Novel R2R3-MYB Transcription Factor LiMYB75 Enhances Leaf Callus Regeneration Efficiency in Lagerstroemia indica

Abstract

Lagerstroemia indica is an important woody ornamental plant worldwide. However, the application of many technologies, such as transgenic breeding and genome editing, has been severely hampered due to the lack of efficient calli induction and regeneration technology. Here, we discussed a reliable and efficient calli induction and regeneration protocol using whole-leaf explants. This protocol’s effectiveness for the calli induction and regeneration systems in crape myrtle were up to 70.33% and 44.33%, respectively. Next, an efficient and stable Agrobacterium-mediated genetic transformation system was created from leaf calli, and the green fluorescent protein (GFP) was able to detect up to 90% of its positive frequency. Meanwhile, two positive lines’ transfer DNA insertion sites and directions were identified using whole genome sequencing. LiMYB75, a novel R2R3-MYB transcription factor, was identified and transferred to the L. indica genome to enhance the leaf calli regeneration frequency. Surprisingly, overexpressing LiMYB75 increased the frequency of calli regeneration in the leaf by 1.27 times and the number of regenerated plantlets per callus by 4.00 times compared to the wild type, by regulating the expression levels of genes involved in callus formation, such as SHOOT MERISTEMLESS (STM). Overall, our findings revealed a simple, reliable, and highly efficient transformation approach and identified the desirable candidate gene LiMYB75, which improves L. indica’s calli regeneration efficiency.

1. Introduction

Lagerstroemia indica L. (crape myrtle) is a deciduous shrub or small tree of the genus Lagerstroemia with significant ornamental and economic value due to its gorgeous flower color, long-lasting summer flowering period, and smooth/clean trunk [1]. The native Asian crape myrtle is found in tropical and northern temperate regions. The creation of unique flower and leaf colors is one of L. indica breeding’s primary objectives due to the continuously increasing demand in China. Although traditional breeding techniques have selected numerous top-notch crape myrtle types with various flower colors (white, red, purple, and combined variants), there are still unclear limitations and drawbacks. Although more and more functional genes have been identified based on transcriptomic and proteomic analyses [2,3,4], their function in L. indica remains unknown. Therefore, a highly efficient genetic transformation system is crucial for gene function analysis and molecular breeding in crape myrtle.

A highly efficient calli induction and regeneration system must be established to advance genetic transformation research and transgenic breeding. Although the calli have been applied for the highly efficient genetic transformation of a few woody plant species, including Camellia sinensis [5], Castanea sativa [6], and Quercus suber [7], the callus induction and regeneration technology for the majority of woody plants is relatively immature due to the genotype limitation [8]. Several factors, including plant hormones, culture conditions, and the source of the explant tissue, impact calli induction and regeneration [9]. Various tissues, including roots, leaves, and petioles, have been employed to induce calli in woody plants over the past few decades [10]. In Populus tomentosa, callus induction occurs more frequently than 97% of the time. The petioles are employed for both callus induction and regeneration [11]. In calli induction and regeneration, the petioles and leaves of triploid poplar (Populus alba × P. glandulosa) × P. tomentosa are better explants than roots [12]. For other tree species [13], a high-frequency regeneration system has been developed for annatto (Bixa orellana) using seed-induced callus and hypocotyl. The latter reduced the duration needed to acquire regenerated plants in 90 d. However, the effective callus induction and regeneration system of L. indica has not yet been widely reported.

Genetic transformation has been used to explore gene functions, and molecular plant breeding has been utilized since the Agrobacterium-mediated leaf disk transformation technique was first applied [14,15]. Little progress has been made in the genetic transformation of woody plants thus far because the success of plant genetic transformation primarily depends on the effectiveness of gene delivery into plant cells and the ability for transgenic calli induction and regeneration [16,17,18]. Therefore, ineffective plant regeneration is a major barrier to genetic engineering of woody plants.

A growing body of information from in vitro Arabidopsis regeneration over the past few decades has revealed that the early stages of callus formation and petal development share many similarities [19,20]. One of the well-known groups of plant transcription factors (TFs) is the MYB gene family. According to the number of highly conserved MYB DNA-binding domains, they can be divided into four categories: MYB-related (1R-MYB), R2R3-MYB (2R-MYB), R1R2R3-MYB (3R-MYB), and 4R-MYB [21]. R2R3-MYBs are the largest MYB TF family found in plants, and their members are involved in several biological processes, including petal development [22]. According to a recent study, two R2R3-MYB TFs redundantly inhibited callus induction in Arabidopsis by suppressing the expression of the target gene [23]. These results suggest that MYB TFs might be essential for callus induction and regeneration.

