Vaccine 33 (2015) 126–132

Contents lists available at ScienceDirect

Vaccine
journal homepage: www.elsevier.com/locate/vaccine

Effect of cationic liposomes on BCG trafficking and vaccine-induced

immune responses following a subcutaneous immunization in mice
Steven C. Derrick a,∗ , Amy Yang a , Marcela Parra a , Kristopher Kolibab b , Sheldon L. Morris a
a
Laboratory of Mycobacterial Diseases and Cellular Immunology, Center for Biologics Evaluation and Research, Silver Spring, MD 20993, United States
b
Center for Drug Evaluation and Research, United States Food and Drug Administration, Silver Spring, MD 20993, United States

a r t i c l e i n f o a b s t r a c t

Article history: While formulating Mycobacterium bovis BCG in lipid-based adjuvants has been shown to increase the
Received 11 July 2014 vaccine’s protective immunity, the biological mechanisms responsible for the enhanced potency of lipid
Received in revised form 9 October 2014 encapsulated BCG are unknown. To assess whether mixing BCG in adjuvant increases its immunogenicity
Accepted 6 November 2014
by altering post-vaccination organ distribution and persistence, mice were immunized subcutaneously
Available online 15 November 2014
with conventional BCG Pasteur or BCG formulated in DDA/TDB adjuvant and the bio-distribution of BCG
bacilli was evaluated in mouse lungs, spleens, lymph nodes, and livers for up to 1 year. Although BCG was
Keywords:
rarely detected in mouse livers, mycobacteria were found in mouse lungs, spleens, and lymph nodes for
BCG
Tuberculosis
at least 1 year post-vaccination. However, at various time points during the 1 year study, the frequency
Immunity of lung and spleen infections and the number of mycobacteria in infected organs of individual mice were
Adjuvants highly variable. In contrast, mycobacteria were nearly always detected in the lymph nodes of vaccinated
mice. While the frequency and extent of lymph node infections generally were not significantly different
between mice vaccinated with adjuvanted or nonadjuvanted BCG preparations, multiparameter flow
cytometry analysis of lymph node cells showed significantly higher frequencies of CD4+ and CD8+ T cells
expressing IFN-␥ and IFN-␥/TNF-␣ in mice immunized with adjuvanted BCG. Overall, our data suggest
that the relationship between lymph node infection and the generation of anti-tuberculosis protective
responses following BCG vaccination should be further investigated.
Published by Elsevier Ltd.

1. Introduction against TB induced by BCG immunization have been highly variable.

More than a century after Mycobacterium tuberculosis was first
identified and decades after antibiotic therapies were initiated,
tuberculosis (TB) still remains a leading cause of morbidity and
mortality in humans and a major public health problem world-
wide [1]. With an estimated one-third of the world’s population
being infected by M. tuberculosis and 1.5 million deaths resulting
from TB infections each year, new control measures are needed to
combat this devastating disease. The only licensed vaccine against
M. tuberculosis, live Mycobacterium bovis BCG, is one of the most
widely used global vaccines, with more than 3 billion people hav-
ing been immunized during its more than 8 decades of use [2].
BCG vaccine is relatively safe and is easy and inexpensive to pro-
duce. However, estimates of the level and persistence of protection
∗ Corresponding author at: Building 52/72, Room 5322 FDA/CBER 10903 New
Hampshire Ave., Silver Spring, MD 20993, United States. Tel.: +1 240 402 9475.
E-mail address: steven.derrick@fda.hhs.gov (S.C. Derrick).
While randomized controlled trials and retrospective case control
studies have shown that BCG immunization is effective in reducing
cases of severe disseminated disease in children, the efficacy of BCG
in preventing tuberculosis in adults has ranged from 0 to 80% [3,4].
Furthermore, the longevity of BCG-induced protection is uncertain
and may be population dependent [5]. Although it has been esti-
mated that BCG’s efficacy wanes about 10 years after vaccination,
a 60 year follow-up of a 1930s clinical trial showed that BCG was
still 52% effective 6 decades after immunization of native Ameri-
cans [6]. Interestingly, several reports have also indicated that BCG
vaccination may impart wide ranging health-related effects that
are unrelated to its anti-tuberculosis activity including an overall
beneficial impact on childhood survival [7,8].
Since BCG is widely used in countries with high TB burdens
and BCG vaccine does reduce the incidence of severe childhood
extrapulmonary TB disease, it is unlikely that BCG vaccine will be
replaced in the near future. However, many investigators are devel-
oping new strategies to improve the protective activity of BCG
[9,10]. One approach has been to amplify BCG-induced immune

