Report

Autophagy Is Required for Neutrophil-Mediated

Inflammation
Graphical Abstract Authors
Abhisek Bhattacharya, Qin Wei, Jin Na
Shin, ..., Diana L. Bonilla, Qian Xiang, N.
Tony Eissa

Correspondence
teissa@bcm.edu

In Brief
Using several mouse models,
Bhattacharya et al. identify autophagy as
an important regulator of neutrophil
degranulation and ROS production by
NADPH oxidase. The authors find that
autophagy is required for inflammatory
activity of neutrophils.

Highlights
d Autophagy regulates neutrophil degranulation

d Autophagy regulates NADPH-oxidase-mediated reactive

oxygen species production

d Autophagy deficiency reduces inflammatory activity of

neutrophils

Bhattacharya et al., 2015, Cell Reports 12, 1731–1739

September 22, 2015 ª2015 The Authors
http://dx.doi.org/10.1016/j.celrep.2015.08.019
Cell Reports
Report
Autophagy Is Required
for Neutrophil-Mediated Inflammation
Abhisek Bhattacharya,1,2 Qin Wei,1 Jin Na Shin,1 Elmoataz Abdel Fattah,1 Diana L. Bonilla,1 Qian Xiang,1
and N. Tony Eissa1,2,*
1Department of Medicine
2Department of Pathology and Immunology
Baylor College of Medicine, Houston, TX 77030, USA
*Correspondence: teissa@bcm.edu
http://dx.doi.org/10.1016/j.celrep.2015.08.019
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
SUMMARY
Autophagy, an intracellular degradation and en-
ergy recycling mechanism, is emerging as an im-
portant regulator of immune responses. However,
the role of autophagy in regulating neutrophil func-
tions is not known. We investigated neutrophil
biology using myeloid-specific autophagy-deficient
mice and found that autophagy deficiency re-
duced neutrophil degranulation in vitro and in vivo.
Mice with autophagy deficiency showed reduced
severity of several neutrophil-mediated inflammatory
and autoimmune disease models, including PMA-
induced ear inflammation, LPS-induced breakdown
of blood-brain barrier, and experimental autoimmune
encephalomyelitis. NADPH oxidase-mediated reac-
tive oxygen species generation was also reduced
in autophagy-deficient neutrophils, and inhibition of
NADPH oxidase reduced neutrophil degranulation,
suggesting NADPH oxidase to be a player at the
intersection of autophagy and degranulation. Over-
all, this study establishes autophagy as an important
regulator of neutrophil functions and neutrophil-
mediated inflammation in vivo.
INTRODUCTION
Neutrophils play a critical role as one of the first lines of innate
immune defense. Activated neutrophils migrate to the site of
inflammation and control microbes by phagocytosis, formation
of neutrophil extracellular traps, and secretion of antimicro-
bials, stored in specialized granular compartments. Neutro-
phils possess at least four types of granules. Among these,
primary, secondary, and tertiary granules develop from the
endoplasmic reticulum (ER)-Golgi network, whereas secretory
vesicles have an endocytic origin (Borregaard, 1997). Primary
(azurophilic) granules contain myeloperoxidase (MPO), b-glucu-
ronidase, elastase, and other antimicrobial factors, whereas
lactoferrin and matrix metalloprotease-9 (MMP-9) are stored
in secondary (specific) granules. Tertiary (gelatinase) granules
also contain MMP-9 but lack lactoferrin. The membranes of
Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors 1731
secretory vesicles contain several important molecules, in-
cluding b2-integrins (CD11b/CD18) and complement receptors,
which play an essential role in neutrophil migration and activation
(Amulic et al., 2012). Upon activation, these granules are mobi-
