Team:SZU-China/Results
Results
We extracted RNAs from the leaves of M. micrantha and got the transcriptome information through RNA-seq. The purity and integrity of RNA samples were checked to ensure the quality of transcriptome (Fig.1). After RNA-seq, we analyzed the transcriptome using many bioinformatical tools, like NCBI, Uniprot and so forth, and also made a heat map to display the expression level of different genes (Fig.2).
Fig.1 Extracted RNAs from the leaves of M. micrantha
Fig.2 Excerpt of heat map
At the very beginning of our project, we doubted that whether the RNAi technology could be used to control M. micrantha, which has such strong survivability that there were almost no effective herbicides that could remove it in just a few days [1]. Hence, we decided to use the most traditional RNAi approach-dsRNA approach. The double-strand RNAs (dsRNAs) targeting different kinds of respiration related genes were synthesized, and introduced into M. micrantha, trying to block its normal metabolism, and destroy it. The results are as follows.
We selected 6 respiration related genes with relatively high expression levels as RNAi target genes (Table 1), These genes encode citrate synthase-related proteins. ~500bp dsDNAs fragments of the selected genes were separately ligated into expression vector pET-28a, the resulting vectors were transformed into the E. coli. DsRNAs were generated by IPTG induction and extracted from E. coli (Fig.3)
Table 1. Information of dsRNA targeted respiration related gene
Fig.3 The electrophoresis map of the dsRNA extracted from E. coli
Real-time fluorescence quantitative PCRs were carried out to test effects of the dsRNAs on silencing their target genes. DsRNA samples and water were introduced into the experiment and control groups respectively. Results showed that these dsRNAs were effective in silencing their targeting genes (Fig.4). From the figure, we could see that the expression levels of these genes were significantly reduced on the very first day, and were reduced to zero at day 3 after treatment.
Fig. 4 The Relative Target Gene Expression after using dsRNA
Besides, dsRNA treated plants exhibit pleiotropic morphology, including yellowish and wilted leaves, which indicated that dsRNA induced gene silencing did impair the normal growth of M. micrantha (Fig.5). The upper line was control group, and the lower line was experiment group. The other operating conditions of the two groups were the same. The control group were applied the solvent of dsRNA (the sterile water), and the experimental group were sprayed the dsRNA mixture every day. We could clearly see that the leaves of experimental group were turning brown and wilting after about three weeks. However, the efficiency of dsRNA was quite low, and we had to spray for many times to get the wilting morphology.
Fig.5 the morphology of the leaves after testing for three weeks
The results of dsRNA approach indicated that the RNAi technology would possibly be used as the specific herbicide for M. micrantha. However, since the dsRNA will be cut into several small interference RNAs (siRNAs), it is not as specific as one single siRNA, we decided to use a new approach, which was to synthesize RNAi nanoparticles as effectors to carry mutiple copies of siRNA into the weed. After discussing with the experts and trying hard, we finally synthesized the RNAi nanoparticles. The process of synthesis was divided into three steps: single-stranded DNA (ssDNA) phosphorylation, cyclization of ssDNA and rolling circle transcription to form the RNAi nanoparticles (Fig.6) [2].
Fig.6 Self-assembled RNAi nanoparticles
The synthesis results were shown below (Fig. 7, 8, 9).
The gel electrophoresis image of cyclized ssDNA and the RNAi nanoparticles are as follows (Fig.8, 9).
Fig.7 The gel electrophoresis image of cyclized ssDNA
Fig.8 The gel electrophoresis image of RNAi nanoparticles
Fig.9 The SEM image of RNAi nanoparticles
What’s more, after discussing with many professors, we selected the siRNAs that could target different kinds of chlorophyll synthesis related gene, which are essential for the survival of M. micrantha as well as helping us better observe the apparent phenotypes of the tested leaves. qRT-PCRs were carried out to test effects of RNAi nanoparticles on their target gene silencing, results are as follows (Fig.10). We could see that all the target genes were silenced three days after treatment, while the negative control nearly had no change, which showed that the siRNA has high specificity.