http://dx.doi.org/10.1016/j.vaccine.2014.11.004
0264-410X/Published by Elsevier Ltd.
 S.C. Derrick et al. / Vaccine 33 (2015) 126–132
responses by boosting with subunit or viral vaccines after an initial
BCG priming vaccination. A potentially simpler and less expensive
alternative approach to increase BCG’s immunogenicity is to formu-
late BCG in lipid-containing mixtures. Lipid encapsulation has been
shown in mice, deer, possums, and cattle to improve the protective
immunity elicited after BCG immunization [11–15]. Our group has
recently demonstrated that the formulation of BCG or the mutant
BCG mmaA4 strain in the cationic liposomal DDA/TDB adjuvant
increased the level of persistence of BCG-induced anti-tuberculosis
protective responses [16]. Significantly increased protective immu-
nity for the adjuvanted BCG preparations was especially seen at
longer post-vaccination and post-challenge time intervals.
Currently, the biological mechanisms associated with the
enhanced potency of lipid encapsulated BCG are uncertain. For
example, it is unclear whether the enhanced viability or persistence
of lipidated BCG or the differential trafficking of lipid encapsulated
BCG contributes to the increased immunogenicity. Undoubtedly,
the protection afforded by BCG is largely dependent on the pres-
ence of live organisms. Killing BCG by heat treatment or antibiotic
therapy markedly reduces (but does not eliminate) its capacity to
induce anti-tuberculosis protective immunity [17–19]. Addition-
ally, the persistence of viable BCG after vaccination has not been
clearly defined. Earlier studies indicated that BCG fails to grow
substantially after vaccination and the bacilli are cleared within 5
months [18–20]. In contrast, a more recent study showed that the
persistence of BCG was strain dependent and following intranasal
immunization BCG Pasteur was detected at low levels in the lung
at 17 months post-vaccination [21]. It is likely that lipid encapsu-
lation of BCG may increase its in vivo survival and alter its organ
distribution. Aldewell et al. [14] have reported that lipidating BCG
extends the survival of orally delivered BCG compared to conven-
tional BCG and increases targeting of the vaccine to the relevant
lymph nodes. Importantly, the biological activities of lipid adju-
vant components may also enhance the potency of lipidated BCG.
Cationic liposomes such as DDA have been shown to act as carriers
which can increase delivery of antigen to antigen-presenting cells
and create depot-like effects where antigen is slowly released over
time [22]. When an innate immune agonist such as TDB is incor-
porated into DDA liposomes antigen-specific immune responses
are often augmented; this immune enhancement by TDB has been
attributed to increases in the migration of monocytes to the injec-
tion site, enhanced T cell differentiation, and improved draining of
vaccine preparations to the lymph nodes22. Therefore, mixing BCG
with an adjuvant (DDA/TDB) containing cationic liposomes and an
immune agonist should result in an amplification of BCG-induced
immunity.
To assess how encapsulating BCG with cationic liposomes
impacts the kinetics of BCG trafficking and persistence in relevant
organs, we compared the biodistribution of conventional BCG with
BCG formulated in DDA/TDB adjuvant after a subcutaneous immu-
nization. We found that the temporal biodistribution of BCG was
highly variable in mouse lungs and spleens but was more consis-
tent in the lymph nodes. Additionally, we showed that adjuvanted
or non-adjuvanted BCG Pasteur given by the subcutaneous (s.c.)
route can persist for more than 1 year in mouse lungs, spleens, and
lymph nodes.
2. Materials and methods
2.1. Animals
C57BL/6 female mice that were 6–8 weeks of age were obtained
from the Jackson Laboratories (Bar Harbor, ME). All mice used in this
study were maintained under appropriate conditions at the Center
for Biologics Evaluation and Research, Bethesda, MD. This study
127
was done in accordance with the guidelines for the care and use of
laboratory animals specified by the National Institutes of Health.
This protocol was approved by the Institutional Animal Care and
Use Committee of the Center for Biologics Evaluation and Research
under Animal Study Protocol 1993–09.
2.2. Preparation of vaccines
BCG Pasteur strain was used in these studies and was origi-
nally obtained from Trudeau Institute (Saranac Lake, NY). BCG was
administered s.c. in the back in PBS or adjuvant at 1 × 106 CFU per
immunization in 0.2 ml. Adjuvanted BCG (BCG + Adj) was prepared
by mixing the BCG with dimethyl dioctadecyl-ammonium bromide
(DDA, Kodak, Rochester, NY) and d-(+)-trehalose 6,6 -dibehenate
(TDB, Avanti Polar Lipids, Alabster, AL). The DDA solution was pre-
pared by heating 25 mg in 10 ml water at 80 ◦ C for 20 min and
vortexing every 5 min. The TDB solution was prepared by adding
1.0 ml of water with 2 ␮l DMSO (0.2% final) to a vial containing
5.0 mg TDB. The TDB suspension was sonicated until it became
homogenous. The adjuvanted vaccines were prepared by mixing
