An aerotaxis receptor influences invasion of Agrobacterium tumefaciens into its host

Bacterial strains, plasmids, media, and growth conditions

All Escherichia coli and Agrobacterium tumefaciens strains and plasmids used in this study are described in Table S1. E. coli strains were grown in the lysogeny broth (LB) medium with or without 1.5% agar at 37 °C or 25 °C. A. tumefaciens C58 and its derivatives were grown in MG/L medium or AB-sucrose medium at 28 °C (Cangelosi et al., 1991; Gelvin, 2006). The final concentrations of antibiotics: for E. coli, ampicillin (Ap) at 100 µg/mL, kanamycin (Km) at 50 µg/mL, carbenicillin (Cr) at 25 µg/mL, tetracycline (Tc) at 15 µg/mL, chloramphenicol (Cm) at 34 µg/mL; for A. tumefaciens, kanamycin (Km) at 100 µg/mL.

In silico sequence analysis and phylogenetic analysis

All sequences utilized in the current investigation were acquired from the National Center for Biotechnology Information (NCBI) database. Likewise, a protein BLAST analysis was employed for the examination of the sequences. The RefSeq Select proteins database was utilized to search for proteins with homology. The Pfam database was utilized to analyze the sequence of Atu1027. Amino acid sequence alignments were conducted using DNAman8 or ClusterW software. The transmembrane sequence of Atu1027 was analyzed using the online tools DeepTMHMM (https://dtu.biolib.com/DeepTMHMM) and ProtScale (https://web.expasy.org/protscale/). Evolutionary analyses were conducted in MEGA11 (Tamura et al., 2021). The evolutionary history was inferred using the Neighbor-Joining method (Saitou & Nei, 1987). The bootstrap consensus tree inferred from 1,000 replicates was taken to represent the evolutionary history of the taxa analyzed (Felsenstein, 1985). Branches corresponding to partitions reproduced in less than 50% bootstrap replicates were collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) was shown next to the branches (Felsenstein, 1985). The evolutionary distances were computed using the JTT matrix-based method (Jones, Taylor & Thornton, 1992) and were in the units of the number of amino acid substitutions per site. This analysis involved 49 amino acid sequences. All positions with less than 50% site coverage were eliminated. There was a total of 502 positions in the final dataset.

DNA manipulations

Molecular manipulations of DNA were conducted according to standard molecular procedures. All primers used in this study are described in Table S2. Purification of PCR products and DNA fragments from agarose gels was performed by using FastPure Gel DNA Extraction Mini Kit (Vazyme Biotech, Nanjing, China). Plasmid DNA was purified with the TIANprep Mini Plasmids Kit (Tiangen Biotech, Beijing, China). According to the Inoue protocol, the competent cells of E. coli were prepared (Inoue, Nojima & Okayama, 1990). The plasmid DNA was transferred by heat shock into E. coli competent cells (Sambrook, Fritsch & Maniatis, 1989). The competent cells of A. tumefaciens were prepared as previously described (Cangelosi et al., 1991). All plasmids were electrotransformed into A. tumefaciens strains using an Eppendorf Eporator (Eppendorf AG, Hamburg, Germany).

Construction of deletion, complementation, or point mutants of A. tumefaciens

Following the principle of homologous recombination, the precise deletion of the atu1027 gene in A. tumefaciens C58 was constructed by the procedures as previously described (Guo, Zhu & Gao, 2009). The gene-deletion strategy was shown in Fig. S1. Briefly, a fusion of 585-bp upstream and 587-bp downstream of the atu1027 open reading frame (ORF) was inserted into the suicide plasmid pEX18Km via the Bam H I and Eco R I restriction sites, resulting in the creation of pEX-d1027. In pEX18Km and its derived plasmids, the kanamycin-resistant gene nptIII was used as a positive selection marker, the suicide gene sacB was used as a counterselection marker (Huang et al., 2018). Subsequently, pEX-d1027 was introduced into A. tumefaciens C58. Two rounds of selection were conducted to identify potential gene-deficient mutants. The colonies were further subjected to PCR screening and DNA sequencing for verification purposes. The avirulent mutant C58 ΔvirA was constructed using the same strategy. The complemented strain of the atu1027-deficient mutant was implemented by the introduction of the plasmid pCB301-1027. The plasmid pCB301-1027 carries a 479-bp promoter region and a 1500-bp intact ORF of the atu1027 gene.