Fig. 10 The Relative Target Gene Expression after Using RNAi Nanoparticles
Meanwhile, we traced the phenotype of tested leaves for 10 days and found black spots on the leaves and some of them had bleached margin (Fig.11). The upper row are leaves before RNAi nanoparticle treatment, and the bottom row are leaves 10 days after RNAi nanoparticle treatment. The results were much better than that of dsRNA.
Fig.11 The Morphology of the Leaves after Spraying RNAi Nanoparticles for 10 Days
We designed a sequence containing several regions that could form hairpin siRNAs and considering the difficulty of hairpin structure synthesis, we just designed four different pairs of regions to form four kinds of hairpin siRNAs linking together, which could help us produce 4 functional siRNAs at one time. In this way, we can produce 4 siRNAs for 4 different target genes in large quantities at the same time, so that we could spray RNAi-based herbicide together to increase silencing efficiency. Plasmid pET-28a (+) containing this sequence was constructed and transformed into ht115(DE3) E. coli (Fig.12).
Fig.12 Construction of plasmid with the hairpin siRNA transcription region
The RNAi molecules transcribed in E Coli after IPTG induction were extracted and characterized (Fig.13). We can clearly see that there are bands at 280bp, which are hairpin siRNAs. Meanwhile the stability was tested by gel electrophoresis (Fig.14).
Fig.13 The electrophoresis map of the hairpin siRNA extracted from E. coli
We tested the stability of hairpin siRNA under different conditions:
-
Exposed to the environment (to simulate the environment when sprayed outside) -
Stored at 4℃ (to simulate the worse storage condition)
-
Stored at -20℃ (to simulate the suggested storage condition)
-
Stored at -80℃ (blank control)
As shown in Fig.14, we can see that when the hairpin siRNAs were exposed outside, they were degraded quickly for one day. However, if they are stored at 4 ℃ over ten days, they will be degraded too. Hence, according to the results, we can conclude that Micrancide is stable when stored at the temperature below -20℃
Fig.14 the stability of hairpin siRNA under different conditions
Besides, we applied a G-quadruplex DNA-based, label-free, and ultrasensitive strategy for microRNA detection (click Measurement to see more) and Northern Blot to test the siRNA existed in the leaves after we introduced the hairpin siRNA into M. micrantha. We tested six days after spraying the Micrancide. However, we failed to use Northern Blot to detect the siRNA (Fig.15). Since the siRNAs were too small, Northern Blot could not detect it as expected although we followed the protocol strictly.
Fig.15 The result of Northern Blot (Nothing could be seen)
Then we tried a new method for this test, and we got the results (Fig.16 ). From the figure, we can see that target genes were significantly silenced after treated for 2 days and recovered after about 4 days according to the RT-qPCR results, while the siRNA in vivo increased dramatically after treated for 1 day and then decreased gradually according to the new method results, from which we can conclude that with the increase of siRNA in vivo, the gene expression was down-regulated as expected and the decrease of siRNA led to the recovery of target gene expression. Comparing to the results of RT-qPCR, these results were probably reliable and could truly reflect the change of siRNA inside the leaves to some extent.
Fig.16 The siRNA detection results vs the RT-qPCR results
From the results of qRT-PCR, we could see that all the gene were silenced two days after treatment, while the negative control remained unchanged (Fig.17). These gene sequences targeted on the chlorophyll synthiesis related gene. Besides, we could clearly see that the gene silence effects of hairpin siRNAs were more stable than that of RNAi nanoparticles.
