​​An Agrobacterium rhizogenes Strain R1000-mediated Efficient Hairy Root Transformation Protocol for Common Bean

Common bean (Phaseolus vulgaris L.) is an important food and fresh vegetable legume in the world. Providing up to 70% of the total protein requirement, common bean is particularly important for people in Latin America and Africa where affordable animal proteins are scarce (Khandual, 2014). In addition to providing proteins, common bean also is an important source of iron, phosphorus, magnesium, manganese and to a lesser degree, zinc, copper and calcium (Stagnari et al., 2017). Thus, especially in developing countries, the excellent dietary microelement source can be complemented from common bean. Globally, the yield and quality of common bean are often constrained by diseases, pests and abiotic stresses such as drought, heat and high salinity (Partap and Godara, 2022; Priyanka et al., 2022; Xu et al., 2022). Conventional breeding is still the mainstream approach for genetic improvement of this crop, which, however, has many disadvantages such as long breeding cycle, labor intensity, low recombination rate of multiple genes and incompatibility of distant crosses (Asfaw and Blair, 2014). With the development of biotechnology, advances in breeding, functional research and targeted genetic modifications have become essential for studying common bean. Thus, efficient transformation systems are required to advance common bean research.
       
Genetic engineering is a powerful approach to accelerate crop improvement and overcomes the major limitations of traditional breeding (Mukeshimana et al., 2013; Varshney et al., 2012). Plant transgenic technology, as a means of genetic engineering, has shown its usefulness in breaking the linkage drag or species barrier in many crop taxa and become a common approach in crop breeding programs (Datta, 2013). Unfortunately, the genus phaseolus vulgaris that includes common bean has known to be recalcitrant to regenerate in vitro, imposing a bottleneck for efficient plant transformation (Malik and Saxena, 1991; Singh and Tiwari, 2012). To date, limited cases of success have been reported in common bean transformation efforts. Some authors reported direct gene transfer based on particle bombardment or electroporation has been documented, but these methods generally yield low transformation frequencies and high rates of chimeras and false positives (Bonfim et al., 2007; Vianna et al., 2004).
       
At present, Agrobacterium-mediated transformation is the method most frequently used for common bean transformation. Compared with Agrobacterium tumefaciens-mediated transformation of common bean, the Agrobacterium rhizogenes-mediated hairy root transformation system has a high transformation efficiency and short transformation period. The process can be completed within one month. Hairy root transformation technique is a commonly used alternative method for the generation of transgenic plants of legumes. Instead of using A.tumefaciens, hairy root transformation uses A.rhizogenes, a gram-negative soil bacterium belonging to the genus Rhizobia that infects most dicotyledonous plants, some monocotyledonous plants and a few gymnosperms (Veena and Taylor, 2017). A. rhizosphere has a root-inciting (Ri) plasmid, which can induce hairy roots formation at the injured site of the infected plant. During this process, the T-DNA in the Ri plasmid can be randomly transferred, integrated and inserted into the genome of the plant cells, thereby generating composite transgenic plants (Chilton et al., 1982). Usually, hairy roots are non-chimeric, because they are derived from single cells and each hairy root consists of uniformly transformed cells. Agrobacterium-mediated hairy root technique has many advantages such as low cost, high reproducibility, high transformation rate and wide range of receptor plants.
       
Successful legumes studies have been performed using Agrobacterium rhizogenes. Aggarwal et al., (2018) demonstrated an efficient, high-throughput and genotype-independent method of root transformation in chickpea using A.rhizogenes K599. The efficiency of transformation was higher (73.33%) using immersion method than the cotyledonary injection method (38.66%). Estrada-Navarrete et al., (2007) described a fast, reproducible and efficient common bean root transformation protocol with the A. rhizogenes strain K599. This method has worked successfully in generating high efficiency (70-90%) of hairy roots in multiple genotypes and landraces of common bean and other phaseolus vulgaris subspecies. To date, the A.rhizogenes-mediated hairy root transformation system has been applied in gene functional analysis, promoter analysis and plant stress response studies in common bean (Carrasco-Castilla et al., 2018; Estrada-Navarrete et al., 2007; Nanjareddy et al., 2017; Yao et al., 2014). However, in most reported cases the A. rhizogenes strain used was K599, which was assumed to be critical by some researchers (Estrada-Navarrete et al., 2007).
       
In this study, to overcome the limitation of A.rhizogenes strain selectivity in current common bean hairy root transformations, we established an alternative, A. rhizogenes strain R1000-mediated hairy root transformation protocol for common bean, which showed up to 60% of transgenic rate in a wide range of receptor genotypes.

