XanFur, a novel Fur protein induced by H2O2, positively regulated by the global transcriptional regulator Clp and required for the full virulence of Xanthomonas oryzae pv. oryzae in rice

ABSTRACT Bacterial leaf blight caused by Xanthomonas oryzae pv. oryzae (Xoo) is one of the most destructive diseases in rice. However, the pathogenic mechanisms of Xoo have not been fully understood. In this study, we identified a novel ferric uptake regulator (Fur) protein, XanFur, in Xoo, which is conserved among the major pathogens of Xanthomonas species, but has no obvious sequence identity with reported Fur proteins. Gene deletion indicated that xanfur is required for the full virulence of Xoo in rice. The expression of xanfur was significantly induced by 0.1 mM H2O2 and its deletion caused a significant decrease in the tolerance of Xoo to H2O2, suggesting that xanfur is likely involved in the resistance of Xoo to the oxidative stress caused by H2O2 in rice. The loss of xanfur also caused significant reduction in extracellular polysaccharide production, biofilm formation, and cell motility, suggesting that xanfur may be also involved in regulating these virulence determinants of Xoo. Furthermore, both in vivo and in vitro experimental results demonstrated that the global transcriptional regulator Clp positively regulated the expression of xanfur by directly binding to its promoter region, especially in the presence of 0.1 mM H2O2, presenting the cellular regulatory pathway for the expression of this gene. These results not only contribute to a better understanding of the interaction mechanisms between rice and Xoo, but also provide reference to developing high-effective bactericides by targeting to the Fur protein in Xoo to control bacterial leaf blight in rice. IMPORTANCE Although Xanthomonas oryzae pv. oryzae (Xoo) has been found to be a bacterial pathogen causing bacterial leaf blight in rice for many years, the molecular mechanisms of the rice-Xoo interaction has not been fully understood. In this study, we found that XanFur of Xoo is a novel ferric uptake regulator (Fur) protein conserved among major pathogenic Xanthomonas species. XanFur is required for the virulence of Xoo in rice, and likely involved in regulating the virulence determinants of Xoo. The expression of xanfur is induced by H2O2, and positively regulated by the global transcriptional regulator Clp. Our results reveal the function and regulation of the novel virulence-related Fur protein XanFur in Xoo, providing new insights into the interaction mechanisms of rice-Xoo.

(Xoo) is one of the most destructive diseases in rice, resulting in up to 50% yield loss in the major rice-growing countries (5).Xoo usually infects rice through leaf hydatho des or wound sites and utilizes the nutritional sources in rice for its generation, resulting in tissue necrosis and wilting (6).The interaction between Xoo and rice has become a vital working model to elucidate how pathogens weaken or block the immune responses of host plants (7).
Recently, plenty of works has been done to study the molecular mechanisms of the rice-pathogens interaction.For example, the fungal pathogens Magnaporthe oryzae and Rhizoctonia solani depend on appressorium to enter the rice tissues, and produce effectors and RS toxin which is a mixture of N-acetyl glucosamine, N-acetyl galactosa mine, glucose, and mannose to cause rice blast and rice sheath blight, respectively (8,9).On the other hand, rice develops varieties of immune responses to prevent the infection of pathogens.For example, rice generates reactive oxygen species (ROS) to cause oxidative stress to the infected pathogens.It has been reported that the H 2 O 2 burst regulated by RbohB (respiratory burst oxidase protein B-like) is required for the resistance of rice to the fungal pathogen M. oryzae (10).It was also reported that H 2 O 2 is toxic to bacteria and causes damages to proteins and cell membranes (11).Therefore, the pathogens need to overcome the oxidative stress that resulted from H 2 O 2 to cause rice diseases.Works on the pathogenic mechanisms of Xoo have revealed that extracellular polysaccharides (EPS) protect Xoo against biotic or abiotic stresses, biofilm enhances the attachment of Xoo to different surfaces, cell motility allows Xoo to avoid toxic compounds, and type III secretion system contributes to the secretion of effectors into rice tissues, all of which are required determinants for the full virulence of Xoo in rice (12)(13)(14)(15).However, the molecular mechanisms of the rice-Xoo interaction have not been fully understood.For example, the mechanisms of which Xoo responses to the oxidative stress caused by H 2 O 2 in rice still remain unknown.
In plants, while the metal element availability is likely a limiting factor to the growth of pathogens, excess intracellular metal element is also toxic to themselves.Therefore, the metal homeostasis is crucial for the survival of pathogens and plays important roles in the plant-pathogen interaction (16,17).The Fur (ferric uptake regulator) family proteins are responsible for the homeostasis of the metal elements in pathogens, which have been divided into several groups, including the Fur, Zur (Zinc uptake regulator), manganese uptake regulator, nickel uptake regulator, and peroxide stress response regulator (18).In addition to their functions in regulating metal homeostasis, these Fur family proteins also regulate the expression of genes involved in the virulence and resistance to oxidative stress in pathogens.As the founding member of the Fur family proteins, the Fur proteins have been studied widely to understand its relationship with the virulence of pathogens.For example, the loss of the Fur protein from Xanthomonas campestris pv.campestris caused significant decrease in its virulence in Chinese cabbage and its tolerance to H 2 O 2 (19).The loss of the Fur protein also caused significant decrease in the virulence of Pectobacterium carotovorum subsp.brasiliense in potato, of Acidovorax citrulli in watermelon, and of Xanthomonas vesicatoria in tomato, and in the EPS production, biofilm formation, and cell motility of these pathogens (20)(21)(22).Although many Fur proteins have been found in pathogens with the development of biochemical, molecular genetics, and wide-genome assays, our understanding on their functions in regulating the virulence of the pathogens is still limited.
As for the Fur proteins in Xoo, Subramoni and Sonti reported that a Fur protein (accession no.AF146830) was required for the full virulence of Xoo in rice (23).In addition, the gene PXO_RS06520 (previously named gene PXO_04819) of Xoo was predicted to encode a Fur family transcriptional regulator, and interacting transcriptomes revealed that its expression in Xoo was significantly induced after the inoculation of Xoo in rice (24).However, the function of gene PXO_RS06520 in regulating the virulence of Xoo in rice still lacks investigation.
In this study, we found that gene PXO_RS06520 (named xanfur in this study) encodes a Fur protein (XanFur) that was highly conserved among the major pathogens of Xanthomonas species and its expression was significantly induced by H 2 O 2 .The loss of xanfur resulted in significant decrease in the full virulence of Xoo in rice, and in its EPS production, biofilm formation, cell motility, and tolerance to H 2 O 2 .Moreover, we found that the global transcriptional regulator Clp (cAMP receptor protein-like protein) positively controlled the expression of xanfur by directly binding to its promoter region.The results indicated that XanFur is a novel H 2 O 2 -induced Fur protein involved in the virulence of Xoo.

