A cbb(3)-type cytochrome C oxidase contributes to Ralstonia solanacearum R3bv2 growth in microaerobic environments and to bacterial wilt disease development in tomato.

Abstract

Ralstonia solanacearum race 3 biovar 2 (R3bv2) is an economically important soilborne plant pathogen that causes bacterial wilt disease by infecting host plant roots and colonizing the xylem vessels. Little is known about R3bv2 behavior in the host rhizosphere and early in bacterial wilt pathogenesis. To explore this part of the disease cycle, we used a novel taxis-based promoter-trapping strategy to identify pathogen genes induced in the plant rhizosphere. This screen identified several rex (root exudate expressed) genes whose promoters were upregulated in the presence of tomato root exudates. One rex gene encodes an assembly protein for a high affinity cbb(3)-type cytochrome c oxidase (cbb(3)-cco) that enables respiration in low-oxygen conditions in other bacteria. R3bv2 cbb(3)-cco gene expression increased under low-oxygen conditions, and a cbb(3)-cco mutant strain grew more slowly in a microaerobic environment (0.5% O(2)). Although the cco mutant could still wilt tomato plants, symptom onset was significantly delayed relative to the wild-type parent strain. Further, the cco mutant did not colonize host stems or adhere to roots as effectively as wild type. These results suggest that R3bv2 encounters low-oxygen environments during its interactions with host plants and that the pathogen depends on this oxidase to help it succeed in planta.

Full text

Page 1

1042 / Molecular Plant-Microbe Interactions
MPMI Vol. 23, No. 8, 2010, pp. 1042–1052. doi:10.1094/MPMI-23-8-1042. © 2010 The American Phytopathological Society
A cbb3-Type Cytochrome C Oxidase Contributes
to Ralstonia solanacearum R3bv2 Growth
in Microaerobic Environments and
to Bacterial Wilt Disease Development in Tomato
Jennifer Colburn-Clifford and Caitilyn Allen
1Department of Plant Pathology, University of Wisconsin-Madison, Madison WI 53706, U.S.A.
Submitted 19 November 2009. Accepted 17 April 2010.
Ralstonia solanacearum race 3 biovar 2 (R3bv2) is an eco-
nomically important soilborne plant pathogen that causes
bacterial wilt disease by infecting host plant roots and colo-
nizing the xylem vessels. Little is known about R3bv2 be-
havior in the host rhizosphere and early in bacterial wilt
pathogenesis. To explore this part of the disease cycle, we
used a novel taxis-based promoter-trapping strategy to iden-
tify pathogen genes induced in the plant rhizosphere. This
screen identified several rex (root exudate expressed) genes
whose promoters were upregulated in the presence of to-
mato root exudates. One rex gene encodes an assembly pro-
tein for a high affinity cbb3-type cytochrome c oxidase
(cbb3-cco) that enables respiration in low-oxygen condi-
tions in other bacteria. R3bv2 cbb3-cco gene expression
increased under low-oxygen conditions, and a cbb3-cco mu-
tant strain grew more slowly in a microaerobic environ-
ment (0.5% O2). Although the cco mutant could still wilt
tomato plants, symptom onset was significantly delayed
relative to the wild-type parent strain. Further, the cco mu-
tant did not colonize host stems or adhere to roots as effec-
tively as wild type. These results suggest that R3bv2 encoun-
ters low-oxygen environments during its interactions with
host plants and that the pathogen depends on this oxidase
to help it succeed in planta.
The root and its immediate surrounding area, the rhizosphere,
are important for plant fitness. Under natural conditions, the
rhizosphere contains vast numbers of microorganisms that inter-
act not only with each other but with the plant root as well
(Rainey 1999). Plants exude nutrients from roots, with up to
20% of available carbon in the rhizosphere coming from roots
(Brencic and Winans 2005; Handelsman 1996). It is this rela-
tively rich, fluctuating, and competitive environment that the
pathogen Ralstonia solanacearum inhabits, growing and mov-
ing toward nutrient sources and signals from host plants that
may be susceptible to disease.
R. solanacearum, the causal agent of bacterial wilt, is an
economically important pathogen affecting crop production
worldwide (Hayward 1991). The R. solanacearum species
complex is a large heterogeneous group of motile, soilborne β-
proteobacteria distributed primarily in tropical and subtropical
regions, although it is also problematic in temperate climates.
R. solanacearum enters the vascular tissue of its host through
natural openings or mechanical wounds. The pathogen multi-
plies in the root cortex and, then, rapidly invades and colonizes
the xylem, ultimately disrupting water flow. Symptoms of bac-
terial wilt include stunting, chlorosis, and sudden wilting of
the plant (Hayward 1991).
The R. solanacearum species complex includes a group his-
torically known as race 3 biovar 2 (R3bv2) and now classified
as phylotype II, sequevar 1 (Fegan and Prior 2005). R3bv2
causes bacterial wilt of tomato, southern wilt of geranium, and
brown rot of potato, diseases that lead to tremendous losses,
particularly in the tropics and subtropics (Allen et al. 2001).
Because it poses a threat to temperate potato production,
R3bv2 is a quarantine pest in Canada and Europe and a select
agent pathogen in the U.S. (Lambert 2002).
R. solanacearum virulence is quantitative and complex, with
many contributing factors, such as a suite of type III-secreted
effectors, bacterial extracellular polysaccharide, several plant
cell wall–degrading enzymes, motility, and energy taxis
(González 2003; Saile et al. 1997; Schell 2000; Tans-Kersten
et al. 2001; Yao and Allen 2006). Most known virulence fac-
tors were identified based on their roles in advanced wilt dis-
ease. We previously described >150 R. solanacearum genes
specifically expressed in mid-phase bacterial wilt disease in
tomato, which offered insight into the physiology of this patho-
gen during pathogenesis (Brown and Allen 2004). However,
despite their epidemiological importance, little is known about
the traits that this soilborne pathogen needs to survive in the
rhizosphere and to begin colonizing host roots.
A significant proportion of the photosynthate that plants
allocate to their roots is released into the rhizosphere (Marschner
1995). These root exudates consist of high–molecular weight
mucilage and acid phosphatase as well as amino acids, pheno-
lics, sugars, and organic acids, the latter two being the most
abundant (Marschner 1995). The total amount of root exudate
and its composition varies according to plant species, nutritional
status, age, presence of microbes, and nitrogen amendments
(Huber and Watson 1974; Lugtenberg; Marschner 1995). Root
exudates form a nutrient source and a habitat signal for micro-
organisms (Marschner 1995). R. solanacearum chemotaxis
toward host-plant root exudates by means of swimming motil-
ity guided by sensor proteins that direct movement (Yao and
Allen 2006). Chemotaxis receptors transmit environmental sig-
nals to the flagellar motor via the core taxis signaling proteins
CheA and CheW (Yao and Allen 2006). Chemotaxis is essen-
Current address for J. Colburn-Clifford: Department of Plant Pathology
and Microbiology, University of California-Riverside, Riverside. CA 92521,
U.S.A.
Corresponding author: C. Allen; E-mail: cza@plantpath.wisc.edu; Tele-
phone: +1.608.262.9568; Fax: +1.608.263.2626.

Page 2

Vol. 23, No. 8, 2010 / 1043
tial for location and invasion of host roots in R. solanacearum;
both cheW and cheA mutants are quantitatively reduced in
virulence on tomato (Yao and Allen 2006). The observation
that R. solanacearum can detect and respond to plant root exu-
dates led to a series of questions. What R3bv2 genes are
induced in response to host root exudates? What is their contri-
bution to life in the rhizosphere and early host-pathogen inter-
actions?
We sought to identify and characterize R3bv2 genes that
affect bacterial behavior in the rhizosphere and early in patho-
genesis. To accomplish this, we used a taxis-based expression
technology strategy to identify pathogen genes induced by
tomato root exudates (TRE). This in vitro approach was fash-
ioned after in vivo expression technology (IVET), a popula-
tion-to-genes approach that takes the ecological context of the
organism into consideration and identifies genes based on gene
induction under specific conditions (Rainey and Preston 2000).
Originally developed for an in planta application (Osbourne et
al. 1987), IVET uses a promoter-probe plasmid containing
genomic DNA fragments that are cloned directly in front of a
marker gene and a reporter gene, both of which are promoter-
less, to create a transcriptional fusion (Rainey and Preston
2000). These constructs are introduced into a mutant strain,
which cannot succeed in a specific selective environment (usu-
ally the eukaryotic host). If a DNA fragment cloned upstream
of the marker gene contains a promoter that is induced in the
selective environment, the wild-type bacterial phenotype is
restored to the mutant and it will be selected in the differentiat-
ing environment. The sequence of the cloned promoter sug-
gests the function of one or more corresponding upregulated
genes. This strategy has successfully identified genes of plant-
associated bacteria induced under various conditions, includ-
ing R. solanacearum genes involved in virulence and adapta-
tion to the host in mid-bacterial wilt disease of tomato (Boch
et al. 2002; Brown and Allen 2004; Dunn et al. 2003; Gal et al.
2003; Marco et al. 2005; Oke and Long 1999).
Traits that improve survival in the rhizosphere environment
are undoubtedly important for R. solanacearum’s success as a
pathogen. We hypothesized that, in addition to acting as
chemoattractants, host root exudates also induce expression of
bacterial rex (root exudate expressed) genes, whose products
may facilitate root invasion and colonization or act directly as
early virulence factors. We identified R3bv2 strain UW551
genes that were upregulated in the presence of chemicals
found in the tomato rhizosphere. These included a locus en-
coding a high-affinity cbb3-type cytochrome c oxidase (cbb3-
cco) that was required for full virulence and for growth of
R3bv2 under the low-oxygen conditions found in the
rhizosphere and plant xylem vessels.
RESULTS
A taxis-based expression technology strategy identified
R. solanacearum R3bv2 genes induced by host root exudates.
Our modified IVET screen selected pathogen promoters that
were active in the presence of TRE. We created a nonchemo-
tactic deletion mutant, UW551ΔcheW, and used it to evaluate
chemotaxis as a selectable trait for this screen. UW551ΔcheW
did not form a halo on taxis motility agar, but it did retain
swimming ability as observed under a microscope (data not
shown). Chemotaxis was restored to UW551ΔcheW by com-
plementation with a plasmid-borne copy of the cheW locus.
To determine if we could select a few tactic cells from a
large pool of nontactic bacteria, we mixed 100 cells of wild-
type UW551 with 105 cells of
was spotted onto taxis agar containing TRE. In response to the
nutrient gradient created by bacterial growth, the small tactic
UW551ΔcheW, and this mixture
subpopulation migrated away from the nontactic majority and
formed an easily distinguishable halo (data not shown). This
demonstrated that restoration of taxis was a selectable trait
under our conditions.
To select promoters active in the presence of TRE, we con-
structed a promoter-trapping integrative plasmid vector,
pJCcheW, which contains the promoterless cheW open reading
frame (ORF) cloned upstream from a promoterless β-glucuroni-
dase (gus) reporter gene. Cloning wild-type genomic UW551
DNA fragments into the XhoI site upstream of the promoter-
less cheW created transcriptional fusions with the potential to
restore chemotaxis and express GUS activity (Fig. 1A).
To confirm that introduced promoters could restore taxis to
the cheW mutant, we cloned various well-characterized pro-
moters into the XhoI site of pJCcheW, transformed the con-
structs into UW551ΔcheW, and measured taxis of the resulting
Fig. 1. Overview of the rex (root exudate expressed genes) screen used to
identify Ralstonia solanacearum race 3 biovar 2 (R3bv2) genes induced in
response to tomato exudate. A, The rex screen vector pJCcheW was con-
structed by cloning the promoterless cheW from Ralstonia solanacearum
UW551 into pVO155. pJCcheW consists of promoterless cheW and uidA
(gus) genes, each with its own translational start (RBS). Random UW551
genomic DNA fragments were cloned into the XhoI restriction site, which
is preceded by a transcriptional terminator (stem loop) and stop codons in
all three reading frames. The oriT allowed for conjugation of the vector
into R. solanacearum and neo (kanamycin resistant) provided a selectable
marker. B, For the rex screen, random fragments of UW551 chromosomal
DNA were cloned into the promoter-trapping vector to create a library.
Strains with restored chemotactic response (active promoters) were selected
by growth on taxis agar supplemented with 1% root exudate. Chemotactic
strains were isolated, and the rex screen vector was recovered by retro-
transfer. Chromosomal DNA inserts were sequenced for bioinformatic
analysis.