5 × 106 CFU of BCG with 0.6 ml of DDA with sufficient PBS to bring
the volume to 0.9 ml. One-tenth milliliters of TDB were added to
the BCG-DDA mixture, vortexed three times and then incubated at
room temperature for 1 h. Mice received one s.c. immunization for
each vaccine formulation.
2.3. Organ CFU determinations
At various time points post-vaccination with either BCG or
BCG + Adj, mice were sacrificed and the lungs, spleen, livers and the
axillary and inguinal lymph nodes (all lymph node homogenates
were pooled for each mouse) were extracted and homogenized
with the barrel of a 3 cc syringe in sterile Petri dishes con-
taining 2.0 ml PBS with 0.04% Tween-80. Organ homogenates
were then plated onto Middlebrook 7H11 agar (Difco) plates
containing 10% OADC enrichment (Becton Dickinson, Sparks,
MD) medium, 10 ␮g/ml ampicillin, 50 ␮g/ml cycloheximide, and
5 ␮g/ml trimethoprim. Plates were incubated at 37 ◦ C for 14–17
days before counting to determine the number of mycobacterial
colony forming units (CFU) per organ. Due to the high degree of
variability in the number of CFU in the lungs and spleens, undi-
luted homogenates were plated for these organs. The upper limit
of CFU determinations for the lungs and spleens was 400 CFU
per organ. The lower limit of detection was set at 2 CFU in
order for an organ to be considered positive for the presence of
BCG bacilli.
2.4. Multiparameter flow cytometry
Axillary and inguinal lymph node homogenates from individ-
ual mice were pooled for intracellular cytokine staining analysis by
multiparameter flow cytometry. Four mice were used per vaccine
group to determine the frequency of CD4+ or CD8+ multifunc-
tional T cells (MFT cells) induced for each vaccine preparation at
different time points post-immunization. Lymph node cells were
isolated by disrupting the lymph nodes with a 3 cc syringe barrel
in complete DMEM (cDMEM) consisting of 10 mM HEPES, 2.0 mM
l-glutamine, 0.1 mM MEM non-essential amino acids with 10%
fetal bovine serum (FBS). After passing the homogenates through
a 70 ␮m cell strainer, the resulting single cell suspension was
washed with cDMEM followed by plating 2 million cells per well
of a 96-well plate in 0.2 ml. For measurement of antigen-specific
responses, BCG Pasteur was added to wells at a multiplicity of
infection (MOI) of 0.02 bacilli per lymph node cell. Wells con-
taining cells without BCG served as unstimulated controls. After
128 S.C. Derrick et al. / Vaccine 33 (2015) 126–132
a 36 h incubation, Golgiplug (BD Biosciences, San Jose, CA) was
added (1 ␮l per well) followed by an additional 4 h incubation.
Unbound cells were then removed from the wells and transferred
to 12 mm × 75 mm tubes, washed with PBS and resuspended in
50 ␮l PBS. Near-IR live-Dead stain (Invitrogen, Carlsbad, CA) (10 ␮l
of a 1:100 dilution) was added to each tube and incubated for
30 min at 4 ◦ C to allow for gating on viable cells. After washing
the cells with PBS with 2% fetal bovine serum (FBS) (PBS-FBS),
antibody against CD16/CD32 (Fc␥III/II receptor, clone 2.4G2) (Fc
block) was added in a volume of 50 ␮l and incubated at 4 ◦ C for
15 min. The cells were then stained for 30 min at 4 ◦ C by adding
antibodies against the CD4 (rat anti-mouse CD4 Alexa Fluor 700
[AF-700] Ab, clone RM4-5), and CD8 (rat anti-mouse CD8 peridinin
chlorophyll protein complex [PerCP] Ab, clone 53-6.7) proteins at
0.1 and 0.4 ␮g per tube respectively. Following the incubation,
the cells were washed twice with PBS and then fixed for 30 min
at 4 ◦ C with 2% paraformaldehyde in PBS. After fixing, the cells
were pelleted, washed twice with PBS-FBS and stored at 4 ◦ C. Fixed
cells were washed twice with perm-wash buffer (1% FBS, 0.01 M
HEPES, 0.1% saponin in PBS) followed by intracellular staining using
the following antibodies at 0.2 ␮g per tube: rat anti-mouse IFN-␥
phycoerythrin [PE] Ab, clone XMG1.2; rat anti-mouse TNF-␣ fluo-
rescein isothiocyanate [FITC] Ab, clone MP6-XT22; rat anti-mouse
IL-2 allophycocyanin [APC] Ab, clone JES6-5H4. The cells were incu-
bated at 4 ◦ C for 30 min, washed twice with perm-wash buffer and
then twice with PBS-FBS. All antibodies were obtained from BD
Biosciences.
The cells were analyzed using an LSRII flow cytometer (Bec-
ton Dickinson) and FlowJo software (Tree Star Inc., Ashland, OR).
We acquired 250,000 events per sample and then, using FlowJo,
gated on live, single cell lymphocytes. To determine the frequency
of different populations of multifunctional T cells (MFT cells),
we gated on CD4+ or CD8+ T cells staining positive for TNF-␣
and IFN-␥, TNF-␣ and IL-2, IFN-␥ and IL-2 or all three cytokines
(TP).
2.5. Statistical analysis
The Graph Pad Prism 5 program was used to analyze the data for
these experiments (Graph Pad Software, San Diego, CA). The organ
CFU data were analyzed using the Mann Whitney test and the flow
cytometry results were evaluated using t-test analysis.
3. Results
3.1. Retrospective analysis of BCG-induced protection in a mouse
model of pulmonary tuberculosis