XanFur encoded by PXO_RS06520 of Xoo represents a new Fur family transcriptional regulator
The encoding region of gene PXO_RS06520 in the genome of Xoo is from 1407439 to 1407945, the genomic sequence length of PXO_RS06520 is 507 bp, and the amino acids size is 169 aa.Meanwhile, PXO_RS06520 encodes a protein containing only a Fur domain based on SMART prediction (http://smart.embl-heidelberg.de/),suggesting that gene PXO_RS06520 encodes a Fur protein (Fig. 1a).
These data indicated that XanFur and its homologs represent a novel Fur protein highly conserved among the major pathogens of Xanthomonas species, whose function is yet unknown.The gene PXO_RS06520 encoding XanFur was named xanfur hereafter for convenience.

The expression of xanfur is significantly induced by H 2 O 2
It has been reported that rice can form H 2 O 2 burst to resist fungal pathogen (10).To investigate whether the expression of xanfur in Xoo is induced by exterior H 2 O 2 , quantitative real-time PCR (qRT-PCR) and western blot were performed.The result of qRT-PCR showed that, compared to that in Xoo untreated with H 2 O 2 , the relative mRNA level of xanfur was significantly increased approximately sixfold in Xoo treated with 0.1 mM H 2 O 2 (Fig. 2a).Consistently, western blot analysis showed that the protein band of XanFur was significantly increased 1.9-fold in Xoo treated with 0.1 mM H 2 O 2 , compared to that in Xoo untreated with 0.1 mM H 2 O 2 .Meanwhile, the protein band of the internal control RNA polymerase (RNAP) exhibited no remarkable increase in Xoo treated with 0.1 mM H 2 O 2 , compared to that in Xoo untreated with 0.1 mM H 2 O 2 (Fig. 2b).These results indicated that the expression of xanfur in Xoo was significantly induced by H 2 O 2 .
In response to pathogen infection, rice can produce a high level of ROS, which is toxic to pathogens (31), and pathogens in turn can use ROS as signals to up-regulate the expression of virulence-associated genes (32).So far, it is yet unknown which genes in Xoo are induced by ROS.Since rice can form H 2 O 2 burst to resist pathogen (10), pathogens may also respond to the increased H 2 O 2 and up-regulate the expression of virulence-associated genes to cope with the oxidative stress caused by H 2 O 2 in their interaction with rice.Here, our result showed that the expression of xanfur in Xoo can be significantly induced by H 2 O 2 , implying that xanfur may be a virulence-related gene of Xoo.

XanFur is required for the full virulence of Xoo in rice
To investigate whether XanFur is involved in the virulence of Xoo in rice, a xanfur deletion mutant strain (Δxanfur) of Xoo and its complementary strain (CpΔxanfur) were success fully constructed (Fig. S1a).The deletion of xanfur caused no significant difference in the growth of wild type (WT) and Δxanfur in both NB liquid medium and MMX (a minimal medium without ammonium sulphate as the nitrogen source) liquid medium (Fig. S1b and c), indicating that xanfur is not necessary for the growth of Xoo.The effect of xanfur on the full virulence of Xoo was then analyzed by comparing the virulence of strains WT, Δxanfur, and CpΔxanfur in rice.The lesion length caused by the inoculation of WT was approximately 12 cm, but that caused by the inoculation of Δxanfur was only approximately 4 cm.The lesion length caused by the inoculation of CpΔxanfur displayed a similar level with that caused by WT (Fig. 3a and b).In addition, the cell population of Δxanfur in rice leaf was significantly lower than that of WT or CpΔxanfur in rice leaf 7 and 14 days post-inoculation (Fig. 3c).These data demonstrated that XanFur is required for the full virulence and the colonization of Xoo in rice, suggesting that xanfur is a virulence-related gene, and that XanFur is likely involved in the regulation of the virulence of Xoo to rice.Some Fur proteins have been reported to be involved in regulating the virulence of plant bacterial pathogens, such as those in Pectobacterium atrosepticum (32), A. citrulli AAC00-1 (22), and S. enterica serovar Typhimurium (27,33).In recent years, some Fur proteins have also been reported to be required for the virulence of Xanthomonas species, including those in X. vesicatoria, X. campestris pv.phaseoli, and X. campestris pv.campestris (19,20,25).Although XanFur exhibits no significant sequence identity to the reported Fur proteins (Fig. 1b), our results indicate that XanFur is also required for the virulence of Xoo in rice.
It has been shown that the reduced virulence of Xoo in rice was related to its growth deficiency caused by the deletion of some genes, such as xrvC, fur, and rnr (23,34,35).However, the loss of xanfur caused no significant deficiency in the growth of Xoo in both NB liquid medium and MMX liquid medium (Fig. S1b and c), but resulted in attenuated virulence in rice.This implies that xanfur may have different pathways in regulating the virulence of Xoo in rice from the reported genes xrvC, fur, and rnr, which, however, awaits further investigation.

