JOURNAL OF BACTERIOLOGY, Apr. 2008, p. 2306–2313
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 190, No. 7
Comparative Genomics and an Insect Model Rapidly Identify Novel
Virulence Genes of Burkholderia mallei?†
Mark A. Schell,1* Lyla Lipscomb,1and David DeShazer2
Department of Microbiology, University of Georgia, Athens, Georgia 30602,1and Bacteriology Division,
United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick,
Received 30 October 2007/Accepted 8 January 2008
Burkholderia pseudomallei and its host-adapted deletion clone Burkholderia mallei cause the potentially fatal
human diseases melioidosis and glanders, respectively. The antibiotic resistance profile and ability to infect via
aerosol of these organisms and the absence of protective vaccines have led to their classification as major
biothreats and select agents. Although documented infections by these bacteria date back over 100 years,
relatively little is known about their virulence and pathogenicity mechanisms. We used in silico genomic
subtraction to generate their virulome, a set of 650 putative virulence-related genes shared by B. pseudomallei
and B. mallei but not present in five closely related nonpathogenic Burkholderia species. Although most of these
genes are clustered in putative operons, the number of targets for mutant construction and verification of
reduced virulence in animal models is formidable. Therefore, Galleria mellonella (wax moth) larvae were
evaluated as a surrogate host; we found that B. pseudomallei and B. mallei, but not other phylogenetically
related bacteria, were highly pathogenic for this insect. More importantly, four previously characterized B.
mallei mutants with reduced virulence in hamsters or mice had similarly reduced virulence in G. mellonella
larvae. Site-specific inactivation of selected genes in the computationally derived virulome identified three new
potential virulence genes, each of which was required for rapid and efficient killing of larvae. Thus, this
approach may provide a means to quickly identify high-probability virulence genes in B. pseudomallei, B. mallei,
and other pathogens.
Burkholderiaceae is a family of the betaproteobacteria that
colonize a variety of ecological niches. Some species are plant
or animal pathogens, while others are important environmen-
tal bacteria, including nitrogen-fixing symbionts, plant growth-
promoting rhizobacteria, chemolithoautotrophs, and bioreme-
diation agents. Nearly all members of the Burkholderiaceae
have large (5- to 9-Mb), multireplicon genomes (http://www
.genomesonline.org/) comprised of a large core of orthologous
genes and smaller subsets of species- or isolate-specific genes
for ecological specialization. Burkholderia pseudomallei and
Burkholderia mallei are aggressive human pathogens (7–9, 13)
categorized as select agent biothreats with a high potential for
misuse (3, 55, 67). The genome sequences of several strains of
B. mallei and B. pseudomallei have revealed that B. mallei is a
deletion clone of B. pseudomallei (27) which lost ?1,000 genes.
Many of the gene deletions appear to have been caused by
insertion sequence-mediated events (49) and likely explain the
many physiological differences that led to the classification of
these two pathogens as separate species (54). Nearly all genes
retained by B. mallei share ?99.5% DNA-DNA sequence
identity with their B. pseudomallei homologs.
B. pseudomallei, which causes melioidosis, is an endemic and
opportunistic pathogen that inhabits the tropical soils and wa-
ters of Southeast Asia and Northern Australia (2, 8, 58, 70),
whereas B. mallei appears to be a zoonotic pathogen that
infects a variety of animals, including equines and humans. The
disease caused by B. mallei, glanders, is not well studied as
there have been few well-documented cases since 1950. For
both pathogens, infection usually occurs through wounds, as-
piration, and possibly inhalation. Studies of intraperitoneally
infected animals have shown that B. pseudomallei and B. mallei
can rapidly migrate to the spleen and liver, where they multiply
extensively within membrane-bound phagosomes and form ab-
scesses (25, 37). Untreated infections are often fatal.
Small-animal models are available for studying the patho-
genesis of B. pseudomallei and B. mallei (15, 34, 40, 42, 68).
These models include a sensitive (50% lethal dose, ?10 cells)
Syrian hamster model utilizing intraperitoneal injection and
two mouse models utilizing C57BL/6 or BALB/c mice and
aerosol, intranasal, or intraperitoneal infection (40, 42). How-
ever, these models are expensive, cumbersome, and hazardous
due to biosafety and biosecurity issues. A Caenorhabditis
elegans nematode infection model for B. pseudomallei has been
described (24, 50, 59), although it may have limited sensitivity
due to the extremely high doses required for killing.
Compared to other bacterial pathogens, little is known
about the virulence factors of B. mallei and B. pseudomallei or
the molecular basis of the pathogenicity of these organisms.
The only verified virulence factors for these pathogens are
those that are well known and shared by most, if not all, animal
and plant pathogens: exopolysaccharide capsule (CAP) (14,
53), type III secretion systems (T3SS) (19, 20, 23, 60, 64, 66),
and lipopolysaccharide O antigen (LPS) (12). Type II secre-
tion, type IV pili, and flagella (10, 11, 18) have been implicated
in B. pseudomallei pathogenesis. Recently, a type VI secretion
* Corresponding author. Mailing address: Department of Microbi-
ology, University of Georgia, Athens, GA 30602. Phone: (706) 542-
2815. Fax: (706) 542-2674. E-mail: email@example.com.
† Supplemental material for this article may be found at http://jb
?Published ahead of print on 25 January 2008.
(T6S) gene cluster was shown to be required for the virulence
of B. mallei (51, 56). Several putative virulence factors, includ-
ing two proteases (10, 41), a hemolysin (1), a rhamnolipid (28),
predicted transporter proteins, and three hypothetical pro-
teins, have been identified using either nematode killing assays
(24) or biochemical analysis. However, these factors are yet to
be verified by animal experiments. Recently, Tuanyok et al.
(62) identified a phospholipase C and a two-component regu-
lator which are required for full virulence of B. pseudomallei.
The genome sequences of ?25 Burkholderiaceae isolates
representing ?12 species and four genera are available (http:
//www.genomesonline.org/). These isolates occupy a wide di-
versity of ecological niches, yet they share a common core
genome comprised of long syntenic regions containing ?2,500
shared “housekeeping” genes whose products share ?60%
amino acid sequence identity. Of greater interest are the spe-
cies- or isolate-specific genes, such as those found only in
human-pathogenic Burkholderia spp. Therefore, we used in
silico genomic subtraction (22) to derive a set of 650 genes that
are present in both B. pseudomallei and B. mallei but are not
present in the sequenced nonpathogenic members of the Burk-
holderiaceae. This set is predicted to be highly enriched in
novel virulence genes and has been designated the Bm-Bp
virulome. We investigated this virulome by selecting genes
likely to encode novel virulence factors, inactivating them, and
testing the corresponding mutants to determine whether they
exhibited reduced virulence in wax moth larvae, a surrogate
model that we adapted and verified as correlating with the
hamster model commonly used to test B. mallei virulence. Our
results show that this is a powerful and easily applicable ap-
proach that can accelerate research on these important patho-
gens and serve as an adjunct to much more expensive and
problematic animal testing.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. Strains and plasmids used
are listed in Table 1. Bacteria were grown at 37°C in LB (45) medium supple-
mented with 2% glycerol (LBG) or in brain heart infusion (BHI) medium. For
some experiments, B. mallei was also grown on minimal medium plates (4) with
5% sucrose as the sole carbon source. When necessary, antibiotics were used at
the following concentrations: kanamycin, 50 ?g/ml; polymyxin B, 20 ?g/ml; and
gentamicin, 20 ?g/ml.
