Development of a multiplex precision gene editing system is highly desirable for pyramiding beneficial alleles in crop improvement. Prime editing is a newly developed genome-editing tool that can precisely enable the installation of all 12 nucleotide substitutions, short insertions, and deletions without exogenous DNA donor repair template and double-strand breaks (Anzalone et al., 2019Anzalone A.V. Randolph P.B. Davis J.R. Sousa A.A. Koblan L.W. Levy J.M. Chen P.J. Wilson C. Newby G.A. Raguram A. et al.Search-and-replace genome editing without double-strand breaks or donor DNA.Nature. 2019; 576: 149-157https://doi.org/10.1038/s41586-019-1711-4Crossref PubMed Scopus (1216) Google Scholar). Among the prime editing systems, prime editor 3 (PE3), which consists of a prime editing guide RNA (pegRNA) and an additional nicking single guide RNA (sgRNA) to nick the non-edited strand, enhances editing efficiency. So far, various strategies have been attempted to test the versability of prime editing of a single endogenous gene and to improve its editing efficiency in plants (Li et al., 2020Li H. Li J. Chen J. Yan L. Xia L. Precise modifications of both exogenous and endogenous genes in rice by prime editing.Mol. Plant. 2020; 13: 671-674https://doi.org/10.1016/j.molp.2020.03.011Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar; Tang et al., 2020Tang X. Sretenovic S. Ren Q. Jia X. Li M. Fan T. Yin D. Xiang S. Guo Y. Liu L. et al.Plant prime editors enable precise gene editing in rice cells.Mol. Plant. 2020; 13: 667-670https://doi.org/10.1016/j.molp.2020.03.010Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar; Lin et al., 2021Lin Q. Jin S. Zong Y. Yu H. Zhu Z. Liu G. Kou L. Wang Y. Qiu J.L. Li J. et al.High-efficiency prime editing with optimized, paired pegRNAs in plants.Nat. Biotechnol. 2021; 39: 923-927https://doi.org/10.1038/s41587-021-00868-wCrossref PubMed Scopus (66) Google Scholar; Xu et al., 2021Xu R. Liu X. Li J. Qin R. Wei P. Identification of herbicide resistance OsACC1 mutations via in planta prime-editing-library screening in rice.Nat. Plants. 2021; 7: 888-892https://doi.org/10.1038/s41477-021-00942-wCrossref PubMed Scopus (12) Google Scholar, Xu et al., 2022Xu W. Yang Y. Yang B. Krueger C.J. Xiao Q. Zhao S. Zhang L. Kang G. Wang F. Yi H. et al.A design optimized prime editor with expanded scope and capability in plants.Nat. Plants. 2022; 8: 45-52https://doi.org/10.1038/s41477-021-01043-4Crossref PubMed Scopus (10) Google Scholar). Multiplex precision gene editing of endogenous genes by prime editing has not been documented in various species, including plants. In this study, we developed three surrogate prime editors, including a hygromycinY46∗-based, an OsALSS627I-based, and a combined double surrogate system, for multiplex prime editing of endogenous genes in rice. We then validated the robustness of these three surrogate systems by precision editing of one, two, or three endogenous genes simultaneously, or a combination of precision editing and gene knockouts, in order to expand the applicability of prime editing in simultaneous improvement of multiple traits in crop plants. Previously, we reported a precise modification of an exogenous defective hptII mutant, to restore its function in hygromycin resistance during calli induction and selection (Li et al., 2020Li H. Li J. Chen J. Yan L. Xia L. Precise modifications of both exogenous and endogenous genes in rice by prime editing.Mol. Plant. 2020; 13: 671-674https://doi.org/10.1016/j.molp.2020.03.011Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). We reasoned that the prime editor-active cells could be enriched along with accurate correction of a null TAG (∗) to Tyr 46 (TAC) in a mutated defective hptII (mhptII) in the vector by prime editing, which will restore rice calli hygromycin resistance, and thus act as a hygromycinY46∗-based surrogate system (Figure 1A ). Besides, considering that the mhptII is an exogenous gene, we also thought of an endogenous allele OsALSS627I-based surrogate system (Figure 1A), which could either act as an endogenous gene or as an indicator and surrogate system to enrich the precisely edited events of other target endogenous genes. A Ser to Ile at position 627 in OsALS will confer rice herbicide resistance to