This transformation procedure generates, with high efficiency (70–90%), hairy roots in cultivars, landraces and accessions of Phaseolus vulgaris (common bean) and other Phaseolus spp. Hairy roots rapidly develop after wounding young plantlets with Agrobacterium rhizogenes, at the cotyledon node, and keeping the plants in high-humidity conditions. Callogenesis always precedes hairy-root formation, and after 15 days, when roots develop at wounded sites, the stem with the normal root is cleaved below the hairy root zone. Transgenic roots and nodules co-transformed with a binary vector can be easily identified using a reporter gene. This procedure, in addition to inducing robust transgenic hairy roots that are susceptible to being nodulated by rhizobia and to fixing nitrogen efficiently, sets the foundation for a high-throughput functional genomics approach on the study of root biology and root–microbe interactions. This protocol can be completed within 30 days.
Grain legumes are one of the most valuable sources of dietary proteins for animal and human consumption, and of these, common beans represent 50% of the grain legume consumed worldwide1. In particular, in Latin America and Eastern Africa, it is the primary source of protein in the human diet2. The legumes, such as bean, have the advantage of fixing atmospheric nitrogen through a symbiotic association with nitrogen-fixing soil bacteria. The production of genetically modified common bean plants has been impeded by the availability of efficient techniques for introducing DNA into the plant genome and regeneration methods. A fast, reproducible and efficient transformation procedure is crucial not only for gene function studies, but also to allow crop improvement.
Common bean is reputed to be a species recalcitrant to transformation, limiting the possibilities for the molecular and physiological analysis of gene function. Particle bombardment-mediated systems have been developed to obtain transgenic plants of P. vulgaris, but are labor intensive3,4. Recently, the transformation system based on the bombardment of meristematic tissue of embryonic axes and the use of an efficient selective agent such as imazapyr, an herbicide of the imidazolinone class, was used to obtain transgenic bean plants, resistant to bean geminivirus. Field tests based on RNA interference (RNAi)-mediated resistance to bean golden mosaic virus are currently being evaluated in Brazil5.
Bean transformation mediated by Agrobacterium tumefaciens has been unsuccessful previously6. However, recent progress has increased the feasibility of transformation as a tool in Phaseolus7. Liu et al.8 developed a protocol based on the combination of sonication and vacuum infiltration to transform common bean with a late embryogenesis abundant gene, conferring abiotic stress tolerance. As the efficiency is 12%, this method looks very promising, and should be further evaluated. The susceptibility of common and tepary bean cultivars to several A. tumefaciens and A. rhizogenes strains has been previously evaluated9. Only one successful report of transformation with A. tumefaciens in a cultivar of Phaseolus acutifolious is available10. A. rhizogenes-mediated root transformation has been described for numerous legumes and composite plants with untransformed shoots and transformed roots that can be nodulated by rhizobia and colonized by micorrhiza11. Composite plants in legumes provide a fast and convenient alternative transformation procedure to generating stable transgenic lines.
Here, we describe a fast, reproducible and efficient common bean root transformation protocol for different cultivars and landraces of P. vulgaris (bean) with A. rhizogenes K599. This protocol gives further details of the method used by Estrada-Navarrete et al.12. This protocol was also successful for other Phaseolus spp.: P. coccineus, P. lunatus and P. acutifolius11. This procedure provides a new tool for functional genomics by enabling the generation of knockdown and gain-of-function composite plants, especially focused on genes involved in root biology and root–microbe interactions.
96% (v/v) ethanol
20% (v/v) commercial bleachCLORALEX (6% sodium hypochlorite solution)/sterile water
Vermiculite (grade 2 or 3)
A. rhizogenes K599 (NCPP2659) fresh culture
Rhizobium tropici strain CIAT899
Solid LB medium (10 g tryptone, 5 g yeast extract, 10 g NaCl, 15 g agar per liter)
Liquid LB medium containing 20% (v/v) glycerol
Liquid PY medium (5 g peptone, 3 g yeast extract per liter)
1 M CaCl2 sterile solution
10 mM MgSO4 sterile solution
50 mg ml−1 (w/v) rifampicin/dimethylsulfoxide stock solution
20 mg ml−1 (w/v) nalidixic acid/NaOH 0.1 N stock solution
0.05% (v/v) plant preservative mixture (PPM)/sterile H2O (Plant Cell Technology Inc.)
Citifluor (Ted Pella Inc., cat. no. 19470).