Page 3

1044 / Molecular Plant-Microbe Interactions
strains. We used the pilA promoter, which drives transcription
of the abundant type IV fimbrial protein (Kang et al. 2002), the
flhDC promoter, which mediates low-level expression of a
flagellar regulator (Tans-Kersten et al. 2004), and cheW’s own
native promoter (Yao and Allen 2006). These three promoters
all restored chemotaxis to UW551ΔcheW, but the cheW pro-
moter did not restore taxis when cloned in the reverse orienta-
tion (data not shown). Thus both native and heterologous pro-
moters could drive expression of the promoterless cheW ORF
in pJCcheW and thereby identify genes in the rex screen.
An overview of the rex screen is shown in Figure 1B. A
library of UW551 DNA fragments was cloned into pJCcheW
and introduced into UW551ΔcheW by triparental mating. At
least 80,000 strains were screened for the restoration of chemo-
taxis, more than the 68,000 required for 95% coverage of R.
solanacearum’s 5.7-Mb genome (Ausubel et al. 1995). Eight
fusions had taxis halo diameters as large or larger than the
parental control strains, with diameters ranging from 0.5 to 3.5
cm larger than wild type. The presence of TRE increased the
expression of these fusions by two- to fivefold (Table 1).
One rex insertion was in a gene cluster encoding synthesis
and assembly of a high oxygen affinity cco.
The sequence adjacent to the rex4 insertion resembled that
encoding a 4Fe4S ferredoxin involved in electron transport in
cbb3, a high affinity oxidase that contributes to aerobic respira-
tion under microaerobic conditions (RRSL_01397; Table 1).
RRSL_01397 encodes a protein with 37 and 35% identity at
the amino acid level to CcoG (also referred to as FixG) from
Bradyrhizobium japonicum and Rhizobium etli, respectively.
CcoG is required for cbb3-Cco expression under oxygen-limit-
ing conditions in B. japonicum (Preisig et al. 1996; Zufferey et
al. 1996). ClustalW alignment of the UW551 CcoG homolog
revealed the conserved ferredoxin motif Cys-X1-X1-Cys-X3-
X4-Cys-X5-X6-X7-Cys-Pro at positions 284 and 308 without
the terminal proline, similar to motifs found in Rhodobacter
sphaeroides RdxA (Neidle and Kaplan 1992), FixG in B. ja-
ponicum, and FixG in Sinorhizobium meliloti. Genes encoding
Cbb3-Cco are typically arranged in two operons, ccoNOQP
(fixNOQP), which encodes structural proteins and is always
found very close to ccoGHIS (fixGHIS), which encodes the
assembly proteins. Both operons are essential for respiration
under microaerobic conditions and symbiosis in B. japonicum
(Preisig et al. 1996).
All genes required for Cbb3-Cco synthesis and assembly
appear to be present in UW551. We hypothesized that the
region tagged by the rex4 insertion encodes a high-affinity Cco
that helps R. solanacearum succeed in hypoxic environments,
such as the rhizosphere and xylem of its host plants. Such an
operon is likely to be upregulated under low-oxygen condi-
tions. We measured expression of the rex4::gus fusion at three
oxygen levels, 21, 5, and 0.5%, corresponding to atmospheric,
slightly hypoxic, and microaerobic conditions, respectively
(Loesche 1969). After 24 h, rex4 expression increased seven-
fold in strains grown under 5% oxygen and more than 30-fold
in strains grown under 0.5% oxygen relative to the strain
grown at 21% oxygen. This induction of gene expression in
hypoxic conditions was consistent with its putative function as
a high-affinity cco.
To further explore the biological role of Cbb3-Cco in R. so-
lanacearum, we mutated the first gene in the operon, which
encodes the CcoN structural protein in the Cbb3-Cco complex
(Fig. 2). Mutation of this ORF was expected to eliminate Cbb3-
Cco function, since the cco genes are coregulated in other sys-
tems and CcoN is necessary for Cco function (Mandon et al.
1994; Marchal et al. 1998; Preisig et al. 1996; Zufferey et al.
1998). The ccoN ORF was disrupted by a site-directed insertion
of a gentamicin (Gm)-resistance cassette, and the expected mu-
tation in strain UW551cco was confirmed by polymerase chain
reaction (PCR) and Southern blot analysis (data not shown).
A cbb3-cco mutant grew more slowly
in microaerobic environments.
We measured the growth rate of UW551cco, its wild-type
parent strain, and a complemented mutant in culture in the
presence of 21 (atmospheric) or 0.5% oxygen. After 24 h, all
three strains grew similarly at 21% oxygen, but at 0.5% oxy-
gen, UW551cco grew on average 53% slower than either
UW551 or the complemented mutant (P = 0.013; Fig. 3A). Af-
ter 5 days, the trend remained the same, with UW551cco at
approximately 73% the population size of the wild-type and
complemented strains (P = 0.012; Fig. 3B). We observed simi-
lar differences under semi-aerobic conditions in which bacte-
rial cultures were grown in large volumes at low rpm to provoke
hypoxia (data not shown). These results demonstrated that a
defined ccoN mutant was quantitatively reduced in ability to
grow in low-oxygen environments and constituted phenotypic
evidence that this region encodes a high-affinity oxidase.

Fig 2. Genomic context of genes encoding structural and assembly proteins of cbb3-type cytochrome c oxidase (cbb3-Cco) in the Ralstonia solanacearum
UW551 genome. Gene annotations are based on conserved cbb3-cco loci in other plant-associated soilborne bacteria. ccoN was disrupted by inserting
aacC1, encoding gentamicin resistance (open triangle), into a unique restriction site. Arrowheads represent predicted open reading frames.
Table 1. Ralstonia solanacearum UW551 genes identified in the rex (root exudate expressed genes) screen
rex ID Genea
Predicted protein
E value Closest match outside Ralstonia Fold inductionb
5 ± 0.9
2 ± 0.2
5 ± 1.2
4 ± 1.1
4 ± 2.8
2 ± 1.0
3 ± 1.1
rex1
rex2
rex3
rex4
rex5
rex6
rex7
RRSL_00505 (hrpG)
RRSL_01880 (pilP)
RRSL_02649 (pehR)
RRSL_01397
RRSL_03240 (dps)
RRSL_03587
RRSL_01657 (vsrA)
Transcriptional regulator
Type IV pili assembly
Transcriptional regulator
Iron-sulfur 4Fe-4S/ferredoxin
Nonspecific DNA binding
Oxidoreductase, glyoxalate metabolism
Sensory transduction, histidine kinase
7e-40
e-32
0.0
0.0
6e-71
6e-77
0.0
Burkholderia graminis
Acidovorax sp.
Janthinobacterium sp.
Bordetella bronchiseptica
Burkholderia oklahomensis
Dechloromonas aromatica
Cupriavidus taiwanensis
a Gene names are based on annotations in the UW551 or GMI1000 genome sequences.
b Fold induction of the rex gene plus or minus standard error, based on comparison of β-glucuronidase (gus) activity in minimal media with 1% tomato root
exudate versus unamended minimal media.