We previously constructed a simple and reliable transformation system and found that its transformation efficiency may reach up to 25.93% using stem segment explants from L. indica [24]. The following steps were taken in the current study to enhance the efficiency of the genetic transformation: (1) a quick, efficient system for callus induction and regeneration was developed using whole-leaf explants; (2) based on the system, a novel and sustainable Agrobacterium tumefaciens-mediated genetic transformation technique was developed; and (3) using our petal transcriptome data (unpublished), a novel MYB TF LiMYB75 was discovered, which increased the callus regeneration efficiency by regulating the RNA levels of genes involved in plant hormone metabolism in crape myrtle. Our findings might offer a useful tool for functional gene verification of important traits and molecular breeding and a new approach to overcoming genotypic limitations on callus induction and regeneration in L. indica.

2. Materials and Methods

2.1. Plant Material and Growth Conditions

The wild relative was accession NL210702. The Carl Whitcomb breeding program produced Lagerstroemia Dynamite (Carl Whitcomb, Lacebark Inc. Stillwater, OK, USA). The Institute of Botany’s Lagerstroemia Resource Nursery; Jiangsu Province; and the Chinese Academy of Sciences, Nanjing, China (32°06′ N, 118°84′ E) were the origins of the two germplasm resources. Sterile seedlings and regenerated shoots were grown in the culture room between 25 °C and 28 °C, 1500 lx light intensity, and 14-h light/10-h dark photoperiod. A temperature range of 25–28 °C, 6000 lx light intensity, and a 12-h light/12-h dark photoperiod were used to develop L. indica plantlets and tobacco (Nicotiana tabacum) plants in an artificial climate chamber. All samples were immediately frozen in liquid nitrogen and stored at −80 °C.

2.2. Callus Induction and Regeneration Using Whole-Leaf Explants

The whole-leaf explants were collected in sterile conditions and transferred to nine different callus induction media (CIM 1–9, 18–20 leaves/CIM, Supplementary Table S1). The calli produced by the whole-leaf explants were counted 20 days after incubating them on the CIMs.

The generated calli were placed onto the callus regeneration medium (CRM1–9, 12 calli/CRM, three repetitions, Supplementary Table S2) for differentiation and shoot regeneration. The calli with regenerated shoots were counted after 24 d of incubation on CRMs. The frequency of plant generation was estimated by dividing the total number of calli cultured by the number of calli with differentiating shoots.

2.3. Shoot Rooting and Transplantation

The generated shoots were transferred to CRM1 to generate adventitious roots (20 shoots/CRM, three repetitions). The cap was removed from the culture chamber to allow the well-rooted plantlets to acclimate for seven days. Subsequently, the plantlets were washed and transferred to pots with peat: perlite (2:1). In the greenhouse, the pots were covered with transparent polybags. These plants were then either transferred to field conditions or hardened in the greenhouse.

2.4. Reconstruction Vector and Agrobacterium Culturing

The binary vectors pBI121 and pBinGFP4 were digested by Sac I and Hind III, respectively, to construct the novel vector pBI121-GFP. In other words, the 35S::GUS fragment in pBI121 was replaced by 2 × 35S::eGFP in pBinGFP4. A selectable marker gene for plants and bacteria, neomycin phosphotransferase (nptII) is similar to the reporter gene eGFP. Agrobacterium tumefaciens strain EHA105 was then given the vector pBI121-GFP.

Centrifugation was used to obtain the overnight Agrobacterium tumefaciens strain EHA105 culture. In resuspension solution (10 mM MES, 10 mM MgCl₂, 200 μM AS, pH = 6.0), Agrobacterium suspensions of various concentrations, determined with optical density (OD₆₀₀ 0.6, 0.8, and 1.0), were prepared. These suspensions were incubated in the dark for 3–4 h and used to inoculate the explants.

