680
Eur. J. Immunol. 2008. 38: 680–694
Christoph Hölscher et al.
Containment of aerogenic Mycobacterium tuberculosis
infection in mice does not require MyD88 adaptor
function for TLR2, -4 and -9
Christoph Hölscher1, Norbert Reiling2, Ulrich E. Schaible3,4, Alexandra Hölscher1,2, Clara Bathmann1,
Daniel Korbel4, Insa Lenz1, Tanja Sonntag1, Svenja Kröger2, Shizuo Akira5, Horst Mossmann6,
Carsten J. Kirschning7, Hermann Wagner7, Marina Freudenberg6 and Stefan Ehlers2,8
1
2
3
4
5
6
7
8
Junior Research Group Molecular Infection Biology, Research Center Borstel, Borstel, Germany
Division of Molecular Infection Biology, Research Center Borstel, Borstel, Germany
Max-Planck-Institute for Infection Biology, Berlin, Germany
London School of Hygiene and Tropical Medicine, ITD Immunology Unit, London, UK
Department of Host Defense, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan
Max-Planck-Institute for Immunobiology, Freiburg, Germany
Department of Medical Microbiology and Hygiene, Technical University of Munich, Munich, Germany
Molecular Inflammation Medicine, Christian-Albrechts-University, Kiel, Germany
The role of Toll-like receptors (TLR) and MyD88 for immune responses to Mycobacterium
tuberculosis (Mtb) infection remains controversial. To address the impact of TLRmediated pathogen recognition and MyD88-dependent signaling events on antimycobacterial host responses, we analyzed the outcome of Mtb infection in TLR2/4/9
triple- and MyD88-deficient mice. After aerosol infection, both TLR2/4/9-deficient and
wild-type mice expressed pro-inflammatory cytokines promoting antigen-specific T cells
and the production of IFN-c to similar extents. Moreover, TLR2/4/9-deficient mice
expressed IFN-c-dependent inducible nitric oxide synthase and LRG-47 in infected lungs.
MyD88-deficient mice expressed pro-inflammatory cytokines and were shown to expand
IFN-c-producing antigen-specific T cells, albeit in a delayed fashion. Only mice that were
deficient for MyD88 rapidly succumbed to unrestrained mycobacterial growth, whereas
TLR2/4/9-deficient mice controlled Mtb replication. IFN-c-dependent restriction of
mycobacterial growth was severely impaired only in Mtb-infected MyD88, but not in
TLR2/4/9-deficient bone marrow-derived macrophages. Our results demonstrate that
after Mtb infection neither TLR2, -4, and -9, nor MyD88 are required for the induction of
adaptive T cell responses. Rather, MyD88, but not TLR2, TLR4 and TLR9, is critical for
triggering macrophage effector mechanisms central to anti-mycobacterial defense.
Received 3/7/07
Revised 20/11/07
Accepted 11/1/08
[DOI 10.1002/eji.200736458]
Key words:
Bacterial infection
Monocyte/
macrophage
Mycobacterium
Rodent TLR
Supporting Information for this article is available at
http://www.wiley-vch.de/contents/jc_2040/2008/36458_s.pdf
Correspondence: Christoph Hlscher, Junior Research Group
Molecular Infection Biology, Research Center Borstel, Borstel,
Germany
Fax: +49-4537-188775
e-mail: choelscher@fz-borstel.de
Abbreviations: CBA: cytometric bead array DexTR: Dextran
Texas Red LAMP-1: lysosome associated membrane protein 1
Mtb: Mycobacterium tuberculosis NOS2: inducible nitric oxide
synthase p.i.: post infection PRR: pattern recognition
receptor RNI: reactive nitrogen intermediates
TB: tuberculosis Tf: transferrin
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Introduction
Human tuberculosis (TB) is a leading global health
threat [1]. The disease caused by Mycobacterium
tuberculosis (Mtb) is currently responsible for 8 million
new cases worldwide and 1.7 million deaths annually
[2]. The increased occurrence of multi-drug resistant
strains of Mtb and the lack of an effective vaccine
emphasize the need for TB research at the interface of
innate and adaptive immunity to develop new prewww.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
ventive and therapeutic strategies. Replication of
mycobacteria takes place inside phagosomes of the
primary host cell, the macrophage, until the onset of
adaptive immunity. Triggered by interleukin (IL)-12,
CD4+ Th1 cells play a pivotal role by secreting interferon
(IFN)-c [3–5], inducing effector immune responses in
macrophages that eventually result in containing
infectious foci [6, 7]. In particular, IFN-c and TNF-a
stimulate the antimicrobial activity of infected macrophages, allowing intracellular bacterial killing through
phagosome maturation [8], the induction of LRG-47, a
member of the 47-kDa guanosine triphosphatase family
probably involved in autophagosome formation [9, 10],
and the effect of reactive nitrogen intermediates (RNI),
which are produced by the inducible nitric oxide
synthase (NOS2) [11].
Activation of the innate immune system is considered
to be a prerequisite for driving a protective adaptive
immune response after Mtb infection. Initial recognition
of mycobacterial components by macrophages and
dendritic cells leads to the release of IL-12 and TNF-a
and involves a number of different pattern-recognition
receptors (PRR) such as Toll-like receptors (TLR).
Several reports have described TLR2-, TLR4- or TLR9dependent activation of macrophages and dendritic cells
by mycobacterial components. For example, the mycobacterial lipoglycans phosphoinositol-capped lipoarabinomannan, phosphatidyl myo-inositol mannosides
(PIM)2 and PIM6 and the 19-kDa mycobacterial
lipoprotein are TLR2 agonists [12–15]. TLR4 mediates
cellular activation in response to not yet fully characterized heat-sensitive cell-associated mycobacterial factors
[16], and the Mtb-induced TNF-a production by murine
macrophages can be blocked by a TLR4 antagonist [17].
The genome of Mtb as well as other mycobacteria
contains highly immunostimulatory CpG motifs [18],
and the cognate receptor TLR9 has also been shown to
be a relevant PRR for Mtb [19]. Together, TLR2, TLR4,
and TLR9 all appear to be involved in the innate
induction of a cell-mediated immune responses after
infection with Mtb.
Despite severely reduced pro-inflammatory responses to Mtb in macrophages from TLR2- or TLR4defective mice in vitro, mice deficient in either or both
receptors are fully capable of containing low-dose
aerosol Mtb infection [20, 21]. However, differences
in experimental models have yielded conflicting results,
ranging from enhanced mortality of TLR2-, TLR4-, or
TLR9 single- or double-deficient mice in response to Mtb
infection to little or no apparent change in resistance
[19–25]. It is unknown whether differential expression
and/or cross-regulation of TLR in the diverse settings
may account for these discrepancies. Regardless, the
prevailing consensus is that resistance to Mtb infection
depends on microbe-sensing through TLR [26, 27]. To
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunity to infection
date, however, mice lacking all TLR to which Mtb
products have been assigned as ligands such as TLR2/4/
9-triple-deficient mice have not yet been subjected to
experimental Mtb infection and analyzed for their
relative susceptibility to infection.
The intracellular adaptor molecule MyD88 transduces signals from all TLR except TLR3 by recruitment of
members of the IL-1R-associated protein kinase family
of proteins, which results in activation of mitogenactivated protein kinases (MAPK), PI3 kinase, and
nuclear factor (NF)-jB [28]. Because MyD88 integrates
signals from multiple TLR, MyD88 deficiency almost
completely ablates macrophage responsiveness to specific bacterial products [29]. In vivo, MyD88-deficient
(–/–) mice are highly susceptible to infection with Listeria
monocytogenes, Staphylococcus aureus and Toxoplasma
gondii [30–32]. Similarly, MyD88–/– mice rapidly
succumb to Mycobacterium avium and Mtb infection
[27, 33, 34]. Whereas some studies imply that both
MyD88 and TLR are critical for the activation of innate
immunity and the differentiation of adaptive defenses
[27], Fremond and colleagues [34] have shown that
MyD88 is not required for the induction of innate proinflammatory and subsequent acquired immune responses.
In the present study, we analyzed the potentially
differential requirement of TLR-mediated recognition
vs. MyD88-mediated signaling in the development and
expression of protective immunity to aerogenic Mtb
infection. Our investigation of TLR2/4/9–/– and
MyD88–/– mice showed that triple TLR-deficient mice
were fully capable of inducing adaptive Th1 immune
responses and surviving low-dose aerosol infection,
while MyD88-deficient mice rapidly succumbed to
infection. Thus, TLR2, -4, and -9 were dispensable for
the development of protective immune responses to Mtb
infection. Moreover, susceptibility of MyD88–/– mice
was not attributable to defective TLR2/4/9-mediated
pattern recognition and instruction of Th1-biased
adaptive immunity. Rather, MyD88, but not TLR2,TLR4
and TLR9, proved critical for shaping macrophage antimycobacterial effector mechanisms.
