Modified carbon dot‐mediated transient transformation for genomic and epigenomic studies in wheat

Summary

Genotype restriction poses a significant bottleneck to stable transformation in the vast majority of plant species, thereby severely impeding advancement in plant bioengineering, particularly for crops. Nanoparticles (NPs) can serve as effective carriers for the transient delivery of nucleic acids, facilitating gene overexpression or silencing in plants in a genotype‐independent manner. However, the applications of NP‐mediated transient systems in comprehensive genomic studies remained underexplored in plants, especially in crops that face challenges in genetic transformation. Consequently, there is an urgent need for efficient NP‐mediated delivery systems capable of generating whole plants or seedlings with uniformly transformed nucleic acids. We have developed a straightforward and efficient modified carbon dot (MCD)‐mediated transient transformation system for delivering DNA plasmids into the seeds of wheat, which is also applicable to other plant species. This system facilitates the generation of whole seedlings that contain the transferred DNA plasmids. Furthermore, our study demonstrates that this system serves as an excellent platform for conducting functional genomic studies in wheat, including the validation of gene functions, protein interactions and regulation, omics studies, and genome editing. This advancement significantly enhances functional genomic research for any plants or crops that face challenges in stable transformation. Thus, our study provides for the first time evidence of new applications for MCDs in functional genomics and epigenomic studies, and bioengineering potentially leading to the improvement of desirable agronomic traits in crops.

Introduction

Nanoparticles (NPs) have been extensively utilized across various fields of science and technology, including physics, chemistry, materials science, life science, human disease diagnosis and therapy, and medicine (Bayda et al., 2019; Pallares et al., 2022; Wang et al., 2020). Additionally, their applications in plants, particularly in modern agricultural science and technology, have garnered significantly more attention. Accumulating evidence shows that NPs can act as potential regulators or mediators in various biological processes, especially in the regulation of essential agronomic traits. These functions include: the regulation of plant growth and development (Aqeel et al., 2022; Giraldo et al., 2014; Hossain et al., 2020; Khodakovskaya et al., 2012; Rahman et al., 2022; Thapa et al., 2019; Wang et al., 2023a; Wu et al., 2017) as well as crop biofortification (Khan et al., 2021); the uptake of macro and micro essential elements (Rizwan et al., 2021; Tiwari et al., 2014); the alleviation of biotic and abiotic stresses (El‐Shetehy et al., 2021; Khan et al., 2023; Li et al., 2020b; Zia‐Ur‐Rehman et al., 2023); and applications in agriculture and food technology, which encompass nano‐fertilizers and biostimulant for promoting plant growth (DeRosa et al., 2010; Garza‐Alonso et al., 2023; Prakash et al., 2022); facilitators for improving quality and preservation time of postharvest fruits and vegetables (Wang et al., 2023b); elicitors for improving secondary metabolites (Lala, 2021; Zhang et al., 2022); and nano‐pesticides and herbicides for the protection of plants and the improvement of crop productivity (Guleria et al., 2023; Wang et al., 2016). Thus, NPs exhibit broad application prospects in sustainable agriculture and agroindustry.

Besides, NPs can serve as carriers for the transient intracellular delivery of nucleic acids or proteins in plants, independent of species or cell/tissue types (Demirer et al., 2019, 2020; Fang and Trewyn, 2012; Komarova et al., 2023; Kwak et al., 2019; Santana et al., 2022; Zhang et al., 2021b; Zhao et al., 2017). Importantly, NPs can facilitate CRISPR/Cas or other genome editing systems in plants without the integration of transgenes. This capability has the potential to produce DNA‐free gene‐modified plants, thereby circumventing the restrictions imposed by GMO (genetically modified organism) regulations and accelerating the application of crop bioengineering (Ghogare et al., 2021; Kocsisova and Coneva, 2023; Laforest and Nadakuduti, 2022). Unlike most metal‐based NPs that raise environmental concerns (Yonathan et al., 2022), carbon dots (CDs) can ultimately be degraded into CO₂ and H₂O. This characteristic makes CDs an environment‐friendly nanoparticle for applications in plants, particularly within agricultural ecosystems (Li et al., 2018, 2019; Schwartz et al., 2020; Wang et al., 2020). Also, CDs have been reported to play a role in regulating plant growth and development, as well as enhancing resistance to biotic and abiotic stress (Arel et al., 2023; Kou et al., 2021; Li et al., 2018, 2019, 2020a, 2022; Waghmare et al., 2020; Wang et al., 2022; Zhu et al., 2023). They are involved in seed germination (Liang et al., 2023), biosensor (Kanwal et al., 2022; Lin et al., 2023; Zulfajri et al., 2020), the regulation of gene expression (Chandrakar et al., 2020; Yan et al., 2021; Zhang et al., 2023) and enzyme activities (Kuang et al., 2023; Li et al., 2021). Additionally, CDs facilitate the delivery of nucleic acids for gene overexpression or silencing (Delgado‐Martin et al., 2022; Demirer et al., 2020; Mitter et al., 2017; Schwartz et al., 2020; Wang et al., 2020; Zhang et al., 2019). This significantly expands their applications in molecular biology, functional genomics, and bioengineering towards the improvement of desirable agronomic traits such as resistance to abiotic or biotic stresses, as well as improvements in crop yield and quality.

Transient or constitutive genetic transformation is a crucial step for the in vitro or in vivo validation of gene functions in plants, particularly in crops. However, constitutive genetic transformation is time‐consuming and labour‐intensive, and it largely depends on the host's genetic background for most crops, including wheat, maize, potato, soybean, and even rice. Currently, the validation of new functional genes from these species relies either on other transformable species or varieties or on transient transformation methods using chloroplasts or tobacco leaves. These methods are primarily mediated by techniques such as VIGS (virus‐induced gene silencing), BIGS (bacterium‐induced gene silencing), gene gun applications and agrobacterium‐mediated transient transformation (Chen et al., 2023; Kolliopoulou et al., 2020; Leissing et al., 2022; Nino‐Sanchez et al., 2021; Panwar and Kanyuka, 2022; Wu and Lai, 2022). Consequently, this reliance restricts the advancement of functional genomic studies. The emergence of species‐independent NP‐mediated transient transformation facilitates the validation of gene functions within the same plant species. This approach enhances functional genomic studies in plants and mitigates potential concerns regarding the reliability of results obtained from different species, particularly for those species that face significant challenges in constitutive genetic transformation. However, most NP‐mediated delivery of nucleic acids has been conducted in specific tissues or plastids, such as roots, calli, leaves, chloroplasts, or particular regions of leaf tissue using local injection or smearing techniques. These methods lead to variable transformation efficiency and a relatively low number of transformed cells for downstream studies (Demirer et al., 2020; Mitter et al., 2017; Schwartz et al., 2020; Wang et al., 2020; Zhang et al., 2019). Therefore, efficient NP‐mediated delivery systems that can uniformly transform whole plants or seedlings with nucleic acids are urgently needed and these systems warrant increased investment.

