Establishing a CRISPR–Cas-like immune system conferring DNA virus resistance in plants
https://ediovision.blogspot.co.uk/2018/04/establishing-crisprcas-like-immune.html
CRISPR–Cas (clustered, regularly interspaced short palindromic repeats–CRISPR-associated proteins) is an adaptive immune system in many archaea and bacteria that cleaves foreign DNA on the basis of sequence complementarity. Here, using the geminivirus, beet severe curly top virus (BSCTV), transient assays performed in Nicotiana benthamiana demonstrate that the sgRNA–Cas9 constructs inhibit virus accumulation and introduce mutations at the target sequences. Further, transgenic Arabidopsis and N. benthamiana plants overexpressing sgRNA–Cas9 are highly resistant to virus infection. Many archaea and bacteria have adaptive immune systems in which CRISPR RNAs direct Cas endonucleases to cleave foreign nucleic acids on the basis of sequence complementarity1. In recent years, CRISPR–Cas-based systems have been developed as a powerful tool for genome editing. In the widely used CRISPR–Cas9 system, a single guide RNA (sgRNA) directs a Cas9 nuclease to make a sequence-specific double-stranded break (DSB) and modify the targeted DNA sequences2. Geminiviruses are circular single-stranded DNA (ssDNA) viruses that replicate within the nuclei of plant cells, causing serious damage to many dicotyledonous crop plants, including tomato, cassava, cotton, sugar beet and pepper3,4. During geminivirus replication, the ssDNA is converted to a double-stranded DNA (dsDNA) intermediate, from which new ssDNA is generated by rolling-circle replication5. Here, we sought to use CRISPR–Cas9 to specifically target the dsDNA of a geminivirus as a way of inhibiting virus replication and conferring virus resistance to host plants (Fig. 1a). To test our approach, we employed the monopartite-genome geminivirus BSCTV and two of its hosts, Arabidopsis thaliana (Col-0) and N. benthamiana6,7. As BSCTV can be inoculated into both of these hosts by agroinoculation, a pCambia T-DNA carrying 1.8 copies of the BSCTV genome was used8. After Agrobacteriummediated transformation of this pCambia–BSCTV construct into plants, the complete circular virus genome is generated by recombination. The BSCTV genome is small (2.9 kb) and encodes only seven proteins (Supplementary Fig. 1a). To identify highly efficient sgRNA–Cas9 target sites, we chose 43 candidate sites (Supplementary Fig. 1 and Table 1), designed sgRNA to target these sites and constructed 43 pHSN401 vectors9 each containing Cas9 driven by a 2 × 35S promoter and one of the sgRNAs driven by an AtU6 promoter (Fig. 1a). All of the candidate sites are within one of three 300-nt regions in coding or non-coding sequences of the BSCTV genome (Supplementary Fig. 1 and Table 1). We first transiently expressed sgRNA–Cas9 in the vectors of pHSN401–sgRNA together with pHSN401 vector (Cas9 alone) as control in N. benthamiana by agroinoculation. Two days later, we infected the same leaves with the pCambia–BSCTV construct also by agroinoculation (Supplementary Fig. 2). Ten days after that, typical BSCTV symptoms, such as leaf curling, were observed on infected control plants (Fig. 1b). Quantification of virus DNA by qPCR showed that all the sgRNA–Cas9 constructs inhibited virus accumulation in the injected (local) leaves to varying levels (Fig. 1b and Supplementary Table 2). Compared with the control vector, 38 of the 43 constructs reduced viral DNA by over 60%, and 20 constructs reduced it by over 80%. Because geminiviruses can be systemically transmitted within plants, symptoms were also observed in non-inoculated (systemic) leaves of the control plants (Fig. 1b). However, in plants containing the highly efficient sgRNA candidates, such as those targeting the A7, B7 and C3 sites, no severe leaf-curling symptoms were observed in systemic leaves, and virus accumulation in local leaves was reduced by 93, 90 and 97%, respectively (Fig. 1b and Supplementary Table 2). These results suggest that we have established an efficient system for screening suitable sgRNA target sites on the viral