Function and regulation of pob genes for 4-hydroxybenzoate catabolism in Agrobacterium tumefaciens

ABSTRACT

Agrobacterium tumefaciens is a pathogen that causes tumors in plants. Phenolic acids present in the soil and rhizosphere may affect the interaction between A. tumefaciens and plants. An important pathway for microorganisms to degrade phenolic acids is the β-ketoadipate pathway, which has been annotated in the genome of A. tumefaciens. The ability of the PobA (atu4544) enzyme to catalyze the conversion of 4-hydroxybenzoate to protocatechuate was essential for cell growth using 4-hydroxybenzoate as the sole carbon source. The pobA gene is located upstream of atu4545, encoding an AraC transcription factor (PobR). Strains with deleted or supplemented atu4545 exhibited similar growth characteristics on common and phenolic acid-containing carbon sources as strains with deleted or supplemented atu4544. Strains with a pobA::lacZ reporter fusion showed that PobR induced pobA expression. In addition, the use of a pobR::lacZ reporter fusion showed that PobR represses its expression. Electromobility shift assay revealed that the PobR regulator can bind specifically to DNA. The binding site was identified as CGTGCGATGGTGGATT. Deletions of atu4544 (pobA) and atu4545 (pobR) decreased A. tumefaciens pathogenicity by infecting carrot roots and kalanchoe leaves, with no effect on virB genes, and decreased bacterial biomass when phenolic acids were present. The collective findings demonstrate how transcriptional regulation by A. tumefaciens controls the metabolism of 4-hydroxybenzoate and imply that PobA and PobR aid in bacterial survival during host plant infection.

IMPORTANCE

Agrobacterium tumefaciens is a widely distributed environmental bacterium and a recognized phytopathogen. Phenolic acids influence the relationship between A. tumefaciens and plants. One of the most important phenolic acids found in soil is 4-hydroxybenzoate, which is generated by plants. Mutants defective in the atu4544 and atu4545 genes inhibit A. tumefaciens tumor development. The atu4544-encoded enzyme, PobA, can metabolize 4-hydroxybenzoate, and the expression of its gene is positively regulated by the transcription factor encoded by atu4545. The atu4545 gene is subject to negative autoregulation. The binding site of atu4545 is CGTGCGATGGTCGGATT. Dual regulation of regulators for phenolic acid catabolism may aid in the maintenance of appropriate quantities of phenolic compounds. These results clarify the pathogenic mechanisms of A. tumefaciens and broaden the understanding of the metabolic control mechanisms of phenolic chemicals.

INTRODUCTION

Agrobacterium tumefaciens is a common gram-negative plant pathogen that is prevalent in the soil, on plant surfaces, and within plants. In the natural environment, A. tumefaciens infects wounded tissues of dicotyledonous plants, forming crown galls (1). The pathogenicity of A. tumefaciens relies on a specific plasmid termed the tumor-inducing (Ti) plasmid (2). In addition, virulence genes located on the chromosome can affect virulence gene expression in A. tumefaciens. For example, chvA and chvB primarily aid in attaching A. tumefaciens to host plants (3). The Ti plasmid is crucial for A. tumefaciens pathogenicity as it facilitates the introduction of T-DNA into the host genome (4). The transfer of T-DNA from A. tumefaciens to host cells is initiated by the activation of virulence (vir) genes in response to plant compounds and signals (5). Acetosyringone and α-hydroxy acetosyringone, which are found exclusively in the exudates of injured plant cells, are commonly used inducers of vir gene expression (6). Other phenols and phenolic acids, such as syringic acid, ferulic acid, and vanillic acid, can affect the interactions between A. tumefaciens and its host plants (7).

Phenolic acids, such as ferulic acid, vanillic acid, 4-hydroxybenzoate, and protocatechuate, are generated through lignin degradation and secretion by plant cells and are often found in the proximity of plant roots (8, 9). The catabolism of phenols and phenolic acids is mostly dependent on the β-ketoadipate pathway, which is found in fungi and soil bacteria (10). The β-ketoadipate pathway comprises two branches, the protocatechuate and catechol routes. Both branches lead to the production of the same metabolic product, 3-oxoadipate enol-lactone (11). Among the various aromatic compounds, 4-hydroxybenzoate is particularly common (12) and is converted to protocatechuate by the microbial hydroxylase encoded by pobA. The degradation pathway of 4-hydroxybenzoate is associated with the pathogenicity of phytopathogens, such as Xanthomonas campestris (Xcc) (13), Fusarium oxysporum f. sp. lycopersici (14), and Cochliobolus heterostrophus (15). The lesion length of Xcc mutant strains ΔpobA and ΔpcaGH is shorter compared to wild-type Xcc in Chinese radish. Invasion of tomato roots was reportedly impaired in a cmle (3-carboxy-cis, cis-muconate lactonizing enzyme) deletion mutant of F. oxysporum f. sp. lycopersici. In addition, β-ketoadipate pathway genes in C. heterostrophus help to detoxify plant phenolic compounds, thus preventing cell death of the fungal pathogen.

The effect of 4-hydroxybenzoate degradation pathway on microbial pathogenicity in A. tumefaciens remains unclear. The atu4544 (pobA) gene encodes 4-hydroxybenzoate hydroxylase, which is involved in the degradation of 4-hydroxybenzoate in the β-ketoadipate pathway. This chromosomally encoded convergent pathway degrades aromatic compounds (10). The pathway is composed of the pob and pca genes organized in a superoperon in A. tumefaciens. Genomic research has revealed the degradation of phenolic acids via protocatechuate routes in A. tumefaciens C58 (Fig. 1). The pca structural genes, organized in two distinct operons, pcaBGHCD (atu4542-atu4538) and pcaIJF (atu4547--atu4549), which are approximately 4 kb apart, are involved in protocatechuate degradation. The structural genes pcaBGHCD encode β-ketoadipate enol-lactone hydrolase, γ-carboxymuconolactone decarboxylase, protocatechuate 3,4-dioxygenase, and β-carboxy-cis,cis-muconate lactonizing enzyme. The pcaIJ genes encode β-ketoadipate succinyl-coenzyme A transferase. The pcaF gene encodes β-ketoadipyl CoA thiolase. These genes have been studied by growth phenotypes of knockout mutants and enzyme activities (16). Three regulatory genes were predicted in the β-ketoadipate pathway, that is, atu4543 (pcaQ), atu4545 (pobR), and atu4546 (pcaR). LysR-type regulatory protein PcaQ (Atu4543) regulates the pcaBGHCD operon (17). Expression of the pcaIJ genes was not affected in the pcaQ: Ω background of strains and was induced by the coeffector β-ketoadipate (16). The pcaIJF genes may be regulated by a LysR-type regulatory protein PcaR (atu4546) based on amino acid sequence homology. Transcriptional regulation of 4-hydroxybenzoate hydroxylase (PobA) is mediated by PobR, which belongs to the AraC family of transcriptional regulatory proteins evolutionarily distant from the other two regulators in the aromatic compound-degradative pathways (16, 17). The operon organization and regulation of the pob-pca genes in A. tumefaciens differ from those of Acinetobacter calcoaceticus and Pseudomonas putida (18). These differences may reflect the ability of these species to adapt to various ecological niches and selection pressures. The molecular genetic characterization of the degradation of 4-hydroxybenzoate in A. tumefaciens provides an excellent model for determining the evolutionary process of catabolic pathways.