To provide a context for the organ distribution studies, we
retrospectively evaluated protection data generated in our labo-
ratory during the past decade using the BCG Pasteur strain as a
positive vaccination control in our mouse model of pulmonary
tuberculosis. During this 10 year time period, we conducted 41
immunization experiments using subcutaneous vaccination with
BCG Pasteur as a control. Our retrospective analysis of the results of
these studies showed that mean CFU reductions of 1.16 ± 0.27 log10
and 1.20 ± 0.33 log10 , respectively, were detected in the lungs and
spleens of BCG vaccinated mice (compared to naïve controls) 1
month following an aerosol challenge with virulent M. tuberculo-
sis Erdman. The BCG-induced protective responses were durable
with significant protective responses seen more than 1 year post-
challenge. Importantly, every mouse that was vaccinated with BCG
and then challenged with M. tuberculosis in the past decade was
significantly protected against the TB infection.
3.2. Temporal organ distribution of mycobacteria after
vaccination with adjuvanted and nonadjuvanted BCG vaccine
preparations
Mice were vaccinated s.c. with 106 CFU of BCG Pasteur or the
same dose of BCG Pasteur mixed with DDA/TDB adjuvant. At
time periods up to 1 year post-vaccination, 5–10 mice per group
were sacrificed, relevant organ were removed, and entire organ
homogenates were plated individually on Middlebrook 7H11 agar
plates. Our range of detection for this assay was 1–400 BCG bacilli
per organ. Figs. 1–3 show the temporal distribution of mycobacte-
ria in the lungs, spleens, and lymph nodes following vaccination.
(These data are representative to two studies which yielded essen-
tially equivalent results.) Mycobacterial trafficking to the liver after
s.c. BCG vaccination was also evaluated but there was a virtual
absence of BCG bacilli detected in the mouse liver at every time
point.
As seen in Fig. 1, mycobacteria were detected in the lung 3 days
after s.c. vaccination with BCG. At this early time point, reduced
trafficking to the lung was seen for the adjuvanted BCG preparation
(p < 0.05). Subsequently, the frequency of animals with viable BCG
organisms and the mycobacterial CFU levels within a lung group
was highly variable. For instance, at day 30 post-vaccination, only
30% of BCG-vaccinated and 60% of the adjuvanted BCG immunized
mice had measurable BCG in the lungs, while at 6 months post-
vaccination BCG was detected in 100% and 70%, respectively, of
the BCG and adjuvanted BCG immunized mice. Surprisingly, high
frequencies of mice with persisting mycobacterial lung infections
were seen in both the BCG (90%) and the adjuvanted BCG (80%)
groups at 1 year post-vaccination, although the median CFU levels
for each group was relatively modest. Overall, BCG infections were
not detected in the lungs of about one-third of vaccinated mice
in this study, and the CFU levels for about 25% of mice exceeded
100 CFU. The highest median for BCG immunized animals was
detected in the lungs at 6 months post-vaccination while no signif-
icant peak values were seen in mice vaccinated with the BCG/Adj
formulation.
The spleen biodistribution results of this 1 year study were
similar to the lung data (Fig. 2). For both immunization groups,
mycobacteria could be detected in some animals by day 3 after
vaccination. Persisting BCG bacilli were detected in the spleen at
1 year post-vaccination in both the BCG (50%) and the adjuvanted
BCG (90%) groups. As seen in mouse lungs, the frequency and lev-
els of bacteria present in vaccinated animals varied considerably
with no mycobacteria being detected in the spleens of about half
of the mice while nearly 15% had splenic CFU levels that reached
the upper detection limit. For BCG immunized mice, the highest
median spleen CFU values were detected at 6 months after the vac-
cination (p < 0.05). In contrast, the peak median CFU value for the
BCG/Adj group was not seen until 12 months post-immunization.
The post-vaccination BCG distribution patterns were consid-
erably different in the lymph nodes (Fig. 3). Here, BCG infections
following vaccination were seen more consistently. Mycobacterial
lymph node infections were observed in more than 90% of vacci-
nated mice and about 40% of these mice had lymph node CFU values
exceeding more than 100 CFU. Most importantly, the overall infec-
tion rate in the lymph nodes exceeded the frequency of lung and
spleen infections with BCG being cultured from the lymph nodes
of 96% of mice vaccinated with the BCG adjuvant formulation and
84% of mice immunized with a conventional BCG Pasteur prepara-
tion during this 1 year study. Interestingly, the median lymph node
CFU value for BCG vaccinated mice was highest at the early day 15
time point (p < 0.05). In contrast, the median lymph node CFU for
the BCG/Adj group did not vary significantly during the 1 year study
but did exceed the medians for BCG vaccinated animals at 1, 4 and
6 months after the immunization.
 S.C. Derrick et al. / Vaccine 33 (2015) 126–132 129