Preparation of electrocompetent B. mallei. An overnight culture was diluted to
obtain an A600of 0.1 in BHI medium and shaken for ?3 h until the A600was 0.8.
Cells were harvested by centrifugation at 6°C at 4,000 ? g and washed with 0.5
culture volume of ice-cold sterile water and then with 0.25 volume of ice-cold
sterile 10% glycerol. Cells were finally resuspended in 0.01 volume of 10%
glycerol and frozen at ?80°C.
Construction of a sacB deletion strain of B. mallei ATCC 23344. Four PCR
primers (see Table S1 in the supplemental material) were used to generate DNA
fragments of the 5? and 3? ends of the sacB gene (BMAA0466) from B. mallei
ATCC 23344, and the fragments were cloned into separate pCR2.1-TOPO vec-
tors. The Sac7/Sac8 product was cut out of pCR2.1-TOPO with EcoRI, and the
Sac5/Sac6 product was cut out with XbaI and BamHI. Both fragments were
cloned stepwise into the corresponding sites of pGRV2 (63), resulting in
TABLE 1. Bacterial strains and plasmids
Strain or plasmid Descriptiona
Reference or source
B. mallei ATCC 23344
E. coli TOP10
E. coli S17-1
B. mallei DD3008
B. mallei RD01
Wild-type pathogen, PmrKmsGmsZeos
General cloning strain
B. mallei derivativesb
?sacB sucrose-resistant derivative
BMA1848 inactivated with pCRXL-ilvI; ilv auxotroph
BMA2786 inactivated with pLCL9; type II secretion
BMA0235 inactivated with pCRXL-A0235
BMAA1900 inactivated with pCRXL-AA1900
BMAA1204 inactivated with pCRXL-AA1204
BMAA1013 inactivated with pCRXL-AA1013
BMA1123 inactivated with pCRXL-A1123
BMAA1111 inactivated with pCRXL-dicty
BMAA1785 inactivated with pCRXL-chitin
BMA0847 inactivated with pCRXL-A0847
BMAA1517 deletion in GRS 23344
BMAA1621 deletion in GRS 23344
Cloning vector, KmrZeor
Cloning vector, KmrAmpr
Gene replacement suicide vector, Gmr
sacB gene deletion allele
Spliced flanking regions of BMAA1517
Spliced flanking regions of BMAA1621
BamHI-XbaI fragment from pCRXL1517SOE
SacI-BamHI fragment from pCRXL1621SOE
aPm, polymyxin B; Gm, gentamicin; Km, kanamycin; Sm, streptomycin; Zeo, zeocin; Amp, ampicillin.
bA gene was inactivated by insertion of a pCR-XL-TOPO vector containing a 400-to 900-bp PCR-amplified internal fragment of the gene.
VOL. 190, 2008NOVEL VIRULENCE GENES OF B. MALLEI2307
pDD159, which was confirmed by sequencing. Subsequently, pDD159 was trans-
formed into Escherichia coli S17-1 (56) and mobilized into B. mallei ATCC
23344. Transconjugants were inoculated into LBG broth without antibiotics, and
10-fold dilutions of the overnight cultures were spread onto LBG plates with 5%
sucrose. Sucrose-resistant colonies were screened for the sacB deletion mutation
by PCR, and one confirmed isolate was designated B. mallei GRS 23344.
Construction of putative virulence mutant. Internal fragments (400 to 900 bp;
lacking both 5? and 3? ends) of target coding sequences (CDSs) were PCR
amplified with appropriate primers (see Table S1 in the supplemental material),
directly cloned into pCR-XL-TOPO (Invitrogen), and transformed into chemi-
cally competent E. coli TOP10. Approximately 0.5 ?g of each of the resultant
plasmids was electroporated into 70 ?l of ice-cold electrocompetent B. mallei
cells with an Eppendorf w-150 electroporator at 2.3 kV, 600 ?, and 10 ?F for 5
ms. Cells were immediately diluted 10-fold with LBG and shaken for 4 h at 37°C,
after which aliquots were spread onto LBG plates containing kanamycin. Colo-
nies that grew after 3 to 4 days were picked, the genomic DNA was isolated, and
the presence of the appropriate cis-merodiploid mutant allele was verified by
PCR. Typically, 25% of the colonies harbored the expected mutant allele.
BMAA1517 and BMAA1621 were refractory to insertional inactivation, so we
used splice overlap extension PCR (32) with a modified sacB-based deletion
method (43, 63). Briefly, four PCR primers (see Table S1 in the supplemental
material) were used to amplify two ?500-bp fragments of DNA flanking each
target gene. Gel-purified amplicons were mixed in a second PCR mixture with
the forward primer of the upstream amplicon and the reverse primer of the
downstream amplicon. The ?1,000-bp chimeric fragments were cloned into
pCR-XL-TOPO, cut out with restriction enzymes (Table 1), and ligated into the
appropriate sites in pGRV2 (63). The resulting plasmids were transformed into
E. coli S17-1 (57) and mobilized into the B. mallei sacB deletion mutant GRS
23344. Transconjugants were spread onto minimal medium sucrose plates. Col-
onies of putative deletion mutants that grew after 5 to 7 days of incubation were
screened by PCR to verify deletion of the target sequence. Approximately 10%
of the colonies tested harbored the deletion allele.
Infection of Galleria mellonella larvae. B. mallei strains were grown at 37°C for
24 to 48 h on LBG plates, resuspended at an A600of 0.2 in BHI medium, and
shaken at 37°C until the A600was between 0.4 and 0.6. Culture aliquots were
diluted in sterile water, and ?10 or 250 cells in 5 ?l were injected under the
carapace into the hemocoel of 10 fifth-instar G. mellonella larvae (Vanderhorst
Wholesale, Marysville OH) using a 50-?l gas-tight Hamilton syringe equipped
with a 30.5-gauge needle (Becton Dickinson 305106). Inoculated larvae were
sealed in 2-oz hinged-lid polypropylene cups (Fisher 03-405-39) with five 2-mm
ventilation holes in the lid. Larval death was scored daily for 1 week; dead larvae
no longer moved when the cup was tapped and/or had turned black.