The experiment was conducted from October 2020 to March 2021 in the Lab of Plant Quality and Safety Biology, College of Life Sciences, China Jiliang University, Hangzhou, China.
 
Plant materials
 
Twenty common bean genotypes provided by Zhejiang Academy of Agricultural Sciences were used for A.rhizogenes strain R1000-mediated transformation. These genotypes include LP089, P331024017, P331024004, LP093, LP074, LP047, LP096, LP020, LP011, LP053, 289332211, 2018334327, Hongjinyuanjia, Honghuaqinjia, Cuiyu, Yudouwang, Yinmanjia, Chunqiujiadouwang, Huazihonghuaqingjia and Pan’an common bean.
 
Seeds disinfection and germination
 
The seeds were rinsed twice with sterile water before being surface-sterilized with 70% ethanol for 30 s. The seeds were thoroughly washed with sterile water. Disinfected seeds were placed on plates containing sterile wet filter paper for germination. The plates were kept in 23 to 26°C temperatures in the dark. After germination, the germinated seedlings were sown into sterilized growing substrate (a mixture of 3:1 peat and vermiculite) and were cultured in a plant growth room of 16 h light/8 h dark cycle at 25±2°C under a 60% relative humidity condition with an intensity 600 μmol m^-1s^-1. Three days after sowing, seedlings with green cotyledons unfolded were selected for hairy roots induction.
 
Plasmid
 
The binary plasmid pBI121 with the reporter gene GUS (pBI121-GUS) or eGFP (pBI121-eGFP) driven by the constitutive promoter cauliflower mosaic virus (CaMV) 35S were used for transformation and rapid identification of the positive hairy roots.
 
Agrobacterium strains and inoculum preparation
 
The plasmid vector was transferred into A.rhizogenes strain K599 and R1000 via electroporation. The strains were grown at 28°C in solidified Luria Bertani (LB) medium with appropriate antibiotics (kanamycin for R1000, kanamycin and rifampin for K599) to produce a lawn in the plate. The plates were cultured for 2 days at 28°C, then the strains were re-streaked onto fresh plates with appropriate antibiotic selection and incubated for 3 days at 28°C before inoculation.
 
Induction of hairy roots
 
Healthy and uniform seedlings with unfolded green cotyledons were selected for agro-infiltration by the soaking and smearing methods as described below:
 
Soaking method
 
The cultures of strains were suspended in MS buffer (10 mM MES-KOH, pH 5.8 and 100 μM acetosyringone) until the OD₆₀₀ value reached 1.2 and then 15 ml of the bacterial suspension was pipetted onto a beaker with a volume of 25 ml. Healthy 7-day-old plants were taken out of chamber and the roots were excised approximately 3 cm below the cotyledons using a sterile razor blade. The plants were soaked into the bacterial suspension in the darkness at 22°C for 1 hour, then were transplanted into seed germinating bag containing 5% 1/2 MS culture liquid. The humidity level was maintained at 95%. Three days later, fluorescent light at 600 μmol m^-1s^-1 of photosynthetic photon flux density (PPFD) in a 16 h photoperiod was resumed. After 5-7 days, calluses production was noticed. After 14 to 15 days after excision, the transformed roots started emerging from the excision site.
 
Smearing method
 
Each plant was gently pricked with a needle tip for six or eight times around the cotyledon node. The cultures of strains that were suspended in the above-mentioned MS buffer were smeared around the wounded sites of the plants. The culture condition of infected plants was the same as that used in the soaking method. Typically, calluses were emerged from the wounding sites 5-7 days after infection, while hairy roots generated about 2 weeks.
 
Detection of transgenic hairy roots
 
According to reporter gene in the designed vector, reporter gene can be used to detect transgenic hairy roots. GFP can be assessed by fluorescence, whereas GUS can be assessed by GUS staining.
 
Visual detection of GFP
 
GFP fluorescence in the transgenic hairy roots were observed by using a portable fluorescence lamp (LUYOR-c3415RG) at the excitation wavelength of 488 nm and an LP 505-nm emission filter. Transgenic hairy roots showed GFP fluorescence labeling.
 
