Accepted Manuscript Activation of the NLRP3 inflammasome by cellular labile iron Kyohei Nakamura, Toru Kawakami, Naoki Yamamoto, Miyu Tomizawa, Tohru Fujiwara, Tomonori Ishii, Hideo Harigae, Kouetsu Ogasawara PII:
S0301-472X(15)00730-4
DOI:
10.1016/j.exphem.2015.11.002
Reference:
EXPHEM 3329
To appear in:
Experimental Hematology
Received Date: 10 April 2015 Revised Date:
28 October 2015
Accepted Date: 1 November 2015
Please cite this article as: Nakamura K, Kawakami T, Yamamoto N, Tomizawa M, Fujiwara T, Ishii T, Harigae H, Ogasawara K, Activation of the NLRP3 inflammasome by cellular labile iron, Experimental Hematology (2015), doi: 10.1016/j.exphem.2015.11.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
15.51.Exp Hematology.Nakamura K et al. 1
ACCEPTED MANUSCRIPT
Immunobiology and immunotherapy
2
Title: Activation of the NLRP3 inflammasome by cellular labile iron.
3
Kyohei Nakamura1,2, Toru Kawakami1,3, Naoki Yamamoto1, Miyu Tomizawa1,
4
Tohru Fujiwara2, Tomonori Ishii2, Hideo Harigae2, and Kouetsu Ogasawara*1
5
1 Department of Immunobiology, Institute of Development, Aging, and Cancer,
6
Tohoku University, Sendai, Japan
7
2 Department of Hematology and Rheumatology, Tohoku University Graduate
8
School of Medicine, Sendai, Japan
9
3 Department of Thoracic Surgery, Institute of Development, Aging and Cancer,
TE D
M AN U
SC
RI PT
1
Tohoku University, Sendai, Japan
11
Address corresponding to K.O; Department of Immunobiology, Institute of
12
Development, Aging and Cancer, Tohoku University, Sendai, 980-8575, Japan;
13
Phone: 81-22-717-8579, Fax: 81-22-717-8452, Email: [email protected]
14
200 words in Abstract, 1892 words in Text, 906 words in Method, 41 References, 5
15
Figures
16
AC C
EP
10
15.51.Exp Hematology.Nakamura K et al. 2
ACCEPTED MANUSCRIPT 17
Abstract
18
Cellular labile iron, which contains chelatable redox-active Fe2+, has been implicated
19
in iron-mediated cellular toxicity leading to multiple organ dysfunction.
20
homeostasis is controlled by monocytes/macrophages through their iron recycling
21
and storage capacities. Furthermore, iron sequestration by monocytes/macrophages
22
is regulated by pro-inflammatory cytokines including IL-1, highlighting the importance
23
of these cells in the crosstalk between inflammation and iron homeostasis. However,
24
a role for cellular labile iron in monocytes/macrophages-mediated inflammatory
25
responses has not been defined. Here we show that cellular labile iron activates the
26
NLRP3 inflammasome in human monocytes. Stimulation of LPS-primed PBMCs with
27
ferric ammonium citrates (FAC) increases the level of cellular Fe2+ levels in
28
monocytes and induces production of IL-1β in a dose-dependent manner. This FAC-
29
induced IL-1β production is caspase-1-dependent and is significantly inhibited by an
30
Fe2+-specific chelator. FAC consistently induced IL-1β secretion in THP1 cells but
31
not in NLRP3-deficient THP1 cells indicating a requirement for the NLRP3
32
inflammasome. Additionally, the inflammasome activation is mediated by potassium
33
efflux,
34
permeabilization. Thus, these results suggest that monocytes/macrophages not only
35
sequestrate iron during inflammation, but also mediate inflammation in response to
36
cellular labile iron, which provides novel insights into the role of iron in chronic
37
inflammation.
38
AC C
EP
TE D
M AN U
SC
RI PT
Iron
ROS-mediated
mitochondrial
dysfunction,
and
lysosomal
membrane
15.51.Exp Hematology.Nakamura K et al. 3
ACCEPTED MANUSCRIPT Introduction
40
Iron is the major transition metal in the body and plays essential roles in synthesis of
41
DNA, mitochondrial ATP, and heme1. Under physiological conditions, the majority of
42
circulating iron exists in complex with transferrin, which maintains iron in a soluble
43
and non-toxic form. The transferrin-iron complex is taken up by cells via receptor-
44
mediated endocytosis and is then utilized for biological processes including
45
hematopoiesis.
46
overwhelmed, non-transferrin bound iron (NTBI), which is bound mainly by citrate,
47
appears in the plasma such as is seen in patients with iron overload conditions
48
including hemochromatosis, thalassemia, and myelodysplasia2. NTBI is rapidly taken
49
up by cells through unknown mechanisms and is subsequently incorporated into the
50
chelatable, redox-active iron pools called labile iron pools, leading to generation of
51
reactive oxygen species (ROS) through the Fenton reaction3. The oxidative stress
52
generated by cellular labile iron is thought to play a central role in the pathogenesis
53
of organ dysfunction associated with iron overload, including that of the heart, liver,
54
and endocrine systems.
55
responses remain unclear.
SC
RI PT
39
EP
TE D
M AN U
However, when the iron-binding capacity of transferrin is
AC C
However, the effects of cellular labile iron on immune
56
Monocytes/macrophages play a crucial role in systemic iron homeostasis where
57
they phagocytize and degrade damaged or senescent erythrocytes and recycle
58
heme-associated-iron4.
59
monocytes/macrophages5 or released into plasma via the iron export protein,
60
ferroportin6. Under inflammatory conditions, pro-inflammatory cytokines including IL-
The recycled iron can either be stored as ferritin in
15.51.Exp Hematology.Nakamura K et al. 4
ACCEPTED MANUSCRIPT
6 and IL-1 stimulate production of hepcidin from hepatocytes, which strongly inhibits
62
the expression of ferroportin on monocytes/macrophages7 leading to iron retention
63
within these cells. Thus, the hepcidin-ferroportin axis is responsible for anemia of
64
chronic inflammation8.
RI PT
61
The pro-inflammatory cytokine IL-1β is predominantly secreted from monocytes
66
/macrophages through a cytosolic protein complex called the inflammasome9. The
67
NOD-like receptor family, pyrin domain containing 3 (NLRP3) protein, is the best
68
characterized subtype, and plays a pivotal role in host defense against infection as
69
well
70
atherosclerosis11, diabetes12, and Alzheimer’s13.
71
assembled in response to various danger signals, leading to activation of caspase-1,
72
which, in turn, catalytically cleaves the inactive IL-1β precursor protein (pro-IL-1β)
73
into biologically-active IL-1β
74
activation of the NLRP3 inflammasome, including cytosolic potassium efflux14,
75
lysosomal damage15, and ROS-dependent and -independent mitochondrial
76
dysfunction16,17.′ Given that excess cellular labile iron could be toxic for many
77
mammalian cells due to its redox-activity, this may be recognized as a danger signal
78
in monocytes/macrophages. In this study we addressed whether cellular labile iron
79
activates the inflammasome.
