Achieving High Expression of Cry in Green Tissues and Negligible Expression in Endosperm Simultaneously via rbcS Gene Fusion Strategy in Rice

To allay excessive public concern about the safety of transgenic foods, and to optimize insect-resistant genes expression to delay the evolution of resistance in pests, we developed a promising strategy to fuse the GOI (gene of interest) with OsrbcS (rice small subunit of ribulose bisphosphate carboxylase/oxygenase) in transgenic rice, which acted as a carrier, driven by the OsrbcS native promoter to sequester its expression in green tissues. Using eYFP as a trial, we reported a high-level accumulation of eYFP in green tissue and almost none in the seed and root of the fused construct compared to the non-fused construct. After applying this fusion strategy in insect-resistant rice breeding, recombinant OsrbcS-Cry1Ab/Cry1Ac expressed rice plants conferred high resistance to leaffolders and striped stem borers, among which two single-copy lines possessed normal agronomic performance in the field. Specifically, Cry1Ab/Cry1Ac protein levels in single-copy construct transgenic lines ranged from 1.8 to 11.5 µg g−1 in the leaf, higher than the Actin I promoter-driven control, T51-1, about 1.78 µg g−1 in the leaf, but negligible (only 0.00012–0.00117 µg g−1) in endosperm by ELISA analysis. Our study provided a novel approach to creating Cry1Ab/Cry1Ac-free endosperm rice with a high level of insect-resistant protein in green tissues through the simultaneous usage of the OsrbcS promoter and OsrbcS as a fusion partner.


Introduction
Insect pests present a major threat to crop productivity and cause annual losses in the order of ten million tonnes [1]. Rice, as the staple food crop that feeds half the world's population [2], is subject to three major detrimental lepidopteran insects throughout the entire period of plant growth: yellow stem borer (Tryporyza incertulas Walker), striped stem borer (Chilo suppressalis Walker), and leaffolder (Cnaphalocrocis medinalis Guenec) [3]. Compared to the use of chemical insecticides, biological control using biopesticides derived from Bacillus thuringiensis, such as crystal toxins (Cry) [4], cytolytic toxins (Cyt) [5], vegetative insecticidal proteins (Vips) [6], and secreted insecticidal protein (Sip) [7], is a more environment-friendly and efficient alternative for the control of insect pests. In particular, the Bt proteins in genetically engineered crops are not toxic to people, other vertebrates, or most beneficial insects. The field performance of genetically modified (GM) Bt crops in 2. Results

Comparison of OsrbcS Fused and OsrbcS Unfused Transgenic Rice Plants
To confine the expression of the GOI strictly in the pest-attacked tissues without any reduction in the amount, we prepared two constructs using eYFP as a trial: pOsrbcS:eYFP for OsrbcS unfused (namely RY) and pOsrbcS:rbcS-eYFP for OsrbcS fused (namely RRY), under the control of the rice green-specific promoter pOsrbcS (Figure 1). For each construct, we randomly chose three transgenic lines for further analysis. As shown in Figure 2, the eYFP expression level was determined by qRT-PCR in four different tissues: leaf, stem, root, and seeds. In the green tissues, leaf and stem, the levels of eYFP in RRY and RY transformed lines were comparable (Figure 2A,B). Compared to the CaMV 35S promoters, the RRY plants were just one-fold lower in leaf tissue. However, in the non-green tissues, root and seed, the levels of the eYFP transcripts in the RRY plants were lower than those in the RY plants ( Figure 2C,D). In seeds especially, the levels of the eYFP transcripts in the RRY plants were 1/13-1/7 to the RY plants.
Therefore, our study provided an enlightening approach to meet biosafety standards and tackle the evolution of pest resistance in insect-resistant rice breeding.

