HYBRIDOMA

Volume 27, Number 4, 2008

© Mary Ann Liebert, Inc.
DOI: 10.1089/hyb.2008.0012

Relationship Between Antigenicity and Pathogenicity for

Burkholderia pseudomallei and Burkholderia mallei
Revealed by a Large Panel of Mouse MAbs

Nianxiang Zou, Shien Tsai, Shaw-Huey Feng, Tamara Newsome, Hyung-Yong Kim, Bingjie Li,
Shimin Zhang, and Shyh-Ching Lo

Abstract

Burkholderia pseudomallei and Burkholderia mallei are two closely related gram-negative bacterial species classi-
fied by the CDC as category B biowarfare agents. To develop monoclonal antibodies (MAbs) that can recog-
nize as many different strains and/or clinical isolates of these two pathogens as possible, we immunized mice
with heat-killed bacterial whole cells and membrane preparations from multiple strains and/or clinical isolates
of B. pseudomallei and B. mallei. More than 100 different hybridoma clones that produced MAbs strongly react-
ing to B. pseudomallei and/or B. mallei have been developed. These MAbs were categorized into eight different
groups according to their reaction specificity against different species of Burkholderia bacteria as well as the
different nature of target antigens (LPS, capsule polysaccharides, proteins, and glycoproteins) on the bacteria
they recognized. Characterization of this large panel of MAbs has demonstrated an interesting pattern of var-
ious antigenic epitopes shared by the different species of Burkholderia bacteria. More importantly, this study
has revealed a pathogenicity-linked antigen epitope(s) on capsule-like polysaccharides found only in the path-
ogenic species of Burkholderia bacteria and a Burkholderia-specific antigen epitope(s) that did not exist in other
gram-negative bacterial species. Our MAbs should prove to be highly valuable in the development of detec-
tion, diagnosis, and therapeutic applications against B. mallei and B. pseudomallei infections.

Introduction However, an earlier study revealed B. mallei could be highly

infectious for humans through the aerosol route.(5) Thus,

B URKHOLDERIA MALLEI AND BURKHOLDERIA PSEUDOMALLEI are

two highly related gram-negative pathogenic bacteria.
Infection by B. pseudomallei causes melioidosis and is often
due to either direct inoculation into wounds and skin abra-
sions or to inhalation of contaminated materials.(1–3) This
would explain the prevalence of the disease among rice farm-
ers as well as helicopter pilots in the Vietnam War who de-
veloped melioidosis due to inhalation of contaminated
dust.(2) The clinical manifestation of melioidosis ranges from
subclinical to acute localized, acute septicemic, and chronic
forms. In recent years, it has been recognized as a major cause
of community-acquired septicemia, resulting in significant
mortality in Thailand and other parts of the world.(2) The
zoonotic disease glanders is caused by B. mallei that do not
normally exist in nature outside of its adapted horse host.(4)
Humans whose occupations put them in close contact with
infected animals would more likely contract the disease.
there is a true concern that B. mallei and B. pseudomallei may
potentially be misused as biological warfare agents.
Both the chronic and subclinical forms of B. mallei and B.
pseudomallei infections could remain undiagnosed until acti-
vated later by a traumatic event or a decrease in host im-
munocompetence.(6) No effective methods of prevention of
either melioidosis or glanders currently exist. Moreover, de-
spite aggressive regimens with antibiotics to treat the infec-
tions, the mortality rate of meliodosis and glanders remains
very high. Thus, it is important to develop B. mallei and B.
pseudomallei specific MAbs for rapid detection and early di-
agnosis. These MAbs could also potentially provide effective
alternative or adjunctive therapies against infections of these
deadly pathogens. Mouse monoclonal antibodies (MAbs)
specific to the capsular polysaccharide, lipopolysaccharide
(LPS), or proteins of B. pseudomallei were developed and
studied for their passive protection efficacy against B. pseudo-

Department of Environmental and Infectious Disease Sciences, American Registry of Pathology, Armed Forces Institute of Pathology,
Washington, D.C.

231
232 ZOU ET AL.

mallei infection using a mouse model.(7) Mouse MAbs spe-
cific to the B. pseudomallei LPS were evaluated for specific
immunological detection and identification of B. pseudoma-
llei.(8–10) However, many MAbs developed apparently have
difficulties in distinguishing LPS between closely related
Burkholderia species, especially between B. pseudomallei and
B. mallei or between B. pseudomallei and B. thailandensis, a
closely related non-pathogenic Burkholderia species.
We previously reported our study using a three-step cross-
screening protocol to develop mouse MAbs by immunizing
mice with the ATCC type strains of B. mallei (23344) and B.
pseudomallei (23343).(11) We selected a total of 10 hybridomas
producing LPS-binding MAbs reacting specifically with B.
mallei (23344 strain) but not with B. pseudomallei. In addition,
we selected a total of 14 hybridomas that produced MAbs
reacting with a high molecular weight (MW) protein anti-
gen(s) present only in B. pseudomallei strain ATCC 23343.
However, this high MW protein antigen was not present in
any other strains of B. pseudomallei tested. Historically de-
velopment of MAbs to effectively differentiate B. pseudoma-
llei from B. mallei or from B. thailandensis was difficult due to
their close antigenic similarity. On the other hand, it may
also be important to note that there are many different clin-
ical and environmental isolates of B. pseudomallei and B.
mallei. Different strains of bacteria within the same species
of B. pseudomallei or B. mallei apparently have a varying de-
gree of antigenic variations. Development of MAbs that ef-
fectively differentiate between B. pseudomallei and B. mallei
and also recognize all strains isolated from different geo-
graphic regions in the world will likely be even more chal-
lenging.
We have continued our effort of developing B. pseudoma-
llei and B. mallei specific MAbs recognizing or reacting to as
many different strains of the respective pathogenic species
of bacteria as possible. We immunized mice with heat-killed
whole bacterial cells and membrane preparations from three
different strains/clinical isolates of B. pseudomallei and two
different strains/clinical isolates of B. mallei. Hybridomas
producing MAbs reacting to B. pseudomallei and B. mallei
were selected from 10 independent fusions. The MAbs were
studied for their reaction specificity against 30 different
strains and/or clinical isolates of B. mallei and B. pseudoma-
llei in our possession as well as against nine non-pathogenic
Burkholderia species. We also characterized the nature of the
target molecules recognized by these MAbs using Western
blot analysis. The MAbs were classified into eight different
groups according to their specificity against different Burk-
holderia species, strains, and/or clinical isolates as well as
the nature of the bacterial antigens they reacted to. The
unique antigen(s) of capsule-like polysaccharides found
specifically in the pathogenic species of Burkholderia bacte-
ria, as well as antigenic epitopes found on LPS, glycopro-
teins, and proteins shared at different levels among differ-
ent species of Burkholderia bacteria, are discussed.
Materials and Methods
Bacterial strains and cultural conditions
B. mallei, B. pseudomallei, and other bacterial species used
in this study are listed in Table 1. Each species and strain/iso-
late of Burkholderia bacterium was confirmed by sequenc-
ing the 16S rRNA gene. Pathogenic strains of B. mallei and
B. pseudomallei were grown in Luria-Bertani medium at 37°C
with shaking overnight in the biosafety level 3 (BSL-3) facil-
ity. Bacterial cells were collected by centrifugation at 4,000
g  30 min and resuspended in 1/10 volume of PBS. The
bacteria in PBS were killed by heating at 65°C for 90 min be-
fore removal from the BSL-3 facility. The non-pathogenic
bacterial species were grown and processed in BSL-2 lab un-
der the same conditions. Each whole bacterial sample was
quantified by total protein concentration with DC Protein
Assay Reagents (Bio-Rad Laboratories, Hercules, CA) to fa-
cilitate future analysis.
Antigen preparation and hybridoma development
Heat-killed whole organisms and lipid-associated mem-
brane preparations (LAMPs) from B. pseudomallei and B.
mallei were used as antigens to immunize mice. To prepare
LAMPs, a heat-killed bacterial pellet from 50 mL of overnight
culture was suspended in 5 mL of Tris-buffered saline (TBS),
solubilized by adding TX-114 to a final concentration of 2%,