XanFur is essential for the tolerance of Xoo to H 2 O 2
Since the expression of xanfur in Xoo was induced by H 2 O 2 (Fig. 2), XanFur may be involved in the tolerance of Xoo to H 2 O 2 .To support this hypothesis, we compared the H 2 O 2 tolerance of strains WT, Δxanfur, and CpΔxanfur.When cultivated on NB solid medium or in NB liquid medium without H 2 O 2 , strains WT, Δxanfur, and CpΔxanfur showed similar growth.However, when 0.1 mM or 0.2 mM H 2 O 2 was presented in NB media, the growth of Δxanfur was significantly inhibited, but that of WT or CpΔxanfur did not (Fig. 4a and b), indicating that the H 2 O 2 tolerance of Δxanfur was much weaker than that of WT.Thus, the deletion of xanfur resulted in a significant decrease in the H 2 O 2 tolerance of Xoo, suggesting that XanFur plays a role in the tolerance of Xoo to H 2 O 2 .
The ability to adapt to adverse environmental stresses is essential for the infection of pathogens.Some Fur proteins have been shown to be able to affect the toler ance of Xanthomonas species to H 2 O 2 .For example, the Fur protein (accession no.WP_005988392) affected the H 2 O 2 tolerance of X. campestris pv.campestris in Chinese cabbage (19), and the Fur protein (accession no.AF146830) affected the H 2 O 2 tolerance of Xoo in rice (23).The deletion of fur from A. citrulli AAC00-1 (22) and P. aeruginosa PA103 (29) caused an increased sensitivity of these pathogens to H 2 O 2 .Consistent with these Fur proteins, XanFur is also required for the tolerance of Xoo to H 2 O 2 , which suggests that XanFur is likely involved in the resistance of Xoo to the oxidative stress caused by H 2 O 2 in rice.

XanFur is essential for some important virulence determinants of Xoo
It has been shown that EPS, biofilm, and cell motility are required determinants for the full virulence of Xoo to rice (12).To investigate whether XanFur is involved in the regulation of these virulence determinants of Xoo, we compared the EPS production, biofilm formation, and cell motility of strains WT, Δxanfur, and CpΔxanfur.Biofilm formation is associated with the attachment of pathogens to different environmental surfaces (36).Thus, we evaluated the effect of xanfur deletion on the biofilm formation of Xoo.The confocal laser scanning microscope observation showed that the thickness of the biofilm formed by strain WT was approximately 80 µM with a well-organized structure, but that formed by strain Δxanfur was only approximately 60 µM with a less organized structure (Fig. 5a and b).Consistently, the biofilm formed by Δxanfur (approximately 0.2 of optical density at 590 nm [OD 590 ]) on the surface of a polystyrene tube was significantly smaller than that formed by WT (approximately 1.25 of OD 590 ) or CpΔxanfur (approximately 1.32 of OD 590 ) (Fig. 5c).Thus, the loss of xanfur caused significant decrease in biofilm formation of Xoo.EPS, a virulence determinant of Xoo, is involved in the biofilm formation of Xoo (12).Therefore, we further investigated the effect of xanfur deletion on the EPS production of Xoo.Strains WT, Δxanfur, and CpΔxanfur were cultured in NB liquid medium to compare their EPS production.The results showed that the EPS produced by Δxanfur (approximately 0.18 mg/mL) was significantly lower than that produced by WT or CpΔxanfur (both approximately 0.25 mg/mL) (Fig. 5d), indicating that the deletion of xanfur significantly weakened the EPS production of Xoo.
Cell motility is another important virulence determinant for Xanthomonas species to avoid themselves from the unfavorable environments.To analyze the effect of xanfur deletion on the cell motility of Xoo, we compared the swimming and swarming motility of strains WT, Δxanfur, and CpΔxanfur.As shown in Fig. 6a, the swarming zone produced by Δxanfur (approximately 2.46 cm 2 ) on the NB solid medium plate with 0.6% agar was significantly smaller than those produced by WT (approximately 6.34 cm 2 ) and CpΔxanfur (approximately 6.18 cm 2 ).Similarly, the swimming zone produced by Δxanfur (approximately 3.20 cm 2 ) was also significantly smaller than those produced by WT (approximately 7.69 cm 2 ) and CpΔxanfur (approximately 7.59 cm 2 ) (Fig. 6b).These data indicated that the loss of xanfur significantly decreased the swimming and swarming motility of Xoo.
Altogether, these results indicated that XanFur is essential for the virulence-related determinants of Xoo, including EPS production, biofilm formation, and cell motility, suggesting that this regulator is involved in the regulation of these virulence-related determinants, directly or indirectly.
It has been shown that the deletion of fur from P. aeruginosa PAO1 had no effect on its biofilm formation (28).However, the deletion of fur from X. vesicatoria (20), A. citrulli AAC00-1 (22), or C. violaceum ATCC 12472 ( 26) not only caused significant decrease in biofilm formation, but also significant decrease in EPS production and cell motility of these strains, suggesting that the fur-encoding proteins are involved in the regula tion of these virulence-related determinants in these strains.Similarly, the deletion of xanfur from Xoo also caused significant decrease in EPS production, biofilm formation, and cell motility.Thus, XanFur also plays an important role in regulating the virulence determinants of Xoo.However, how XanFur and other Fur proteins regulate the virulence determinants of these pathogens awaits further investigation.