Fig 1.

4-Hydroxybenzoate degradation pathway in A. tumefaciens. (A) Proposed 4HBA metabolic pathway. (B) Genetic map of pca and pob genes in A. tumefaciens. pcaB(atu4538), pcaG(atu4539), pcaH(atu4540), pcaC(atu4541), pcaD(atu4542), pcaQ(atu4543), pobA(atu4544), pobR(atu4545), pcaR(atu4546), pcaI(atu4547), pcaJ(atu4548), and pcaF(atu4549).

This study investigated the role of the metabolic gene pobA (atu4544) in the degradation of phenolic acids. Upstream of atu4544, there is a gene, atu4545, which encodes a transcription factor. This transcription factor was found to have dual regulation properties. The effects of atu4544 and atu4545 on the tumorigenicity of A. tumefaciens were assessed, and a preliminary investigation assessed the pathogenic mechanism. The study findings expand previous research on the β-ketoadipate pathway in A. tumefaciens and open up new possibilities for understanding the pathophysiology of A. tumefaciens.

RESULTS

Function of A. tumefaciens atu4544

Of the variety of aromatic compounds found in soil and produced by plants, 4-hydroxybenzoate is one of the most common (12). This enzyme, encoded by pobA, catalyzes the hydroxylation of the C3 carbon atom of 4-hydroxybenzoate on the benzene ring, converting it to protocatechuate. Subsequently, protocatechuate is metabolized through the β-ketoadipate pathway (Fig. 1). In A. tumefaciens C58, the atu4544 gene on the linear chromosome was designated to encode 4-hydroxybenzoate 3-monooxygenase. To study the atu4544 gene and its encoded protein, an atu4544 knockout strain was created by homologous recombination with the suicide plasmid pEX18Km. The complementation strain designated C-Δatu4544 was obtained by expressing atu4544 in the strain Δatu4544 using the plasmid pUCA19. Cell growth of the wild-type C58, knockout mutant, and the complementation strain C-Δatu4544 was investigated on common sugars (sucrose and arabinose) and phenolic acids (4-hydroxybenzoate and protocatechuate) as the sole carbon source. As shown in Fig. S1, the three strains exhibited similar growth trends on common carbon sources, suggesting that atu4544 had no influence on the central carbon metabolism of A. tumefaciens. A. tumefaciens C58 can utilize 4-hydroxybenzoate as the sole carbon source. When atu4544 was deleted, 4-hydroxybenzoate was not utilized as a carbon source. Growth deficiency was restored when atu4544 was expressed in Δatu4544 (Fig. 2A). When protocatechuate was used as the sole carbon source, A. tumefaciens C58, strain Δatu4544, and C-Δatu4544 showed comparable growth trends, and their biomass reached nearly OD₆₀₀ of 1.0 at the stationary phase (Fig. 2B). Cell growth determined in the presence of phenolic acids indicated the involvement of atu4544 in 4-hydroxybenzoate catabolism, rather than protocatechuate metabolism, in the β-ketoadipate pathway.

Fig 2.

Cell growth of A. tumefaciens strains (C58, Δatu4544, and C-Δatu4544) on AB minimal medium with phenolic acids as carbon sources. 5 mM 4-hydroxybenzoate (A) and 10 mM protocatechuate (B) as the carbon source. Every point represents a standard deviation from the mean of a minimum of three replicates.

To determine whether atu4544 encoded 4-hydroxybenzoate 3-monooxygenase, its protein expression and enzymatic activity were examined. The atu4544 nucleotide sequence was used to construct an isopropyl β-D-1-thiogalactopyranoside (IPTG)-inducible clone in Escherichia coli BL21 with plasmid pGex-4t-1, and the glutathione (GST)-tagged recombinant protein was extracted using an affinity chromatography resin. Figure S2A shows the expression of GST-atu4544, which has a molecular mass of approximately 70 kDa. The empty plasmid pGEX-4T1 was also purified (Fig. S2B) and used for enzymatic activity assays. The retention periods in the high-performance liquid chromatography (HPLC) trace (Fig. S3A and B) of 17.3 and 8.4 min for 4-hydroxybenzoate and protocatechuate, respectively, indicated that the mobile phase utilized is capable of effectively separating 4-hydroxybenzoate and protocatechuate. Only 4-hydroxybenzoate was observed when the purified GST-tag protein was introduced; no protocatechuate was produced (Fig. S3C). The peak area of 4-hydroxybenzoate as the substrate was reduced with the addition of purified atu4544 protein and NADPH (or NADH) using the flavin cofactor FAD, and the concentration decreased similarly (Fig. S3D and E). The peak area and concentration of protocatechuate progressively increased as the enzymatic reaction proceeded (Fig. S3D and E). Atu4544 (PobA) catalyzed the hydroxylation of 4-hydroxybenzoate to produce protocatechuate. Enzyme activities of PobA were 1.96 and 4.53 U/mg when NADH and NADPH, respectively, were used as the electron donor for the enzymatic assay.