Page 4

Vol. 23, No. 8, 2010 / 1045
The Cbb3-cco is required for full virulence on tomato.
Tomato plants soil-soak inoculated with UW551cco devel-
oped bacterial wilt symptoms more slowly than did plants in-
oculated with the wild-type parental strain (P < 0.0001; Fig.
4A). Beginning 4 days after inoculation, disease development
in UW551cco-inoculated plants lagged behind that induced by
the wild-type strain, although, by 13 days after inoculation,
both strains caused full mortality. A similar difference in dis-
ease progress was observed when approximately 200 bacteria
were introduced directly into plant vasculature through a cut
petiole (P = 0.010; Fig. 4B).
UW551cco had reduced ability to colonize tomato stems.
To quantify the cell density of R. solanacearum strains in
plant stems, we inoculated cut tomato petioles with approxi-
mately 200 cells of either UW551 or UW551cco and meas-
ured bacterial population sizes in stems over time. By day 4,
the wild-type strain had colonized 53% of inoculated plants
with a mean population of 4.3 × 109 CFU per gram of stem,
while the cco mutant colonized 40% of plants with a mean
population of 2.8 × 108 CFU per gram of stem. At 6 days
after inoculation, 89% of stems from plants inoculated with
either strain contained detectable bacteria. However, mean
cell populations of UW551 were 4.3 × 1010 CFU per gram of
stem, while those for UW551cco were more than 30-fold
lower at 1.3 × 109 CFU/g. The median bacterial population
size differed significantly between plants inoculated with
UW551 (1.1 × 10
(P = 0.027; Fig. 5A). There were no significant differences
between strain stem populations by 8 days after inoculation,
but 80% of tomato plants inoculated with wild type had died,
while only 40% of plants inoculated with the cco mutant had
died (data not shown). These results indicate that R. solana-
cearum needs a functional Cbb3-Cco for rapid colonization
of tomato vasculature.
9 CFU/g) and UW551cco (8.2 × 10
7 CFU/g)
The cco mutant had reduced competitive fitness
on tomato plants.
We measured the ability of the cco mutant to invade and
colonize tomato plants in competition with its wild-type parent
strain. Plants were soil-soak inoculated with a 1:1 mixture of
wild-type UW551-rif and the UW551cco mutant; when wilt-

Fig. 3. UW551 cytochrome c oxidase (cco) mutant grew more slowly than wild type in oxygen-limiting conditions. A, Bars represent mean population sizes
after 24 and B, 120 h of growth in buffered minimal medium supplemented with 0.2% glucose at either 21% (atmospheric) or 0.5% (microaerobic) O2. At
both timepoints, growth of UW551cco (gray bars) was significantly reduced at 0.5% O2 when compared to the wild-type parent (black bars) and the
complemented mutant strain UW551cco(pUFJcco) (white bars). Averages are of five independent assays for the 24-h timepoint and three independent assays
for the 120-h timepoint. Asterisks represent significantly different means (P < 0.05). Bars represent standard error of the mean.

Fig. 4. Ralstonia solanacearum UW551cco caused delayed bacterial wilt disease on tomato. A, ‘Bonny Best’ plants (21 to 22 days old) were inoculated in a
naturalistic soil-soak inoculation or B, directly into tomato stems via cut petioles. Plants were rated daily on a disease index scale of 0 to 4. Disease
development in plants inoculated with UW551cco mutant (squares) lagged significantly relative to its wild-type parent (circles). Results shown are the means
of three independent assays with 16 plants per treatment. Bars represent standard error of the mean.

Page 5

1046 / Molecular Plant-Microbe Interactions
ing symptoms appeared, bacterial populations in the upper
crown were quantified. The median population of wild-type
cells in tomato stems was 2.2 × 109 CFU/g, while that of the
cco mutant strain was 120-fold lower at 1.0 × 108 CFU/g (P =
0.075; Fig 5B). Mean population sizes of UW551cco in to-
mato stems also trended lower than those of UW551-rif.
The cco mutant had reduced ability
to adhere to tomato roots.
To determine the contribution of cbb3-cco to root attach-
ment and colonization, we measured the relative ability of the
wild-type and cco mutant strains to adhere to and colonize to-
mato seedling roots. Roots incubated 1.5 h with a suspension
of UW551-GFP (green fluorescent protein) carried adhering
populations 10-fold greater than those on roots inoculated with
UW551cco-GFP (P = 0.042; Fig. 5C). After 4 h of incubation,
UW551cco adhering populations increased slightly, but the
average adhering population of the wild-type strain remained
more than twice as large (P = 0.158; Fig. 5C). There was no
visible difference between strains in root colonization, as as-
sessed by fluorescence microscopy (data not shown).
DISCUSSION
We developed a novel strategy that identified seven R. solana-
cearum genes that are upregulated in response to TRE. Several
of these rex genes were previously known to be plant-induced,
providing a reassuring internal control that the screen functioned
as expected. For example, HrpG, which activates hrp gene ex-
pression in response to a plant signal (Brito et al. 1999), is
required for successful colonization of tomato roots and for bac-
terial wilt virulence (Vasse et al. 2000). The two-component
regulator PehSR positively controls expression of the extracellu-
lar plant cell wall–degrading enzyme polygalacturonase and
twitching and swimming motility, which are all virulence factors
(Allen et al. 1997; Kang et al. 2002). PehSR expression in-
creases when the bacterium grows in tobacco leaves, suggesting
it responds to one or more undefined plant signals (Allen et al.
1997). VsrA is the sensor kinase of the two-component regulator
VsrAD, which controls diverse components of R. solanacearum
virulence, including motility, exopolysaccharide production, and
the type three secretion system (Schell et al. 1994) (J. Yao and
C. Allen unpublished results). PilP, which is involved in assem-
bly of type IV pili, is important in R. solanacearum twitching
motility and virulence (Kang et al. 2002; Liu et al. 2001).
Other rex genes apparently contribute to bacterial fitness and
environmental survival. The rex5 insertion identified a locus
that encodes a homolog of Dps, a nonspecific DNA binding
protein widely present in prokaryotes. Dps protects Esche-
richia coli from nutrient and oxidative stress, especially in sta-
tionary phase (Almirón et al. 1992; Martinez 1997). The rex6
insertion identified a gene annotated as a 2-hydroxy-3-oxopro-
prionate reductase involved in glyoxylate and dicarboxylate
metabolism. This enzyme may have been induced by a com-
pound in TRE such as glycerate, which can be shuttled into the
glycolytic pathway (White 2000). Interestingly, the operon im-
mediately upstream encodes proteins involved in sulfur assimi-
lation; both these loci may be important for rhizosphere nutrient
uptake and catabolism.
The rex4 insertion created a fusion with a gene encoding a
4Fe4S ferredoxin that mediates electron transport to O2 as part
of a cbb3-type Cco. Cbb3-Cco belongs to a family of proton-
pumping, heme-copper terminal oxidases with a high affinity
for oxygen, and it is associated with oxygen use in microaero-
bic environments. Cbb3-Cco functions in aerobic respiration; it
is the terminal oxidase in the electron transport chain that runs
from cytochrome c to O2, creating ATP in the process. Cbb3-
cco is an evolutionarily ancient enzyme, well-suited for the
low levels of oxygen in the earlier earth’s atmosphere and is
more closely related to nitric oxide reductase than to other Cco
(Sakamoto and Sone 2004; van der Oost et al. 1994). This
high-affinity Cco may be important in the rhizosphere, where
oxygen levels can be quite low, depending on soil type, mois-
ture levels, competitive consumption by the soil community,
and oxygen consumption or output by the plant roots. For ex-
ample, in the rhizosphere of Juncus effusus L. grown in a
flooded rhizotron, oxygen levels were 10 to 70 μM per liter
around root tips and 0.9 μM per liter in bulk soil (Blossfeld
Fig. 5. The Ralstonia solanacearum race 3 biovar 2 (R3bv2) cytochrome c oxidase (cco) mutant strain does not colonize tomato stems as well as the wild-type
strain. A, Cut petioles of 21- or 22-day-old tomato plants were inoculated with approximately 200 CFU of either strain. At 6 days postinoculation, plant stems
were ground and dilution-plated to determine bacterial population sizes. Colony counts were normalized to CFU per gram of tissue, and data were log-trans-
formed for analysis. Colonization differed significantly when median populations were analyzed using the Mann-Whitney comparison (P = 0.026). Nineteen
plants from four assays were analyzed per strain, and bars represent the standard error of the means. B, Wild-type R3bv2 was generally more effective than the
cco mutant at invading and colonizing tomato plants. Tomato plants (21 or 22 days old) were soil-soak inoculated with a 1:1 mixture of UW551-rif and the
UW551cco mutant in a competition assay. A stem section was dilution-plated onto selective medium when plants showed symptoms. A scatter plot shows the log-
transformed populations of each strain from individual plants. The boxed ‘X’ indicates median population size; bars indicate the 95% confidence interval of the
mean. Mean stem populations of UW551cco were lower than UW551-rif, according to the paired t-test (P = 0.075). C, Ralstonia solanacearum UW551cco did
not adhere to roots as well as wild type. Sterile seedlings of tomato cultivar ‘Bonny Best’ were incubated without aeration for 1.5 or 4 h in 1.0 × 107 CFU/ml sus-
pensions of either UW551 or UW551cco. Rinsed roots were ground and dilution plated. At 1.5 h, the average adhering populations of UW551cco (white bars)
were lower than those of UW551 (dark bars) (P = 0.061). After 4 h of incubation, populations of UW551 and UW551cco were not significantly different (P =
0.158), but wild-type cell densities were still double those of the cco mutant. Bars represent the standard error of the mean.