Clp positively regulates the expression of xanfur by directly binding to its promoter region
Clp is a conserved global regulator among the pathogenic Xanthomonas species.It has been shown that the expression of some virulence-related genes can be regulated by Clp (37,38).However, no fur gene has been reported to be regulated by Clp.To investigate whether the expression of xanfur could be regulated by Clp in Xoo, qRT-PCR and western blot were performed.The result of qRT-PCR showed that when strains were cultured in NB liquid medium without H 2 O 2 , the relative mRNA level of xanfur in Δclp was decreased approximately threefold compared to that in WT.However, when strains were cultured in NB liquid medium containing 0.1 mM H 2 O 2 , the relative mRNA level of xanfur in Δclp was decreased approximately 18-fold compared with that in WT (Fig. 2a).Consistently, the result of western blot also showed that the protein band of XanFur was decreased 2.3-fold in Δclp cultured in NB liquid medium without H 2 O 2 , and 4.4-fold in Δclp cultured in the presence of 0.1 mM H 2 O 2 , compared to that in WT with the same culture condition.Meanwhile, the protein band of the internal control RNAP exhibited no remarkable reduction in Δclp cultured with or without 0.1 mM H 2 O 2 , compared to that in WT with the same culture condition (Fig. 2b).These results indicated that the expression of xanfur was positively regulated by Clp, especially in the presence of 0.1 mM H 2 O 2 .
It has been shown that the deletion of clp from Xoo resulted in decrease in cell motility, EPS production, and virulence in rice (38).To further determine whether xanfur is a downstream component of clp, we constructed the xanfur deletion mutant strain of Δclp (Δxanfur&clp) and its complementary strain [CpΔxanfur(Δclp)] (Fig. S1a), and compared the growth, cell motility, EPS production, and virulence of strains WT, Δclp, Δxanfur&clp, and CpΔxanfur(Δclp).The growth of strains Δclp, Δxanfur&clp, and CpΔxanfur(Δclp) in NB liquid medium and MMX liquid medium was only a little weaker compared with that of WT under the same conditions (Fig. S1b and c).While strains Δclp, Δxanfur&clp, and CpΔxanfur(Δclp) all showed a significantly decreased virulence in rice compared with that caused by strain WT, the virulence caused by Δxanfur&clp was not significantly weaker than that caused by Δclp, indicating that the deletion of xanfur from Δclp did not cause a further significant decrease in the virulence (Fig. S2).In addition, the deletion of xanfur from Δclp also caused no significant decrease in biofilm formation, EPS production, or cell motility (Fig. S3).All the data demonstrated that xanfur is a downstream component of Clp in Xoo.
To verify the result of B1H, we purified the GST (glutathione-S-transferase) protein, and the GST-tagged Clp protein (Clp-GST), and analyzed their interactions with the free DNA (biotin-labeled promoter of xanfur) and the cold DNA (biotin unlabeled promoter of xanfur) by using EMSA (electrophoretic mobility shift assay).The result showed that while the band of complex representing the combination of the GST protein with the free DNA was not detectable, the band of complex representing the combination of the Clp-GST protein with the free DNA was detected and exhibited remarkable increase with an increased application of the Clp-GST protein from 0 µM to 0.15 µM.Meanwhile, the band of the free DNA exhibited remarkable decrease with the increased application of the Clp-GST protein.In addition, when excess cold DNA was present, the band of complex representing the combination of the Clp-GST protein with the free DNA exhibited remarkable decrease and the band of free DNA exhibited remarkable increase with an increased application of the cold DNA (100-, 200-, and 400-folds of free DNA) (Fig. 7b).Moreover, the interaction between Clp and the promoter of xanfur was confirmed by using MST (microscale thermophoresis), in which Clp-GST was found to bind to the promoter of xanfur with a high affinity (K d , 0.687 ± 0.0491 nM).In contrast, no affinity between GST and the promoter of xanfur was found (Fig. 7c).These results further demonstrated that Clp can bind the promoter region of xanfur.
Studies have reported that Clp can up-regulate the expression of enzyme genes related to the virulence of X. campestris pv.campestris, including pmeA encoding a pectin methylesterase, manA encoding a endo-1,4-β-mannosidase, and pelA1 encoding a pectate lyase (39)(40)(41).In addition, it has been reported that Clp regulated the expression of virulence-related genes by binding to their promoter regions, such as hshB encoding a hydrolase in X. oryzae pv.oryzicola (42).Studies also have reported that the Clp binding site exhibits perfect twofold sequence symmetry, such as the 5´-ATCC-N8-GGAT-3´ and 5´-ATCG-N8-CGAT-3´ motifs of the heat-stable antifungal factor (HSAF) biosynthesis operon in Lysobacter enzymogenes, the 5´-CAC-N8-GTG-3´ motif of PXO_03177 which is a virulence-related gene in Xoo (37,43).Our results in this study showed that Clp also positively regulated the expression of xanfur by binding to its promoter region, but there is no 5´-ATCC-N8-GGAT-3´, 5´-ATCG-N8-CGAT-3´, or 5´-CAC-N8-GTG-3´ motif found in the promoter region.Then a 5´-TGGG-CAAGGTCG-CCCA-3´ motif was found in the promoter region of xanfur, exhibiting 93.75% similarity to the Clp binding site (5´-TGGG-CAAGGTGG-CCCA-3´) of zur gene in X. campestris pv.campestris (44).This indicated that the 5´-TGGG-N8-CCCA-3´ motif maybe the potential Clp binding site.Thus, it seems that Clp regulates a variety of virulence-related genes through different binding sites in pathogenic Xanthomonas species.

