Mapping of the Interaction Between Agrobacterium tumefaciens and Vanda Kasem’s Delight Orchid Protocorm-Like Bodies

Abstract

Physical contact between A. tumefaciens and the target plant cell walls is essential to transfer and integrate the transgene to introduce a novel trait. Chemotaxis response and attachment of Agrobacterium towards Vanda Kasem’s Delight (VKD) protocorm-like bodies (PLBs) were studied to analyse the interaction between Agrobacterium and PLB during the transformation event. The study shows that initially A. tumefaciens reversibly attached to PLB surface via polar and lateral mode of adherence followed by the irreversible attachment which involved the production of cellulosic fibril by A. tumefaciens. Cellulosic fibril allows formation of biofilm at the tip of trichome. Contrarily, attachment mutant Escherichia coli strain DH5α was significantly deficient in the attachment process. Spectrophotometric GUS assay showed the mean value of attachment by A. tumefaciens was 8.72 % compared to the negative control E. coli strain DH5α that produced 0.16 %. A. tumefaciens swarmed with sharper and brighter edge when severe wounding was applied to the PLBs producing the highest swarming ratio of 1.46 demonstrating the positive effect of the plant exudates on bacterial movement. The study shows that VKD’s PLBs are the suitable explants for Agrobacterium-mediated transformation since the bacteria expressed higher competency rate.

Electronic supplementary material

The online version of this article (doi:10.1007/s12088-015-0519-7) contains supplementary material, which is available to authorized users.

Introduction

Agrobacterium tumefaciens is widely utilized for delivering foreign gene into a plant genome because it offers the advantage of simple methodology that enhances specific and stable transgene integration [1, 2]. In addition, usage of various target explants, high efficiency of transformation and a high percentage of single T-DNA insertion make Agrobacterium-meditated transformation the widely used method for gene transfer [3].

Initial step in the Agrobacterium-mediated genetic transformation of crops will be the target tissue-Agrobacterium interaction. Chemotactically attracted to and attachment of A. tumefaciens to target plant cell walls is essential to transfer and integrate the transgene to introduce a novel property. Bacterial chemotaxis is the response by bacteria to actively modulate their direction of movement towards gradients of chemoattractants (positive chemotaxis) or away from chemorepellents (negative chemotaxis).

Protocorm-like bodies (PLBs) are the commonly used explants for Agrobacterium-mediated transformation studies. This is partly due to their ability to proliferate rapidly [4]. Furthermore, PLBs can regenerate into complete plantlets in plant growth regulator free medium due to their bipolar nature which develop into shoot and root meristems [5].

The aim of the study were to quantify and analyse attachment of A. tumefaciens strain LBA4404 to PLBs and study the chemotactic response of A. tumefaciens strain LBA4404 towards plant exudates from PLBs.

Materials and Methods

Plant Materials

In vitro culture of protocorm-like bodies (PLBs) of Vanda Kasem’s Delight (VKD) (Supplementary Fig. 1) were maintained on modified Vacin and Went medium [6] with 15 % coconut water and 30 % tomato extract. The pH of Vacin and Went media in this study was adjusted from 4.8 to 5.0 (CyberScan PC 510 pH/mV/Conductivity/TDS/°C/°F Bench Meter, Eutech 73 Instruments, Singapore) prior to autoclaving (STURDY SA-300VFA-F-A505, Sturdy Industrial Co. Ltd, Taiwan). The culture was incubated at 25 °C under 16 h semi-photoperiod with cool white fluorescent light (supplied by Philips TLD fluorescent light tubes of 36 W, 150 μmol m⁻² s⁻¹). The PLBs were then subcultured for every 4 weeks on modified Vacin and Went medium supplemented with 15 % coconut water and 30 % tomato extract to produce large quantities of explants for transformation. Healthy, greenish and rapidly growing PLBs were used for transformation.

Bacterial Strains and Plasmid Constructs

A. tumefaciens strain LBA4404 harbouring disarmed plasmid pCAMBIA 1304 plasmid with gusA and nptII genes) was used for attachment, quantification and chemotaxis studies. Plasmid pCAMBIA 1304 was provided by Dr. Richard Bretell from CSIRO, Australia. Escherichia coli strain DH5α harbouring pMRC 1301 plasmid consisting of gusA and nptII genes was used as a control throughout the experiment since it is attachment deficient strain.

