Cathepsin G and Neutrophil Elastase
Contribute to Lung-Protective Immunity
against Mycobacterial Infections in Mice
This information is current as
of June 14, 2017.
Kathrin Steinwede, Regina Maus, Jennifer Bohling, Sabrina
Voedisch, Armin Braun, Matthias Ochs, Andreas Schmiedl,
Florian Länger, Francis Gauthier, Jürgen Roes, Tobias
Welte, Franz C. Bange, Michael Niederweis, Frank Bühling
and Ulrich A. Maus
Supplementary
Material
References
Subscription
Permissions
Email Alerts
http://www.jimmunol.org/content/suppl/2012/03/28/jimmunol.110334
6.DC1
This article cites 43 articles, 20 of which you can access for free at:
http://www.jimmunol.org/content/188/9/4476.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2012 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
J Immunol 2012; 188:4476-4487; Prepublished online 28
March 2012;
doi: 10.4049/jimmunol.1103346
http://www.jimmunol.org/content/188/9/4476
The Journal of Immunology
Cathepsin G and Neutrophil Elastase Contribute to
Lung-Protective Immunity against Mycobacterial Infections
in Mice
Kathrin Steinwede,* Regina Maus,* Jennifer Bohling,* Sabrina Voedisch,†
Armin Braun,† Matthias Ochs,‡ Andreas Schmiedl,‡ Florian Länger,x Francis Gauthier,{
Jürgen Roes,‖ Tobias Welte,# Franz C. Bange,** Michael Niederweis,†† Frank Bühling,‡‡
and Ulrich A. Maus*
T
uberculosis remains a global health threat. Approximately
8.8 million people are newly infected with Mycobacterium
tuberculosis, and ∼1.4 million people die annually from
tuberculosis (1). In the mouse model, the early phase of myco-
*Department of Experimental Pneumology, Hannover Medical School, Hannover
30625, Germany; †Department of Immunology, Allergology, and Immunotoxicology,
Fraunhofer Institute of Toxicology and Experimental Medicine, Hannover 30625,
Germany; ‡Institute of Functional and Applied Anatomy, Hannover Medical School,
Hannover 30625, Germany; xInstitute of Pathology, Hannover Medical School, Hannover 30625, Germany; {INSERM U618, Proteases et Vectorisation Pulmonaires,
Université François Rabelais de Tours, Tours 37032, France; ‖Department of Immunology and Molecular Pathology, University College London Medical School, London W1T 4JF, United Kingdom; #Clinic for Pneumology, Hannover Medical School,
Hannover 30625, Germany; **Institute of Medical Microbiology and Hospital Epidemiology, Hannover Medical School, Hannover 30625, Germany; ††Department of
Microbiology, University of Alabama at Birmingham, Birmingham, AL 35294; and
‡‡
Institute of Laboratory Medicine, Carl-Thiem-Klinikum, Cottbus 03048, Germany
Received for publication November 22, 2011. Accepted for publication February 23,
2012.
This work was supported by grants from the German Research Foundation (to U.A.M.
and T.W.) and the Society of Lower Saxony for the Control of Tuberculosis (to
U.A.M.).
Address correspondence and reprint requests to Prof. Ulrich A. Maus, Department of
Experimental Pneumology, Hannover Medical School, Feodor-Lynen-Strasse 21,
Hannover 30625, Germany. E-mail address: [email protected]
The online version of this article contains supplemental material.
Abbreviations used in this article: BAL, bronchoalveolar lavage; BCG, bacillus
Calmette-Guérin; CG, cathepsin G; CtsB, cathepsin B; DC, dendritic cell; dLN,
draining lymph node; FRET, fluorescence resonance energy transfer; FSC, forward
light scatter; KO, knockout; NE, neutrophil elastase; SSC, side light scatter.
Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103346
bacterial infections of the lung is characterized by nearly exponential growth of mycobacteria within alveolar macrophages (2).
Control of mycobacterial replication is initiated by adequate
T cell-dependent activation of macrophages to overcome mycobacteria-induced arrest of phagosomal maturation (3, 4).
Professional phagocytes are a heterogeneous group of phagocytic cells of myeloid origin, including neutrophils and mononuclear phagocytes, which subdivide in organ-specific macrophages
and myeloid dendritic cell (DC) subsets. These cells are differentially equipped with an array of lysosomal proteolytic and antimicrobial effector molecules, including cysteine, aspartate, and
serine proteases (cathepsins), as well as noncathepsin serine proteases like neutrophil elastase (NE) and proteinase-3, and various
defensins (3, 5). The two neutrophil serine proteases cathepsin G
(CG) and NE are classically expressed in azurophilic granules of
neutrophils, where they are stored in their active form. Both serine
proteases are synthesized as zymogens and are activated after cleavage and removal of the N-terminal prosequence of two amino acid
residues (6). Besides their involvement in various biological processes like degradation of extracellular matrix components or the
regulation of immune responses through processing of chemokines and cytokines, CG and NE exert antimicrobial activity
against various pathogens (6, 7). It has been shown that CG is
implicated in the antimicrobial defense against Staphylococcus
aureus, and, together with NE, also against Aspergillus fumigatus
infection in mice (8, 9). Recently, work of our group showed that
neutrophil-derived CG and NE are crucial for effective lung innate
immune responses against Streptococcus pneumoniae infections in
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
The neutrophil serine proteases cathepsin G (CG) and neutrophil elastase (NE) are involved in immune-regulatory processes and
exert antibacterial activity against various pathogens. To date, their role and their therapeutic potential in pulmonary host defense
against mycobacterial infections are poorly defined. In this work, we studied the roles of CG and NE in the pulmonary resistance
against Mycobacterium bovis bacillus Calmette-Guérin (BCG). CG-deficient mice and even more pronounced CG/NE-deficient
mice showed significantly impaired pathogen elimination to infection with M. bovis BCG in comparison to wild-type mice.
Moreover, granuloma formation was more pronounced in M. bovis BCG-infected CG/NE-deficient mice in comparison to CGdeficient and wild-type mice. A close examination of professional phagocyte subsets revealed that exclusively neutrophils shuttled
CG and NE into the bronchoalveolar space of M. bovis BCG-infected mice. Accordingly, chimeric wild-type mice with a CG/NEdeficient hematopoietic system displayed significantly increased lung bacterial loads in response to M. bovis BCG infection.
Therapeutically applied human CG/NE encapsulated in liposomes colocalized with mycobacteria in alveolar macrophages, as
assessed by laser scanning and electron microscopy. Importantly, therapy with CG/NE-loaded liposomes significantly reduced
mycobacterial loads in the lungs of mice. Together, neutrophil-derived CG and NE critically contribute to deceleration of
pathogen replication during the early phase of antimycobacterial responses. In addition, to our knowledge, we show for the first
time that liposomal encapsulated CG/NE exhibit therapeutic potential against pulmonary mycobacterial infections. These findings
may be relevant for novel adjuvant approaches in the treatment of tuberculosis in humans. The Journal of Immunology, 2012,
188: 4476–4487.
The Journal of Immunology
Materials and Methods
Animals
Female control mice (129 S2/SvPasCrl) were initially purchased from
Charles River Laboratories. CG-deficient mice and CG/NE double-knockout
(KO) mice on a 129 S2 background were generated, as described elsewhere
(9), and were obtained from the Medical Research Council (Oxford, U.K.).
All animals were used for the described experiments at 8–12 wk of age, and
matched for sex and genetic background. This study was carried out in
strict accordance with the guidelines of the Animal Care and Use Committee of the Central Animal Facility at the Hannover School of Medicine.
Animal experiments were approved by the Lower Saxony State Office for
Consumer Protection and Food Safety (Hannover, Germany).
Cy5.5, CD8a allophycocyanin, CD19 PE, and CD45 PE Cy7, were purchased from BD Biosciences (Heidelberg, Germany). MACS kit and
CD11c, CD4, CD8a, CD19, and Ly-6G beads were all purchased from
Miltenyi Biotec (Bergisch Gladbach, Germany). FITC-labeled BSA
(FITC-BSA) and lipophilic PKH26-GL fluorescent cell linker kit were
purchased from Sigma-Aldrich (Steinheim, Germany).
Preparation of liposome-encapsulated CG and NE
Liposome encapsulation of CG and NE was done as previously reported,
with some modifications (18–20). Briefly, 4 mg cholesterol was added to 43
mg egg phosphatidylcholine, and the chloroform phase was evaporated
with a rotary evaporator (Büchi, Flawil, Switzerland). The CG- and NEcontaining solution was prepared by dissolving 3 mg of each enzyme in 2.5
ml sterile PBS, and this solution was added to the lipid film and mixed
thoroughly. Empty liposomes were made by addition of PBS alone. This
solution was sonicated and centrifuged at 13,000 rpm for 1 h at 4˚C. The
supernatant was discarded, and the liposomal pellet was resuspended twice
in sterile PBS and centrifuged again. Subsequently, liposomes were
resuspended in 2.5 ml sterile PBS, kept at 4˚C, and used within 24 h after
preparation. In selected experiments, liposomes were labeled with the lipophilic dye, PKH26 (10 mM), according to the manufacturer’s instructions, or were loaded with FITC-BSA (1 or 5 mg) to control uptake of
liposomes by alveolar macrophages in vivo using confocal laser-scanning
microscopy and flow cytometric analysis, respectively.
