Identification of virulence tal gene in the cotton pathogen, Xanthomonas citri pv. malvacearum strain Xss-V2-18

Background Bacterial blight of cotton (BBC), which is incited by Xanthomonas Xcm ), is a destructive disease in cotton. Transcription activator-like effectors (TALEs), encoded by tal -genes, play critical roles in the pathogenesis of xanthomonads. Characterized strains of cotton pathogenic Xcm harbor 6-13 different tal genes and only one of them is functionally decoded. Further identification of novel tal genes in Xcm strains with virulence contributions are prerequisite to decipher the Xcm -cotton interactions Results In this study, we identified six tal genes in Xss-V 2 -18, a highly-virulent strain of Xcm from China, and assessed their role in BBC. RFLP-based Southern hybridization assays indicated that Xss-V 2 -18 harbors the six tal genes on a plasmid. The plasmid-encoded tal genes were isolated by cloning Bam HI fragments and screening clones by colony hybridization. The tal genes were sequenced by inserting a Tn 5 transposon in the DNA encoding the central repeat region (CRR) of each tal gene. Xcm TALome evolutionary relationship based on TALEs CRR revealed relatedness of Xss-V 2 -18 to MSCT1 and MS14003 from the United States. However, Tal2 of Xss-V 2 -18 differs at two repeat variable diresidues (RVDs) from Tal6 and Tal26 in MSCT1 and MS14003, respectively, inferred functional dissimilarity. The suicide vector pKMS1 was then used to construct tal deletion mutants in Xcm Xss-V 2 -18. The mutants were evaluated for pathogenicity in cotton based on symptomology and growth in planta . Four mutants showed attenuated virulence and all contained mutations in tal2 . One tal2 mutant designated M2 was further investigated in complementation assays. When tal2 was introduced into Xcm M2 and expressed in trans , the mutant was complemented for both symptoms and growth in planta , thus indicating that tal2 functions as a virulence factor in Xcm Xss-V 2 -18. Conclusions Overall, the results demonstrated that Tal2 is

. For example, the TALEs PthXo1 and PthXo2 from X. oryzae pv. oryzae (Xoo) were shown to enhance the expression of rice genes OsSWEET11 and OsSWEET13, which are required for susceptibility to bacterial leaf blight [33,34]. However, rice cultivars were resistant to Xoo when they contained OsSWEET11 and OsSWEET13 alleles lacking PthXo1 and PthXo2 EBEs [35][36][37]. Recently, a new rice S gene (OsERF#123) was shown to be targeted by TalB in African strains of Xoo [38]. Other examples of TALEs include AvrBs3 that targets the pepper resistance gene Bs3 and AvrXa10, AvrXa23 and AvrXa27 that interact with rice R genes Xa10, Xa23 and Xa27, respectively [29,[39][40][41]. Recently, L Cai, Y Cao, Z Xu, W Ma, M Zakria, L Zou, Z Cheng and G Chen [21] reported that Tal7 from Xoo binds and activates the expression of the rice gene Os09g29100, an interaction that suppresses avrXa7-Xa7-mediated resistance in rice. A number of truncated TALEs (truncTALEs) and interfering TALEs (iTALEs) have also been reported in Xoo that function as suppressors of Xa1-mediated defense in rice [42,43].
Resistance to Xcm has been identified primarily in G. hirsutum. The genetic nature of resistance to BBC was first revealed in 1939, and efforts to breed cotton plants for resistance ensued shortly thereafter [44]. About 20 major R genes or polygene complexes (B genes) participate in resistance to BBC in cotton [7,8]. Based on their virulence phenotype in differential cotton hosts, Xcm strains have been classified into 22 races that are named 1-22 [7]. Race 18 is the most common variant and was first isolated in 1973 [45,46]. In some cases, the outcome of interactions between Xcm strains and differential cotton varieties is dependent on the avrBs3/pthA gene family in Xcm, which indicates that The aim of the current study is to identify a novel virulent tal-gene encoding TALE protein in a highly virulent cotton pathogen, Xcm strain Xss-V 2 -18 (from China).

