Unlocking the power and potential of the Cas12 family members and their ancestors for crop genome editing

Abstract

Genome editing technologies have revolutionized plant research and offer great potential for crop improvement. Among them, CRISPR/Cas, as a versatile, simple, and inexpensive system for genetic manipulation, has dominated the genome editing field over the past decade. The applications of CRISPR/Cas in both plant biological sciences and crop improvement are particularly important in the context of global climate change, as well as in addressing current agricultural, environmental, and ecological challenges. Members of the Cas12 family, including Cas12a, Cas12b, Cas12i1, Cas12i2, Cas12i3, Cas12j, and Casδ, which belong to the Class II Type V CRISPR system, as well as their ancestors Fanzors and TnpBs, have received extensive attention in recent years due to their relatively smaller protein size, and alternative protospacer adjacent motifs (PAMs) or target adjacent motifs (TAMs), compared to the widely used SpCas9, representing promising alternative genome editing tools for crop improvement. This special section compiles a set of 19 articles, including 7 research papers and 12 brief communications, covering the latest advances in unlocking the potential of Cas12a and novel Cas12 family members as well as their ancestors for crop genome editing.

TnpBs, as the RNA-guided endonucleases ancestral to the CRISPR-Cas12 family, are components of the insertion sequences (IS) in the IS200/IS605 and IS607 transposon families. TnpBs belong to the obligate mobile element-guided activity (OMEGA) system and can specifically cleave double-stranded DNA by coordinating with their cognate obligate mobile element-guided activity RNAs (ωRNAs) or the so-called right-end element RNAs (reRNAs). TnpBs are promising alternatives to the larger CRISPR-Cas systems, due to their smaller sizes and hypercompact nature, which are advantageous for viral delivery in crop genome editing. So far, a number of TnpB family members have been exploited. However, most of them showed relatively lower editing efficiencies and generally produced a higher proportion of chimerism in the T₀ generation, complicating the recovery of stable edited lines for crop improvement. In this issue, to overcome these limitations, on the basis of a previously reported ISDra2-based TnpBmax, Li et al. (2026) developed an efficient ISDra2-based TnpBmax system for crop genome editing by integrating ωRNA truncation, T5 exonuclease fusion, and a surrogate selection strategy. This engineered ISDra2-based TnpBmax system significantly improved the editing efficiencies by up to 81.5% and 34.9% in rice and tomato, respectively. Furthermore, it drastically reduced chimerism from 68.6% to 21% and from 83.3% to 38.6% in rice and tomato, respectively, facilitating the recovery of transgene-free stable edited homozygous lines in the progeny for either functional genomics or crop improvement.

Fanzors (Fz), as the eukaryotic orthologs of prokaryotic TnpBs, are the programmable RNA-guided DNA endonucleases belonging to the obligate mobile element-guided activity (OMEGA) system as well. As the highly compact miniature RNA-guided endonucleases, Fz represent promising alternatives for genome editing in eukaryotes. However, the editing activities of the Fanzor systems in plants remain limited. In this issue, by using a single-transcript unit (STU) expression system, Ji et al. (2026) established an efficient editing system with NlovFz2, achieving editing efficiencies of up to 50.0%. Furthermore, Cao et al. (2026) engineered an optimized Fanzor2 system for plant genome editing by introducing a newly identified K416R amino acid mutation through near-saturation mutagenesis into the enNlovFz2 endonuclease and further fusing a self-cleaving ribozyme (HDV) to the ωRNA expression tandem array. This system, designated as enNlovFz2(K416R)/ωRNA-Rz, demonstrated significantly enhanced editing efficiency in soybean hairy roots and achieved robust mutagenesis in stable transgenic rice lines with efficiencies of up to 93.3%. They also established the first Fanzor-based cytosine base editor for plants, enabling C-to-T base transitions, albeit the efficiency of C-to-T base editing was very low.

