引导编辑（Prime Editing, PE）是一种可实现精准碱基替换、插入和删除的基因组编辑技术，然而其在植物中的应用受到SpCas9对NGG型PAM（Protospacer Adjacent Motif）的严格依赖性所限制。目前，植物中通过将SpCas9替换为SpCas9 PAM变体或者识别不同PAM的Cas蛋白拓宽引导编辑的靶向范围，但通常存在编辑活性低的问题。为克服上述瓶颈，本研究提出了一种多pegRNA顺序编辑策略：首先由pegRNA1引入临时PAM序列，随后pegRNA2识别该临时PAM并完成目标编辑，同时去除先前引入的临时PAM，从而实现负位或远距离的基因组编辑。在此基础上，结合本实验室前期开发的高效引导编辑器，构建了水稻（Oryza sativa）基因组DNA“打印”系统（Prime editing unpalpable nucleotide targeted editor, Printer），拓展了引导编辑系统的靶向范围。主要研究结果如下：

首先，在水稻原生质体中利用荧光报告系统验证了基于PPE2的Printer系统，平均编辑效率为12.07%，成功实现+27位A>G替换，显示出拓宽编辑范围的潜力。进一步基于PPE2探究了引导编辑的有效编辑窗口，证明单pegRNA可以实现+1~+20位的精准编辑，+20位平均编辑效率为3.15%。同时证明了pegRNA1通过插入GG创制PAM的能力，+20位插入GG的平均编辑效率为3.09%；以及pegRNA2实现目的编辑GG同时去除创制PAM GG的能力，+20位插入GG & +5~+6删除GG的平均编辑效率为3.22%。由此确定引导编辑的高效窗口为+1~+20位，Printer系统在此基础上可将编辑范围扩展至−15~+35位。

其次，利用单pegRNA在8个靶点验证了实验室自主开发的PEproV1和PEproV2的高效性，与PPE2相比，PEproV1和PEproV2的平均编辑效率分别提升2.93倍和3.20倍。进一步，PEproV1和PEproV2实现了+20位（5.20%）的有效编辑；证明了pegRNA1创制PAM的功能性，+19位的平均编辑效率为4.67%；验证了pegRNA2实现目的编辑同时去除PAM的功能性，+20位的平均编辑效率为7.65%。将高效编辑器PEproV1和PEproV2、已报道的复合型启动子（CaMV35S-CmYLCV-U6）驱动的tRNA-pegRNA-HDV表达系统与Printer策略相结合，在内源靶位点现实+15~+16 TG删除，平均编辑效率为2.22%；对于切口上游−8~−9 AG删除的负位编辑，PEproV2的平均编辑效率为0.21%。

第三，在原生质体系统中，通过系统优化Printer参数，揭示影响编辑效率的三个因素：（1）破坏pegRNA1 PAM策略可使编辑效率从0.00%提升至2.24%；（2）双pegRNA空间解耦设计揭示了pegRNA1和pegRNA2间互补性导致的效率下降；（3）pegRNA1创制的临时PAM存在显著的位置效应。在非靶向链的不同位置创制临时PAM，效率为2.48~3.95%。在靶向链和非靶向链创制临时PAM，效率为0.00~2.04%。进一步研究发现，通过碱基替换创制临时PAM的效率优于插入策略。在荧光报告系统中，通过碱基替换创制临时PAM最终实现+27位A>G目的编辑的平均编辑效率最高可达34.50%优于插入（32.70%）。并在水稻原生质体中成功实现内源靶位点+30 G>C替换，效率为0~0.31%，通过插入的方式未能实现目的编辑。

最后，在T0代水稻再生植株中，Printer系统在五个内源靶点Site 1~5上实现了18.46~59.02%的编辑效率，成功获得+29 G>C超远距离替换（Site1）及−4 C>G负位编辑（Site2）的纯合突变体。针对Site2靶点，尽管在T0代中使用PEproV1仅获得18.46%的较低编辑效率，研究进一步对其T1代植株进行了检测。结果表明，Printer系统在T1代仍保持良好的活性，在Site2靶点实现了53.85~55.00%的编辑效率，并成功获得−14至−16位GTC删除的负位编辑纯合体。并证明PEproV1是更适宜Printer系统的编辑器。该技术突破了传统引导编辑对NGG PAM位点的依赖，在负位区域实现高效编辑，为作物精准育种提供了普适化基因编辑工具。

