) for 2 h

919 at room temperature. All samples and dilutions were tested in triplicate. After washing, 100 ml=well of goat anti-human IgGg-chain Abs conjugated with alkaline phosphatase (Biosource Int., Camarillo, CA, U.S.A.) (diluted 1:10000 in dilution buffer) was added, and the plates were incubated for 1 h at room temperature. The plates were then washed, and alkaline phosphatase activity was detected by incubation with a solution of p-nitrophenyl phosphate (Sigma Chemical Co., Milan, Italy) (1 mg=ml) in 1 M diethanolamine buffer (containing 0.5 mM MgCl2, adjusted to pH 9.8 with HCl). After stopping the color reaction with NaOH, the plates were read at 410 nm by an automated ELISA reader (DiaSorin, Saluggia, Italy). Cell cultures and cytokine assay PBMCs were isolated from heparinized venous blood of twenty HCVþ patients by Lympholyte (Cedarlane Laboratories, Hornby, Ontario, Canada) density gradient centrifugation. The cell suspensions recovered at the interface were washed and resuspended in RPMI 1640 (Sigma) supplemented with penicillin (200 IU=ml), streptomycin (100 mg=ml), L-glutamine (2 mM) and 10% heat-inactivated fetal calf serum (FCS) (complete medium). The monocyte concentration in PBMC suspensions was approximately 15%, as evaluated by morphological (Giemsa staining) and cytochemical (non-specific esterase) criteria. PBMCs (2 105 cells=well) were incubated for 7 days at 37  C, 5% CO2, in 96-well round-bottom microtiter plates (Costar, Cambridge, MA, U.S.A.) in the presence of medium alone, R9-CMV, or 2R9-CMV (both used at 10 mg=ml). Culture supernatants were harvested after 3 and 7 days of incubation, frozen and stored at
80  C until assayed for interferon (IFN)-g levels by a commercially available ELISA kit (Euroclone, Paignton, Devon, UK; sensitivity: 1.2 pg=ml). All reagents were LPS-free, as assessed by the Limulus amebocyte lysate assay (PBI International, Milan, Italy). Statistical analysis Student’s t-test was used to determine statistical significance. Since cytokine concentration values were not normally distributed among stimulated cells, we performed loge transformation to normalize data and to allow parametric statistical analysis. Values of p< 0.05 were considered statistically significant.

Results Machine learning techniques based prediction of CTL epitopes In the current work, the prediction of CTL epitopes, within a series of HCV-derived antigenic

920

sequences, was performed by highly sensitive and specific machine learning techniques, allowing accurate discrimination between T-cell epitopes and non-epitope MHC binders, as well as between Tcell epitopes and non-binders [5]. A preliminary survey was carried out on numerous consensus sequences (‘‘mimotopes’’), synthetically derived from the HVR-1 of many natural variants of HCV [12, 33]. The best results were obtained when R10 was analysed, as its sequence was predicted to contain five epitopes (in the region corresponding to amino acids 7–20) with very high scores (maximum values of 0.94 and 0.72, as calculated by the ANN and SVM techniques, respectively). Data obtained with the R9 sequence were positive as well: three epitopes were identified in the region 8–20, with maximum ANN and SVM values of 0.93 and 0.49, respectively. Expression and properties of chimeric CMVs Chimeric R10-, 2aR9- and 2R9-CMV were inoculated on Xanthi tobacco plants. IC-RT-PCR analyses of systemically infected leaves showed the presence of fragments of the expected size for all nucleic acids, confirming their encapsidation in the

M. Nuzzaci et al.

corresponding chimeric viruses (R10-CMV: 209 bp; 2aR9-CMV: 209 and 573 bp; 2R9-CMV: 209 and 429 bp) (see Fig. 2A). Seven days after the first plant infection, tissues from systemically infected leaves were used for subsequent re-inoculations. Analysing R10-CMV-infected tissues, we found that the band corresponding to the R10 gene was no longer detectable following the first passage on tobacco (Fig. 2B). At the same time, when two copies of R9 were inserted in positions 529 and 248 of the CMV CP gene, the resulting chimera (2aR9-CMV) was demonstrated to stably maintain the insert only in position 529, whereas it lost the insert in position 248 following the first passage on Xanthi plants (see Fig. 2B). Different results were obtained with 2R9-CMV, as the two expected amplification bands of 209 and 429 bp were detected in systemically infected leaves up to the fifteenth passage (Fig. 2B and C), demonstrating the stability of the R9 insert in both the 529 and 392 positions. Western blot analysis of plants inoculated with CMV-D=S or chimeric viruses confirmed the abovedescribed results (Fig. 3). Following the first passage, it revealed in all infected plants the presence of a 29-kDa protein (Fig. 3A–C), confirming the 3-kDa increase due to the R9 or R10 peptide, or a