TALEs of Xss-V 2 -18
Restriction fragment length polymorphism (RFLP) analysis was conducted to estimate the number and size of tal genes in Xcm Xss-V 2 -18. Since most tal genes retain two BamHI sites, Xcm Xss-V 2 -18 plasmid and genomic DNAs were digested with BamHI and analyzed by Southern blotting as described above. Six bands hybridized to the probe in BamHI-digested genomic and plasmid DNA, indicating that Xss-V 2 -18 contained six plasmid-encoded tal genes (Fig. 1A).
The six tal genes were cloned in pBluescript as BamHI fragments, giving rise to pB-tal1, pB-tal2, pB-tal3, pB-tal4, pB-tal5 and pB-tal6 (Fig. 1B) and confirmed by colony hybridization and sequence analysis. To obtain the complete DNA sequence of each tal gene, we inserted the Tn5 transposome into the CRR region and used primer sets tal-F/RP and FP/tal-R to obtain the sequences (Fig. 1C). The  Nearly identical RVD sequences were observed for the six TALEs in Xss-V 2 -18, MSCT1, H1005, MS14003 and AR81009 (Fig. 2B). Differences of two RVDs between Tal2 of Xss-V2-18 and TAL6 of   MSCT1, Tal26 of MS14003 indicate that they are functionally different from each other and may target   a different EBE. The predicted theoretical EBE box for Tal2, Tal6 and Tal26 of Xss-V2-18, MSCT1 and MS14003, respectively, are mentioned in Figure S1. RVDs in Xcm strains included NI, NG, NS, HD and NN; the latter RVD was absent in Tal1, Tal2, Tal3 and Tal4.

Xss-V 2 -18 tal deletion mutants
To assess the role of tal genes in the virulence of Xss-V 2 -18, we generated tal deletion mutants by homologous recombination using the suicide vector pKMS1. Fragments a (580 bp) and b (350 bp) were amplified on the left and right sides of DNA encoding the CRR, respectively, and cloned as a fused fragment in pKMSA1 (Fig. 3A, B). Construct pKMSA1 was introduced into Xcm Xss-V 2 -18; after homologous recombination, 41 putative mutants were selected for PCR amplification using primers pKMSA1-5F/pKMSA1-3R. Four putative mutants designated M1, M2, M3 and M4 contained a 930-bp PCR product, which is consistent with the size of the insert in pKMSA1 (Fig. 3C). Southern hybridization indicated that one or more tal genes were deleted in the four mutants (Fig. 3D). M1 and M2 were lacking tal3 and tal2, respectively, M3 was missing tal2 and tal4, and M4 lacked tal2, tal4, tal5, and tal6. These results indicated that four tal loci underwent homologous exchange via pKMSA1, and copies of the plasmid pKMSA1 functioned to delete multiple tal genes simultaneously in M3 and

M4.
A second round of deletion mutagenesis was conducted with plasmid pKMSA2, which contains a fusion of fragments c (150 bp) and d (300 bp) on the left and right sides of the DNA encoding the CRR, respectively (Fig. 3A). Construct pKMSA2 was used to generate new deletions in the M4 mutant, and potential new mutants were analyzed by PCR with primer pairs pKMSA2-5F/pKMSA2-3R. Two mutants designated M5 and M6 contained a 450-bp PCR product that is consistent with the size of the insert in pKMSA2. In addition to tal2, tal4, tal5, and tal6, Southern hybridization indicated that mutant M5 contained a deletion in tal3. M6 was lacking both tal1 and tal3 (Fig. 3F); thus, M6 lacked all six tal genes and can be considered a tal-free mutant of Xss-V 2 -18.