Regulatory role of A. tumefaciens atu4545

Regulation of pobA expression by the transcription factor PobR has been documented in several bacterial species, including A. calcoaceticus, Azotobacter chroococcum, Cupriavidus necator, Streptomyces coelicolor, and Xanthomonas campestris (19–23). In A. calcoaceticus, the transcriptional activator PcaU shares structural similarities with PobR (24). In P. putida, both PobC and PobR regulate pobA expression (25). In A. tumefaciens C58, the gene atu4545 (pobR) is located upstream of atu4544 and has been annotated as an AraC family transcriptional activator. The strains Δatu4545 and C-Δatu4545 were constructed and cultivated on common carbon sources and phenolic acids. The growth trends of these atu4545 mutants (Fig. 3) were consistent with the atu4544 mutants on certain carbon sources, indicating a correlation between atu4544 and atu4545. To investigate the atu4545-encoded transcription factor that may act on atu4544 expression, an atu4544-lacZ fusion on the A. tumefaciens chromosome was created to represent atu4544 expression by detecting β-galactosidase activity. Theoretically, the function of atu4544 is not affected by lacZ insertion because the lacZ gene is linked after the stop codons of atu4544 by a strong ribosome-binding site sequence. Two recombinant strains, A. tumefaciens atu4544-lacZ and A. tumefaciens Δatu4545 atu4544-lacZ, respectively, displayed similar growth phenotypes to the wild-type and A. tumefaciens Δatu4545 (Fig. S4). Subsequently, β-galactosidase activities of the two recombinant strains were measured on an AB-arabinose mineral medium (Fig. 4). Compared with A. tummefaciens atu4544-lacZ, β-galactosidase activity of A. tummefaciens Δatu4545 atu4544-lacZ decreased significantly by 74%, indicating that the atu4544 expression could be triggered by the transcription factor atu4545. The effect of external substances on the expression of atu4544 was studied. 4-hydroxybenzoate, protocatechuate, or adipic acid was tested as PobR effectors in AB-arabinose growth medium based on earlier studies of pobR homologous genes in other bacterial species and metabolites in the β-ketoadipate pathway in A. tumefaciens. Adipic acid, as a nonmetabolizable analog of β-ketoadipate, was added into arabinose medium to investigate the enzyme activities of the β-ketoadipate pathway in A. tumefaciens (26). Figure 4 shows that the addition of 4-hydroxybenzoate, protocatechuate, and adipic acid boosted atu4544 expression by 2.01, 1.17, and 1.64 times, respectively. 4-Hydroxybenzoate and adipic acid had the potential to greatly increase lacZ expression (P < 0.05). These phenotypic studies and enzyme activities of related mutants revealed that atu4545 positively regulated the expression of atu4544 and that 4-hydroxybenzoate and adipic acid can increase its regulatory function more than protocatechuate as inducers.

Fig 3.

Cell growth of A. tumefaciens strains (C58, Δatu4545, and C-Δatu4545) on AB minimal medium with different sole carbon sources. (A) 15 mM sucrose as the sole carbon source, (B) 15 mM arabinose as the sole carbon source, (C) 5 mM 4-hydroxybenzoate as the sole carbon source, and (D) 10 mM protocatechuate as the sole carbon source. Every point represents a standard deviation from the mean of a minimum of three replicates.

Fig 4.

Effects of atu4545 on atu4544 promoter activity using β-galactosidase activity. C58 and Δatu4545 correspond to A. tumefaciens atu4544-lacZ and A. tumefaciens Δatu4545 atu4544-lacZ. They were grown in AB-arabinose (Ara) medium. Protocatechuate (PCA), 4-hydroxybenzoate (4HBA), and adipic acid (AA) were, respectively, added to the Ara medium. Data are presented as averages of three independent experiments. The error bars indicate the standard deviation of the mean. **P < 0.01; *P < 0.05; n.s., not significant.

Autoregulation of PobR has been documented in A. calcoaceticus (27) and X. campestris (23). The atu4545 promoter was ligated to the lacZ gene using plasmid pCB301 to examine whether atu4545 influences its own expression in A. tumefaciens. The pCB301 PpobR-lacZ was transformed into A. tumefaciens C58 and A. tumefaciens Δatu4545 to obtain A. tumefaciens lacZ and A. tumefaciens Δatu4545 lacZ, respectively. PpobR-dependent β-galactosidase activities in the strain Δatu4545 lacZ were all higher than those in A. tumefaciens lacZ under the same condition (Fig. 5). On AB-arabinose minimum medium, the β-galactosidase activity of strain Δatu4545 was 1.44 times higher than that of A. tumefaciens lacZ. The enhanced promoter activity of atu4545 following its deletion indicated that the expression of pobR was negatively controlled by its own gene products. After adding 4-hydroxybenzoate, protocatechuate, or adipic acid, the ratio of lacZ activity in A. tumefaciens Δatu4545 lacZ to lacZ activity in A. tumefaciens lacZ was 3.14, 0.99, and 0.98 times, respectively, of the lacZ activity ratio grown on AB-arabinose minimum medium. 4-Hydroxybenzoate enhanced the self-regulation of PobR compared to protocatechuate and adipic acid. There was no significant increase in lacZ activity after adding protocatechuate and adipic acid to strain Δatu4545, which indicated that protocatechuate and adipic acid had no obvious effect on enhancing pobR gene self-regulation. 4-Hydroxybenzoate, rather than protocatechuate or adipic acid, enhanced the autoregulation of the pobR gene.

Fig 5.

Effects of atu4545 on atu4545 promoter activity using β-galactosidase activity. C58 and Δatu4545 correspond to A. tumefaciens lacZ and A. tumefaciens Δatu4545 lacZ, and they were grown in AB-arabinose (Ara) medium. Protocatechuate (PCA), 4-hydroxybenzoate (4HBA), and adipic acid (AA) were, respectively, added to Ara medium. Data shown are the average of three independent experiments. The error bars indicate the standard deviations of the mean. *P < 0.05; n.s., not significant.

PobR protein binds directly to the intergenic region of atu4544 and atu4545

Genome analysis allowed predicting that the intergenic region of atu4544 and atu4545 had two divergent promoters. Transcription factor binding sites were predicted using the MEME suite (28) (Fig. 6A). To test whether atu4545 can bind to the intergenic region of atu4544 and atu4545, an electrophoretic mobility shift assay (EMSA) was performed (Fig. 6B). The intergenic regions were labeled with FAM oligonucleotides. The labeled intergenic region did not form shifted bands with the bovine albumin protein used as a negative control (lane 7). When the concentration of purified His-atu4545 protein (Fig. S5) was increased to 20 and 30 µg, the incubation of the labeled intergenic region and the purified atu4545 protein formed retarded bands (lanes 2–5). These shifts indicated the interaction between the atu4545 protein and natural DNA probes, which were not influenced by unrelated DNA fragments (lane 5). The possible binding motifs were validated by replacing the motif with the internal sequences of atu4544 of the same length (17 bp). The purified atu4544 did not form retarded bands (lanes 8–10) with the mutated DNA probe. The purified PobR protein was precisely attached to the upstream sequences of atu4544 and atu4545; the atu4545-binding sites were CGTGCGATGGTCGGATT, according to the EMSA results of the original and modified DNA probes. In addition, EMSA was conducted using varying concentrations of 4-hydroxybenzoate and adipic acid. After the addition of 4-hydroxybenzoate, the shifted band intensity strengthened, indicating that a given mass of 4-hydroxybenzoate (0.25, 0.5, and 1.0 mM) increased the formation of the PobR/DNA probe complex (Fig. S6A). When a small amount of adipic acid was added, the shifted band remained almost unchanged. The increased adipic acid seemed to hinder the binding of the PobR/DNA probe (Fig. S6B).