The Cas12 family shows significant evolutionary divergence, ranging from complex multi-domain architectures to streamlined, minimalist structures. The well-characterized Cas12a and Cas12b are relatively larger in protein size, harboring different domains essential for their canonical functions. However, the editing performances of the Cas12a- or Cas12b-mediated base editors still lag behind these of nCas9-based base editors. In this issue, Cheng et al. (2026) enhanced the dead Cas12a (dCas12a)-mediated base editing by integrating the LbCas12a-RRV variant with introns to confer herbicide resistance in rice, while Wei et al. (2026) achieved efficient C-to-T editing (up to 29.10%) in stable rice lines by using a dead Cas12b (dAaCas12b) and a tRNA–crRNA tandem array. On the other hand, the Cas12i subfamily represents a trend toward minimalism, having evolved distinct folds and streamlined domains likely to adapt to specific host defense niches. To fully exploit the compact features of the Cas12 subfamily members, various strategies have been attempted to facilitate their versatility and feasibility for crop improvement. For example, Casδ is a recently identified novel evolutionary transitional Type V CRISPR system characterized by its compact size (~900 amino acids), broad temperature tolerance, and guided by a short crRNA without the requirement of a tracrRNA. However, the low editing efficiency of Casδ in eukaryotic cells limits its wider applications. In this issue, through a hierarchical engineering strategy, Ge et al. (2026) recently incorporated nine synergistic amino acid mutations to improve its binding affinities to the crRNA, the PAM duplex, and DNA substrates. The resulting enhanced enCasδ variant showed significantly improved editing efficiency. Compared to the wild-type Casδ-1, the enCasδ variant showed an average 8.5- and 5.3-fold increase in editing efficacies in both human cells and maize protoplasts, respectively. Furthermore, Zhong et al. (2026) engineered CasY7 as an efficient Cas12 system tailored for genome editing of monocot crops. By integrating codon optimization, a T5 exonuclease fusion, and an MS2-stabilized crRNA, the resulting pCasY7e3 editor achieved the highest efficiencies of 87.7% in maize, 82.9% in rice, and 58.8% in wheat.

Notably, CRISPR/Cas12i3, which belongs to the Type V–I Cas system, has attracted extensive attention recently due to its smaller protein size and its less restricted canonical “TTN” protospacer adjacent motif (PAM). Its engineered variant, Cas-SF01 (also known as Cas12i3-5M), has been successfully applied in rice, soybean, and wheat genome editing, but not in maize, oilseed rape, and orphan crop species as yet. In this issue, Hao et al. (2026) developed an efficient Cas12i3 genome editing system for allotetraploid oilseed rape (Brassica napus), successfully generating desirable agronomic traits such as multilocular siliques and male sterility. In maize, Liu et al. (2026a) established an optimized Cas12i3–TREX2 system that achieved high editing efficiencies of up to 86.96% in a commercial maize inbred JING724. By targeting the ZmCCT gene, they created an early-flowering germplasm with significantly reduced anthesis–silking interval. In broomcorn millet, Bai et al. (2026) utilized CRISPR/Cas12i3 to knock out PmBADH2a and PmBADH2b, creating the first fragrant broomcorn millet. They identified PmBADH2a as the dominant gene for aroma, and the resulting double mutants accumulated 2-acetyl-1-pyrroline (2-AP) up to 6.83 mg/kg without compromising key agronomic traits. Furthermore, Zhang et al. (2026b) developed an enhanced Cas12i3-5M-CP system for rice genome editing, achieving an impressive editing efficiency of 89.57%. By targeting the OsAUX3 promoter, this system significantly increased grain length without causing adverse pleiotropic effects.

By simultaneously targeting multiple loci within a single generation, multiplex gene editing could facilitate the efficient stacking and precise design of complex traits to substantially shorten breeding cycles and enhance comprehensive agronomic performance. In this issue, Chai et al. (2026) developed a multiplex editing platform in rice using Cas12i2 fused with a 5′ exonuclease (Exo-Cas12i2). This chimeric system successfully targeted miniature inverted-repeat transposable elements (MITEs) in OsDOG and OsMHZ1, highlighting the potential of minimalist Cas12 variants for modulating key agronomic traits, such as plant height and altered ethylene response in rice. Soybean (Glycine max L.) is a globally important crop, providing oil and protein to human diets and animal feeds. To facilitate trait stacking and germplasm innovation in soybean, Lin et al. (2026) developed a robust CRISPR/Cas12i3-5M system capable of editing up to eight target loci simultaneously in soybean and generated a quintuple mutant with ultra-low beany flavor and high oleic acid content in the elite cultivar DS89. Similarly, Xie et al. (2026) utilized CRISPR/Cas-SF01 to edit five genes simultaneously in soybean, creating a high-oleic, flavor-improved germplasm free of grassy and beany flavors.