250 ml Erlenmeyer flasks
Forceps, dressing, 16 cm (Phytotechnology Lab. cat. no. F950)
Pots 3.5″ × 3.5″ (T.O. Plastics Inc., cat. no. 114150-1)
Pots 15 cm diameter (T.O. Plastics Inc., cat. no. 127142)
Greenhouse and/or growth chamber adjustable to 12 h light/8 h dark, 25 /28 °C equipped with cool-white lighting (40–50 μmol m−2 s−1)
“Humid chamber” (see EQUIPMENT SETUP)
Petri dishes 100/15 (Lab SyM, cat. no. NX7171B)
Bent glass rod in a triangular shape
Eppendorf tubes 1.5 ml (Axygen Scientific Inc., cat. no. MCT-150-C)
3 ml syringe
Needle 21G × 1.25 Prime; (0.8 × 32 mm)
Zeiss LSM 510 Meta confocal microscope attached to an Axiovert 200 M
B&D solution Make stock solutions of A, B, C and D respectively: solution A (2 M CaCl2), solution B (1 M KH2PO4 pH 7.0), solution C (20 mM Fe-citrate; this should be kept away from light), solution D (0.5 M MgSO4, 0.5 M K2SO4, 2 mM MnSO4, 4 mM H3BO3, 1 mM ZnSO4, 4 mM CuSO4, 0.2 mM CoSO4, 0.2 mM Na2MoO4). Mix 0.5 ml per liter of each stock solution A, B, C and D. Add 8 ml of 1 M KNO3 per liter, to obtain a B&D solution supplemented with 8 mM of nitrogen. The complete solution can be stored at room temperature (23–25 °C).
“Humid chamber” We use a plastic tray with transparent lid humidome: for example, Plastic tray (Acroplastics Ltd, cat. no. 113030-1); transparent lid acrodome 6″H × 21″L × 11″W (Acroplastics Ltd, cat. no. 143850-1). The tray should be sterilized by spraying with 70% (v/v) ethanol.
Sterilization of seeds (using ethanol/hypochlorite)
Timing: 30 min
Place around 100 seeds into a sterile 250 ml Erlenmeyer flask and wash seeds twice with 100 ml of sterile water. Remove water.
Add 100 ml of 96% ethanol while gently mixing and leave for 1 min.
Discard the ethanol and rinse the seeds four times with excess sterile water.
Add 100 ml of 20% sodium hypochlorite and leave for 5 min.
Discard the sodium hypochlorite and rinse four times with excess (200 ml) of sterile water.
Germination of bean seeds
Timing: 2 days
Distribute the sterilized seeds with forceps, placing them ∼2 cm apart into a metallic tray on top of a wet sterilized paper towel.
Cover the tray with aluminum foil, incline the tray slightly (∼2.5 cm) and incubate in the dark for 2 days in the growth chamber at 25–28 °C. All germinates should grow in the same direction owing to the inclination of the tray. Seed of a good quality should give 95–100% germination (Fig. 1a).
Place the germinated seeds into wet vermiculite (B&D) in small pots. Use one seedling per pot with autoclaved vermiculite (or alternatively into each cell of a 30- to 50-well germination tray). Leave seedlings to develop for 3 days in a growth chamber at 25–28 °C (Fig. 1b).
Inoculum preparation (A. rhizogenes K599)
Timing: 3 days
On the same day that Step 8 is performed (3 days before plantlets infection, Step 11), spread 50–100 μl of the Agrobacterium strain harboring the binary vector directly from a −80 °C glycerol stock (LB medium containing 20% (v/v) glycerol) onto LB plates with appropriate antibiotic selection.
Incubate the Agrobacterium plate for 2 days at 28 °C, then re-streak a single colony onto a fresh plate. Prepare a backup stock (kept at −80 °C) and a working stock (also at −80 °C), which are used to inoculate freshly poured plates. Working and backup stocks are obtained from fresh fully grown plates prepared as described in Step 9 by suspending the bacteria in 3 ml of LB with 20% glycerol before freezing. By using the freshly prepared working stock glycerol, better bacterial growth and high cell density to inoculate the plants can be assured.
Induction of hairy roots
Timing: 8 days
After keeping the plants for 3 days in the growth chamber (the plants incubated in Step 8), select healthy and robust plantlets to use in the transformation experiments (Fig. 1c).
Sterilize the glass rod by using 96% ethanol and burn it under the flame in sterile conditions.
Inoculation of bacterial suspension into the cotyledonary nodes
Add 1 ml of sterile water to the Agrobacterium from the plate described in Step 10, resuspending the bacteria in the plate with a glass rod at the edge of the Petri dish to obtain the bacterial inoculum (this will have a milky nature and be rose-colored). Collect the bacterial suspension into an Eppendorf tube (Fig. 1d).
Collect the suspension with the syringe and slightly wound the cotyledonary node with the needle tip, and carefully prick each plant two or three times at different positions around the node (Fig. 1e).
Inject 5–10 μl drops of the inoculum into the wound.
After the plantlets are infected, immediately wet the vermiculite with B&D supplemented with nitrogen (8 mM KNO3).
Cover the tray containing the little pots with the transparent lid acrodome sterilized with 70% ethanol. Close and seal to maintain humidity (>90%), which is essential for callus formation and the development of hairy roots. Callus should be observed at the wounding sites 5–7 days after infection (Fig. 1f).