lized, fuse with phagosomes or plasma membrane, and release
their contents in the respective environments. Secondary gran-
ules are of interest, as the membranes of these granules contain
flavocytochrome b558, a component of the NADPH oxidase ma-
chinery (Amulic et al., 2012). Thus, fusion of secondary granules
with phagosomes or plasma membrane results in formation of a
functional NADPH oxidase complex, which then produces reac-
tive oxygen species (ROS). Degranulation, a process of regu-
lated exocytosis of granules, is one of the major mechanisms
of inflammatory response by neutrophils. However, the underly-
ing regulatory pathways and their respective roles remain incom-
pletely understood.
Autophagy plays important roles in the immune responses and
abnormalities in this pathway have been linked to several dis-
eases (Levine et al., 2011). More recently, autophagy has been
implicated in regulating secretion (Cadwell et al., 2008; Patel
et al., 2013; Ushio et al., 2011; Bhattacharya et al., 2014b),
though the underlying mechanisms and in vivo relevance are
not fully understood. The role of autophagy in neutrophil func-
tions remains largely unexplored.
Here, we show that deficiency of autophagy reduced degran-
ulation of neutrophils. Autophagy-deficient neutrophils had re-
duced NADPH oxidase-mediated ROS generation. Moreover,
inhibition of NADPH oxidase reduced neutrophil degranulation,
implicating NADPH oxidase in mediating effects of autophagy
on neutrophil degranulation. The in vivo relevance of these find-
ings was shown in the context of autoimmune and inflammatory
processes. These findings establish autophagy as an important
regulator of neutrophil functions and suggest that targeting auto-
phagy pathway could have therapeutic value during infection,
inflammation, or neutrophil-mediated autoimmune diseases.
RESULTS
Deficiency of Autophagy Reduced Neutrophil
Degranulation
To evaluate the role of autophagy in neutrophils, we generated
mice with myeloid-specific deletion of Atg7 (called Atg7MD mice
hereafter) by breeding mice bearing floxed Atg7 alleles (Atg7f/f
Figure 1. ATG7 Deficiency Reduced Neutrophil Degranulation
(A) Representative immunoblot of ATG7 using lysates from purified peritoneal macrophages (PM), bone marrow-derived macrophages (BMDM), spleen-derived
macrophages (SpDM), and bone marrow neutrophils (BMN). Lanes marked f/f and MD denote lysates from control and Atg7MD mice, respectively.
(B) BMN from control or Atg7MD mice were primed with cytochalasin B and then stimulated with 1 mM fMLF for 15 or 30 min. Gelatinase zymography was
performed on supernatants and on lysates from unstimulated BMN; n = 6/group. Densitometry analysis and a representative image, in which each lane or symbol
represents one mouse, are shown.
(C) Lysates from unstimulated BMN, and supernatant collected following cytochalasin B and 1 mM fMLF treatment of Atg7f/f and Atg7MD neutrophils were used for
immunoblotting by lactoferrin antibody. n = 6–9/group. Each lane or symbol represents one mouse.
(D) BMN from control or Atg7MD mice were stimulated as above. Supernatant was collected and incubated overnight with p-nitrophenyl-b-D-glucuronide (for b-D-
glucuronidase). OD values were measured at 405 nm.
(E and F) BMN were stimulated with cytochalasin B and 1 mM fMLF, 1 mg/ml LPS, or 1 mg/ml PMA for 15 min and evaluated by flow cytometry using antibodies
against Ly6G and CD11b (E) or LAMP-1 (F). n = 6/group. Each symbol represents one mouse.
Data are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001. MFI, mean fluorescence intensity.