To substitute the conserved histidine residue at position 100 within the globin region of Atu1027 with alanine, two single-residue mutation primers (cp1027-H100A-2 and cp1027-H100A-3) were designed (Table S2). Two primers for amplifying the atu1027 expression cassette, in conjunction with the two single-residue mutation primers, were utilized to procure two DNA fragments. The specific site of the two base mutations is situated within the region of overlap between the two aforementioned DNA fragments. Using overlap PCR, the atu1027 expression cassette containing the two base mutation sites was generated and subsequently inserted into the plasmid pCB301 via the Bam H I and Eco R I restriction sites. The recombinant plasmid pCB-1027H100A was then introduced into the atu1027-deficient mutant via electroporation.

Bacterial two-hybrid assay

The bacterial two-hybrid system (Stratagene, Agilent Technologies Inc., Santa Clara, CA, USA) was used to detect potential protein–protein interactions. The ORF of atu1027 was amplified and inserted into the bait plasmid pBT to express λcI-Atu1027 fusion protein. The ORFs of cheW₁ and cheW₂ were amplified and inserted into the target plasmid pTRG to express CheW1–RNAP and CheW2–RNAP fusion proteins, respectively. Then the recombinant target plasmid and bait plasmid were collectively transferred into E. coli XL1-Blue MR competent cells by heat shock and spread on LB-CTCK plates for screening positive clones. The LB-CTCK medium is made up of LB medium, 25 µg/mL carbenicillin, 15 µg/mL tetracycline, 34 µg/mL chloramphenicol, and 50 µg/mL kanamycin. The positive clones in the LB-CTCK plates were picked out and streaked onto a new LB-TCK plate that contains 15 µg/mL tetracycline, 34 µg/mL chloramphenicol, 50 µg/mL kanamycin, 80 µg/mL X-gal, and 0.2 mM galactosidase inhibitor. The LB-TCK plate with positive clones was placed in a lightless biochemical incubator at 37 °C for 16 h. The galactosidase activity of different groups was quantified as previously described (Ye et al., 2021).

Protein expression, pull-down assay, and immunoblotting

E. coli BL21 (DE3) was employed as the host organism for the overproduction of Atu1027-His, Atu1027H100A-His, CheW1, CheW2. E. coli cells containing pET-30a and its derived recombined plasmids were cultivated until reaching a cell density of approximately 5 × 10⁸ cell/mL. Subsequently, isopropylthio-β-D-galactoside (IPTG) was added to the cultures at a final concentration of 0.2 mM to induce the expression of the fusion proteins. After being cultivated at 25 °C for 6 h, the cells were centrifuged 5,000 × g for 10 min, washed twice and resuspended in PBS buffer (10 mM, pH 7.4). The pull-down assay was conducted following the methodology described in a previous study (Huang et al., 2018). The cell suspension was subjected to sonication until it reached a state of near clarity, and the supernatant from this sonicated cell suspension was collected through centrifugation at a force of 12,000 × g for a duration of 30 min. Equal amounts of supernatant that from different BL21(DE3) cell lysates were then respectively introduced into one mL Ni-IDA Resins (Sangon Biotech, Shanghai, China). The supernatant obtained from the lysates of BL21(DE3) containing pET30a was used as the negative control. The resins were subjected to five washes with PBS buffer and subsequently incubated with two mL of cell crude extract derived from a 50 mL cell culture that overproduced either CheW1 or CheW2. Following a 1-hour incubation at 4 °C with gentle shaking, the resins were washed with 10 volumes of PBS buffer containing increasing concentrations of imidazole (10 mM, 30 mM, and 50 mM) to remove nonspecifically bound proteins. The specifically bound proteins were then eluted from the resins using 500 µL of PBS buffer containing 250 mM imidazole. 5 µL of eluent obtained from distinct samples were subjected to SDS-PAGE. For immunoblotting, proteins were transferred onto hydrophilic polyvinylidene difluoride membranes (Merck Millipore, Billerica, MA, USA) through electrophoresis. The visualization of the transferred proteins was achieved using the BCIP/NBT alkaline phosphatase color development kit (Beyotime Biotechnology, Shanghai, China), following the recommended protocol provided by the manufacturer. To obtain antibodies capable of distinguishing between two CheWs, peptide fragments corresponding to the variable sequences of CheW1 (amino acid residues 142-155) and CheW2 (amino acid residues 146-159) were synthesized artificially and utilized as antigens for antibody generation in New Zealand rabbits (Huang et al., 2018).

