CasY7: An optimized Cas12i system for enhanced genome editing in monocot crops

ABSTRACT

The CRISPR–Cas12 family nucleases, particularly the Cas12i subtypes, are considered promising alternatives to Cas9 for genome editing in plants. We previously developed a new Cas12i variant, CasY7, which has been successfully applied in clinical trials; its performance in plants remains to be investigated. Initial testing in stable transgenic maize and rice showed that the codon‐optimized CasY7 (pCasY7e1) achieved average editing efficiencies of 58.7% and 62.3% across five target sites, respectively, outperforming the typical Cpf1 (pCpf1) control that targets the same sites. To further enhance activity, we fused T5 exonuclease to CasY7 (pCasY7e2), which shifted mutation profiles toward larger deletions, and subsequently integrated an MS2 aptamer into the crRNA scaffold (pCasY7e3). The optimized pCasY7e3 system increased editing efficiencies to 87.7% in maize and 82.9% in rice—approximately 2.7‐fold higher than pCpf1. We further demonstrated multiplexed editing in maize, generating biallelic dwarf mutants, and validated functionality in hexaploid wheat with editing efficiencies up to 58.8%. Overall, our comprehensive validation across 942 transgenic plants confirmed robust editing in maize, rice, and wheat, establishing CasY7 as a high‐efficiency addition to the CRISPR toolkit.

The Cas12i‐derived nuclease CasY7 system achieved high editing efficiency in rice, maize, and wheat, notably enabling multiplex editing. This establishes CasY7 as a robust CRISPR tool for crop improvement.

INTRODUCTION

Monocot crops, particularly cereals such as maize (Zea mays) and rice (Oryza sativa), serve as fundamental pillars of global food security, providing the primary caloric source for a majority of the world's population. In light of increasing food demands and environmental challenges, the utilization of advanced technologies like CRISPR/Cas systems for precision genetic improvement in crops is of paramount importance (Gao, 2021; Li et al., 2025; Tuncel et al., 2025). Among Class 2 CRISPR systems, the Type II nuclease Cas9 and the Type V nuclease Cas12 have emerged as cornerstone tools for plant genome editing (Zhang et al., 2024; Tuncel et al., 2025). Notably, Cas12 offers distinct advantages, including recognition of T‐rich protospacer adjacent motif (PAM) sequences, DNA cleavage mediated by a single RuvC domain resulting in staggered ends, and an inherent ability to process its own precursor crRNA, thereby significantly expanding the targeting scope and operational convenience for genome editing (Yamano et al., 2016; Zhang et al., 2024).

The Cas12i subfamily within the Type V CRISPR systems has garnered considerable attention due to its more compact protein size, minimal “TTN” PAM requirement, and autonomous pre‐crRNA processing capability (Zhang et al., 2020, 2022; Lv et al., 2023; Duan et al., 2024). However, the initial editing activity of wild‐type Cas12i orthologs is often limited in both animal and plant systems, presenting a major constraint for their broad application (Zhang et al., 2022; Duan et al., 2024). To overcome this limitation, several optimization strategies have been developed. One approach involves the rational engineering of the Cas protein sequence itself, where key amino acid substitutions are introduced to improve nuclease activity or alter PAM recognition (Kong et al., 2023; Duan et al., 2024; Han et al., 2024; Chen et al., 2025a). For example, the engineered variant Cas‐SF01, which contains several critical residue changes, exhibits an expanded PAM range and achieves high editing efficiencies in rice (Duan et al., 2024). Alternatively, fusion of exonucleases such as T5 to the Cas nuclease provides a distinct mechanism for enhancing editing outcomes (Wu et al., 2020; Wang et al., 2024, 2025; Devkar et al., 2025; Liao et al., 2025). T5 exonuclease processes the DNA ends generated by Cas cleavage by resecting the 5′ overhangs, thereby eliminating terminal complementarity and microhomology that facilitate precise repair. This shifts DNA repair toward the error‐prone non‐homologous end joining (NHEJ) pathway, resulting in significantly increased indel frequencies and large deletions. The efficacy of this strategy has been demonstrated in systems including Cas12i, where T5 fusion boosted editing efficiency in wheat (Wang et al., 2024), as well as in Cas9 and other Cas12 orthologs (Devkar et al., 2025; Wang et al., 2025; Chen et al., 2025b). These advancements underscore that systematic engineering represents a crucial approach to fully realizing the editing potential of Cas12i nucleases. Concurrently, the stability and proper folding of the crRNA constitute another critical factor influencing editing efficiency. The native crRNA architecture of Cas12i systems is relatively simple and can be susceptible to nuclease degradation or the formation of non‐functlional secondary structures within the cellular environment (Nishimasu et al., 2014; Yamano et al., 2016; Zhang et al., 2020). Previous studies have indicated that optimizing the direct repeat (DR) sequence can enhance the structural stability of crRNA, thereby improving the editing efficiency of the Cas12 system (Zhang et al., 2020; Han et al., 2025). MS2 is a stable RNA stem‐loop structure commonly employed as an RNA aptamer to recruit MCP‐fused proteins (Bertrand et al., 1998; Chen et al., 2025b); however, its potential role in stabilizing crRNA has not been previously tested.

