Prime editing using paired pegRNAs targeting NG- or NGG-PAM in rice

Prime editing (PE) enables precise genome modification, i.e., all 12 types of base substitution, as well as designed insertion and deletion. Previously, we developed an efficient PE system using a pair of engineered pegRNAs (epegRNAs), appending an RNA pseudoknot sequence to the 3'ends of pegRNAs to enhance stability and prevent degradation of the 3'extension. Using a wild-type nSpCas9-based PE system (PE-wt) recognizing an NGG-protospacer adjacent motif (PAM) in this approach, two NGG-PAMs (NGG and CCN) adjacent to the target site are required for targeting by paired pegRNAs; however, this is not the PAM configuration available at most target sites. Using an nSpCas9-NG variant recognizing NG-PAM in PE (PE-NG) can expand applicability. Here, we compare the PE efficiency of PE-wt with paired epegRNAs targeting a distal NGG-PAM <i>versus</i> PE-NG with paired epegRNAs targeting NG-PAMs adjacent to the target site. By introducing substitution and designated deletion mutations into target genes via PE-wt and PE-NG with paired epegRNAs, we demonstrated that PE-wt could edit the target site efficiently despite targeting the distal PAM site when either of the paired epegRNAs for PE-NG targets PGC-PAM. If epegRNAs for PE-NG are designed to recognize NGA and NGT-PAM, there is no significant difference in frequency between PE-NG and PE-wt. These findings indicate that PE efficiency via PE-NG is particularly low at the NGC-PAM in rice.