INFECTION AND IMMUNITY, Nov. 2008, p. 5173–5180
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 76, No. 11
Mycobacterium bovis BCG Immunization Induces Protective Immunity
against Nine Different Mycobacterium tuberculosis Strains in Mice?†
Bo Young Jeon,1‡ Steven C. Derrick,1JaeHyun Lim,1Kristopher Kolibab,1Veerabadran Dheenadhayalan,1§
Amy Li Yang,1Barry Kreiswirth,2and Sheldon L. Morris1*
Laboratory of Mycobacterial Diseases and Cellular Immunology, Center for Biologics Evaluation and Research, United States Food
and Drug Administration, Bethesda, Maryland,1and Public Health Research Institute, Tuberculosis Center,
International Center for Public Health, Newark, New Jersey2
Received 7 January 2008/Returned for modification 2 March 2008/Accepted 13 August 2008
Recent preclinical and epidemiologic studies have suggested that certain Mycobacterium tuberculosis genotypes (in
particular, Beijing lineage strains) may be resistant to Mycobacterium bovis BCG vaccine-induced antituberculosis
protective immunity. To investigate the strain specificity of BCG-induced protective responses in a murine model
of pulmonary tuberculosis, C57BL/6 mice were vaccinated with BCG vaccine and then challenged 2 months later
with one of nine M. tuberculosis isolates. Four of these strains were from the W-Beijing lineage (HN878, N4, NHN5,
and ChS) while four were non-Beijing-type isolates (C913, CDC1551, NY669, and NY920). As a control, the WHO
standard M. tuberculosis Erdman strain was evaluated in these vaccination/challenge experiments. To assess the
protective responses evoked by BCG immunization, organ bacterial burdens and lung pathology were assessed in
the aerosol challenge with each of these strains, significantly reduced bacterial growth in the lungs and spleens and
significantly improved lung pathology were seen in all vaccinated animals compared to naïve controls. After 12
weeks, reduced organ bacterial burdens were detected in vaccinated animals infected with six of nine challenge
strains. Although lung CFU values were lower in vaccinated mice for only three of nine groups at 20 weeks
postchallenge, significantly decreased lung inflammation was seen in all immunized animals relative to controls at
20 weeks postchallenge. Taken together, these data demonstrate that BCG vaccination protects against infection
with diverse M. tuberculosis strains in the mouse model of pulmonary tuberculosis and suggest that strain-specific
resistance to BCG-induced protective immunity may be uncommon.
As one of the world’s most devastating infectious diseases,
tuberculosis (TB) is responsible for approximately 2 million
deaths per year (41). This high rate of mortality has persisted
despite the availability of an attenuated vaccine (using Myco-
bacterium bovis BCG) for more than 50 years. The reasons for
the overall ineffectiveness of the BCG vaccine in controlling
global TB are complex and certainly not well understood (4, 5,
11). Potential factors which may contribute to the varying ef-
fectiveness of BCG vaccination in controlling TB include the
limited durability of BCG-induced protective immunity, ge-
netic variations among vaccinated individuals, and interference
with vaccine activity by host presensitization with environmen-
tal mycobacterial strains.
Recently, it has been hypothesized that the anti-TB protec-
tive immunity induced by BCG may be Mycobacterium tuber-
culosis strain specific and that this strain specificity may also
contribute to the limited effectiveness of this vaccine (1). The
proposed strain specificity is consistent with the increasing
awareness in recent years of the genetic heterogeneity and
phenotypic differences among M. tuberculosis strains (13, 17,
24). Epidemiologic studies have suggested that the differences
in virulence may be associated with the various genetic back-
grounds of clinical M. tuberculosis isolates (20). Preclinical
experiments have clearly shown that M. tuberculosis isolates
exhibit different levels of virulence and induce various immune
responses in animal models (9, 10, 19, 21, 28, 29, 35). Differ-
ential host-pathogen interactions caused by phenotypic differ-
ences among M. tuberculosis isolates could alter BCG-induced
immunity, resulting in reduced anti-TB protective responses to
specific pathogenic strains (25). Interestingly, previous studies
have shown that genetic variability among Plasmodium and
Streptococcus pneumoniae strains can reduce the effectiveness
of vaccines against these pathogens in specific geographic re-
Among the most prominent of the M. tuberculosis genotypes
are the W-Beijing lineage strains which have been associated
with worldwide TB outbreaks for more than a decade (2, 3, 12,
14). Recent epidemiological studies have suggested that mass
vaccination with BCG may have been a selective force for the
emergence of the W-Beijing genotype (15, 38). In these stud-
ies, W-Beijing family strains were isolated more frequently
from BCG-vaccinated TB patients than from nonimmunized
patients. Based on these data, Abebe and Bjune have sug-
gested that BCG immunization may actually increase the risk
of developing tuberculous disease in areas with a high preva-
lence of W-Beijing strain infections (1). The overall conclusion
from these studies is that W-Beijing strains may be resistant to
BCG-induced protective immunity and that BCG vaccination
* Corresponding author. Mailing address: FDA/CBER, Building 29,
Room 502, 29 Lincoln Drive, Bethesda, MD 20892. Phone: (301)
496-5978. Fax: (301) 435-5675. E-mail: email@example.com.
‡ Present address: Department of Microbiology, Yonsei University
College of Medicine, Seoul, South Korea.
§ Present address: Aeras Global TB Vaccine Foundation, Rockville,
† Supplemental material for this article may be found at http://iai
?Published ahead of print on 18 August 2008.
may actually promote the spread of these isolates. Previous
results of preclinical studies have supported the strain-specific
BCG resistance hypothesis. Lopez et al. have reported that
BCG vaccination of mice provided less protection against in-
fection with an M. tuberculosis W-Beijing isolate than against
challenge with the H37Rv M. tuberculosis laboratory strain
(19). Moreover, Tsenova and colleagues have shown in rabbits
that BCG vaccination confers poor protection against central
nervous disease caused by infection with the M. tuberculosis
HN878 W-Beijing strain (36). It should be noted, however, that
only a limited number of comparative preclinical studies have
been designed to examine postvaccination protection induced
against virulent W-Beijing isolates. Additionally, other epide-
miological studies which directly compared isolates from BCG-
vaccinated and nonimmunized patients have found minimal
association between BCG vaccination and the prevalence of
W-Beijing strains (2). Consequently, the linkage between the
emergence of W-Beijing strains and the overall efficacy of
BCG immunization is still uncertain and remains to be proven
Clearly, the importance of the phenotypic variations among
M. tuberculosis isolates on the design and development of new
TB vaccines has not been adequately assessed. Here, we de-
scribe initial investigations aimed at examining the effect on
TB vaccine efficacy of using different M. tuberculosis challenge
strains in a mouse model of pulmonary TB. In these studies,
mice were vaccinated with the licensed BCG vaccine and then
were challenged with either the standard laboratory M. tuber-
culosis Erdman strain, Beijing lineage M. tuberculosis clinical
isolates, or non-Beijing M. tuberculosis strains. Five of these
strains (C913, CDC1551, HN878, N4, and NHN5) have been
associated with outbreaks of TB (24, 26, 34, 37). In this report,
we show by evaluating mycobacterial growth in relevant organs
and comparing lung pathology in vaccinated and naïve mice
that BCG immunization induces protective immune responses
that limit the proliferation of aerosol infections by each of the
nine M. tuberculosis strains tested.
