Improvement of plant regeneration and Agrobacterium-mediated genetic transformation of Stylosanthes guianensis Mejoramiento de la regeneración de plantas y transformación genética mediada por Agrobacterium en Stylosanthes guianensis

As a pioneer tropical pasture legume, stylo (Stylosanthes guianensis) is well adapted to growth-limiting factors in acid soils. Considering the importance of stylo, there is a need to improve Agrobacterium-mediated genetic transformation to enable development of elite cultivars. In this study, S. guianensis cv. RY5 was used to systematically optimize Agrobacteriummediated transformation based on its plant regeneration. Results showed that Murashige and Skoog (MS) medium containing 0.2 mg/L 2,4-dichlorophenoxyacetic acid (2,4-D) and 2 mg/L 6-benzylaminopurine (6-BA) was the optimal callus induction medium. MS medium supplemented with 2 mg/L 6-BA was suitable for shoot regeneration from cotyledonderived calluses, and 0.5 mg/L indole-3-acetic acid (IAA) and 0.5 mg/L indole-3-butyric acid (IBA) applications were beneficial for rooting. The highest transformation efficiency (67%) was obtained at an Agrobacterium concentration of optical density = 0.6 combined with an infection time of 15 min and 3 days of co-cultivation. Furthermore, 200 mg/mL carbenicillin (Carb) and 0.6 mg/L Basta supplements were effective in eliminating excess bacterial growth and selecting transgenic plants, respectively. Subsequent polymerase chain reaction (PCR) analysis confirmed that the β-glucuronidase (GUS) and BAR genes were successfully integrated into the stylo genome. Wider testing of this improved protocol as a means of enhancing genetic improvement and gene function analysis of stylo seems warranted.


Introduction
Species of Stylosanthes, referred to as stylos, are important pasture legumes and widely distributed across the tropical and subtropical areas of the Americas, Africa and Asia. The genus includes more than 40 species, such as S. guianensis (common stylo), S. hamata (Caribbean stylo), S. scabra (shrubby stylo), S. viscosa (sticky stylo) and S. seabrana (Caatinga stylo) (Chandra 2013). Stylo is regarded as the 'tropical alfalfa' due to its multiple uses in tropical agricultural systems, including sown pastures, feed for livestock and improvement of soil properties, natural grassland and orchard mulching. Considering the importance of this pasture legume, it was introduced to China from Colombia and Australia in the 1960s (Tang et al. 2009). Stylosanthes guianensis cv. RY5, a single plant isolated from the population of its parent CIAT 184, exhibits the traits of earlier flowering, higher seed production, anthracnose resistance and cold tolerance compared with its parent, and is considered an elite stylo cultivar that is widely grown in South China (Liu et al. 2001;Tang et al. 2009).
With the development of technology, stylo has attracted large research interest in many aspects, including ecological studies (Vander Stappen et al. 1999;Sawkins et al. 2001), physiological and biochemical analyses (Zhou et al. 2005;Sun et al. 2014;Chen et al. 2015;Wang et al. 2017;Liu et al. 2018) and genetic diversity studies (Vander Stappen et al. 2000;Jiang et al. 2005;Ding et al. 2015). For example, stylo has been recognized as a pioneer tropical pasture with extensive adaptation to growth-limiting factors in acidic infertile soils, such as phosphorus (P) deficiency and aluminum (Al) and manganese (Mn) toxicity (Chen et al. 2015;Jiang et al. 2018;Liu et al. 2019). It has been demonstrated that malate synthesis in roots and exudation from roots of S. guianensis are the critical tolerance strategies for Al and Mn toxicity in acidic soils (Sun et al. 2014;Chen et al. 2015). Recently, a rootassociated purple acid phosphatase, SgPAP23, was characterized as a primary mediator of extracellular phytate-P utilization in stylo (Liu et al. 2018).
Although much work has been conducted as described above, extensive studies of stylo are restricted by some challenges, such as limited available germplasm resources, narrow genetic variability, lack of genome information and susceptibility to chilling stress as well as anthracnose disease (Chandra 2013;Wang et al. 2017). To solve these issues, genetic improvement through biotechnological techniques is one of the most effective and useful approaches that will help to improve the desirable traits of plant adaptation to biotic or abiotic stresses, contributing to agricultural production (Mittler and Blumwald 2010).
Over the past decades, significant development in efficient genetic transformation methods has been made in many plants. Among them, Agrobacterium-mediated transformation based on plant tissue culture is the most useful method for genetic transformation in various crops, such as rice (Oryza sativa) (Ozawa 2012), maize (Zea mays) (Ishida et al. 2007), sorghum (Sorghum bicolor) (Gurel et al. 2012), soybean (Glycine max) (Wang et al. 2009) and alfalfa (Medicago sativa) (Tesfaye et al. 2001). Although great efforts have been made to develop an Agrobacterium-mediated genetic transformation system for the genus Stylosanthes, such as S. hamata (Iji et al. 1995;Kumar and Chandra 2010), S. humilis (Manners and Way 1989), S. seabrana (Kumar and Chandra 2009; and S. guianensis (Quecini et al. 2006;Wang et al. 2008;Yuan et al. 2011;Bao et al. 2016), the genetic transformation procedure has not been adequately improved. Furthermore, combination analyses of Agrobacterium-mediated transformation based on plant regeneration systems are rarely performed on stylo. Only a few studies have successfully developed transgenic stylo plants thus far (Wang et al. 2008;Bao et al. 2016;Chen et al. 2016). Therefore, optimization of plant regeneration and an Agrobacterium-mediated transformation system are critical for the genetic improvement of stylo cultivars and basic research.
The efficiency of Agrobacterium-mediated genetic transformation is influenced by many factors, such as plant species and explants, the process of callus induction and differentiation, and Agrobacterium infection (Cheng et al. 2004;Trivellini et al. 2015). In this regard, the S. guianensis cultivar RY5 was used to optimize the plant regeneration system in this study, including callus induction, shoot regeneration and root induction. Subsequently, factors influencing the Agrobacteriummediated transformation system, including bacteriostatic antibiotics, Agrobacterium concentrations, infection and co-cultivation duration and Basta ® concentrations, were analyzed according to β-glucuronidase (GUS) staining analysis. Furthermore, polymerase chain reaction (PCR) analysis was performed to examine the putative transformed stylo plants.

