1 Introduction

It is a great challenge to meet the growing food demands under the circumstance of increasing world population and rapid global climate change [1]. To address this challenge, plant genetic engineering is viewed with high hopes and demonstrates great potential [2, 3]. In general, plant genetic engineering techniques includes transgene, RNA interference (RNAi) and CRISPR/Cas9 genome editing. Transgene is involved heterologous expression of exogenous gene(s) in host plants [4]. RNAi, a post-transcriptional gene silencing (PTGS) process uses small interfering RNAs (siRNAs) to silence target gene(s) [5]. CRISPR/Cas9 employs a single guide RNA (sgRNA) guiding the Cas9 nuclease to cut specific plant genomic DNA fragment(s) [6]. However, effective and convenient approaches to deliver these genetic engineering components into the plant cells are lacking especially for those woody plants, e.g., species of Camellia and Eucommia [7].

The extensively used genetic transformation methods in plants (i.e. Arabidopsis, tobacco, tomato, rice and poplar) include Agrobacterium and virus infection, and particle bombardment (gene gun) [8]. Although widely used, there are still drawbacks to overcome: (1) Agrobacterium and virus vectors are host specific, infecting only sensitive plant species. (2) The viral vector usually achieves transient transfection, and thus unable to be inherited to the offsprings. (3) Agrobacterium and particle bombard methods require the plant regeneration process via tissue culture, which is heavily genotype dependency, time consuming and labor intensive [9, 10]. Therefore, developing efficient and convenient genetic transformation methods is not only a practical need but also will revolutionize the development of plant genetic engineering.

Nanoparticles are widely used in the medical field recently years for their excellent properties of being small-sized (c. 1–100 nm) and large surface areas [10]. Advancements in nanomedicine have paved the way for plant nanobiology, an emerging filed of plant sciences that especially investigates the application of nanotechnology in agriculture [11, 12]. A key focus is the utilization of nanomaterials as carriers for gene delivery in plants, which promises a novel and more efficient approach to plant genetic engineering [13]. During last two decades, many types of nanoparticles have been developed as gene vectors in plants, including carbon-, silicon- and metal-based nanoparticles [14]. In this work, the available plant nanodelivery systems of DNA and RNA are firstly summarized. And then the current challenges that nanoparticles-mediated plant gene delivery faces are discussed. Eventually, a potential and promising natural nanodelivery platform of CRISPR/Cas9 is proposed for plant genetic engineering.

2 Nanoparticle-mediated DNA delivery in plants

In plant genetic engineering, exogenous genes are usually cloned into expression plasmids, such as pCAMBIA-series vectors, for delivering into intact cells to archive transient or stable expression [15]. There are several nanoparticles including magnetic nanoparticles (MNPs), carbon-based nanoparticles, mesoporous silica (MSNs), and peptide-based nanoparticles have been developed as plasmid DNA (pDNA) delivery vehicles (Fig. 1).

Fig. 1
figure 1

Nanoparticle-mediated DNA delivery in plants to achieve transient expression of exogenous genes in nucleus and chloroplast

2.1 Magnetofection of DNA

Magnetic nanoparticle-based transformation (magnetofection) employs magnetic forces to deliver complexs of magnetic nanoparticles and DNAs (MNPs-DNAs) into cells, which has gained popularity in animal systems [16]. This technique has also been adapted to plants recently. Hao and colleagues firstly demonstrated the applicability of MNPs as a gene carrier for plant cells [17]. They synthesized magnetic gold nanoparticles (mGNPs) with diameters ranging from c. 20–30 nm and functionally modified mGNPs surface by adding free amino groups from polyethylene glycol (PEG). Fluorescein isothiocyanate (FITC) was conjugated to the mGNPs, which were then delivered into canola (Brassica napus) protoplasts using an external magnetic field. A high delivery efficiency of up to 95% was reported for FITC-labelled mGNPs. Additionally, plasmids encoding GUS reporter gene were covalently bound to the surface of mGNPs and also delivered into canola protoplasts and walled cell suspensions leading to transient GUS expression. Magnetofection could also be used for transformation of tobacco mesophyll protoplasts [18]. Specifically, the plasmid pCAMBIA1303 was delivered by polyethyleneimine (PEI) modified Fe3O4 MNPs resulting in transient GUS gene expression within the protoplasts.

Zhao et al. [19] further developed a simple, rapid and efficient pollen magnetofection technology to generate transgenic seeds, thereby bypassing the labor-intensive and time-consuming tissue culture process. In this method, Fe3O4 MNPs were positively charged and coated with PEI to bind negatively charged pDNAs through electrostatic attraction. The MNP/DNA complexes were then mixed with pollens and guided by an external magnetic field to direct them into the pollen via germination apertures. The transformed pollens were then used for artifical pollination and the resulting seeds were sown to screen for transgenic plants. The authors selected cotton (Gossypium hirsutumcotton) as a model species to test this technique, successfully obtained transgenic cotton plants. In addition, they further demonstrated the versatility of pollen magnetofection using the pBI121 GUS plasmid as a reporter in four other flowering plants: three eudicots- cocozelle (Cucurbita pepo), pepper (Capsicum annuum) and pumpkin (Cucurbita moschata), and one monocot- lily (Lilium brownii). Transgenic pepper and pumpkin plants were also obtained with the GUS gene integrated into the host genomes. This pioneering work combined magnetofection with pollen-mediated genetic transformation, offering a tissue culture-free and host-independent approach that may be applicable for genetic transformation in a wide range of plant species.