Bioinformatic methods. In silico genomic subtraction was performed as pre-
viously described (22). Briefly, using BLASTP, the amino acid sequences en-
coded by all ?6,000 Glimmer-predicted CDSs of B. mallei ATCC 23344 were
individually aligned with the best hits for predicted CDSs of five closely related
nonpathogenic strains: Burkholderia sp. strain 383 (? ATCC 17660 ? R-18194),
Ralstonia eutropha JMP134, Burkholderia ambifaria AMMD, Burkholderia xeno-
vorans LB400, and Burkholderia vietnamiensis G4. The Burkholderia strains were
chosen because they are nonclinical (environmental) isolates with a long history
of use and no reports of human infection. Sequences of their CDS products and
contigs were obtained from JGI/ORNL (http://compbio.ornl.gov/channel). For
each CDS alignment, an “orthology score” was calculated by using amino acid
sequence identity, the ratio of alignment and query lengths, and the ratio of
query and hit lengths. Utilizing an orthology score cutoff value based on the
average amino acid sequence identities of the 100 most similar ortholog pairs of
each, a set of B. mallei CDS products lacking orthologs in any of the closely
related nonpathogens was derived. TBLASTN was then used to compare each
CDS in this set to the genomic DNA sequence of the same five bacteria to find
and remove orthologous CDSs of B. mallei present in the DNA sequence but not
in the called CDS sets. This method provided results similar to those obtained
with the reciprocal best BLAST hit method but, in contrast, can be used with raw
unfinished genome sequences (22).
Select agent handling. GmrB. mallei strains constructed during this work were
generated before the Centers for Disease Control Select Agent Program (CDC-
SAP) informed us that, despite the fact that Gmrhad been widely used in B.
mallei before 2006, introduction of Gmrinto B. mallei is now considered a
“restricted experiment” that requires prior CDC-SAP approval. Thus, all GmrB.
mallei strains that we constructed have been destroyed. Otherwise, all regulations
and procedures mandated by CDC-SAP were followed during this work.
RESULTS AND DISCUSSION
Derivation of a virulome for B. mallei and B. pseudomallei.
Despite the extensive ecological diversity of the organisms,
genome comparisons of members of the Burkholderiaceae have
revealed a core genome comprised of long syntenic regions
containing thousands of shared “housekeeping” genes. Of
greater interest are the hundreds of species- or isolate-specific
genes that confer on each isolate its ability to colonize and
survive in a specific niche or host. To find genes that are
specific to B. mallei and B. pseudomallei and also have in-
creased probability of major involvement in virulence, we used
an in silico genomic subtraction method in which the amino
acid sequences encoded by all B. mallei CDSs were individually
compared to the amino acid sequences encoded by all the
CDSs of four closely related nonpathogenic Burkholderia
strains, and all CDS products found in “orthologous” pairs
were removed. This yielded a set of 650 CDSs designated the
Bm-Bp virulome (see Table S2 in the supplemental material),
all of which are present in both B. mallei and B. pseudomallei
but whose products have ?45% amino acid sequence identity
with the product of any other CDS in the genomes of the
nonpathogens. In contrast, the average amino acid identity
observed for most orthologous CDS pairs is ?65%. This viru-
lome comprises ?12% of the genome of B. mallei and appears
to be highly enriched in CDSs/genes encoding probable viru-
lence and pathogenicity factors since it contains most of the
genes in the animal pathogen-like T3SS (T3SSAP), CAP, T6S,
and LPS biosynthetic gene clusters, the only published loci
experimentally proven to be critical for causing disease in the
hamster model (14, 56, 64). When we attempted to use the
total CDSs of Burkholderia thailandensis E264, a strain previ-
ously considered a nonpathogen (26, 47), in a subsequent in
silico subtraction experiment to further reduce and refine the
virulome, we found orthologs for 70% of the virulome CDSs in
the nonpathogen, including the majority of the CDSs in the
three large gene clusters (T3SS, T6S, and CAP) that are es-
sential for B. mallei virulence. In contrast, orthologs for ?10%
of the virulome CDSs were found in the genome of Burkhold-
eria cenocepacia J2315 or B. cenocepacia AU1054, two patho-
genic strains isolated from cystic fibrosis patients (5, 44, 65).
This is perhaps not surprising considering that B. thailandensis
E264 is very closely related to B. pseudomallei and that at
relatively high doses it has the ability to kill nematodes (50)
and hamsters (2) and has the ability to survive in macrophages
(F. Gherardini, personal communication).
Most of the 650 virulome CDSs occur in ?40 operonlike clus-
ters of four or more contiguous genes (only 100 genes are single-
tons). Not surprisingly, most clusters are on the smaller replicon,
which carries the majority of nonhousekeeping genes (31, 49).
Particularly high concentrations of the CDSs are found between
BMAA0345 and BMAA0410, between BMAA0729 and BMAA
0758, and between BMAA1517 and BMAA1565. Very few of the
CDSs in clusters have Karlin score deviations (38), G?C con-
tents, or GC skews that significantly differ from those of the
flanking regions or the genome averages. Nearly 130 CDSs are
predicted by SignalP V.3.0 (48) to encode signal peptides for
secretion, a value which is about four times higher than the pre-
dicted genome-wide average. In the clusters with ?10 genes are
two complete sets of T3SS genes and several predicted secondary
2308SCHELL ET AL.J. BACTERIOL.
metabolite-producing clusters encoding nonribosomal peptide or
polyketide synthetases possibly involved in the cytotoxic activities
ascribed to B. pseudomallei (6, 29, 30). Also present are CDSs
encoding major parts of both the CAP and LPS biosynthesis
clusters and two fimbria/pilus biogenesis pathways, in addition to
?100 CDSs encoding predicted membrane proteins. The pres-
ence of CDSs encoding proteins with predicted enzymatic func-
tions in other clusters suggests that these CDSs may encode parts
of metabolic pathways. However, many of the clustered genes
have no predicted function; ?200 CDSs lack any functional de-
scription, and ?100 do not have PFAM domains.
G. mellonella larvae are a valid surrogate host for B. mallei.
The ultimate goal of deriving the virulome was to extract a
comprehensive set of potentially novel virulence genes that
contained few enough genes that the majority could be indi-
vidually knocked out and mutants evaluated to determine
whether the knockouts reduced virulence in animal models.
Uncertainty about the pathogenic potential of B. thailandensis
precluded using in silico subtraction to reduce the virulome to
a size at which we could realistically test a large proportion of
the genes in mammalian models under cumbersome BSL3
conditions. In contrast, higher-throughput surrogate host sys-
tems can be very simple and inexpensive, yet quite accurate for
study of human pathogens (61). G. mellonella (wax moth)
larvae have served as an insect host and simple pathogenesis
model for several mammalian pathogens (39); they are easy to
acquire and contain without special facilities. Most impor-
tantly, G. mellonella has a circulatory system and a complex
innate and mammal-like immune response that effectively
deals with a wide range of microbial pathogens (39). In some
cases the virulence of a pathogen and attenuated mutants in
animal hosts has been accurately reflected in G. mellonella (33,
46). Therefore, we tested G. mellonella as a pathogenesis
model for verification of B. mallei virulence genes in the
When ?10 wild-type B. mallei ATCC 23344 (or B. mallei
GRS 23344) cells were injected into the hemocoel of larvae,
the majority of the larvae were killed within 4 days. In ?20
experiments we inoculated more than 300 larvae with any-
where from 3 to 200 cells of wild-type B. mallei and always
observed ?90% killing within 6 days. Injecting 10 B.