bispyribac-sodium (BS) (Kawai et al., 2007Kawai K. Kaku K. Izawa N. Shimizu T. Fukuda A. Tanaka Y. A novel mutant acetolactate synthase gene from rice cells, which confers resistance to ALS-inhibiting herbicides.J. Pestic. Sci. 2007; 32: 89-98https://doi.org/10.1584/jpestics.g06-40Crossref Google Scholar), which could be used as a selection marker during calli induction and selection. We also hypothesized that, when both hygromycinY46∗-based and OsALSS627I-based surrogate prime editors were used synergistically as a double surrogate primer editor, stronger selection of prime editor-active cells will result in more efficient enrichment of the precisely edited events of endogenous genes than using a single surrogate prime editor alone, and thus the double surrogate system will be more labor saving and cost-effective. To test this possibility, we selected several endogenous genes as target genes for prime editing (Figure 1A), including OsSPL14 (also known as OsIPA1) (Jiao et al., 2010Jiao Y. Wang Y. Xue D. Wang J. Yan M. Liu G. Dong G. Zeng D. Lu Z. Zhu X. et al.Regulation of OsSPL14 by OsmiR156 defines ideal plant architecture in rice.Nat. Genet. 2010; 42: 541-544https://doi.org/10.1038/ng.591Crossref PubMed Scopus (883) Google Scholar; Wang et al., 2018Wang J. Zhou L. Shi H. Chern M. Yu H. Yi H. He M. Yin J. Zhu X. Li Y. et al.A single transcription factor promotes both yield and immunity in rice.Science. 2018; 361: 1026-1028https://doi.org/10.1126/science.aat7675Crossref PubMed Scopus (171) Google Scholar), OsDHDPS (a dihydrodipicolinate synthase gene) (Yang et al., 2021Yang Q.Q. Yu W.H. Wu H.Y. Zhang C.Q. Sun S.S.M. Liu Q.Q. Lysine biofortification in rice by modulating feedback inhibition of aspartate kinase and dihydrodipicolinate synthase.Plant Biotechnol. J. 2021; 19: 490-501https://doi.org/10.1111/pbi.13478Crossref PubMed Scopus (11) Google Scholar), and OsNR2, a gene encoding an NADH/NADPH-dependent NO3- reductase (NR) (Gao et al., 2019Gao Z. Wang Y. Chen G. Zhang A. Yang S. Shang L. Wang D. Ruan B. Liu C. Jiang H. et al.The indica nitrate reductase gene OsNR2 allele enhances rice yield potential and nitrogen use efficiency.Nat. Commun. 2019; 10: 5207https://doi.org/10.1038/s41467-019-13110-8Crossref PubMed Scopus (81) Google Scholar), and OsEPSPS gene (Li et al., 2020Li H. Li J. Chen J. Yan L. Xia L. Precise modifications of both exogenous and endogenous genes in rice by prime editing.Mol. Plant. 2020; 13: 671-674https://doi.org/10.1016/j.molp.2020.03.011Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). The detailed design of a set of pegRNA and sgRNA for each endogenous gene is as indicated in Figure 1A and the section “materials and methods” in the supplemental information. In order to enable multiplex prime editing, we used the polycistronic tRNA strategy to simultaneously produce a tandem array of pegRNAs and nicking sgRNAs (Xie et al., 2015Xie K. Minkenberg B. Yang Y. Boosting CRISPR/Cas9 multiplex editing capability with the endogenous tRNA-processing system.Proc. Natl. Acad. Sci. U S A. 2015; 112: 3570-3575https://doi.org/10.1073/pnas.1420294112Crossref PubMed Scopus (677) Google Scholar; Li et al., 2020Li H. Li J. Chen J. Yan L. Xia L. Precise modifications of both exogenous and endogenous genes in rice by prime editing.Mol. Plant. 2020; 13: 671-674https://doi.org/10.1016/j.molp.2020.03.011Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar) (Figure 1B). We first cloned a cassette containing the Actin promoter, polyA, and the Nos terminator into our prime editor-basic vector pCXUN-Ubi-NLS-nCas9(H840A)-Linker(33aa)-M-MLV-RT-NLS-PolyA-E9 (Li et al., 2020Li H. Li J. Chen J. Yan L. Xia L. Precise modifications of both exogenous and endogenous genes in rice by prime editing.Mol. Plant. 