In this study, we developed a highly efficient and robust MCD‐mediated transient transformation system for delivering DNA plasmids into the seeds of wheat, which can also be applied to other plant species. This approach leads to the entire germinated seedlings containing the transferred alien DNA plasmids. Notably, our study demonstrated that this innovative system serves as an excellent platform for conducting various functional genomic and epigenomic studies in wheat. These studies include the validation of gene functions related to biotic and abiotic stresses, protein interactions and regulation, omics studies, and genome editing. Consequently, this system significantly enhances functional genomic studies for any plants or crops that face challenges with constitutive DNA transformation, such as wheat, horticultural crops, and their wild relatives.

Results

Modified carbon dot (MCD)‐mediated transient transformation for wheat and other plant species

Our previous study showed that CD priming with rice seeds influences the growth of the resulting seedlings and roots (Zhang et al., 2023). However, it remains largely unknown whether CDs can directly deliver nucleic acids into seeds and subsequently generate transiently or constitutively transformed plants. To investigate this, we mixed newly synthesized modified carbon dots (MCDs) (Figure S1) with DNA plasmids containing the hygromycin resistance gene, along with reporter genes for LUC (luciferase), GFP (green fluorescent protein) or GUS (β‐glucuronidase) to treat seeds of various wheat varieties, as indicated in Figure 1a. The main procedures for the MCD‐mediated transformation system are illustrated in Figure S2. We compared the phenotype of seedlings at the two‐leaf stage and found that no significant phenotypic change occurred in seedlings germinated from seeds treated with MCDs alone or MCDs plus DNA plasmids as compared with CK (the control). This suggests that MCD treatment does not significantly affect seed germination or the subsequent growth of seedlings (Figure S3).

Figure 1.

MCD‐mediated transient transformation using seeds from various wheat varieties. (a) Seeds from the following wheat varieties and wild relatives were transiently transformed with MCDs mixed with or without (CK, control) DNA plasmids containing hygromycin resistance gene plus LUC (luciferase). They included Y15 (Yangmai 15), Y29 (Yangmai 29), Z12 (Zhenmai 12), Z366 (Zhengmai 366), AK58 (Aikang 58), BN607 (Bainong 607), N13 (Ningmai 13), Z168 (Zhenmai 168), SJZ407 (Shijiazhuang 407), S3 (Sumai 3), N9 (Ningmai 9), HS2 (Hesheng 2), Y158 (Yangmai 158), and CS (Chinese Spring) wheat varieties and wild relatives (AA, BB, AABB and DD). (b) Detection of the LUC signal from the 6th to the 10th day (D) of seedlings germinated from transiently transformed seeds, as shown in (a). (c) Seeds from the CS, C22, Y158, and Fielder wheat varieties were transiently transformed with MCDs, with or without (CK) DNA plasmids containing the hygromycin resistance gene and the SAMMT gene (a newly identified gene with functions annotated as SAM methyltransferase), NLR1 (a newly identified resistance gene belonging to the NLR family), Pm21 (a powdery mildew resistance gene) fused with the GFP (green fluorescent protein) reporter gene. Scale bar = 20 μm. (d) The relative expression levels of the SAMMT, NLR1 genes in seedlings of the CS and C22 wheat varieties, with or without (CK) DNA plasmids containing the SAMMT/NLR1gene fused with the GFP reporter gene, were analysed. The Relative expression levels of the SAMMT/NLR1 gene were calculated relative to the beta‐actin gene (an internal control) in the leaves of CS and C22 seedlings with or without (CK) DNA plasmids containing gene SAMMT/NLR1 fused with GFP reporter gene. The expression levels were then normalized to the corresponding CK, which was set as 1. The significance test was determined by using one‐way ANOVA followed by Tukey's test. *P < 0.05; **P < 0.01. (e) Western blotting analysis for the detection of GFP protein in seedlings of CS, C22, Y158, and Fielder wheat varieties as shown in (c). (f) The intensity of the GFP fluorescent signal from seedlings of the CS, C22, Y158, and Fielder wheat varieties is shown in (c). AU, Arbitrary Units.

For wheat plants transformed with a DNA plasmid containing the LUC gene driven by the 35S promoter, we observed that the whole transformed plants exhibited a clear LUC signal, while the LUC signal was undetectable in the CK group treated only with MCDs (control) (Figure 1a). We recorded the LUC signal every 2 days, starting from the 4th day of growth, and found that the LUC signal remained strong until the 8th day, after which it dramatically decreased by the 10th day (Figure 1b). For wheat plants transformed with DNA plasmids containing the GFP gene, we observed a clear GFP signal in the vast majority of cells under confocal microscopy. In contrast, the GFP signal was completely absent in the CK group treated only with MCDs (Figure 1c). As expected, we observed elevated relative expression levels of the transferred SAMMT gene (a novel gene we identified, annotated as a SAM methyltransferase) and the NLR1 gene (a newly identified resistance gene belonging to the NLR family) fused with the GFP gene (Figure 1d). Consequently, a distinct band corresponding to GFP was detected in the Western blotting analysis of samples transformed with DNA plasmids, in contrast to the CK samples (Figure 1e). This finding is consistent with the mean value of GFP signal intensity (Figure 1f). These results confirm that MCDs can efficiently deliver DNA plasmids into the cells. Additionally, MCDs can facilitate the transfer of DNA plasmids containing the GUS (Figure S4) or Basta (phosphinothricin acetyltransferase) (Figure S5) resistance genes into wheat seeds. GUS signal can also be detected in wheat seeds (Figure S4a) and in various tissues, such as roots and sprouts (Figure S4b). Similarly, the MCD‐mediated transfer of a DNA plasmid containing the LUC or GFP reporter gene is effective in maize seeds (Figure S6), and in eight other plant species that present challenges for constitutive gene transformation, including soybean, potato, cucumber, Agropyron mongolicum Keng, and Medicago sativa L. (Figure S7).

To interrogate if autonomous replication occurs for DNA plasmids within cells, we conducted a PCR assay using plasmid DNA extracted from the root, stem, and leaf tissues of seedlings, both with and without (the control, CK) transformed DNA plasmids. A distinct DNA band corresponding to the transferred plasmid DNA was visible in all three tissues on the 3rd day, and in the stem and leaf tissues on the 5th day for transformed seedlings growing in biotin‐dA/CTP‐containing water. In contrast, no DNA band was detected in either the CK or the transformed seedlings growing in d A/G/C/GTP‐containing water (Figure S8). We also conducted a PCR assay using plasmid DNA extracted from transformed seedlings at the 4th, 6th, 8th, and 10th day. We observed a gradual decrease in the intensity of the plasmid DNA band, with a particularly dramatic reduction occurring on the 10th day. This result indicates a time‐dependent degradation of plasmid DNA within the cells (Figure S9). These results can partly explain the transmission of DNA plasmids between cells and their time‐dependent degradation during cell division. To further evaluate the contributions of endophytes or bacteria within grains to the expression levels of a specific gene through MCD‐mediated plasmid transfer, we conducted a comparison of the relative expression levels of the gene (TraesCS3B02G218700), which contains an intron (Figure S10a). This analysis utilized RNA extracted from wheat seedlings subjected to MCD‐mediated transfer of a plasmid containing its complete CDS or genomic DNA sequences (Figure S10b). Our results indicated no significant difference in the relative expression levels between the CDS and the genomic DNA expression units (Figure S10c). Moreover, we conducted RT‐PCR with cDNA from genomic DNA sequences as a template by using primer F1R1 and F2R1 as shown in Figure S10d. We found that the intensity of DNA band amplified by using primer F1R1 (Figure S10e, left) or F2R1 (Figure S10e, right), corresponding to unspliced cDNA (marked by the red arrow), was much weaker than the one corresponding to spliced cDNA (marked by the asterisk) (Figure S10e). This finding suggests that the presence of endophytes or bacteria within grains does not significantly affect the expression levels of the transferred genes.