Results
Cytokine production by TLR2/4/9–/– macrophages
is impaired in response to TLR agonists and
Mtb infection
Bone marrow-derived macrophages (BMMU) were
generated from wild-type, MyD88–/– and TLR2/4/9–/–
mice. The release of TNF-a and IL-12/IL-23p40 was
determined 24 h after challenge with lipopeptide (LP),
LPS, or CPG, or infection with Mtb H37Rv at different
www.eji-journal.eu
681
682
Eur. J. Immunol. 2008. 38: 680–694
Christoph Hölscher et al.
Figure 1. After stimulation with TLR agonists or after Mtb infection, cytokine production is impaired in macrophages from
MyD88–/– and TLR2/4/9–/– mice. BMMU from wild-type, MyD88–/– and TLR2/4/9–/– mice were stimulated with medium, different TLR
agonists, or were incubated with the indicated MOI of viable Mtb H37Rv. Supernatants were harvested 24 h p.i. and measured for
TNF-a and IL-12/IL-23p40 concentrations. Each point indicates mean and SD of triplicate values from one out of two representative
experiments. Statistical analysis was performed by ANOVA defining differences between wild-type and MyD88–/– (*p0.05;
**p0.01; ***p0.001) or TLR2/4/9–/– (+p0.05; ++p0.01; +++p0.001) mice as significant.
doses (Fig. 1). In response to all three TLR agonists, both
MyD88–/– and TLR2/4/9–/– BMMU released significantly less of either cytokine. Whereas MyD88–/– cells
released no TNF-a after infection with Mtb, responsiveness in TLR2/4/9–/– macrophages was impaired but not
abolished. In contrast, even high amounts of mycobacteria did not induce release of substantial amounts of
IL-12/IL-23p40 from MyD88–/– and TLR2/4/9–/–
macrophages. Together, our analysis revealed that
TLR2, -4, and -9 are required to induce an efficient
inflammatory immune response in macrophages in
vitro.
In contrast to TLR2/4/9–/– mice, MyD88–/–
animals are highly susceptible to Mtb infection
Bacterial loads in the lungs of TLR2-, TLR4- and TLR9single-deficient mice, as well as TLR2/4-double-deficient mice, were similar in organs from wild-type mice
(Table 1). On days 21 and 82 post infection (p.i.) with
Mtb, CFU in lungs, of TLR2/4/9–/– mice were indistinguishable from those found in wild-type mice
(Fig. 2A; Table 1). CFU in liver and spleen from
triple-deficient mice were also comparable to bacterial
loads in organs from infected wild-type mice (Table 1),
indicating that TLR2/4/9–/– mice were fully capable of
chronically restricting mycobacterial growth at a plateau
Table 1. Bacterial loads in Mtb-infected TLR–/– and MyD88–/– micea)
Lungb)
Mice
C57BL/6
21 days p.i.
82 days p.i.
21 days p.i.
82 days p.i.
21 days p.i.
82 days p.i.
6.10.2
6.70.2
4.00.5
5.90.2
3.70.5
4.50.3
5.80.2
7.00.2
3.60.5
6.00.2
3.70.3
4.70.3
–/–
6.00.2
6.60.1
2.70.6
5.70.2
2.60.3
4.10.2
–/–
6.10.2
6.70.4
3.90.7
6.00.3
3.40.5
4.50.8
6.40.2
6.90.3
3.30.7
5.90.4
3.40.5
4.40.5
6.10.3
6.80.2
3.60.4
6.00.3
3.30.1
4.00.7
21 days p.i
35 days p.i
21 days p.i
35 days p.i
21 days p.i
35 days p.i
6.60.3
6.50.2
5.00.4
6.00.2
4.70.5
4.40.5
10.10.2***
5.10.9
7.10.1*
4.60.9
6.00.3***
TLR4
TLR9
–/–
TLR2/4
–/–
TLR2/4/9
C57BL/6
–/–
MyD88
b)
Liverb)
–/–
TLR2
a)
Spleenb)
7.80.4*
–/–
–/–
C57BL/6, TLR and MyD88 mice were infected with 100 CFU Mtb H37Rv via the aerosol route. For mycobacterial colony
enumeration assays, organs from infected mice were removed at the indicated time points and CFU were counted 21 days after
plating of serial diluted organ homogenates
The data are reported as log10 CFU per organ. Means and SD of four mice per group are shown. Statistical analysis was performed
by ANOVA defining differences between wild-type and mutant mice as significant (*p0.05; ***p0.001).
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
Immunity to infection
Figure 2. In contrast to MyD88–/– mice, TLR2/4/9–/– animals efficiently control Mtb infection. Wild-type, TLR2/4/9–/–, and MyD88–/–
mice were infected with 100 CFU Mtb H37Rv via the aerosol route. For mycobacterial colony enumeration assays, lungs from
infected (A) TLR2/4/9–/– or (B) MyD88–/– mice and respective wild-type animals were removed at the indicated times and CFU were
counted 21 days after plating of serial diluted organ homogenates. Means and SD of four mice per group are shown. (B) During the
course of infection, survival of ten infected wild-type and MyD88–/– mice was monitored. Animals that lost more than 25% of their
original body weight were sacrificed. Statistical analysis was performed by ANOVA defining differences between wild-type and
mutant mice as significant (*p0.05; ***p0.001). Statistical analysis of the resulting survival curve was performed using the log
rank test. Differences in survival kinetics between wild-type and MyD88–/– mice were significant (p0.001).
level, similar to wild-type mice. Thus, all three TLR that
have been implicated in Mtb sensing are dispensable for
inducing protective cell-mediated immune responses to
aerosol Mtb infection.
In contrast, at 3 and 5 weeks following aerosol
infection with 100 CFU Mtb, the bacterial load in the
lungs of MyD88–/– mice was dramatically higher owing
to exponential bacterial growth (Fig. 2B; Table 1)
accompanied by enhanced dissemination into peripheral
organs such as the liver and spleen (Table 1). Compared
to infected wild-type mice, the bacterial load in the lungs
of MyD88–/– mice was increased 15-fold (21 days p.i.)
and 4600-fold (35 days p.i.). As a consequence of
uncontrolled mycobacterial growth, all infected
MyD88–/– mice died during the first 6 weeks of infection
with Mtb (Fig. 2B; Table 1), whereas infected wild-type
and TLR2/4/9 triple-deficient mice survived for at least
82 days (Fig. 2A). By day 35 p.i., the lungs of MyD88–/–
mice harbored massive cellular infiltrations that were
accompanied by extensive necrosis of granulomas,
showing abundant mycobacteria within macrophages
in the extracellular space, and erupting into bronchi
(data not shown). In conclusion, MyD88, but not
TLR2,TLR4 and TLR9, is critical for restricting mycobacterial growth.
Innate immune responsiveness is unimpaired
in TLR2/4/9–/– mice
TLR2, TLR4, and TLR9 mediate cellular recognition of
Mtb components [35]. To exclude potential crosscompensation by any one receptor in the absence of
one or two other receptors, we performed infection
experiments with Mtb in TLR2/4/9 triple-deficient mice.
Comparative analysis of wild-type and TLR2/4/9–/– mice
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
after aerosol infection with Mtb revealed that in the latter
mice the expression of TNF-a and IL-12/IL-23p40 was
not reduced during the course of infection either on the
RNA (Fig. 3A) or on the protein (Fig. 3B) level. Because
MyD88 mediates signals from multiple TLR, we expected
that the induction of innate pro-inflammatory and T cell
responses might be comparable in TLR2/4/9–/– and
MyD88–/– mice. For analysis, expression of pro-inflammatory cytokines in the lung was determined in wildtype and MyD88–/– mice after aerosol infection with Mtb
(Fig. 3C and D). As determined by quantitative RT-PCR,
TNF-a and IL-12/IL-23p40 mRNA accumulation was
reduced in MyD88–/– mice during the first 4 weeks of
infection (Fig. 3C). However, MyD88–/– mice were
capable of inducing TNF-a and IL-12/IL-23p40 gene
expression, albeit transcript levels were reduced compared to those measured in wild-type mice. The amount
of IL-12/IL-23p40 protein was also found to be reduced
in lung homogenates during the course of infection as
revealed by cytometric bead array (CBA), although
differences were not statistically significant (Fig. 3D).
Taken together, our data indicate that TLR2/4/9 signals
are dispensable for the induction of an innate proinflammatory immune response in murine tuberculosis.
Adaptive immune responses are normal in
Mtb-infected TLR2/4/9–/– animals
In the lung, IL-12-driven CD4+ T cells are the most
prominent lymphocyte population responsible for
protective effector responses to Mtb infection [5].
However, antigen-specific CD8+ T cells are required to
maintain adaptive immune responses [36]. At the
infection site, both T cell types produce IFN-c as a
master mediator of host defense against Mtb.
www.eji-journal.eu
683
684
Christoph Hölscher et al.