Conclusion
Although Xoo has been found to be a bacterial pathogen causing bacterial leaf blight in rice for many years, the molecular mechanisms of the rice-Xoo interaction has not been fully understood.In this study, we found a novel Fur protein, XanFur, conserved in pathogenic Xanthomonas species, which is required for the full virulence of Xoo in rice.Moreover, exterior H 2 O 2 significantly induces the expression of xanfur in Xoo, and xanfur is essential for the tolerance of Xoo to H 2 O 2 , suggesting that XanFur is likely involved in the response and resistance of Xoo to the oxidative stress caused by H 2 O 2 in rice.xanfur is also essential for some important virulence determinants of Xoo, including EPS production, biofilm formation, and cell motility, further demonstrating that XanFur is likely involved in regulating the virulence of Xoo.In addition, it was found that Clp positively regulates the expression of xanfur by directly binding to its promoter region, especially in the presence of H 2 O 2 .These findings not only contribute to a better understanding of the interaction mechanisms between rice and Xoo, but also provide reference to developing high-effective bactericides by targeting to the Fur protein in Xoo to control bacterial leaf blight in rice.

Strains, plasmids, and culture conditions
The bacterial strains and plasmids used in this study were listed in Table S1.The WT Xoo strain PXO99 A was preserved in our lab, which was named strain WT Xoo in this study.The clp deletion mutant strain of WT Xoo was constructed previously in our lab, which was named strain Δclp in this study.The xanfur deletion mutant strains of WT Xoo and Δclp were constructed in this study, which were named strains Δxanfur and Δxanfur&clp, respectively.The Δxanfur complementary strains of Δxanfur and of Δxanfur&clp were constructed in this study, which were named strains CpΔxanfur and CpΔxanfur(Δclp), respectively.All these strains were grown in NB liquid medium or MMX liquid medium or on NB solid medium at 28°C (45).Escherichia coli was grown in Luria-Bertani (LB) medium or on M9-based medium at 37°C (43).Antibiotics of kanamycin (50 µg/mL), spectino mycin (25 µg/mL), and gentamicin (25 µg/mL) were added to the growth medium as appropriate for selection.

Generation of the gene xanfur deletion mutant and complementary strains
Generation of the in-frame deletion mutants of gene xanfur was carried out by using the strains WT and Δclp as the parental strains via allelic homologous recombina tion, according to the previously described protocol (45).Here, two flanking regions were amplified by PCR using the corresponding primer pairs (Table S2), which then were digested with the related restriction enzymes, and were integrated into plasmid pK18mobsacB.The integrated plasmid was electroporated into the strains WT and Δclp.Transformants were selected on NB solid medium plates containing 50 µg/mL kanamycin but without sucrose for the first cross-over event.The well-grown colonies were further cultured on nutrient broth with agar (NA) solid medium plate containing 10% (wt/vol) sucrose to obtain cells undergoing the secondary cross-over event.After two rounds of recombination, the resulting xanfur deletion mutant strains, Δxanfur and Δxanfur&clp, were confirmed by using PCR.For complementation, the gene xanfur with its predicted promoter region was amplified by using PCR with specific primers (Table S2), and then was cloned into plasmid pUFR047.The integrated plasmid was electroporated into the mutant strains Δxanfur and Δxanfur&clp to generate the complemented strains CpΔxanfur and CpΔxanfur(Δclp), respectively.

Bacterial growth assays in NB liquid medium and MMX liquid medium
The Xoo strains [WT, Δxanfur, Δclp, Δxanfur&clp, CpΔxanfur, and CpΔxanfur(Δclp)] were cultured in NB liquid medium to OD 600 = 1.0, and then 500 µL bacterial suspensions were transferred into 250 mL flasks containing 50 mL sterilized NB liquid medium, which were incubated at 28℃ with shaking at 220 rpm.The value of OD 600 of the bacterial suspensions was recorded every 2 h for 36 h.In addition, the cells of the Xoo strains were collected after the centrifugation at 6,000 g for 10 min when OD 600 of the bacterial suspensions were 1.0.Then, the collected cells were suspended with MMX liquid medium to OD 600 = 1.0, and 5 mL bacterial suspensions were transferred into 250 mL flasks containing 50 mL sterilized MMX liquid medium, which were incubated at 28℃ with shaking at 220 rpm.The value of OD 600 of the bacterial suspensions was recorded every 8 h for 5 days.Each treatment was repeated at least three times.

Virulence analysis in rice
The virulence of Xoo in rice was analyzed by using the leaf-clipping method (45).Briefly, the susceptible rice cultivar IR24 was planted in a growth chamber under a cycle of 16 h of light at 28°C and 8 h of dark at 25°C.The Xoo strains [WT, Δxanfur, Δclp, Δxanfur&clp, CpΔxanfur, and CpΔxanfur(Δclp)] were cultured in NB liquid medium to OD 600 = 0.1, and then the cells of the Xoo strains were collected after the centrifugation at 6,000 g for 10 min.Then the collected cells were suspended with ddH 2 O to OD 600 = 0.1, which were used for inoculation.Five-week-old rice leaves were inoculated with the bacterial suspensions by using scissors.Lesion lengths of the inoculated rice leaves were measured after 14 days.Meanwhile, 1 cm 2 rice leaves were clipped from the inoculation sites and were ground in a 10 mL tube with 5 mL sterile distilled water.The ground suspensions were diluted 10 5 -folds, and 100 µL diluted suspensions were coated on NB solid medium plates, which were cultured at 28°C for 3 days.Then, the population of the Xoo strains (WT, Δxanfur, and CpΔxanfur) on NB solid medium plates were investigated.Fifty rice leaves were inoculated and four of them were clipped in each treatment.Each treatment was repeated at least three times.