These results pinpoint the high efficiency of MCD‐mediated transformation of DNA plasmids within seeds. Consequently, MCD‐mediated seed transformation could serve as a powerful system for effectively conducting functional genomic studies in wheat and other plant species that face challenges of constitutive gene transformation.

Validation of proteins interacting with proteins or cis‐regulatory elements

To validate previously reported protein interactions between TaHAG1 (histone acetyltransferase of the GNAT family 1) and TaBZR1 (BRASSINAZOLE‐RESISTANT 1) from wheat (Chu et al., 2024), as well as between Rf2b (a transcription factor belonging to basic leucine zipper protein) and PRE6 (paclobutrazol resistance) from cotton (Zhang et al., 2021a) in N. benthamiana leaves, we conducted MCD‐mediated transient transformation using seeds from Chinese Spring (CS). At the one‐leaf stage, a clear YFP signal was observed in leaves germinated from seeds transformed with a mixture of TaHAG1‐cYFP and TaBZR1‐nYFP (Figure 2a) or Rf2b‐cYFP and PRE6‐nYFP (Figure 2b) instead of TaHAG1/Rf2b‐cYFP or TaBZR1/PRE6‐nYFP alone. This finding confirms the interactions between two TFs. Additionally, interactions between TaHAG1 and TaBZR1 were further validated by using the Split‐LUC assay (Figure 2c). A YFP signal in Figure 2b was further validated by using a Western blotting assay, where a clear GFP band was observed in samples containing a mixture of Rf2b‐cYFP and PRE6‐nYFP, as opposed to samples containing only Rf2b‐cYFP or PRE6‐nYFP (Figure 2d). Additionally, we detected an increased LUC signal (highlighted with the red rectangular box) from leaves transformed with Rf2b and the promoter sequences of PRE6 as compared with the background signal from leaves treated solely with MCDs or transformed with either PRE6 or Rf2b alone. This finding confirms the regulatory relationship between the R2fb and PRE6 genes (Figure 2e). Vector construction is illustrated in Figure 2f. Furthermore, this transformation system can be applied to validate enhancer functions, which can drive higher expression of LUC as compared with the mini 35 promoter‐driven LUC expression (negative control) (Figure 2g). Additionally, it can be employed to validate PM (powdery mildew, E26) inducible enhancers, which become active once leaves are infected with PM (E26), as indicated by the elevated GFP signal (Figure 2h) and the relative expression levels of the GFP reporter gene before and post‐PM treatment (Figure 2i). Consequently, this transformation system demonstrates efficacy in examining interactions between proteins with proteins or cis‐regulatory elements (CREs).

Figure 2.

Proteins interacting with other proteins or cis‐regulatory elements using the MCD‐mediated seed transient transformation system. (a) BiFC (bimolecular fluorescence complementation) validation of previously reported protein interactions between TaBZR1 (BRASSINAZOLE‐RESISTANT 1) and TaHAG1 (histone acetyltransferase of the GNAT family 1) in wheat seedlings. Scale bar = 20 μm. (b) BiFC validation of previously reported protein interactions between Rf2b (a transcription factor belonging to basic leucine zipper protein) and PRE6 (paclobutrazol resistance), which were identified in cotton, in wheat seedlings. (c) Split‐LUC validation of previously reported protein interactions between TaBZR1 and TaHAG1 in wheat seedlings. (d) Western blotting analysis for the detection of YFP (yellow fluorescent protein) in seedlings, as illustrated in (b). (e) Transcriptional activity assay examining the interaction between Rf2b and the promoter sequences of PRE6 in wheat seedlings. (f) The construction of vectors is detailed in (e). (g) The validation of the enhancer was conducted using a DNA plasmid containing the mini 35 promoter, the candidate enhancer sequences, and the LUC gene in wheat seedlings. (h) The validation of the PM‐inducible enhancer was performed using a DNA plasmid that includes the mini 35 promoter, the candidate enhancer (EN) sequences, and the GFP gene in wheat seedlings treated with or without PM (E26) for 8 h. Scale bar = 20 μm. (i) The relative expression levels of the GFP gene were calculated in seedlings that were transiently transformed with the candidate enhancer (EN) sequences, as shown in (h). The relative expression levels of the GFP gene were calculated relative to the beta‐actin gene (an internal control) in the leaves of Chinese Spring (CS) wheat, both with (CS‐PM) or without (CS‐CK) PM inoculation (E26). The expression levels of GFP in CS‐PM were then expressed relative to GFP expression level in CS‐CK, which was set as 1. Significance test was determined by using one‐way ANOVA followed by Tukey's test. *P < 0.05.

Functional verification of candidate genes for biotic or abiotic resistance

We further employed the transformation system to validate the functions of candidate genes in response to biotic or abiotic stresses in the transformed T0 plants. As expected, we observed that seedlings with overexpression of the previously reported Pm21 (powdery mildew) gene, which confers resistance to powdery mildew (Cao et al., 2011) exhibited significantly higher resistance to PM (E26) infection as compared with CK or seedlings treated solely with MCDs (Figure S11). Similarly, we found that the overexpression of an uncharacterized NLR1 gene, identified from CS, also conferred resistance to PM (E26) in the seedlings of Fielder and Yangmai 158 (Y158) when compared with CK or seedlings treated solely with MCDs (Figure 3a). MCD‐treated seeds with or without (control, CK) overexpression of NLR1 genes exhibited similar germination rates (Figure S12). Consistent with the resistant phenotype, we observed that the overexpression of NLR1 can dramatically inhibit the production of PM (E26) spores, as indicated by the Coomassie Brilliant Blue R250 staining of leaves (Figure 3b). Moreover, we observed that the overexpression of another uncharacterized bHLH (basic helix–loop–helix) gene identified from wheat line P527, which exhibits resistance to Fusarium pseudograminearum, can enhance resistance to Fusarium pseudograminearum as compared with the CK, as indicated by the stem colour following inoculation with Fusarium pseudograminearum (Figure 3c, Figure S13). In addition, we identified two uncharacterized HSF (heat stress transcription factor) genes (HSF1 and 2) from TAM107 wheat subjected to heat treatment. After transiently transforming both plasmids into CS seeds, we observed that CS seedlings with overexpression of HSF1 (HSF1‐OE) continued to grow robustly with upright leaves. In contrast, seedlings in the CK group, treated only with MCDs or overexpression of HSF2 (HSF2‐OE) exhibited dropping and wilting leaves under heat stress (Figure 3d,e). These results indicate that HSF1‐OE significantly enhances heat resistance (Figure 3d), while HSF2‐OE markedly reduce heat resistance in plants (Figure 3e).

Figure 3.