Recently, pollen magnetofection technology has successfully been applied to achive efficient genetic transformation in maize (Zea mays) [20]. Plasmid DNA encoding red fluorescent protein (RFP), enhanced green fluorescent protein (EGFP), GUS or bialaphos resistance (bar) was successfully delivered into pollen grains for six inbred lines of maize (178, B73, HZ178, Jing2416, Jing92 and Zheng58). Expression of EGFP or GUS was detected in 5–16% of T1 transfected seedlings, while glufosinate resistance was observed in 1.41% of T1 plants transformed with the bar gene. Notably, pretreating maize pollens at 8 ℃ for 10 min prior to magnetofection is a critical step that significantly enhances the transformation efficiency.

Moreover, the pollen magnetofection system has also been successfully employed for the genetic transformation of tree peonies (Paeonia ostii and P. rockii) using GUS gene as the reporter [21]. Additionally, the plasmid vector pYBA1132 containing EGFP gene was also delivered into the protoplasts of a green algae (Chlorella ellipsoidea) resulting in a transient expression of EGFP [22]. Notably, this technology has also been applied to deliver CRISPR/Cas9 plasmid for genome editing in pepper [23]. The presence of CRISPR/Cas9 pDNA was reported in both T0 and T2 generation plants, suggesting the potential of pollen magnetofection for precise genome editing.

2.2 Carbon-based nanocarriers of DNA

Carbon-based nanoparticles, including carbon nanotubes (CNTs), carbon dots (CDs), graphene, nanodiamond, etc., exhibit exceptional electrical, optical, mechanical and thermal properties [24]. CNTs are elongated, cylindrical nanoparticles characterized by a high aspect ratio and are typically categorized into two types: single-walled carbon nanotubes (SWCNTs) and multiwalled carbon nanotubes (MWCNTs) [25]. SWCNTs are small diameter sized (1–3 nm), whereas MWCNTs are larger (5–40 nm). CDs, another class of carbon nanomaterials, are measured in c. 5–10 nm and noted for their excellent biocompatibility [26]. Both CNTs and CDs by far have been developed as nanocarriers for delivering foreign DNAs in plant cells.

2.2.1 Carbon nanotubes

Liu et al. [27] firstly used oxidized SWCNTs (o-SWCNTs) to archive DNA delivery into intact Bright Yellow (BY-2) cells of Nicotiana tobacum, demonstrating the potential of SWCNTs as gene nanotransporters capable of penetrating the plant cell wall and cell membrane. Following this, Burlaka and colleagues [28] advanced the application of CNTs in plant genetic transformation. They used commercial SWCNTs and MWCNTs as nanocarriers to deliver the pGreen0029 plasmid into callus cells, mesophyll protoplasts and leaf explants of N. tabacum. Their results showed transient expression of the YFP reporter gene in the protoplasts using both SWCNTs (20 µg/mL) and MWCNTs (15 µg/mL). Notably stable expression of the nptII gene was only seen in the regenerated plants derived from N. tabacum callus and leaf explants when using SWCNTs, not MWCNTs. This study underscored the suitability of SWCNTs-based nanocarriers for genetic transformation in both plant protoplasts and walled cells. Subsequently, Golestanipour et al. [29] assessed the efficiency of the arginine functionalized SWCNTs (Arg-SWCNTs) for gene delivery in plants. They suggested that the cylindrical nanostructure and cationic groups of engineered Arg-SWCNTs facilitated the transfer of pDNAs into tobacco (N. tabacum) root cells. Successful expression of GFP was confirmed through fluorescence microscopy and western blotting analysis.

Remarkably, Demirer et al. [30] and Kwak et al. [31] successfully employed polymer-functionalized SWCNTs to deliver genes into intact plants without external assistance. Demirer and colleges [30] utilized carboxylic acid-functionalized SWCNTs as raw materials, covalently attaching PEI to these nanotubes to impart a net positive charge, resulting PEI-SWCNTs. The negatively charged GFP-encoding plasmids were subsequently attached to the PEI-SWCNTs via electrostatic attraction. This system enabled efficient DNA delivery into mature leaves of three eudicots species- N. benthamiana, G. hirsutum and Eruca sativa, and one monocot species- Triticum aestivum. Robust GFP expression was detected at both mRNA and protein levels, though no integration of the transgene into the host genome was detected. This method achieved high transformation efficiency reaching up to 85% in the arugula protoplasts. Furthermore, Kwak and colleagues [31] provided evidence of gene delivery using SWCNTs as nanocarrier in mature N.tabacum, E.sativa, Nasturtiumofficinale and Spinaciaoleracea plants, as well as in the isolated A.thaliana protoplasts. They used chitosan, an alternative cationic polymer to covalently modify SWCNTs (CS-SWCNTs), achieving chloroplast-selective gene delivery. It was proposed that the acidic pH value of plant cell cytosol (~ 5.5) facilitates the binding of pDNAs to CS-SWCNTs, whereas the alkaline pH value of chloroplast (~ 8) weakens the electrostatic interactions between pDNAs and CS-SWCNTs, leading to the release of pDNAs from CS-SWCNTs.