80
in
a
variety
of
sterile
inflammatory
diseases
including
gout10,
The NLRP3 inflammasome is
TE D
as
M AN U
SC
65
AC C
EP
Several mechanisms have been proposed for
15.51.Exp Hematology.Nakamura K et al. 5
ACCEPTED MANUSCRIPT Results
82
Cellular labile iron induces IL-1β production in human monocytes.
83
Non-transferrin bound Fe3+ can rapidly enter cells and increases intracellular
84
chelatable iron, known as labile iron pools, which contain redox-active Fe2+. We first
85
investigated whether excess cellular labile iron induces IL-1β production. To this end,
86
we treated LPS-primed PBMCs with ferric ammonium citrate (FAC) as a source of
87
non-transferrin bound Fe3+, and evaluated the levels of IL-1β in culture supernatants.
88
FAC induced IL-1β production in a dose-dependent manner in LPS-primed PBMCs
89
(Figure 1A) and in LPS-primed monocytes (Figure 1B). FAC also decreased cell
90
viability of monocytes in a dose dependent manner (Figure 1C). To evaluate the
91
labile Fe2+ pools within FAC-treated monocytes, we performed calcein-acetoxymethyl
92
ester (calcein-AM) experiments18. We observed that FAC increased the amount of
93
labile Fe2+ within monocytes in a dose-dependent manner (Figure 1D). Furthermore,
94
FAC-induced IL-1β production was inhibited not only by a Fe3+-specific chelator,
95
deferiprone (L1), but also by a Fe2+-specific chelator, 2,2'-bipyridyl (BIP) (Figure 1E),
96
whereas these chelators did not inhibit the IL-1β production by nigericin (Figure 1F).
97
Taken together, cellular Fe2+, namely cellular labile iron, is involved in FAC-induced
98
IL-1β production.
99
Cellular labile iron activates the NLRP3 inflammasome.
100
AC C
EP
TE D
M AN U
SC
RI PT
81
The activation of caspase-1, formally known as IL-1-converting enzyme, is required
15.51.Exp Hematology.Nakamura K et al. 6
ACCEPTED MANUSCRIPT
for processing of pro-IL-1β into the mature form of IL-1β19. As shown in Figure 2A,
102
IL-1β production is significantly inhibited by the pan-caspase inhibitor Z-VAD-FMK.
103
In addition, IL-1β production is completely abrogated by the Z-WEHD-FMK caspase-
104
1 inhibitor, indicating that caspase-1 is required for FAC-induced IL-1β production. In
105
human monocytes, active caspase-1 constitutively exists, which allows monocytes to
106
produce IL-1β in response to priming signal alone20. Consistently, we found that LPS
107
alone stimulated the production of IL-1β in a caspase-1-dependent manner, whereas
108
FAC markedly augmented IL-1β production (Figure 1A, 1B and 2A). To confirm the
109
involvement of the inflammasome-mediated caspase-1 activation, we next performed
110
western blot analysis. As shown in Figure 2B, the mature form of IL-1β and activated
111
form of caspase-1 (p10 subunit) were detectable in culture supernatant, indicating
112
that inflammasome-mediated caspase-1 activation occurs in response to cellular iron.
113
FAC stimulation alone failed to increase the mRNA of IL-1B, suggesting that it chiefly
114
act as an activation signal, but not priming signal (Figure 2C).
TE D
M AN U
SC
RI PT
101
Of the members of the inflammasome family, the NLRP3 inflammasome in
116
particular is known to be activated in response to various sterile stimuli21. Thus, we
117
hypothesized that cellular labile iron specifically activates the NLRP3 inflammasome.
118
To this end, we incubated LPS-primed PBMCs with the NLRP3 inflammasome
119
inhibitor glyburide22 and found significant inhibition of FAC-induced IL-1β production
120
(Figure 3A).
121
activation14,23 and as expected, FAC-induced IL-1β production was completely
122
abrogated in high potassium media (Figure 3A). To further define the involvement of
AC C
EP
115
Potassium efflux is a common trigger of NLRP3 inflammasome
15.51.Exp Hematology.Nakamura K et al. 7
ACCEPTED MANUSCRIPT
NLRP3, we used PMA-differentiated macrophage-like THP1 cells and THP1 cells
124
deficient for NLRP3 (THP1-defNLRP3 cells) (Figure3B). As shown in Figure 3C,
125
FAC treatment increased the level of cellular labile Fe2+ pools in both control-
126
transfected THP1 cells and THP1-defNLRP3 cells; however, FAC-induced IL-1β
127
production was found in PMA-differentiated THP1 cells, but not in THP1-defNLRP3
128
cells (Figure 3D).
129
activates the NLRP3 inflammasome.
130
Cellular labile iron induces ROS production and mitochondrial dysfunction
131
leading to NLRP3 inflammasome activation.
132
Several studies have shown that ROS-dependent and -independent mitochondrial
133
dysfunction activates the NLRP3 inflammasome16,17. Because cellular labile iron is
134
known to catalyze ROS production3, it is possible that ROS-mediated mitochondrial
135
dysfunction is involved in NLRP3 inflammasome activation mediated by cellular labile
136
iron. To address this possibility, we evaluated the ROS generation in FAC-treated
137
monocytes.
138
dependent manner (Figure 4A and 4B).
139
percentage of monocytes with decreased mitochondrial membrane potential (Figure
140
4C and 4D).
141
FAC-induced IL-1β production (Figure 4E)′ and mitochondrial dysfunction (Figure
142
4F).
143
involved in the activation of the NLRP3 inflammasome by cellular labile iron.
RI PT
123
EP
TE D
M AN U
SC
Taken together, these results indicated that cellular labile iron
AC C
We found that FAC treatment increased the ROS level in a doseSimilarly, FAC treatment increased the
Furthermore, pretreatment with N-acetyl-L-cysteine (NAC) inhibited
Collectively, we show that ROS-dependent mitochondrial dysfunction is
15.51.Exp Hematology.Nakamura K et al. 8
ACCEPTED MANUSCRIPT
ROS-mediated lysosomal membrane permeabilization is involved in cellular
145
labile iron-mediated activation of the NLRP3 inflammasome
146
The NLRP3 inflammasome has been reported to be activated by lysosomal damage
147
following phagocytosis of crystalline and particulate materials11,24. ′ Thus, we
148
investigated the involvement of phagocytosis in FAC-induced IL-1β production. We
149
found that cytochalasine D (CytoD), an inhibitor of actin polymerization, failed to
150
inhibit IL-1β production from FAC-treated LPS-primed PBMCs (Figure 5A).
