• Open Access

Induction of toxin-specific neutralizing immunity by molecularly uniform rice-based oral cholera toxin B subunit vaccine without plant-associated sugar modification

Authors

  • Yoshikazu Yuki,

    Corresponding author
    1. International Research and Development Center for Mucosal Vaccine, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    • Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Mio Mejima,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Shiho Kurokawa,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Tomoko Hiroiwa,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Yuko Takahashi,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Daisuke Tokuhara,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Tomonori Nochi,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Yuko Katakai,

    1. Corporation for Production and Research of Laboratory Primates, Ibaraki, Japan
    Search for more papers by this author
  • Masaharu Kuroda,

    1. Crop Development Division, NARO Agricultural Research Center, Niigata, Japan
    Search for more papers by this author
  • Natsumi Takeyama,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Koji Kashima,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Michiyo Abe,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    2. MG Pharma Inc., Osaka, Japan
    Search for more papers by this author
  • Yingju Chen,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Ushio Nakanishi,

    1. Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    2. MG Pharma Inc., Osaka, Japan
    Search for more papers by this author
  • Takehiro Masumura,

    1. Laboratory of Genetic Engineering, Graduate School of Life and Environmental Sciences, Kyoto Prefectural University, Kyoto, Japan
    2. Biotechnology Research Department, Kyoto Prefectural Agriculture, Forestry and Fisheries Technology Research Center, Kyoto, Japan
    Search for more papers by this author
  • Yoji Takeuchi,

    1. Hytoculture Control Co. Ltd., Osaka, Japan
    Search for more papers by this author
  • Hiroko Kozuka-Hata,

    1. Medical Proteomics Laboratory, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Hiroaki Shibata,

    1. Tsukuba Primate Research Center, National Institute of Biomedical Innovation, Ibaraki, Japan
    Search for more papers by this author
  • Masaaki Oyama,

    1. Medical Proteomics Laboratory, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Kunisuke Tanaka,

    1. Laboratory of Genetic Engineering, Graduate School of Life and Environmental Sciences, Kyoto Prefectural University, Kyoto, Japan
    2. Hytoculture Control Co. Ltd., Osaka, Japan
    Search for more papers by this author
  • Hiroshi Kiyono

    Corresponding author
    1. International Research and Development Center for Mucosal Vaccine, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    • Division of Mucosal Immunology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author

Correspondence (Tel +81-3-5449-5274; fax +81-3-5449-5411; emails yukiy@ims.u-tokyo.ac.jp and kiyono@ims.u-tokyo.ac.jp)

Summary

Plants have been used as expression systems for a number of vaccines. However, the expression of vaccines in plants sometimes results in unexpected modification of the vaccines by N-terminal blocking and sugar-chain attachment. Although MucoRice-CTB was thought to be the first cold-chain-free and unpurified oral vaccine, the molecular heterogeneity of MucoRice-CTB, together with plant-based sugar modifications of the CTB protein, has made it difficult to assess immunological activity of vaccine and yield from rice seed. Using a T-DNA vector driven by a prolamin promoter and a signal peptide added to an overexpression vaccine cassette, we established MucoRice-CTB/Q as a new generation oral cholera vaccine for humans use. We confirmed that MucoRice-CTB/Q produces a single CTB monomer with an Asn to Gln substitution at the 4th glycosylation position. The complete amino acid sequence of MucoRice-CTB/Q was determined by MS/MS analysis and the exact amount of expressed CTB was determined by SDS-PAGE densitometric analysis to be an average of 2.35 mg of CTB/g of seed. To compare the immunogenicity of MucoRice-CTB/Q, which has no plant-based glycosylation modifications, with that of the original MucoRice-CTB/N, which is modified with a plant N-glycan, we orally immunized mice and macaques with the two preparations. Similar levels of CTB-specific systemic IgG and mucosal IgA antibodies with toxin-neutralizing activity were induced in mice and macaques orally immunized with MucoRice-CTB/Q or MucoRice-CTB/N. These results show that the molecular uniformed MucoRice-CTB/Q vaccine without plant N-glycan has potential as a safe and efficacious oral vaccine candidate for human use.

Introduction

One of the major practical obstacles to vaccination in the field, especially in developing countries, is the need to transport and store vaccines under refrigeration (i.e. the need to maintain a cold chain) (Giudice and Campbell, 2006). Other considerations include the need for skilled medical professionals to inject the vaccine via needle and syringe and the need to safely dispose of used needles and syringes from mass vaccinations; the latter is now a major concern, as medical waste affects our environment (Giudice and Campbell, 2006; Yuki and Kiyono, 2009). Furthermore, it should be noted that the aim of injectable vaccines is to induce protective systemic immunity, not mucosal immunity, as a first line of defence against mucosal pathogens that invade the host via inhalation and ingestion (Mason and Herbst-Kralovetz, 2012; Yuki and Kiyono, 2009). To overcome these critical concerns, of production, storage, and delivery, a rice-based vaccine (MucoRice) has recently attracted interest as an alternative approach for mucosal immunization (Nochi et al., 2007a, 2009; Tokuhara et al., 2010; Yuki et al., 2009, 2012). Our findings show that rice seed is suitable for the expression, accumulation and mucosal delivery of vaccine antigens; such vaccines are stable at room temperature for several years without loss of immunogenicity (Nochi et al., 2007a; Tokuhara et al., 2010). Successful oral vaccination of mice with MucoRice expressing the B subunit of cholera toxin (CTB) has induced antigen-specific serum IgG and mucosal IgA responses with toxin-neutralizing immunity (Nochi et al., 2007a; Tokuhara et al., 2010; Yuki et al., 2009). Furthermore, oral MucoRice-CTB induces cholera toxin (CT) neutralizing antibodies in non-human primates (Nochi et al., 2009).

