CRISPR/Cas9-mediated targeted T-DNA integration in rice.

KEY MESSAGE: Combining with a CRISPR/Cas9 system, Agrobacterium-mediated transformation can lead to precise targeted T-DNA integration in the rice genome. Agrobacterium-mediated T-DNA integration into the plant genomes is random, which often causes variable transgene expression and insertional mutagenesis. Because T-DNA preferentially integrates into double-strand DNA breaks, we adapted a CRISPR/Cas9 system to demonstrate that targeted T-DNA integration can be achieved in the rice genome. Using a standard Agrobacterium binary vector, we constructed a T-DNA that contains a CRISPR/Cas9 system using SpCas9 and a gRNA targeting the exon of the rice AP2 domain-containing protein gene Os01g04020. The T-DNA also carried a red fluorescent protein and a hygromycin resistance (hptII) gene. One version of the vector had hptII expression driven by an OsAct2 promoter. In an effort to detect targeted T-DNA insertion events, we built another T-DNA with a promoterless hptII gene adjacent to the T-DNA right border such that integration of T-DNA into the targeted exon sequence in-frame with the hptII gene would allow hptII expression. Our results showed that these constructs could produce targeted T-DNA insertions with frequencies ranging between 4 and 5.3% of transgenic callus events, in addition to generating a high frequency (50-80%) of targeted indel mutations. Sequencing analyses showed that four out of five sequenced T-DNA/gDNA junctions carry a single copy of full-length T-DNA at the target site. Our results indicate that Agrobacterium-mediated transformation combined with a CRISPR/Cas9 system can efficiently generate targeted T-DNA insertions.

Identifying Factors that Determine Effectiveness of Delivery Agents in Biolistic Delivery Using a Library of Amine-Containing Molecules.

Biolistic transfection is a popular and versatile tool for plant transformation. A key step in the biolistic process is the binding of DNA to the heavy microprojectile using a delivery agent, usually a positively charged molecule containing amine groups. Currently, the choice of the commercial delivery agent is mostly limited to spermidine. In addition, the detailed delivery mechanism has not been reported. To help broaden the selection of the delivery agent and reveal the fundamental mechanisms that lead to high delivery performance, a library of amine-containing molecules was investigated. A double-barrel biolistic delivery device was utilized for testing hundreds of samples with much-improved consistency. The performance was evaluated on onion epidermis. The binding and release of DNA were measured via direct high-performance liquid chromatography analysis. This study shows that the overwhelming majority of the amine library performed at the same level as spermidine. To further interpret these results, correlations were performed with thousands of molecular descriptors generated by chemical modeling. It was discovered that the overall charge is most likely the key factor to a successful binding and delivery. Furthermore, even after increasing the amount of the DNA concentration 50-fold to stress the binding capacity of the molecules, the amines in the library continued to deliver at a near identical level while binding all the DNA. The increased DNA was also demonstrated with a Cas9 editing test that required a large amount of DNA to be delivered, and the result was consistent with the previously determined amine performance. This study greatly expands the delivery agent selection for biolistic delivery, allowing alternatives to a commercial reagent that are more shelf-stable and cheaper. The library also offers an approach to investigate more challenging delivery of protein and CRISPR-Cas via the biolistic process in the future.

An improved biolistic delivery and analysis method for evaluation of DNA and CRISPR-Cas delivery efficacy in plant tissue.

Biolistic delivery is widely used for genetic transformation but inconsistency between bombardment samples for transient gene expression analysis often hinders quantitative analyses. We developed a methodology to improve the consistency of biolistic delivery results by using a double-barrel device and a cell counting software. The double-barrel device enables a strategy of incorporating an internal control into each sample, which significantly decreases variance of the results. The cell counting software further reduces errors and increases throughput. The utility of this new platform is demonstrated by optimizing conditions for delivering DNA using the commercial transfection reagent TransIT-2020. In addition, the same approach is applied to test the efficacy of multiple gRNAs for CRISPR-Cas9-mediated gene editing. The novel combination of the bombardment device and analysis method allows simultaneous comparison and optimization of parameters in the biolistic delivery. The platform developed here can be broadly applied to any target samples using biolistics, including animal cells and tissues.