Eur. J. Immunol. 2008. 38: 680–694
Figure 3. Instructive immune responses are normal in Mtb-infected TLR2/4/9–/– mice. Wild-type, TLR2/4/9–/– and MyD88–/– mice
were infected with 100 CFU Mtb H37Rv via the aerosol route. At the indicated time points, expression of TNF-a and IL-12/IL-23p40
was determined in lung homogenates from (A) TLR2/4/9–/– or (C) MyD88–/– mice and respective wild-type animals by RT-PCR based
on gene expression in uninfected mice. Results represent means and SD of three mice. The content of IL-12/IL-23p40 protein was
determined in lung homogenates from uninfected and infected (B) TLR2/4/9–/– or (D) MyD88–/– mice using a CBA. Results represent
means and SD of four mice. Statistical analysis was performed by ANOVA defining differences between wild-type and mutant
mice as significant (*p0.05; **p0.01).
To determine the frequency of these IFN-c-producing
T cells in the absence of TLR-dependent instructive
mechanisms, lymphocytes isolated from lungs of
infected animals were stimulated with plate-bound
anti-CD3/CD28 21 days after aerosol infection with
Mtb. As shown by intracellular cytokine staining,
suspensions of unstimulated cells already contained
substantial amounts of IFN-c+ CD4+ and CD8+ T cells
(Fig. 4A and B). After stimulation with anti-CD3/CD28,
an increased proportion of T cells were found to be
IFN-c+ (Fig. 4A and B). Compared to stimulated wildtype lung cells, the frequencies of IFN-c+ CD4+ and
CD8+ T cells did not differ from those of TLR2/4/9–/–
lung lymphocytes (Fig. 4A). In contrast, unstimulated
CD4+ and CD8+ T cells from Mtb-infected MyD88–/–
mice already produced significantly lower amounts of
IFN-c (Fig. 4B). After stimulation with anti-CD3/CD28,
the amount of IFN-c+ CD4+ T cells was still reduced in
MyD88–/– mice. However, MyD88–/– mice were capable
of inducing a substantial proportion of IFN-c+ CD4+
T cells, and in response to anti-CD3/CD28 the amount of
IFN-c+ CD8+ T cells did not differ from their frequency
in wild-type mice (Fig. 4B).
To analyze the number of antigen-specific T cells,
enriched CD4+ and CD8+ Tcell suspensions with a purity
of >90% were prepared from lungs of Mtb-infected mice
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
by magnetic cell sorting. The frequency of IFN-c+ antigenspecific CD4+ and CD8+ T cells was determined by an
IFN-c ELISPOT assay after restimulation with ESAT61–20
or Mtb3293–102, respectively (Fig. 4C and D). The number
of IFN-c-producing antigen-specific CD4+ and CD8+
T cells was not affected by TLR2, -4 and -9 deficiency
(Fig. 4C). In contrast, the frequency of IFN-c+ CD4+ and
CD8+ T cells was reduced in the lungs of infected
MyD88–/– mice (Fig. 4D). However, the difference of the
numbers of IFN-c+ cells in infected MyD88–/– and wildtype mice was not significant. Together, as a result of the
efficient innate pro-inflammatory cytokine response in
TLR2/4/9–/– mice, we concluded that adaptive immunity
in murine tuberculosis is induced independently of TLR2/
4/9-mediated signaling. MyD88 contributes tothe overall
frequency of antigen-specific T cells in a TLR2/4/9independent fashion.
IFN-c production and expression of
IFN-c-dependent NOS2 and LRG-47 are normal
in Mtb-infected TLR2/4/9–/– mice
MyD88–/– mice have previously been shown to be highly
susceptible to mycobacterial infection [23, 27, 33]. If,
however, TLR and to some extent MyD88 are dispensable for innate pro-inflammatory and adaptive T cell
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
responses to Mtb infection, what effector functions
might MyD88 be involved in that are pivotal to
protection from experimental TB? To address this
question, we analyzed the relevance of TLR and
Immunity to infection
MyD88 in IFN-c-mediated macrophage effector functions after infection with Mtb in vivo.
IFN-c induces efficient macrophage activation and
subsequent containment of intracellular mycobacterial
Figure 4. Adaptive immune responses are normal in Mtb-infected TLR2/4/9–/– mice. Wild-type, TLR2/4/9–/– and MyD88–/– mice were
infected with 100 CFU Mtb H37Rv via the aerosol route. After 3 weeks of infection, single-cell suspensions from lungs of (A, C)
TLR2/4/9–/– or (B, D) MyD88–/– mice and respective wild-type animals were prepared. (A, B) Intracellular IFN-c production by antiCD3/CD28-stimulated T cells was analyzed by flow cytometry of gated CD44+CD4+ and CD44+CD8+ cells. Representative
histograms of three mice per group are shown. Numbers in histograms indicate means and SD of three mice. (C, D) After
enrichment of CD4+ or CD8+ T cells, the frequency of responding CD4+ or CD8+ T cells was determined after restimulation with
ESAT61–20 or MTB3293–102 in an IFN-c ELISPOT assay. Results represent means and SD of five mice. Statistical analysis was
performed by ANOVA defining differences between wild-type and mutant mice as significant (*p0.05; **p0.01).
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eji-journal.eu
685
Eur. J. Immunol. 2008. 38: 680–694
Christoph Hölscher et al.
·
686
growth. Therefore, we quantified IFN-c expression in
lung homogenates of Mtb-infected mice by CBA (Fig. 5A
and D). Owing to the unimpaired development of
antigen-specific CD4+ T cell-mediated immune responses following infection with Mtb, expression of IFN-c
in lung homogenates from triple TLR-deficient mice was
comparable to wild-type counterparts (Fig. 5A). IFN-c
production was found to be reduced in MyD88–/– mice
during the first 2 weeks of aerosol Mtb infection when
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 5. TLR2/4/9–/– mice efficiently express IFN-c, LRG-47 and
NOS2. Wild-type, TLR2/4/9–/– and MyD88–/– mice were infected
with 100 CFU Mtb H37Rv via the aerosol route. At the indicted
time points, expression of IFN-c and the IFN-c-dependent NOS2
and LRG-47 was determined in lung homogenates from (A, B)
TLR2/4/9–/– or (D, E) MyD88–/– mice and respective wild-type
animals by CBA or RT-PCR based on gene expression in
uninfected mice, respectively. Statistical analysis was performed by ANOVA defining differences between wild-type and
mutant mice as significant (*p0.05; ***p0.001). (C, F) For
immunohistological detection of NOS2, 2–3-lm sections were
prepared from (C) TLR2/4/9–/– or (F) MyD88–/– mice and
respective wild-type animals at the indicated time points.
Staining was performed with a polyclonal rabbit anti-mouse
NOS2 antibody. Representative results of four mice per group
are shown.
compared to protein levels of IFN-c found in homogenates from wild-type mice (Fig. 5D). Thereafter, IFN-c
production was rather increased in MyD88–/– mice. In
line with the efficient induction of inflammatory innate
and subsequent adaptive immune responses in Mtbinfected TLR2/4/9–/– mice, the mRNA expression of
IFN-c-dependent effector molecules such as LRG-47 and
NOS2 were also not impaired (Fig. 5B) and these mice
exhibited substantial NOS2 protein expression in
pulmonary granulomatous lesions after 82 days of
infection (Fig. 5C). In contrast to triple-deficient mice,
IFN-c-dependent macrophage effector functions were
affected in Mtb-infected MyD88–/– mice to some extent.
After aerosol infection with Mtb, RT-PCR of LRG-47 and
NOS2 expression revealed that LRG-47 was induced in
both wild-type and MyD88–/– mice to a substantial
degree (Fig. 5E). In contrast, mRNA expression of NOS2
was significantly reduced when compared to wild-type
mice (Fig. 5E). However, transcripts of NOS2 were found
to be induced in MyD88–/– mice in a delayed fashion
and, as shown by immunohistology, no difference in the
expression of NOS2 on the protein level was apparent in
lung sections of either mouse strain 21 days after aerosol
infection with Mtb (Fig. 5F). At day 35 p.i., however,
severe necrosis in lungs from MyD88–/– mice precluded
any meaningful comparison of NOS2 expression in wildtype and mutant mice (Fig. 5F).
In addition to the TLR-independent induction of
adaptive immune responses after Mtb infection, our data
demonstrate that the expression of two critical
IFN-c-dependent macrophage effector molecules, LRG47 and NOS2, are not affected by the absence of TLR2, -4
and -9.
Restriction of Mtb growth is impaired in MyD88–/–
but not in TLR2/4/9–/– macrophages
It has recently been shown that MyD88 is involved in
macrophage effector mechanisms induced by IFN-c
[37]. This suggested to us that the exacerbated
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
susceptibility of Mtb-infected MyD88–/– mice might be
due to a defect at the level of IFN-c-mediated antibacterial macrophage effector functions.
IFN-c-mediated anti-mycobacterial effector mechanisms are dependent on efficient maturation of the
phagosomal compartment within macrophages [8, 9].