Although many plant-based vaccines have been reported, there have been no detailed published examinations of the structure or amino acid sequence of these vaccines (Cardi et al., 2010; Paul and Ma, 2010). Because the use of plant-based expression systems sometimes results in unexpected molecular modifications to the vaccine, such as plant-associated sugar-chain attachment or blocking of N-terminal amino acid sequence by acetylation (Bosch and Schots, 2010; Twyman et al., 2003), it is essential to elucidate the exact structure and amino acid sequences of plant-expressed vaccines that are intended for human clinical application. Although the MucoRice expression and production system has advantages over other expression systems, such as plant-expression and bacterial fermentation or mammalian cell-culture systems, the two key issues of detailed molecular characterization and quality control remain to be resolved in order to advance the development of MucoRice-CTB for clinical study. In the initial development of MucoRice-CTB as a vaccine, we used the glutelin B promoter and signal peptide in an overexpressing vaccine cassette (Nochi et al., 2007a; Yuki et al., 2009) and found by SDS-PAGE analysis that MucoRice-CTB produced two CTB monomers (12 and 14 kDa), most likely the result of addition of the N-terminal signal peptide (28 amino acids) used for CTB or because of plant-based glycosylation. Because the native CTB monomer spontaneously forms a pentamer (Yuki et al., 2001), the presence of two kinds of monomeric CTB could yield at least six kinds of CTB pentamer, meaning that the composition and biological activity of the vaccine are difficult to quantify. To address this issue, we generated an overexpression cassette by fusing the rice 10 kDa prolamin signal peptide with the CTB gene and then connecting them with the 13 kDa prolamin promoter and terminator. In this overexpression cassette, the 10 kDa prolamin signal peptide is reported to be spliced correctly (Kuroda et al., 2010; Yuki et al., 2012). We also introduced an RNAi cassette into the T-DNA vector, together with the vaccine cassette, to suppress the expression of rice 13 kDa prolamin and glutelin A. This strategy was aimed at increasing the expression of CTB, because suppression of the expression of internal storage proteins such as prolamin and glutelin is known to enhance the accumulation of foreign proteins (Kuroda et al., 2010; Yuki et al., 2012). Although SDS-PAGE analysis of the freshly harvested original MucoRice-CTB/N vaccine showed the presence of two monomers 14 and 12.5 kDa in size, protein sequence analysis revealed that the signal sequence of 10 kDa prolamin was spliced correctly between Ala and Ser and suggested that a sugar group was attached to the sixth amino acid (Asn) of the 14 kDa CTB monomer but not the 12.5 kDa CTB monomer. The molecular heterogeneity of MucoRice-CTB was attributed to the above two factors, the results of addition of N-terminal blocking signal peptide used for CTB and plant-based glycosylation. Using the T-DNA vector system together with the RNAi suppression system, we established a new version of this vaccine, MucoRice-CTB/Q, that consists of a single monomer with an Asn to Gln substitution at one (the 4th) of two glycosylation-sequence positions (the other being the 89th) in the CTB sequence. In this study, we established MucoRice-CTB/Q as a new-generation plant-based sugar modification-free oral cholera vaccine in which the complete amino acid sequence is known. We here demonstrate that molecular uniform and plant–sugar modification-free MucoRice-CTB/Q are highly immunogenic and produce protective immunity in mice and macaques. This finding highlights the potential clinically applicability of MucoRice-CTB/Q as a new-generation oral cholera vaccine and its suitability for safety and efficacy testing in a clinical trial.

Results

MucoRice-CTB/N produces two monomers of CTB

In our original MucoRice system, the glutelin B-1 signal peptide gene was fused to the 5′end of the codon-optimized CTB gene linked to the ER retention signal peptide (KDEL) for rice seed under the control of an endosperm-specific promoter, glutelin GluB-1 (Nochi et al., 2007a; Yuki et al., 2009). MucoRice-CTB produced two CTB monomers (13- and 15 kDa). We used Western blot analysis with the rabbit Ab to glutelin B-1 signal peptide to confirm that these CTBs in rice seed contained the glutelin signal peptides (Data not shown). We showed that the glutelin B-1 signal peptide sequence was not spliced correctly when CTB was overexpressed in rice seed. Because the signal sequence of 10 kDa prolamin is spliced correctly between Ala (A) and Ser (S) when overexpressed in rice (Kuroda et al., 2010), we used the endosperm-specific promoter and signal peptide genes from prolamin in these constructs (Figure 1). In addition, we introduced a combination cassette for RNAi suppression of the 13 kDa prolamin and glutelin A rice storage proteins into the T-DNA vector to increase the expression of CTB (Figure 1). To examine the effects of this RNAi suppression, we initially compared rice expressing CTB containing no RNAi triggers (pZH2B-CTB/N-iK) with that containing the RNAi trigger (pZH2B-CTB/N-ik45-G1B). After we had established new versions of MucoRice-CTB, named MucoRice-CTB/N line 1 (MucoRice-CTB/N-1 with no RNAi trigger) and line 2 (MucoRice-CTB/N-2 plus RNAi triggers), SDS-PAGE analysis showed that both MucoRice CTB/N lines still generated two bands with molecular weights 12.5- and 14 kDa (Figure 2a). Densitometric analysis of the SDS-PAGE gels for the upper 14 kDa band generated by both MucoRice-CTB/N lines could not be performed accurately because of heavy contamination by rice proteins. Because the amount of CTB in the lower band was determined by densitometry, we used Western blot analysis to estimate the amount of CTB only in the upper band. The average expression levels for the N-1 and N-2 lines were estimated by SDS-PAGE and Western blot analysis to be approximately 0.33 mg/g seed and 1.34 mg/g seed, respectively. The total amount of CTB for the N-1 and N-2 lines reached 0.61 and 2.82 mg/g seed, respectively, after addition of the amount for the lower band for the N-1 line (an average of 0.28 mg/g seed) and the N-2 line (an average of 1.48 mg/g seed).

Figure 1.