TABLE 1. BACTERIAL SPECIES AND STRAINS/ISOLATES USED IN THIS STUDY

B. pseudomallei B. mallei Non-pathogenic Burkholderia Other gram-negative bacteria

ATCC 23343 ATCC 23344 B. thailandensis (ATCC 700388) P. aeruginosa (ATCC15152)

AFIP BP1 AFIP BM 1 B. cepacia (ATCC 70070) E. coli XL1 Blue
AFIP BP2 AFIP BM 2 B. vietnamiensi (ATCC BAA-248)
AFIP BP3 AFIP BM 3 B. stabilis (ATCC BAA-67)
AFIP BP4 AFIP BM 4 B. ambifaria (ATCC BAA-244)
AFIP BP5 AFIP BM 5 B. caledonica (ATCC BAA-462)
AFIP BP6 AFIP BM 6 B. kurureinsis (ATCC 700977)
AFIP BP7 AFIP BM 7 B. multivorans (ATCC BAA-247)
AFIP BP8 AFIP BM 8 B. fungorum (ATCC BAA-463)
AFIP BP9 AFIP BM 9
AFIP BP10 AFIP BM 10
AFIP BP11 AFIP BM 11
AFIP BP12 AFIP BM 12
AFIP BP13 AFIP BM 13
AFIP BM 14
AFIP BM 15
ANALYSIS OF MOUSE MAbs AGAINST B. MALLEI AND B. PSEUDOMALLEI 233

