Acetosyringone treatment duration affects large T-DNA molecule transfer to rice callus

Construction of binary vectors

The binary vector BIBAC2 [27] is a single-copy plasmid that harbors a GUS::NPTII fusion gene and a hygromycin B resistant gene (HYG) located near the left and right border of the T-DNA, respectively. To examine the relationship between the T-DNA size of BIBAC2 and the transformation efficiency, two recombinant binary vectors were constructed. A random piece of 50-kb yeast genomic DNA and a 5-kb Arabidopsis genomic DNA fragment were individually cloned into BIBAC2, and the resulting recombinant binary vectors were named pB50 and pB5, respectively (Fig. 1a). In addition, the same 5-kb Arabidopsis genomic DNA was cloned into a multi-copy (10–20 per cell) binary vector pCAMBIA1301 [32, 33] with HYG and GUS gene located near the left and right border of the T-DNA, respectively, and the resulting vector was named pC5 (Fig. 1b). Comparison between pC5 and pB5 could indicate whether and how the copy number of the binary vector affects transformation efficiency. pB50, pB5 and pC5 were individually introduced into Agrobacterium strain AGL1, and the resultant strains are named B50, B5 and C5, respectively.

T-strand formation and accumulation in Agrobacterium cells is affected by T-DNA size and the copy number of the binary vector

T-strand formation in Agrobacterium cells is a critical step in Agrobacterium-mediated plant transformation. To examine the factors that have impact on the T-strand formation in Agrobacterium cells, we performed T-strand immuno-precipitation (IP) using the anti-VirD2 antibody and qPCR assay to measure T-strands formation inside the Agrobacterium cells harboring pB50, pB5, or pC5 following AS induction.

We first confirmed the functionality of the anti-VirD2 antibody. As shown in Fig. 2a, the anti-VirD2 antibody was able to detect GST-VirD2 fusion protein but not GST only. Next, we examined the expression level of VirD2 after AS induction. Samples with similar number of bacterial cells were collected at 0, 6, 9, and 24 h after AS induction. The expression of VirD2 was hardly detected by 9 h but became very obvious at 24 h after AS induction (Fig. 2b).

Excessive amount of the anti-VirD2 antibody was used in T-strand-IP in order to maximize immature T-complex binding and capture. The fold increases of the T-strands were quantified by qPCR by measuring the two transgenes near the right and left border of the T-DNA (GUS and HYG), respectively, and normalized against Agrobacterium chromosomal marker gene dnaK. Our data showed that the amount of T-strands was highly significantly increased after AS induction within 24 h in all three strains of Agrobacterium cells (Fig. 2c-f). As shown in the figures, at 6 h, significant differences were already seen between B5 and B50 as well as between C5 and B5. From 9 to 12 h, the amount of T strands in B5 and C5 increased near-exponentially while fold increase in B50 just started. From 12 to 24 h, the increase rates in C5 and B50 were down slightly whereas B5 still maintained exponential-like fold increase. By 24 h, the amount of T-strands in C5 strain increased by around 3000 fold, while the accumulation in B5 and B50 strains rose to ~ 1500 and 100–250 fold, respectively. The results showed that the amount of T-strands increased remarkably upon AS induction in all three strains, and the increase was negatively impacted by the T-DNA size and positively affected by the copy number of the binary vector (Fig. 2c-f). The results also demonstrated that longer AS induction (such as 24 h) was beneficial to T-strand formation and accumulation, and thus might facilitate improvement of transformation efficiency.

Longer AS induction improves large DNA fragment transfer

AS is a naturally occurring phenolic compound upon plant wounding, and induces expression of vir genes in Agrobacterium [18, 34]. It is a common practice to add AS in bacterial culture to activate the Agrobacterium virulence for monocot transformation [4]. However, the literature is ambiguous with regards to an optimal AS treatment duration. Our results indicate that the longer duration (24 h) of the treatment leads to significantly higher accumulation of the T-strands in the Agrobacterium cells (Fig. 2c-f). We hypothesize that longer AS induction may improve the transformation efficiency, particularly for the single-copy vectors and the larger size T-DNAs since longer time of AS induction produced hundreds of fold more T-strands (Fig. 2c, d). To validate this hypothesis, AS-treated B50 strain collected at three different time points (3, 9 and 24 h) was used to transform rice calli, and the transiently-expressed GUS transcripts were quantified by qRT-PCR following 3-day co-cultivation. Our data clearly showed that by 24 h, the GUS transcript level increased by ~ 3.5 fold from 3 h, which was highly significant when compared to 3 and 9 h (p = 0.002 and 0.008, respectively, Fig. 3), an indicator of the improved efficiency of transient transformation, suggesting that extended AS induction time could be a valuable approach to improve large size T-DNA transfer to rice cells.