mice) with transgenic LysM-cre mice (Abdel Fattah et al., 2015; ATG7, and the efficiency of genetic deletion was comparable to
Bonilla et al., 2013; Shin et al., 2013). Bone marrow neutrophils that observed in peritoneal macrophages or macrophages gener-
(BMNs) from Atg7MD mice showed substantial reduction of ated from bone marrow or spleen from Atg7MD mice (Figure 1A).

1732 Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors

As autophagy has been recently implicated in the secretory
pathway, we hypothesized that autophagy deficiency would
affect neutrophil degranulation. Among the granules origi-
nating from ER-Golgi apparatus, tertiary granules are the first
to release their contents upon activation. To test the effect of
autophagy on degranulation of tertiary granules, we stimu-
lated BMN with cytochalasin B and bacterial tripeptide
fMLF, and used the supernatant for gelatinase zymography.
Loss of ATG7 reduced the levels of MMP-9 secreted after
fMLF treatment (Figure 1B). Gelatinase zymography, per-
formed on the lysates of unstimulated BMN, showed that
baseline levels of MMP-9 in Atg7MD neutrophils were para-
doxically elevated (Figure 1B), suggesting that the reduction
in MMP-9 secretion was due to retention, rather than a defect
in its synthesis.
To determine whether autophagy was involved in the secre-
tion of other types of neutrophil granules, we stimulated
BMN with cytochalasin B and fMLF and evaluated the secre-
tion of b-D-glucuronidase (stored in primary granules) and
lactoferrin (stored in secondary granules). Deficiency of
ATG7 reduced degranulation from both primary and second-
ary granules (Figures 1C and 1D). In contrast, upregulation
of CD11b in response to various stimuli remained unaltered
in Atg7MD neutrophils (Figure 1E). Secretory lysosomes were
also unaffected, as evidenced by similar LAMP-1 levels in
ATG7-deficient and control neutrophils following PMA stimula-
tion (Figure 1F).
To rule out autophagy-independent roles of Atg7 in neutrophil
degranulation, we generated mice with myeloid cell-specific
deletion of Atg5 (Atg5MD mice) (Abdel Fattah et al., 2015). We
found that deletion of Atg5 reduced degranulation of primary,
Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors 1733
Figure 2. ATG5 Deficiency Reduced Neutro-
phil Degranulation
BMN from control or Atg5MD mice were primed
with cytochalasin B and then stimulated with 1 mM
fMLF for 15 or 30 min.
(A and B) Supernatant was collected, incubated for
15 min with tetramethyl-benzidine to determine
levels of MPO (A), or incubated overnight with
p-nitrophenyl-b-D-glucuronide to determine b-D-
glucuronidase (B).
(C) In addition, supernatant was used for immu-
noblotting by lactoferrin antibody. One represen-
tative blot in which each lane corresponds to one
mouse and densitometry analysis is shown.
(D) Gelatinase zymography was performed on
supernatant, and densitometry analysis is shown.
For all panels, each symbol represents one mouse.
n = 6/group. Data are mean ± SEM; *p < 0.05, **p <
0.01.
secondary, and tertiary granules from
BMNs (Figure 2). These data suggested
that autophagy was required for neutro-
phil degranulation.
We then wanted to determine whether
there was a global dysfunction in auto-
phagy-deficient neutrophils. We have
previously shown that zymosan phago-
cytosis was not altered in autophagy-deficient macrophages
(Bonilla et al., 2013). To test whether autophagy deficiency
altered zymosan uptake by neutrophils, we incubated BMN
from control or Atg7MD mice with FITC-labeled zymosan A
and checked phagocytosis by flow cytometry. There was no
significant difference in zymosan A uptake, or surface levels
of CD11b, between control and Atg7MD neutrophils (Figures
S1A and S1B). However, consistent with our previous findings,
we observed substantial reduction of MMP-9 in the super-
natant collected after incubation of zymosan A with Atg7MD
neutrophils (Figure S1C). Thus, the defect in neutrophil degran-
ulation in Atg7MD mice was not part of a global dysfunction in
these cells.
Reduced In Vivo Neutrophil-Mediated Inflammation in
Atg7MD Mice
We hypothesized that autophagy deficiency would reduce
severity of neutrophil-mediated inflammatory diseases in vivo.
To test this hypothesis, we utilized a model of neutrophil-medi-
ated PMA-induced ear inflammation (Gross et al., 2009). Atg7MD
mice showed substantial reduction in ear swelling (Figure 3A).
After PMA application, degranulation of MPO, detectable by
in vivo bioluminescence imaging, requires fusion of primary
granules with phagosomes (Gross et al., 2009). Neutrophil
MPO in Atg7MD mice was substantially reduced (Figure 3B).
These data suggested that autophagy deficiency reduced
degranulation of primary granules by reducing their fusion with
phagosomes.
We then extended our findings to another in vivo model of
inflammation. Intraperitoneal LPS injection increases the perme-
ability of blood-brain barrier, an effect dependent on MMP-9 (Aid