151
Furthermore, CytoD did not inhibit the increase in cellular Fe2+ in FAC-treated
152
monocytes (Figure 5B). These results indicate that actin-mediated phagocytosis is
153
not required for the generation of cellular Fe2+ pools or NLRP3 inflammasome
154
activation. In contrast, FAC-induced IL-1β production was significantly inhibited by
155
the cathepsin B-specific inhibitor, CA-074ME (Figure 5A), suggesting the involvement
156
of phagocytosis-independent lysosomal damage. ROS is known to induce lysosomal
157
membrane permeabilization (LMP), leading to release of cathepsin B25,26. Thus, we
158
evaluated whether cellular labile iron induces LMP through ROS generation by using
159
PMA-differentiated THP1 cells stained with LysoTracker Red.
160
dramatically induced LMP in THP1 cells (seen as decreased fluorescence)
161
pretreatment with NAC partially attenuated FAC-induced LMP (Figure 5C).
162
results suggest that cellular labile iron activates the NLRP3 inflammasome through
163
ROS-dependent LMP and subsequent release of cathepsin B.
164
AC C
EP
TE D
M AN U
SC
RI PT
144
Although FAC
These
15.51.Exp Hematology.Nakamura K et al. 9
ACCEPTED MANUSCRIPT Discussion′ ′
166
In this study, we demonstrate that excess cellular labile iron activates the NLRP3
167
inflammasome leading to secretion of IL-1β in human monocytes. The generation of
168
ROS has been considered to be an upstream event for NLRP3 inflammasome
169
activation in response to a wide range of stimuli16,27. Moreover, cellular labile iron is
170
known to be a potent inducer of ROS through the Fenton reaction28. Consistently,
171
we found that cellular labile iron activates the NLRP3 inflammasome through ROS-
172
dependent mitochondrial dysfunction.
173
activated in response to lysosomal damage following phagocytosis of crystalline and
174
particulate materials including silica, alum, and cholesterol crystals11,24; however, we
175
found that phagocytosis was not required for activation of the NLRP3 inflammasome
176
by cellular iron. Instead, ROS-dependent LMP was involved in this process. Multiple
177
pathways have been implicated for NTBI uptake such as those including stimulator of
178
Fe transport (SFT), DMT-1, ZIP14, and the L-type Ca2+ channel29,30, which might
179
contribute to generation of cellular labile Fe2+ pools in monocytes. Since our results
180
are based on in vitro study, using PBMCs and THP1 cell line with chemical inhibitors,
181
there are limitations in our study. However, our results suggest a novel link between
182
iron and inflammation.
SC
RI PT
165
AC C
EP
TE D
M AN U
The NLRP3 inflammasome is reportedly
183
IL-1β together with IL-6 plays key roles in iron homeostasis during inflammation. In
184
response to IL-1β and IL-6, hepatocytes produce hepcidin, which downregulates the
185
iron exporter channel ferroportin on monocytes/macrophages leading to iron
186
retention within these cells7,8. The role of cellular iron during innate immune reposes
15.51.Exp Hematology.Nakamura K et al. 10
ACCEPTED MANUSCRIPT
has been studied using mice with a myeloid lineage-specific ferroportin deficiency.
188
These mice show iron accumulation in reticuloendothelial macrophages of liver,
189
spleen and bone marrow, and produce higher levels of TNF-α in response to LPS
190
than do control mice31. Conversely, low intracellular iron levels impair TRAM/TRIF-
191
dependent TLR4 signaling in murine macrophages32.
192
therapy
193
encephalomyelitis33 and animal models of rheumatoid arthritis34, highlighting the
194
importance of iron in inflammatory responses. However, the possibility that excess
195
cellular iron may be recognized by an intracellular danger sensor, namely, the
196
inflammasome, had not previously been addressed.
197
production of hepcidin from hepatocytes, iron-mediated NLRP3 inflammasome
198
activation, together with the hepcidin-ferroportin axis, might comprise a positive
199
feedback loop that augments inflammation. Further study is necessary to understand
200
the crosstalk between hepcidin-ferroportin axis and the NLRP3 inflammasome
201
activation by cellular iron.
202
Iron-mediated organ dysfunction is a main cause of the mortality and morbidity in
203
patients with thalassemia major, which causes iron overload due to increased
204
intestinal iron absorption and/or transfusion-dependency35.
205
levels in patients with thalassemia major are higher compared to asymptomatic
206
carriers and heathy controls36.
207
replacement therapy stimulates production of pro-inflammatory cytokines including
208
IL-1β in both healthy subjects and patients receiving hemodialysis37,38. Our results
disease
progression
in
Moreover, iron chelation
experimental
autoimmune
M AN U
SC
ameliorates
RI PT
187
AC C
EP
TE D
Given that IL-1β stimulates
Notably, serum IL-1β
Moreover, several studies have found that iron
15.51.Exp Hematology.Nakamura K et al. 11
ACCEPTED MANUSCRIPT
raise a possibility that NLRP3 inflammasome activation might be involved in the
210
production of pro-inflammatory cytokines in patients with iron-overload. However, it
211
should be noted that plasma NTBI level in patients with iron overload patients are
212
less than 20 µM39, which is considerably lower than the concentration we used in this
213
study. Given that iron deposition in reticuloendothelial system is commonly found in
214
patients with iron overload, NLRP3 inflammasome activation might occur in
215
reticuloendothelial cells that are exposed to locally high concentration of iron.
216
Because inflammasome
217
fibrosis40,41, it is imperative to investigate the role of inflammasome activation in iron-
218
mediated
219
endocrinopathies. Overall, activation of the NLRP3 inflammasome by cellular labile
220
iron might have broad implications for iron overload and chronic inflammation and
221
future study of this system is warranted.
dysfunction
M AN U
plays a critical role for inflammation-associated organ
including
AC C
EP
TE D
organ
SC
RI PT
209
liver
cirrhosis,
cardiomyopathy
and
15.51.Exp Hematology.Nakamura K et al. 12
ACCEPTED MANUSCRIPT Methods
223
Human blood samples
224
Human peripheral blood mononuclear cells (PBMCs) were isolated from healthy
225
donors with informed consent. The study was approved by the Ethical Committee of
226
Tohoku University Graduate School of Medicine.
227
Isolation and stimulation of mononuclear cells
228
The
229
SEPARATE-L (Muto chemical, Tokyo, Japan). Monocytes were purified by negative
230
selection using the EasySep™ Human Monocyte Enrichment Kit (STEMCELL
231
Technologies, Vancouver, Canada). These cells were cultured in RPMI1640 media
232
supplemented with 10% fetal bovine serum, 1 mM sodium pyrubate, 2 mM glutamine,
233
100 U/mL penicillin, and 100 µg/mL streptomycin. After 4 hr priming with 10 ng/ml
234
LPS (from Salmonella Minnesota R595, Enzo Life sciences, Farmingdale, NY), cells
235
were stimulated with ferric ammonium citrate (FAC) (Wako, Osaka, Japan) at the
236
indicated doses for 4 hr or nigericin(20µM, Invivogen, San Diego, CA) for 30 min.