MATERIALS AND METHODS
Bacterial strains. The M. tuberculosis Erdman strain was prepared as a pre-
clinical standard for TB vaccine testing under a collaborative agreement between
the World Health Organization, the Food and Drug Administration’s Center for
Biologics Evaluation and Research, and the Aeras Global TB Vaccine Founda-
tion. The C913 and N4 M. tuberculosis strains were obtained from the strain
collections at the Public Health Research Institute in Newark, NJ, while the
HN878 and the NHN5 strains were kindly provided by Clifton Barry of the
National Institutes of Health in Rockville, MD. The CDC1551, ChS, NY669, and
NY920 strains had been previously characterized at the Food and Drug Admin-
istration’s Center for Biologics Evaluation and Research. Each strain was grown
in Middlebrook’s 7H9 medium (Difco, Detroit, MI) supplemented with 10% of
Middlebrook’s OADC (oleic acid, albumin, dextrose, and catalase) enrichment
medium (BBL, Sparks, MD) until late log phase. The cells were frozen at a
concentration of 2 ? 108CFU/ml.
Genetic analyses of M. tuberculosis isolates. For DNA isolation, M. tuberculosis
isolates were inoculated in 10 ml of Middlebrook 7H9 medium and incubated at
37°C with constant shaking until the mid-log phase was reached. The bacterial
cells were collected by centrifugation at 10,000 ? g and resuspended with dis-
tilled water. Resuspended bacterial cells were autoclaved at 121°C for 15 min,
and the supernatant was used directly for PCR.
The W-Beijing and non-W-Beijing-type strains were differentiated using
unique primer sets as described earlier (39). To identify a strain belonging to the
Beijing evolutionary lineage, the following primers were used: ACCGAGCTG
ATCAAACCCG (forward) and ATGGCACGGCCGACCTGAATGAACC (re-
verse). Using these primers, a 239-bp region of an IS6110 element (which is
unique for Beijing lineage isolates) was amplified (39). Non-Beijing isolates were
identified with a primer set complementary to the Rv2819c gene: GGTGCGA
GATTGAGGTTCCC (forward) and TCTACCTGCAGTCGCTTGTGC (re-
verse). The principal genetic groups and the cluster number were determined as
described previously (13, 34). For further genetic analysis based on differences in
variable-number tandem repeats (VNTR), three primer sets (VNTR0580,
VNTR1955, and QUB1895) were selected to discriminate the isolates of M.
tuberculosis (32, 33). The specific primer sequences used were the following:
VNTR 0580, CTGCGGTCAAACAGGTCA (forward) and CATACATCGGT
ACCCGAC (reverse); VNTR1955, AGACGTCAGATCCCAGTT (forward)
and ACCCGACAACAAGCCCA (reverse); and QUB1895, GGTGCACGGCC
TCGGCTCC (forward) and AAGCCCCGCCGCCAATCAA (reverse). Each
PCR mixture contained 10 mM Tris-HCl, pH 9.0, 50 mM KCl, 1.5 mM MgCl2,
a 200 ?M concentration of each deoxynucleoside triphosphate, a 0.5 ?M con-
centration of each primer, and 2.5 U of Taq DNA polymerase (Perkin Elmer,
Emeryville, CA). Two microliters of template DNA was added to each reaction
mixture. For negative controls, 2 ?l of sterile distilled water was added to the
PCR mixtures. After an initial denaturation at 94oC for 5 min, a touch-down
PCR was undertaken with 8 cycles consisting of a denaturation step at 94°C for
30s, an annealing step from 65oC to 57oC (each cycle at each temperature) for
1 min, and an elongation step at 72°C for 2 min, followed by 30 cycles of 94°C for
30 s, 52°C for 1 min, and 72°C for 2 min. A final elongation step at 72°C for 10
min terminated the program. The amplified products were analyzed by agarose
Evaluation of BCG-induced protective immunity using a murine aerogenic
infection model. The vaccination/challenge studies were performed as described
earlier (7). Pathogen-free C57BL/6 mice were obtained from the Jackson Lab-
oratories (Bar Harbor, ME). For the mycobacterial growth experiments, five
mice were evaluated for each group. Initially, mice were vaccinated once subcu-
taneously with BCG Pasteur (106bacteria in 0.1 ml of phosphate-buffered saline
[PBS]). Eight to 10 weeks after the immunization, mice were aerogenically
challenged with the M. tuberculosis isolates suspended in PBS at a concentration
known to deliver 200 CFU in the lungs over a 30-min exposure time in a
Middlebrook chamber (GlasCol, Terre Haute, IN). To assess the level of pul-
monary exposure during the aerosol challenge, the number of CFU in the lung
were measured at 4 h after the tuberculous infection. To determine the extent of
bacterial growth in the lungs and spleens, the mice were sacrificed at 4, 12, and
20 weeks postchallenge. The lungs and spleens were then removed aseptically
and homogenized separately in PBS using a Seward Stomacher 80 blender
(Tekmar, Cincinnati, OH). The lung and spleen homogenates were diluted
serially in 0.4% PBS–Tween 80, and 50-?l aliquots were placed on Middlebrook
7H11 agar (Difco) plates supplemented with 10% OADC enrichment (Becton
Dickinson, Sparks, MD) medium, 2 ?g/ml 2-thiophenecarboxylic acid hydrazide
(TCH) (Sigma), 10 ?g/ml ampicillin, and 50 ?g/ml cycloheximide (Sigma). The
addition of TCH to the agar plates inhibits the growth of BCG but not M.
tuberculosis. All plates were incubated at 37°C for 14 to 17 days in sealed plastic
bags, and the colonies were counted to determine the organ bacterial burdens.
