Wild-type Agrobacterium rhizogenes-mediated gene transfer in plants: Agrobacterium virulence and selection of transformants

Agrobacterium rhizogenes ATCC 15834 wild type strain was transformed with the binary vector pBI121 using the heat shock method. The transformed Agrobacterium was then tested for virulence through tobacco leaf explant transformation. Compared to the non-transformed Agrobacterium, the transformed Agrobacterium showed reduced virulence, producing signifi cantly lower number of hairy roots in tobacco leaf explants. Although the transformed Agrobacterium showed reduced virulence, it was able to transfer the T-DNA of the binary vector into the plant genome, resulting in stable GUS expression in the generated hairy roots. This indicated that in addition to the transfer DNA (T-DNA) from its root inducing (Ri) plasmid, the transformed Agrobacterium is also capable of transferring the binary vector T-DNA and allowing the integration of a foreign gene. Results also showed that hairy root generation effi ciency of the transformed Agrobacterium varied with the concentration of the selection agent (kanamycin). Hairy root generation effi ciency (hairy roots·explant-1) progressively increased with decreasing concentrations of kanamycin; and the effi ciency was highest in the absence of kanamycin. Generated hairy roots showed very strong to tiny GUS expression even those that grew under the highest concentration of the kanamycin (50 mg·L-1). This indicated that co-transformation and effi cient transgene expression does not always occur. Research Article Wild-type Agrobacterium rhizogenesmediated gene transfer in plants: Agrobacterium virulence and selection of transformants Mohammad M Rana1,2, Muhammad Abdullah1, Ferdinand L Shamalla1 and Shu Wei1* 1State Key Laboratory of Tea Plant Biology and Utilization, Anhui Agricultural University, 130 Changjiang Ave W Hefei, Anhui 230036, China 2Bangladesh Tea Research Institute, Srimangal-3210, Moulvibazar, Bangladesh *Address for Correspondence: Shu Wei, State Key Laboratory of Tea Plant Biology and Utilization, Anhui Agricultural University, 130 Changjiang Ave W, Hefei, Anhui, 230036, China, Tel & Fax: +86-551-65783941; Email: weishu@ahau.edu.cn Submitted: 27 May 2017 Approved: 09 June 2017 Published: 12 June 2017 Copyright: 2017 Rana MM, et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


INTRODUCTION
Agrobacterium-mediated genetic transformation is the most widely used method of transferring genes into plants [1]. Besides being cheaper and simpler than most direct gene transfer methods, it reduces the rearrangement of transgenes, and ef icient integration of the transgene into the plant genome [2]. Agrobacterium rhizogenes is a Gram-negative, rod-shaped soil bacterium [3], which is a close relative of the best-characterized species A. tumefaciens among the genus Agrobacterium [4]. All A. rhizogenes strains are characterized by the presence of a large root inducing (Ri) plasmid [5]. Following A. rhizogenes infection, hairy root formation occurs directly from the infection site in the form of large numbers of small protruding roots resembling ine hairs [6]; the genomic integration of a set of genes (encoding enzymes involved in auxin and cytokinin biosynthesis) is mediated through a speci ic transfer DNA (T-DNA) from the Ri plasmid [7]. A. rhizogenes-induced roots have the unique property of being able to grow in vitro without exogenous plant growth regulators [8,9].
Hairy-root cultures have several properties that have promoted their use for plant biotechnological applications. These roots are also characterized by their genetic stability [10]. Their fast growth, short doubling time, ease of maintenance, and ability to synthesize a range of chemical compounds and proteins offer hairy roots advantages over plant cell suspension cultures as a continuous source for the production of valuable secondary metabolites and foreign proteins [4]. Hairy root cultures are usually able to produce the same compounds found in wild-type roots of the parent plant, but without the loss of concentration frequently observed with callus or cell suspension cultures [11]. Using these features of A. rhizogenes, a broad range of dif iculties in in vitro plant tissue and organ cultures were eliminated, and it became possible to generate fast growing organs with the capability of producing more metabolites even at higher concentrations than in the mother plant [12,13].
Apart from this, the ability of A. rhizogenes to transfer the disarmed T-DNA region from the A. tumefaciens-based binary vector along with the T-DNA from Ri plasmid has been exploited to develop co-transformed transgenic hairy root culture [14]. The co-transformed hairy roots are an extremely powerful tool for rapid and reproducible research in diverse areas including the establishment of gene function [15], analysis of promoter activity [16], heterologous protein expression [17], antibody production [18], modi ication of plant metabolic pathways [19] and functional genomics studies [20]. In addition, composite plants can be developed [21]; transformed hairy roots can also be used to produce stable transgenic plants, which have already been reported in some plants including Crotalaria [22], Medicago [23], Alyssum [24] and Plumbago [25].
Importantly, the T-DNA from Ri plasmid and the disarmed T-DNA region from the A. tumefaciens-based binary vector independently integrates into the plant genome.
Thus, the transgenic plant regenerated from a transgenic hairy root culture allows segregation of the two T-DNAs during meiosis and transgenic plant lines with only the binary vector derived T-DNA can be obtained in the ilial progeny [26]. The abovementioned advantages have made A. rhizogenes a promising agent for plant genetic engineering; at least for those plants that are recalcitrant or less-responsive to the A. tumefaciens-mediated plant transformation. However, insertion of a foreign plasmid into the wild-type A. rhizogenes strain might have some impact on its virulence due to the involvement of two plasmid simultaneously during the transformation process. In addition, some other events can also happen like T-DNA from either the plasmids or any of the plasmids can be transferred and integrated into the plant genome that will lead different consequences. Therefore, this basic study was conducted using the model plant tobacco and wild type A. rhizogenes ATCC 15834 strain to understand the impact of insertion of a foreign plasmid into this wild-type strain and to optimize the selection process of the transformants.