Fig 6.

PobR protein binds to the intergenic region of atu4544 and atu4545. (A) Predicted overlapping promoters between atu4544 and atu4545. The −35 and −10 elements of the promoter were labeled and predicted by the BPROM online tool. The sequences in the box are possible binding sites. Two arrows represent the start of the respective ORFs. (B) EMSA results of PobR protein with the FAM-labeled DNA. Unlabeled intergenic sequences or mutated binding sites or unrelated DNA were added to determine the specificity of the binding. BSA protein was only added in lane 7 as a negative control.

Deletion of atu4544/atu4545 impairs the infection of host plants

Effects of atu4544/atu4545 on plant infection were evaluated by measuring tumorigenesis on carrot root discs and kalanchoe leaves (Fig. 7A and B). The inoculated strains were A. tumefaciens NT1, C58, Δatu4544, C-Δatu4544, Δatu4545, and C-Δatu4545. After 4 weeks, the tumors were scraped and weighed for statistical analysis. On the carrot plates, the NT1 strain lacking the Ti plasmid was almost completely non-tumorigenic. The tumor weighing results showed that the deletion of atu4544 and atu4545 resulted in the tumor weight reduction of 74% and 76% on carrot root discs, and 62% and 59% on kalanchoe leaves, respectively. Tumor weights of C-Δatu4544 and C-Δatu4545 could approach 100% and 94% of the wild type, respectively, indicating that gene expression of atu4544/atu4545 could regain the tumorigenesis of the respective knockout mutants (Fig. 7C and D). Bacteria collected from different tumors were diluted and counted in AB-sucrose medium. The number of colonies per 0.1 g tumor in Δatu4544 and Δatu4545 was only 20% and 6% for carrot root discs and 32% and 21% for kalanchoe leaves, respectively, of those in A. tumefaciens C58 (Fig. 7E and F). The reduction in the number of colonies in tumors suggests that the loss of atu4544/atu4545 affects the survival and growth of A. tumefaciens in infected plants. 4-hydroxybenzoate concentrations in carrot root discs were assessed using tumors that had grown after 2, 3, or 4 weeks. Water and methanol extraction did not detect any 4HBA. To determine 4HBA concentrations in tumors, heterologous 4-hydroxybenzoate was added and quantified. Tumor weights (Fig. S7A) and colony development (Fig. S7B) in the tumors of the strains were comparable to those without exogenous 4HBA. The loss of atu4544/atu4545 resulted in a more than twofold increase in 4HBA levels relative to the wild-type C58 (Fig. S7C). Following supplementation with atu4544/atu4545, the 4HBA levels decreased considerably, indicating that A. tumefaciens’ 4HBA metabolic ability was restored. Both tumor weight and colony development in tumors showed that deletion of atu4544/atu4545 reduced A. tumefaciens pathogenicity.

Fig 7.

Effects of atu4544 and atu4545 on tumorigenesis of A. tumefaciens. (A) Tumorigenesis of A. tumefaciens strains on carrot root discs. (B) Tumorigenesis of A. tumefaciens strains on kalanchoe leaves. (C) Tumor weights after infecting the carrots for 4 weeks. (D) Tumor weights after infecting kalanchoe for 4 weeks. (E) The number of colonies in tumors on carrots infected by A. tumefaciens strains after 4 weeks. (F) The number of colonies in tumors from kalanchoe leaves infected by A. tumefaciens strains after 4 weeks. * denotes significant difference; ***P < 0.001; ****P < 0.0001.

pobA/pobR affects pathogenicity, rather than virB gene expression, of A. tumefaciens by affecting growth

Effects of atu4544/atu4545 on virB expression were first evaluated in A. tumefaciens. The previously constructed A. tumefaciens virB-lacZ as–chassis strains, atu4544 and atu4545, were individually deleted. The activity of β-galactosidase was measured to determine whether the two gene knockouts would change the expression of the virB gene. As shown in Fig. 8, acetosyringone increased the relative expression of the virB gene in A. tumefaciens. The atu4544 and atu4545 gene deletion strains did not significantly differ from that of wild-type C58 when acetosyringone was added, suggesting that atu4544 and atu4545 genes did not affect the expression of virB. The addition of 4-hydroxybenzoate significantly reduced the expression of the virB gene. Low β-galactosidase activity in the presence of 4-hydroxybenzoate was linked to cellular states suppressed by phenolic compounds. For example, vanillic acid inhibits virB gene expression (7). The wild-type C58, Δatu4544, and Δatu4545 strains still showed no difference in virB gene expression when 4-hydroxybenzoate was present. The underlying reason for atu4544/atu4545 influencing the host plant infection is thought to affect the cell growth of A. tumefaciens. Pot assays of A. tumefaciens strains on plates (Fig. 9) showed that cell growth was inhibited with increasing 4-hydroxybenzoate concentrations. At the same concentration of 4-hydroxybenzoate, the colonies of the knockout mutants were fewer than those of the wild-type and complemented strains. Similar trends were also seen in growth curves on liquid medium (Fig. 10). Lower colony development in tumors of atu4544/atu4545 mutant tumors was linked to the decreased A. tumefaciens growth. Decreased tumorigenesis in host plants could be related to the decrease in bacterial proliferation caused by the absence of atu4544/atu4545.

Fig 8.

Effects on virB gene expression of atu4544 and atu4545 in A. tumefaciens. Strains used here were engineered based on the virB-lacZ reporter fusion strain (A. tumefaciens virB-lacZ). AS denotes acetosyringone (100 µM), and 4HBA (5 mM and 0.5 mM) denotes 4-hydroxybenzoate. “ns” means no significance.

Fig 9.

Pot assays of A. tumefaciens growth on plates with different 4-hydroxybenzoate concentrations.

Fig 10.

Growth curves of A. tumefaciens at different 4-hydroxybenzoate concentrations on 0.05% AB-sucrose medium. (A) no 4HBA, (B) 0.02 mM 4HBA, (C) 0.2 mM 4HBA, (D) 2 mM 4HBA, (E) 4 mM 4HBA, (F) 6 mM 4HBA, (G) 8 mM 4HBA, (H) 10 mM 4HBA, (I) 12 mM 4HBA, (J) 14 mM 4HBA, and (K) 16 mM 4HBA.

