Characterization of transgenic rice expressing fusion protein Cry1Ab/Vip3A for insect resistance

Management of resistance development of insect pests is of great importance for continued utilization of Bt crop. The high-dose/refuge and pyramid (gene stacking) strategy are commonly employed to delay the evolution of insect resistance. Due to the anticipated difficulty for deployment of mandatory refuge for transgenic crops in China, where the size of farmer is quite small, stacking of genes with different modes of action is a more feasible strategy. Here we report the development of transgenic rice expressing a fusion protein of Cry1Ab and Vip3A toxin. Analysis of trypsin proteolysis suggested that the fusion protein is equivalent to the combination of Cry1Ab and Vip3A protein. The transgenic plants expressing the fusion protein were found to be highly resistant to two major rice pests, Asiatic rice borer Chilo suppressalis (Lepidoptera: Crambidae) and rice leaf folder Cnaphalocrocis medinalis (Lepidoptera: Crambidae), while their agronomic performances showed no significant difference compared to the non-transgenic recipient rice. Therefore, the transgenic rice may be utilized for rice pest control in China.

In China, rice (Oryza sativa L.) is the staple food for most people. Although transgenic rice had not been commercially planted in China, researches on Bt transgenic rice has lasted for over 20 years. The transgenic line KMD1 expressing a synthetic cry1Ab gene was highly resistant to eight lepidopteran rice pest species 25 . Another case was Bt shanyou-63 containing a chimeric cry1Ab/cry1Ac gene, which showed high protection against rice leaffolder and yellow stem borer 26 . All these lines were single-toxin Bt events. To date, there is still no report on Bt transgenic rice expressing single Vip3A toxin. As to the high probability of insect resistance, tactics for pest management must be updated. Considering the breeding pattern of rice in China, the exploration on transgenic rice lines fusing two or more toxins seems to be a more convenient method for insect resistance management 27 .
Here we reported the development of a transgenic rice line expressing a fusion protein of Cry1Ab and Vip3A. The truncated and active cry1Ab gene, encoding N-terminal 651 amino acid residues of Cry1Ab, was fused in reading frame to the 5′ end of the synthetic vip3A gene encoding 790 amino acid residues 28 . Proteolysis of the fusion protein by trypsin suggested that it would have an equivalent activity with individual Cry1Ab and Vip3A toxin in insect midgut. Bioassay results on transgenic events revealed that the selected event A1L3 had strong insecticidal activities against two major rice pests in China, Asiatic rice borer Chilo suppressalis (Lepidoptera: Crambidae) and rice leaf folder Cnaphalocrocis medinalis (Lepidoptera: Crambidae). Moreover, the insect resistance trait of A1L3 was found to be stable among plants of different generations. Thus, the transgenic line A1L3 could be a good candidate for rice pest control in China.