Fig. 2. Detection of chimeric viruses in N. tabacum cv. Xanthi by IC-RT-PCR. A, First-passage analysis. M, GeneRulerTM 100 bp DNA Ladder Plus; 1, DNA fragment of 209 bp derived from tissues infected with R9-CMV; 2, DNA fragment of 209 bp derived from R10-CMV-infected tissues; 3, DNA fragments of 209 bp and 573 bp derived from 2aR9-CMV-infected tissues; 4, DNA fragments of 209 bp and 429 bp derived from 2R9-CMV-infected tissues; 5, CMV-D=S; 6, healthy control. B, Second-passage analysis. M, GeneRulerTM 100-bp DNA Ladder Plus; 1, DNA fragment derived from tissues infected with R9-CMV; 2, tissues infected with R10-CMV; 3, DNA fragment of 209 bp derived from 2aR9-CMV-infected tissues; 4, DNA fragments of 209 bp and 429 bp, derived from 2R9-CMV-infected tissues; 5, CMV-D=S; 6, healthy control. C, Serial passages analysis. M, GeneRulerTM 100-bp DNA Ladder Plus; 1, 2, 3, DNA fragments of 209 bp and 429 bp derived from 2R9-CMV-infected tissues, after the fifth, tenth, and fifteenth passage, respectively; 4, CMV-D=S; 5, healthy control

Double expression of an HCV-derived epitope in CMV

921

Fig. 3. Western blot analysis of Xanthi tobacco plants infected with CMV-D=S and chimeric viruses. First-passage analysis. A, Membrane probed with CMV-D=S polyclonal antiserum. B, Membrane probed with R10 mimotope polyclonal antiserum. C, Membrane probed with R9 mimotope polyclonal antiserum. Second-passage analysis. D, Membrane probed with R10 mimotope polyclonal antiserum. E, Membrane probed with R9 mimotope polyclonal antiserum. Fifteenth-passage analysis. F, Membrane probed with R9 mimotope polyclonal antiserum. M, Prestained Protein Marker, Broad Range (New England BioLabs). CP, CMV-D=S; CP1, R9-CMV; CP2, R10-CMV; CP3, 2aR9-CMV; CP4, 2R9-CMV; HC, healthy control

32-kDa protein (Fig. 3A and C), confirming the 6-kDa increase due to the 2 R9 insertions, with respect to the 26-kDa CMV protein. In contrast, Fig. 3D and E show that, at the second passage, R10- and 2aR9-CMV-infected tissues lost the R10 peptide and one of the inserted R9s, respectively. Conversely, 2R9-CMV was found to preserve both inserts after the second passage (Fig. 3E) and continued to do so right up to the fifteenth passage (Fig. 3F). Symptoms induced by 2R9-CMV consisted of the appearance of vein clearing and strong mosaic within five days from inoculation. They remained unaltered up to the 15th passage and, interestingly, were more severe than those induced by R9-CMV. 2R9-CMV particles were purified and quantified by measuring the optical density of the virus suspension at 260 nm wavelength. Virus extraction

yielded an average of 10 mg=100 g of fresh tissue, similar to that achieved with R9-CMV. Computer-assisted analysis To acquire more information on the ability of R9 and=or R10 mimotopes to affect CP stability, we calculated the pIs of the two peptides and of the corresponding chimeric CPs. The total charges of the chimeric CPs were also calculated in order to compare the pI=charge ratio of each chimeric CP with that of the wild-type CMV-D=S (Table 1). We found that the R9 and R10 mimotopes had comparable pI values and that their insertion in CMV-D=S CP did not result in a significant modification of the respective CP pI=charge ratio. Moreover, analysis of the two chimeric forms of CMV bearing a second copy of R9 on each CP subunit showed that