Determination of the tolerance to H 2 O 2
The tolerance of Xoo to H 2 O 2 was determined with the method described previously (37).Briefly, the Xoo strains (WT, Δxanfur, and CpΔxanfur) were cultured in NB liquid medium to to OD 600 = 1.0, and then 80 µL bacterial suspensions were transferred into 8 mL sterilized NB liquid medium containing 0 mM, 0.1 mM, or 0.2 mM H 2 O 2 , which were incubated at 28°C with shaking at 220 rpm.The growth of the Xoo strains was observed every 12 h after incubation for 2 days.Meanwhile, the bacterial suspensions were diluted fivefolds (OD 600 = 0.2) or 25-folds (OD 600 = 0.04), and then 3 µL diluted bacterial suspensions were dropped onto the surface of NB solid medium plates containing 0 mM, 0.1 mM, or 0.2 mM H 2 O 2 , which were incubated at 28°C without shaking.The growth of Xoo strains were observed after incubation for 2 days.Each treatment was repeated at least three times.

Measurement of EPS production
The EPS production of Xoo was measured with the method described previously (48).Briefly, the Xoo strains [WT, Δxanfur, Δclp, Δxanfur&clp, CpΔxanfur, and CpΔxanfur(Δclp)] were cultured in NB liquid medium to OD 600 = 1.0, and then 500 µL bacterial suspensions were transferred into 250 mL flasks containing 50 mL sterilized NB liquid medium, which were incubated at 28°C with shaking at 220 rpm.After 5-day incubation, the bacterial suspensions were centrifuged at 12,000 g for 10 min.The collected supernatants of the Xoo strains were mixed with 100 mL ethanol, and incubated at room temperature for 24 h to precipitate the EPS.The precipitated EPS was dried at 70°C, and then weighed using a digital analytical balance.Each treatment was repeated at least three times.

Biofilm formation assay
The biofilm formation of Xoo was measured with the method described previously (48).Briefly, the Xoo strains [WT, Δxanfur, Δclp, Δxanfur&clp, CpΔxanfur, and CpΔxanfur(Δclp)] were cultured in NB liquid medium to OD 600 = 1.0, and then 40 µL bacterial suspensions were transferred into sterilized polystyrene tubes containing 4 mL NB liquid medium and incubated at 28°C for 24 h with shaking at 220 rpm.The polystyrene tubes were subsequently kept in a growth chamber at 28°C for 7 days without shaking, and then were gently washed three times with water after discarding the bacterial suspensions.The biofilm formed on the wall of the polystyrene tubes was stained with 0.1% crystal violet for 20 min.After that, the stained biofilm on the polystyrene tubes was dissolved in 4 mL acetic acid/ethanol (1:4, vol/vol), and the value of OD 590 of the solutions were measured by using an Agilent 8453 UV-visible spectrophotometer (Agilent Technologies, USA).The value of OD 590 was used as an indicator of the production of biofilm.
The 3D structure of bacterial biofilm was analyzed by using confocal laser scanning microscope, according to the previously described protocol (49).The plasmid pUFZ75 previously constructed for the expression of the green fluorescent protein (GFP) (50) was electroporated into the Xoo strains (WT and Δxanfur).The GFP-labeled Xoo strains were cultured in NB liquid medium with 10 mM kanamycin to OD 600 = 1.0, and then 200 µL bacterial suspensions were transferred into the flow chambers covered with glass slides.The flow chambers were kept in a humidified condition at 28°C for 48 h without shaking.Biofilm formation was visualized by using a confocal laser scanning microscope (Leica Microsystems Inc., USA) with an excitation wavelength of 488 nm and an emission wavelength of 500 to 545 nm.The images were analyzed by using the software of LAS_X_Small_2.0.0_14332.Each treatment was repeated at least three times.

Measurement of cell motility
The cell motility was measured with the method described previously (48).The Xoo strains (WT, Δxanfur, Δclp, Δxanfur&clp, CpΔxanfur and CpΔxanfur(Δclp)) were cultured in NB liquid medium to OD 600 = 1.0, and then 3µL bacterial suspensions were drop ped onto the surface of NB solid medium plates containing 0.6% (wt/vol) agar (for swarming motility) or 0.3% (wt/vol) agar (for swimming motility).The swarming zones and swimming zones were calculated after the incubation of the Xoo strains on the plates at 28°C for 2 or 3 days without shaking.The calculated zone was used as an indicator of the cell motility.Each treatment was repeated at least three times.

qRT-PCR analysis
The mRNA level expression of gene in bacteria was measured with the method described previously (51).The Xoo strains (WT and Δclp) were cultured in NB liquid medium to OD 600 = 1.0, and then 80 µL bacterial suspensions were transferred into 8 mL sterilized NB liquid medium with or without 0.1 mM H 2 O 2 and incubated at 28°C for 24 h with shaking at 220 rpm.The cells of the Xoo strains were collected by centrifugating at 10,000 g for 10 min.Total RNA was extracted from the collected Xoo cells by using the TRIzol reagent (ThermoFisher, USA), and cDNA was synthesized from 100 ng total RNA by using the TransScript All-in-One First-Strand cDNA Synthesis SuperMix Kit (Trans Gen Biotech, China).qRT-PCR was performed in a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystem, USA).The gene relative expression ratio was calculated by the method of 2 -ΔΔCt .The 16S rRNA gene was used as the endogenous control.Each treatment was repeated at least three times.

Western blot analysis
Western blot analysis of the protein expression in Xoo was performed with the method described previously (37).Briefly, the Flag tag (GAT TAC AAG GAT GAC GAC GAT AAG) was fused to the C terminus of gene xanfur, and then the integrated fragment was cloned into the plasmid pUFR047 which was previously constructed (52).The integrated plasmid pUFR047-xanfur-Flag was electroporated into the strains of Xoo (WT and Δclp).Then, the cells of the Flag tagged Xoo strains were cultured in 50 mL sterilized NB medium with or without the presence of 0.1 mM H 2 O 2 to OD 600 = 1.0, and then the cells were collected by centrifugation at 6,000 g for 10 min.The total proteins were extracted from the collected cells by using 1 mL radioimmunoprecipitation assay lysis buffer containing 3 µL phenylmethanesulfonyl fluoride and 3 µL protease inhibitor cocktail.To obtain the soluble proteins, the supernatant was harvested after centrifugation at 10,000 g for 10 min.The soluble proteins were separated by SDS-PAGE and immobi lized onto a polyvinylidene difluoride membrane by using the semi-dry blot machine (Bio-Rad, USA).The membrane was incubated in the blocking buffer containing the anti-Flag mouse primary antibody (1:5,000, Abmart, China), followed by an incubation in new blocking buffer containing the horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody (1:10,000, Abmart, China).After that, the membrane was stained using the Super ECL Plus kit and imaged using FluorChem M (Alpha Innotech, USA).The α-subunit of RNA polymerase was used as a control for sample loading.Each treatment was repeated at least three times.