2.5. L. indica Leaf Transformation and Regeneration

Whole-leaf explants were continuously shaken at room temperature for varying times (10, 15, or 20 min, Supplementary Table S4) while being infected with Agrobacterium resuspensions (OD₆₀₀ 0.6, 0.8, or 1.0, Supplementary Table S3). The infected whole-leaf explants were dried by blotting them on sterile filter paper before being placed vertically into the co-cultivation medium Driver and Kuniyuki walnut basal medium supplemented with 1.0 mg·L⁻¹ of 6-benzylaminopurine (6-BA), 0.1 mg·L⁻¹ of 1-naphthylacetic acid (NAA), 30 g·L⁻¹ of sucrose, and cultured at 22–25 °C in the dark for 24, 48, or 72 h (Supplementary Table S5). The putatively transformed whole-leaf explants were placed onto CIM3 for callus induction and CRM1 for regeneration.

2.6. Fluorescence Screening and DNA Detection of Transgenic Plantlets

The leaves were collected and observed under a fluorescence imaging system (Tanon-5200 Multi, Shanghai, China) based on GFP fluorescence to screen the transgenic plantlets. The positive rate was obtained by dividing the total number of plantlets tested by the number of plantlets that showed GFP fluorescence. After being isolated from GFP-positive lines, subculturing regenerated plants established different transgenic lines on a proliferation medium.

A plant genomic DNA kit (DP3211, Bioteke Biotech, Wuxi, China) was used to extract DNA from the leaves of regenerated plantlets to verify the positive plantlets further. To determine the quantity of whole genomic DNA, 1.0% agarose gel electrophoresis was used. Polymerase chain reaction (PCR) testing using eGFP-specific primers helped confirm the putative transgenic plantlets (Supplementary Table S6).

2.7. Genome Resequencing Using Next-Generation Sequencing (NGS)

Genomic DNA was isolated from wild-type plants and two transgenic lines, following the manufacturer’s instructions, using a plant DNA extraction kit (DP305, Tiangen Biotech, Beijing, China). Nextomics Bioscience Co. Ltd. (Wuhan, China) carried out library construction, genome resequencing using NGS, and bioinformatics analysis. The MGISEQ-2000 sequencing platform (MGI, Shenzhen, China) was applied to construct the libraries and perform sequencing. An Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) and Qubit 2.0 (Life Technologies, Carlsbad, CA, USA) were used to assess the concentration and quality of the library.

The raw data were screened following the sequencing process to eliminate adaptor sequences and low-quality reads (Q < 20) using Fastp software v0.23.1 (https://github.com/OpenGene/fastp (accessed on 17 August 2022)) and provide clean, high-quality data. The clean reads were mapped to the Lagerstroemia Jinhuang genome and T-DNA sequences to identify junction reads using TBtools software [25]. Following identification of the junction reads that aligned with both the reference genome sequence and the vector sequence, the T-DNA insertion sites and orientations were established using the alignment data associated with the junction reads. The structures of genes positioned close to the insertion sites (10 kb) were identified based on the Lagerstroemia Jinhuang genome annotation and were also predicted by the online tool FGENESH (http://www.softberry.com/berry.phtml?topic=fgenesh&group=programs&subgroup=gfind (accessed on 17 August 2022)). The results were visualized using IGV-GSAman software (https://gitee.com/CJchen/IGV-sRNA (accessed on 17 August 2022)). The genome DNA-sequence dataset is available at the NCBI Sequence Read Archive with accession number PRJNA883013.

2.8. Cloning, Sequencing, and Structural Analysis of the LiMYB75 Gene

Lagerstroemia Dynamite’s petals were used to extract the total RNA. The gene was cloned using a specific primer and complementary DNA (cDNA) as a template, followed by sequence verification after gel recovery (Supplementary Table S6). Sequences of the MYB gene from Vitis vinifera, Arabidopsis thaliana, Theobroma cacao, Eucalyptus grandis, Populus euphratica, Gossypium hirsutum, Gossypium raimondii, and Citrus sinensis were downloaded from the Plant Transcription Factor Database (http://planttfdb.gao-lab.org/family.php?sp=Ath&fam=MYB (accessed on 20 August 2022)).

