• Open Access

Transplastomic Nicotiana benthamiana plants expressing multiple defence genes encoding protease inhibitors and chitinase display broad-spectrum resistance against insects, pathogens and abiotic stresses

Authors


Summary

Plastid engineering provides several advantages for the next generation of transgenic technology, including the convenient use of transgene stacking and the generation of high expression levels of foreign proteins. With the goal of generating transplastomic plants with multiresistance against both phytopathogens and insects, a construct containing a monocistronic patterned gene stack was transformed into Nicotiana benthamiana plastids harbouring sweet potato sporamin, taro cystatin and chitinase from Paecilomyces javanicus. Transplastomic lines were screened and characterized by Southern/Northern/Western blot analysis for the confirmation of transgene integration and respective expression level. Immunogold localization analyses confirmed the high level of accumulation proteins that were specifically expressed in leaf and root plastids. Subsequent functional bioassays confirmed that the gene stacks conferred a high level of resistance against both insects and phytopathogens. Specifically, larva of Spodoptera litura and Spodoptera exigua either died or exhibited growth retardation after ingesting transplastomic plant leaves. In addition, the inhibitory effects on both leaf spot diseases caused by Alternaria alternata and soft rot disease caused by Pectobacterium carotovorum subsp. carotovorum were markedly observed. Moreover, tolerance to abiotic stresses such as salt/osmotic stress was highly enhanced. The results confirmed that the simultaneous expression of sporamin, cystatin and chitinase conferred a broad spectrum of resistance. Conversely, the expression of single transgenes was not capable of conferring such resistance. To the best of our knowledge, this is the first study to demonstrate an efficacious stacked combination of plastid-expressed defence genes which resulted in an engineered tolerance to various abiotic and biotic stresses.

Introduction

Plastid transformation technology has been well established and widely utilized in plant transgenic research. In comparison with conventional nuclear gene transformation techniques, plastid engineering offers several potential advantages such as (i) more than 10–100 times greater expression levels than the conventional nuclear transformation system, (ii) a more convenient methodology for transferring multiple genes into plants via gene stacking methods, (iii) elimination of position effects in chloroplasts, which thereby reduces the chances for transgene silencing, (iv) minimal chance for transgene flow by pollen contamination due to maternal inheritance (Verma and Daniell, 2007).

One of the key aspects for the plastid transformation system is to employ a plastidic sequence to exchange exogenous genes into the chloroplast genome via homologous recombination. The usage of genetic markers, which enable the selective enrichment of ptDNA copies, is also a critical component for plastid transformation (Dudas et al., 2012; Maliga and Bock, 2011). One of the most commonly used selection markers in the plastid transformation system is the aadA gene (aminoglycoside-3-adenyltransferase), which enables transgenic plants to inactivate two antibiotics: spectinomycin and streptomycin. At the present time, tobacco has been extensively studied as a model plant for plastid transformation systems (Bock, 2007). There have been many reports of successful plastid transformation, such as the transformation of the Bacillus thuringiensis cry9Aa2 gene into tobacco plastids to generate tolerance to the potato tuber moth (Chakrabarti et al., 2006); Lee et al. (2011) demonstrated resistance to Pectobacterium carotovorum subsp. carotovorum and tobacco mosaic virus (TMV) via expressing two important antimicrobial peptides: retrocyclin-101 (RC101) and protegrin-1 (PG1); and Jin et al. (2012) reported that transplastomic plants expressing the agglutinin gene from Pinellia ternata exhibited broad-spectrum resistance against insects and phytopathogens. Besides the tobacco system, a plastid transformation system has been successfully established in tomato (Ruf et al., 2001), potato (Valkov et al., 2011), soybean (Dufourmantel et al., 2004) and lettuce (Kanamoto et al., 2006; Ruhlman et al., 2010). The potential utility of chloroplast transformation has been recognized for metabolic engineering, bioreactor use for biopharmaceutical production and to drive transgenic strategies with the goal to increase stress tolerance in many crops. Thus far, tomato plants have been successfully engineered via plastid transformation to alter carotenoid biosynthesis and bolster nutritional quality (Apel and Bock, 2009; Wurbs et al., 2007). The investigation on transcription, translation and processing of the heterologous multigene operon in transgenic chloroplast facilitated the progress for multigene engineering via plastids (Quesada-Vargas et al., 2005). Furthermore, a multigene engineering by inserting six transgenes of the entire cytosolic mevalonate pathway into the tobacco chloroplast genome was reported. The result showed that transplastomic plants accumulating higher levels of mevalonate, carotenoids, squalene, sterols and triacylglycerol were obtained (Kumar et al., 2012). Several efforts have also focused on chloroplast engineering to fuel the production of edible vaccines and pharmaceutical proteins (Bock and Warzecha, 2010; Boyhan and Daniell, 2011; Davoodi-Semiromi et al., 2010; Kanagaraj et al., 2011; Lakshmi et al., 2013; Maliga and Bock, 2011). Especially, recent progresses of the expression of exendin-4 (EX4) in tobacco chloroplast, and autoantigens and vaccine antigen in lettuce chloroplast were successfully achieved (Kwon et al., 2013a,b). This indicates that the platform concept by developing chloroplast as bioreactor to produce biopharmaceuticals/antigens is promising in the future (Kwon et al., 2013b). As a result of the aforementioned benefits and successful transplastomic examples, chloroplast transformation has been viewed as a next generation of transgenic technology and its potential utility in plant biotechnology is widely recognized.