Functional verification of candidate genes for biotic or abiotic resistance using the MCD‐mediated seed transient transformation system. (a) Validation of an uncharacterized NLR1 gene demonstrating resistance to PM (Powdery Mildew). Detached leaves from Y158 (Left) and Fielder (Right) seedlings, with or without overexpression of the NLR1 gene, were treated with or without PM (E26) inoculation. (b) Coomassie Brilliant Blue R250 staining of Fielder leaves, with or without the overexpression of the NLR1 gene, treated with or without PM (E26) inoculation. Scale bar = 10 μm. (c) Validation of an uncharacterized bHLH gene exhibiting resistance to Fusarium pseudograminearum. Whole seedlings of Sumai 3, with or without overexpression of the bHLH gene, were treated with or without Fusarium pseudograminearum inoculation. Scale bar = 1 cm. (d) Validation of an uncharacterized HSF1 (heat stress transcription factor) gene showing resistance to heat. Whole seedlings of CS (Chinese Spring), with or without overexpression of the HSF1 gene, were treated at 25 °C (CK) and 40 °C (heat). Scale bar = 1 cm. (e) Validation of an uncharacterized HSF2 gene demonstrating sensitivity to heat. Whole seedlings of CS, with or without overexpression of the HSF2 gene, were treated at 25 °C (CK) and 40 °C (heat). Scale bar = 1 cm.

Thus, these results indicate that this transformation system can be effectively utilized for the study of gene functions in plants.

Co‐immunoprecipitation and ChIP assays

Through yeast two‐hybrid (Y2H) screening using the reported TaSG1(TaSG) gene (semispherical grain, encoding a Ser/Thr protein kinase glycogen synthase kinase3) (Cheng et al., 2020) as bait, we identified potential interactions between TaSIP2/3/4 (SIP2/3/4)(SG interacting protein) and TaSG1 (Figure 4a). The interaction was further verified by using a yeast two‐hybrid point‐to‐point protein–protein interaction assay (Figure 4a). Following MCD‐mediated transient co‐transformation of two DNA constructs, one containing TaSG1 fused with 6 × HIS (histine) and the other containing TaSIP2/3/4 fused with 3 × FLAG, we then conducted co‐immunoprecipitation (Co‐IP) using anti‐HIS or anti‐FLAG antibodies followed by Western blotting assays based on anti‐FLAG‐ or anti‐HIS detection (Figure 4b). As shown in Figure 4b, a clear Western blotting signal was observed in seedlings co‐transformed with TaSG1 and an individual TaSIP2, TaSIP3, or TaSIP4. This finding confirms the physical interactions between the TaSG1 and TaSIP2/3/4. Additionally, these interactions were validated through a split‐LUC experiment as shown in Figure S14.

Figure 4.

Co‐immunoprecipitation and ChIP assays utilizing the MCD‐mediated transient transformation system. (a) Validation of interactions between TaSG1(TaSG) (semispherical grain, encoding a Ser/Thr protein kinase glycogen synthase kinase3) and TaSIP2/3/4(SIP2/3/4) (SG interacting protein) using yeast two‐hybrid point‐to‐point protein–protein validation. (b) Validation of interactions between TaSG1 and TaSIP2/3/4 through co‐immunoprecipitation (Co‐IP), which was conducted with anti‐HIS (histine) or anti‐FLAG antibodies, followed by Western blotting assay using anti‐FLAG or anti‐HIS antibodies. (c) Western blotting detection of H3K4me3 levels in seedlings with (H3K4M‐OE) or without (CK) overexpression of H3K4M, where the 4th lysine (K) amino acid from the N‐terminal of histone H3 is mutated into methionine (M), using anti‐H3K4me3 antibody. (d) Distributions of normalized read counts from H3K4me3 ChIP‐seq (chromatin immunoprecipitation coupled with sequencing) data in seedlings with (H3K4M‐OE) or without (CK) overexpression of H3K4M across ± 1.0 kb from the transcription start sites (TSSs) to transcription termination sites (TTSs) of genes. (e) Venn diagram illustrating the overlap of H3K4me3 peaks between CK and H3K4M‐OE samples. (f) Genomic distributions of CK/H3K4M‐biased and common H3K4me3 peaks within the wheat genome. The entire genome was partitioned into 5 functionally annotated subdomains, including promoters, introns, exons, downstream and intergenic regions.

Next, we conducted a ChIP‐seq (chromatin immunoprecipitation coupled with sequencing) experiment using native wheat leaf tissues with overexpression of histone H3K4M (H3K4M‐OE). In this variant, the 4th lysine (Lys, K) amino acid from the N‐terminal of histone H3 is mutated to methionine (Met, M) (Figure S15). We observed that the overexpression of H3K4M resulted in a global reduction of H3K4me3 enrichment levels (Figure 4c). Consistently, we found that seedlings with H3K4M‐OE exhibited lower normalized H3K4me3 ChIP‐seq read counts as compared with CK (Figure 4d). After peak calling, we identified 17 773 and 35 483 H3K4me3 peaks in seedlings with H3K4M‐OE and CK, respectively, with 8671 peaks common to both groups (Figure 4e). This finding is similar to previous research on H3K36M, where the 36th Lys (K) amino acid from the N‐terminal of histone H3 is mutated into Met (M), and H3K27M, with the 27th Lys (K) amino acid from the N‐terminal of histone H3 mutated into Met (M), in both humans and plants (Herz et al., 2014; Lin et al., 2018; Sanders et al., 2017). H3K4M‐OE‐biased H3K4me3 peaks were more prevalent in intergenic regions (~25% vs. 24%) but less distributed in exons (~ 44% vs. 46%) as compared with CK‐biased H3K4me3 peaks (Figure 4f). This indicates subtle differences in genomic distributions between H3K4M‐OE‐ and CK‐biased peaks.

Taken together, these results indicate that this transformation system can be efficiently applied to co‐immunoprecipitation and ChIP assays, which are essential steps in understanding gene functions or epigenetic mechanisms responsible for important agronomic traits in plants.

Potential applications for testing the efficiency of genome editing

MCD‐mediated transient transformation system holds promise as a vector‐free method for creating genome‐edited plants, despite the current lack of successful case studies. Increased investment is essential, as its success could circumvent restrictive policies regarding GMOs, thereby accelerating the release of bioengineered plants aimed at enhancing agronomic traits in crops. To evaluate the mutation efficiency of Cas9, we conducted seed‐based transient transformation using MCDs in conjunction with CRISPR‐Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats‐CRISPR‐Associated protein 9) vectors, which specifically target the aforementioned NLR1 and NLR2 genes, respectively, with or without the assistance of a gene gun. After sequencing two biological replicates for each vector using the Hi‐TOM (high‐throughput tracking of mutations) platform (Liu et al., 2019) and conducting data analyses, we found that for samples from seeds treated only with MCDs, we detected an average of 1.16% base A insertions, 1.71% SNP (single‐nucleotide polymorphism for T to C) for the gene NLR1, 3.48% deletion and 1.69% SNPs (T to C) for the gene NLR2. In contrast, for samples from seeds treated with both MCDs and the gene gun, we detected an average of 1.54% SNPs (T to C) and 2.35% base A insertions for the NLR1 gene, as well as 9.85% deletions and 1.18% base A insertions for the NLR2 gene (Figure 5). The total mutation rate of NLR1 increased from 2.87% to 3.89%, while the total mutation rate of NLR2 increased from 7.02% to 11.66% after the gene gun treatment. This suggests that gene gun treatment tends to elevate the occurrence of large DNA fragment deletion. Our study indicates that this system cannot be directly applied to genome editing studies due to the low efficiency of the resulting targeted mutagenesis. More efforts need to be invested in improving the genome editing efficiency or producing inheritable mutations, which could potentially lead to the generation of vector‐free genome‐edited plants in the future.