Following the above breakthroughs, Lew et al. [32] further synthesized imidazolium-functionalized SWCNTs (IM-SWCNTs) and efficiently delivered the GFP reporter gene into oil palm (Elaeis guineensis) pollens without external physical aid. IM-SWNTs displays enhanced biocompatibility with oil palm pollens compared to the previously used PEI-SWCNTs [30] and CS-SWCNTs [31]. The study highlighted the zeta potential of IM-SWNTs as a critical factor influencing pollen transformation efficiency. Moreover, Ghaghelestany et al. [33] explored the application of PEI-SWCNTs for gene transformation in German chamomile callus cells, revealing that combining ultrasound and PEI-SWCNTs markedly improved the transformation efficiency.

2.2.2 Carbon dots

Wang et al. [34] firstly demonstrated that CDs are a promising vehicle to deliver foreign DNAs into intact plants. They synthesized CDs using an electrochemical etching method and subsequently modified these CDs with PEI to create a positively charged nanocomposite (CDP). This CDP was used to successfully deliver two pDNA vectors i.e. pJIT163-mCherry and pJIT163-hGFP into the leaves of rice (Oryza sativa), wheat (T. aestivum) and mung bean (Phaseolus radiatus), as well as into rice roots. Furthermore, the CDP delivery system was effective in transforming rice callus cells achieving both hygromycin resistance and GUS expression. The study suggested that the CDP protects DNA from DNase degradation and facilitates its transport into plant cells, thereby enhancing the efficiency of genetic transformation.

Notably, a straightforward method for spray-on gene editing was developed for plants in the preprint paper of Doyle et al. [35]. The researchers prepared PEG functionalised CDs (PEG-CDs) and conjugated pDNAs to these nanoparticles via electrostatic attraction. Thanks to their innate fluorescence at 475 nm, the CDs are easily tractable suggesting their diffusion throughout the somatic cells, but not in the chloroplasts. The GFP expression was observed by directly spraying the pDNAs/PEG-CDs formulations onto leaves of four monocots: wheat, maize, barley and sorghum. Furthermore, the study successfully delivered a CRISPR/Cas9 plasmid expressing Cas9 protein and sgRNAs targeting two regions of the SPO11 gene into wheat leaves, achieving a deletion mutation in the SPO11 gene. This approach enabled CRISPR/Cas9-mediated gene editing in cereal crops without integrating exogenous DNA into the host genome.

2.3 Peptide-based nanocarriers of DNA

Peptides with approximately 30 amino acids in length have been employed as nanocarieers to transfer DNA and protein into animal cells [36]. To date, numerous cell-penetrating peptides (CPPs) have also been used for gene delivery in plant cells for their native cell wall/membrane penetrating capability. These polycationic peptides can encapsulate negatively charged DNAs, protecting them from enzymatic degradation and facilitating their delivery into plant cytoplasm and specific organelles, e.g. mitochondria and chloroplasts [37].

In 2004, Rosenbluh and colleagues [38] demonstrated for the fist time that histone peptides could penetrate the plasma membrane of petunia (Petunia hybrida) protoplasts. Following this seminal work, arginine-rich intracellular delivery (AID) peptides were subsequently used to efficiently deliver GFP protein into onion epidermal and root tip cells, tomato roots and tobacco BY-2 cells [39]. The CPP of TAT2 was also developed to archive high transient GUS gene expression in the wheat immature embryos [40]. Moreover, a peptide-based gene delivery system was designed by combing CPPs (such as BP100 or TAT2) with a polycation (like R9 or KH9) to deliver pDNA into intact leaf cells of A. thaliana and N. benthamiana [41]. Transient GFP expression was observed in the cytoplasm using this composite peptides transporter. These studies collectively demonstrated that peptides serve as efficient cell-penetrating agents for delivering exogenous DNA into intact plant cells.

Notably, Chuah et al. [42] were the first to achieve plant mitochondria transfection using peptides-based nanocarriers. They combined CPP with mitochondria-targeting peptide (MTP) to form CPPKH-MTPKH-pDNA complexes. This approach led to observable transient GFP expression in the mitochondria of A. thaliana leaves in a short time of 12 h of transfection. Furthermore, the same research group created stable mitochondrial transgenic lines of A. thaliana and N. tabacum in 2018 [43], where exogenous genes were integrated into the host mitochondrial genomes. More recently, Thagun et al. [44] developed a technique for transient plastid transformation using peptide-based gene carrier. They formulated peptide/pDNA complexes by integrating a CPP with a chloroplast-targeting peptide (CTP). These complexes enabled efficient delivery of pDNA into plastids through simple leaf infiltration, offering a valuable tool for plastid genetic engineering.