Results
Fusion protein expression and its insecticidal activity. The truncated cry1Ab and the full-length vip3A gene were fused by a 24 base-pair nucleotide linker in reading frame to generate C1V3 gene. This fusion gene was inserted into pET28a vector and then transformed into E. coli BL21(DE3) strain for protein over expression. E. coli expressed protein was examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis (SDS-PAGE, Fig. 1). The result showed that the C1V3 protein was expressed at high level as inclusion body. The molecular weight of expressed C1V3 was about 160-kDa as expected. When digested with trypsin, active Cry1Ab and Vip3A protein generated trypsin-resistant core of about 60-kDa and 65-kDa in size, respectively 2,10 . These trypsin-resistant cores are the hallmark of active Bt toxins. A commercial trypsin and the insect midgut juice was prepared to investigate if C1V3 protein retained these trypsin-resistant cores, and then Western blot analysis was performed with Cry1Ab and Vip3A polyclonal antiserum respectively (Fig. 2). We found that C1V3 protein did generate a ~60-kDa Cry1Ab and a ~65-kDa Vip3A trypsin-resistant core, exactly same as individual Cry1Ab and Vip3A were digested by trypsin and insect midgut juice (Fig. 2). Therefore, once the fusion protein is ingested by insects, it could be processed to activated cores and works like a combination of two individual proteins of Cry1Ab and Vip3A. Rice transformation and screening. A binary vector was constructed for rice transformation based on vector pCambia1300. Two functional cassettes were constructed into T-DNA of the vector. One is the fusion Bt gene for insect resistance and the other is the EPSPS gene for glyphosate tolerance (Fig. 3). Approximately 100 transgenic T 0 lines were obtained by agrobacterium-mediated transformation. All of T0 lines was screened by enzyme-linked immunosorbent assay of Cry1Ab (ELISA, Data not shown). Transgenic lines with C1V3 concentration lower than 4.0 μg per gram of fresh leaf weight were discarded. Ten transgenic lines were selected for further analysis with expression level from 4.0 to 28 μg per gram of fresh leaf weight. The T-DNA border sequence of most of these lines were then determined. Based on the insertion site and insect resistance bioassay, a line named A1L3 was selected for further characterization and evaluation.
Protein immunoblot analysis and quantification in generations of transgenic rice. The expression level of T3, T4 and T5 progenies of A1L3 at tilling stage were determined by ELISA method (Fig. 4). Different    tissues of each plant were sampled and measured. The C1V3 protein was expressed at 4.691~5.483 μg/g fresh leaf weight in leaves, 3.005~3.563 μg/g in stems, 1.369~1.780 μg/g in roots and 1.096~1.181 μg/g in seeds. We also found that the C1V3 protein expressed stably in different generation of transgenic rice. We also analyzed C1V3 protein by Western blot analysis (Fig. 5). Six A1L3 samples from different plants of the T5 generation were selected for detection of the fusion protein. We detected the fusion protein in plant leaves and found that it has almost identical size of the E. coli expressed C1V3 protein, about ~162-kDa ( Supplementary  Fig. S1).
T-DNA integration in transgenic rice genome. The insertion site and copy number of T-DNA were analyzed by hiTail-PCR and Southern blot, respectively. Sequencing results suggested that an intact T-DNA was inserted at the Chromosome 3. The Southern blot analysis showed only one band for each enzyme-digested sample, indicating that there was only one T-DNA copy in the rice genome (Fig. 6).  Evaluation of insect resistance of A1L3 plants. Two Table 1). The tested insects were all observed dead after then. The mortality of both neonate larvae fedding on the non-transgenic control was less than 5% throughout the bioassay. The leaves and stems of non-transgenic control infested by C. suppressalis or C. medinalis were significantly damaged.
Agronomic performance of transgenic line A1L3. A1L3 line was planted in the experimental plots for two consecutive years (2016~2017) to evaluate the agronomic traits, including plant height, panicles per plant, grains per panicle, weight per 100-grain and seed-set rate. Compared to the non-transgenic line Xiushui-134, A1L3 demonstrated no significant difference (P > 0.05) from the control plants (Table 2).

Discussion
Pyramiding strategy seems to be a feasible plan for rice pest resistance management in area with very small farm size. In this study, we developed a fusion strategy to express two Bt insecticidal toxins with different modes of action in transgenic rice simultaneously. A fusion gene of cry1Ab/vip3A with linker was utilized. Our study suggested that the fusion protein can be reconstituted into two individual toxins upon trypsin digestion. Western blot analysis revealed that C1V3 could be digested into a ~60-kDa Cry1Ab product and a ~65-kDa Vip3A product respectively, which were identical to that of the individual protein of Cry1Ab and Vip3A. The digested products by commercial trypsin and Asiatic rice borer midgut extraction indicated that C1V3 protein could successfully be activated in vivo. These trypsin-generated products were the hallmark of activated Bt toxins. Therefore, we suggested that C1V3 protein could be successfully activated and work equivalently to individual Cry1Ab and Vip3A. Expressing fusion protein has some advantages over expressing individual proteins. The fusion method equalizes the expression level of the two proteins, and also is more convenient in trait integration into different crop elite lines as no stacking is required. The transgenic rice event A1L3 was selected for possible commercial development after a comprehensive evaluation of about a hundred events on insertion sites, expression levels, insect bioassay and agronomic traits. Nearly 40 events of the T0 transgenic events were discarded due to the low expression level of C1V3 (less than 4.0 μg per gram of fresh leaf weight). The C1V3 expression level of the left events was determined at the range of 4.0~28.0 μg/g-fwt. Integration of T-DNA in rice genome and bioassay on insects would be considered for events selection. Only four events were kept for further evaluation because their T-DNA didn't integrate into any annotated loci of rice genome and the T-DNA copy number was single. A1L3 was not the highest expressers but it already demonstrated high resistance to the two major rice pests, C. suppressalis and C. medinalis. We believe that the expression level in A1L3 is high enough to control the target pests effectively. The amount of the insecticidal protein C1V3 in A1L3 was almost triple of the chimeric Cry1Ab/Cry1Ac protein expressed in previously reported transgenic events Bt Huahui-1, in which the Bt protein was determined at 1.88~2.36 μg/g fresh leaf weight 26 . We did not observe any negative effects on agronomic performances or on disease resistance in A1L3 for two  consecutive years planting in fields. However, we observed in two seasons that the events with significant higher expression of C1V3 appeared to have more disease spots on leaves. We will continue our study to check if there is any relationship between the expression levels of C1V3 and the disease spots in their leaves.