Culture and quantification of M. bovis BCG
GFP-expressing M. bovis BCG (GFP-BCG) and M. bovis BCG (both strain
Pasteur) were cultured in Middlebrook 7H9 medium until midlog phase
(15) and then frozen at 280˚C until use. For quantification, mycobacteria
were serially diluted in Middlebrook 7H9 medium and plated on Middlebrook 7H10 agar plates. After 3 wk of incubation at 37˚C, CFU were
determined.
Infection of mice
Mice were anesthetized by i.m. application of tetrazoline hydrochloride (2.5
mg/kg; Rompun, Bayer, Leverkusen, Germany) and ketamine (50 mg/kg;
Ketamin Gräub, Albrecht, Aulendorf, Germany), followed by low-dose
infection of mice with M. bovis BCG (2 3 105 CFU/mouse), which was
done via intratracheal application, as described recently in detail (15).
Intratracheal instillation of LPS
Intratracheal instillation of LPS (20 mg/mouse) into the lungs of mice was
performed in the same manner as described for the intratracheal instillation
of M. bovis BCG. In selected experiments, mice received intratracheal
applications of 20 mg LPS/mouse for 12–24 h to trigger neutrophil recruitment into the bronchoalveolar space. Subsequently, mice were euthanized, and elicited neutrophils (purity .95%) contained in bronchoalveolar lavage (BAL) fluids were collected for further experiments
in vitro.
Intratracheal instillation of liposome-encapsulated CG and NE
At 24 h post-M. bovis BCG infection, wild-type mice were anesthetized
with desfluran and received a single (day 1 postinfection) or a repetitive
application (days 1 and 7 postinfection) of liposomal CG/NE (100 ml per
mouse) or PBS-containing liposomes as control. Intratracheal instillation
of liposome-encapsulated CG/NE into the lungs of mice was performed in
the same manner as described for the intratracheal instillation of M. bovis
BCG.
Reagents
BAL and preparation of lung, spleen, and lung draining lymph
nodes for quantification of bacterial loads
Purified CG from human sputum was purchased from Sigma-Aldrich
(Steinheim, Germany) and Biocentrum (Krakow, Poland). Purified cathepsin B (CtsB) from human liver was obtained from Calbiochem
(Darmstadt, Germany). Cholesterol and egg phosphatidylcholine were
purchased from Sigma-Aldrich (Steinheim, Germany) and Avanti Polar
Lipids (Alabaster, AL), respectively. Goat anti-mouse CG polyclonal Ab
(clone I-19) was obtained from Santa Cruz Biotechnology (Santa Cruz,
CA). Mouse anti-mouse b-actin mAb was obtained from Sigma-Aldrich
(München, Germany; clone AC 15), and peroxidase-conjugated donkey
anti-goat polyclonal IgG (H+L) and peroxidase-conjugated goat antimouse polyclonal IgG (H+L) were from Jackson ImmunoResearch Laboratories (Suffolk, U.K.). All Abs used in the current study for flow
cytometry analysis, including anti-CD11c PE-Cy5.5, anti-CD11b PE-Cy7,
anti-MHC II PE, anti–Ly-6G/Ly-6C (GR-1) FITC, CD3 FITC, CD4 PerCP
BAL was performed, as described recently (15, 21, 22). Total BAL fluid
cells were counted, followed by preparation of cytospins and Pappenheim
staining used for determination of leukocyte subset differentials. Remaining BAL cells were lysed in HBSS containing 0.1% saponin (pH 7.2) for
10 min at 37˚C, and then plated in 10-fold serial dilutions on Middlebrook
7H10 agar plates and incubated at 37˚C for 3 wk for CFU determination.
Lung lobes, spleen, and lung draining lymph nodes (dLN) were carefully
isolated, while avoiding any contamination of organs with conducting
airways. Lung and spleen tissue was homogenized using a tissue homogenizer (IKA, Staufen, Germany), the resulting homogenates were further
lysed in HBSS containing 0.1% saponin (pH 7.2), and 10-fold serial
dilutions of lung or spleen homogenates were plated on 7H10 agar plates
for determination of CFU. Determination of mycobacterial loads in lung
dLN was done by lysis of the lung dLN in 0.1% saponin (pH 7.2) for
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
mice (10). In line with this finding, CG, NE, and proteinase-3 were
recognized to contribute to human neutrophil killing of S. pneumoniae in vitro (11). In addition, NE is important for the elimination of Klebsiella pneumoniae and Escherichia coli as well as
Candida albicans and Pseudomonas aeruginosa (8, 12–14).
In contrast, the role of CG and neutral serine protease NE in
chronic infections of the lung elicited by mycobacterial pathogens
is poorly defined. We recently demonstrated that among various
cathepsins, particularly CG, but not cathepsin K, L, S, D, or B,
mRNA was strongly upregulated in Mycobacterium bovis bacillus
Calmette-Guérin (BCG)-infected alveolar macrophages in mice
at 7 d postinfection. CG expression triggered in vivo by lung
mycobacterial infections may therefore be part of an endogenous
host defense program against mycobacteria (15). This suggestion
is further supported by a recent study demonstrating that in NO
synthase 2-deficient mice CG contributes to control of M. tuberculosis in hypoxic lung granulomas in mice (16). In contrast,
previous work demonstrated that CG was downregulated in a
human monocytic THP-1 cell line in response to M. tuberculosis
in vitro, suggesting a possible escape mechanism of mycobacteria
to overcome mycobacterial killing by macrophages (17). However, there are currently no data available to directly specify roles
for CG and NE in the lung host defense against mycobacterial
infections in vivo. Therefore, CG-deficient mice, CG/NE-deficient
mice, and respective wild-type controls were infected via the airways with M. bovis BCG, and the course and severity of infection
were analyzed. Our study provides evidence that CG and NE are
involved in early containment of mycobacterial growth. Capitalizing on this finding, we then explored a novel therapeutic antimycobacterial approach by treating M. bovis BCG-infected mice
with CG and NE encapsulated in liposomes to achieve a targeted
uptake of exogenously administered proteolytic activities specifically by alveolar macrophages, the primary target cells of mycobacterial infections. To our knowledge, the data show for the first
time that therapeutically applied liposomal encapsulated human
CG/NE colocalize with mycobacteria within the same compartment of alveolar macrophages and improve lung-protective immunity against mycobacterial infections.
4477
4478
NEUTROPHIL SERINE PROTEASES IN MYCOBACTERIAL INFECTIONS
10 min at 37˚C, followed by incubation for 10 min in an ultrasonic bath.
Afterward, lung dLN were crushed with a pistil, and homogenates were
10-fold serially diluted and plated onto 7H10 agar plates.
Determination of antimycobacterial activity of CG
For determination of the antimycobacterial activity of CG in vitro, M. bovis
BCG cultures were grown to midlog phase, and 100-ml aliquots containing
∼1 3 106 mycobacteria were incubated with human CG or human CtsB
serving as a control at different concentrations (1, 10, 25, 50, or 100 mg/ml)
for 72 h at 37˚C both under aerobic or anaerobic conditions in 96-well plates.
Maintenance of anaerobic culture conditions was achieved by using the
anaerocult system (Merck, Darmstadt, Germany). Briefly, plates and
anaerogen sachets (Oxoid, Cambridge, U.K.) were placed in an anaerobic
jar (Merck). Anaero-test strips were fixed in the container to control for
anaerobic conditions. Subsequently, the anaerobic jar was sealed and incubated at 37˚C. After 72 h of culture, 10-fold serial dilutions of samples
were prepared in 7H9 medium and plated onto 7H10 agar plates for determination of CFU.
Isolation of peripheral blood monocytes, lung CD11c+
mononuclear phagocyte subsets, and lung neutrophils
Isolation and digestion of lung dLN
To analyze numbers of DCs and the proliferation of lymphocyte subsets
in dLN under various experimental conditions, we also collected dLN of
the lung in 1 ml digestion solution, followed by disruption of the tissue
with forceps. Subsequently, dLN were incubated at 37˚C in a water bath for
15 min with repeated shaking. A 1-ml digestion solution was added, and
lymph nodes were incubated for additional 15 min. Thereafter, 12 ml
FACS buffer was added, and the samples were centrifuged at 400 3 g, 4˚C,
for 5 min. The supernatant was discarded, and the pellet was resuspended
in 2 ml FACS buffer and filtered through a 40-mm cell strainer. Cells were
counted, followed by FACS analysis of dLN cellular constituents.
Immunophenotypic analysis of mononuclear phagocyte
and lymphocyte subsets in lung parenchymal tissue and dLN
CD11c+ mononuclear phagocyte subsets as well as lymphocyte subsets isolated from lung parenchyma and lung dLN were immunophenotypically analyzed according to their cell surface Ag expression profiles, as described
previously (24, 25). First, the respective CD11c+ mononuclear phagocyte
subsets purified from lung homogenates of CG/NE-deficient and wild-type
mice were gated according to their forward light scatter (FSC)-A versus side
light scatter (SSC)-A characteristics and FSC-A versus autofluorescence
properties. Low autofluorescent cells were further characterized as conventional DCs (CD11bpos/CD11cpos/MHCIIpos), or as highly autofluorescent lung
macrophages (F4/80pos/CD11bneg/CD11cpos/MHCIIneg), or as autofluorescent
exudate macrophages (F4/80pos/CD11bpos/CD11cpos/MHCIIneg-low). Immunophenotypic analysis of lymphocyte subsets in lung dLN was done by gating
according to their FSC-A versus SSC-A characteristics, followed by characterization of CD4+ T cells (CD3pos/CD4pos/CD8neg), CD8+ T cells (CD3pos/
CD4neg/CD8pos), and B cells (CD3neg/CD19pos).