Histochemical GUS assay
 
The histochemical assay for the reporter gene GUS activity was performed following the established method (Jefferson, 2007). Briefly, the hairy roots were cut off and placed in a centrifuge tube containing GUS staining buffer [50 mmol L^-1 sodium phosphate, 0.5 mmol L^-1 potassium ferrocyanide, 0.5 mmol L^-1 potassium ferricyanide, 0.5 mg mL^-1 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (X-Gluc), 0.1% Triton X-100 and 20% methanol, pH = 7.0] for 12 h. The reaction mixture was incubated at 37°C overnight. Expression of the GUS gene in hairy roots was visually assayed and the percentage of transformed roots was calculated.

Development and efficiency of the soaking method using R1000
 
We first tested the efficiency of R1000 in inducting hairy roots from 6-day-old seedlings with root below cotyledons excised by the soaking method according to Ali et al., (2012) with little modifications. A total of 20 genotypes were screened in this test. In general, after about 5-7 days of growth following inoculation, callus generated around the excision points. After 14 days of growth, small hairy roots emerged from the callus (Fig 1). From the 20 genotypes, transgenic hairy roots with green fluorescence of eGFP were observed from five genotypes (LP089, P331024017, LP074, Cuiyu and Honghuaqinjia) under the irradiation of a fluorescence detector. The genotype ‘Cuiyu’ had the highest transformation rate, up to 33%, followed by ‘LP089’. Our attempt to improve the transformation efficiency in five randomly selected genotypes namely LP089, Pan’an common bean, Hongjinyuanjia, honghuaqinjia and LP074 by adjusting the infection concentration of A. rhizogenes from OD₆₀₀ 0.8 to 1.2 failed to raise the positive transformation rate.
 

 
Development and efficiency of the smearing method using R1000
 
To investigate if the infection method had a major effect on transformation rate, we then established a smearing-based transformation protocol using R1000 according to a published method with the K599 strain (Estrada-Navarrete et al., 2007). Four genotypes, namely LP089, P331024004, LP074 and Honghuaqinjia, were used in this analysis. After 5-7 days, callus generated around the smearing point and after 14 days of growth, hairy roots were developed from the callus. Through GUS staining, transgenic hairy roots that expressed b-glucuronidase were identified. As follow, the transformation efficiency of genotype ‘Honghuaqinjia’ increased drastically from 4% by soaking infection to 59% by smearing infection and that of the genotype ‘LP089’ increased from 25% to 47%, respectively. Under the soaking condition, the transformation efficiency of ‘P331024004’ was 0%, but under the smearing condition, the efficiency was 43%. However, the genotype ‘LP074’ had a transformation rate of 0% by smearing, whereas showed a rate of 17% by soaking. Overall, the transformation efficiency by using the smearing infection method was higher than that using the soaking infection. These results taken together suggest that the efficiency of A. rhizogenes strain R1000-mediated hairy root transformation is generally satisfying with the smearing method but is dependent on the genotype of the common bean plant.
 
Comparison of using the strains R1000 and K599
 
We next compared the transformation efficiency in the common bean genotype ‘LP089’ of using the two A. rhizogenes strains (R1000 and K599) containing the empty-plasmid pBI121-GUS. In our experiment, K599 appeared to be no better than R1000 to induce positive transgenic hairy roots. As show in Table 1, there was no obvious difference in the transformation rate between the two strains when the soaking infection method was used. Under the smearing infection method, the strain R1000 even exhibited a higher transformation efficiency of 53% as compared to that with the strain K599, which was only 17%. Therefore, for this specific common bean genotype, R1000 represents a good choice of A. rhizogenes strain.
 

       
To further test the superiority of the combination of strain R1000 and the smearing method, this hairy root induction protocol was then used in the genotype ‘Cuiyu’ as well (Fig 2). The transformation efficiency was found to be similar (~50%) to that in ‘LP089’. Therefore, the combination of the strain R1000 and smearing infection method could be considered as a quite versatile, useful system for hairy root transformation in common bean.
 

In the present study, the Agrobacterium rhizogenes strain R1000 harboring pBI121 was proved to be able to effectively induce hairy roots in common bean. Our data demonstrate that using an optimized method and appropriate common bean genotype such as ‘LP089’ and ‘Cuiyu’, a high efficiency of transgenic hairy root formation (~60%) can be achieved, which offers a new choice of common bean genetic transformation protocol. Further efforts to improve this protocol could be made by screening more genotypes of common bean for higher sensitivity to R1000 infection.

This work was supported by the National Key Research and Development Program of China (2021YFE19800), Key Research Program of Zhejiang Province (2021C02041) and the Natural Science Foundation of Zhejiang Province (LQ21C150004).

None.