Rotavirus is the main cause of gastroenteritis in children worldwide, and the World Health Organisation has recommended that a rotavirus vaccine should be included in all infant immunization programmes. VP6 is the most immunogenic rotavirus subunit and is a potential target for an oral subunit vaccine. VP6 accumulated at up to 3% of total soluble protein in the young leaves of transplastomic tobacco plants, but the protein was unstable and was lost as the leaves aged. The aim of this study was to alter the 5′-untranslated region (5′-UTR) and the 5′ end of the coding region of VP6 cDNA in an attempt to increase the expression and stability of VP6 protein in tobacco chloroplasts. The inclusion of the 5′-UTR from gene 10 of bacteriophage T7 (T7g10) and the addition of 15 nucleotides, encoding five additional amino acid residues, at the 5′ end of the coding region increased the expression to >15% of total leaf protein and stabilized the protein in ageing leaves. Plants containing VP6 expression constructs with the rbcL 5′-UTR and with the native VP6 5′ end of the coding region produced VP6 protein at only 1.9% of total leaf protein. Both the T7g10 5′-UTR and the additional 15 nucleotides increased transcript accumulation and translational efficiency compared with VP6 constructs containing the rbcL 5′-UTR. The VP6 protein produced from all gene constructs appeared to be susceptible to proteolytic processing at its N-terminal region. However, in all transplastomic lines, VP6 proteins assembled into the trimeric form found in the rotavirus capsid.
Rotavirus is the leading cause of severe gastroenteritis in children worldwide, causing an estimated annual death rate of 500 000 children under 5 years of age (Parashar et al., 2009). Rotavirus particles were first identified by electron microscopy in duodenal epithelial cells from children with acute gastroenteritis (Bishop et al., 1973) and have subsequently been examined structurally in great detail (Prasad et al., 1988; Zhang et al., 2008; Chen et al., 2009; Li et al., 2009; McClain et al., 2010; Settembre et al., 2011). The viral particles have a triple-layered structure surrounding a central cavity containing 11 segments of double-stranded RNA encoding the individual components of the virus (Newman et al., 1975; Estes and Kapikian, 2007). The outer layer consists of 60 trimers of the VP4 spike protein embedded in 260 trimers of the VP7 glycoprotein, the middle layer consists of 260 trimers of VP6, and the inner layer consists of 120 copies of VP2 (Prasad et al., 1988; Zhang et al., 2008). The VP4 and VP7 proteins are the targets for neutralizing and protective antibodies, and two live attenuated oral vaccines, RotaTeq and Rotarix, are currently being licensed worldwide (Matson, 2006; O’Ryan, 2007; Dennehy, 2008). The World Health Organization (WHO) recommended in 2009 that rotavirus vaccine for infants should be included in all national immunization programmes. However, a live tetravalent vaccine, RotaShield, licensed in 1998, was withdrawn after 14 months because of a potential risk of intussusception, an intestinal inversion (Bines, 2006; Dennehy, 2008). Because of possible safety limitations with the use of live rotavirus vaccines, alternative strategies including inactivated rotavirus strains, virus-like particles and subunit vaccines are currently being investigated (Ward and McNeal, 2010).
VP6 makes up more than 50% of the mass of proteins in the rotavirus particle and is a highly conserved protein (Tang et al., 1997). The VP6 protein is also highly antigenic, and the majority of antibodies generated on rotavirus infection are directed against VP6 (Svensson et al., 1987). Oral administration of recombinant VP6 is able to induce serum IgG and mucosal IgA antibodies and to protect mice and neonatal calves against rotavirus infection (Choi et al., 1999; Dong et al., 2005a,b; Gonzalez et al., 2010; Zhou et al., 2010). IgA antibodies against VP6 have a protective effect against rotavirus infection in mice, even though they are not neutralizing (Burns et al., 1996) and this has been shown to be due to the inhibition of rotavirus replication at an early stage of infection (Feng et al., 2002). Priming with VP6 has also been shown to enhance the production of neutralizing antibodies to VP4 and VP7 in mice (Esquivel et al., 2000). VP6 therefore has many desirable features as a rotavirus vaccine candidate.
The aim of the work described in this paper was to increase the accumulation and stability of the VP6 protein produced in tobacco chloroplasts by altering the 5′ untranslated region (UTR) and the 5′ end of the coding region in the VP6 transcripts. The ribosome-binding site and 5′-UTR from gene 10 of bacteriophage T7 (T7g10) have been shown to be very effective in tobacco chloroplasts (Kuroda and Maliga, 2001a; Herz et al., 2005), and this may improve translation compared with transcripts containing the ribosome-binding site and 5′-UTR from tobacco chloroplast rbcL used previously (Birch-Machin et al., 2004). The VP6 protein produced earlier was predicted to contain two additional amino acids (asparagine and serine) following the initiating methionine at the N-terminus, owing to the use of an EcoRI restriction site to link the VP6 cDNA to the 5′-UTR and promoter (Birch-Machin et al., 2004). This alteration to the native amino acid sequence may be responsible for the instability of the protein. To examine this possibility, the expression of constructs encoding the native VP6 sequence and a VP6 protein with five additional amino acid residues at the N-terminus has been compared to the previous constructs encoding VP6 with two additional amino acid residues at the N-terminus. The addition of 15 nucleotides encoding five amino acid residues (Met-Ala-Ser-Ile-Ser) at the N-terminus has been shown to increase markedly the accumulation of β-glucuronidase (GUS) protein in tobacco chloroplasts (Herz et al., 2005). Inclusion of the T7g10 5′-UTR and a 15-nucleotide sequence encoding an N-terminal 5-amino acid extension in the VP6 construct has resulted in the stable accumulation of VP6 protein to >15% total leaf protein in transplastomic tobacco plants.
