IAI Accepts, published online ahead of print on 18 February 2014 Infect. Immun. doi:10.1128/IAI.00096-14 Copyright © 2014, American Society for Microbiology. All Rights Reserved.
1
NADPH oxidase promotes neutrophil extracellular trap formation in
2
pulmonary aspergillosis
3 4
Running title: NADPH oxidase and NETs in pulmonary aspergillosis
5 6 7
Marc Röhm1,2,3, Melissa J. Grimm4, Anthony C. D’Auria4, Nikolaos G. Almyroudis4,5,
8
Brahm H. Segal4,5,6,*,#, Constantin F. Urban1,2,3,*,#
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
1
Department of Clinical Microbiology, Umeå University, Umeå, Sweden Laboratory for Molecular Infection Medicine Sweden, Umeå University, Umeå, Sweden 3 Umeå Centre for Microbial Research, Umeå University, Umeå, Sweden 4 Department of Medicine, Roswell Park Cancer Institute, Buffalo, New York, United States of America 5 Department of Medicine, School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, New York, United States of America 6 Department of Immunology, Roswell Park Cancer Institute, Buffalo, New York, United States of America 2
*
Equally contributing senior authors
#
Correspondence: Constantin F. Urban,
[email protected] Brahm H. Segal,
[email protected] 29 30 31 32 33 34 35
Keywords
36
Aspergillus, NADPH oxidase, chronic granulomatous disease, mouse, p47phox-/-,
37
neutrophil extracellular traps
38 1
39
Abstract
40
NADPH oxidase is a crucial enzyme in antimicrobial host defense and in
41
regulating inflammation. Chronic granulomatous disease (CGD) is an inherited
42
disorder of NADPH oxidase in which phagocytes are defective in generation of ROIs.
43
Aspergillus species are ubiquitous, filamentous fungi, which can cause invasive
44
aspergillosis, a major cause of morbidity and mortality in CGD, reflecting the critical
45
role for NADPH oxidase in antifungal host defense. Activation of NADPH oxidase in
46
neutrophils can be coupled to release of proteins and chromatin that co-mingle in
47
neutrophil extracellular traps (NETs), which can augment extracellular anti-microbial
48
host defense. NETosis can be driven by NADPH oxidase-dependent and –
49
independent pathways. We therefore undertook an analysis of whether NADPH
50
oxidase was required for NETosis in Aspergillus fumigatus pneumonia.
51
Oropharyngeal instillation of live Aspergillus hyphae induced neutrophilic pneumonitis
52
in both wild-type and NADPH oxidase-deficient (p47phox-/-) mice which had resolved in
53
wild-type mice by day 5, but progressed in p47phox-/- mice. NETs, identified by
54
immunostaining, were observed in lungs of WT mice, but were absent in p47phox-/-
55
mice. Using bona fide NETs and nuclear chromatin decondensation as an early
56
NETosis marker, we found that NETosis required a functional NADPH oxidase in vivo
57
and ex vivo. Additionally, NADPH oxidase increased the proportion of apoptotic
58
neutrophils. Together, our results show that NADPH oxidase is required for
59
pulmonary clearance of Aspergillus hyphae and generation of NETs in vivo. We
60
speculate that dual modulation of NETosis and apoptosis by NADPH oxidase
61
enhances antifungal host defense and promotes resolution of inflammation upon
62
infection clearance.