ClustalW in MEGA5.0 was used to align all sequences with the default parameters. MEGA5.0 produced a phylogenetic tree using the neighbor-joining method and 1000 bootstrap replicates. ProtComp version 9.0 software (http://linux1.softberry.com/berry.phtml?topic=protcomppl&group=programs&subgroup=proloc (accessed on 25 August 2022)) was used to analyze the possible subcellular localization. The ProtParam program was used to calculate the molecular weight, theoretical isoelectric point (pI), amino acid sequence length, instability index, and grand average of hydropathicity. (http://web.expasy.org/protparam/ (accessed on 25 August 2022)). The self-optimized prediction method with multiple alignments software (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_sopma.html (accessed on 25 August 2022)) was used to predict the helices, sheets, and coils in the protein’s secondary structure.

2.9. Subcellular Localization

The full-length coding sequence (CDS) of LiMYB75 was cloned into pBinGFP4 between the KpnI and BamHI restriction sites to create the LiMYB75-eGFP construct. In contrast to the empty vector used as a control, the recombined vector was transformed into Agrobacterium tumefaciens strain EHA105 and transiently expressed in epidermal cells of tobacco (Nicotiana benthamiana) leaves via agroinfiltration. Next, the tobacco leaves were observed under a confocal laser scanning microscope (LSM900, Zeiss, Germany) after being cultured for 48 h in the dark. Tobacco leaf cells were injected with 10 μg·mL⁻¹ 4′,6-diamidino-2-phenylindole for 15–30 min before observation to stain the nuclei.

2.10. Overexpression of LiMYB75 in L. indica

The CDS of LiMYB75 was inserted into the KpnI and BamHI sites of the vector pBI121-GFP to create the construct 2 × 35S::LiMYB75-eGFP. Agrobacterium tumefaciens strain EHA105 (optical density [OD]₆₀₀ 0.6, 0.8, or 1.0) was transformed into wild-type crape myrtle (L. indica NL210702) after receiving the vector 2 × 35S::LiMYB75-eGFP. The Agrobacterium tumefaciens strain EHA105 carrying empty vector pBI121-GFP (OD₆₀₀ = 0.6, 0.8, or 1.0) was applied to the control mock.

2.11. Quantitative Real-Time (qPCR) Analysis

The FastPure Plant Total RNA Isolation Kit (RC401, Vazyme Biotech, Najing, China) was applied to extract total RNA from wild-type and transgenic plants. The HiScript III 1st Strand cDNA Synthesis Kit (+gDNA wiper) (R312, Vazyme Biotech, Najing, China) was used to generate the cDNA library. SYBR Premix Ex Taq II (RR820, Takara, Beijing, China) was used for qPCR in the Applied Biosystems Step One Plus TM RT-PCR System (Applied Biosystems, San Francisco, CA, USA) according to the manufacturer’s instructions. The 20 µL reaction mixture for qPCR included 10 µL of 2X SYBR Green Master Mix, 2 µL of diluted cDNA, 0.4 µL of ROX Reference Dye, 0.8 µL of each primer (10 µM), and 6.0 µL of RNAase-free water. The following are the details of the RT-PCR program: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s, then 60 °C–95 °C to produce the melting curve.

LiEF1α was amplified to normalize the quantity of the gene-specific qPCR product. Each assay was performed in triplicate to ensure the reproducibility of the results. The 2^−ΔCT and 2^−ΔΔCT methods (CT method) were used to analyze gene expression. All primers are listed in Supplementary Table S6.

2.12. Statistical Analysis

Microsoft Excel 2019 was used to organize the data. Statistical analyses were performed using IBM SPSS Statistics 26 (https://www.ibm.com/cn-zh/spss (accessed on 21 July 2022)).

3. Results

3.1. Callus Induction and Regeneration Using Whole-Leaf Explants in L. indica

The fresh leaves from accession NL210702 were transferred to callus induction media (Figure 1a,b). Interestingly, the white-green callus appeared on leaf petioles after 14–20 days of incubation in callus induction media CIM1–CIM9 (Figure 1c,d). These calli turned dark brown and regenerated shoots after 28 days of being transferred into callus regeneration media CRM1–CRM9 (Figure 1e,f). The regenerated shoots were allowed to multiply and grow roots on CRM1, with an adventitious rooting rate of 86.00% ± 3.67 (Figure 1g,h). When transferred to a greenhouse, the rooted plantlets exhibited a 90%–95% survival rate (Figure 1i,j).