Fig. 1. BCG infections of mouse lungs following subcutaneous immunization. At day 3, 15, 30, 120, 180 and 1 year post-vaccination, the lungs from mice (10 per time point)
immunized with 1 × 106 CFU BCG or BCG formulated with DDA + TDB adjuvant (BCG + Adj) were removed, homogenized and aliquots were added to 7H11-OADC agar plates
to determine the number of BCG CFU. Horizontal bars indicate the median CFU per organ. Percentages correspond to the frequency of mice that exhibited detectable CFU in
the organ. *Significantly different median organ CFU relative to mice immunized with BCG/Adj (p < 0.05).

3.3. Flow cytometric analysis of lymph node cells from BCG of T cells expressing all three cytokines were not detected at any
vaccinated mice study time point for the BCG/Adj group compared to BCG alone.

Since mycobacteria were consistently detected with high fre-

quency in lymph nodes (and not other organs tested) of BCG
vaccinated mice, we evaluated the relative immune lymph node T
cell responses up to 180 days post-vaccination using multiparame-
ter flow cytometry. At specific time points after BCG immunization,
the lymph nodes were removed, the cells were incubated with BCG
for 36 h, the cells were stained for T cell markers and intracellular
cytokine expression, and the stained cells were analyzed using flow
cytometry. At 15 days post-vaccination, elevated frequencies of
CD4+ and CD8+ single positive T cells expressing IFN-␥ and double
positive cells expressing IFN-␥ and TNF-␣ were detected for both
vaccination groups (Fig. 4). Among these T cell subtypes, only the
frequency of adjuvanted BCG CD8+ T cells expressing IFN-␥ (1.9%)
statistically exceeded the levels of similar cells (1.0%) from mice
vaccinated with conventional BCG. For the 30 and 120 day time
points, statistically higher frequencies of CD4+ and CD8+ T cells
expressing IFN-␥ and IFN-␥/TNF-␣ were detected from the adju-
vanted BCG immunized mice. It should be noted that at 120 days
post-vaccination, exceedingly high levels of CD4+ (3.2%) and CD8+
(5.2%) IFN-␥+ lymph node T cells were observed in the adjuvanted
BCG group. Even at 180 days post-vaccination, elevated levels of
CD4+ and CD8+ T cells expressing IFN-␥ and IFN-␥/TNF-␣ were seen
in both vaccination groups. Overall, during this 6 month study, con-
sistently higher frequencies of CD4+ and CD8+ T cells expressing
either IFN-␥ or IFN-␥ and TNF-␣ were detected in lymph node cells
from mice vaccinated with BCG formulated in DDA/TDB adjuvant
relative to nonadjuvanted BCG. Significantly elevated frequencies
4. Discussion
In recent years, the formulation of BCG in lipid-containing com-
pounds has been shown to increase the vaccine’s immunogenicity
[11–16]. However, the mechanisms associated with the increased
protection induced by lipidated BCG have not been clearly defined.
It has been shown previously that the anti-tuberculosis immunity
evoked by BCG immunization is largely dependent on the presence
of live organisms [17–19]. Further, the prolonged persistence of
BCG bacilli is likely critical for the development of strong adaptive
immune responses. To develop a greater understanding of why
BCG formulated in lipid-containing compounds is generally more
effective than BCG alone, we evaluated the organ distribution of
BCG bacilli for 1 year after vaccination with adjuvanted and non-
adjuvanted BCG preparations. Our organ biodistribution studies
yielded three major findings. First, we showed that BCG Pasteur
persists at low levels for at least 1 year post-vaccination with either
conventional BCG or adjuvanted BCG preparations. Between 50 and
100% of mouse spleens, lungs, and lymph nodes remained infected
with BCG 1 year following subcutaneous BCG immunization. While
some earlier studies indicated that BCG was cleared 4–5 months
post-vaccination, other investigators have reported that BCG
infections can persist in mice for at least 1 year after vaccination
[18,20,21,23]. For example, persistence of orally administered
lipidated BCG in the lymph nodes of whitetail deer was found up to
12 months [23]. Additionally, Leversen et al. [21] recently reported
130 S.C. Derrick et al. / Vaccine 33 (2015) 126–132

Fig. 2. BCG infections of mouse speens following subcutaneous immunization. At day 3, 15, 30, 120, 180 and 1 year post-vaccination, the spleens from mice (10 per time
point) immunized with 1 × 106 CFU BCG or BCG formulated with DDA + TDB adjuvant (BCG + Adj) were removed, homogenized and aliquots were added to 7H11-OADC agar
plates to determine the number of BCG CFU. Horizontal bars indicate the median CFU per organ. Percentages correspond to the frequency of mice that exhibited detectable
CFU in the organ. *Significantly different median organ CFU relative to mice immunized with BCG/Adj (p < 0.05).

