Polyamine signaling communications play a key role in regulating the pathogenicity of Dickeya fangzhongdai

ABSTRACT Dickeya fangzhongdai is a devastating bacterial pathogen infecting a wide range of crops and ornamental plants worldwide. As a newly identified bacterial species in 2016, the regulatory mechanisms that govern its virulence are still a mystery. In this study, we explored the potential roles of polyamine-mediated cell-to-cell communication in regulation of D. fangzhongdai virulence. Null mutation of speA and speC in D. fangzhongdai strain ZXC1, which encodes polyamine biosynthesis through arginine and ornithine pathways, respectively, dramatically reduced bacterial motility, decreased production of plant cell wall degradation (PCWD) enzymes, and attenuated the bacterial virulence on taro and potato. We then tested the effect of various polyamine molecules in the restoration of the mutant phenotypes and showed that putrescine was the most potent signal in the regulation of virulence traits in strain ZXC1. In addition, we found that taro extract contained active signals to rescue putrescine-deficient phenotypes. High-performance liquid chromatography mass spectrometry analysis validated the speA was essential for production of putrescine in D. fangzhongdai ZXC1. We further showed that the putrescine transporters PotF and PlaP are required for putrescine-mediated cell-to-cell communication and virulence against taro and potato tubers. quantitative reverse transcription-PCR analysis demonstrated that putrescine influences the pathogenicity of D. fangzhongdai ZXC1 by regulating the expression of PCWD enzymes, bacterial chemotaxis, and flagellar-related genes. The findings from this study shed a new light for elucidating the pathogenic mechanisms of D. fangzhongdai and present useful clues for developing relevant disease control strategies. IMPORTANCE Dickeya fangzhongdai is a newly identified plant bacterial pathogen with a wide host range. A clear understanding of the cell-to-cell communication systems that modulate the bacterial virulence is of key importance for elucidating its pathogenic mechanisms and for disease control. In this study, we present evidence that putrescine molecules from the pathogen and host plants play an essential role in regulating the bacterial virulence. The significance of this study is in (i) demonstrating that putrescine signaling system regulates D. fangzhongdai virulence mainly through modulating the bacterial motility and production of PCWD enzymes, (ii) outlining the signaling and regulatory mechanisms with which putrescine signaling system modulates the above virulence traits, and (iii) validating that D. fangzhongdai could use both arginine and ornithine pathways to synthesize putrescine signals. To our knowledge, this is the first report to show that putrescine signaling system plays a key role in modulating the pathogenicity of D. fangzhongdai.

Most studies on the pathogenic mechanisms of Dickeya pathogens were focused on Dickeya dadantii and Dickeya oryzae.D. dadantii infects and cause various plant soft rot diseases by producing an array of virulence factors, including cell wall-degrading enzymes, bacterial motility, extracellular polysaccharide, blue pigment indigoidine, iron assimilation system, and type III secretion system (23).The above virulence traits are also well conserved in D. oryzae, which, in addition, also produces a family of phytotoxins known as zeamines (24,25).Noticeably, zeamines and bacterial motility were identified as the two key virulence determinants of D. oryzae.The zemine-minus mutants of D. oryzae could hardly cause infection in rice (25,26), whereas its motility mutants were significantly retarded in invasion and systemic infection of the same plant (27)(28)(29).Virulence factor production was controlled by several cell-to-cell communication systems, including the acyl-homoserine lactone-mediated quorum sensing (QS) system (26,(30)(31)(32), Vfm QS system (33,34), and putrescene signaling system (27).Significantly, D. oryzae could detect and respond to both host and itself produced putrescene as an intraspecies and interkingdom cell-to-cell communication signal to activate expression of various virulence genes (27).In contrast, as a newly identified bacterial pathogen, hardly any work has been done in characterization of the virulence and regulatory mechanisms in D. fangzhongdai.
Taro is an edible tropical crop belonging to the Araceae family, which is widely distributed throughout the world and has high nutritional, ornamental, medicinal, and economic values (35).However, its plantation has been severely affected by bacterial soft rot disease caused by various bacterial pathogens (36).We recently found that D. fangzhongdai is one of the major pathogens causing taro soft rot disease in Guangdong Province of China (13).Bioinformatics analysis showed that polyamine synthesis and transporter genes are highly conserved in D. fangzhongdai, suggesting that putrescine might also play a role in regulating the virulence and pathogenic mechanisms of this newly discovered Dickeya species.In this study, to explore the role of putrescine in the regulation of D. fangzhongdai virulence, we performed a systematic deletion analysis of the genes involved in polyamine biosynthesis and transportation by using the taro isolate ZXC1 as the parental strain.The results unveiled the key roles of putrescine biosynthesis in supporting bacterial growth, production of plant cell wall-degrading enzymes, and bacterial motility.In addition, we also demonstrated that D. fangzhongdai ZXC1 was able to influx exogenous putrescine through its specific transporters, PotF and PlaP, to compensate for the defected phenotypes due to deletion of the genes responsible for putrescine biosynthesis.The findings from this study add a new insight on the pathogenic mechanisms of D. fangzhongdai, which may facilitate developing new strategies to curb the rampage of this important bacterial pathogen.