Eighteen different induction media with cytokinin and auxin were examined to determine how they influenced callus induction and regeneration. The majority of callus induction (70.33%) was seen in Woody Plant Medium (WPM) supplemented with 0.8 mg·L⁻¹ 6-BA and 0.5 mg·L⁻¹ NAA (CIM3), where the cytokinin/auxin ratio reached 1.60 (Figure 1k and Supplementary Table S1). Of the initial explants, 22.00%–70.33% responded to cytokinin and auxin and formed calluses. Callus induction rates gradually decreased as cytokinin and auxin levels increased. The minimum in medium CIM4 was 22.00%. The calluses from the CIM media were transferred into CRM after three weeks of induction. In the medium CRMs, 16.67%–44.33% of the callus-regenerated sprouts appeared after three weeks of incubation (Figure 1l and Supplementary Table S2). The frequency of regeneration decreased as cytokinin concentration increased, and the optimal concentration for callus regeneration was 1.0 mg·L⁻¹ 6-BA (Supplementary Table S2). The rates of callus regeneration gradually decreased with the decrease in the ratio of cytokinin and auxin concentration when the concentration of 6-BA was 1.0 mg·L⁻¹. CRM1 (WPM + 1.0 mg·L⁻¹ 6-BA and 0.1 mg·L⁻¹ NAA) had the highest callus regeneration rate among the nine callus regeneration media. These findings led us to select CRM1 as the subsequent proliferation medium.

3.2. Transformation Efficiency of L. indica

After exposure to varied Agrobacterium concentrations, infection times, and co-culturing times, a total of 270 crape myrtle leaves were transferred to CIM3 and CRM1. The leaves of transgenic plants that were successful after nine weeks of culture showed green fluorescence under a fluorescence imager (Figure 2a,b). Meanwhile, PCR-based confirmation revealed an extremely significant positive correlation with the fluorescence results (Figure 2c,d). These findings demonstrated the efficacy of fluorescence detection as a rapid and nondestructive method for identifying transgenic plants in L. indica.

In earlier studies, we found that the concentration of Agrobacterium was crucial for the infection of L. indica [24]. We initially attempted to screen transgenic plants without antibiotics due to the reduced rate of callus regeneration. Surprisingly, the positive rates reached a maximum value of 90% and were not significantly influenced by Agrobacterium concentration, suspended duration, or co-culturing time (Figure 2e–g). These findings suggest that future transgenic research of L. indica might not require antibiotics. In the case of L. indica, which had a maximum transformation efficiency of 37.71%, OD₆₀₀ = 0.8 was the most effective (Figure 2e, Supplementary Table S3). These results might have resulted from the Agrobacterium’s robust proliferation and significant infection activity at this bacterial concentration (Supplementary Figure S1). Therefore, the optimal bacterial concentration for L. indica’s genetic transformation was 6 × 10⁸ cells·mL⁻¹.

Additionally, infection time had no significant impact on how effectively transgenic shoots were transformed (Figure 2f, Supplementary Table S4). However, suppose the infection lasts too long. In that case, the Agrobacterium will adhere to the leaves too much and cause excessive damage to the wound, making it difficult to introduce a healthy callus and produce positive transgenic shoots. This will also significantly increase the contamination rate of the affected plants. Therefore, a 15 min infection period was optimal for L. indica transformation.

According to Figure 2g, the positive rate increased as the co-cultivation time increased, reaching a maximum frequency of 86.67% (Supplementary Table S5). However, the factor of co-cultivation time did not significantly impact the transformation frequency (Supplementary Table S5). During the experiment, it was observed that the leaves showed signs of withering and yellowing, as well as contamination with Agrobacterium with the expansion of the co-cultivation duration. These possibilities explain how co-cultivation time affects the rate of improvement in genetic transformation. Thus, when submerged in EHA105 resuspension, a 48-h co-cultivation duration was optimal.

3.3. Verification of T-DNA Insertion by Whole-Genome Resequencing

We performed genome sequencing on two positive lines (#OE1, #OE2) and the wild-type plants to further verify that the T-DNA sequence was inserted into the L. indica genome. Clean readings of 12.36–13.43 Gb, or more than 30× coverage of the L. indica reference genome, were left after the adaptor sequence, and low-quality reads were removed (Supplementary Table S7). The Q30 values of all the libraries were >85%, confirming the accuracy of our data (Supplementary Table S7).