Virulence assays
Xss-V 2 -18 and mutants M1-M6 were inoculated into cotton leaves and phenotypes were observed 3-5 days post-inoculation (Fig. 4A). Xss-V 2 -18, M1, and M4 produced substantial water-soaked lesions in the inoculation sites; however, water-soaking was reduced in leaves inoculated with M2, M3, and M5 ( Fig. 4A). In contrast, the region inoculated with the tal-free mutant M6 showed cell death and necrosis (Fig. 4A). On the second day post-inoculation, the populations of the M2 and M6 mutants were significantly lower than Xss-V 2 -18, M1, M3, M4 and M5 (Fig. 4B). On days 4 and 6 postinoculation, the growth of Xss-V 2 -18 was significantly higher than mutants M1-M6 with no significant difference among the mutants. These results indicated that some of the tal genes are involved in Xss-V 2 -18 virulence, and the absence of selected tal genes impacted growth of the pathogen in planta.
Mutant M2, which lacks tal2, exhibited reduced symptomology and bacterial growth when compared to wild-type Xss-V2-18 (Fig. 4A,B). Based on these observations, we speculated that tal2 might be involved in virulence; this was addressed by constructing pHZW-tal2 (Table 1) for complementation analysis. The pHZW-tal2 construct was introduced into Xcm M2, and the empty vector (ev, pHM1) was used as a negative control. Western blot analysis indicated that the Tal2 protein was produced in Xcm M2 (Fig. 4D). The wild-type Xss-V 2 -18, mutant M2, M2(ev), and M2(tal2) were inoculated into cotton leaves; phenotypes were observed at 5-7 days post-inoculation (Fig. 4C), and bacterial growth was measured at 0, 2, 4, and 6 days post-inoculation (Fig. 4E). Both water-soaking and bacterial growth in planta were restored to wild-type levels in Xcm M2 containing pHZW-tal2 (Fig. 4C, E). Based on results shown in Fig. 5, we conclude that Tal2 is major virulence factor in Xss-V 2 -18.  To assess the role of tal genes in Xss-V 2 -18, we generated deletions in Xss-V 2 -18 by homologous recombination with pKMS1 [67], which was previously used to generate deletion mutants in the rice pathogen, Xoc [68][69][70]. This is the first report where pKMS1 was used to generate tal deletion mutants in Xcm, and the basic strategy was to replace the CRR (encoded by 102-bp repeat units) with up-and downstream fragments flanking the tal genes. Using construct pKMSA1, we obtained four mutants; M1 and M2 lacked tal3 and tal2, M3 had deletions in tal2 and tal4, and M4 lacked tal2, tal4, tal5 and tal6. We speculate that tal5 and tal6 might be located in the same gene cluster. The second knockout was obtained using pKMSA2 where up-and downstream flanking fragments (homology arms) were located closer to the CRR. Mutant M4 was used as a parental strain for the deletions generated with pKMSA2, and we recovered two new mutants designated M5 and M6. In addition to tal2, tal4, tal5 and tal6, mutant M5 also lacks tal3, whereas M6 contains deletions in all six tal genes.

Discussion
The symptoms induced by M2, M3, M5 and M6 were significantly reduced relative to the wild-type, thus indicating that one or more tal genes contribute to symptom development in Xss-V 2 -18. Mutants M2, M3, M5 and M6 all lack the tal2 gene; thus the potential contribution of tal2 to symptom development was further investigated. Expression of tal2 in trans restored symptoms and growth in planta to the M2 mutant, thus confirming that Tal2 is a virulence factor.
TALEs functionally resemble eukaryotic transcription factors that target and regulate the expression of host genes by binding to their promoter sequences. TALE-triggered susceptibility has been wellstudied, and the contribution of TALEs to virulence has been evaluated in many Xanthomonas spp.
[21-23, 28, 71-75]. For example, the TALEs PthXo1 and PthXo2 from Xoo were shown to enhance the expression of rice genes OsSWEET11 and OsSWEET13, which are required for susceptibility to bacterial leaf blight in rice [33,34]. However, rice cultivars were resistant to Xoo when they contained OsSWEET11 and OsSWEET13 alleles lacking PthXo1 and PthXo2 EBEs [35][36][37]. In Xcm, Avrb6 was the first TALE shown to be important for virulence [25]. Recently, the Xcm effector Avrb6 was shown to target and induce the expression of the cotton S gene, GhSWEET10, thus enhancing virulence and promoting disease [28]. The present study provides an important foundation for identifying potential S genes that interact with Tal2, which will ultimately help us develop better control strategies for BBC.