Isolation of alveolar macrophages and treatment with
liposomal CG and NE
Alveolar macrophages were recovered from untreated mice by BAL, and 1
million BAL cells (.99% alveolar macrophages) were seeded into cell
culture dishes in RPMI 1640 supplemented with 10% FCS, 1% glutamine,
Generation of chimeric wild-type mice
To better dissect the roles of CG/NE activity of resident lung cells versus
circulating leukocyte subsets to contribute to mycobacterial killing, we
subjected wild-type mice to whole body irradiation, and then transplanted
these mice with bone marrow collected from CG/NE double-mutant mice to
generate chimeric wild-type mice with a CG/NE-deficient hematopoietic
system. For isolation of bone marrow cells, tibias and femurs of sexmatched CG/NE KO and wild-type donor mice were flushed with RPMI
1640 containing 10% FCS, and single-cell suspensions were prepared. Cell
aggregates were removed by careful filtration of the cell suspensions
through 40 mm nylon meshes. Bone marrow cell suspensions were washed
twice in Leibovitz L15 medium. Recipient wild-type mice received a total
body irradiation of 8 Gy delivered by a linear accelerator (21). Within 16 h
after irradiation, recipient wild-type mice received ∼107 CG/NE KO bone
marrow cells or wild-type bone marrow cells (transplant controls) suspended in Leibovitz medium without supplements via lateral tail vein
injections. Resulting chimeric CG/NE wild-type mice exhibiting a CG/NEdeficient hematopoietic system, but wild-type alveolar and lung parenchymal macrophages, were housed under specific pathogen-free conditions
with free access to autoclaved food and water for 5 wk, before M. bovis
BCG infections were initiated.
Western blot analysis
For cell-specific detection of CG and NE protein in BAL cells or monocytes
or CD11c+ lung mononuclear phagocytes or neutrophils, purified cell
populations were lysed in ice-cold lysis buffer (100 ml/1–2 3 106 cells),
and total protein concentrations of lysates were determined by BCA Pierce
assay (Thermo Scientific, Waltham, MA), according to the manufacturer’s
instructions (26). For SDS-PAGE, protein samples diluted 1:4 with 43
reducing Roti-Load buffer (Roth, Karlsruhe, Germany) were heated for 15
min at 99˚C, and then cooled at room temperature prior to loading of equal
amounts of protein (15 mg) and Precision Plus Protein Standard (Dual
Color; Bio-Rad, München, Germany) onto a 12.5% SDS-polyacrylamide
gel. Separated proteins were then electrophoretically transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA) via semidry
blotting. Detection of CG/NE immunoreactivity was done using an anti-CG
Ab (clone I-19; Santa Cruz Biotechnology) with specificity for CG and
cross-reacting with NE (K. Steinwede and U. Maus, unpublished observations) using chemiluminescence ECL substrate (GE Healthcare, Buckinghamshire, U.K.). Stripped blots were reprobed with anti–b-actin mAb to
verify equal sample loading.
Measurement of liposomal encapsulated neutrophil serine
protease content using fluorescence resonance energy transfer
substrates
Amounts of liposomal encapsulated CG and NE were determined, as described recently (10, 27, 28). Briefly, 100 ml CG/NE-loaded liposomes
were centrifuged at 13,000 3 rpm at 4˚C for 30 min. Subsequently, liposome pellets were resuspended in 200 ml ice-cold PBS supplemented
with 0.15% Brij 35 (Sigma-Aldrich, Steinheim, Germany), followed by
one freeze/thaw cycle. Then, liposome preparations were sonicated for 3 3
10 s with 1-min break on ice (UW 2200; Bandelin Electronics, Berlin,
Germany). Liposome lysates were centrifuged at 13,000 rpm at 4˚C for 25
min, resulting in supernatants and pellets, respectively. Supernatants were
diluted in 50 mM PBS (pH 7.4) to adjust Brij 35 concentrations to final
0.05%. Pellets of lysed liposomes were again resuspended in 200 ml icecold PBS supplemented with Brij 35. Subsequently, supernatants and
pellets of lysed CG/NE liposomes were subjected to kinetic measurements
of peptidase activities to determine their CG/NE content using highly
sensitive Abz-peptidyl-EDDnp fluorescence resonance energy transfer
(FRET) substrates, which fully discriminate between these two neutrophil
serine proteases, as described recently (10, 27, 28).
Lung histopathology
Histopathological examination of formalin-fixed, paraffin-embedded, and
H&E-stained lung tissue sections (3 mm) from wild-type mice, CG KO
mice, and CG/NE double-KO mice infected with M. bovis BCG was done
under blinded conditions by an experienced lung pathologist (F.L.) using
a Zeiss Axiovert 200 M microscope (29).
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
Anticoagulated blood collected from the vena cava of untreated or M. bovis
BCG-infected mice was diluted 1:4 in PBS, layered over LymphoLyte
(BIOZOL, Eching, Germany), and then centrifuged at 800 3 g for 20 min at
room temperature. The interphase containing monocytes and lymphocytes
was carefully removed and resuspended in PBS. Cells were counted, and
then centrifuged again at 300 3 g for 10 min. Cell pellets were resuspended
in 90 ml MACS buffer and 25 ml anti-CD4 beads, anti-CD8a beads, and
anti-CD19 beads on ice for 15 min. For negative selection of peripheral
blood monocytes, cells were washed, resuspended in MACS buffer, and
loaded onto prepared MACS columns. Monocytes were collected from
the flow-through washes of the column. Subsequently, viable monocytes
(.85% purity) were lysed for determination of CG/NE and b-actin protein.
In addition, we also determined CG/NE protein expression in purified
CD11c+ mononuclear phagocyte subsets and neutrophils collected from the
lungs of mice challenged with M. bovis BCG, following recently published
protocols for the purification of CD11c+ mononuclear phagocytes (15, 23).
Neutrophils were purified from lung homogenates of M. bovis BCGinfected mice using a Ly-6G magnetic cell separation (MACS) kit following the instructions of the manufacturer (Miltenyi Biotec).
1% penicillin, and streptomycin. Alveolar macrophages were allowed to
adhere to the plastic surface for 1 h. Then, cells were washed three times
with cell culture media without FCS to remove nonadherent cells, and 100
ml liposomal CG/NE or empty liposomes as control were added. After 3
and 6 h of liposomal phagocytosis, alveolar macrophages were lysed for
Western blot analysis of immunoreactive CG/NE.
The Journal of Immunology
4479
Confocal laser-scanning microscopy
ELISA
Confocal laser-scanning microscopy was performed using an inverted
LSM 510 META microscope (Zeiss, Jena, Germany) equipped with a
C-Apochromat 403/1.2 W corr water immersion objective, an argon ion
laser (488 nm excitation), and a helium-neon laser (543 nm excitation).
GFP (green) and PKH26-GL (red) emissions were detected with 505/530
and 560/615 band pass filters, respectively. The transmitted light photomultiplier was used for bright field images of samples. Images were analyzed using Imaris (Bitplane) software.
TNF-a and CCL2 release in BAL fluids of uninfected or M. bovis BCGinfected CG-deficient, CG/NE double-mutant, or wild-type mice was determined using commercially available ELISA, according to the manufacturer’s instructions (R&D Systems).
Electron microscopy
FIGURE 1. Determination of mycobacterial loads in
the lung, spleen, and lung dLN of CG-deficient and
wild-type mice infected with M. bovis BCG. Wild-type
mice (black bars) and CG-deficient mice (white bars)
were infected with M. bovis BCG (2 3 105 CFU/
mouse), and mycobacterial loads were determined in
BAL cells (A), lung tissue (B), spleen (C), and lung
dLN (D). Values are shown as mean 6 SEM for n = 9
mice per time point and treatment group. **p , 0.03,
***p , 0.001, significant difference relative to wildtype mice.
All data are given as mean 6 SEM. Differences between controls and respective treatment groups over time were analyzed by ANOVA, followed by
post hoc Dunnett test. Significant differences between groups were analyzed
by Levene’s test for equality of variances, followed by Student t test using
SPSS for Windows software package. Statistically significant differences
between various treatment groups were assumed when p values were ,0.05.
Results
Effect of CG deficiency on pathogen elimination in M. bovis
BCG-infected mice
Our recently made observation that CG, but not cathepsin B, K, L, D,
or S, mRNA is strongly upregulated in sorted alveolar macrophages
of mice challenged with M. bovis BCG in vivo (15) prompted us to
analyze whether selective deletion of CG would influence protective
immunity in mice in response to M. bovis BCG infection. As shown
in Fig. 1, CG KO mice infected with M. bovis BCG showed significantly increased mycobacterial loads in their BAL cells on day
28 postinfection and, even more pronounced, in lung parenchymal
tissue on days 14 and 28 postinfection (Fig. 1A, 1B). CG deficiency
had no significant impact on dissemination of M. bovis BCG into
extrapulmonary organ systems such as the spleen and lung dLN
during an observation period of 28 d (Fig. 1C, 1D).