Page 6

Vol. 23, No. 8, 2010 / 1047
and Gansert 2007). Oxygen in xylem is also reported to be
low, around 2 ppm (Pegg 1985). R. solanacearum lives in both
environments and so must adapt to niches in which oxygen
may not be readily available.
Cbb3-Cco is present in several plant-associated soilborne
bacteria including B. japonicum, R. etli, Azorhizobium cauloino-
dans, Azospirillum brasilense, Pseudomonas stutzeri, P. aerugi-
nosa, and photosynthetic Rhodobacter spp. (Comolli and
Donohue 2004; Koch et al 1998; Mandon et al. 1994; Marchal
et al. 1998; Presig et al. 1996; Urbani et al. 2001). The rex4
4Fe4S, annotated ccoG, is in an apparent operon encoding pro-
teins for synthesis and assembly of a subgroup of Cbb3-Cco
found almost exclusively in the proteobacteria. Several proteo-
bacteria that have cbb3-cco can reportedly adapt to microaero-
bic conditions when they are dependent on aerobic respiration
(Oh and Kaplan 2004). R. solanacearum is considered an obli-
gate aerobe (Holt 1994). Consistent with this adaptation hy-
pothesis, expression of the UW551 rex4::gus fusion increased
more than 30-fold when the bacterium grew in 0.5% rather
than 21% oxygen. In other bacteria, this operon is also highly
upregulated under low-oxygen tensions (Comolli and Donohue
2004; Mandon et al. 1994; Preisig et al. 1996).
The R. solanacearum cbb3-cco locus encodes apparent ho-
mologs to other Cbb3-cco structural and assembly proteins,
most notably from B. japonicum. The R. solanacearum CcoG
contains the conserved ferredoxin motif, CcoN contains the
conserved histidine residues required for full function in B. ja-
ponicum, and CcoI contains the conserved ATPase catalytic
domains (Zufferey et al. 1998). In the proteobacteria, ccoGHIS
is always near ccoNOQP, and the Ralstonia species complex is
no exception. Interestingly, a rearrangement of the cbb3-cco
gene clusters is conserved among Ralstonia spp., and it appears
that ccoIS was translocated upstream of ccoNOQP in R. so-
lanacearum.
We hypothesized that R. solanacearum needs a functioning
cbb3-type Cco to sustain respiration in the rhizosphere and in
the xylem, microaerobic environments relevant to bacterial
wilt pathogenesis. To test this hypothesis, we constructed a R.
solanacearum ccoN mutant, UW551cco. CcoN accepts elec-
trons from CcoP and CcoO, which are membrane-bound
mono- and dihemes, respectively (Oh and Kaplan 2004); a
ccoN mutant of B. japonicum was unable to form the oxidase
(Zufferey et al. 1996). The genomic arrangement of the cbb3-
cco region suggested that ccoN, ccoO, and ccoP might be co-
transcribed, such that disrupting ccoN would have polar effects
on ccoO and ccoP. However, this does not appear to be the
case, since the DNA fragment that successfully complemented
UW551cco to wild-type behavior contained only a complete
copy of the ccoN ORF, its immediate upstream region, and
part of the ccoO ORF. The UW551cco mutant’s phenotype
offered biological evidence supporting the annotation of the
cbb3-cco locus. At the atmospheric oxygen concentration of
21%, UW551cco grew as well as its parent strain, but its
growth was significantly impaired at the microaerobic concen-
tration of 0.5% O2. The decrease in growth was even more pro-
nounced after 5 days of incubation, indicating that this oxidase
is important for R. solanacearum R3bv2 respiration when oxy-
gen tensions are low.
Collectively, our tomato experiments indicate that R. solana-
cearum uses this high-affinity oxidase to effectively colonize
host plants and cause disease. The cco mutant caused delayed
wilt disease on tomato in both a naturalistic soil-soak assay
and after direct inoculation into the vasculature, indicating that
R. solanacearum encounters low-oxygen environments during
growth in host xylem tissue and possibly also during the root
colonization and infection process. However, all inoculated
plants did eventually wilt. R. solanacearum may have an alter-
native process for respiration when oxygen is limiting. Alter-
natively, the mutant’s delayed wilting phenotype could be due
to slower multiplication in the low-oxygen environment of the
tomato xylem, similar to its poor growth under low-oxygen
conditions in culture. This effect may be additive, with re-
duced fitness during the first stages of infection in the fluctuat-
ing oxygen conditions of the rhizosphere. Consistent with this
hypothesis, cco mutant population sizes in tomato stems were
smaller than those of the wild type during the period when dis-
ease progress was delayed. By 8 days after inoculation, popu-
lations of the two strains inside tomato stems were similar,
although tomato plants inoculated with UW551cco still had
markedly lower disease severity than those inoculated with
wild type. We speculate that, at this point, alternative mecha-
nisms for dealing with reduced oxygen may compensate for
lack of a functional Cbb3-Cco or that UW551cco simply takes
longer to multiply but, given time, can eventually attain wild-
type levels. Under natural field conditions the cco mutant
would probably suffer markedly reduced fitness. The ability of
the cco mutant to invade and colonize host roots would depend
on the oxygen tension of the surrounding soil and rhizosphere.
The rhizosphere is characterized by high rates of oxygen de-
pletion that depend on water volume, soil temperature, and
respiring biomass (roots and microbes) (Drew and Stolzy
1996; Hawkes et al 2007).
To better understand the role of Cbb3-Cco in R. solanacea-
rum rhizosphere fitness, we coinoculated plants with the wild
type and cco mutant in a competition assay. The wild-type
strain generally outcompeted the cco mutant strain at the onset
of wilting symptoms, although stem populations varied consid-
erably. This could be due to the fluctuating oxygen levels in
the rhizosphere, where increased oxygen levels could restore
wild-type respiration, or possibly to compensation by other
oxidases in R. solanacearum that can transport electrons under
reduced oxygen availability.
R. solanacearum has other oxidases useful in respiration, in-
cluding the aa3-type, which has a relatively low affinity for
oxygen (Sakamoto and Sone 2004). In R. sphaeroides, biosyn-
thesis of Cbb3-Cco increases under microaerobic conditions
while aa3 oxidase expression is reduced, so that Cbb3-Cco be-
comes the predominant Cco (Oh and Kaplan 2004). R. solana-
cearum UW551 does, however, possess a cytochrome bd-type
of oxidase, which is believed to be important in aerobic respi-
ration (Richardson 2000) and functions with Cbb3-Cco to sup-
port hypoxic growth of A. caulinodans (Kaminski et al. 1996).
Such a terminal oxidase could explain the ability of the patho-
gen to overcome a low-oxygen environment during coloniza-
tion and wilting. Some R. solanacearum phylotype II strains,
such as R3bv2 UW551, cannot use nitrate as an alternative
electron acceptor because they lack NosZ (C. Allen and E. T.
Gonzalez, unpublished data). R. solanacearum UW551 did
not grow under strictly anaerobic conditions (data not shown),
suggesting the importance of the high oxygen-affinity oxidases
in its aerobic respiration. R. solanacearum is a microaerophilic
organism that tends to move toward moderate rather than high
oxygen levels (Yao and Allen 2007). This behavior may reflect
adaptation to the microaerobic environments of the plant
rhizosphere and xylem (Yao and Allen 2007). Pathogen attach-
ment is generally required for biofilm formation on host roots,
but the role of biofilms in bacterial wilt is uncertain. We specu-
late that when this pathogen adheres to plant roots and forms
aggregates, it is forced to respire while embedded in a low-
oxygen tension matrix. UW551 did not form biofilms in an
artificial PVC well assay when tested at temperatures ranging
from 4 to 28°C over a period of 10 days (J. Colburn-Clifford
and C. Allen, unpublished results), although R. solanacearum
strain K60 does form aggregates on tomato roots (Yao 2006).

Page 7

1048 / Molecular Plant-Microbe Interactions
A CcoO mutant of P. aeruginosa was reduced in late-stage
biofilm formation, suggesting that Cbb3-Cco could also be im-
portant in R. solanacearum aggregation on roots (Southey-
Pillig et al. 2005). Significantly more wild-type than cco mu-
tant bacteria adhered to tomato roots following incubation in
R. solanacearum suspensions. The quantifiable differences in
population densities along the root could explain some of the
reduced virulence of UW551cco. In situ gene expression analy-
ses could determine if Cbb3-Cco supports aerobic respiration
when cells are aggregated on host roots.
The rex screen identified pehR and pilP, twitching motility
genes that potentially play a role in biofilm formation. The cco
mutant also had reduced twitching motility (data not shown),
but we suspect this was a consequence of the mutant’s reduced
fitness in the low-oxygen environment of our twitching assay.
Additional evidence supports the importance of this high-
affinity Cco in R. solanacearum. The ccoNOQP genes are con-
sidered ‘core’ genes (Guidot et al. 2007), meaning that these
are found in the genomes of all or nearly all R. solanacearum
strains studied to date. Our previous IVET screen found the
rex4 ferredoxin (ipx38) was up-regulated during mid-phase
disease in R. solanacearum phylotype II strain K60 (Brown
and Allen 2004). Moreover, recent microarray analysis of
UW551 showed that ccoG was up-regulated 10.3-fold in
planta (P = 0.019). Similarly, the log2 (absolute) expression
levels of ccoNOQP were moderately high in planta, 11.5, 9.7,
9.4, and 11.8, respectively, during active bacterial wilt disease
(P = 0.024, 0.022, 0.013, and 0.022, respectively) (F. Meng, J.
Jacobs, L. Babujee, and C. Allen unpublished results).
This study identified R. solanacearum R3bv2 genes involved
in rhizosphere fitness and bacterial wilt virulence. While not at
all exhaustive, this gene set provides insight into this pathogen’s
behavior in the rhizosphere of its tomato host. The screen,
which used TRE in a soft agar medium rather than the natural
tomato rhizosphere environment, certainly affected the traits
identified. A complementary although technically daunting
approach would be to use direct transcriptomic analysis to find
R. solanacearum genes that are expressed in the rhizosphere
before infection but not during pathogenesis in the stem.
Our results also suggest a potential approach to bacterial
wilt disease management. Can we identify one or more com-
ponents of root exudate that induce pathogen genes critical for
wilt disease initiation? Such relationships exist in other soil-
borne pathogens. For example, sorgolactone and sesquiterpene
derivatives exuded from sorghum roots induce germination of
seeds of the root parasites belonging to genus Striga (Ejeta
2007). Sorghum genotypes that produced very low levels of
these stimulating chemicals were resistant to colonization
(Ejeta 2007). Can we similarly exploit the tomato–R. solana-
cearum rhizosphere interaction? Potential approaches include
inactivating the putative plant signals through breeding for
host lines that do not produce signal, the addition of com-
pounds that inactivate plant signals, or selecting for signals
that inhibit attachment to roots (Whipps 2001). Breeding has
also been used to improve the effectiveness of biocontrol agent
colonization in the rhizosphere (Smith et al. 1997). These
points should be considered as we strive to translate our find-
ings into application.
MATERIALS AND METHODS
Bacterial strains, media, and growth conditions.
All R. solanacearum strains used in this study were derived
from the wild-type R3bv2 strain UW551 (Williamson et al.
2002) (Table 2). R. solanacearum strains were cultured at 28°C
in casamino peptone glucose (CPG) broth or CPG solid medium
Table 2. Strains and plasmids used in this study
Strain or plasmid Description Source
Escherichia coli
DH5α
HB101
Ralstonia solanacearum
UW551
UW551cheW
PPilA-forward
PPilA-reverse
PflhDC-forward
PfldDC-reverse
PcheW-forward
PcheW-reverse
UW551cheW rex4
UW551cco
UW551cco(pUFcco) UW551cco complemented with pUFJ10cco, Gmr, Kmr
UW551-rif UW551 spontaneous mutant, Rifr
UW551-GFP UW551 constitutively expressing GFP, Tcr
UW551cco-GFP UW551-2 constitutively expressing GFP, Tcr
Plasmids