pseudomallei K96243 cells into the larvae killed them nearly
twice as fast (?80% mortality by 2 days). Similar to hamsters
infected with B. mallei or B. pseudomallei, infected larvae
showed extensive paralysis 12 h before death. At the onset of
paralysis, typically ?104B. mallei cells/ml were present in the
hemolymph; just prior to death ?106B. mallei or B. pseudoma-
llei cells/ml were found. These results indicate that B. mallei
and B. pseudomallei can multiply to very high numbers in the
hemocoel of wax moth larvae and are highly pathogenic for
this insect. Injection of up to 105cells of the pathogenic cystic
fibrosis epidemic strain B. cenocepacia J2315, Pseudomonas
putida, B. xenovorans, or E. coli W3110 killed ?15% of the
larvae by day 6; using an inoculum containing 104cells reduced
the percentage of dead larvae at day 6 to ?10%, which is
identical to the level observed if water was injected. Injection
of ?105B. thailandensis cells killed ?35% of the larvae after 7
days. Although this is somewhat inconsistent with the “non-
pathogenic” label for B. thailandensis, as mentioned above, this
species harbors ?70% of the genes in the Bm-Bp virulome and
at relatively high doses (?105cells) it also can kill hamsters
and nematodes. Moreover, at least one opportunistic infection
of a human by B. thailandensis has been reported (26).
To show correspondence between the wax moth larva model
and mammalian models, we first tested B. mallei RD01, which
has a polar insertion in the T3SSAPgene cluster (64). This
T3SS is essential for the virulence of B. mallei and B.
pseudomallei in mice and hamsters, respectively (64, 66). Com-
pared to the wild type, the ability of B. mallei RD01 to kill wax
moth larvae was dramatically reduced (Fig. 1A, compare WT
and T3SS); at lower infection levels (?20 cells/larva), RD01
was largely avirulent (data not shown). When B. mallei strain
FIG. 1. Killing curves for G. mellonella larvae inoculated with wild-type and mutant B. mallei cells. (A) Time course of larval killing by B. mallei
wild type (WT) and mutants with defects in characterized virulence genes. (B) Time course of larval killing by wild type and mutants with defects
in genes selected from the Bm-Bp virulome. Assays were performed at least three independent times by inoculating 10 larvae with each mutant.
The actual number of cells (CFU) injected into the larvae was determined by plating duplicate aliquots of the injectant on LBG plates. The
percentages of larvae killed are the averages of three assays; the standard deviations are ?15%. Wild-type B. mallei and water-inoculated controls
were included with each set of assays.
VOL. 190, 2008 NOVEL VIRULENCE GENES OF B. MALLEI2309
DD3008, a CPS-deficient mutant with dramatically reduced
virulence in hamsters (10), was used, the killing of wax moth
larvae was delayed and reduced (Fig. 1A, compare WT and
CAP). Next, we constructed an isoleucine-valine auxotroph of
B. mallei by insertional inactivation of ilvI (BMA1848) encod-
ing a subunit of acetolactate synthase III. The resultant mutant
was not able to grow on minimal glucose medium and, similar
to what has been observed with mice (63), showed attenuated
virulence in larvae (Fig. 1A, compare WT and IlvI). Thus, the
behavior of the three mutants whose virulence in animal mod-
els is attenuated correlated well with their virulence in larvae.
Inactivation of type II secretion affects the virulence of P.
aeruginosa (17, 69), some plant pathogens (36, 43), and B.
pseudomallei (10). Therefore, we insertionally inactivated gspD
(BMA2786), a key component of type II secretion, and found
that the killing of wax moth larvae by the resultant GspD
mutant was reduced and delayed (Fig. 1A), but not to the
extent caused by the T3SS mutation. These data suggest that
the wax moth model of infection is an efficient predictor of
potential virulence genes in B. mallei.
To verify the stability of the mutant genotype in the wax
moth larvae, bacterial cells were recovered from the hemo-
lymph at days 2 and 5 by plating on nonselective LBG agar. All
of the 250 resultant colonies tested retained the antibiotic
resistance encoded by the plasmid insertion. When auxotro-
phic ilvI mutant cells were recovered from larval hemolymph
and plated on both LBG and minimal media, we found that ?1
in 105of the ex vivo cells had reverted to prototrophy, further
demonstrating the high level of stability of the cis-merodiploid
Use of the wax moth larva model to screen putative viru-
lence mutants. Based on a manual review of bioinformatic data
for virulome CDSs, 10 CDSs were evaluated to determine
whether they have a role in virulence (Table 2). Some of these
CDSs were chosen because their best BLAST hits were in phy-
logenetically distant pathogens or in a eukaryote, some were cho-
sen because they appeared to be potential virulence gene regu-
lators, and some were chosen because they appeared to encode
polyketide or peptide synthetases that may be involved in cyto-
toxin production. Knockout mutants for each CDS were con-
structed and verified by PCR, and their virulence in wax moth
larvae was tested as described above. B. mallei strains with mu-
tations in BMA0847, BMAA1111, BMA1123, BMAA1013, and
BMAA1621 had marginal or no differences in virulence com-
pared with the wild type (Fig. 1A and data not shown), whereas
with strains harboring insertions in BMA0235, BMAA1204,
BMAA1517, BMAA1785, and BMAA1900 there was reduced
and/or delayed killing of the larvae (Fig. 1B).
Of all the mutants tested, inactivation of BMAA1517, en-
coding an AraC-type transcriptional regulator, resulted in the
most dramatic reduction in virulence in the larvae. At 7 days
postinfection 80% of the larvae infected with ?200 CFU of the
BMAA1517 mutant remained asymptomatic, similar to the
behavior of the T3SSAPmutant. Whole-genome expression
profiling of a B. mallei strain overexpressing BMAA1517 has
shown that a cluster of genes found in the Bm-Bp virulome
(BMAA0727 to BMAA0744) is exclusively upexpressed on
average 4-fold; the levels of some genes are elevated ?10-fold
(56). This previous study also showed that these genes encode
components of a T6S system which, when inactivated, ren-
dered B. mallei avirulent in hamsters. When we tested a mutant
lacking a critical component of the T6S system (BMAA0739),
its virulence in larvae was reduced to the same extent as that of
BMAA1517 mutants (not shown). These results further sup-
port the hypothesis that the wax moth larva model is an accu-
rate predictor of B. mallei virulence genes.
Inactivation of BMA0235 in B. mallei dramatically delayed
and reduced killing of larvae in the wax moth infection model.