T-DNA vector for the expression of MucoRice-CTB/N and MucoRice-CTB/Q. We constructed a tandem T-DNA plasmid containing a CTB/N or a CTB/Q overexpression cassette with authentic CTB or CTB (N4Q) sequences, respectively, controlled by the rice 13 kD prolamin promoter, and a combination cassette of RNAi triggers for suppression of the major rice endogeneous storage proteins, 13 kDa prolamin and glutelin, controlled by the ubiquitin promoter. We also constructed a CTB/N or CTB (N4Q) overexpression cassette without RNAi triggers for suppression of major rice endogeneous storage proteins. CaMV35S-P, cauliflower mosaic virus 35S promoter; mHPT, mutant hygromycin phosphotransferase; 13P-P, 13 kDa prolamin promoter; S-10P, signal sequence of 10 kDa prolamin; 13P-T, 13 kDa prolamin terminator; Nos-T, nos terminator; RAP intron, rice aspartic protease intron; Ubi-P, ubiquitin promoter; LB, T-DNA left border; RB, T-DNA right border.

Figure 2.

SDS-PAGE analysis of MucoRice-CTBN/Q. (a) SDS-PAGE analysis of protein expression on MucoRice-CTB/N lines and MucoRice-CTB/Q lines containing two T DNA plasmids: #N-1 or #Q-1, no RNAi suppression; #N-2 or #Q-2, RNAi suppression of 13 kDa prolamin and glutelin A. WT indicates wild-type rice, rSR-rCTB/Q (12.5 kDa) indicates recombinant CTB (N4Q) in addition to SR amino acids at N-terminal from B. choshinensis expression system, and rCTB (12 kDa) indicates recombinant CTB from B. choshinensis system. The results showed predominant expression of MucoRice-CTB/N (N-1 or N-2) with two monomeric bands of approximately 14 and 12.5 kDa, and MucoRice-CTB/Q (Q-1 or Q-2) with a monomeric band of approximately 12.5 kDa. Western blot analysis revealed that two transgenic proteins of 12.5 and 14 kDa specifically reacted with the anti-CTB Ab. (b) Each band 1, 2 or 3 on SDS-PAGE was analysed by MS/MS or a protein sequencer and was confirmed as a partial or complete amino acid sequence. The N-terminal part of the upper band 1 of MucoRice-CTB/N (line N-2) was not detected. Yellow indicates confirmed sequence and black indicates where the sequence is not confirmed. Red indicates substitution of amino acids in the authentic CTB.

A restriction enzyme (Xbal) site introduced between the 10 kDa prolamin signal peptide and the CTB N-terminal genes produces two amino acids Ser and Arg (SR) when spliced correctly (Figure 1). MS/MS analysis showed that the lower band (12.5-kDa) of MucoRice-CTB/N-2 contained the complete amino acid sequence in addition to the sequence of the N-terminal SR peptide as a restriction enzyme site. By comparison, MS/MS analysis showed that the 14 kDa band of the N-2 line contained the complete amino acid sequence but missed the N-terminal part of the native CTB sequence, SRTPQNITDLCAEYHNTQIHTLNDK (Figure 2b bands 1&2, Tables S1, S2). However, N-terminal sequence analysis of the 14 kDa band of MucoRice-CTB/N-2 only confirmed the presence of SRTPQX, but not below the position 6 (Asn) (data not shown). Taken together, these results suggested that the 14 kDa band of MucoRice-CTB/N attaches an N-linked sugar chain at position 6 (Asn) on the N-terminal sequence of the native protein. In addition, using Western blot analysis with the anti-HRP Ab, which recognizes plant-specific N-glycoproteins carrying β1,2-xylose and/or α1,3-fucose residues (Kajiura et al., 2012), we confirmed that MucoRice-CTB/N (only the 14 kDa monomeric band), but not MucoRice-CTB/Q, contained β1,2-xylose or α 1,3-fucose residues, or both (Figure S1a).

MucoRice-CTB/Q produces an authentic CTB monomer

To confirm the production of rice-based CTB vaccine without an N-linked sugar chain, we expressed CTB/Q (N4Q) in rice seeds as MucoRice-CTB/Q line 1 (MucoRice-CTB/Q-1 with no RNAi trigger) and line 2 (MucoRice-CTB/Q-2 plus RNAi triggers). As expected, SDS-PAGE analysis showed that MucoRice-CTB/Q-1 and MucoRice-CTB/Q-2 were monomers with a molecular weight of 12.5 kDa (Figure 2a). Because all peptides from trypsin digestion of the MucoRice-CTB/Q-2 band 3 were obtained by MS/MS analysis (Figure 2b band 3, Tables S3 and S4), we were able to confirm the complete sequence of MucoRice-CTB/Q and show that the CTB sequence was authentic, apart from the amino acid change to Gln and the addition of the N-terminal SR peptide (Figure 2b). To further confirm our result, we created recombinant SR-CTB/Q (N4Q) by using the Brevibacillus choshinensis expression system. The SDS-PAGE mobility of MucoRice-CTB/Q-1 and MucoRice-CTB/Q-2 (Figure 2) was consistent with that of recombinant SR-CTB/Q (12.5 kDa), and the mobility of authentic rCTB (12.0 kDa) showed only a slight difference from that of MucoRice-CTB/Q. This result confirmed that MucoRice-CTB/Q produced the complete CTB sequence without plant-based modification. We quantified the CTB produced by MucoRice-CTB/Q by using SDS-PAGE densitometric analysis with recombinant SR-CTB/Q as the standard. The amount of SR-CTB/Q produced by MucoRice-CTB/Q-1 and MucoRice-CTB/Q-2 was estimated by SDS-PAGE densitometric analysis to be an average of 0.42 and 2.35 mg/g seed weight, respectively. Taken together with the results for MucoRice-CTBN (Figure 2), these data suggested that introduction of the RNAi system (N-2 and Q-2 lines) minimized the production of rice storage proteins (prolamines and glutelins) and thus expanded the space available for expression and accumulation of the vaccine antigen in the rice seed. Therefore, we used MucoRice-CTB/N-2 seed (2.82 mg/g seed) and MucoRice-CTB/Q-2 seed (2.35 mg/g seed) as the original and new-generation MucoRice-CTB, respectively, for all further immunological assessments.