Numerous studies have utilized nuclear transformation techniques to increase pest resistance in transgenic plants via the overexpression of defence genes such as chitinase and trypsin inhibitors (Amian et al., 2011; Christou et al., 2006; Outchkourov et al., 2004). Related attempts using plastid engineering technology, however, have not been as commonly employed. In the present study, three defence genes, sporamin, cystatin and chitinase, were selected for chloroplast transformation experiments. They were chosen based on their reported role in plant defence response and had shown potential to enhance plant tolerance in our previous nuclear transformation studies. Sporamin is a storage protein in the tuber of sweet potato (Ipomoea batatas) and is a strong trypsin inhibitor (Yeh et al., 1997a). Its function in pest resistance was confirmed by overexpression studies in tobacco, sugar beets and Brassica (Cai et al., 2003; Chen et al., 2006; Yeh et al., 1997b). CeCPI, a cysteine protease inhibitor (cystatin), was isolated from taro corm (Colocasia esculenta), and its antifungal activity and role in insect resistance was previously demonstrated (Wang et al., 2008; Yang and Yeh, 2005). The gene coding for an endo-chitinase was isolated from Paecilomyces javanicus, and its antifungal activity was also previously demonstrated (Chen et al., 2007). Ectopic expression of CeCPI and a chitinase in transgenic tomato confirmed that overexpression of both genes enhanced nematode resistance (Chan et al., 2010a,b). CeCPI inhibits early root-knot nematode (RKN) infection and the production of nematode offspring. Chitinase can effectively reduce the production of egg masses in Meloidogyne incognita and repress embryonic development. Senthilkumar et al. (2010) reported that overexpression of sporamin and CeCPI in tobacco using gene stacking resulted in dual resistance to both insects and phytopathogens.

The long-term goal of our present research is to use plastid engineering technology to develop new vegetable cultivars with broad-spectrum stress tolerance. The plastid transformation vector (pYEH), comprised of an intergenic region, rrn16 operon promoter (Prrn), ribosome binding site (RBS) and psbA terminator from the brassica genomic DNA, was constructed and utilized in the present study. Transplastomic Nicotiana benthamiana plants expressing these important defence proteins either individually or in three-stacking combinations were produced to evaluate construct expression and efficacy. Transplastomic N. benthamiana simultaneously expressing these three genes exhibited synergistic and enhanced effects compared with plants expressing the genes individually. Collectively, the results included in this study demonstrate that the pyramiding and expression of these three defence genes in the plastome produced multiple and broad-spectrum resistance to insects, phytopathogens and abiotic stresses.

Results

Plastid expression vectors and generation of transplastomic N. benthamiana plants

A novel plastid transformation vector, pYEH-scec, was constructed containing backbone sequences from Brassica oleracea for targeting and integrating the desired genetic elements into the intergenic region between rrn16 and rrn23 (Figure 1), located in the inverted repeats (IRs) region of the plastome (Liu et al., 2007). In pYEH-scec, the selection marker gene (aadA) and genes of interest (sporamin, CeCPI and chitinase) are transcribed as a monocistronic RNA transcript driven by the rrn operon promoter (Prrn). Plasmid DNA was utilized to biolistically transform N. benthamiana leaf explants. MS medium containing 500 mg/L spectinomycin was utilized to select for transplastomic N. benthamiana plants. Twelve transformants were obtained from 33 calli that were induced from 50 bombarded leaf samples (data not shown). Primary transformants were transferred and subcultured three times onto fresh medium containing spectinomycin to obtain homoplasmy of the plastid transformants. After rooting, shoots were transferred to pots. No visible phenotypic changes were observed in T0 plants (data not shown).

Figure 1.

Diagrammatic illustration of the plastid transformation vector pYEH-scec. (a) Schematic representation of the chloroplast flanking sequences used for homologous recombination. (b) pYEH-scec vector is based on the backbone of the plasmid pYEH, which was designed for the integration of foreign genes between trnI and trnA in the inverted-repeat region of the chloroplast genome (ptDNA). Transgenes sporamin, CeCPI and chitinase along with the selection marker gene (aadA) were stacked in order. Prrn, rRNA operon promoter; RBS, ribosome binding site; TpsbA, 3′UTR of psbA gene; aadA, selectable marker gene; probe, indicating the DNA fragment used for determining homoplasmic lines by hybridization.