Protein expression and purification
Protein expression and purification were performed with the method described previously (43).Briefly, the coding region of clp was amplified from Xoo by PCR, and then was cloned into the plasmid pGEX-6P-1 which was previously constructed (53).The integrated plasmid pGEX-6p-1-Clp was transformed into E. coli strain BL21 (DE3) by heat shock for protein expression.The transformed strain was cultured in 400 mL LB liquid medium in the presence of 100 µg/mL ampicillin at 37°C with shaking at 220 rpm to OD 600 = 0.4.Then, isopropyl β-D-1-thiogalactopyranoside at a final concentration of 0.4 mM was added to the LB medium, followed by a further incubation at 28°C with shaking at 220 rpm for 4 h.The E. coli cells were then lysed by sonication in 30 mL phosphate-buffered saline (PBS) lysis buffer containing 1 mM phenylmethanesulfonyl fluoride.The crude extraction was centrifuged at 13,000 g for 5 min.The collected supernatant was mixed with Glutathione Sepharose 4B (GE) at 4°C for 12 h.The Clp-GST protein was eluted using the elution buffer.The concentration of the purified Clp-GST protein was determined using a Bradford protein assay kit (Bio-Rad, USA).

B1H assay
The in vivo interaction between the Clp protein and the promoter of xanfur was analyzed by using EMSA, according to the previously described protocol (43).The coding region of clp and the promoter region of xanfur was, respectively, cloned into plasmids pTRG and pBXcmT which were previously constructed (54).The two integrated plasmids pTRG-clp and pBXcmT-xanfur were transformed into the E. coli strain XL1-Blue MRF' kan by heat shock, and then the recombinant cells were cultured on M9-based medium plates containing 12.5 µg/mL tetracycline, 34 µg/mL chloramphenicol, 30 µg/mL kanamycin, 0 or 5 mM/L 3-amino-1,2,4-triazole and 0 µg/mL or 8 µg/mL streptomycin at 28°C.The well-grown cells indicated that direct physical interaction occurred between Clp and the promoter of xanfur.The cell of E. coli containing the plasimd pBXcmT-xanfur and the empty plasmid pTRG was used as the negative control.The cell of E. coli contain ing plasmids pBXcmT-R2031 and pTRG-R3133 was used as the positive control.Each treatment was repeated at least three times.

EMSA assay
The in vitro interaction between the Clp protein and the promoter of xanfur was analyzed by using EMSA, according to the previously described protocol (43).The amplified biotin-labeled fragment of the xanfur promoter and the purified protein of Clp-GST were mixed and incubated according to the instruction of the LightShift EMSA Optimization & Control Kit (ThermoFisher, USA).The binding mixture was electrophoresed into the polyacrylamide gel, and then was transferred to a nylon membrane for nucleic acid blotting by using a semi-dry blot machine (Bio-Rad, USA).The biotinylated fragment of xanfur and its complex with Clp-GST protein on the membrane were detected by chemiluminescence using the VersaDoc imaging system (Bio-Rad, USA).The biotin-unla beled fragment of the xanfur promoter and the purified GST protein were used as the negative controls.Each treatment was repeated at least three times.

MST measurement
Binding of the Clp-GST fusion protein to the promoter of xanfur was determined by MST using Monolith NT.115 (NanoTemper Technologies, Germany), according to the previously described protocol (43).Briefly, FAM (5-carboxyfluorescein)-labeled fragment of the xanfur promoter was amplified by PCR.A constant concentration (10 µM) of the FAM-labeled xanfur promoter in the MST buffer (50 mM Tris pH 7.5, 150 mM NaCl, 10 mM MgCl 2 , 0.05% Tween 20) was titrated against increasing concentration of the purified Clp-GST protein dissolved in the MST buffer.MST premium-coated capillaries (Monolith NT.115 MO-K005, Germany) were used to upload the samples into the MST instrument at 25℃ using 40% MST power and 20% light-emitting diode (LED) power.The time of laser on was 30 s and laser off was 5 s.Data were analyzed by using Nanotemper Analysis software v.1.2.101 (NanoTemper Technologies, Germany).The purified GST protein was used as the negative control.The value of K d represented the binding ability of Clp-GST or GST with the fragment of the xanfur promoter.Each treatment was repeated at least three times.

FIG 2
FIG 2 The effect of H 2 O 2 on the expression of XanFur in Xoo strains WT and Δclp.(a) The relative mRNA level of xanfur in Xoo strains WT and Δclp detected by using qRT-PCR.(b) The representative detection of XanFur expressed in Xoo strains WT and Δclp by using western blot.The upper bands corresponding to the predicted size of xanfur-Flag fusion were detected by anti-Flag monoclonal antibody, and the lower bands corresponding to the predicted size of RNA polymerase α-subunit were detected by the specific antibody anti-RNAP.WT, the wild-type Xoo strain PXO99 A .Δclp, the clp deletion mutant strain of WT.Strains were cultured in nutrient broth (NB) liquid medium with or without 0.1 mM H 2 O 2 to OD 600 = 1.0.The experiment was independently repeated three times.Values were the means ± SDs from three independent experiments.The asterisks above the error bars indicated significant differences compared with the expression in strain WT which was incubated without 0.1 mM H 2 O 2 (t-test, **P < 0.01).