In our earlier work, we developed a new Cas12i variant, termed CasY7. It was discovered by metagenomic data collected within China. This variant exhibits high activity in mammalian cells and has advanced to clinical trials (Ren Ji Hospital, 2024; Lu et al., 2024). However, its performance in plants remains unexplored. Building upon the foundation described above, this study aimed to develop an efficient CasY7‐based editing platform for major monocot crops. Through successive rounds of systematic engineering, we developed a series of progressively enhanced editors, designated pCasY7e1 to pCasY7e3. These optimizations encompassed three primary aspects: implementation of codon optimization to enhance protein expression levels; fusion with T5 exonuclease to promote the formation of indel mutations; and structural remodeling of the crRNA using an MS2 aptamer to improve its intracellular stability. The final optimized system, pCasY7e3, demonstrated robust editing in maize, rice, and wheat and provides a new tool for functional genomics and precision breeding in crops.

RESULTS

Screening of a codon‐optimized CasY7 variant

The newly identified Cas12i nuclease CasY7 (1,021 amino acids) exhibits only ~22% amino acid sequence identity with LbCpf1. This low degree of sequence identity with characterized Cas proteins supports its classification as a new subtype (Figure S1). It utilizes a minimal 47‐nt crRNA and recognizes a simple TTN PAM, making it an attractive candidate for genome editing. However, the initial test using the mammalian‐optimized variant (CasY7v1) showed negligible editing activity in maize protoplasts, confirming the need for systematic optimization (Figure S2). To enhance its performance in monocot crops, we recognized that optimizing its coding sequence would be critical (Fallahpour et al., 2025; Ravi et al., 2025). Using an optimized monocot codon optimization algorithm, we designed three new coding sequences (V2, V3, V4) to improve expression efficiency. A dual‐luciferase reporter system (pCasY7‐RP) for rapid screening of these variants was developed (Figure 1A). In this system, each CasY7 variant was fused to the HiBiT peptide and expressed under the maize ubiquitin promoter, with Firefly luciferase as an internal control. Transient expression assays revealed substantial differences in protein accumulation among the variants. The optimized CasY7v4 variant demonstrated the highest HiBiT/Fluc activity, showing a 41‐fold increase compared to CasY7v1 (Figure 1B). We further confirmed these expression patterns using GFP fusion constructs in maize immature embryos, where CasY7v4 again produced the strongest fluorescent signal (Figure 1C).

Figure 1.

Screening of a codon‐optimized CasY7 variant using a HiBiT/Fluc reporter system

(A) Schematic diagram of the reporter system constructs (pCasY7‐RP) used for transient expression in maize protoplasts. The cassette includes the maize ubiquitin promoter (ZmUBI), codon‐optimized CasY7 variants (V1 to V4), and a nuclear localization signal (NLS; indicated by a black‐gray box). (B) Quantification of CasY7 variant expression levels based on relative HiBiT/Fluc luminescence activity. (C) Representative fluorescence microscopy images showing the expression efficiency of each CasY7 variant in transfected maize embryos. Top panel: fluorescence channel; bottom panel: untransformed control. Scale bar = 0.5 mm. (D) Comparison of editing efficiencies of CasY7 variants and the LbCpf1 control in maize protoplasts. Editing frequency was determined by deep sequencing. (B–D) Data are presented as means ± SD (n = 3). Statistical analyses were performed using Student's t‐test: ** indicates P < 0.01, ns indicates P > 0.05.

The maize gene ZmBGLU11 was selected for an editing test in maize protoplasts. Analysis at the target site (crRNA1) confirmed that improved protein expression directly enhanced editing efficiency. The V3 and V4 variants achieved about 13% editing efficiency in protoplasts, significantly outperforming other variants and the LbCpf1 control (Figure 1D). These results demonstrate the critical importance of codon optimization and validate our optimization approach for developing efficient genome editing tools. Based on its superior performance, we selected CasY7v4 for all subsequent experiments.