Assessment of lung inflammation. To evaluate the level of inflammation in the
lungs of mice infected with M. tuberculosis, lung sections stained with hematox-
ylin and eosin were photographed using a Nikon Optiphot 2 microscope fitted
with a camera which was connected to a computer. The photos were taken at a
magnification of ?5 or ?10. Spot Advanced software was used to save the
computer images. The Image Pro Plus program (Media Cybernetics, Silver
Spring, MD) was utilized to objectively assess the level of inflammation present
in each image. In these images, the inflamed areas stained a more intense purple
than the noninflamed areas. For these analyses, colors were assigned as follows:
red to represent the inflamed areas, green to represent noninflamed areas, and
yellow to represent the background. After the color assignments were estab-
lished, the computer software identified inflamed and noninflamed sections on
each slide. The percentage of the lung sections staining red, green, or yellow was
then determined by the computer software. To quantitate the percent area
inflamed, we determined the mean percent red area from three to five lung
sections of each of the different groups.
Statistical analyses. The protection results and the lung inflammation data
were analyzed using the GraphPad Prism, version 4, program.
M. tuberculosis clinical isolates. To assess the effectiveness of
BCG vaccination against an aerogenic challenge by different
M. tuberculosis clinical isolates, we selected nine M. tuberculosis
strains (described in Table S1 in the supplemental material).
5174JEON ET AL.INFECT. IMMUN.
The Erdman strain is a common laboratory strain which has
been designated by the World Health Organization as a stan-
dard infection strain for the preclinical evaluation of new TB
vaccines. The other eight strains are drug-susceptible clinical
isolates; four of these isolates were previously shown to be
Beijing family strains (24, 26, 34, 37). To verify their genotypes,
the strains were evaluated using a PCR assay developed to
detect strains from the Beijing evolutionary lineage (39). The
results of this assay confirmed that the HN878, N4, NHN5, and
ChS isolates were Beijing family strains while the CDC1551,
C913, NY669, and NY920 isolates did not have the W-Beijing
genotype. To further characterize these strains, the strains
were assigned to a major genetic cluster based on its single
nucleotide polymorphism profile and to a principal genetic
group as defined by polymorphisms at the katG codon 463 and
the gyrA codon 95 (13, 24, 34). Finally, molecular typing was
done using VNTR PCR primers designed to detect tandem
repeats of interdispersed repetitive units within mycobacteria.
Using this methodology, novel PCR profiles were detected for
each strain. These analyses confirmed that each of these nine
strains was unique.
Mycobacterial growth in naïve mice after aerosol infections.
To evaluate whether BCG vaccination restricted the growth of
the nine M. tuberculosis strains after a low-dose challenge, mice
were initially vaccinated with 106CFU of BCG Pasteur. Two
months later, the relative mycobacterial growth rates in rele-
vant organs of naïve and BCG-vaccinated mice were deter-
mined by infecting C57BL/6 mice via the aerosol route and
then sacrificing the infected animals at 4, 12, and 20 weeks
postchallenge. The 200-CFU infection dose was verified by
sacrificing mice and plating lung homogenates at 4 h after an
aerogenic challenge with M. tuberculosis bacilli. Representative
mycobacterial growth data from two to three experiments with
each of these strains are shown in Fig. 1. Direct comparisons of
the bacterial growth of these strains are shown in Fig. S1 in the
supplemental material. Interestingly, the growth profiles of the
different M. tuberculosis strains in the lungs of naïve mice were
generally similar for the first 3 months postchallenge. During
the first month, rapid growth of the TB organisms was seen for
all strains, with a 4 to 5 log10increase in the lung infection
being observed in naïve mice. After the lung bacterial burdens
peaked at 4 weeks, the pulmonary concentrations of M. tuber-
culosis declined between 4 and 12 weeks postchallenge, with
significant 0.4 to 1.2 log10decreases in the number of myco-
bacterial CFU (relative to 4-week CFU values) detected for
seven of nine test strains. Surprisingly, the pulmonary bacterial
burdens for seven of the clinical isolates and the standard
Erdman strain were not different at 20 weeks following the
aerogenic challenges. At this time point, mice infected by aero-
sol with these eight strains had chronic infections with lung
CFU values of approximately 6.0 log10. In contrast, by 20
weeks postchallenge, all of the NY669-infected naïve mice had
died, and their deaths were associated with elevated pulmo-
nary bacterial burdens.
Mycobacterial growth is limited in the lungs of BCG-vacci-
nated mice infected by the aerosol route. Substantial growth of
FIG. 1. Growth of nine different M. tuberculosis strains in the lungs of BCG-vaccinated and naïve mice. Mice were challenged with about 200
CFU of each of these M. tuberculosis isolates and then were sacrificed 4, 12, and 20 weeks postchallenge to determine pulmonary bacterial burdens.
To determine the aerosol infection dose, groups of mice were also sacrificed at 4 h postchallenge. The results are representative of two to three
experiments. Significant differences between the numbers of lung CFU for vaccinated (open circles) and control mice (closed squares) are shown
with asterisks. ?, P ? 0.05; ??, P ? 0.01; and ???, P ? 0.001. In these studies, the ChS, HN878, N4, and NHN5 strains are Beijing (B) lineage strains.
VOL. 76, 2008 BCG INDUCES PROTECTION AGAINST DIFFERENT TB STRAINS5175
these M. tuberculosis strains (3 to 3.5 log10) in the lungs was
also seen in the BCG-vaccinated mice during the first 4 weeks
(Fig. 1). However, the extent of mycobacterial growth was
limited relative to naïve mice. Significantly decreased pulmo-
nary concentrations of TB organisms were seen at this time
point in immunized mice challenged with each M. tuberculosis
isolate. Table 1 shows the mean protective values (log10CFU
for naïve mice minus log10CFU for immunized mice) at 4 and
12 weeks postchallenge for two to three experiments. As seen
in Table 1, BCG vaccination induced significant protective
responses in the lungs against each M. tuberculosis challenge
strain at 4 weeks after the tuberculous infection. Average levels
of protection between 0.75 and 1.26 log10CFU were detected
in BCG-immunized mice at this early time after the infection.
The levels of protection at 4 weeks postchallenge were not
significantly different among these isolates (including the stan-
dard Erdman strain).