Figure 5.

Hi‐TOM for the detection of CRISPR‐Cas9‐based genome editing efficiency using the MCD‐mediated transient transformation system. (a) Summary of the mutation rate of sequence mutations within genomic DNA, which was extracted from seedlings germinated from wheat seeds that were transiently transformed with a CRISPR‐Cas9 vector specifically targeting the NLR1 and NLR2 genes. (b) Summary of sequence mutations from Hi‐TOM (high‐throughput tracking of mutations) data. (c) Detailed information on sequence mutations within the two target regions specific to the NLR1 and NLR2 genes. KO, knock‐out; PAM, protospacer adjacent motif; T, gene gun bombardment.

Discussion

It has been reported that nanomaterials can efficiently deliver nucleic acids such as DNA vectors or siRNA to the leaf, root, pollen, or callus tissues of various plants, including Arabidopsis, Trifolium repens, rice, wheat, tomato, and tobacco. This delivery can be mediated by various types of nanomaterials, including clay nanosheets, chitosan nanoparticles, polymer nanoparticles, peptide assemblies, mesoporous silica nanoparticles (MSNs), and carbon dots (CDs) (Cai et al., 2024; Christiaens et al., 2018; Delgado‐Martin et al., 2022; Demirer et al., 2019, 2020; Golestanipour et al., 2018; Kwak et al., 2019; Li et al., 2019; Liu et al., 2020; Mitter et al., 2017; Nino‐Sanchez et al., 2022; Numata et al., 2014; Schwartz et al., 2020; Torney et al., 2007; Wang et al., 2020; Zhang et al., 2010, 2019; Zhao et al., 2017). For instance, carbon dots (CDs), MSNs or BioClay™ nanomaterials have been successfully utilized to deliver RNAi system for gene silencing in plants, as well as for crop protection against viruses, fungi, and phytoparasitic nematodes (Cai et al., 2024; Mitter et al., 2017; Nino‐Sanchez et al., 2022; Opdensteinen et al., 2024; Schwartz et al., 2020). However, the applications of the aforementioned systems in comprehensive genomic and epigenomic studies are still largely uncharacterized. Our study further advanced the MCD‐mediated delivery of DNA plasmids into seeds, which were subsequently germinated into seedlings of various plant species that face challenges in achieving stable genetic transformation. This technical progress opens new avenues for the potentially practical application of nanomaterials in functional genomics and epigenomics research, as well as in the enhancement of agronomic traits in crops.

The study presents, for the first time, evidence indicating that DNA plasmids are capable of autonomous replication following their introduction into seeds (Figure S8). As a result, the replicated plasmid DNA can be disseminated from one cell to another cell during mitosis, leading to the development of germinated seedlings in which a significant proportion of cells harbour plasmid DNA. This may explain the findings regarding the LUC or GFP signal detected throughout the entire leaf tissue. However, our study also showed that the transferred DNA plasmids remain stable for only 8 to 10 days (Figure S9). One possible explanation is as follows: the presence of CDs may protect plasmid DNA from degradation within the cells. It has been reported that carbon nanocarriers, MSNs, DNA nanostructures, and LDHs (layered double hydroxides) can protect RNA from RNase A cleavage in vitro or in plant cell lysates (Cai et al., 2024; Demirer et al., 2020; Mitter et al., 2017; Zhang et al., 2019). Conversely, CDs are also unstable within cells; overtime, they degrade into smaller particles or into CO₂ and H₂O within cellular environments (Li et al., 2019), ultimately compromising their ability to protect plasmid DNA. It has been reported that 5 nm CDs can degrade into 3 nm CDs within 20 days in vitro (Li et al., 2019). LDHs was found to degrade into a biocompatible residue after exposure to atmospheric CO₂ and moisture (Mitter et al., 2017; Xu et al., 2006), leading to the release of bound dsRNA (double‐stranded RNA) (Mitter et al., 2017). Therefore, naked DNA plasmids may be vulnerable to in vivo DNase, leading to DNA degradation.

Additionally, our study provides evidence that MCD‐mediated seed transient transformation is an efficient system for genomic and epigenomic studies in wheat and other plant species that face challenges in stable transformation. This method is straightforward and time‐saving; furthermore, it does not require tissue culture and is independent of species or genotypes. However, this system still requires further investment to be modified for the generation of stably transformed plants. Here, we propose several potential strategies to enhance the success of nanomaterial‐mediated stable transformation in plants, particularly in crops. They include: (i) The selection of different tissues or cells for transformation, with or without regeneration, under long‐term antibiotic selection. This includes germ cells like pollen (Zhao et al., 2017) and microspores (Cho et al., 2020), as well as meristematic cells like calli or apical meristem cells. (ii) The use of asexual reproduction for transformation, with or without regeneration, under long‐term antibiotic selection; and (iii) Combinations of different transformation methods, such as the integration of MCDs with a gene gun or a production of various nanomaterials.

Materials and methods

Synthesis of MCDs

Carbon dots (CDs) were synthesized by using an electrochemical etching method, as previously reported by Li et al. (2019). Two high‐purity graphite rods were positioned in parallel in 1 L ultrapure water. Electrolysis was conducted at a direct voltage of 20 V (volt) for a duration of 4 weeks. The resulting solution was centrifuged to recover the supernatant, which was then dialysed with deionized water using a cellulose filter membrane (2.5kD) for 2 days. NEthyl‐N‐(3‐dimethylaminopropyl) carbodiimide hydrochloride (EDC) was added to 10 mL of CDs solution (1 mg/mL) for activating the carboxyl group for 2 h. Following this, 5 ML (5 mg/mL) of PEI600 (polyethyleneimine, Cat# 408719; Sigma, St. Louis, MO, USA) was added to generate modified CDs (MCDs). After 4 h of reaction, the solution was dialysed for 2 days. A FEI/Philips Tecnai G2 F20 transmission electron microscope was used to obtain transmission electron microscopy (TEM) images. The synthesized MCDs can remain stable at 4–8 °C for more than 6 months.