2
63 64
Introduction
65 66
NADPH oxidase is a crucial enzyme in antimicrobial host defense and in
67
regulating inflammation. Patients with chronic granulomatous disease (CGD), an
68
inherited disorder of NADPH oxidase in which phagocytes are defective in generation
69
of reactive oxidant intermediates (ROIs), suffer from life-threatening bacterial and
70
fungal infections (1). CGD is also associated with severe inflammatory complications,
71
such as Crohn’s-like inflammatory bowel disease and obstructive granulomata in the
72
genitourinary system (2). Among CGD patients, the degree of impairment of NADPH
73
oxidase in neutrophils correlates with mortality (3). NADPH oxidase is rapidly
74
activated by conditions that, in nature, are associated with infectious threat, such as
75
ligation of specific pathogen recognition receptors by microbial products (e.g.,
76
formylated peptides and fungal cell wall beta-glucans), opsonized particles, and
77
integrin-dependent adhesion (4-6). Activation of the phagocyte NADPH oxidase
78
(NOX2) requires translocation of cytoplasmic subunits p47phox, p67phox, and p40phox
79
and Rac to a membrane-bound heterodimer cytochrome comprised of gp91phox and
80
p22phox. Molecular oxygen is converted to superoxide anion, which can be converted
81
to downstream metabolites, including H2O2, hydroxyl anion, and hypohalous acid that
82
damage microbes.
83
Invasive mould diseases, most commonly aspergillosis, are major causes of
84
morbidity and mortality in CGD (7-11), a reflection of the critical role for NADPH
85
oxidase in mediating antifungal host defense. Invasive aspergillosis in CGD is a
86
disorder of host defense and of dysregulated neutrophilic inflammation. “Mulch
87
pneumonitis” in CGD is a life-threatening hyperinflammatory response to inhalation of
88
a large fungal inoculum, and is treated with both systemic corticosteroids to limit 3
89
inflammation and antifungal therapy (12). Administration of heat-killed hyphae and
90
fungal cell wall extracts lead to robust and persistent inflammation in NADPH
91
oxidase-deficient mice, but only mild self-limited inflammation in wild-type (WT) mice
92
(13-15), thereby demonstrating an intrinsic role for NADPH oxidase in regulating
93
neutrophilic inflammation in response to fungal cell wall motifs.
94
The kinetics and mode of neutrophil death are important for host defense,
95
persistence versus resolution of inflammation, and the injurious phenotype of
96
neutrophils. Mature neutrophils are likely committed to autocrine death by their
97
constitutive co-expression of cell-surface Fas and FasL via a mechanism that is
98
mediated by the activation of caspases and suppressed by pro-inflammatory
99
cytokines (16). Neutrophil extracellular traps (NETs) are defined by extracellular
100
release of nuclear, cytosolic and granular proteins and chromatin that co-mingle in
101
filamentous structures (17, 18). NETosis is modulated by complex intracellular
102
signaling, which can include activation of the Raf-MEK-ERK pathway, upregulation of
103
anti-apoptotic proteins, and autophagy (19, 20). While neutrophil apoptosis
104
represents non-inflammatory physiological cell death, NETosis results in the release
105
of antimicrobial products that likely amplify extracellular host defense, but can be
106
injurious to host tissue. The balance between NETosis versus apoptosis of
107
neutrophils is likely to be important for defense against pathogens while averting
108
excessive inflammatory injury (21).
109
There is in vitro evidence that, depending on the nature of the stimulus,
110
neutrophil NADPH oxidase can stimulate both apoptosis and NETosis. Impaired
111
neutrophil apoptosis and clearance likely contribute to persistent inflammation in
112
CGD (22-24). Activation of neutrophil NADPH oxidase is also coupled to activation of
113
neutrophil serine proteases in primary granules (25) and to the generation of NETs.
114
NET products bind to and kill bacteria, degrade bacterial virulence factors, and target 4
115
fungi (18, 26, 27). Seen in this light, NADPH oxidase in neutrophils may defend
116
against pathogens both through the direct antimicrobial effect of ROIs and by the
117
subsequent activation of proteases and release of NETs.
