The evolving landscape of precise DNA insertion in plants

The ability to accurately insert new genetic information into a specific genomic location represents the pinnacle of crop biotechnology. Since the 1980s, genetic engineering has transformed agriculture, primarily by enabling the integration of genetic material (Fig. 1), achieving novel traits unattainable by conventional breeding¹. This foundational approach, encompassing transgenesis and cisgenesis/intragenesis, demonstrated the immense potential of genetic modification^2,3. However, the unpredictable nature of random insertion of foreign DNA fragments frequently causes undesired outcomes, such as multiple insertion sites^4,5, transgene silencing, and unstable expression^6,7,8 or rarer phenomena like epigenetic suppression⁹ or large-scale chromosome rearrangement^{10,11,12,13,14}, fundamentally limiting its precision and reliability.

To overcome the limitations of random DNA integration methods, a precise DNA integration method called site-specific recombination was developed using recombinases like Cre and FLP^15,16,17,18. This method requires a pre-existing recombination landing site within the plant genome for DNA sequence exchange with donors carrying transgenes, catalyzed by a recombinase. So far, there has been a limited number of reports showing efficient and routine applications of recombinase-based DNA insertion in plants^19,20,21. Alternatively, recombinases could be used to eliminate selection markers in the random DNA insertion approaches^22,23,24,25. This action is taken to address safety concerns and prevent regulatory issues²⁰.

Precise DNA integration can also be achieved through gene targeting (GT), a technique that utilizes homologous recombination (HR) to replace genomic DNA with homologous DNA donors that carry desired modifications^26,27,28. The pioneering GT work in plants was the directed integration of transgenes within a predictable site in the tobacco genome²⁸. Subsequently, GT introduced a waxy allele replacement to improve rice grain quality²⁹. However, due to the low efficiency of HR in somatic cells, the gene transfer method has seen limited applications in plants^26,30. A significant breakthrough occurred when it was discovered that GT efficiency could be dramatically enhanced by inducing a targeted double-stranded break (DSB) at the recombination site^30,31.

Recently, the invention of customizable molecular scissors such as homing nucleases (HMs) (also known as meganucleases), zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), and the clustered regularly interspaced short palindromic repeats (CRISPR)-CRISPR-associated protein (Cas) (CRISPR-Cas) systems (Fig. 2) have revolutionized the DNA integration field by assisting GT via site-specific DSB formation^32,33,34 or upgrading the precise DNA insertion tool with novel approaches such as prime editing (PE)^35,36,37 for short DNA insertion³⁸ and PE-recombinase-mediated large sequence insertion^39,40,41,42 (Fig. 3A). More importantly, the CRISPR-Cas-based methods offer much greater accuracy and specificity of DNA insertion, enabling single-copy, site-directed insertion of DNA fragments up to kilobase scales^39,40,42. Alternatively, large DNA sequence insertion at target sites of choice could also be possible with a transposase‑assisted target‑site integration (TATSI) system⁴³. These methods have transformed precise DNA insertion for crop breeding, creating more opportunities for plant synthetic biology and biotechnology applications that require controllable and customizable transgene integration and expression to produce bioactive compounds in plants.

In the context of our rapidly changing technological landscape, this Review synthesizes and discusses the principal methods of DNA insertion in plants. We outline the progression from traditional, albeit imprecise, random integration techniques to the highly controlled and programmable systems introduced by the CRISPR-Cas era. Additionally, we evaluate recent findings, identify ongoing challenges that hinder routine application, and explore future prospects that could make precise DNA insertion a universally applicable tool in plant science.

Transgenesis

Transgenesis refers to molecular techniques that introduce genetic materials, known as transgenes, from one species into another. This process allows the recipient organism to acquire traits that cannot be achieved through conventional breeding methods. The foreign genetic material may include gene expression cassettes and regulatory elements such as promoters, terminators, and genetic insulators like matrix attachment regions (MARs)⁴⁴. Once delivered into the cell’s nucleus, transgenes can either function temporarily or become stably integrated into the host genome, enabling long-lasting expression.

To effectively express foreign DNA, it is crucial to use delivery methods that align with the biological characteristics of the host organism. The successful introduction of foreign DNA depends on various transformation techniques. Common techniques for gene delivery include Agrobacterium-mediated transformation (AMT)⁴⁵, particle bombardment⁴⁶, polyethylene glycol (PEG)-mediated transformation, electroporation⁴⁷, microinjection⁴⁷, liposome-mediated transformation⁴⁸, viral vector-mediated delivery⁴⁹, and sonoporation⁵⁰. Among the transformation methods used in plant systems, AMT, particle bombardment, and PEG-mediated transformation stand out due to their compatibility with plant cell structures and regeneration processes.

AMT is commonly used in dicot plants and typically involves explants such as leaf discs or callus tissues. In contrast, particle bombardment is often applied to monocots or difficult-to-transform species, utilizing tissues like immature embryos or meristems. PEG-mediated transformation is widely used for plant protoplasts, allowing for the direct uptake of DNA. The choice of transformation method largely depends on the type of plant tissue or cell being used, whether intact tissue, callus, or protoplast, as well as the species-specific regeneration protocols⁵¹ (Fig. 1). Plant transgenesis has enabled various biotechnological applications, including studying gene function, introducing specific traits, molecular farming, and enhancing crop production^52,53,54 (Supplementary Data 1).

In prokaryotic organisms, particularly bacteria, foreign DNA is usually introduced using plasmid vectors. These plasmids replicate independently of the host’s genome. This allows for stable gene expression without requiring integration into the genome. Additionally, because bacteria have short generation times, it is relatively easy to remove or replace these vectors⁵⁵. In eukaryotic systems, gene expression can be classified as either transient or stable, depending on how the transgene is delivered and whether it integrates into the genome. Transient expression occurs when the DNA remains in an episomal form. This type of expression is short-lived, as it is subject to degradation or dilution during cell division. Therefore, transient expression is useful for short-term gene expression studies or functional assays. In contrast, stable expression happens when foreign DNA integrates into the genome. This integration allows for long-term expression, the creation of transgenic organisms, and heritable transmission of the traits.

The molecular mechanisms that enable the stable integration of transgenes are still under investigation. While canonical non-homologous end joining (cNHEJ) has traditionally been considered the primary pathway for this process^56,57,58, recent studies indicate that alternative non-homologous end joining (alt-NHEJ), also known as microhomology-mediated end joining (MMEJ), also plays a significant role^59,60. MMEJ may facilitate the integration of transgenes through short homologous sequences or microhomologies, ranging from 2 to 20 nucleotides. However, it lacks specificity and cannot identify the precise insertion site⁶¹.

Transgenesis has greatly impacted several fields, but it faces two significant challenges that restrict its wider applications. First, organisms altered through transformation are categorized as genetically modified organisms (GMOs) (Supplementary Data 1). This classification requires a costly and lengthy evaluation process before these organisms can be released into the market or the environment. Second, conventional transgenesis techniques often rely on the endogenous DSB repair mechanisms, such as MMEJ and cNHEJ, for transgene integration. This process depends on the spontaneous occurrence of double-strand breaks, making it impossible to predict or customize the site, copy number, and orientation of the inserted transgene.

The significant lack of control, along with public and regulatory concerns about “foreign” DNA, not only drove the quest for genuine site-specific integration but also encouraged the exploration of alternative strategies for modifying plants using genetic material sourced from their own breeding pools.

Cisgenesis and intragenesis

To address the regulatory and public perception challenges related to the use of “foreign” DNA in genetic engineering, two alternative approaches have emerged: cisgenesis and intragenesis. Both methods involve the insertion of DNA solely from sexually compatible species, setting them apart from transgenesis (Supplementary Data 1), which permits the incorporation of genes from organisms that are not closely related phylogenetically⁶². Both techniques aim to introduce specific genetic traits from within the conventional breeding pool, but differ in how they utilize the available genetic pool².

In cisgenesis, the entire gene, including its native promoter, coding region, and terminator, is transferred intact, maintaining the same sequence and orientation as in the donor organism. In intragenesis, by contrast, functional elements such as promoters, coding sequences, and terminators can be assembled, as long as they are all derived from sexually compatible species⁶³. This modularity facilitates the creation of customized gene constructs and enables the adjustment of gene expression levels, resulting in greater phenotypic variability compared to cisgenesis. These methods offer advantages over traditional breeding by allowing the introduction of targeted genetic materials without the risk of linkage drag, which can lead to the unintentional inclusion of undesirable alleles during cross-breeding⁶⁴.

Importantly, because the inserted sequences originate from crossable species, cisgenic and intragenic plants are generally perceived as more natural and thus face less public resistance than transgenic plants. These are potentially more acceptable regarding ecological safety and potential human health impacts, which positions them more favorably under GMO regulations⁶². This, in turn, may facilitate a more favorable path toward commercialization for such technologies. There are notable limitations to consider. The available genes for use are limited to species that can be crossed, and in the processes of cisgenesis and intragenesis, the insertion site cannot be precisely targeted. This leads to random integration of genes into the genome. Furthermore, when transformation is mediated by Agrobacterium, the T-DNA border sequences remain as foreign elements flanking the inserted DNA. This undermines the ideal of introducing only native sequences in cisgenic or intragenic approaches.

Indeed, many cisgenic plants have been shown to retain T-DNA border sequences^65,66. By definition, cisgenic/intragenic plants should not include sequences derived from outside the sexually compatible gene pool. Thus, strictly speaking, the presence of T-DNA border sequences means that these plants cannot be considered true cisgenic/intragenic plants. Although some studies have suggested that T-DNA border sequences can be omitted during AMT, these border sequences do not have significant effects on the plant phenotype, and T-DNA border-like sequences naturally exist in plant genomes^67,68. To achieve a genuine cisgenesis or intragenesis approach, additional strategies are required. This involves the development of methods to produce plants that are free from any exogenous DNA residues. Often, this includes techniques like co-transformation, followed by genetic segregation, or employing various molecular excision systems to remove unwanted DNA sequences after they have been integrated.

Several plants have been developed using cisgenesis and intragenesis approaches. The first reported cisgenic plant was an apple engineered to resist Venturia inaequalis, the cause of apple scab⁶⁶. The scab resistance gene HcrVf2, derived from the wild apple Malus floribunda 82, was introduced into the susceptible cultivar Gala, providing enhanced resistance (Supplementary Data 1). Other notable examples include a marker-free potato (Solanum tuberosum) harboring blight resistance genes Rpi-sto1 and Rpi-vnt1.1, derived from S. stoloniferum and S. venturii, respectively⁶⁵; an intragenic potato showing reduced enzymatic browning in the tubers⁶⁹; a cisgenic barley (Hordeum vulgare) with enhanced grain phytase activity via introduction of a HvPAPhya gene from a sexually compatible species⁷⁰; and bread wheat with constitutive expression of class I chitinase gene, resulting in increased resistance to leaf rust⁷¹.

Some plants have incorporated genes derived from sexually compatible species, but removing antibiotic resistance markers could not be achieved. For example, several apple cultivars developed in previous studies for scab resistance⁷², a grape (Vitis vinifera) expressing the thaumatin-like protein gene that confers resistance to fungal disease⁷³, as well as the Amflora potato developed by BASF Plant Science, contain antibiotic resistance genes. Since these markers are exogenous, the resulting plants are classified as transgenic GMOs rather than cisgenic/intragenic.