Air trap assay for aerotaxis

The procedure for the air trap assay was followed as described by Arrebola & Cazorla (2020) with some modifications (Arrebola & Cazorla, 2020). To establish an air trap, a modified Pasteur pipette was initially placed within a glass tube, covered with a lid, and subjected to autoclaving. Next, sterile AB-sucrose medium containing 0.1% agar was heated and carefully poured into the glass tube. The liquid level of the medium was maintained three cm below the top of the Pasteur pipette. Following this, two mL of paraffin oil was added to the glass tube but was prevented from falling into the Pasteur pipette. Consequently, the paraffin oil acted as a barrier, effectively isolating the external medium of the Pasteur pipette from the surrounding air and creating a localized hypoxic environment. Only the medium contained within the Pasteur pipette was able to interact with the air and undergo air exchange. Subsequently, A. tumefaciens strains that grew to the middle-logarithmic growth phase were gently centrifuged at 3,500 × g for 3 min and resuspended in fresh AB-sucrose medium to achieve an OD₆₀₀ of 0.6. 100 µL of bacterial cells were slowly injected into AB-sucrose medium with a disposable sterilized syringe but outside of the Pasteur pipette. The inoculation site was positioned two cm above the base of the Pasteur pipette. If the bacteria possess oxygen-sensing capabilities, they would exhibit movement toward the interior of the Pasteur pipette. Finally, the air trap devices were placed in a biochemical incubator without any disturbance and observed every 12 h, until the bacterial population was visible.

Aerotaxis behavior observation

The aerotaxis behavior observation was conducted as previously described with some modifications (Arrebola & Cazorla, 2020). A. tumefaciens strains were grown in AB-sucrose medium to middle-logarithmic growth phase, centrifuged at 3,500 × g for 3 min, and suspended at the same cell density (1 ×10⁸ cfu/mL) in chemotaxis buffer (10 mmol/L KH₂PO₄, 0.1 mmol/L EDTA, pH 7.0) with 14.6 mM sucrose as the energy source. 3 µL drop of bacterial cells was deposited on a coverslip, the coverslip was inverted on an excavated slide. As a result, the drop was surrounded by air. The bacterial behavior was monitored by optical microscopy at 1,000 × magnification. The photographs of bacterial behavior at the edges of the droplets were taken after 5, 10, and 15 min. To prevent bacterial motility from being orientated by the light during the photo interval, the microscope light was turned off.

Chemotaxis assays

Chemotaxis to different carbon sources in a spatial gradient was assayed in the soft agar plates containing AB-medium, 0.2% Bacto agar (Beyotime Biotech, Shanghai, China), and 15 mM single carbon source. The carbon sources included mannitol, xylose, fructose, sucrose, glucose, and galactose. A. tumefaciens cells were grown and collected as indicated above for aerotaxis. The tested cell cultures were normalized to an OD₆₀₀ of 0.6. 3 µL of tested cell suspension was inoculated into the soft agar plates and incubated for 36 h without disturbance. The diameter of the swimming halo was measured and used to quantify the chemotaxis to different carbon sources.

Biofilm formation assays

A. tumefaciens cells in the middle-logarithmic growth phase were collected and normalized to an OD₆₀₀ of 0.1. two mL of culture was added to a sterile glass tube. The tubes were incubated for 72 h. For crystal violet (CV) visualization, the tube was washed thoroughly with double-distilled H₂O and stained with 0.15% (wt/vol) crystal violet staining solution for 15 min. After the staining procedure, the tubes were washed five times with double-distilled H₂O to remove the crystal violet staining solution. The formation of purple ring-like structures on the tube was observed. Finally, the crystal violet is eluted off with 30% acetic acid (vol/vol). The eluent is measured at 570 nm to determine the strength of biofilm formation. A normalization of absorbance values to culture growth was achieved by dividing the solubilized CV OD_nm value by the planktonic cell OD₆₀₀ value. All strains were performed in three replicates.