118
Despite strong evidence from in vitro studies that NADPH oxidase activation is
119
required for NETosis in stimulated neutrophils (17, 18), the role for NADPH oxidase
120
in driving NETosis in vivo is less clear. We undertook an analysis of whether NADPH
121
oxidase was required for NETosis in Aspergillus fumigatus pneumonia. Pulmonary
122
challenge with conidia (spores) at an inoculum sufficient to cause even mild
123
neutrophilic inflammation in WT mice is rapidly fatal in NADPH oxidase-deficient
124
(p47phox-/-) mice (28). In contrast, we found that oropharyngeal instillation of a live
125
hyphal suspension resulted in hyphal deposition in alveoli and focal neutrophilic
126
pneumonitis in both genotypes at early time points, thus permitting a comparison of
127
the kinetics of neutrophilic accumulation and death. Both live Aspergillus hyphal
128
challenge (29) and administration of killed hyphae (15, 30) have been used in murine
129
models to understand immune responses to hyphae. In contrast to fungal conidia that
130
are phagocytosed, host defense against hyphae relies on extracellular pathways.
131
Use of hyphae as a model of fungal pneumonia enables evaluation of NADPH
132
oxidase-dependent host defense targeted to the tissue invasive form of the
133
pathogen, as distinguished from pathways that phagocytose and kill spores or limit
134
their germination (15, 29, 30).
135
We found that NADPH oxidase was required for hyphal clearance from lungs
136
and resolution of inflammation. Neutrophils in lungs of wild-type mice showed
137
evidence of NETosis, identified by laminar extracellular DNA stretches that co-
138
localized with neutrophil myeloperoxidase and histone immunostaining. In contrast,
139
NET formation was not observed in the lungs of p47phox-/- mice. Neutrophils showing
140
nuclear decondensation, an early marker of NETosis, and no activation of caspase-3 5
141
were detected in WT mice, but were virtually absent in p47phox-/- mice. In addition, the
142
proportion of apoptotic neutrophils was similar between genotypes on day 1 after
143
infection, but significantly greater in wildtype mice on day 3. Consistent with these in
144
vivo results, exposure of neutrophils ex vivo to A. fumigatus hyphae led to NADPH
145
oxidase-dependent nuclear decondensation and NET generation. Together, our
146
results show that NADPH oxidase is required for pulmonary clearance of Aspergillus
147
hyphae and that NADPH oxidase modulates both NETosis and apoptosis of
148
neutrophils in vivo. We speculate that NADPH oxidase-driven NETosis augments
149
host defense against extracellular hyphae while stimulation of neutrophil apoptosis
150
promotes resolution of inflammation and averts excessive tissue injury.
151
6
152
Materials and Methods
153 154
Ethical statement
155
All procedures performed on animals were approved by the Institutional
156
Animal Care and Use Committee at Roswell Park Cancer Institute and complied with
157
the U.S. Department of Health and Human Services’ Guide for the Care and Use of
158
Laboratory Animals.
159 160 161
Mice Targeted disruption of the p47phox gene leads to a defective NADPH oxidase.
162
Phagocytes of mice with this gene defect are not capable of producing measurable
163
amounts of superoxide (31). p47phox-/- mice were derived from C57BL/6 and 129
164
intercrosses and backcrossed 14 generations (N14) in the C57BL/6 background. Age
165
(8-15 weeks) and sex matched C57BL/6 WT mice were used as controls. Mice were
166
bred and maintained under specific pathogen free conditions at the animal care
167
facility at Roswell Park Cancer Institute, Buffalo, NY.
168 169 170
Administration of A. fumigatus A. fumigatus strain Af293 was used. Hyphae were generated in the lab of Dr.
171
Tobias Hohl (Fred Hutchinson Cancer and Research Center, Seattle, WA) by
172
growing the strain in Sabouraud liquid broth (5 × 105 conidia/ml) for 48 h at 37 °C.