Our results further confirmed that MucoRice-CTB/Q had a pentameric structure and GM1 receptor-binding activity similar to those of MucoRice-CTB/N after the purification of CTB/N and CTB/Q from MucoRice-CTB/N-2 and MucoRice-CTB/Q-2 seeds, respectively (Figure S1b and c). Because CTB/N binds to a plant N-glycan, its apparent molecular weight of 52 kDa was estimated to be higher than that of CTB/Q (42 kDa) when assessed by HPLC using a Superose 12 column in PBS (Figure S1b). These results provide further supportive evidence for the applicability of MucoRice-CTB/Q as a vaccine in humans.

MucoRice-CTB/Q is effectively taken up by the mucosal inductive site

We previously showed that MucoRice-CTB expressing the two types of CTB monomer was effectively taken up by M cells associated with the PPs, which function as the mucosal inductive site or gut-associated lymphoid tissue (GALT), to induce antigen-specific mucosal immunity (Nochi et al., 2007a). To confirm uptake of MucoRice-CTB/Q by GALT, we administered a seed powder suspension of MucoRice-CTB/Q, MucoRice-CTB/N, or WT rice into ligated small intestine loops including PPs in naïve mice. Histological analysis with CTB-specific rabbit Ab and M-cell-specific Rat mAb NKM-16-2-4 using the appropriate fluorescence detection system showed that the original MucoRice-CTB/N was taken up by M cells located in the GALT epithelium, thus confirming the results of our previous study (Nochi et al., 2007a). Of note, the improved MucoRice-CTB/Q was equally taken up by the M cells associated with the PPs (Figure 4). These results demonstrate that MucoRice-CTB/Q is an effective vaccine delivery vehicle that targets vaccine to mucosal inductive sites such as PPs and GALT.

Oral immunogenicity of MucoRice-CTB/Q for the induction of antigen-specific immune responses in systemic and mucosal compartments, providing protective immunity against CT-induced diarrhoea

In our previous study, we demonstrated that oral immunization of mice with MucoRice-CTB expressing the two types of CTB monomer (or MucoRice-CTB/N) induced antigen-specific immune responses in both the systemic and the mucosal compartment (Nochi et al., 2007a; Tokuhara et al., 2010; Yuki et al., 2009). To examine whether MucoRice-CTB/Q could induce similar antigen-specific mucosal and systemic immune responses to that of MucoRice-CTB/N, we orally immunized mice with 22–130 mg of MucoRice-CTB/N or 26–155 mg of MucoRice-CTB/Q containing 60–360 μg CTB. Systemic CTB-specific IgG Ab and mucosal CTB-specific IgA Ab immune responses were induced by both MucoRice-CTB/Q and MucoRice-CTB/N in the same dose-response manner (Figure 4a,b). The systemic and mucosal Ab immune responses were not significantly different between mice immunized with MucoRice-CTB/Q and those immunized with MucoRice-CTB/N (Figure 4a,b). Furthermore, to examine the vaccine efficacy of the induced CTB-specific Ab, we challenged immunized mice with MucoRice-CTB/Q, MucoRice-CTB/N, or WT rice with CT. Mice immunized with MucoRice-CTB/Q or MucoRice-CTB/N were protected against CT-induced diarrhea, whereas those immunized with WT rice developed severe diarrhoea (Figure 4c) There was no significant difference in the level of protection against CT-induced diarrhoea between mice immunized with MucoRice-CTB/Q and those immunized with MucoRice-CTB/N. These findings demonstrated that MucoRice-CTB/Q provided a high level of oral immunogenicity and had the capacity to induce protective immunity.

Oral MucoRice-CTB/Q induces antigen-specific immune responses with neutralizing antibodies in non-human primates

Our previous study demonstrated that oral immunization of non-human primates with MucoRice-CTB (or MucoRice-CTB-N) induced antigen-specific Ab immune responses with toxin-neutralizing activity (Nochi et al., 2009). To examine whether MucoRice-CTB/Q could induce similar antigen-specific Ab responses in non-human primates to that of MucoRice-CTB/N, we orally immunized macaques 5 times at 2-week intervals with MucoRice-CTB/Q (600 mg powder) or MucoRice-CTB/N (500 mg powder), both containing 1.4 mg of CTB, or WT rice (600 mg powder). After three doses of the primary immunization, the level of CTB-specific IgG with inhibition of GM1-binding was increased in all macaques immunized with MucoRice-CTB/Q or MucoRice-CTB/N, but not in the control (Figure 5a,b). One macaque from each group (#1 and #3) showed a high serum antigen-specific IgG immune response (Figure 5a). There was no difference in the antigen-specific Ab immune responses between macaques immunized with MucoRice-CTB/Q and those immunized with MucoRice-CTB/N. Similar to our previous study (Nochi et al., 2009), examination of pre-immunization faecal samples showed that CTB-specific faecal IgA Abs were already present in the macaques used for this study (Figure 5c). Furthermore, consistent with our previous findings (Nochi et al., 2009), the level of pre-existing CTB reactive IgA Abs did not alter following immunization (Figure 5c). In addition, no rice protein-specific IgE immune responses were found in immunized macaques (data not shown). These results showed that MucoRice-CTB/Q is a potent and safe oral vaccine that can induce toxin-specific neutralizing Abs in non-human primates.