237
For the inhibition assay, LPS-primed PBMCs were pretreated with the following
238
reagents for 30 min before stimulation: Fe3+ chelator deferiprone (400 µM; Sigma),
239
Fe2+-specific chelator 2,2'-bipyridine (400 µM; Sigma), pan caspase Inhibitor Z-VAD-
240
FMK, caspase1/ICE inhibitor Z-WEHD-FMK (100 µM; R&D Systems, Minneapolis,
241
MN), glyburide (200 µM; Sigma-Aldrich), N-acetyl-L-cysteine
was
SC
M AN U
fraction
isolated
by
density-gradient
centrifugation
using
AC C
EP
TE D
PBMC
RI PT
222
(5 mM or 25 mM;
15.51.Exp Hematology.Nakamura K et al. 13
ACCEPTED MANUSCRIPT
Wako), cytochalasin D (10 µM; Wako) and CA074-ME (40 nM; Enzo). The amount of
243
IL-1β in cell culture supernatants was measured with a Human IL-1β ELISA kit
244
(eBioscience, San Diego, CA) according to the manufacturer’s instructions.′ Cell
245
viability was measured by propidium iodide (2µg/ml, Sigma), using FACSCantoII (BD
246
biosciences, San Jose, CA).
RI PT
242
SC
247
Cell lines and transfectants
249
THP1 cells were cultured in RPMI1640 media supplemented with 10% fetal bovine
250
serum, 1 mM sodium pyrubate, 2 mM glutamine, 100 U/mL penicillin, and 100 µg/mL
251
streptomycin).
252
purchased from Invivogen. These cells were differentiated by phorbol 12-myristate
253
13-acetate (PMA) (5 ng/ml; Sigma-Aldrich) 48 h before stimulation on 12 well plate
254
(106 cells / well). After overnight priming with LPS (100 µg/ml), cells were stimulated
255
with the indicated doses of FAC for 6 h.
256
supernatants was measured.
EP
TE D
THP1 cells with reduced NLRP3 activity (THP1-defNLRP3) were
The amount of IL-1β in cell culture
AC C
257
M AN U
248
258
Immunoblot analysis
259
LPS (10 ng/mL)-primed PBMCs were stimulated with the indicated doses of FAC for
260
4 hr. As a positive control, LPS-primed PBMCs were stimulated with ATP (2.5mM,
15.51.Exp Hematology.Nakamura K et al. 14
ACCEPTED MANUSCRIPT
Sigma) for 30 min. Culture supernatants and total cell lysates were collected and
262
then clarified by centrifugation. Proteins in culture supernatants were precipitated
263
with Strataclean Resin (Stratagene, La Jolla, CA). Anti-human caspase-1 p10 Ab (C-
264
20, Santa Cruz), anti-human NLRP3 mAb (D2P5E, Cell signaling technology, Boston,
265
MA), anti-IL-1β mAb (D3U3EH, Cell signaling technology), and anti-α tubulin antibody
266
(DM1A, Santa Cruz) were used for detection.
267
Quantification of intracellular labile iron pools
268
Intracellular labile Fe2+ pools were quantitated using the iron-sensing fluorescent
269
probe, calcein-acetoxy-methyl-ester (calcein-AM) as previously reported18. Briefly,
270
FAC-loaded cells were stained with calcein-AM (250 nM, Dojin Chemical Co,Tokyo,
271
Japan) for 15 min. These cells were washed twice with PBS, and then treated with
272
or without 2,2′-bipyridyl (400 µM, Sigma) for 1 hr.
273
defined as the difference of MFI (FL-1 channel) in the presence or absence of 2,2’-
274
bipyridyl, using FACSCantoII.
275
Quantitative PCR
276
LPS-primed or unprimed PBMCs were stimulated with or without FAC (20mM) for 4hr.
277
Total RNA from PBMCs was extracted using the RNeasy lipid tissue kit (QIAGEN,
278
Hilden, Germany) after stimulation. Total RNA was reverse transcribed into cDNA
279
using SuperScript III with Oligo (dT)12–18 (Invitrogen, Carlsbad, CA, USA). RT-PCR
280
was performed in a DNA chromo 4 (Bio-Rad Laboratories, Hercules, CA, USA).
Intracellular Fe2+ levels were
AC C
EP
TE D
M AN U
SC
RI PT
261
15.51.Exp Hematology.Nakamura K et al. 15
ACCEPTED MANUSCRIPT 281
Following PCR primers were used: IL-1B: sense 5
282
TGA C-3
283
sense 5′-GGA GAG ACC TTT ATG AGA AAG CAA-3′, antisense 5′-GCT GTC TTC
284
CTG GCA TAT CAC A-3′; ACTB: sense 5′-ATT GCC GAC AGG ATG CAG AA -3′,
285
antisense 5′-GCT GAT CCA CAT CTG CTG GAA -3′. SYBR green (QIAGEN) was
286
used for quantification of the amplified DNA. Data were analyzed by Opticon Monitor
287
version 3 software (Bio-Rad laboratories), according to manufacturer’s instruction.
288
ROS detection assay
289
PBMCs (5×105) were stimulated with the indicated doses of FAC for 4 hr, and then
290
cells were stained with APC-conjugated anti-CD14 mAb (Biolegend). The level of
291
ROS in CD14-gated population was measured by total ROS detection kit (Enzo life
292
sciences) according to the manufacturer’s instruction. Mean fluorointensity (MFI) of
293
the FL-1 channel in the CD14-gated population was determined by FACSCantoII.
294
Mitochondrial membrane potential assay
295
To evaluate mitochondrial damage, JC-10 assay (Abcam, Cambridge, UK) was used
296
according to manufacturer’s instruction. Briefly, PBMCs (5×105) were treated with
297
different doses of FAC for 4 hr. After staining with APC-conjugated anti-CD14 mAb,
298
cells were stained with JC-10 dye loading solution for 30 min. The percentage of
299
CD14-positive cells with depolarized mitochondria (green fluorescence) was
300
determined by FACSCantoII.
-GTC GGA GAT TCG TAG CTG GAT-3
; NLRP3:
AC C
EP
TE D
M AN U
SC
RI PT
, antisense 5
-TTA CAG TGG CAA TGA GGA
15.51.Exp Hematology.Nakamura K et al. 16
ACCEPTED MANUSCRIPT Fluorescence microscopy′ ′ ′
302
To evaluate lysosomal membrane permeabilization, PMA-differentiated THP1 cells
303
were loaded with Lysotracker Red (200 nM; Invivogen) and Hoechst 33342 (1 µg/ml;
304
Dojin Chemical Co) for 15 min. After being washed twice with PBS, THP1 cells
305
pretreated with or without 25 mM NAC were incubated with various doses of FAC for
306
the indicated periods. Fluorescent images were acquired using an Olympus IX81
307
microscope (Olympus, Tokyo, Japan) with a 20× objective lens. Merged images
308
were obtained by using Lumina Vision software (Mitani Corporation, Fukui, Japan).