To analyze whether TLR- and MyD88-mediated signals
influence Mtb phagosome maturation, the intracellular
trafficking of Mtb was studied in resting, IFN-c- or IFN-c/
TNF-a-activated BMMU from wild-type, TLR2/4/9–/–
and MyD88–/– mice. Cells were infected with Mtb-GFP
[38] (MOI 5:1) for either 4 h or 2 days. Macrophages
were labeled with tracers/antibodies for either early or
late endosomal/lysosomal compartments, i.e., transferrin (Tf-Cy3), dextran (DexTR), lysosome associated
membrane protein 1 (LAMP-1) or the vesicular proton
pump (H+ATPase), respectively, and analyzed by
confocal microscopy. At 4 h p.i., around 50% of the
phagosomes were associated with Tf but less frequently
with the late markers in resting cells (Fig. 6, Supporting
Information Fig. 1). At the same time point, IFN-c
promoted Mtb phagosome maturation towards LAMP-1and H+ATPase-positive late endosomal compartments in
cells from all three mouse strains (Supporting Information Fig. 1). Loss of Tf association was prominent in wildtype and MyD88–/– macrophages (Fig. 6; see graphs for
quantification), whereas in IFN-c-activated TLR2/4/9–/–
macrophages more phagosomes became positive for
both Tf and LAMP-1, also indicating transition towards a
later stage (Fig. 6, Supporting Information Fig. 1).
Importantly, association with the lysosomal tracer for
the lysosomal stage, DexTR, was reduced at 4 h p.i. in
IFN-c-activated MyD88–/– macrophages when compared
to wild-type and TLR2/4/9–/– cells (Fig. 6). In contrast,
activation by both IFN-c and TNF-a promoted phagosome maturation towards the late endosomal/lysosomal
stage in cells from all three mouse strains (Fig. 6,
Supporting Information Fig. 1). Although in fully
activated MyD88–/– cells some phagosomes were
positive for both LAMP-1 and Tf, the overall number
of phagosomes was within late compartments (Fig. 6;
Supporting Information Fig. 1). At day 2 p.i.,
IFN-c-mediated phagosome maturation was less efficient in MyD88–/– but also to some extent in TLR2/4/9–/–
cells (Supporting Information Fig. 2 and 3). Our data
indicate that, early in infection, IFN-c-dependent
phagosome maturation in Mtb-infected macrophages
was affected by the absence of MyD88. However, at later
stages mycobacterial trafficking was impaired in both
TLR2/4/9–/– and MyD88–/– macrophages. In addition to
efficient maturation of the phagosomal compartment
within macrophages, the IFN-c-dependent production of
RNI through NOS2 is central to anti-mycobacterial
effector responses in IFN-c-activated macrophages [11].
After infection, we found little induction of nitrite in the
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunity to infection
supernatants of infected BMMU from wild-type,
MyD88–/– and TLR2/4/9–/– mice (Fig. 7). In contrast,
simultaneous IFN-c activation and Mtb infection enhanced nitrite production in MyD88–/– cells, albeit to a
significantly lower extent when compared with wildtype cells. The addition of IFN-c to TLR2/4/9–/– BMMU
also led to an increased nitrite content that was still
significantly reduced. However, addition of TNF-a
restored the ability of MyD88–/– and TLR2/4/9–/–
macrophages to produce nitrite at amounts that were
comparable to those generated by wild-type macrophages (Fig. 7).
Finally, we determined the growth of Mtb H37Rv in
BMMU derived from wild-type, TLR2/4/9–/– and
MyD88–/– mice that were incubated with medium or
IFN-c 24 h prior to infection (Fig. 8). IFN-c-primed
macrophages from wild-type and TLR2/4/9–/– mice
were both able to reduce mycobacterial growth to a
similar extent after 72 h of infection with H37Rv
(Fig. 8). In contrast, IFN-c-incubated MyD88–/– macrophages were unable to reduce the growth of Mtb (Fig. 8),
showing a substantially increased bacterial load as
compared to that of activated macrophages from wildtype mice 72 h p.i. (Fig. 8). To evaluate whether IFN-cor TNF-a/NFjB-dependent pathways are affected in
MyD88–/– macrophages, infected cells were additionally
stimulated with TNF-a (Fig. 8). In contrast to unstimulated macrophages, cells from wild-type, TLR2/4/9–/–
and MyD88–/– mice were able to restrict bacterial
growth when stimulated with TNF-a.
Discussion
Based on in vitro studies TLR2, TLR4, and TLR9 have all
been suggested to be involved in the innate induction of
a Th1 cell-mediated immune responses after infection
with Mtb [13, 15, 16, 18, 22, 39, 40]. We now
demonstrate that TLR2/4/9–/– mice are capable of
efficiently controlling Mtb infection.
The different and highly variable outcome of
infection in Mtb-infected TLR2-, TLR4-, TLR9-singledeficient and TLR2/4- or TLR2/9-double-deficient mice
in this and a number of other studies [20–25, 41, 42]
may be due to different housing conditions of animals
reflecting different levels of preactivation, and/or to
infection with different Mtb strains. For example,
TLR9–/– mice that were infected with a high-dose
inoculum (500 CFU) of H37Rv were highly susceptible
to Mtb infection and died between days 45 and 75 upon
infection [19]. However, using our own H37Rv lab
strain, an aerosol infection even with 2000 CFU did not
lead to premature death in TLR9–/– mice (data not
shown).
www.eji-journal.eu
687
688
Christoph Hölscher et al.
Eur. J. Immunol. 2008. 38: 680–694
Figure 6. Endosome and lysosome maturation in TLR2/4/9–/– and MyD88–/– macrophages early after Mtb infection. Cells from wildtype, TLR2/4/9–/– and MyD88–/– mice were either left resting or activated overnight with 1000 U/mL IFN-c, or 1000 U/mL IFN-c and
20 ng/mL TNF-a. Macrophages were infected with Mtb-GFP at an MOI of 5:1 for 2 h, washed and further incubated for 2 h. Early
compartments in live cells were labeled using Tf-Cy3 (early endosomes). Lysosomes were labeled in live cells using DexTR
and investigated by confocal fluorescence microscopy (A). Macrophage activation decreases the amount of Tf uptake
independently of the mouse strain analyzed. Pictures represent representative examples. Phagosomes are indicated by arrows,
yellow = colocalization, white = no colocalization (magnification: 1000). Inserts represent Mtb-GFP (magnification: 5000).
(B) Quantification of Mtb-GFP phagosome association with intracellular markers. By confocal LSM, Mtb-GFP phagosomes from 10
to 30 cells from independent cover slips were evaluated double-blindly for colocalization with either Tf-Cy3 or DexTR and
expressed as % means SD.
The use of TLR triple-deficient mice supported our
previous contention that TLR are dispensable for innate
and Th1 responses and subsequent macrophage effector
functions against Mtb. Hence, TLR2/4/9-independent
but MyD88-mediated mechanisms are necessary for the
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
induction and effective expression of protective immune
responses. Despite only moderately reduced numbers of
IFN-c-producing T cells, survival kinetics in Mtb-infected
MyD88–/– mice are similar to infected IFN-c–/– mice [43,
44]. This indicates that anti-mycobacterial effector
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
Figure 7. RNI production is impaired in IFN-c-stimulated and
Mtb-infected macrophages from MyD88–/– and TLR2/4/9–/–
mice. BMMU from wild-type, TLR2/4/9–/– and MyD88–/– mice
were either incubated with medium, Mtb H37Rv, Mtb H37Rv
and 1000 U/mL IFN-c, or Mtb H37Rv, 1000 U/mL IFN-c and
20 ng/mL TNF-a. Supernatants were harvested after 48 h and
the nitrite contend was measured. Each point indicates mean
and SD of triplicate values from one out of two representative
experiments. Statistical analysis was performed by ANOVA
defining differences between wild-type and MyD88–/– (**p0.01)
or TLR2/4/9–/– (+++p0.001) mice as significant.
mechanisms – most likely in macrophages – rather than
instructive innate immune responses are induced or
recruited independently of TLR2/4/9 but are severely
affected by the absence of MyD88.
A pivotal role of MyD88 for the expression of effector
immune responses in IFN-c-activated macrophages had
originally been highlighted by microarray studies
comparing gene expression in wild-type, TLR- and
Immunity to infection
MyD88-deficient macrophages. While TLR-induced
gene expression in macrophages in response to Mtb
was largely independent of MyD88 [37], responses to
IFN-c were severely compromised in MyD88-deficient
cells. This was recently attributed to a direct involvement of MyD88 in the IFN-c signal transduction pathway
[45, 46]. Inefficient expression of TNF-a and IL-12/
IL-23p40 in Mtb-infected MyD88–/– mice is probably
aggravated by a defective IFN-c-signaling pathway. In
addition, genes central to macrophage effector responses against Mtb may also be stabilized through
IFN-cR1/MyD88-mediated p38 activation.