Discussion

As illustrated in our previous study, SDS-PAGE analysis of the first version of MucoRice-CTB showed that it produced two produced two CTB monomers (Nochi et al., 2007a). The amino acid sequence of the N-terminal region of this MucoRice-CTB vaccine could not be determined because of the presence of an N-terminal blocking group (Nochi et al., 2007a; Yuki et al., 2009). The original T-DNA vector had the GluB-1 signal peptide gene fused to the 5′end of the CTB gene and placed under the control of the endosperm-specific promoter, glutelin GluB-1 (Nochi et al., 2007a; Tokuhara et al., 2010; Yuki et al., 2009). Using a different variation of the T-DNA vector, which instead included the prolamin promoter and signal peptide genes (Kuroda et al., 2010), we generated MucoRice-CTB/N, the second-generation MucoRice-CTB vaccine (Figure 2a). Analysis of MucoRice-CTB/N revealed that it consisted of an authentic CTB amino acid sequence but still produced two CTB monomers (Figure 2a). Further modification led to the generation of MucoRice-CTB/Q, which expressed CTB/Q (N4Q) with an Asn to Gln substitution at the 4th glycosylation-sequence position of CTB and resulted in the production of a single CTB monomer (Figure 2b). This finding suggested that, for MucoRice-CTB/N, the upper 14 kDa band, but not the lower 12.5 kDa one, had an additional sugar chain at position No 6. Structural analysis of the original MucoRice-CTB with the glutelin GluB-1 signal peptide showed that most of the high-molecular-weight CTB monomer (71%) contained α1,3-fucosylated or β1,2-xylose N-glycans, or both (Kajiura et al., 2012). Consistent with these findings, we used Western blot analysis with anti-HRP Ab to show that the 14 kDa MucoRice-CTB/N monomeric band, but not the 12.5 kDa one, contained α1,3-fucosylated or β1,2-xylose N-glycans, or both (Figure S1a). We were able to confirm the complete amino acid sequence of MucoRice-CTB/Q by MS/MS analysis and show that, in contrast, it had no modifications apart from the addition of the N-terminal SR peptide as a restriction enzyme site (Table S4). Because carbohydrate-specific IgE Abs that predominantly bind non-mammalian typed xylose or fucose on the N-glycans of plant have been found in patients allergic to pollen allergens (van Ree, 2002), the molecularly improved MucoRice-CTB/Q has much less safety concern than the original MucoRice-CTB/N, since the advanced MucoRice-CTB/Q lacks allergenicity-associated plant N-glycans such as α1,3-fucose or β1,2-xylose (Figure S1a). Thus, the molecularly advanced form of MucoRice-CTB/Q is now suitable for future clinical development.

The plant transgenic systems have been used for expression of CTB or LTB (Arakawa et al., 1998; Daniell et al., 2001; Jiang et al., 2007; Tacket et al., 1998, 2004). Both edible (e.g., potato and tomato)- and bio-factory (e.g., tobacco)-type plants are useful systems for the expression of CTB and LTB, which can be advanced to plant-based oral vaccine candidates. MucoRice-CTB/Q, the second-generation MucoRice-CTB vaccine, has the potential to move the development of rice-based cold-chain-free oral vaccines toward clinical investigation. Because we previously demonstrated that MucoRice-CTB was stable at room temperature for 3 years in the state of rice seed (Nochi et al., 2007a; Tokuhara et al., 2010), we plan to use a polished rice powder preparation of MucoRice-CTB without any further purification for future clinical studies to maintain the cold-chain-free feature of this rice-based vaccine. The precise amounts of CTB in MucoRice seed or powder must be quantified for clinical application. Because quantification by Western blot analysis depends on the efficacy of blotting and visualization of the vaccine antigen by using a specific Ab-enzyme detection system (Towbin et al., 1979), it is difficult to accurately measure the amount of CTB expressed in MucoRice seed or powder, especially if the preparation contains both native and modified CTB, owing to the presence of the plant glycosylation system. By comparison, the amount of CTB expressed from MucoRice-CTB/Q in rice seed can be accurately determined by SDS-PAGE densitometric analysis of the single CTB monomer (Figure 2a).

In our original design of the transgene expression system for the MucoRice-CTB vaccine, we used an ER retention signal peptide (KDEL) gene fused to the 3′end of the CTB gene to increase expression of the transgene (Nochi et al., 2007a; Yuki et al., 2009). A potential safety concern is the possibility that orally administered vaccine antigen linked to KDEL can be redirected to the plasma membrane by retrograde transport via the ER in gut epithelial cells (Hagiwara et al., 2006). In addressing this issue, by using a BoHc overexpression system together with RNAi technology to suppress the production of both major endogenous storage proteins, prolamin and glutelin A, we reported recently that a heavy chain fragment of botulinum neurotoxin (BoHc) was produced at high levels in rice seed (Yuki et al., 2012). In the present study, we adopted the RNAi approach to develop MucoRice-CTB/Q without the KDEL peptide. Our introduction of the RNAi suppression cassette into the T-DNA vector greatly improved the production of CTB/Q (2.35 mg/g seed) when compared with using the T-DNA vector without the RNAi cassette (0.42 mg/g seed, Figure 2b). Our findings here suggest that the advanced MucoRice system with RNAi technology can be extended to the production of other universal vaccine antigens.

To induce effective mucosal immunity in the digestive tract, vaccine antigens administered orally need to be taken up by the PPs via the antigen-sampling cells or M cells located in the follicle-associated epithelium (Nochi et al., 2007b). To examine whether MucoRice-CTB/Q was effectively taken up by mucosal inductive sites such a PPs, we investigated whether MucoRice CTB/Q or MucoRice CTB/N was taken up by the M cells associated with the PPs. We used a murine intestinal loop assay to show that administration of MucoRice CTB/Q or MucoRice CTB/N into the intestinal loop resulted in uptake by the M cells of both forms of MucoRice-CTB (Figure 3). Because mammalian epithelial cells, including those of mice, express GM1, a receptor of CTB or CT (Mekalanos et al., 1983), it is expected that MucoRice-CTB could bind to these epithelial cells and be taken up. To this end, our previous study revealed trace binding of CTB onto the intestinal epithelium of ligated murine intestine treated with the original form of MucoRice-CTB (Nochi et al., 2007a). Taken together, our current results demonstrate MucoRice CTB/Q possesses clinically applicable characteristics and can be considered as a next-generation plant-derived effective mucosal vaccine that can deliver vaccine antigen to the inductive site.

Figure 3.