Molecular characterization of transplastomic plants

To confirm the correct integration of the transgenes into the plastid genome, and to test for homoplasmy, transplastomic T0 plants putatively containing the three stacked genes were randomly selected for Southern blotting (Figure 2a). Total leaf genomic DNA was extracted and digested with BamHI. Two BamHI recognition sites are located in the flanking sequence. Digestion with BamHI created two different-sized DNA fragments, a 4.3-kb fragment from the wild-type DNA and a 8.5-kb fragment from the transplastomic DNA when hybridized to a probe corresponding to trnI/trnA (Figure 1a). Data indicated that all the transplastomic T0 plants (line 2 to line 6) were heteroplasmic, with the exception of line 1 which was homoplasmic (Figure 2a). The strong signal obtained for the 4.3-kb fragment suggested that it was not caused by the presence of promiscuous plastid DNA in the nuclear genome. Based upon these results, T0 seeds of line 1 and line 6 were germinated on MS medium containing 500 mg/L spectinomycin. Green seedlings were obtained from all of the T0 offspring of line 1, whereas offspring of line 6 exhibited many bleached seedlings (~12.8 ± 5.3%) (data not shown). These data further confirmed that line 6 was heteroplasmic. The surviving green seedlings were selected, considered as T1 plants and grown in pots. Genomic DNA was subsequently extracted from leaves of T1 plants and subjected to Southern blot analysis to again check for homoplasmy. As shown in Figure S1, data confirmed that several T1 transplastomic N. benthamiana plants of line 1 were homoplasmic because the 4.3-kb band signal was absent (Figure S1). Several T1 offspring of line 6, such as 3, 5, 6 and 7, displaying a similar homoplasmic genomic pattern were also selected for further study. After fruiting, seeds from the selected line 1 and 6 plants were harvested and germinated again on MS medium containing 500 mg/L spectinomycin. None of the T2 seedlings exhibited any signs of bleaching, and their homoplasmy was confirmed by Southern hybridization, indicating that homoplasmic lines could be obtained after two generations of selection using antibiotic selection (spectinomycin). Similarly, the other transplastomic lines (2–5) were treated in an identical manner to obtain homoplastic plants (data not shown). Subsequently, T2 plants of line 1 and line 6 (#3, 5, 6 and 7) were employed as experimental materials for biochemical and molecular analysis and in stress tolerance assays. RNA transcript abundance was monitored to determine the level of transgene expression. Northern blot analysis showed that all the inserted defence genes were expressed at a higher level in leaves than in roots (Figure 2b,c). More than one transcriptional signal bands were evident for all of the three genes (sporamin, CeCPI and chitinase).

Figure 2.

Southern and Northern blot analyses of transformed plants. (a) Hybridization of BamHI-digested total genomic DNA with the α-32P-labelling trnI/trnA probe. A 4.3-kb signal is representative of wild type, and an approximately 8.5-kb signal is representative of integrated transgene DNA in transplastomic lines. (b) and (c) Hybridization of total RNA with each of three individual α-32P-labelling transgene probes. RNA samples (5 μg) extracted from leaves (b) and roots (c) each were used. Band sizes larger than those expected from the respective transgenes (sporamin 0.7 kb, CeCPI 0.6 kb and chitinase 1.3 kb) indicate a read-through transcriptional event in the gene cassette. WT: wild type; Line 1–6: pYEH-scec-transformed Nicotiana benthamiana.

Protein expression, enzyme activity and immunolocalization

Total soluble protein was extracted from leaf and root tissue and subjected to immunoblot analysis in order to confirm that expressed genes were being translated. Sporamin and CeCPI were immunologically detected in all transplastomic lines using their corresponding antiserum (Figure S2). Densitometer analysis revealed that recombinant sporamin accounted for 0.85–1% of total soluble protein (data not shown). Estimation of the other two proteins suggested that their levels were also similar. Protein levels in leaf plastids, however, were higher than in root plastids (Figure S2c,f).

Protein activity assays were performed using total soluble proteins extracted from leaf tissue. Sporamin in all of the tested transplastomic plants exhibited 70–90% trypsin inhibitory compared with WT where inhibitory activity was not detected (Figure 3a). The cystatin activity in all transplastomic plants exhibited 70–80% cysteine proteinase inhibitory compared with WT where inhibitory activity was not detected (Figure 3b). Additionally, the level of endochitinase activity in transplastomic plants increased between 30- and 40-fold in relative comparison with WT plants (Figure 3c).

Figure 3.

Enzyme activity assay in leaves of the T2 generation of transplastomic plants. (a) and (b) Trypsin inhibitor activity and cysteine protease activity were quantified as inhibition percentage (%) defined and described in Materials and Methods. (c) Measurement of endochitinase activity in transplastomic plants. All measurements of enzyme activity were taken in triplicate, and all of the experiments were repeated three times.

An immunolocalization experiment was conducted using leaf and root tissue to determine the subcellular localization of the recombinant defence proteins. Gold particle labelling was primarily observed in leaf chloroplasts (Figure 4a,b) and root leucoplasts (Figure 4d,e). These results confirmed that the transgenes were highly expressed and that the translated proteins were copiously produced and localized to plastids. Protein levels in root leucoplasts were relatively low compared with chloroplasts (Figure 4d,e), suggesting that gene expression levels were higher in chloroplasts than in leucoplasts.

Figure 4.

Immunogold labelling of sporamin and CeCPI in transplastomic plants. (a) Immunogold labelling of sporamin in a leaf plastid. (b) Immunogold labelling of CeCPI in leaf plastid. (d) Immunogold labelling of sporamin in a root plastid. (e) Immunogold labelling of CeCPI in a root plastid. (c) (f) Negative controls with pre-immune serum. Arrows indicate labelling of targeted proteins within the plastids. P: plastoglobuli; SG: starch granule; T: thylakoid.