922

M. Nuzzaci et al.

Table 1. Effect of the insertion of R9 and R10 mimotopes on CMV CP isoelectric point (pI) and total charge (modelled at pH 7) CP

Peptide

Peptide pI

CP pI

CP total charge

PI=charge ratio

R9-CP R10-CP 2R9-CP 2aR9-CP CP wt

QTTVVGGSQSHTVRGLTSMFSPGASQN TTHTTGGSASHQTSRLVSLFSPGAQQN QTTVVGGSQSHTVRGLTSMFSPGASQN

10.05 9.96 10.05

9.79 9.79 9.87

þ10.55 þ10.72 þ11.72

0.93 0.91 0.84

–

–

9.79

þ10.39

0.94

Fig. 4. Ribbon models showing wild-type CP (A), chimeric CP with R9 located in bC region (B), in bE-aEF region (C), in bG-aGH region (D). Chimeric CP with R10 located in bG-aGH region (E). a-helices and b-strands are coloured red and yellow, respectively. The epitope is displayed in grey

Double expression of an HCV-derived epitope in CMV

they were characterized by a lower pI=charge ratio value than CMV-D=S. To investigate the structural factors typical of the foreign insert and the corresponding chimeric CPs potentially accounting for the different stability of R9- and R10-CMV during serial passages on tobacco, a prediction of the protein secondary structure type was made for each residue in the respective mimotope amino acid sequences. By this approach, the two peptides were shown to display different structural features, since higher amounts of b strand elements were observed in the R9 as compared with the R10 mimotope. The ribbon models corresponding to wild-type and chimeric CPs are displayed in Fig. 4. The electrostatic potential of the CMV CP protein subunit was also evaluated. Its representation is shown in Fig. 5, where regions with a potential less than
1.8 kT are coloured red and those with a potential greater than þ1.8 kT are coloured blue (k: Boltzmann’s constant; T: absolute temperature). We found that the position between Asp176 and

923

Ile177 in the bG-aGH region, where the first R9 had been inserted, and the position between Ser131 and 132 in the bE-aEF, where the so-called 2R9CMV has the second R9, were both characterized by positive domains. In contrast, the position between Gly83 and Ser84 in the bC region, where the so-called 2aR9-CMV has the second R9, was characterized by the absence of charge. Reactivity of HCV patient sera to chimeric R9-CMV and 2R9-CMV Natural anti-HVR1 antibodies of patients infected with different HCV genotypes have been demonstrated to be more strongly recognized by the R9 mimotope expressed on the surface of CMV CP with respect to the mimotope in its soluble form [24]. In order to evaluate whether the effectiveness of the Ag recognition was influenced by doubling the number of copies of R9 on the surface of each chimeric particle of CMV, we made a comparative evaluation of the reactivity of sera from thirty HCVþ patients against the two forms of chimeric CMV. Accordingly, ELISA tests were performed by coating multi-well plates with purified chimeric particles of R9- or 2R9-CMV; wells coated with purified particles of CMV-D=S represented our negative controls and served to calculate the nonspecific background signal ascribable to the virus itself. We found that the two chimeras were recognized by serum samples from the same patients, with the exception of one sample that recognized only the 2R9-CMV; in particular, 24 sera (80%) reacted against R9-CMV and 25 (83.3%) against 2R9-CMV (Fig. 6). No reactivity was detected when testing sera from healthy donors. IFN-
release by PBMC from patients with chronic HCV infection

Fig. 5. Electrostatic density map of wild-type CP. i.p. ¼ insertion point. i.p. 1 ¼ Gly83-Ser84. i.p. 2 ¼ Ser131Ser132. i.p. 3 ¼ Asp176-Ile177. B H-I ¼ bH-bI loop. Areas with a potential less than
1.8 kT or greater than þ1.8 kT are coloured red and blue, respectively

Finally, we decided to perform a functional comparison between the two chimeric viruses, R9- and 2R9-CMV, by evaluating their effects on PBMC cytokine production. Taking into account that in vitro R9-CMV stimulation of PBMC from HCVþ and healthy subjects had been demonstrated previously to induce a dominant T helper-1 cytokine

924

Fig. 6. Evaluation of serum sample reactivity (n ¼ 30) to purified chimeric particles of R9- and 2R9-CMV. Results are expressed as the percentage of positive serum samples. Each sample was tested in triplicate and mean O.D. values were calculated. The reaction was considered positive when the mean O.D. value differed by more than 3smax (p < 0.003) from the background signal observed with purified particles of CMV-D=S. HD, Healthy donors