BMA0235 is in an operonlike cluster of six genes that is con-
served in all sequenced B. pseudomallei and B. mallei strains
but is not present in any other Burkholderia genome or in the
600 microbial genomes at Integrated Microbial Genomes
(http://img.jgi.doe.gov/cgi-bin/pub/main.cgi), with the excep-
tion of the genome of Photorhabdus luminescens. In fact, P.
luminescens, which can be an insect pathogen or nematode
symbiont (21), has similarly organized orthologs of the five
genes flanking BMA0235 (Fig. 2). The average amino acid
identity between the products of the six CDS pairs from these
phylogenetically distant bacteria is 66%, which is quite remark-
able considering that the G?C content of the P. luminescens
CDSs is ?38%, while the average G?C content of the B.
mallei orthologs is 68%. The BMA0237 protein has a domain
TABLE 2. Potential virulence genes targeted for mutagenesis
B. mallei CDSProducta
Rationale for targeting
Virulence of mutant
BMA02353-Phosphoshikimate 1-carboxyvinyltransferaseBest BLASTP hit in phylogenetically
Best BLASTP hit in eukaryote
Possible role in cytotoxin production
Potential virulence gene regulator
Best BLASTP hit in eukaryote
Possible role in cytotoxin production
Potential virulence gene regulator
Potential virulence gene regulator
Best BLASTP hit in phylogenetically
Best BLASTP hit in phylogenetically
Galactose oxidase-related protein
ECF sigma factor
AraC family transcriptional regulator
Regulatory protein HrpB
Chitin binding domain protein
BMAA1900 Pentapeptide repeat family protein Reduced
aGene product description in Integrated Microbial Genomes (http://img.jgi.doe.gov/).
bSee text and Fig. 1B.
2310SCHELL ET AL. J. BACTERIOL.
similar to a domain in chorismate mutase (PFAM1042), while
the BMA0236a and BMA0235 proteins are both assigned to
the enoylpyruvate carboxytransferase family (PFAM 00275)
and are related to 3-phosphoshikimate 1-carboxyvinyl trans-
ferases. Both of the latter proteins were originally annotated as
AroA (EC 126.96.36.199), which is involved in aromatic amino acid
biosynthesis. However, these proteins show ?25% identity to
each other and to E. coli AroA. The AroA function of B. mallei
more likely resides in the BMA0430 protein, annotated as a
bifunctional prephenate dehydrogenase/3-phosphoshikimate
1-carboxyvinyl transferase enzyme whose C-terminal and N-
terminal halves show ?50% amino acid identity to full-length
E. coli AroA and TyrA, respectively. Thus, the BMA0235 and
BMA0236a proteins are more likely novel carboxyvinyl trans-
ferases that transfer a vinyl group from phosphoenolpyruvate
to an unknown substrate. The BMA0234 protein is annotated
as a putative asparagine synthase B (EC 188.8.131.52), as are the
products of two other CDSs in the B. mallei genome
(BMAA1158 and BMAA1921). The products of these three
CDSs do not show ?30% amino acid identity to E. coli AsnB.
The BMA0234 protein is likely involved in an AsnB-like reac-
tion, transfer of the amido group of glutamine to an aspartate-
like substrate. The BMA0233a protein may be a carboxylase,
while the BMA0232 protein is a putative drug/metabolite
transporter (PFAM 05297). It is plausible that the gene clus-
ters shown in Fig. 2 synthesize a secondary metabolite or toxin
that plays a role in insect pathogenesis in P. luminescens and
perhaps in B. mallei. This metabolite could be derived from
carboxyvinylation of 3,4-dihydroxy-2-butanone 4-phosphate
made by the adjacent ribA gene, followed by amidation of the
carboxyvinyl group. P. luminescens and its close relatives are
well known for production of insect toxins.
BMAA1204 mutants took twice as long as the wild type to
start killing larvae; after 7 days only 60% of the mutant-in-
fected larvae were dead, compared to the 100% mortality
caused by the wild type. BMAA1204 is a ?4,200-residue CDS
annotated as encoding a putative polyketide synthase (PKS) in
COG family 0332. PKSs are giant multidomain proteins ubiq-
uitous in the prokaryotes. In a manner analogous to fatty acid
biosynthesis, they condense or “polymerize” organic acids
(e.g., acetate and malonate) into linear polyketides. The re-
sultant polyketides are often cyclized and modified, yielding
300- to 1,500-Da molecules that exhibit diverse antibiotic, an-
titumor, or even immunosuppressive activities. BMAA1204 is
in a large gene cluster along with genes encoding two other
putative PKSs and several CDSs whose products are in protein
families associated with polyketide synthesis and/or modifica-
tion. In fact, most CDSs in this region (BMAA1201 to
BMAA1222) are in the virulome. It is possible that BMAA
1204 and/or the adjacent PKS genes encode production of one
of the “cytotoxin” activities found previously in culture super-
natants of B. pseudomallei (16, 41, 42), whose characteristics
(e.g., heat stability and molecular weight) are consistent with a
polyketide origin. PKSs have been implicated in pathogenesis
by several other bacteria (e.g., Mycobacterium ulcerans ). B.
mallei has seven putative PKS genes; three of these genes
(BMAA1021, BMAA1204, and BMAA1451) are present in the
virulome but are not present in B. thailandensis.
Inactivation of BMAA1785, whose product is annotated as a
“chitin binding domain protein,” reduced virulence quantita-
tively to an extent very similar to the extent observed after
inactivation of the PKS gene BMAA1204. Although nonpatho-
genic Burkholderia spp. lack orthologs of the BMAA1785 pro-
tein, Yersinia pestis and Yersinia pseudotuberculosis have or-
thologs with 55% amino acid sequence identity. The actual
function of the BMAA1785 protein is not clear, but its so-
called chitin binding domain 3 (PFAM03067) is related to
baculoviral spheroidin and spindolin capsid proteins. It has a
predicted signal sequence and is present at moderately high
levels in culture supernatants of B. pseudomallei and B. mallei
(52; data not shown). We speculate that it binds to some type
of chitin-like oligosaccharide moiety and has an as-yet-un-
Inactivation of BMAA1900 caused only a moderate delay (1
day) in the onset of killing and a minor reduction in B. mallei-
related mortality of wax moth larvae. This gene is in a cluster
(BMAA1897 to BMAA1915) that is predicted to encode one
of the four predicted T6S systems in B. mallei (56). BMAA1900
encodes a putative pentapeptide repeat protein (COG family
1357) that is 38% identical to a putative exported protein of
Bordetella bronchiseptica, which is located in an analogous T6S
system cluster. While one can speculate that BMAA1900 is
involved in T6S system-mediated secretion of a virulence fac-
tor(s), it seems that this T6S system plays a relatively minor
role in virulence for wax moth larvae.
Conclusion. To date, relatively few virulence factors of B.
mallei have been identified, and most of the factors that have
been found are widespread in diverse pathogens (e.g., T3SS
and capsule). Some of the methods used to identify these
virulence factors, such as subtractive hybridization and trans-
poson mutagenesis, are time-consuming and especially cum-
bersome with BSL3-requiring select agent pathogens. With the
availability of hundreds of bacterial genome sequences, it
makes sense to exploit these databases to use new bioinfor-
matic methods to direct functional genomic studies. In silico
genomic subtraction provided a set of B. mallei genes that
contained genes encoding all known virulence factors and
FIG. 2. Organization and amino acid sequence identity for ortholo-
gous genes in B. mallei (BMA) and P. luminescens (PLU) that may
encode an insect toxin. Orthologous genes have the same arrow pat-
tern. The location of the B. mallei CDS (BMA0235) that was disrupted
by insertion is indicated. DMT, drug/metabolite transporter.