Insect resistance in transplastomic plants

Transplastomic plants obtained from line 1 and 6, whose expression of sporamin, cystatin and chitinase in plastids had been confirmed, were selected for insect resistance assays. Previous studies indicated that nuclear transformation of either sporamin or CeCPI in tobacco and cabbage (Brassica oleracea) conferred strong insect, bacterial and fungal resistance (Chen et al., 2006; Senthilkumar et al., 2010; Yeh et al., 1997b). In the present study, the second-instar larvae of Spodoptera litura and Spodoptera exigua were used for the insect resistance bioassay. Seven days after the initiation of the feeding test, the weight of larvae fed detached leaves of two transplastomic plants lines (T2) was significantly less than those that were fed leaves from WT plants. Lethal time (LT50) for larvae of S. litura ranged from 3.4 to 4.7 days, and 3.1 to 4.6 days for S. exigua (Table S1) fed leaves from transplastomic plants, whereas larvae fed leaves from WT plants all survived for at least 7 days. The amount of leaf area consumed by larvae also varied between WT and transplastomic plants. WT plants were severely damaged by S. litura (Figure 5a,b) and S. exigua (Figure 5i,j) after 3 and 5 days of infestation, whereas transplastomic plants exhibited only minor damage to leaves (Figure 5c–f, k–n). The average leaf fresh weight of control plants following 5 days of feeding was much less than in transplastomic plants, indicating that transplastomic plants were more resistant to insect attack (Table S1). Larvae fed transplastomic plant leaves also displayed growth retardation, a brownish body or died, while those fed control leaves were active and alive (Figure 5g,o). Insect resistance in line 1 appeared to be greater than in line 6. These data suggest that insect resistance is correlated with the level of protein activity in homoplasmic plants (Figure 3a–c). A whole plant insect bioassay was also conducted using 30 second-instar larvae of S. litura or S. exigua distributed on each plant. The whole plant feeding test lasted for 4 days. Results were similar to the detached leaf feeding assay. Transplastomic lines were only slightly damaged, whereas control plants were seriously damaged (Figure 5h,p).

Figure 5.

Insect feeding bioassay. (a–h) Detached leaf and whole plant insect feeding assay using Spodoptera litura. (i–p) detached leaf and whole plant insect feeding assay using Spodoptera exigua. Second-instar larvae of S. litura and S. exigua were fed with either T2 transplastomic Nicotiana benthamiana (line 1 and 6) or WT leaves and observed 3 or 5 days after feeding. (*) indicates dead larvae in g and o. Scale bar = 1 cm. Five independent plants of each line were used per assay, and data were obtained from the average value of three assays.

Bacterial resistance in transplastomic plants

The bacterium, Pectobacterium carotovorum ssp. carotovorum (Pcc), which is responsible for causing soft rot disease, was injected into T2 transplastomic plants to assay for bacterial resistance. Veins of the third and fourth leaves were injected with a bacterial suspension (10CFU/mL). After 24 h, Pcc produced an obvious necrotic area in leaves of WT plants, which increased in size with time (Figure 6a). In contrast, transplastomic plant line 1 and 6 exhibited a clear hypersensitive-like response without any further evidence of necrosis after 2 days (Figure 6b,c). Data indicated that the population of Pcc was <5.0 × 102 cfu/cm2 in all T2 transplastomic N. benthamiana, but was 1.0 × 105 cfu/cm2 in WT plants at 2 dpi (Figure 6d).

Figure 6.

Bacterial resistance of T2 transplastomic and WT Nicotiana benthamiana plants. Leaves from intact plants were infiltrated with P. carotovorum ssp. carotovorum (Pcc; 1.0 × 10CFU/mL) and observed 2 days postinfection. (a) WT leaf showing necrotic symptoms at 2 dpi. (b,c) line 1 and 6 leaves showing a hypersensitive – like response. Scale bar = 1 cm. (d) Bacterial population density of Pcc in treated leaves of transplastomic (line 1 and 6) and WT plants. Leaf discs were macerated, and bacterial titres were calculated as the CFUs per leaf disc as determined by serial dilution plating on LB agar. Two leaves of each plant were assayed, and ten independent plants of each line were used in the study, all experiments were repeated three times.

Fungal resistance in transplastomic plants

Resistance to the fungal pathogen, Alternaria alternate, was evaluated in T2 transplastomic N. benthamiana plants using a detached leaf test. An agar disc (0.2 cm diameter) containing mycelia of Alternaria alternata, a widespread pathogen causing leaf spots, rots and blights on many plant species, was placed on the adaxial surface of leaves. Seven days after inoculation, symptoms on transplastomic lines consisted of mild chlorosis or a hypersensitive-like necrosis with little sporulation (Figure 7b,c). In contrast, leaves from WT plants showed extensive areas of necrosis symptoms surrounded by chlorotic halos (Figure 7a). The leaf lesion area in transplastomic plants was significantly restricted compared with WT plants (Figure 7d). Data indicate that transplastomic plants were resistant to fungal pathogen attack by A. alternata.

Figure 7.