Comparison of OsrbcS Fused and OsrbcS Unfused Transgenic Rice Plants
To confine the expression of the GOI strictly in the pest-attacked tissues without any reduction in the amount, we prepared two constructs using eYFP as a trial: pOsrbcS:eYFP for OsrbcS unfused (namely RY) and pOsrbcS:rbcS-eYFP for OsrbcS fused (namely RRY), under the control of the rice green-specific promoter pOsrbcS (Figure 1). For each construct, we randomly chose three transgenic lines for further analysis. As shown in Figure  2, the eYFP expression level was determined by qRT-PCR in four different tissues: leaf, stem, root, and seeds. In the green tissues, leaf and stem, the levels of eYFP in RRY and RY transformed lines were comparable (Figure 2A,B). Compared to the CaMV 35S promoters, the RRY plants were just one-fold lower in leaf tissue. However, in the non-green tissues, root and seed, the levels of the eYFP transcripts in the RRY plants were lower than those in the RY plants ( Figure 2C,D). In seeds especially, the levels of the eYFP transcripts in the RRY plants were 1/13-1/7 to the RY plants.  Similar results were observed under confocal laser scanning analysis. As shown in Figure 3, the positive control line p35S:eYFP exhibited high expression in all tissues (leaf, root, and seed), while the RY plants showed stronger signals in the leaf, but only weak signals in the root and the aleurone cells of seeds (Figure 3 arrow). The RRY plants showed similar fluorescence intensity compared to the RY plants in the leaf tissues, but were barely detected in the roots and seeds, similar to the negative controls, which contained no eYFP protein. Similar results were observed under confocal laser scanning analysis. As shown in Figure 3, the positive control line p35S:eYFP exhibited high expression in all tissues (leaf, root, and seed), while the RY plants showed stronger signals in the leaf, but only weak signals in the root and the aleurone cells of seeds (Figure 3 arrow). The RRY plants showed similar fluorescence intensity compared to the RY plants in the leaf tissues, but were barely detected in the roots and seeds, similar to the negative controls, which contained no eYFP protein.

Generation and Protein Quantification of RRC Lines
According to the above, using OsrbcS as a fusion partner presents high expression in green tissue and reduced expression in seed organs. The validation trial results satisfied the requirements of null expression in endosperm in insect-resistant rice breeding. Therefore, we took this fusion strategy in insect-resistant breeding using the fusion gene Cry1Ab/Cry1Ac, which was characterized before [11]. Transgenic lines (namely RRC) were generated using the EHA105 Agrobacterium-mediated transformation method, and eight RRC lines without significant phenotypic changes were selected based on visual observation in the T 0 generation. The copy number of T-DNA insertion in the transgenic plants was determined by Southern blot analysis. As shown in Figure 4A, there were five single-copy lines. The Cry1Ab/Cry1Ac protein concentrations of these five single-copy lines ranged from 2.54 µg g −1 leaf fresh weight in RRC-12 to 10.19 µg g −1 in RRC-11 at the tillering stage ( Figure 4B), and ranged from 1.80 µg g −1 leaf fresh weight in RRC12 to 11.56 µg g −1 in RRC-1 at the filling stage ( Figure 4B). All lines possessed higher protein levels compared to the elite transgenic insect-resistant rice T51-1 [11], in which the Cry1Ab/Cry1Ac gene was driven by the constitutively active rice Actin I promoter [28] in the leaf, both at the tillering and filling stage ( Figure 4B). In contrast, the protein amount in the stem is around 1 µg g −1 at the tillering and filling stage in both T51-1 and RRC lines ( Figure 4C). However, the Cry1Ab/Cry1Ac protein level was nearly zero in the endosperm. The minimum concentration was only 0.00012 µg g −1 in the RRC-12 line. Even the highest concentration was only 0.00117 µg g −1 in the RRC-3 line ( Figure 4D, Table S1).

Generation and Protein Quantification of RRC Lines
According to the above, using OsrbcS as a fusion partner presents high expression in green tissue and reduced expression in seed organs. The validation trial results satisfied the requirements of null expression in endosperm in insect-resistant rice breeding. There fore, we took this fusion strategy in insect-resistant breeding using the fusion gene Cry1Ab/Cry1Ac, which was characterized before [11]. Transgenic lines (namely RRC) wer generated using the EHA105 Agrobacterium-mediated transformation method, and eigh RRC lines without significant phenotypic changes were selected based on visual observa tion in the T0 generation. The copy number of T-DNA insertion in the transgenic plant was determined by Southern blot analysis. As shown in Figure 4A, there were five single the tillering and filling stage ( Figure 4B). In contrast, the protein amount in the stem is around 1 µg g −1 at the tillering and filling stage in both T51-1 and RRC lines ( Figure 4C). However, the Cry1Ab/Cry1Ac protein level was nearly zero in the endosperm. The minimum concentration was only 0.00012 µg g −1 in the RRC-12 line. Even the highest concentration was only 0.00117 µg g −1 in the RRC-3 line ( Figure 4D, Table S1).