genome. In the above experiment, viral infection was achieved by agroinoculation. Because plasmids are double-stranded DNA, it seemed possible that Cas9 might cleave the pCambia–BSCTV plasmid before it generated a single-stranded circular viral genome. To exclude this possibility, we developed a method to directly determine whether sgRNA–Cas9 could inhibit actively replicating virus (Fig. 1c). In this method, we selected the sgRNAs targeting the A7, B7 and C3 sites, which substantially inhibited virus accumulation in the targeting site screening assays (Fig. 1b). Five groups of constructs (experimental vectors) were tested: pHSN401– sgRNA (A7, B7 and C3), pHSN401 (Cas9 alone), pHSN401–A7/ B7/C3%Cas9 (sgRNA alone), pHSN401–mA7/mB7/mC3 (mutated sgRNA) and pHSN401–A7/B7/C3–dCas9 (dead Cas9). The mock vector pHSN401%Cas9%sgRNA (without Cas9 and sgRNA) was used as the control (Supplementary Fig 3a), and we simultaneously injected Agrobacteria containing a given experimental vector and the control vector together with pCambia–BSCTV into individual 30-day-old N. benthamiana leaves (Fig. 1c). The pCambia–BSCTV construct was injected into the top part of the leaf and one of the experimental vectors and the control vector were injected into separate areas of the bottom part of the leaf (Fig. 1c). Because of virus movement, the BSCTV generated by the pCambia–BSCTV constructs in the top part of the leaf should be able to move to the bottom part and replicate. Therefore, if the BSCTV reaching that part of the leaf could be inactivated, this would show that the cutting acts on actively replicating virus rather than on incoming plasmids. Six days after injection, we harvested leaf punches from areas in the bottom of the leaves injected with Agrobacteria containing a given experimental vector or the control vector and quantified the virus in these areas. The qPCR results showed that virus accumulation was reduced by 65, 66 and 70% in the regions containing sgRNA–Cas9 targeting the A7, B7 and C3 sites (Fig. 1c), respectively, while virus accumulation in the regions containing the other experimental vectors remained at a level
CRISPR–Cas (clustered, regularly interspaced short palindromic repeats–CRISPR-associated proteins) is an adaptive immune system in many archaea and bacteria that cleaves foreign DNA on the basis of sequence complementarity. Here, using the geminivirus, beet severe curly top virus (BSCTV), transient assays performed in Nicotiana benthamiana demonstrate that the sgRNA–Cas9 constructs inhibit virus accumulation and introduce mutations at the target sequences. Further, transgenic Arabidopsis and N. benthamiana plants overexpressing sgRNA–Cas9 are highly resistant to virus infection. Many archaea and bacteria have adaptive immune systems in which CRISPR RNAs direct Cas endonucleases to cleave foreign nucleic acids on the basis of sequence complementarity1. In recent years, CRISPR–Cas-based systems have been developed as a powerful tool for genome editing. In the widely used CRISPR–Cas9 system, a single guide RNA (sgRNA) directs a Cas9 nuclease to make a sequence-specific double-stranded break (DSB) and modify the targeted DNA sequences2. Geminiviruses are circular single-stranded DNA (ssDNA) viruses that replicate within the nuclei of plant cells, causing serious damage to many dicotyledonous crop plants, including tomato, cassava, cotton, sugar beet and pepper3,4. During geminivirus replication, the ssDNA is converted to a double-stranded DNA (dsDNA) intermediate, from which new ssDNA is generated by rolling-circle replication5. Here, we sought to use CRISPR–Cas9 to specifically target the dsDNA of a geminivirus as a way of inhibiting virus replication and conferring virus resistance to host plants (Fig. 1a). To test our approach, we employed the monopartite-genome geminivirus BSCTV and two of its hosts, Arabidopsis thaliana (Col-0) and N. benthamiana6,7. As BSCTV can be inoculated into both of these hosts by agroinoculation, a pCambia T-DNA carrying 1.8 copies of the BSCTV genome was used8. After Agrobacteriummediated transformation of this