Production of transplastomic plants
In an attempt to improve accumulation and stability of the VP6 protein in tobacco chloroplasts, four new gene constructs with alterations of the 5′-UTR or the 5′ end of the VP6 coding region, compared to the constructs used by Birch-Machin et al. (2004), have been introduced into the transformation vector pZSloxP (Zhou et al., 2008) for insertion into the plastid genome between the rbcL and accD genes (Figure 1). In all constructs, the bovine VP6 cDNA, under the control of the tobacco plastid rrn promoter (Prrn) and the E. coli rrnB terminator (TrrnB), was inserted downstream of and in the opposite orientation to the aadA selectable marker gene. The VP6-A constructs contained the tobacco plastid rbcL 5′-UTR linked either directly (VP6-AD construct) or by an introduced EcoRI site (VP6-AE construct, as used by Birch-Machin et al., 2004) to the VP6 coding region. The VP6-B constructs contained the T7g10 5′-UTR linked directly (VP6-BD construct) or by an introduced EcoRI site (VP6-BE construct) to the VP6 coding region. The VP6-CD construct contained the T7g10 5′-UTR and a 15-nt sequence encoding an additional five amino acid residues (Met-Ala-Ser-Ile-Ser) linked directly to the VP6 coding region.
Spectinomycin-resistant transplastomic plants were produced by bombardment of 35–43 tobacco leaves with each construct, as described by Svab and Maliga (1993), yielding 10–24 primary transformants for each construct. Primary transformants containing PCR-detectable VP6 in total cellular DNA extracted from the shoots 8–12 weeks after bombardment were taken through two more rounds of selection on spectinomycin and grown to maturity.
DNA-gel blot analysis was carried out on total leaf DNA digested with SacII and EcoRV to confirm insertion of the VP6 expression cassettes into the intergenic region between rbcL and accD in the plastid genome. The 32P-labelled VP6 probe hybridized to a band at 6.5 kb in the DNA from all transplastomic plants, but not to DNA from untransformed wild-type plants (Figure 2). A 6.5-kb band is expected for the insertion of the transforming DNA between rbcL and accD and re-creation of the EcoRV recognition site at the end of the accD region in the transformation vector. The rbcL probe also hybridized to a 6.5-kb band from transplastomic plants and to the expected 2.8-kb band in wild-type plants (Figure 2), confirming insertion between rbcL and accD. The absence of the hybridizing 2.8-kb band in the transplastomic plants indicates that the plants were homoplasmic. The minor hybridizing bands in all the transplastomic plants have been observed previously with pZS197-based vectors (Newell et al., 2003; Birch-Machin et al., 2004; Tangphatsornruang et al., 2011) and are consistent with recombination between the introduced psbA 3′-UTR sequence at the end of the aadA selectable marker and the identical endogenous sequence. This would be expected to generate a 3.6-kb SacII-EcoRV fragment hybridizing to the rbcL probe, and a 3.8-kb EcoRV fragment hybridizing to the VP6 probe, similar to those detected in Figure 2. The origins of other bands hybridizing weakly to the VP6 probe are not known, but they may result from additional recombination events.
VP6 expression in transplastomic tobacco seedlings
Protein immunoblot blot analysis performed on total soluble protein extracted from 10-day-old seedlings of 3–7 individual lines of each of the new VP6 expression constructs showed that individual lines with the same construct contained similar amounts of VP6 protein, detected by a polyclonal anti-VP6 antiserum (Figure 3). Comparison of the amounts of VP6 protein in seedlings containing the VP6 expression constructs with the VP6-AE line 7A (Birch-Machin et al., 2004) and wild-type seedlings showed marked differences in the amounts of VP6 observed in seedlings containing different VP6 expression constructs (Figure 3). The amount of the 55-kDa RbcL (large subunit of Rubisco) in an identical Coomassie-stained gel was used to standardize the amount of VP6, determined from the intensity of the bands on the immunoblot, in the seedling extracts. The amounts of VP6 produced in the seedlings containing the new constructs were expressed relative to the amount in the VP6-AE seedlings (Birch-Machin et al., 2004), where the VP6/RbcL ratio was set at 1.0 (Figure 3b).
Seedlings containing the VP6-CD construct, which contains the T7g10 5′-UTR and adds five amino acid residues to the N-terminus of VP6, produced 3.7 ± 0.9-fold more VP6 than the original VP6-AE line, whereas seedlings containing the VP6-AD construct, which contains the rbcL 5′-UTR and encodes VP6 with its native N-terminus, produced approximately 10× less VP6 than the original VP6-AE line. The lines with the VP6-BE and VP6-BD constructs were intermediate between these extremes. There appeared to be two influences on the amount of VP6 protein in the seedlings: constructs containing the T7g10 5′-UTR produced more protein than constructs containing the rbcL 5′-UTR and constructs encoding VP6 proteins with additional amino acids at the N-terminus produced more protein than comparable constructs encoding the native VP6 protein. The native protein, produced by constructs VP6-AD and VP6-BD, migrated as a protein with a molecular mass of 46 kDa, whereas the protein with two additional amino acid residues at the N-terminus, produced by the VP6-AE and VP6-BE constructs, migrated mainly as a protein with a molecular mass of 40 kDa. The smear of lower-mobility protein above the main 40-kDa band in the VP6-AE and VP6-BE extracts suggests possible degradation. The broad band of VP6 protein in the extract of the VP6-CD seedlings also suggests possible degradation of a lower-mobility form.