Plant material and explant preparation
In this study, the stylo cultivar RY5 was used. Seeds of RY5 were soaked in hot water at 80 ℃ for 3 min, followed by surface sterilization using 75% ethanol (v/v) and 2% sodium hypochlorite (v/v), and were subsequently germinated in Petri dishes containing basal Murashige and Skoog (MS) solid medium containing 30 g/L sucrose and 7 g/L agar (Liu et al. 2018). The pH of the medium was adjusted to 5.8 using 1 mol/L KOH or H2SO4. After 6-8 days germination, explants of euphylla, cotyledon and hypocotyl, approximately 10 mm in length, were used for callus induction.

Callus induction and plant regeneration
For embryogenic callus induction, the different explants were inoculated on MS solid medium containing different concentrations of 2,4-dichlorophenoxyacetic acid (2,4-D) (0-0.5 mg/L) and 6-benzylaminopurine (6-BA) (0-2 mg/L) supplemented alone or in combination. All treatments had 4 biological replicates, and each replicate included 10 explants. The explants were sub-cultured every 10 days on fresh MS medium. After 30-40 days of sub-culture, a granular callus approximately 5 mm in size was recorded for callus induction.
For shoot regeneration, the embryogenic callus from the explants of the cotyledon was selected and transferred onto MS medium supplemented with different concentrations of 6-BA (0-4 mg/L) for shoot regeneration. The cultures were incubated at 28 ℃ under an irradiance of 80 mol/m/s (16 h of light:8 h of dark). All treatments had 3 biological replicates and each replicate included 10 calluses. After 8-10 weeks of culture, the shoots with 2-4 leaves that elongated from the callus were recorded.

Carbenicillin (Carb) concentration determination
To detect the effects of Carb on the growth of Agrobacterium tumefaciens strain EHA105, A. tumefaciens with an initial absorbance of 0.02 at 600 nm was grown in YEB liquid medium (5 g/L tryptone, 1 g/L yeast extract, 5 g/L beef extract, 5 g/L sucrose, 0.5 g/L MgSO4 . 7H2O) supplied with different concentrations (0-400 mg/mL) of Carb at 28 ℃. After 48 h of growth, the absorbance of A. tumefaciens was recorded. To detect the effects of Carb on callus induction, cotyledon explants were incubated in the optimal callus induction MS solid medium (MS medium supplied with 0.5 mg/L 2,4-D and 2.0 mg/L 6-BA) containing different concentrations (0-1,000 mg/mL) of Carb. The numbers of embryogenic calluses were recorded after 30-40 days of culture. All experiments had 5 biological replicates and each replicate included 100 explants.