Effect of CG and NE deficiency on pulmonary protective
immunity against M. bovis BCG
Because CG is part of a serine protease network, the members of
which are all stored in similar amounts in neutrophil azurophil
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
Alveolar macrophage suspensions were centrifuged at 10,000 rpm and
fixed by immersion in fixation solution composed of 1.5% glutaraldehyde
(Agar Scientific, Essex, U.K.) and 1.5% formaldehyde freshly prepared
from depolymerized paraformaldehyde (Merck Chemicals, Darmstadt, Germany) in 0.15 M HEPES buffer (Sigma-Aldrich, Hamburg, Germany; total
osmolarity of 800 mosmol/L and a vehicle osmolarity of 300 mosm/kg at pH
7.35) for at least 3 h. The suspensions were centrifuged at 10,000 rpm for 5
min after each of the following processing steps. After repeated rinsing in
0.15 M HEPES buffer and in 0.1 mmol/L cacodylate buffer (Plano, Wetzlar,
Germany), stabilization of predominantly unsaturated lipids was achieved
by postfixation with 1% osmium tetroxide (Plano, Wetzlar, Germany) in 0.1
M cacodylate buffer, followed by rinsing in 0.1 M cacodylate buffer (2 3 5
min). After twice rinsing in distilled water, specimens were stained en bloc
overnight (12–18 h) with a mixture of equal portions of uranyl-acetate and
water (half-saturated aqueous uranyl acetate solution, 1:1; Agar Scientific,
Stansted, Essex, U.K.) at 4–8˚C. Cell suspensions were dehydrated in an
ascending series of acetone (70, 90, and 100%) (J. T. Baker, Deventer, The
Netherlands) and embedded in epon (SERVA Feinbiochemica, Heidelberg,
Germany). Ultrathin sections (70 nm) were cut by an ultra microtome
(Reichert Ultracut S, München, Germany), collected on nickel grids
(Plano), and stained with lead citrate (Merck, Darmstadt, Germany) and
uranyl acetate. Ultrathin sections were examined with an electron microscope (Morgagni 268; FEI, Eindhoven, The Netherlands). Electron micrographs of representative areas were taken by a digital camera (Veleta TEM
camera; Olympus Europe Holding, Hamburg, Germany).
Statistics
4480
NEUTROPHIL SERINE PROTEASES IN MYCOBACTERIAL INFECTIONS
FIGURE 3. Effect of CG/NE deficiency on lung-protective immunity
against M. bovis BCG. Wild-type, CG-deficient, and CG/NE-deficient
mice were infected with M. bovis BCG (2 3 105 CFU/mouse). At indicated time points, whole lungs were removed, and lung tissue sections
were prepared for histopathologic examination (HE stain) (A) and lung
morphometric examination (B). Representative photographs are original
magnification 32.5 (A). (C) TNF-a levels in BAL fluids of wild-type
(black bars), CG-deficient (white bars), and CG/NE-deficient mice (gray
bars) postinfection with M. bovis BCG. Values are shown as mean 6 SEM
for n = 9 mice per time point and treatment group. *p , 0.05,
***p , 0.001, significant difference relative to wild-type mice.
Leukocyte subset recruitment profiles in wild-type mice,
CG-deficient mice, and CG/NE double-mutant mice infected
with M. bovis BCG
FIGURE 2. Determination of mycobacterial loads in the lung, spleen,
and lung dLN of CG/NE-deficient and wild-type mice infected with
M. bovis BCG. Wild-type and CG/NE-deficient mice were infected with
M. bovis BCG (2 3 105 CFU/mouse). At the indicated time points, mycobacterial loads were determined in BAL cells (A), lung tissue (B), spleen
(C), and dLN (D) of wild-type mice (black bars) and CG/NE-deficient mice
(white bars). Values are shown as mean 6 SEM for n = 9 mice per time
point and treatment group. **p , 0.03, ***p , 0.001, significant difference relative to wild-type mice.
Previous reports have suggested a role for CG in leukocyte extravasation (6, 30, 31). The described differences between wildtype mice, CG KO, and CG/NE KO mice in terms of mycobacterial
pathogen elimination and induction of lung granuloma formation
might thus be explained by differences in lung leukocyte subset
recruitment in response to mycobacterial infection. Therefore, we
analyzed leukocyte subset profiles in BAL fluids of mice of the
various treatment groups postinfection with M. bovis BCG. As
shown in Supplemental Fig. 1, no significant differences in
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
granules, we addressed the question whether additional KO of NE,
a serine protease with strong homology to CG, would further attenuate pulmonary resistance of double-mutant mice to challenge
with M. bovis BCG. Relative to wild-type mice, mice deficient of
both serine proteases CG and NE demonstrated significantly increased mycobacterial loads both in BAL cells and lung parenchymal tissue as early as 14 d postinfection with M. bovis BCG
(Fig. 2A, 2B), which were also significantly increased when compared with CG KO mice at this time point (compare with Fig. 1),
thus demonstrating that additional knockout of NE further attenuated pulmonary resistance to mycobacterial infection in mice.
Moreover, significantly increased bacterial loads were also noted
in the spleens of CG/NE-deficient mice relative to wild-type mice
by day 28 postinfection with M. bovis BCG (Fig. 2C). In contrast,
no differences in mycobacterial loads were observed in dLN between groups (Fig. 2D). Histopathological examination of lung
tissue sections collected from M. bovis BCG-infected wild-type
mice, CG KO mice, and CG/NE double-KO mice revealed that,
particularly on day 28 postinfection, lung granuloma formation was
significantly increased in CG KO mice and even further increased
in CG/NE KO mice, relative to wild-type mice (Fig. 3A, 3B),
whereas on day 56 postinfection, no difference in the degree of lung
granuloma formation was noted between groups (data not shown).
In addition, significantly increased TNF-a protein levels in BAL
fluids were observed in CG/NE KO, but not CG KO mice after
28 d postinfection (Fig. 3C), indicating an increased proinflammatory response in these mice.
The Journal of Immunology
numbers of exudate macrophages, neutrophils, or lymphocytes
were observed in BAL fluids of CG-deficient, CG/NE-deficient, or
wild-type mice in response to M. bovis BCG infection, with peak
alveolar neutrophil and exudate macrophage numbers observed
by day 28 postinfection, followed by peak alveolar lymphocyte
numbers noted on day 56 postinfection (Supplemental Fig. 1A–
D). Moreover, numbers of lung mononuclear phagocyte subsets
and CD45pos/Gr-1pos neutrophils in lung parenchymal tissue were
not significantly different between M. bovis BCG-infected CG/
NE-deficient mice relative to wild-type mice (data not shown).
Also, FACS analysis did not reveal any differences in mobilization of conventional lung DC toward lung dLN, or numbers
of lymphocyte subsets within lung dLN in response to M. bovis
BCG infection in CG or CG/NE KO mice when compared with
wild-type mice by day 7 up to day 56 postinfection (data not
shown). Together, these data demonstrate that neutral serine proteases CG/NE are not required for either leukocyte recruitment
into the lungs or lung dLN or numbers of lymphocytes within lung
dLN of mice infected with M. bovis BCG.
The finding that deficiency of CG significantly decreased the
containment of M. bovis BCG in the lungs of mice prompted us to
analyze whether purified CG would exert direct antimycobacterial
effects in vitro. As shown in Fig. 4A, using aerobic culture conditions, incubation of M. bovis BCG with purified human CG dose
FIGURE 4. Effect of CG on mycobacterial growth
in vitro. M. bovis BCG was plated in triplicates into
cell culture dishes in the absence (white bars) or
presence of CG (black bars) or CtsB (gray bars) for
72 h at 37˚C either under aerobic conditions (A) or
anaerobic conditions (B), as indicated. Subsequently,
mycobacteria were collected and plated in serial
dilutions on 7H10 agar plates for CFU determination.
Values are shown as mean 6 SEM from triplicate
experiments. **p , 0.03, ***p , 0.001, significant
difference relative to untreated control at 72 h. Data
represent at least n = 3 independent experiments.
dependently impaired the mycobacterial outgrowth during an
observation period of 72 h in vitro, with lowest significantly active
CG protein levels amounting to 10 mg/ml culture medium. In
contrast, using anaerobic culture conditions, in which metabolic
activities of mycobacteria are drastically affected, significantly
reduced mycobacterial outgrowth in vitro was only achieved in the
presence of CG applied at 50 and 100 mg/ml (Fig. 4B). Importantly, under both aerobic and anaerobic experimental conditions,
addition of CtsB as a control protease did not impair mycobacterial growth in vitro (Fig. 4).