Fungal resistance of T2 transplastomic and WT Nicotiana benthamiana plants infected with Alternaria alternata. Detached leaves were inoculated with mycelia of A. alternata and observed 7 days postinoculation (dpi). (a) Wild-type leaf showing typical necrosis symptoms; (b,c) pYEH-scec line 1 and 6 leaves exhibiting an HR-like necrosis. (d) Measurement of lesion area at 7 dpi. The necrosis area measured in WT leaves at 7 dpi was counted as 100%. The necrotic area measured in the independent transplastomic lines was calculated as the ratio of necrosis in transplastomic leaves relative to the necrosis in WT leaves. Five independent plants of each line were used, and all the experiments were repeated three times.

Abiotic stress tolerance in transplastomic plants

Salt and osmotic stress

Recent studies have reported that overexpression of chitinase or cystatin in plants increases abiotic stress tolerance to salt or drought (Dana et al., 2006; Zhang et al., 2008b). Therefore, salt and osmotic stress tolerance was also assayed in the current study. Seven-day-old transplastomic and WT plants were transferred to 1/2 MS medium or 1/2 MS medium amended with either 200 mM NaCl or 3% PEG. No significant differences in root length were observed between WT and transplastomic lines on the control medium (Figure 8a). In the presence of 200 mM NaCl (Figure 8b) or 3% PEG, however (Figure 8c), the root growth was significantly inhibited in WT plants after 14 days, but was only moderately inhibited in the transplastomic lines (Figure 8b–e). Data indicated that the transplastomic plants had a higher level of abiotic stress tolerance and were able to maintain greater root growth activity owing to transgene expression in the leucoplasts of roots (Figure 4).

Figure 8.

Salt and osmotic stress tolerance of transplastomic and WT Nicotiana benthamiana seedlings. (a–c) Seven-day-old T2 transplastomic (line 1 and 6) and seedlings were transferred to half-strength MS solid agar medium either nonamended or amended with 200 mM NaCl, or 3% PEG for 2 weeks. Transplastomic plants exhibited more active growth than WT seedlings in response to the stress treatments. (d,e) Plant biomass measured at 14 days after stress treatment. Every treatment had 3 replicates, and each replicate included 30 seedlings. The experiment was repeated three times.

Effect of methyl viologen (MV) and sodium chloride (NaCl) on leaf bleaching and oxidative damage to membrane lipids

We performed multiple evaluations to assess the protective function of transgenes against the oxidative stress caused by MV and NaCl. Specifically, we performed visual observation of leaf discs, determined chlorophyll content, assessed membrane damage with conductivity measurements and quantified lipid peroxidation as determined by malondialdehyde measurements. All of the measurements were markedly different in transplastomic vs. WT plants. Leaf bleaching in leaf discs of WT plants caused by MV (10 μm) or NaCl (400 mm) treatment 5 days was greater than in leaf discs obtained from transplastomic plants (Figure 9). Membrane damage in leaf discs of WT and transplastomic plants subjected to MV treatment was measured by electrolyte leakage. Results indicated a lower percentage of electrolyte leakage in transplastomic plants compared with WT plants (Figure S3a). Transplastomic plants also exhibited a lower level of lipid peroxidation than WT plants, as determined by malondialdehyde levels (Figure S3b).

Figure 9.

Salt and oxidative stress tolerance in leaf discs of transplastomic and WT Nicotiana benthamiana plants. Leaf discs were obtained from leaves of 4-week-old N. benthamiana plants and floated on solutions of either 400 mM NaCl or 10 μM methyl viologen solution for 5 days. (a and d) WT in NaCl and MV, respectively. (b, c, e, and f) pYEH-scec line 1 and line 6 in NaCl and MV, respectively, showing chlorophyll. (g) chlorophyll content of leaf discs measured 5 days after treatments. Significantly less chlorophyll degradation was observed in transplastomic lines compared with the WT. Triplicates plates were prepared for each plant line (WT, pYEH-scec line 1 and 6). Each Petri plate contained five leaf discs, and the experiment was repeated three times.

A comparison of expression and effect of single gene and gene stacking of sporamin, cystatin and chitinase in transplastomic plants

Stacking of sporamin, cystatin and chitinase genes using the pYEH-scec vector for plastid transformation is a novel approach obtaining transplastomic plants with a broad spectrum of stress tolerance. We were interested to evaluate the level of stress tolerance contributed by the individual genes. As a result, two other constructs, harbouring a single defence gene, pYEH-s (sporamin) or pYEH-ce (CeCPI), were constructed (Figure S4a). Both were individually introduced into the N. benthamiana plastid genome by particle bombardment as previously described. Homoplasmic lines were identified by Southern blot analysis as described above (Figure S4b). Expression levels of the recombinant proteins were verified by Western blot analysis probed with either antisporamin or anti-CeCPI antiserum. The protein level of sporamin (~24 kDa) and CeCPI (~23 kDa) in single-gene transplastomic plants was not significantly different from that obtained in pYEH-scec plants containing all three defence-related genes stacked in a single construct (Figure S2a–c for sporamin; d-f for CeCPI). Likewise, the protein level of sporamin and CeCPI expressed in leaf plastids was higher than in root plastids (Figure S2c,f).

The insect resistance assay indicated that transplastomic plants containing the pYEH-scec construct were the most effective; however, plants containing the pYEH-s construct were similar to those with pYEH-scec and much better than plants with the pYEH-ce construct which only displayed minor resistance to insect attack (Figure S5) (Table S2). These results indicated that sporamin was the contributor to insect resistance in transplastomic plants; however, cystatin and chitinase can enhance insect resistance when they are expressed in combination with sporamin.