Overall, cisgenesis and intragenesis offer a significant improvement over traditional transgenesis by allowing for safer trait introduction without the need to cross species barriers. However, they do not address the main technical challenge of random integration, which means they still carry the same risks of position effects and unpredictable gene expression. This important limitation indicates that a genuine paradigm shift would involve not just managing the source of the inserted DNA but also mastering the specific location of its integration. This necessity paved the way for the first generation of true site-specific insertion technologies.

The pursuit of true genomic precision began long before the advent of CRISPR-Cas. This “pre-CRISPR-Cas era” spanning roughly from the late 1980s to 2012⁷⁴, was characterized by the development of foundational technologies that, while often inefficient or cumbersome, established the core principles of site-specific DNA integration. Two principal strategies dominated this period: the use of site-specific recombinases (SSRs) to catalyze DNA exchange at pre-defined “landing sites” (Fig. 3B) and the stimulation of the cell’s native HR pathway, a process that came to be known as GT.

SSRs were among the first tools used for controlled DNA integration. SSRs are categorized into two types: tyrosine recombinases (YR) and serine recombinases (SR). This classification is based on the catalytic amino acid residue that forms a covalent bond with the DNA during the recombination process⁷⁵. YR use identical recognition sites for bidirectional recombination, while SR recognize two distinct sites (e.g., attB and attP) and carry out unidirectional recombination⁷⁶. In genetic engineering, SSRs are utilized to reversibly recombine DNA sequences that are flanked by specific recombination sites, leading to the exchange of DNA sequences between donor and acceptor organisms^3,77. The SSR systems, including Cre-lox, FLP-FRT, and PhiC31, were used to remove selection marker genes from transgenic plants^23,24,76. Additionally, site-specific integration of transgenes using SSRs was investigated to resolve transgene silencing issues caused by random insertion^15,21. Nonetheless, the effectiveness of this method for de novo gene insertion was fundamentally limited by its reliance on pre-existing recombination sites within the host genome, which restricts its application to specially engineered plants.

In parallel, GT provided a more flexible approach for precise modifications by utilizing the cell’s own HR machinery to accurately insert or replace alleles using a homologous donor template^28,29,31. This was typically shown by restoring a truncated marker gene, leading to observable traits such as antibiotic resistance or fluorescence^28,78. However, due to its low efficiency (~ 10⁻⁴), routine applications in plant breeding were limited and unpopular^3,32,79.

The most critical finding of this period was the discovery that a targeted DSB at the recombination site could boost GT efficiency by several orders of magnitude, immediately highlighting the potential of site-directed nucleases. While early experiments used rare-cutting HMs like I-SceI on pre-inserted recognition sites to study HR mechanics^30,31, the true breakthrough was the invention of programmable nucleases. The first generations of these tools, ZFNs and TALENs, were engineered proteins capable of inducing DSBs at specific genomic loci^79,80. TALENs provide improved specificity and reduced cytotoxicity compared to ZFNs. They have been used to enhance plant genome editing by creating targeted DSBs in various plant species, including tobacco, tomato, barley, and rice. The editing frequency achieved was up to 3% in barley leaf cells, 6.3% in rice transformants, 9.65% in tomato calli, and 14% in tobacco protoplasts^33,81,82,83.

An efficient transformation and selection of GT cells, such as the positive-negative or allele-associated selection, facilitated the success of some cases in rice^29,84,85. However, this selection system left a positive selection marker within the genome of the targeted plant. Another strategy to enhance GT in plants is modulating the HR machinery, which may enhance GT frequency at several orders of magnitude (10⁻² to 10⁻¹ frequency)⁸⁶.

By the early 2010s, the field had made significant progress. The principles of precise DNA insertion were well-established, and with the use of TALENs, GT efficiencies were reaching levels suitable for practical application in major crops. However, a major bottleneck remained: the design, assembly, and validation of ZFNs and TALENs were complex, costly, and labor-intensive, which hindered their widespread adoption. As a result, there was a growing need for a new technology that could provide the same targeting capability but with much greater ease and programmability. This need would soon be spectacularly met by CRISPR-Cas.

CRISPR-Cas overview

The challenge of programming protein-based nucleases like ZFNs and TALENs was decisively overcome by the repurposing of a prokaryotic adaptive immune system: CRISPR-Cas. Initially, the CRISPR-Cas complexes were discovered as an adaptive immune system in bacteria. The first report of CRISPR arrays in 1987 in Escherichia coli⁸⁷ marked the beginning of the study and understanding of CRISPR-Cas systems and their role in bacterial immunity. Bacteria store fragments of viral DNA (called spacers) in CRISPR arrays and use them as guide RNAs (gRNAs) to direct Cas proteins for sequence-specific cleavage of invading foreign DNA⁸⁸.

The transformative breakthrough came in 2012 when researchers harnessed the CRISPR-Cas9 system from Streptococcus pyogenes as a simple, programmable genome editing tool⁷⁴. By designing a guide RNA to target a complementary DNA sequence, Cas9 could induce a DSB at the desired site (Fig. 2). This DSB is the critical starting point for DNA insertion, as it is resolved by one of two major cellular repair pathways: the error-prone NHEJ pathway or the high-fidelity HR pathway. This simple, RNA-guided programmability opened the door to vast applications across biology and medicine^89,90.

The true versatility of CRISPR-Cas lies in its modularity. More recently, scientists have fused a deaminase with Cas9 nickase to make base editors, allowing direct conversion of single DNA bases such as C to T or A to G without inducing double-strand breaks^91,92. CRISPR-Cas has also been modified to control gene activity without changing the DNA itself. Using a “dead” Cas9 (dCas9) fused with other proteins, scientists can turn genes on or off, edit epigenetic marks, or even target RNA. Most relevant to this review, the CRISPR-Cas system evolved to advanced insertion platforms like PEs and CRISPR-transposase systems.

The applications of this revolutionary technology have been rapid and widespread, impacting diverse fields such as biomedical research and agriculture. In plant biotechnology, the key promise of CRISPR-Cas is to make precise DNA insertion an efficient and routine process. In the following sections, we will examine the primary strategies developed to achieve this goal. We will explore how the CRISPR-Cas platform has been utilized to harness different cellular mechanisms for DNA insertion, starting with DSB repair pathways (NHEJ and HR), and later advancing to more sophisticated, DSB-independent systems like PE and programmable recombinases.

CRISPR-Cas-based NHEJ-mediated precise sequence insertion

Plant DSB repair pathways mediate not only simple rejoining but also the capture of exogenous DNA. In a foundational 1998 study, Salomon and Puchta demonstrated that DSB repair in tobacco somatic cells sometimes incorporates genomic or T-DNA fragments at break sites⁵⁷. Subsequent work further revealed that double-stranded T-DNA preferentially integrates at DSBs via NHEJ pathways (cNHEJ or MMEJ), generating junctions with microhomologies, filler sequences, and occasional concatemers^56,58. These findings highlight that plant DSB repair can drive both targeted and ectopic sequence integration, explaining the junction heterogeneity commonly observed in engineered DSB-mediated DNA insertion systems. Therefore, the most direct strategy for CRISPR-Cas-mediated DNA insertion is to hijack the cell’s dominant DSB repair pathway: NHEJ. While classically considered “error-prone”, this pathway’s core function is the rapid ligation of broken DNA ends. The key insight was that by flooding the site of a CRISPR-Cas-induced DSB with a high concentration of a donor DNA fragment, this ligation machinery could be co-opted to directly insert the donor into the genome. This approach was first enabled by the observation that the CRISPR-SpCas9 complex has long been thought to generate a predicatable blunt-end DSB at the −3 position, counting from the NGG PAM⁷⁴. It was therefore natural to harness the cleaved pattern to obtain precise sequence insertion or replacement with blunt-ended donors via the canonical NHEJ (cNHEJ) mechanism (Fig. 2).

By co-delivery of chemically modified blunt-end dsDNA donors (dsODNs) and CRISPR-Cas9 complexes that could generate single or double cleavage at targeted sites, precise donor sequence bidirectional insertion and replacement were achieved at up to 25% in rice⁹³ and 6–61% in maize⁹⁴ (Fig. 2 and Supplementary Data 2). Notably, this approach outperformed the GT-based sequence insertion under the same experimental conditions⁹³ and up to four target multiplexing was possible⁹⁴, albeit it also introduced small sequence damages at the junction sites^93,94. In a subsequent work, Tian and coworkers co-introduced the chemically modified dsODNs with a surrogate wheat dwarf virus-based replicon system for gRNA amplification into stably expressed SpCas9 calli, achieved in-locus tagging of a 3 × FLAG tag at three targets⁹⁵. Tagging with the luminescent HiBiT peptide (TagBIT), a method for in-locus tagging and dynamic tracking of endogenous protein localization and interaction, was developed in rice based on the CRISPR-Cas-based cNHEJ-mediated precise sequence insertion approach³⁸. Inspired by the work of Lu and coworkers⁹³, Li and colleagues developed a CRISPR-Cas9-mediated seamless gene replacement system using tobacco protoplasts, enabling precise and marker-free replacement of the tobacco mosaic virus (TMV) N′ resistance gene with a chemically modified 1819-bp fragment, obtaining TMV-U1 resistance⁹⁶. Reasoning that the CRISPR-Cas9 complex may also induce 5′ 1-bp-staggered end cleavage, Kumar and coworkers harnessed the CRISPR-Cas9 system for directional dsODN-based targeted insertion (DOTI) (Fig. 2) using donors containing 5′ 1-nt overhangs. Small-scale (up to 34-bp) sequence insertion was possible at high frequencies in green foxtail (Setaria viridis) protoplasts and rice⁹⁷.

Once the cNHEJ fails to repair a DSB, the cells may switch the damage repair machinery to an alternative NHEJ (also known as MMEJ) mechanism, a backup plan for direct end ligation⁹⁸. This DSB repair pathway utilizes microhomologies flanking the broken ends to anneal and mend the ends, resulting in intervening sequence deletions⁹⁹. While the CRISPR-Cas9-based MMEJ-mediated precise sequence insertion technique has been shown to be feasible in mammalian cells^100,101, its application remains at a nascent stage in plants, especially with the AMT system¹⁰².

CRISPR-Cas-based GT for precise sequence insertion

In contrast to the direct ligation approach of NHEJ, GT represents the gold standard for scarless, high-fidelity DNA insertion, a method revitalized and made practical by CRISPR-Cas^26,32.

Briefly, once a DSB is formed and signaled/redirected to be repaired by HR mechanism, the broken ends are first resected to generate 3′ ssDNA overhangs that can then anneal with a homologous arm of a ssDNA/dsDNA donor, and thereby trigger DNA-dependent DNA polymerization that copies the information from the donor template into the genomic site (Fig. 2). In plant somatic cells, the HR mechanism may end up with noncrossover products via synthesis-dependent strand annealing³¹. Theoretically, DNA sequences of any length flanked by homologous arms can be inserted via this method, provided that they can be introduced to the target site and the HR process has been initiated.

A DSB was shown to be critical for the initiation and success of a GT experiment³⁰. Initially, there was no tool to introduce a DSB at a target site to initiate DSB repair, and thus, GT events were likely obtained by the random DSB formation at or near the target region^28,29. Since the customizable nucleases such as ZFNs, TALENs, and CRISPR-Cas were invented, the GT method has entered a new era of exploitation and application^26,103,104. By integrating DNA sequences within the homologous donor, researchers could precisely insert up to several kilobases into a genome site of interest in plants via the assistance of customizable nucleases that could induce DSBs at the GT-targeted sites^105,106.