Infection and co-culture assay
The binary vector pCAMBIA3301, containing the GUS reporter gene driven by the cauliflower mosaic virus 35S (CaMV 35S) promoter, was used as the transformation vector in this study. The BAR gene was used as a selection marker under the control of the CaMV 35S promoter. Agrobacterium tumefaciens strain EHA105 harboring pCAMBIA3301 was grown in 50 mL of YEB liquid medium containing 50 mg/L kanamycin and 50 mg/L rifampicin at 28 ℃ for 48 h. After centrifugation at 5,000 rpm for 3 min, the bacterial cells were harvested and further re-suspended in YEB liquid medium.
The cotyledon explants were incubated in 50 mL of Agrobacterium suspension cells harboring the pCAMBIA3301 vector with different absorbances at 600 nm (from 0.2 to 1.0), which contained 50 mg/L acetosyringone (AS). The mixtures were then cultured for 5-25 min with gentle shaking. Subsequently, the Agrobacterium-infected explants were blotted dry on sterile filter paper and were then transferred onto cocultivation medium (MS medium containing 0.5 mg/L 2,4-D and 2.0 mg/L 6-BA) supplied with 50 mg/L AS for 1-5 days. After that, the explants were transferred onto the callus induction MS medium (MS medium supplied with 0.5 mg/L 2,4-D and 2.0 mg/L 6-BA) containing 200 mg/mL Carb and 0.2-0.8 mg/L Basta ® . After 30-40 days of culture, the transformation efficiency was determined based on GUS staining analysis as described by Qin et al. (2012). Briefly, the putative transformed callus was incubated in the GUS staining solution containing 0.2 mol/L Na2HPO4-NaH2PO4 and 1 mmol/L 5-bromo-4chloro-3-indolyl--D-glucuronic acid (pH 7.0) at 37 ℃ for 12 h, followed by washing in 70% ethanol (v/v). All experiments had 3 biological replicates and each replicate included more than 40 explants.

Putative transformed plant generation
Under the optimal procedure of infection, co-culture and selection, the basta-resistant callus was transferred onto selection medium (MS medium containing 2.0 mg/L 6-BA, 0.6 mg/L Basta ® and 200 mg/mL Carb) for shoot regeneration for 8-10 weeks. Subsequently, the regenerated plantlets were transferred onto the optimal rooting medium (MS medium containing 0.5 mg/L IAA, 0.5 mg/L IBA, 20 g/L sucrose and 0.8 g/L activated charcoal; pH 5.8) for 3 weeks of root growth. All of these cultures were incubated at 28 ℃ under an irradiance of 80 mol/m/s (16 h of light:8 h of dark). Finally, the putative transformed plantlets with sufficient root systems were transplanted and cultured in a 1:1 soil:sand mix for normal growth in the greenhouse.

Transgenic plant determination
For polymerase chain reaction (PCR), total genomic DNA was isolated from leaves of putative transgenic stylo plants using the cetyltrimethyl ammonium bromide (CTAB) extraction method modified by Ding et al. (2015).

Statistical analysis
The callus induction rate (%) was calculated as the ratio of the numbers of induced calluses and the numbers of explants used. Shoot regeneration efficiency (%) was calculated as the ratio of the numbers of regenerated shoots and the numbers of calluses used. Rooting efficiency (%) was calculated as the ratio of the numbers of regenerated roots and the numbers of shoot buds used. The frequency of GUS-stained calluses (%) was calculated as the ratio of the numbers of GUS-stained calluses and the numbers of calluses used. All statistical analyses were performed by one-way ANOVA using the SPSS program (SPSS Institute, USA, v. 13.0).