and incubated at 4°C for 1 h. The solution was centrifuged SDS polyacrylamide gel electrophoresis (PAGE) and
at 8,000 g for 20 min to remove insoluble components. The immunoblot analysis
supernatant was then subjected to 37°C for 10 min. After cen-
Whole bacteria lysates were prepared by adding SDS gel
trifugation at 10,000 g for 20 min at room temperature, the
loading buffer into the cell suspension and incubating at
upper aqueous phase was removed. Cold TBS was added to
95°C for 5 min. Proteinase K (PK)-treated bacterial lysates
the lower TX-114 phase to the original volume, and incu-
were prepared by adding PK to a final concentration of 20
bated at 4°C for 10 min, and the procedures for phase frac-
g/mL into bacterial lysate and incubating for 1 h at 37°C.
tionation was repeated twice in order to obtain clean mem-
Sodium periodate (SP)-treated samples were prepared as fol-
brane fractions. The final TX-114 phase was suspended in
lows. Heat-killed bacteria were suspended in a fresh-made
cold TBS to the original volume, and 2.5 volumes of ethanol
solution containing 20 mM sodium periodate and 50 mM so-
was added to precipitate membrane components at 20°C
dium acetate (pH 4.5). The suspension was incubated at
overnight. After centrifugation, the pellet was suspended in
room temperature for 60 min. The bacteria were pelleted at
5 mL of 0.2% SDS in PBS and used as the LAMP fraction.
14,000 rpm for 2 min to remove supernatant. The pelleted
BALB/c mice were immunized according to the following
bacteria were washed once with 50 mM sodium acetate (pH
two procedures: (1) five times (i.p.) at 2-week intervals with
4.5) and incubated with 50 mM NaBH4 in PBS for 30 min at
heat-killed whole bacteria containing 107 cfu in 0.2 mL of
room temperature. The bacterial cells were pelleted and re-
PBS mixed with 0.2 mL of the MPL  TDM adjuvant sys-
suspended with 1 SDS sample buffer. Samples were sepa-
tems (Sigma Chemicals, St. Louis, MO); (2) first immunized
rated by 10% SDS-polyacrylamide gel (14 cm  12 cm  0.75
(i.p.) with 0.2 mL of the LAMP/LPS fraction (equivalent to
cm) electrophoresis and transferred onto nitrocellulose mem-
5  106 cfu) mixed with 0.4 mL of CpG containing the mon-
branes (Bio-Rad). The membranes were blocked with 1%
tanide adjuvant system (kindly provided by Dr. James Shih
BSA in PBS and incubated with designated monoclonal an-
at NIH) and followed by four boosts with 0.2 mL of the
tibodies diluted 1:5 in 1% BSA in PBS either overnight at 4°C
LAMP/LPS fraction mixed with 0.2 mL of the MPL  TDM
or for 2 h at room temperature.
Adjuvant systems.
The membranes were washed five times with PBS. HRP-
After the fourth immunization, mice usually had titers
labeled goat-anti-mouse IgAIgGIgM (HL) was used as
ranging between 1:50,000 and 1:100,000. They received a fi-
secondary antibody. The secondary antibody was diluted in
nal boost 3 days before sacrifice for spleen harvest. Mixed
1% BSA in PBS (1:2000) and incubated with the membranes
spleen cells and myeloma cells (P3X63Ag8.653) were fused
at room temperature for 2 h. The membranes were washed
in the presence of 50% polyethylene glycol (PEG MW 1500).
five times with PBS and antigen antibody reactions were re-
Cells were cultured in selecting medium containing hypox-
vealed by 4CN substrate (Kirkegaard & Perry). To prepare
anthine, aminopterin, and thymidine in 96-well culture
immunostrips, one single sample in the amount equivalent
plates.(12)
to the original whole protein level of 100 g was loaded on
a 10% SDS gel with a flat surface. After electrophoresis and
Hybridoma screening and characterization of antigen transfer of separated antigens to the NC membrane, the
specificity by enzyme-linked immunosorbent assay membrane was block with 1% BSA in PBS and cut into im-
munostrips. Immunoblot analysis of intact membranes was
Supernatants from hybridoma lines were screened against
done in a similar fashion.
heat-killed whole bacteria or LAMPs used as the immuno-
gen for mouse immunization. Positive lines were further
Results
tested for cross-reactions with other strains and/or clinical
isolates of B. mallei and B. pseudomallei as well as other Burk- Development of hybridomas producing MAbs against
holderia species, as shown in Table 1. 100 L of selected heat- B. mallei and/or B. pseudomallei
killed bacteria were suspended in PBS at 10 ng/mL and
To obtain hybridomas producing MAbs that could react
added to each well and incubated at room temperature for
with as many different strains or clinical isolates of B. mallei
2 h. The wells were washed once with PBS and overcoated
and B. pseudomallei as possible, we immunized mice sepa-
with 1% BSA in PBS for 2 h. Primary antibody was added to
rately with either whole cells or membrane preparations
each well at 100 L/well and incubated for 2 h at room tem-
from two different strains of heat-killed B. mallei (ATCC
perature. ELISA plates were washed with PBST five times
23344, AFIP BM4) and three different strains of B. pseudo-
with an EL403H Microplate autowasher (Bio-Tek Instru-
mallei (ATCC 23343, AFIP BP2, AFIP BP5). All together 108
ments, Winooski, VT). Secondary antibody interaction was
clones of stable hybridomas that produced MAbs reacting
done by adding to each well 100 L of horse radish peroxi-
strongly to whole cell lysates of B. mallei and/or B. pseudo-
dase conjugated goat anti-mouse IgAIgGIgM (HL)
mallei were developed from a total of 10 fusions using spleen
(Kirkegaard & Perry Lab, Gaithersburg, MD) diluted at
cells from the hyperimmuned mice.
1:5000 in PBS1% BSA. After incubation for 1 h at room tem-
perature and washing with PBST, 100 L of 2,2-azino-bis(3-
Grouping of MAbs by specificity to different species or
ethyl-benzthiazoline-6-sulphonic acid (ABTS) peroxidase
strains of Burkholderia bacteria
substrate solution was added to each well. After incubation
at room temperature for 1 h, the absorbance was read at 405 B. mallei- or B. pseudomallei-reacting MAbs produced by
nm on a V max kinetic microplate reader (Molecular Devices, the selected hybridomas were first characterized for their re-
Sunnyvale, CA). Isotypes of each MAb were determined by activity to the following bacterial species and/or strains as
the Mouse MonoAB ID kit (Invitrogen, Carlsbad, CA) fol- the target antigens in ELISA: 15 different strains and/or iso-
lowing the procedures provided. lates of B. mallei in our possession plus the ATCC type strain
234
23344; 13 different strains and/or isolates of B. pseudomallei
in our possession plus the ATCC type strain 23343; 9 non-
pathogenic species of Burkholderia bacteria, Pseudomonas
aeruginosa, and Escherichia coli (Table 1).
According to the reaction specificity to the whole cell
lysates of different isolates and different species of Burk-
holderia bacteria in ELISA, we could effectively classify the
MAbs into eight groups, groups A to H (Table 2). All MAbs
in group A were developed by immunizing mice with either
heat-killed whole bacteria or the membrane preparation
(LAMPS) of the particular ATCC 23343 type strain of B.
pseudomallei as described in our previous study.(11) We
showed MAbs of group A could only react to the immuniz-
ing strain (ATCC 23343) and not to any other strains and/or
clinical isolates of B. pseudomallei. This group of MAbs also
failed to recognize any other species of Burkholderia bacte-
ria tested. Group B MAbs were separately developed from
mice immunized by the heat-killed whole bacteria of B.
pseudomallei strains AFIP BP2, AFIP BP5, and B. mallei strain
AFIP BM4 in five separate immunization procedures. Every
MAb in group B reacted positively to all 16 different strains
of B. mallei and all 14 different strains of B. pseudomallei tested.
These MAbs did not react to any of the other nine non-path-
ogenic species of Burkholderia bacteria tested, including B.
thailandensis (Table 2). As expected, they also did not react
to non-Burkholderia bacteria P. aeruginosa and E. coli.
All 12 MAbs in group C reacted specifically to B. mallei.
All of group C MAbs failed to react with three specific B.
mallei isolates, AFIP BM4, AFIP BM11, and AFIP BM15. Ten
of the 12 MAbs in group C were developed using the B. mallei
strain ATCC 22344 membrane preparations as immunizing
antigens. Two MAbs in group C (BMW 2D5 and BM1 3G11)
were developed using the B. mallei strain ATCC 22344 whole
bacteria as immunizing antigens.
All MAbs in group D were developed in subsequent at-
tempts to obtain MAbs reactive to the three B. mallei isolates
not recognized by the group C antibodies generated from
the heat-killed whole bacteria. These group D MAbs could
indeed react strongly with all 16 different strains of B. mallei
tested, including isolates AFIP BM4, AFIP BM11, and AFIP
BM15. However, they also reacted weakly to 2–4 B. pseudo-
mallei isolates (Table 2). Like MAbs of group C, MAbs of
group D did not react with the nine non-pathogenic species
of Burkholderia bacteria tested, including B. thailandensis.
MAbs in groups E and F were developed by immunizing
mice with the whole cells of heat-killed bacteria or LAMPs
from two isolates of B. pseudomallei (strains AFIP BP2 and
AFIP BP5). Group E MAbs reacted positively to all 14 dif-
ferent strains of B. pseudomallei tested as well as the non-path-
ogenic B. thailandesis (Table 2). These MAbs did not react
with any of the 16 different strains of B. mallei and 8 other
non-pathogenic Burkholderia species tested. In comparison,
MAbs in group F reacted positively to all 16 isolates of B.
mallei and all 14 different strains of B. pseudomallei and B.
thailandesis, but would not react with any of the other eight
non-pathogenic species of Burkholderia bacteria tested. As
expected, they also did not react to P. aeruginosa and E. coli.
MAbs in groups G and H were developed from fusions of
spleen cells from mice immunized by various preparations
of different strains of B. mallei and B. pseudomallei as immu-
nizing antigens. Typically, MAbs of these two groups cross-
reacted to many other non-pathogenic species of Burk-
ZOU ET AL.
holderia bacteria (Table 2). Group G turned out to be the
largest group of MAbs obtained in this study. All MAbs in
group G reacted positively to all 16 different strains of B.
mallei, all 14 different strains of B. pseudomallei and eight non-
pathogenic species of Burkholderia bacteria tested, but none
of the group G MAbs reacted with B. thailandensis (Table 2).
MAbs of group G did not react to non-Burkholderia bacte-
ria P. aeruginosa or E. coli. In comparison, MAbs in group H
reacted positively to all of the species of Burkholderia bac-
teria tested, including B. thailandensis. They also reacted to
non-Burkholderia bacteria P. aeruginosa and weakly to E. coli.
There were 16 MAbs that could not be classified into the
above eight groups. We later found that they either reacted
to single protein bands or conformational epitopes related to
only certain strains or species. These 16 MAbs are not dis-
cussed here. Immunoglobulin subtypes were determined for
each MAbs to assure the clonality of hybridomas. The heavy
chains included subtypes of 1, 2a, 2b, 3, , and A, while
the light chains included both  and  (Table 2).
Characterization of antigens recognized by each group of
MAbs using Western blot analysis
Western blot analysis was performed for all of the MAbs
we obtained to confirm the ELISA results and to character-
ize the nature of antigens recognized by them. Characteri-
zation of the nature of target molecules recognized by MAbs
of group A was described previously.(11) All MAbs of group
A reacted with proteinase K-sensitive high MW protein(s)
found only in the immunizing strain B. pseudomallei ATCC
23344, but not in any other isolates of B. pseudomallei or other
Burkholderia species. In comparison, MAbs of group B re-
acted to high MW, proteinase K-resistant target molecule(s)
(Fig. 1). We also pre-treated the bacterial antigens with so-
dium periodate, an oxidizing reagent that disrupts carbohy-
drate moieties.
The antigenic epitopes recognized by group B MAbs
showed a varying degree of sensitivity to treatment of so-
dium periodate. Figure 1 reveals the representative Western
blot (WB) analysis result using group B MAb BP1 1E4. The
high MW antigenic epitope(s) were partially affected by a 1
h sodium periodate treatment and abolished completely
with a longer (3 h) treatment with sodium periodate. We con-
sidered the proteinase K resistant high MW molecule(s) with
slight aberrant migration and varying degrees of sodium pe-
riodate sensitivity capsule-like surface polysaccharides.
All MAbs of group C shared the same reaction character-
istic and produced a highly similar WB reaction profile. Fig-
ure 2 shows the representative WB results of group C MAbs
using BML 5D11 as an example, against samples prepared
from B. mallei (BM), B. pseudomallei (BP), and B. thailandensis
(BT). Western blot study revealed group C MAbs reacted
strongly with B. mallei, but only minimally with B. pseudo-
mallei or B. thailandensis. The antigenic epitopes recognized
by BML 5D11 were clearly resistant to both pretreatments of
proteinase K and sodium periodate. The typical LPS ladder
pattern, seen between molecular weights 20 and 90 kDa, be-
came even more prominent when the whole bacteria anti-
gens were treated with proteinase K. The Western blot pat-
tern produced by BMW2 14B2 of group D was essentially
identical to the pattern produced by group C MAb BML
5D11 (data not shown). Consistent with the ELISA result,
 TABLE 2. GROUPING OF MOUSE MONOCLONAL ANTIBODIES AGAINST BURKHOLDERIA