Keep the plants in the sealed trays for 16 h light/8 h dark in a growth chamber for 8 days at 25/28 °C. During this time, water and spray the plants with an atomizer every second day using sterile B&D solution containing 2–8 mM KNO3 as the nitrogen source, depending on further interest. Hairy roots should be approximately 3–5 cm in length (Fig. 1g,h) after this period. Plantlets injected with sterile water from neither calli nor hairy roots.
When hairy roots are fully grown from wounded sites, approximately 15 days after infection, remove the primary root by cutting the stem 2 cm below the hairy roots.
Transfer plants to a new pot with wet sterile vermiculite. Place 1–3 plants into a 15 cm diameter pot and cover hairy roots up to 1 cm with sterile vermiculate. Keep plants in a growth chamber (16 h light/8 h dark at 25–28 °C; Fig. 1i) until hairy roots emerge.
Innoculate recently emerged hairy roots
Timing: 10–12 days
Inoculate Rhizobium etli or R. tropici strain CIAT899 into a 250 ml Erlenmeyer flask containing 100 ml of PY medium supplemented with 100 μl of 7 mM CaCl2, 50 μg ml−1 rifampicin and 20 μg ml−1 nalidixic acid.
Incubate at 30 °C for 24 h with shaking at 300 r.p.m.
Centrifuge for 3 min at 2,300g at room temperature and discard the supernatant.
Resuspend cells in 100 ml of 10 mM MgSO4, measure A600 of the resulting suspension and calculate the number of cells per ml assuming that one A600 unit equals 5 × 108 cells ml−1.
With a pipette, inoculate emerged hairy roots when they are ∼0.5–2 cm long with about 5–8 × 108 bacteria per ml per plant.
Return the plants to the growth chamber (16 h light/8 h dark at 25–28 °C) and stop watering for 2 days after hairy-root inoculation with Rhizobium, but keep plants in covered trays, to allow infection threads to get going. Reassume watering with sterile B&D solution with 2 mM KN03 for 5 days while nodules develop.
Collect nodulated hairy roots to determine the phenotype and process samples for activity of reporter genes or microscopic analysis (Fig. 1j–l).
Troubleshooting advice can be found in Table 1.
It takes 30 min to sterilize seeds, growing seedling requires 5 days, hairy root development requires 15 days and fully grown nodules in hairy roots take between 3 and 4 weeks to develop.
The frequency of co-transformation obtained using this procedure is high (75–90%) when cultivars and landraces of Phaseolus spp. were infected with A. rhizogenes K599 containing the binary vector p35SGFPGUS+ (ref. 12). Transgenic hairy roots induced are robust, grow rapidly and are susceptible to be nodulated by inoculation with Rhizobium.
Although a positive selectable marker, such as an antibiotic or herbicide, has not been proven to be necessary (co-transformation efficiencies are in general higher than 70%), co-transformation frequencies may vary with different binary vectors. To screen for co-transformed hairy roots, a GFP or equivalent reporter protein-expressing binary vector can be used. We are currently using a construction (pTdTRNAi) with a double tandem tomato fluorescent protein (tdT) TFP reporter gene, under the NOS promoter13 (see Fig. 1k).
In our group, we have successfully used an antisense construction (ASN30) directed against a conserved region shared by a common bean nodulin gene family encoding nodulin 30 (NPV30) (ref. 14). Interestingly, although this nodulin family is expressed in normal root nodules after the onset of nitrogen fixation15, in the downregulated nodules, the integrity of infected cells is highly impaired when antigen was detected 50% below normal protein levels, as determined by western blotting (J.-E.O. et al., unpublished observations, data not shown). Another example that was tested in the bean hairy root system was the subcellular localization of NPV30 by expressing its signal peptide (PSN30) fused to a GFP-β-glucuronidase (GFP5-gusA) cloned into the pCAMBIA 1304 (pCAMBIA-PSN30GFP-GUS) vector. We analyzed bean root-nodule sections by confocal microscopy. In Figure 1l, it can be observed that this hydrophobic signal peptide directs the GFP-GUS fusion protein to the membranes that surround rhizobia-infected cells, forming intercellular aggregates or patches (Fig. 1l, white arrows); it is also confined to the cytoplasm of uninfected and epidermal cells.
We are currently testing antisense and RNAi constructs to downregulate, at the translational and mRNA silencing level, several receptor-like kinases and key signaling proteins, involved in perception of R. etli, and in nodule development and metabolism.
This research was partially supported by CONACYT 42562-Q and by Dirección General de Asuntos del Personal Académico IN-215805-2 grants. We thank Dr. José Luis Reyes and Oswaldo Valdés L. for pTdTRNAi vector construction and Olivia Santana for technical assistance.