At the early stage, the GT experiments were mostly focused on the demonstration of the GT’s feasibility, investigation of the underlined mechanisms, and improvement of GT efficiency. In most GT experiments, the homologous donor template contained an antibiotic selection or fluorescent marker to enrich the GT events, and for better comparison of the GT efficiency among treatments^32,107. The allele-associated marker gene can later be excised via the employment of the transposon-mediated excision mechanism^108,109.

The majority of the GT data were reported with monocot plants such as rice and maize via particle bombardment delivery of the editing tools. The inserted sequence length varied from a dozen to thousands of nucleotides, and with varied frequencies (Supplementary Data 2). The bombardment method helps deliver a high dose of donor templates and thus significantly enhances GT efficiency. Despite some advantages, particle bombardment has notable disadvantages, including the occurrence of random DNA breaks and chromosomal rearrangements¹¹. When high copies of donor DNA are delivered alongside DSBs, it can also result in random insertions at these break sites within the plant genome. Alternatively, GT tools could also be introduced into plant cells via the AMT system. However, compared to particle bombardment, AMT has a limitation in the number of T-DNA copies that can be delivered into plant cells^4,110. Therefore, the number of homologous donor templates delivered by T-DNA is also limited and becomes one of the bottlenecks in achieving high GT efficiency. GT-mediated sequence insertion efficiency was recently shown to be significantly enhanced when homologous donors were delivered and amplified by geminiviral replicon vectors^33,34,111, taking advantage of the autonomous amplification capacity of the replicons. Replicon-based tools significantly enhanced GT-mediated sequence insertion efficiency in plants (Supplementary Data 2). In Arabidopsis, since a geminiviral replicon system has not been established for GT, a sequential transformation may elevate the achievement of GT plants in the T2 generation¹¹².

Tethering the donor to the CRISPR-Cas complex or other donor recruitment methods to provide the homologous DNA templates for GT reactions at the target sites is another way to enhance GT efficiency. Ali and colleagues fused Agrobacterium VirD2 relaxase to Cas9 and included phosphorothioate‑modified donor DNA to anchor the repair template, achieving 8.7% efficiency in precisely inserting an HA tag at the OsHDT site, a significant improvement over untethered systems¹¹³. Similarly, Nagy and coworkers combined LbCas12a with the HUH endonuclease from Faba Bean Necrotic Yellow Virus to tether single-stranded donor DNA to the CRISPR-Cas complex, boosting local donor concentration at a target site (D5) in soybean. This strategy led to site-specific integration of a 70‑nt oligonucleotide donor with 3.3% precise sequence insertion efficiency, higher than 1.3% in the case of the untethered system (Supplementary Data 2)¹¹⁴. Recently, by fusing Cas9 with a geminiviral Rep protein, Zhou and coworkers enabled covalent tethering of donor DNA via rolling‑circle replicons in vivo. This method increased knock‑in efficiencies 4 ~ 7.6‑fold compared to Cas9 alone, achieving up to 72.2% of T₀ rice plants carried scar-free in‑frame insertions of 33 ~ 519 bp tags, transmitted faithfully and without off‑target events¹¹⁵. Though the method was not used to precisely insert the gene-scale sequences, it is promising for the same to be evaluated and exploited for large-scale sequence insertion in plants.

PE enables precise sequence insertions

Seeking to circumvent the cellular dependency on DSBs and donor DNA templates altogether, PE emerged as a revolutionary approach that “writes” new genetic information directly into the genome³⁵. Theoretically, PE allows almost all types of DNA modifications, such as base substitutions, deletions, and insertions, without a DSB^36,37,116. PE utilizes a nickase SpCas9 (nCas9) to create an R-loop at the target site and single-stranded breaks (SSBs) on the non-target strand. The SSB releases a 3′ single-stranded end that anneals to the primer binding site (PBS) of a pre-designed pegRNA, which contains the PBS and reverse transcriptase template (RTT) sequences with desired bases to be inserted into the genome³⁵ (Fig. 3A). An RT enzyme fused to nCas9 adds deoxynucleotides to the 3′-OH of the nicked end, resulting in a 3′ flap that competes with the original sequence during DNA repair^35,117, potentially involving mismatch repair (MMR) pathways¹¹⁸. The technique is efficient and precise in animals³⁵ and plants^36,37, albeit different loci often bring various editing efficiencies.

The frequency of DNA insertion or protein tagging in plants using PE was found to be inversely related to the length of the DNA. Reflecting this inverse relationship, the efficiency of PE drops sharply for insertions larger than a few hundred base pairs, with most achievable edits being under 500 bp in animals and monocotyledonous plants (Supplementary Data 2). In mammalian cells, the efficiency of PE-based sequence insertions can be improved by using a pair of pegRNAs that generate overlapping 3′ flaps, such as the TwinPE⁴¹ approach and its modified variants^119,120. Similarly, the GRAND-PE method¹²⁰ can achieve efficient insertion of immunogenic tags (such as 6 × His, 1 × HA, or 3 × FLAG tag, up to 72-bp in size) in rice, which allows for deleting genomic sequences at the insertion sites^121,122. Recently, PE-mediated gene tagging has been upgraded for convenient insertion of up to 204-bp sequences^122,123 and has hardly reached up to 403-bp insertion using a template-jumping PE approach¹²⁴. In dicotyledon plants, however, protein tagging has been extremely inefficient and has not been inheritable due to low PE efficiency^125,126. With the breakthrough in dicot PE³⁷ and a recent upgrade of the PE system with evolved PE variants¹²⁷ in monocots^123,128, we can expect more efficient small-scale sequence insertion in dicots to be reported soon.

Recently, a powerful strategy combining PE with SSRs has enabled the precise insertion of kilobase-scale DNA sequences into predefined genomic loci in both mammalian and plant cells^{39,41,129,130} (Fig. 3B). In this two-step method, the PE system is first used to install specific recombination target sequences, such as loxP, lox66, or attB, into selected genomic regions. This step relies on the high precision of PE to introduce these recombination motifs without generating double-strand breaks. In the second step, exogenous donor DNAs flanked by matching recombination sites, such as lox71 or attP, are introduced into the edited cells, and site-specific recombination is catalyzed by co-delivered SSRs (e.g., Cre or phiC31 integrase), leading to precise and unidirectional integration of the donor fragment into the pre-installed recombination site^41,129,130 (Fig. 3B). In plants, this concept has been significantly advanced by the development of PrimeRoot, a paired-PE-based platform that enables targeted insertion of large DNA fragments into any genomic site of interest with high precision^39,40 (Fig. 3B and Supplementary Data 2). PrimeRoot employs paired pegRNAs to introduce a recombination site, such as the lox66 sequence, into a genomic site of interest, thus creating a docking site for recombinase-mediated integration of a large DNA donor. A donor DNA containing a corresponding recombination motif, such as lox71, flanking the cargo DNA is then introduced, and the Cre recombinase facilitates directional recombination between the lox66 and lox71 sites. This results in a stable insertion of the donor DNA into the plant genome³⁹. PrimeRootv3 demonstrated its potential by successfully inserting a 4.9-kb gene expression cassette into a characterized GSH site³⁹. Recently, the PrimeRoot system was upgraded to programmable chromosome engineering (PCE) tool with retrofitted recombination sites and AI-based engineered recombinases that enabled precise insertion of up to 18.8 kb in the rice genome and scarless kilobase-to-megabase genome editing in plants and human cells⁴² (Supplementary Data 2). This cutting-edge platform provides a transformative tool for plant biotechnology by enabling the precise stacking of multiple genes or alleles at defined loci, thereby facilitating accelerated breeding of elite traits and supporting de novo domestication strategies tailored for next-generation crop improvement^40,42.

CRISPR-Cas-based transposase-mediated sequence insertion

A fourth major paradigm for DNA insertion involves merging the programmability of CRISPR-Cas with the powerful “cut-and-paste” mechanism of mobile genetic elements. Transposons have long been harnessed for functional genomics, such as gene tagging (mutating or activating), via their characteristics of random insertion and the capability of carrying DNA cargos to the inserted sites in a number of plants¹³¹. Recently, Liu and colleagues developed a transposase‑assisted target‑site integration (TATSI) system (Fig. 3C) by fusing the rice Pong transposase (ORF2) with CRISPR‑Cas9 (or Cas12a), enabling precise “cut‑and‑paste” insertions at CRISPR-Cas‑specified loci in Arabidopsis and soybean. They demonstrated targeted integration of various cargos, including a ~ 444 bp enhancer (heat shock element), a ~ 1 kb open‑reading‑frame (bar CDS), and full gene expression cassettes (up to 8994 bp), achieving insertion efficiencies ranging from 6 ~ 36% depending on sequence lengths and species (Supplementary Data 2)⁴³. The TATSI system marks a major advancement in plant genetic engineering and crop breeding. By converting a native transposon into a programmable toolkit, TATSI provides breeders with a reliable and precise method for stacking agronomic traits, enhancing nutritional profiles, and accelerating the introduction of traits in major crops. This approach reduces random insertions and minimizes the need for extensive screening efforts. However, despite this remarkable improvement, ectopic insertions were still detected alongside site-specific events in the tested plants, indicating that complete control over transposase activity has not yet been achieved. This caveat highlights the need for further optimization to restrict integration strictly to target loci, potentially through the use of engineered transposase variants or temporal control of expression. Addressing these residual off-target integrations will be essential to fully harness TATSI’s potential for stable and predictable DNA insertion on plants.

Transposons such as Pong, when precisely guided by CRISPR-Cas systems, provide an effective platform for precise and site-specific gene insertion in plants, eliminating the need for homology-directed repair. This approach enables the effective integration of extensive DNA sequences, making it particularly well-suited for stacking complex traits within crop genomes. Unlike traditional transposon systems, programmable variants reduce random insertions and enhance genomic stability. These tools are set to play a vital role in next-generation plant breeding and pan-genome engineering strategies¹³¹

Comparative insights into CRISPR-Cas-based DNA insertion strategies

The recent data (Supplementary Data 2) demonstrate that the current CRISPR-Cas-based insertion platforms, including NHEJ-based DNA insertion, GT, PE, PE-recombinase-based DNA insertion, and TATSI, each have unique balances between precision, efficiency, and versatility in plants (Table 1).

Among the various methods, NHEJ-mediated strategies, such as DOTI, demonstrate a simple system design, the widest range of species applicability and the highest efficiencies with moderate DNA lengths, achieving up to 71.75% in cereals and several dicots^{38,93,94,95,96,97,132} (Table 1 and Supplementary Data 2). These methods do not require homology arms and can function through straightforward donor ligation. However, they typically result in the introduction of short indels and small sequence scars at the junctions. In contrast, GT facilitates scarless, allele-specific integration of large fragments, up to 11.1 kilobases¹³³. Nonetheless, it is limited by the low activity of HR in plant somatic cells and high efficiency variation among plant species and delivery methods (Supplementary Data 2). PE offers a newer, DSB-independent method that directly “writes” new sequences into the genome. It is most effective for short to moderate insertions (less than 500 base pairs, Supplementary Data 2). When combined with recombinases, such as those used in the PrimeRoot platform, it can accommodate payloads extending to 18.8 kilobases⁴². Transposase systems, such as TATSI, enable the insertion of gene-scale fragments (up to 8.994 kilobases) with efficiencies ranging from 6% to 36% in Arabidopsis and soybean⁴³, all without relying on HR. This makes them a promising option that strikes a balance between precision and payload capacity.