et al., 2010; Bennett et al., 2010). After LPS administration,
we observed that Atg7MD mice showed substantially reduced
permeability of the blood-brain barrier (Figure 3C). Moreover,
MMP-9 release, following stimulation of BMN with LPS, was
also reduced from autophagy-deficient neutrophils (Figure 3D).
These data suggested that autophagy-deficiency reduced over-
all inflammatory potential of neutrophils.
Increased Circulating Neutrophils in Atg7MD Mice
To rule out defects in neutrophil migration, we used a peritonitis
model in which peritoneal exudate cells were collected 4 hr after
intraperitoneal thioglycollate injection. Surprisingly, peritoneal
neutrophils were increased in Atg7MD mice, though no differ-
ences in CD11b levels were observed in these neutrophils (Fig-
ures 3E and 3F). In vitro transwell migration assay showed no
gross defect in Atg7MD neutrophil migration in response to
different stimuli (Figure S1D). Complete blood analysis of Atg7MD
mice and their littermate controls showed similar total number of
white blood cells (Figure S2), but total and differential neutrophil
counts were higher in the Atg7MD mice (Figures 3G and S2).
The increase in circulating neutrophils in Atg7MD mice was
likely the result of increased neutrophil content in the bone
marrow of Atg7MD mice (Figure 3H). These data suggested that
myeloid-specific autophagy deficiency increased the number
of circulating neutrophils and, consequently, their recruitment
to the site of inflammation. However, because of their reduced
effector functions, these neutrophils could not efficiently mediate
inflammation.
1734 Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors
Figure 3. Reduced Neutrophil-Mediated
Inflammation in Atg7MD Mice In Vivo
(A and B) Ear inflammation was induced by topical
application of PMA for 24 hr, and ear thickness
was then measured; n = 7–8/group (A). Biolumi-
nescence imaging was performed. A representa-
tive image and quantitative analysis are shown (B).
Similar symbols denote matched pairs. n = 7/
group.
(C) LPS was injected intra-peritoneally (5 mg/kg, at
0 and 48 hr), and blood-brain barrier permeability
was measured.
(D) BMN were isolated from control and Atg7MD
mice and stimulated with 1 mg/ml LPS for 30 min,
and supernatant was subjected to gelatinase zy-
mography. A representative zymogram, in which
each lane denotes one mouse, and quantitative
analyses are shown. n = 6/group.
(E and F) Peritoneal exudate cells were collected
4 hr after thioglycollate injection. Neutrophils were
isolated and stained with Ly6G (E) and CD11b (F).
(G) Blood was collected by cardiac puncture, and
complete blood count was performed.
(H) BMN were isolated and counted. For all panels,
each symbol represents one mouse.
Data are mean ± SEM; *p < 0.05, **p < 0.01. MFI,
mean fluorescence intensity.
Autophagy Deficiency in Myeloid Cells Reduced
Severity of Experimental Autoimmune
Encephalomyelitis
To further test the inflammatory potential of autophagy-deficient
neutrophils, we used experimental autoimmune encephalomy-
elitis (EAE), an autoimmune model, which mimics several as-
pects of multiple sclerosis disease in humans. Neutrophils
have been implicated as a critical mediator of CNS inflammation
during the effector phase of EAE and neutrophil depletion at
different phases of EAE inhibits or abolishes EAE (McColl et al.,
1998; Rumble et al., 2015; Steinbach et al., 2013). Secretion of
proteases, such as MMP-9, by neutrophils facilitates breakdown
of blood-CNS barrier during EAE (Opdenakker et al., 2001). We
found that, after EAE induction by active immunization, both
control and Atg7MD mice showed 100% incidence of EAE. How-
ever, the severity of EAE was lower in Atg7MD mice (Figure 4A).
Additionally, weight loss and cumulative disease score (area
under the curve) were attenuated in Atg7MD mice (Figures 4B
and 4C).
Consistent with the roles of MMP-9 in breakdown of blood-
CNS barrier, we observed an increase in MMP-9 in spinal cord
lysates, correlating with severity of EAE (Figure S3A). Previous
reports showed that blood-spinal cord barrier, rather than the
blood-brain barrier, was predominantly affected in mouse model
of EAE (Bennett et al., 2010). At the peak of EAE, we observed
reduced amount of Evan’s blue dye in the spinal cord, but not
in organs without permeability barrier, in Atg7MD mice (Figures
4D and S3B). Consistent with attenuation of breakdown of