Through our analysis of BMMU in vitro, we show
here that the ability of Mtb-infected macrophages to
induce NOS2 and to produce RNI in response to IFN-c
was largely dependent on MyD88, in line with a previous
report [37]. In addition, our in vitro experiments
revealed that the defective RNI production and mycobacterial growth restriction in Mtb-infected and
IFN-c-stimulated MyD88–/– macrophages could be
restored by TNF-a in vitro, similar to findings reported
by Shi et al. [37]. Therefore, IFN-c- and not TNF-a/
NFjB-mediated effector mechanisms appeared to be
specifically impaired in the absence of MyD88. However,
both signaling pathways are functional in TLR2/4/9–/–
mice in vivo possibly explaining the differential outcome
of Mtb infection in both mutant mice. The lack of MyD88
expression may have a strictly local effect by being
restricted to macrophages at the infection site, which
eludes systemic detection. Therefore, a defect in
generation of RNI may still account for ablation of
macrophage mycobacteriostatic functions in MyD88–/–
mice in vivo.
Figure 8. Restriction of Mtb growth is not impaired in TLR2/4/9–/– macrophages. To analyze the bacterial growth in IFN-c-stimulated
macrophages, BMMu from wild-type, TLR2/4/9–/– and MyD88–/– mice were incubated with medium, IFN-c, or IFN-c/TNF-a for 24 h
before infection with Mtb H37rv at an MOI of 1:1. Cells were lysed at different time points, serial dilutions were plated onto agar
plates and CFU were determined after 19–21 days. Data represent means and SD of triplicate cultures. Statistical analysis was
performed by ANOVA defining differences between medium and IFN-c-stimulated cells (*p0.05; **p0.01) or medium and IFN-c/
TNF-a-incubated macrophages (#p0.05; ##p0.01; ###p0.001) as significant.
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eji-journal.eu
689
690
Eur. J. Immunol. 2008. 38: 680–694
Christoph Hölscher et al.
Upon IFN-c activation of macrophages, mycobacteria-containing phagosomes mature from an early to a late
endosomal stage concomitant with reduced mycobacterial growth and viability [8]. Although both TLR and
MyD88 signaling were initially suggested to influence
phagosome maturation [47], a recent report describes
phagosome maturation to be entirely independent of
TLR2 and 4 but dependent on MyD88 [48]. The
differential influence of MyD88 and TLR2/4/9 on
maturation of Mtb phagosomes, however, has not yet
been directly investigated. Our results suggest that
IFN-c-promoted Mtb phagosome maturation requires
MyD88 in the initial infection phase. Although fewer
mycobacteria trafficked to phagolysosomes in
IFN-c-activated MyD88–/– cells as compared to TLR2/
4/9–/– and wild-type macrophages (most obvious early
after infection), deficient phagosome maturation on its
own is unlikely to account for the unrestricted growth in
these cells. However, the impaired killing of Mtb in
activated MyD88–/– macrophages may find its explanation in combination with reduced RNI production. Taken
together, our results show the striking susceptibility of
MyD88–/– mice to Mtb to be caused by a combination of
defects such as reduced TNF-a production and impaired
IFN-c-mediated RNI generation and phagosome maturation. Importantly, all of these effector mechanisms
were still in place in TLR2/4/9–/– mice.
In addition to TLR signal transduction, MyD88 serves
an equally critical role in IL-1 and IL-18 receptor/IRAK
signal-transduction. Therefore, contribution of the latter
signaling pathway to defective immune responses
observed in MyD88–/– mice cannot be excluded.
Whereas a deficiency in IL-18 is associated only with
minor defects in anti-TB immunity in mice [49],
infection of IL-1R–/– mice with Mtb H37Rv has very
recently been shown to result in exacerbated bacterial
loads and rapid death [50]. Importantly, the only
moderately reduced innate and adaptive immune
responses, uncontrolled mycobacterial growth, histopathology and survival kinetics in Mtb H37Rv-infected
IL-1R1–/– mice were highly similar to the kinetics
observed in Mtb-infected MyD88–/– mice [50], implying
that a common defect downstream of adaptive immune
responses – possibly in macrophages – must account for
the detrimental outcome of infection in both strains of
mice. It appears plausible that TLR-independent, but
IL-1R1-mediated MyD88-dependent signaling compensates for the absence of all three TLR and promotes the
induction of an inflammatory immune response during
Mtb infection. In fact, similar to TLR, activation of the socalled inflammasome, which is involved in processing
IL-1b, is thought to occur through the recognition of
pathogen associated molecular patterns (PAMP) by the
leucine-rich repeat of the NACHT-, LLR-, and pyrindomain (NALP)-containing proteins [51]. In line with
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
the cytoplasmic localization of the inflammasome, van
der Wel and colleagues [52] have recently shown that
Mtb is capable of translocating from the phagolysosome
to the cytosol where it may encounter cytosolic PRR
different from TLR. This putative TLR-independent
recognition mechanism in Mtb infection may bypass the
TLR-mediated signaling pathway, but still rely on IL-1R
signaling via MyD88.
In conclusion, our results demonstrate that after Mtb
infection, TLR-mediated pattern recognition and
MyD88-dependent adaptor signaling have little impact
on the induction and expression of Th1 immune
responses. In contrast, MyD88 shapes anti-bacterial
effector mechanisms in macrophages in response to
IFN-c independently of its function as a TLR signal
transducer. Our results not only underscore the dual
function of MyD88 as a TLR-coupled signaling adaptor
on the one hand and a master regulator of macrophage
activation on the other, they also assign critical
relevance for in vivo protection against TB only to the
latter.
Materials and methods
Mice and macrophages
MyD88–/– (Research Center Borstel, Germany) [29], TLR2–/–,
TLR4–/–, TLR9–/–, TLR2/4–/–, and TLR2/4/9–/– (from the
Technical University of Munich, Germany) mice on a C57BL/6
genetic background were maintained under specific-pathogenfree conditions. As wild-type controls, C57BL/6 mice were
purchased from Charles River (Sulzfeld, Germany). All
experimental mice were between 8 and 16 weeks old. In
any given experiment, mice were matched for age, sex and
genetic background. For infection experiments, mice were
maintained under barrier conditions in the BSL 3 facility at the
Research Center Borstel in individually ventilated cages. All
experiments performed were in accordance with the German
Animal Protection Law and were approved by the Animal
Research Ethics Board of the Ministry of Environment, Kiel,
Germany.
BMMU were obtained after flushing of femora from the
mice described above. To generate BMMU, bone marrow cells
were cultivated in L-929 conditioned medium as source for MCSF activity for 7 days [8].
Bacteria and infection
For all experiments, Mtb H37rv were used. For confocal
microscopy of mycobacterial trafficking within BMMU, GFPexpressing Mtb were applied (kindly provided by Tanya Parish,
Centre for Infectious Diseases, Institute for cell and Molecular
Science, London, UK) [38]. Mtb was grown in Middlebrook
7H9 broth (Difco, Detroit, MI) supplemented with Middlebrook OADC enrichment medium (Life Technologies, Gaithersburg, MI), 0.002% glycerol, and 0.05% Tween 80. Midlog
phase cultures were harvested, aliquoted, and frozen at –80 C.
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
After thawing, viable cell counts were determined by plating
serial dilutions of the cultures on Middlebrook 7H10 agar
plates followed by incubation at 37 C.
For in vitro experiments, Mtb aliquots were thawed, spun at
835 g and subsequently resuspended in PBS. BMMU were
infected with different multiplicities of infection (MOI) of Mtb.
For bacterial growth inhibition assays, determination of RNI
and intracellular localization of mycobacteria, BMMU were
incubated with 200 U/mL IFN-c or 200 U/mL IFN-c and
20 ng/mL TNF-a (both from Peprotech) 24 h before infection.
Lipopeptide (LP; 10 ng/mL Pam3Csk4; EMC Microcollections,
Tbingen, Germany), bacterial LPS of Salmonella enterica,
serotype friedenau H909 (10 ng/mL; kindly provided by
H. Brade, Research Center Borstel, Germany) and CpG (1 lM
ODN1648; a kind gift of A. Dalpke, University of Heidelberg,
Germany) were used as control stimuli.
Before infection of experimental animals, stock solutions of
Mtb were diluted in sterile distilled water and pulmonary
infection was performed using an inhalation exposure system
(Glas-Col, Terre-Haute, IN). To infect mice with a dose of
100 CFU/lung, animals were exposed for 40 min to an aerosol
generated by nebulizing approximately 5.5 mL of a suspension
containing 107 live bacteria. Inoculum size was checked 24 h
p.i. by determining the bacterial load in the lung of infected
mice. Mice were regularly weighed before and after infection.
In accordance with the Animal Research Ethics Board of the
Ministry of Environment, mice that lost 25% of their original
weight during the course of infection had to be killed.
Colony enumeration assay and immunohistology
Bacterial loads in lungs, spleen and liver were evaluated at
different time points p.i. for Mtb to follow the course of
infection. Organs from sacrificed animals were removed
aseptically, weighed and homogenized in PBS containing a
proteinase inhibitor cocktail (Roche Diagnostics, Mannheim,
Germany) prepared according to the manufacturer's instructions. Tenfold serial dilutions of organ homogenates were
plated in duplicates onto Middlebrook 7H10 agar plates
containing 10% OADC and incubated at 37 C for 19–21 days.