MAb groups WB resultsa ELISA resultsb

A BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli

BP E19A2 IgG2a, K High MW proteins P N N N N N N
BPL 12H7 IgG1, K High MW proteins P N N N N N N
BPL 17D10 IgG3, K High MW proteins P N N N N N N
BP Z3D2 IgG2a, K High MW proteins P N N N N N N
BP Z5D6 IgG2a, K High MW proteins P N N N N N N
BP Z10B3 IgG2a, K High MW proteins P N N N N N N
BP Z11F3 IgG2a, K High MW proteins P N N N N N N
BP M29E3 IgG2a, K High MW proteins P N N N N N N
BP M8C11 IgG2a, K High MW proteins P N N N N N N
BPL 15D8D7 IgG1, K High MW proteins P N N N N N N
BPL 19H10F6 IgG1, K High MW proteins P N N N N N N
BP 195 IgG2a, K High MW proteins P N N N N N N
B WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
BP1 1E4 IgM, K Capsule polysaccharides P P P N N N N
BP7 1C10 IgM, K Capsule polysaccharides P P P N N N N
BP1 14F4 IgM, K Capsule polysaccharides P P P N N N N
BP2 D2 IgG2a, K Capsule polysaccharides P P P N N N N
BP1 6A8 IgM, K Capsule polysaccharides P P P N N N N
BP1 10F11 IgG3, K Capsule polysaccharides P P P N N N N
BP7 3E10 IgG1, K Capsule polysaccharides P P P N N N N
BP2 D17 IgG1, K Capsule polysaccharides P P P N N N N
BP7 2F4 IgG2b, K Capsule polysaccharides P P P N N N N
BP2 I67 IgG1, K Capsule polysaccharides P P P N N N N
BMW2 1B5 IgM, K Capsule polysaccharides P P P N N N N
BP1 9G2 IgG3, K Capsule polysaccharides P P P N N N N
BP7 4G9 IgM, K Capsule polysaccharides P P P N N N N
BP7 2C6 IgG2a, K Capsule polysaccharides P P P N N N N
BP1 2E7 IgG3, K Capsule polysaccharides P P P N N N N
BP1 14A7 IgG3, K Capsule polysaccharides P P P N N N N
BPW5 1D9 IgM, K Capsule polysaccharides P P P N N N N
BP7 7B2 IgM, K Capsule polysaccharides P P P N N N N
C WB results ELISA results
BP ATCC 13 BPs 13/16 BMs* BT 8 BK sp. P. aeruginosa E. coli
BML 7F10 IgA, L LPS N N P N N N N
BML 5D11 IgG2b, L LPS N N P N N N N
BML 18F8 IgG1, L LPS N N P N N N N
BML 20F1 IgG1, L LPS N N P N N N N
BML 8C8 IgG1, L LPS N N P N N N N
BML 15G9 IgG1, L LPS N N P N N N N
BML 4C10 IgG1, L LPS N N P N N N N
BML 11G6 IgG1, L LPS N N P N N N N
BML 11H8 IgG1, L LPS N N P N N N N
BM1 3G11 IgG3, K LPS N N P N N N N
BML1 3F11 IgG2a, K LPS N N P N N N N
BMLW 2D5 IgG1, L LPS N N P N N N N
D WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
BMW2 11F2 IgG1 LPS N (2/11) P N N N N
BMW2 14B2 IgG3, L LPS N (4/9) P N N N N
BMW2 3C8 IgG3, L LPS N (2/11) P N N N N
BMW2 6C12 IgG3, L LPS N (2/11) P N N N N
BMW2 14H3 IgG3, L LPS N (2/11) P N N N N
BMW2 9C8 IgG3, L LPS N (2/11) P N N N N
E WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
BP7 6G2 IgM, L LPS P P N P N N N
BP L30D11 IgG2a, K LPS P P N P N N N
BP7 2G6 IgG2a, L LPS P P N P N N N
BP1 7F7 IgG3, K LPS P P N P N N N
BP L5F7 IgG3, K LPS P P N P N N N
BP A2 IgG1, K LPS P P N P N N N
BP7 1H7 IgG1, L LPS P P N P N N N
F WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
(continued)
236 ZOU ET AL.