2020; 13: 671-674https://doi.org/10.1016/j.molp.2020.03.011Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar), to generate a pCXUN-Ubi-NLS-nCas9(H840A)-Linker(33aa)-M-MLV-RT-NLS-PolyA-E9-Actin-PolyA-Nos vector (hereafter referred to as the PE3 vector-Original). To establish the hygromycinY46∗-based surrogate system, we mutated the hptII gene in our PE3 vector-original at position Tyr 46 (TAT) to TAG, to generate hygromycinY46∗ (mhptII) (Figure 1A). We then designed a pegRNA and a nicking sgRNA for correction of mhptII by prime editing (Figure 1A). We cloned the tRNA-pegRNA-tRNA-sgRNA-tRNA tandem array for correction of mhptII into the HindⅢ digested mhptII vector to generate PE3-HS vector (Figure 1B). For the OsALSS627I-based surrogate system, we designed a pegRNA and a nicking sgRNA (Figure 1A), and assembly the tRNA-pegRNA-tRNA-sgRNA-tRNA complex for OsALSS627I editing into PE3 vector-Original to generate a PE3-based OsALSS627I surrogate system, PE3-AS vector (Figure 1B). For the double surrogate system, the tRNA-pegRNA-tRNA-sgRNA-tRNA complex for OsALSS627I editing was introduced into PE3-HS vector to generate double surrogate system, PE3-DS vector (Figure 1B). Then the tRNA-pegRNA-tRNA-sgRNA-tRNA tandem array for prime editing of each endogenous target gene (Figure 1B) was introduced into PE3-HS vector, PE3-AS vector, and PE3-DS vector, respectively, while the same tandem array for prime editing of the target endogenous gene in the PE3 vector-Original was used as a parallel control during rice transformation. We then tested the effectiveness of these three surrogate systems for prime editing of endogenous OsSPL14, OsDHDPS, and OsNR2, respectively. The aforementioned vectors were introduced into rice calli by biolistic transformation, and then to generate stable lines following the tissue culture process as indicated in Figure 1C, respectively. For the prime editing of a single endogenous gene, OsSPL14, as indicated in Figure 1D, nearly no plants were recovered for the calli transformed with PE3-Original with mhptII (CK) on induction medium added with either hygromycin or both hygromycin and BS. In comparison to PE3-Original with wild-type hptII, we observed the growth of fewer calli transformed with the surrogate editors PE3-HS, PE3-AS, and PE3-DS on induction medium added with either hygromycin or both hygromycin and BS after 4 weeks of selection, but vigorous growth of plantlets after 4–6 weeks of regeneration, respectively, due to the correction of mhptII or/and precise editing of OsALSS627I in vivo (Figure 1D). Among the 200 calli bombarded for each parallel experiment, treating the plantlets regenerated from each original bombarded callus as a single event, we recovered 101, 95, 20, and 73 independent transgenic lines derived from the calli transformed with PE3-Original, PE3-HS, PE3-AS, and PE3-DS vectors, respectively (Figure 1E). The genotype of each independent line was determined by Sanger sequencing of the PCR amplicons using the primer set listed in Supplemental Table 1, followed by decoding. We achieved one, two, two and nine independent lines with precise edits at efficiencies of 1.0%, 2.1%, 10.0%, and 12.3%, respectively (Figure 1E). The surrogate system could stimulate the prime editing efficiency by ∼2-, ∼10-, and ∼12-fold, respectively, and the double surrogate system was particularly effective (Figure 1E), in which among nine independent lines with precise edits, three were heterozygous lines, six were bi-allelic lines with one strand having the accurate edits, while another strand had partial precise edits (Supplemental Figure 1). Segregation analysis of one heterozygous line indicated that the precisely edited OsSPL14 could be inherited by the following generation (Supplemental Table 2). Furthermore, genotyping of mhptII and OsALS indicated that these lines had accurately edited functional hptII or/and OsALSS627I either at a homozygous or heterozygous status (Supplemental Figure 1). For OsDHDPS, among the 200 calli bombarded for each parallel experiment, we obtained 94, 80, 35, and 24 independent lines derived from the calli transformed with PE3-Original, PE3-HS, PE3-AS, and PE3-DS vectors, respectively. After genotyping, we obtained zero, one, five, and 13 independent lines with desired edits at efficiencies of 0.0%, 1.3%, 14.3%, and 54.2%, respectively (Figure 1E). Thus, the surrogate system could stimulate the prime editing efficiency by ∼2-, ∼14-, and ∼50-fold, respectively, again the double surrogate system was more effective (Figure 1E). For OsNR2, among the 200 calli bombarded for each parallel experiment, we obtained 102, 48, 41, and 31 independent lines derived from the calli transformed with PE3-Original, PE3-HS, PE3-AS, and PE3-DS vectors, respectively. After genotyping, although no precisely