173
The mycelia were collected on a Buchner funnel, extensively washed with
174
phosphate-buffered saline, and sonicated to an approximate hyphal fragment size of
175
10 - 100 micrometers. Hyphal suspensions (~750 CFU/mouse) were administered by
176
oropharyngeal instillation. We found that this approach leads to similar degrees of
177
fungal pneumonia in p47phox-/- mice (32), and avoids surgery. Briefly, mice were 7
178
anesthetized by isofluorane inhalation using an approved chamber. Following
179
anesthesia, mice were suspended by their upper incisors from a suture thread on a
180
90° incline board. The tongue was gently extended, and a liquid volume (50 μl) was
181
delivered into the distal part of the oropharynx. With the tongue extended, the animal
182
was unable to swallow, and the liquid volume was aspirated into the lower respiratory
183
tract. Just prior to liquid delivery, the chest was gently compressed and then released
184
just after deposition of liquid into the oropharynx to enhance aspiration of the liquid
185
into the lung. Mice recovered within 5 minutes of the procedure, and were observed
186
until they resumed normal activity.
187 188 189
Histopathology After sacrifice, mouse lungs were infused with 10% neutral buffered formalin
190
via the trachea. Paraffin-embedded blocks were prepared and sections were stained
191
with Hematoxylin and Eosin (H&E) to assess inflammation and Grocott-Gomori
192
methenamine-silver stain (GMS) to visualize fungi. Tissues were microscopically
193
examined for pulmonary injury, vascular invasion, and structural changes in
194
Aspergillus hyphae. All slides were analyzed using 40× magnification without formal
195
morphometric analysis, and blinded to genotype. The percentage of lung involved by
196
inflammation was scored in each mouse as follows: 0%, 5%, 10%, and then by 10%
197
increments (e.g., 20%, 30%, 40%, etc.). The predominant inflammatory cell type was
198
scored.
199 200 201
Immunostaining of mouse tissue For immunostaining, specimens were processed similarly as described
202
previously (33). Briefly, samples were deparaffinized, rehydrated in decreasing
203
concentrations of EtOH, and subjected to antigen retrieval by cooking in 10 mM 8
204
citrate buffer, pH 6.0, for 10 min. Specimens were blocked with 2% BSA and 36 μl/ml
205
mouse Ig blocking reagent (Vector Laboratories) in PBS/0.1% Triton for 1 h at room
206
temperature. For visualization of NETs, primary antibodies directed against
207
myeloperoxidase (A 0398, Dako) and histone H1 (clone AE-4, Acris Antibodies)
208
diluted in blocking solution were applied over night at 4°C. For detection of apoptotic
209
neutrophils, specimens were blocked with 2%BSA, PBS/0.1% Triton, and primary
210
antibodies directed against cleaved caspase-3 (#9661, Cell Signaling Technology)
211
and the Ly-6B.2 alloantigen (clone7/4, AbD Serotec) were used. Primary antibodies
212
were detected with Alexa Fluor 568- and 488-conjugated secondary antibodies (Life
213
Technologies) diluted in 2% BSA in PBS/0.1% Triton, respectively. DNA was
214
visualized with 4',6-diamidino-2-phenylindole (DAPI) (Life Technologies) and slides
215
were mounted with fluorescence mounting medium (Dako). An Eclipse C1 plus
216
confocal microscope (Nikon Instruments) using a 100× oil immersion objective was
217
used for image acquisition. Wavelengths of 405 nm (diode), 488 nm (Argon), and 543
218
nm (HeNe) were used to excite DAPI, Alexa Fluor 488 (and transmission images
219
(trans)), and Alexa Fluor 568, respectively. Images were captured in separate passes
220
to avoid cross talk and are presented as maximum intensity projections from Z-stacks
221
if not otherwise stated. All images were adjusted for background fluorescence and
222
enhanced slightly for signal intensity in NIS elements software (Nikon Instruments).
223
Slides were analyzed for NET formation blinded to genotype.