309
Statistical analysis
310
Statistical analysis was performed using the unpaired two-tailed Student's t-test: *P
S0301-472X(15)00730-4
DOI:
10.1016/j.exphem.2015.11.002
Reference:
EXPHEM 3329
To appear in:
Experimental Hematology
Received Date: 10 April 2015 Revised Date:
28 October 2015
Accepted Date: 1 November 2015
Please cite this article as: Nakamura K, Kawakami T, Yamamoto N, Tomizawa M, Fujiwara T, Ishii T, Harigae H, Ogasawara K, Activation of the NLRP3 inflammasome by cellular labile iron, Experimental Hematology (2015), doi: 10.1016/j.exphem.2015.11.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
15.51.Exp Hematology.Nakamura K et al. 1
ACCEPTED MANUSCRIPT
Immunobiology and immunotherapy
2
Title: Activation of the NLRP3 inflammasome by cellular labile iron.
3
Kyohei Nakamura1,2, Toru Kawakami1,3, Naoki Yamamoto1, Miyu Tomizawa1,
4
Tohru Fujiwara2, Tomonori Ishii2, Hideo Harigae2, and Kouetsu Ogasawara*1
5
1 Department of Immunobiology, Institute of Development, Aging, and Cancer,
6
Tohoku University, Sendai, Japan
7
2 Department of Hematology and Rheumatology, Tohoku University Graduate
8
School of Medicine, Sendai, Japan
9
3 Department of Thoracic Surgery, Institute of Development, Aging and Cancer,
TE D
M AN U
SC
RI PT
1
Tohoku University, Sendai, Japan
11
Address corresponding to K.O; Department of Immunobiology, Institute of
12
Development, Aging and Cancer, Tohoku University, Sendai, 980-8575, Japan;
13
Phone: 81-22-717-8579, Fax: 81-22-717-8452, Email: [email protected]
14
200 words in Abstract, 1892 words in Text, 906 words in Method, 41 References, 5
15
Figures
16
AC C
EP
10
15.51.Exp Hematology.Nakamura K et al. 2
ACCEPTED MANUSCRIPT 17
Abstract
18
Cellular labile iron, which contains chelatable redox-active Fe2+, has been implicated
19
in iron-mediated cellular toxicity leading to multiple organ dysfunction.
20
homeostasis is controlled by monocytes/macrophages through their iron recycling
21
and storage capacities. Furthermore, iron sequestration by monocytes/macrophages
22
is regulated by pro-inflammatory cytokines including IL-1, highlighting the importance
23
of these cells in the crosstalk between inflammation and iron homeostasis. However,
24
a role for cellular labile iron in monocytes/macrophages-mediated inflammatory
25
responses has not been defined. Here we show that cellular labile iron activates the
26
NLRP3 inflammasome in human monocytes. Stimulation of LPS-primed PBMCs with
27
ferric ammonium citrates (FAC) increases the level of cellular Fe2+ levels in
28
monocytes and induces production of IL-1β in a dose-dependent manner. This FAC-
29
induced IL-1β production is caspase-1-dependent and is significantly inhibited by an
30
Fe2+-specific chelator. FAC consistently induced IL-1β secretion in THP1 cells but
31
not in NLRP3-deficient THP1 cells indicating a requirement for the NLRP3
32
inflammasome. Additionally, the inflammasome activation is mediated by potassium
33
efflux,
34
permeabilization. Thus, these results suggest that monocytes/macrophages not only
35
sequestrate iron during inflammation, but also mediate inflammation in response to
36
cellular labile iron, which provides novel insights into the role of iron in chronic
37
inflammation.
38
AC C
EP
TE D
M AN U
SC
RI PT
Iron
ROS-mediated
mitochondrial
dysfunction,
and
lysosomal
membrane
15.51.Exp Hematology.Nakamura K et al. 3
ACCEPTED MANUSCRIPT Introduction
40
Iron is the major transition metal in the body and plays essential roles in synthesis of
41
DNA, mitochondrial ATP, and heme1. Under physiological conditions, the majority of
42
circulating iron exists in complex with transferrin, which maintains iron in a soluble
43
and non-toxic form. The transferrin-iron complex is taken up by cells via receptor-
44
mediated endocytosis and is then utilized for biological processes including
45
hematopoiesis.
46
overwhelmed, non-transferrin bound iron (NTBI), which is bound mainly by citrate,
47
appears in the plasma such as is seen in patients with iron overload conditions
48
including hemochromatosis, thalassemia, and myelodysplasia2. NTBI is rapidly taken
49
up by cells through unknown mechanisms and is subsequently incorporated into the
50
chelatable, redox-active iron pools called labile iron pools, leading to generation of
51
reactive oxygen species (ROS) through the Fenton reaction3. The oxidative stress
52
generated by cellular labile iron is thought to play a central role in the pathogenesis
53
of organ dysfunction associated with iron overload, including that of the heart, liver,
54
and endocrine systems.
55
responses remain unclear.
SC
RI PT
39
EP
TE D
M AN U
However, when the iron-binding capacity of transferrin is
AC C
However, the effects of cellular labile iron on immune
56
Monocytes/macrophages play a crucial role in systemic iron homeostasis where
57
they phagocytize and degrade damaged or senescent erythrocytes and recycle
58
heme-associated-iron4.
59
monocytes/macrophages5 or released into plasma via the iron export protein,
60
ferroportin6. Under inflammatory conditions, pro-inflammatory cytokines including IL-
The recycled iron can either be stored as ferritin in
15.51.Exp Hematology.Nakamura K et al. 4
ACCEPTED MANUSCRIPT
6 and IL-1 stimulate production of hepcidin from hepatocytes, which strongly inhibits
62
the expression of ferroportin on monocytes/macrophages7 leading to iron retention
63
within these cells. Thus, the hepcidin-ferroportin axis is responsible for anemia of
64
chronic inflammation8.
RI PT
61
The pro-inflammatory cytokine IL-1β is predominantly secreted from monocytes
66
/macrophages through a cytosolic protein complex called the inflammasome9. The
67
NOD-like receptor family, pyrin domain containing 3 (NLRP3) protein, is the best
68
characterized subtype, and plays a pivotal role in host defense against infection as
69
well
70
atherosclerosis11, diabetes12, and Alzheimer’s13.
71
assembled in response to various danger signals, leading to activation of caspase-1,
72
which, in turn, catalytically cleaves the inactive IL-1β precursor protein (pro-IL-1β)
73
into biologically-active IL-1β
74
activation of the NLRP3 inflammasome, including cytosolic potassium efflux14,
75
lysosomal damage15, and ROS-dependent and -independent mitochondrial
76
dysfunction16,17.′ Given that excess cellular labile iron could be toxic for many
77
mammalian cells due to its redox-activity, this may be recognized as a danger signal
78
in monocytes/macrophages. In this study we addressed whether cellular labile iron
79
activates the inflammasome.