Initial editing test in maize and rice

Building on the superior performance of the CasY7v4 variant, we constructed the plant editor of CasY7, pCasY7e1. This construct employs the maize ubiquitin promoter to drive the expression of the CasY7v4 and contains a U6 promoter‐driven crRNA expression unit (Figure 2A). To evaluate its editing efficacy, we designed five crRNAs targeting four maize genes, including ZmBGLU11 (crRNA1, crRNA2), ZmWaxy (crRNA3), ZmLAC1 (crRNA4), and ZmRAVL (crRNA5), and five crRNAs targeting three rice genes, including OsGBSS1 (crRNA6, crRNA7), OsIPA1 (crRNA8), and OsGS3 (crRNA9, crRNA10). LbCpf1 with an identical vector backbone was used as a control (pCpf1). The results demonstrated that pCasY7e1 maintained high editing efficiency across all tested sites. In T0 maize plants, the average editing efficiency for the five crRNAs (crRNA1 to crRNA5) reached 58.7%, significantly higher than the 33.2% observed with the pCpf1 control (Figure 2D; Table 1). Similarly, in rice, pCasY7e1 achieved an average efficiency of 62.3% for the 5 crRNAs (crRNA6 to crRNA10), compared to 30.9% for pCpf1 (Figure 2D; Table 1). This represents a 1.8‐fold and 2.0‐fold higher over LbCpf1 in maize and rice, respectively.

Figure 2.

Editing performance of CasY7‐based constructs in maize and rice

(A) Schematic diagrams of the four editor constructs (pCpf1, pCasY7e1, pCasY7e2, and pCasY7e3). The black‐gray box represents NLS. (B) A model illustrating the proposed molecular mechanism through which T5 exonuclease enhances CasY7 editing efficiency and generates larger deletions. The red triangles indicate cleavage sites. DSB, double‐strand break. (C) Sequence alignment showing the crRNA and the cognate target DNA. (D) Editing efficiencies of the four editors at 10 target sites (five in maize, five in rice) in T0 plants and distribution of mutation types (including homozygous, heterozygous, biallelic, and chimeric mutations).

Table 1.

Summary of the CasY7‐mediated editing efficiency

Species	Editor	Locus	crRNA	PAM	Total T0 plants	Edited plants	Efficiency (%)
Maize	pCpf1	ZmBGLU11	crRNA1	TTC	15	7	46.7
		ZmBGLU11	crRNA2	TTG	17	6	35.3
		ZmWaxy	crRNA3	TTC	21	5	23.8
		ZmLAC1	crRNA4	TTC	13	5	38.5
		ZmRAVL	crRNA5	TTC	23	5	21.7
	pCasY7e1	ZmBGLU11	crRNA1	TTC	12	10	83.3
		ZmBGLU11	crRNA2	TTG	16	10	62.5
		ZmWaxy	crRNA3	TTC	30	12	40
		ZmLAC1	crRNA4	TTC	12	8	66.7
		ZmRAVL	crRNA5	TTC	17	7	41.2
	pCasY7e2	ZmBGLU11	crRNA1	TTC	18	16	88.9
		ZmBGLU11	crRNA2	TTG	12	10	83.3
		ZmWaxy	crRNA3	TTC	31	27	87.1
		ZmLAC1	crRNA4	TTC	15	13	86.7
		ZmRAVL	crRNA5	TTC	17	10	58.8
	pCasY7e3	ZmBGLU11	crRNA1	TTC	12	12	100
		ZmBGLU11	crRNA2	TTG	8	8	100
		ZmWaxy	crRNA3	TTC	16	15	93.8
		ZmLAC1	crRNA4	TTC	23	21	91.3
		ZmRAVL	crRNA5	TTC	15	8	53.3
Rice	pCpf1	OsGBSS1	crRNA6	TTG	12	1	8.3
		OsGBSS1	crRNA7	TTC	15	2	13.3
		OsIPA1	crRNA8	TTC	16	7	43.8
		OsGS3	crRNA9	TTT	14	5	35.7
		OsGS3	crRNA10	TTA	15	8	53.3
	pCasY7e1	OsGBSS1	crRNA6	TTG	14	5	35.7
		OsGBSS1	crRNA7	TTC	20	8	40
		OsIPA1	crRNA8	TTC	18	14	77.8
		OsGS3	crRNA9	TTT	21	16	76.2
		OsGS3	crRNA10	TTA	22	18	81.8
	pCasY7e2	OsGBSS1	crRNA6	TTG	14	6	42.9
		OsGBSS1	crRNA7	TTC	21	9	42.9
		OsIPA1	crRNA8	TTC	20	17	85
		OsGS3	crRNA9	TTT	24	23	95.8
		OsGS3	crRNA10	TTA	25	24	96
	pCasY7e3	OsGBSS1	crRNA6	TTG	15	8	53.3
		OsGBSS1	crRNA7	TTC	18	11	61.1
		OsIPA1	crRNA8	TTC	12	12	100
		OsGS3	crRNA9	TTT	16	16	100
		OsGS3	crRNA10	TTA	10	10	100