DISCUSSION

In addition to growing independently, A. tumefaciens is a plant pathogen that coexists with plants. During normal growth and development, plants produce a wide range of phenolic compounds in both root and shoot tissues (29). The β-ketoadipate pathway breaks down 4-hydroxybenzoate and other phenolic chemicals (30). The enzymes of the β-ketoadipate pathway are conserved among bacteria, but their regulation and organization have evolved significantly (10)(31). The study of atu4544 and atu4545 extends the knowledge of the β-ketoadipate pathway in A. tumefaciens. The β-ketoadipate pathway in A. tumefaciens includes atu4543 (pcaQ), atu4545 (pobR), and atu4546 (pcaR). The LysR regulatory protein PcaQ has been validated as an activator of pcaBGHCD (17). The study of the other two regulatory genes depended only on BLAST analysis. atu4546 shared 42.7% amino acid sequence identity with PcaR of P. putida. The atu4545 gene is homologous to a putative pobR gene in Rhizobium leguminsarum biovar; however, it does not show homology with PobR of A. calcoaceticus (18). The ability of atu4544 (pobA) to catalyze the conversion of 4-hydroxybenzoate to protocatechuate was examined in this study, as it was essential for cell growth on AB-4-hydroxybenzoate minimum medium. The AraC regulatory gene, atu4545 (pobR), was predicted to be located on the nucleotide sequence upstream of atu4544. The ability of PobR to bind to the promoter site CGTGCGATGGTCGGATT and activate atu4544 expression was confirmed.

The β-ketoadipate pathway of A. tumefaciens was similar to the most thoroughly studied bacterial species (P. putida and A. calcoaceticus), but had its distinctive patterns of transcriptional control (32). Similar to many gram-negative bacteria, the pobA gene is linked to the pobR gene in A. tumefaciens. The intergenic region between pobA and pobR, a bidirectional promoter, drives coordinated transcription in opposite directions. The regulatory role of PobR in the pobA gene can be divided into positive and negative. The combination of in situ knockout and reporter gene assay suggests that atu4545 activates the expression of atu4544. Despite the differences in protein sequence similarity, for example, A. chroococcum PobR has no homology to the PobR protein in Acinetobacter strain ADP1, and positive regulation by PobR is currently found in most bacteria, such as X. campestris (23), A. chroococcum (20), A. calcoaceticus (27), and C. necator (21). S. coelicolor is an exception, where PobR (SCO3209) negatively regulates pobA transcription (22). In P. putida WCS358, the regulatory protein PobC, which has a low identity with other PobR proteins, activates pobA expression (25). In addition to controlling pobA, PobR was also confirmed to regulate its expression in A. tumefaciens. This function was previously observed only in A. calcoaceticus and not in other bacteria. Furthermore, the binding sequences were compared with those of other known regulatory motifs using TOMTOM. This low similarity indicated a distinct regulatory role of PobR in A. tumefaciens (Fig. S8). Divergence in transcriptional regulation mechanisms might represent evolutionary flexibility and nuanced features of niche adaptation. Bacteria use different inducers for the β-ketoadipate pathway. For example, protocatechuate activates all of the protocatechuate branch genes in A. calcoaceticus but not any of the pca genes in Agrobacterium (10). In Pseudomonas cepacia, β-ketoadipate and 4-hydroxybenzoate induce pcaB expression (33). 4-hydroxybenzoate could enhance atu4544 expression for 4HBA degradation and also induce pobR negative autoregulation, which may be linked to maintaining appropriate levels of 4-hydroxybenzoate. Adipic acid, a nonmetabolizable analog of β-ketoadipate, increases atu4544 expression, indicating that pathway intermediates may affect 4-hydroxybenzoate metabolism in A. tumefaciens (16). Bacterial species exhibit diverse physiological regulation due to the inducibility of β-ketoadipate pathway genes by different pathway intermediates.

Deletion of atu4544 and its transcriptional activator atu4545 reduced tumor weight and bacterial colonies in tumors when the strains were incubated on carrot root discs and kalanchoe leaves. Similarly, strains ΔpobR/ΔpobA of Xcc exhibited compromised virulence in Chinese radish (13). Additional genes in the β-ketoadipate pathway were linked to the infection of host plants in F. oxysporum f. sp. Lycopersici (14) and C. heterostrophus (15). The underlying mechanism of the β-ketoadipate pathway on microbial pathogenicity remains unknown. Expression of virulence genes is thought to be important for the pathogenicity of A. tumefaciens C58 (34). The promoter activity of vir genes in the atu4544/atu4545 knockout mutants was equivalent to that in the C58 strain, indicating that 4-hydroxybenzoate hydroxylation did not affect vir gene expression. The compromised virulence of atu4544/atu4545 knockout mutants may be related to cell growth inhibition in the presence of 4-hydroxybenzoate in pot assays.

In addition to 4-hydroxybenzoate, cell growth is also inhibited by other aromatic compounds, such as p-coumaric acid, ferulic acid, and vanillic acid (7) (Fig. S9 and S10). Plant roots generate phenolic compounds and serve dual functions of repelling and attracting different organisms to the plant environment, which could promote rhizobial nodulation and agrobacterial pathogenicity (35). Appropriate quantities of phenolic compounds may be most favorable to the infection of A. tumefaciens in plant hosts (36). 4-Hydroxybenzoate is one of the most significant aromatic compounds formed from plant-derived molecules and is present in soil (12). Atu4544 could convert 4-hydroxybenzoate to the structurally similar protocatechuate, which is subsequently metabolized to tricarboxylic acid intermediates via the β-ketoadipate pathway (10). Atu4545 harbors 4-hydroxybenzoate binding sites that are conserved with PobR sites of Xcc, including H8, R15, H19, W21, R27, Q33, Y36, H67, and R134. These sites were required for 4-hydroxybenzoate binding and the degradation of 4-hydroxybenzoate (13). Simultaneously, the negative autoregulation of atu4545 allows self-modification of its level to control the breakdown of 4-hydroxybenzoate. Another regulatory gene in the β-ketoadipate pathway, pcaQ (atu4543), was autorepressed and served as an activator of pcaBGHCD (atu4542-atu4538) (17). The atu4546 regulatory gene was validated by dual regulation (unpublished data). Regulators frequently regulate the expression of several genes by binding to transcription factor binding sites upstream of their target genes or transcription units (37). Several transcription factors act as activators and repressors. One basic example is transcription factors that bind to a single location in the intergenic region between divergently transcribed units and regulate each in a unique manner (38). The metabolic function and transcriptional regulation of 4-hydroxybenzoate metabolism were adapted to A. tumefaciens infection of the host plant.