80
in
a
variety
of
sterile
inflammatory
diseases
including
gout10,
The NLRP3 inflammasome is
TE D
as
M AN U
SC
65
AC C
EP
Several mechanisms have been proposed for
15.51.Exp Hematology.Nakamura K et al. 5
ACCEPTED MANUSCRIPT Results
82
Cellular labile iron induces IL-1β production in human monocytes.
83
Non-transferrin bound Fe3+ can rapidly enter cells and increases intracellular
84
chelatable iron, known as labile iron pools, which contain redox-active Fe2+. We first
85
investigated whether excess cellular labile iron induces IL-1β production. To this end,
86
we treated LPS-primed PBMCs with ferric ammonium citrate (FAC) as a source of
87
non-transferrin bound Fe3+, and evaluated the levels of IL-1β in culture supernatants.
88
FAC induced IL-1β production in a dose-dependent manner in LPS-primed PBMCs
89
(Figure 1A) and in LPS-primed monocytes (Figure 1B). FAC also decreased cell
90
viability of monocytes in a dose dependent manner (Figure 1C). To evaluate the
91
labile Fe2+ pools within FAC-treated monocytes, we performed calcein-acetoxymethyl
92
ester (calcein-AM) experiments18. We observed that FAC increased the amount of
93
labile Fe2+ within monocytes in a dose-dependent manner (Figure 1D). Furthermore,
94
FAC-induced IL-1β production was inhibited not only by a Fe3+-specific chelator,
95
deferiprone (L1), but also by a Fe2+-specific chelator, 2,2'-bipyridyl (BIP) (Figure 1E),
96
whereas these chelators did not inhibit the IL-1β production by nigericin (Figure 1F).
97
Taken together, cellular Fe2+, namely cellular labile iron, is involved in FAC-induced
98
IL-1β production.
99
Cellular labile iron activates the NLRP3 inflammasome.
100
AC C
EP
TE D
M AN U
SC
RI PT
81
The activation of caspase-1, formally known as IL-1-converting enzyme, is required
15.51.Exp Hematology.Nakamura K et al. 6
ACCEPTED MANUSCRIPT
for processing of pro-IL-1β into the mature form of IL-1β19. As shown in Figure 2A,
102
IL-1β production is significantly inhibited by the pan-caspase inhibitor Z-VAD-FMK.
103
In addition, IL-1β production is completely abrogated by the Z-WEHD-FMK caspase-
104
1 inhibitor, indicating that caspase-1 is required for FAC-induced IL-1β production. In
105
human monocytes, active caspase-1 constitutively exists, which allows monocytes to
106
produce IL-1β in response to priming signal alone20. Consistently, we found that LPS
107
alone stimulated the production of IL-1β in a caspase-1-dependent manner, whereas
108
FAC markedly augmented IL-1β production (Figure 1A, 1B and 2A). To confirm the
109
involvement of the inflammasome-mediated caspase-1 activation, we next performed
110
western blot analysis. As shown in Figure 2B, the mature form of IL-1β and activated
111
form of caspase-1 (p10 subunit) were detectable in culture supernatant, indicating
112
that inflammasome-mediated caspase-1 activation occurs in response to cellular iron.
113
FAC stimulation alone failed to increase the mRNA of IL-1B, suggesting that it chiefly
114
act as an activation signal, but not priming signal (Figure 2C).
TE D
M AN U
SC
RI PT
101
Of the members of the inflammasome family, the NLRP3 inflammasome in
116
particular is known to be activated in response to various sterile stimuli21. Thus, we
117
hypothesized that cellular labile iron specifically activates the NLRP3 inflammasome.
118
To this end, we incubated LPS-primed PBMCs with the NLRP3 inflammasome
119
inhibitor glyburide22 and found significant inhibition of FAC-induced IL-1β production
120
(Figure 3A).
121
activation14,23 and as expected, FAC-induced IL-1β production was completely
122
abrogated in high potassium media (Figure 3A). To further define the involvement of
AC C
EP
115
Potassium efflux is a common trigger of NLRP3 inflammasome
15.51.Exp Hematology.Nakamura K et al. 7
ACCEPTED MANUSCRIPT
NLRP3, we used PMA-differentiated macrophage-like THP1 cells and THP1 cells
124
deficient for NLRP3 (THP1-defNLRP3 cells) (Figure3B). As shown in Figure 3C,
125
FAC treatment increased the level of cellular labile Fe2+ pools in both control-
126
transfected THP1 cells and THP1-defNLRP3 cells; however, FAC-induced IL-1β
127
production was found in PMA-differentiated THP1 cells, but not in THP1-defNLRP3
128
cells (Figure 3D).
129
activates the NLRP3 inflammasome.
130
Cellular labile iron induces ROS production and mitochondrial dysfunction
131
leading to NLRP3 inflammasome activation.
132
Several studies have shown that ROS-dependent and -independent mitochondrial
133
dysfunction activates the NLRP3 inflammasome16,17. Because cellular labile iron is
134
known to catalyze ROS production3, it is possible that ROS-mediated mitochondrial
135
dysfunction is involved in NLRP3 inflammasome activation mediated by cellular labile
136
iron. To address this possibility, we evaluated the ROS generation in FAC-treated
137
monocytes.
138
dependent manner (Figure 4A and 4B).
139
percentage of monocytes with decreased mitochondrial membrane potential (Figure
140
4C and 4D).
141
FAC-induced IL-1β production (Figure 4E)′ and mitochondrial dysfunction (Figure
142
4F).
143
involved in the activation of the NLRP3 inflammasome by cellular labile iron.
RI PT
123
EP
TE D
M AN U
SC
Taken together, these results indicated that cellular labile iron
AC C
We found that FAC treatment increased the ROS level in a doseSimilarly, FAC treatment increased the
Furthermore, pretreatment with N-acetyl-L-cysteine (NAC) inhibited
Collectively, we show that ROS-dependent mitochondrial dysfunction is
15.51.Exp Hematology.Nakamura K et al. 8
ACCEPTED MANUSCRIPT
ROS-mediated lysosomal membrane permeabilization is involved in cellular
145
labile iron-mediated activation of the NLRP3 inflammasome
146
The NLRP3 inflammasome has been reported to be activated by lysosomal damage
147
following phagocytosis of crystalline and particulate materials11,24. ′ Thus, we
148
investigated the involvement of phagocytosis in FAC-induced IL-1β production. We
149
found that cytochalasine D (CytoD), an inhibitor of actin polymerization, failed to
150
inhibit IL-1β production from FAC-treated LPS-primed PBMCs (Figure 5A).