edited lines were obtained by using the PE3-Original vector, we obtained one, one, and one independent lines with desired edits at efficiencies of 2.1%, 2.4%, and 3.2% by using the surrogate prime editors, respectively (Figure 1E). The relative lower editing efficiency of OsNR2 may indicate that the prime editing efficiency also depends on the innate nature of the target endogenous gene. However, the surrogate system enables the prime editing of non-editable target such as OsDHDPS and OsNR2, which is otherwise impossible when using the original PE3. Together, these results indicated that the surrogate systems, especially the double surrogate system, were robust, labor saving, and cost-effective in generation of precisely edited lines by prime editing. Encouraged by the prime editing of single target endogenous genes using the surrogate prime editors, we then investigated the feasibility of multiplex prime editing for precision editing of multiple endogenous genes simultaneously and its efficacy. In addition to the above target genes for concurrent prime editing of two or three endogenous genes, we also added another endogenous gene, OsEPSPS (Figure 1A), and performed a combination of precision editing and gene knockouts, taking advantages of the nature of nCas9, which enables the cuts in two strands upon two sgRNAs being used. The employment of either a single or double surrogate system enabled us to precisely edit two or three endogenous genes simultaneously. For example, OsSPL14, OsDHDPS, and OsNR2 were precisely edited along with the precise modification of endogenous OsALS, at a simultaneous editing efficiency of 9.6% (7 out of 73), 45.8% (11 out of 24), and 3.2% (1 out of 31), respectively (Figure 1F). Furthermore, as well as endogenous OsALS, we also precisely edited another two endogenous genes simultaneously by using this double surrogate system, including a combination of OsSPL14 and OsDHDPS, OsSPL14 and OsEPSPS, OsSPL14 and OsVQ25, OsSPL14 and OsCYP71A1, and OsDHDPS and OsVQ25, respectively (Figure 1F). For both the endogenous OsALS and the above two endogenous gene combinations, the simultaneous editing efficiencies could reach 4.7%–4.8% (3 out of 64; 3 out of 62), 3.8% (2 out of 52), 1.8% (1 out of 56), 6.7% (1 out of 15), and 4.0% (1 out of 25), respectively (Figure 1F). For the prime editing of the OsSPL14 and OsDHDPS combination, again, we observed that the double surrogate system was more effective compared with the single OsALSS627I-based surrogate system (Figure 1F). Besides, we also obtained some independent lines with only one target endogenous gene (excluding OsALS) having desired edits at efficiencies of 8.1%–14.1% when using the double surrogate system (Supplemental Table 3; Supplemental Figure 1). Together, these results reinforced that we successfully established an efficient surrogate system for prime editing of multiple endogenous genes. Notably, generation of novel germplasm with improved multiple agronomic traits along with increased resistance to the OsALS-inhibiting herbicides certainly are highly desirable in breeding practice. To evaluate the specificity of each surrogate prime editor in rice, we examined the off-target possibility for each on-target site. No off-target effects were found at the potential off-target sites (CRISPR-GE; http://skl.scau.edu.cn/) in the tested lines (Supplemental Table 4). In summary, we developed three surrogate prime editors (hygromycinY46∗-based, OsALSS627I-based, and a combined double surrogate system, respectively) for prime editing of endogenous genes with substantially improved editing efficiency in rice stable lines. While the hygromycinY46∗-based and OsALSS627I-based surrogate prime editors could increase the editing efficiencies by ∼2- to 14-fold, the double surrogate system could stimulate the prime editing efficiencies up to ∼50-fold. Furthermore, we precisely edited several endogenous genes simultaneously and obtained stable lines by using the double surrogate system. The surrogate prime editor systems we established in this study enable the prime editing with improved efficiency and multiplex precision editing in rice, which will greatly expand the applicability of prime editing in simultaneous improvement of multiple traits in crop plants.