Although all the VP6 constructs contain the tobacco plastid rrn promoter and the E. coli rrnB 3′ UTR, the effect of changes in the 5′-UTR and the 5′ end of the coding region on transcript abundance was examined by RNA-gel blot analysis of total RNA extracted from 10-day-old transplastomic seedlings (Figure 4). Total RNA extracted from 200 mg of seedlings was separated by electrophoresis in a 0.8% agarose gel under denaturing conditions and transferred to a nylon membrane by capillary blotting. Hybridization with a 32P-labelled probe for VP6 detected a close doublet of 1.4–1.6 kb in each of the transplastomic lines (Figure 4). This doublet has been observed previously in lines containing the E. coli rrnB 3′ UTR (Newell et al., 2003; Birch-Machin et al., 2004; Zhou et al., 2008) and is because of the presence of two termination sites in the rrnB 3′ UTR (Amann and Brosius, 1985). The abundance of the VP6 transcripts was assessed in relation to the intensity of the signal from hybridization of the same membrane with a 32P-labelled probe for plastid 16S rRNA (Figure 4).
There was marked variation in the amounts of VP6 transcripts accumulated in seedlings containing different VP6 expression constructs (Figure 4b). VP6 transcript accumulation in VP6-CD lines was twice that in VP6-AE seedlings. VP6-BE and VP6-BD lines also accumulated more VP6 transcripts than VP6-AE seedlings. As all the VP6 expression cassettes contain the same promoter (Prrn) and terminator (TrrnB), and the native coding region of VP6 is common to all constructs, the differences in VP6 transcript accumulation are likely to be caused by the different sequences in the 5′-UTRs and around the 5′ end of the coding region.
VP6 expression in leaves of transplastomic tobacco plants
Protein immunoblot analysis was carried out on extracts of all the leaves of 44-day-old plants to investigate the accumulation and stability of VP6 in leaves (Figure 5). Birch-Machin et al. (2004) showed that the amounts of VP6 protein declined with age in the leaves of VP6-AE plants, with VP6 present in the youngest leaves of the plant, but not in the older leaves. Total soluble protein (TSP) was extracted from 100 mg of leaf tissue from the tip of all eight leaves of 44-day-old plants containing each of the VP6 expression constructs and was subjected to SDS-PAGE and blotting onto nitrocellulose membranes. Protein loading was assessed from Coomassie-stained gels, and the ratio of the intensity of the VP6 signal to the stained 55-kDa RbcL band was used to compare the amounts of VP6 accumulated in different leaves. The VP6/RbcL ratio, standardized to a value of 1.0 for leaf 7 of each plant, is given below the gel images in Figure 5. The leaves were numbered from the bottom of the plant upwards towards the youngest leaf.
VP6 protein was detected in only the youngest leaves of the VP6-AE line, in accord with the results obtained by Birch-Machin et al. (2004). No VP6 was detected in the five oldest leaves (leaves 1–5), and only a faint VP6 band was detected in leaf 6. However, in all the transplastomic lines containing the new VP6 expression constructs, VP6 protein was detected in all the leaves of the plants (Figure 5). Although the youngest leaf (leaf 8) generally contained the highest amounts of VP6, expressed relative to RbcL, comparable accumulation of VP6 was observed in the oldest leaves (leaf 1) of plants containing the VP6-AD, VP6-BE and VP6-CD constructs. The results indicate that the introduced changes to the VP6 expression constructs have resulted in VP6 protein accumulation in older leaves, in comparison with plants containing the VP6-AE construct.
The stability of the VP6 protein was examined by re-sampling leaf 4 of individual transplastomic plants containing each of the VP6 expression constructs over the period of 27–39 days following germination (Figure 6). VP6 accumulation was assessed by protein immunoblot and standardized to the amount of the 55-kDa RbcL protein determined from a Coomassie-stained gel. The VP6/RbcL ratio, standardized to a value of 1.0 for leaf samples at day 27 of each plant, is given below the gel images in Figure 6. This revealed that the amount of VP6 remained essentially constant in leaf 4 of plants containing the VP6-BE, VP6-BD and VP6-CD constructs, whereas the amount of VP6 declined in VP6-AE and VP6-AD plants. A half-time for the disappearance of VP6 was obtained by plotting the natural logarithm of the VP6/RbcL ratio against time. VP6 in the VP6-AE line 7A had a half-time of 3.2 ± 0.4 days over the period of 31–39 days, whereas the half-time in the VP6-AD line 40 was 11.6 ± 1.9 days. These values are not true half-lives of the VP6 protein because no attempt was made to inhibit further synthesis of the VP6 protein during the sampling period. However, they provide an estimate of the rate of disappearance of VP6 from ageing leaves.