Analysis of the allelic mutation patterns in T0 plants revealed that compared with LbCpf1, the proportions of homozygous and biallelic editing in pCasY7e1 increased significantly, while the proportion of chimeric editing decreased (Figure 2D). These results indicate that pCasY7e1 exhibits superior editing performance compared to LbCpf1 in monocot crops. The reliable editing capability demonstrated across multiple target sites provides a foundation for further optimization of the editing system.

Optimization of CasY7 editor using T5 exonuclease

To improve the editing efficiency of the CasY7 system, we constructed an advanced editor, pCasY7e2, by fusing a codon‐optimized T5 exonuclease to the C‐terminus of CasY7v4 (Figure 2A). This design was based on the premise that T5 exonuclease, which exhibits 5′ → 3′ exonuclease activity, could degrade the 5′ overhangs generated by CasY7‐mediated DNA cleavage (Wu et al., 2020; Devkar et al., 2025; Liao et al., 2025; Wang et al., 2025). We hypothesized that this processing would generate longer single‐stranded DNA regions, thereby interfering with cellular DNA repair pathways and ultimately increasing the indel frequency while promoting larger genomic deletions (Figure 2B). We systematically evaluated the performance of pCasY7e2 using the same 10 target sites described above (five in maize and five in rice). The results demonstrated that the fusion of T5 exonuclease significantly enhanced editing efficiency. In maize, pCasY7e2 achieved an average editing efficiency of 81% across the five target sites, representing a 1.4‐fold increase over pCasY7e1 (58.7%) (Figure 2D; Table 1). Among these sites, the crRNA3 target in the ZmWaxy gene showed the highest efficiency, reaching 87.1%. Similarly, in rice, pCasY7e2 attained an average efficiency of 72.5%, outperforming pCasY7e1 (62.3%) (Figure 2D; Table 1). The crRNA9 and crRNA10 targets in the OsGS3 gene exhibited the highest efficiencies, at 95.8% and 96%, respectively. Further genotyping analysis of T0 plants revealed that pCasY7e2 significantly improved editing quality. Compared to pCpf1, pCasY7e2 generated a higher proportion of homozygous and biallelic edits, with a lower incidence of chimeric editing, showing a similar pattern of edit‐type distribution to pCasY7e1 (Figure 2D). Specifically, in maize, pCasY7e2 induced 13.2% (10/76) homozygous edits, 60.5% (46/76) biallelic edits, 19.7% (15/76) heterozygous edits, and only 6.6% (5/76) chimeric edits. In rice, the corresponding proportions were 24.1% (19/79) homozygous, 50.6% (40/79) biallelic, 19% (15/79) heterozygous, and 6.3% (5/79) chimeric edits.

Notably, the incorporation of T5 exonuclease also markedly altered the indel pattern. As shown in Figure 3A, while pCpf1 and pCasY7e1 predominantly produced short deletions of 1–20 bp, pCasY7e2 substantially increased the proportion of large deletions, with a significant fraction ranging from 21–60 bp and even exceeding 200 bp. Sanger sequencing also confirmed the presence of these larger‐fragment deletion events (Figures 3B, S3). These results demonstrate that the fusion of T5 exonuclease to CasY7 not only enhances editing efficiency but also alters the indel pattern by promoting larger‐fragment deletions and increasing the frequency of biallelic modifications.

Figure 3.

Analysis of mutation types and deletion patterns of the optimized CasY7 system

(A) Size distribution of deletion events generated by pCasY7e1 to pCasY7e3 and a pCpf1 control. (B) Representative Sanger sequencing chromatograms of larger deletions induced by pCasY7e2. The TTN PAM is indicated in bold and the 23‐nucleotide spacer sequence is underlined.

MS2‐mediated crRNA optimization improves editing efficiency

Building on the T5‐enhanced pCasY7e2, we tried to optimize the crRNA architecture to further increase editing efficiency. In the Cas12i system, the minimal crRNA structure comprises a direct repeat (DR) and a spacer sequence, which form the less stable stem‐loop structure observed in sgRNA of Cas9. (Nishimasu et al., 2014; Yamano et al., 2016; Zhang et al., 2024). This simplicity may limit crRNA stability and compromise editing consistency. To address this, we introduced an MS2 RNA aptamer upstream of the DR sequence, generating the pCasY7e3 vector (Figures 2A, C, S4A).