blood-CNS barrier permeability, we found a significant reduction
in inflammatory cells in the CNS of the Atg7MD mice at the peak
of EAE (Figure 4E). Atg5MD mice also showed a significant reduc-
tion in severity and cumulative disease score of EAE compared
to their littermate controls (Figures 4F and 4G). Six out of ten
Atg5f/f mice reached score 5 compared to one out of 12 Atg5MD
mice (*p = 0.02, chi-square test). Thus, deficiency of either Atg5
or Atg7 in myeloid cells resulted in attenuated EAE phenotype.
Finally, to rule out global immunosuppression in these mice,
we utilized models of hapten-induced contact hypersensitivity
(CHS). CHS is mediated primarily by CD8 T cells (Vocanson
et al., 2009). The severity of hapten-induced CHS in response
to Oxazolone or DNFB remained undiminished in Atg7MD mice
(Figure 4H). Thus, the reduction in severity of EAE in Atg7MD
mice was not due to a global reduction in immune functions.
Peripheral Antigen-Specific T Cell Generation in EAE
Was Unaltered in Atg7MD Mice
To rule out reduced peripheral T cell activation in Atg7MD mice,
we examined MOG-specific T cell proliferation at different
phases of EAE. There was no significant reduction in antigen-
specific T cell proliferation in spleen of Atg7MD mice at any of
the doses or time points tested (Figure S4A). To test whether
macrophage-mediated antigen presentation was defective in
Atg7MD mice, we used DOBW cells, which recognize oval-
bumin323–339 peptide in the context of MHC-II and produce IL-
2 upon proliferation (Harding et al., 1991). We could not detect
any significant differences in antigen presentation between resi-
dent peritoneal macrophages from control and Atg7MD mice
(Figure S4B).
Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors 1735
Figure 4. Autophagy Deficiency in Myeloid
Cells Reduced Severity of EAE
(A–C) EAE was induced in Atg7MD mice and
littermate controls, and daily clinical scores (A),
daily weight changes (B), and cumulative disease
scores (C) were recorded. n = 12–13/group.
(D) Evaluation of brain and spinal cord barrier
permeability at day 16 after EAE induction.
(E) Mice were sacrificed at day 16 after EAE in-
duction, and CNS-infiltrating cells were isolated by
Percoll gradient.
(F and G) EAE was induced in Atg5MD mice and
littermate controls, and daily clinical scores (F) and
cumulative disease scores (G) were monitored. n =
10–12/group.
(H) Contact hypersensitivity was induced in
Atg7MD and control mice by sensitization and
challenge with Oxazolone (OXA) or DNFB. Ear
swelling was measured 24 hr after challenge. Each
symbol represents one mouse. AUC, area under
the curve.
Data are mean ± SEM; *p < 0.05, **p < 0.01.
We also ruled out any gross defects in
T cell compartments in Atg7MD mice
that might account for the phenotype
observed (Figures S4C–S4J). Moreover,
we found that use of neutrophil-depleting
1A8 antibody, but not isotype control
antibody, after the onset of EAE abolished the difference be-
tween Atg7f/f and Atg7MD mice in disease severity, cumulative
disease score, and weight loss (Figure S5). Taken together, we
concluded that the protective effect in the Atg7MD mice was
caused by defective neutrophil function and not by changes in
peripheral T cell activation.
Autophagy Deficiency Led to Reduced NADPH Oxidase-
Mediated ROS Production
Flavocytochrome b558, a component of the NADPH oxidase ma-
chinery, is present on the membrane of secondary granules of
neutrophils (Jesaitis et al., 1990). Therefore, NADPH oxidase-
mediated ROS production in neutrophils depends on fusion of
granules with plasma membrane/phagosome. Our in vivo data
suggested a fusion defect between granules and phagosome
as a cause of reduced degranulation. To further examine how
autophagy affected ROS generation by NADPH oxidase, we stim-
ulated control or Atg7MD BMN with fMLF and measured extracel-
lular ROS generation by cytochrome c reduction assay. Atg7MD
neutrophils showed significant reduction in ROS production in
response to fMLF (Figure 5A). Similarly, FcOxyburst assay also re-
vealed reduced phagosomal ROS production by Atg7MD neutro-
phils (Figure 5B). Similar results were found in Atg5MD BMN, in
which both phagosomal ROS generation and extracellular ROS
generation, measured by OxyBURST Green H2 HFF BSA, were
substantially reduced (Figures 5C and 5D). We did not find any
gross difference with respect to fMLF-induced F-actin organiza-
tion between control and Atg7MD neutrophils (data not shown).
Collectively, our data suggested that autophagy deficiency re-
duced NADPH oxidase-mediated ROS production by neutrophils.
Figure 5. Autophagy Deficiency Reduced NADPH Oxidase-Mediated ROS Production in BMN
(A) BMN from control or Atg7MD mice were incubated with cytochrome c and extracellular ROS generation in response to 1 mM fMLF was monitored continuously
by measuring changes in absorbance (OD550). Data are from four independent experiments.
(B) BMN from control or Atg7MD mice were incubated with 125 mg/ml FcOxyburst reagent, and phagosomal ROS generation was measured by flow cytometry. A
representative plot and quantitative analyses are shown. Each symbol represents one mouse. Data are from three independent experiments.
(C) BMN from control or Atg5MD mice were incubated with OxyBURST H2HFF Green BSA, and extracellular ROS generation in response to 1 mM fMLF was
monitored by measuring changes in fluorescence. Time point 0 is the first time point measured after addition of fMLF. One representative curve for each genotype
and quantitative analyses are shown. n = 4/genotype.
(D) BMN from control or Atg5MD mice were incubated with 125 mg/ml FcOxyburst reagent, and phagosomal ROS generation was measured by flow cytometry. A
representative plot and quantitative analyses are shown. Each symbol represents one mouse. n = 4/genotype.
(E–H) BMN from wild-type mice were pre-treated with DPI (10 mM, 1 hr) and then stimulated with cytochalasin B and fMLF. Supernatant was collected and
b-glucuronidase was measured; n = 8/group. Each pair of symbols represents one mouse (E). Supernatant was also subjected to immunoblotting using lactoferrin
antibody (F) or gelatinase zymography (G). Representative immunoblots and zymogram are shown. (H) Supernatant was analyzed colorimetrically for LDH activity
to detect cell death.
Data are representative of three independent experiments and presented as mean ± SEM; *p < 0.05, **p < 0.01. AUC, area under the curve.