VOL. 190, 2008NOVEL VIRULENCE GENES OF B. MALLEI 2311
therefore was very valuable as a starting point for functional
analyses to discover new genes with a role in pathogenesis.
G. mellonella larvae have successfully been used as a model
for infection by several pathogenic bacteria (39), as well as in
screening for bacterial mutants with reduced virulence and/or
altered host colonization ability (33, 46). We have shown that
G. mellonella also is a good model for B. mallei pathogenesis,
because four B. mallei strains with mutations in different char-
acterized virulence gene clusters (T3SS, CAP, type II secretion
system, and T6S system) showed reduced virulence that cor-
related well with their behavior in the hamster model of infec-
tion. The larva model is a facile and inexpensive method that
can be used to test potential virulence mutants of B. mallei
before hamster or mouse models are used. The value of the
larva model is enhanced by the observation that G. mellonella
has many sophisticated defenses analogous to those found in
mammals, such as circulating hemocytes utilizing surface re-
ceptors similar to those found in mammalian phagocytes to
help encapsulate and destroy microbial invaders (39). Other
antimicrobial responses of these insects to pathogens are also
similar to those of mammals, including oxidative burst and
inducible production of lysozyme and other antibacterial pro-
teins, such as cecropins, defensins, and proline-rich peptides.
Besides demonstrating the utility and validity of the G. mel-
lonella insect model for screening mutants for virulence phe-
notypes, some of our results (i.e., the killing of larvae by very
low infectious doses and the presence of multiple insect patho-
gen-associated genes in the virulome [Fig. 2; see Table S2 in
the supplemental material]) suggest that some soil-dwelling
insects may be natural hosts and reservoirs of B. pseudomallei
(and perhaps B. mallei). However, if this is true, the use of the
larva model may result in a bias toward discovery of genes for
virulence in insects that may or may not play as significant a
role in mammalian pathogenesis.
This work was supported by National Institutes of Health grant
R21-AAI69081 to M.A.S.
We thank Michele Sturgeon and Erin Raybon for their technical
assistance and Benjamin Hasselbring and Joanne Rue for critiquing
1. Ashdown, L. R., and J. M. Koehler. 1990. Production of hemolysin and other
extracellular enzymes by clinical isolates of Pseudomonas pseudomallei.
J. Clin. Microbiol. 28:2331–2334.
2. Brook, M. D., B. Currie, and P. M. Desmarchelier. 1997. Isolation and
identification of Burkholderia pseudomallei from soil using selective culture
techniques and the polymerase chain reaction. J. Appl. Microbiol. 82:589–
3. Christopher, G. W., T. J. Cieslak, J. A. Pavlin, and E. M. Eitzen, Jr. 1997.
Biological warfare. A historical perspective. JAMA 278:412–417.
4. Clough, S. J., M. A. Schell, and T. P. Denny. 1994. Evidence for involvement
of a volatile extracellular factor in Pseudomonas solanacearum virulence
gene expression. Mol. Plant-Microbe Interact. 7:621–630.
5. Coenye, T., T. Spilker, A. Van Schoor, J. J. LiPuma, and P. Vandamme.
2004. Recovery of Burkholderia cenocepacia strain PHDC from cystic fibrosis
patients in Europe. Thorax 59:952–954.
6. Colling, M., C. Nigg, and R. J. Heckly. 1958. Toxins of Pseudomonas
pseudomallei. I. Production in vitro. J. Bacteriol. 76:422–426.
7. Currie, B. J., D. A. Fisher, D. M. Howard, J. N. Burrow, D. Lo, S. Selva-
Nayagam, N. M. Anstey, S. E. Huffam, P. L. Snelling, P. J. Marks, D. P.
Stephens, G. D. Lum, S. P. Jacups, and V. L. Krause. 2000. Endemic
melioidosis in tropical northern Australia: a 10-year prospective study and
review of the literature. Clin. Infect. Dis. 31:981–986.
8. Dance, D. A. 2000. Ecology of Burkholderia pseudomallei and the interactions
between environmental Burkholderia spp. and human-animal hosts. Acta
9. Dance, D. A. 2002. Melioidosis. Curr. Opin. Infect. Dis. 15:127–132.
10. DeShazer, D., P. J. Brett, M. N. Burtnick, and D. E. Woods. 1999. Molecular
characterization of genetic loci required for secretion of exoproducts in
Burkholderia pseudomallei. J. Bacteriol. 181:4661–4664.
11. DeShazer, D., P. J. Brett, R. Carlyon, and D. E. Woods. 1997. Mutagenesis
of Burkholderia pseudomallei with Tn5-OT182: isolation of motility mutants
and molecular characterization of the flagellin structural gene. J. Bacteriol.
12. DeShazer, D., P. J. Brett, and D. E. Woods. 1998. The type II O-antigenic
polysaccharide moiety of Burkholderia pseudomallei lipopolysaccharide is
required for serum resistance and virulence. Mol. Microbiol. 30:1081–1100.
13. DeShazer, D., and D. M. Waag. 2004. Glanders: new insights into an old
disease, p. 209–237. In L. Lindler, F. Lebada, and G. W. Korch (ed.),
Biological weapons defense: principles and mechanisms for infectious dis-
eases counter-bioterrorism. Humana Press, Inc., Totowa, NJ.
14. DeShazer, D., D. M. Waag, D. L. Fritz, and D. E. Woods. 2001. Identification
of a Burkholderia mallei polysaccharide gene cluster by subtractive hybrid-
ization and demonstration that the encoded capsule is an essential virulence
determinant. Microb. Pathog. 30:253–269.
15. DeShazer, D., and D. Woods. 1999. Animal models of melioidosis, p. 199–
203. In O. Zak and M. Sande (ed.), Handbook of animal models of infection.
Academic Press Ltd., London, United Kingdom.
16. Dharakul, T., and S. Songsivilai. 1999. The many facets of melioidosis.
Trends Microbiol. 7:138–140.
17. Durand, E., A. Bernadac, G. Ball, A. Lazdunski, J. N. Sturgis, and A. Filloux.
2003. Type II protein secretion in Pseudomonas aeruginosa: the pseudopilus
is a multifibrillar and adhesive structure. J. Bacteriol. 185:2749–2758.
18. Essex-Lopresti, A. E., J. A. Boddey, R. Thomas, M. P. Smith, M. G. Hartley,
T. Atkins, N. F. Brown, C. H. Tsang, I. R. A. Peak, and J. Hill. 2004. A type
IV pilin, PilA, contributes to adherence of Burkholderia pseudomallei and
virulence in vivo Infect. Immun. 73:1260–1264.
19. Finlay, B. B., and P. Cossart. 1997. Exploitation of mammalian host cell
functions by bacterial pathogens. Science 276:718–725.
20. Finlay, B. B., and S. Falkow. 1997. Common themes in microbial pathoge-
nicity revisited. Microbiol. Mol. Biol. Rev. 61:136–169.