Fig. 7. Individual trends of IFN-g production in R9-CMVand 2R9-CMV-stimulated PBMCs from HCV-infected patients (n ¼ 20). PBMC (2  105 cells=well) were stimulated with R9-CMV or 2R9-CMV, both used at 10 mg=ml. Culture supernatants were harvested after 3 days (T3) and 7 days (T7) of incubation and assayed for IFN-g levels by ELISA. Horizontal bars indicate mean values.  Significance versus R9-CMV: p < 0.05

M. Nuzzaci et al.

profile only in the HCVþ subset [30], we analyzed R9- or 2R9-CMV-dependent IFN-g release from peripheral lymphocytes isolated from twenty patients with chronic hepatitis C. Cell culture supernatants were harvested after 3 and 7 days of incubation. As shown in Fig. 7, 2R9-CMV was found to induce a significantly greater IFN-g release as compared with R9-CMV. Strikingly, this property was evident early, as the difference between cytokine amounts detected in R9- and 2R9CMV-stimulated cell culture supernatants was statistically significant from the 3rd day of cell incubation onward (Fig. 7). Discussion In recent years, the possibility of developing epitope presentation systems in which antigenic sequences are expressed on the surface of non-pathogenic vectors has aroused considerable interest. In this regard, plant virus coat proteins have proven to be suitable carriers for the presentation of immunogenic peptides to the immune system, above all because the chimeric virus particles resulting from their self-assembly in the host plant are not infectious in humans and animals and have the potential for cost-effective manufacture. The identification of CTL epitopes in short peptide sequences to be used in such presentation systems has strategic importance in that it would allow the design of synthetic vaccines against many infectious agents [7]. In particular, immune-based prevention strategies against a persistent and highly variable pathogen like HCV are required to stimulate a strong, persistent CTL response directed against multiple viral epitopes. Recently, synthetic HCV HVR1 mimotopes, incorporating the variability found in a great number of viral variants, have been selected and demonstrated to mimic both Band T helper-cell epitopes [12, 33]. In the current work, in order to identify potential CTL-inducing mimotopes for insertion in a plant-derived vector in a novel vaccine strategy against HCV, a computerassisted prediction of CTL epitopes was made of all the mimotopes selected by Puntoriero et al. [33] and Frasca et al. [12]. By this approach, it was firstly possible to identify, within the R9 mimotope

Double expression of an HCV-derived epitope in CMV

sequence, three CTL epitopes that likely accounted for the previously reported R9 mimotope CTL stimulating activity [30]. The results with the R10 mimotope were equally promising. Its sequence was, in fact, predicted to contain a high number of epitopes (five), associated with the highest ANN and SVM scores. Based on the previous results obtained using CMV as the carrier of the R9 peptide, it seemed interesting to verify the characteristics of chimeric CMV particles carrying this new R10 mimotope. Furthermore, the evidence that the efficiency of short peptides as immunogenic molecules can be improved by presenting them as multiple peptides on the surface of a large carrier molecule [11] prompted us to double the expression of R9 on each chimeric viral particle. For the construction of the chimeric R10-CMV, insertion of the foreign sequence in the bH-bI loop of the CP subunits was carefully avoided because several CMV structural data [38, 45] converge advising against its use as an insertion area. The bHbI loop, negatively charged due to the presence of acidic residues, forms a phosphorylation site [13] and is remarkably conserved among strains of CMV as well as in other cucumoviruses [38], suggesting the importance of keeping this area unchanged. Moreover, it is reported to play an essential role in virus aphid transmission [29]. The recently reported deletions of one to several insert amino acids occurring when epitopes derived from Newcastle disease virus were expressed in the bH-bI loop of a chimeric CMV Ixora strain CP [47] seem to provide further support of our hypothesis. Hence, R10CMV was prepared by insertion of the mimotope in the same position previously chosen for R9 insertion, between Asp176 and Ile177, i.e., in the bG-aGH region. This area is less structured than the a-helix and b-sheet regions between which it is located [38], and such a feature is likely to provide virus stability following the insertion of foreign epitopes. However, the resultant chimeric R10-CMV was found to be stable only for a single passage on tobacco, and then the insert was lost. This behaviour recalled the problems encountered with chimeras containing inserts of considerable size (approximately 30 residues) [32], suggesting that we should verify the impact of R10 pI on virion stability and