Although the lung bacterial burdens in naïve mice decreased
substantially after 4 weeks postchallenge, the lung CFU levels
in vaccinated mice generally increased during these later time
periods. However, the rates of growth of the bacilli in the lungs
of BCG-vaccinated mice slowed considerably between the 4-
and 20-week time points, with increases of less than 0.5 log10
CFU generally being detected. Significantly elevated rates of
growth were only detected for the NY669 strain, where a 1
log10increase was seen in the 4- to 12-week time interval and
the HN878, N4, and NY920 strains which increased by at least
0.5 log10CFU in the lungs during the 12- to 20-week time
period (P ? 0.05). Interestingly, significant levels of protection
in the lungs of BCG-vaccinated mice (relative to naïve mice)
were detected at 12 weeks after the challenge in mice infected
with all of the test strains except for the NY669, HN878, and
N4 isolates (Fig. 1 and Table 1). However, significant differ-
ences in the lung bacterial CFU values of BCG-vaccinated and
control mice were only observed in the NHN5, ChS, and
NY920 groups at the 20-week time point (Fig. 1). It should be
emphasized that significantly increased BCG-induced protec-
tive responses, compared to the standard Erdman strain, were
only seen at 12 weeks postinfection for the ChS and NHN5
Beijing lineage isolates (P ? 0.05).
Postinfection dissemination to the spleen in naïve and BCG-
vaccinated mice. After a low-dose aerosol infection of
C57BL/6 mice with M. tuberculosis, dissemination of the or-
ganisms to the spleen is detectable about 2 to 3 weeks post-
challenge, and by 4 weeks the concentrations of splenic myco-
bacteria generally reach a chronic steady-state level for naïve
mice (31). For most of the test strains, similar overall splenic
growth profiles were seen in naïve mice with no substantial
changes in mycobacterial concentrations detected between 4
and 20 weeks (Fig. 2; see also Fig. S1 in the supplemental
material). Significant increases in splenic TB levels between 4
and 20 weeks postchallenge were only detected for naïve mice
infected with the NY669, NY920, and HN878 isolates. Impor-
tantly, relative to the naïve controls, BCG vaccination delayed
or reduced dissemination to the spleen of all of the different M.
tuberculosis challenge strains. As seen in Table 1, significant
decreases of about 60 to 95% in splenocyte CFU values (0.48
to 1.32 log10) were seen in immunized animals at 4 weeks
following infection with all of the different M. tuberculosis
isolates. At 12 weeks, significant levels of protection in the spleen
(relative to naïve mice) were detected in BCG-vaccinated mice
aerogenically challenged with the Erdman, CDC1551, C913, ChS,
NY669, and NY920 isolates. By 20 weeks postchallenge, spleen
CFU values were not statistically different between vaccinated
animals and naïve controls after infection with all of the nine M.
tuberculosis test strains.
Improved lung pathology in BCG-vaccinated mice com-
pared to naïve controls after aerogenic M. tuberculosis infec-
tions. The relative postinfection lung pathology in control and
immunized mice is a critical parameter in the evaluation of the
effectiveness of new TB vaccines. To assess lung pathology
after challenge, lung sections were removed after sacrifice, and
these sections were stained with the hematoxylin and eosin
reagent. Representative lung sections from mice infected with
the Erdman and HN878 strains are shown in Fig. S2 to S5
in the supplemental material. At 4 weeks after the challenge
with the different TB strains, three types of lung pathology
patterns were observed in naïve mice. For animals infected
with the C913, ChS, N4, NHN5, and NY920 strains, modestly
sized, nonconsolidated lesions containing moderate lympho-
cyte infiltration were seen in lung sections. At this time point,
interstitial and peribronchial swelling were common. For the
Erdman, HN878, and NY669 strains, substantially more in-
flammation was apparent, and large coalescing inflammatory
lesions were detected in the lungs at 28 days postchallenge.
Interestingly, the CDC1551 aerogenic infections induced exag-
gerated early pathological responses. Substantial consolidation
with lymphocyte aggregates was apparent, and considerable
phagocytic infiltrate was seen in CDC1551 infected lungs at 4
weeks postchallenge. This CDC1551 strain has been previously
associated with early vigorous inflammatory responses in mice
and elevated rates of tuberculin conversion in humans (22, 37).
TABLE 1. Protective responses in the lungs and spleens of BCG-
vaccinated mice at 4 and 12 weeks postchallenge with
different M. tuberculosis strains
Mean (? SEM) protective response in tissue for the indicated
time postchallenge (CFU)b
4 wk12 wk 4 wk 12 wk
1.26 ? 0.13
0.87 ? 0.20
1.02 ? 0.21
0.75 ? 0.27
1.25 ? 0.11
0.92 ? 0.12
1.11 ? 0.13
1.15 ? 0.42
1.11 ? 0.22
0.49 ? 0.03
0.57 ? 0.10
0.62 ? 0.15
0.82 ? 0.02d
0.40 ? 0.11c
0.37 ? 0.05c
0.95 ? 0.10d
0.34 ? 0.25c
0.80 ? 0.19
1.27 ? 0.40
0.81 ? 0.10
1.32 ? 0.08
0.55 ? 0.04
0.84 ? 0.26
0.77 ? 0.25
0.48 ? 0.10
0.93 ? 0.02
0.75 ? 0.02
0.71 ? 0.17
0.89 ? 0.20
0.74 ? 0.20
0.58 ? 0.25
0.35 ? 0.02e
0.14 ? 0.05e
0.33 ? 0.12e
0.63 ? 0.14
0.68 ? 0.20
aIn this study, the ChS, HN878, N4, and NHN5 isolates are Beijing-lineage
bMean protection is the average protective response (number of CFU in naı ¨ve
mice ? number of CFU in vaccinated mice) for two to three experiments.
cAll of the BCG-induced protective responses in the lung at 4 and 12 weeks
postchallenge were significant relative to the naı ¨ve controls except for the 12-
week protection against the HN878, N4, and NY669 infections.
dImmunization with BCG induced significantly better protection against the
ChS and NHN5 infections than against the Erdman challenge at 12 weeks
eAll of the BCG-induced protective responses in the spleen at 4 and 12 weeks
post-challenge were significant relative to the naı ¨ve controls except for the 12
week protection against the HN878, N4 and NHN5 infections.