151
Furthermore, CytoD did not inhibit the increase in cellular Fe2+ in FAC-treated
152
monocytes (Figure 5B). These results indicate that actin-mediated phagocytosis is
153
not required for the generation of cellular Fe2+ pools or NLRP3 inflammasome
154
activation. In contrast, FAC-induced IL-1β production was significantly inhibited by
155
the cathepsin B-specific inhibitor, CA-074ME (Figure 5A), suggesting the involvement
156
of phagocytosis-independent lysosomal damage. ROS is known to induce lysosomal
157
membrane permeabilization (LMP), leading to release of cathepsin B25,26. Thus, we
158
evaluated whether cellular labile iron induces LMP through ROS generation by using
159
PMA-differentiated THP1 cells stained with LysoTracker Red.
160
dramatically induced LMP in THP1 cells (seen as decreased fluorescence)
161
pretreatment with NAC partially attenuated FAC-induced LMP (Figure 5C).
162
results suggest that cellular labile iron activates the NLRP3 inflammasome through
163
ROS-dependent LMP and subsequent release of cathepsin B.
164
AC C
EP
TE D
M AN U
SC
RI PT
144
Although FAC
These
15.51.Exp Hematology.Nakamura K et al. 9
ACCEPTED MANUSCRIPT Discussion′ ′
166
In this study, we demonstrate that excess cellular labile iron activates the NLRP3
167
inflammasome leading to secretion of IL-1β in human monocytes. The generation of
168
ROS has been considered to be an upstream event for NLRP3 inflammasome
169
activation in response to a wide range of stimuli16,27. Moreover, cellular labile iron is
170
known to be a potent inducer of ROS through the Fenton reaction28. Consistently,
171
we found that cellular labile iron activates the NLRP3 inflammasome through ROS-
172
dependent mitochondrial dysfunction.
173
activated in response to lysosomal damage following phagocytosis of crystalline and
174
particulate materials including silica, alum, and cholesterol crystals11,24; however, we
175
found that phagocytosis was not required for activation of the NLRP3 inflammasome
176
by cellular iron. Instead, ROS-dependent LMP was involved in this process. Multiple
177
pathways have been implicated for NTBI uptake such as those including stimulator of
178
Fe transport (SFT), DMT-1, ZIP14, and the L-type Ca2+ channel29,30, which might
179
contribute to generation of cellular labile Fe2+ pools in monocytes. Since our results
180
are based on in vitro study, using PBMCs and THP1 cell line with chemical inhibitors,
181
there are limitations in our study. However, our results suggest a novel link between
182
iron and inflammation.
SC
RI PT
165
AC C
EP
TE D
M AN U
The NLRP3 inflammasome is reportedly
183
IL-1β together with IL-6 plays key roles in iron homeostasis during inflammation. In
184
response to IL-1β and IL-6, hepatocytes produce hepcidin, which downregulates the
185
iron exporter channel ferroportin on monocytes/macrophages leading to iron
186
retention within these cells7,8. The role of cellular iron during innate immune reposes
15.51.Exp Hematology.Nakamura K et al. 10
ACCEPTED MANUSCRIPT
has been studied using mice with a myeloid lineage-specific ferroportin deficiency.
188
These mice show iron accumulation in reticuloendothelial macrophages of liver,
189
spleen and bone marrow, and produce higher levels of TNF-α in response to LPS
190
than do control mice31. Conversely, low intracellular iron levels impair TRAM/TRIF-
191
dependent TLR4 signaling in murine macrophages32.
192
therapy
193
encephalomyelitis33 and animal models of rheumatoid arthritis34, highlighting the
194
importance of iron in inflammatory responses. However, the possibility that excess
195
cellular iron may be recognized by an intracellular danger sensor, namely, the
196
inflammasome, had not previously been addressed.
197
production of hepcidin from hepatocytes, iron-mediated NLRP3 inflammasome
198
activation, together with the hepcidin-ferroportin axis, might comprise a positive
199
feedback loop that augments inflammation. Further study is necessary to understand
200
the crosstalk between hepcidin-ferroportin axis and the NLRP3 inflammasome
201
activation by cellular iron.
202
Iron-mediated organ dysfunction is a main cause of the mortality and morbidity in
203
patients with thalassemia major, which causes iron overload due to increased
204
intestinal iron absorption and/or transfusion-dependency35.
205
levels in patients with thalassemia major are higher compared to asymptomatic
206
carriers and heathy controls36.
207
replacement therapy stimulates production of pro-inflammatory cytokines including
208
IL-1β in both healthy subjects and patients receiving hemodialysis37,38. Our results
disease
progression
in
Moreover, iron chelation
experimental
autoimmune
M AN U
SC
ameliorates
RI PT
187
AC C
EP
TE D
Given that IL-1β stimulates
Notably, serum IL-1β
Moreover, several studies have found that iron
15.51.Exp Hematology.Nakamura K et al. 11
ACCEPTED MANUSCRIPT
raise a possibility that NLRP3 inflammasome activation might be involved in the
210
production of pro-inflammatory cytokines in patients with iron-overload. However, it
211
should be noted that plasma NTBI level in patients with iron overload patients are
212
less than 20 µM39, which is considerably lower than the concentration we used in this
213
study. Given that iron deposition in reticuloendothelial system is commonly found in
214
patients with iron overload, NLRP3 inflammasome activation might occur in
215
reticuloendothelial cells that are exposed to locally high concentration of iron.
216
Because inflammasome
217
fibrosis40,41, it is imperative to investigate the role of inflammasome activation in iron-
218
mediated
219
endocrinopathies. Overall, activation of the NLRP3 inflammasome by cellular labile
220
iron might have broad implications for iron overload and chronic inflammation and
221
future study of this system is warranted.
dysfunction
M AN U
plays a critical role for inflammation-associated organ
including
AC C
EP
TE D
organ
SC
RI PT
209
liver
cirrhosis,
cardiomyopathy
and
15.51.Exp Hematology.Nakamura K et al. 12
ACCEPTED MANUSCRIPT Methods
223
Human blood samples
224
Human peripheral blood mononuclear cells (PBMCs) were isolated from healthy
225
donors with informed consent. The study was approved by the Ethical Committee of
226
Tohoku University Graduate School of Medicine.
227
Isolation and stimulation of mononuclear cells
228
The
229
SEPARATE-L (Muto chemical, Tokyo, Japan). Monocytes were purified by negative
230
selection using the EasySep™ Human Monocyte Enrichment Kit (STEMCELL
231
Technologies, Vancouver, Canada). These cells were cultured in RPMI1640 media
232
supplemented with 10% fetal bovine serum, 1 mM sodium pyrubate, 2 mM glutamine,
233
100 U/mL penicillin, and 100 µg/mL streptomycin. After 4 hr priming with 10 ng/ml
234
LPS (from Salmonella Minnesota R595, Enzo Life sciences, Farmingdale, NY), cells
235
were stimulated with ferric ammonium citrate (FAC) (Wako, Osaka, Japan) at the
236
indicated doses for 4 hr or nigericin(20µM, Invivogen, San Diego, CA) for 30 min.