925

compare this effect with the one induced by the R9 mimotope, demonstrated to be stably tolerated by CMV. The importance of the pI of the insert and of its effects on the hybrid CP pI=charge ratio value in affecting successful epitope expression as well as virus assembly and infectivity has already been pointed out in the case of TMV and CPMV [4, 32]. Here, interference with the normal virus infection cycle, often occurring as a result of insertion of a peptide with a high pI value (>8) [32], was averted if the pI=charge of the chimeric CP had a value close to that of the wild-type virus [4]. The evaluation of these ratios in the present study showed that the R9 and R10 mimotopes had similar pIs, both being higher than 8; at the same time, the pI=charge ratio calculated for each of the chimeric CPs was found to be similar to that of the wild-type CMVD=S. These findings indicated that the different stability of R9- and R10-CMV during serial passages on tobacco could not be accounted for by an adverse (CP) pI=charge ratio. The possibility that structural factors typical of the foreign insert might be implicated was thus considered. To address this hypothesis, we decided to perform a prediction of the secondary structure type for each residue in the amino acid sequence of the R9 and R10 mimotopes. Interestingly, this analysis revealed that the R9 sequence was characterized by a larger number of b-structure-inducing amino acids than the R10 mimotope. The importance of this finding lies in the fact that CMV is a spherical virus and that its stability could be seriously affected by any variation of its basic architecture, such as an increased size of the protein subunits that interact with the encapsidated RNA. It might perhaps be speculated that the presence in the R9 peptide of b-structureinducing amino acids facilitates the creation of binding pockets in viral protein subunits. These have been described to bind unpaired RNA bases and to stabilize protein-RNA complexes through conformational rearrangements [8]. On the other hand, the R10 mimotope lacks this peculiar structural feature, and this fact might crucially contribute to its poor stability within R10-CMV. These results prompted us to focus our interest on the attempt to double the number of R9 copies expressed on each CMV CP subunit.

926

When testing the ability of R9-CMV to accept a second R9, continuing to avoid the bH-bI loop for the above-mentioned reasons, we decided to insert the second mimotope in the bC or bE-aEF regions of each subunit of the coat protein. In this way we produced two clones encoding CP subunits that contained the second insert between Gly83 and Ser84 in the bC region (designated 2aR9-CMV) and between Ser131 and Ser132 in the bE-aEF region (2R9-CMV). The pI=charge ratio for both of these chimeras was calculated to be lower than that of the other chimeric forms of CMV. However, their viability appeared to be quite different. In fact, 2aR9-CMV was found to be unstable, as it lost the second insert (R9 in position 248) following the first passage on tobacco. Conversely, 2R9-CMV showed results very similar to those of R9-CMV, showing comparable stability and similar extraction yields. It is worth noting that although 2R9-CMV produced more severe symptoms on N. tabacum Xanthi plants, no reduction in local and=or systemic accumulation of chimeric particles was observed. This is a crucial aspect when plant viruses are thought of as biological systems that are designed to function in medical molecular farming. Usually, the insertion of large inserts on the surface of the virus has an adverse effect on viral assembly, leading to loss of systemic infection and=or induction of a necrotic response in the systemic host [3, 31, 41], all effects that might dramatically limit their use as vaccines. The differences in stability between 2aR9-CMV and 2R9-CMV prompted us to further investigate the charge densities at the selected insertion points. It is well known that the electrostatic potential of proteins plays an important role in protein folding, stability, protein-protein, and protein-nucleic acid interactions. Specifically in the case of CMV, which is stabilized by strong ionic protein-RNA interactions, any change in the electrostatic potentials associated with the inner and outer sides of capsids could result in virus instability. Our data suggest that a situation of charge equivalency in the bC region, which loses the foreign insert after the first passage in tobacco, might contribute to the instability of 2aR9-CMV. In fact, the two positions used for peptide insertion in 2R9-CMV, both character-

