Similar to the mammalian RNA editors, we synthesized human m6A demethylase ALKBH5 catalytic domain (66–292 aa) and fused it to the C-terminus of inactive dLwaCas13a (R474A and R1046A)-msfGFP structure using the unstructured 16-residue peptide XTEN as a linker(Fig. 1A). Nuclear localization signal (NLS) peptides were added to the N-terminus of dLwaCas13a and C-terminus of ALKBH5. Catalytically inactive dLwaCas13a can be used as a programmable RNA binding protein. msfGFP was used to enhance the stability of the dLwaCas13a8. We expressed the dLwaCas13a-msfGFP-XTEN-ALKBH5 fusion sequence under the CMV promoter and the CRISPR RNA (crRNA) sequence under the Arabidopsis RNA polymerase III promoter AtU6. RNA polymerase III-transcribed crRNAs carry 3’ terminal poly U sequences, which are immediately adjacent to protospacer sequences involved in RNA recognition12. We thus speculated that sequences of the crRNA might be highly specific and critical for Cas13a activity. To meet the sequence specificity, we also used a double ribozyme system that precisely processes the crRNAs (Fig. 1A)13.
To test the PMEs system, we chose WUS, STM, FT, SPL3, and SPL9 as the target genes (Fig. 1A). Several m6A sites have been identified in the transcript of the five genes5,14. FT, SPL3, and SPL9 are key activator of flowering5. m6A methylation of FT, SPL3, and SPL9 mRNAs affects floral transition. STM and WUS are two key shoot apical meristem (SAM) regulators14. m6A methylation on STM and WUS determines soot stem cell fate in plants. Studies suggested that the secondary structure of crRNA are critical for the editing efficiency of Cas13a/crRNA8. Therefore, selection of target sequences should avoid disrupting the secondary structure of crRNA. In this study, the guide sequences targeting the m6A sites of the chosen genes were predicted with the software RNAfold (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi). In this study, we used two classes of structure. Four units of crRNAs expression cassettes under four AtU6 promoters were ligated in tandem (PME-WS and PME-FSS), and four units of crRNA ribozyme cassettes under one AtU6 promoter were ligated in tandem (PME-WS-H and PME-FSS-H) (Fig. 1A). The four structures together with control vector (PME-MCS) were transformed into Arabidopsis (Col-0) via Agrobacterium Tumefaciens-mediated transformation.
We then verified the effect of PMEs on m6A modification of the selected genes in T3 Arabidopsis transgenic plants. qRT-PCR showed that the four structures targeting the selected genes can increase mRNA levels of the genes compared to the control vector (PME-MCS) (Fig. 1B). Notably, for structures with the double ribozyme system, the mRNA levels of the selected genes were significantly elevated up to 5 times. SELECT qPCR further revealed that PME-FSS and PME-FSS-H decreased the methylation levels of SPL9 mRNA (Fig. 1C-D). Indeed, the PME-FSS and PME-FSS-H lines gave an obvious early-flower phenotype (Fig. 1E-F). Taken together, these data show that targeted demethylation of functional gene transcripts can be efficiently generated using the PMEs system.
It is known that m6A methylation exists and affects RNA stability and translational regulation in plants and animals3. Given that the impact of m6A methylation on specific RNA could not be addressed in a methylase and demethylase activity-dependent manner5, we envisage that the PMEs system will be widely adopted to accelerate plant m6A methylation research. This technology allows for targeted RNA demethylation, thus avoiding board epigenetic changes. In addition, multiplex target RNA editing using a double ribozyme system provides a powerful tool for multiple transcripts and multiple sites in single transcript. A further improvement of this system could be achieved by integrating nuclear export signal (NES), which changes the subcellular location of Cas13a-ALKBH fusion protein and enhances the editing efficiency9,10. Meanwhile, the use of full-length ALKBH5 or a more active RNA demethylase such as Cas13b in plants will also greatly improve the efficiency of demethylation9,10. The PMEs system described here hold great promise to change the game play for future RNA regulation research.