Bioinformatic analysis of the genes involved in polyamine synthesis and transportation
We firstly conducted bioinformatic analysis of the putative genes associated with polyamine biosynthesis in D. fangzhongdai ZXC1.Five polyamine biosynthesis genes have been identified and characterized in the well-studied Escherichia coli, including speA, encoding an arginine decarboxylase [National Center for Biotechnology Infor mation (NCBI) accession number NP_417413.1];speB, encoding an agmatinase (NCBI accession number NP_417412.1);speC, encoding an ornithine decarboxylase (NCBI accession number NP_417440.4);speD, encoding a S-adenosylmethionine decarboxylase (NCBI accession number NP_414662.1);and speE, encoding a spermidine synthase (NCBI accession number NP_414663.1)(37).In addition, in Yersinia pestis, aguA, encoding an agmatine deiminase (NCBI accession number AJJ88030.1),and aguB, encoding an N-carbamoylputrescine amidase (NCBI accession number AJJ88068.1),are also known to be involved in polyamine biosynthesis (37).Among them, SpeA and SpeB are involved in conversion of arginine to produce putrescine (38); SpeC uses ornithine as a substrate to generate putrescine (38).In Y. pestis and Pseudomonas aeruginosa, where SpeB is missing, arginine is firstly converted to agmatine by SpeA, and then agmatine deiminase AguA catalyzes formation of N-carbamoylputrescine, which is further converted into putrescine by amidinohydrolase AguB.SpeE catalyzes formation of spermidine by addition of the aminopropyl moiety derived from decarboxylated S-adenosylmethionine (dSAM), which is synthesized by SpeD (39,40).A Blast search using the above seven genes, respectively, against the genome sequence of D. fangzhongdai ZXC1 (NCBI number CP119773.1)led to identification of all the genes encoding polyamine biosynthesis homologs except speB.These homologs at amino acid level are most similar to their counterparts in D. dadantii 3937, followed by D. oryzae EC1, two Yersinia pestis strains, P. aeruginosa PAO1, and E. coli K-12 MG 1655 (Table 1).It seems that D. fangzhongdai ZXC1 contains all the genes required for polyamine biosynthesis, and its biosynthetic pathway is reminiscent of that of Y. pestis (37) and P. aeruginosa PAO1 (41).
To date, four putrescine-specific transporters (PotE, PotFGHI, PlaP, and PuuP) and one spermidine/putrescine transporter (PotABCD) have been identified and characterized in E. coli, in which PotE, PotF, PlaP, PuuP, and PotD are corresponding substrate-bind ing proteins (42)(43)(44)(45).Among them, PotF and PlaP are involved in transportation of putrescine signal molecules in D. oryzae EC1 (27).While bioinformatics analysis did not reveal PotE and PuuP homologs, highly conserved homologs of PotF and PlaP from D. fangzhongdai ZXC1 were identified, which displayed high levels of similarities to the corresponding counterparts from D. dadantii 3937, D. oryzae EC1, Y. pestis strains, P. aeruginosa PAO1, and E. coli K-12 MG 1655 (Table 1).In addition, three copies of the potD gene encoding potential spermidine/putrescine transporter are present in the genome of D. fangzhongdai ZXC1, whose products are similar to the PotD proteins from D. dadantii 3937 and D. oryzae EC1, respectively, but displayed low levels of similarities to  S4).