pSTBlue-1 Cloning vector, Apr, Kmr
pLAFR3 Broad-host range cosmid vector, Tcr
pVO155 Cloning vector, Kmr, oriT, uidA
pRK600 pRK2013 ntp::Tn9, Cmr, Nms
pJCcheW pVO155 with promoterless UW551 cheW
pJYdcheWGm K60 cheW::aacC1 in pSTJYoT, Apr, Gmr, Kmr
pLJYcheW2.0 pLAFR3 containing cheW with promoter, Tcr
pUCGm Plasmid with aacC1, Gmr
pSTcco PCR-amplified 1.6-kb fragment of UW551 RRSL_01401 in pSTBlue-1, Apr, Kmr
pSTcco::Gm 0.85-kb Gmr cassette in NcoI site of RRSL_01401in pSTcco, Apr, Kmr, Gmr
pUFJ10 Cosmid vector, Gmr, Kmr
pUFJ10::cco PCR-amplified 2.4-kb fragment of UW551 RRSL_01401 with promoter in pUFJ10, Gmr, Kmr This study

F– endA1 relA Φ80 lacZΔM15 hsdR17 supE44 thi-1 recA1 gyrA96
F– thi-I hsd20 supE44 recA13 ara-14 leuB6 proA2 lacY1 galK2 psdL20 xyl-5 mtl-1

Invitrogen, Carlsbad, CA, U.S.A.
Boyer and Rouillan-Dussoix 1969

Wild-type race 3, biovar 2, isolated from geranium
UW551cheW::aacC1, Gmr
UW551cheW; pilA promoter (forward orientation)::cheW; Kmr
UW551cheW; pilA promoter (reverse orientation)::cheW; Kmr
UW551cheW; flhDC promoter (forward orientation)::cheW
UW551cheW; flhDC promoter (reverse orientation)::cheW
UW551cheW; cheW promoter (forward orientation)::cheW
UW551cheW; cheW promoter (reverse orientation)::cheW
UW551cheW::aacC1, Gmr rex4(4Fe4S::gus), Gmr, Kmr
UW551cco::aacC1, Gmr
Williamson et al. 2002
This study
This study
This study
This study
This study
This study
This study
This study
This study
This study
Swanson et al. 2005
This study
This study

Novagen, Madison, WI, U.S.A.
Staskawicz et al. 1986
Oke and Long 1999
Finan et al. 1986
This study
Yao 2006
Yao 2006
Schweizer 1993
This study
This study
Gabriel et al. 2006
a Ap
respectively, and Nms = neomycin sensitivity. GFP = green fluorescent protein, PCR = polymerase chain reaction.
r, Cmr, Gmr, Kmr, Rifr, Tcr, and Cmr = ampicillin, chloramphenicol, gentamicin, kanamycin, rifampicin, tetracycline, and chloramphenicol resistance,

Page 8

Vol. 23, No. 8, 2010 / 1049
(Hendrick and Sequeira 1984; Kelman 1954). Boucher’s mini-
mal medium (BMM) (Boucher et al. 1985) was used when
minimal medium was required. Taxis agar (TA) consisted of
BMM buffered with 20 mM morpholineethanesulfonic acid to
pH 6.5, supplemented with 4 mM aspartic acid as the carbon
source. TA was semisolidified with 0.2% noble agar and TRE
were added at 1% concentration, when required.
E. coli strains were grown at 37°C, using Luria-Bertani
(Ausubel et al. 1995) solid or liquid media supplemented with
an appropriate antibiotic. Antibiotics were added to media for
final concentrations (milligrams per liter) as follows: kanamy-
cin (Km), 25; Gm, 5; chloramphenicol, 15; naladixic acid, 4;
tetracycline, 25; and rifampicin, 25. Unless otherwise noted,
media components were obtained from Difco Laboratories
(Detroit) and chemicals from Sigma-Aldrich (St. Louis).
DNA manipulation and sequence analysis.
Standard methods were used to isolate chromosomal and
plasmid DNA and to perform cloning and PCR (Ausubel et al.
1995). E. coli and R. solanacearum were transformed by elec-
troporation as previously described (Allen et al. 1991). Tripar-
ental mating was used to introduce plasmid constructs with
donor plasmid pJCcheW and helper plasmid pRK600 in E. coli
HB101 (Finan et al. 1986). Oligonucleotide synthesis was per-
formed by Integrated DNA Technologies (Coralville, IA,
U.S.A.). Sequencing was performed at the University of Wis-
consin-Madison Biotechnology Center (Madison, WI, U.S.A.),
using automated fluorescent sequencing. Biology Workbench,
Softberry, and NEBcutter databases, the UW551 genome data-
base, and National Center for Biotechnology Information and
the Integrated Microbial Genomics databases were used to
view and analyze DNA sequences.
Construction of UW551Δ ΔcheW.
A UW551cheW deletion mutant was constructed by double
recombination, using plasmid pJYdcheWGm (Yao and Allen
2006) in which the entire cheW gene was replaced in the
UW551 genome by the Gm-resistance cassette (aacC1).
Proper deletion of cheW and replacement by aacC1 was con-
firmed by PCR and Southern blot analysis. Wild-type swim-
ming motility of the mutant was confirmed by microscopy, and
lack of taxis was confirmed on taxis agar. To complement this
mutant, pLJYcheW2.0 (Yao and Allen 2006) was electropo-
rated into the cheW mutant strain. Restoration of chemotaxis
was confirmed on taxis agar.
Construction of pJCcheW.
The promoter-trapping vector pJCcheW was derived from
pVO155 (Oke and Long 1999). pJCcheW carries the promoter-
less cheW ORF sequence cloned directly upstream of the pro-
moterless uidA (gus) reporter gene as well as the Km- and
ampicillin-resistance genes used for selection. The promoter-
less cheW gene was PCR-amplified from R. solanacearum
UW551 (cheW forward and reverse primers, respectively, 5′-
GCTCTAGATGATTGATTGAGGGAGCA-3′ and 5′-GCTCT
AGATTGTGTCCGTCGTTGTTC-3′). Both primers contained
XbaI restriction sites to aid in cloning. The sense primer con-
tained stop codons in all three reading frames and a putative
ribosome binding site to generate a functional transcriptional
fusion. To ensure that pJCcheW functioned correctly, we con-
structed three control plasmids containing known promoters.
Plasmids carrying the native cheW promoter in both the for-
ward (pPcheW-forward) and reverse (pPcheW-reverse) orien-
tation with respect to cheW were created by inserting a PCR-
amplified, 517-bp DNA fragment (forward primer 5′-TCAGC
TCATCAATCTGGTCG-3′ and reverse primer 5′-AGGTAATG
TCGGAAAACGGA-3′) containing the putative cheW pro-
moter region directly upstream of the promoterless cheW ORF.
Additional control plasmids carried the highly expressed pilin
structural gene promoter PpilA as well as the promoter of the
flagellar regulator FlhDC, PflhDC (Brown and Allen 2004;
Kang et al. 2002). Promoter fragments were obtained by re-
striction digestion, were treated with Klenow (NEB), and were
ligated into pJCcheW at the XhoI site. Constructs were ana-
lyzed by restriction digestion to confirm the orientation of the
promoter regions. All constructs were tested for the ability to
restore UW551ΔcheW chemotaxis on taxis agar, and gene ex-
pression was determined quantitatively by a GUS activity assay
(Jefferson 1986).
Construction of the promoter-trapping library.
R. solanacearum UW551 genomic DNA libraries containing
potential promoter sequences were created as described (Oke
and Long 1999) and were ligated into pJCcheW. Two separate
libraries were constructed with an average insert size of 800
bp, as assessed by gel analysis. Over 65,000 colonies from the
libraries were introduced into UW551ΔcheW in eight separate
triparental matings. Integration of the plasmid into the cheW
mutant strain occurs through single recombination at the site
of homology between the chromosome and the cloned DNA
fragment; the plasmid could not recombine at the cheW locus
because it was deleted from UW551ΔcheW. Strains with inte-
grated pJCcheW were selected on CPG agar amended with
Km and were used in subsequent rex screening.
Preparation of TRE.
Seeds of the wilt-susceptible tomato cultivar ‘Bonny Best’
were surface-sanitized as previously described, with some
modification. Briefly, seeds were agitated for 30 min in sterile
water, 5 min in 95% ethanol, 5 min with sterile water, 5 min in
20% bleach solution, and then, rinsed five times for 5 min with
sterile water. Seeds were incubated overnight at 4°C for ver-
nalization and were then germinated either on 1% water agar
or directly on sterile cellulose filters at 28°C for 48 to 72 h,
when rootlets and stems began to emerge. For each extraction,
20 to 30 seedlings were transferred to a sterile 1L filter appara-
tus (0.2 µm, Corning Glass, Corning, NY, U.S.A.) and were
allowed to grow an additional 13 to 16 days, or until secondary
roots began to form. Seedlings were maintained on one-tenth
strength filter-sterilized Hoagland’s solution. For the TRE col-
lection, 5 to 10 ml of sterile Hoagland’s solution was added to
the filter and allowed to sit overnight, and the solution was col-
lected by vacuum filtration and stored at –80°C in 1-ml aliquots.
Root exudates were tested for sterility and the ability to support
UW551 growth on TA and were characterized for protein con-
tent, using a bicinchoninic acid assay (Pierce, Rockford, IL,
U.S.A.), and for reducing sugar content, using a modified Prado
assay (Prado et al. 1998).
Screening the library: the rex screen.
Strains carrying the integrated pJCcheW were pooled and
plated onto CPG agar amended with Km and Gm. Cells were
washed, were collected by centrifugation, and were adjusted to
an optical density at 600 nm (OD600) of 0.05 (approximately 5
× 107 CFU/ml). Bacterial suspension (2 μl, 1 × 105 CFU) were
placed into the center of a taxis agar plate supplemented with 4
mM aspartic acid and 1% (vol/vol) TRE and were incubated at
28°C until measurable taxis halos formed. For each library, the
number of inoculated plates ranged from 30 to 45. Any strain
that carried a DNA fragment capable of restoring chemotaxis
to the nontactic cheW mutant formed a halo by moving up the
nutrient gradient created by bacterial growth on the taxis me-
dia. Cells that had migrated (che+) were recovered by cutting
the halo out of the taxis plate. Resulting colonies were