Absolute amounts of VP6 in the leaves of the different transplastomic lines were determined by protein immunoblot of total protein extracted from the tips of young leaves of 42- to 58-day-old plants, together with standard amounts of VP6 protein (Figure 7). Figure 7a shows an immunoblot of protein extracted from the 7th leaf of 42-day-old plants, which were chosen because, at this growth stage, leaf 7 was the youngest leaf large enough (∼4 cm2) to sample repeatedly. Similar analyses of the 7th leaf of 44-day-old plants and the 10th leaf of 58-day-old plants are included in Figure 7b. The relative accumulation of VP6 between different transplastomic lines was similar in leaves to that observed previously with 10-day-old seedlings (Figure 3). The greatest accumulation of VP6 was observed in VP6-CD plants and the lowest amount was detected in VP6-AD plants, with the VP6-BE and VP6-BD plants containing intermediate amounts of VP6 (Figure 7). By reference to the standards, the amount of VP6 present in the young leaves of VP6-AE plants was determined to be 3.3 ± 1.6% of total protein. In comparison, Birch-Machin et al. (2004) reported VP6 accumulation to 3% of total soluble protein in 7-day-old seedlings containing the VP6-AE construct. VP6-CD plants accumulated VP6 protein to 16.2 ± 5.9% of total protein, which represents ∼five-fold increase over the amount of VP6 in the original VP6-AE line. VP6-BE and VP6-BD plants contained VP6 protein at 7.8 ± 1.2% and 6.7 ± 1.7% of total protein, >two-fold increases over the amount of VP6 in the original VP6-AE line. The only plants to accumulate less VP6 than VP6-AE plants were VP6-AD plants, which contained VP6 at only 2.0 ± 0.7% of total protein. The three lines accumulating the greatest amounts of VP6 contained constructs with the T7g10 5′UTR, whereas the two lowest accumulating lines (VP6-AE and VP6-AD) contained constructs with the rbcL 5′-UTR. In addition, lines containing constructs encoding VP6 with N-terminal extensions of two amino acid residues (VP6-AE and VP6-BE) accumulated more VP6 than comparable lines (VP6-AD and VP6-BD) encoding the native VP6 protein.
In order to investigate whether changes in VP6 transcript accumulation, previously observed in seedlings (Figure 4), were able to account for differences in the patterns of VP6 protein accumulation in older leaves, RNA-gel blot analysis was carried out on total RNA extracted from the tips of the 4th and 6th leaves of 44-day-old transplastomic plants (Figure 8). A doublet of VP6 transcripts was detected in the RNA extracted from both leaves of all the transplastomic lines, although there were clear differences in the amounts of VP6 transcripts in the different lines (Figure 8). The differences in VP6 transcript abundance between the 4th and 6th leaves of each line were much smaller. The VP6-AE and VP6-BE lines showed very little difference between the 4th and 6th leaves, whereas the VP6-AD, VP6-BD and VP6-CD lines all showed higher accumulation of VP6 transcripts in the 6th leaf compared with the 4th leaf (Figure 8).
The relative abundance of VP6 transcripts in the leaves of different transplastomic lines (Figure 8) was similar to the pattern observed for VP6 transcripts in 10-day-old seedlings (Figure 4). The greatest accumulation of VP6 transcripts in leaves was observed with VP6-CD plants and the fewest VP6 transcripts were observed in the leaves of VP6-AD plants, as observed previously for seedlings. VP6 transcripts were observed in leaves 4 and 6 of VP6-AE plants, even though very little VP6 protein accumulated in these leaves (Figure 5). This indicates that the lack of VP6 protein accumulation in the older leaves of VP6-AE plants was not due to the absence of VP6 transcripts.
Multimers of VP6 protein in chloroplasts
VP6 trimers form the middle layer of the three-layered rotavirus capsid (Gorziglia et al., 1985; Mathieu et al., 2001). Using native PAGE, Birch-Machin et al. (2004) demonstrated the formation of VP6 trimers, which migrated as a single band of ∼135 kDa, in leaves of transplastomic tobacco VP6-AE line 7A. To investigate whether alterations to the N-terminal amino acid residues of VP6 affected assembly of VP6 into trimers, total protein extracts of the seventh leaf of 50-day-old transplastomic plants were separated by native PAGE and VP6 was detected by immunoblot analysis (Figure 9). VP6 was detected in the lanes of all the transplastomic lines and migrated with a molecular mass of ∼135 kDa. This indicates that five additional amino acid residues at the N-terminus do not prevent the assembly of VP6 into trimeric structures. Most of the VP6 protein from VP6-AD plants migrated slightly more slowly than the proteins from the other transplastomic lines, although there was a small amount of protein with the same mobility. The VP6 protein from VP6-AD plants does not contain any additional N-terminal amino acid residues, and this may influence the mobility of the trimers in a nondenaturing gel where charge and shape, rather than just molecular mass, influence electrophoretic mobility. The VP6 protein from VP6-BD plants also does not contain any additional N-terminal amino acid residues and migrates as a broad band, part of which has a mobility similar to the protein in VP6-AD plants. No VP6 was detected at the position expected for monomeric VP6 (∼40–46 kDa) in any of the transplastomic lines, suggesting that assembly into trimers produces the preferred quaternary structure of the VP6 protein.
The accumulation and stability of rotavirus VP6 protein in the leaves of transplastomic tobacco plants can be increased markedly by changes to the 5′-UTR and the 5′ end of the coding region. The inclusion of the 5′-UTR from gene 10 of bacteriophage T7 and a 15-bp sequence encoding five additional amino acid residues at the N-terminus of the VP6 protein produced stable VP6 protein at yields of >15% of total leaf protein. This is about 5× higher than obtained previously with a construct containing the 5′-UTR from the tobacco rbcL gene and two additional amino acid residues at the N-terminus (Birch-Machin et al., 2004) and about 30× higher than achieved by transient expression from viral vectors (O’Brien et al., 2000; Zhou et al., 2010) or in transgenic plants containing a codon-optimized VP6 construct (Dong et al., 2005a,b). These transplastomic plants should therefore be an excellent source of VP6 for oral vaccination to protect against rotavirus infection. Oral administration of leaf extracts containing lower levels of VP6 protein has been shown to induce mucosal IgG and serum IgA and to reduce significantly the symptoms of rotavirus infection in mice (Dong et al., 2005a,b; Zhou et al., 2010).