Constructs, bacterial strain, antibodies and recombinant proteins

A random piece of 50-kb yeast genomic DNA was isolated using CHEF Yeast Genomic DNA Plug Kit (Cat#170–3593, Bio-Rad, Hercules, CA) and a 5 kb Arabidopsis genomic DNA fragment was individually cloned into BIBAC2 between the two NotI sites as single copy binary vectors pB50 and pB5, respectively. The same 5 kb Arabidopsis genomic DNA fragment was cloned into pCAMBIA1301 at the HindIII site to serve as a multi-copy binary vector pC5. Each binary vector was introduced into Agrobacterium strain AGL1 individually. AGL1 is a recA-deficient derivative of A. tumefaciens strain C58 [39] and was acquired from Dr. Lynn Dahleen, USDA, ARS, Fargo, ND. Binary vectors BIBAC2 and pCAMBIA1301 were acquired from Cornell University, Center for Technology Enterprise and Commercialization (CCTEC), and CAMBIA (Canberra, Australia), respectively. The antibody against VirD2 was kindly provided by Dr. Zambryski at UC-Berkeley [40] and Dr. C. Baron at the University of Montreal (Montreal, Canada).

For expression and purification of the recombinant protein, a full-length gene of virD2 (amplified with primers virD2Fwd: 5’-GGGTCCATGGATATCGGGATGCCCG ATCGCGCTCA AG-3′, and virD2Rev: 5’-TGCTCGAGTGCGGCCGCACTAGGTC CCCCCGCGCC CA-3′) was cloned into the bacterial expression vector pET42b (Novogen, Hornsby Westfield, Australia) between BamHI and HindIII using Gibson assembly (New England Biolabs, Ipswich, MA). 6xHis-tagged GST fusion protein was expressed in Escherichia coli strain BL21 (DE3) (Agilent Technologies, Santa Clara, CA). The cell pellets were lysed by French press in the lysis buffer containing 50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole, and protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO). The cell extract was prepared by centrifugation at 20,000 g for 20 min at 4 °C. The cleared cell extract was incubated with His60 Ni Superflow Resin (Clontech, Mountain View, CA) equilibrated in the lysis buffer at 4 °C for 2 h. The recombinant proteins were subsequently purified in the gravity-flow column following the manufacturer’s instructions (Clontech).

Sample preparation

For Agrobacterium samples, Agrobacterium strain AGL1 harboring various binary vectors was grown overnight in 150 ml YEP medium containing 25 mg/L rifampicin and 50 mg/L kanamycin at 28 °C. Cells were then centrifuged and re-suspended in 150 ml AB induction medium [41]. The OD₆₀₀ reading of the culture was adjusted to 0.6, and six ml from each culture was collected at this point as non-induced samples. Then acetosyringone (AS) was added (final concentration 200 μM) to each of the remaining cultures. Cultures were grown at room temperature (RT). Agrobacterium samples were collected at 6, 9, 12 and 24 h after AS addition, immediately washed by 20 mM sodium phosphate buffer (pH 6.8) and pelleted for lysate extraction. When sampling, the OD₆₀₀ value of each sample was measured and adjusted to 0.6.

Lysate extraction from Agrobacterium cells was prepared after sampling [42]. Briefly, cells were re-suspended in 200 μl TES buffer (50 mM Tris-HCl pH 6.8; 2 mM EDTA; 1% β-mercaptoethanol; 1% SDS) and shaken for 30 min at 37 °C. Then, 900 μl of NP1 buffer (150 mM Tris-HCl pH 8.0, 0.5 M sucrose, 10 mM EDTA) containing 1 mg/ml lysozyme was added. The mixture was incubated on ice for at least 1 h, followed by 30 min shaking at 37 °C. triton X-100 was then added to a 4% final concentration and the mixture was incubated for 15 min at RT. Three hundred and forty-three μl of 5× EDTA-free protease inhibitors cocktail (in 25 mM MgCl₂, Roche Applied Science, Penzberg, Germany) was added and the mixture was rotated for 15 min at 37 °C followed by 2–3 h at 4 °C. The insoluble material was removed by centrifugation at 18,400 g for 15 min and the supernatant was collected. The protein concentration of the total soluble lysate was determined by the Bradford reagent (Bio-Rad) and diluted to a final concentration of 0.5 mg/ml. One mL of the diluted lysate was used for T-strand immune-precipitation (T-strand-IP). Another 100 μL of the same diluted lysate was treated with DNase-free RNase (Thermo Fisher Scientific, Waltham, MA), Proteinase K (Thermo Fisher Scientific) and then precipitated by ethanol to obtain the genomic DNA, which would be used as the “input” in the T-strand-IP and qPCR assays described later.

For rice calli samples, rice (Oryza sativa L.) cultivar ‘Taipei 309″ seeds were used for callus induction as previously described [5]. The calli were sub-cultured every two weeks. Two- to three-month-old healthy friable rice calli were selected for Agrobacterium transformation. Agrobacterium culture was used to infect 100–120 pieces of rice calli according to Patel et al. [5]. Infected calli were blotted onto three layers of sterile filter paper to remove the excessive Agrobacterium suspension and then put on clean sterile filter papers for sample collection. Healthy rice calli without Agrobacterium infection were used as controls.