All clean reads were mapped to the sequence of the pBI121-GFP vector and the Lagerstroemia Jinhuang reference genome to determine potential insertion sites of exogenous fragments. The potential integration sites of transgenic events were identified based on junction reads, with one end mapped to the vector sequence and the other to the reference genome. Two junction reads in the #OE1 genome, and one in the #OE2 genome were obtained after sequence alignment. However, T-DNA sequences were not detected in the wild-type genome. These findings showed that the #OE2 genome had a single T-DNA insertion site, while the #OE1 genome had two (Figure 3). According to the physical positions of the junction reads for #OE1, one insertion site is at 7,591,779 bp on chromosome 21, where one gene was found within 10 kb of the insertion site, and the other was at 16,383,307  bp on chromosome 22, where no functional gene was found (Figure 3, Supplementary Table S8). The insertion site was located at 8,786,870 bp on chromosome 13 based on the physical positions of the junction readings of #OE2; however, no functional gene was detected (Figure 3, Supplementary Table S8).

3.4. LiMYB75 Belonging to the R2R3 Subgroup Is a Nuclear Protein

Our earlier research found a novel MYB TF called LiMYB75 in petal transcriptome data unpublished. It was thought to be involved in petal development and to improve the effectiveness of callus regeneration. LiMYB75 (GenBank ID: OP566496) was consequently cloned from the strain Lagerstroemia Dynamite, and its open reading frame (ORF) was 1962 bp in length. The protein sequence of LiMYB75 was 654 amino acids in length. The theoretical pI of LiMYB75’s protein sequence was 7.76, and its theoretical molecular weight was 73.27 KDa. It appears to be a stable hydrophilic protein based on the instability index and grand average of hydropathicity, which were 38.29 and −0.51, respectively. The LiMYB75 protein has a secondary structure that is made up of a 40.58% α-helix, a 43.19% random coil, an 11.49% extended strand, and a 4.75% β-turn (Supplementary Figure S2).

We performed phylogenetic analysis and multiple sequence alignment based on the 13 MYB TF sequences from nine species. LiMYB75, Arabidopsis thaliana (AT3G52250.1), and Eucalyptus grandis (Eucgr.G02815.1.p) developed a branch, as depicted in Figure 4a, and LiMYB75 is the closest relative of Arabidopsis thaliana. LiMYB75 belonged to the R2R3 subgroup family and had two C-terminal SANT/MYB domains, according to multiple sequence alignment (Figure 4b).

We assumed that the nucleus was where the LiMYB75 gene was located. We fused LiMYB75 with eGFP (35S::LiMYB75-eGFP) to study its subcellular distribution to support the hypothesis. As shown in Figure 5a, confocal images showed that in a transient expression assay, the LiMYB75-eGFP fusion protein was localized exclusively in the nuclei of Nicotiana tabacum leaf epidermal cells. The eGFP protein was located in the nucleus and cytoplasm as a control. LiMYB75′s nuclear localization is consistent with its predicted role as a TF.

QPCR was used to examine the transcript abundance of LiMYB75 in various tissues and the stages of callus induction and regeneration to measure the expression levels of this gene. All four tissues had the LiMYB75 transcript, with the lead having the highest levels (Figure 5b). As shown in Supplementary Figure S3, callus activation (0 days after cutting, DAC), callus formation (5 DAC), callus expansion (10 DAC), and shoot regeneration (35 DAC) are the four main stages of the development process of leaf callus induction and regeneration. The relative expression of LiMYB75 was significantly lower during callus formation and callus expansion than during callus activation. However, the LiMYB75 transcript was significantly higher than that in callus activation (Figure 5c). These findings suggest that gene LiMYB75 might be involved in L. indica callus induction and regeneration.

3.5. LiMYB75 Promotes Callus Regeneration Efficiency in L. indica

We hypothesized that the LiMYB75 gene aided the callus induction and regeneration processes. The ORF of LiMYB75 was transformed into wild-type crape myrtle under the control of a 2× CaMV 35S promoter to support the hypothesis. After 14 days of infection, there was no significant difference in the callus induction rates between the control mock and LiMYB75-overexpressing whole-leaf explants across three concentrations of Agrobacterium tumefaciens (Figure 6a–c). Surprisingly, the frequency of callus regeneration of the LiMYB75-overexpressing whole-leaf explants increased from 23.33% (OD₆₀₀ = 0.6) to 56.67% (OD₆₀₀ = 0.8) (Figure 6d–f). LiMYB75-overexpression levels also significantly increased the number of callus regeneration shoots by 8–9 plants (OD₆₀₀ = 0.8) per piece (Figure 6e), which was much greater than the 1–3 shoots per callus regeneration with leaf infection simply using the empty vector (Figure 6d,g). Moreover, the phenotypes of the three LiMYB75-overexpressing lines did not significantly differ from the wild-type line (Supplementary Figure S4). These findings indicate that the LiMYB75 gene might be important for callus regeneration but does not affect L. indica development.