5176 JEON ET AL.INFECT. IMMUN.
By 20 weeks postchallenge, the inflammatory responses had
increased but remained moderate in naïve mice infected with
the C913, NHN5, and NY920 strains. Also, the inflammatory
responses had subsided at this time in the CDC1551-infected
animals. For these mice, large lesions containing lymphocyte
aggregates were present in lung sections, but areas of relatively
noninflamed lung tissue were also seen. In contrast, substantial
disease progression was observed in lung sections of mice in-
fected with the Erdman, HN878, NY669, ChS, and N4 strains
at the 20-week time point. For these isolates, nearly complete
consolidation of some lung regions was observed. Also, mac-
rophage and neutrophil influx and occasional necrosis were
seen in specific lesions.
In contrast, at 4 weeks postchallenge, substantially less in-
flammation was observed in the lungs of BCG-vaccinated
animals. The granulomatous-type structures were more con-
densed, mature, and lymphocyte rich in the lungs of BCG-
vaccinated mice than the larger, immature granulomas seen in
naïve mice at this early time point. For the BCG-vaccinated
animals at 20 weeks after the aerosol infections, smaller lesions
and overall less inflammation were also observed (relative to
naïve mice). The central pathological features within the lungs
of BCG-vaccinated mice at later time points were the large
areas of aggregated lymphocytes often surrounded by macro-
phages within the inflammatory lesions.
Reduced lung inflammation values in BCG-vaccinated mice
following aerogenic M. tuberculosis infections. To quantita-
tively compare the pathological immune responses postinfec-
tion, the lung sections were evaluated using the Image Pro Plus
analysis system (see Fig. S2 to S5 in the supplemental mate-
rial). With this imaging system, the proportion of the lung
section that is inflamed can be quantitatively defined. Previous
analyses to validate the system had shown that the inflamma-
tion value for lung sections of moribund CD4?/?mice was
80% at 28 days postinfection while the degree of inflammation
was 30% for BCG-vaccinated CD4?/?mice (S. Derrick, un-
published results). These inflammation values correlated with
the mean survival times postinfection since the BCG-vacci-
nated CD4?/?mice (156 ? 22 days) survived fivefold longer
than naïve CD4?/?mice (33 ? 6 days) in these earlier vacci-
nation/challenge studies (8). The mean inflammation values
for the lung sections of mice infected 4 or 20 weeks earlier with
the different isolates of M. tuberculosis are listed in Table 2.
These computer-generated inflammation values were generally
consistent with the lung pathology observations described
above. The levels of inflammation in the lungs of naïve mice at
4 weeks postchallenge with the different strains varied between
34 to 72%. Similar to our the lung pathology observations, the
highest computer-generated inflammation values (72%) were
seen after infection with the CDC1551 isolate at 4 weeks after
the aerosol infection. Consistent with the significantly lower
mycobacterial burdens and improved lung pathology detected
in BCG-vaccinated mice at 4 weeks, the pulmonary inflamma-
tion values (13 to 35%) of immunized animals were signifi-
cantly decreased, relative to naïve controls, for all nine strains
tested (P ? 0.05).
At 16 or 20 weeks postchallenge, significantly elevated lung
inflammation values (relative to 4 weeks) were seen for five
FIG. 2. Growth of nine different M. tuberculosis strains in the spleens of BCG-vaccinated and naïve mice. Mice were sacrificed at 4, 12, and 20
weeks following a 200-CFU aerogenic challenge with each of these M. tuberculosis isolates. To determine the aerosol infection dose, groups of mice
were also sacrificed at 4 h postchallenge. The results are representative of two to three experiments. Significant differences between the numbers
of lung CFU for vaccinated (open circles) and control (closed squares) mice are shown with asterisks. ?, P ? 0.05; ??, P ? 0.01; and ???, P ? 0.001.
In these studies, the ChS, HN878, N4, and NHN5 strains are Beijing (B) lineage strains.
VOL. 76, 2008 BCG INDUCES PROTECTION AGAINST DIFFERENT TB STRAINS5177
strains (Erdman, ChS, HN878, N4, and NY669). Considerable
disease progression had been noted for all of these strains in
visual pathological assessments. Importantly, our evaluation of
the relative lung pathology at 20 weeks postinfection showed
that the decreased inflammation in the BCG-vaccinated ani-
mals persisted for each strain tested. At this time point, the
inflammation values were generally two- to threefold lower for
vaccinated animals than values in the naïve controls. For the
highly virulent NY669 strain, lung pathology assessments were
done at 16 weeks because of the rapid mortality rate of naïve
mice infected with this strain. Although the naïve animals
challenged with NY669 were highly inflamed at 16 weeks
(56.1%), a significant reduction in the inflammation value for
BCG-vaccinated mice (29.9%) was detected.
The genetic and phenotypic differences that have been iden-
tified among M. tuberculosis isolates during the past decade
have raised concerns that the protective responses elicited by
new TB vaccines may be strain specific and, therefore, that
these novel TB vaccine preparations may not protect against
all M. tuberculosis strains. In fact, it has been speculated that
the geographic variability in the efficacy of BCG vaccination
may be due to BCG’s inability to protect against the various
types of M. tuberculosis strains that are endemic in specific
regions of the world (1). Consistent with this hypothesis, in-
triguing studies in mice and rabbits have suggested that BCG
vaccination is not effective at controlling infections by M. tu-
berculosis W-Beijing strains (19, 36). Selected epidemiologic
data have also predicted that the W-Beijing strains may be
resistant to BCG-induced protective immunity (1). However,
in contrast to these findings, the results of our experiments did
not support the hypothesis that specific M. tuberculosis strains
are resistant to the anti-TB immunity evoked by BCG. In our
studies, mice immunized with BCG were protected following
aerosol infections with a classic laboratory strain, four Beijing
clinical isolates, and four non-Beijing strains. After aerosol
infections with all nine of these strains, statistical differences in
the lung bacillary burdens and the lung inflammation values
were detected between vaccinated and naïve mice at the
4-week time point. Relative growth of the infecting organisms
was also statistically lower in BCG-vaccinated animals, relative
to naïve mice, for six of the strains (including the Erdman
strain and three of the W-Beijing isolates) at 12 weeks post-
challenge. Interestingly, BCG immunization induced better
protective responses against two of the Beijing lineage strains
than against the standard Erdman strain at the 12-week time
point. Although pulmonary bacterial burdens were reduced in
vaccinated mice only when they were challenged with three of
these strains at 20 weeks postchallenge, significant differences
in the lung inflammation values and lung pathology between
vaccinated and naïve animals were detected at the end of the
study for all of the test strains. Importantly, the BCG-induced
protection against two of these strains, as measured by de-
creases in pulmonary mycobacterial growth and reductions in
lung pathology, correlated with extended survival periods for
the vaccinated animals reported in earlier studies. Mice immu-
nized with BCG and then aerogenically challenged with either
the virulent M. tuberculosis HN878 or Erdman strains survived
significantly longer than naïve controls infected with the same
virulent strains (6, 16). Additionally, in this study, while all
naïve mice infected with the NY669 strain died by 20 weeks
postchallenge, all of their BCG-immunized counterparts sur-
vived until the 20-week time point.