Evaluation of Insect Resistance and Agronomic Performance in the Field
Five transgenic lines above were field tested for their resistance to striped stem borers by artificial and natural infection combined and leaffolders by natural infection. The representative phenotype was shown in Figure 5. Wild-type Nipponbare suffered serious damage caused by the pests, while the RRC lines were nearly unaffected. The statistical results of insect damage are shown in Table 1. It was discovered that 68.03% of tillers of the wild-type Nipponbare were severely damaged by striped stem borers, showing dead hearts or white heads, while only slight damage (2.88% at the highest) was observed in these five transgenic lines. Regarding the damage caused by leaffolders, the number of leaves damaged per plant was 0-0.37, while that of the negative controls was 24 (Table 1). These data thus demonstrated that the RRC transgenic lines were highly resistant to striped stem borers and leaffolders under field conditions.

Evaluation of Insect Resistance and Agronomic Performance in the Field
Five transgenic lines above were field tested for their resistance to striped stem borers by artificial and natural infection combined and leaffolders by natural infection. The representative phenotype was shown in Figure 5. Wild-type Nipponbare suffered serious damage caused by the pests, while the RRC lines were nearly unaffected. The statistical results of insect damage are shown in Table 1. It was discovered that 68.03% of tillers of the wild-type Nipponbare were severely damaged by striped stem borers, showing dead hearts or white heads, while only slight damage (2.88% at the highest) was observed in these five transgenic lines. Regarding the damage caused by leaffolders, the number of leaves damaged per plant was 0-0.37, while that of the negative controls was 24 (Table 1). These data thus demonstrated that the RRC transgenic lines were highly resistant to striped stem borers and leaffolders under field conditions. At last, three of the five single-copy transgenic lines, RRC-1, -2, and -3, were selected for agronomic performance evaluation. The data collected from the field experiment for agronomic performance are shown in Table 2. There were two lines, i.e., RRC-1 and -2 showed no significant differences from the wild-type, indicating that RRC-1 and -2 were the ideal line for developing transgenic rice against lepidopteran pests. All the results above followed the original intention of designing transgenic rice transformants with null expression in the endosperm and high expression Bt protein level in insect-attacked parts. Figure 5. Field performance of RRC transgenic lines against an artificial infestation of striped stem borers and natural infection combined, and natural infestation of leaffolders (red and yellow arrows indicate white heads and fold leaves that were damaged by striped stem borers and leaffolders, respectively). Wild-type Nipponbare was used as the negative control.   At last, three of the five single-copy transgenic lines, RRC-1, -2, and -3, were selected for agronomic performance evaluation. The data collected from the field experiment for agronomic performance are shown in Table 2. There were two lines, i.e., RRC-1 and -2 showed no significant differences from the wild-type, indicating that RRC-1 and -2 were the ideal line for developing transgenic rice against lepidopteran pests. All the results above followed the original intention of designing transgenic rice transformants with null expression in the endosperm and high expression Bt protein level in insect-attacked parts.