Identification of neutrophils as the main cellular source of CG
and NE in mice
We further aimed at identifying the cellular source mediating CG/
NE-dependent innate immune responses in mice after mycobacterial challenge. CG and NE are localized in azurophilic granules of
neutrophils (6, 7). Increased CG mRNA levels were also found in
alveolar macrophages of wild-type mice infected with M. bovis
BCG (15), suggesting that both types of professional phagocytes
might contribute to CG bioactivity in mice in response to mycobacterial challenge. In initial experiments, BAL fluid cells collected from the lungs of M. bovis BCG-infected wild-type mice
were subjected to Western blot analysis of CG/NE protein on days
14, 28, and 56 postinfection. As shown in Fig. 5A, both untreated
wild-type mice and mice infected for 14 d with M. bovis BCG did
not show detectable CG/NE protein in their BAL cells, whereas
CG/NE protein was easily detectable in whole BAL cell lysates
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
Effect of purified CG on M. bovis BCG growth under aerobic
and anaerobic conditions in vitro
4481
4482
NEUTROPHIL SERINE PROTEASES IN MYCOBACTERIAL INFECTIONS
lungs of these mice (Supplemental Fig. 1C). In contrast, no immunoreactivity for CG/NE was detectable in BAL cell lysates of
M. bovis BCG-infected CG/NE double-mutant mice (Fig. 5A).
Interestingly, neither peripheral blood monocytes nor CD11c+
mononuclear phagocytes (alveolar and lung macrophages and
myeloid DC subsets) purified from lung parenchymal tissue of
mice infected with M. bovis BCG demonstrated any detectable
CG/NE protein expression by Western blot analysis (Fig. 5B, 5C).
In striking contrast, lung parenchymal tissue neutrophils collected
from mice at day 28 postinfection with M. bovis BCG were found
to express CG/NE protein (Fig. 5C). This indicates that recruited
neutrophils are the most likely cellular shuttle transporting neutral
serine protease bioactivity into the bronchoalveolar compartment
of mice in response to infection with M. bovis BCG.
Hematopoietic CG/NE bioactivity is an important factor in
lung-protective immunity against mycobacterial pathogens
collected from mice at 28 and 56 d postinfection, thereby strongly
correlating with peak alveolar neutrophil influx observed in the
Having demonstrated that alveolar neutrophil recruitment underlies increased CG immunoreactivity in the bronchoalveolar
compartment of mice postinfection with M. bovis BCG, we hypothesized that selective hematopoietic deletion of CG/NE bioavailability would attenuate lung-protective immunity against
M. bovis BCG. To test this hypothesis, wild-type mice were irradiated and reconstituted with bone marrow cells from wild-type
mice (transplant controls) or CG/NE double-KO mice. As expected,
transfer of wild-type bone marrow led to strong CG/NE immunoreactivity in BAL fluid cells collected at 24 h post-LPS challenge (Fig. 6A), whereas mice receiving bone marrow from the
dual mutants showed .90% reduced CG/NE immunoreactivity in
their BAL cells at 24 h posttreatment (Fig. 6A). We found no CG/
NE immunoreactivity in BAL fluid cells 28 d postinfection of
FIGURE 6. Effect of a hematopoietic CG/NE deletion on CG/NE protein content in BAL fluids of LPS-treated or M. bovis BCG-infected mice and
mycobacterial pathogen elimination in the lung and spleen of chimeric wild-type mice. Wild-type mice were irradiated, followed by reconstitution of their
hematopoietic system with bone marrow cells collected from CG/NE double-mutant mice (chimeric wild type) or with bone marrow cells collected from
wild-type mice (transplant control). (A) At 5 wk post-BMT, mice were challenged intratracheally with 20 mg LPS. At 24 h post-LPS treatment, transplant
control mice (left lane) and chimeric wild-type mice (right lane) were subjected to BAL for collection of BAL fluid cellular constituents (.95% neutrophils). (B) At 5 wk post-BMT, mice were infected with M. bovis BCG (2 3 105 CFU/mouse), and BAL fluid cellular constituents of transplant control
mice (left lane) and chimeric wild-type mice (right lane) collected at day 28 postinfection were subjected to analysis of CG/NE immunoreactivity. (C and
D) Mycobacterial loads in transplant control mice (black bars) and chimeric wild-type mice (white bars) in BAL cells (C) and lung tissue (D). The given
Western blots are representatives of at least three independent determinations. Values are shown as mean 6 SEM for n = 9 mice per time point and
treatment group. **p , 0.03, significant difference relative to transplant control mice.
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
FIGURE 5. Analysis of CG/NE protein expression in BAL cells, peripheral
blood monocytes, CD11c+ mononuclear phagocytes, or neutrophils (PMN)
from lung parenchymal tissue of M. bovis BCG-infected mice. Wild-type mice
and CG/NE-deficient mice were either mock infected or infected with M. bovis
BCG (2 3 105 CFU/mouse) for 14, 28, and 56 d. Subsequently, mice were
subjected to BAL for isolation of BAL cells (A), or monocytes were purified
from peripheral blood of wild-type mice (B), or CD11c+ mononuclear phagocyte subsets and neutrophils were isolated from lung parenchymal tissue of
wild-type mice (C) at the indicated time points. Subsequently, CG/NE protein
expression was determined by Western blot analysis. Control represents purified
human CG (left lane) and purified human NE (right lane). Each Western blot
analysis is representative of at least n = 3 independent experiments.
The Journal of Immunology
mutant chimeras with M. bovis BCG (Fig. 6B). Notably, hematopoietic deletion of CG/NE in chimeric wild-type mice provoked
significantly increased mycobacterial loads in the bronchoalveolar
and lung parenchymal compartment of these mice 14 and
28 d postinfection, when compared with M. bovis BCG-infected
transplant control mice (Fig. 6C, 6D). In addition, depletion of
neutrophils with anti-Ly6G Ab 1A8 applied once daily starting by
day 24 after M. bovis BCG infection in wild-type mice resulted
in increased lung bacterial loads at day 28 postinfection when
compared with control IgG-treated M. bovis BCG-infected wildtype mice (data not shown). Together, these data demonstrate that
elicited neutrophils harboring CG/NE contribute to early resistance of the lung to aerogenic mycobacterial infections.
Therapeutic treatment of mice with liposome-encapsulated
CG/NE improves lung innate immunity against M. bovis BCG
infection
The earliest steps in lung mycobacterial infections include invasion
of alveolar macrophages by mycobacteria, followed by phagosomal
arrest induction in macrophages to prevent phagolysosomal fusion
and proteolytic degradation of the pathogen. We asked whether
an increased intracellular uptake of serine protease activities by
mycobacteria-infected alveolar macrophages would increase their
antimycobacterial activities and reduce bacterial loads in distal air
spaces during the early phase of infection. To avoid that CG and NE
released within the lungs of mice would mediate unspecific lung
injury or would be inactivated by endogenous serine protease inhibitors, we encapsulated CG/NE into liposomes, which, according
to our previous studies, are very efficiently phagocytosed by alveolar
macrophages (18). Successful encapsulation of CG and NE into
liposomes was verified by Western blotting, as shown in Fig. 7A.
Subsequent FRET analysis of proteolytic activities of CG and NE in
lysed liposomal preparations (separated into liposomal pellets and
supernatants) revealed the greatest portion of CG enzymatic activity
to be located in the pellets of liposomal lysates, whereas NE was
primarily found in the supernatant of lysed liposomes (Fig. 7B).
This different distribution most probably comes from the strongly
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
FIGURE 7. CG and NE encapsulation in liposomes and liposome phagocytosis by noninfected
or GFP-BCG–infected alveolar macrophages. (A)
CG/NE liposomes or empty PBS-containing liposomes were lysed and analyzed for immunoreactive CG/NE by Western blot. (B) A total of 100 ml
CG/NE liposomes was lysed and centrifuged, and
the amount of enzymatically active CG (black bars)
or NE (white bars), either released into the supernatant or still associated with the pellet, was determined using protease-specific FRET substrates.
(C) Alveolar macrophages were isolated from BAL
of untreated wild-type mice and cultured in the
presence of liposomal CG/NE or empty PBS liposomes. After 3 and 6 h of phagocytosis, alveolar
macrophages were analyzed for immunoreactive
CG/NE by Western blotting. (D) Wild-type mice
were first infected with GFP-BCG (3 3 106 CFU/
mouse) and then received intratracheal application
of PKH26-labeled liposomes containing CG/NE
(100 ml liposomes per mouse) 24 h postinfection.
In vivo uptake of PKH26-labeled liposomes by
GFP-BCG–infected alveolar macrophages was
then determined by confocal laser-scanning microscopy at 96 h postinfection (D). Upper lane,
Three-dimensional reconstruction of a Z stack;
lower lane, single sections of the same stack.
Colocalization is indicated by merge of red and
green color, resulting in yellow. Scale bars, 5 mm.
(E) Wild-type mice were infected with M. bovis
BCG (3 3 106 CFU/mouse) and then received
an intratracheal application of CG/NE-containing
liposomes 24 h postinfection. Colocalization of
mycobacteria (black arrow) and CG/NE-loaded
liposomes (black arrowhead) within alveolar macrophages was analyzed by electron microscopy at
96 h postinfection. Scale bar, 5 mm.