M. Nuzzaci et al.

ized by positive domains, were found to be able to tolerate a peptide carrying the same positive charge, as they stably maintained the insert during several passages on plants. Conversely, the position between Gly83-Ser84 in the bC region, where 2aR9-CMV has its second R9, is prevalently neutral and unlikely to be able to tolerate the deviation from neutrality caused by peptide insertion. It should also be stressed that, as illustrated in Fig. 5, due to the folding of protein subunits, the bC region appears to be situated very close to the bH-bI loop. This further supports our choice to avoid this particular area when inserting a foreign peptide, all the more so because the negative potential on the hexameric and pentameric CMV subunits appears to be associated with bH-bI charged amino acid residues, stressing the essential role of this loop in determining the electrostatic nature of the viral surface [18, 28]. Once a new chimeric virus actively replicating in plants had been obtained, the final step was to verify whether this 2R9-CMV preserved, or even increased, its immunological properties, as compared with R9-CMV. In this regard, the comparable seroreactivity displayed by patients with chronic hepatitis C against the two chimeric forms of CMV demonstrated that the presence of a large number of foreign peptides on the carrier surface does not pose any steric hindrance to the antigen=antibody binding reaction. At the same time, the enhanced effectiveness of 2R9-CMV in inducing PBMC IFNg production as compared to R9-CMV, testified to the realistic possibility of strengthening the functional activity of a chimeric plant virus by doubling the epitope copies on each CP subunit surface. This last effect certainly represents a result that might have important clinical implications in the field of oral vaccination. Due to the risk of degradation of proteins in the gastrointestinal tract, the presentation of vaccine antigens by the oral route requires much higher levels of immunogen than by parenteral delivery [1, 34]. Considering that an extension in the length of the epitope would likely be deleterious for virus viability, the importance of increasing the expression level of the antigen on each chimeric particle is evident, as it would enhance the effectiveness of the infected plant or=fruit tissue as a source of the immunogen.

Double expression of an HCV-derived epitope in CMV

Our results demonstrate the ability of CMV to accommodate foreign peptides of significant size without any interference with overall virus functions (infectivity, cell-to-cell and long-distance movement). Taking into account the extreme versatility of CMV as an immunological vector and its wide range of edible hosts, chimeric forms of this virus might constitute ideal vectors for the creation of potentially effective oral vaccines.

927

10.

11. 12.

Acknowledgments

13.

Paper supported by a grant from Ministero dell’Istruzione, dell’Universita e della Ricerca (MIUR), Italy (prot. 2003070899).

14.

References

15.

1. Aizpurua HJD, Russell-Jones GJ (1988) Oral vaccination: identification of classes of proteins that provoke an immune response upon oral feeding. J Exp Med 167: 440–451 2. Arazi T, Huang PL, Huang P, Zhang L, Shiboleth YM, Gal-On A, Lee-Huang S (2002) Production of antiviral and antitumor proteins MAP30 and GAP31 in cucurbits using the plant virus vector ZYMV-AGII. Biochem Biophys Res Commun 292: 441–448 3. Beachy RN, Fitchen JH, Hein MB (1996) Use of plant viruses for delivery of vaccine epitopes. Ann NY Acad Sci 792: 43–49 4. Bendahamane M, Koo M, Karrer E, Beachy RN (1999) Display of epitopes on the surface of tobacco mosaic virus: impact of charge and isoelectric point of the epitope on virus-host interactions. J Mol Biol 290: 9–20 5. Bhasin M, Raghava GPS (2004) Prediction of CTL epitopes using QM, SVM and ANN techniques. Vaccine 22: 3195–3204 6. Brennan FR, Bellaby T, Helliwell SM, Jones TD, Kamstrup S, Dalsgaard K, Flock JI, Hamilton WDO (1999) Chimeric plant virus particles administered nasally or orally induce systemic and mucosal immune responses in mice. J Virol 73(2): 930–938 7. Brunak S, Buus S (2000) Identifying cytotoxic T cell epitopes from genomic and proteomic information: ‘‘The human MHC project’’. Rev Immunogenet 2: 477–491 8. Draper DE (1999) Themes in RNA-protein recognition. J Mol Biol 293: 255–270 9. El Attar AK, Shamloul AM, Shalaby AA, Riad BY, Saad A, Mazyad HM, Keith JM (2004) Expression of chimeric HCV peptide in transgenic tobacco plants infected with recombinant alfalfa mosaic virus for

16.

17.

18.

19.

20.

21.