GT and PE provide the highest sequence fidelity, making them ideal for in-locus tagging, precise allele replacement, or the insertion of regulatory elements. On the other hand, NHEJ and transposase-based systems are more accommodating of small junctional scars and potential off-target insertions at spontaneous DSB sites within the plant genome (Table 1). These systems excel in throughput and fragment size, making them valuable for applications in functional genomics, stacking of synthetic pathways’ genes, and trait pyramiding. Currently, NHEJ-based technologies such as TagBIT and DOTI, PrimeRoot/PCE, and TATSI are showing promise for plant research and breeding applications. However, GT and PE still face challenges related to efficiency, particularly with large DNA fragments. PrimeRoot/PCE and TATSI systems may be more complex to design (Table 1). For future advancements, integrating these technologies with efficient in planta or replicon-based delivery systems will be essential for achieving scalability.

Collectively, these systems represent a continuum of methods for DNA integration, ranging from simple ligation-based integration (NHEJ-based DNA insertion) to high-fidelity template-based repair (such as GT and PE), PE-recombinase-mediated insertion (PrimeRoot/PCE), and transposase-assisted insertion (like TATSI). Each method addresses the limitations of the others: NHEJ-based DNA insertion offers versatility, GT ensures precision, PE minimizes genomic disruption, and PrimeRoot/PCE and TATSI enable greater cargo capacity. A hybrid approach that combines the programmability of CRISPR nucleases, the accuracy of GT and PE, and the cargo-handling capabilities of PrimeRoot/PCE and TATSI may ultimately create a unified platform for predictable and heritable DNA insertion in plants.

The advancement of precise DNA insertion technologies has evolved them from theoretical concepts into practical tools, enabling a wide range of applications in both fundamental plant biology and applied biotechnology. These applications include subtle modifications aimed at studying gene function, as well as the integration of entire multi-gene pathways to enhance crop improvement. In this section, we will examine three key areas where these technologies are having a transformative impact: in-locus protein tagging, cis-regulatory element engineering, and the targeted insertion of beneficial genes.

In-locus tagging for gene functioning

Precise in-locus tagging of endogenous proteins in plants has advanced rapidly, offering tools for real-time tracking, quantification, and functional analysis under native expression contexts¹³⁴. Tian and colleagues made a significant contribution by developing the TagBIT system, which allows for the insertion of a small HiBiT luminescent peptide tag into the coding sequences of endogenous genes in rice. This system facilitates the sensitive detection of tagged proteins through luminescence-based assays, immunoblotting, and immunoprecipitation (Fig. 4A). The tagged lines were consistently passed down through generations, demonstrating the system’s reliability and usefulness for quantitative proteomics in crops³⁸.

PE has also emerged as a promising method for small epitope tag insertion (Supplementary Data 2). Using improved prime editor variants, researchers efficiently inserted immunogenic tags such as 6 × His or HA tags precisely at the C-terminus of rice genes, enabling in-locus gene functioning in rice. This approach avoids DSBs and donor DNA templates, offering a precise and efficient strategy for seamless protein tagging in plants^122,123. However, PE-based gene tagging has not been efficient with long sequences and in dicots, thereby demanding significant improvement in the systems^125,126.

For inserting larger tags, such as full-length fluorescent proteins (e.g., GFP), CRISPR-Cas-based GT remains the primary method for enabling heritable, in-locus tagging, though its efficiency still requires improvement (Supplementary Data 2). This approach enables protein localization and dynamics to be visualized directly under endogenous regulation. While some proof-of-concept studies were limited to transient expression or used alternative luminescence reporters, the ultimate goal is to generate stably transformed plants with functional, tagged proteins, laying a foundation for advanced proteomics and cell biology in planta.

In summary, recent advancements in in-locus protein tagging and functional modification techniques for plants emphasize the expanding toolkit available for researchers. These approaches include small epitope tags, luciferase reporters, and enhancer elements, offering precise, heritable, and biologically relevant methods to study plant proteins and gene regulation directly within their natural context. As delivery and editing technologies continue to advance, in-locus tagging is set to become an essential practice in plant functional genomics, crop improvement, and synthetic biology.

Cis-regulatory element insertion

The targeted insertion of cis-regulatory elements (CREs), such as enhancers and promoters, into plant genomes is emerging as a transformative strategy for fine-tuning gene expression without altering coding sequences. Unlike traditional overexpression methods, the insertion of CREs enables the endogenous regulation of gene activity, allowing for stable and contextually appropriate gene activation. This approach also differs from conventional CRISPR-Cas activation systems, which rely on transient dCas-effector proteins. A recent groundbreaking study by Yao and colleagues developed a precise strategy for inserting short transcriptional enhancers (STEs) upstream of endogenous genes in rice (Fig. 4B). This approach led to a significant increase in transcriptional activity, achieving up to an 869-fold enhancement while maintaining the native gene structure. By simultaneously inserting multiple short enhancer modules, they realized strong, stable, and heritable gene activation across generations T0 to T3. This method enables precise adjustment of gene expression strength, which is crucial for optimizing traits related to growth, metabolism, and stress responses¹³⁵. Shen et al. improved gene expression in rice by inserting the viral-derived translational enhancer AMVE into the 5′ untranslated region (5′ UTR) of target genes¹³⁶. This enhancement increased reporter activity by approximately 8.5-fold in rice protoplasts. When AMVE was inserted upstream of the WRKY71 gene, it boosted protein levels by up to 2.8-fold without altering mRNA abundance, resulting in improved resistance to bacterial blight. Furthermore, incorporating AMVE into the SKC1 gene enhanced salt tolerance in rice seedlings, demonstrating its potential for crop improvement¹³⁶. In maize, the insertion of a trimer of a palindromic 12-bp plant enhancer sequence (5′-GTAAGCGCTTAC-3′) into the core promoter of eleven target genes resulted in consistent and significant increases in the expression of these genes in both roots and leaves⁹⁴. Kumar and colleagues employed the DOTI method to insert transcriptional enhancers, specifically distinct TAL effector-binding elements, into the promoter of the rice recessive xa23 allele. These insertions allowed for pathogen-inducible expression of xa23, resulting in heritable resistance to bacterial blight⁹⁷. The results show that targeted enhancer insertion can modify gene expression in response to environmental factors, providing an effective strategy for engineering disease resistance in crops.

Recently, using a CRISPR-Cas-based GT method, Ke and colleagues focused on stress-responsive genes in rice that are related to drought and heat tolerance by inserting synthetic cis-regulatory elements (40–120 bp) upstream of the native promoters. The scarless insertion efficiency reached up to 10%¹³⁷. Functional tests demonstrated that plants with specific genetic insertions showed significantly increased expression of target genes in response to stress, resulting in improved performance under abiotic stress conditions. This research offers a promising strategy for fine-tuning the expression of endogenous genes to enhance crop resilience without the need for foreign gene introduction. Advances in gene technology build upon the groundbreaking work related to replicon-based delivery systems. For instance, the geminiviral replicon approach facilitates the precise insertion of regulatory sequences in important crop species, such as tomatoes. Cermak and his team have illustrated that replicon-enhanced homology-directed repair (HDR) can be achieved using the remarkable TALEN and CRISPR-SpCas9 technologies. They successfully inserted a CaMV 35S promoter before the SlANT1 open reading frame (ORF), which significantly increased anthocyanin accumulation in tomato fruits³³. Vu and colleagues also demonstrated a similar approach using LbCas12a³⁴. These foundational platforms, initially designed for GT enhancement experiments, demonstrate that the concept can also be adapted for inserting regulatory elements into plants.

In-locus beneficial gene/allele insertion

The insertion of genes or alleles into specific locations within the plant genome is a crucial aspect of precise plant genetic engineering. This technique enables the introduction of beneficial traits, such as pest resistance, stress tolerance, and improved nutritional content. Traditional methods primarily relied on AMT and random T-DNA integration, which often resulted in unpredictable gene expression due to position effects or gene silencing. Recent advancements using CRISPR-Cas technology have focused on achieving precise, efficient, and marker-free insertion of genes into predetermined genomic locations, thereby enhancing both predictability and safety in the genetic engineering process. Dong and colleagues made significant progress by inserting a 5.2 kb carotenoid biosynthesis cassette into a GSH locus in rice using the CRISPR-Cas9 gene editing method¹³⁸ (Fig. 4C). The system successfully produced homozygous, marker-free rice lines with high carotenoid levels and no reduction in yield. This study demonstrated that large gene cassettes can be integrated stably and passed on to the next generations¹³⁸.

The GT-based DNA insertion method holds promise for inserting genes and alleles; however, its low and inconsistent efficiency across various species restricts its broader use in plant breeding. This limitation has prompted the development of alternative strategies for inserting large DNA fragments. One notable example is the PrimeRoot system, which utilizes the high precision of PE to first install a recombination site at the targeted site. This initial step allows for the subsequent, efficient integration of a large donor cassette through an SSR (Fig. 3B)³⁹. Using the PrimeRoot method, Sun and colleagues successfully integrated a PigmR gene cassette into a designated GSH site, with the cassette being driven by an Act1 promoter to provide resistance against rice blast disease. Importantly, the rice plants that contained this precise insertion showed improved resistance to blast disease, demonstrating a significant agronomic benefit³⁹. This research marks a significant advancement in crop genome engineering by providing a powerful alternative method for precisely stacking genes or regulatory elements to achieve desirable traits.

Despite the impressive advancements and various applications of precise DNA insertion in plants, there are several significant challenges that hinder its routine and widespread adoption. These obstacles range from the molecular complexities involved in the editing process to the difficulties associated with plant regeneration and the overarching regulatory approval framework. Tackling these issues is the main focus of ongoing research and is crucial for unlocking the full potential of these technologies.

DNA insertion with scars or scarlessness

A key factor at the molecular level is the characteristics of the final edited allele, specifically whether the insertion is seamless or leaves behind small sequence “scars.” Achieving scarless DNA insertion is the ultimate goal in precise plant genome engineering. This is particularly important for applications such as in-locus tagging, gene fusion, and partial gene replacement, where it is essential to preserve the reading frame.

Scarless editing is achievable through methods like GT, DOTI, and PE (Table 1). However, each approach poses specific technical challenges: GT and DOTI require the precise selection of CRISPR-Cas cleavage sites and, in the case of DOTI, the synthesis of chemically modified long dsODNs; PE requires the design of longer and more complex RTTs and pegRNAs.

In contrast, DNA insertions accompanied by minor scars, such as small indels at the junctions, are often acceptable, particularly when full gene cassettes or regulatory elements are inserted at genomic GSH sites or upstream regulatory regions. These scars may even be tolerated in in-locus tagging applications using the GRAND-PE approach, where short deletions at the insertion site do not impair gene function. Indeed, GRAND-PE often results in higher insertion efficiency and allows for longer sequence insertions compared to the scarless paired pegRNA strategy at identical loci^120,121,123.

The challenges of insertion efficiency and payload size

One of the most significant technical barriers is the interconnected challenges of insertion efficiency and the size of the DNA payload that can be integrated. These two factors are often inversely correlated, representing the core limitations of current platforms.