Discussion
In this study, we took a strategy to fuse a foreign gene with OsrbcS, which served as a carrier under the control of its native promoter to sequester its expression exclusively in green tissue. The Cry1Ab/Cry1Ac-rbcS recombinant protein retained the toxin activity, conferring high resistance to the insects in the field test even under the harsh conditions where each rice plant was infested with about 80 first-instar larvae of striped stem borers and natural leaffolders, attacking at the filling stage. A similar strategy was proposed in nutrition fortified lysine-rich (LR) transgenic rice, in which the recombination protein composed of a lysine-rich protein (LRP)-encoding region derived from the winged bean and the glutelin (Gt1) of rice was driven by the GT1 promoter. The safety assessment test via animal feeding indicated that no adverse effects were observed in rats fed transgenic rice compared to non-transgenic rice [29,30]. Fusing foreign genes with endogenous genes may be an attractive alternative to sustaining an optimal abundance in an appropriate organ.
Using an rbcS transit peptide coding sequence to import foreign protein into chloroplast has been reported in many insect-resistant transgenic rice [31][32][33]. However, using the full-length OsrbcS as a chaperone has not been reported yet. According to our practice, the fusion construct also exhibited strong resistance to striped stem borers and leaffolders in field experiments. The protein amount assessed by ELISA in green leaf varied between 1.80 and 11.50 µg/g, and was almost 100% effective against striped stem borers and leaffolders. Compared to the previous similar work, our OsrbcS fusion construct showed a higher content of Bt protein in the rice leaves. The evidence was listed as follows: The resistant gene, cry1C*, driven by the rice rbcS promoter, was introduced into Zhonghua 11 (Oryza sativa L. ssp. japonica) by Agrobacterium-mediated transformation. The cry1C* protein level was only from 0.87 to 3.13 µg g −1 in the leaves [21]. The Cry1Ab gene was conjugated to the rice rbcS promoter to express the Bt toxin in specific tissues of transgenic plants. The Cry1Ab protein level was merely 1.66 to 3.31 µg g −1 in the leaves [22]. A novel cry2AX1 gene consisting of a sequence of the cry2Aa and cry2Ac genes driven by the rice rbcS promoter was introduced into a rice cultivar, ASD16. The expression of the Cry2AX1 protein in transgenic rice plants was only from 0.00595 to 0.12240 µg g −1 of fresh leaf tissues [18]. A fusion protein of Cry1Ac and Cry1I-Like driven by a green tissue-specific promoter was introduced into a rice cultivar, Xiushui-134. The expression of the fusion protein in transgenic rice plants was merely 1.050 to 1.510 µg g −1 of fresh leaf tissue [34]. Even when compared to the rice Actin I promoter-driven control, T51-1, all five RRC transgenic lines are higher than the control in the leaves ( Figure 4B). Meanwhile, the protein amount assessed by ELISA in the endosperm ranged between 0.000117 and 0.001173 µg g −1 . The highest amount is lower than the minimum line of 0.0026 µg g −1 , as reported by Ye et al. [21]. In addition, three lines in our study showed no declining trend of the Cry1Ab/Cry1Ac protein concentration between the tillering and filling stages in the leaves. However, this declining trend was reported in previous studies [21]. The OsrbcS fusion protein seemed more stable than the unfused form. It is worth emphasizing again that the expression level of the Cry1Ab/Cry1Ac protein is much higher than our previous study [23] and the other transformants with OsrbcS unfused studies [21,22], since the high-level expression of the Cry1Ab/Cry1Ac protein in leaves and stems is an effective way to delay the evolution of pest resistance.
Here, we tried to explain why the OsrbcS fusion protein seemed more stable than the unfused form. Firstly, OsrbcS may facilitate the recombinant protein translocating to the chloroplast. The chloroplast is a key hub in converting solar energy into chemical energy to sustain life on earth. The number of chloroplast-targeted nuclear-encoded proteins largely exceeds the number of chloroplast-encoded proteins. The proteome of today's chloroplasts consists of ∼3000 proteins, most of which are nucleus-encoded and posttranslationally imported into the organelle [35,36]. Accordingly, we concluded that chloroplasts provide an in-built remarkable capacity to allow foreign protein expression.
Moreover, chloroplast lacks the gene silencing modification that precludes sustainable expression over generations [37]. Secondly, it is well accepted that transgenic proteins suffer degradation more rapidly in free cytosol in comparison to their presence in chloroplasts [38]. A fusion protein between the target protein and rbcS was linked with rbcL to make an L8S8 hexadecameric complex structure [39]. rbcS, which is known to be stably present in the form of a protein complex in a chloroplast, may increase the expression of a target protein in plant cells, thereby highly increasing resistance to protein degradation [40].
The over-expression of insect-resistant proteins might lead to severe negative effects and abnormal morphology to the host [41,42]. This problem can be circumvented by targeting the insect-resistant protein to the chloroplast. The outer and inner envelope membranes of chloroplast encapsulate the heterologous expression limited to the compartment, without affecting the biology of the rest of the cell [43,44]. Regarding the recalcitrant features for the manipulation of cereal crop plastid transformation [45][46][47], our OsrbcS fusion strategy is not only considered a potential avenue for the expression of the foreign protein to develop null expression in endosperm with higher expression in insect-damaged parts of GM rice, but can also be used as a model to regulate the appropriate expression of other GOI by using endogenous genes fused in GM crops.