MATERIALS AND METHODS

Strains, plasmids, and growth conditions

The bacterial strains and plasmids used in this study are listed in Table 1. The primers and synthetic DNA fragments are listed in Table S1. E. coli strains were grown in Luria-Bertani (LB) medium at 37°C (39). A. tumefaciens strains were grown in MG/L medium or Agrobacterium (AB) minimal medium at 28°C (40). The sole carbon sources added to the AB minimal medium were 5 mM 4-hydroxybenzoate, 10 mM protocatechuate, 15 mM sucrose, and 15 mM L-arabinose for A. tumefaciens. Cultivation of plasmid-bearing strains required the addition of 50 or 100 mg/L kanamycin, 100 mg/L ampicillin, and 100 mg/L carbenicillin.

TABLE 1.

Strains and plasmids used in this study^a

Strains and plasmids	Characteristics	Source
Strains
E. coli DH5a	Host for DNA cloning	(41)
E. coli BL21 (DE3）	Expression vector host	(42)
E. coli Atu4544	BL21(DE3) strain transformed using pGEX::GST-atu4544	This study
E. coli Atu4545	BL21(DE3) strain transformed using pET30a::HIS -atu4545	This study
A. tumefaciens C58	Wild type, nopaline-type pTiC58 plasmid	(43)
A. tumefaciens NT1	Lack of the Ti plasmid and no production of HSL	Integrative Microbiology Research Center South China Agricultural University
A. tumefaciens Δatu4544	atu4544 gene deleted in A. tumefaciens C58	This study
A. tumefaciens Δatu4545	atu4545 gene deleted in A. tumefaciens C58	This study
A. tumefaciens C-Δatu4544	Complementary strain obtained by transforming pUCA19-atu4544 into Δatu4544 strain	This study
A. tumefaciens C-Δatu4545	Complementary strain obtained by transforming pUCA19-atu4545 into Δatu4545 strain	This study
A. tumefaciens atu4544-lacZ	lacZ gene inserted in situ at atu4544 downstream on the chromosome of A. tumefaciens C58	This study
A. tumefaciens Δatu4545 atu4544-lacZ	lacZ gene inserted in situ at atu4544 downstream on the chromosome of A. tumefaciens Δatu4545	This study
A. tumefaciens lacZ	pCB301:: promoteratu4545:: lacZ transformed into A. tumefaciens C58	This study
A. tumefaciensΔatu4545 lacZ	pCB301:: promoteratu4545:: lacZ transformed into A. tumefaciens Δatu4545	This study
A. tumefaciens virB-lacZ	lacZ gene inserted in situ at virB downstream (located at the virB initiation codon +162 bp) of A. tumefaciens C58	Constructed by our lab (unpublished)
A. tumefaciens virB-lacZ Δatu4544	atu4544 gene deleted in A. tumefaciens virB-lacZ	This study
A. tumefaciens virB-lacZ Δatu4545	atu4545 gene deleted in A. tumefaciens virB-lacZ	This study
Plasmids
pEX18Km	Gene replacement vector carrying a counter selectable marker sacB, oriT, KmR	(44)
pEX18Km-atu4544	500 bp upstream and 500 bp downstream fragments of atu4544 integrated into the suicide plasmid pEX18Km to delete the atu4544 gene; KmR	This study
pEX18Km-atu4545	500 bp upstream and 631 bp downstream fragments of atu4545 integrated into the suicide plasmid pEX18Km to delete the atu4544 gene; KmR	This study
pEX18Km-atu4544-lacZ	Three fragments integrated into the suicide plasmid pEX18Km to insert lacZ gene behind atu4544 on the chromosome; KmR	This study
pUCA19	pUC19 carrying an agrobacterial replicon; ApR, CrR	(44)
pUCA19-atu4544	Plasmid pUCA19 carrying encoding sequence of atu4544; ApR, CrR	This study
pUCA19-atu4545	Plasmid pUCA19 carrying encoding sequence of atu4545; ApR, CrR	This study
pCB301	Minimum binary vector plasmid; KmR	(45)
pCB301:: promoteratu4545:: lacZ	Upstream of atu4545 ligated to the lacZ gene integrated into pCB301	This study
pET30a	Protein-expressing plasmid; KmR	Novagen
pET30a:: HIS-atu4545	N-terminal HIS-tag of the pET30a ligated with 924 bp atu4545 ORF; KmR	This study
pGEX-4T-1	Protein-expressing plasmid; Apr, Crr	GE Healthcare
pGEX::GST-atu4544	N-terminal GST-tag of the pGEX-4T-1 ligated with 1,173 bp atu4544 ORF; Apr, Crr	This study

^a Km: kanamycin; Ap: ampicillin; Cr: carbenicillin.

Construction of A. tumefaciens deletion mutants and complementary strains

Atu4544 and atu4545 were deleted in both A. tumefaciens C58 and A. tumefaciens virB-lacZ. The upstream and downstream sequences of the two genes were amplified using Gene-U-F/R and Gene-D-F/R primers (see Table S1). The gene knockout cassette was produced by overlapping PCR with Gene-U-F and Gene-D-R primers and detected by 1% agarose gel electrophoresis. Purified PCR products obtained using an agarose gel DNA extraction kit (TaKaRa Bio, Shiga, Japan) and the pEX18Km enzyme-digested suicide vector were ligated to convert E. coli DH5α. After PCR and DNA sequencing, the generated plasmids were transformed into A. tumefaciens C58 and A. tumefaciens virB-lacZ. After kanamycin resistance and sucrose sensitivity screening, gene deletion mutants were obtained and validated by DNA sequencing.

To construct the atu4544 and atu4545 expression plasmids, the coding sequences of these genes were amplified using C-atu4544-F/R and C-atu4545-F/R and inserted into the pUCA19 plasmid expression vector harboring the lacZ promoter. The correctly constructed plasmids were used to transform the corresponding deletion mutant strains of A. tumefaciens. A. tumefaciens C-Δatu4544 and A. tumefaciens C-Δatu4545 were verified by detecting kanamycin resistance and DNA sequencing.

Construction of A. tumefaciens lacZ reporter strains

The A. tumefaciens strain harboring a genome containing the in situ inserted reporter gene lacZ was used to measure atu4544 promoter activity in the presence or absence of atu4545. The lacZ gene was programmed to follow atu4544 using the strong RBS sequence 5′-TTTCTCCTCTTT-3′ (46). The lacZ gene was amplified using the atu4544-lacZ-F/R primer. The atu4544 coding region (666–1,173 bp) and its downstream 507 bp sequence were amplified from A. tumefaciens C58. The three fragments were ligated to the enzyme-digested suicide vector pEX18Km via homologous recombination. The correctly constructed pEX18Km-atu4544-lacZ chosen from E. coli DH5α was transformed into A. tumefaciens C58 and A. tumefaciens Δatu4545. A. tumefaciens atu4544-lacZ and A. tumefaciens Δatu4545 atu4544-lacZ generated by the gene replacement system were verified by PCR and DNA sequencing.