Conclusions
In this study, we identified genes encoding TALEs in the highly-virulent Xcm strain, Xss-V 2 -18 (from China), and assessed TALE roles in BBC. We found that Xss-V 2 -18 encodes six plasmid-borne tal genes. Knockout mutagenesis of Xss-V 2 -18 tal genes and complementation analysis demonstrated that Tal2 is required for full virulence of Xss-V 2 -18 on cotton. The identification of the Tal2 target in cotton will ultimately provide new avenues for developing BBC-resistant varieties.

Bacterial strains, growth conditions, and plasmids
The bacterial strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were grown in Luria-Bertani (LB) medium at 37˚C. Xcm strains were grown at 28˚C on TSA agar (1% tryptone, 1% sucrose, 0.1% sodium glutamate, 1.6% agar) or in nutrient broth (NB) (1 g yeast extract, 3 g beef extract, 5 g polypeptone and 10 g sucrose/L). Xcm transformants containing the first crossover event were grown on NAN (nutrient agar without sucrose) or NBN (NAN without agar) medium. For the second crossover event, transformants were plated on NAS agar (NAN with 10% sucrose). When appropriate, antibiotics were added at the following concentrations (µg/mL): ampicillin, 100; kanamycin, 20; spectinomycin, 25; and rifampicin, 50. The pH of both solid and liquid media was adjusted to 6.8.

DNA preparation
Total genomic DNA of Xss-V 2 -18 was isolated using the Bacterial Genomic DNA Extraction Kit (TaKaRa, China). The isolated gDNA pellet was re-suspended in double-distilled water and stored at -20˚C. Bacterial plasmid DNA was isolated using the Plasmid Miniprep Kit (Omega, USA).

Isolation, cloning and sequencing of Xss-V2-18 tal genes
The isolation and cloning of tal genes from Xcm strain Xcc-V 2 -18 followed the procedure as described previously [72] with minor modifications. Plasmid DNA (50 µg) was isolated from Xcm, digested with BamHI, and separated on 1.2% agarose gels. Specific tal DNA fragments were then gel-purified and ligated into pBluescriptII SK(-) that was digested with BamHI and treated with calf intestinal phosphatase (CIP). The ligated products were introduced into competent E. coli cells by the heat shock method according to the manufacturer's protocol (Bio-Rad, USA). The successful cloning of tal genes in pBluescript II was validated by restriction digestion, colony hybridization and sequence analysis.
The repeat units in tal genes complicate abilities to sequence the genes using conventional approaches. Thus, after cloning into pBluescript II SK(-), we used the EZ-Tn5 TM <KAN-2>Tnp Transposome TM Kit to insert Tn5 into each tal gene as recommended by the manufacturer (Epicentre, Madison, WI). Clones with Tn5 insertions in the middle of the repeat region were selected by SphI digestion and sequenced using primers pair tal-F/RP and FP/tal-R (Table S1).