For medium-sized insertions (up to several hundred base pairs), several competing strategies exist. The PE-based method is promising for scarless editing, but its efficiency is often dramatically reduced when the insertion length is increased (Supplementary Data 2). This is believed to stem from the increasing complexity and secondary structures (e.g., internal stem-loops) of the required RTT, which can hinder the reverse transcription process⁴⁰. Recent advances, such as the template-jumping PE (TJ-PE) strategy, have pushed the limit to ~400-bp in rice¹²⁴, but this is still short of the ~800-bp achieved in human cells¹¹⁹, indicating room for improvement. Potential solutions being actively explored include engineered PE variants (e.g., PE6) that can better navigate complex RTTs¹²⁷, AI-based engineering of PE components or small molecules, which can interfere with the MMR, such as the MLH1 binder¹³⁹, and strategies like GRAND-PE that tolerate small junctional scars to achieve higher efficiency with longer sequences^123,128.

As an alternative to PE, NHEJ-based methods such as DOTI or MMEJ can also be employed. These strategies are often more efficient for generating insertions with small junctional scars but are generally inefficient for obtaining scarless products¹³⁸. The DOTI strategy, for example, can be enhanced by using chemically modified dsODNs (e.g., with phosphorylation and phosphorothioate linkages) to improve insertion frequency, but this approach is ultimately limited by the high cost and size constraints of synthetic donors^{93,97,135,140}.

Integrating large DNA fragments, such as entire gene cassettes (> 1 kb), remains a paramount challenge. The “gold standard” GT-based DNA insertion method allows scarless insertion up to ten kilobases, but its reliance on the HR pathway results in low and highly variable efficiency among plant species and delivery methods (Supplementary Data 2). Newer platforms that combine with SSRs (e.g., PrimeRoot)³⁹ and its upgraded version⁴² or CRISPR–transposase systems (e.g., TATSI) have successfully enabled kilobase-scale insertion⁴³. However, these may leave recombination scars and still be limited by the efficiency of the initial editing step, or leave recombination scars or rely on pre-installed docking sites, making true seamless integration difficult.

These molecular challenges are further compounded by the universal bottleneck of DNA delivery. While particle bombardment can deliver high doses of donor DNA, it is impractical for many species and can cause unwanted genomic damage. Conversely, the widely used AMT method is often limited in the dose of donor DNA it can deliver, a key bottleneck for both GT and NHEJ-mediated strategies. Several strategies can enhance AMT, such as increasing the binary vector copy number or pre-activating Agrobacterium cells with acetosyringone, which is particularly relevant for large binary vectors used in GT^141,142. Alternatively, the geminiviral replicon system integrated into AMT can amplify donor concentration, although it hardly delivers overly long DNA fragments due to constraints associated with the viral genome structure^33,34,95,111. However, even with replicons, temporal coordination remains a significant hurdle; achieving the synchronized cleavages at both the genomic and donor sites required for some NHEJ-mediated strategies, or ensuring the genomic DSB coincides with the maximum local concentration of the donor template for GT, is challenging¹⁰².

Ultimately, to enhance insertion rates, a holistic and multi-faceted approach is essential: concurrently improving core editors, manipulating cellular DNA repair pathways, and optimizing delivery and plant regeneration systems.

Tissue culture and delivery dependencies

In addition to the molecular events involved, a significant practical challenge is the reliance on tissue culture methods for transformation and regeneration. The process of inserting DNA into plants is often constrained by the dependence on these methods, which typically limit applicability to a narrow range of genotypes that are capable of forming callus and regenerating somatic shoots^143,144. Protoplast-based techniques present challenges such as low regeneration efficiency and poor integration rates¹⁴⁵. Furthermore, traditional methods such as AMT and particle bombardment are dependent on the genotype and may result in unwanted genomic alterations, including random double-strand breaks or chromosomal rearrangements^146,147.

To address these challenges, several innovative delivery strategies are emerging. Nanocarriers and carbon nanotubes provide genotype-independent, non-integrative delivery systems^148,149. Viral vectors, such as Potato virus X, enable transient expression and facilitate the efficient amplification of donor templates through replicons¹⁵⁰. Alternative methods, such as de novo meristem induction and in planta transformation, aim to completely eliminate the need for tissue culture^144,151. Moreover, methods such as pollen magnetofection and floral dip enhance the editing capabilities across various genotypes, offering new flexibility and efficiency in plant genome engineering^151,152, albeit concerns regarding repeatability remain an issue¹⁵³.

Regulatory concerns

Finally, even a well-engineered plant faces the challenges posed by a complex and fragmented global regulatory landscape. The classification of genome-edited crops, which contain precisely inserted DNA sequences developed using techniques such as GT, PE, PrimeRoot, DOTI, or TATSI, primarily depends on the size, origin, and function of the inserted allele. Regulatory frameworks initially categorized these edits as either SDN-2 or SDN-3. SDN-2 refers to small, precise edits made using a homologous repair template that does not involve the integration of foreign DNA. In contrast, SDN-3 includes larger or transgenic insertions and is generally subject to regulations governing GMOs^1,154. The difference between SDN-2 and SDN-3 can be ambiguous, especially for insertions that incorporate synthetic or cisgenic sequences resembling natural alleles.

Regulatory authorities employ various approaches to evaluating genetic products. Initially, the European Union (EU) classified all genetic edits that involve inserted DNA under the GMO law, regardless of their precision or the origin of the inserted sequence. However, a new category for “new genomic techniques” is currently being proposed, which could relax restrictions on modifications that involve fewer than 20 nucleotides. Notably, the EU Commission is considering the possibility of deregulating gene insertions that originate from the gene pool of the same species (cisgenesis) and do not disrupt an endogenous gene¹⁵⁵. Countries such as the United States, Japan, and China take a more nuanced approach to genetic modification. For instance, Japan allows small insertions without vector or marker sequences to bypass GMO regulations, as long as the product does not contain any transgenes. In the United States, the focus is on the risk profile of the final product rather than the specific editing method used. Similarly, China’s 2023 guidelines establish a streamlined approval process for gene-edited crops that do not contain foreign DNA, although larger insertions face stricter scrutiny¹⁵⁶.

The varying interpretations of regulations create uncertainty for crops that have precisely inserted alleles, even when these edits resemble natural variants or do not involve the use of transgenes. With the emergence of new tools like CRISPR-Cas-based GT, PE, PrimeRoot, and DOTI, regulators are challenged to find a balance between scientific accuracy and consistent policy. Furthermore, there is still no clear, science-based international consensus, which complicates the global adoption of edited crops in agriculture and trade.

The field of plant genome engineering is rapidly transitioning from random, transgene-based integration to precise, marker-free DNA insertion enabled by CRISPR-Cas systems and related tools such as GT, PE, PrimeRoot, DOTI, and TATSI. These approaches are unlocking broad applications in functional genomics, synthetic biology, and sustainable agriculture, including in-locus protein tagging, regulatory element tuning, and trait enhancement through gene stacking. However, efficiently inserting large DNA fragments into complex plant genomes remains a formidable challenge, hindered by low locus-specific efficiency, technical design complexities, and delivery constraints. Overcoming these barriers will require integrated strategies, such as advancing next-generation PE systems, applying AI to evolve DNA-insertion enzymes, manipulating repair pathways, and optimizing transformation and regeneration. As these technologies mature and align with biosafety and regulatory standards, CRISPR-Cas-based precision DNA insertion will underpin the next generation of crop improvement, driving the creation of climate-resilient, nutritionally enhanced, and publicly accepted elite varieties.

In the near future, efforts will likely concentrate on developing high-efficiency, programmable systems that enable precise integration of genes, regulatory elements, and synthetic circuits. To address the bottlenecks of efficiency and payload size, emerging platforms such as CRISPR-transposase fusions⁴³, PE-based recombinase-assisted insertions^39,42, and novel systems, such as the programmable bridge recombination^157,158, offer new possibilities for kilobase-scale, heritable DNA insertions without the need for homology arms or selection markers. Furthermore, alternative systems of CRISPR-Cas, such as the tandem interspaced guide RNA (TIGR)-TIGR-associated (Tas) systems (TIGR-Tas)¹⁵⁹, may be harnessed for precise DNA insertion in plants. Furthermore, emerging AI-assisted protein design and engineering can be employed to further evolve PE components (nCas9 and RT enzymes) and PE-facilitating peptides, such as MLH1 small binder¹³⁹, NC^37,160, and La¹⁶¹. AI-based protein evolution can also be explored to improve transposases for more efficient TATSI. These innovations are vital for advancing precise, broad-scale DNA insertion in plants^{162,163,164,165,166,167,168}.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

1. Vu, T. V., Das, S., Hensel, G. & Kim, J.-Y. Genome editing and beyond: what does it mean for the future of plant breeding?. Planta 255, 130 (2022).

   Article  PubMed  PubMed Central  Google Scholar 

2. Mullins, E. et al. Updated scientific opinion on plants developed through cisgenesis and intragenesis. EFSA J. 20, e07621 (2022).

   PubMed  PubMed Central  Google Scholar 

3. Dong, O. X. & Ronald, P. C. Targeted DNA insertion in plants. Proc. Natl. Acad. Sci. USA 118, e2004834117 (2021).

   Article  CAS  PubMed  PubMed Central  Google Scholar 

4. Thomson, G., Dickinson, L. & Jacob, Y. Genomic consequences associated with Agrobacterium-mediated transformation of plants. Plant J. 117, 342–363 (2024).

   Article  ADS  CAS  PubMed  Google Scholar 

5. Olhoft, P. M., Flagel, L. E. & Somers, D. A. T-DNA locus structure in a large population of soybean plants transformed using the Agrobacterium-mediated cotyledonary-node method. Plant Biotechnol. J. 2, 289–300 (2004).

   Article  CAS  PubMed  Google Scholar 

6. Daxinger, L. et al. Unexpected silencing effects from T-DNA tags in Arabidopsis. Trends Plant Sci. 13, 4–6 (2008).

   Article  CAS  PubMed  Google Scholar 

7. Kirchhoff, J. et al. Gene expression variability between randomly and targeted transgene integration events in tobacco suspension cell lines. Plant Biotechnol. Rep. 14, 451–458 (2020).

   Article  Google Scholar 

8. Schubert, D. et al. Silencing in Arabidopsis T-DNA transformants: the predominant role of a gene-specific RNA sensing mechanism versus position effects. Plant Cell 16, 2561–2572 (2004).

   Article  CAS  PubMed  PubMed Central  Google Scholar 

9. Gao, Y. & Zhao, Y. Epigenetic suppression of T-DNA insertion mutants in Arabidopsis. Mol. Plant 6, 539–545 (2013).

   Article  MathSciNet  CAS  PubMed  Google Scholar 

10. Ruprecht, C., Carroll, A. & Persson, S. T-DNA-induced chromosomal translocations in feronia and anxur2 mutants reveal implications for the mechanism of collapsed pollen due to chromosomal rearrangements. Mol. Plant 7, 1591–1594 (2014).

    Article  CAS  PubMed  Google Scholar 

11. Liu, J. et al. Genome-scale sequence disruption following biolistic transformation in rice and maize. Plant Cell 31, 368–383 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

12. Fossi, M., Amundson, K., Kuppu, S., Britt, A. & Comai, L. Regeneration of Solanum tuberosum plants from protoplasts induces widespread genome instability. Plant Physiol. 180, 78–86 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

13. Raabe, K., Sun, L., Schindfessel, C., Honys, D. & Geelen, D. A word of caution: T-DNA-associated mutagenesis in plant reproduction research. J. Exp. Bot. 75, 3248–3258 (2024).