Two features of the VP6 expression constructs appear to affect the accumulation of VP6 protein in both seedlings and leaves. The first is the 5′-UTR sequence. The inclusion of the T7g10 5′-UTR resulted in increased accumulation of VP6 protein compared with constructs containing the rbcL 5′-UTR (Figures 3 and 7). Comparison of the VP6-BD and VP6-BE constructs (containing the T7g10 5′-UTR) with the VP6-AD and VP6-AE constructs (containing the rbcL 5′-UTR) shows 3.3- to 5.5-fold increases in VP6 protein in seedlings and leaves because of the changed 5′-UTR sequence. The second feature is the presence of additional nucleotides at the 5′ end of the coding sequence, which extend the N-terminal sequence of the protein by two or five amino acid residues. Comparison of the VP6-AE and VP6-BE constructs (containing an EcoRI site encoding two additional amino acid residues) with the VP6-AD and VP6-BD constructs (encoding the native VP6 protein) shows 1.7- to 10-fold increases in VP6 protein in seedlings and leaves because of the presence of the additional nucleotides. The inclusion of 15 nucleotides at the 5′ end of the coding region, in the VP6-CD construct, resulted in 2.4- to 5.0-fold increases in VP6 protein in seedlings and leaves compared with the VP6-BD construct.
The mechanisms responsible for the increased accumulation of VP6 are not easily discerned from the available data, in part because of differences in the amounts of VP6 transcripts accumulated in the different transplastomic lines. Although all the VP6 expression constructs contained the same promoter (Prrn) and terminator (TrrnB) sequences, there were marked differences in the amounts of VP6 transcripts accumulated in seedlings (Figure 3) and in leaves (Figure 7). Plants containing VP6 expression constructs with the T7g10 5′-UTR (VP6-BD and VP6-BE) accumulated more VP6 transcripts than plants with the rbcL 5′-UTR (VP6-AD and VP6-AE), and the addition of 15 nucleotides at the 5′ end of the coding region (in VP6-CD) further increased VP6 transcript accumulation. These differences are most likely due to changes in the stability of the VP6 transcripts, but it is currently not possible to exclude effects on transcription initiation. However, the 5′-UTR and downstream box sequences have been shown previously to influence transcript stability and accumulation in tobacco chloroplasts (Eibl et al., 1999; Kuroda and Maliga, 2001a,b; Zou et al., 2003).
The relative translation efficiencies of different expression constructs can be estimated from the ratio of the relative protein accumulation to the relative transcript abundance for each construct (Eibl et al., 1999). Using this approach with the quantitative data in Figures 3b and 4b for seedlings, and in Figures 6b and 7b for leaves, it appears that each of the elements (5′-UTR, 6-bp (EcoRI site) and 15-bp additions at the 5′ end of the coding region) examined in the VP6 expression constructs affected the translation of the VP6 transcripts. The T7g10 5′-UTR increased the translation efficiency by 3.4 ± 1.3-fold compared with the rbcL 5′UTR, the presence of the EcoRI site increased the translation efficiency 6.3 ± 3.9-fold compared with its absence, and the 15-nt sequence increased the translation efficiency 2.1 ± 1.0-fold compared with its absence. These values are means ± standard errors of all the relevant comparisons in seedlings and leaves, but should be regarded as only indicators of the effects on translation efficiency, because the amounts of protein and RNA were not measured in exactly the same samples of plant material. Although the estimates were based on two different samples of 10-day-old seedlings, the leaf estimates were made on leaf 7 of 42-day-old plants for VP6 protein and on leaf 6 of 44-day-old plants for VP6 transcripts. However, the amounts of VP6 transcripts, relative to those from the VP6-AE construct, were similar in leaf 4 and leaf 6 of 44-day-old plants (Figure 8), providing some justification for estimates of translation efficiency made on the leaves of different ages.
Measurements of the amount of VP6 in a single leaf over a time course of 12 days (Figure 6) indicated that changes to the 5′-UTR and the 5′ end of the coding region affected the apparent stability of the protein. The protein produced from the VP6-AE construct disappeared as the leaf aged, as reported previously (Birch-Machin et al., 2004). A half-time of the disappearance of VP6 protein of 3.2 ± 0.4 days was estimated from measurements on leaf 4 over the period of 32–39 days (Figure 6). It is important to point out that this is not a true estimate of the half-life of the protein, because no attempt was made to prevent the synthesis of new VP6 protein. The disappearance of VP6 is therefore likely to reflect the balance between synthesis and degradation over this period; with this gene construct, the rate of VP6 degradation must have exceeded the rate of synthesis, at least over the last 8 days of the time course (Figure 6). However, the protein produced from the VP6-BE construct, which is predicted to contain two additional amino acid residues at the N-terminus and be identical to the protein from the VP6-AE construct, did not decline over the same 12-day period (Figure 6). This suggests that the rates of VP6 degradation did not exceed the rates of synthesis in the leaves of these plants. This may result from increased rates of synthesis because of the presence of the T7g10 5′-UTR in the VP6-BE plants. However, it cannot be assumed that the rates of synthesis, even from constructs containing the T7g10 5′-UTR, are constant in all leaves; it is possible that the translation efficiency of transcripts changes with leaf age. Translation efficiency of transcripts containing the rbcL 5′-UTR (as in the VP6-AE plants) may decrease with age, whereas the translation efficiency of transcripts containing the T7g10 5′-UTR (as in the VP6-BE plants) may increase with age. Although the amounts of VP6 protein accumulated must be a reflection of the balance between synthesis and degradation, the stability of VP6 protein in the older leaves of plants with the VP6-AD construct, which also contain the rbcL 3′-UTR, is not easily explained. A full understanding of the mechanisms regulating the amounts of VP6 protein accumulated will require detailed individual measurements of rates of synthesis and of degradation of the protein in the leaves of different ages.