22.

development of a plant derived vaccine against HCV. Afr J Biotech 3: 588–594 Fernandez-Fernandez MR, Martinez-Torrecuadrada JL, Casal JI, Garcia JA (1998) Development of an antigen presentation system based on plum pox potyvirus. FEBS Lett 427: 229–235 Francis MJ (1990) Peptide vaccines for viral diseases. Sci Progr 74: 115–130 Frasca L, Scotta C, Del Porto P, Nicosia A, Pasquazzi C, Versace I, Masci AM, Racioppi L, Piccolella E (2003) Antibody-selected mimics of hepatitis C virus hypervariable region 1 activate both primary and memory Th lymphocytes. Hepatology 38: 653–662 Gellert A, Sala´nki K, Na´ray-Szab o G, Bala´zs E (2006) Homology modelling and protein structure based functional analysis of five cucumovirus coat proteins. J Mol Graph Model 24: 319–327 Guex N, Peitsch MC (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modelling. Electrophoresis 18: 2714–2723 Joelson T, Akerblom L, Oxelfelt P, Strandberg B, Tomenius K, Morris TJ (1997) Presentation of a foreign peptide on the surface of tomato bushy stunt virus. J Gen Virol 78: 1213–1217 Johnson J, Lin T, Lomonossoff G (1997) Presentation of heterologous peptides on plant viruses: genetics, structure, and function. Annu Rev Phytophathol 35: 67–86 Kneller DG, Cohen FE, Langridge R (1990) Improvements in protein secondary structure prediction by an enhanced neural network. J Mol Biol 214: 171–182 Konecny R, Trylska J, Tama F, Zhang D, Baker NA, Brooks CL, McCammon JA (2005) Electrostatic properties of cowpea chlorotic mottle virus and cucumber mosaic virus capsids. Biopolymers 82: 106–120 Koo M, Bendahmane M, Lettieri GA, Paoletti AD, Lane TE, Fitchen JH, Buchmeier MJ, Beachy RN (1999) Protective immunity against murine hepatitis virus (MHV) induced by intranasal or subcutaneous administration of hybrids of tobacco mosaic virus that carries an MHV epitope. Proc Natl Acad Sci USA 96: 7774–7779 Lot H, Marrou J, Quiot JB, Esvan C (1972) Contribution a l’etude du virus de la mosaique du cocombre (CMV). I. Methode de purification rapide du virus. Ann Phytopathol 14: 25–38 Marusic C, Rizza P, Lattanzi L, Mancini C, Spada M, Belardelli F, Benvenuto E, Capone I (2001) Chimeric plant virus particles as immunogens for inducing murine and human immune responses against human immunodeficiency virus type1. J Virol 75: 8434–8439 McLain L, Durrani Z, Wisniewski LA, Porta C, Lomonossoff GP, Dimmock NJ (1996) Stimulation of neutralizing antibodies to human immunodeficiency virus type 1 in three strains of mice immunized with a 22 amino acid peptide of gp41 expressed on the surface of a plant virus. Vaccine 14: 799–810