2.4 Silicon-based nanocarriers of DNA

Silicon is considered as one of the beneficial elements for plants and silica-based nanoparticles have been utilized to enhance plant tolerance to abiotic stresses [24]. In 2007, Torney and colleagues firstly demonstrated that MSNs could serve as effective nanocarriers for pDNA in plants [45]. In their study honeycomb gold-capped MSNs with 3 nm pores were designed, which successfully delivered a GFP-encoding plasmid into intact leaves of N. tabacum by gene gun. Subsequently, Martin-Ortigosa et al. [46] showed that gold-plated MSNs with a large pore diameter of 10 nm could co-deliver pDNA and protein to the epidermal tissues of Allium cepa. Notably, gold-plated MSNs were also used to deliver site-specific recombinase, such as the Cre protein, for gene editing in maize [47]. However, it is important to note that this gold-modified MSNs delivery system still require the use of an expensive gene gun to penetrate the cell wall, highlighting a limitation in terms of cost and accessibility.

Furthermore, Chang et al. [48] developed organically functionalized MSNs measuring 40–50 nm and successfully delivered pDNA-mCherry into intact roots of A. thaliana. This modified MSN-based gene delivery system can passively penetrate the cell wall without the aid of gene gun and has a high transformation efficiency (up to 46.5%). Recently, aminopropyl-triethoxysilane functionalized MSNs (APTES-MSNs) with a similar size range of ~ 40 nm were synthesized and used to deliver pDNA-GUS into tomato leaves and shoots via simple spray or injection [49]. Successful transfer of the resistance gene cry1Ab of Tuta absolutawas was also accomplished in tomato plants by APTES-MSNs. These studies demonstrate that MSNs can deliver pDNA into plant tissues without external assistance, providing a promising direction for modifying MSNs as efficient tools for gene delivery in plants.

Based on the comparison of aforementioned nanodelivery systems of DNAs, it is evident that significant refinement of existing methodologies is required to suit specific plant species. For instance, Vejlupkova et al. [50] observed no transient expression of pDNA-GFP in three monocot species (lily, maize, and sorghum) when using pollen magnetofection system developed by Zhao et al. [19]. Two key factors likely contribute to these discrepancies:

Pollen structure variability Pollen strucuture varies significantly among different plant species. Pollens of lily, maize and sorghum used in Vejlupkova et al. [50] are monosulcate, whereas pollents of cotton, pepper and pumpkin examined by Zhao et al. [19] are tricolpate. Monosulcate pollens may present greater challenges for transformation compared to tricolpate pollens [19, 50].

Pollen germination aperture size The size and state of the pollen germination aperture can also influence the transformation efficiency. Wang et al. [20] found that maize pollens collected from greenhouse-grown plants were often capped with an operculum, which prevented the entry of MNP-DNA complexes. In contrast, pollens collected from field-grown maizes, after low-temperature (8 °C, 10 min) pretreatment, turned to be open and could be successfully transformed.

3 Nanoparticle-mediated RNA delivery in plants

RNAi is a potent tool in plant genetic engineering for regulating the expression of endogenous genes [13, 51]. Commonly, in the RNAi pathway, Dicer-like proteins (DCL) cleave double-stranded (dsRNA) into siRNAs sized with 21–22 bp. These siRNAs are then incorporated into RNA-induced silencing complex (RISC) which further degrades the target mRNA or inhibits its translation [5, 14]. Traditionally plant RNAi involves coding siRNAs into DNA vectors, which are then delivered into plant cells using the tobacco rattle virus (TRV). However, this approach is limited to certain crops and can be challenging to apply broadly [52]. Recently, advancements in nanotechnology have introduced various nanocarriers for delivering siRNA directly into plant cells for gene silencing (Fig. 2), expanding the scope of RNAi applications.

Fig. 2
figure 2

Nanoparticle-mediated RNA delivery in plants to achieve gene knockdown via RNA interference (RNAi) pathway

3.1 Carbon-based nanocarriers of RNA

Demirer et al. [53] revealed that intracellular transport and protection of siRNAs from degradation are crucial for the effectiveness of SWCNTs-based siRNA delivery platform in plants. Their study demonstrated that SWNTs could effectively carry Cy3-RNA into N. benthamiana leaves via needleless syringe infiltration. This method successfully silenced both the GFP transgene and an endogenous gene ROQ1 using corresponding siRNA/SWCNTs complexes. Notably, leaves infiltrated with the siRNA/SWNTs complex showed a reduction of over 90% in the GFP mRNA transcript levels in 24 h and one single infiltration maintained this silencing effect for up to three days. This underscores the potential of SWCNTs as robust carriers for siRNAs delivery.