Page 9

1050 / Molecular Plant-Microbe Interactions
screened simultaneously on TA and TA with 1% TRE, with the
wild-type strain as the control. Those strains that formed halos
as large or larger than the wild-type control on TA plus TRE
were saved for further analysis.
Analysis of the rex library.
To characterize active rex promoters, we transferred the inte-
grated pJCcheW+insert from UW551ΔcheW using retrotrans-
fer (Rainey et al. 1997), a conjugation procedure that recreates
the original rex fusion in E. coli DH5α. The inserted DNA
fragment in purified pJCcheW+insert was sequenced from the
5′ end of the transcriptional fusion in the recovered rex fusions,
using primer 5′-GTTCCTTGACTTCCATTGC-3′. These se-
quences were compared with the UW551 genome database to
determine the likely promoter driving cheW expression. The
expression levels of rex fusion strains were determined by
measuring their relative GUS activity following growth in
minimal media with and without 1% TRE.
Quantitative GUS activity assays.
GUS activity was measured as described (Jefferson 1986).
To validate promoter activities in pJCcheW and to measure
expression of rex fusions, cells were collected from overnight
culture and were resuspended and extracted for GUS activity.
Total cell number was determined by dilution plating an ali-
quot. For samples grown under low-oxygen conditions, 1 ml of
suspension was transferred immediately into GUS extraction
buffer and an additional 100 µl was removed to determine cell
numbers. GUS activity was thus determined as nanomole 4-
methylumbelliferone released per minute per cell.
Selection and mutagenesis of a cbb3-type cco in UW551.
Mutagenesis of cbb3-cco. We mutated RRSL_01401, the
first ORF in an operon encoding for structural proteins of a
cbb3-type Cco (cbb3-cco), using marker-assisted gene replace-
ment. Briefly, an approximately 1.8-kb region of RRSL_
01401 was amplified using primers 5′-CAGACCTTCAATTA
CCGCGT-3′ and 5′-CTTGCAGATAGGCCACCAC-3′. The
product was A-T cloned into pSTBlue-1 (Kmr) (Novagen, Inc.,
Madison, WI, U.S.A.) to create pSTcco. pUCGm (Schweizer
1993) was digested with BamH1 to liberate the aacC1 Gm-
resistance gene cassette. The insert of pSTcco was digested
with NcoI, and aacC1 was blunt-end cloned into the site using
Klenow, creating pSTcco::Gm. pSTcco::Gm was introduced
into UW551 by electroporation, and transformants were selected
for Gm resistance and Km sensitivity. The mutation in the re-
sulting strain, UW551cco, was confirmed by PCR and Southern
blot analysis.
Complementation of UW551cco. A 2.4-kb region of
RRSL_01401 including the likely upstream promoter sequence
was amplified using primers 5′-CTTTGTCGCTGATGGCAG
TG-3′ and 5′-GAGTAGTGGCCGTAGCGTTC-3′ and were A-
T cloned into pSTBlue-1 (Kmr) (Novagen) to create pSTcco-
comp. This insert was liberated by EcoRI digestion and cloned
into EcoR1-digested pUFJ10 (Gmr, Kmr) (Gabriel et al. 2006)
to create the complementation plasmid pUFcco (Kmr). pUFcco
was electroporated into competent UW551cco to create the
complemented strain UW551cco(pUFcco).
Virulence assays.
We tested the virulence of the cbb3-cco mutant on ‘Bonny
Best’ tomato plants, using either a naturalistic soil-soak or by
direct inoculation through a cut petiole as described (Tans-
Kersten et al. 1998). Briefly, for soil-soak, a bacterial suspen-
sion was poured directly into individual pots containing 21-
day-old unwounded tomato plants to a final cell density of 5 ×
107 to 1 × 108 CFU/g soil. For petiole inoculations, approxi-
mately 200 cells were applied to the cut petiole of the first true
leaf of 21- to 22-day-old tomato plants. Bacterial wilt progress
was measured as previously described (Yao and Allen 2007).
All plant assays had 16 plants per treatment and were performed
three times.
Population studies.
Plant colonization. To quantify tomato-stem colonization by
wild-type and mutant strains, we measured bacterial population
sizes in plants following petiole inoculation, as described above.
At 4, 6, and 8 days after inoculation, approximately 1 cm of
tomato stem was harvested from above and just below the
inoculation point. The plant tissue was ground in 1 ml of sterile
water and the dilution was plated onto CPG and was amended
with antibiotic when necessary, and was then incubated for 48
h at 28°C. Colonies were counted to determine number of cells
per gram of stem tissue. Each experiment contained at least
three plants per treatment and was repeated four times, for a
total of 19 to 20 plants per strain.
Competition study. Plants (21 to 22 days old) were soil-soak
inoculated with a 1:1 mixture of UW551-rif and the cco mu-
tant strain. When each plant developed wilting symptoms, a 1-
cm portion of the stem was cut at approximately 1 cm above
the soil line to avoid soil contamination. The stem tissue was
ground and a dilution was plated as described above, to deter-
mine the number of colony-forming units per gram of stem tis-
sue of each strain. Each experiment contained at least five
plants per treatment and was repeated four times on a total of
27 plants.
Growth and survival under low-oxygen conditions.
To measure bacterial growth in defined oxygen levels, initial
inoculum was prepared as described above. Each strain was
adjusted to an OD600 = 0.01 (1 × 107 CFU/ml) in 125-ml side-
arm flasks and was then subjected to vacuum for 5 min with
swirling every 30 s, to release and remove oxygen from the
medium. The suspension was immediately dispensed in 25-ml
volumes into 50 ml of vented-cap tissue-culture flasks (Falcon;
Franklin Lakes, NJ, U.S.A.), and a flask containing each strain
was placed in plastic storage bags designed for vacuum (Rey-
nolds, Richmond, VA, U.S.A.). Air was immediately removed
from the bags by applying vacuum and then replaced with a
mixture of 21, 5, or 0.5% oxygen gas with nitrogen gas bal-
ance (Airgas North Central, Carol Stream, IL, U.S.A.). Vacuum
purging and oxygen replacement was repeated twice more for
each treatment to fully purge the bags. Suspensions were then
incubated at 22 to 24°C with gentle rotation at 100 rpm to
facilitate the gas exchange within the bag. Resauzrin strips
(Oxoid, Hants, U.K.) confirmed the expected oxygen levels
inside each bag. After 24 or 120 h, each bottle was sampled
and dilution was plated as described above.
Visualization and colonization
of R. solanacearum strains on tomato roots.
GFP-expressing strains of UW551 and UW551cco were
prepared by natural transformation using genomic DNA from
K60-GFP, a phylotype II sequevar 7 (race 1 biovar 1) strain of
R. solanacearum that expresses GFP (Bertolla et al. 1997; Yao
and Allen 2007). UW551-GFP was previously shown to have
no growth or virulence defect compared with its parent
UW551 (J. Yao and C. Allen unpublished results). Tetracycline-
resistant transformants were visualized under a fluorescent
microscope to confirm the expected green fluorescent pheno-
type.
Inoculation, observation, and quantification of R. solanacea-
rum on tomato roots. ‘Bonny Best’ seeds were surface-steril-
ized and germinated on 1% water agar amended with strepto-