The mobility of the protein in SDS-PAGE provides an indication that processing or limited degradation of the VP6 protein occurs in all plants. The plants with the highest rates of VP6 disappearance, from the leaves with the VP6-AE construct, show a predominant band at ∼40 kDa, with only smaller amounts of protein migrating with higher apparent molecular masses. However, the more stable VP6 proteins in the leaves containing the other constructs show bands at ∼46 and ∼43 kDa, in addition to the 40-kDa band. No other higher-mobility VP6-immunoreactive bands were detected in any of the plants, suggesting that further degradation of the 40-kDa protein does not produce any stable intermediate forms. It appears most likely that this initial degradation occurs at the N-terminus, because the N-terminal region of VP6 is exposed in the trimer, whereas the C-terminus is tucked away inside the trimer (Mathieu et al., 2001). In general, the pattern of bands observed in the different transplastomic lines appears to be related to the predicted N-terminus of the protein, although there is some variability in the bands observed in seedlings and different leaves of individual lines (Figures 3, 5 and 6). The native VP6 proteins without any additional amino acid residues, produced from the VP6-AD and VP6-BD constructs, show predominantly one band at 46 kDa, or two bands at 46 and 40 kDa, whereas VP6 proteins with two additional residues, produced from the VP6-AE and VP6-BE constructs, show more of the 40-kDa band (Figures 3 and 5). However, additional 43- and 46-kDa bands in different proportions were observed in the extracts of the VP6-BE plants (Figures 5 and 6). The VP6 protein with five additional amino acid residues, from the VP6-CD construct, shows the lowest mobility band, as expected from the higher molecular mass of the protein. This is most clearly seen in the seedling extracts (Figure 3a), but in leaf extracts the VP6 protein normally showed a very broad band, although a triplet of bands was observed when lower amounts of protein were loaded (Figure 5).
The processing of the different forms of VP6 does not appear to conform to the N-end rule (Varshavsky, 1996), which has recently been shown to occur in chloroplasts (Apel et al., 2010). The native VP6 protein with an N-terminal sequence MDVLY, produced from constructs VP6-AD and VP6-BD, has a destabilizing amino acid (Asp) following the initiating methionine, although the lowest mobility band predominates in most leaf and seedling extracts. In contrast, the VP6 protein produced from the VP6-AE and VP6-BE constructs, with an N-terminal sequence of MNSMD, appears to be the most rapidly processed form, with little of the lowest mobility band occurring in most extracts, despite asparagine (Asn) being a stabilizing amino acid for GFP accumulation in tobacco chloroplasts (Apel et al., 2010). The VP6 protein with five additional amino acid residues, MASIS, appears to be intermediate in displaying a pattern of three VP6 bands, which may accord with the intermediate position of alanine (Ala), being neither stabilizing nor destabilizing for GFP (Apel et al., 2010). The determination of the N-terminal amino acid residues of the VP6 bands produced from the different constructs may help our understanding of the determinants of this protein processing.
The assembly of the VP6 protein into trimers in the leaves of all transplastomic lines indicates that variation in the N-terminal sequence of the protein is not important for assembly. The absence of any monomeric form of VP6 in any of the plants indicates that alterations to the N-terminus do not provide a kinetic constraint to assembly. It is currently not clear whether the trimeric form of VP6 is more stable to proteolytic degradation than the newly synthesized monomeric forms. More work is needed on the chronology and mechanism of VP6 degradation to determine whether it is possible to prevent any loss of the VP6 protein in tobacco chloroplasts. Notwithstanding this VP6 degradation, the present study has shown that it is possible to manipulate the VP6 expression constructs so that VP6 protein accumulates in all the leaves of transplastomic tobacco plants, providing a ready source of VP6 protein as a vaccine candidate.
Plants of Nicotiana tabacum cv. Petite Havana were grown axenically in Magenta boxes (Sigma-Aldrich, Poole, UK) on solid MS(B5) medium consisting of MS salts (Murashige and Skoog, 1962), B5 vitamins (Gamborg et al., 1968), sucrose (30 g/L) and agar (7 g/L, Duchefa, Haarlem, The Netherlands) at pH 6.0, in a tissue culture room at 24 °C with a 16-h photoperiod of 50 μmol photons/m2/s1. Prior to sowing, seeds were surface-sterilized in 10% bleach for 10 min and washed three times with sterile water. Leaves from 4- to 6-week-old plants were used for bombardment. Plants grown for seed collection only were grown in a growth room on Levington M3 compost mixed with 20% (v/v) vermiculite.