Uptake of MucoRice-CTBQ by M cells in mice. MucoRice-CTB/N (panel a & d), MucoRice-CTB/Q (panel b & e), or WT rice (panel c) was administered into ligated small intestine loops including PPs in naïve mice. Immunohistochemical analysis of PPs 4 h after administration revealed that the CTBs of both MucoRice-CTB preparations (panels a and b) but not of WT-rice (panel c) were bound by NKM 16-2-4-positive M cells or columnar epithelial cells and taken up into the cells (arrows in panels a and b). The negative control for CTB staining is shown in panels d and e.

Although the N-terminal part (positions 1–7) of CTB does not contain B- or T-cell epitopes (Cong et al., 1996; Jacob et al., 1983), we investigated whether minor modification of the N-terminal region of MucoRice-CTB/Q [SR-CTB (N4Q)] could affect the induction of antigen-specific immune-responses when the vaccine was administered via the oral route. The plant N-glycan possessed by MucoRice-CTB/N may affect the antigen-specific immune responses by a class of C-type lectins serving as mannose-binding receptors (Keler et al., 2004; Lam et al., 2007). In fact, Tobacco-based vaccine consisting of a fusion protein HIV peptide and CTB or rice-based native CTB possesses N-glycan with a high-mannose-type glycans in the N-terminal region of CTB (Kajiura et al., 2012; Matoba et al., 2008, 2009). The plant-based CTB-HIV peptide or CTB vaccines retained affinity to GM1 ganglioside and the critical antigenicity of the peptide moiety; it is therefore possible that the presence of high-mannose glycans in the majority of plant-based vaccines could lead to enhanced immune responses to the CTB or CTB fusion protein (Matoba et al., 2009). We examined the immunogenicity of MucoRice-CTB/Q (SR-CTB/Q) and compared it with that of MucoRice-CTB/N [a mixture of SR-CTB with or without a plant–sugar chain] by orally immunizing two groups of mice with various doses of the two preparations. Similarly high titres of CTB-specific systemic IgG and mucosal IgA antibodies were induced in mice orally immunized with MucoRice-CTB/Q or MucoRice-CTB/N (Figure 4a,b). We also examined the efficacy of the CTB-specific Abs induced by orally administered MucoRice-CTB/Q. Mice immunized with MucoRice-CTB/Q and challenged with CT were protected against CT-induced diarrhea, and there was no significant difference in vaccine efficacy between MucoRice-CTB/Q and MucoRice-CTB/N (Figure 4c). These results show that MucoRice-CTB/Q and MucoRice-CTB/N have equivalent immunogenicity and suggest that removal of the plant-based sugar chain does not affect the ability of MucoRice-CTB/Q to induce CTB-specific Abs. Taken together, our results show that the improved MucoRice-CTB/Q vaccine overcomes the initial concerns with the original MucoRice-CTB, including plant-related sugar modification and molecular non-uniformity, and thus has potential as an oral cholera vaccine for future clinical development.

Figure 4.

Comparison of oral immune-responses and protective immunity induced with MucoRice-CTBN/Q in mice. (a, b) When mice were orally immunized a total of four times at 2-week intervals with MucoRice-CTB/N, MucoRice-CTB/Q or nontransgenic rice dissolved in water as a control, equal CTB-specific serum IgG and faecal IgA responses were induced in mice immunized with MucoRice-CTB/N or MucoRice-CTB/Q, but not in mice receiving WT rice. (c) In contrast to mice receiving WT rice, mice orally vaccinated with MucoRice-CTB/N or MucoRice-CTB/Q showed no signs of diarrhoea and a low intestinal water content. *, P < 0.05, MucoRice-CTB/N or MucoRice-CTB/Q versus WT-Rice. (d) After CT challenging, typical small intestines from mice receiving orally vaccinated with rice-expressed CTB/N or CTB/Q showed no symptoms of diarrhoea when compared with that from mice receiving WT rice.

In preparation for future human clinical studies with MucoRice-CTB/Q, we examined its oral immunogenicity in non-human primates. When one group of macaques was orally immunized with MucoRice-CTB/Q and one with MucoRice-CTB/N, both MucoRice-CTB preparations induced similarly high antigen-specific serum IgG Ab immune responses with neutralizing activity against CT (Figure 5a,b). In addition, neither MucoRice-CTB preparation induced rice protein-specific IgE Ab responses in macaques (data not shown). As we reported previously (Nochi et al., 2009), macaques naturally possess CTB-specific faecal IgA Abs; thus, all pre-immunization samples in this study contained pre-existing CTB-specific faecal IgA Abs (Figure 5c). It has further been shown that the level of pre-existing CTB-reactive faecal IgA antibodies does not change after oral immunization with MucoRice-CTB (Nochi et al., 2009). In all macaques immunized in the present study, the CTB-specific IgA titre remained unchanged after oral immunization with MucoRice-CTB/Q and MucoRice-CTB/N (Figure 5c); this result is in complete agreement with those of our former investigation (Nochi et al., 2009). These results suggest that the pre-existing CTB-reactive intestinal humoral immunity (or CT-reactive SIgA) that had developed in the conventional housing environment may have already reached immunological plateau levels, as reported previously (Nochi et al., 2009), and thus could not be augmented by oral vaccination with MucoRice-CTB/Q. Although we still could not logically explain why macaques naturally acquire CTB-specific faecal IgA Abs, CTB is known to share a high level of identity with the B subunit of heat-label toxin (LT) produced by enterotoxigenic Escherichia coli (Spangler, 1992) meaning that CTB-specific faecal IgA Abs can cross-react with LTB (Clemens et al., 1988). Thus, it is possible that LT-producing enterotoxigenic E. coli or an as-yet unknown homologous bacterium could induce pre-existing cross-reactive IgA Abs following natural infection. It is also possible that these bacteria are part of the commensal flora of macaques and thereby stimulate only the mucosal, and not the systemic immune system, leading to the induction of faecal antigen-specific IgA, but not serum IgG Ab responses (Macpherson and Uhr, 2004). Our findings in macaques further demonstrate that MucoRice-CTB/Q is a safe and immunogenic oral vaccine preparation and is a suitable candidate for human clinical studies.