237
For the inhibition assay, LPS-primed PBMCs were pretreated with the following
238
reagents for 30 min before stimulation: Fe3+ chelator deferiprone (400 µM; Sigma),
239
Fe2+-specific chelator 2,2'-bipyridine (400 µM; Sigma), pan caspase Inhibitor Z-VAD-
240
FMK, caspase1/ICE inhibitor Z-WEHD-FMK (100 µM; R&D Systems, Minneapolis,
241
MN), glyburide (200 µM; Sigma-Aldrich), N-acetyl-L-cysteine
was
SC
M AN U
fraction
isolated
by
density-gradient
centrifugation
using
AC C
EP
TE D
PBMC
RI PT
222
(5 mM or 25 mM;
15.51.Exp Hematology.Nakamura K et al. 13
ACCEPTED MANUSCRIPT
Wako), cytochalasin D (10 µM; Wako) and CA074-ME (40 nM; Enzo). The amount of
243
IL-1β in cell culture supernatants was measured with a Human IL-1β ELISA kit
244
(eBioscience, San Diego, CA) according to the manufacturer’s instructions.′ Cell
245
viability was measured by propidium iodide (2µg/ml, Sigma), using FACSCantoII (BD
246
biosciences, San Jose, CA).
RI PT
242
SC
247
Cell lines and transfectants
249
THP1 cells were cultured in RPMI1640 media supplemented with 10% fetal bovine
250
serum, 1 mM sodium pyrubate, 2 mM glutamine, 100 U/mL penicillin, and 100 µg/mL
251
streptomycin).
252
purchased from Invivogen. These cells were differentiated by phorbol 12-myristate
253
13-acetate (PMA) (5 ng/ml; Sigma-Aldrich) 48 h before stimulation on 12 well plate
254
(106 cells / well). After overnight priming with LPS (100 µg/ml), cells were stimulated
255
with the indicated doses of FAC for 6 h.
256
supernatants was measured.
EP
TE D
THP1 cells with reduced NLRP3 activity (THP1-defNLRP3) were
The amount of IL-1β in cell culture
AC C
257
M AN U
248
258
Immunoblot analysis
259
LPS (10 ng/mL)-primed PBMCs were stimulated with the indicated doses of FAC for
260
4 hr. As a positive control, LPS-primed PBMCs were stimulated with ATP (2.5mM,
15.51.Exp Hematology.Nakamura K et al. 14
ACCEPTED MANUSCRIPT
Sigma) for 30 min. Culture supernatants and total cell lysates were collected and
262
then clarified by centrifugation. Proteins in culture supernatants were precipitated
263
with Strataclean Resin (Stratagene, La Jolla, CA). Anti-human caspase-1 p10 Ab (C-
264
20, Santa Cruz), anti-human NLRP3 mAb (D2P5E, Cell signaling technology, Boston,
265
MA), anti-IL-1β mAb (D3U3EH, Cell signaling technology), and anti-α tubulin antibody
266
(DM1A, Santa Cruz) were used for detection.
267
Quantification of intracellular labile iron pools
268
Intracellular labile Fe2+ pools were quantitated using the iron-sensing fluorescent
269
probe, calcein-acetoxy-methyl-ester (calcein-AM) as previously reported18. Briefly,
270
FAC-loaded cells were stained with calcein-AM (250 nM, Dojin Chemical Co,Tokyo,
271
Japan) for 15 min. These cells were washed twice with PBS, and then treated with
272
or without 2,2′-bipyridyl (400 µM, Sigma) for 1 hr.
273
defined as the difference of MFI (FL-1 channel) in the presence or absence of 2,2’-
274
bipyridyl, using FACSCantoII.
275
Quantitative PCR
276
LPS-primed or unprimed PBMCs were stimulated with or without FAC (20mM) for 4hr.
277
Total RNA from PBMCs was extracted using the RNeasy lipid tissue kit (QIAGEN,
278
Hilden, Germany) after stimulation. Total RNA was reverse transcribed into cDNA
279
using SuperScript III with Oligo (dT)12–18 (Invitrogen, Carlsbad, CA, USA). RT-PCR
280
was performed in a DNA chromo 4 (Bio-Rad Laboratories, Hercules, CA, USA).
Intracellular Fe2+ levels were
AC C
EP
TE D
M AN U
SC
RI PT
261
15.51.Exp Hematology.Nakamura K et al. 15
ACCEPTED MANUSCRIPT 281
Following PCR primers were used: IL-1B: sense 5
282
TGA C-3
283
sense 5′-GGA GAG ACC TTT ATG AGA AAG CAA-3′, antisense 5′-GCT GTC TTC
284
CTG GCA TAT CAC A-3′; ACTB: sense 5′-ATT GCC GAC AGG ATG CAG AA -3′,
285
antisense 5′-GCT GAT CCA CAT CTG CTG GAA -3′. SYBR green (QIAGEN) was
286
used for quantification of the amplified DNA. Data were analyzed by Opticon Monitor
287
version 3 software (Bio-Rad laboratories), according to manufacturer’s instruction.
288
ROS detection assay
289
PBMCs (5×105) were stimulated with the indicated doses of FAC for 4 hr, and then
290
cells were stained with APC-conjugated anti-CD14 mAb (Biolegend). The level of
291
ROS in CD14-gated population was measured by total ROS detection kit (Enzo life
292
sciences) according to the manufacturer’s instruction. Mean fluorointensity (MFI) of
293
the FL-1 channel in the CD14-gated population was determined by FACSCantoII.
294
Mitochondrial membrane potential assay
295
To evaluate mitochondrial damage, JC-10 assay (Abcam, Cambridge, UK) was used
296
according to manufacturer’s instruction. Briefly, PBMCs (5×105) were treated with
297
different doses of FAC for 4 hr. After staining with APC-conjugated anti-CD14 mAb,
298
cells were stained with JC-10 dye loading solution for 30 min. The percentage of
299
CD14-positive cells with depolarized mitochondria (green fluorescence) was
300
determined by FACSCantoII.
-GTC GGA GAT TCG TAG CTG GAT-3
; NLRP3:
AC C
EP
TE D
M AN U
SC
RI PT
, antisense 5
-TTA CAG TGG CAA TGA GGA
15.51.Exp Hematology.Nakamura K et al. 16
ACCEPTED MANUSCRIPT Fluorescence microscopy′ ′ ′
302
To evaluate lysosomal membrane permeabilization, PMA-differentiated THP1 cells
303
were loaded with Lysotracker Red (200 nM; Invivogen) and Hoechst 33342 (1 µg/ml;
304
Dojin Chemical Co) for 15 min. After being washed twice with PBS, THP1 cells
305
pretreated with or without 25 mM NAC were incubated with various doses of FAC for
306
the indicated periods. Fluorescent images were acquired using an Olympus IX81
307
microscope (Olympus, Tokyo, Japan) with a 20× objective lens. Merged images
308
were obtained by using Lumina Vision software (Mitani Corporation, Fukui, Japan).
309
Statistical analysis
310
Statistical analysis was performed using the unpaired two-tailed Student's t-test: *P