The factors that have contributed to the different BCG im-
munization protection results seen in our study compared to
earlier published reports remain uncertain but may include
differences in animal models, M. tuberculosis challenge meth-
ods, lung pathology analyses, and strains used for infection and
vaccination. Regarding the strains, various production meth-
ods can yield mycobacterial strain preparations with contrast-
ing immunogenic activities because of different ratios of live
and dead organisms, different bacterial concentrations, and
altered surface compositions. For example, the failure to stan-
dardize production protocols can result in M. tuberculosis chal-
lenge strains with inconsistent levels of virulence. In a previous
comparative study, the use of a subpotent M. tuberculosis
Erdman preparation for murine infections led to improper
initial assessments of the virulence of the CDC1551 strain (18).
By contrast, the impact of immunizing with various live atten-
uated vaccine strains (including different BCG preparations) is
unclear. Although different BCG strains have been shown to
induce unique immune responses in animal models and hu-
mans, both preclinical study results and data from clinical trials
have strongly suggested that different BCG preparations usu-
ally yield similar levels of protection when given at equivalent
doses by the same route of administration (4, 40, 42). In fact,
we have shown that the BCG Pasteur and SSI BCG strains
induce statistically indistinguishable protective responses
against the M. tuberculosis HN878 and M. tuberculosis Erdman
strains (S. Derrick, unpublished data). To reduce this strain
heterogeneity with the aim of improving the TB vaccine testing
process, our laboratory has collaborated with the World
Health Organization to provide standard BCG vaccine prepa-
rations and M. tuberculosis challenge strains to researchers
throughout the world. Overall, the availability of these refer-
TABLE 2. Lung inflammation in naı ¨ve and BCG-vaccinated mice
at 4 and 20 weeks postchallenge with nine different
M. tuberculosis strains
Mean inflammation value (% of lung section inflamed) for
the group at:a
4 wk postchallenge20 wk postchallenge
Naı ¨ve mice
Naı ¨ve mice
48.8 ? 3.0
36.4 ? 4.2
72.3 ? 2.5
34.4 ? 2.5
49.3 ? 4.0
39.6 ? 3.4
35.3 ? 3.7
47.5 ? 7.0
38.9 ? 5.8
13.1 ? 2.1**
24.8 ? 2.4*
35.9 ? 1.8**
11.7 ? 4.3**
16.5 ? 1.4**
15.4 ? 3.8*
24.7 ? 2.8*
16.0 ? 0.6**
19.4 ? 1.2**
38.7 ? 1.5
34.0 ? 0.8
37.8 ? 1.7
53.9 ? 7.7
43.6 ? 5.2
47.3 ? 9.0
42.7 ? 2.9
56.1 ? 0.2b
36.3 ? 6.2
25.5 ? 1.1**
18.0 ? 1.7**
23.6 ? 2.5**
24.4 ? 1.3**
23.2 ? 2.7*
15.2 ? 2.5*
16.8 ? 1.4**
29.9 ? 3.5*
20.1 ? 4.4*
aMean percentage of the area of inflammation ? standard error of the mean.
Significantly reduced lung inflammation in the BCG-vaccinated mice relative to
naı ¨ve controls is indicated as follows: *, P ? 0.05; **, P ? 0.01.
bSince naı ¨ve mice infected with NY669 died prior to the 20-week time point,
these mice were sacrificed at 16 weeks postchallenge.
5178 JEON ET AL.INFECT. IMMUN.
ence strains has facilitated and improved the comparative eval-
uation of new TB vaccines.
In recent reports, phenotypically different M. tuberculosis
isolates have been shown to have various levels of virulence in
animal models (9, 10, 21, 28, 30). In our experiments, the
growth profiles for eight of nine M. tuberculosis strains were
similar, and the pulmonary CFU values for most of these
isolates at 20 weeks postchallenge were nearly equivalent.
However, the importance of the number of culturable M. tu-
berculosis bacilli in the lungs at several months postinfection is
unclear, and whether the lung CFU values correlate with vir-
ulence (as measured in survival studies) is uncertain. In an
earlier report, North et al. concluded that the growth rate of
mycobacteria in mice is not a reliable indicator of mycobacte-
rial virulence (27). More recently, Palanisamy et al. showed
that organ pathology is a better correlate of virulence than the
number of viable organisms in animal tissues (30). At 16 or 20
weeks postinfection in our study, five of the test strains had
elevated lung inflammation values including three strains—
HN878, NY669, and Erdman strains—that have been shown to
be virulent in mice in this and earlier studies (6, 23). Moreover,
the modestly virulent CDC1551 strain was among the isolates
with low inflammation values at the later time points (18, 30).
Therefore, our data suggest that virulence may correlate with
the extent of postinfection lung pathology but is not necessarily
directly related to pulmonary CFU levels. Obviously, survival
studies involving all of the strains in this study are needed to
further support the association between virulence and lung
In sum, we have evaluated nine different M. tuberculosis
strains in a mouse model of pulmonary TB and have shown
that while organ mycobacterial growth profiles were generally
similar for eight of nine strains, various lung pathological re-
sponses were induced after the infection. Most importantly, we
showed that BCG vaccination induced significant protective
immune responses against all of these strains including four
W-Beijing strains. The levels of BCG-induced protective im-
munity against the eight clinical isolates and the standard
Erdman strain were generally similar, especially at the 4-week
time point. Overall, we could not demonstrate the strain-spe-
cific resistance to BCG-induced protective immunity in our
mouse model that has been suggested by other studies. It
should be noted, however, that the specific protective immune
responses induced by live attenuated vaccines such as BCG
vaccine may differ from the protective immunity induced by
protein-based, viral vectored, or DNA vaccines against TB.
Further studies are needed to assess whether the anti-TB im-
mune responses induced by nonliving TB vaccines also protect
against various M. tuberculosis phenotypes.
1. Abebe, F., and G. Bjune. 2006. The emergence of Beijing family genotypes of
Mycobacterium tuberculosis and low-level protection by bacille Calmette-
Guerin (BCG) vaccines: is there a link? Clin. Exp. Immunol. 145:389–397.