FIG 3
FIG 3 The effect of xanfur deletion on the full virulence of Xoo in rice.(a) The representative lesion lengths in rice leaves inoculated with the Xoo strains WT, Δxanfur, and CpΔxanfur after 14 days.(b) The calculated lesion lengths in rice leaves inoculated with the Xoo strains WT, Δxanfur, and CpΔxanfur after 14 days.(c) The investigated bacterial population in rice leaves inoculated with the Xoo strains WT, Δxanfur, and CpΔxanfur after 0, 7, and 14 days.WT, the wild-type Xoo strain PXO99 A .Δxanfur, the xanfur deletion mutant strain of WT.CpΔxanfur, the Δxanfur complementary strain of WT.DPI, day post-inoculation.The experiment was independently repeated three times.Values are the means ± SDs from three independent experiments.The asterisks above the error bars indicate significant differences compared with the wild-type strain (t-test, **P < 0.01).

FIG 4
FIG 4 The effect of xanfur deletion on the tolerance of Xoo to H 2 O 2 .(a) The representative growth of Xoo strains WT, Δxanfur, and CpΔxanfur on NB solid medium plates with 0 mM, 0.1 mM, or 0.2 mM H 2 O 2 for 2 days.1×, 5×, and 25×, the bacterial suspension was diluted 1-, 5-, and 25-fold.(b) The representative growth of Xoo strains WT, Δxanfur, and CpΔxanfur in 8 mL NB liquid medium with 0 mM, 0.1 mM, or 0.2 mM H 2 O 2 for 36 h.WT, the wild-type strain of Xoo PXO99 A .Δxanfur, the xanfur deletion mutant strain of WT.CpΔxanfur, the Δxanfur complementary strain of WT.The experiment was independently repeated three times.

FIG 5
FIG 5 The effect of xanfur deletion on the biofilm formation and EPS production of Xoo.(a and b) The representative biofilm formed by Xoo strains WT and Δxanfur.The representative thickness of the biofilm (up, from the angles of X-and Z-axis) and the representative organized structure of the biofilm (down, from the angles of X and Y axis) were analyzed by using confocal laser scanning microscopy.Images were obtained by using a 20× objective.(c) The representative biofilm formed by Xoo strains WT, Δxanfur, and CpΔxanfur on polystyrene tubes stained with crystal violet (left) and the calculated biofilm production (right).The values of OD 590 stand for biofilm production.(d) The EPS production of Xoo strains WT, Δxanfur, and CpΔxanfur which were cultured in NB liquid medium for 5 days.WT, the wild-type strain of Xoo PXO99 A .Δxanfur, the xanfur deletion mutant strain of WT.CpΔxanfur, the Δxanfur complementary strain of WT.The experiment was independently repeated three times.Values are the means ± SDs from three independent experiments.The asterisks above the error bars indicate significant differences compared with the wild-type strain (t-test, **P < 0.01).

FIG 6
FIG 6 The effect of xanfur deletion on the cell motility of Xoo.(a) The representative swarming zone (up) and the calculated area of the swarming zone (down) produced by Xoo strains WT, Δxanfur, and CpΔxanfur which were cultured on NB solid medium plates containing 0.6% agar for 3 days.(b) The representative swimming zone (up) and the calculated area of the swimming zone (down) produced by Xoo strains WT, Δxanfur, and CpΔxanfur which were cultured on NB solid medium plate containing 0.3% agar for 2 days.WT, the wild-type strain of Xoo PXO99 A .Δxanfur, the xanfur deletion mutant strain of WT.CpΔxanfur, the Δxanfur complementary strain of WT.The experiment was independently repeated three times.Values are the means ± SDs from three independent experiments.The asterisks above the error bars indicate significant differences compared with the wild-type strain (t-test, **P < 0.01).

FIG 7
FIG 7 Detection of the interaction between Clp and the promoter of xanfur in vivo and in vitro.(a) The representative in vivo interaction between Clp and the promoter of xanfur analyzed by using B1H in E. coli.pTRG-R2031/pBXcmT-R3133, the cells containing plasmids pBXcmT-R2031 and pTRG-R3133 used as the positive control.pTRG/pBXcmT-xanfur, the cells containing the plasmid pBXcmT-xanfur and the empty plasmid pTRG used as the negative control.pTRG-clp/ pBXcmT-xanfur, the cells containing plasmids pTRG-clp and pBXcmT-xanfur.−3-AT − Str, the M9-based solid medium plates without 3-amino-1,2,4-triazole and streptomycin.+3-AT + Str, the M9-based solid medium plates with 5 mM/L 3-amino-1,2,4-triazole and 8 µg/mL streptomycin.(b) The representative in vitro interaction between Clp and the promoter of xanfur analyzed by using electrophoretic mobility shift assay.The arrows indicate the complex bands of free DNA with Clp (up) and the bands of free DNA (down).Free DNA, the biotin-labeled promoter of xanfur.Complex, Clp interacting with the biotin-labeled promoter of xanfur.Cold DNA, the biotin-unlabeled promoter of xanfur.100×, 200×, and 400×, the cold DNA used to competitively inhibit the interaction between Clp and free DNA was 100-, 200-, or 400-folds of the free DNA concentration.The glutathione-S-transferase (GST) protein was used as the negative control.(c) The representative in vitro interaction between Clp and the 5-carboxyfluorescein-labeled promoter of xanfur by using microscale thermophoresis.The value of K d represents their binding capacity.The GST protein was used as the negative control.The experiment was independently repeated three times.Values are the means ± SDs from three independent experiments.