TABLE 2. GROUPING OF MOUSE MONOCLONAL ANTIBODIES AGAINST BURKHOLDERIA (CONT’D)

MAb groups WB results ELISA resultsb

BP4 10C9 IgM, L LPS P P P P N N N

BP4 17G6 IgM, L LPS P P P P N N N
BP4 5D7 IgM, L LPS P P P P N N N
BP1 14H2 IgM, L LPS P P P P N N N
BP1 1H4 IgM, K LPS P P P P N N N
G WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
BPL 19B12 IgG1, K Glycoproteins P P P N P N N
BMW6 1E7 IgG2B, K Glycoproteins P P P N P N N
BPL 23E10 IgG2B, K Glycoproteins P P P N P N N
BMW2 13A12 IgM, K Glycoproteins P P P N P N N
BMW 3H8 IgG1, K Glycoproteins P P P N P N N
BMW5 4E11 IgM, K Glycoproteins P P P N P N N
BMW5 5A1 IgM, K Glycoproteins P P P N P N N
BML 9E2 IgG1, K Glycoproteins P P P N P N N
BMW2 1C4 IgM, K Glycoproteins P P P N P N N
BMW5 3B9 IgA, K Glycoproteins P P P N P N N
BMW5 5H5 IgM, K Glycoproteins P P P N P N N
BMW5 7C10 IgA, K Glycoproteins P P P N P N N
BPL 23F11 IgG3, K Glycoproteins P P P N P N N
BP7 10B11 IgG1, K Glycoproteins P P P N P N N
BMW5 2C8 IgM, K Glycoproteins P P P N P N N
BP7 1C3 IgM, K Glycoproteins P P P N P N N
BM1 1A5 IgA, K Glycoproteins P P P N P N N
BPL 17A2 IgG2b, K Glycoproteins P P P N P N N
BP7 3B10 IgG1, K Glycoproteins P P P N P N N
BP7 10D5 IgG2b, K Glycoproteins P P P N P N N
BMW2 11A5 IgM, K Glycoproteins P P P N P N N
BMW5 1G12 IgM, K Glycoproteins P P P N P N N
BM1 6G4 IgG2a, K Glycoproteins P P P N P N N
H WB results ELISA results
BP ATCC 13 BPs 16 BMs BT 8 BK sp. P. aeruginosa E. coli
BML 1F3 IgG1, K Proteins, multiple bands P P P P P P weak
BP2 C19 IgG1, K Proteins, multiple bands P P P P P P weak
BP1 16H3 IgM, K Proteins, multiple bands P P P P P P weak
BMW2 12H10 IgG2a, K Proteins, multiple bands P P P P P P weak
BM2 2D12 IgM, K Proteins, multiple bands P P P P P P weak
BM2 9D12 IgM, K Proteins, multiple bands P P P P P P weak
BM2 17G6 IgM, K Proteins, multiple bands P P P P P P weak
BP4 18H7 IgM, K Proteins, multiple bands P P P P P P weak
BMW2 5H8 IgG2a, K Proteins, multiple bands P P P P P P weak
aWB results: the nature of antigenic epitopes recognized by each MAb was characterized by Western blot analysis against whole bacterial

cell lysates with or without proteinase K treatment (against protein epitopes), with or without sodium periodate treatment (against carbohy-
drate epitopes).
bELISA results: ELISA assays were done against lysates of specified bacterial species and strains/isolates. BP ATCC indicates B. pseudmallei

ATCC 23344 strain. 13 BPs indicates 13 B. pseudomallei strains/isolates tested. 16 BMs indicates B. mallei ATCC 23343 strain and 15 B. mallei
strains/isolates. BT indicates B. thailandensis (ATCC 700388). 8 BK sp indicates eight non-pathogenic Burkholderia species tested: B. cepacia
(ATCC 700070), B. stabilis (ATCC BAA-244), B. vietnamiensis (ATCC BAA-248), B. ambifaria (ATCC-BAA-244), B. multivorans (ATCC BAA-
247), B. caledonica (ATCC BAA-462), B. fungorum (ATCC BAA-463), and B. kururiensis (ATCC 700977). 13/16 BMs: all 16 strains of B. mallei
tested, except strains AFIP BM4, AFIP BM11, and AFIP BM15.