2. Anh, D. D., M. W. Borgdorff, L. N. Van, N. T. Lan, T. van Gorkom, K.
Kremer, and D. van Soolingen. 2000. Mycobacterium tuberculosis Beijing
genotype emerging in Vietnam. Emerg. Infect. Dis. 6:302–305.
3. Bifani, P., B. Mathena, N. Kurepina, and B. N. Kreiswirth. 2002. The global
dissemination of the W family of Mycobacterium tuberculosis strains. Trends
4. Brewer, T. F. 2000. Preventing tuberculosis with bacillus Calmette-Guerin
vaccine: a meta-analysis of the literature. Clin. Infect. Dis. 31:S64–S67.
5. Colditz, G. A., T. F. Brewer, C. S. Berkey, M. E. Wilson, E. Burdick, H. V.
Fineberg, and F. Mosteller. 1994. Efficacy of BCG vaccine in the prevention
of tuberculosis. Meta-analysis of the published literature. JAMA 271:698–
6. Delogu, G., A. Li, C. Repique, F. Collins, and S. L. Morris. 2002. DNA
vaccine combinations expressing either tissue plasminogen activator signal
sequence fusion proteins or ubiquitin-conjugated antigens induce sustained
protective immunity in a mouse model of pulmonary tuberculosis. Infect.
7. Derrick, S. C., A. L. Yang, and S. L. Morris. 2004. A polyvalent DNA vaccine
expressing an ESAT6-Ag85B fusion protein protects mice against a primary
infection with Mycobacterium tuberculosis and boosts BCG-induced protec-
tive immunity. Vaccine 23:780–789.
8. Derrick, S. C., C. Repique, P. Snoy, A. L. Yang, and S. L. Morris. 2004.
Immunization with a DNA vaccine cocktail protects mice lacking CD4 cells
against an aerogenic infection with Mycobacterium tuberculosis. Infect. Im-
9. Dormans, J., M. Burger, D. Aguilar, R. Hernandez-Pando, K. Kremer, P.
Roholl, S. M. Arend, and D. van Soolingen. 2004. Correlation of virulence,
lung pathology, bacterial load, and delayed-type hypersensitivity responses
after infection with different Mycobacterium tuberculosis genotypes in a
BALB/c mouse model. Clin. Exp. Immunol. 137:460–468.
10. Dunn, P. L., and R. J. North. 1995. Virulence ranking of some Mycobacte-
rium tuberculosis and Mycobacterium bovis strain according to their ability to
multiply in the lungs, induce lung pathology, and cause mortality in mice.
Infect. Immun. 63:3428–3437.
11. Fine, P. E. 2001. BCG: the challenge continues. Scand. J. Infect. Dis. 33:
12. Glynn, J. R., J. Whitely, P. J. Bifani, K. Kremer, and D. van Soolingen. 2002.
Worldwide occurrence of Beijing/W strains of Mycobacterium tuberculosis: a
systemic review. Emerg. Infect. Dis. 8:843–849.
13. Gutacker, M., B. Mathena, H. Soini, E. Shashkina, B. N. Kreiswirth, E. A.
Graviss, and J. M. Musser. 2006. Single-nucleotide polymorphism-based
population genetic analysis of Mycobacterium tuberculosis strains from 4
geographic sites. J. Infect. Dis. 193:121–128.
14. Hanekom, M., G. D. van der Spuy, E. Streicher, S. L. Ndabambi, C. R. E.
McEvoy, M. Kidd, N. Beyers, T. C. Victor, P. D. van Helden, and R. M.
Warren. 2007. A recently evolved sublineage of the Mycobacterium tubercu-
losis Beijing strain family was associated with an increased ability to spread
and cause disease. J. Clin. Microbiol. 45:1483–1490.
15. Hermans, P. W., M. F. Messadi, H. Guebrexabher, D. van Soolingen, P. E.
de Haas, H. Heersma, H. de Neeling, A. Ayoub, F. Portaels, D. Frommel, M.
Zribi, and J. D. A. van Embden. 1995. Analysis of the population structure
of Mycobacterium tuberculosis in Ethiopia, Tunisia, and the Netherlands:
usefulness of DNA typing for global tuberculosis epidemiology. J. Infect.
16. Hinchey, J., S. Lee, B. Y. Jeon, R. J. Basaraba, M. M. Venkataswamy, B.
Chen, J. Chan, M. Braunstein, I. M. Orme, S. C. Derrick, S. L. Morris, W. R.
Jacobs, Jr., and S. A. Porcelli. 2007. Enhanced priming of adaptive immunity
by a proapoptotic mutant of Mycobacterium tuberculosis. J. Clin. Investig.
17. Hirsh, A. E., A. G. Tsolaki, K. DeRiemer, M. W. Feldman, and P. M. Small.
2004. Stable association between strains of Mycobacterium tuberculosis and
their human host populations. Proc. Natl. Acad. Sci. USA 101:4871–4876.
18. Kelley, C. L., and F. M. Collins. 1999. Growth of a highly virulent strain of
Mycobacterium tuberculosis in mice of differing susceptibility to tuberculous
challenge. Tuber. Lung Dis. 79:367–370.
19. Lopez, B., D. Aguilar, H. Orozco, M. Burger, C. Espitia, V. Ritacco, L. Barrera,
K. Kremer, R. Hernandez-Pando, K. Huygen, and D. van Soolingen. 2003. A
marked difference in the pathogenesis and immune response induced by differ-
ent Mycobacterium tuberculosis genotypes. Clin. Exp. Immunol. 133:30–37.
20. Malik, A. N. J., and P. Godfrey-Fausett. 2005. Effects of genetic variability of
Mycobacterium tuberculosis strains on the presentation of disease. Lancet
Infect. Dis. 5:174–183.
21. Manabe, Y. C., A. M. Dannenberg, S. K. Tyagi, C. L. Hatem, M. Yoder, S. C.
Woolwine, B. C. Zook, M. L. Pitt, and W. R. Bishai. 2003. Different strains
of Mycobacterium tuberculosis cause various spectrums of disease in the
rabbit model of tuberculosis. Infect. Immun. 71:6004–6011.
22. Manca, C., L. Tsenova, C. E. Barry III, A. Bergtold, S. Freeman, P. A. J.
Hackett, J. M. Musser, V. H. Freedman, and G. Kaplan. 1999. Mycobacte-
rium tuberculosis CDC1551 induces a more vigorous host response in vivo
and in vitro, but is not more virulent than other clinical isolates. J. Immunol.