Bacterial strains and transformation vector
The binary vector pBI121 used in this study was purchased from Clontech (Beijing, China). The vector contained gusA and nptII, as reporter and selective genes, respectively. Escherichia coli strain DH5α was used to amplify the plasmid pBI121 in small scale. Agrobacterium rhizogenes strain ATCC 15834 was used for the transformation study.
Both bacterial strains were obtained from Lab 263, Sate Key Laboratory of Tea Plant Biology & Utilization, Anhui Agricultural University, Hefei, Anhui, China.

Bacterial transformation
The binary vector pBI121 was irst introduced into E. coli strain DH5α following standard protocol to amplify the plasmid in small scale. The plasmid was then extracted from E. coli and inally introduced into the Agrobacterium following heat shock method. Successful transformation of Agrobacterium was con irmed through PCR. For the PCR, 1 μL bacterial culture was used as template for each reaction mixture (25 μL). Positive and negative controls consisted of 1 μL plasmid DNA (∼50 ng) and 1 μL ddH 2 O, respectively. The primer sequences of 35S (Forward), 5′-CAATCCCACTATCCTTCG CAAGACCC-3′ and NOS (Reverse), 5′-GATCTAGTAACATAGATGAC ACCG-3′ were used to amplify a fragment length of 2239-bps from pBI121. PCR was performed in a 25 μL reaction volume containing 1 Unit of Easy-Taq DNA Polymerase, 1 μL of bacterial culture, 1 μL of each primer (10 mM·L -1 ), 2.5 μL10×buffer, 2 μL dNTP (2.5 mM·L -1 ), and 17.5 μL water. Ampli ication was performed in a programmable Thermal cycler (Bio-Rad S1000) as follows: initial denaturation at 95°C for 5 min, 30 cycles of ampli ication (denaturation at 95°C for 30s, annealing at 55°C for 30s, and extension at 72°C for 2min 30s) and a inal extension of 72°C for 10 min. After ampli ication, the samples were resolved by DNA gel electrophoresis ( Figure 1). Finally, the veri ied positive Agrobacterium (transformed with pBI121) culture was preserved at -80°C for further use.

Tobacco transformation
To compare the virulence of the transformed A. rhizogenes with wild-type, and to check the ability of the transformed Agrobacterium to transfer and integrate the binary vector T-DNA into the plant genome, the tobacco leaf explant transformation experiment was carried out. Tobacco leaf explants were inoculated separately using both wild-type and transformed Agrobacterium. Hairy root generation ef iciency of the transformed Agrobacterium was also tested under different selective pressures (i.e. different concentrations of the antibiotic kanamycin). In brief, the treatments included are: (T0) Wild-type Agrobacterium+0 mg·L -1 kanamycin; (T1) Transformed

Experimental design
The experiment was carried out in a completely randomized design with three replications. Each replication consisted of a single Petri plate containing 10-15 leaf explants.