Plant materials and growth conditions

A total of 20 to 30 seeds from various plant species was used, including cultivated wheat varieties: Y15 (Yangmai 15), Y29 (Yangmai 29), Z12 (Zhenmai 12), Z366 (Zhengmai 366), AK58 (Aikang 58), BN607 (Bainong 607), N13 (Ningmai 13), Z168 (Zhenmai 168), SJZ407 (Shijiazhuang 407), S3 (Sumai 3), N9 (Ningmai 9), HS2 (Hesheng 2), Y158 (Yangmai 158), and CS (Chinese Spring), as well as their wild relatives (AA, BB, AABB and DD). Additionally, maize (Zea mays L.) B73, Hordeum vulgare var. coeleste L., Agropyron mongolicum Keng, Medicago sativa L.cv. Zhongmu NO. 3, Cucumber‐Gy14, Brassica rapa ssp. pekinensis (Chinese cabbage), Raphanus sativus L., C44, Solanum commersonii, a wild (2n), hybrid of S. phureja and S. chacoense (4n), and William 82 Glycine max (L.) Merr., were placed in a 5 mL test tube containing 2–3 mL of MCDs mixed with different types of DNA plasmids for transient transformation.

About 20–30 seeds, with or without (control, CK) treatment of MCDs mixed with or without DNA plasmids, were pre‐germinated in a petri dish (10 × 10 cm, EO sterilized, dust‐free). The dish contained autoclaved ddH₂O‐saturated filter paper (Brand: Cytiva, Standard No.: No: GB/T1914‐2017, medium speed ash ≤13%, Marlborough, MA, USA), at room temperature (RT). Germinated seeds were subsequently transferred into pots filled with artificial soil, consisting of 3:1 (v/v) mixture of nutrient soil and vermiculite. The plants were cultivated in a controlled growth chamber at day/night temperature of 25/20 °C, with a light–dark cycle of 12 h each, and maintained at a relative humidity of 80%. Leaf tissues or whole seedlings, with or without transient transformation using DNA plasmids of interest, were utilized for various experiments. They included biotic/abiotic treatments, recording of LUC, GFP, and GUS signals, as well as Western blotting, co‐IP (co‐immunoprecipitation), ChIP (chromatin immunoprecipitation coupled with sequencing), and the extraction of RNA and DNA for RT‐qPCR (Quantitative Reverse Transcription‐Polymerase Chain Reaction) or standard PCR assays.

MCD‐mediated seed transient transformation

About 20–30 seeds of interest were completely submerged in 2–3 mL MCD solution per tube, which contained 20 μg/μL of MCDs, 10 ng/μL of plasmid DNA, 0.5% Tween‐20, and 10 mm MES (2‐(N‐morpholino)ethanesulfonic acid) buffer (pH 5.8). The tubes were then placed into a fridge at 4 °C overnight. The following day, the tubes were removed and allowed to reach room temperature on the bench before being returned to 37 °C water bath for an additional 30 min. This was followed by two rounds of vacuum treatment for 5 min each, after which the tubes were placed back in the fridge at 4 °C overnight. On the third day, the MCD‐treated seeds were thoroughly washed with autoclaved ddH₂O for 3–4 times, and then put into a petri dish lined with two‐layered wet qualitative filter paper (Brand: Cytiva, Standard No.: GB/T1914‐2017, medium speed ash ≤13%) for germination at RT.

Wheat seedlings with biotic or abiotic treatment

For the treatment of PM (E26), wheat leaves, measuring 4.0 cm in size cut from 5‐to 6‐day‐old wheat seedlings, were placed flatly on a plate containing a culture medium, composed of 0.02 mg/L 6BA (N‐(Phenylmethyl)‐9H‐purin‐6‐amine, Cat.# 13 151; Sigma, St. Louis, MO, USA), 10 mg/mL Agar (Cat.# A8190; Solarbio, Beijing, China), and adjusted to a pH of 5.8. Spores of E26, which had been cultivated under optimal growth conditions, were evenly distributed onto the detached wheat leaves and inoculated leaves. The inoculated leaves were then incubated for 6 days at 22 °C, after which phenotypic observations and recording were conducted. To examine the structures of E26 on the 6th day post‐inoculation, wheat leaves with and without overexpression of the NLR gene were stained with Coomassie brilliant blue for 90s, and subsequently were submerged in autoclaved ddH₂O for 5 min. After washing three times with autoclaved ddH₂O, the signals were recorded by using a fluorescent microscope under bright field illumination.

For the treatment of Fusarium pseudograminearum, 2–3‐day‐old wheat seedlings (with sprouts measuring 0.5–1.0 cm in length) were inoculated with a liquid culture medium (sterilized mung bean soup) containing Fusarium pseudograminearum at a concentration of 1 × 10⁶ spores/mL. The inoculated seedlings were placed in an incubator to grow for 10 days at 22 °C, followed by phenotypic observation and data recording.

For heat treatment, 5‐day‐old wheat seedlings grown in soil were divided into two groups. One group continued to grow under the previously described normal conditions, while the other group was placed in an incubator at 40 °C as heat treatment for 8 h, then recovery at RT for 1 h before phenotypic observation and data recording.

Observation and recording of LUC, GFP, and GUS signal

And 5‐day‐old wheat seedlings, germinated from seeds transformed with DNA plasmids containing the CaMV (cauliflower mosaic virus) 35S promoter‐driven LUC reporter gene, were treated with 1 mm D‐Luciferin sodium salt (Cat.# 40901ES01; YESEN, Shanghai, China) as a substrate for signal development for 5 min followed by recording using a Plant in vivo imaging system (Berthold LB 985, Bad Wildbad, Baden‐Württemberg, Germany).

A leaf segment was excised from 3‐ to 5‐day‐old wheat seedlings that had germinated from seeds transformed with DNA plasmids containing the CaMV 35S promoter‐driven GFP reporter gene. The GFP signal was observed and recorded using a confocal microscope (LSM 900; Carl Zeiss, Oberkochen, Baden‐Württemberg, Germany). Signal intensity was quantified by using ImageJ, with three replicates per sample employed for the measurement of signal intensity. Briefly, the image was first converted into grayscale by using the analysis option of ‘Image‐Type‐8bit’ from the menu bar. Next, the background of the image was eliminated by adjusting the threshold followed by selecting the fluorescence area, which was conducted by using the analysis option of ‘Image→Adjust→Threshold’, and confirmed by clicking ‘Set’ and ‘OK’. Subsequently, the results were obtained by using the analysis option of ‘Analyse‐Measure’. The mean value was calculated as the final result.

A piece of roots and shoots was cut from wheat seedlings that were 2‐ to 3‐day old, having germinated from seeds transformed with pCAMBIA1300‐GUS plasmids for the purpose of GUS staining. The staining procedure was conducted by using a kit purchased from Coolaber (Cat.# Coolaber/SL7160, Beijing,China), in accordance with the manufacturer's instructions. Following a de‐staining process with 75% ethanol for a duration of 24 h, the stained tissues were examined and documented for GUS signal using a stereomicroscope.