The rotavirus expression cassettes were initially generated in pNtcC1, a derivative of pSP73 in which the EcoRV restriction site was replaced by a polylinker containing restriction sites for NotI, NheI and NcoI (Hibberd et al., 1998). PrrnB, the rrn promoter with T7g10 5′-UTR, was generated by PCR amplification from pZSloxP (Zhou et al., 2008), with primers Prrn-F (5′ ATAAGAATGCGGCCGCGCTCCCCCGCCGTCGTTC), which contains a NotI site, and Prrn-R1 (5′ GCGAATTCATATGTATATCTCCTTCTTAAAGTTAAACAAAATTATTTCTAGTGGGAAACCGTTGTGGTCTCCCGTATCCAAGCGCTTCGATTCGCCCGG), which contains the 62-nt T7g10 5′-UTR sequence (Kuroda and Maliga, 2001a,b) in front of the ATG start codon and an EcoRI site for cloning. PrrnC, the rrn promoter with T7g10 5′-UTR and 15 nt at the 5′ end of the coding region, was generated by PCR on the PrrnB amplification product using primers Prrn-F and Prrn-R2 (5′ GCGAATTCATGGAAATGCTAGCCATATGTATATCTCCTTCTTAAAG), which anneals to the T7g10 5′-UTR sequence and contains the 15-nt downstream sequence (encoding Met-Ala-Ser-Ile-Ser, including the ATG start codon) and an EcoRI site for cloning. The PCR amplification products PrrnB and PrrnC were digested with NotI and EcoRI and inserted into pNtcC1-rrnB, which contains the E. coli rrnB terminator in the SacI and SalI sites (Birch-Machin et al., 2004), producing pNtcC1-PrrnB/TrrnB and PNtcC1-PrrnC/TrrnB.
The 1.2-kb VP6 cDNA was amplified by PCR from plasmid pAT153-VP6 (Tarlow and McCrae, 1990) with forward primer VP6-FD (5′-GATGTCCTGTACTCCTTGTC), for blunt-end cloning, or primer VP6-FE (5′-CATGGAATTCAGATGTCCTGTACTCCTTGTC, for cloning in the EcoRI site, and reverse primer VP6-R (5′-AGGTGAGCTCATTTGACAAGCATGCTTCTAATGG), for cloning in the SacI site. The PCR product obtained with primers VP6-FE and VP6-R was digested with EcoRI and SacI and ligated into pNtcC1-rrn/rrnB (Birch-Machin et al., 2004), pNtcC1-PrrnB/TrrnB and PNtcC1-PrrnC/TrrnB that had been digested with EcoRI and SacI, to give pNtcC1-VP6-AE, pNtcC1-VP6-BE and pNtcC1-VP6-CE. The PCR product obtained with primers VP6-FD and VP6-R was digested with SacI and ligated into pNtcC1-rrn/rrnB (Birch-Machin et al., 2004), pNtcC1-PrrnB/TrrnB and PNtcC1-PrrnC/TrrnB that had been cut with EcoRI, followed by digestion with mung bean nuclease (New England Biolabs, Ipswich, MA) and digestion with SalI, to give pNtcC1-VP6-AD, pNtcC1-VP6-BD and pNtcC1-VP6-CD. The VP6 expression constructs were removed from the pNtcC1 plasmids by digestion with NotI and HpaI and transferred to the tobacco chloroplast transformation vector pZSBB, which contains a polylinker with NotI and HpaI sites in the unique Bsu36I site of pZSloxP (Zhou et al., 2008).
All PCRs were carried out using Kod Hifi Polymerase (Novagen, Feltham, Middlesex, UK) and the cloned products verified by sequencing. All transformation plasmids were verified by digestion with diagnostic restriction enzymes and sequencing across restriction sites used for cloning.
Chloroplast transformation was carried out according to the method of Svab and Maliga (1993), as described by Birch-Machin et al. (2004). Biolistic bombardment of 5-cm-long tobacco leaves with tungsten (M5; Bio-Rad, Hemel Hempstead, UK) or gold particles (M10; Bio-Rad) coated with DNA was carried out using a PDS-1000/He gun (Bio-Rad) with a 1100 psi rupture disc and a vacuum of 28 in Hg. Following bombardment, the leaves were treated as described by Birch-Machin et al. (2004) and transformants selected on NBM medium [MS(B5) medium supplemented with NAA (naphthalene acetic acid; 0.1 mg/L) and BAP (benzyladenine; 1.0 mg/L), pH 5.9)] solidified with Phytagar (7 g/L) containing spectinomycin (500 mg/L). Leaves from shoots that tested positive for VP6 cDNA by PCR were cut into pieces and taken through two more rounds of regeneration before rooting on MS(B5) medium containing spectinomycin (500 mg/L).
Putative transformant shoots were analysed by PCR using the Red-Extract-N-Amp plant PCR kit (Sigma-Aldrich) following the manufacturer’s instructions. In brief, 50 μL extraction solution was added to a 5-mm2 piece of leaf tissue from the putatively transformed shoot and heated at 94 °C for 10 min prior to the addition of 50 μL dilution solution. PCR were carried out using Kod Hifi Polymerase (Novagen) in a Techne thermocycler (Techgene; Krackeler Scientific, Albany, NY). Each 10-μL PCR mixture contained 2 μL extracted DNA, 5 μL Red-Extract-N-Amp PCR mix, 0.4 μL each of the forward and reverse primers and 2.2 μL H2O. PCR conditions were 3 min at 94 °C followed by 30 cycles of 30 s at 94°C, 30 s at 58 °C and 40–80 s at 72 °C, followed by 10 min at 72 °C. The VP6 cDNA was detected using primers VP6-F (5′-CTTGATGGGTACGATGTGGC) and VP6-R (5′-AGGTGAGCTCATTTGACAAGCATGCTTCTAATGG), which should amplify a 679-bp region. The presence of aadA was detected using primers aadA-F (5′-ACTATCAGAGGTAGTTGGCG) and aadA-R (5′-ACTACCTTGGTGATCTCGCC), which should amplify a 747-bp region.