Page 10

Vol. 23, No. 8, 2010 / 1051
mycin to avoid seed-borne contaminants. Germinated seeds
were transferred to sterile Magenta boxes and were maintained
under light on one-tenth strength Hoagland’s solution. Seed-
lings (12 to 16 days old) were incubated without aeration in
suspensions of GFP-labeled strains (1 × 107 CFU ml–1) for 1.5
h, to allow bacteria to adhere to roots but not colonize them.
Roots were then excised, gently rinsed and blotted on tissue,
measured, ground in sterile water, and dilution-plated to deter-
mine the populations of adhering cells.
To observe root colonization progress, roots incubated with
R. solanacearum for 4 h were excised, rinsed, gently blotted
on tissue, and visualized under a fluorescent microscope as de-
scribed (Yao and Allen 2006). To compare R. solanacearum
population sizes on these root surfaces, a matched set of seed-
lings from the 4-h timepoint were treated similarly and meas-
ured and were then ground in 1 ml of sterile water and dilu-
tion-plated.
Statistical analysis.
Strain growth was analyzed by mixed model or general lin-
ear model analysis of variance with Tukey’s significant differ-
ence test. Virulence assay data were analyzed by repeated
measures analysis of variance (Tans-Kersten et al. 1998). Log
transformation was used to analyze tomato stem populations.
The Mann-Whitney median test was used to further analyze
population data from single strain inoculations, and the paired
t-test was used to analyze population data from the competi-
tion assay. Data analyses were conducted in JMP Ver. 7.0.2
and Ver. 8 (SAS Institute, Raleigh, NC, U.S.A.) or in Minitab
Release 14.13 (State College, PA, U.S.A.).
ACKNOWLEDGMENTS
This work was supported by the National Research Initiative of the
United States Department of Agriculture-Cooperative State Research Edu-
cation and Extension Service (Plant Biosecurity Project 2006-04560) and
by the University of Wisconsin-Madison College of Agricultural and Life
Sciences and Graduate School. We thank N. Masuda for technical assis-
tance with TRE and expression analysis and Agdia Inc. (Elkhart, IN,
U.S.A.) for donated research materials. We also acknowledge D. Waller
and N. Keuler for help with statistical analysis.
LITERATURE CITED
Allen, C., Huang, Y., and Sequeira, L. 1991. Cloning of genes affecting
polygalacturonase production in Pseudomonas solanacearum. Mol.
Plant-Microbe Interact. 4:147-154.
Allen, C., Gay, J., and Simon-Buela, L. 1997. A regulatory locus, pehSR,
controls polygalacturonase production and other virulence functions in
Ralstonia solanacearum. Mol. Plant-Microbe Interact. 10:1054-1064.
Allen, C., Kelman, A., and French, E. R. 2001. Brown rot of potatoes.
Pages 11-13 in: Compendium of Potato Diseases. W. R. Stevenson, R.
Loria, G. D. Franc, and D. P. Weingartner, eds. American Phytopa-
thological Society Press St. Paul, MN, U.S.A.
Almirón, M., Link, A., Furlong, D., and Kolter, R. 1992. A novel DNA-
binding protein with regulatory and protective roles in starved Esche-
richia coli. Genes Dev. 6:2646-2654.
Ausubel, F., Brent, R., Kingston, R., Moore, D., Seidman, J., Smith, J.,
and Struhl, K. 1995. Short Protocols in Molecular Biology. John Wiley
and Sons, New York.
Bertolla, F., Van Gijsegem, F., Nesme, X., and Simonet, P. 1997.
Conditions for natural transformation of Ralstonia solanacearum. J.
Bacteriol. 180:27-34.
Blossfeld, S., and Gansert, D. 2007. A novel non-invasive optical method
for quantitative visualization of pH dynamics in the rhizosphere of
plants (Abstr.). Plant Cell Environ. 30:176-186.
Boch, J., Joardar, V., Gao, L., Robertson, T. L., Lim, M., and Kunkel, B.
N.2002. Identification of Pseudomonas syringae pv. tomato genes
induced during infection of Arabidopsis thaliana. Mol. Microbiol.
44:73-88.
Boucher, C., Barberis, P., Trigalet, A., and Demery, D. 1985. Transposon
mutagenesis of Pseudomonas solanacearum: isolation of Tn5-induced
avirulent mutants. J. Gen. Microbiol. 131:2449-2457.
Boyer, H. W., and Rouillan-Dussoix, D. 1969. A complementation analysis
of the restriction and modification of DNA in Escherichia coli. J. Mol.
Biol. 53:459-472.
Brencic, A., and Winans, S. C. 2005. Detection of and response to signals
involved in host-microbe interactions by plant-associated bacteria.
Microbiol. Mol. Biol. Rev. 69:155-194.
Brito, B., Marenda, M., Barberis, P., Boucher, C., and Genin, S. 1999.
prhJ and hrpG, two new members of the plant signal-dependent regula-
tory cascade controlled by PrhA in Ralstonia solanacearum. Mol. Mi-
crobiol. 31:237-251.
Brown, D., and Allen, C. 2004. Ralstonia solanacearum genes induced
during growth in tomato: an inside view of bacterial wilt. Mol. Micro-
biol. 53:1641-1660.
Drew, M., and Stolzy, L. 1996. Growth under oxygen stress. Oages 397-
414 in: Plant Roots: The Hidden Half. 2nd ed. A. Eshel, T. Beekman, Y.
Waisal, and U. Kafkafi, eds. Marcel Dekker, New York.
Dunn, A. K., Klimowicz, A. K., and Handelsman, J. 2003. Use of a
promoter trap to identify Bacillus cereus genes regulated by tomato
seed exudate and a rhizosphere resident, Pseudomonas aureofaciens.
Appl. Environ. Microbiol. 69:1197-1205.
Ejeta, G. 2007. Breeding for Striga resistance in sorghum: Exploitation of
an intricate host-parasite biology. Crop Sci. 47:S216-S227.
Fegan, M., and Prior, P. 2005. How complex is the “Ralstonia solanacea-
rum” species complex? Pages 449-461 in: Bacterial Wilt Disease and
the Ralstonia solanacearum species complex. C. Allen, P. Prior and A.
C. Hayward, eds. American Phytopathological Society Press, St Paul,
MN, U.S.A.
Finan, T. M., Kunkel, B. N., DeVos, G. F., and Signer, E. R. 1986. Second
symbiotic megaplasmid of Rhizobium meliloti carrying exopolysaccha-
ride and thiamine synthesis genes. J. Bacteriol. 167:66-72.
Gabriel, D. W., Allen, C., Schell, M., Denny, T. P., Greenberg, J. T., Duan,
Y. P., Flores-Cruz, Z., Huang, Q., Clifford, J. M., Presting, G., González,
E. T., Reddy, J., Elphinstone, J., Swanson, J., Yao, J., Mulholland, V.,
Liu, L., Farmerie, W., Patnaikuni, M., Balogh, B., Norman, D., Alvarez,
A., Castillo, J. A., Jones, J., Saddler, G., Walunas, T., Zhukov, A., and
Mikhailova, N. 2006. Identification of open reading frames unique to a
select agent: Ralstonia solanacearum race 3 biovar 2. Mol. Plant-Microbe
Interact. 19:69-79.
Gal, M., Preston, G. M., Massey, R. C., Spiers, A. J., and Rainey, P. B.
2003. Genes encoding a cellulosc polymer contribute toward the
ecological success of Pseudomonas fluorescens SBW25 on plant
surfaces. Mol. Ecol. 12:3109-3121.
Gonzalez, E. T., and Allen, C. 2003. Characterization of a Ralstonia
solanacearum operon required for polygalacturonate degradation and
uptake of galacturonic acid. Mol. Plant-Microbe Interact. 16:536-544.
Guidot, A., Philippe, P., Schoenfeld, J., Carrère, S., Genine, S., and
Boucher, C. 2007. Genomic structure and phylogeny of the plant patho-
gen Ralstonia solanacearum inferred from gene distribution analysis. J.
Bacteriol. 189:377-387.
Handelsman, J., and Stabb E. V. 1996. Biocontrol of soilborne pathogens.
Plant Cell 8:1855-1869.
Hawkes, C. V., deAngelis, K. M., and Firestone, M. K. 2007. Root
interactions with soil microbial communities and processes. Pages 1-30
in: The Rhizosphere: An Ecological Perspective. Z. G. Cardon and J. L.
Whitbeck, eds. Elsevier Academic Press, London.
Hayward, A. C. 1991. Biology and epidemiology of bacterial wilt caused
by Pseudomonas solanacearum Ann. Rev. Phytopathol. 29:65-87.
Hendrick, C., and Sequeira, L. 1984. Lipopolysaccharide-defective mu-
tants of the wilt pathogen Pseudomonas solanacearum. Appl. Environ.
Microbiol. 48:94-101.
Holt, J. G. 1994. Bergey's Manual of Determinative Bacteriology. 9th ed.
Lippincott Williams & Wilkins, New York.
Huber, D. and Watson, R. 1974. Nitrogen form and plant disease. Annu.
Rev. Phytopathol. 12 :139-165.
Jefferson, R. A., Burgess, S. M., and Hirsh, D. 1986. Beta-glucuronidase
from Escherichia coli as a gene-fusion marker. Proc. Natl. Acad. Sci.
U.S.A. 83:8447-8451.
Kaminski, P. A., Kitts, C. L., Zimmerman, Z., and Ludwig, R. A. 1996.
Azorhizobium caulinodans uses both cytochrome bd (quinol) and cyto-
chrome cbb3 (cytochrome c) terminal oxidases for symbiotic nitrogen
fixation. J. Bacteriol. 178:5989-5994.
Kang, Y., Liu, H., Genin, S., Schell, M. A., and Denny, T. P. 2002. Ralsto-
nia solanacearum requires type 4 pili to adhere to multiple surfaces,
and for natural transformation and virulence. Mol. Microbiol. 46:427-
437.
Kelman, A. 1954. The relationship of pathogenicity of Pseudomonas so-
lanacearum to colony appearance in a tetrazolium medium. Phytopa-
thology 44:693-695.
Koch, H.-G., Hwang, O., and Daldal, F. 1998. Isolation and characteri-
zation of Rhodobacter capsulatus mutants affected in cytochrome cbb3