Inoculation and Co-cultivation with A. tumefaciens

Three mL cultures from the glycerol stock were transferred into Luria–Bertani (LB) broth containing 50 mg/L kanamycin and incubated at 28 °C and 120 rpm overnight. Bacterial suspensions were then streaked on solid LB medium containing 50 mg/L kanamycin using inoculation loop and incubated at 28 °C for 2–3 days until a visible single colony forms. Single colony were then inoculated to LB broth containing 50 mg/L kanamycin and incubated at 28 °C and 120 rpm for 16 h to reach an optimal density of 0.9 units at 600 nm (OD_600nm).

Microscopical Study of Bacterial Attachment

Microscopical analysis was carried out to validate the competence of A. tumefaciens strain LBA4404 to adhere onto PLBs. Attachment mutant E. coli strain DH5α (pMRC 1301) was included as negative control in all the conducted experiments. Grown bacterial cultures were pelleted (3400×g for 10 min) and resuspended in 25 mM phosphate buffer (pH 7.5) to infect PLBs. To study bacterial attachment to PLBs, the explants were co-cultivated with 250 μL of bacterial suspension and incubated in a rotary shaker at 28 °C at 25 rpm for 2 h. After this period, unbound bacteria were removed by washing the PLB once with 1 mL fresh buffer and gently tapping the bottom of the microtube several times to discard unattached bacteria. Explants were then directly observed using optical microscopy or freeze dried for observation using scanning electron microscope (SEM).

Quantification of Bacterial Attachment to VKD’s PLBs Via Spectrophotometric Measurement of GUS Expression

Quantification of bacterial attachment was carried out to confirm the binding capacity of A. tumefaciens strain LBA4404 according to the pre established method [7]. Attachment mutant Escherichia coli strain DH5α (pMRC 1301) was maintained as described above. PLBs were maintained in 1 mL of 25 mM phosphate buffer (pH 7.5) in 1.5 mL microtubes. For infection, the microtubes were loaded with 50 μL aliquots of buffer suspended bacteria. Microtubes were then incubated in a rotary shaker at 28 °C at 25 rpm for 2 h. After this period, unbound bacteria were removed by washing the PLBs once with 1 mL fresh buffer and vortexed 30 s each time to discard unattached bacteria. β-glucuronidase activity in the samples was measured following the method outlined previously [8]. Washed explants were transferred to 1 mL of extraction buffer (50 mM sodium phosphate (pH 7), 10 mM dithiothreitol, 1 mM sodium EDTA, 0.1 % (v/v) sodium lauryl sarcosine, 0.1 % (v/v) triton X-100), vortexed and incubated at 37 °C for 10 min. The GUS enzyme substrate p-nitrophenyl β-D-glucuronide was added at a final concentration of 1 mM. After incubation at 37 °C for 30 min reactions were stopped by the addition of 400 µL of 400 mM Na₂CO₃ solution. GUS activity was quantified by measuring light absorbance at 415 nm in spectrophotometer (Hitachi U-1900 UV/VIS, Japan) [9]. Absorbance A₄₁₅ was also measured from PLBs-containing uninfected microtubes to determine light absorption by plant release compounds, as well as from inoculated microtubes in the absence of PLBs to measure the total enzymatic activity in the inoculum used for infections. Finally, the percentage of inoculated bacteria that remained attached to the different tissues was calculated using the formula:

$$Attachment\left(\%\right)=\frac{X-Y}{Z}\times 100\%$$

where by the variables are the A₄₁₅ values corresponding to infected tissues (X), uninfected tissues (Y), and total bacterial inoculum (Z) for each individual combination of explants type and bacterial strain.

Chemotaxis Assay

Chemotaxis assays were carried out based on modified swarm agar plate method [10]. A sterile inoculation loop was used to inoculate bacteria in the middle of a Petri dish (9 cm diameter) containing chemotactic media (CM: 10 mM phosphate buffer, pH 7.0; 1 mM ammonium sulfate; 1 mM magnesium sulfate; 0.1 mM potassium-EDTA) partially solidified with 0.5 % (w/v) nutritional agar. Some treatments were used for this assay including, intact PLBs, mild wounding using a needle, severe wounding using a scalpel and PLBs excised into small pieces.