Western blotting assays

Five‐day‐old wheat seedlings, germinated from seeds transformed with DNA plasmids containing the CaMV 35S promoter‐driven GFP reporter gene or the H3K4M overexpression plasmid, were utilized for the extraction and purification of total protein or nucleoprotein. This sample was subsequently analysed via Western blotting with anti‐GFP (Cat.# ab290; abcam, Cambridge, MA, USA) or anti‐H3K4me3 (Cat.# A223146; ABclonal, Wuhan, China) antibodies. Briefly, total protein and nucleoprotein were isolated through the application of 12% SDS‐PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis), after which the proteins were transferred to an Amersham Hybond‐N⁺‐nylon membrane (PVDF, pore size: 0.2 μm; Millipore, Billerica, MA, USA) by using an electro‐blotting apparatus. The membrane was initially blocked with 5% milk (non‐fat powdered milk, Cat. #A600669; BBI, Shanghai, China) for 30 min at RT followed by an overnight incubation at 4 °C with either the anti‐GFP or anti‐H3K4me3 antibody at a dilution of ~1:5000, with gentle rotating. After washing with 1 × PBST (phosphate‐buffered saline +0.1% Tween‐20) three times and incubating with an anti‐HRP (horseradish peroxidase)‐conjugated rabbit polyclonal antibody (Cat. # D110011; BBI), the membrane was used for immune‐signal development by using the SuperPico ECL Chemiluminescence kit (Cat.# E422‐01; Vazyme, Nanjing, China), which was conducted following the standard procedures.

Co‐immunoprecipitation assay

DNA plasmid pMWB110‐6 × HIS which contains the TaSG gene was mixed with the pCAMBIA1305‐3 × FLAG plasmid, which harbours the uncharacterized genes SIP2, SIP3, or SIP4 in a 1:1 ratio for MCD‐mediated seed transient transformation. 5‐day‐old wheat seedlings, with or without the co‐transformation of the mixed DNA plasmids, were used for total protein extraction. The extracted proteins were then incubated with either an anti‐6 × HIS Tag rabbit polyclonal antibody (Cat.# D110002; BBI, Shanghai, China) or an anti‐FLAG Tag rabbit polyclonal antibody (Cat.# D110005; BBI) for co‐immunoprecipitation. The proteins immunoprecipitated with anti‐HIS or anti‐FLAG antibodies were subsequently analysed using Western blotting with anti‐FLAG or anti‐HIS antibodies (1:5000), following the previously mentioned protocol.

ChIP‐seq sequencing and data analyses

Wheat seedlings with (H3K4M‐OE) or without (control, CK) overexpression of H3K4M were utilized for nuclei purification and H3K4me3‐related ChIP experiments, which were conducted as previously described (Zheng et al., 2019). Input and ChIPed DNA were recovered for preparation of the ChIP‐seq library, which was constructed by using the NEBNext Ultra II DNA Library Prep Kit for Illumina (Cat.# E7645S; NEB, County Road, Ipswich, MA, USA) and finally sequenced on the Illumina NovaSeq Xplus platform with 150 bp paired‐end mode.

For the analysis of ChIP‐seq data, raw reads were processed using fastp (v0.21.0) (Chen et al., 2018) for the removal of adapters and low‐quality bases. The clean reads were then aligned to the wheat reference genome (IWGSC RefSeq v1.0) using Bowtie2 (v2.5.1) (Langmead and Salzberg, 2012). All aligned reads were sorted and filtered using Samtools (v1.5) (Danecek et al., 2021) with the parameter ‘‐q 20’. Picard (v2.23.3) (https://broadinstitute.github.io/picard/) was employed to remove duplicated reads. The de‐duplicated BAM files were converted into BigWig files using bamCoverage from Deeptools (v3.5.0) (Ramirez et al., 2016) with the parameters ‘‐‐bs 10 ‐‐effectiveGenomeSize 14600000000 ‐‐normalizeUsing RPKM ‐‐smoothLength 50’. The BigWig files were visualized using Deeptools (v3.5.0) (Ramirez et al., 2016) and IGV (v2.8.9) (Thorvaldsdottir et al., 2013). Macs2 (v2.2.7.1) (Zhang et al., 2008) was used for H3K4me3 peak calling using the BAM files with the parameters ‘‐f BAMPE ‐g 1.42e+10 ‐q 0.01’.

Validation of enhancers

DNA sequences were amplified from the candidate enhancer by using PCR. The amplified DNA fragment was ligated into a modified pJIT163‐hGFP vector or an LUC reporter gene vector containing the mini 35S promoter for enhancer validation by using the MCD‐mediated seed transient transformation system mentioned above. The LUC signal was detected as previously mentioned. For the inducible enhancer experiment, 3‐ to 5‐day‐old wheat seedlings were divided into two equal halves. One half was evenly inoculated with E26 spores for 8 h, while the other half without inoculation with E26 served as the control. A leaf segment was excised from each wheat seedling with or without inoculation of E26 for recording the GFP signal using the ZEISS confocal LSM 900 (Oberkochen, Baden‐Württemberg, Germany).

RT‐qPCR assay

Total RNA was extracted from 5‐day‐old wheat seedlings with or without E26 inoculation for 8 h by using TRIzol reagent (Cat.# 15596018CN; ThermoFisher Scientific, Waltham, MA, USA). After the complete removal of genomic DNA contamination, the purified RNA was used for the synthesis of first‐stranded cDNA by using HiScript II Q RT SuperMix (Cat.# R223‐01; Vazyme, Nanjing, China). The cDNA was mixed with Taq Pro Universal SYBR qPCR Master Mix (Cat.# Q712; Vazyme) for qPCR assay, which was conducted in C1000 Touch Real‐Time PCR system (Bio‐Rad, Hercules, CA, USA) under the following conditions: 95 °C × 5 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. qPCR results were calculated by using the 2^−ΔΔCT method (Schmittgen and Livak, 2008) and expressed as relative expression levels of the beta‐actin gene (an internal control). The primer sequences are provided in Table S1.

Bimolecular fluorescence complementation (BiFC) assay

In the BiFC assay, the complete CDS (coding DNA sequences) sequences of two target genes were inserted into vectors containing Pxy‐104 (cYFP) and Pxy‐106 (nYFP), respectively to facilitate MCD‐mediated seed transient transformation. Leaf tissue samples were excised from wheat seedlings aged 3–5 days, which had been transformed with either a single gene vector or a co‐transformation of both vectors for the purpose of measuring the YFP signal using the ZEISS confocal LSM 900.

Autonomous replication of transformed DNA plasmids

A total of 10–20 seeds, which were incorporated with a DNA plasmid through MCD‐mediated transformation, were pre‐germinated in a petri dish (10 × 10 cm, EO sterilized, dust‐free). This dish contained sterilized ddH₂O saturated filter paper (Brand: Cytiva, Standard No.: GB/T1914‐2017, medium speed ash ≤13%), and the pre‐germination process was conducted at RT. Once the sprout reached a size of 1 cm, 4–5 germinated seeds were placed into 1.5 mL Eppendorf tube containing a solution of normal dATP, dGTP, dCTP, dTTP or a combination of dGTP, and dTTP plus biotin‐11‐dCTP (Cat.# orb1785980; Biorbyt, Cambridge, Cambridgeshire, UK) and biotin‐14‐dATP (Cat.# orb533181; Biorbyt). The seeds were then allowed to grow for 3 and 5 days in a growth chamber. Tissues from the roots, leaves, or stems of 3‐ to 5‐day‐old wheat seedlings were collected for the extraction of genomic DNA. DNA, with or without (control, CK) incorporation of biotin‐dCTP/dATP was enriched by using Dynabeads M280 streptavidin (Cat.# 11206D; ThermoFisher Scientific) for a PCR assay using primers specific to the transformed DNA plasmid.