Page 11

1052 / Molecular Plant-Microbe Interactions
oxidase activity. J. Bacteriol. 180:969-978.
Lambert, C. D. 2002. Agricultural Bioterrorism Protection Act of 2002.
Possession, use, and transfer of biological; agents and toxins; interim
and final rule. (7 CFR Part 331). Fed. Regist. 67:76908-76938.
Liu, H., Kang, Y., Genin, S., Schell, M. A., and Denny, T. P. 2001. Twitch-
ing motility of Ralstonia solanacearum requires a type IV pilus system.
Microbiology 147:3215-3229.
Loesche, W. J. 1969. Oxygen sensitivity of various anaerobes. (Abstr.)
Appl. Environ. Microbiol. 18:723-727.
Lugtenberg, B. J. J. , Kravchenko, L.V., and Simons, M. 1999. Tomato
seed and root exudate sugars: composition, utilization by Pseudomonas
biocontrol strains and role in rhizosphere colonization. Environ.
Microbiol. 1:439-446.
Mandon, K., Kaminski, P. A., and Elmerich, C. 1994. Functional analysis
of the fixNOQP region of Azorhizobium caulinodans. J. Bacteriol.
176:2560-2568.
Marchal, K., Sun, J., Keijers, V., Haaker, H., and Vanderleyden, J. 1998. A
cytochrome cbb3 (cytochrome c) terminal oxidase in Azospirillum
brasilense Sp7 supports microaerobic growth. J. Bacteriol. 180:5689-
5696.
Marco, M. L., Legac, J., and Lindow, S. E. 2005. Pseudomonas syringae
genes induced during colonization of leaf surfaces. Environ. Microbiol.
7:1379-1391.
Marschner, H. 1995. Mineral nutrition of higher plants. Academic Press,
London.
Martinez, A.a.K., R. 1997. Protection of DNA during oxidative stress by
the nonspecific DNA-binding protein Dps. J. Bacteriol. 179:5188-5194.
Neidle, E. L., and Kaplan, S. 1992. Rhodobacter sphaeroides rdxA, a ho-
molog of Rhizobium meliloti fixG, encodes a membrane protein which
may bind cytoplasmic [4Fe-4S] clusters. J. Bacteriol. 174:6444-6454.
Oh, J.-I., and Kaplan, S. 2004. Oxidases as redox sensors in pigment syn-
thesis. Pages 287-319 in: Respiration in Archea and Bacteria Vol. I. D.
Zannoni, ed. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Oke, V., and Long, S. R. 1999. Bacterial genes induced within the nodule
during the Rhizobium-legume symbiosis. Mol. Microbiol. 33:837-849.
Osbourne, A. E., Barber, C. E., and Daniels, M. J. 1987. Identification of
plant-induced genes of the bacterial pathogen Xanthomonas campestris
pathovar campestris using a promoter-probe plasmid. EMBO (Eur. Mol.
Biol. Organ.) J. 6:23-28.
Pegg, G. F. 1985. Life in a black hole: The microenvironment of the vas-
cular pathogen. Trans. Brit. Mycol. Soc. 85:1-20.
Prado, F. E., González, J. A., Boero, C., and Sampietro, A. R. 1998. A sim-
ple and sensitive method for determining reducing sugars in plant tis-
sues. Application to quantify the sugar content in quinoa (Chenopodium
quinoa Willd.) seedlings. Phytochem. Anal. 9:58-63.
Preisig, O., Zufferey, R., and Hennecke, H. 1996. The Bradyrhizobium ja-
ponicum fixGHIS genes are required for the formation of the high-affin-
ity cbb3-type cytochrome c oxidase. Arch. Microbiol. 165:297-305.
Rainey, P. B. 1999. Adaptation of Pseudomonas fluorescens to the plant
rhizosphere. Environ. Microbiol. 1:243-257.
Rainey, P. B., and Preston, G. M. 2000. In vivo expression technology
strategies: Valuable tools for biotechnology. Curr. Opin. Biotechnol.
11:440-444.
Rainey, P. B., Heithoff, D. M., and Mahan, M. J. 1997. Single-step conju-
gative cloning of bacteria gene fusions involved in microbe-host inter-
actions. Mol. Gen. Genet. 256:84-87.
Richardson, D. J. 2000. Bacterial respiration: A flexible process for a
changing environment. Microbiol. 146:551-571.
Saile, E., McGarvey, J. A., Schell, M. A., and Denny, T. P. 1997. Role of
extracellular polysaccharide and endoglucanase in root invasion and
colonization of tomato plants
Phytopathology 87:1264-1271.
Sakamoto, J., and Sone, N. 2004. Biochemical and molecular features of
terminal oxidases. Pages 87-113 in: Respiration in Archaea and Bacte-
ria Vol. I, D. Zannoni, ed. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Schell, M. A. 2000. Control of virulence and pathogenicity genes of
Ralstonia solanacearum by an elaborate sensory network. Annu. Rev.
Phytopathol. 38:263-292.
Schell, M. A., Denny, T. P., and Huang, J. 1994. vsrA, a second two-com-
ponent sensor regulating virulence genes of Pseudomonas solanacea-

by
Ralstonia solanacearum.
rum. Mol. Microbiol. 11:489-500.
Schweizer, H. P. 1993. Small broad-host-range gentamycin resistance gene
cassettes for site-specific insertion and deletion mutagenesis. BioTech-
niques 15:831-833.
Smith, K. P., Handelsman, J., and Goodman, R. M. 1997. Modeling dose-
response relationships in biological control: Partitioning host responses
to the pathogen and biocontrol agent. Phytopathology 87:720-729.
Southey-Pillig, C. J., Davies, D. G., and Sauer, K. 2005. Characterization
of temporal protein production in Pseudmonas aerugnosa biofilms. J.
Bacteriol. 187:8114-8126.
Staskawicz, B., Dahlbeck, D., Keen, N., and Napoli, C. 1986. Molecular
characterization of cloned avirulence genes from race 0 and race 1 of
Pseudomonas syringae pv. glycinea. J. Bacteriol. 169:5789-5794.
Swanson, J. K., Yao, J., Tans-Kersten, J. K., and Allen, C. 2005. Behavior
of Ralstonia solanacearum race 3 biovar 2 during latent and active
infection of geranium. Phytopathology 95:136-143.
Tans-Kersten, J., Guan, Y., and Allen, C. 1998. Ralstonia solanacearum
pectinmethylesterase is required for growth on methylated pectin but
not for bacterial wilt virulence. Appl. Environ. Microbiol. 64:4918-
4923.
Tans-Kersten, J. K., Brown, D., and Allen, C. 2004. Swimming motility, a
virulence factor of Ralstonia solanacearum, is regulated by FlhDC and
by the plant host environment. Mol. Plant-Microbe Interact. 17:686-
695.
Urbani, A., Gemeinhardt, S., Warne, A., and Saraste, M. 2001. Proprties of
the detergent solubilized cytochrome c oxidase (cytochrome cbb3.
purified from Pseudomonas stutzeri. FEBS (Fed. Eur. Biochem. Soc.)
Lett. 508:29-35.
van der Oost, J., de Boer, A. P. N., deGier J.-W. L. Zumft, W. G.,
Stouthamer, A. H., and van Spanning, R. J. M. 1994. The heme-copper
oxidase family consists of three distinct types of terminal oxidases and
is related to nitric oxide reductase. FEMS (Fed. Eur. Microbiol. Soc.)
Microbiol. Lett. 121:1-10.
Vasse, J., Genin, S., Frey, P., Boucher, C., and Brito, B. 2000. The hrpB
and hrpG regulatory genes of Ralstonia solanacearum are required for
different stages of the tomato root infection process. Mol. Plant-
Microbe Interact. 13:259-267.
Whipps, J.M. 2001. Microbial interactions and biocontrol in the
rhizosphere. J. Expt. Bot. 52:487-511.
White, D. 2000. The physiology and biochemistry of prokaryotes. Oxford
University Press, Inc., New York.
Williamson, L., Nakaho, K., Hudelson, B., and Allen, C. 2002. Ralstonia
solanacearum race 3, biovar 2 strains isolated from geranium are patho-
genic on potato. Plant Dis. 86:987-991.
Yao, J., and Allen, C. 2006. Chemotaxis is required for virulence and com-
petitive fitness in the bacterial wilt pathogen Ralstonia solanacearum. J.
Bacteriol. 188:3697-3708.
Yao, J., and Allen, C. 2007. The plant pathogen Ralstonia solanacearum
needs aerotaxis for normal biofilm formation and interactions with its
tomato host. J. Bacteriol. 189:6415-6424.
Zufferey, R., Preisig, O., Hennecke, H., and Thöny-Meyers, L. 1996. As-
sembly and function of the cytochrome cbb3 oxidase subunits in Bra-
dyrhizobium japonicum. J. Biol. Chem. 271:9114-9119.
Zufferey, R., Arslan, E., Thöny-Meyer, L., and Hennecke, H. 1998. How
replacements of the 12 conserved histidines of subunit I affect assem-
bly, cofactor binding, and enzymatic activity of the Bradyrhizobium ja-
ponicum cbb3-type oxidase. J. Biol. Chem. 273:6452-6459.
AUTHOR-RECOMMENDED INTERNET RESOURCES
Integrated Microbial Genomes database:
img.jgi.doe.gov/cgi-bin/pub/main.cgi
myCAP(my Collaborative Annotation Portal UW551 genome database:
vision.biotech.ufl.edu/mycap
National Center for Biotechnology Information website:
www.ncbi.nlm.nih.gov
New England Biolabs’s NEBcutter tools.neb.com/NEBcutter2/index.php
San Diego Supercomputer Center’s Biology Workbench:
workbench.sdsc.edu
Softberry website: linux1.softberry.com/berry.phtml