Colonies on plates were enumerated and results expressed as
log10 CFU per organ. For immunohistology, one lung lobe per
mouse were fixed in 4% formalin-PBS, set in paraffin blocks,
and sectioned (2–3 lm). Tissue sections were prepared and
stained with a polyclonal rabbit anti-mouse NOS2 antiserum
(Biomol, Hamburg, Germany) as previously described [20].
Preparation of single-cell suspensions from
infected lungs
For antigen-specific restimulation and flow cytometric analysis, single-cell suspensions of lungs were prepared from Mtbinfected mice at different time points. Mice were anesthetized
and injected intraperitoneally with 150 U heparin (Ratiopharm, Ulm, Germany). Lungs were perfused through the right
ventricle with warm PBS. Once lungs appeared white, they
were removed and sectioned. Dissected lung tissue was then
incubated in collagenase A (0.7 mg/mL; Roche Diagnostics,
Mannheim, Germany) and DNase (30 lg/mL; Sigma) at 37 C
for 2 h. Digested lung tissue was gently disrupted by
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunity to infection
subsequent passage through a 100 lm pore size nylon cell
strainer. Recovered vital lung cells were counted using an
automatic cell counter (ViCellJ; Beckman Coulter, Krefeld,
Germany), diluted in complete Iscove's-modified Dulbecco's
medium (IMDM; Gibco) supplemented with 10% FCS (Gibco),
0.05 mM b-mercaptoethanol (Sigma, Deisenhofen, Germany),
and penicillin and streptomycin (100 U/mL and 100 lg/mL;
Gibco) and used for further experiments.
RT-PCR
Before and at different time points after aerosol infection with
Mtb, weighed lung samples were homogenized in 5 mL 4 M
guanidinium isothiocyanate buffer and total RNA was
extracted by acid phenol extraction. cDNA was obtained using
murine Moloney leukemia virus (MMLV) reverse transcriptase
(Superscript II, Invitrogen, Karlsruhe, Germany) and oligo-dT
(12–18mer; Sigma) as a primer. Quantitative PCR was
performed on a Light Cycler (Roche Diagnostics Corporation,
Indianapolis, IN) as previously described [53]. Data were
analyzed employing the “Fit Points” and “Standard Curve
Method” using b-2-microglubulin (b2m) or hypoxanthineguanine phosphoribosyl transferase (hprt) as housekeeping
genes to calculate the level of gene expression in relation to
b2m or hprt, respectively The following primer sets were
employed: b2m: sense 50 -TCA CCG GCT TGT ATG CTA TC-30 ,
antisense 50 -CAG TGT GAG CCA GGA TAT AG-30 ; hprt: sense 50 GCA GTA CAG CCC CAA AAT GG-30 , antisense 50 -AAC AAA GTC
TGG CCT GTA TCC AA-30 ; TNF-a: sense 50 -TCT CAT CAG TTC
TAT GGC CC-30 , antisense 50 -GGG AGT AGA CAA GGT ACA AC30 , IL-12/23p40: sense 50 -CTG GCC AGT ACA CCT GCC AC-30 ,
antisense 50 -GTG CTT CCA ACG CCA GTT CA-30 , NOS2: sense
50 -AGC TCC TCC CAG GAC CAC AC-30 , antisense 50 -ACG CTG
AGT ACC TCA TTG GC-30 , LRG-47: sense 50 -AGC CGC GAA GAC
AAT AAC TG-30 , antisense 50 -CAT TTC CGA TAA GGC TTG G-30 .
Quantification of TNF-a, IL-12/23p40, and IFN-c
To determine cytokine production after infection of BMMU,
supernatants were analyzed in threefold serial dilutions using
the sandwich ELISA system OptEiaTM (BD Biosciences). After
incubation with horseradish peroxidase (HRP) coupled to
avidin and developing with TMB substrate reagent, the
absorbance was read on a microplate reader (Sunrise; Tecan,
Mnnedorf, Switzerland). Samples were compared to appropriate recombinant cytokine standards using a test wavelength
of 450 nm and a reference wavelength of 630 nm. The
detection limits for all cytokines were 5 pg/mL.
The concentrations of IL-12/IL-23p40 and IFN-c in lung
homogenates from uninfected and infected mice were
determined by CBA (BD Bioscience). IFN-c was analyzed
using a ready CBA mouse-flex-set (BD Bioscience). To
determine IL-12/IL-23p40, beads were conjugated with a
purified mouse anti-IL-12/IL-23p40 antibody (BD Bioscience)
using a functional bead-conjugation buffer set following the
manufacturer's instructions (BD Bioscience). In a 96-well
plate, 50 lL lung homogenates containing a proteinase
inhibitor cocktail (Roche Diagnostics) and serial dilutions of
cytokine standards were incubated with 50 lL of respective
beads diluted in capture-bead dilution buffer. After 1 h at room
www.eji-journal.eu
691
692
Eur. J. Immunol. 2008. 38: 680–694
Christoph Hölscher et al.
temperature in the dark, beads were washed and 50 lL of a
biotinylated mouse anti-IL-12/IL-23p40 antibody was added
and incubated for another 1 h. A PE-detection reagent mix was
added and plates were incubated at room temperature for 1 h
in the dark. Once plates were washed in wash buffer, 500
events/cytokine were acquired on a FACSCalibur (BD
Biosciences) gating on the total bead population identified
by FSC-SSC-profile. Data and cytokine concentrations were
analyzed using the FCAPArray software (BD Bioscience).
ated using an ELISPOT reader (EliSpot 04 XL; AID, Strassberg,
Germany). The frequency of responding CD4+ and CD8+
T cells was determined. Neither CD4+ or CD8+ T cells cultured
in the absence of ESAT61–20 or Mtb3293–102 nor cells from
uninfected mice produced detectable spots.
Intracellular IFN-c staining
To analyze the anti-bacterial activity of macrophages, BMMU
were cultured 24 h before infection with Mtb in DMEM
supplemented with 10% FCS, 2 mM L-glutamine, 1 mM
sodium pyruvate, 10 mM HEPES (all from PAA, Pasching,
Austria) in the presence of medium, or stimulated with either
of 200 U/mL IFN-c or 200 U/mL IFN-c and 20 ng/mL TNF-a
(both from Peprotech). Cells were inoculated with Mtb H37Rv
at an MOI of 1:1 for 4 h. BMMU were washed vigorously with
warm HBSS (PAA) to remove extracellular bacteria. To
determine bacterial uptake at different time points, macrophages were lysed by addition of 0.1% saponin (Sigma).
Lysates were serially diluted in 0.05 % Tween 80 (EMD
Biosciences, Darmstadt, Germany) and plated on Middlebook
7H10 agar containing 10% OADC. To determine the production
of RNI, supernatants were harvested to quantify the release of
nitrite.
To quantify the production of RNI, the content of nitrite in
supernatants was determined after adding Griess reagents by
photometric measurement reading the absorbance at 540 nm
as previously described [54].
Intracellular localization of Mtb within IFN-c-activated
macrophages was analyzed in BMMU after infection with MtbGPF at an MOI of 5:1 for 2 h [8]. After 4 h or 2 days, cells were
labeled with tracers/antibodies for either early or late
endosomal/lysosomal compartments, i.e., Tf-Cy3, DexTR
(Invitrogen, Paisley, UK), LAMP-1 (IB4D, Developmental
Studies Hybridoma Bank, University of Iowa, IA) or the
vesicular proton pump (H+ATPase, a kind gift from Mhairi
Skinner, University of Guelph, Ontario, Canada), respectively,
and analyzed by confocal microscopy using a confocal
fluorescence microscope (LSM 510, Carl Zeiss Inc., Jena,
Germany). Mtb-GFP phagosomes from 30 cells from independent cover slips were evaluated double-blindly for colocalization with either tracer.
For detection of intracellular IFN-c, an intracellular cytokine
staining kit (BD Biosciences) was employed. Briefly, single-cell
suspensions were prepared at 21 days p.i. and 2 106 cells
were incubated with IMDM or stimulated with plate-bound
anti-CD3/CD28 mAb (clone 2C11 and clone 37/51 at 10 lg/
mL, respectively ) for 4 h in the presence of GolgiPlugTM (BD
Biosciences). Nonspecific antibody binding was blocked by
incubation with a cocktail containing anti-FccRIII/II mAb
(clone 2.4G2), mouse and rat serum. Cells were washed and
incubated with optimal concentrations of anti-CD44-FITC and
anti-CD4-APC, or anti-CD8-APC (all from BD Biosciences).
After staining, cells were fixed and permeabilized with
Cytofix/CytopermTM (BD Biosciences) and intracellularly
accumulated IFN-c was stained with a PE-labeled anti-IFN-c
mAb (BD Biosciences). Fluorescence intensity was analyzed on
a FACSCalibur (BD Biosciences) gating on CD44+CD4+ or
CD44+CD8+ lymphocytes identified by FSC-SSC profile.