To evaluate the effects of atu4545 on its promoter activity, A. tumefaciens lacZ and A. tumefaciensΔatu4545 lacZ strains were constructed. Using diluted A. tumefaciens C58 as the template, the promoter sequences of atu4545 were amplified using Pr-atu4545-F/R primers (see Table S1). The promoter-free pCB301 plasmid was digested with the restriction enzymes HindIII and BamHI. The purified promoter and lacZ reporter were ligated with the pCB301 plasmid and transferred into E. coli DH5α. The validated pCB301::promoter_atu4545::_lacZ plasmid was transformed into both A. tumefaciens C58 and A. tumefaciens Δatu4545.

Strain cultivation

To investigate A. tumefaciens growth on different carbon sources, a single colony of each A. tumefaciens strain retrieved from MG/L agar was incubated overnight in a tube of MG/L liquid medium at 28°C and 200 rpm. The bacteria grown to logarithmic phase were washed twice and adjusted to an OD₆₀₀ of 0.5 using sterile water. One milliliter aliquots of each dilute bacterial suspension was transferred to 50 mL of AB minimal medium in a 250 mL shake flask containing 5 mM 4-hydroxybenzoate, 10 mM protocatechuate, 15 mM sucrose, or 15 mM L-arabinose as the sole carbon source. The absorbance value of OD₆₀₀ was measured to monitor cell growth in aliquots of each culture collected every 4 h over time.

For protein expression, single colonies of E. coli atu4544 and E. coli atu4545 activated from LB solid plates were cultivated overnight at 37°C in LB tubes. Ampicillin (100 mg/L) or kanamycin (50 mg/L) was added as required. E. coli atu4544 was inoculated into 150 mL LB medium (5% v/v inoculum), and E. coli atu4545 was inoculated into 100 mL LB medium (2% vol/vol inoculum). Both were shaken at 200 rpm. Following the respective addition of 0.1 mL of 0.1 M IPTG and 0.7 mL of 0.1 M IPTG, when the OD₆₀₀ reached 0.6, E. coli Atu4544 and E. coli Atu4545 were cultivated at 16°C for 16 h and 25°C for 4 h at 160 rpm.

To evaluate A. tumefaciens tumorigenicity, a single colony of A. tumefaciens strains (A. tumefaciens C58, Δatu4544, Δatu4545, C-Δatu4544, and C-Δatu4545) that grew on MG/L agar was retrieved and incubated overnight in a tube of MG/L liquid medium at 28°C and 200 rpm. Aliquots from each culture were transferred to AB-sucrose medium and grown to the logarithmic growth phase. The OD₆₀₀ of each bacterial suspension was adjusted to 0.5 with sterile water for inoculation into the host plants.

To examine the growth of A. tumefaciens C58, Δatu4544, Δatu4545, C-Δatu4544, and C-Δatu4545 under varying concentrations of 4-hydroxybenzoate, a single colony of each A. tumefaciens strain recovered from MG/L agar was cultured overnight in a tube of MG/L liquid medium at 28°C and 200 rpm. After two rounds of washing, the bacteria that had reached the logarithmic phase were adjusted to an OD₆₀₀ of 0.5 using sterile water. 50 mL of AB minimum medium in a 250 mL shake flask with 0.5 g/L sucrose and 0, 0.02, 0.2, 2, 4, 6, 8, 10, 12, 14, or 16 mM of 4-hydroxybenzoate was filled with one milliliter aliquot of each diluted bacterial suspension. To track cell development in aliquots of each culture that was taken, the absorbance value of OD₆₀₀ was recorded.

Pot assays

The growth of A. tumefaciens C58, Δatu4544, Δatu4545, C-Δatu4544, and C-Δatu4545 on 4HBA-containing plates was investigated. Cells recovered from overnight cultures in MG/L medium were washed twice with AB buffer and adjusted to an OD₆₀₀ of 0.1. Each suspension was diluted by 10⁻⁶, and aliquots were spotted onto AB-sucrose agar plates containing 2, 4, 6, 8, or 10 mM of 4-hydroxybenzoate. The growth phenotype of the A. tumefaciens strains on the 4HBA-containing agar was observed after 48 h of cultivation at 28°C.

Protein expression and purification

The atu4544 and atu4545 genes were amplified from diluted A. tumefaciens C58. The atu4544 gene was cloned into the pGEX-4T-1 expression vector with an N-terminal GST-tag. The atu4545 gene was cloned into the pET30a vector with an N-terminal 6-histidine tag. The pET30a::HIS-atu4545 and pGEX::GST-atu4544 were, respectively, transformed in E. coli BL21 (DE3) to generate E. coli atu4544 and E. coli atu4545. After recombinant protein expression in the two strains was induced by IPTG, the cells were collected, washed, and resuspended in phosphate-buffered saline (PBS; 10 mM, pH 7.4). After lysis by sonication in cycles of 4 s ON and 8 s OFF for 40 min, each cell suspension was centrifuged (12,000 × g, 4°C for 20 min). Each collected supernatant was passed through a 0.45 µm microfiltration membrane for protein purification. The supernatants of E. coli atu4544 and E. coli atu4545 were, respectively, incubated with BeyoGold GST-tag Purification Resin and BeyoGold His-tag Purification Resin for 1 h at 4°C with gentle rocking. The resins were washed several times with washing solution (His-tag resin: 1 L containing Tris-base 6.06 g, NaCl 29.22 g, pH 8.0; GST-tag resin: 1 L containing NaCl 8.19 g, KCl 201 mg, Na₂HPO₄ 1.42 g, KH₂PO₄ 245 mg, pH 8.0) to remove unbound and nonspecifically bound proteins. GST-atu4544 and HIS-atu4545 were eluted using 250 mM imidazole and 10 mM GST, respectively. Protein samples were analyzed using 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Electrophoretic mobility shift assays