Moreover, Schwartz et al. [54] demonstrated successful delivery of siRNA into N. benthamiana and Solanum lycopersicum using CDs. In this study, GFP transgene was effectively repressed through a simple low-pressure spraying method, with no need of leaf infiltration. Additionally, two endogenous genes, CHLH and CHLI were also knocked-down. When comparing silencing efficiency, the researchers found that CDs produced from 5kD branched PEI showed the highest silencing efficiency (> 90%). This work suggests the potential of CDs as efficient nanocarriers for RNAi applied in plants, offering an non-invasive and highly effective method for gene knock-down.

Polymer-functionalized graphene oxide nanoparticles (GONs) were also developed to deliver siRNA into N. benthamiana [55]. Sheet-like GONs were fabricated using PEI and PEG, and loaded siRNA to form a near-spheroidal GONs-siRNA complex with sizes ranging from 15 to 50 nm. The GONs-siRNA formulation exhibited a remarkable silencing efficiency of 97.2% in intact leaves within 24 h of treatment. This high efficiency likely attributed to the enhanced ability of the GONs to penetrate cell walls in addition to efficient cellular uptake of the siRNAs. GONs appears to play critical roles in facilitating both penetration and absorption, largely improving the efficiency of gene silencing.

3.2 Silicon-based nanocarriers of RNA

More recently, an advanced multi-gene silencing nanodelivery platform has been developed using MSN-siRNA spraying [56]. This novel delivery method enables efficient siRNAs uptake in N. benthamiana leaves via topical application, eliminating the need for mechanical methods such as infiltration. The silencing efficiency achieved was remarkable, reaching up to 98% at the molecular level for GFP transgene. Additionally, this platform successfully silenced five endogenous genes i.e. PDS, ChlH, SOS, HHL1 and FtsH2 leading to observable phenotypic changes. Notably, MSNs were capable of concurrently delivering siRNAs targeting FtsH2, HHL1, SOS and GFP, resulting simultaneous multi-gene silencing within the leaves. Moreover, sustained gene silencing lasting for 14 days was detected following sequential deliveries of siRNAs targeting SOS, FtsH2, HHL1 and GFP genes.

The MSN-siRNA delivery platform offers a versatile approach to enhance seed shattering resistance in wild rice (O. alta) [57]. Foliar spraying of MSN/siRNA complexes resulted in up to 70% gene silencing for the endogenous gene PDS and 75% silencing for the Ruby transgene. Furthermore, four candidate genes (SH01-SH04) of O. alta were analyzed by using the MSN-siRNA gene silencing technique, resulting that SH01, SH03 and SH04 were shattering genes of O. alta. This study showed the pivotal role of MSN-based siRNA delivery in elucidating gene functions in non-model plants.

3.3 Gold-based nanocarriers of RNA

Gold nanoparticles (AuNPs) were synthesized and functionalized by 25kD branched PEI, resulting in PEI-AuNPs nanoparticles [58]. These nanoparticles could efficiently integrate siRNAs to form stable formulations capable of delivering siRNAs into Arabidopsis cells. An 80% expression reduction of NPR1 gene was achieved using the AuNPs-siRNANPR1 formulation. As anticipated, Columbia-0 Arabidopsis plants treated with AuNPs-siRNANPR1, exhibited yellow leaves, similar to the phenotype observed in npr1 mutants.

Furthermore, PEI-functionalized gold nanoclusters (PEI-AuNCs) were developed for siRNA delivery in plants [59]. PEI-AuNCs could effectively protect siRNAs from RNase degradation and and facilitate its passage through the plant cell wall. Efficiency of gene knockdown carried using PEI-AuNCs exceeded 76% for both GFP and ROQ1 genes in N. benthamiana. Notably, the size and shape of gold AuNPs significantly influence the delivery efficiency of siRNAs [60]. In general, smaller and rod-shaped AuNPs are internalized into plant cells more efficiently. The reductions in GFP expression were as follows: 8% for 20 nm gold nanospheres, 29% for 15 nm gold nanospheres, 42% for 5 nm gold nanospheres, 99% for 10 nm gold nanospheres and 39% for gold nanorods. These results suggested that 10 nm spherical AuNPs are the most efficient for siRNAs delivery. Interestingly, while the AuNPs assist in delivering siRNAs to the cell wall, it is only the siRNAs that enter the plant cell, leaving the AuNPs outside the cell membrane.

3.4 LDH-based nanocarriers of RNA

Layered double hydroxide (LDH) nanosheets belong to a family of bio-compatible and degradable clay materials that characterized by their layered structure [61]. The positively charged hydroxide layers make LDH nanosheets an ideal carrier for RNAs delivery. In 2017, Mitter et al. [62] demonstrated that LDH nanosheets measuring 48 nm in diameter and 5.3 nm in thickness successfully loaded dsRNA to form dsRNA/LDH complexes which they named as BioClay. The dsRNA within BioClay exhibited resistance to nuclease degradation and remained stable on the leaf surface for up to 20 days following a single spray application. Importantly, topical application of BioClay protected N. tabacum from CMV virus infection via RNAi pathway. Furthermore, Yong et al. [63] using S. lycopersicum pollen as a model system found that LDH nanosheets with a diameter of 50 nm could internalize dsRNA into pollen within 2–4 h of incubation. The functional dsRNAs delivered by LDH efficiently induced RNAi of the reporter gene GUS, leading to an 89% reduction in mRNA levels.