Fig. 3. BCG infections of mouse lymph nodes following subcutaneous immunization. At day 3, 15, 30, 120, 180 and 1 year post-vaccination, the axillary and inguinal lymph
nodes were pooled from each mouse (4–10 per time point) immunized with 1 × 106 CFU BCG or BCG formulated with DDA + TDB adjuvant (BCG + Adj). Homogenates were
then added to 7H11-OADC agar plates to determine the number of BCG CFU. Horizontal bars indicate the median CFU per organ. Percentages correspond to the frequency of
mice that exhibited detectable CFU in the organ. *Significantly different median organ CFU relative to mice immunized with nonadjuvanted BCG (p < 0.05).
 S.C. Derrick et al. / Vaccine 33 (2015) 126–132 131

Fig. 4. Multiparameter flow cytometry analysis of CD4+ (A) and (B) or CD8+ (C) and (D) lymph node T cells. Axillary and inguinal lymph nodes were pooled from each mouse
at day 15, 30, 120 and 180 post-immunization and incubated with BCG in vitro. The frequencies (%) of T cells producing only IFN-␥ (A) and (C) or cells producing both IFN-␥
and TNF-␣ (IFN/TNF) (B) and (D) were determined. *Significant differences relative to nonadjuvanted BCG (p < 0.05).