Optimization of the plant regeneration system of stylo
In this study, 3 explants, hypocotyl, cotyledon and euphylla were used to investigate the effects of different hormone supplements on embryogenic callus induction. The results (Table 1) showed that after 30-40 days of culture, callus induction rates were approximately 10% for the 3 tested explants under control treatment (without hormone addition), in which the callus displayed a pale yellow or white and watery texture. Callus induction rates were significantly increased by 2,4-D and 6-BA treatment, alone or in combination. The callus induction rates were 100, 87.5-90 and 95-100% for hypocotyl, cotyledon and euphylla, respectively, under 2,4-D treatment. Treatment with 6-BA enhanced the callus induction rates with more than 500, 500 and 380% increases in hypocotyl, cotyledon and euphylla, respectively, as compared with their respective controls. Furthermore, 2.0 mg/L 6-BA in combination with 0.2-0.5 mg/L 2,4-D supplement significantly increased the callus induction rates (approximately 100%), especially for cotyledon and euphylla, compared with their respective controls. However, no significant differences in callus induction were observed among the 3 tested explants under the same hormone treatment. Therefore, the optimal callus induction medium was MS medium containing 2 mg/L 6-BA and 0.2 mg/L 2,4-D, in which granular callus displayed a light green and compact texture (Table 1, Figures 1a-1d).
The light green and compact texture granular callus induced from the cotyledon became green and then produced a lot of tubers after cultured on the medium containing 6-BA for 50-70 days, whereas control calluses which appeared white, sticky and watery were unable to generate shoots on the differentiation medium (Figures 1e-1i). The effects of different levels of 6-BA on shoot differentiation from the cotyledon-derived callus of cv. RY5 were demonstrated as shoot regeneration rates and were significantly increased with increasing 6-BA concentrations; highest shoot regeneration rates (approximately 80.0%; P<0.05) were observed under 2 and 4 mg/L 6-BA supplements (Figure 2a). Subsequent rooting efficiency from the regenerated shoot after 3 weeks of root induction was higher in M2 and M3 media with different hormone supplements than in M1 medium without hormone addition (Figure 2b). The highest rooting efficiency (100%) was observed in M2 medium containing 0.5 mg/L IAA and 0.5 mg/L IBA (Figure 1j, Figure 2b).

Improvement of Agrobacterium-mediated transformation
Growth of Agrobacterium was significantly inhibited by increasing Carb concentrations compared with the control without Carb addition (Figure 3a) with decreases of 69.6 and 98.8% under 50 and 200 mg/mL Carb treatments, respectively, compared with the control. Callus induction rate was significantly (P<0.05) decreased when Carb concentrations were higher than 600 mg/mL with decreases of 29.2 and 66.6% under 800 and 1,000 mg/mL Carb treatments, respectively, compared with the control (Figure 3b).  Subsequently, the effects of Agrobacterium concentration, infection time and co-cultivation duration on callus transformation efficiency were evaluated based on GUS staining analysis (Table 2). A callus with blue coloration was considered a positive transformed callus after GUS staining. Highest transformation efficiency (67.5%) was observed at an Agrobacterium concentration of optical density (OD600) of 0.6 and transformation efficiency decreased when the Agrobacterium concentration (OD600) exceeded 0.6. Similarly, transformation efficiency increased with increasing Agrobacterium infection time from 5 to 20 min, and the highest transformation efficiency (59.3%; P<0.05) was achieved with infection times of 15 and 20 min. However, transformation efficiency was decreased significantly (by 93.8%) after 25 min of infection compared with that after 15 min of infection. The transformation efficiency was also affected by different co-cultivation durations. The maximum transformation efficiency was obtained under 3 days of co-cultivation, when it reached 69.4%.
Additionally, the effects of Basta ® on transformed callus selection were also examined and showed that transformation efficiencies varied among different Basta ® treatments (Table 2). Transformation efficiency increased with increasing Basta ® concentrations, peaking at 0.6 mg/L at a value of 62.1%, but then decreased at Basta ® concentrations greater than 0.6 mg/L. Therefore, 0.6 mg/L Basta ® was chosen for appropriate callus selection.

Transformed plant production
Subsequently, cotyledon explants were used for transformation under the optimized procedure of an Agrobacterium concentration (OD600) of 0.6 with 15 min of Agrobacterium infection and 3 days of co-cultivation ( Figure 4). The infected explants were transferred onto MS medium containing 0.2 mg/L 2,4-D, 2.0 mg/L 6-BA and 200 mg/L Carb for 1 week of recovery growth. After that, the explants were transferred onto callus induction medium containing 0.2 mg/L 2,4-D, 2.0 mg/L 6-BA, 0.6 mg/L Basta ® and 200 mg/mL Carb for 30-40 days. A mass of basta-resistant calluses was generated and then transferred onto shoot regeneration medium containing 2.0 mg/L 6-BA, 0.6 mg/L Basta ® and 200 mg/mL Carb for 8-10 weeks. The regenerated plantlets were cut and further transferred onto rooting medium containing 0.5 mg/L IAA and 0.5 mg/L IBA for 3 weeks. Finally, the putative transformed plantlets with a sufficiently vigorous root system were transplanted and cultured into a soil and sand mixture (1:1) in a greenhouse for normal growth (Figure 4, Figure 1k).