Plant Materials and Plasmid Construction
Rice cultivar Oryza sativa ssp. japonica cv. Nipponbare was used as a recipient for all transformations included in this study. Fragments were cloned into MCS (multiple cloning sites) in the T-DNA of the commercially available binary vector pSB130 via the restrictionligation cloning method [23]. pSB130 is a standard binary vector, which contains two T-DNA regions and multiple cloning sites (MCS having restriction sites of Hind III, Sal I, Xba I, BamH I, Sma I, Kpn I, Sac I, and EcoR I), and is used to load the target fragments in T-DNA1, while the hygromycin B phosphotransferase (HPT) gene in T-DNA2 functioned as a transformation selectable marker. The restriction sites of T-DNA1 used to construct the pRY, pRRY, and pRRC plasmids are shown in Figure 1. The expression cassette was assembled in three steps. Firstly, the Nos terminator was introduced into the MCS in T-DNA1. Secondly, the GOI (eYFP or Cry1Ab/Cry1Ac fusion gene) was ligated with the Nos terminator. Thirdly, the promoter pOsrbcS or pOsrbcS:rbcS was ligated with the GOI. More specifically, the pOsrbcS sequence (3179 bp) was obtained from the genomic DNA of wild-type Nipponbare by PCR amplification with forward 5 -CCCAAGCTTCACTTAAATTTTGGTGACAGG-3 and reverse 5 -ACGCGTCGACCTCTGCAGCTCACCAAGCTC-3 primers (restriction sites are underlined). The pOsrbcS:rbcS sequence (3807 bp) was obtained from the genomic DNA of wildtype Nipponbare by PCR amplification with forward 5 -CCCAAGCTTCACTTAAATTTTG-GTGACAGG-3 and reverse 5 -ACGCGTCGACGTTGCCACCAGACTCCTCGCA-3 primers (restriction sites are underlined). The Cry1Ab/Cry1Ac fusion gene, consisting of the 1,344 bp encoding the N terminus of Cry1Ab and 486 bp encoding the C terminus of Cry1Ac, was derived from our lab's previous work by PCR amplification with forward 5 -ACGCGTCGACA-TGGAC AACAACTGCAGGCC-3 and reverse 5 -CGGGGTACCTTATTCAGCCTCGAGT-GTTG-3 primers (restriction sites are underlined) [11,23]. A fusion of two toxin genes was used to assure the durability and sustainability of pest resistance.

RNA Extraction and qRT-PCR Analysis
Total RNA was isolated from fresh plant tissues using TRIzol reagent (Takara, Japan). Each sample was a mixture of at least 20 plants. According to the manufacturer's instructions, the first-strand cDNA was generated using the Perfect Real Time Primescript RT reagent (TaKaRa, Japan). Real-time quantitative PCR was performed on an optical 96-well plate with a LightCycler 96 Real-Time PCR System (Roche, Switzerland) using SYBR Premix Ex Taq (TaKaRa, Japan), with three technical replicates. The PCR thermal cycling protocol was as follows: 95 • C for 10 s, followed by 40 cycles at 95 • C for 5 s, and 60 • C again for 30 s. The Actin gene was used as the internal reference to assay the relative expression levels of eYFP. The gene-specific primers used for qRT-PCR are listed in Table S2.