Figure 5.

Comparison of oral immune responses and neutralizing Ab in non-human primates. (a, c) Following immunization of macaques five times at 2-week intervals, MucoRice-CTB/N and MucoRice-CTB/Q, but not WT rice, induced CTB-specific serum IgG but not faecal IgA Abs after the fifth immunization. The serum IgG immune response of one macaque from each group (#1 and #3) dramatically increased after the third oral dose. (b) The serum collected from immunized macaques, but not the control macaque, inhibited the binding of CT to GM1 ganglioside at a level corresponding to the Ab titre.

Experimental procedures

DNA construction and transformation of rice plants

The sequences encoding the CTB/N (original amino acid sequence) and CTB/Q (N4Q) were synthesized with optimized codon usage for rice (Nochi et al., 2007a) and inserted into a binary T-DNA vector (pZ2028) with an overexpression cassette for CTB/N or CTB/Q and a combination cassette for RNAi suppression of the major rice endogenous storage proteins 13 kDa prolamin and glutelin A (pZH2Bik45-G1B) as shown in Figure 1 (Kuroda et al., 2010). The RNAi trigger sequence for the gene encoding 13 kDa prolamin was a 45-bp fragment of rice 13 kDa prolamin gene (RM1: Os07 g0206500) comprising coding sequence (CDS) 1–45. The RNAi trigger sequence for glutelin genes was a 129-bp fragment of the rice glutelin gene (GluA: Os10 g0400200) comprising coding sequence 142–270. The Acs I–Mul I fragment of the CTB expression cassette was subcloned into pZH2Bik45-G1B. The resultant expression vectors for the expression of CTB/N and CTB/Q together with the combination RNAi cassette were named pZH2B-CTB/N-ik45-G1B and pZH2B-CTB/Q-ik45-G1B, respectively. We also designed and subcloned overexpression cassettes for CTB/N and CTB/Q into an RNAi cassette without the sequence triggering expression of the rice endogenous storage proteins (pZH2BiK); we named these pZH2B-CTB/N-iK and pZH2B-CTB/Q-iK, respectively. We transformed a japonica rice plant cultivar, Nipponbare, with these plasmids by using an Agrobacterium-mediated method, as described previously (Nochi et al., 2007a). Several independent transgenic rice lines were generated for each of the four types of MucoRice, and the CTB accumulation levels in seeds were determined by SDS-PAGE densitometric analysis or Western blotting, or both. For each of the four types of MucoRice, the plant line with the highest levels of CTB antigen accumulated in the seed was selected and advanced to the T4 generation by self-crossing to obtain homozygous lines.

Protein expression, purification and analyses

For recombinant CTB and SR-CTB/Q expression, we prepared the secretion expression vector pNU212 containing a synthetic SR-CTB/Q gene (pNU212-SR-CTB/Q) in accordance with a method described previously (Yuki et al., 2001); we then introduced it into Brevibacillus choshinensis 47K. After concentration of the culture supernatant containing pNU212-CTB or pNU212-SR-CTB/Q, the rCTB or rSR-CTB/Q was purified by using a galactose-immobilized gel (Pierce, Rockford, IL) column followed by gel filtration on a Sephadex G-100 column (GE Healthcare Biosciences, Pittsburgh, PA).

Total seed proteins were extracted from transgenic rice plant seeds as described previously (Nochi et al., 2007a). Seeds of rice plants were ground to a fine powder using a Multibeads shocker (Yasui Kikai, Osaka, Japan), and total seed proteins were extracted from 20 mg rice powder in 1 mL of the sample buffer [2% (wt/vol) SDS, 5% (wt/vol) β-mercaptoethanol, 50 mm Tris-HCl (pH 6.8) and 20% (wt/vol) glycerol] before being separated by SDS/PAGE with a 12% polyacrylamide gel (NuPAGE 12% Bis-Tris Gel, Life Technologies, Carlsbad, CA). SDS/PAGE was carried out using 10 μL of sample, and the protein was subsequently transferred to PVDF membranes (GE Healthcare) for Western blot analysis with HRP-conjugated rabbit anti-CTB Ab (500 ng/mL) prepared in our laboratory as described previously (Nochi et al., 2007a). A serial dilution of recombinant CTB (0.05, 0.1, 0.2, 0.4 and 0.8 μg/mL) was used as a standard for the quantitation western blot analysis. For SDS-PAGE densitometric analysis, SDS-PAGE gel was stained using Invitrogen LC6065 Simply Blue Safe Stain kit (Life Technologies) for 12 h at room temperature, and the stained spot intensity of CTB on SDS-PAGE was measured on GS-800 calibrated densitometer (Bio-Rad Laboratories, Hercules, CA) using a BioRad Quantity One software. A serial dilution of recombinant CTB or SR-CTB/Q (0.1, 0.2, 0.4, 0.8, and 1.6 μg/mL) was used as a standard for the densitometric analysis.

Mass spectrometry and N-terminal amino acid analyses

To identify the CTB expressed in transgenic rice, we performed a mass spectrometric analysis as described previously, with some modifications (Oyama et al., 2009). MucoRice-CTB powder (20 mg) was lysed in 1 mL of sample buffer (2% SDS, 5% β-mercaptoethanol, 50 mm Tris-HCl (pH 6.8) and 20% glycerol) and separated by using SDS-PAGE on a 12% NuPAGE Bis-Tris Gel (Life Technologies). The CTB bands with molecular weight of 12–14 kDa were cut out of the gel, digested at 37 °C overnight with 0.2 μg of trypsin (Sequence grade, Promega, Madison, WI), desalted and then concentrated to a volume of 20 μL. The samples were then injected into a direct nanoflow liquid chromatography system (Dina; KYA Technologies) and sprayed into a quadrupole time-of-flight tandem mass spectrometer (QSTAR Elite; Applied Biosystems, Carlsbad, CA) or a linear ion trap Orbitrap mass spectrometer (LTQ Orbitrap; Thermo Fischer scientific, Suwanee, GA). Analysis of the data using the Mascot search server confirmed the complete amino acid sequences of the CTB bands expressed in transgenic rice, with the exception of that of the14 kDa monomeric band of MucoRice-CTB/N. To determine the N-terminal amino acid sequence of this band, the band was excised from the membrane after electroblotting and CCB staining and subjected to a gas-phase protein sequencer model 494A (Applied Biosystems).