We then investigated the effects of NADPH oxidase-mediated
ROS generation on neutrophil degranulation. We pre-treated
BMN from wild-type mice with vehicle only or with diphenyle-
neiodonium chloride (DPI, 10 mM, 1 hr), a specific inhibitor of
NADPH oxidase, and then stimulated these neutrophils with
cytochalasin B and fMLF. Treatment with DPI reduced degranu-
lation of primary, secondary, and tertiary granules and did not
alter cell viability (Figures 5E–5H). Taken together, these data
showed that autophagy deficiency led to reduced NADPH oxi-
dase-mediated ROS production and reduced ROS production
by NADPH oxidase further contributed to reduced degranulation
from Atg7MD neutrophils.
DISCUSSION
Despite extensive characterization of neutrophil phagocytic
activity and granules, the underlying mechanisms governing
functions of neutrophils remain incompletely understood. In
this regard, the current study provided several important findings
and established autophagy as a regulator of neutrophil func-
tions. It showed that inhibition of autophagy reduced degranula-
tion from neutrophils. Such reduction was not due to decreased
production of proteins but rather to a secretion defect. Moreover,
deficiency of either Atg5 or Atg7 led to similar defects in de-
granulation from neutrophils, suggesting a role of autophagy,
rather than autophagy-independent ATG functions, in mediating
degranulation from neutrophils. The in vivo relevance of these
findings was confirmed using several mouse models of neutro-
phil-mediated inflammatory and autoimmune processes, which
showed that autophagy deficiency reduced the inflammatory
potential of neutrophils.
Autophagy deficiency produced maximal effects on degran-
ulation of tertiary and secondary granules. In contrast, we
did not find any defect in upregulation of CD11b, suggesting