224
For quantification analyses, fluorescence images were captured with an
225
Eclipse 90i microscope (Nikon Instruments) and a Hamamatsu Orca-ER digital
226
camera at 60× magnification as Z-stacks covering the complete specimen depth. In
227
overlay images of all three channels captured (DAPI, Alexa Fluor 488 and 568), the
228
proportion of neutrophils with the following immunostaining phenotypes was
229
quantified by manually scrolling through Z-planes: (i) decondensed nucleus; (ii) 9
230
cleaved capsase-3-positive; (iii) decondensed nucleus and cleaved capsase-3-
231
positive; and (iv) decondensed nucleus and cleaved capsase-3-negative (a marker of
232
early stage NETosis). Neutrophils were counted manually from at least two digitized
233
images per site (bronchioles and alveoli) and animal (n=3) in NIS elements software
234
(Nikon Instruments) covering between 770 and 3600 cells per site and animal (total:
235
between 3500 and 6800 cells per genotype and site). Results are expressed as
236
number of positive cells/1000 evaluated neutrophils.
237 238
Analysis of neutrophil NET generation and apoptosis following hyphal
239
stimulation ex vivo
240
Bone marrow neutrophils from WT and p47phox-/- mice were purified by density
241
centrifugation as described (34). The purity of neutrophils is 85% based on
242
cytology. Neutrophils (1x106 cells) were seeded onto glass cover slips (22x22 mm) in
243
6-well cell culture plates in 500μL RPMI and incubated in a 5% CO2 incubator at 37°C
244
for 1 h. Neutrophils were then exposed to A. fumigatus hyphae (~750 CFU/50 μl).
245
Following either 2 or 6h, cells were fixed with 2% paraformaldehyde overnight at 4°C.
246
Cells were gently washed in PBS, then incubated for 1 h with anti-Ly6G (clone RB6-
247
8C5, eBioscience; 1 μg/mL), then permeabilized for 5 min in permeabilization buffer
248
(eBioscience), followed by 1 h incubation with cleaved caspase-3 (clone D175, Cell
249
Signaling; 1 μg/mL). After washing, cells were incubated with TRITC-conjugated anti-
250
rat IgG (eBioscience) and Cy 2-conjugated anti-rabbit IgG (Jackson
251
ImmunoResearch) secondary antibodies. Cover slips were washed and mounted on
252
to slides using mounting media with DAPI (Vectashield, 1.5 μg/mL). Fluorescence
253
images were obtained from a TCS SP2 AOBS Spectral Confocal Scanner mounted
254
on a Leica DM IRE2 inverted fluorescent microscope using a 63x oil immersion
255
objective. Wavelengths of 405 nm (diode), 488 nm (Argon), and 543 nm (HeNe) were 10
256
used to excite DAPI (+ bright field image (BF)), FITC, and TRITC, respectively. Prism
257
spectrophotometric filters ranging from 415-460, 495-550, and 570-640 nm were
258
used to collect fluorescence emission of DAPI, FITC, and TRITC respectively in
259
sequential acquisition mode. Leica Confocal Software was used to optimize laser
260
settings for saturation and eliminate cross talk. Imaging fields were chosen at random
261
and Z-sections were optimized for number of cells. Cells were assessed by
262
overlaying DAPI stain with BF to assess if a cell was intact or fragmented. Intact cells
263
were quantified as neutrophils if they were both Ly6G+ and contained nuclear
264
material by overlaying DAPI and TRITC stains. Three hundred (300) neutrophils per
265
sample were counted. Nuclei of these cells were determined to be decondensed or
266
hypersegmented by comparison with unstimulated neutrophils, using the same
267
criteria as we used in in vivo studies. These cells were then assessed for caspase-3
268
activity by overlaying DAPI and FITC stains. The major endpoints evaluated were:
269
presence of NETosis (yes or no) and proportion of neutrophils with decondensed
270
nuclei and cleaved caspase-3-negative per 300 evaluated neutrophils.
271 272
Statistical analysis
273
Comparison of the proportion of neutrophils with specific immunostaining
274
phenotypes between genotypes was assessed by chi square test. Percent lung
275
inflammation over time was compared between genotypes by 2-way ANOVA with
276
Bonferroni post-test. Analysis was performed using Prism 5.0 software (GraphPad
277
Software). Differences were considered significant for a two-sided P value of p