    Article  CAS  PubMed  Google Scholar 

14. Kleinboelting, N. et al. The structural features of thousands of T-DNA insertion sites are consistent with a double-strand break repair-based insertion mechanism. Mol. Plant 8, 1651–1664 (2015).

    Article  CAS  PubMed  Google Scholar 

15. Sauer, B. & Henderson, N. Targeted insertion of exogenous DNA into the eukaryotic genome by the Cre recombinase. N. Biol. 2, 441–449 (1990).

    CAS  Google Scholar 

16. Golic, K. G. & Lindquist, S. The FLP recombinase of yeast catalyzes site-specific recombination in the Drosophila genome. Cell 59, 499–509 (1989).

    Article  CAS  PubMed  Google Scholar 

17. Onouchi, H. et al. Operation of an efficient site-specific recombination system of Zygosaccharomyces rouxii in tobacco cells. Nucleic Acids Res. 19, 6373–6378 (1991).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

18. Gottfried, P. et al. Site-specific recombination in Arabidopsis plants promoted by the Integrase protein of coliphage HK022. Plant Mol. Biol. 57, 435–444 (2005).

    Article  CAS  PubMed  Google Scholar 

19. Rubtsova, M. et al. Expression of active Streptomyces phage phiC31 integrase in transgenic wheat plants. Plant Cell Rep. 27, 1821–1831 (2008).

    Article  CAS  PubMed  Google Scholar 

20. Ow, D. W. GM maize from site-specific recombination technology, what next?. Curr. Opin. Biotechnol. 18, 115–120 (2007).

    Article  CAS  PubMed  Google Scholar 

21. Albert, H., Dale, E. C., Lee, E. & Ow, D. W. Site-specific integration of DNA into wild-type and mutant lox sites placed in the plant genome. Plant J. 7, 649–659 (1995).

    Article  CAS  PubMed  Google Scholar 

22. Russell, S. H., Hoopes, J. L. & Odell, J. T. Directed excision of a transgene from the plant genome. Mol. Gen. Genet. 234, 49–59 (1992).

    Article  CAS  PubMed  Google Scholar 

23. Dale, E. C. & Ow, D. W. Gene transfer with subsequent removal of the selection gene from the host genome. Proc. Natl. Acad. Sci. USA 88, 10558–10562 (1991).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

24. Hare, P. D. & Chua, N.-H. Excision of selectable marker genes from transgenic plants. Nat. Biotechnol. 20, 575–580 (2002).

    Article  CAS  PubMed  Google Scholar 

25. Hohn, B., Levy, A. A. & Puchta, H. Elimination of selection markers from transgenic plants. Curr. Opin. Biotechnol. 12, 139–143 (2001).

    Article  CAS  PubMed  Google Scholar 

26. Vu, T. V. et al. Challenges and perspectives in homology-directed gene targeting in monocot plants. Rice 12, 95 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

27. Puchta, H. & Fauser, F. Gene targeting in plants: 25 years later. Int. J. Dev. Biol. 57, 629–637 (2013).

    Article  CAS  PubMed  Google Scholar 

28. Paszkowski, J., Baur, M., Bogucki, A. & Potrykus, I. Gene targeting in plants. EMBO J. 7, 4021–4026 (1988).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

29. Terada, R., Urawa, H., Inagaki, Y., Tsugane, K. & Iida, S. Efficient gene targeting by homologous recombination in rice. Nat. Biotechnol. 20, 1030–1034 (2002).

    Article  CAS  PubMed  Google Scholar 

30. Puchta, H., Dujon, B. & Hohn, B. Homologous recombination in plant cells is enhanced by in vivo induction of double strand breaks into DNA by a site-specific endonuclease. Nucleic Acids Res. 21, 5034–5040 (1993).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

31. Puchta, H., Dujon, B. & Hohn, B. Two different but related mechanisms are used in plants for the repair of genomic double-strand breaks by homologous recombination. Proc. Natl. Acad. Sci. USA 93, 5055–5060 (1996).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

32. Fauser, F. et al. In planta gene targeting. Proc. Natl. Acad. Sci. USA 109, 7535–7540 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

33. Cermak, T., Baltes, N. J., Cegan, R., Zhang, Y. & Voytas, D. F. High-frequency, precise modification of the tomato genome. Genome Biol. 16, 232 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

34. Vu, T. V. et al. Highly efficient homology-directed repair using CRISPR/Cpf1-geminiviral replicon in tomato. Plant Biotechnol. J. 18, 2133–2143 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

35. Anzalone, A. V. et al. Search-and-replace genome editing without double-strand breaks or donor DNA. Nature 576, 149–157 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

36. Lin, Q. et al. Prime genome editing in rice and wheat. Nat. Biotechnol. 38, 582–585 (2020).

    Article  CAS  PubMed  Google Scholar 

37. Vu, T. V. et al. Optimized dicot prime editing enables heritable desired edits in tomato and Arabidopsis. Nat. Plants 10, 1502–1513 (2024).

    Article  CAS  PubMed  Google Scholar 

38. Tian, Y. et al. Rapid and dynamic detection of endogenous proteins through in locus tagging in rice. Plant Commun. 5, 101040 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

39. Sun, C. et al. Precise integration of large DNA sequences in plant genomes using PrimeRoot editors. Nat. Biotechnol. 42, 316–327 (2024).

    Article  CAS  PubMed  Google Scholar 

40. Vu, T. V., Nguyen, N. T., Kim, J., Hong, J. C. & Kim, J.-Y. Prime editing: mechanism insight and recent applications in plants. Plant Biotechnol. J. 22, 19–36 (2024).

    Article  CAS  PubMed  Google Scholar 

41. Anzalone, A. V. et al. Programmable deletion, replacement, integration and inversion of large DNA sequences with twin prime editing. Nat. Biotechnol. 40, 731–740 (2022).

    Article  CAS  PubMed  Google Scholar 

42. Sun, C. et al. Iterative recombinase technologies for efficient and precise genome engineering across kilobase to megabase scales. Cell 188, 4693–4710.e4615 (2025).

    Article  CAS  PubMed  Google Scholar 

43. Liu, P. et al. Transposase-assisted target-site integration for efficient plant genome engineering. Nature 631, 593–600 (2024).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

44. Holmes-Davis, R. & Comai, L. Nuclear matrix attachment regions and plant gene expression. Trends Plant Sci. 3, 91–97 (1998).

    Article  Google Scholar 

45. Li, S. et al. Optimization of Agrobacterium-mediated transformation in soybean. Front. Plant Sci. 8, 246 (2017).

    ADS  PubMed  PubMed Central  Google Scholar 

46. Kikkert, J. R., Vidal, J. R. & Reisch, B. I. Stable transformation of plant cells by particle bombardment/biolistics. Methods Mol. Biol. 286, 61–78 (2005).

    CAS  PubMed  Google Scholar 

47. Gehl, J. Electroporation: theory and methods, perspectives for drug delivery, gene therapy and research. Acta Physiol. Scand. 177, 437–447 (2003).

    Article  CAS  PubMed  Google Scholar 

48. Deshayes, A., Herrera-Estrella, L. & Caboche, M. Liposome-mediated transformation of tobacco mesophyll protoplasts by an Escherichia coli plasmid. Embo J. 4, 2731–2737 (1985).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

49. Zaidi, S. S. & Mansoor, S. Viral vectors for plant genome engineering. Front. Plant Sci. 8, 539 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

50. Zarnitsyn, V. G. & Prausnitz, M. R. Physical parameters influencing optimization of ultrasound-mediated DNA transfection. Ultrasound Med. Biol. 30, 527–538 (2004).

    Article  PubMed  Google Scholar 

51. Bélanger, J. G., Copley, T. R., Hoyos-Villegas, V., Charron, J.-B. & O’Donoughue, L. A comprehensive review of in planta stable transformation strategies. Plant Methods 20, 79 (2024).

    Article  PubMed  PubMed Central  Google Scholar 

52. Pereira, A. A transgenic perspective on plant functional genomics. Transgenic Res. 9, 245–260 (2000).

    Article  CAS  PubMed  Google Scholar 

53. Ye, X. et al. Engineering the provitamin A (β-carotene) biosynthetic pathway into (carotenoid-free) rice endosperm. Science 287, 303–305 (2000).

    Article  ADS  CAS  PubMed  Google Scholar 

54. Paul, M. & Ma, J. K. Plant-made pharmaceuticals: leading products and production platforms. Biotechnol. Appl. Biochem. 58, 58–67 (2011).

    Article  CAS  PubMed  Google Scholar 

55. Park, C. H. et al. Comparison of plasmid curing efficiency across five lactic acid bacterial species. J. Microbiol. Biotechnol. 34, 2385–2395 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

56. Tzfira, T., Frankman, L. R., Vaidya, M. & Citovsky, V. Site-specific integration of Agrobacterium tumefaciens T-DNA via double-stranded intermediates. Plant Physiol. 133, 1011–1023 (2003).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

57. Salomon, S. & Puchta, H. Capture of genomic and T-DNA sequences during double-strand break repair in somatic plant cells. EMBO J. 17, 6086–6095 (1998).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

58. Chilton, M. D. & Que, Q. Targeted integration of T-DNA into the tobacco genome at double-stranded breaks: new insights on the mechanism of T-DNA integration. Plant Physiol. 133, 956–965 (2003).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

59. van Kregten, M. et al. T-DNA integration in plants results from polymerase-θ-mediated DNA repair. Nat. Plants 2, 16164 (2016).

    Article  PubMed  Google Scholar 

60. Kralemann, L. E. M. et al. Distinct mechanisms for genomic attachment of the 5’ and 3’ ends of Agrobacterium T-DNA in plants. Nat. Plants 8, 526–534 (2022).

    Article  CAS  PubMed  Google Scholar 

61. Deng, S. K., Gibb, B., de Almeida, M. J., Greene, E. C. & Symington, L. S. RPA antagonizes microhomology-mediated repair of DNA double-strand breaks. Nat. Struct. Mol. Biol. 21, 405–412 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

62. Schouten, H. J., Krens, F. A. & Jacobsen, E. Cisgenic plants are similar to traditionally bred plants: international regulations for genetically modified organisms should be altered to exempt cisgenesis. EMBO Rep. 7, 750–753 (2006).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

63. Holme, I. B., Wendt, T. & Holm, P. B. Intragenesis and cisgenesis as alternatives to transgenic crop development. Plant Biotechnol. J. 11, 395–407 (2013).

    Article  CAS  PubMed  Google Scholar 

64. Holme, I. et al. A cisgenic approach for improving the bioavailability of phosphate in the barley grain. ISB News Report, 3, 8–11 (2012).

65. Jo, K. R. et al. Development of late blight resistant potatoes by cisgene stacking. BMC Biotechnol. 14, 50 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

66. Vanblaere, T. et al. The development of a cisgenic apple plant. J. Biotechnol. 154, 304–311 (2011).

    Article  CAS  PubMed  Google Scholar 

67. Rommens, C. M. et al. Crop improvement through modification of the plant’s own genome. Plant Physiol. 135, 421–431 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

68. Rommens, C. M. et al. Plant-derived transfer DNAs. Plant Physiol. 139, 1338–1349 (2005).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

69. Rommens, C. M., Ye, J., Richael, C. & Swords, K. Improving potato storage and processing characteristics through all-native DNA transformation. J. Agric. Food Chem. 54, 9882–9887 (2006).