The MS2 aptamer was expected to improve crRNA stability by reducing degradation and preventing non‐specific secondary structure formation (Bertrand et al., 1998; Chen et al., 2025b), while an HDV sequence at the 3′ end promoted accurate processing under U6 promoter regulation. Evaluation of the MS2‐optimized system showed improved editing efficiency at most target sites. In maize, pCasY7e3 achieved an average efficiency of 87.7%, surpassing both pCasY7e2 (81%) and pCasY7e1 (58.7%). Notably, the ZmBGLU11 crRNA1 and crRNA2 sites reached 100% editing efficiency. Similarly, in rice, the average editing efficiency rose to 82.9% with pCasY7e3, compared to 72.5% for pCasY7e2. Three rice targets (crRNA8, crRNA9, and crRNA10) showed 100% editing efficiency in T0 plants (Figure 2D; Table 1).

The MS2 modification consistently enhanced editing efficiency across most tested targets, likely by improving crRNA stability and preventing non‐specific secondary structure formation. The mutation profile remained similar to that of pCasY7e2, with a high proportion of larger‐fragment deletions and biallelic edits (Figures 2D, 3A), indicating that the MS2 aptamer elevates efficiency without altering the editing features of the CasY7‐T5 complex. The incorporation of the MS2 aptamer into the crRNA represents an effective strategy to enhance editing efficiency. And pCasY7e3 demonstrates overall superior performance and serves as the optimized editor of choice for CasY7‐mediated genome editing.

Multiplexed genome editing using CasY7

Targeting a gene with multiple crRNAs may enhance editing efficiency by increasing the likelihood of successful DNA cleavage. To test this hypothesis and evaluate the capacity of the optimized pCasY7e3 system for coordinated multiplex editing, we constructed three T‐DNA vectors, each harboring a distinct crRNA array designed to target different regions of the same gene (Figure S4B). The targeted genes included ZmGA20ox3, ZmBGLU11, and ZmNL4, with three specific crRNAs targeting exonic regions of each respective gene (Figures 4A, S5). Each crRNA array consisted of three expression units driven by independent Pol III promoters (ZmU6, ZmU3, and ZmU6‐2) within a single T‐DNA construct (Figure 4A).

Figure 4.

Multiplexed editing in maize using pCasY7e3

(A) Schematic diagram of multiplex crRNA expression cassettes. (B) Summary of multiplex editing efficiency in T0 maize plants. (C) Distribution of mutation types (homozygous, heterozygous, biallelic, and chimeric) in the three target genes. (D) Size distribution of deletion events generated by pCasY7e3 in three target genes.

Analysis of a total of 162 T0 maize plants showed that the pCasY7e3 system demonstrated efficient simultaneous editing of all three target sites. The individual crRNA editing efficiencies ranged from 38% to 94%, with an average efficiency of 72.9% across all nine targets (Figure 4B). Analysis of editing patterns revealed that 74.1% of all T0 plants contained mutations in at least two target sites, while 56.8% exhibited simultaneous editing at all three target sites of their respective gene (Figure 4B), demonstrating the robust multi‐target editing capability of the pCasY7e3 system. Notably, the multiplex editing approach also resulted in a high proportion of homozygous edits. The percentages of homozygous edited plants for ZmGA20ox3, ZmBGLU11, and ZmNL4 reached 40%, 50%, and 19.4%, respectively, indicating that multi‐target editing facilitates the generation of homozygous mutants more effectively than single‐target editing (Figures 2D, 4C).

The pCasY7e3 system also generated significantly larger‐fragment deletions in this multiplex editing configuration. Sequence analysis revealed distinct profiles for each target gene: ZmGA20ox3 primarily accumulated deletions shorter than 50 bp, ZmBGLU11 predominantly showed deletions of 401–500 bp, and ZmNL4 mainly contained deletions of 101–200 bp (Figure 4D). Importantly, these characteristic larger‐fragment deletions predominantly occurred in plants harboring mutations at all three target sites. The deletion sizes also correlated with the distance between the sites. Sanger sequencing further confirmed these representative editing events (Figures 5A–C, S6). The functional relevance of successful multiplex editing was further validated by distinct phenotypic changes, including a notable dwarfing phenotype in line ZmGA20ox3#7‐T1‐5 (Figure 5D; Paciorek et al., 2022). These results demonstrate that the CasY7 system enables efficient multiplexed genome editing, with the capacity to generate coordinated fragment deletions across multiple target sites.