4483
4484
NEUTROPHIL SERINE PROTEASES IN MYCOBACTERIAL INFECTIONS
cationic character of CG (pI ∼ 12) that favors ionic interactions with
liposomal membrane fragments. Next, incubation of alveolar macrophages with CG/NE-containing liposomes in vitro resulted in
increased immunoreactive CG/NE in alveolar macrophages after 3
and 6 h of phagocytosis, whereas no CG/NE protein was detectable
in alveolar macrophages incubated with empty PBS-containing
liposomes (Fig. 7C), demonstrating that liposome-encapsulated
CG/NE indeed accumulates in alveolar macrophages.
To visualize the process of liposomal uptake by alveolar macrophages in vivo, wild-type mice received an intratracheal instillation with FITC-BSA–containing liposomes. Alveolar macrophages collected from mice at 24 h after FITC-BSA liposome
instillation demonstrated strong green fluorescence, indicative of
liposome uptake (Supplemental Fig. 2). Additionally, we exploited
laser-scanning microscopy of alveolar macrophages collected
from mice that were previously infected with GFP-expressing M.
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
FIGURE 8. Therapeutic treatment of M. bovis
BCG-infected mice with liposomal CG and NE.
Wild-type mice were infected with M. bovis
BCG (2 3 105 CFU/mouse) and were then
therapeutically treated with a single application
of CG/NE-containing liposomes or PBS liposomes at day 1 postinfection (A) or a repetitive
application of CG/NE-containing liposomes or
PBS liposomes at day 1 and 7 postinfection (D).
Mycobacterial loads were determined in BAL
cells (B, E) and lung tissue (C, F) at the indicated
time points. (G and H) Numbers of alveolar macrophages and lymphocytes in BAL fluids of M.
bovis BCG-infected mice therapeutically treated
with repetitive applications of CG/NE liposomes
or PBS liposomes, as indicated. Values are
shown as mean 6 SEM for n = 6 mice per time
point and treatment group. *p , 0.05, **p ,
0.03, ***p , 0.001, significant difference relative to wild-type mice treated with PBS liposomes.
bovis BCG, followed by therapeutic treatment of mice with CG/
NE-loaded liposomes that were stained with the lipophilic red
fluorescent dye, PKH26. We were able to show that red fluorescent
CG/NE liposomes and GFP-expressing M. bovis BCG colocalized
within the same macrophage compartments, as indicated by
a yellow overlay fluorescence shown in Fig. 7D. Finally, successful colocalization of mycobacteria with CG/NE liposomes
within alveolar macrophages collected from mice previously
infected with M. bovis BCG and therapeutically treated with CG/
NE-containing liposomes was further confirmed by electron microscopy (Fig. 7E).
We next treated M. bovis BCG-infected wild-type mice therapeutically with a single dose of liposomal CG/NE on day 1 after
mycobacterial infection (Fig. 8A). This resulted in significantly
decreased mycobacterial loads in the bronchoalveolar and lung
parenchymal compartment of mice after 3 and 7 d up until
The Journal of Immunology
4485
Discussion
This study was aimed at elucidating the role of the two neutrophil
serine proteases CG and NE in lung-protective immunity against
mycobacterial infections induced by M. bovis BCG in mice. We
found evidence for a critical contribution of the two closely related
neutrophil serine proteases, CG and NE, in early resistance to
mycobacterial infection. CG mutant mice and, even more pronounced, CG/NE double-mutant mice responded with significantly
increased mycobacterial loads both in the bronchoalveolar and
lung parenchymal compartment to infection with M. bovis BCG
relative to wild-type mice. Importantly, neutrophils were identified
as the primary cellular source and shuttle to transport CG/NE
bioactivity into the lungs of mice during mycobacterial infection.
Along those lines, selective hematopoietic KO of CG/NE protein
availability in chimeric mice provoked strongly increased mycobacterial loads in the lungs relative to transplant controls. Collectively, these data support our view that the observed CG/NE
bioactivity is derived from the hematopoietic system and most
probably confined to circulating neutrophils. Most importantly, to
our knowledge, we demonstrate for the first time that therapeutic
application of liposomal CG/NE targeting BCG-infected alveolar
macrophages led to significantly reduced mycobacterial loads in
distal airspaces of mice, accompanied by substantially increased
numbers of DC and CD4pos T cells in lung dLN.
Despite upregulated CG mRNA levels in response to mycobacterial challenge in vivo (15), we were not able to detect CG
protein in alveolar macrophages of naive or M. bovis BCGinfected mice. Also, CG was not detected in circulating monocytes or pulmonary CD11c+ mononuclear phagocyte subsets collected from mice infected with M. bovis BCG, strongly suggesting
that CG is not expressed in these cells in mice. Previous studies
reported that, in humans, CG protein is produced by peripheral
blood monocytes, albeit at very low levels, just containing ∼5% of
the content as observed in human neutrophils (32), whereas CG and
NE could not be detected by histochemistry in alveolar macrophages
FIGURE 9. Numbers of DC and T lymphocyte subsets in lung dLN after
therapeutic treatment of M. bovis BCG-infected mice with liposomal CG
and NE. Wild-type mice were infected with M. bovis BCG (2 3 105 CFU/
mouse) and then received therapeutic applications of CG/NE-containing
liposomes or PBS-containing liposomes at days 1 and 7 postinfection. At
indicated time points, lung dLN were removed and processed for quantification of DC (A), CD4pos T cells (B), and CD8pos T cells (C) by flow
cytometry. Values are shown as mean 6 SEM for n = 8 mice per time point
and treatment group and n = 3 mice for untreated control. *p , 0.05,
significant difference relative to wild-type mice treated with PBS-containing liposomes.
(33). As such, we cannot fully exclude that very low amounts of CG
and/or NE protein are also present in alveolar macrophages in mice,
yet were not detectable with the currently employed Western blotting techniques. However, when considering the results obtained
from chimeric wild-type mice lacking CG and NE bioactivity in
their hematopoietic system, but not in their lungs, it appears very
unlikely that possible trace amounts of CG and/or NE protein derived from alveolar macrophages would exert any major impact on
lung-protective immunity against mycobacterial infections.
Two aspects support our concept that increased CG/NE activity
observed in the lungs of mice in response to mycobacterial challenge is primarily derived from immigrating blood neutrophils.
First, detection of CG and also NE in the lung coincides with peak
neutrophil influx 28 d after mycobacterial infection, and CG and
NE protein expression was directly detected in purified neutrophils
collected from the lungs of M. bovis BCG-infected mice. Second,
chimeric mice lacking CG and NE in their hematopoietic system
recruited similar numbers of neutrophils into their lungs upon
mycobacterial infection, but no CG/NE antimicrobial activity was
detected in their bronchoalveolar compartment and significantly
increased mycobacterial loads were observed. Some residual im-
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
10 d postinfection (Fig. 8B, 8C). As on day 10 postinfection, we
found mycobacterial loads to again increase in mice treated once
with liposomal CG/NE therapeutically (Fig. 8B, 8C); we next
explored a repetitive liposomal CG/NE therapy in mice infected
with M. bovis BCG (Fig. 8D). Repeated application of liposomal
CG/NE on day 1 and 7 postinfection led to further significantly
reduced mycobacterial loads within distal airspaces of mice particularly on day 10 postinfection (Fig. 8E, 8F), with improved
antibacterial effects still observed on day 14 postinfection. At the
same time, repetitive liposomal CG/NE application into the lungs
of M. bovis BCG-infected mice triggered increased numbers of
alveolar macrophages and lymphocytes on day 10 and 14 postinfection, respectively (Fig. 8G, 8H), whereas recruitment of
neutrophils and inflammatory monocytes was not affected (data not
shown). Such increased numbers of alveolar macrophages were
accompanied by elevated CCL2 levels in BAL fluids in mice receiving repetitive CG/NE liposome therapy (PBS versus CG/NE
liposome treated: 18 6 9 versus 164 6 116 pg/ml CCL2, day 10
postinfection). Moreover, repeated application of liposomal CG/
NE led to increased accumulation of DCs and significantly increased numbers of CD4pos T cells, but not CD8pos T cells in lung
dLN on day 10 postinfection (Fig. 9). Collectively, the data support our concept that therapeutic application of liposomal CG/NE
is an effective novel therapeutic approach to diminish mycobacterial loads within distal airspaces of mice. Therapeutic efficacy
wanes after ∼4 (to 7) days, suggesting that longer-lasting antimycobacterial effects would require repetitive intrapulmonary
applications of liposomal encapsulated proteases.
4486
NEUTROPHIL SERINE PROTEASES IN MYCOBACTERIAL INFECTIONS
ditionally, Reece et al. (16) demonstrated that serine protease
activity of CG functions as a protective mechanism within hypoxic lung granulomas during M. tuberculosis infection. Thus, in
addition to direct antimicrobial activities of CG and NE, immunomodulatory activities of the examined serine proteases may
have contributed to the improved immune response against mycobacterial infections of wild-type mice as opposed to CG KO and
CG/NE KO mice.