3.5 DNA-based nanocarriers of RNA

Based on base-pairing principles, DNA nanostructures were synthesized with precisely programmed sizes and shapes, such as a 2.4 nm tetrahedron and 16 nm hairpin-tile monomer [64]. These DNA nanostructures can deliver siRNA into plant leaves without external assistance. Using transgenic mGFP5 N. benthamiana plants as a model, leaves infiltrated with the DNA-nanostructures/siRNA complex showed over a 40% reduction in GFP fluorescence intensity compared with untreated leaves. Notably, the gene silencing efficiency was influenced by the size, shape, compactness, and stiffness of DNA nanostructure.

3.6 Peptide-based nanocarriers of RNA

In addition, a fusion peptide KH9-BP100 with the amino acid sequences of KHKHKHKHKHKHKHKHKHKKLFKKILKYL has been developed as an efficient siRNA carrier into plant cells [65]. This siRNA/peptide complex was positively charged and 100–300 nm in diameter. When infiltration of the complex into intact A. thaliana leaves, RNAi of transgene YFP and endogenous gene CHS was observed, respectively. Recently, the KH9-BP100 peptide was fused with CTP (OEP34) and successfully delivered siRNA into chloroplast [66]. Spray of siGFPS1/KH9-OEP34/BP100 complex onto a transplastomic N. tabacum leaded to an efficient gene knockdown of eGFP.

As summarized in Table 1, a great variety of nanoparticles can successfully deliver siRNA into plant cells compared to pDNA. This disparity may primarily stem from differences in cargo size. siRNA molecules, typically 21–22 bp in size, are considerably smaller than pDNAs that usually range from c. 5000 to 10,000 bp. The diminutive size of siRNAs enhances their ability to penetrate the troublesome barriers of cell wall and cell membrane [9, 36]. In addition, the distinct mechanisms by which siRNAs and pDNAs work in plant cells likely affect their respective nanodelivery success rates. siRNAs are destined for the cytoplasm where they induce degradation of target mRNAs via the RNAi pathway [5]. In contrast, pDNAs have to experience a more intricate journey before performing their functions: they first enter the nucleus for transcription and then proceed to the cytoplasm for translation [4]. Failure at transcription or translation stage can result in the silencing of the reporter gene.

Table 1 Nanocarriers used for DNA and RNA delivery in plant genetic engineering

MSNs mesoporous silica nanoparticles, MNPs magnetic nanoparticles, SWCNTs single-walled carbon nanotubes, CDs carbon dots, GONs graphene oxide nanoparticles, LDH layered double hydroxide, CPP cell-penetrating peptide, CTP chloroplast-targeting peptide, AuNPs gold nanoparticles, AuNCs gold nanoclusters, pDNA plasmid DNA, siRNA small interfering RNA.

4 Challenges of using nanoparticles for plant genetic engineering

As previously reported, the utilization of nanotechnology in various aspects of agricultural production has demonstrated promising benefits [3, 11]. Whereas, the development of nanodelivery system for nucleic acids in plant genetic engineering is still in its infancy and faces numerous challenges.

4.1 Phytotoxicity and loading capacity of nanoparticles

The studies mentioned above predominantly utilized artificially manufactured nanomaterials as vehicles for nucleic acids (Table 1). However, recent observations have highlighted the potential inhibitory effect of artificial nanoparticles e.g., CNTs and CDs on plant growth [25, 26]. For example, in studies on rice, MWCNTs were found to exert toxic effect on rice suspension culture, inducing ROS accumulation and oxidative stress in cells [68]. In maize, exposure to CDs at concentrations of 1000 and 2000 mg/L for four weeks resulted in significant reduction in fresh shoot weights by 38% and 72%, and root weights by 57% and 68% [69]. Similarly, CDs with concentrations exceeding 125 mg/L significantly reduced root elongation in A. thaliana, and comparative transcriptome analysis showed that 715 genes in roots were up-regulated following treatment of 1000 mg/L CDs [70].

It is worthnoting that there may be a point of contradiction in the phytotoxicity and loading capacity due to the concentration dependency [71]. To minimize phytotoxicity, concentration of nanoparticles should ideally be at a lower dose [25], however higher concentrations are necessary to load an adequate amount of DNAs or RNAs [56]. For instance, when using Fe3O4 MNPs as a DNA carrier system, it was found that a mass ratio of pDNA to Fe3O4 MNPs ranging from 0.1:1 to 1:1 (w/w) was optimal for pDNA loading [19]. Nevertheless, Fe3O4 MNPs at such concentrations showed a potential phytotoxic effects to cotton, likely leading to a low efficiency of gene transformation [72].