Eight genes involved in callus formation were assessed for expression levels 48 h after Agrobacterium infection (OD₆₀₀ = 0.8) to further investigate the role of LiMYB75 in callus regeneration. The expression levels of two genes, STM and LATERAL ORGAN BOUNDARIES DOMAIN 18 (LBD18), were significantly lower in the LiMYB75-overexpressing whole-leaf explants than in the wild type. At the same time, LiMYB75-overexpressing whole-leaf explants showed a considerable increase in the expression of four TEOSINTEBRANCHED1/CYCLOIDEA/PROLIFERATINGCELLFACTOR (TCP) genes: TCP2, TCP3, TCP4-1, and TCP4-2 (Figure 6h). Moreover, the expression levels of two ARABIDOPSIS RESPONSE REGULATOR (ARR) genes, ARR1 and ARR12, did not alter significantly between wild-type and LiMYB75-overexpressing whole-leaf explants, indicating that these ARR genes might not play a key role in LiMYB75-induced callus regeneration (Figure 6h).

4. Discussion

Crape myrtle is a valuable woody ornamental shrub that generates about $300 million annually in China. Genetic transformation technology is now one of the key technologies for the functional testing of plant genes and molecular breeding with the rapid development of biotechnology. Plant genetic transformation frequently uses the Agrobacterium-mediated technique [26]. We first described callus induction in this study and established a successful genetic transformation system in L. indica.

4.1. The Cytokinin/Auxin Ratio Is Important for Callus Induction and Regeneration

The callus induction rate is the first step in this experiment to build an effective genetic transformation system using isolated leaves. Plant growth regulators are crucial for the initiation, maintenance, and maturation of regeneration, as well as the introduction of calli [27]. In most species investigated, cytokinin and auxin are crucial plant growth regulators associated with the onset and growth of callus [28]. The most often utilized auxins and cytokinins to induce calli are 2,4-D, NAA, BA, and kinetin, which are artificial plant growth regulators [6,29,30]. In many woody plants, including Catalpa bungei [8], a combination of BA and NAA has been reported to effectively induce and enhance callus formation. Similar results were also observed in the present study (Figure 1).

According to earlier research, a balanced cytokinin to auxin ratio promotes callus induction, whereas high and low ratios promote shoot and root formation, respectively [31]. The callus induction rate appeared to be a single-peak curve with an increase in the ratio of BA to NAA concentrations in Paeonia ostia, peaking at 3:1 [32]. In Sapindus mukorossi [33] and Paeonia lactiflora [34], callus induction and regeneration frequencies also varied with the concentration ratio of cytokinin to auxin.

According to the results of the present study, CIM3 had the highest callus induction rates (70.33%) and the highest cytokinin to auxin ratio (1.6:1) (0.8 mg·L⁻¹ 6-BA/0.5 mg·L⁻¹ NAA, Supplementary Table S1). CRM1 had the highest rates of callus regeneration (44.33%), and the ratio of cytokinin to auxin was 10:1 (1.0 mg·L⁻¹ 6-BA/0.1 mg·L⁻¹ NAA, Supplementary Table S2). Cytokinin and auxin concentrations also had an impact on callus regeneration frequency. As the concentration increased, the callus regeneration rate significantly decreased (Figure 1l). Therefore, we developed efficient protocols for L. indica callus induction and regeneration by regulating the cytokinin to auxin ratio.

4.2. The Established Agrobacterium-Mediated Transformation System Provides a Powerful Tool for Functional Gene Analysis in L. indica

Genetic transformation efficiency is the most crucial factor in genetic transformation protocols. The frequency of transformation in woody plants varies, ranging from 4% obtained by Agrobacterium tumefaciens-mediated transformation with embryogenic calli of Q. suber [35] to 7% obtained by Agrobacterium tumefaciens-mediated transformation with internodal stem segments from P. alba plantlets [36] and 32.2% obtained by Agrobacterium-mediated transformation with leaf explants in the P. alba × P. glandulosa [37], while 92.31% were obtained by Agrobacterium-mediated transformation with embryogenic callus explants in C. bungei [8].