Western blot study revealed group C MAb BML5D11 could
not and group D MAb BMW2 14B2 could react to the LPS
of B. mallei isolates AFIP BM4, AFIP BM11, and AFIP BM15.
Western blot analysis was similarly performed for groups
E-H. Data from immunostrips again confirmed the reaction
specificity of these MAbs from the ELISA results presented
in Table 2. MAbs of group E reacted positively to all 14 dif-
ferent strains of B. pseudomallei as well as to B. thailandensis,
but not to any strain of B. mallei tested. On the other hand,
MAbs of group F reacted positively to all of the 14 different
strains of B. pseudomallei, B. thailandensis, as well as to all 16
different strains of B. mallei tested. Figure 3 shows the reac-
tion patterns of two representative MAbs from groups E and
F. MAb BP7 6G2 of group E reacted to a ladder pattern LPS,
with or without proteinase K treatment, of B. pseudomallei
and B. thailandensis (lanes BP PK, BT PK). It clearly would
not react with the samples from B. mallei treated in parallel
(Fig. 3, left panel). MAb BP4 17G6 of group F reacted to a
ladder-pattern LPS antigen(s) in proteinase K-treated and
un-treated cell lysates of B. pseudomallei, B. mallei, and B.
thailendensis (Fig. 3, right panel). Both MAbs still reacted pos-
itively to the sodium periodate treated samples, indicating
ANALYSIS OF MOUSE MAbs AGAINST B. MALLEI AND B. PSEUDOMALLEI
FIG. 1. Western blot analysis and characterization of tar-
get molecules recognized by MAbs of Group B. Bacteria
lysate of B. pseudomallei (AFIP BP2 strain) (BP) was treated
with proteinase K (PK) or sodium periodate (SP, 1 or 3 h),
separated by SDS-PAGE, transferred onto nitrocellulose
membrane, which was then cut into strips. Group B MAb
(BP1 1E4) from hybridoma culture supernatant was diluted
at 1:10 in PBS1% BSA and incubated with membrane strips
at 4°C overnight.
the breaking of sugar structure did not abolish the antibody-
antigen interaction (Fig. 3, lanes BM-SP, BP-SP, and BT-SP).
Using immunostrips, we found all MAbs in group G pro-
duced multiple bands with all of the different species of
Burkholderia bacteria tested, except B. thailandensis, consis-
tent with the ELISA results. MAb BPL 23E10 was used in
Western blot as an example for group G MAbs (Fig. 4, left
panel). It reacted strongly to multiple bands between MW
15 and 40 kDa from the whole cell lysates of both B. mallei
and B. pseudomallei (lanes BM WC and BP WC), but not from
B. thailandensis (lane BT WC). Treatment with proteinase K
or sodium periodate abolished the antibody-antigen reaction
(lanes BM PK, BM SP, and BP PK, BP SP). MAbs of group G
appeared to recognize antigen molecule(s) containing both
protein and carbohydrate moieties, most likely glycopro-
tein(s).(13)
MAbs of group H reacted with lysates of different species
of Burkholderia bacteria producing multiple bands with a
237
wide molecular weight range in Western blot. Figure 4 (right
panel) shows the representative WB reaction of group H
MAbs, using BMW2 5H8 as an example. It recognized a ma-
jor band at MW 80 kDa as well as other minor bands. These
target molecule(s) recognized by the MAbs of group H were
clearly protein in nature, very sensitive to treatment with
proteinase K (lanes BM PK, BP PK, and BT PK), but resis-
tant to treatment with sodium periodate (lanes BM SP, BP
SP, and BT SP). The protein-related antigenic epitope(s) were
present in all of the species, pathogenic and non-pathogenic,
of Burkholderia bacteria as well as P. aeruginosa. Very weak
cross-reaction with proteins of E. coli could also be observed
for the group H MAbs (data not shown).
Discussion
Our laboratory developed B. pseudomallei- and B. mallei-
specific MAbs. For diagnostic use, these MAbs should dif-
ferentiate between two closely related pathogenic species of
Burkholderia. However, if the MAbs were developed for im-
mune therapeutics against either B. mallei or B. pseudomallei,
the MAbs reacting strongly to both B. mallei and B. pseudo-
mallei would be highly desirable. It is also important to note
there are many different clinical isolates or isolates of B.
pseudomallei and B. mallei. These clinical isolates within ei-
ther of the species could have varying degrees of antigenic
variations. Thus, for practical applications, development of
MAbs reacting to as many different strains/clinical isolates
of B. mallei and/or B. pseudomallei would also be critical. In
our previous attempt,(11) none of the B. mallei- or B. pseudo-
mallei-specific MAbs developed reacted with all the different
Burkholderia clinical isolates of these pathogenic species of
bacteria archived in our institute’s culture collection.
In this study, we immunized mice with heat-killed whole
bacteria and membrane preparations from different strains
and/or clinical isolates of B. mallei and three different strains
or clinical isolates of B. pseudomallei, respectively. A total of
108 MAbs that reacted strongly to B. mallei and/or B. pseudo-
mallei were developed. The isotypes of all the MAbs obtained
were determined. We demonstrated that most of these MAbs
could be classified into eight distinct groups based on ywo
completely different criteria (Table 2): their reactivity to the
different clinical isolates and the different species of Burk-
holderia bacteria tested as well as the nature of antigens or
epitopes. Interestingly, MAbs that had the same reaction
specificity against the different isolates or species of Burk-
holderia bacteria tested (the first grouping criteria) were re-
acting to the same kind of antigenic molecule(s) (the second
grouping criteria) from specific species of Burkholderia bac-
teria. Thus, MAbs of a particular group were reacting specif-
ically to the similar antigenic epitopes present on the same
target molecules (e.g., LPS, capsules, proteins, or glycopro-
teins) found in certain specific strains or species of Burk-
holderia bacteria.
We defined the MAb-targeted antigen molecules as pro-
teins in nature when the epitopes were sensitive to pre-
treatment of the bacterial lysates with proteinase K but not
sodium periodate (groups A and H, Table 2). The MAb-tar-
geted antigen molecules were considered to be glycoproteins
in nature when the epitopes were sensitive to both proteinase
K and sodium periodate (group G, Table 2). In addition, gly-
cosylated proteins destroyed by periodic acid oxidation
238 ZOU ET AL.