Fig.17 The Relative Target Gene Expression after Using hairpin siRNA
Meanwhile, we had a collaboration with Jiangnan-China team and test the effectiveness of their biosurfactant in promoting hairpin siRNA transaction. We set four tested groups, which were sprayed with surfactant-hairpin-siRNA and hairpin-siRNA-only; and two control group, which were sprayed with sterile-water or surfactant (to test whether surfactant itself would do harm to the plants). Then, we traced the morphology of tested leaves for 10 days and found the leaves were wilting and turning black (Fig.18, 19). The photos were taken after 7 and 10 days after treatment, and they were sprayed with hairpin siRNAs or control solutions every day. The yellowing and wilting leaf phenotypes were much more significant than before. However, there were no significant difference between the surfactant-hairpin-siRNA group and hairpin-siRNA-only group, which meant that the surfactant did no significant help to our herbicide
Fig.18 The Morphology of the Leaves after Spraying Hairpin siRNA every day for 10 Days
Fig.19 Morphological change of leaves for 7 days
We carried out statistical analysis on the average etiolation rate of each group on the tenth day, and the figure below is the result of the statistical analysis (Fig.20). The model we built based on the Matlab script statistics (click Measurement to see more) predicted that M. micrantha would die about 13 days after treatment. And finally, the tested M. micrantha were dead 15 days after treatment in fact (Fig.21), which meant that the model was reliable to some extent. For more information of Models, click Model.
Fig.20 The Etiolation Rate Model
Fig.21 The results of Micrancide-treated plants
We then sprayed the hairpin siRNA on the Arabidopsis thaliana and Nicotiana tabacum L. to the specificity of the hairpin siRNA. As expected, results showed that the Arabidopsis and Nicotiana tabacum L. were unaffected by the hairpin siRNA (Fig.22, 23).
Fig.22 Specificity detection on Arabidopsis thaliana
Fig.23 Specificity detection on Nicotiana tabacum L.
Meanwhile, we had a collaboration with Jiangnan-China team. They mailed their biosurfactant to us and SZU-China team helped them test the efficiency and safety of it. The result showed that the surfactant was quite safe to the plants but the promotion effect on the herbicide was not significant (Fig.24).
Fig.24 M. micrantha treated with the bio-surfactant
We built a self-cracking mechanisms to guarantee the safety as well as the yield of RNAi molecules, the refractile body (R-body) controlled by the change of pH, for our system, which can poke the plasma membrane of E. coli, causing the inclusion to flow out, so that we can get the RNAi molecules transcribed by E. coli.
We constructed plasmid pET-28a (+) with the cds of R-body and transformed it into E. coli (Fig.25). The R-body would be in a coiled stage when pH7~pH8 and would be unrolled when pH turned to 3. (click BBa_K2912017 to see more)
Fig.25 R-body in pH=3 and pH=6
[1] Zhang, L Y , et al. "Mikania micrantha H. B. K. in China – an overview." Weed Research 44.1(2004).
[2] Lee J B , Hong J , Bonner D K , et al. Self-assembled RNA interference microsponges for efficient siRNA delivery[J]. NATURE MATERIALS, 2012, 11(4):316-322.
[3] Robinson K E , Worrall E A , Mitter N . Double stranded RNA expression and its topical application for non-transgenic resistance to plant viruses[J]. Journal of Plant Biochemistry and Biotechnology, 2014, 23(3):231-237.
[4] Yan L, Yan Y, Pei L, et al. A G-quadruplex DNA-based, Label-Free and Ultrasensitive Strategy for microRNA Detection[J]. Sci Rep, 2014, 4:7400.
[5] Pond FR, Gibson I, Lalucat J, et al. R-body-producing Bacteria.[J]. Microbiol Rev, 1989, 53(1): 25-67.
[6] Matsuoka JI, Ishizuna F, Kurumisawa K, et al. Stringent Expression Control of Pathogenic R-body Production in Legume Symbiontazorhizobium Caulinodans[J]. Mbio, 2017, 8(4): 0-17.
[7] Heruth DP, Pond FR, Dilts JA, et al. Characterization of Genetic Determinants for R Body Synthesis and Assembly in Caedibacter Taeniospiralis 47 and 116.[J]. Journal of Bacteriology, 1994, 176(12): 3559-3567.