Figure 5.

pCasY7e3 produced large fragment deletions in maize

(A–C) Schematics of the target sites in ZmBGLU11 (A), ZmGA20ox3 (B), and ZmNL4 (C) genes with representative Sanger sequencing chromatograms of edited T0 maize plants. The “A” position in the start codon “ATG” is set to “+1”. The red triangles indicate the corresponding crRNA targeting sites, and the dotted lines represent the dividing lines for larger fragment deletions. (D) Dwarfing phenotype of ZmGA20ox3#7‐T1‐5 line. Scale bar = 20 cm.

Application of CasY7 in wheat genome editing

To further evaluate the versatility of our optimized editing system, we applied the pCasY7e3 to the polyploid crop wheat. Three target sites were selected in the wheat genome, including the TaFT‐D gene, which regulates flowering time (Figure 6A; Li et al., 2024). Analysis of 95 T0 wheat plants showed that the pCasY7e3 achieved efficient genome editing across all tested loci, with editing efficiencies ranging from 25.7% to 58.8% (Figure 6A). The highest efficiency of 58.8% was observed at the crRNA19 target site, demonstrating the robust activity of pCasY7e3 in the complex hexaploid genome of wheat. Sanger sequencing also clearly showed the induced mutations (Figure 6B, C). Compared to wild‐type plants, TaFT‐D mutant lines exhibited significant alterations in spike morphology during the late heading stage (Figure 6D), confirming the functional impact of the introduced mutations. These phenotypic modifications demonstrate the potential of CasY7‐mediated editing for rapid trait modification in wheat, and maintain high editing efficiency in the challenging context of a polyploid crop.

Figure 6.

Genome editing in wheat using the pCasY7e3 system

(A) Summary of editing efficiency at three target sites (crRNA18, crRNA19, and crRNA20) in T0 wheat plants. (B) Schematic of the TaFT‐D gene feature and the target site. The TTN PAM is indicated in bold; the 23‐nucleotide spacer sequence is underlined. (C) Sanger sequencing chromatograms of a representative wheat mutant at the crRNA18 targeted site. (D) Spike morphology of wild‐type and TaFT‐D mutant line at the late heading stage. Scale bar = 1.5 cm.

DISCUSSION

In this study, we established CasY7, a new member of the Cas12i nuclease family, as an efficient and versatile genome editing tool for monocot crops. Through systematic engineering, including codon optimization, fusion with T5 exonuclease, and architectural refinement of the crRNA, we developed a series of editors with progressively enhanced performance. The optimized pCasY7e3 system achieved high average editing efficiencies of 87.7% in maize and 82.9% in rice, significantly outperforming the commonly used LbCpf1 nuclease. Notably, it exhibited low off‐target activity (Table S1), consistent with previous reports (Wang et al., 2024). Furthermore, we demonstrated its applicability in multiplexed editing and, importantly, successfully extended its use to the genetically complex polyploid crop wheat, where it reached editing efficiencies of up to 58.8%.

Our stepwise optimization strategy addressed several key factors known to limit the efficacy of CRISPR systems in plants. The initial limitation of low protein expression of the mammalian‐optimized CasY7v1 was overcome by screening codon‐optimized variants using a sensitive HiBiT/Fluc reporter system. The identification of the highly expressive CasY7v4 variant underscores the critical importance of codon optimization for achieving sufficient nuclease accumulation, a prerequisite for efficient editing (Wu et al., 2024; Fallahpour et al., 2025; Ravi et al., 2025). The fusion of T5 exonuclease to CasY7 enhances editing efficiency and shifts mutation profiles toward larger deletions, expanding the system's functional utility. While beneficial for generating complete loss‐of‐function alleles—especially in polycistronic or redundant genes—this feature is particularly valuable for regulatory element engineering. The ability of pCasY7e2 and pCasY7e3 to induce deletions spanning tens of base pairs enables targeted promoter editing. Unlike single‐nucleotide changes that subtly modulate expression, large indels can reconfigure transcription factor binding sites, eliminate repressor elements, or alter chromatin accessibility, thereby generating diverse promoter variants with graded or novel expression patterns. Thus, CasY7‐based systems serve not only as efficient tools for gene knockout but also as practical platforms for synthetic promoter engineering. The MS2 aptamer was incorporated into the crRNA scaffold to create the pCasY7e3 editor, demonstrating improved editing efficiency. This enhancement aligns with the hypothesis that structured RNA aptamers like MS2 can protect crRNAs from degradation and prevent their misfolding (Bertrand et al., 1998; Zhang et al., 2020; Han et al., 2025; Chen et al., 2025b), though the precise mechanism underlying this improvement warrants further investigation.