Moreover, it is important to understand the maximum length of DNA that can be effectively delivered by nanoparticles [73]. To date, PEI-SWCNTs have successfully delivered DNA fragments ranging from 4.2 to 5.4 kb into plant cells [30]. Additionally, CS-SWCNTs have facilitated the transfer of larger plasmids of up to 8.8 kb in size [31]. However, the upper-limit for the maximum plasmid size that can be delivered by nanoparticles remains undetermined. Loading pDNA as large as 15 kb onto nanoparticles is challenging and still debatable [9]. To our knowledge, there are scarce reports on nanocarrier-mediated CRISPR-Cas9 genome editing in plants. One main reason for this gap is that DNA plasmids encoding for the CRISPR-Cas9 system are approximately 15 kb, substantially larger than the pCAMBIA-series reporter plasmids (c.10 kb) that are commonly used in proof-of-principle nanodelivery studies. In the future, it is imperative to optimize existing nanodelivery systems to accommodate larger CRISPR-Cas9 pDNA in plant genetic engineering.

4.2 The uptake and translocation of nanoparticles

The cell wall poses a primary barrier to the delivery of nucleic acids into plant cells due to its limited pore sizes (5–20 nm) [74]. In canonical scenario, nanocarrier-DNA or nanocarrier-RNA complexes go through the cell wall and membrane to reach the cytoplasm where they release their cargoes. However, the mechanism by which nanoparticles enter plant cells remains poorly understood [75, 76]. Theoretical and experimental studies on mammals suggested that nanoparticles internalization occurs via lipid membrane exchange [77]. Recent studies on plants suggested that similar mechanism may also be involved in nanoparticle uptake by plant cells [31]. Besides, energy-dependent endocytic pathways, channel-mediate transport and pore formation may play roles in the cellular internalization in plants [7].

Moreover, a recent study by Zhang et al. [60] reported an unusual scenario in which AuNPs did not need to enter cell to deliver siRNA. In this case, AuNPs intercalated with individual plant cell walls, protecting the siRNA cargo from nuclease degradation. Although AuNPs per se was not internalized by the cells, they served as reservoirs for siRNA to provide more opportunity for siRNA to enter plant cells. The uptake mechanism may vary depending on the type of nanoparticles and plant species involved [78]. Therefore, considerable effort is needed to investigate the interactions between nanoparticles and plant cells. Such research will help to guide design strategies for nanoparticle-based gene delivery in plants.

To date, the translocation of nanomaterials between plant cells, tissues and organs remains poorly understood. Both phloem and xylem are likely involved in nanoparticle distribution throughout the plant [76, 79], which is particularly concerning for the development of RNAi-based pesticides. For pests that feed on phloem sap, such as aphids, it is optimal to use nanoparticles that transport through the phloem as dsRNA nanocarrier. Conversely, nanoparticles that travel through xylem are suitable for insects that feed on xylem-sap [80]. In addition, when nanocarrier-DNA or nanocarrier-RNA solutions are sprayed onto leaves they may be transported to flowers, seeds and fruits, potentially leading to unexpected accumulation of nanoparticles in the food chain [81]. This poses potential risk to biodiversity, an issue that has been largely overlooked. For example, high concentrations of Fe3O4 MNPs (10.0 mg/kg) could accumulate in mycorrhizal fungi and destroy the mutualistic symbiosis between fungi and plants [82]. The accumulation results in a significant reduction in mycorrhizal infection rate, lower levels of glomalin-related protein in the soil and decreased alkaline phosphatase activity. Therefore, it is crucial to address the issue of nanoparticles accumulation in non-target organisms during cargo delivery to mitigate these risks.

5 Plant-derived exosome-like nanoparticles as natural nanodelivery platform for CRISPR/Cas9 toolkit

Plant-derived exosome-like nanoparticles (PELNs) are plant-secreted small extracellular vesicles, characterized by a bilayer phospholipid membrane and typically measuring between 30 and 300 nm in diameter [83]. These nanovesicles are similar to animal exosomes and facilitate inter-cellular and inter-species communications through the transport of proteins, lipids and nucleic acids. Owning to their biocompatibility, low immunogenicity, and natural origin, PELNs present promising prospects as drug delivery systems [84]. Previous studies have showed the capability of PELNs to convey therapeutic nucleic acids into mammalian cells, offering potential treatment of various diseases [85]. Likewise, PELNs are probably used as natural nanocarriers for nucleic acids delivery in plant cells. Despite of this, exploration of PELNs as gene delivery system within plants remains to be investigated.