ESAT61–20- and Mtb3293–102-specific IFN-c ELISPOT
assay
For measuring antigen-specific CD4+ and CD8+ T cells in lungs
from infected mice, single-cell suspensions were resuspended
in IMDM. To enrich CD4+ T cells, single cell suspensions were
incubated with magnetic CD4 microbeads (Miltenyi, Bergisch
Gladbach, Germany) and separated from other cells on an
MACS separation unit (Miltenyi). Separated CD4+ and CD8+
T cells were collected in IMDM, counted using a cell counter
(ViCellJ; Beckman Coulter), diluted in IMDM and used for
further experiments. Purity of enriched CD4+ T cells was
>90% as determined by flow cytometry. Detection of antigenspecific IFN-c-producing CD4+ and CD8+ T cells from infected
lungs was conducted using an ELISPOT assay kit (AID
Diagnostika, Strassberg, Germany). In brief, purified CD4+
or CD8+ T cells from lungs of infected mice were seeded in
wells of anti-mouse IFN-c-coated and blocked 96-well
multitest plates at an initial concentration of 1 105 cells/
well in IMDM. After making doubling dilutions of these cells,
mitomycin-D (Sigma)-inactivated splenocytes from uninfected
wild-type mice were used as APC at a concentration of 1 106
cells/well. CD4+ and CD8+ T cells were restimulated with the
Mtb ESAT61–20 or Mtb3293–102 peptides, respectively (Research
Center Borstel, Germany) at a concentration of 10 lg/mL in
the presence of 10 U/mL recombinant mouse IL-2 (Peprotech).
After 20 h of incubation in 5% CO2 at 37 C, plates were
washed, and biotinylated anti-mouse IFN-c was used to detect
the captured cytokine. Spots were visualized using streptavidin-HRP and as substrate. Spots were automatically enumerf 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
IFN-c-induced bacterial growth inhibition,
RNI production and phagosome maturation in
macrophages
Statistical analysis
Quantifiable data are expressed as the means and SD of
individual determinations. Statistical analysis was performed
by ANOVA defining different error probabilities as significant
(*,+p0.05), (**,++p0.01), (***,+++p0.001).
Acknowledgements: The present study was financially
supported in part by the German Research Foundation
(Research Grants HO 2145/2–1 to C.H., SFB 415-C7 to S.E.
and N.R., and SP “Innate Immunity” FR 448/4–2 to M.F.,
SPP1131–2 Scha 514–2 to U.S.), the National Genome
Research Network (Workpackage Tuberculosis, NIES05T22) to S.E. and Royal Society Wolfson Research
www.eji-journal.eu
Eur. J. Immunol. 2008. 38: 680–694
Merit Awards to U.S. The authors thank Stefanie Pfau,
Susanne Metken and Manfred Richter for excellent
technical assistance; Ilka Monath, Sven Mohr and Claus
Mller for organizing the animal facility and taking care
of the mice at the Research Center Borstel; Hella Stbig
and Helga Kochanowski for taking care of the mice at
the Max-Planck-Institute for Immunobiology; Tanya
Parish at the Centre for Infectious Disease, Institute for
Cell and Molecular Science, London, United Kingdom
and Mhairi Skinner, University of Guelph, Ontario,
Canada for providing Mtb-GFP and anti-H+ATPase,
respectively; and Kristine Hagens for her expert
technical assistance at the Max-Planck-Institute for
Infection Biology.
Conflict of interest: The authors declare no financial or
commercial conflict of interest.
References
1 Kaufmann, S. H., Tuberculosis: Back on the immunologists' agenda.
Immunity 2006. 24: 351–357.
2 Maher, D. and Raviglione, M., Global epidemiology of tuberculosis. Clin.
Chest Med. 2005. 26: 167–182, v.
3 Kaufmann, S. H., How can immunology contribute to the control of
tuberculosis? Nat. Rev. Immunol. 2001. 1: 20–30.
4 North, R. J. and Jung, Y. J., Immunity to tuberculosis. Annu. Rev. Immunol.
2004. 22: 599–623.
5 Flynn, J. L. and Chan, J., Immunology of tuberculosis. Annu. Rev. Immunol.
2001. 19: 93–129.
6 Hsieh, C. S., Macatonia, S. E., Tripp, C. S., Wolf, S. F., O'Garra, A. and
Murphy, K. M., Development of TH1 CD4+ T cells through IL-12 produced
by Listeria-induced macrophages. Science 1993. 260: 547–549.
7 Trinchieri, G., Interleukin-12 and its role in the generation of TH1 cells.
Immunol. Today 1993. 14: 335–338.
8 Schaible, U. E., Sturgill-Koszycki, S., Schlesinger, P. H. and Russell, D. G.,
Cytokine activation leads to acidification and increases maturation of
Mycobacterium avium-containing phagosomes in murine macrophages.
J. Immunol. 1998. 160: 1290–1296.
9 MacMicking, J. D., Taylor, G. A. and McKinney, J. D., Immune control of
tuberculosis by IFN-gamma-inducible LRG-47. Science 2003. 302: 654–659.
10 Gutierrez, M. G., Master, S. S., Singh, S. B., Taylor, G. A., Colombo, M. I.
and Deretic, V., Autophagy is a defense mechanism inhibiting BCG and
Mycobacterium tuberculosis survival in infected macrophages. Cell 2004. 119:
753–766.
11 MacMicking, J. D., North, R. J., LaCourse, R., Mudgett, J. S., Shah, S. K.
and Nathan, C. F., Identification of nitric oxide synthase as a protective
locus against tuberculosis. Proc. Natl. Acad. Sci. USA 1997. 94: 5243–5248.
12 Gilleron, M., Quesniaux, V. F. and Puzo, G., Acylation state of the
phosphatidylinositol hexamannosides from Mycobacterium bovis bacillus
Calmette Guerin and Mycobacterium tuberculosis H37Rv and its implication
in Toll-like receptor response. J. Biol. Chem. 2003. 278: 29880–29889.
13 Jones, B. W., Means, T. K., Heldwein, K. A., Keen, M. A., Hill, P. J., Belisle,
J. T. and Fenton, M. J., Different Toll-like receptor agonists induce distinct
macrophage responses. J. Leukoc. Biol. 2001. 69: 1036–1044.
14 Means, T. K., Lien, E., Yoshimura, A., Wang, S., Golenbock, D. T. and
Fenton, M. J., The CD14 ligands lipoarabinomannan and lipopolysaccharide
differ in their requirement for Toll-like receptors. J. Immunol. 1999. 163:
6748–6755.
15 Thoma-Uszynski, S., Stenger, S., Takeuchi, O., Ochoa, M. T., Engele, M.,
Sieling, P. A., Barnes, P. F. et al., Induction of direct antimicrobial activity
through mammalian toll-like receptors. Science 2001. 291: 1544–1547.
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunity to infection
16 Means, T. K., Wang, S., Lien, E., Yoshimura, A., Golenbock, D. T. and
Fenton, M. J., Human toll-like receptors mediate cellular activation by
Mycobacterium tuberculosis. J. Immunol. 1999. 163: 3920–3927.
17 Means, T. K., Jones, B. W., Schromm, A. B., Shurtleff, B. A., Smith, J. A.,
Keane, J., Golenbock, D. T. et al., Differential effects of a Toll-like receptor
antagonist on Mycobacterium tuberculosis-induced macrophage responses.
J. Immunol. 2001. 166: 4074–4082.
18 Krieg, A. M., CpG motifs in bacterial DNA and their immune effects. Annu.
Rev. Immunol. 2002. 20: 709–760.
19 Bafica, A., Scanga, C. A., Feng, C. G., Leifer, C., Cheever, A. and Sher, A.,
TLR9 regulates Th1 responses and cooperates with TLR2 in mediating
optimal resistance to Mycobacterium tuberculosis. J. Exp. Med. 2005. 202:
1715–1724.
20 Reiling, N., Holscher, C., Fehrenbach, A., Kroger, S., Kirschning, C. J.,
Goyert, S. and Ehlers, S., Cutting edge: Toll-like receptor (TLR)2- and
TLR4-mediated pathogen recognition in resistance to airborne infection with
Mycobacterium tuberculosis. J. Immunol. 2002. 169: 3480–3484.
21 Shi, S., Blumenthal, A., Hickey, C. M., Gandotra, S., Levy, D. and Ehrt, S.,
Expression of many immunologically important genes in Mycobacterium
tuberculosis-infected macrophages is independent of both TLR2 and TLR4
but dependent on IFN-alphabeta receptor and STAT1. J. Immunol. 2005.
175: 3318–3328.
22 Abel, B., Thieblemont, N., Quesniaux, V. J., Brown, N., Mpagi, J., Miyake,
K., Bihl, F. and Ryffel, B., Toll-like receptor 4 expression is required to
control chronic Mycobacterium tuberculosis infection in mice. J. Immunol.