FIG. 2. Western blot analysis and characterization of target molecules recognized by MAbs of Group C. Left three lanes
contained bacterial antigens of B. mallei strain ATCC 23344 (BM). Middle three lanes contained bacterial antigens of B.
pseudomallei strain AFIP BP2 (BP). Right three lanes contained bacterial antigens of B. thailandensis strain ATCC 700388 (BT).
WC, whole bacterial lysate; PK, proteinase K-treated bacterial lysate; SP, sodium periodate-treated (1 h) bacterial lysate.
Group C MAb (BML 5D11) from hybridoma culture supernatant was diluted at 1:10 in PBS1% BSA and incubated with
WB membranes at 4°C overnight.

would often exhibit an aberrant migration on SDS-PAGE gels
(1) (Fig. 4, left panel). In the Western blot study, the antigen
molecules that were highly resistant to both proteinase K and
sodium periodate showed a characteristic ladder pattern
(groups C-F, Table 2) typical of gram-negative bacterial
LPS.(14) By contrast, the high MW antigen molecules resis-
tant to proteinase K but somewhat sensitive to the prolonged
treatment of sodium periodate did not produce characteris-
tic LPS ladder pattern in Western blot. These molecules were
believed to be capsule-like polysaccharides in nature (group
B, Table 2). All of the LPS epitopes recognized by MAbs in
groups C, D, E, and F were highly resistant to proteinase K

FIG. 3. Western blot analysis and characterization of target molecules recognized by MAbs of groups E and F. Left three
lanes contained bacterial antigens of B. mallei strain ATCC 23344 (BM). Middle three lanes contained bacterial antigens of
B. pseudomallei strain AFIP BP2 (BP). Right three lanes contained bacterial antigens of B. thailandensis strain ATCC 700388
(BT). WC, whole bacterial lysate; PK, proteinase K-treated bacterial lysate; SP, sodium periodate-treated (1 h) bacterial
lysate. Group E MAb (BP7 6G2) in the left panel and Group F MAb (BP4 17G6) in the right panel were from the respec-
tive hybridoma culture supernatants diluted at 1:10 in PBS1% BSA and incubated with WB membranes at 4°C overnight.
ANALYSIS OF MOUSE MAbs AGAINST B. MALLEI AND B. PSEUDOMALLEI 239

FIG. 4. Western blot analysis and characterization of target molecules recognized by MAbs of groups G and H. Left three
lanes contained bacterial antigens of B. mallei strain ATCC 23344 (BM). Middle three lanes contained bacterial antigens of
B. pseudomallei strain AFIP BP2 (BP). Right three lanes contained bacterial antigens of B. thailandensis strain ATCC 700388
(BT). WC, whole bacterial lysate; PK, proteinase K-treated bacterial lysate; SP, sodium periodate-treated (1 h) bacterial
lysate. Group G MAb (BPL 23E10) in the left panel and group H MAb (BMW2 5H8) in the right panel were from the re-
spective hybridoma culture supernatants diluted at 1:10 in PBS1% BSA and incubated with WB membranes at 4°C
overnight.