that BCG Pasteur reached a low level persistence in the lungs of mice
at 8 months post-vaccination and persisted until the end of the 17
month experiment. Interestingly, in Leversen’s study, BCG Russia
was cleared from the lungs by 8 months post-vaccination. Cer-
tainly, the inconsistency of the BCG persistence data in mice may be
largely due to strain differences. While the biological mechanisms
responsible for enhanced persistence are unknown, current data
suggests that BCG Pasteur is among the most persistent BCG strains
[21,24]. It should be emphasized that this long-term persistence
of BCG in mouse models is consistent with clinical reports where
disseminated BCG disease has been seen in immunocompromised
individuals at least 30 years after BCG vaccination [25].
Second, the frequency of detection of bacilli in mouse lungs and
spleens at varying time periods after subcutaneous immunization
was highly variable and generally did not differ significantly
among the adjuvanted and nonadjuvanted BCG vaccination groups
during the 1 year study. Surprisingly, at specific time points
post-vaccination, BCG bacilli were not recovered from 70% of lungs
and all spleens tested. The variable biodistribution profiles and the
absence of detectable BCG bacilli in a significant fraction of mouse
lungs and spleens post-vaccination are similar to the mycobacterial
biodistribution data reported in earlier BCG vaccination studies
[21,26]. Unexpectedly, we consistently observed peak median
CFU values in the lungs and spleens of BCG immunized mice at
6 months post-vaccination such that 90–100% of these mice had
elevated numbers of bacilli in these organs. Similar peak CFU values
have been reported at 3–4 months post-vaccination. It is unclear
whether these high CFU values resulted from immune-related
mechanisms (T cell exhaustion, etc.) or possibly the selection
and growth of more persistent BCG organisms [27]. It should be
noted most of the initial BCG inoculum is not recovered from the
organs of vaccinated mice. Investigators who have reported similar
findings have suggested that most of the vaccinating bacilli are sim-
ply eliminated by physical or innate immune mechanisms [18,24].
Third, in contrast to the variable lung and spleen results,
mycobacteria were detected in most lymph nodes analyzed for
at least 1 year post-vaccination with either the non-adjuvanted
or adjuvanted BCG preparations. Also, the numbers of BCG
bacilli detected in the lymph nodes after immunization with the
adjuvanted and conventional BCG formulations generally were ele-
vated relative to the BCG CFU found in lung, spleen, and liver
homogenates. The consistency of the lymph node infections was
not unexpected. In fact, more than 70 years ago, the noted pathol-
ogist Arnold Rich stated that mycobacteria multiply more freely
in the lymph nodes than at the site of primary infection [28].
Since then, reports of mycobacterial lymph node infections of mice,
cattle, deer, monkeys, and humans have been widespread. For
example, cervical lymphadenitis has been shown to be a com-
mon disease of humans caused by Mycobacterium avium infections
and M. bovis lymph node infection of cattle and deer have been
frequently reported [29–31]. After a bronchoscopic infection of
cynomolgus monkeys with M. tuberculosis, all test animals had
lymph node involvement while less than half had detectable lung
infections [32]. Given the extensive literature on mycobacterial
infections of lymph nodes, Behr and Waters [33] speculated that
tuberculosis is actually a lymphatic disease with a pulmonary portal
for the infection.
The high frequency of BCG lymph node infections seen after
vaccination is relevant because Wolf et al. [34] have demonstrated
that the induction of immunity after a M. tuberculosis infection of
mice correlated with the delivery of mycobacteria to the appropri-
ate lymph node. In our study, although peak median CFU values
were detected in the lymph nodes of BCG vaccinated mice early
after vaccination (day 15), significantly higher median values were
132 S.C. Derrick et al. / Vaccine 33 (2015) 126–132
seen in the BCG/Adj group at several later time points. To further
examine the cellular mechanisms associated with the increased
protective responses induced by adjuvanted BCG, lymph node cells
from BCG vaccinated mice were analyzed by multiparameter flow
cytometry. Our intracellular cytokine staining results indicated that
the adjuvanted BCG mixture was more effective than conventional
BCG at inducing antigen-specific CD4+ and CD8+ T cells. During a
6 month post-vaccination study period, increased frequencies of
CD4+ and CD8+ T cells expressing IFN-␥ and IFN-␥/TNF-␣ were
detected in lymph node cells recovered from mice vaccinated with
the adjuvanted BCG preparation (relative to the conventional BCG
controls). In particular, after vaccination with the adjuvanted BCG
mixture, the frequencies of CD4+ and CD8+ T cells expressing IFN-␥
was exceedingly high at 120 and 180 days post-vaccination (3–5%
of the total CD4+ or CD8+ T cells). The enhanced CD8+ T cell
responses in the adjuvanted BCG group were especially interest-
ing because conventional BCG has been shown to have a limited
capacity to stimulate CD8-based anti-tuberculosis immunity [35].
The elevated CD8+ T cell responses observed after vaccination with
adjuvanted BCG is consistent with the enhanced long-term pro-
tection seen with this formulation because in mice CD8+ T cells
seem to be important for controlling the chronic phase of M. tuber-
culosis infections [36]. It is likely that the higher median BCG CFU
values detected in the lymph nodes of mice immunized with the
BCG/Adjuvant formulation at 1–6 months post-vaccination may
contribute to the elevated pro-inflammatory immune responses
and the increased long-term protective immunity seen in these
animals (relative to mice vaccinated with conventional BCG).
Overall, we have demonstrated that while the levels and fre-
quencies of BCG infections in mouse lungs and spleens after
a subcutaneous immunization with either adjuvanted or non-
adjuvanted BCG preparations were highly variable, BCG bacilli
were consistently detected in the lymph nodes for at least 1 year.
Given the consistency of lymph node infections and the obser-
vation that every BCG vaccinated mouse has been significantly
protected against an aerosol M. tuberculosis challenge during a
decade of experiments in our laboratory, lymph node involvement
after BCG vaccination is likely critical in the development of a per-
sistent and strong anti-tuberculosis response. The elevated CD4+
and CD8+ T cell responses detected in the lymph nodes of mice vac-
cinated with the potent adjuvanted BCG preparations support the
importance of lymph node T cell activation in vaccine-induced anti-
tuberculosis protective immunity. Clearly, further studies focusing
on the relationship between post-BCG vaccination lymph node