Tumorigenesis assays

The tumorigenesis assays on the leaves of the Kalanchoe plant were implemented as described previously with some modifications (Yang et al., 2020). Different A. tumefaciens strains were prepared as in the air trap assay described previously. An avirulent strain C58 ΔvirA was used as a negative control. The agrobacterial cells were resuspended in AB-sucrose liquid medium at 5 ×10⁸ cfu/mL for inoculation. Kalanchoe plants’ leaves that grew for more than three weeks were selected for the infection. Before inoculation, several wound lines were made in the Kalanchoe leaf by using a sterile scalpel. Each wound line was inoculated with 3 µL of different cell suspensions. The infected Kalanchoe plants were grown in the pots at room temperature for 30–40 days. The tumours in the wound lines were photographed and weighed. Each inoculation was repeated at least three times on separate leaves.

Statistical analysis

The data were presented as the mean ± standard error of three replicates using analysis of variance (ANOVA). To assess the significance of differences between various groups, the Student-Newman-Keuls (SNK) multiple comparison test was employed at a significance level of P < 0.05.

Atu1027 is a cytoplasmic chemoreceptor containing a globin domain

The globin domain of the heme-containing transducers (HemAT-Bs from Bacillus subtilis and HemAT-Hs from Halobacterium salinarum) can bind diatomic oxygen and mediate aerotactic reactions in bacteria and archaea (Hou et al., 2000; Hou et al., 2001a). A. tumefaciens C58 has 20 genes coding putative chemoreceptors on its genome (Wood et al., 2001). Among these genes, atu1027, which is located on the circular chromosome, is predicted to encode a chemoreceptor (accession number: WP_010971342.1) with a globin domain (Fig. 1A). The Pfam database revealed that the peptide chain of the globin domain within Atu1027 encompasses 159 amino acid residues. A prior investigation proposed that the globin domains of Atu1027, HemAT-Bs, and HemAT-Hs evolved from a shared ancestral globin (Freitas et al., 2005). In the present study, amino acid sequence alignment (Fig. 1B) indicated that the globin domain of Atu1027 shares 17.65% sequence identity with that of HemAT-Bs and 18.64% sequence identity with that of HemAT-Hs. The alignment of the sequence also demonstrated that the globin domain of Atu1027 retains two strictly conserved residues present in all globin domains (Hou et al., 2001a). In addition, the analysis conducted using DeepTMHMM and ProtScale indicated that Atu1027 does not possess a transmembrane domain. These findings suggested that Atu1027 potentially functions as a cytoplasmic chemoreceptor involved in mediating aerotaxis.

The initial comprehensive alignment of the complete sequence indicated a significant degree of conservation of Atu1027 across various Agrobacterium species (Fig. S2). Consequently, a more extensive alignment of the entire sequence and subsequent construction of a phylogenetic tree were undertaken to investigate the evolutionary lineage of Atu1027. A total of 48 sequences, exhibiting a sequence identity of over 50% with Atu1027, were carefully selected for the construction of the phylogenetic tree. Figure 2 demonstrates the division of 49 proteins from 11 bacterial genera into three distinct groups. Notably, the clustering of proteins within the same bacterial genus is readily apparent. These proteins primarily originate from Agrobacterium, Shinella, Neorhizobium, and Ciceribacter. However, the proteins derived from Rhizobium in the tree were not effectively clustered. Notably, the protein WP 082184016.1 of Rhizobium oryzihabitans and the protein WP 045017035.1 of Rhizobium nepotum were grouped together in group 2, alongside Atu1027 and five other proteins from Agrobacterium. Furthermore, the protein comparison revealed that WP 082184016.1 and WP 045017035.1 exhibit a higher degree of similarity with Atu1027. This suggests that these two proteins may potentially possess similar functional characteristics to Atu1027.

Atu1027 interacts with CheW1 and CheW2

In the chemotaxis system of E. coli, the formation of the ternary signaling complex relies on the interaction of chemoreceptors and the unique coupling protein CheW (Parkinson, Hazelbauer & Falke, 2015). A. tumefaciens C58 two coupling proteins, namely CheW1 and CheW2 (Huang et al., 2018; Wood et al., 2001). In light of the hypothesis that Atu1027 serves as a cytoplasmic chemoreceptor, it is crucial to validate the potential interaction between Atu1027 and two CheWs. To assess the protein-protein interactions, the bacterial two-hybrid and in vitro His-tag pull-down assays were employed. In the bacterial two-hybrid assay, the presence of blue colonies on the designated plates serves as an indication of the interaction between the target protein and the bait protein. The intensity of the blue color corresponds to the strength of the protein-protein interaction and the level of β-galactosidase activity. As illustrated in Figs. 3A and 3B, the experimental groups (CheW1+Atu1027 and CheW2+Atu1027) displayed a higher β-galactosidase activity compared to the negative control group. In the His-tag pull-down assay (Fig. 3C), both CheW1 and CheW2 were identified in the eluate from the resin bound with Atu1027-His, while they were absent in the negative control. This outcome serves as confirmation that both CheW1 and CheW2 can interact with Atu1027 in an in vitro setting. Consequently, the results indicated that Atu1027 interacts with either CheW1 or CheW2.