Chemotaxis was quantified after 24, 48 and 72 h of incubation at 28 °C. The swarming distances from the point of bacterial inoculation towards (T mm) and backwards (B mm) from the sources of tissue exudates were measured and used to obtain a ratio (R) of the bacterial movement using the following formula:

$$R=\frac{T}{B}$$

Thus, R values over or under 1.00 represent positive or negative chemotaxis, respectively. Data were analysed using one way ANOVA and the differences contrasted using Tukey’s multiple comparison test.

Results

Microscopical Studies of Bacterial Attachment

SEM observations of inoculated PLBs showed four predominant ways of A. tumefaciens colonization on the PLB surface. A. tumefaciens can be seen polarly (Supplementary Fig. 3) or laterally (Supplementary Fig. 4) adhered to the PLB surface to form bigger clusters. In contrast, no bacterium was observed on the surface of PLB co-cultivated with attachment-mutant E. coli strain DH5α (Supplementary Fig. 2). Attachment prone sites including crevices, stomata (Supplementary Fig. 2b) and trichome (Supplementary Fig. 2c) were free of infection.

One of the dominant modes of A. tumefaciens attachment was via a single pole of bacteria in contact with the PLBs surface (Supplementary Fig. 3). Several individual Agrobacterium cells polarly bound to the surface of PLB especially on the side of trichome. No cellulosic fiber was seen to hold the Agrobacterium cells (Supplementary Fig. 3). This indicates that Agrobacterium cells flagellated at the opposite pole for swimming to allow motility while the other pole is bold to permit attachment to surfaces. During polar mode of attachment, Agrobacterium cells will be positioned vertically as though they are standing on the PLBs surface.

However, A. tumefaciens cells were also laterally adhered along the stretch of trichome (Supplementary Fig. 4). Lateral attachment requires A. tumefaciens cells to adhere horizontally parallel to the PLB surface (Supplementary Fig. 4) suggesting that there are multiple points of bacterial attachment to the host cells. Unlike polar attachment, lateral attachment makes the whole surface of bacteria in contact with PLB.

In addition, SEM showed that bacteria were surrounded by cellulosic fibrillar material (Fig. 1). The fibrillar-material was probably produced by the bacteria attached to the PLB as it was not noticed on PLBs infected with attachment mutant E. coli strain DH5α (Supplementary Fig. 2). Arrows on Fig. 1a indicate the fibril at the sideway of bacteria, suggesting it is a bacterial product. Cellulosic fibrillar net aids massive bacterial cluster formation by entrapping more bacteria cells into the cluster (Fig. 1b). This allows planktonic cells to indirectly get attached to the PLB. Moreover, cellulosic fibrils twisted to form larger filaments that hold the cluster of bacteria stronger in place (Fig. 1c). Getting trapped into bacteria cluster prevents Agrobacterium from being washed away during mechanical forces such as tapping, washing or drying procedures. Extensive colonization aided by cellulosic fibril was profoundly observed at wounded parts of PLB (Fig. 1). Clusters of fibril producing bacteria formed at the wounded part allow successful infection of PLBs. PLBs surface wounded during explant preparation will produce phenol compounds such as acetosyringone that acts as chemoattractant to A.tumefaciens. Thus, large size of bacterial clusters is centered at wounded part of PLBs.

Fig. 1.

Formation of cellulosic fibrillar during bacterial attachment. a Arrows indicate the synthesis of fibril at the sideway of A. tumefaciens; b Formation of a larger A. tumefaciens cluster; c Arrows indicate intertwined cellulosic fiber

Similarly, the progress in the development of Agrobacterium clusters was observed on the tip of trichome via formation of a cap-like structure. Phase contrast studies of trichome confirmed that this area was extensively colonized by bacteria (Supplementary Fig. 5). A. tumefaciens adhered to the perimeter of tip of trichome on PLBs appears as thick black lining (Supplementary Fig. 5a). Massive aggregation of A. tumefaciens on the trichome forms a cap-like structure covering the tip of trichome of PLB (Supplementary Fig. 5b). Higher magnification of trichome on Supplementary Fig. 5c shows the substantial adhesion of individual planktonic bacteria in the progress of forming cap-like structure on the tip of trichome.

Quantification of Bacterial Attachment to VKD’s PLBs Via Spectrophotometric Measurement of GUS Expression

Besides the qualitative assessment of the A. tumefaciens binding aptitude to PLBs, bacterial attachment to PLBs was quantified through spectrophotometric measurement (A₄₁₅) of GUS expression to confirm and verify the ultrastructural results. The mean value of attachment of A. tumefaciens strain LBA4404 appeared to be higher (8.72 %) compared to the negative control (E. coli strain DH5α) (0.16 %) (Supplementary Fig. 6).