Animals and immunization

All experiments were performed according to the guidelines provided by the Animal Care and Use Committee of The University of Tokyo and NIBIO.

Female BALB/c mice (5–7 weeks old) purchased from Japan Clea, (Tokyo, Japan) were used for oral immunization at the Institute of Medical Science of The University of Tokyo. Eight-week-old female mice (five per group) were orally given with 21, 42, 85 or 128 mg of MucoRice-CTB/N with a corresponding dose of CTB at 60, 120, 240 or 360 μg, respectively; or with 25, 51, 102 or 150 mg of MucoRice-CTB/Q with a corresponding dose of CTB at 60, 120, 240 or 360 μg, respectively, by stomach tube a total of four times at 2-week intervals. In the control group, mice (five per group) were orally given 100 mg of powdered non-transgenic wild-type (WT) rice in distilled water.

Five female naive cynomolgus macaques, Macaca fascicularis, aged 5 years and with a body weight of approximately 3 kg, were used for the immunization and were maintained at the Tsukuba Primate Research Center for Medical Science at the National Institute of Biomedical Innovation (NIBIO, Ibaraki, Japan). Two cynomolgus macaques (nos. 001, 002) and other two macaques (nos. 003, 004) were orally immunized with 500 mg of powdered MucoRice-CTB/N or 600 mg of powdered MucoRice-CTB/Q containing 1.4 mg of CTB, respectively. One macaque (no. 005) was given the same amount (600 mg) of powdered non-transgenic WT rice. The rice powder was suspended in 5–6 mL of physiologic saline and administered on five occasions at 2-week intervals under ketamine anaesthesia.

Uptake of MucoRice-CTB/N and MucoRice-CTB/Q by M cells

After the mice were anaesthetized with ketamine (5 mg/mL, Sigma, Tokyo, Japan), we injected 20 mg (56 μg of CTB) of MucoRice-CTB/N or 24 mg (56 μg of CTB) MucoRice-CTB/Q into intestinal loops containing Peyer's patches (PPs). The mice were killed 4 h after the inoculation, and frozen sections (7 μm) of intestinal loop containing PPs were stained with 0.33 μg/100 μL of NKM16-2-4 [M-cell-specific monoclonal Ab produced by our laboratory (Nochi et al., 2007b)], followed by 200 dilution of Alexa 488-conjugated rat IgG in 100 μL (Jackson Immuno Research, West Grove, PA) or 1 μg/100 μL of anti-CTB rabbit Ab (protein A purified, our laboratory) and then 200 dilution of Alexa 647-conjugated rabbit IgG in 100 μL (Jackson Immuno Research). Normal rabbit IgG (1 μg/100 μL, protein A purified in our laboratory) followed by 200 dilution of Alexa 647-conjugated rabbit IgG in 100 μL (Jackson Immuno Research) was used as a negative control for CTB staining.

Detection of CTB- Ab production by ELISA

CTB-specific Ab responses were determined as described previously (Nochi et al., 2007a). The wells of 96-well microtiter plates (BD Falcon, Franklin Lakes, NJ) were coated with 100 μL CTB at 5 μg/mL and incubated overnight at 4 °C. Two-fold serial dilutions of serum and nasal wash were added for 2 h at room temperature (RT) after being blocked with 1% BSA. After washing of the plates, HRP-conjugated goat anti-mouse IgG or HRP-conjugated goat anti-mouse IgA, each diluted 1:4000 (SouthernBiotech, Birmingham, AL), was added and the plates were kept for 1 h at RT. The reaction was finally developed with a TMB Microwell Peroxidase Substrate System (XPL, Caithersburg, MD). Endpoint titres were expressed as the reciprocal log2 of the last dilution, which gave an OD450 of 0.1 greater than the negative control.

Toxin challenge

An in vivo oral CT challenge test was used as described previously (Tokuhara et al., 2010). After being fasted for 12 h, mice (10 per group) were orally challenged with 20 μg of CT (List Biological Laboratories, Campbell, CA). Nine to twelve hours after the challenge, the mice were sacrificed. The small intestine and colon were removed for clinical diarrhoea observation and collection of intestinal contents. After centrifugation of the samples, the volume of intestinal water was measured.

GM1-binding inhibition assay

A GM1-binding inhibition assay was performed using a GM1-ELISA as described previously, with some modifications (Nochi et al., 2009). The serum of macaques (10%, v/v) was treated with CT at a final concentration of 50 ng/mL for 1 h at RT and then incubated in 96-well plates (Nunc #439454, Roskilde, Denmark) coated with monosialoganglioside GM1 (5 μg/mL, Sigma) for 1 h at RT. After being washed, the plates were incubated for 1 h at RT with HRP-conjugated rabbit anti-CTB Ab (500 ng/mL) prepared in our laboratory (Nochi et al., 2007a), and TMB Substrate was used for detection of the HRP reaction.

Data analysis

Data are expressed as means ± SD. All analyses for statistically significant differences were performed with Student's t-test.

Acknowledgments

We are grateful to H. Hatai and A. Cyubachi for their technical support. This work was supported by grants from the Programs of Special Coordination Funds for Promoting Science and Technology and a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (Y.Y., H.K.); the Ministry of Health, Labor and Welfare of Japan (Y.Y., H.K.); New Energy and Industrial Technology Development Organization (NEDO) (H.K.); Adaptable and Seamless Technology Transfer Program through Target-driven R&D (A-step) and the Research and Development Program for New Bio-industry Initiatives of the Bio-oriented Technology Research Advancement Institution (Y.Y.).

Disclosures

The authors have no financial conflict of interest.

Ancillary