Construction of Xss-V 2 -18 tal deletion mutants
The tal genes in Xcm Xss-V 2 -18 were deleted by homologous recombination using the suicide vector pKMS1 [67]. The 5' and 3' fragments that flank the CRR repeat in tal genes are conserved [61] and were used as sites for homologous recombination. The left-and right-flanking fragments of each tal gene were PCR-amplified using genomic DNA of Xcc strain Xcc049 (Table 1) as the template, and ligated into the MCS of pKMS1 [4], resulting in constructs pKMSA1 and pKMSA2, respectively. The new constructs were verified by restriction digestion and sequence analysis (TaKaRa, China). Constructs pKMSA1 and pKMSA2 were introduced into Xcm strain Xss-V 2 -18 by electroporation; cells were then plated on NAN medium supplemented with kanamycin and incubated at 28˚C for 4 days. Single colonies were then cultured in NBN broth at 28˚C to OD 600 ≤ 0.2, inoculated to NAS agar medium, and incubated for two days at 28˚C. Single colonies that grew on NAS were then transferred to NA and NA containing kanamycin. Colonies that grew on NA, but not on NA Km , were selected as potential deletion mutants. The mutants were then analyzed by Southern blot hybridization and PCR with primer pairs pKMSA1-5F/pKMSA1-3R and pKMSA2-5F/pKMSA2-3R (Table S1).

Southern hybridizations
Xcm plasmid and genomic DNA were extracted as described above. After BamHI digestion, DNA was separated on 1.2% agarose gels and then transferred onto Hybond N + nylon membranes (Roche, Germany). The 2898-bp internal SphI fragment of pthXo1 (GenBank accession no: AY495676) from Xoo [78] was labeled with digoxigenin (DIG) and used as a hybridization probe to detect the tal genes.
Probe labeling and Southern blotting were performed using the DIG Probe Synthesis Kit as recommended by the manufacturer's instructions (Roche, Sweden).

Virulence assays
Cotton cultivar TM-1 (G. hirsutum) was used in this study. Plants were grown in a greenhouse at 23˚C with a 12-h light/dark photoperiod and ~80% RH. Two-week-old plants were used in virulence assays.
Single colonies of Xcm were inoculated to 4 ml NA broth and cultured overnight at 28˚C. Bacterial cells were harvested by centrifugation (5000 rpm, 3 min); pellets were washed twice in sterile 10 mM MgCl 2 and then re-suspended in 10 mM MgCl 2 buffer to OD 600 =0.1. The suspensions were inoculated to the abaxial surface of leaves by infiltration with a sterile needleless syringe. Leaf phenotypes were examined 4-5 days after inoculation. All experiments were repeated three times.

Expression of tal2 gene in Xcm M2
The plasmid pZWavrXa7 (supplied by Dr. Bing Yang) was used to construct the plasmid for expression of tal2 in Xss-V 2 -18 strain. Plasmid pZWavrXa7 contains a FLAG-tag epitope immediately downstream of the second SphI site in the C-terminus of AvrXa7. The central SphI fragment of avrXa7 was replaced with the SphI fragment of Xss-V 2 -18 tal2 gene to generate pZW-tal2 ( Table 1). The recombinant plasmid was then fused with broad-host-range vector pHM1 at the HindIII site giving rise to pHZW-tal2. The constructs were transformed into Xcm M2 (∆tal2 strain, see below) by electroporation (2.5 kv, 4 ms).
The expression of tal2 in M2 was confirmed by western blotting with flag-labelled antisera. Briefly, the M2 strain containing pHZW-tal2 was cultured in NB to the logarithmic phase and harvested by centrifugation. The pellets were washed twice, and re-suspended in 1X PBS buffer to OD 600 = 1.0. SDS loading buffer (5X) was added to the bacterial suspensions and boiled in a water bath for 10 min.
Proteins were separated on 8% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes for immunoblotting using anti-FLAG (TransGene, Beijing, China) as the primary antibody.

Acknowledgments
We are thankful to Dr. Bing Yang, Iowa State University, for providing pHZWavrXa7.

Funding
This research work was supported by the National Natural Science Foundation of China (grant numbers; 31830072, 31230059).

Availability of data and materials
All the dataset generated or analyzed during this study are included in this published article. The nucleotide sequences have been deposited in GenBank under the following accession numbers (MK654746-MK654751). The plasmids are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
Not Applicable.

Consent for publication
Not Applicable.