and to sodium periodate treatment. There were two possi-
ble explanations. (1) If the epitopes reside on the poly-
saccharides of the LPS, the polysaccharides were apparently
different from those carbohydrates in capsules or in glyco-
proteins for their different sensitivity to oxidizing disruption
by sodium periodate. (2) The MAbs were reacting to the epi-
topes on LPS with little carbohydrate involved.
Characterization of a large panel of randomly selected
MAbs reacting against B. mallei and B. pseudomallei has elu-
cidated some critical biological properties as well as patho-
genic and antigenic relationships among different species of
Burkholderia bacteria. The data show several protein and
glycoprotein antigens are shared among the different species
of Burkholderia bacteria. Other than the common Burk-
holderia antigenic epitopes found on the glycoproteins/pro-
teins, we demonstrated that Burkholderia uniquely share
high MW carbohydrate or polysaccharide epitope(s) recog-
nized specifically by the MAbs of group B. The unique cap-
sule-like polysaccharide epitope(s) could be found in all 30
different isolates of B. mallei and B. pseudomallei tested, but
not in any of the nine non-pathogenic Burkholderia bacteria
species (Table 2). Although the capsule-like epitopes have
not been shown to be directly associated with virulence of
Burkholderia bacteria, this epitope(s) found only in the path-
ogenic species of Burkholderia bacteria could be important
pathogenicity-associated epitopes. Similarly, B. pseudomallei
and the non-pathogenic B. thailandensis are also closely re-
lated species. Our study showed the LPS of B. pseudomallei
and B. thailandensis share highly unique epitopes recognized
by the MAbs of group E. The specific LPS epitopes could not
be found in any other species of Burkholderia bacteria tested.
Moreover, the LPS of B. mallei, B. pseudomallei, and B. thai-
landensis shared other unique antigenic epitopes recognized
specifically by the MAbs of group F that were not found in
other species of Burkholderia bacteria. All these findings are
consistent with the extremely high DNA homology among
these three species of Burkholderia bacteria.(15,16)
Two earlier studies(17,18) used subtractive hybridization of
genomic DNAs from B. pseudomalei, B. mallei, and B. thailan-
densis to reveal B. thailandensis had truncations in the “cap-
sule gene cluster” that is present in both B. mallei and B.
pseudomallei. The inactivated gene(s) was shown to be in-
volved in the production of a major surface polysaccharide.
In this study, we showed all MAbs in group B were specif-
ically reacting to a large MW molecule(s) with carbohydrate
property that was uniquely present in both B. mallei and B.
pseudomallei but missing in B. thailandensis. The high MW
molecules we tentatively termed “capsule-like polysaccha-
rides” could very well be the “major surface polysaccharide”
described by these earlier studies. In this context, we also
found another group of MAbs (group G) that reacted with
the carbohydrate moieties of glycoprotein antigen(s) present
in all the different species of Burkholderia bacteria tested,
except B. thailandensis. It is likely that in addition to being
responsible for the synthesis of a major surface polysaccha-
ride(s) in B. mallei and B. pseudomallei, the capsule gene clus-
ter lacking in B. thailandensis could also affect the synthesis
of other carbohydrate moieties in glycoproteins. It is impor-
tant to note this glycoprotein antigen(s) was clearly absent
in P. aeruginosa, a phylogenically most closely related species
to the genus of Burkholderia bacteria (Table 2). Thus, MAbs
of group G might indeed be recognizing an antigen(s) spe-
cific to all the members of Burkholderia bacteria. However,
B. thailandensis failed to produce this Burkholderia bacteria-
specific glycoprotein antigen due to truncation of genes re-
lated to glycosylation.
We have developed a large panel of MAbs with different
reaction specificity against different species of Burkholderia
bacteria. These MAbs will be good candidates for the de-
velopment of immunologic diagnosis of B. mallei and B.
240
pseudomallei infections. We believe that by using a combina-
tion of MAbs from different groups, one will be able to ef-
fectively differentiate between B. mallei and B. pseudomallei,
as well as other non-pathogenic species of Burkholderia bac-
teria. Many of these MAbs could also prove to be valuable
in the development of immune therapeutics against infec-
tions of B. mallei and B. pseudomallei. Studies are in progress
in both directions.
Acknowledgment
The authors thank Mr. Jose Rodriguez for assisting in the
preparation of the figures. This work was supported in part
by JSTO-CBD/DTRA Grant # 2.10014_05_AF_B.
References
1. Chaowagul W, White NJ, Dance DA, Wattanagoon Y,
Naigowit P, Davis TM, Looareesuwan S, and Pitakwatchara
N: Melioidosis: a major cause of community-acquired sep-
ticemia in northeastern Thailand. J Infect Dis 1989;159:
890–899.
2. Howe C, Sampath A, and Spotnitz M: The pseudomallei
group: a review. J Infect Dis 1971;11:413–425.
3. Leelarasamee A, and Bovornkitti S: Melioidosis: review and
update. Rev Infect Dis 1989;11:413–425.
4. Howe C: Glanders. In: The Oxford Medicine. Christian HA
(Ed.). Oxford University Press, New York, 1950, pp. 185–202.
5. Howe C, and Miller WR: Human glanders: report of six
cases. Ann Intern Med 1947;26:93–115.
6. Ip M, Osterberg LG, Chau PY, and Raffin TA: Pulmonary
melioidosis. Chest 1995;108:1420–1424.
7. Jones SM, Ellis JF, Russell P, Griffin KF, and Oyston PC: Pas-
sive protection against Burkholderia pseudomallei infection in
mice by monoclonal antibodies against capsular polysac-
charide, lipopolysaccharide or proteins. J Med Microbiol
2002; 51:1055–1062.
8. Ho M, Schollaardt T, Smith MD, Perry MB, Brett PJ,
Chaowagul W, and Bryan LE: Specificity and functional ac-
tivity of anti-Burkholderia pseudomallei polysaccharide anti-
bodies. Infect Immun 1997;65:3648–3653.
9. Pongsunk S, Thirawattanasuk N, Piyasangthong N, and
Ekpo P: Rapid identification of Burkholderia pseudomallei in
blood cultures by a monoclonal antibody assay. J Clin Mi-
crobiol 1999;37:3662–3667.
10. Steinmetz I, Reganzerowski A, Brenneke B, Haussler S,
Simpson A, and White NJ: Rapid identification of Burk-
holderia pseudomallei by latex agglutination based on an ex-
opolysaccharide-specific monoclonal antibody. J Clin Mi-
crobiol 1999;37:225–228.
11. Feng SH, Tsai S, Rodriguez J, Newsome T, Emanuel P, and
Lo SC: Development of mouse hybridomas for production
of monoclonal antibodies specific to Burkholderia mallei and
Burkholderia pseudomallei. Hybridoma 2006;25:193–201.
ZOU ET AL.
12. Yokoyama WM: Antibody detection and preparation. In:
Current Protocols in Immunology. Coligan JE, Kurisbeek
AM, Margulies DH, Shevach EM, and Strober W (Eds.). John
Wiley & Sons, New York, 1991.
13. Benz I, and Schmidt MA: Never say never again: protein
glycosylation in pathogenic bacteria. Mol Microbiol 2002;45:
267–276.
14. Raetz CR, and Whitfield C: Lipopolysaccharide endotoxins.
Annu Rev Biochem 2002;71:635–700.
15. Holden MT, Titball RW, Peacock SJ, Cerdeño-Tárraga AM,
Atkins T, Crossman LC, Pitt T, Churcher C, Mungall K, Bent-
ley SD, Sebaihia M, Thomson NR, Bason N, Beacham IR,
Brooks K, Brown KA, Brown NF, Challis GL, Cherevach I,
Chillingworth T, Cronin A, Crossett B, Davis P, DeShazer
D, Feltwell T, Fraser A, Hance Z, Hauser H, Holroyd S, Jagels
K, Keith KE, Maddison M, Moule S, Price C, Quail MA, Rab-
binowitsch E, Rutherford K, Sanders M, Simmonds M, Song-
sivilai S, Stevens K, Tumapa S, Vesaratchavest M, White-
head S, Yeats C, Barrell BG, Oyston PC, and Parkhill J:
Genomic plasticity of the causative agent of melioidosis,
Burkholderia pseudomallei. Proc Natl Acad Sci USA 2004;101:
14240–14245.
16. Nierman WC, DeShazer D, Kim HS, Tettelin H, Nelson KE,
Feldblyum T, Ulrich RL, Ronning CM, Brinkac LM, Daugh-
erty SC, Davidsen TD, Deboy RT, Dimitrov G, Dodson RJ,
Durkin AS, Gwinn ML, Haft DH, Khouri H, Kolonay JF,
Madupu R, Mohammoud Y, Nelson WC, Radune D, Romero
CM, Sarria S, Selengut J, Shamblin C, Sullivan SA, White O,
Yu Y, Zafar N, Zhou L, and Fraser CM: Structural flexibil-
ity in the Burkholderia mallei genome. Proc Natl Acad Sci USA
2004;101:14246–14251.
17. Reckseidler SL, DeShazer D, Sokol PA, and Woods DE: De-
tection of bacterial virulence genes by subtractive hy-
bridization: identification of capsular polysaccharide of
Burkholderia pseudomallei as a major virulence determinant.
Infect Immun 2001;69:34–44.
18. DeShazer D, Waag DM, Fritz DL, and Woods DE: Identifi-
cation of a Burkholderia mallei polysaccharide gene cluster by
subtractive hybridization and demonstration that the en-
coded capsule is an essential virulence determinant. Microb
Pathog 2001;30:253–269.
Address reprint requests to:
Shyh-Ching Lo
Armed Forces Institute of Pathology
Bldg. 54, Room 3025
6825 16th Street, NW
Washington, D.C. 20306
E-mail: los@afip.osd.mil
Received: February 28, 2008
Accepted: April 24, 2008