Tissue Localization of pOsrbcS:eYFP and pOsrbcS:OsrbcS-eYFP Line
Seedlings were grown in Yoshida solution for approximately 2 weeks in a growth chamber [48] (temperature 25 • C; humidity 75%; light ≥ 20,000 lux; 12 h of light and 12 h of the night). The leaves and roots of the young seedlings at the 2-3 leaf stage and the cross-section of the newly germinated seeds derived from transgenic eYFP lines were prepared in this study. The fluorescence signals were visualized under confocal microscopy (Zeiss LSM710, Germany ).

Quantitative Estimation of Cry1Ab/Cry1Ac Protein Expressed in Transgenic Plants
A modified procedure using the ELISA kit AP003 CRBS (EnviroLogix, Portland) was implemented to detect the expression level of the Cry1Ab/Cry1Ac protein. Approximately 20 mg of fresh sample from the leaf and stem at the tillering and filling stages, and endosperm at the mature stage, were collected and ground into powder in liquid nitrogen. The powder was then suspended in an extraction buffer according to the 20 mg powder/500 µL extraction buffer proportion and diluted to an appropriate concentration using dilution buffer. The enzyme-linking reaction was performed according to the manufacturer's instructions. The final concentrations of Cry1Ab/Cry1Ac in the samples (µg·g −1 tissue) were calculated based on the standard curve.

Southern Blot
Southern blot analysis was conducted to screen the copy number, and 10 µg of genomic DNA from each sample was digested with Hind III, separated on a 1% agarose gel, and then transferred to a nylon membrane using 20 × SSC solution followed by standard capillary transfer protocol (GM Healthcare, UK). The probes were prepared from a PCR-amplified fragment of Cry1Ab/Cry1Ac (Table S1), which was labeled with DIG-dUTP using a PCR DIG Probe Synthesis Kit (Roche, Switzerland). The transferred DNA was cross-linked by a UV crosslinker at 1200 mJ for 1 min. Hybridization, membrane washing, and subsequent chemical staining with BCIP/NBT (Roche, Switzerland) were conducted as described [49].

Field Tests for Insect Resistance and Agronomic Performance
All the transgenic lines and controls (wild-type Nipponbare) for insect resistance and agronomic traits were planted in the experimental farm of Zhejiang University in Hangzhou, China. The seeds were sown in a seedling bed in early June 2018, and the seedlings were transplanted to the experimental fields approximately 25 days later. The field layout followed a randomized complete block design with three replications. Each plot consisted of 20 plants in 2 rows with distances of 20.0 cm between plants within a row, 15.0 cm between rows, and 40.0 cm between plots. The experimental plots were bordered by three rows of non-transgenic Nipponbare rice plants. In the verification test carried out, no chemical insecticides targeted against lepidopteran pests were applied throughout the growth period. Therefore, natural infestations of leaffolders were obtained. The reaction of the plants to the natural infestation of leaffolders was scored 5 to 7 days after peak damage appeared. Leaves with visible scrapes and tillers with visibly scraped leaves were scored as damaged leaves and tillers, respectively. Both natural and manual infestations of striped stem borers were used. For manual infestations under field conditions, each rice plant was infested with 80 first-instar larvae of striped stem borers at the late booting stage. The insect eggs were bought from China Taobao (https://item.taobao.com/item.htm? spm=a1z10.3-c-s.w4002-24362899300. 10.42566b3cA2jvBM&id=698368036197), accessed on 1 September 2018, and hatched in a plant growth chamber at 25 • C and 75% humidity under a photoperiod of 14 h light/10 h dark for 3-5 days to incubate the first-instar larvae. Damage symptoms were checked 7 to 15 days after infestation. Dead hearts and white heads caused by striped stem borers were counted together to calculate the infection rate.
To test the agronomic performance of the transgenic plants in the field, the plot was designed and used as above, although chemical pesticides were applied throughout the growth period for crop protection. At maturity, the plant height was measured in the field, and all plants from each plot (15 individual plants per line were divided into 3 repetitions) were collected to measure the panicle length, number of panicles per plant, number of grains per panicle, seed-set rate, 1000-grain weight, and yield per plant.  Data Availability Statement: Data will be made available upon request.