The interaction of purified HIS-atu4545 with the intergenic sequences of atu4544 and atu4545 was evaluated by EMSA. Unlabeled intergenic sequences and unrelated DNA (similar length) were, respectively, amplified from diluted A. tumefaciens C58 using int-4544-F/R and pobR-inter-F/R primers. The recovered PCR products provided templates for amplifying the FAM [5(6)-carboxyfluorescein]-labeled intergenic sequence and unrelated sequences with the primer pairs int-4445-F/FAM and pobR-inter-F/FAM. The sites for atu4545 protein binding to the atu4544 promoter were predicted using the Motif Alignment and Search Tool from a set of motifs for PobR binding in other species (20, 22, 23, 27). A FAM-labeled intergenic sequence containing mutant binding sites was synthesized (Table S1). Unrelated DNA and bovine serum albumin were added to some lanes for the binding assay as negative controls. The 50 mL of binding buffer contained 0.394 g Tris-base, 0.186 g KCl, 0.019 g EDTA•2Na, 10% glycerol, and 0.008 g dithiothreitol, pH 8.0. The 20 µL reaction system that was added contained 20 ng DNA probes, 0.1 µg salmon sperm DNA, an increasing amount of HIS-atu4545 protein (0, 1, 4, and 6 µM), and a binding buffer. Each reaction volume was then incubated at 25°C for 40 min. Pre-electrophoresis using 6% native PAGE gels was performed at 20 V for 30 min, followed by protein separation at 100 V for 1.5 h. Each gel was scanned using a Typhoon FLA 9500 device (GE Healthcare, Uppsala, Sweden).

Enzymatic activity assay

4-Hydroxybenzoate hydroxylase activity was determined as previously described (47). The reaction products and substrates were detected by HPLC. The 3 mL reaction mixture in 50 mM Tris∙Cl (pH 7.4) buffer contained 5 µM purified GST-atu4544 protein, 0.5 mM 4-hydroxybenzoate, 0.4 mM NADPH (or 0.4 mM NADH), and 5 µM FAD. Each mixture was incubated at 30°C. In the reaction, 4-hydroxybenzoate consumption and protocatechuate formation were catalyzed by 4-hydroxybenzoate hydroxylase encoded by the atu4544 protein. The 500 µL reaction mixtures collected at 0, 0.5, and 1 h were put in 100°C boiling water for 5 min to quench the reaction. After removing precipitated protein using Ultra 30K cutoff centrifugal filters (Millipore, Billerica, MA, USA), 20 µL aliquots of each reaction mixture were measured by HPLC at 250 nm using a ZORBAX SB 80 Å C18 column, eluted with acetonitrile-water (9:91, vol/vol, with the pH adjusted to 2.5 using phosphoric acid) at a flow rate of 1 mL min⁻¹. One unit of 4-hydroxybenzoate hydroxylase-specific activity was defined as the amount of enzyme that generated 1 µmol min⁻¹ protocatechuate per mg protein at 30°C.

β-galactosidase activity was determined as previously described (48) with some modifications. Briefly, a single colony of each A. tumefaciens strain on MG/L agar was incubated overnight in a tube of MG/L liquid medium at 28°C and 200 rpm. After washing twice using sterile water, two procedures were performed depending on the sample. In the first procedure, 100 µL of the suspension of A. tumefaciens atu4544-lacZ, A. tumefaciens Δatu4545 atu4544-lacZ, A. tumefaciens lacZ, and A. tumefaciens Δatu4545 lacZ was inoculated in 5 mL AB-arabinose medium, followed by the addition of 10 mM protocatechuate, 5 mM 4-hydroxybenzoate, or 15 mM adipic acid as needed. The OD₆₀₀ values of each bacterial suspension were measured when the cells reached the logarithmic growth stage. In the second procedure, 100 µL of the suspension of A. tumefaciens virB-lacZ Δatu4544 and A. tumefaciens virB-lacZ Δatu4545 was inoculated in 5 mL AB-sucrose medium or 5 mL AB-sucrose medium containing 5 mM 4-hydroxybenzoate. After 4 h of cultivation, 100 µM acetosyringone was added. After continuous cultivation for 9 h, the OD₆₀₀ of each bacterial suspension was measured. Five hundred microliters of each bacterial suspension was successively received 500 µL Z-buffer, 50 µL chloroform, and 20 µL 0.1% SDS. The reaction was started by adding 200 µL of o-nitrophenyl-β-galactoside and was terminated after the solution turned yellow by adding 1 M Na₂CO₃. Reaction time is the time it takes for a batch of experimental samples to turn yellow from the start of the reaction. After centrifugation, the OD₄₂₀ of each supernatant was measured. One unit of β-galactosidase activity was defined as the amount of enzyme required to degrade o-nitrophenyl-β-galactoside to produce 1 µmol min⁻¹ o-nitrophenol at 37°C.

Tumorigenesis assay

A tumorigenesis assay was performed to determine tumor formation and bacterial abundance in tumors using A. tumefaciens infections of carrot plates and kalanchoe leaves (49, 50). The carrots were cleaned and their middle parts were peeled. Cylindrical disks with an inner diameter of 1.5 cm and a thickness of 5 mm were made from the clean and peeled middle parts of the carrots. After soaking in 1.05% sodium hypochlorite for 30 min, the disks were placed on 1.5% agar solid plates at regular intervals. Five microliter aliquots of each bacterial suspension was inoculated in the center of each carrot disk. Fifteen replicates were analyzed for each bacterial strain. The infected carrot tuber discs were incubated at 25°C for 2–4 weeks. Kalanchoe leaves were wiped using a 75% alcohol swab and scratched with a hypodermic needle to create wounds of similar length and depth. Five microliter aliquots of each bacterial suspension was inoculated on each wound line of kalanchoe leaves at room temperature in a cool and shaded area. A. tumefaciens C58, NT1, Δatu4544, Δatu4545, C-Δatu4544, and C-Δatu4545 were individually injected on the same leaf to rule out the possibility of leaf age-related tumorigenesis. The tumors were carefully scraped off the discs or incisions after 2–4 weeks and weighed. In addition, 0.1 g of tumors from carrots or kalanchoe leaf was put in 500 µL sterile normal saline and ground for as close to the same duration and with the same force as possible. One hundred microliter aliquots of each mixture was diluted by 10⁻⁸, and 100 µL of each bacterial suspension was spread onto AB-sucrose agar. The colonies on the plates were counted after 48 h at 28°C. To determine 4-hydroxybenzoate concentrations in tumors, on each carrot disk, 50 µL of 15 mM 4-hydroxybenzoate was injected into the tumors once they had grown out (after approximately 10 days). These disks were incubated for 4 weeks. The tumors were collected, weighed, colonies were enumerated, and 4-hydroxybenzoate concentrations were determined by HPLC.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00255-25.

Supplemental material. aem.00255-25-s0001.docx.

Table S1; Figures S1 to S10.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.