The antibacterial assay revealed that both pYEH-scec and pYEH-ce T2 transplastomic plants displayed an HR-like response to bacterial inoculation (Figure S6b,d). pYEH-s transplastomic plants displayed a level of necrosis similar to WT plants (Figure S6c). pYEH-ce had a slightly higher bacterial population than pYEH-scec plants (Figure S6e). These data indicate that the bacterial resistance observed in transplastomic plants was primarily due to CeCPI and a slight contribution by chitinase and no contribution from sporamin.

In the fungal resistance assay, pYEH-ce T2 transplastomic plants displayed faint chlorosis, an HR-like response, and a lesion area that was slightly greater than in line 1 pYEH-scec transplastomic plants (Figure S7). pYEH-s transplastomic plants displayed a necrosis response similar to WT plants (Figure S7c). These data indicate that the fungal resistance was mainly due to CeCPI and chitinase.

The assay of abiotic stress resistance indicated that pYEH-s transgenic lines were as stress-sensitive as WT plants (Figure S8a,c), whereas pYEH-ce transgenic lines exhibited a significant increase in abiotic stress tolerance (Figure S8b,d), and pYEH-scec transgenic lines exhibited the greatest increase in abiotic stress tolerance (Figure S8a–d). A measurement of plant biomass confirmed the results of the phenotypic observations (Figure S8e,f). Fresh weights of plants from the pYEH-scec transgenic lines were higher than those from the pYEH-ce transgenic lines, while fresh weights of plants from the pYEH-s transgenic lines were similar to WT plants (Figure S8e,f). Results indicate that abiotic stress tolerance is mainly due to CeCPI. Chitinase may have a synergistic effect when expressed with CeCPI.

Lastly, the assay of chlorophyll levels in leaf discs treated with either NaCl (400 mM) or MV (10 μm) indicated that discs obtained from pYEH-scec and pYEH-ce transplastomic plants were more tolerant than pYEH-s transplastomic and WT plants (Figure S9a–d,i and Figure S9e–h,i). These data indicate that CeCPI was mainly responsible for the increase in abiotic stress tolerance, whereas sporamin was not effective.

Discussion

Plastid transformation has proven to be an effective technology for improving plant resistance to insects (Jin et al., 2012), diseases (DeGray et al., 2001) and abiotic stresses, such as drought (Zhang et al., 2008a), chilling (Craig et al., 2008) and salt (Kumar et al., 2004). In the present study, the plastid transformation vector, pYEH-scec, containing two plastid genes, rrn16 and rrn23 of B. oleracea, was used to stack three individual gene cassettes to produce a monocistronic structure (Figure 1). Initially, only a few homoplasmic lines were obtained, even after multiple rounds of antibiotic selection (Verma et al., 2008). The low number of homoplasmic plants obtained may have been due to the low homologous recombination frequency of trnI/trnA sequences between Brassica and N. benthamiana in T0 plants (DeGray et al., 2001; Sidorov et al., 1999). Although we were capable of identifying homoplasmic clones, this required a large-scale screening effort of offspring in the successive generation (Figure S1).

Our approach was successful in producing functional proteins of sporamin, CeCPI and chitinase in both leaf chloroplast and root leucoplast. Expression in both tissues was possible due to the use of a Prrn promoter, which is functional in both light and dark environments, to drive transcription (Hanson et al., 2013; Zhang et al., 2012). Generally, multigene engineering strategies in plastids have utilized a polycistronic structure. For example, three tocopherol biosynthetic genes were stacked and used to transform tobacco and tomato where the polycistronic transcripts were processed into translatable monocistronic units through the triggering of IEE (intercistronic expression element) (Lu et al., 2013). In our study, three genes were expressed independently from a monocistronic gene cassette. The three transgenes were translatable and represented up to ~1% of the total soluble protein. Therefore, our study has provided evidence that high levels of functional proteins can be produced from multiple transgenes without the use of a polycistronic gene cassette.

The whole plant and the detached leaf assay of T2 transplastomic lines demonstrated that pYEH-s (sporamin) has higher resistance to S. exigua than lines containing pYEH-ce (CeCPI) (Figure S5). This result indicates that sporamin plays an active role in insect resistance. This observation is in agreement with previous reports (Senthilkumar et al., 2010). Data from the present study also indicate that cystatin and chitinase play lesser roles in insect resistance. In contrast, pYEH-ce was more effective than pYEH-s in producing plants with bacterial (Figure S6) and fungal resistance (Figure S7). This result was consistent with our previous report that CeCPI was functional in impeding microbial growth (Yang and Yeh, 2005). Martinez et al. (2005) have proposed that alterations in the fungal membrane permeability could be the origin of the antifungal properties of plant cystatins. Recently, it has been reported that plants overexpressing cystatin acquire increased abiotic stress tolerance as a result in an alteration of plant stress metabolism by upregulating several stress-related genes (Munger et al., 2012). Plants containing pYEH-scec were more resistant in the antimicrobial and insect feeding tests than either YEH-s or pYEH-ce, indicating that stacking the genes allowed plants to obtain the benefits of all three of the individual defence genes.