Preparation of media
The half-strength MS media used in this study was solidi ied with 6.5g·L -1 agar and the pH was adjusted to 5.7 before autoclaving at 121°C for 20 min. Antibiotics were ilter-sterilized and added when the autoclaved media was cooled to about 50°C. For the induction of hairy roots in tobacco leaf explants, lager culture of the transformed A. rhizogenes was prepared in LB broth containing antibiotic rifampicin and kanamycin (both at 50 mg·L -1 ). Agrobacterium culture was grown to an OD 600 value of around 0.6. The cells were then pelleted by centrifugation at 2100×g for 10 min, and the pellet was re-suspended in MS liquid medium to get a inal OD 600 value of 0.6 before inoculation.

Agrobacterium inoculation of tobacco leaf explants and induction of hairy roots
Young tobacco leaves were excised from the in vitro grown plantlets and soaked in MS liquid media (4.4g·L -1 MS salts, 30g·L -1 sucrose, pH5.7) in a sterile Petri dish to avoid de-hydration. The leaves were irst punctured at several places using a sterile scalpel to facilitate Agrobacterium infection, and then cut into about 8×10 mm rectangular pieces and placed into liquid MS media in a Petri dish. The explants were submerged in the Agrobacterium culture for 5 min. Inoculated leaf pieces were blotted on sterile ilter paper to remove excess Agrobacterium, and then plated to the halfstrength MS media supplemented with 30 g·L -1 sucrose for co-cultivation (25°C in the dark for 2days). Following co-cultivation, the leaf explants were rinsed 2 times with sterile distilled water to remove Agrobacterium, and then subjected to one additional rinses with MS liquid containing 500 mg·L -1 carbenicillin. Afterwards, the explants were blotted on ilter paper, and plated (adaxial side down, 10-15 explants·plate -1 ) into the media with different concentrations of kanamycin in combination with 300 mg·L -1 carbenicillin. Petri dishes were covered and well-sealed with a strip of para ilm. The cultures were incubated at 25±1°C with a 16 h photoperiod for the generation of hairy roots. The non-transformed wild-type Agrobacterium was used as control, and in this case, no kanamycin was used for the respective culturing.

Data collection
After four weeks of culturing, the individual hairy root lines were counted from each explant through destructive sampling. The mean data from all explants within a plate represented the data for one replication. In this way the data from replications of all treatments were collected. For this study, the treatments were compared by determining the mean number of hairy root lines generated per explant (i.e. hairy roots·explant -1 ).

β-glucuronidase (GUS) expression assay in the hairy roots
The histochemical assay for gusA gene expression was performed as previously described by Jefferson et al. [27]. In brief, the hairy roots were separated from the explants, submerged into GUS staining solution (1mM X-Gluc, 50mM phosphate buffer, 10mM EDTA, and 0.1% Triton x-100) and incubated for 12 h at 37 °C. The explants were soaked in 70% ethanol and examined for blue colour development.

STATISTICAL ANALYSIS
Results of hairy root generation were expressed as the mean value±standard deviation (SD). Data were analyzed using ANOVA, and statistical differences between means were compared by Duncan's multiple range test (DMRT), using the statistical package MSTAT [28].

Transformation of wild-type A. rhizogenes
After several attempts, the binary vector pBI121 was successfully introduced into the CaCl 2 competent A. rhizogenes ATCC15834 through heat shock method, which was con irmed by PCR and gel electrophoresis.

Comparative effi ciency of the wild-type and transformed Agrobacterium in hairy root generation
In order to compare the transformation ef iciency of the wild-type and transformed A. rhizogenes, both the wild-type and transformed Agrobacterium were used to generate hairy roots from tobacco leaf explants. Hairy roots started to appear after about 9 days of inoculation. After about four weeks of inoculation, hairy roots were almost covered all the explants in T0 (Figure 2). Results showed that when the wildtype A. rhizogenes ATCC15834 was transformed with a second plasmid (pBI121), its virulence signi icantly decreased (p<0.01), showing 6.95±0.76 hairy roots·explant -1 in T1 compared to the T0 where the wild-type Agrobacterium generated 12.70±1.41 hairy roots·explant -1 (Figure 2 and 3). This suggested that transfer of T-DNA from the Ri plasmid was signi icantly affected by the presence of the second plasmid and determination of the underlying reason will require further study.