DNA-gel blot analysis
The GenElute Plant Genomic DNA Miniprep kit (Sigma-Aldrich) was used for the extraction of genomic plant DNA from leaf tissue, following the manufacturer’s instructions. Plant genomic DNA (1.5 μg) was digested with SacII and EcoRV overnight at 37 °C. Fragments were separated by electrophoresis in a 0.8% (w/v) agarose gel and transferred to a GeneScreen Plus nylon membrane (NEN Life Science Products, Boston, MA) by overnight salt transfer (Southern, 1975). A full-length 1.2-kb VP6 probe was PCR-amplified from pAT153-VP6 (Tarlow and McCrae, 1990) with primers VP6-FD and VP6-R (see above) using Taq polymerase (Bioline, London, UK). A 1.2-kb rbcL probe was excised from pZSBB by digestion with BamHI and AvrII and purified by agarose gel electrophoresis. DNA probes were labelled with [α-32P]dATP by the random primer method of Feinberg and Volgestein (1983), following instructions from the Prime It TmT random primer labelling kit (Stratagene, Santa Clara, CA) manual. The membrane was prehybridized, hybridized with the 32P-labelled probe overnight at 65 °C and washed according to the manufacturer’s instructions. The membrane was wrapped in clingfilm and autoradiographed using a storage phosphor screen and scanned using a Typhoon 8600 Imager (GE Healthcare, Chalfont St Giles, UK).
RNA-gel blot analysis
RNA was extracted from plant tissue using TriPure isolation reagent (Roche, Burgess Hill, UK) following the manufacturer’s instructions. RNA (10 μg) was then separated by electrophoresis in a 1% agarose gel in MOPS-formaldehyde buffer, as described by Lehrach et al. (1977). Following electrophoresis, the RNA was transferred onto a GeneScreen Plus nylon membrane (Perkin Elmer, Waltham, MA) as described by Southern (1975) and Sambrook and Russell (2001). Probes were the full-length 1.2-kb VP6 cDNA used for DNA-gel blot analysis and a 1.4-kb rrn16 probe amplified from tobacco chloroplast DNA with the primers rrn16-1 (5′-TGGAAACGGCTGCTAATACC) and rrn16-2 (5′-ATACCCAACAAGCATTA GCTCC). 32P-labelling of DNA probes for RNA-gel blot analysis was carried out as described for DNA-gel blot analysis. The membrane was prehybridized, hybridized with the 32P-labelled probe overnight at 42 °C and then washed according to the manufacturer’s instructions. The membrane was wrapped in clingfilm and exposed to a storage phosphor screen before being scanned in a Typhoon 8600 Imager. Band intensities were quantified using ImageQuant TL software (GE Healthcare).
Protein immunoblot analysis
Total soluble protein (TSP) extracts from plant tissue were prepared by grinding 5–150 mg of tissue in 50–150 μL extraction buffer (50 mm HEPES–KOH pH 7.5, 1 mm EDTA, 10 mm potassium acetate, 5 mm magnesium acetate, 1 mm DTT, 2 mm PMSF and 10 mm sodium metabisulphite) with a spatula tip of acid-washed sand in a microfuge tube (Khan and Maliga, 1999). The debris was pelleted by a 10-min centrifugation at 12 800 g at 4 °C, and the supernatant was transferred to a fresh tube. Protein was quantified by Bradford assay (Bradford, 1976). For the preparation of total protein (TP) extracts from leaf tissue, a 1-cm2 disc of leaf tissue was ground directly in 4× protein loading buffer (0.5 m Tris–HCl pH 6.8, 20% (w/v) SDS, 10% (v/v) glycerol, 5% (v/v) 2-mercaptoethanol and 0.025% (w/v) bromophenol blue) with a spatula tip of acid-washed sand in a microfuge tube and was heated at 95 °C for 5 min.
SDS-polyacrylamide gel electrophoresis (SDS-PAGE) of proteins was performed following the method described by Laemmli (1970) with a Dual Mini Slab Chamber (ATTO, Tokyo, Japan) using a 12% resolving gel and a 5% stacking gel. Nondenaturing polyacrylamide gel electrophoresis was carried out at 4 °C in a 0.75-mm-thick 10% resolving gel and 5% stacking gel, using the buffers of Davis (1964). The proteins were then either stained with Coomassie brilliant blue or transferred to a Hybond nitrocellulose membrane (GE Healthcare) using a semi-dry blotting apparatus (Yrdimes Semidry Transfer System; Wealtec Corporation, Sparks, NV). The membrane was incubated with rabbit anti-VP6 (provided by Malcolm McCrae; Department of Biological Sciences, University of Warwick, UK), diluted 1 : 2000, and donkey anti-rabbit IgG linked to horseradish peroxidase (NA934; GE Healthcare), diluted 1 : 5000, prior to detection with enhanced chemiluminescence (ECL Detection Hyperfilm; GE Healthcare). The film was then developed using the image system X-OGraph Compact X2 (Xograph Imaging Systems, Tetbury, UK). For the quantification of band intensities on stained gels and Western blots, images were obtained using an HP PSC 1350 scanner (Hewlett Packard, Palo Alto, CA), saved as jpeg files and analysed using GeneTools version 4.0 software (Syngene, Cambridge, UK). Area scans, which fully encompassed all the VP6-immunoreactive material or the stained RbcL band, were carried out with background subtraction based on a rolling ball algorithm. The integrated signal intensity was used as the quantitative output.
We thank Malcolm McCrae (University of Warwick) for the VP6 cDNA and antiserum, Jesus Agustin Badillo-Corona for help in making the constructs and Christine Newell, Amanda Cottage, Juliette Jouhet and Sue Aspinall for help and advice. This work was supported by grants from the Studienstiftung des deutschen Volkes and the Frank Smart Fund (University of Cambridge) to AMIB and from the Gates Cambridge Trust, CONACyT Mexico and the EU PharmaPlanta consortium to NGR.