21. Fischer-Le Saux, M., V. Viallard, B. Brunel, P. Normand, and N. E.
Boemare. 1999. Polyphasic classification of the genus Photorhabdus and
proposal of new taxa: P. luminescens subsp. luminescens subsp. nov., P.
luminescens subsp. akhurstii subsp. nov., P. luminescens subsp. laumondii
subsp. nov., P. temperata sp. nov., P. temperata subsp. temperata subsp. nov.,
and P. asymbiotica sp. nov. Int. J. Syst. Bacteriol. 49:1645–1656.
22. Gabriel, D. W., C. Allen, M. Schell, T. P. Denny, J. T. Greenberg, Y. P. Duan,
Z. Flores-Cruz, Q. Huang, J. M. Clifford, G. Presting, E. T. Gonzalez, J.
Reddy, J. Elphinstone, J. Swanson, J. Yao, V. Mulholland, L. Liu, W. Farm-
erie, M. Patnaikuni, B. Balogh, D. Norman, A. Alvarez, J. A. Castillo, J.
Jones, G. Saddler, T. Walunas, A. Zhukov, and N. Mikhailova. 2006. Iden-
tification of open reading frames unique to a select agent: Ralstonia
solanacearum race 3 biovar 2. Mol. Plant-Microbe Interact. 19:69–79.
23. Galan, J. E., and A. Collmer. 1999. Type III secretion machines: bacterial
devices for protein delivery into host cells. Science 284:1322–1328.
24. Gan, Y. H., K. L. Chua, H. H. Chua, B. Liu, C. S. Hii, H. L. Chong, and P.
Tan. 2002. Characterization of Burkholderia pseudomallei infection and iden-
tification of novel virulence factors using a Caenorhabditis elegans host sys-
tem. Mol. Microbiol. 44:1185–1197.
25. Gauthier, Y. P., R. M. Hagen, G. S. Brochier, H. Neubauer, W. D.
Splettstoesser, E. J. Finke, and D. R. Vidal. 2001. Study on the pathophys-
iology of experimental Burkholderia pseudomallei infection in mice. FEMS
Immunol. Med. Microbiol. 30:53–63.
26. Glass, M. B., J. E. Gee, A. G. Steigerwalt, D. Cavuoti, T. Barton, R. D. Hardy,
D. Godoy, B. G. Spratt, T. A. Clark, and P. P. Wilkins. 2006. Pneumonia and
septicemia caused by Burkholderia thailandensis in the United States. J. Clin.
27. Godoy, D., G. Randle, A. J. Simpson, D. M. Aanensen, T. L. Pitt, R.
Kinoshita, and B. G. Spratt. 2003. Multilocus sequence typing and evolu-
tionary relationships among the causative agents of melioidosis and glanders,
Burkholderia pseudomallei and Burkholderia mallei. J. Clin. Microbiol. 41:
28. Haussler, S., M. Nimtz, T. Domke, V. Wray, and I. Steinmetz. 1998. Purification
and characterization of a cytotoxic exolipid of Burkholderia pseudomallei. Infect.
29. Heckly, R. J. 1964. Differentiation of exotoxin and other biologically active
substances in Pseudomonas pseudomallei filtrates. J. Bacteriol. 88:1730–1736.
30. Heckly, R. J., and C. Nigg. 1958. Toxins of Pseudomonas pseudomallei. II.
Characterization. J. Bacteriol. 76:427–436.
31. Holden, M. T., R. W. Titball, S. J. Peacock, A. M. Cerdeno-Tarraga, T.
Atkins, L. C. Crossman, T. Pitt, C. Churcher, K. Mungall, S. D. Bentley, M.
Sebaihia, N. R. Thomson, N. Bason, I. R. Beacham, K. Brooks, K. A. Brown,
N. F. Brown, G. L. Challis, I. Cherevach, T. Chillingworth, A. Cronin, B.
Crossett, P. Davis, D. DeShazer, T. Feltwell, A. Fraser, Z. Hance, H. Hauser,
S. Holroyd, K. Jagels, K. E. Keith, M. Maddison, S. Moule, C. Price, M. A.
Quail, E. Rabbinowitsch, K. Rutherford, M. Sanders, M. Simmonds, S.
Songsivilai, K. Stevens, S. Tumapa, M. Vesaratchavest, S. Whitehead, C.
2312SCHELL ET AL.J. BACTERIOL.
Yeats, B. G. Barrell, P. C. Oyston, and J. Parkhill. 2004. Genomic plasticity
of the causative agent of melioidosis, Burkholderia pseudomallei. Proc. Natl.
Acad. Sci. USA 101:14240–14245.
32. Horton, R. M., S. N. Ho, J. K. Pullen, H. D. Hunt, Z. L. Cai, and L. R. Pease.
1993. Gene-splicing by overlap extension. Methods Enzymol. 217:270–279.
33. Jander, G., L. G. Rahme, and F. M. Ausubel. 2000. Positive correlation
between virulence of Pseudomonas aeruginosa mutants in mice and insects. J.
34. Jeddeloh, J. A., D. L. Fritz, D. M. Waag, J. M. Hartings, and G. P. Andrews.
2003. Biodefense-driven murine model of pneumonic melioidosis. Infect.
35. Jenkin, G. A., T. P. Stinear, P. D. Johnson, and J. K. Davies. 2003. Subtrac-
tive hybridization reveals a type I polyketide synthase locus specific to
Mycobacterium ulcerans. J. Bacteriol. 185:6870–6882.
36. Jha, G., R. Rajeshwari, and R. V. Sonti. 2007. Functional interplay between
two Xanthomonas oryzae pv. oryzae secretion systems in modulating viru-
lence on rice. Mol. Plant-Microbe Interact. 20:31–40.
37. Jones, A. L., T. J. Beveridge, and D. E. Woods. 1996. Intracellular survival of
Burkholderia pseudomallei. Infect. Immun. 64:782–790.
38. Karlin, S. 2001. Detecting anomalous gene clusters and pathogenicity islands
in diverse bacterial genomes. Trends Microbiol. 9:335–343.
39. Kavanagh, K., and E. P. Reeves. 2004. Exploiting the potential of insects for
in vivo pathogenicity testing of microbial pathogens. FEMS Microbiol. Rev.
40. Leakey, A. K., G. C. Ulett, and R. G. Hirst. 1998. BALB/c and C57Bl/6 mice
infected with virulent Burkholderia pseudomallei provide contrasting animal
models for the acute and chronic forms of human melioidosis. Microb.
41. Lee, M. A., and Y. Liu. 2000. Sequencing and characterization of a novel
serine metalloprotease from Burkholderia pseudomallei. FEMS Microbiol.
42. Liu, B., G. C. Koo, E. H. Yap, K. L. Chua, and Y. H. Gan. 2002. Model of
differential susceptibility to mucosal Burkholderia pseudomallei infection.
Infect. Immun. 70:504–511.
43. Liu, H., S. Zhang, M. A. Schell, and T. P. Denny. 2005. Pyramiding un-
marked deletions in Ralstonia solanacearum shows that secreted proteins in
addition to plant cell-wall-degrading enzymes contribute to virulence. Mol.
Plant-Microbe Interact. 18:1296–1305.