Comparison of CRISPR-Cas9/Cas12a Ribonucleoprotein Complexes for Genome Editing Efficiency in the Rice Phytoene Desaturase (OsPDS) Gene.

BACKGROUND: Delivery of CRISPR reagents into cells as ribonucleoprotein (RNP) complexes enables transient editing, and avoids CRISPR reagent integration in the genomes. Another technical advantage is that RNP delivery can bypass the need of cloning and vector construction steps. In this work we compared efficacies and types of edits for three Cas9 (WT Cas9 nuclease, HiFi Cas9 nuclease, Cas9 D10A nickase) and two Cas12a nucleases (AsCas12a and LbCas12a), using the rice phytoene desaturase (PDS) gene as a target site. FINDINGS: Delivery of two Cas9 nucleases (WT Cas9, and HiFi Cas9) and one Cas12a nuclease (LbCas12a) resulted in targeted mutagenesis of the PDS gene. LbCas12a had a higher editing efficiency than that of WT Cas9 and HiFi Cas9. Editing by Cas9 enzymes resulted in indels (1-2 bp) or larger deletions between 20-bp to 30-bp, which included the loss of the PAM site; whereas LbCas12a editing resulted in deletions ranging between 2 bp to 20 bp without the loss of the PAM site. CONCLUSIONS: In this work, when a single target site of the rice gene OsPDS was evaluated, the LbCas12a RNP complex achieved a higher targeted mutagenesis frequency than the AsCas12a or Cas9 RNPs.

High-frequency random DNA insertions upon co-delivery of CRISPR-Cas9 ribonucleoprotein and selectable marker plasmid in rice.

An important advantage of delivering CRISPR reagents into cells as a ribonucleoprotein (RNP) complex is the ability to edit genes without reagents being integrated into the genome. Transient presence of RNP molecules in cells can reduce undesirable off-target effects. One method for RNP delivery into plant cells is the use of a biolistic gun. To facilitate selection of transformed cells during RNP delivery, a plasmid carrying a selectable marker gene can be co-delivered with the RNP to enrich for transformed/edited cells. In this work, we compare targeted mutagenesis in rice using three different delivery platforms: biolistic RNP/DNA co-delivery; biolistic DNA delivery; and Agrobacterium-mediated delivery. All three platforms were successful in generating desired mutations at the target sites. However, we observed a high frequency (over 14%) of random plasmid or chromosomal DNA fragment insertion at the target sites in transgenic events generated from both biolistic delivery platforms. In contrast, integration of random DNA fragments was not observed in transgenic events generated from the Agrobacterium-mediated method. These data reveal important insights that must be considered when selecting the method for genome-editing reagent delivery in plants, and emphasize the importance of employing appropriate molecular screening methods to detect unintended alterations following genome engineering.

Cytoplasmic inclusion cistron of Soybean mosaic virus serves as a virulence determinant on Rsv3-genotype soybean and a symptom determinant.

Soybean mosaic virus (SMV; Potyvirus, Potyviridae) is one of the most widespread viruses of soybean globally. Three dominant resistance genes (Rsv1, Rsv3 and Rsv4) differentially confer resistance against SMV. Rsv1 confers extreme resistance and the resistance mechanism of Rsv4 is associated with late susceptibility. Here, we show that Rsv3 restricts the accumulation of SMV strain G7 to the inoculated leaves, whereas, SMV-N, an isolate of SMV strain G2, establishes systemic infection. This observation suggests that the resistance mechanism of Rsv3 differs phenotypically from those of Rsv1 and Rsv4. To identify virulence determinant(s) of SMV on an Rsv3-genotype soybean, chimeras were constructed by exchanging fragments between avirulent SMV-G7 and the virulent SMV-N. Analyses of the chimeras showed that both the N- and C-terminal regions of the cytoplasmic inclusion (CI) cistron are required for Rsv3-mediated resistance. Interestingly, the N-terminal region of CI is also involved in severe symptom induction in soybean.