Chemotaxis Assay

Generally, A. tumefaciens strain LBA4404 exhibited strong and positive chemotactic response to wounded PLBs compared to intact PLBs (independently of the incubation period). The overall swarming ratio of A. tumefaciens tested in the presence of wounded VKD PLBs ranged between 1.03 and 1.46 (Fig. 2), indicating a positive effect of the plant wound exudates on bacterial movement. The study also proved that swarming of A. tumefaciens was significantly accelerated (p < 0.05) by severe wounding compared to mild wounding. For example, at 24 h of observation, swarming of Agrobacterium towards the severely and mildly wounded explants were 1.46 and 1.2 units respectively. Moreover, chemotactic response was on the peak in the first 24 h and then followed by a downfall at 48 and 72 h (Fig. 2). Such a pattern of chemotactic response at various incubation periods is similar for all the treatments and controls analysed.

Fig. 2.

Chemotaxis ratios of A. tumefaciens towards plants exudates from intact and wounded VKD’s PLBs and controls at various incubation period. Controls are filter papers dipped in chemotaxis medium (CM), chemotaxis medium supplemented with 10 mM glucose (Glu) and chemotaxis medium supplemented with 15 % (v/v) acetic acid (AA). Data were analyzed using one way ANOVA and the differences contrasted using Tukey’s multiple comparison test. Different letters indicate values are significantly different (p < 0.05)

Negative chemotactic response of Agrobacterium towards intact PLBs and filter papers soaked in chemotactic medium supplemented with acetic acid indicates that they have restricted the swarming of Agrobacterium (Fig. 2).

Similarly no significant difference in chemotactic response was observed between filter papers soaked in chemotactic medium supplemented with glucose and chemotactic medium alone. As for glucose, the response function is confirmed to be biphasic and perfectly adapted, that is, the chemotaxis response is neither positive nor negative. Similar properties were found for the filter paper soaked in chemotaxis medium alone (Fig. 2). Swarming ratios remained at 1.00 when glucose and chemotactic medium was used as negative control at all incubation period except for the latter at 72 h showed a negative response (Fig. 2). The strong negative response at longer incubation period could be caused by the reduced level of chemoattractants.

Discussion

VKD orchid was selected as a model system for this study to analyze the potential of this plant in terms of genetic modification. Furthermore, PLBs was chosen as the target material since it carries the similar characteristics such as rapid micropropagation and totipotent in nature.

Bacterial attachment to the host is necessary for transformation and it is mediated by chromosomally encoded Agrobacterium genes and not by the Ti plasmid [11]. Based on the results, failure of attachment mutant E. coli strain DH5α to adhere to the PLBs surface indicated a specific cell–cell contact were maintained by A. tumefaciens in order to establish stable irreversible binding.

Generally, bacteria explore surface for reversible attachment and subsequent detachment or stable attachment via polar mode of attachment. SEM analysis showed that single A. tumefaciens cells intensively attached in polar orientation to the PLBs (Supplementary Fig. 3). Polar mode of attachment is made possible by features that are localized at the cell poles of Agrobacterium. There are a number of structures that naturally confined to cell poles of Agrobacterium such as T-pilus [12, 13], protein comprising T4S system [14] and unipolar polysaccharides [15]. Vir proteins localized at single cell poles [14] concentrate at the tip of T-pilus in the Agrobacterium, which is an extracellular appendage [12, 13]. The vir proteins are responsible for the binding of the T-pillus to the receptors on the plant cell [13]. The attached pole is open to make contact with plant surface in order to transfer T-DNA. T-DNA transport apparatus (VirD4–VirB complex) assembled at the cell pore [16] located at the pole. Therefore, as the bacteria adhere to the plant surface via polar mode attachment, it forms a T-DNA transport apparatus at the contact side to safely transfer T-DNA without getting degraded by nucleases [17].

A. tumefaciens was also found to attach to PLBs predominantly by making lateral contacts along their sides. This suggests the probability of distribution of attachment related proteins around the perimeter of A. tumefaciens cells. Lateral attachment results in efficient contact with the PLB for the maximal transfer of the transgene. Lateral attachment may also classify as the reversible mode of attachment to initiate colonization. A. tumefaciens attachment is reported as a stepwise dynamic process where they observed pole- and subpole-attached A. tumefaciens in transition to become laterally attached or in transition from detaching away from the host cell following transfer of DNA and proteins [18].