44. Liu, L., T. Spilker, T. Coenye, and J. J. LiPuma. 2003. Identification by
subtractive hybridization of a novel insertion element specific for two wide-
spread Burkholderia cepacia genomovar III strains. J. Clin. Microbiol. 41:
45. Miller, J. H. 1992. A short course in bacterial genetics: a laboratory manual
and handbook for Escherichia coli and related bacteria. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
46. Miyata, S., M. Casey, D. W. Frank, F. M. Ausubel, and E. Drenkard. 2003.
Use of the Galleria mellonella caterpillar as a model host to study the role of
the type III secretion system in Pseudomonas aeruginosa pathogenesis. In-
fect. Immun. 71:2404–2413.
47. Moore, R. A., S. Reckseidler-Zenteno, H. Kim, W. Nierman, Y. Yu, A.
Tuanyok, J. Warawa, D. DeShazer, and D. E. Woods. 2004. Contribution of
gene loss to the pathogenic evolution of Burkholderia pseudomallei and
Burkholderia mallei. Infect. Immun. 72:4172–4187.
48. Nielsen, H., J. Engelbrecht, S. Brunak, and G. von Heijne. 1997. Identifica-
tion of prokaryotic and eukaryotic signal peptides and prediction of their
cleavage sites. Protein Eng. 10:1–6.
49. Nierman, W. C., D. DeShazer, H. S. Kim, H. Tettelin, K. E. Nelson, T.
Feldblyum, R. L. Ulrich, C. M. Ronning, L. M. Brinkac, S. C. Daugherty,
T. D. Davidsen, R. T. Deboy, G. Dimitrov, R. J. Dodson, A. S. Durkin, M. L.
Gwinn, D. H. Haft, H. Khouri, J. F. Kolonay, R. Madupu, Y. Mohammoud,
W. C. Nelson, D. Radune, C. M. Romero, S. Sarria, J. Selengut, C. Shamblin,
S. A. Sullivan, O. White, Y. Yu, N. Zafar, L. Zhou, and C. M. Fraser. 2004.
Structural flexibility in the Burkholderia mallei genome. Proc. Natl. Acad. Sci.
50. O’Quinn, A. L., E. M. Wiegand, and J. A. Jeddeloh. 2001. Burkholderia
pseudomallei kills the nematode Caenorhabditis elegans using an endotoxin-
mediated paralysis. Cell. Microbiol. 3:381–393.
51. Pilatz, S., K. Breitbach, N. Hein, B. Fehlhaber, J. Schulze, B. Brenneke, L.
Eberl, and I. Steinmetz. 2006. Identification of Burkholderia pseudomallei
genes required for the intracellular life cycle and in vivo virulence. Infect.
52. Rainbow, L., M. C. Wilkinson, P. J. Sargent, C. A. Hart, and C. Winstanley.
2004. Identification and expression of a Burkholderia pseudomallei collage-
nase in Escherichia coli. Curr. Microbiol. 48:300–304.
53. Reckseidler, S. L., D. DeShazer, P. A. Sokol, and D. E. Woods. 2001. De-
tection of bacterial virulence genes by subtractive hybridization: identifica-
tion of capsular polysaccharide of Burkholderia pseudomallei as a major
virulence determinant. Infect. Immun. 69:34–44.
54. Redfearn, M. S., N. J. Palleroni, and R. Y. Stanier. 1966. A comparative
study of Pseudomonas pseudomallei and Bacillus mallei. J. Gen. Microbiol.
55. Rotz, L. D., A. S. Khan, S. R. Lillibridge, S. M. Ostroff, and J. M. Hughes.
2002. Public health assessment of potential biological terrorism agents.
Emerg. Infect. Dis. 8:225–230.
56. Schell, M. A., R. L. Ulrich, W. J. Ribot, E. E. Brueggemann, H. B. Hines, D.
Chen, L. Lipscomb, H. S. Kim, J. Mrazek, W. C. Nierman, and D. Deshazer.
2007. Type VI secretion is a major virulence determinant in Burkholderia
mallei. Mol. Microbiol. 64:1466–1485.
57. Simon, R., U. Priefer, and A. Puhler. 1983. A broad host range mobilization
system for in vivo genetic engineering: transposon mutagenesis in gram
negative bacteria. Bio/Technology 1:784–791.
58. Smith, M. D., V. Wuthiekanun, A. L. Walsh, and N. J. White. 1995. Quan-
titative recovery of Burkholderia pseudomallei from soil in Thailand. Trans.
R. Soc. Trop. Med. Hyg. 89:488–490.
59. Smith, M. P., T. R. Laws, T. P. Atkins, P. C. Oyston, D. I. de Pomerai, and
R. W. Titball. 2002. A liquid-based method for the assessment of bacterial
pathogenicity using the nematode Caenorhabditis elegans. FEMS Microbiol.
60. Stevens, M. P., M. W. Wood, L. A. Taylor, P. Monaghan, P. Hawes, P. W.
Jones, T. S. Wallis, and E. E. Galyov. 2002. An Inv/Mxi-Spa-like type III
protein secretion system in Burkholderia pseudomallei modulates intracellu-
lar behaviour of the pathogen. Mol. Microbiol. 46:649–659.
61. Tan, M. W. 2002. Cross-species infections and their analysis. Annu. Rev.
62. Tuanyok, A., M. Tom, J. Dunbar, and D. E. Woods. 2006. Genome-wide
expression analysis of Burkholderia pseudomallei infection in a hamster
model of acute melioidosis. Infect. Immun. 74:5465–5476.
63. Ulrich, R. L., K. Amemiya, D. M. Waag, C. J. Roy, and D. DeShazer. 2005.
Aerogenic vaccination with a Burkholderia mallei auxotroph protects against
aerosol-initiated glanders in mice. Vaccine 23:1986–1992.
64. Ulrich, R. L., and D. DeShazer. 2004. Type III secretion: a virulence factor
delivery system essential for the pathogenicity of Burkholderia mallei. Infect.
65. Vandamme, P., B. Holmes, T. Coenye, J. Goris, E. Mahenthiralingam, J. J.
LiPuma, and J. R. Govan. 2003. Burkholderia cenocepacia sp. nov.—a new
twist to an old story. Res. Microbiol. 154:91–96.
66. Warawa, J., and D. E. Woods. 2005. Type III secretion system cluster 3 is
required for maximal virulence of Burkholderia pseudomallei in a hamster
infection model. FEMS Microbiol. Lett. 242:101–108.
67. Wheelis, M. 1998. First shots fired in biological warfare. Nature 395:213.
68. Woods, D. E. 2002. The use of animal infection models to study the patho-
genesis of melioidosis and glanders. Trends Microbiol. 10:483–484. (Discus-
69. Wretlind, B., A. Bjorklind, and O. R. Pavlovskis. 1987. Role of exotoxin A
and elastase in the pathogenicity of Pseudomonas aeruginosa strain PAO
experimental mouse burn infection. Microb. Pathog. 2:397–404.
70. Zanetti, F., G. De Luca, and S. Stampi. 2000. Recovery of Burkholderia
pseudomallei and B. cepacia from drinking water. Int. J. Food Microbiol.
VOL. 190, 2008NOVEL VIRULENCE GENES OF B. MALLEI 2313