Atu1027 is an aerotaxis receptor in A. tumefaciens C58

To ascertain the functions of Atu1027, a precise deletion of atu1027 in A. tumefaciens C58 was accomplished via homologous recombination, as outlined in the Materials and Methods section. The resulting atu1027-deficient mutant was denoted as C58 Δ1027. Subsequently, the atu1027 gene, along with its native promoter (located 479 bp upstream of atu1027), was reintroduced into the C58 Δ1027 mutant using a plasmid, thereby restoring the expression of Atu1027 protein in C58 Δ1027. This complemented strain was designated as C58 Δ1027+. The bioinformatic analysis and protein interaction experiments provided indications that Atu1027 could potentially function as an aerotaxis receptor in A. tumefaciens. To validate this hypothesis, the aerotaxis of various A. tumefaciens strains was assessed using air trap assays and optical microscope observations.

The results of the air trap assays are presented in Fig. 4. At 72 h post-inoculation, the device containing the complemented strain C58 Δ1027+ exhibited a significant aggregation of bacterial cells on the surface of the semi-solid medium in the Pasteur pipette. In contrast, no such phenomenon was observed in the device containing the wild-type C58 or the atu1027-deficient mutant C58 Δ1027. At 96 h post-inoculation, the wild-type C58 cells displayed visible aggregation on the surface of the semi-solid medium, while the medium inoculated with C58 Δ1027 remained clear in the Pasteur pipette. At 120 h post-inoculation, the medium inoculated with C58 Δ1027 remained clear on the surface of the medium in the Pasteur pipette. The results above indicated that the deletion of atu1027 eliminates the aerotactic response of A. tumefaciens to atmospheric air. Furthermore, we conducted growth curve experiments in AB-sucrose liquid medium to determine if the deletion of atu1027 affects the growth rate of A. tumefaciens. The growth rates of C58 Δ1027 and C58 Δ1027+ were comparable to that of the wild-type C58 (Fig. S4), and the difference was not statistically significant (P value > 0.05). Hence, the deletion of atu1027 did not affect the normal growth of A. tumefaciens.

Subsequently, an optical microscope was employed to perform continuous monitoring of the aerotaxis exhibited by the tested agrobacterial strains. The bacterial cells were deposited onto a coverslip, which was promptly inverted onto an excavated slide. To prevent air infiltration, Vaseline was applied to all four sides of the coverslip. The slide was then positioned on the objective table of the optical microscope and observed at 5-minute intervals. The findings are illustrated in Fig. 5. At 5 min post-inoculation, C58 and C58 Δ1027+ cells began to accumulate at the boundary of bacterial suspension and air. The quantity of C58 and C58 Δ1027+ cells was significantly higher compared to that of C58 Δ1027. At 10 and 15 min post-inoculation, there was further accumulation of C58 and C58 △1027+ cells at the boundary between the bacterial suspension and the air, while the number of C58 △1027 cells did not show a significant increase. These results are consistent with the results obtained from the air trap assays. Collectively, it can be concluded that Atu1027 is an aerotaxis receptor that modulates in the aerotaxis of A. tumefaciens under the test condition.

Atu1027 is involved in the chemotaxis of A. tumefaciens

In A. tumefaciens C58, CheW1 and CheW2 are involved in the formation of ternary signaling complexes at the cell poles (Huang et al., 2018). Atu1027 interacts with both CheW1 and CheW2, indicating its role as a component of the chemotaxis system in A. tumefaciens. The deletion of the atu1027 gene may have an impact on the chemotactic abilities of A. tumefaciens. Therefore, in order to assess the chemotaxis of A. tumefaciens, six common carbon sources (mannitol, xylose, fructose, sucrose, glucose, galactose) were tested on the swim agar plates. As shown in Fig. 6, in six different kinds of semi-solid AB-medium containing different sugars, the swimming halo of the atu1027-deficient mutant C58 Δ1027 is slightly smaller than that of wild-type C58, demonstrating that the deletion of atu1027 slightly reduced the chemotactic response to multiple carbon sources in A. tumefaciens. Additionally, when the atu1027 gene was re-introduced into the atu1027-deficient mutant C58 Δ1027, the swimming halo of the complemented strain C58 Δ1027+ was bigger than that of wild-type C58. This may be caused by an increased copy number of the atu1027 gene in the complemental strain C58 Δ1027+.