The pioneering work by Kim et al. [86] demonstrated that cancer-derived exosomes mediated CRISPR/Cas9 delivery in human cells. Through electroporation (1000 V, 10 ms, 2 pulses), they successfully encapsulated Cas9-/sgRNA-expressing plasmids within exosomes isolated from HEK293 and SKOV3 cells. This approach led to the suppression of the target gene PARP-1, resulting apoptosis in ovarian cancer. Since then, exosome-liposome hybrid nanoparticles were developed to deliver CRISPR-Cas9 encoded plasmid vectors into the mesenchymal stem cells, achieving knocked-out of the target gene hCTNNB1 [87]. Additionally, exosomes isolated from HEK293T cells were used to carry CRISPR/Cas9 pDNA using commercial transfection reagents Exo-Fect [88]. These exosomes encapsulating genome editing machinery were cable of being absorbed by pancreatic cancer cells, thereby enabling editing of the oncogenic allele KrasG12D. Recently, serum-derived exosomes have been shown to effectively deliver CRISPR-Cas9 ribonucleoprotein (RNP) for modifying the dystrophin gene hDMD [89]. By incorporating CRISPR-Cas9 RNP into exosomes via electroporation, researcher formed EV@RNP complexes to realize cardiac-specific genome editing [90]. Furthermore, exosomes loaded with CRISPR-Cas9 RNP have been employed to inhibit herpes simplex virus1 infection too [91], highlighting the versatility and potential of exosome-mediated delivery system in various therapeutic applications.

In recent studies on plant-pathogen communication, cross-kingdom RNA trafficking by exosome-like extracellular vesicles was observed as well [92, 93]. Notably, the fungus Botrytis cinerea that causes grey mould in hundreds of crops can transport siRNAs into its host plant A. thaliana [94]. These transferred siRNAs further utilized host RNAi machinery to silence immunity-relevant genes of A. thaliana. Fascinatingly, this siRNAs exchange is bidirectional, that is A. thaliana also send siRNAs to B. cinerea to silence its virulence-related genes [95]. In addition, A. thaliana also used exosome-like extracellular vesicles to deliver mRNAs into B. cinerea, leading to the production of proteins within the fungus that reduce its pathogenicity [96]. Moreover, PELNs isolated from citrus juice showed the capability to package exogenous dsRNAs [97]. For instance, citrus PELNs loaded with dsCrcB and dsFUM21 inhibited the expression of Penicillium italicum CrcB gene and Fusarium proliferatum FUM21 gene, respectively. In a study conducted by our group, siRNA-loaded turmeric PELNs were shown to induce the silencing of GFP transgene in N. benthamiana (data unpublished).

Collectively, the above findings suggest that PELNs hold significant promise as a next-generation nanocarrier of CRISPR/Cas9 toolkit to advance genomic editing in plants (Fig. 3). It is probable to generate heritable genetic variation via utilizing PELNs to introduce CRISPR/Cas9 expression pDNA into plant seeds, buds or germ cells such as pollens and ovules. Such applications could accelerate the development of new crop varieties with desirable agronomic traits [98, 99]. Furthermore, PELNs also can serve as carriers for siRNAs, enabling the modulation of genetic circuits that control flowering time, vernalization requirement and drought tolerance [100]. This is beneficial for transient reprogramming of agronomic performance in existing crop varieties, providing a valuable tool to tackle the inevitable extreme weather conditions under global climate change.

Fig. 3
figure 3

Plant-derived exosome-like nanoparticles (PELNs) mediated delivery of CRISPR/Cas9 plasmid DNA to achieve gene knock-out in plants. During this process, double-strand break of genomic DNAs generated by Cas9 protein is commonly repaired by non-homologous end joining (NHEJ), leading to loss-of-function mutation. If donor DNAs provide the repair template, then the genome can undergo homologous recombination repair (HDR) to realize DNA replacement

6 Conclusion and outlook

As discussion above, the intersection of nanoscience and plant science has promoted the development of plant nanobiology in recent years. The use of nanoparticles for the delivery of DNAs and RNAs in plants is rapidly advancing plant functional genomics and genomic engineering by overcoming limitations associated with traditional methods, such as Agrobacterium-mediated transformation, viral vectors and particle bombardment. Certain nanocarriers can even selectively deliver foreign DNAs and RNAs to chloroplast via simple foliar spray, this not only simplifies the process but also enhances the efficiency and specificity of plastid genetic engineering.

Despite rapid advances in nanodelivery of nucleic acids to plants, several challenges remain to be addressed. Most notably, there is a gap in nanoparticle-mediated genome editing in plants, because the majority of nanodelivery studies have focused on small-sized pDNA and siRNA molecules (20–10,000 bp), with limited explorations into more complex genomic editing tools. During the long-term co-evolution, plants have evolved a mechanism to use PELNs for the trafficking of mRNA and siRNA between organisms. Similarly, fungal cells can secret siRNA-loaded exosome-like nanoparticles into the extracellular space, which further entered plant cells to regulate target gene expression. Moreover, exosomes have been successfully employed to mediate the delivery of CRISPR/Cas9 pDNA or RNP complex for genome editing in human cells. Given these advancements, we propose to develop PELNs as biocompatible and efficient agent to introduce genomic editing tools into plant cells, thereby facilitating precise genetic modifications. Therefore, future research should focus on optimizing PELNs as vehicles for CRISPR/Cas9 components, potentially revolutionizing plant genetic engineering.