Developments of Cas12i3-mediated base editing systems for either C-to-T or A-to-G transitions will expand the editing scope and enrich the plant base editing toolkits for crop improvement. However, while the Cas12i3-based cytosine base editor (CBE) only shows very low editing efficiency in plants, its adenine base editor (ABE) has not been documented as yet. To overcome these limitations, Zhang et al. (2026a) engineered CBE-V5 and ABE-V5 by fusing a transactivation module (VP64) and a double-stranded DNA-binding domain (HMG-D) to the complex composed of the dead Cas12i3-5M (dCas12i3-5M, also known as dCas-SF01) and the cytosine or adenine deaminases, respectively, boosting C-to-T and A-to-G editing in rice by up to 4.78- and 3.35-fold in rice stable lines, respectively, with maximum efficiencies reaching 32.35% for C-to-T and 38.24% for A-to-G transitions. Similarly, Jia et al. (2026) developed base editors based on Cas-SF01, and its IKRR variant (T208I/D293K/V460R/E494R) for rice base editing, with the IKRR variant showing base editing efficiencies up to 73.2% and 15.0% at target sites with relaxed 5′-NTN PAMs. Liu et al. (2026b) further improved ABE efficiency in rice and soybean by fusing dimeric TadA-8e to dCas-SF01, achieving robust precision editing with a higher proportion of lines in homozygosity. Hao et al. (2026) successfully established a Cas-SF01-derived CBE for allotetraploid oilseed rape (Brassica napus), achieving approximately 13%–15% editing efficiency in stable lines. Collectively, these advancements in Cas12i3-mediated or its engineered variants-mediated base editing systems have substantially improved their precision editing performances, establishing diverse versatile and precise base editing platforms that span from diploid monocots to polyploid crops, thereby facilitating crop improvement.

Development of efficient, genotype-independent delivery systems is highly valuable in genome editing for crop improvement due to the recalcitrant nature of some crop varieties to either Agrobacterium-mediated genetic transformation or regeneration of the transformants/edited plants. In this issue, Jing et al. (2026) optimized the Agrobacterium-mediated transformation workflow in soybean by combining a spectinomycin resistance gene (SpecR) with the visual RUBY reporter system. This dual-selection strategy effectively minimized chimerism and non-transgenic escapes, facilitating the robust delivery of the CRISPR/Cas-SF01 reagent and achieving relatively higher editing efficiencies ranging from 73.5% to 95.2% across 13 elite varieties. Furthermore, to address the challenges of genome editing in recalcitrant woody species, Suo et al. (2026) developed a tissue culture-independent delivery system in loquat (Eriobotrya japonica Lindl.) by magnetic nanoparticles (MNPs). By assembling CRISPR/Cas9 reagent with polyethyleneimine-coated Fe₃O₄ nanoparticles, they leveraged natural pollen apertures for pollen magnetofection. This pollen magnetofection system successfully bypassed the in vitro regeneration bottleneck of woody plant species, resulting in stable transgenic lines and targeted editing of the EjSLD gene. Although this study used the CRISPR/Cas9 system as a proof of concept, the established platform could be extended to the Cas12 family members and their ancestors for genome editing in woody plant species or other transformation- or regeneration-recalcitrant crop varieties.

The latest advancements presented in this issue unlock the power and potential of novel Cas12 family members and their ancestors in crop genome editing for both functional genomics and genetic improvement. However, it is worth mentioning that both Cas12 family members and their ancestors such as ISDra2 TnpB only possess RuvC domains for double-stranded DNA cleavage, lacking the typical HNH domains present in Cas9 orthologs, substantially limiting the developments of highly efficient base editing and prime editing toolkits. In the future, uploading the HNH domains from other Cas orthologs or some modules through AI-assisted rational design to induce nicks on either targeting strand or the non-targeting strand will be highly desirable to facilitate the development of Cas12 family members- or their ancestors-based efficient precision editing toolkits.