928

M. Nuzzaci et al.: Double expression of an HCV-derived epitope in CMV

23. Memon ML, Memon MA (2002) Hepatitis C: an epidemiological review. J Viral Hepatol 9: 84–100 24. Natilla A, Piazzolla G, Nuzzaci M, Saldarelli P, Tortorella C, Antonaci S, Piazzolla P (2004) Cucumber mosaic virus as a carrier of a hepatitis C virus-derived epitope. Arch Virol 149: 137–154 25. Natilla A, Hammond RW, Nemchinov LG (2006) Epitope presentation system based on cucumber mosaic virus coat protein expressed from a potato virus X-based vector. Arch Virol 151(7): 1373–1386 26. Nemchinov LG, Liang TJ, Rifaat MM, Mazyad HM, Hadidi A, Keith JM (2000) Development of a plantderived subunit vaccine candidate against hepatitis C virus. Arch Virol 145: 2557–2573 27. Nicholas BL, Brennan FR, Martinez-Torrecuadrada JL, Casal JI, Hamilton WD, Wakelin D (2002) Characterization of the immune response to canine parvovirus induced by vaccination with chimaeric plant viruses. Vaccine 20: 2727–2734 28. Pacios LF, Garcia-Arenal F (2006) Comparison of properties of particles of cucumber mosaic virus and Tomato aspermy virus based on the analysis of molecular surfaces of capsids. J Gen Virol 87: 2073–2083 29. Perry KL, Zhang L, Shintaku MH, Palukaitis P (1994) Mapping determinants in cucumber mosaic virus for transmission by Aphis gossypii. Virology 205: 591–595 30. Piazzolla G, Nuzzaci M, Tortorella C, Panella E, Natilla A, Boscia D, De Stradis A, Piazzolla P, Antonaci S (2005) Immunogenic properties of a chimeric plant virus expressing a hepatitis C virus (HCV)-derived epitope: new prospects for an HCV vaccine. J Clin Immunal 25: 142–152 31. Porta C, Spall VE, Loveland J, Johnson JE, Barker PJ (1994) Development of cowpea mosaic virus as a highyielding system for the presentation of foreign peptides. Virology 202: 949–955 32. Porta C, Spall VE, Findlay KC, Gergerich RS, Farrance CE, Lomonossoff G (2003) Cowpea mosaic virus-based chimaeras. Effects of inserted peptides on the phenotype, host range, and transmissibility of the modified virus. Virology 310: 50–63 33. Puntoriero G, Meola A, Lahm A, Zucchelli S, BruniErcole B, Tafi R, Pezzanera M, Mondelli MU, Cortese R, Tramontano A, Galfre’ G, Nicosia A (1998) Toward a solution for hepatitis C virus hypervariability: mimotopes of the hypervariable region 1 can induce antibodies cross-reacting with a large number of viral variants. EMBO J 13: 3521–3533

34. Richter LJ, Thanavala Y, Arntzen CJ, Mason HS (2000) Production of hepatitis B surface antigen in transgenic plants for oral immunization. Nat Biotechnol 18: 1167–1171 35. Rosemberg S (2001) Recent advances in the molecular biology of hepatitis C virus. J Mol Biol 313: 451–464 36. Sambrook J, Russell DW (2001) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York 37. Scheuer PJ (1991) Classification of chronic viral hepatitis: a need for reassessment. J Hepatol 13: 372–374 38. Smith TJ, Chase E, Schmidt T, Perry KL (2000) The structure of cucumber mosaic virus and comparison to cowpea chlorotic mottle virus. J Virol 74: 7578–7686 39. Tacket CO, Mason HS (1999) A review of oral vaccination with transgenic vegetables. Microbes Infect 1: 777–783 40. Tacket CO, Sztein MB, Losonsky GA, Wasserman SS, Estes MK (2003) Humoral, mucosal, and cellular immune responses to oral Norwalk virus-like particles in volunteers. Clin Immunol 108: 241–247 41. Takamatsu N, Watanabe Y, Yanagi H, Meshi T, Shiba T, Okada Y (1990) Production of enkephalin in tobacco protoplasts using tobacco mosaic virus RNA vector. FEBS Lett 269: 73–76 42. Thanavala Y, Mahoney M, Pal S, Scott A, Richter L, Natarajan N, Goodwin P, Arntzen CJ, Mason HS (2005) Immunogenicity in humans of an edible vaccine for hepatitis B. Proc Natl Acad Sci USA 102(9): 3378–3382 43. Turpen TH, Reini SJ, Charoenvit Y, Hoffman SL, Fallarme V, Grill LK (1995) Malarial epitopes expressed on the surface of recombinant tobacco mosaic virus. Biotechnology 13: 53–57 44. Wasley A, Alter MJ (2000) Epidemiology of hepatitis C: geographic differences and temporal trends. Semin Liver Dis 20: 1–16 45. Wikoff WR, Tsai CJ, Wang G, Baker TS, Johnson JE (1997) The structure of cucumber mosaic virus: cryoelectron microscopy, X-ray crystallography, and sequence analysis. Virology 232: 91–97 46. Yusibov V, Hooper DC, Spitsin SV, Fleysh N, Kean RB, Mikheeva T, Deka D, Karasev A, Cox S, Randall J, Koprowski H (2002) Expression in plants and immunogenicity of plant virus-based experimental rabies vaccine. Vaccine 20: 3155–3164 47. Zhao Y, Hammond RW (2005) Development of a candidate vaccine for newcastle disease virus by epitope display in the cucumber mosaic virus capsid protein. Biotechnol Lett 27: 375–382