23. Manca, C., L. Tsenova, A. Bergtold, S. Freeman, M. Tovey, J. M. Musser,
C. E. Barry, V. H. Freedman, and G. Kaplan. 2001. Virulence of a Myco-
bacterium tuberculosis isolate in mice is determined by failure to induce Th1
type immunity and is associated with induction of IFN ?/?. Proc. Natl. Acad.
Sci. USA 98:5752–5757.
24. Mathema, B., N. Kurepina, P. Bifani, and B. N. Kreiswirth. 2006. Epidemi-
ology of Mycobacterium tuberculosis: current insights. Clin. Microbiol. Rev.
25. McShane, H. 2003. Susceptibility to tuberculosis—the importance of the
pathogen as well as the host. Clin. Exp. Immunol. 133:20–21.
26. Milan, S. J., K. A. Hauge, N. E. Kurepina, K. H. Lofy, S. V. Goldberg, M.
VOL. 76, 2008BCG INDUCES PROTECTION AGAINST DIFFERENT TB STRAINS5179
Narita, C. M. Nolan, P. D. McElroy, B. N. Kreiswirth, and G. A. Cangelosi. Download full-text
2004. Expanded geographical distribution of the N family of Mycobacterium
tuberculosis strains within the United States. J. Clin. Microbiol. 42:1064–
27. North, R. J., L. Ryan, R. LaCourse, T. Mogues, and M. E. Goodrich. 1999.
Growth rate of mycobacteria in mice as an unreliable indicator of mycobac-
terial virulence. Infect. Immun. 67:5483–5485.
28. Ordway, D., M. Henao-Temayo, M. Harton, G. Palanisamy, J. Troudt, C.
Shanley, R. J. Basaraba, and I. M. Orme. 2007. The hypervirulent Myco-
bacterium tuberculosis strain HN878 induces a potent TH1 response followed
by rapid down-regulation. J. Immunol. 179:522–531.
29. Ordway, D. J., M. G. Sonnenberg, S. A. Donahue, J. T. Belisle, and I. M.
Orme. 1995. Drug-resistant strains of Mycobacterium tuberculosis exhibit a
range of virulence for mice. Infect. Immun. 63:741–743.
30. Palanisamy, G. S., E. E. Smith, C. A. Shanley, D. J. Ordway, I. M. Orme, and
R. J. Basaraba. 5 March 2008, posting date. Disseminated disease severity as
a measure of virulence of Mycobacterium tuberculosis in the guinea pig
model. Tuberculosis. doi:10.1016/j.tube.2007.12.003.
31. Repique, C. J., A. Li, F. M. Collins, and S. L. Morris. 2002. DNA immuni-
zation in a mouse model of latent tuberculosis: effect of DNA vaccination on
reactivation of disease and on reinfection with a secondary challenge. Infect.
32. Roring, S., A. Scott, D. Brittain, I. Walker, G. Hewinson, S. Neill, and R.
Skuce. 2002. Development of variable-number tandem repeat typing of My-
cobacterium bovis: comparison of results with those obtained by using exist-
ing exact tandem repeats and spoligotyping. J. Clin. Microbiol. 40:2126–
33. Smittipat, N., P. Billamas, M. Palittapongarnpim, A. Thong-On, M. M.
Temu, P. Thanakijcharoen, O. Karnkawinpong, and P. Palittapongarnpim.
2005. Polymorphism of variable-number repeats at multiple loci in Mycobac-
terium tuberculosis. J. Clin. Microbiol. 43:5034–5043.
34. Sreevatsan, S., X. Pan, K. E. Stockbauer, N. D. Connell, B. N. Kreiswirth,
T. S. Whittam, and J. M. Musser. 1997. Restricted structural gene polymor-
phism in the Mycobacterium tuberculosis complex indicates evolutionarily
recent global dissemination. Proc. Natl. Acad. Sci. USA 94:9869–9874.
35. Tsenova, L., E. Ellison, R. Harbacheuski, A. L. Moreira, N. Kurepina, M. B.
Reed, B. Mathema, C. E. Barry, and G. Kaplan. 2005. Virulence of selected
Mycobacterium tuberculosis clinical isolates in the rabbit model is dependent
on the phenolic glycolipid produced by the bacilli. J. Infect. Dis. 192:98–106.
36. Tsenova, L., R. Harbacheuski, N. Sung, E. Ellison, D. Fallows, and G.
Kaplan. 2007. BCG vaccination confers poor protection against M. tubercu-
losis HN878-induced central nervous system disease. Vaccine 25:5126–5132.
37. Valway, S. E., M. P. Sanchez, T. F. Shinnick, I. Orme, T. Agerton, D. Hoy,
J. S. Jones, H. Westmoreland, and I. M. Onorato. 1998. An outbreak in-
volving extensive transmission of a virulent strain of Mycobacterium tubercu-
losis. N. Engl. J. Med. 338:633–639.
38. van Soolingen, D., L. Quan, and P. E. de Han. 1995. Predominance of a
single genotype of Mycobacterium tuberculosis in countries of East Asia.
J. Clin. Microbiol. 33:3234–3238.
39. Warren, R. M., T. C. Victor, E. M. Streicher, M. Richardson, N. Beyers, N. C.
Gey van Pittius, and P. D. van Helden. 2004. Patients with active tuberculosis
often have different strains in the same sputum specimen. Amer. J. Respir.
Crit. Care Med. 165:610–614.
40. Wedlock, D. N., M. Denis, H. M. Vordermeier, R. G. Hewinson, and B. M.
Buddle. 2007. Vaccination of cattle with Danish and Pasteur strains of
Mycobacterium bovis BCG induce different levels of IFN-? postvaccination,
but induce similar levels of protection against bovine tuberculosis. Vet.
Immunol. Immunopathol. 118:50–58.
41. World Health Organization. 2007. Tuberculosis fact sheet no.104. World
Health Organization, Geneva, Switzerland. http://www.who.imt/mediacentre
42. Wu, B., C. Huang, L. Garcia, A. Ponce de Leon, J. S. Osnorio, M. Bobadilla-
del-Valle, L. Ferreira, S. Canizales, P. Small, M. Kato-Maeda, A. M. Krensky,
and C. Clayberger. 2007. Unique gene expression profiles in infants vaccinated
with different strains of Mycobacterium bovis bacille Calmette-Gue ´rin. Infect.
Editor: J. L. Flynn
5180JEON ET AL.INFECT. IMMUN.