pCambia–BSCTV construct into plants, the complete circular virus genome is generated by recombination. The BSCTV genome is small (2.9 kb) and encodes only seven proteins (Supplementary Fig. 1a). To identify highly efficient sgRNA–Cas9 target sites, we chose 43 candidate sites (Supplementary Fig. 1 and Table 1), designed sgRNA to target these sites and constructed 43 pHSN401 vectors9 each containing Cas9 driven by a 2 × 35S promoter and one of the sgRNAs driven by an AtU6 promoter (Fig. 1a). All of the candidate sites are within one of three 300-nt regions in coding or non-coding sequences of the BSCTV genome (Supplementary Fig. 1 and Table 1). We first transiently expressed sgRNA–Cas9 in the vectors of pHSN401–sgRNA together with pHSN401 vector (Cas9 alone) as control in N. benthamiana by agroinoculation. Two days later, we infected the same leaves with the pCambia–BSCTV construct also by agroinoculation (Supplementary Fig. 2). Ten days after that, typical BSCTV symptoms, such as leaf curling, were observed on infected control plants (Fig. 1b). Quantification of virus DNA by qPCR showed that all the sgRNA–Cas9 constructs inhibited virus accumulation in the injected (local) leaves to varying levels (Fig. 1b and Supplementary Table 2). Compared with the control vector, 38 of the 43 constructs reduced viral DNA by over 60%, and 20 constructs reduced it by over 80%. Because geminiviruses can be systemically transmitted within plants, symptoms were also observed in non-inoculated (systemic) leaves of the control plants (Fig. 1b). However, in plants containing the highly efficient sgRNA candidates, such as those targeting the A7, B7 and C3 sites, no severe leaf-curling symptoms were observed in systemic leaves, and virus accumulation in local leaves was reduced by 93, 90 and 97%, respectively (Fig. 1b and Supplementary Table 2). These results suggest that we have established an efficient system for screening suitable sgRNA target sites on the viral genome. In the above experiment, viral infection was achieved by agroinoculation. Because plasmids are double-stranded DNA, it seemed possible that Cas9 might cleave the pCambia–BSCTV plasmid before it generated a single-stranded circular viral genome. To exclude this possibility, we developed a method to directly determine whether sgRNA–Cas9 could inhibit actively replicating virus (Fig. 1c). In this method, we selected the sgRNAs targeting the A7, B7 and C3 sites, which substantially inhibited virus accumulation in the targeting site screening assays (Fig. 1b). Five groups of constructs (experimental vectors) were tested: pHSN401– sgRNA (A7, B7 and C3), pHSN401 (Cas9 alone), pHSN401–A7/ B7/C3%Cas9 (sgRNA alone), pHSN401–mA7/mB7/mC3 (mutated sgRNA) and pHSN401–A7/B7/C3–dCas9 (dead Cas9). The mock vector pHSN401%Cas9%sgRNA (without Cas9 and sgRNA) was used as the control (Supplementary Fig 3a), and we simultaneously injected Agrobacteria containing a given experimental vector and the control vector together with pCambia–BSCTV into individual 30-day-old N. benthamiana leaves (Fig. 1c). The pCambia–BSCTV construct was injected into the top part of the leaf and one of the experimental vectors and the control vector were injected into separate areas of the bottom part of the leaf (Fig. 1c). Because of virus movement, the BSCTV generated by the pCambia–BSCTV constructs in the top part of the leaf should be able to move to the bottom part and replicate. Therefore, if the BSCTV reaching that part of the leaf could be inactivated, this would show that the cutting acts on actively replicating virus rather than on incoming plasmids. Six days after injection, we harvested leaf punches from areas in the bottom of the leaves injected with Agrobacteria containing a given experimental vector or the control vector and quantified the virus in these areas. The qPCR results showed that virus accumulation was reduced by 65, 66 and 70% in the regions containing sgRNA–Cas9 targeting the A7, B7 and C3 sites (Fig. 1c), respectively, while virus accumulation in the regions containing the other experimental vectors remained at a level