The replacement of His100 with alanine completely abolished the function of the globin domain of Atu1027

In the globin domains of HemAT-Bs and HemAT, the proximal histidine residue (His-123) plays a crucial role in both oxygen sensing and heme binding (Freitas et al., 2005; Hou et al., 2000; Hou et al., 2001a). Replacing the His-123 with alanine eliminates the function of the globin domains (Hou et al., 2001a). The globin domain of Atu1027 contains a histidine residue at position 100 (His100), which has been predicted to serve as the heme binding site (Fig. 1). Mutation on His100 may also affect the function of Atu1027. Therefore, to assess the importance of His100 for the function of Atu1027, a site-directed mutagenesis technique was employed to replace His100 with an alanine residue. The H100A mutant of Atu1027 was expressed in the C58 Δ1027 strain, resulting in the creation of the mutant strain C58 Δ1027H100A. Chemotaxis assay and air trap assay for aerotaxis were conducted to test the phenotype of the mutant strain C58 Δ1027H100A. As anticipated, the C58 Δ1027H100A strain exhibited a compromised chemotactic response to sucrose (Figs. 7A and 7B) and did not accumulate in the Pasteur pipette after 120 h of inoculation (Fig. 7C). These phenotypic characteristics of C58 Δ1027H100A were consistent with those of the parental strain C58 Δ1027. At the same time, mutation in His100 did not affect the interactions between Atu1027 and two CheWs (Fig. S5). These results indicated that His100 plays a crucial role in the functionality of the globin domain within Atu1027. Additionally, this finding also provides evidence to support the point that Atu1027 acts as an aerotaxis receptor.

Deletion of atu1027 impaired the biofilm formation of A. tumefaciens in static culture

Biofilm formation is an integral aspect of the interactions between plants and microbes, playing a critical role in the establishment of infections (Bogino et al., 2013; Danhorn & Fuqua, 2007). In the soilborne pathogen Ralstonia solanacearum, the normal formation of biofilms necessitates the involvement of aerotaxis (Yao & Allen, 2007). Atu1027 serves as an aerotaxis receptor to modulate chemotaxis in A. tumefaciens, the deletion of atu1027 may impair the biofilm formation of A. tumefaciens. Hence, the biofilm formation assays of atu1027 mutants were performed in static culture. As shown in Fig. 8, at 72 h post-inoculation, all strains under examination possess the ability to generate biofilms. However, it is noteworthy that the biofilms produced by strains C58 and C58 Δ1027+ exhibit considerably greater strength compared to those formed by C58 Δ1027. This observation serves to validate the notion that the absence of Atu1027 diminished the biofilm formation capacity of A. tumefaciens in static culture.

atu1027 is required for the full pathogenicity of A. tumefaciens

The deletion of the atu1027 gene impairs not only chemotaxis and aerotaxis of A. tumefaciens but also its ability to form biofilm in static culture. Nevertheless, it remains unknown the impact of deletion of atu1027 on the pathogenicity of A. tumefaciens. To investigate the potential effect of atu1027 deficiency on the pathogenicity of A. tumefaciens, three different strains were used: the wild type C58, the atu1027-deficient mutant C58 Δ1027, and the complemented strain C58 Δ1027+. These strains were individually inoculated onto the leaves of Kalanchoe plants. All tested strains were precultured in AB-sucrose liquid medium to mid-log phase, normalized to the same cell density, and inoculated on the Kalanchoe plants’ leaves. Based on the tumour photograph presented in Fig. 9A, it can be observed that, except the avirulent strain C58 ΔvirA, the other three tested strains were capable of inducing tumours on the wound sites of Kalanchoe leaves after 30 days of inoculation. However, the tumour induced by the C58 Δ1027 was smaller compared to those induced by the C58 and C58 Δ1027+. The data presented in Fig. 9B further supports this observation, as it demonstrated that the tumour weight induced by the strain lacking atu1027 was significantly lower than that induced by the other tested strains. Hence, it can thus be concluded that deletion of atu1027 gene attenuates the pathogenicity of A. tumefaciens.