Effect of null mutation of polyamine synthesis genes on extracellular enzyme production and bacterial motility
To study the role of polyamines in D. fangzhongdai ZXC1 physiology and pathogenicity, polyamine synthase genes speA, aguA, aguB, speC, speD, and speE were deleted in-frame and named as △A, △uA, △uB, △C, △D, and△E, respectively.Given that both speA and speC are involved in the synthesis of putrescine, their double-deletion mutant △AC was also generated for phenotype analysis.The growth assay showed that the growth rate of the single mutant △A, △uA, △uB, △C, △D, and △E in Luria-Bertani (LB) medium or minimal medium (MM) cultures was consistent with that of wild-type D. fangzhongdai ZXC1 (Fig. S1a through d).In addition, the growth rate of double-mutant △AC in the LB medium was consistent with that of wild-type D. fangzhongdai ZXC1, while the growth rate of the double-mutant △AC in the MM was delayed compared to that of wild-type strain ZXC1 (Fig. S1c and d).
Similarly, the single mutants △A, △uA, △uB, △C, △D, and △E cultured in correspond ing substrate media (Table S2) produced basically the same levels of plant cell wall degradation (PCWD) enzymes, including cellulase (Cel), pectinase (Pel), and protease (Prt), as their wild-type strain ZXC1 (Fig. 1a; Fig. S2a).In contrast, the double-mutant △AC showed significantly decreased activity of three PCWD enzymes, compared with wild-type D. fangzhongdai ZXC1.In trans expression of either speA or speC in the double-mutant △AC could fully rescue the mutant phenotypes (Fig. 1a).
We then analyzed the cell motility of D. fangzhongdai ZXC1 and its derivatives.The single-deletion mutants △C, △D, and △E did not show obvious difference in both swimming and swarming motility compared with strain ZXC1, and mutants △uA and △uB showed a moderately weakened swimming motility (Fig. S2b).Intriguingly, however, the mutant △A showed significantly decreased swimming motility and slightly decreased swarming motility (Fig. 1b) when compared with wild-type D. fangzhongdai ZXC1 and its complement △A(A).Deletion of speC in the background of △A further decreased the bacterial swimming and swarming motility (Fig. 1b).Taken together, the above data suggest the pathogen could use both SpeA and SpeC pathways to synthe size putrescine to modulate bacterial growth, PCWD enzyme production, and bacterial motility.

Impact of exogenous addition of polyamines and taro tissue extract on bacterial motility
To verify the regulatory role of polyamines in D. fangzhongdai ZXC1, we tested the effect of polyamines on restoration of swimming motility in double-mutant △AC.As shown in Fig. 2, putrescine showed the best effect, which could fully restore the swimming motility of double-mutant △AC to the wild-type level at a final concentration of 0.01 mM (Fig. 2a).However, further increasing putrescine level appeared detrimental, and the swimming motility of strain ZXC1 and double-mutant △AC was completely inhibited at a final concentration of 0.5-mM putrescine (Fig. 2a).Agmatine was at least over 100 times less effective than putrescine in restoration of the mutant phenotype, but no side effect was noted on both wild-type ZXC1 and the double-mutant △AC even at a concentration as high as 10 mM (Fig. 2b).Spermidine and spermine had no effect or merely a minor effect in activation of swimming motility and could block bacterial motility at 1.0 and 0.2 mM, respectively (Fig. 2c and d).The above results indicate that putrescine is the key cell-to-cell communication signal produced by D. fangzhongdai ZXC1 in regulation of its physiology and virulence.
We then tested whether taro tissue may contain sufficient signal to restore the defected phenotype of the double-mutant △AC.Since taro extract contains rich nutrients, strain ZXC1 and mutant in the swimming plate supplemented with taro extract grew faster than on the control plate without taro extract, and hence the bacterial motility was recorded 10 h instead of normally 16 h post inoculation (Fig. S3).For the convenience of comparison, the mutant motility was expressed as a relative value to that of wild type, which was arbitrarily set as 100 (Fig. 2e).With the exogenous addition of 20% taro tissue extract, the swimming motility of double-mutant △AC was recovered to about 98% of that of D. fangzhongdai ZXC1 (Fig. 2e).In conclusion, the above data suggest that putrescine is the key signal regulating bacterial swimming motility in D. fangzhongdai ZXC1, and that the pathogen appeared able to tape the host signal to activate its pathogenic mechanisms.

Disruption of speA and speC abolishes putrescine production in strain ZXC1
The above data suggest that putrescine is the cell-to-cell communication signal in modulation of bacterial physiology and production of PCWD enzymes.We therefore determine the cellular concentration of polyamines in D. fangzhongdai ZXC1 and its derivatives.The cell-free extracts of bacterial strains were prepared for derivation of polyamine molecules using benzoyl chloride (46,47).The concentration of benzoylated polyamines were determined using high-performance liquid chromatography coupled with mass spectrometry (HPLC-MS) with reference to standard benzoylated polyamines.The results showed that deletion of speA led to about 40%-50% decrease in putres cine level compared to wild-type D. fangzhongdai ZXC1 at different bacterial growth stages, which could be partially rescued by in trans expression of wild-type speA (Fig. 3).Intriguingly, null mutation of speC did not seem to significantly alter the putrescine level, whereas double deletion of both speA and speC almost fully abolished putrescine production (Fig. 3).Quantitative reverse transcription-PCR (qRT-PCR) analysis showed that expression of speA was increased in the mutant △C except at low cell density with OD 600 at 0.5 (Fig. S4a).In contrast, speC expression was increased from OD 600 at 0.5-1.5, but then decreased with OD 600 at 2.0 in the mutant △A (Fig. S4b).
Considering that putrescine is also a host signal for host-pathogen cell-to-cell communication (27), we then determine its level along with that of other polyamines in taro tuber tissues.The results showed that the concentrations of putrescine and spermidine in the taro tissue extracts were about 16.3 and 8.7 nmol/g, respectively, while spermine could not be detected.The findings indicate that putrescine is likely to be the major polyamine compound in taro tubers.

Effect of null mutation of polyamine transporter genes on PCWD enzyme production and bacterial motility
Bacteria not only can synthesize polyamines by themselves but also can also absorb exogenous polyamines through specific polyamine transporters.The bioinformatics analysis showed that D. fangzhongdai ZXC1 genome contains the genes encoding putrescine transporters PotF and PlaP, respectively (Table 1).To verify the roles of PotF and PlaP in putrescine-mediated cell-to-cell communication among D. fangzhongdai ZXC1 cells and between the bacterial cells and host organisms, single-and double-dele tion mutants of potF and plaP were generated in the genetic background of wild-type D. fangzhongdai ZXC1 and the speA mutant △A and named as △F, △P, △FP, △AF, △AP, and △AFP, respectively.In addition, a quadruple-deletion mutant △ACFP was also generated by deleting speC in the genetic background of the triple-deletion mutant △AFP.The growth curve showed that the growth rates of double-mutant △AC and triple-mutant △AFP in LB or MM were similar to that of wild-type D. fangzhongdai ZXC1, while the growth rate of quadruple mutant △ACFP in LB and MM was substantially retarded compared to that of strain ZXC1 (Fig. S1c and S1d).Intriguingly, when 0.1-mM putrescine was added into the LB or MM, the growth rate of mutant △ACFP was partially rescued (Fig. S1e and S1f ), suggesting that there could be an unidentified putrescine transporter in D. fangzhongdai ZXC1.
While production of PCWD enzymes in the △F, △P, △FP, △AF, △AP, and △AFP mutants in the MM was similar to that of wild-type ZXC1 and their parental mutant △A (Fig. S5a and b), PCWD enzyme activities in quadruple mutant △ACFP were significantly decreased compared with wild-type ZXC1 and its complemented strains △ACFP(F) and △ACFP(P) supplemented with 0.1-mM putrescine (Fig. 4a).In the motility assay, we found that the bacterial swimming motility of △F, △P, and △FP mutants were similar to that of wild-type ZXC1 (Fig. S5c).In contrast, the bacterial motility of quadruple mutant △ACFP was significantly reduced compared to the wild-type D. fangzhongdai ZXC1 and mutants △A and △AC (Fig. S6).While the bacterial swimming motility of mutant △AFP was significantly decreased compared with that of wild-type D. fangzhongdai ZXC1, exogenous addition of 0.1 mM putrescine to its complemented △AFP(F) and △AFP(P) could partially restore the defected motility (Fig. 4b).Similarly, we found that deletion of transporter genes potF and plaP in the genetic background of speA mutant △A abolished its response to taro extract in restoration of defected motility (Fig. 4c).In conclusion, the above data suggest that PotF and PlaP are essential for D. fangzhongdai ZXC1 to uptake exogenous putrescine.

Quantitative RT-PCR analysis of the virulence genes regulated by putrescine
To understand how putrescine could affect PCWD enzyme production and cell motility of D. fangzhongdai ZXC1, qRT-PCR analysis was performed to determine the expression levels of the genes associated with PCWD enzymes and bacterial motility.The genes encoding PCWD enzyme production and bacterial motility and associated regulators were identified by Blast searches using the corresponding homologs from the well-char acterized D. dadantii 3937 (48).This led to identification of 43 genes in D. fangzhongdai ZXC1, which share >83% identity at amino acid level with corresponding homologs in D. dadantii 3937 (Table S5).
We then compared the transcriptional expression levels of the genes encoding regulatory functions.As shown in Fig. 5a, pecS encodes a negative transcriptional regulator of PCWD enzyme production (56,57), and its transcript level in mutant △AC was two times higher than that in wild-type ZXC1 or the mutant △AC supplemented with 0.1-mM putrescine.Interestingly, although the growth rate of △pecS was significantly lower than that of wild-type D. fangzhongdai ZXC1 in MM, production of PCWD enzymes was much higher than wild-type D. fangzhongdai ZXC1 (Fig. S7a and S7b), suggesting that the negative regulator plays a key role in curbing the expression of PCWD enzyme genes under the in vitro conditions used in this study.It is worth noting that this study found for the first time that PecS also negatively regulated the expression of the genes encoding proteases in D. fangzhongdai ZXC1.The cheB encodes a chemotaxis response regulator protein-glutamate methylesterase and is known to play a role in regulation of methylation of bacterial chemotaxis genes (58).The growth rate of △cheB was consistent with that of wild-type D. fangzhongdai ZXC1 in MM, while its motility was decreased by 50% (Fig. S7a and S7c).Expression levels of the genes encoding transcriptional regulators of the flagellar operon (rpoD, flhC, flhD, fliA, and flgM) in mutant △AC were decreased by 70%-90% compared to those in mutant △AC supplemented with 0.1-mM putrescine (Fig. 5c), suggesting their roles in positive regulation of flagellar biogenesis.Taken together, the above findings suggest that putrescine modulates the expression of the genes involved in PCWD enzyme produc tion by regulating the expression of negative transcriptional regulator gene pecS and regulates bacterial motility by affecting the expression of genes related to bacterial chemotaxis, flagellar biosynthesis, and flagellar motors in D. fangzhongdai ZXC1.

The mutants defective in putrescine-mediated cell-to-cell communication were much attenuated in virulence against taro and potato
Bacterial virulence assay was performed on taro and potato using wild-type D. fangz hongdai ZXC1, putrescine-related mutants, and complemented strains.The bacterial cells were cultured in MM overnight with OD 600 adjusted to about 0.5, and 2 µL of which was added to the center of taro or potato tubers.The results showed that D. fangzhongdai ZXC1 could infect both dicotyledonous (potato) and monocotyledonous (taro) plants (Fig. 6).Compared with the wild-type strain, the macerated zones of △A in potato and taro were reduced by about 70% and 53%, respectively, which could be partially rescued by in trans expression of the wild-type speA in the mutant (Fig. 6a).To validate the role of putrescine transporters in planta, we then compared the virulence of △AFP and strain ZXC1.The results showed that the macerated zones of △AFP were reduced by about 80% and 58%, respectively, compared to wild-type ZXC1, and complementation with potF or plaP could restore the mutant virulence (Fig. 6b).The bacterial virulence of quadruple mutant △ACFP was further reduced compared with the single-deletion mutant △A, double-deletion mutant △AC, and triple-deletion mutant △AFP (Fig. 6c).These results demonstrate the critical role of putrescine-mediated cell-to-cell communi cation in coordination of D. fangzhongdai virulence.

DISCUSSION
D. fangzhongdai is a newly emerged and defined plant pathogen which was found capable of infecting numerous host plants (5,(13)(14)(15)(18)(19)(20)(21)(22), but its pathogenic and corresponding regulatory mechanisms are mostly unknown.In this study, we used the D. fangzhongdai strain ZXC1 isolated from taro, as a model organism to investigate the role of putrescine in D. fangzhongdai physiology and virulence.The findings from this study demonstrated that the pathogen could tape to the putrescine molecules produced by itself and those from host organisms as a signal to regulate production of PCWD enzymes and bacterial motility.The D. fangzhongdai mutants defective in putrescine signaling communication were drastically attenuated in its virulence against taro and potato.Our data also outlined the molecular mechanisms by which the putrescine signaling system modulates the bacterial motility and PCWD enzyme production.
In contrast to D. oryzae EC1, which produces putrescine mainly through the argi nine pathway (27), we found that D. fangzhongdai ZXC1 could use both arginine and ornithine pathways to synthesize putrescine signals.Bioinformatics analysis indicated that D. fangzhongdai ZXC1 encodes highly conserved arginine decarboxylase SpeA and ornithine decarboxylase SpeC, which share over 95% identity with D. oryzae at amino acid level (Table 1).HPLC-MS analysis results showed that the intracellular putrescine level of the speA deletion mutant △A was reduced by 50% when compared with D. fangzhongdai ZXC1 (Fig. 3), whereas double deletion of both speA and speC in D. fangzhongdai ZXC1 abolished putrescine biosynthesis (Fig. 3).Interestingly, transcrip tional expression of speA was upregulated in the speC deletion mutant, and vice versa, deletion of speA also led to increased expression of speC (Fig. S4).Highly agreeable with the key role of putrescine in keeping with bacterial growth (Fig. S1) and virulence (Fig. 6), these findings suggest that D. fangzhongdai might have a compensation mechanism to balance the expression of speA and speC so as to maintain the cellular level of putrescine, which is worthy of further investigations.
Spermidine, spermine, and putrescine constitute a group of ubiquitous aliphatic small polycationic molecules known as polyamines, which are widely distributed from bacteria to plants and animals.Bacterial pathogens could tape on to these polyamines through polyamine transporters or sensors (59,60).Among these polyamine molecules, putrescine was the most effective signal in rescuing the speA mutant phenotypes (Fig. 2).In this regard, it is interesting to note that putrescine was the most abundant polyamine molecule in taro extract.Further analysis found that two highly conserved putrescine transporters in D. fangzhongdai ZXC1, i.e., PotF and PlaP (Table 1), play a crucial role in influx of exogenous putrescine signals into the bacterial cells, including those from taro tissues (Fig. 4).Similarly, PlaP and PotF were also found in transporting putrescine into D. oryzae cells (27), suggesting that these putrescine transportation systems could be well conserved in the Dickeya genus.However, unlike D. oryzae EC1 that relies solely on PlaP and PotF in efflux of putrescine signals (27), our data suggest that D. fangzhongdai ZXC1 might contain additional polyamine transporters, which could uptake both spermidine and putrescine in the absence of PlaP and PotF (Fig. 4c).Our preliminary work seemed to preclude the PotABCD transporter, which is known for uptaking both spermidine and putrescine (61), as the swimming motility of the speA/plaP/potF/potD deletion mutant was still rescuable by exogenous addition of 0.1-mM putrescine (Fig. S8).This intriguing potential transporter needs to be further identified and characterized.
Putrescine signaling system was firstly found regulating D. oryzae virulence and systemic infection through modulating motility and biofilm formation (27).Similarly, deletion of speA and speC in D. fangzhongdai ZXC1 drastically attenuated its swimming and swarming motility (Fig. 1).However, we were not able to determine whether putrescine system could regulate biofilm formation in D. fangzhongdai ZXC1 as the pathogen did not form biofilms under the same experimental conditions used for D. oryzae analysis (27).Interestingly, in contrast to D. oryzae in which putrescine signal was not involved in the regulation of PCWD enzyme production (27), deletion of speC and speA significantly reduced the production of PCWD enzymes, and the defected phenotypes could be rescued via expression of the corresponding wild type genes in mutants or by exogenous addition of putrescine or taro tissue extract (Fig. 1 and 2).In addition, deletion of speA in D. oryzae did not alter the bacterial growth rate (27), whereas putrescine was needed for supporting D. fangzhongdai growth (Fig. S1).Our findings from this study thus add new functions to the list of putrescine regulatory spectrum in plant bacterial pathogens (Fig. 7).
The evidence of polyamines as signaling molecules in microorganisms has been accumulating in recent years, including spermidine from mammalian host that induces the expression of the genes encoding type III secretion system in a human bacterial pathogen Pseudomonas aeruginosa (60,62); putrescine that regulates bacterial motility in Proteus mirabilis (45,63), E. coli K-12 (44), D. oryzae EC1 (64), and D. zeae MS3 (65); and biofilm formation and disassembly in Yersiniapestis and Shewanella oneiden sis, respectively (66,67).However, the molecular mechanisms by which polyamine signals modulate various phenotypes remain largely unknown.Bearing this in mind, we specifically analyzed the genes encoding and regulating PCWD enzyme production and bacterial motility using qRT-PCR.The results suggest that putrescine might modulate the production of PCWD enzymes through downregulating the expression of PecS (Fig. 5a), which is a negative regulator known for curbing the expression of PCWD enzyme genes (56), and that putrescine could promote D. fangzhongdai motility by upregulating the genes encoding bacterial chemotaxis and flagellar biogenesis (28) (Fig. 5b and c), because bacterial motility is mainly guided and driven by chemotaxis and flagellar (68,69).
In summary, the findings from this study unveiled that D. fangzhongdai has evolved a four-component putrescine signaling system containing two signal synthases and two signal transporters to regulate bacterial virulence by inducing the expression of the genes encoding PCWD enzymes and bacterial motility (Fig. 7).In brief, the putrescine signal molecules produced by the SpeA and SpeC pathways in D. fangzhongdai and those from host plants activate PCWD enzyme production by downregulation of the negative regulator gene pecS and induce the transcriptional expression of chemotaxis and flagellar synthesis genes to modulate bacterial motility, and hence influence the bacterial virulence.In this signaling process, the bacterial transporter PotF and PlaP play a key role in uptaking extracellular putrescine signal into the bacterial cells for modulation of target gene expression (Fig. 7).This is highly aggregable with the critical roles of putrescine signaling system for the bacterial survival and competition, as such dual configurations would offer the pathogen sufficient flexibility and plasticity to accommodate complicated and adverse circumstances.We further present evidence of how the putrescine signaling system could modulate PCWD enzyme production and bacterial motility.These original findings add a new insight on the biological functions of polyamine signaling systems and present a solid basis and useful clues for further elucidating the putrescine signaling network and mechanisms of action in the regulation of D. fangzhongdai physiology and virulence.

Bacterial strains, growth conditions, and reagents
The bacterial strains and plasmids used and constructed in this study are listed in Table S1.The medium formula used in this study is shown in Table S2.Escherichia coli strains were maintained in LB medium at 37°C, and D. fangzhongdai ZXC1 and its derivatives were grown at 28°C in LB or MM and used for comparison of bacterial growth rate.Putrescine was purchased from Macklin; agmatine sulfate salt, spermidine, and spermine were purchased from Sigma-Aldrich.Antibiotics were added at the following final concentrations when required: kanamycin, 50 µg/mL; streptomycin, 50 µg /mL; and tetracycline, 10 µg/mL.

Generation of in-frame deletion mutants and complemented strains
In-frame deletion mutants of speA, aguA, aguB, speC, speD, and speE were generated through homologous recombination (25,26) and named as △A, △uA, △uB, △C, △D, and△E, respectively.In addition, double-deletion mutants △AC, △AF, and △AP were generated by deleting speC, potF, and plaP in the genetic background of △A, respec tively.A triple-deletion mutant △AFP was generated by deleting plaP in the genetic background of the double-deletion mutant △AF.A quadruple-deletion mutant △ACFP mutant was generated by deleting speC in the genetic background of the triple-deletion mutant △AFP.Detailed descriptions of the strains and mutants are provided in Table S1.All the primers used in this study are listed in Table S3.Firstly, fragments contain ing about 500 bp upstream (primers: gene-1 and gene-2) and downstream regions (primers: gene-3 and gene-4) of the corresponding target gene were amplified using D. fangzhongdai ZXC1 chromosomal DNA as template, respectively.Secondly, the upstream and downstream fragments of the target gene were fused together by using primer pair gene-1 and gene-4.The fusion fragment was then ligated to the suicide plasmid pKNG101 digested with restriction enzymes BamH I, and transformed into competent cells of E. coli CC118λ.The recombinant constructs were verified by DNA sequencing, and introduced into D. fangzhongdai ZXC1 by tri-parental mating to generate in-frame mutants as previously described (25).
To generate complemented strains, the encoding regions of speA, speC, potF, and plaP were amplified from D. fangzhongdai ZXC1 genomic DNA using the primers C-gene-F and C-gene-R by the primers listed in Table S3, respectively.The PCR products were ligated to expression vector PLAFR3 digested with restriction enzyme BamH I and transformed into competent cells of E. coli DH5α.The recombinant constructs were verified by DNA sequencing and introduced into corresponding mutants by tri-parental mating and confirmed by PCR analysis.

Bacterial extracellular enzyme activity assay
Cel, Pel, and Prt enzymes were assayed using carboxymethyl cellulose sodium, polygalac turonic acid, and non-fat-dried milk as substrates (Table S2), respectively, as described previously (70).Briefly, assay plates were prepared by pouring about 40 mL of substrate medium into a 13 × 13 cm square Petri dishes and allowing it to set at room temperature until completely solidified.The wells with 5 mm in diameter were punched in the assay plates with a metal puncher.The bacteria were inoculated in MM and cultured at 28°C and 200 rpm until population density reached about OD 600 = 1.0.Bacterial cells were removed by centrifugation with 12,000 rpm at 4°C for 5 min.Then, 20 µL of bacterial supernatants was taken and added into the wells of the assay plates, which were air-dried and invertedly incubated at 28°C.After incubation for about 17 h (Cel) and 32 h (Pel and Prt), respectively, the plates were stained with dye as follows: Cel assay plates were stained with 0.1% Congo red for 25 min and then decolored with NaCl (1 M) for 30 min twice; Pel assay plates were treated with HCl (1 M) for coloration; and the transparent zones surrounding the wells in Prt assay plates were directly recorded.The experiments were repeated three times in triplicates.

Bacterial motility assay
Bacterial motility assay was carried out as follows: 1.5-µL fresh cultures of D. fangzhong dai ZXC1 and its derivatives were spotted, respectively, on the center of a 90-mm semisolid plate containing 15-mL swimming medium (Table S2) and were incubated at 28°C.Swarming motility was assayed under the same conditions, except that bacterial cells were spotted onto a 90-mm semisolid swarming plate containing 15-mL swarming medium (Table S2) and incubated at 28°C.The diameters of swimming or swarming motility were measured 16 h post inoculation.To test the effect of polyamines and taro extract on restoration of the defected swimming motility of the polyamine biosynthesis and transportation mutants, taro extracts were prepared by processing the peeled taro with a juicer, filtered by eight layers of gauze, centrifuged by 12,000 rpm at 4°C for 1 h, and then filtered by a 0.45-µm filter to obtain taro extracts.The swimming motilities of the D. fangzhongdai ZXC1 and its mutants were quantified using the method described above by adding putrescine, spermidine, spermine, and taro extract to the swimming medium at a final concentration of 0.1, 0.5, and 0.1 mM and 20%, respectively.The above experiments were repeated three times in triplicate.

Polyamine derivatization and quantification
The cellular levels of polyamines in D. fangzhongdai ZXC1 and its derivatives were measured based on previous described methods with minor modifications (27,46,47,71,72).Briefly, bacterial cells were grown in MM to OD 600 = 0.5, 1.0, 1.5, and 2.0, respectively, at 28°C with shaking at 200 rpm.Aliquots of 10 mL of the bacterial cell suspension were centrifuged for 15 min at 4°C and 4,000 rpm; the bacterial pellets were collected and the wet weight of bacterial cells was recorded.The cell pellets were washed once with 10-mL ddH 2 O, resuspended in 1-mL lysis buffer (xTractor Buffer and xTractor Buffer Kit, purchased from Takara), and then incubated for 3 h at 4°C.The supernatants were collected by centrifugation and used for derivatization.To 400 µL of bacterial supernatants, 2 mL of 2-M NaOH solution and 14-µL of benzoyl chloride were added, shaken for 2 min, and incubated for 30 min at 37°C (27,47).The benzoylated mixtures were added to 4 mL of saturated NaCl solution and shaken for 20 s, which were then extracted by adding 4-mL petroleum ether.The petroleum ether phase was dried at room temperature.The samples were dissolved by adding 500-µL methanol filtrated with a 0.22-µm filter.The benzoylated polyamines were assayed with HPLC-MS using a 1.8-µm Eclipse Plus C18 (Agilent) column fitted with a 100 × 2.1-mm guard column at a flow rate of 0.2 mL/min.Mass spectroscopy (Agilent 6540B Q-TOF) was used to verify the identity of each peak observed in the HPLC fractions.A standard curve was generated using various concentrations of benzoylated putrescines.The polyamine levels in the taro extract were measured using the same methods.The experiments were repeated three times in triplicate.

qRT-PCR assay
Bacterial strains were cultured in fresh MM (with or without 0.1-mM putrescine) at 28°C to OD 600 = 1.0.RNA samples were prepared using an SV total RNA isolation system kit (Promega).RNA quantity was measured using a NanoDrop (Wilmington, DE, USA) ND-100 spectrophotometer, and RNA integrity was determined by using agarose gel electrophoresis.Total RNA samples were treated with DNase I to remove DNA contam inations and then reverse transcribed into double-strand cDNA using HiScript III RT SuperMix for qPCR (Vazyme).Quantitative PCR (qPCR) was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme) with the qPCR primers listed in Table S3.The experiments were repeated three times in triplicate.

Bacterial pathogenicity assay
Pathogenicity assays of D. fangzhongdai and derivatives were performed using potato and taro tubers.Briefly, potato and taro tubers were washed and air dried, and slices about 5 mm in thickness were prepared.A small hole was made using a sterilized needle in the center of potato and taro tuber slices, to which 2 µL of bacterial suspension with OD 600 at 0.5 about 4.35 × 10 7 CFU/mL was added using a sterilized pipette tip.The inoculated plant tuber slices were incubated at 28°C and 80% relative humidity, and soft rot symptoms were recorded at 24 or 48 h post inoculation as indicated.The experiments were repeated three times in triplicate.The sizes of maceration zones in plant slices were measured using ImageJ.

FIG 6
FIG 6 Virulence of D. fangzhongdai ZXC1 and derivatives on taro and potato.(a) Virulence assay with mutant △A and strain ZXC1 on potato and taro tubers; (b) virulence assay with mutant △AFP and strain ZXC1 on potato and taro tubers; (c) virulence assay with single-, double-, triple-, and quadruple-deletion mutant strain ZXC1 on potato and taro tubers.The data shown are the mean ± standard error (n = 3).Statistical significance: **P ＜ 0.01, ***P ＜ 0.001, ****P ＜ 0.0001 (by two-way analysis of variance with multiple comparisons).

FIG 7
FIG 7 Putrescine signal regulates the PCWD enzyme production and bacterial motility of D. fangzhongdai.

TABLE 1
Analysis of the genes encoding polyamine biosynthesis and transportation in D.