Hairy root generation effi ciency under selection pressure
To observe the hairy root generation ef iciency of the transformed Agrobacterium under different selection pressures, different concentrations of kanamycin was used in the selection media. It was observed that number of hairy roots gradually decreased with increasing concentrations of the selection agent kanamycin (Figure 2). Signi icantly (p<0.01) higher hairy roots per explant (6.95±0.76) was obtained in T1 when no kanamycin was used (Figure 2 and 3). It was followed by 4.94±0.49 hairy roots·explant -1 in T2 (12.5mg·L -1 kanamycin), 2.86±0.80 hairy roots·explant -1 in T3 (25 mg·L -1 kanamycin), 1.83±0.39 hairy roots·explant -1 in T4 (37.5mg·L -1 kanamycin), and the lowest of 0.753±0.150 hairy roots·explant -1 in T5 (50 mg·L -1 kanamycin). However, the differences of ef iciencies between T3 and T4, and in between T4 and T5 were not statistically signi icant (p>0.05). In general, when the T-DNA of the Ri plasmid is transferred into a plant cell and integrated into the genome, the hairy roots are supposed to be generated only from that transformed cell. If the same cell is also transformed simultaneously with a resistant gene from the binary vector, then the cotransformed cell will be capable of generating hairy roots under selection pressure. The results showing signi icantly reduced hairy roots in T2-T5 under selection pressure compared to that in T1 without any selection pressure, suggesting that cotransformation might has not always occurred. Crane et al. [23], also reported that not all hairy roots were the result of a co-transformation event. Again, Collier et al. [29], reported that between 20 and 60% of hairy roots produced on shoots inoculated with a wild-type A. rhizogenes containing a binary plasmid will contain both hairyroot-inducing and binary T-DNA. Considering these, the generated hairy roots were then tested for GUS expression to screen for genomic integration of the T-DNA from the pBI121 vector. Some hairy roots showed highly variable (very weak to strong) GUS expression ( Figure 4) throughout the whole root; some showed very low levels only at the root tip, and others did not show any expression, even under variable selection pressures. The probable cause is that plant transformation does not always result in ef icient transgene expression due to integration of T-DNA into transcriptionally silent regions of the plant genome [30]. These results suggest that luorescent markers could be a better choice as an alternative to antibiotic selection, which seems to be more reliable than kanamycin resistance for screening the transgenic hairy roots where the transgene is ef iciently expressed.

CONCLUSION
In this study, wild-type A. rhizogenes ATCC 15834 was successfully transformed with the binary vector pBI121, but compared to the wild-type, the transformed Agrobacterium showed reduced virulence. Hairy root generation ef iciency of the transformed Agrobacterium was highest without any kanamycin in the media and lowest with 50 mg·L -1 kanamycin. Some hairy roots showed GUS expression, con irming the genomic integration of transcriptionally active T-DNA, althought this was not always the case. Results of this study also suggested that luorescent markers could be a better alternative to antibiotic selection, which seems to be more reliable than kanamycin resistance for screening the transgenic hairy roots with ef icient transgene expression. Figure 3: Effect of Agrobacterium type and kanamycin concentration of the selection media on the generation of hairy roots from tobacco leaf explants. T0, Wild-type Agrobacterium + 0 mg·L −1 kanamycin; T1, transformed Agrobacterium+0mg·L -1 kanamycin; T2, transformed Agrobacterium +12.5 mg·L -1 kanamycin; T3, transformed Agrobacterium+25.0 mg·L -1 kanamycin; T4, transformed Agrobacterium +37.5mg·L -1 kanamycin; T5, transformed Agrobacterium+50 mg·L -1 kanamycin. Data show the mean±SD (n=3). Statistical signifi cance was analyzed using ANOVA. Means followed by the same letter are not signifi cantly different (p>0.01).