Transcriptional activity assay

The promoter sequences of the PRE6 gene were ligated into the pGreenII 0800‐LUC vector, while the full coding sequences (CDS) of the Rf2b gene were ligated into the pGreenII 62‐SK vector. Both vectors were combined for MCD‐mediated seed transformation. 3‐ to 5‐day‐old wheat seedlings, germinated from seeds transformed with the mixed DNA plasmids or an empty vector (control, CK), were treated with 1 mm D‐Luciferin sodium salt as a substrate for signal development for 5 min, followed by imaging using the Plant in vivo Imaging System (Berthold LB 985, Bad Wildbad, Baden‐Württemberg, Germany).

Validation of protein–protein interactions by yeast two‐hybrid assay

The full length of cDNA from the TaSG gene was cloned into the pGBKT7 (BD) vector, while the full length of cDNA from the TaSIP2/3/4 gene was cloned into the pGADT7 (AD) vector. The expression plasmids encoding the bait and prey were mixed and co‐transformed into the yeast (Saccharomyces cerevisiae) strain AH109. Transformants were initially screened using a synthetic complete medium (SD‐L/W) that lacked leucine (Leu) and tryptophan (Trp). Interactions were subsequently detected using a synthetic complete medium (SD‐L/W/H/A) without Leu, Trp, histidine (HIS), and adenine (Ade). The pGBKT7‐p53 and pGADT7‐T constructs served as a positive control, while the pGBKT7‐Lam and pGADT7‐T constructs were utilized as a negative control (Lin and Lai, 2024). Positive clones were selected and re‐suspended in autoclaved H₂O with serial dilutions (1 ×, 10 × and 100 ×). 10 μL of the diluted cultures were spotted onto plates containing synthetic dropout (SD) yeast medium (without leucine, tryptophan, and histidine, SD/‐T/‐L/‐H) (Lin and Lai, 2024). After air‐drying, plates were sealed with parafilm and inverted before being placed in an incubator for the culture of 2–3 days at 30 °C. The primers used in this study are listed in Table S1. Detailed information regarding the candidate genes used in this study is provided in Table S2. The components of each medium used for yeast culture are detailed in Table S3.

CRISPR‐Cas9‐based editing and sequencing

Sequences of gRNA specific for the genes NLR1 and NLR2 were ligated into the CRISPR‐Cas9 system for MCD‐mediated seed transformation, with or without the assistance of a gene gun. 5‐day‐old wheat seedlings were used for the extraction of genomic DNA, which was subjected to Hi‐TOM sequencing according to the published protocol (Sun et al., 2024). Two biological replicates were performed to assess mutation efficiency.

The ballistic transfer of MCD‐bound DNA plasmids

About 10 μL of MCDs were mixed with the aforementioned CRISPR‐Cas9 DNA plasmids and introduced into wheat seeds using a helium‐driven biolistic transformation system (PDS‐1000/He Particle Delivery System device, Bio‐Rad). The process was conducted at a vacuum pressure of 88–94.8 KPa, in accordance with the manufacturer's instructions. Approximately 20–30 bombarded seeds were germinated on a petri dish containing wetted filter paper and 100 ng/μL of hygromycin at RT for 48 h. After washing the seeds three times with ddH₂O, the germinated seeds were transferred into nutrient soil for growing in a controlled growth chamber, maintained at a day/night temperature of 25/20 °C, with a 12 h light/12 h dark cycle and a relative humidity of 80%. Genomic DNA was extracted from 5‐day‐old seedlings that were transformed with or without the CRISPR‐Cas9 vector. This DNA was them amplified using PCR and sequenced on the Hi‐TOM platform as previously described (Sun et al., 2024).

Funding

This research was supported by grants from the National Natural Science Foundation of China (U23A20179, 82371544); Fundamental Research Funds for the Central University (No. XUEKEN2022012); the Major Program of the National Agricultural Science and Technology of China (NK2022060101); Research on Key Technologies for Multi‐dimensional Precision Identification and Germplasm Creation of Salt alkali tolerant and Suitable Crops (2024BBF02002); Jiangsu Agricultural Science and Technology Innovation fund (CX(19)1001); the Natural Science Foundation of Jiangsu Province (BK20210383); Seed Industry Revitalization Project of Jiangsu Province (JBGS2021006); The R&D Foundation of Jiangsu province, China (BE2022425).

Conflict of interest

The authors declare no conflicts of interest.

Author contributions

W.L.Z. conceived and designed the study. L.W.S. and S.M.S. performed the experiments. X.J.C. and C.Y.C. assisted with material preparation and supervised the experiments. X.M.Y. assisted with F. pseudograminearum inoculation. F.X.D., J.X., and X.E.W. assisted with PM (E26) inoculation and vector preparation. H.J., M.T., and Z.H.K. provided MCDs and edited the manuscript. L.W.S., H.J., Z.H.K., and W.L.Z. interpreted the results. W.L.Z. wrote the manuscript with contributions from all other authors.

Supporting information

Figure S1 Electron microscope observation of modified carbon dots (MCDs).

Figure S2 The diagram illustrates the primary procedures involved in MCD‐mediated transient transformation with seeds, resistance screening and signal recording.

Figure S3 Phenotypic comparisons of seed germination and subsequent seedling development at various stages, with or without treatment using MCDs containing with or without DNA plasmids.

Figure S4 Detection of the GUS signal from wheat seeds, as well as root, and sprout tissues germinated from seeds treated with MCDs, with or without the CaMV 35S promoter‐driven GUS gene.

Figure S5 Basta screening of wheat seedlings germinated from seeds treated with MCDs, with or without the CaMV 35S promoter‐driven Basta gene.

Figure S6 Detection of LUC and GFP signals from maize B73 seedlings germinated from B73 seeds treated with MCDs with or without the CaMV 35S promoter‐driven LUC gene.

Figure S7 Detection of LUC signal from seedlings germinated from seeds of different species.

Figure S8 PCR detection of autonomous replication of a DNA plasmid within wheat seedlings germinated from seeds treated with MCDs.

Figure S9 PCR analysis demonstrating the time‐dependent degradation of plasmid DNA within wheat seedlings following MCD‐mediated transfer.

Figure S10 Relative expression levels of an intronized gene from wheat seedlings containing a MCD‐mediated transferred plasmid with its complete CDS (without intron) or genomic sequences (with intron).

Figure S11 MCD‐mediated seed transient transformation of the Pm21 gene fused with the GFP reporter gene.

Figure S12 Germination of wheat seeds treated with MCDs containing with or without the overexpression of NLR1, followed by screening with or without hygromycin.

Figure S13 Phenotype of wheat seedlings with (bHLH‐OE) or without (WT) overexpression of bHLH.

Figure S14 Split‐LUC experiment demonstrating the interactions between TaSG and SIP2, 3 or 4, respectively.

Figure S15 Phenotype of wheat seedlings with (H3K4M‐OE) or without (WT) the overexpression of H3K4M, which were transiently transformed using MCDs.

Table S1 Summary of Primer sequences used in this study.

Table S2 Vector information used in this study.

Table S3 Summary of media for yeast culture.

Data availability statement

All the data generated in this study are available upon request. The data that supports the findings of this study are available in the supplementary material of this article.