2002. 169: 3155–3162.
23 Drennan, M. B., Nicolle, D., Quesniaux, V. J., Jacobs, M., Allie, N., Mpagi,
J., Fremond, C. et al., Toll-like receptor 2-deficient mice succumb to
Mycobacterium tuberculosis infection. Am. J. Pathol. 2004. 164: 49–57.
24 Shim, T. S., Turner, O. C. and Orme, I. M., Toll-like receptor 4 plays no role
in susceptibility of mice to Mycobacterium tuberculosis infection. Tuberculosis
(Edinb.) 2003. 83: 367–371.
25 Sugawara, I., Yamada, H., Li, C., Mizuno, S., Takeuchi, O. and Akira, S.,
Mycobacterial infection in TLR2 and TLR6 knockout mice. Microbiol.
Immunol. 2003. 47: 327–336.
26 Ryffel, B., Fremond, C., Jacobs, M., Parida, S., Botha, T., Schnyder, B.
and Quesniaux, V., Innate immunity to mycobacterial infection in mice:
Critical role for toll-like receptors. Tuberculosis (Edinb.) 2005. 85: 395–405.
27 Scanga, C. A., Bafica, A., Feng, C. G., Cheever, A. W., Hieny, S. and Sher,
A., MyD88-deficient mice display a profound loss in resistance to
Mycobacterium tuberculosis associated with partially impaired Th1 cytokine
and nitric oxide synthase 2 expression. Infect. Immun. 2004. 72: 2400–2404.
28 Medzhitov, R., Preston-Hurlburt, P., Kopp, E., Stadlen, A., Chen, C.,
Ghosh, S. and Janeway, C. A. Jr., MyD88 is an adaptor protein in the hToll/
IL-1 receptor family signaling pathways. Mol. Cell 1998. 2: 253–258.
29 Kawai, T., Adachi, O., Ogawa, T., Takeda, K. and Akira, S., Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity 1999. 11:
115–122.
30 Scanga, C. A., Aliberti, J., Jankovic, D., Tilloy, F., Bennouna, S., Denkers,
E. Y., Medzhitov, R. and Sher, A., Cutting edge: MyD88 is required for
resistance to Toxoplasma gondii infection and regulates parasite-induced
IL-12 production by dendritic cells. J. Immunol. 2002. 168: 5997–6001.
31 Seki, E., Tsutsui, H., Tsuji, N. M., Hayashi, N., Adachi, K., Nakano, H.,
Futatsugi-Yumikura, S. et al., Critical roles of myeloid differentiation factor
88-dependent proinflammatory cytokine release in early phase clearance of
Listeria monocytogenes in mice. J. Immunol. 2002. 169: 3863–3868.
32 Takeuchi, O., Hoshino, K. and Akira, S., Cutting edge: TLR2-deficient and
MyD88-deficient mice are highly susceptible to Staphylococcus aureus
infection. J. Immunol. 2000. 165: 5392–5396.
33 Feng, C. G., Scanga, C. A., Collazo-Custodio, C. M., Cheever, A. W., Hieny,
S., Caspar, P. and Sher, A., Mice lacking myeloid differentiation factor 88
display profound defects in host resistance and immune responses to
Mycobacterium avium infection not exhibited by Toll-like receptor 2 (TLR2)and TLR4-deficient animals. J. Immunol. 2003. 171: 4758–4764.
34 Fremond, C. M., Yeremeev, V., Nicolle, D. M., Jacobs, M., Quesniaux, V. F.
and Ryffel, B., Fatal Mycobacterium tuberculosis infection despite adaptive
immune response in the absence of MyD88. J. Clin. Invest. 2004. 114:
1790–1799.
www.eji-journal.eu
693
694
Christoph Hölscher et al.
Eur. J. Immunol. 2008. 38: 680–694
35 Quesniaux, V., Fremond, C., Jacobs, M., Parida, S., Nicolle, D., Yeremeev,
V., Bihl, F. et al., Toll-like receptor pathways in the immune responses to
mycobacteria. Microbes Infect. 2004. 6: 946–959.
44 Cooper, A. M., Dalton, D. K., Stewart, T. A., Griffin, J. P., Russell, D. G.
and Orme, I. M., Disseminated tuberculosis in interferon gamma genedisrupted mice. J. Exp. Med. 1993. 178: 2243–2247.
36 Rausch, A., Hessmann, M., Holscher, A., Schreiber, T., Bulfone-Paus, S.,
Ehlers, S. and Holscher, C., Interleukin-15 mediates protection against
experimental tuberculosis: A role for NKG2D-dependent effector mechanisms of CD8+ T cells. Eur. J. Immunol. 2006. 36: 1156–1167.
46 Sun, D. and Ding, A., MyD88-mediated stabilization of interferon-gammainduced cytokine and chemokine mRNA. Nat. Immunol. 2006. 7: 375–381.
37 Shi, S., Nathan, C., Schnappinger, D., Drenkow, J., Fuortes, M., Block, E.,
Ding, A. et al., MyD88 primes macrophages for full-scale activation by
interferon-gamma yet mediates few responses to Mycobacterium tuberculosis. J. Exp. Med. 2003. 198: 987–997.
38 Carroll, P., Muttucumaru, D. G. and Parish, T., Use of a tetracyclineinducible system for conditional expression in Mycobacterium tuberculosis
and Mycobacterium smegmatis. Appl. Environ. Microbiol. 2005. 71:
3077–3084.
39 Brightbill, H. D., Libraty, D. H., Krutzik, S. R., Yang, R. B., Belisle, J. T.,
Bleharski, J. R., Maitland, M. et al., Host defense mechanisms triggered by
microbial lipoproteins through toll-like receptors. Science 1999. 285:
732–736.
40 Underhill, D. M., Ozinsky, A., Smith, K. D. and Aderem, A., Toll-like
receptor-2 mediates mycobacteria-induced proinflammatory signaling in
macrophages. Proc. Natl. Acad. Sci. USA 1999. 96: 14459–14463.
41 Kamath, A. B., Alt, J., Debbabi, H. and Behar, S. M., Toll-like receptor 4defective C3H/HeJ mice are not more susceptible than other C3H substrains
to infection with Mycobacterium tuberculosis. Infect. Immun. 2003. 71:
4112–4118.
42 Branger, J., Leemans, J. C., Florquin, S., Weijer, S., Speelman, P. and van
der Poll, T., Toll-like receptor 4 plays a protective role in pulmonary
tuberculosis in mice. Int. Immunol. 2004. 16: 509–516.
43 Flynn, J. L., Chan, J., Triebold, K. J., Dalton, D. K., Stewart, T. A. and
Bloom, B. R., An essential role for interferon gamma in resistance to
Mycobacterium tuberculosis infection. J. Exp. Med. 1993. 178: 2249–2254.
f 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
45 Han, J., MyD88 beyond Toll. Nat. Immunol. 2006. 7: 370–371.
47 Blander, J. M. and Medzhitov, R., Regulation of phagosome maturation by
signals from toll-like receptors. Science 2004. 304: 1014–1018.
48 Yates, R. M. and Russell, D. G., Phagosome maturation proceeds
independently of stimulation of toll-like receptors 2 and 4. Immunity
2005. 23: 409–417.
49 Sugawara, I., Yamada, H., Kaneko, H., Mizuno, S., Takeda, K. and Akira,
S., Role of interleukin-18 (IL-18) in mycobacterial infection in IL-18-genedisrupted mice. Infect. Immun. 1999. 67: 2585–2589.
50 Fremond, C. M., Togbe, D., Doz, E., Rose, S., Vasseur, V., Maillet, I.,
Jacobs, M. et al., IL-1 receptor-mediated signal is an essential component of
MyD88-dependent innate response to Mycobacterium tuberculosis infection.
J. Immunol. 2007. 179: 1178–1189.
51 Martinon, F. and Tschopp, J., NLRs join TLR as innate sensors of pathogens.
Trends Immunol. 2005. 28: 447–454.
52 van der Wel, N., Hava, D., Houben, D., Fluitsma, D., van Zon, M., Pierson,
J., Brenner, M. and Peters, P. J., M. tuberculosis and M. leprae translocate
from the phagolysosome to the cytosol in myeloid cells. Cell 2007. 129:
1287–1298.
53 Holscher, C., Holscher, A., Ruckerl, D., Yoshimoto, T., Yoshida, H., Mak,
T., Saris, C. and Ehlers, S., The IL-27 receptor chain WSX-1 differentially
regulates antibacterial immunity and survival during experimental tuberculosis. J. Immunol. 2005. 174: 3534–3544.
54 Holscher, C., Kohler, G., Muller, U., Mossmann, H., Schaub, G. A. and
Brombacher, F., Defective nitric oxide effector functions lead to extreme
susceptibility of Trypanosoma cruzi-infected mice deficient in gamma
interferon receptor or inducible nitric oxide synthase. Infect. Immun. 1998.
66: 1208–1215.
www.eji-journal.eu