An analysis of the impact of salt stress and osmotic stress, as measured by MDA content and ion leakage, revealed that CeCPI played a pivotal function in conferring abiotic stress tolerance (Figures S3, 8 and 9). In contrast, sporamin had little impact on abiotic stress tolerance but functioned primarily in improving insect resistance. The single gene construct of chitinase was not employed and analysed in the present study, but is believed to be capable of enhancing both insect resistance and abiotic stress tolerance when used in the plastid system. It has been reported that plants overexpressing cystatin acquire increased abiotic stress tolerance as a result in an alteration of plant stress metabolism by upregulating several stress-related genes (Munger et al., 2012). Overexpression of chitinase in plants has also been shown to enhance tolerance to both biotic and abiotic stress (Ahmed et al., 2012; Nakamura et al., 2008). However, details pertaining to the protective mechanisms of cystatin and chitinase for plants exposed to abiotic stress still remain unclear.

Although important advances have been made in the application of plastid engineering to improve stress tolerance in plants, a comprehensive understanding of the mechanism by which improved stress tolerance is achieved remains to be elucidated. Despite the noteworthy benefits for this approach, this technology still has limitations to many economic crops. Therefore, there is justifiable need to improve chloroplast transformation of economically important crop species.

In summary, we were successful in demonstrating that the stacking of multiple genes in a single cassette and the use of plastid transformation provide a novel approach for the efficient production of defence proteins and transplastomic plants with an increased tolerance to a broad range of biotic and abiotic stresses. The novel vector used in this study has good potential for improving stress tolerance in economically important crops species.

Experimental procedures

Plant material

Seeds of N. benthamiana were kindly provided by Dr. C.P. Cheng, Institute of Plant Biology, College of Life Science, National Taiwan University, Taiwan. Seeds were sterilized with 1% NaOCl for 10 min, rinsed once with sterile water and then germinated on Murashige and Skoog (MS) basal medium (Murashige and Skoog, 1962) using a 16-h photoperiod (54 μmol/s/m2) at 22 °C.

Transformation vectors

pYEH-scec was designed to express sporamin, CeCPI and chitinase genes in N. benthamiana through plastid transformation. The intergenic region of B. oleracea was amplified as previously described (Liu et al., 2007) and ligated to pUC19 to form the pYEH vector. The multiple cloning site (MCS) from pUC19 was amplified using the forward and reverse primer containing a Pme1 restriction enzyme site and ligated to pGEMT to form the pMCS vector. The Prrn promoter and psbA terminator were both amplified from B. oleracea and compared to the Arabidopsis chloroplast genome sequence. The Prrn promoter was blunt-ligated to the synthetic ribosome binding site (GGAGG) to form the PR promoter. The PR promoter and psbA terminator were blunt-ligated to the pMCS vector to form the pPT vector. The aadA (amplified from the PCR8 vector, Invitrogen), sporamin, CeCPI and chitinase genes were blunt-ligated to pPT to form pA (Prrn:aadA:psbA), pS (Prrn:sporamin:psbA), pCe (Prrn:CeCPI:psbA) and pC (Prrn:chitinase:psbA). The pA, pS, pCe and pC vectors were digested with Pme1 enzyme to release gene cascades and were blunt-ligated to pMCS one by one to form the pMCSA-scec, pMCSA-s and pMCSA-ce vectors. The pMCSA-scec, pMCSA-s and pMCSA-ce vector were digested with Pme1 enzyme to release gene cascades and blunt-ligated to the pYEH vector to form pYEH-scec, pYEH-s and pYEH-ce vectors, respectively.

Plastid transformation and selection of homoplasmic transplastomic N. benthamiana lines

Plastid transformation in N. benthamiana was carried out as previously described (Davarpanah et al., 2009). Fully expanded leaves of in vitro cultured N. benthamiana plants were bombarded with 50 mg of 0.6-μm gold particles (Bio-Rad) coated with 10 μg of plasmid DNA using 1,100 psi rupture discs (Bio-Rad, Hercules, CA). Spectinomycin-resistant lines were selected on RMOP regeneration medium (Svab et al., 1990) containing 500 mg/L spectinomycin dihydrochloride. The independent transplastomic lines were subjected to four additional rounds of regeneration on ROMP/spectinomycin to obtain homoplasmic lines. After root regeneration, plants were transferred to soil and grown in a greenhouse with a 16-h light photoperiod (100–300 μmol/s/m2) at 26 °C day/19 °C night.

Southern and Northern blot analyses

Genomic DNA was extracted using a CTAB DNA mini-prep procedure (Clarke, 2009). Total RNA was extracted utilizing the Pine-Tree protocol (Chang et al., 1993). For Southern blotting, DNA was digested by restriction enzymes, resolved on an agarose gel and transferred onto an Immobilon-N+ membrane (Millipore, Bedford, MA). The gene-specific fragment (trnI/trnA), generated by random primer labelled with α-P32-dCTP [DecaLabel™ DNA Labeling Kit; Fermentas (Thermo Fisher Scientific), Vilnius, Lithuania], was used to hybridize the blotted membranes. Membranes were washed following a standard protocol and exposed on a fluorescent plate for 12 h [Typhoon 9400; Amersham Biosciences (GE Healthcare), Pittsburgh, PA]. Northern blotting was performed as previously described by Chang et al. (2010).