    Article  CAS  PubMed  Google Scholar 

70. Holme, I. B. et al. Cisgenic barley with improved phytase activity. Plant Biotechnol. J. 10, 237–247 (2012).

    Article  CAS  PubMed  Google Scholar 

71. Maltseva, E. R. et al. Assessment of cisgenic bread wheat lines carrying class I chitinase gene to leaf rust. J. Biotechnol. 280, S80–S81 (2018).

    Article  Google Scholar 

72. Joshi, S. G. et al. Functional analysis and expression profiling of HcrVf1 and HcrVf2 for development of scab resistant cisgenic and intragenic apples. Plant Mol. Biol. 75, 579–591 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

73. Dhekney, S. A., Li, Z. T. & Gray, D. J. Grapevines engineered to express cisgenic Vitis vinifera thaumatin-like protein exhibit fungal disease resistance. Vitr. Cell. Dev. Biol. -Plant 47, 458–466 (2011).

    Article  CAS  Google Scholar 

74. Jinek, M. et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337, 816–821 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

75. Grindley, N. D. F., Whiteson, K. L. & Rice, P. A. Mechanisms of site-specific recombination. Annu. Rev. Biochem. 75, 567–605 (2006).

    Article  CAS  PubMed  Google Scholar 

76. Thomson, J. G., Chan, R., Thilmony, R., Yau, Y.-Y. & Ow, D. W. PhiC31 recombination system demonstrates heritable germinal transmission of site-specific excision from the Arabidopsis genome. BMC Biotechnol. 10, 17 (2010).

    Article  PubMed  PubMed Central  Google Scholar 

77. Ow, D. W. The long road to recombinase-mediated plant transformation. Plant Biotechnol. J. 14, 441–447 (2016).

    Article  CAS  PubMed  Google Scholar 

78. Offringa, R. et al. Extrachromosomal homologous recombination and gene targeting in plant cells after Agrobacterium mediated transformation. EMBO J. 9, 3077–3084 (1990).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

79. Weinthal, D., Tovkach, A., Zeevi, V. & Tzfira, T. Genome editing in plant cells by zinc finger nucleases. Trends Plant Sci. 15, 308–321 (2010).

    Article  CAS  PubMed  Google Scholar 

80. Malzahn, A., Lowder, L. & Qi, Y. Plant genome editing with TALEN and CRISPR. Cell Biosci. 7, 21 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

81. Zhang, Y. et al. Transcription activator-like effector nucleases enable efficient plant genome engineering. Plant Physiol. 161, 20–27 (2013).

    Article  ADS  CAS  PubMed  Google Scholar 

82. Budhagatapalli, N., Rutten, T., Gurushidze, M., Kumlehn, J. & Hensel, G. Targeted modification of mene function exploiting homology-directed repair of TALEN-mediated double-strand breaks in barley. G3 (Bethesda) 5, 1857–1863 (2015).

    Article  CAS  PubMed  Google Scholar 

83. Li, T., Liu, B., Chen, C. Y. & Yang, B. TALEN-mediated homologous recombination produces site-directed DNA base change and herbicide-resistant rice. J. Genet Genom. 43, 297–305 (2016).

    Article  CAS  Google Scholar 

84. Endo, M. et al. Molecular breeding of a novel herbicide-tolerant rice by gene targeting. Plant J. 52, 157–166 (2007).

    Article  CAS  PubMed  Google Scholar 

85. Wang, X. H. et al. Positive–negative selection for homologous recombination in Arabidopsis. Gene 272, 249–255 (2001).

    Article  CAS  Google Scholar 

86. Shaked, H., Melamed-Bessudo, C. & Levy, A. A. High-frequency gene targeting in Arabidopsis plants expressing the yeast RAD54 gene. Proc. Natl. Acad. Sci. USA 102, 12265–12269 (2005).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

87. Ishino, Y., Shinagawa, H., Makino, K., Amemura, M. & Nakata, A. Nucleotide sequence of the iap gene, responsible for alkaline phosphatase isozyme conversion in Escherichia coli, and identification of the gene product. J. Bacteriol. 169, 5429–5433 (1987).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

88. Mojica, F. J. M., Díez-Villaseñor, C. S., García-Martínez, J. & Soria, E. Intervening sequences of regularly spaced prokaryotic repeats derive from foreign genetic elements. J. Mol. Evol. 60, 174–182 (2005).

    Article  ADS  CAS  PubMed  Google Scholar 

89. Doudna, J. A. The promise and challenge of therapeutic genome editing. Nature 578, 229–236 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

90. Li, B., Sun, C., Li, J. & Gao, C. Targeted genome-modification tools and their advanced applications in crop breeding. Nat. Rev. Genet. 25, 603–622 (2024).

    Article  CAS  PubMed  Google Scholar 

91. Komor, A. C., Kim, Y. B., Packer, M. S., Zuris, J. A. & Liu, D. R. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature 533, 420–424 (2016).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

92. Gaudelli, N. M. et al. Programmable base editing of A•T to G•C in genomic DNA without DNA cleavage. Nature 551, 464–471 (2017).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

93. Lu, Y. et al. Targeted, efficient sequence insertion and replacement in rice. Nat. Biotechnol. 38, 1402–1407 (2020).

    Article  CAS  PubMed  Google Scholar 

94. Claeys, H. et al. Coordinated gene upregulation in maize through CRISPR/Cas-mediated enhancer insertion. Plant Biotechnol. J. 22, 16–18 (2024).

    Article  CAS  PubMed  Google Scholar 

95. Tian, Y. et al. High-throughput genome editing in rice with a virus-based surrogate system. J. Integr. Plant Biol. 65, 646–655 (2023).

    Article  CAS  PubMed  Google Scholar 

96. Li, Y. et al. CRISPR/Cas9-mediated seamless gene replacement in protoplasts expands the resistance spectrum to TMV-U1 strain in regenerated Nicotiana tabacum. Plant Biotechnol. J. 21, 2641–2653 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

97. Kumar, J. et al. Efficient protein tagging and cis-regulatory element engineering via precise and directional oligonucleotide-based targeted insertion in plants. Plant Cell 35, 2722–2735 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

98. Sfeir, A. & Symington, L. S. Microhomology-mediated end joining: a back-up survival mechanism or dedicated pathway?. Trends Biochem. Sci. 40, 701–714 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

99. Vu, T. V. et al. CRISPR/Cas-based precision genome editing via microhomology-mediated end joining. Plant Biotechnol. J. 19, 230–239 (2021).

    Article  PubMed  Google Scholar 

100. Sakuma, T., Nakade, S., Sakane, Y., Suzuki, K. T. & Yamamoto, T. MMEJ-assisted gene knock-in using TALENs and CRISPR-Cas9 with the PITCh systems. Nat. Protoc. 11, 118–133 (2016).

     Article  CAS  PubMed  Google Scholar 

101. Nakade, S. et al. Microhomology-mediated end-joining-dependent integration of donor DNA in cells and animals using TALENs and CRISPR/Cas9. Nat. Commun. 5, 5560 (2014).

     Article  ADS  CAS  PubMed  Google Scholar 

102. Vu, T.V. et al. CRISPR-Cas-based microhomology-mediated end joining for precise gene replacement in plant. Preprint at bioRxiv https://www.biorxiv.org/content/10.1101/2022.08.27.505510v3 (2024).

103. Huang, T. K. & Puchta, H. CRISPR/Cas-mediated gene targeting in plants: finally a turn for the better for homologous recombination. Plant Cell Rep. 38, 443–453 (2019).

     Article  CAS  PubMed  Google Scholar 

104. Miki, D. et al. Gene targeting facilitated by engineered sequence-specific nucleases: potential applications for crop improvement. Plant Cell Physiol. 62, 752–765 (2021).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

105. Schiml, S., Fauser, F. & Puchta, H. The CRISPR/Cas system can be used as nuclease for in planta gene targeting and as paired nickases for directed mutagenesis in Arabidopsis resulting in heritable progeny. Plant J. 80, 1139–1150 (2014).

     Article  ADS  CAS  PubMed  Google Scholar 

106. Nekrasov, V., Staskawicz, B., Weigel, D., Jones, J. D. & Kamoun, S. Targeted mutagenesis in the model plant Nicotiana benthamiana using Cas9 RNA-guided endonuclease. Nat. Biotechnol. 31, 691–693 (2013).

     Article  CAS  PubMed  Google Scholar 

107. Peterson, D. et al. Advances in Agrobacterium transformation and vector design result in high-frequency targeted gene insertion in maize. Plant Biotechnol. J. 19, 2000–2010 (2021).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

108. Nishizawa-Yokoi, A., Endo, M., Ohtsuki, N., Saika, H. & Toki, S. Precision genome editing in plants via gene targeting and piggyBac-mediated marker excision. Plant J. 81, 160–168 (2015).

     Article  ADS  CAS  PubMed  Google Scholar 

109. Endo, M., Iwakami, S. & Toki, S. Precision genome editing in plants via gene targeting and subsequent break-induced single-strand annealing. Plant Biotechnol. J. 19, 563–574 (2021).

     Article  CAS  PubMed  Google Scholar 

110. Oltmanns, H. et al. Generation of backbone-free, low transgene copy plants by launching T-DNA from the Agrobacterium chromosome. Plant Physiol. 152, 1158–1166 (2010).

     Article  CAS  PubMed  Google Scholar 

111. Wang, M. et al. Gene targeting by homology-directed repair in rice using a geminivirus-based CRISPR/Cas9 system. Mol. Plant 10, 1007–1010 (2017).

     Article  CAS  PubMed  Google Scholar 

112. Miki, D., Zhang, W., Zeng, W., Feng, Z. & Zhu, J.-K. CRISPR/Cas9-mediated gene targeting in Arabidopsis using sequential transformation. Nat. Commun. 9, 1967 (2018).

     Article  ADS  PubMed  PubMed Central  Google Scholar 

113. Ali, Z. et al. Fusion of the Cas9 endonuclease and the VirD2 relaxase facilitates homology-directed repair for precise genome engineering in rice. Commun. Biol. 3, 44 (2020).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

114. Nagy, E. D. et al. Site-directed integration of exogenous DNA into the soybean genome by LbCas12a fused to a plant viral HUH endonuclease. Plant J. 111, 905–916 (2022).

     Article  ADS  CAS  PubMed  Google Scholar 

115. Zhou, Z. et al. Cas9-Rep fusion tethers donor DNA in vivo and boosts the efficiency of HDR-mediated genome editing. Plant Biotechnol. J. 23, 2006–2017 (2025).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

116. Anzalone, A. V., Koblan, L. W. & Liu, D. R. Genome editing with CRISPR-Cas nucleases, base editors, transposases and prime editors. Nat. Biotechnol. 38, 824–844 (2020).

     Article  CAS  PubMed  Google Scholar 

117. Chen, P. J. & Liu, D. R. Prime editing for precise and highly versatile genome manipulation. Nat. Rev. Genet. 24, 161–177 (2023).

     Article  CAS  PubMed  Google Scholar 

118. Chen, P. J. et al. Enhanced prime editing systems by manipulating cellular determinants of editing outcomes. Cell 184, 5635–5652.e5629 (2021).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

119. Zheng, C. et al. Template-jumping prime editing enables large insertion and exon rewriting in vivo. Nat. Commun. 14, 3369 (2023).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

120. Wang, J. et al. Efficient targeted insertion of large DNA fragments without DNA donors. Nat. Methods 19, 331–340 (2022).