To the best of our knowledge, the current study for the first time
directly investigated the role of neutrophil serine proteases on the
lung host defense against mycobacterial infections in mice. The
data show that antimicrobial activities of CG and NE particularly
contribute to control early mycobacterial outgrowth in the acute
phase of mycobacterial infection. Repetitive therapeutic application of liposomal encapsulated human CG/NE significantly reduced mycobacterial loads within the lung. Notably, we observed
that .95% of alveolar macrophages participated in uptake of liposomal CG/NE, as judged by flow cytometry and confocal laserscanning microscopy. We believe this rapid phagocytosis of CG/
NE liposomes by macrophages is the major advantage of the
chosen liposomal CG/NE-based therapeutic approach, because
alveolar macrophages at the same time represent the primary
target cell of mycobacterial infection of the lung. As such, this
therapeutic approach is cell specific, whereby the encapsulated
components are immediately released in target cells within close
vicinity to the invading pathogen. Moreover, given that CG and
NE are both neutral serine proteases exhibiting their maximal
proteolytic activity at neutral pH values, inhibition of macrophage
acidification of the phagosomal compartment by engulfed mycobacteria most probably would favor the proteolytic activity of
these two proteases in vivo. The liposomal encapsulation of the
proteases is an important aspect in safety considerations for the
treatment of mycobacteria-infected individuals with liposomal
CG/NE to avoid induction of acute lung injury by uncontrolled
release of the proteases in the bronchoalveolar compartment, and
also to achieve a macrophage-targeting therapeutic approach. Of
note, intratracheal application of CG without encapsulation led to
significant hemorrhage in mice without any effect on lung mycobacterial loads (data not shown).
Collectively, to our knowledge, we, in this study, for the first
time provide evidence that neutral serine proteases CG and the
closely related NE critically contribute to resistance against mycobacterial infections in mice. We further successfully established
a novel liposome-based therapeutic approach to control mycobacterial outgrowth in the lungs of mice during the early phase of
infection in vivo. This strategy bears promise as an adjuvant
treatment of life-threatening multiresistant or extensively drugresistant mycobacterial infections in humans.
Acknowledgments
We are grateful for the excellent technical assistance of Susanne Faßbender
in preparing sections of alveolar macrophages for electron microscopy.
Christophe Epinette is acknowledged for support in determination of CG
and NE activities using FRET substrates. We thank Stefan Ehlers for
discussion and critical reading of the manuscript.
Disclosures
The authors have no financial conflicts of interest.
References
1. WHO. 2011. Global tuberculosis control. WHO Report 2011.
2. North, R. J., and Y. J. Jung. 2004. Immunity to tuberculosis. Annu. Rev. Immunol.
22: 599–623.
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
munoreactive CG/NE protein detected in chimeric mice receiving
mutant bone marrow is most probably due to incomplete eradication of host hematopoiesis (21). In addition, we recently demonstrated in a S. pneumoniae infection model that depletion of
circulating neutrophils abrogated immunoreactive CG and NE
protein expression in BAL fluids of infected mice (10).
The early phase of mycobacterial infections in mouse lungs is
characterized by exponential growth of the mycobacteria within
distal lung airspaces (2). Recently, it has been shown that virulent
M. tuberculosis is able to turn the apoptotic process of infected
macrophages into a necrotic pathway by blocking the formation
of an apoptotic envelope, thereby causing bacterial escape and
mycobacterial spread throughout the lung (34). During this early
phase of infection, extracellular located mycobacteria may be
attacked directly by newly recruited neutrophils carrying a substantial CG/NE cargo into the bronchoalveolar space. This scenario is strongly supported by the data of the current study.
However, during later phases of mycobacterial infection, most
of the mycobacteria are considered to reside within the arrested
phagosomal compartment of macrophages and DCs, partly within
granulomas (15). Therefore, CG and NE may also contribute to
the killing of mycobacteria, once the pathogen has been engulfed
by macrophages. One concept supports the idea that the proteolytic activity of phagocytosed apoptotic neutrophils is shuttled
into mycobacteria-infected macrophages, thereby possibly supporting the antimycobacterial activity of alveolar macrophages. In
line with this concept, Tan et al. (35) recently showed that alveolar macrophages killed M. tuberculosis more efficiently after
uptake of apoptotic neutrophils, and, in that report, neutrophil
granules were found to colocalize with M. tuberculosis in macrophages, whereas the molecular mechanism of increased M. tuberculosis killing in macrophages was not addressed. Although
future experiments are required to further specify a role for apoptotic neutrophil-derived CG/NE-containing granules in mycobacterial killing by infected macrophages, the aforementioned
scenario, taken together with the data of the current study,
strongly supports the concept that neutrophil-derived CG and NE
participate in the early pathogen elimination process developing
during the acute phase of this chronic infection. This concept is
further supported by the current study showing colocalization
of GFP-expressing M. bovis BCG with red fluorescent CG/NEcontaining liposomes within alveolar macrophages of infected
mice subjected to CG/NE therapy.
The results of the current study support the notion that neutrophils serve a protective role in lung host defense against mycobacterial infections. Recently published reports also suggested
either direct or immunomodulatory roles for neutrophils in mycobacterial pathogen elimination (36–41). In tuberculosis, a strict
regulation of protective immune responses to mycobacterial
challenge is important to limit excessive inflammatory responses,
and hence, irreversible loss of lung structure and function (42).
Seiler et al. (38) showed that neutrophils regulate early granuloma
formation via CXCR3 signaling. Moreover, TNF-a is important in
granuloma formation and maintenance (43). Both CG and NE
have been described to degrade mature TNF-a (44), thereby
limiting neutrophil activation and possibly regulating the overall
inflammatory process, including the formation and maintenance of
granulomas. In line with this observation, CG- and NE-deficient
mice showed increased TNF-a protein levels in their lungs in
response to mycobacterial infections, and granuloma formations
were found to develop earlier in the lungs of mycobacteriainfected CG/NE KO as compared with CG KO and wild-type
mice, despite the fact that similar numbers of alveolar recruited
neutrophils were observed in the various treatment groups. Ad-
The Journal of Immunology
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
overexpress CC chemokine ligand 2 in their lungs show increased protective
immunity to infection with Mycobacterium bovis bacille Calmette-Guérin. J.
Infect. Dis. 198: 1044–1054.
Steinwede, K., O. Tempelhof, K. Bolte, R. Maus, J. Bohling, B. Ueberberg,
F. Länger, J. W. Christman, J. C. Paton, K. Ask, et al. 2011. Local delivery of
GM-CSF protects mice from lethal pneumococcal pneumonia. J. Immunol. 187:
5346–5356.
Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein
measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265–275.
Korkmaz, B., S. Attucci, M. A. Juliano, T. Kalupov, M. L. Jourdan, L. Juliano,
and F. Gauthier. 2008. Measuring elastase, proteinase 3 and cathepsin G activities at the surface of human neutrophils with fluorescence resonance energy
transfer substrates. Nat. Protoc. 3: 991–1000.
Kalupov, T., M. Brillard-Bourdet, S. Dadé, H. Serrano, J. Wartelle, N. Guyot,
L. Juliano, T. Moreau, A. Belaaouaj, and F. Gauthier. 2009. Structural characterization of mouse neutrophil serine proteases and identification of their substrate specificities: relevance to mouse models of human inflammatory diseases.
J. Biol. Chem. 284: 34084–34091.
Winter, C., K. Taut, M. Srivastava, F. Länger, M. Mack, D. E. Briles, J. C. Paton,
R. Maus, T. Welte, M. D. Gunn, and U. A. Maus. 2007. Lung-specific overexpression of CC chemokine ligand (CCL) 2 enhances the host defense to
Streptococcus pneumoniae infection in mice: role of the CCL2-CCR2 axis. J.
Immunol. 178: 5828–5838.
Kessenbrock, K., L. Fröhlich, M. Sixt, T. Lämmermann, H. Pfister, A. Bateman,
A. Belaaouaj, J. Ring, M. Ollert, R. Fässler, and D. E. Jenne. 2008. Proteinase 3
and neutrophil elastase enhance inflammation in mice by inactivating antiinflammatory progranulin. J. Clin. Invest. 118: 2438–2447.
Wittamer, V., B. Bondue, A. Guillabert, G. Vassart, M. Parmentier, and
D. Communi. 2005. Neutrophil-mediated maturation of chemerin: a link between innate and adaptive immunity. J. Immunol. 175: 487–493.
Campbell, E. J., E. K. Silverman, and M. A. Campbell. 1989. Elastase and cathepsin G of human monocytes. Quantification of cellular content, release in
response to stimuli, and heterogeneity in elastase-mediated proteolytic activity.
J. Immunol. 143: 2961–2968.
Kargi, H. A., E. J. Campbell, and C. Kuhn, III. 1990. Elastase and cathepsin G of
human monocytes: heterogeneity and subcellular localization to peroxidasepositive granules. J. Histochem. Cytochem. 38: 1179–1186.
Gan, H., J. Lee, F. Ren, M. Chen, H. Kornfeld, and H. G. Remold. 2008. Mycobacterium tuberculosis blocks crosslinking of annexin-1 and apoptotic envelope formation on infected macrophages to maintain virulence. Nat. Immunol. 9:
1189–1197.
Tan, B. H., C. Meinken, M. Bastian, H. Bruns, A. Legaspi, M. T. Ochoa,
S. R. Krutzik, B. R. Bloom, T. Ganz, R. L. Modlin, and S. Stenger. 2006.
Macrophages acquire neutrophil granules for antimicrobial activity against intracellular pathogens. J. Immunol. 177: 1864–1871.