Formation of a large cluster of A. tumefaciens to orchid PLB is a two step process. The initial step was the direct attachment of the single A. tumefaciens cell to the PLB surface. Once permanently attached, bacteria start to synthesize insoluble exopolysaccharide (EPS) fibrils that encase the adherent bacteria in a three-dimensional matrix [17]. Freely floating bacteria that entrapped by cellulosic fibrils [19, 20] into the 3 dimensional network (Fig. 1) will develop biofilm. Such biofilm was observed covering the tip of trichome forming a cap-like structure (Supplementary Fig. 5).

The cellulosic fibrils arose from the side of the virulent A. tumefaciens (Fig. 1a) attached to PLBs. This is proven by the absence of such fibers on PLBs infected by attachment-mutant E. coli strain DH5α (Supplementary Fig. 2). PLBs do not readily digest the fibrils because both the plant cell wall and microfibrils are made of cellulose. Microfibrils are frequently appeared twisted together into wide filaments (Fig. 1c).

The specificity of the cell–cell contact is demonstrated precisely by a quantitative measurement of the adherent capacity of attachment-competent A. tumefaciens strain LBA4404 to VKD PLBs (Supplementary Fig. 6). Competence of A. tumefaciens strain LBA4404 towards PLBs is comparable to that of attachment mutant E. coli strain DH5α. The binding affinity of A.tumefaciens strain LBA4404 towards PLBs is 54.5 fold higher than the attachment mutant E. coli strain DH5α. This supports the finding from microscopy analysis that PLBs is a suitable target explants for Agrobacterium-mediated transformation studies.

A. tumefaciens incubated on semisolid agar swarmed outward from the point of inoculation, in response to the gradient created by the diffused chemicals or plant wound exudates at the edges of the Petri plates. Visible bacterial swarming to naked eye allowed the quantification of chemotactic response of A. tumefaciens.

Negative chemotaxis response towards intact PLBs (Fig. 2) suggests that intact PLBs do not produce inducers or phenolics compounds that act as chemoattractant. Wounding is an integral step in Agrobacterium-mediated transformation as the wounded tissue often produces inducers of the T-DNA transfer process to allow the bacterium to infect the target tissue [21]. A. tumefaciens strain LBA4404 exhibited positive chemotactic response to wounded PLBs comparative to intact PLBs (Fig. 2) and severely wounded PLBs accelerated migration of A. tumefaciens compared to the mildly wounded PLBs (Fig. 2). The findings showed that severe wounding may enhance the accessibility of plant cells for Agrobacterium.

Besides, chemotactic response of both control and treatments were scored the highest in the first 24 h and then followed by a downfall at 48 and 72 h (Fig. 2). Probable stationary phase of Agrobacterium and toxicity due to phenolic exudates from wounded PLBs at extended incubation period [22] caused the downfall in terms of chemotactic response.

Negative chemotactic response of Agrobacterium towards filter papers soaked in chemotactic medium supplemented with 15 % acetic acid (Fig. 2) caused by the excessive acidic content. Acidic conditions repressed genes involved in motility, chemotaxis, and cellular metabolism and flagellin proteins [23]. Similarly, strong negative chemotactic response of A. tumefaciens towards filter paper soaked in chemotactic medium may caused by the irreversible effect of the medium itself on the chemotactic machinery of A. tumefaciens.

A. tumefaciens responded biphasically perfectly adapted towards filter paper soaked in glucose and chemotaxis medium alone (Fig. 2). Increasing extracellular hexoses cause a transient attractant response by increasing the rate of autophosphorylation to activate chemotaxis cascade of event [24]. Thus, allowing the bacteria to move towards the glucose to produce a biphasic chemotactic response.

Conclusion

It can be concluded that A. tumefaciens cells adopted a polar, lateral, and 3D cluster mode of attachment onto PLBs surfaces. Agrobacterium tumefaciens strain LBA4404 has demonstrated the specific cell–cell contact and stable binding capacities of attachment competent bacteria towards PLBs. Agrobacterium is chemotactically attracted to the exudates of PLBs. This suggests that there is no blocking step in Agrobacterium-mediated transformation of orchid plant which is naturally not a host plant. It is also notable that chemotactic response of A. tumefaciens to VKD’s PLB-released compound was always higher than to glucose control which is easily metabolizable form of carbon sources.