综上所述，本研究部分解决了基于SpCas9为底盘的引导编辑受PAM限制的难题，通过多pegRNA顺序编辑拓宽了植物引导编辑系统的编辑范围，突破了传统PE无法实现负位编辑以及远距离编辑效率低下的技术瓶颈，部分解决了植物中因受到PAM限制而无法获得精准编辑种质资源的问题，为作物精准育种提供了更加普适的基因组编辑工具。

Prime editing (PE) enables precise base substitutions, insertions, and deletions, but its use in plants is limited by SpCas9’s strict NGG PAM requirement. Existing strategies to expand target sites via Cas variants often yield low or no editing activity. To address this, we developed a sequential multi-pegRNA strategy: pegRNA1 creates a temporary PAM, which pegRNA2 uses to install the desired edit while removing the temporary PAM. This enables efficient editing upstream and distally downstream of the nCas9 (H840A) nick site. By combining this strategy with high-efficiency prime editors from our lab, we established a PAM-independent editing system in rice, named Printer (Prime editing unpalpable nucleotide targeted editor), offering precise and flexible genome modification.

Firstly, the Printer system based on PPE2 was validated using a fluorescence reporter assay in rice protoplasts, achieving an average editing efficiency of 12.07% and successfully introducing a +27 A>G substitution, demonstrating its potential to expand the editing window. Further analysis revealed that a single pegRNA could mediate efficient editing from positions +1 to +20 (with +20 site averaging 3.15%). Additionally, pegRNA1 was shown to be capable of generating a PAM via GG insertion (+20 position, 3.09%), and pegRNA2 effectively introduced the desired edit while deleting the created PAM (combined editing efficiency of 3.22%). Thus, the effective editing window of PE was defined as +1 to +20, and the Printer system extended this range to −15 to +35.

Secondly, using single pegRNA at eight target sites, the editing efficiencies of two newly developed editors, PEproV1 and PEproV2, were evaluated. Compared to PPE2, their average efficiencies increased by 2.93-fold and 3.20-fold, respectively. PEproV1 and PEproV2 successfully mediated editing at +20 (5.20%), validated the functionality of PAM generation at +19 (4.67%), and confirmed PAM removal along with target editing at +20 (7.65%). Incorporating these editors with a composite promoter-driven tRNA-pegRNA-HDV expression system, the Printer strategy enabled endogenous +15~+16 TG deletion (average efficiency 2.22%). Notably, PEproV2 also achieved −8~−9 AG deletion (0.21% efficiency), whereas PPE2 and PEproV1 did not produce detectable edits.

Thirdly, optimization of the Printer system in protoplasts revealed three key factors affecting efficiency: (1) Disruption of the pegRNA1 PAM site increased editing efficiency from 0.00% to 2.24%; (2) Spatial decoupling of pegRNA1 and pegRNA2 mitigated efficiency loss due to sequence complementarity; (3) The position of the temporary PAM significantly influenced editing outcomes, with non-target strand modifications achieving 2.48~3.95% efficiency, while target and non-target strand combinations ranged from 0.00~2.04%. Base substitution proved superior to insertion for PAM creation, achieving up to 34.50% efficiency for +27 A>G editing in the reporter assay, compared to 32.70% by insertion. At endogenous sites, +30 G>C editing was accomplished (0~0.31% efficiency) only via base substitution.

In T0 generation transgenic rice, the Printer system achieved editing efficiencies of 18.46%~59.02% across five endogenous targets (Sites 1~5), generating homozygous mutants with a +29 G>C substitution at Site 1 and a −4 C>G edit at Site 6. Although PEproV1 showed a modest 18.46% efficiency at Site 2 in T0 plants, it retained strong activity in the T1 generation, with efficiencies rising to 53.85%~55.00%. Homozygous mutants with 14~16 bp (GTC) deletions upstream of the nCas9 (H840A) nick site were successfully obtained. Stable transformation data further support that PEproV1 is a more effective editor for the Printer system. This approach overcomes the NGG PAM constraint of conventional prime editors, enabling efficient upstream editing and offering a versatile platform for precision crop improvement.

In conclusion, This study partially addresses the challenge of PAM sequence constraints in guide RNA-directed editing based on the SpCas9 platform.—for rice, which expands the editable window of plant prime editing systems through a multi-pegRNA sequential editing strategy. It overcomes limitations of conventional prime editors in performing on the upstream and distal downsteam sequence of the nick site created by nCas9 (H840A) and provides a more versatile genome editing platform to accelerate the development of precisely edited crop germplasm.