Despite these successes, certain limitations of the current system should be acknowledged. Editing efficiency varied across target sites, indicating persistent sequence‐dependent effects. A comprehensive, genome‐wide assessment of CasY7 specificity remains to be conducted. Future efforts to enhance the CasY7 system could focus on several fronts. First, optimizing nuclear localization signals (NLS) could further improve nuclear accumulation of the nuclease (Bircheneder and Parniske, 2025). Second, more extensive engineering of the crRNA represents a promising direction, including the design of target‐specific crRNA variants and the incorporation of novel stabilizing architectures such as circular crRNAs. Specifically, target‐specific design can be achieved by optimizing spacer length to balance binding energy, or by performing structure‐guided scaffold modifications to stabilize the ternary complex at recalcitrant loci. Such strategies, along with chemical modifications and circularization, have been reported to significantly enhance the efficiency and overcome the target‐dependency of other Cas12i systems (Zhang et al., 2023, 2025; Han et al., 2025; Liang et al., 2025). Third, testing stronger RNA polymerase II promoters to drive crRNA expression may potentially improve editing outcomes at recalcitrant loci (Jiang et al., 2020; Wang et al., 2024). Finally, extending the application of CasY7 to a broader range of monocot species and rigorously profiling its off‐target effects will be essential for establishing its broader utility and safety.

In conclusion, our work introduces CasY7 as a potent and adaptable genome editing tool for monocot crops. By implementing a rational engineering pipeline, we have unlocked the potential of this Cas12i nuclease, enabling high‐efficiency gene editing, multiplexing, and the generation of larger fragment deletions in both diploid and polyploid species. The CasY7 system provides the plant research and breeding community with a new option that favorably balances physical size, targeting simplicity, and editing efficacy.

MATERIALS AND METHODS

Construction of the plasmids

The CasY7 coding sequence was generated by an optimized algorithm. The coding sequences of CasY7v1 to v4, fused in‐frame with HiBiT, were synthesized and subsequently cloned into the pCXB053 backbone via BamHI/NcoI (NEB) digestion and Gibson Assembly (Vazyme, Nanjing, Jiangsu, China), replacing the original Cas9 gene, to generate four intermediate vectors. The UBI–CasY7 variant—HiBiT–NOS—was then excised from these intermediates and inserted into the pDual‐Luc01 vector (from our lab) via BamHI/SbfI digestion, replacing the 35S‐Nluc‐polyA cassette, resulting in the final reporter plasmid pCasY7‐RP. For Agrobacterium‐mediated transformation of maize immature embryos, the CasY7v1 to v4 sequences were subcloned from the pCasY7‐RP vector into the BsrGI‐linearized pCXBGFP backbone, generating the pCasY7‐GFP expression vectors. The CRISPR/CasY7 editing vectors (pCasY7e1 to e3) were constructed based on the pCXB053 backbone (Figure S4A). The codon‐optimized CasY7 sequence was cloned to replace Cas9 via BamHI/NcoI digestion and Gibson Assembly. Simultaneously, a synthesized DR‐ccdb–HDV was amplified and used to replace the ccdb–tracrRNA via SbfI/BsaI digestion and Gibson Assembly, yielding pCasY7e1. The pCpf1 vector was obtained from our lab (Wang et al., 2018). To construct pCasY7e2, a codon‐optimized T5 exonuclease was cloned and inserted between CasY7 and the NOS terminator in the pCasY7e1 backbone via BamHI digestion and Gibson Assembly. The pCasY7e3 vector was further modified from pCasY7e2 by replacing the original DR‐ccdb fragment with a synthesized MS2–DR‐ccdb cassette via BsaI digestion and Gibson Assembly. For crRNA design, 26‐bp targeting sequences (including the PAM) were selected within the genomic regions of interest (Table S2). Their targeting specificity was assessed using CRISPR‐P 2.0 (Liu et al., 2017). The crRNA expression cassettes were constructed as previously described (Lu et al., 2020). The crRNA array was assembled into the BsaI‐linearized pCasY7e3 backbone using Golden Gate assembly to generate the final multiplex editing vector (Figure S4B). The key carrier structure sequence is available in the Supplementary Data (supplementary sequence).

Maize protoplast transfection

Protoplasts were isolated from maize seedlings at the three‐leaf stage using a previously described protocol (Tu et al., 2020). For each transfection, approximately 5 × 10⁵ protoplasts were transfected with 10 μg of plasmid DNA. After transfection, the protoplasts were incubated in the dark at 25°C for 24 h and subsequently harvested by centrifugation at 100 g for 3 min for further analysis.