Fulton, S. A., S. M. Reba, T. D. Martin, and W. H. Boom. 2002. Neutrophilmediated mycobacteriocidal immunity in the lung during Mycobacterium bovis
BCG infection in C57BL/6 mice. Infect. Immun. 70: 5322–5327.
Pedrosa, J., B. M. Saunders, R. Appelberg, I. M. Orme, M. T. Silva, and
A. M. Cooper. 2000. Neutrophils play a protective nonphagocytic role in systemic Mycobacterium tuberculosis infection of mice. Infect. Immun. 68: 577–
583.
Seiler, P., P. Aichele, S. Bandermann, A. E. Hauser, B. Lu, N. P. Gerard,
C. Gerard, S. Ehlers, H. J. Mollenkopf, and S. H. Kaufmann. 2003. Early
granuloma formation after aerosol Mycobacterium tuberculosis infection is
regulated by neutrophils via CXCR3-signaling chemokines. Eur. J. Immunol. 33:
2676–2686.
Blomgran, R., and J. D. Ernst. 2011. Lung neutrophils facilitate activation of
naive antigen-specific CD4+ T cells during Mycobacterium tuberculosis infection. J. Immunol. 186: 7110–7119.
Feng, C. G., M. Kaviratne, A. G. Rothfuchs, A. Cheever, S. Hieny, H. A. Young,
T. A. Wynn, and A. Sher. 2006. NK cell-derived IFN-gamma differentially
regulates innate resistance and neutrophil response in T cell-deficient hosts
infected with Mycobacterium tuberculosis. J. Immunol. 177: 7086–7093.
Martineau, A. R., S. M. Newton, K. A. Wilkinson, B. Kampmann, B. M. Hall,
N. Nawroly, G. E. Packe, R. N. Davidson, C. J. Griffiths, and R. J. Wilkinson.
2007. Neutrophil-mediated innate immune resistance to mycobacteria. J. Clin.
Invest. 117: 1988–1994.
Ehlers, S. 1999. Immunity to tuberculosis: a delicate balance between protection
and pathology. FEMS Immunol. Med. Microbiol. 23: 149–158.
Flynn, J. L., M. M. Goldstein, J. Chan, K. J. Triebold, K. Pfeffer,
C. J. Lowenstein, R. Schreiber, T. W. Mak, and B. R. Bloom. 1995. Tumor
necrosis factor-alpha is required in the protective immune response against
Mycobacterium tuberculosis in mice. Immunity 2: 561–572.
Scuderi, P., P. A. Nez, M. L. Duerr, B. J. Wong, and C. M. Valdez. 1991.
Cathepsin-G and leukocyte elastase inactivate human tumor necrosis factor and
lymphotoxin. Cell. Immunol. 135: 299–313.
Downloaded from http://www.jimmunol.org/ by guest on June 14, 2017
3. Flannagan, R. S., G. Cosı́o, and S. Grinstein. 2009. Antimicrobial mechanisms of
phagocytes and bacterial evasion strategies. Nat. Rev. Microbiol. 7: 355–366.
4. Flynn, J. L., and J. Chan. 2001. Immunology of tuberculosis. Annu. Rev.
Immunol. 19: 93–129.
5. Colbert, J. D., S. P. Matthews, G. Miller, and C. Watts. 2009. Diverse regulatory
roles for lysosomal proteases in the immune response. Eur. J. Immunol. 39:
2955–2965.
6. Pham, C. T. 2006. Neutrophil serine proteases: specific regulators of inflammation. Nat. Rev. Immunol. 6: 541–550.
7. Korkmaz, B., M. S. Horwitz, D. E. Jenne, and F. Gauthier. 2010. Neutrophil
elastase, proteinase 3, and cathepsin G as therapeutic targets in human diseases.
Pharmacol. Rev. 62: 726–759.
8. Reeves, E. P., H. Lu, H. L. Jacobs, C. G. Messina, S. Bolsover, G. Gabella,
E. O. Potma, A. Warley, J. Roes, and A. W. Segal. 2002. Killing activity of
neutrophils is mediated through activation of proteases by K+ flux. Nature 416:
291–297.
9. Tkalcevic, J., M. Novelli, M. Phylactides, J. P. Iredale, A. W. Segal, and J. Roes.
2000. Impaired immunity and enhanced resistance to endotoxin in the absence of
neutrophil elastase and cathepsin G. Immunity 12: 201–210.
10. Hahn, I., A. Klaus, A. K. Janze, K. Steinwede, N. Ding, J. Bohling,
C. Brumshagen, H. Serrano, F. Gauthier, J. C. Paton, et al. 2011. Cathepsin G
and neutrophil elastase play critical and nonredundant roles in lung-protective
immunity against Streptococcus pneumoniae in mice. Infect. Immun. 79: 4893–
4901 .
11. Standish, A. J., and J. N. Weiser. 2009. Human neutrophils kill Streptococcus
pneumoniae via serine proteases. J. Immunol. 183: 2602–2609.
12. Belaaouaj, A., K. S. Kim, and S. D. Shapiro. 2000. Degradation of outer
membrane protein A in Escherichia coli killing by neutrophil elastase. Science
289: 1185–1188.
13. Belaaouaj, A., R. McCarthy, M. Baumann, Z. Gao, T. J. Ley, S. N. Abraham, and
S. D. Shapiro. 1998. Mice lacking neutrophil elastase reveal impaired host defense against gram negative bacterial sepsis. Nat. Med. 4: 615–618.
14. Hirche, T. O., R. Benabid, G. Deslee, S. Gangloff, S. Achilefu, M. Guenounou,
F. Lebargy, R. E. Hancock, and A. Belaaouaj. 2008. Neutrophil elastase mediates
innate host protection against Pseudomonas aeruginosa. J. Immunol. 181: 4945–
4954.
15. Srivastava, M., A. Meinders, K. Steinwede, R. Maus, N. Lucke, F. Bühling,
S. Ehlers, T. Welte, and U. A. Maus. 2007. Mediator responses of alveolar
macrophages and kinetics of mononuclear phagocyte subset recruitment during
acute primary and secondary mycobacterial infections in the lungs of mice. Cell.
Microbiol. 9: 738–752.
16. Reece, S. T., C. Loddenkemper, D. J. Askew, U. Zedler, S. Schommer-Leitner,
M. Stein, F. A. Mir, A. Dorhoi, H. J. Mollenkopf, G. A. Silverman, and
S. H. Kaufmann. 2010. Serine protease activity contributes to control of Mycobacterium tuberculosis in hypoxic lung granulomas in mice. J. Clin. Invest.
120: 3365–3376.
17. Rivera-Marrero, C. A., J. Stewart, W. M. Shafer, and J. Roman. 2004. The downregulation of cathepsin G in THP-1 monocytes after infection with Mycobacterium tuberculosis is associated with increased intracellular survival of bacilli.
Infect. Immun. 72: 5712–5721.
18. Maus, U. A., M. Srivastava, J. C. Paton, M. Mack, M. B. Everhart,
T. S. Blackwell, J. W. Christman, D. Schlöndorff, W. Seeger, and J. Lohmeyer.
2004. Pneumolysin-induced lung injury is independent of leukocyte trafficking
into the alveolar space. J. Immunol. 173: 1307–1312.
19. Koay, M. A., X. Gao, M. K. Washington, K. S. Parman, R. T. Sadikot,
T. S. Blackwell, and J. W. Christman. 2002. Macrophages are necessary for
maximal nuclear factor-kappa B activation in response to endotoxin. Am. J.
Respir. Cell Mol. Biol. 26: 572–578.
20. Maus, U. A., M. A. Koay, T. Delbeck, M. Mack, M. Ermert, L. Ermert,
T. S. Blackwell, J. W. Christman, D. Schlöndorff, W. Seeger, and J. Lohmeyer.
2002. Role of resident alveolar macrophages in leukocyte traffic into the alveolar
air space of intact mice. Am. J. Physiol. Lung Cell. Mol. Physiol. 282: L1245–
L1252.
21. Taut, K., C. Winter, D. E. Briles, J. C. Paton, J. W. Christman, R. Maus,
R. Baumann, T. Welte, and U. A. Maus. 2008. Macrophage turnover kinetics in
the lungs of mice infected with Streptococcus pneumoniae. Am. J. Respir. Cell
Mol. Biol. 38: 105–113.
22. Maus, U. A., M. Backi, C. Winter, M. Srivastava, M. K. Schwarz, T. Rückle,
J. C. Paton, D. Briles, M. Mack, T. Welte, et al. 2007. Importance of phosphoinositide 3-kinase gamma in the host defense against pneumococcal infection. Am. J. Respir. Crit. Care Med. 175: 958–966.
23. Maus, U. A., S. Janzen, G. Wall, M. Srivastava, T. S. Blackwell, J. W. Christman,
W. Seeger, T. Welte, and J. Lohmeyer. 2006. Resident alveolar macrophages are
replaced by recruited monocytes in response to endotoxin-induced lung inflammation. Am. J. Respir. Cell Mol. Biol. 35: 227–235.
24. Schreiber, O., K. Steinwede, N. Ding, M. Srivastava, R. Maus, F. Länger,
J. Prokein, S. Ehlers, T. Welte, M. D. Gunn, and U. A. Maus. 2008. Mice that
4487