involvement and the generation of an effective anti-tuberculosis
protective response are warranted.
References
[1] WHO. Global tuberculosis report 2013. Geneva: World Health Organization;
2013. http://www.who.int/tb/publications/global report/en/
[2] Orme IM. The Achilles heel of BCG. Tuberculosis 2010;90:329–32.
[3] Colditz GA, Brewer TF, Berkey CS, Wilson ME, Burdick E, Fineberg HV, et al.
Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the
published literature. JAMA 1994;271:698–702.
[4] Colditz GA, Berkey CS, Mosteller F, Brewer TF, Wilson ME, Burdick E, et al.
The efficacy of bacillus Calmette–Guerin vaccination of newborns and infants
in the prevention of tuberculosis: meta-analyses of the published literature.
Pediatrics 1995;96:29–35.
[5] Fine PE. BCG: the challenge continues. Scand J Infect Dis 2001;33:243–5.
[6] Aronson NE, Santosham M, Comstock GW, Howard RS, Moulton LH, Rhoades
ER, et al. Long-term efficacy of BCG vaccine in American Indians and Alaska
Natives: a 60-year follow-up study. JAMA 2004;291:2086–91.
[7] Kristensen I, Aaby P, Jensen H. Routine vaccinations and child survival: follow
up study in Guinea-Bissau, West Africa. BMJ 2000;321:1435–8.
[8] Roth A, Gustafson P, Nhaga A, Djana Q, Poulsen A, Garly ML, et al. BCG vacci-
nation scar associated with better childhood survival in Guinea-Bassau. Int J
Epidemiol 2005;34:540–7.
[9] Ottenhoff TH. Vaccines against tuberculosis: where are we and where do we
need to go? PLoS Pathog 2012;8:e1002607.
[10] Kaufmann SH. Tuberculosis vaccine development at a divide. Curr Opin Pulm
Med 2014;20:294–300.
[11] Aldwell FE, Baird MA, Fitzpatrick CE, McLellan AD, Cross ML, Lambeth MR, et al.
Oral vaccination of mice with lipid-encapsulated Mycobacterium bovis BCG:
anatomical sites of bacterial replication and immune activity. Immunol Cell
Biol 2005;83:549–53.
[12] Buddle BM, Aldwell FE, Skinner MA, de Lisle GW, Denis M, Vordermeier
HM, et al. Effect of oral vaccination of cattle with lipid-formulated BCG
on immune responses and protection against bovine tuberculosis. Vaccine
2005;23:3581–9.
[13] Wedlock DN, Aldwell FE, Keen D, Skinner MA, Buddle BM. Oral vaccination
of brushtail possums (Tichosurus vulpecula) with BCG: immune responses,
persistence of BCG in lymphoid organs and excretion in faeces. N Z Vet J
2005;53:301–6.
[14] Aldwell FE, Cross ML, Fitzpatrick CE, Lambeth MR, de Lile GW, Buddle BM. Oral
delivery of lipid-encapsulated Mycobacterium bovis BCG extends survival of the
bacillus in vivo and induces a long-term protective immune response against
tuberculosis 2006;24:2071–8.
[15] Ancelet LR, Aldwell FE, Rich FJ, Kirman JR. Oral vaccination with lipid-
formulated BCG induces a long-lived, multifunctional CD4(+) T cell memory
immune response. PLoS ONE 2012;7:e45888.
[16] Derrick SC, Dao D, Yang A, Kolibab K, Jacobs WR, Morris SL. Formulation
of a mmaA4 gene deletion mutant of Mycobacterium bovis BCG in cationic
liposomes significantly enhances protection against tuberculosis. PLoS ONE
2012;7:e32959.
[17] Tuberculosis Program, Public Health Service, USA. Experimental studies of vac-
cination, allergy, and immunity in tuberculosis: 3. Effect of killed BCG vaccines.
Bull World Health Organ 1955;12:47–62.
[18] Olsen AW, Brandt L, Agger EM, van Pinxteren LA, Andersen P. The influence
of remaining live BCG organisms in vaccinated mice on the maintenance of
immunity to tuberculosis. Scand J Immunol 2004;60:273–7.
[19] Parra M, Yang AL, Lim J, Kolibab K, Derrick S, Cadieux N, et al. Development of
a murine mycobacterial growth inhibition assay for evaluating vaccine against
Mycobacterium tuberculosis. Clin Vaccine Immunol 2009;16:1025–32.
[20] Saxena RK, Weissman D, Simpson J, Lewis DM. Murine model of BCG lung infec-
tion: dynamics of lymphocyte subpopulations in lung interstitium and tracheal
lymph nodes. J Biosci 2002;27:143–53.
[21] Leversen NA, Sviland L, Wiker HG, Mustafa T. Long-term persistence of BCG Pas-
teur in lungs of C57BL/6 mice following intranasal infection. Scand J Immunol
2012;75:489–99.
[22] Henriksen-Lacey M, Bramwell VW, Christensen D, Agger E-M, Andersen P, Per-
rie Y. Liposomes based on dimethyldioctadecylammonium promote a depot
effect and enhance immunogenicity of soluble antigen. J Control Release
2010;142:180–6.
[23] Palmer MV, Thacker TC, Waters WR, Robe-Austerman S, Aldwell FE. Persistence
of Mycobacterium bovis bacillus Calmette–Guerin (BCG) Danish in white-tailed
deer (Odocoileus virginianus) vaccinated with a lipid-formulated oral vaccine.
Transbound Emerg Dis 2014;61:266–72.
[24] Lagranderie MR, Balazuc AM, Deriaud E, Leclerc CD, Gheorghiu M. Comparison
of immune responses of mice immunized with five different Mycobacterium
bovis BCG vaccine strains. Infect Immun 1996;64:1–9.
[25] Armbruster C, Junker W, Vetter N, Jaksch G. Disseminated Bacille
Calmette–Guerin infection in an AIDS patient 30 years after BCG vaccination. J
Infect Dis 1990;162:1216.
[26] Irwin SM, Goodyear A, Keyser A, Christensen R, Troudt JM, Taylor JL, et al.
Immune response induced by three Mycobacterium bovis BCG substrains with
diverse regions of deletion in a C57BL/6 mouse model. Clin Vaccine Immunol
2008;15:750–6.
[27] Domingue Sr GJ, Woody HB. Bacterial persistence and expression of disease.
Clin Microbiol Rev 1997;10:320–44.
[28] Rich AR. The pathogenesis of tuberculosis. 1st ed. Springfield, IL: Charles C.
Thomas; 1944.
[29] Engbaek HC. Lymph gland processes caused by atypical mycobacteria and
Mycobacterium avium. Bacteriologically verified cases in Denmark from 1935
to 1961. Acta Tuberc Pneumol Scand 1964;44:108–37.
[30] Waters WR, Palmer MV, Nonnecke BJ, Thacker TC, Scherer CF, Estes DM, et al.
Efficacy and immunogenicity of Mycobacterium bovis deltaRD1 against aerosol
M. bovis infection in neonatal calves. Vaccine 2009;27:1201–9.
[31] Palmer MV, Waters WR, Thacker TC. Lesion development and immunohis-
tochemical changes in granulomas from cattle experimentally infected with
Mycobacterium bovis. Vet Pathol 2007;44:863–74.
[32] Lin PL, Rodgers M, Smith L, Bigbee M, Myers A, Bigbee C, et al. Quantitative com-
parison of active and latent tuberculosis in the cynomolgus macaque model.
Infect Immun 2009;77:4631–42.
[33] Behr MA, Waters WR. Is tuberculosis a lymphatic disease with a pulmonary
portal. Lancet Infect Dis 2014;14:250–5.
[34] Wolf AJ, Desvignes L, Linas B, Banaiee N, Tamura T, Takatsu K, et al. Initia-
tion of the adaptive immune response to Mycobacterium tuberculosis depends
on antigen production in the local lymph node, not the lungs. J Exp Med
2008;205:105–15.
[35] Kaufmann SH. Is the development of a new tuberculosis vaccine possible. Nat
Med 2000;6:955–60.
[36] Flynn JL, Goldstein MM, Triebold KJ, Koller B, Bloom BR. Major histocom-
patibility complex class I-restricted T cells are required for resistance to
Mycobacterium tuberculosis infection. Proc Natl Acad Sci 1992;89:12013–7.
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.