Immunoblot analysis

Concentration of the total soluble proteins was determined by the Bradford reagent (Bio-Rad). Equal amounts of total proteins (20 μg) from each sample were subjected to SDS-PAGE and subsequently transferred to polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA). After blocking with 2% non-fat milk in TBST buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 0.05% Tween-20), the membrane was incubated in the same buffer containing 1/1000 dilution of rabbit anti-VirD2 antibodies. After three times of washing with TBST buffer, the membrane was incubated with 2% non-fat milk in TBST buffer containing 1/5000 dilution of the goat anti-rabbit HRP (Thermo Fisher Scientific). The signal was detected with the SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher Scientific).

T-strand-immunoprecipitation and qPCR assay of Agrobacterium cells

One mL of diluted lysates (0.5 mg/mL) extracted from Agrobacterium cells were incubated with the antibodies against VirD2 at 4 °C overnight. Meanwhile, the Dynabeads™ Protein G magnetic beads (Thermo Fisher Scientific) were blocked with 20 μg BSA (New England BioLabs) and 20 μg glycogen (Thermo Fisher Scientific) at 4 °C overnight. The next day, the magnetic beads were washed twice using the extraction buffer NP1, and then re-suspended in appropriate volume of NP1 buffer. Forty μL of equilibrated magnetic beads were then added to the mixture of lysate and antibody against VirD2, and the total mixture was incubated for four hours at 4 °C. After washing with NP1 buffer for four times, freshly prepared elution buffer (1% SDS, 0.168 g NaHCO₃/20 ml buffer) was used to elute the products of immuno-precipitation at 65 °C. The eluate was digested by Proteinase K at 55 °C overnight and then purified using phenol-chloroform extraction. The final products (T-strand DNA) were precipitated by ethanol.

qPCRs were performed with iTaq Universal SYBR Green Supermix (Bio-Rad) to analyze T-strand formation in Agrobacterium cells on a Real-Time PCR Instrument (Agilent). Agrobacterium chromosome marker dnaK (with primers of dnaKFwd: 5’-TACCTTCCTCGGTGGTGAAG-3′, and dnaKRev: 5’-CGAGGACGAAAGTTC GATC-3′) was amplified from the 100 μL diluted lysate prepared previously and used as an internal control. T-strand marker gene HYG (HYGFwd: 5’-GGTCGCCAACATCTT CTTCT-3′, HYGRev: 5’-CGAAATTGCCGTCAACCAAG-3′) or GUS (GusFwd: 5’-ACGTCTGGTATCAGCGCGAAGTC-3′, GusRev: 5’-TATAGCCGCCCTGATGCTCC ATC-3′) was amplified from the purified T-strand-IP products. Relative quantification using comparative CT calculation method was applied in qPCR data analysis.

Transient GUS transcripts analysis in the infected calli

Agrobacterium cells harboring the binary vector of pB50 were grown overnight in 15 ml of YEP medium containing 25 mg/L rifampicin and 50 mg/L kanamycin at 28 °C. The next day, the overnight culture was divided into 3 equal aliquots (5 ml each). Forty-five ml of fresh infection medium containing 200 μM AS was added to each aliquot for induction at RT. The cultures were sampled at three time points (3, 9, and 24 h). The OD₆₀₀ of each 50 ml Agrobacterium suspension was examined and adjusted to 0.6 before rice calli infection. The rice transformation was performed as previously described [5]. After a three-day co-cultivation, samples were collected, washed thoroughly with 200 mg/L timentin (GlaxoSmithKline, Research Triangle Park, NC) and stored at − 80 °C.

Total RNA was isolated from the infected rice calli using the Quick-RNA MicroPrep (Zymo Research, Irvine, CA), and then treated with DNase I (RNase-free DNase set; Zymo Research) to clean up potential contamination of the genomic DNA. Concentration of the total RNA was quantified using a NanoDrop 2000C spectrophotometer. The first-strand cDNAs were synthesized using iScript cDNA Synthesis Kit (Bio-Rad). Subsequent real-time RT-PCRs were performed with iTaq Universal SYBR Green Supermix (Bio-Rad) to analyze GUS transcripts on a Real-Time PCR Instrument (Agilent). The housekeeping gene UBQ5 was used as an internal control [43]. The relative fold change of GUS transcripts is calculated against the GUS expression level with 3-h AS induction before transformation.

Statistical analysis

Two-tailed student t-test was carried out to evaluate significant differences in values of qPCR and qRT-PCR. “*” indicates significant differences: ***p < 0.001, **p < 0.01, *p < 0.05. Data presented are means ± standard errors (n = 3).