The main benefit of GFP as a reporter gene of transgenic plants is that it makes chemical staining very convenient as a reporter molecule for direct imaging detection. Yang et al. [38] used GFP as a reporter gene to detect foreign genes in transgenic tobacco. Ma et al. [39] used the Agrobacterium tumefaciens-mediated method to insert the green fluorescent protein gene into rice (Oryza sativa) callus, and then transgenic rice calli and seedling roots exhibited high-intensity green fluorescence. In the present study, the total number of positive transgenic lines identified by GFP was consistent with those identified by PCR (Figure 2). These results also demonstrated the validity of using GFP as a reporter gene to identify transgenic-positive crape myrtle plants.

We established the leaf callus’s Agrobacterium-mediated genetic transformation system and improved the positive frequency to 90% in L. indica (Figure 2). Bacterial concentration, co-cultivation time, and infection time were three variables evaluated in this method to see how they affected transformation efficiency (Supplementary Tables S3–S5). Fluorescence imager research revealed that different bacterial concentrations significantly impacted the effectiveness of leaf transformation with an increase in bacterial concentration. Additionally, it was discovered that the leaves had high fluorescence intensity when the bacterial liquid concentration was 6 × 10⁸ cells·mL⁻¹, which might be the reason for their high genetic transformation efficiency (Supplementary Figure S1).

4.3. LiMYB75 Is a Novel Regulator to Modulate In Vitro Callus Formation in L. indica

Plant regeneration and genetic transformation are vital and imperative for genetically engineered plants. However, the establishment of genetic transformation has been limited to selecting well-studied woody species because of the low rate of plant regeneration [40]. Over the past few decades, hormone-induced callus formation from pericycle and pericycle-like cells has indicated a specific cell fate transition in which somatic cells acquire pluripotency [23]. A variety of developmental regulators, including WUSCHEL (WUS), SHOOT MERISTEMLESS, LEAFY COTYLEDON 1, and SOMATIC EMBRYOGENESIS RECEPTOR KINASE1, can control and switch the fate of plant cells [41,42,43,44].

Recent research has demonstrated that the ability of callus formation under in vitro culture conditions might be improved by the overexpression of developmental regulators [15]. High rates of monocot species transformation were induced by overexpressing the two developmental regulators, BABY BOOM and WUS2 [42]. The transformation effectiveness and transformable genotypes in wheat (Triticum aestivum) and rice (Oryza sativa) were significantly enhanced by overexpressing a fusion protein combining wheat (Triticum aestivum) GROWTH-REGULATING FACTOR 4 (GRF4) and its cofactor GRF-INTERACTING FACTOR 1 [45]. Therefore, it is crucial to identify novel developmental regulators to improve the callus regeneration ability and increase the variety of transformable species.

Notably, in Arabidopsis, two MYB TFs, MYB94, and MYB96, were determined to have negative roles in callus formation by suppressing LATERAL ORGAN BOUNDARIES DOMAIN 29 expression by directly binding to the gene’s promoter [23]. Instead, the frequency of callus regeneration increased by 23.33%–56.67% in the current study, and the number of callus regeneration shoots also significantly increased up to 8–9 per piece in LiMYB75-overexpressing leaves. This was significantly higher than the 1–3 shoots per callus regeneration with the leaf infection only using the resuspension solution (Figure 6d–g). Interestingly, LiMYB75 also impacted the expression level of LiLBD18, suggesting that LiMYB75 might share a similar regulatory mechanism with AtMYB94/MYB96. Moreover, LiMYB75 impacted LiSTM expression levels under in vitro culture. However, more research is still needed to determine the molecular regulation mechanism through which LiMYB75 promotes callus formation.

5. Conclusions

In summary, we established an efficient and rapid callus induction and regeneration system using whole leaf explants of L. indica. On this basis, a new Agrobacterium-mediated genetic transformation system was developed. Overexpression of a novel MYB TF LiMYB75 gene, which using our petal transcriptome data unpublished significantly enhanced the regeneration ability of L. indica callus. The results indicate that LiMYB75 increased the callus regeneration efficiency by regulating the RNA levels of genes involved in plant hormone metabolism in crape myrtle. Which can be used as a candidate gene for further study of plant regeneration.