Methods
Fusion gene assembling. The coding sequences of cry1Ab and vip3A gene were both optimized according to the codon bias of monocots and synthesized by Sangon Biotech (Shanghai, China). A BamHI and SmaI restriction site were introduced into start and end codon of cry1Ab respectively, as SmaI and SacI into vip3A while synthesizing. Meanwhile, a 24 base pair nucleotide linker was introduced to conjugate cry1Ab and vip3A gene. The linker encodes a peptide with 8 amino acids (GGAGGAGG). A pair of reverse complementary oligonucleotides linker-F (5′-GGTGGAGCAGGTGGAGCAGGTGGA-3′) and linker-R (5′-TCCACCTGCTCCACCTGCTCCACC-3′) were used to generate linker sequence with 72 °C extension processing for 15 mins in thermocycler. Linker was assembled to the blunt ends of cry1Ab (BamHI-SmaI) and vip3A (SmaI-SacI) fragment. The E. coli expressed protein was digested by trypsin (Trypsin from bovine pancreas, Sigma-Aldrich, Catalog number T8802) and insect midgut juice. Protein suspension was incubated with trypsin in a final concentration of 0.5 μg/μl at 37 °C for 0.5 min, 5 min, 15 min, 30 min, 1 h and 6 h to estimate the digestion process. In our assay, 1 h was enough to convert all of the full-length ~160-kDa protein into presumed ~60-kDa active cores of Cry1Ab and Vip3A. The insect midgut juice was extracted from five-star larvae of Asiatic rice borer. Proteins were incubated with the midgut juice extraction at a ratio of 3:1 (e.g. 75 μl protein suspension with 25 μl midgut juice, 100 μl in total) at 37 °C for 2 h. After incubation, digestion reaction was immediately terminated by protease inhibitor cocktail (MedChemExpress, Shanghai). Western blot analysis was utilized to identify the digestion products using specific polyclonal antisera. All of the polyclonal antisera were personalized and manufactured by GenScript Biotech Corp. (Service No. SC1030, Nanjing, China).

Southern blot analysis.
To detect T-DNA insertion in transgenic rice genome, Southern blot was utilized. Firstly, we amplified the flanking sequence to identity the border sequence of T-DNA using hiTAIL-PCR method 30 . Depending on the flanking sequence, the location of T-DNA on rice genome was found. Thus we predicted the length of DNA containing probe sequence (cry1Ab gene and g10evo gene respectively) digested by appropriate restriction enzymes. For g10evo nucleotide probe, rice genome was digested by four restrict enzymes (XhoI, EcoRI, BamHI and ApaLI) respectively. For cry1Ab nucleotide probe, rice genome was digested by three restrict enzymes (XbaI, PstI and KpnI) respectively. Primers used to amplify the fragments of probes was displayed in Supplementary Table S1. Protocol was demonstrated as Users' manual (Version 13) of DIG High Prime DNA Labeling and Detection Starter Kit II (Roche).
Insect bioassays. Insecticidal activities were evaluated for transgenic rice and carried out as described by Ye et al. 31 . Bioassay would performed 30 replicates for both transgenic samples and non-transgenic control respectively. Insect eggs were provided by Genralpest Biotech (Beijing, China), which colony was reared on artificial diet for 10-20 generations. For bioassay of transgenic rice against Asiatic rice borer C. suppressalis (Lepidoptera: Crambidae), neonate larvae (1-2 hr old) were fed on fresh leaves at stem elongation stage, while non-transgenic rice Xiushui-134 at the same growth period was used as negative control. 75 mm-diam petri dishes with a filter paper on the bottom were prepared for cultivation. For each replicate, 10 larvae were introduced into each petri dish with a piece of 50-60 mm leaf, and 200 μl ddH 2 O was dropped on the filter paper to maintain humidity. Subsequently, parafilm membrane was used to seal the petri dishes to prevent larvae from escaping. All petri dishes were stored in a hermetic box without light at 28 °C. For each replicate of bioassay against rice leaf folder C. medinalis (Lepidoptera: Crambidae), 15 neonate larvae were placed on the living plant leaves of transgenic rice and the non-transgenic control respectively. All samples were placed in the cage in greenhouse.
Evaluating the agronomic performance. Transgenic rice lines were planted in paddy field at the

Data Availability
All data generated or analysed during this study are included in this published article.