Luciferase reporter assay

Luciferase activity was measured using the Nano‐Glo® HiBiT Dual‐Luciferase® Reporter System on a GloMax® 96 Microplate Luminometer (Promega, Madison, Wisconsin, USA) according to a previously described method (Tian et al., 2022). Briefly, 200 μL of maize protoplasts were transfected with the respective dual‐luciferase reporter plasmid (pCasY7‐RP). After transfection and incubation, protoplasts were lysed with 100 μL of ONE‐GIo™ EX reagent containing LgBiT protein, which can lyse protoplasts, provide the Fluc substrate, and convert the HiBiT tag into NanoBiT® luciferase. Fluc was quantified after a 10‐min incubation. Next, 100 μL of NanoDLR® Stop & Glo® Reagent was added to the samples to quench the firefly signal and detect the now‐complementary NanoBiT® (HiBiT + LgBiT) signal. Finally, the ratio of the NanoBiT® and firefly signals was calculated.

Agrobacterium‐mediated transformation of maize, rice, and wheat

For maize Agrobacterium‐mediated transformation, immature ears were harvested at 10 d after pollination from the transformation maize receptor line (KN5585). Corn husks were removed and immature embryos (about 1.5 mm) were isolated in sterile conditions. Before transformation, immature embryos were placed with the scutellum side up on the medium and incubated in the dark at 32°C for two to 3 d of cultivation. After transformation, embryos were transferred to the resting medium and cultured for 5 to 7 d in the dark at 28°C (If the pCasY7‐GFP vector is transformed, the GFP fluorescence is observed under a fluorescence microscope.) Then, calli were transferred to the selection medium (glufosinate‐ammonium 10 mg/L), incubated in the dark at 28°C for 1 week. Positive calluses obtained were placed on the regeneration medium, incubated under 5000 lux at 25°C for 3 weeks. Regenerated shoots were transferred to rooting medium under 5,000 lux at 25°C for 14 d. Finally, transfer the regenerated seedlings to the soil. For rice, Agrobacterium‐mediated transformation of callus cells of japonica rice variety Zhonghua 11 (ZH11), Hygromycin B (50 mg/L) was used to select hygromycin‐resistant calli. Plantlets were regenerated from hygromycin‐resistant calli. The Agrobacterium‐mediated transformation of wheat variety Fielder was efficiently achieved using 1.5–2.0 mm immature embryos subjected to heat shock at 45.8°C, followed by infection with EHA105 harboring the vector. After co‐culture, resting, and single‐step selection‐regeneration, transgenic plants were obtained within 60 d. The transformation experiments were conducted by Wimi Biotechnology.

Genotyping

Genomic DNA was isolated using the CTAB method. PCR amplification was carried out with Taq polymerase (Vazyme) using 100 ng of genomic DNA as template. Genotyping of maize protoplasts was conducted through next‐generation sequencing (NGS). T0 transgenic positive plants are screened by PCR from tissue culture‐regenerated seedlings for subsequent target editing efficiency analysis. For precise mutation characterization, all plants were individually genotyped by PCR followed by Sanger sequencing, and the resulting chromatograms were analyzed using DsDecode (Liu et al., 2015). To confirm mutagenesis events, selected PCR products were cloned into a TA cloning vector (Vazyme), and eight positive colonies per sample were subjected to sequencing. The relevant primers can be found in the supplementary data (Table S3).

Statistical analysis

The relevant statistical test, sample size, and replicate type for each figure and table are found in the figure or table and/or the corresponding figure legends.

Generative AI usage statement

During the preparation of this work, the authors utilized DeepSeek for language polishing. After using this tool, the authors reviewed and revised the content where necessary and took full responsibility for the final version of the publication.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.

AUTHOR CONTRIBUTIONS

D.Z. and Y.L. designed the research; D.Z., Y.D., H.P., Y.F., Y.Z., and W.Y. performed most of the experiments; S.R., W.Y., and Q.Y. performed the plant transformation; D.Z., Y.D., Y.F., Y.W., Y.H., and Y.Z. genotyped the transgenic lines; D.Z., J.S., H.Z., Y.W., and J.X. analyzed the data; D.Z. wrote and revised the manuscript; Y.L. revised the manuscript. All authors read and approved the manuscript.

Supporting information

Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.70181/suppinfo

Figure S1. Phylogenetic tree of Cas12i effector

Figure S2. Performance of the codon‐optimized CasY7v1 in maize protoplasts

Figure S3. Representative editing outcomes at multiple target sites in maize and rice using the pCasY7e2 and pCasY7e3

Figure S4. Schematic of the CRISPR/CasY7 vector system

Figure S5. Target design for multiplexed editing

Figure S6. Representative genotyping results of maize genes edited by the CasY7 crRNA array system

Table S1. Mutation analysis of putative crRNA off‐target sites

Table S2. Summary of targeted loci in this study

Table S3. The primers used in this study

Biographies