Determination of putative transgenic stylo plants
As shown in Figure 5, clear GUS staining was observed in callus, leaf, root and stem of transgenic plants, suggesting that GUS gene was successfully expressed in different tissues of stylo plants. In this study, a total of 850 explants were used for Agrobacterium-mediated transformation according to the optimized procedure, and 51 putative transformed lines were obtained. To further verify the integration of foreign genes into the stylo genome, genomic DNA extracted from leaves of 15 putative transformed lines was used for PCR analysis. The results showed that 426 and 533 bp of the expected DNA fragments of the respective BAR and GUS genes were successfully amplified from 9 out of 15 randomly selected lines ( Figures  6a-6c), suggesting the successful integration of foreign genes into the stylo genome. Furthermore, the expression of the GUS gene in the leaves of the 15 tested putative transgenic stylo lines was also detected using RT-PCR analysis. The results showed that the foreign GUS gene was stably expressed in 9 out of 15 tested transgenic stylo lines (Figures 6d and 6e). In total, 33 out of 51 putative lines were proved to be positive transgenic lines, resulting in a transformation efficiency of 3.88%.

Discussion
The Agrobacterium-mediated genetic transformation approach is widely used for the development of transgenic plants (Cheng et al. 2004). Despite the importance of the Stylosanthes genus for pasture, only a few transgenic plants have been generated and used for gene function analysis of Stylosanthes species (Wang et al. 2008;Bao et al. 2016), suggesting that the existing protocols might not be efficient for the development of transgenic stylo plants. Therefore, it is essential to improve the Agrobacteriummediated genetic transformation procedure for stylo based on its plant regeneration. This study has shown that MS medium containing 0.2 mg/L 2,4-D and 2 mg/L 6-BA was the optimal callus induction medium, especially for cotyledon and euphylla explants ( Table 1). As growth regulators, 2,4-D, 6-BA and NAA have been generally used for callus induction of stylo (Wang et al. 2008;Yuan et al. 2011;Bao et al. 2016). Yuan et al. (2011) found that the optimal condition for callus induction from cotyledons of S. guianensis cv. Reyan 2 was MS medium containing 3.0 mg/L 2,4-D with highest rates of callus induction reaching 74%, which is lower than that measured in this study. On the other hand, Godwin et al. (1987) showed the optimal callus induction medium for S. scabra leaves was MS medium supplemented with both 2,4-D and 6-BAP. These variations in appropriate hormone application can be attributed to different stylo species, cultivars or genotypes. Similar variations have also been observed in other plant species. For instance, a combination of 0.5 mg/L 2,4-D and 2.0 mg/L 6-BA resulted in better callus induction from Vitis vinifera flowers (Dai et al. 2015), while 2.0-3.0 mg/L 2,4-D was the optimal concentration for promoting callus induction in leaves of Urochloa (Yaguinuma et al. 2018).
In addition to callus induction, the differentiation of shoots from the callus is also essential for plant regeneration. Shoot organogenesis induced from the callus can be improved by application of appropriate hormone (Cheng et al. 2004;Yaguinuma et al. 2018). In this study, although shoot differentiation from the cotyledon-derived callus of cv. RY5 was observed in hormone-free MS medium, the percentage of shoot regeneration was only 3.3% (Figure 2a). Our findings that MS medium with 2 mg/L 6-BA applications was most efficient for shoot regeneration (Figure 2a), producing a shoot regeneration rate of approximately 80%, is consistent with earlier findings that 2-4 mg/L 6-BA applications significantly increased the regeneration of shoots from the cotyledon-derived callus of cv. Reyan 2 (Yuan et al. 2011). Similarly, a high frequency of shoot regeneration was observed from the leaf-derived callus of S. scabra on MS medium containing 2.0 mg/L BAP (Godwin et al. 1987) and from the cotyledon-derived callus of S. macrocephala on MS medium supplemented with 0.1 mg/L NAA and 0.4 mg/L BAP (Vieira et al. 1990). However, other factors, such as callus derived from different types of explants and callus states (e.g. color and texture) at different hormone treatments, may influence the regeneration capacity of shoots in stylo, which merits further study.