     Article  CAS  PubMed  Google Scholar 

121. Li, J. et al. Prime editing-mediated precise knock-in of protein tag sequences in the rice genome. Plant Commun. 4, 100572 (2023).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

122. Li, X. et al. Efficient in situ epitope tagging of rice genes by nuclease-mediated prime editing. Plant Cell 37, koae316 (2025).

     Article  PubMed  Google Scholar 

123. Xu, R. et al. Engineering PE6 prime editors to efficiently insert tags in rice. Plant Biotechnol. J. 22, 3383–3385 (2024).

     Article  PubMed  PubMed Central  Google Scholar 

124. Li, F. et al. Targeted insertion of large DNA fragments through template-jumping prime editing in rice. Plant Biotechnol. J. 23, 2645–2647 (2025).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

125. Wang, L. et al. Spelling changes and fluorescent tagging with prime editing vectors for plants. Front. Genome Ed. 3, 617553 (2021).

     Article  PubMed  PubMed Central  Google Scholar 

126. Hong, C., Han, J. H., Hwang, G. H., Bae, S. & Seo, P. J. Genome-wide in-locus epitope tagging of Arabidopsis proteins using prime editors. BMB Rep. 57, 66–70 (2024).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

127. Doman, J. L. et al. Phage-assisted evolution and protein engineering yield compact, efficient prime editors. Cell 186, 3983–4002 e3926 (2023).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

128. Cao, Z. et al. PE6c greatly enhances prime editing in transgenic rice plants. J. Integr. Plant Biol. 66, 1864–1870 (2024).

     Article  CAS  PubMed  Google Scholar 

129. Merrick, C. A., Zhao, J. & Rosser, S. J. Serine integrases: advancing synthetic biology. ACS Synth. Biol. 7, 299–310 (2018).

     Article  CAS  PubMed  Google Scholar 

130. Yarnall, M. T. N. et al. Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases. Nat. Biotechnol. 41, 500–512 (2023).

     Article  CAS  PubMed  Google Scholar 

131. Tao, X.-Y. et al. Harnessing transposable elements for plant functional genomics and genome engineering. Trends Plant Sci. 30, 1130–1146 (2025).

     Article  CAS  PubMed  Google Scholar 

132. Li, J. et al. Gene replacements and insertions in rice by intron targeting using CRISPR-Cas9. Nat. Plants 2, 16139 (2016).

     Article  CAS  PubMed  Google Scholar 

133. Gu, X., Lang, J., Chang, Y. & Zhang, M. Cleavable donor-assisted CRISPR/Cas9 system significantly improves the efficiency of large DNA insertion in Physcomitrium patens. Plant J. 121, e70020 (2025).

     Article  CAS  PubMed  Google Scholar 

134. Lu, Y., Ronald, P. C., Han, B., Li, J. & Zhu, J.-K. Rice protein tagging project: a call for international collaborations on genome-wide in-locus tagging of rice proteins. Mol. Plant 13, 1663–1665 (2020).

     Article  CAS  PubMed  Google Scholar 

135. Yao, Q. et al. Efficient and multiplex gene upregulation in plants through CRISPR-Cas-mediated knockin of enhancers. Mol. Plant 17, 1472–1483 (2024).

     Article  CAS  PubMed  Google Scholar 

136. Shen, R. et al. Targeted insertion of regulatory elements enables translational enhancement in rice. Front. Plant Sci. 14, 1134209 (2023).

     Article  PubMed  PubMed Central  Google Scholar 

137. Ke, Y. et al. Precise knock-in of stress-responsive cis-regulatory elements using gene targeting for improving abiotic stress tolerance in plants. N. Phytol. 247, 2147–2162 (2025).

     Article  CAS  Google Scholar 

138. Dong, O. X. et al. Marker-free carotenoid-enriched rice generated through targeted gene insertion using CRISPR-Cas9. Nat. Commun. 11, 1178 (2020).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

139. Park, J.-C. et al. AI-generated MLH1 small binder improves prime editing efficiency. Cell 188, 1–16 (2025).

     Article  Google Scholar 

140. Yin, Y., Arneson, R., Yuan, Y. & Fang, S. Long oligos: direct chemical synthesis of genes with up to 1728 nucleotides. Chem. Sci. 16, 1966–1973 (2025).

     Article  CAS  PubMed  Google Scholar 

141. Szarzanowicz, M. J. et al. Binary vector copy number engineering improves Agrobacterium-mediated transformation. Nat. Biotechnol. 43, 1708–1716 (2025).

     Article  CAS  PubMed  Google Scholar 

142. Xi, J., Patel, M., Dong, S., Que, Q. & Qu, R. Acetosyringone treatment duration affects large T-DNA molecule transfer to rice callus. BMC Biotechnol. 18, 48 (2018).

     Article  PubMed  PubMed Central  Google Scholar 

143. Altpeter, F. et al. Advancing crop transformation in the era of genome editing. Plant Cell 28, 1510–1520 (2016).

     CAS  PubMed  PubMed Central  Google Scholar 

144. Maher, M. F. et al. Plant gene editing through de novo induction of meristems. Nat. Biotechnol. 38, 84–89 (2020).

     Article  CAS  PubMed  Google Scholar 

145. Lin, C. S. et al. Application of protoplast technology to CRISPR/Cas9 mutagenesis: from single-cell mutation detection to mutant plant regeneration. Plant Biotechnol. J. 16, 1295–1310 (2018).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

146. Gelvin, S. B. Integration of Agrobacterium T-DNA into the plant genome. Annu. Rev. Genet. 51, 195–217 (2017).

     Article  CAS  PubMed  Google Scholar 

147. Zhang, Y., Malzahn, A. A., Sretenovic, S. & Qi, Y. The emerging and uncultivated potential of CRISPR technology in plant science. Nat. Plants 5, 778–794 (2019).

     Article  PubMed  Google Scholar 

148. Demirer, G. S. et al. High aspect ratio nanomaterials enable delivery of functional genetic material without DNA integration in mature plants. Nat. Nanotechnol. 14, 456–464 (2019).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

149. Kwak, S.-Y. et al. Chloroplast-selective gene delivery and expression in planta using chitosan-complexed single-walled carbon nanotube carriers. Nat. Nanotechnol. 14, 447–455 (2019).

     Article  ADS  CAS  PubMed  Google Scholar 

150. Ellison, E. E. et al. Multiplexed heritable gene editing using RNA viruses and mobile single guide RNAs. Nat. Plants 6, 620–624 (2020).

     Article  CAS  PubMed  Google Scholar 

151. Wang, Z. P. et al. Efficient and genotype independent maize transformation using pollen transfected by DNA-coated magnetic nanoparticles. J. Integr. Plant Biol. 64, 1145–1156 (2022).

     Article  CAS  PubMed  Google Scholar 

152. Zhao, X. et al. Pollen magnetofection for genetic modification with magnetic nanoparticles as gene carriers. Nat. Plants 3, 956–964 (2017).

     Article  CAS  PubMed  Google Scholar 

153. Vejlupkova, Z. et al. No evidence for transient transformation via pollen magnetofection in several monocot species. Nat. Plants 6, 1323–1324 (2020).

     Article  PubMed  Google Scholar 

154. Schmidt, S. M., Belisle, M. & Frommer, W. B. The evolving landscape around genome editing in agriculture. EMBO Rep. 21, e50680 (2020).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

155. Puchta, H. Regulation of gene-edited plants in Europe: from the valley of tears into the shining sun?. aBIOTECH 5, 231–238 (2024).

     Article  PubMed  Google Scholar 

156. Yang, F., Zheng, K. & Yao, Y. China’s regulatory change toward genome-edited crops. Trends Biotechnol. 42, 801–806 (2024).

     Article  CAS  PubMed  Google Scholar 

157. Durrant, M. G. et al. Bridge RNAs direct programmable recombination of target and donor DNA. Nature 630, 984–993 (2024).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

158. Perry, N. T. et al. Megabase-scale human genome rearrangement with programmable bridge recombinases. Science 0, eadz0276 (2025).

     Article  Google Scholar 

159. Faure, G. et al. TIGR-Tas: a family of modular RNA-guided DNA-targeting systems in prokaryotes and their viruses. Science 388, eadv9789 (2025).

     Article  CAS  PubMed  PubMed Central  Google Scholar 

160. Zong, Y. et al. An engineered prime editor with enhanced editing efficiency in plants. Nat. Biotechnol. 40, 1394–1402 (2022).

     Article  CAS  PubMed  Google Scholar 

161. Yan, J. et al. Improving prime editing with an endogenous small RNA-binding protein. Nature 628, 639–647 (2024).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

162. Svitashev, S. et al. Targeted mutagenesis, precise gene editing, and site-specific gene insertion in maize using cas9 and guide RNA. Plant Physiol. 169, 931–945 (2015).

     Article  PubMed  PubMed Central  Google Scholar 

163. Li, Z. et al. Cas9-guide RNA directed genome editing in soybean. Plant Physiol. 169, 960–970 (2015).

     Article  ADS  PubMed  PubMed Central  Google Scholar 

164. Zhao, Y. et al. An alternative strategy for targeted gene replacement in plants using a dual-sgRNA/Cas9 design. Sci. Rep. 6, 23890 (2016).

     Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

165. Begemann, M. B. et al. Precise insertion and guided editing of higher plant genomes using Cpf1 CRISPR nucleases. Sci. Rep. 7, 11606 (2017).

     Article  ADS  PubMed  PubMed Central  Google Scholar 

166. Xu, Y. et al. Intron-targeted gene insertion in rice using CRISPR/Cas9: a case study of the Pi-ta gene. Crop J. 8, 424–431 (2020).

     Article  Google Scholar 

167. Zhao, Y. et al. Precise deletion, replacement and inversion of large DNA fragments in plants using dual prime editing. Nat. Plants 11, 191–205 (2025).

     Article  CAS  PubMed  Google Scholar 

168. Lou, H. et al. Engineering source-sink relations by prime editing confers heat-stress resilience in tomato and rice. Cell 188, 530–549.e520 (2025).

     Article  CAS  PubMed  Google Scholar 

169. Vu, T. V. et al. Enhancing CRISPR-Cas-based gene targeting in tomato using a dominant-negative ku80. Hortic. Res. 12, uhae294 (2024).

     Article  PubMed  PubMed Central  Google Scholar 

Download references

Author notes

1. These authors contributed equally: Tien Van Vu, Ngan Thi Nguyen.

Authors and Affiliations

1. Division of Applied Life Science (BK21 Four Program), Plant Molecular Biology and Biotechnology Research Center, Gyeongsang National University, Jinju, Republic of Korea

   Tien Van Vu, Ngan Thi Nguyen, Jihae Kim, Yeon Woo Sung, Woo Sik Chung & Jae-Yean Kim

2. Division of Life Science, Gyeongsang National University, Jinju, Republic of Korea

   Jae-Yean Kim

3. Nulla Bio Inc., Jinju, Republic of Korea

   Jae-Yean Kim

Contributions

Writing–original draft, T.V.V., N.T.N, J.K., Y.W.S., W.S.C., and J.Y.K.; writing–review & editing, T.V.V. and J.Y.K.; funding acquisition, J.Y.K.; supervision, T.V.V. and J.Y.K. All authors read and approved the manuscript.

Corresponding authors

Correspondence to Tien Van Vu or Jae-Yean Kim.

Competing interests

J.Y.K. is the founder and CEO of Nulla Bio Inc. The remaining authors declare no competing interests.

Peer review information

Nature Communications thanks Jian-Kang Zhu and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions