948 / Molecular Plant-Microbe Interactions
MPMI Vol. 24, No. 8, 2011, pp. 948–957. doi:10.1094/MPMI-02-11-0031. © 2011 The American Phytopathological Society
Xanthomonas campestris Diffusible Factor
Is 3-Hydroxybenzoic Acid and Is Associated
with Xanthomonadin Biosynthesis, Cell Viability,
Antioxidant Activity, and Systemic Invasion
Ya-Wen He,1,2 Ji’en Wu,1 Lian Zhou,2 Fan Yang,1 Yong-Qiang He,3 Bo-Le Jiang,3 Linquan Bai,2 Yuquan Xu,2
Zixin Deng,2 Ji-Liang Tang,3 and Lian-Hui Zhang1
1Institute of Molecular and Cell Biology, 61 Biopolis Drive, Singapore 138673; 2State Key Laboratory of Microbial Metabolism
and National Center for Molecular Characterization of GMOs, School of Life Science and Biotechnology, Shanghai Jiao Tong
University, Shanghai 200240, China; 3State Key Laboratory for Conservation and Utilization of Subtropical Agro-bioresources,
Guangxi University, Nanning 530004, China
Submitted 6 February 2011. Accepted 8 April 2011.
Xanthomonas campestris pv. campestris produces a mem-
brane-bound yellow pigment called xanthomonadin. A dif-
fusible factor (DF) has been reported to regulate xantho-
monadin biosynthesis. In this study, DF was purified from
bacterial culture supernatants using a combination of sol-
vent extraction, flash chromatography, and high-perform-
ance liquid chromatography. Mass spectrometry and nu-
clear magnetic resonance analyses resolved the DF chemical
structure as 3-hydroxybenzoic acid (3-HBA), which was
further confirmed by synthetic 3-HBA. Significantly, bioas-
say and in silico analysis suggest that DF production is
widely conserved in a range of bacterial species. Analysis of
DF derivatives established the hydroxyl group and its posi-
tion as the key structural features for the role of DF in xan-
thomonadin biosynthesis. In addition, we showed that DF
is also associated with bacterial survival, H2O2 resistance,
and systemic invasion. Furthermore, evidence was also pre-
sented that DF and diffusible signaling factor have overlap-
ping functions in modulation of bacterial survival, H2O2
resistance, and virulence. Utilization of different mecha-
nisms to modulate similar virulence traits may provide X.
campestris pv. campestris with plasticity in response to
various environmental cues.
The genus Xanthomonas is one of the most ubiquitous
groups of plant-associated bacterial pathogens. Members of
this genus have been shown to infect at least 124 monocotyle-
donous and 268 dicotyledonous plant species, many of which
are economically important crops or plants (Leyns et al. 1984).
A characteristic feature of the genus Xanthomonas is the pro-
duction of yellow, membrane-bound pigments called xantho-
monadin (Starr 1981). The presence of the pigments has be-
come a useful chemotaxonomic and diagnostic marker for
Xanthomonas spp. (Schaad and Stall 1988; Starr and Stephens
1964). The xanthomonadins were initially thought to be caro-
tenoids. However, further characterization has demonstrated
that they represent a unique group of halogenated, aryl-
polyene, water-insoluble pigments that are associated exclu-
sively with the outer membrane of the bacterial cell wall
(Andrewes et al. 1976; Stephens et al. 1963). Moreover, xan-
thomonadins from different Xanthomonas spp. differ in bromi-
nation and methylation patterns, and this is useful for identify-
ing members of the same genus (Starr et al. 1977).
Although xanthomonadins are not essential for the bacterial in
planta growth (Durgapal 1996; Poplawsky et al. 1993; Tsuchiya
et al. 1982), several lines of evidence show that xanthomo-
nadins protect the pathogen from photooxidative damage. Pig-
ment-deficient mutants from several Xanthomonas spp., includ-
ing Xanthomonas juglandis, X. oryzae pv. oryzae, and X. cam-
pestris pv. campestris, were found to be more susceptible to
photooxidative damage than their corresponding wild-type
parental strains (Jenkins and Starr 1982; Poplawsky et al.
2000; Rajagopal et al. 1997). In addition, xanthomonadins
were found to protect egg-phophatidylcholine lipsomes against
peroxidation damage (Rajagopal et al. 1997). Furthermore,
disruption of xanthomonadin production in X. campestris pv.
campestris B-24 by transposon mutagenesis caused an approxi-
mately 100-fold decrease in bacterial epiphytic survival rate
following exposure to high-intensity light (Poplawsky et al.
2000). Taken together, these findings have established the role
of xanthomonadins in maintaining the ecological fitness of
Xanthomonas spp. by protecting bacterial cells against photo-
Early genetic analysis reported that xanthomonadin biosyn-
thesis is encoded by a 25.4-kb genomic region that contains
seven transcriptional units (pigA to pigG) in X. campestris pv.
campestris (Poplawsky and Chun 1997; Poplawsky et al.
1993). A similar pig cluster was also reported in X. oryzae pv.
oryzae (Goel et al. 2002). Subsequent genome sequence analy-
sis of X. campestris pv. campestris strains ATCC33913, 8004,
and B100 showed that the pig cluster consists of 22 open read-
ing frames (ORF), which may constitute part of a novel type II
polyketide synthase pathway involved in biosynthesis of xan-
thomonadins (da Silva et al. 2002; Qian et al. 2005; Vorhölter
et al. 2008).
Evidence has emerged that a diffusible factor (DF) is impli-
cated in the regulation of xanthomonadin and exopolysaccha-
Y.-W. He and J. Wu contributed equally to this work.
Corresponding authors: Y.-W. He; E-mail: firstname.lastname@example.org; L.-H.
Zhang; E-mail: email@example.com
*The e-Xtra logo stands for “electronic extra” and indicates that seven
supplementary figures and two supplementary tables are published online.
Also, Figure 6 appears in color online.
Vol. 24, No. 8, 2011 / 949
ride (EPS) production in X. campestris pv. campestris. Null
mutation of pigB in X. campestris pv. campestris B24 abol-
ished the production of xanthomonadins and DF, and resulted
in impaired production of EPS (Poplawsky and Chun 1997).
Subsequent study with the pigB mutants detected a significant
reduction in epiphytic survival rate, and a decrease in infection
of natural host via hydathodes (Poplawsky and Chun 1998).
Particularly noteworthy is that xanthomonadin and EPS pro-
duction in the pigB mutants were restored in the presence of
DF-producing strains, or by addition of partially purified DF,
suggesting that DF might act as a signal for activation of xan-
thomonadin and EPS production.
DF was predicted to be a butyrolactone based on mass spec-
trometry (MS) analysis (Chun et al. 1997). However, a recent
study showed that synthetic butyrolactone had no effect on
xanthomonadin or EPS production, indicating that natural DF
produced by X. campestris is not butyrolactone (Yajima et al.
2010). In addition, our preliminary analysis with a range of
butyrolactones from Streptomyces spp. also did not identify
any reported DF-like activity. To determine the biological
functions of DF, we set out to purify and characterize the
chemical structure of DF. Based on the findings that the gene
xanB2, which encodes a putative pteridine-dependent deoxy-
genase, is needed for the production of DF signal (Poplawsky
et al. 2005), we developed a reporter strain for detection and
semiquantitative analysis of DF activity, which facilitated the
purification and characterization of DF. In addition, functional
analysis showed that DF plays a key role in regulating xantho-
monadin biosynthesis, bacterial survival, H2O2 resistance, and
bacterial systemic invasion. Furthermore, we showed that DF
and diffusible signaling factor (DSF) signals have a cumulative
effect on modulation of X. campestris pv. campestris survival,
oxidative stress resistance, and systemic invasion.
X. campestris pv. campestris strains 8004 and XC1
produce the diffusible signal DF
to regulate xanthomonadin production.
Previously, Xcc4014 (xanB2) was shown to be responsible
for the biosynthesis of DF in X. campestris pv. campestris B24
(Poplawsky et al. 2005). In this study, the xanB2 deletion mu-
tant (designated as xanB2) of X. campestris pv. campestris
wild-type strain XC1 was generated. The mutant xanB2 did
not produce xanthomonadins and this mutant phenotype was
restored by growing xanB2 in the vicinity of a wild-type
strain, or by adding the supernatant extracts from the overnight
culture of wild-type strain XC1 (Fig. 1). Similarly, deletion of
xanB2 from X. campestris pv. campestris wild-type strain
8004 also resulted in a phenotype almost identical to that of
the mutant xanB2 (data not shown). These findings confirm
that DF signaling is conserved in different X. campestris pv.
campestris strains and isolates, and that DF has an important
role in xanthomonadin biosynthesis.
Purification and structural characterization of DF.
DF was extracted from bacterial culture supernatants using
ethyl acetate and the crude extracts were separated using flash-
column chromatography and high-performance liquid chroma-
tography (HPLC). A single peak associated with strong DF
activity was detected in HPLC elute (Fig. 2A). Approximately
1 mg of pure DF was obtained from 30 liters of culture super-
natants, and electrospray ionization (EMI)-MS showed the
(molecular weight – 1) of DF to be 137.07 (Supplementary
Fig. S1A). Subsequent nuclear magnetic resonance (NMR)
analysis suggested the structure of DF to be 3-hydroxybenzoic
acid (3-HBA) (Fig. 2C). Further comparison showed that the
1H and 13C NMR spectra of purified DF were virtually indis-
tinguishable from those of commercially available 3-HBA.
Consistent with the above findings, addition of commercially
available 3-HBA and the purified DF to cultures of xanB2
induced similar levels of xanthomonadin biosynthesis (Fig. 1).
DF production during bacterial growth and induction
of xanthomonadin biosynthesis.
Production of DF was not detected until 12 h postinocula-
tion (Fig. 3A). A sharp DF level increase was observed from
16 to 20 h of bacterial growth, with the highest level (5.75 µM)
being detected 28 h after inoculation (Fig. 3A). Consistent
with the trend of DF production, exogenous addition of DF to
the culture of DF-minus mutant xanB2 at a final concentra-
tion of 6.2 µM resulted in approximately 82% of wild-type
pigment production (Fig. 3B). Exogenous addition of DF at
12.5 µM fully restored the pigment production in xanB2 (Fig.
3B). Further results showed that early xanthomonadin produc-
tion could be induced at an optical density at 600 nm (OD600) =
1.0 or 1.5 by exogenous addition of DF at a final concentration
of 12.5 µM (Fig. 3C). A similar trend was observed in the pres-
ence of higher concentrations of DF (data not shown).
Induction of xanthomonadin by DF derivatives.
To identify the structural features of DF that are important
for its biological activity, a variety of commercially available
DF analogues were assayed for their abilities to induce xantho-
Fig. 1. Diffusible factor (DF) activity in modulation of xanthomonadin
biosynthesis. A, Representative diffusion plate assay showing the restora-
tion of xanthomonadin production in DF-deficient mutant xanB2 follow-
ing exposure to Xanthomonas campestris pv. campestris wild-type strain
XC1, purified DF, and synthetic DF (25 µM) as indicated. B, Quantitative
analysis of xanthomonadin production. Purified DF (pDF) and synthetic
DF (sDF) were added at a final concentration of 10 µM. Data are the
means of two independent repeats, and error bar indicates standard devia-
950 / Molecular Plant-Microbe Interactions
monadin production. Based on the observation that DF com-
pletely restores pigment expression at a concentration of
approximately 12.5 µM (Fig. 3B), DF and its analogues were
tested at a final concentration of 10 µM or higher when neces-
sary. Among them, 3-methoxybenzoic acid and 4-amino-3-
HBA showed approximately 80 and 5%, respectively, DF-like
activity based on assay of pigment restoration (Table 1). In
contrast, benzoic acid and salicylic acid (2-hydroxybenzoic
acid) had no effect on pigment restoration, even at a concentra-
tion of 500 µM (Table 1). These data suggest that the hydroxyl
group at position 3 of the phenol ring is important for the
activity of DF, and amino group substitution at position 4 of
the phenol ring compromises its activity.
DF production is widely conserved.
To determine whether other bacterial pathogens also produce
DF, 19 bacterial strains belonging to 15 bacterial species were
assayed. DF-like activity was detected in 10 strains of the fol-
lowing eight bacterial species: X. campestris pv. campestris, X.
axonopodis pv. citri, X. oryzae pv. oryzae, X. albilineans, Xy-
lella fastidiosa, Xylophilus ampelinus, Burkholderia cenoce-
pacia J2315, and B. xenovorans LB400 (Supplementary Table
S1). Interestingly, majority of these strains produce yellow or
pink pigments. Furthermore, a Blast search revealed that a
range of bacterial species, including Dechloromonas aromatica
RCB, Nitrosococcus oceani ATCC 19707, Methylococcus cap-
sulatus Bath, Thiobacillus denitrificans ATCC 25259, B. ceno-
cepacia PC184, Roseobacter sp. strain MED193, Streptomyces
hygroscopicus, Granulibacter bethesdensis CGDNIH1, and
Verrucomicrobium spinosum DSM 4136 contain the proteins
with >30% identical amino acids and similar size to XanB2
(Supplementary Fig. S2). The roles of the XanB2 homologues
in these bacterial species remains to be investigated.
EPS production by DF-deficient mutant is marginally
decreased in the late stationary phase of bacterial growth.
Previous studies of Xanthomonas campestris pv. campestris
B24 showed that the DF signal also affects the production of
EPS. The DF-deficient mutant derived from strain B24 pro-
duced fourfold less EPS than its parent strain on solidified nu-
trient starch agar (NSA) medium (Poplawsky and Chun 1997).
In this study, however, we found that EPS production by DF-
deficient strain xanB2 was not significantly different from
that by X. campestris pv. campestris wild-type strains XC1 or
8004 on the same NSA medium (Supplementary Fig. S3A).
EPS production by X. campestris pv. campestris wild-type
strains XC1 and 8004 were further compared with their corre-
sponding DF-deficient mutants using the sucrose-containing
rich YEB medium as previously described (He et al. 2006,
2009). The results showed that both wild-type strains and mu-
tant strains produced similar levels of EPS in the early station-
ary phase of growth (OD600 = 2.0, 2.1, and 2.3). However, at
the late stationary phase (OD600 = 2.5), the EPS levels of DF-
deficient mutants of XC1 and 8004 were approximately 11.5
and 18.5%, respectively, lower than their corresponding parent
strains. Consistent with the above findings, reverse-transcrip-
tion polymerase chain reaction (PCR) analysis did not reveal
any significant differences in the transcriptional levels of
gumB and gumK, two genes of the gum operon encoding the
key enzymes required for the final step of EPS biosynthesis
and secretion (Vorhölter et al. 2008) in X. campestris pv.
campestris wild-type and its DF-deficient mutant.
DF production is associated
with bacterial survival and H2O2 resistance.
To further explore the biological roles of DF, cell viability of
wild-type X. campestris pv. campestris and the mutant xanB2
was compared by counting CFU. No significant difference in
CFU was observed between wild-type and DF-deficient mu-
tants during log phase and early stationary phase of bacterial
growth (OD600 = 1.5, 2.0, 2.1, and 2.3) (Supplementary Fig.
S5; Fig. 4A). However, the CFU value of the DF-deficient mu-
tant at the late stationary phase (OD600 = 2.5) was significantly
lower than that of wild-type XC1 (Fig. 4A). The finding was
validated by exogenous addition of DF to cultures of xanB2,
and the results showed that its CFU value was restored to the
wild-type level (Fig. 4A).
Considering the role of DF in production of the pigment
xanthomonadin, which is known for their activity in protecting
bacteria against photooxidative stress, we used a pigment-defi-
cient mutant as a control to determine whether DF contributed
to bacterial viability directly or indirectly through xanthomo-
nadin. Xcc4015, which encodes a putative AMP ligase, is one
of the pig genes required for xanthomonadin biosynthesis in X.
campestris pv. campestris (da Silva et al. 2002). The Xcc4015
deletion mutant (designated as xcc4015) was xanthomonadin-
deficient (Supplementary Fig. S4). Importantly, this mutant
produced a similar level of DF as X. campestris pv. campestris
wild-type in liquid culture. It was noticed that the CFU values
of strain xcc4015 at different growth stages were similar to
those of DF-deficient strain xanB2, suggesting that the low
cell viability phenotype of the DF-minus mutant is probably
due to lack of the yellow pigment xanthomonadin.
DF was also found to affect the in vitro survival of bacteria.
On sucrose-containing YEB plates (He et al. 2010), the aver-
Fig. 2. Diffusible factor (DF) purification and characterization. A, High-
performance liquid chromatography purification of DF following flash-
column chromatography. The single peak at 4.2 min showed a strong DF
activity in bioassay. B, 1H nuclear magnetic resonance (NMR) spectrum of
purified DF. C, 13C NMR spectrum of purified DF and the chemical struc-
ture of DF (3-hydroxybenzoic acid).
Vol. 24, No. 8, 2011 / 951
age survival time of wild-type strain XC1 was 24 days at room
temperature in darkness whereas the DF-deficient mutant
xanB2 only survived for 16 days (Fig. 4B). When the culture
medium was supplemented with 10 µM DF, the survival time
of xanB2 was substantially increased (Fig. 4B). The survival
time of the mutant xcc4015 was similar to that of DF-defi-
cient strain xanB2.
It is known that, along with bacterial growth, active oxygen
species accumulate which might cause damage to bacterial
cells. To test whether DF production is associated with oxida-
tive stress resistance, the survival rate of X. campestris pv.
campestris wild-type strain XC1 and its mutant xanB2 after
H2O2 treatment were compared. Bacterial strains were grown
in YEB medium until they reached the early stationary phase
Fig. 3. Time-course analysis of diffusible factor (DF) production and xanthomonadin induction. A, Time course of DF production by strain XC1. B, Dosage
effect of DF on xanthomonadin production by DF-deficient mutant xanB2. C, Time course of xanthomonadin production by DF-deficient strain xanB2 in
the presence of 12.5 µM DF. Data presented are the means of two independent repeats and error bars represent standard deviations.
952 / Molecular Plant-Microbe Interactions
(OD600 = 2.1). The bacterial cells were precipitated and the
pellet was re-suspended in fresh YEB medium. H2O2 was then
added to the cultures at final concentrations of 176, 528, and
880 µM. After 30 min of H2O2 treatment, the CFU values of
each strain were compared. The results showed that abolish-
ment of DF production substantially decreased bacterial resis-
tance to oxidative stress, and addition of DF to cultures of
xanB2 restored H2O2 resistance close to the wild-type level
(Fig. 5). As a control, strain xcc4015 was more sensitive to
H2O2 treatment than the X. campestris pv. campestris wild-type
strain but less than the DF-deficient strain xanB2 (Fig. 5).
DF-deficient mutant is attenuated in systemic invasion.
The virulence of wild-type, DF-deficient, and complemented
strains was evaluated by measuring lesion lengths 14 days
postinoculation on leaves of Chinese radish. Deletion of xanB2
resulted in a substantial reduction in virulence, which was
restored to wild-type levels by in trans expression of xanB2
(Fig. 6A and B).
X. campestris pv. campestris is a vascular pathogen, and its
systemic infection of plant tissues correlates with its ability to
move and proliferate along the xylems of leaves and stems of
infected plants. Given that null mutation of DF production
does not affect the production of extracellular enzymes (un-
published data), we speculated that the mutation might affect
the bacterial ability in systemic infection. To this end, bacterial
progression along the xylem of Chinese cabbage (Wongbok)
was analyzed. One week after inoculation, the average distance
of xanB2 migration along the infected leaves was approxi-
mately 7.5 cm, compared with 10.5 cm of the wild-type strain
(Fig. 6C). As expected, in trans expression of xanB2 in the DF-
deficient mutant resumed its systemic infection ability to the
level of wild-type strain (Fig. 6C). The migration ability of
strain xcc4015 was similar to that of xanB2 (Fig. 6C).
DF regulates xanthomonadin production independent of
the DSF signaling system.
DSF is an important quorum-sensing signal involved in the
regulation of diverse biological functions such as EPS and
exoenzyme production, multidrug resistance, detoxification,
aerobic respiration, and flagellar biosynthesis in X. campestris
pv. campestris (He and Zhang 2008). To determine whether
cross-talk occurs between DF and DSF signaling systems, DF
biosynthesis and DF-dependent xanthomonadin production
were analyzed in the DSF-deficient strain rpfF and in the
DSF-overproduction strain rpfC (He et al. 2006). The results
Fig. 4. Effect of diffusible factor (DF) on bacterial viability. A, Viability of
Xanthomonas campestris pv. campestris XC1, mutant xanB2, and
xcc4015 in the absence or presence of DF at different growth stages in
YEB medium. B, In vitro survival on YEB plates. Data are the means of
representative duplicates from three independent experiments.
Fig. 5. Diffusible factor (DF) is required for Xanthomonas campestris pv.
campestris resistance to oxidative stress. CFU of X. campestris pv. cam-
pestris wild-type (solid circle), xanB2 (open circle), xanB2 + DF (open
square), and xcc4015 (solid diamond) after H2O2 treatment. Data are the
means of two repeats and the experiment was independently repeated three
Table 1. Biological activity of diffusible factor (DF) and its derivatives
Name Structure DF-like activity
Vol. 24, No. 8, 2011 / 953
showed that deletion of rpfF or rpfC did not affect the produc-
tion of DF or the DF-dependent pigmentation induction (Sup-
plementary Fig. S6A and B). Furthermore, an rpfF and xanB2
double-deletion mutant (rpfFxanB2) was generated. rpfF
xanB2 strains were deficient in production of DF and DSF
signals and did not produce pigment. When rpfFxanB2
strains were treated with DF, the pigmentation phenotype was
restored in the absence of the DSF signal, suggesting that in-
duction of pigment is DSF independent. Similarly, disruption
of DF production did not affect DSF biosynthesis r DSF-
dependent EPS induction (Supplementary Fig. S7).
DF and DSF signaling system have cumulative effects
on bacterial viability, H2O2 resistance, and systemic invasion.
Using DF- and DSF-deficient mutants and the double-dele-
tion mutant rpfFxanB2, the contribution of DF and DSF
signals to bacterial viability, H2O2 resistance, and virulence
were analyzed. In bacterial viability assays, the wild-type
strain XC1 and its derivatives were grown in YEB medium
until late stationary phase before counting CFU. The results
showed that disruption of either DF or DSF production re-
sulted in a decrease in bacterial cell viability (Fig. 7A). The
cell viability was further decreased in the double-deletion
mutant rpfFxanB2, which was defective in DF and DSF
biosynthesis (Fig. 7A). The mutant rpfFxanB2 also exhib-
ited the highest sensitivity to H2O2 compared with the single-
deletion mutants rpfF or xanB2 (Fig. 7B). Consistent with
the above findings, rpfFxanB2 was less virulent than the
rpfF or xanB2 when assayed using Chinese Radish (Fig.
7C). The data support the notion that the DF and DSF signal-
ing system exert cumulative effects in modulating bacterial
viability, H2O2 resistance, and systemic invasion.
A previous study showed that X. campestris pv. campestris
B24 produces a DF to regulate xanthomonadin production
Fig. 6. Disruption of diffusible factor production impaired Xanthomonas
campestris pv. campestris virulence and systemic invasion ability. Chinese
radish and cabbage were used in A and B, virulence and C, invasion
assays, respectively. Values shown are the mean standard deviation from
Fig. 7. Diffusible factor and diffusible signaling factor signaling system have
cumulative effect on A, bacterial cell viability; B, H2O2 resistance; and C,
virulence. Values shown are the means standard deviation from two
954 / Molecular Plant-Microbe Interactions
(Poplawsky and Chun 1997). Based on MS analysis, it was
speculated that DF is a butyrolactone (Chun et al. 1997). In
this study, we showed that X. campestris pv. campestris strains
8004 and XC1 also produced DF (Fig. 1). The MS and NMR
analyses indicated that DF is 3-HBA, which was further con-
firmed by using synthetic compound and evaluation of biologi-
cal activity (Fig. 1). Identification of DF chemical structure
would facilitate further investigation of its molecular mecha-
nisms in modulation of various biological functions.
DF-like activity has been detected in more than 40 gram-
negative bacterial species within the genera Xanthomonas, Xy-
lella, Xylophilus, Streptomyces, and Burkholderia (Poplawsky
and Chun 1997; Poplawsky et al. 2005). This list could be fur-
ther expanded because an in silico analysis identified xanB2
homologues in a range of other bacterial species, including D.
aromatica, N. oceani, M. capsulatus, and T. denitrificans.
These data suggest that production of DF is widely conserved
in the bacterial kingdom.
The results from this study showed that DF is associated
with diverse biological functions. The DF-deficient mutant
xanB2 was nonpigmented, impaired in survival ability, and
became less virulent than wild-type strain XC1 when assayed
against Chinese radish (Fig. 7A). These data are in good agree-
ment with previous findings that DF-deficient pigB mutants
derived from X. campestris pv. campestris B24 cause fewer
lesions on cabbage, and are impaired in epiphytic survival
(Poplawsky and Chun 1997, 1998; Poplawsky et al. 1998). In
addition, the present study also identified several novel DF-
dependent biological functions. For example, the DF-deficient
mutant was less viable in the late stationary phase (Fig. 4),
more sensitive to exogenous H2O2 (Fig. 5), and attenuated in
systemic infection along the xylem tissues of Chinese cabbage
(Fig. 6B). It remains unknown how DF is involved in the regu-
lation of these biological functions. Given that there is an aryl
group in the structure of xanthomonadin, we speculate that DF,
instead of acting as a signal, might be a biosynthetic interme-
diate for xanthomonadin biosynthesis. Xanthomonadin might
account for most, if not all, the observed phenotype changes in
the DF-minus mutant, including cell viability, bacterial sur-
vival, H2O2 resistance, and systemic invasion. This hypothesis
was supported by several lines of evidences. First, mutation of
the gene encoding DF production abolished pigment produc-
tion (Fig. 1), which is more typical of inactivation of a gene in
a biosynthetic pathway than of inactivation of a regulatory
gene. Second, although DF production was experimentally
confirmed to be associated with EPS production, cell viability
and survival, H2O2 resistance, and systemic invasion, microar-
ray analysis revealed that transcription of the genes encoding
these functions were not affected in the DF-deficient mutant
(Supplementary Table S2); Third, the Xcc4015 mutant, which
was pigment deficient but produced a level of DF similar to
that of the wild type, showed patterns in cell viability, H2O2
resistance, and systemic invasion similar to the DF-deficient
mutant (Figs. 3 to 5). Further investigations are needed to elu-
cidate the role of DF in xanthomonadin biosynthesis.
A previous study showed that the DF-deficient pigB mutants
derived from X. campestris pv. campestris B24 produced four-
fold less EPS than their parental wild-type strain on NSA
plates (Poplawsky and Chun 1997). However, we did not ob-
serve any significant difference in EPS production in X.
campestris pv. campestris wild-type and DF-deficient strains
on NSA plates. By using sucrose-containing YEB medium, we
demonstrated that disruption of DF biosynthesis in X. campes-
tris pv. campestris strains 8004 and XC1 only marginally af-
fected EPS production at the late stationary phase. Further
analysis showed that mutation of the gene for DF production
did not affect the transcriptional level of gum operon. Al-
though the mechanism behind this difference deserves further
investigation, this type of strain-specific variation may not be a
total surprise. For example, our previous study showed that
null mutation of DSF production in X. campestris pv. campes-
tris 8004 caused extensive cell aggregation, whereas the same
mutation in strain XC1 only led to a moderate formation of
cell aggregates (He et al. 2006).
Oxidative stress is a common environmental cue that bacterial
pathogens encounter in various niches, particularly during the
systemic invasion of host organisms. Increasing production of
reactive oxygen species, including superoxide, H2O2, and OH·,
is associated with aerobic respiration and with active plant de-
fense responses (Bestwick et al. 1997; Bindschedler et al. 2006;
Sutherland 1991). X. campestris pv. campestris is a vascular
pathogen and is normally restricted to the xylem tissues of
infected plants. Therefore, the ability of X. campestris pv.
campestris to survive oxidative stress is of critical importance
for successful colonization in host plants. The DSF-based quo-
rum-sensing system is known to play a role in response to oxi-
dative stress (He et al. 2006, 2007). The results of this study
showed that DF also has a role in modulating oxidative stress
(Fig. 5). Consistent with the above findings, the double-deletion
mutant rpfFxanB2, which is deficient for DF and DSF sig-
nals, was substantially impaired in H2O2 resistance, systemic
invasion, and virulence in Chinese radish (Fig. 6). The previous
findings showed that the DSF signal activates the genes encod-
ing multidrug resistance, acriflavin resistance, superoxide dis-
mutase, and catalase (He et al. 2006). In contrast, transcriptional
profiling analysis showed that the DF had no effect on expres-
sion of the genes encoding the above functions. These results
suggest that DF and DSF signals might use different mecha-
nisms to modulate resistance to oxidative stress.
A previous study on X. campestris pv. campestris B24 indi-
cates that DF and DSF signaling system are two separate sys-
tems with overlapping roles in regulation of EPS production
(Poplawsky et al. 1998). This study has provided further evi-
dence for the roles of DF and the DSF signaling system in co-
modulation of bacterial viability, survival, and H2O2 resistance
(Fig. 7). Bacterial pathogens commonly evolve and adopt an
array of mechanisms to modulate virulence. Utilization of multi-
ple signaling systems by X. campestris pv. campestris may
confer considerable flexibility in expressing specific sets of
genes required for adaptation and survival in response to
changing environmental conditions. Identification of the
chemical structure of DF would facilitate future studies on this
intriguing and widely conserved molecule; much remains to be
done to understand the roles and mechanisms of DF in X.
campestris pv. campestris physiology and virulence.
In general, the findings in this study have not only laid down
a valuable framework for designing and developing novel strate-
gies against bacterial infections, such as quorum quenching or
competitive inhibition, but also presented many intriguing ques-
tions for further investigations. For example, little is known of
how DF is synthesized in X. campestris pv. campestris, and it
remains unclear how DF is involved in the regulation of diverse
functions; in particular, xanthomonadin biosynthesis. In addi-
tion, the fascinating findings that production of DF is conserved
in a wide range of bacterial species, especially in human patho-
gens and ocean microorganisms, may substantially broaden the
scope of investigation on DF.
MATERIALS AND METHODS
Bacterial strains and growth conditions.
X. campestris pv. campestris wild-type strains XC1 and
8004, and their derivatives were grown at 28C in NYG (5 g of
peptone, 3 g of yeast extract, and 20 g of glycerol, pH 7.0) or
Vol. 24, No. 8, 2011 / 955
YEB (5 g of yeast extract, 10 g of tryptone, 5 g of sodium
chloride, 5 g of sucrose, and 0.5 g of MgSO4) media in the
dark. Escherichia coli strains were grown at 37C in Luria-
Bertani medium. Antibiotics were added at the following con-
centrations when required: kanamycin at 50 mg ml–1, gen-
tamicin at 50 mg ml–1, rifampicin at 25 mg ml–1, ampicillin at
100 mg ml–1, and tetracycline at 5 mg ml–1; 5-bromo-4-chloro-
3-indolyl -D-glucopyranoside at 60 mg ml–1 was included for
detection of -glucuronidase activity.
Gene deletion and functional complementation analysis.
Xcc4014 in-frame deletion mutants of X. campestris pv.
campestris strains 8004 and XC1 were generated according to
He and associates (2006) using the following primers: 4014-
FOR1: cgaggacgccggcaacatca, 4014-REV1: gggatcccagtggcac
gcgcaaatagc, 4014-FOR2: cgggatccgacggcgtgcacggctgagt, and
4014-REV2: ccgccgctggtaggcatcgtc. Xcc4014 and rpfF double-
deletion mutants were generated using mutant rpfF as the
parental strain and the same primers listed above. For comple-
mentation analysis, the coding region of Xcc4014 was ampli-
fied by PCR using: Xcc4014-FOR (cgggatcccgtgttccgcgccatct
attc) and Xcc4014-REV (cccaagctttcagccgtgcacgccgtcgat), and
was cloned under the control of lac promoter in pLAFR3. The
resulting construct was transferred into X. campestris pv. cam-
pestris strains through triparental mating. The Xcc4014 deletion
mutant was further verified by PCR and sequencing analysis.
Extraction and quantification of pigment.
The procedure described previously (Irey and Stall 1982) for
extraction of pigment from X. juglandis was used in this study
with minor modifications. Briefly, X. campestris pv. campes-
tris cultures were grown until the stationary phase (OD600 =
2.3), and the cells from 5 ml of culture were collected by cen-
trifugation. Pigments were extracted with 1 ml of methanol by
shaking for 5 min at room temperature. The amount of pig-
ment produced was expressed as the absorbance (OD at 445
nm) of crude pigment extracts (Poplawsky and Chun 1997).
Bioassay of DF.
The presence of DF was indicated by its effect on the resto-
ration of pigment in DF-deficient mutant xanB2. Two methods
were developed to detect the DF signal in this study. For quali-
tative detection of DF, 15 ml of NYG medium containing 0.8%
agarose were poured into a sterilized plastic petri dish (90 mm;
Sterilin, Newport, U.K.). After solidification, the agarose
plates were aseptically cut into separated rectangular bars of 1
cm in width by removing approximately 2-mm slices between
the bars. Bacterial cells, or 10-µl extracts or purified DF dis-
solved in methanol, were added to the upper end of agarose
bars. Drops of xanB2 cell culture (OD600 = 0.05) were then
spotted at progressively further distances from the end of aga-
rose bars loaded with the DF sample or bacterial strain to be
tested. The resulting plates were sealed and incubated at 30C
for 2 days. The presence of yellow-colored bacterial spots
indicated the activity of DF. For quantitative analysis of DF,
mutant xanB2 was grown in 5 ml of NYG liquid medium un-
til an OD600 of 1.0. DF samples were then added as indicated
and the bacterial cultures were grown for another 8 h to an
OD600 of 2.3. Cells were collected by centrifugation at 10,000
rpm for 5 min and pigments were extracted using 1 ml of
methanol (Irey and Stall 1982). The amount of pigment pro-
duced was determined by measuring absorbance at 445 nm
(Poplawsky and Chun 1997). The percentage (P) of pigment
restored by DF was defined as the OD445 of DF-treated
xanB2 cultures divided by that of the wild-type strain and
multiplied by 100. The quantity of DF was calculated using
the formula DF (µM) = 0.0741 e0.0507P, which was derived from
a dose–response plot of the biosensor xanB2 to DF, with a
correlation coefficiency (R2) of 0.98.
Purification and structural analysis of DF.
X. campestris pv. campestris XC1 was cultured in NYG me-
dium for 32 h. Bacterial supernatants (30 liters total) were col-
lected by centrifugation (3,800 rpm for 30 min at 4C). The
supernatants were adjusted to pH 4.0 by addition of 1 M hy-
drochloride prior to extraction with an equal volume of ethyl
acetate. The extraction was repeated twice and the extracts
were combined and dried using rotary evaporation at 40C.
The resulting residues were dissolved in 20 ml of methanol.
The crude extracts, divided into four batches, were subjected
to flash-column chromatography using a silica gel column (12
by 150 mm; Biotage Flash 12M cartridge) eluted with ethyl
acetate-hexane (25:75, vol/vol, 0.05% acetic acid) mixture.
The active fractions were combined and were further separated
by using HPLC with a C18 reverse-phase column (4.6 by 250
mm; Phenomenex Luna) eluted with methanol-water (30:70,
vol/vol, 0.05% acetic acid) at a flow rate of 1 ml/min in a
Waters 2695 system with 996 PDA detector. Purified DF was
subjected to ESI-MS analysis (Finnigan LCQ system). For
NMR analysis, the purified DF was dissolved in methanol-d4
and 1H and 13C spectra were recorded on a Bruker DRX400
400 MHz spectrometer at room temperature. The chemical
shifts () were quoted in parts per million (ppm), and the cou-
pling constant (J) values were recorded in Hertz (Hz).
Time-course analysis of DF production.
The relationship between bacterial population density and DF
production was examined by sampling cell culture at different
stages of bacterial growth. X. campestris pv. campestris XC1
was grown in 1,000 ml of NYG broth with shaking at 28C.
Samples (50 ml) were collected at 4, 8, 12, 16, 20, 24, 28, 32,
and 36 h after inoculation and DF was extracted as described
above. After evaporation, the residues were dissolved in 100 µl
of methanol, and 10-µl aliquots were added to 5 ml of xanB2
cell cultures for quantitative analysis of DF as described above.
H2O2 sensitivity and bacterial viability test.
X. campestris pv. campestris strains were grown in YEB me-
dium with or without DF to an OD600 of 2.1. The cultures were
then diluted with fresh YEB medium to an OD600 of 0.1. Bac-
terial suspensions were treated with 176, 528, and 880 µM
H2O2 for 30 min. Bacterial cells were then spun down and
washed once in phosphate buffer (50 mM, pH 7.4). Cells were
serially diluted and spotted onto YEB agar plates, then incu-
bated at 30C for 2 days before counting CFU. The experiment
was repeated three times.
For bacterial viability testing, the wild type and mutant
xanB2 were subcultured in YEB liquid medium with or with-
out DF as indicated. Bacterial viability at OD600 = 1.5, 2.0, 2.3,
and 2.5 was determined by plate-counting of CFU. For deter-
mination of survival time, the wild type and xanB2 mutant
were subcultured on YEB agar plate with or without DF as
indicated. After growth at 30C for 2 days, the agar plates
were maintained at room temperature in the dark. From the
seventh day onward, the viability of cells was tested every 2
days by subculture of three randomly selected colonies from
each treatment on fresh YEB plate. Growth of bacterial cells
was monitored for 2 days at 30C. Loss of viability was fur-
ther confirmed by growing five colonies per treatment in liquid
YEB medium, separately.
Bioassays and virulence test.
DSF bioassay was performed as previously described (Wang
et al. 2004). Extracellular cellulase and protease activity were
956 / Molecular Plant-Microbe Interactions
analyzed as described by He and associates (2009). To analyze
the production of EPS, the supernatants of overnight bacterial
culture in YEB medium (10 ml, OD600 of approximately 2.3)
were collected by centrifugation at 10,000 rpm for 10 min.
Two volumes of absolute ethanol were added the supernatants
and the mixtures were kept at –20C for a half-hour. The pre-
cipitated EPS were spun down and dried in a 55C oven over-
night before determination of dry weights. Cell aggregation
assay and virulence tests were conducted according to the scis-
sors-clipping method described previously (He et al. 2006,
2009). Oligomicroarray analysis was conducted according to
the methods described previously (He et al. 2006). The total
RNAs were extracted from bacterial cells at OD600 = 2.3.
Assay for bacterial systemic invasion.
Fully matured Chinese cabbage (Wongbok) leaves were
used for the assay. The bottom part (approximately 1 cm) of
each leaf was cut off and the resulting wound was immediately
immersed in fresh XC1 bacterial culture (OD600 = 0.1) for 5 s.
The inoculated leaf was then wrapped with plastic film and
kept in an incubator at 25C with 80% humidity for a week.
The infected leaf was marked every 2 cm from the inoculated
end, and a slice of tissue (0.2 by 1.0 cm) was taken out at each
mark for sap extraction. The extracted sap was serially diluted
and 100 µl of each dilution was spread on an agar plate con-
taining rifampicin (25 µg/ml) for colony counting. For each
treatment, 10 leaves were used and three independent experi-
ments were performed.
This research was supported by funding from the Agency for Science,
Technology and Research (A*STAR), Singapore and Shanghai Jiao Tong
Andrewes, A. G., Jenkins, C. L., Starr, M. P., Shepherd, J., and Hope, H.
1976. Structure of xanthomonadin I, a novel dibrominated aryl-polyene
pigment produced by the bacterium Xanthomonas juglandis. Tetrahe-
dron Lett. 45:4023-4024.
Bestwick, C. S., Brown, I. R., Bennett, M. H. R., and Mansfield, J. W.
1997. Localization of hydrogen peroxide accumulation during the hy-
persensitive reaction of lettuce cells to Pseudomonas syringae pv.
phaseolicola. Plant Cell 9:209-221.
Bindschedler, L. V., Dewdney, J., Blee, K. A., Stone, J. M., Asai, T.,
Plotnikov, J., Denoux, C., Hayes, T., Gerrish, C., Davies, D. R., et al.
2006. Peroxidase dependent apoplastic oxidative burst in Arabidopsis
required for pathogen resistance. Plant J. 47:851-863.
Chun, W., Cui, J., and Poplawsky, A. R. 1997. Purification, characteriza-
tion and biological role of a pheromone produced by Xanthomonas
campestris pv. campestris. Physiol. Mol. Plant Pathol. 51:1-14.
da Silva, A. C., Ferro, J. A., Reinach, F. C., Farah, C. S., Furlan, L. R.,
Quaggio, R. B., Monteiro-Vitorello, C. B., Van Sluys, M. A., Almeida,
N. F., Alves, L. M., do Amaral, A. M., Bertolini, M. C., Camargo, L. E.,
Camarotte, G., Cannavan, F., Cardozo, J., Chambergo, F., Ciapina, L.
P., Cicarelli, R. M., Coutinho, L. L., Cursino-Santos, J. R., El-Dorry,
H., Faria, J. B., Ferreira, A. J., Ferreira, R. C., Ferro, M. I., Formighieri,
E. F., Franco, M. C., Greggio, C. C., Gruber, A., Katsuyama, A. M.,
Kishi, L. T., Leite, R. P., Lemos, E. G., Lemos, M. V., Locali, E. C.,
Machado, M. A., Madeira, A. M., Martinez-Rossi, N. M., Martins, E.
C., Meidanis, J., Menck, C. F., Miyaki, C. Y., Moon, D. H., Moreira, L.
M., Novo, M. T., Okura, V. K., Oliveira, M. C., Oliveira, V. R., Pereira,
H. A., Rossi, A., Sena, J. A., Silva, C., de Souza, R. F., Spinola, L. A.,
Takita, M.A., Tamura, R. E., Teixeira, E. C., Tezza, R. I., Trindade dos
Santos, M., Truffi, D., Tsai, S. M., White, F. F., Setubal, J. C., and
Kitajima, J. P. 2002. Comparison of the genomes of two Xanthomonas
pathogens with differing host specificities. Nature 417:459-463.
Durgapal, J. C. 1996. ‘Albino’ mutation in rice bacterial blight pathogen.
Curr. Sci. 70:15.
Goel, A. K., Rajagopal, l., Nagesh, N., and Sonti, R. V. 2002. Genetic locus
encoding functions involved in biosynthesis and outer membrane local-
ization of xanthomonadin in Xanthomonas oryzae pv. oryzae. J. Bacte-
He, Y. W., and Zhang, L. H. 2008. Quorum sensing and virulence regula-
tion in Xanthomonas campestris. FEMS (Fed. Eur. Microbiol. Soc.) Mi-
crobiol. Rev. 32:842-857.
He, Y. W., Xu, M., Lin, K., Ng, Y. J., Wen, C. M., Wang, L. H., Liu, Z.
D., Zhang, H. B., Dong, Y. H., Dow, J. M., and Zhang, L. H. 2006.
Genome scale analysis of diffusible signal factor regulon in Xantho-
monas campestris pv. campestris: identification of novel cell-cell
communication-dependent genes and functions. Mol. Microbiol.
He, Y. W., Ng, A. Y., Xu, M., Lin, K., Wang, L. H., Dong, Y. H., and
Zhang, L. H. 2007. Xanthomonas campestris cell-cell communication
involves a putative nucleotide receptor protein Clp and a hierarchical
signalling network. Mol. Microbiol. 64:281-292.
He, Y. W., Boon, C., Zhou, L., and Zhang, L. H. 2009. Co-regulation of
Xanthomonas campestris virulence by quorum sensing and a novel two-
component regulatory system RavS/RavR. Mol. Microbiol. 71:1464-
He, Y. W., Wu, J. E., Cha, J. S., and Zhang, L. H. 2010. Rice bacterial
blight pathogen Xanthomonas oryzae pv. oryzae produces multiple
DSF-family signals in regulation of virulence factor production. BMC
Microbiol. 10: 187.
Irey, M. S., and Stall, R. E. 1982. Value of xanthomonadins for identifica-
tion of pigmented Xanthomonas campestris pathovars. Pages 85-95 in:
Proc. Fifth International Conference on Plant-Pathogenic Bacteria. Cali,
Jenkins, C. L., and Starr, M. P. 1982.The pigment of Xanthomonas populi
is a nonbrominated aryl-heptaene belonging to xanthomonadin pigment
group II. Curr. Microbiol. 7:195-198.
Leyns, F., De Cleene, M., Swings, J., and de Ley, J. 1984. The host range
of the genus Xanthomonas. Bot. Rev. 50:308-355.
Poplawsky, A. R., and Chun, W. 1997. pigB determines a diffusible factor
needed for extracellular polysaccharide slime and xanthomonadin pro-
duction in Xanthomonas campestris pv. campestris. J. Bacteriol.
Poplawsky, A. R., and Chun, W. 1998. Xanthomonas campestris pv.
campestris requires a functional pigB for epiphytic survival and host
infection. Mol. Plant-Microbe Interact. 11:466-475.
Poplawsky, A. R., Kawalek, M. D., and Schaad, N. W. 1993. A xanthomo-
nadin-encoding gene cluster for the identification of pathovars of Xan-
thomonas campestris. Mol. Plant-Microbe Interact. 6:545-552.
Poplawsky, A. R., Chun, W, Slater, H., Daniels, M. J., and Dow, J. M.
1998. Synthesis of extracellular polysaccharide, extracellular enzymes,
and xanthomonadin in Xanthomonas campestris: evidence for the in-
volvement of two intercellular regulatory signals. Mol. Plant-Microbe
Poplawsky, A. R., Urban, S. C., and Chun, W. 2000. Biological role of
xanthomonadin pigments in Xanthomonas campestris pv. campestris.
Appl. Environ. Microbiol. 66:5123-5127.
Poplawsky, A. R., Walters, D. M., Rouviere, P. E., and Chun, W. 2005. A
gene for a dioxygenase-like protein determines the production of the
DF signal in Xanthomonas campestris pv. campestris. Mol. Plant
Qian, W., Jia, Y., Ren, S. X., He, Y. Q., Feng, J. X., Lu, L. F., Sun, Q.,
Ying, G., Tang, D. J., Tang, H., Wu, W., Hao, P., Wang, L., Jiang, B. L.,
Zeng, S., Gu, W. Y., Lu, G., Rong, L., Tian, Y., Yao, Z., Fu, G., Chen,
B., Fang, R., Qiang, B., Chen, Z., Zhao, G. P., Tang, J. L., and He, C.
2005. Comparative and functional genomic analyses of the pathogenic-
ity of phytopathogen Xanthomonas campestris pv. campestris. Genome
Rajagopal, L., Sundari, C. S., Balasubramanian, D., and Sonti, R. V. 1997.
The bacterial pigment xanthomonadin offers protection against photo-
damage. FEBS (Fed. Eur. Biochem. Soc.) Lett. 415:125-128.
Schaad, N. W., and Stall, R. E. 1988. Xanthomonas. Pages 81-94 in: Labo-
ratory Guide for Identification of Plant Pathogenic Bacteria, 2nd ed. N.
W. Schaad, ed. American Phytopathological Society Press, St. Paul,
Starr, M. P. 1981. The genus Xanthomonas. Pages 742-763 in: The Pro-
karyotes, Vol. 1. M. P. Starr, H. Stolp, G. H. Truper, A. Balows, and G.
H. Schlegel, eds. Springer-Verlag, Berlin.
Starr, M. P., and Stephens, W. L. 1964. Pigmentation and taxonomy of the
genus Xanthomonas. J. Bacteriol. 87:293-302.
Starr, M. P., Jenkins, C. L., Bussey, L. B., and Andrewes, A. G. 1977.
Chemotaxonomic significance of the xanthomonadins, novel brominated
aryl-polyene pigments produced by bacteria of the genus Xanthomonas.
Arch. Microbiol. 113:1-9.
Stephens, W. L., and Starr, M. P. 1963. Localization of carotenoid pigment
in the cytoplasmic membrane of Xanthomonas juglandis. J. Bacteriol.
Sutherland, M. W. 1991. The generation of oxygen radicals during host
Vol. 24, No. 8, 2011 / 957 Download full-text
plant responses to infection. Physiol. Mol. Plant Pathol. 39:79-93.
Tsuchiya, K., Mew, T. W., and Wakimoto, S. 1982. Bacteriological and
pathological characteristics of wild-type and induced mutants of Xan-
thomonas campestris pv. oryzae. Phytopathology 72:43-46.
Vorhölter, F. J., Schneiker, S., Goesmann, A., Krause, L., Bekel, T., Kaiser,
O., Linke, B., Patschkowski, T., Rückert, C., Schmid, J., Sidhu, V. K.,
Sieber, V., Tauch, A., Watt, S. A., Weisshaar, B., Becker, A., Niehaus,
K., and Pühler, A. 2008. The genome of Xanthomonas campestris pv.
campestris B100 and its use for the reconstruction of metabolic path-
ways involved in xanthan biosynthesis. J. Biotechnol. 134:33-45.
Wang, L. H., He, Y. W., Gao, Y., Wu, J. E., Dong, Y. H., He, C., Wang, S.
X, Weng, L. X., Xu, J. L. Tay, L., Fang R. X., and Zhang, L. H. 2004. A
bacterial cell-cell communication signal with cross-kingdom structural
analogues. Mol. Microbiol. 51:903-912.
Yajima , A., Imai, N., Poplawsky, A. R., Nukada, T., and Yabuta, G. 2010.
Synthesis of a proposed structure for the diffusible extracellular factor
of Xanthomonas campestris pv. campestris. Tetrahedron Lett. 51:2074-