Protein activity assay

Total soluble protein extracted from leaves was assayed for inhibitory activity (Beyene et al., 2006; Botelho-Junior et al., 2008). Activity was quantified as percentage inhibition (%) following the equation described by Chen et al. (2006). For the endo-chitinase activity assay, total soluble protein was assayed as described by Maximova et al. (2006).

Immunolocalization of sporamin and CeCPI in transplastomic N. benthamiana plants

Leaves and roots of pYEH-scec transplastomic line 1 plants were collected at 14 days after germination (DAG) and treated as described in Lin et al. (2012). Sectioned tissues were incubated with antisporamin or anti-CeCPI antiserum. The antisporamin antiserum was obtained from a previous study (Cai et al., 2003), while the anti-CeCPI antiserum was identical to that used by Yang and Yeh (2005). Ultrathin sections were observed on a Philips CM 100 transmission electron microscope (Philips Electron Optics, Cambridge, UK). Controls were performed by the replacement of the primary antibody with pre-immune serum.

Insect bioassays

Insect resistance of T2 plants was assayed in both detached leaves and whole plants using second-instar larvae of S. litura and S. exigua following the method of Senthilkumar et al. (2010). Leaves of similar size and weight were placed in 9-cm Petri dishes containing 2% (w⁄ v) agar and covered with filter paper to retain proper moisture. Six early second-instar larvae of S. litura and S. exigua were placed in the Petri dishes to feed on the leaves. A fresh leaf disc was supplied every 24 h. Larval body weight was measured after a 7-day period of feeding. To determine the lethal time (LT) values, mortality of second-instar larvae of S. litura was checked daily for a period of 18 days. Assay of the insecticidal activity in whole plants was carried out as previously reported (Yeh et al., 1997b). Plants were grown in a greenhouse until plants were approximately 30 cm tall. Thirty early second-instar larvae of S. litura or S. exigua were administered to each plant. Treatments were carried out in duplicate. Larvae were evenly distributed on leaves, and the plants were put in a large plastic box and covered with a nylon mesh.

Antibacterial assays

To assess bacterial resistance in planta, plants were washed with 1% bleach to remove soil, surface-sterilized with 70% alcohol and air-dried. Leaves were then inoculated with bacterial suspensions of Pectobacterium carotovorum ssp. carotovorum (10CFU/mL) by injecting the midvein of the third and fourth leaf. Sterile demineralized water was used as a negative control. Bacterial multiplication was calculated in macerated tissue of T2 transgenic lines and WT plants. Several leaf discs were pooled per treatment in each experiment, homogenized in sterile water and serially diluted, and appropriate dilutions were spread on nutrient broth agar plates. Infiltrated leaves were monitored for disease resistance as reflected by the extent of necrosis after 1 and 2 days.

Antifungal assays

Fungal resistance was assayed as described by Zhang et al. (2009). Mycelia of A. alternata were cultured on potato dextrose agar at 28 °C. When the mycelia reached the edge of the plate, 0.2-cm-diameter agar discs were excised from the edges of growing colonies using a cork borer and placed onto the detached leaves. All leaves were placed on wet filter paper in Petri dishes containing 2% water agar and incubated at 28 °C to permit normal disease development under high humidity. After 7 days, the area of measured necrosis in WT plants was given a value of 100%. The area of necrosis was also measured in the independent transgenic lines and the level of disease resistance calculated as the ratio of necrosis in the independent lines relative to WT.

Salt and osmotic stress

Sterilized N. benthamiana seeds were germinated on half-strength Murashige and Skoog (MS) medium. One-week-old seedlings were then transferred to fresh half-strength MS medium amended with either 200 mM NaCl or 3% polyethylene glycol 6000 (PEG). Experiments were performed in a growth chamber set at 27 °C and a 16-h/8-h light/dark photoperiod.

Leaf discs obtained from 2-week-old plants were floated on a solution of either 400 mM NaCl or 10 μm methyl viologen (MV) in 9-cm Petri dishes. After 5 days of treatment, leaf discs were collected and assayed for chlorophyll content as previously described (Minocha et al., 2009). Alternatively, leaf discs were floated on a solution of 10 μm MV in the light for 48 h, and MDA (malondialdehyde) content and electrolyte leakage were subsequently determined.

Statistical analysis

All data are presented as a mean and standard deviation. Data were subjected to a one-way analysis of variance (anova) using SAS v.9.2 (SAS Institute, Cary, NC). < 0.05 was considered statistically significant.

Acknowledgements

The authors would like to thank Dr. S. H. Kao and Dr. C. C. Tzeng (Department of Biopesticides, Taiwan Agricultural Chemicals and Toxic Substance Research Institute, Taichung, Taiwan) for kindly providing the second-instar larvae of S. litura and S. exigua. We would also like to thank Dr. S.H. Wu (Institute of Plant and Microbial Biology, Academia Sinica, Taiwan) for assistance on the immunogold labelling assay. All authors have declared no existing conflict of interests. Financial support was provided by the National Science Council, Taiwan, to Dr Kai-Wun Yeh under the project NSC-101-2324-B-002-023-CC2, and partly from Council of Agriculture under the project 100AS-1.1.1-AD-Z1(5).

Ancillary