The increases in rooting efficiency we observed with hormone applications support the increased rooting efficiencies of S. seabrana and S. hamata in response to the application of IAA and IBA alone or in combination compared with untreated controls (Kumar and Chandra 2010). MS medium containing NAA and IBA is beneficial for rooting of S. guianensis (Kelemu et al. 2005). In contrast, it has been reported that hormone-free medium is suitable for rooting of S. scabra and S. guianensis (Valarini et al. 1997;Quecini et al. 2006;Bao et al. 2016). We found that the appropriate rooting medium was MS medium supplemented with 0.5 mg/L IAA and 0.5 mg/L IBA, in which rooting efficiency was significantly higher than that of hormone-free MS medium (Figure 2b). Similar results have been found for other plants, such as white ash (Fraxinus americana) and grass pea (Lathyrus sativus), where IAA and IBA applications promoted rooting efficiency (Barpete et al. 2014;Palla and Pijut 2015). Taken together, the results indicate that stable plantlet regeneration was optimized in cv. RY5 using cotyledons as explants.
Carb is used for terminating rudimentary Agrobacterium growth at suitable doses after the co-culture period during T-DNA insertion into the plant chromosome (Li et al. 2015). In our study, a dose-response experiment showed that 200 mg/mL Carb efficiently inhibited excess bacterial growth but did not affect callus induction. Optimal Carb concentrations that inhibit excess Agrobacterium growth range from 200 to 1,000 mg/L (Figure 3). For example, it has been demonstrated that 200, 500 and 1,000 mg/L Carb are efficient for the elimination of Agrobacterium in grapes (Vitis vinifera) (Dai et al. 2015), loblolly pine (Pinus taeda) (Tang et al. 2004) and tobacco (Nicotiana tabacum) (Nauerby et al. 1997), respectively, during bacterial infection.
Agrobacterium concentration, infection and cocultivation duration are important factors that affect transformation efficiency. An inappropriate concentration of Agrobacterium may produce toxic contamination events or may not be effective for the callus, and a long or short infection time may result in a low frequency of transformation (Cheng et al. 2004;Sharma et al. 2011;Trivellini et al. 2015). In this study, the highest transformation efficiency (67%) was obtained under the conditions of an Agrobacterium concentration of OD600 = 0.6, combined with an Agrobacterium infection time of 15 min and 3 days of co-cultivation based on GUS staining analysis (Table 2). This finding is similar to the Agrobacterium concentration of OD600 = 0.4-0.6, an infection time of 10 min and a co-cultivation duration of 2-3 days found to be optimal conditions for transformation of baby bamboo (Pogonatherum paniceum) (Li et al. 2015) and white ash (Fraxinus americana) (Palla and Pijut 2015). Furthermore, Agrobacterium infection for 2-10 min with 2-4 days of co-cultivation has been shown beneficial for increasing transformation efficiency in other stylo species, e.g. S. humilis and S. guianensis (Manners 1987;Manners and Way 1989;Bao et al. 2016;Chen et al. 2016). Basta ® is commonly used for the selection of transgenic plants that contain the herbicide-resistant selection marker BAR gene, which is resistant to Basta ® (Lin et al. 2009;Mayavan et al. 2015). Optimization of the Basta ® concentration for selection pressure is important for increasing transformation efficiency. In this study, 0.6 mg/L Basta ® was found to be suitable for transformed callus selection (Table 2), which is similar to the level recommended by Bao et al. (2016). It is noteworthy that Agrobacterium-mediated genetic transformation efficiency is influenced by many factors, such as phytohormone, antibiotic, growth medium and bacterial infection process (Nandakumar et al. 2004;Sharma et al. 2011). Therefore, in this study, the transformation efficiency of stylo may have been potentially affected by different hormone supplements and culture media in the regeneration and rooting processes, which need to be studied further.
The total of 33 positive transgenic plants we generated from 850 explants of the cotyledon, representing a transformation efficiency of 3.88%, was higher than that observed in S. humilis cv. Paterson, where only 0.3% of explants generated transgenic plants (Manners 1988). However, it is similar to the transformation efficiency of 3.47% in S. guianensis cv. Mineirão using microparticle bombardment reported by Quecini et al. (2006). Besides, only 25% of the putative transformed S. guianensis cv. RY5 lines generated from hypocotyl-derived calluses by Chen et al. (2016) proved to be positive transgenic plants, which is lower than the frequency of 64.7% obtained from cotyledons herein ( Figure 6). It appears that the transformation efficiency of stylo was improved using the optimal protocol in the current study.
In conclusion, this study has developed an optimal strategy for the Agrobacterium-mediated genetic transformation of cv. RY5 based on its plant regeneration system. This optimal procedure provides an appropriate platform to investigate the molecular mechanisms underlying stylo adaptation to environmental stresses. Furthermore, based on the optimized protocol, we can modify the specific traits of various stylo varieties by biotechnological and molecular approaches, which are beneficial for the genetic improvement of stylo.