1736 Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors
intact degranulation of secretory vesicles in absence of auto-
phagy. In neutrophils, secretory vesicles degranulate immedi-
ately following mild stimulation. Tertiary and secondary granules
follow secretory vesicles, in that order, and primary granules are
the most difficult to mobilize. It is possible that the differential
effect of autophagy deficiency on different granules is a direct
consequence of ease of mobilizing these granules.
We predominantly used fMLF, which is among the strongest
inducers of neutrophil degranulation. It is possible that fMLF in-
duction masked any potentially small defect in degranulation of
secretory vesicles. However, defective degranulation is unlikely
to be specific for fMLF, as both LPS and zymosan resulted in
reduced MMP-9 release by autophagy-deficient neutrophils.
Thus, our data suggested an intrinsic problem in the degranula-
tion mechanism in absence of autophagy.
Another important finding in this study is the role of NADPH ox-
idase and ROS in mediating effects of autophagy on degranula-
tion. Our data are consistent with a recent report, which showed
similar effects of autophagy in mucus secretion by goblet cells
(Patel et al., 2013). However, the potential roles of NADPH oxi-
dase and ROS in mediating neutrophil degranulation remain
incompletely elucidated. Studies by different groups using neu-
trophils isolated from patients of chronic granulomatous dis-
ease, in which NADPH oxidase-mediated ROS generation is
impaired, provided contrasting observations showing impaired,
augmented, or unaltered degranulation (Baehner et al., 1969;
Gold et al., 1974; Pak et al., 2007). Such a wide spectrum of re-
sults could be attributed to demography of patients studied,
types of stimuli used, methods employed, and granular contents
measured as an index of degranulation. Although, in our study,
inhibition of NADPH oxidase resulted in significant reduction
of degranulation, the effects were modest compared to that
observed in autophagy-deficient neutrophils, suggesting that
additional factors contributed to the effect of autophagy on
degranulation.
Assembly of NADPH oxidase requires fusion of secondary
granules, the membrane of which contains b558 subunit, with
phagosomal or plasma membrane (Jesaitis et al., 1990). Auto-
phagy proteins are important for fusion of enzyme-containing
lysosomes with plasma membrane in osteoclasts (DeSelm
et al., 2011). The role of autophagy in mediating granule fusion
is also observed in our study, using in vivo model of PMA-
induced ear inflammation. In this model, bioluminescence im-
aging of neutrophilic MPO by luminol depends on proper fusion
of MPO-containing granules with phagosome (Gross et al.,
2009). We found substantial reduction in MPO bioluminescence
from the PMA-treated ear of Atg7MD mice, suggesting a poten-
tial defect in fusion of these granules with phagosome. Thus,
our study supports a model in which autophagy mediates
both fusion of granules with other membrane compartments
and degranulation. Autophagy-mediated granule fusion helps
in the assembly of an active NADPH oxidase complex and
NADPH oxidase-mediated ROS generation, in turn, further fa-
cilitates degranulation.
Despite normal transwell migration in response to different
stimuli, Atg7MD mice showed increased neutrophils in the perito-
neal cavity following thioglycollate administration. This finding
was likely the result of a higher number of circulating and
Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors 1737
bone-marrow neutrophils in Atg7MD mice. We have previously
shown increased proliferative activity of leukocytes in bone
marrow of Atg7MD mice (Abdel Fattah et al., 2015). Moreover,
ROS are critical for several neutrophil death mechanisms
and reduced ROS generation in Atg7MD neutrophils may
also contribute to better survival (Geering and Simon, 2011).
Together, these factors could explain higher numbers of circu-
lating neutrophils in Atg7MD mice.
Emerging evidence suggests that neutrophils play a broad role
in the immune system (Kumar and Sharma, 2010). Neutrophils
have been implicated in autoimmunity and as one of the major
effector cells in EAE (Rumble et al., 2015). Autophagy has also
been implicated in several autoimmune diseases (Bhattacharya
and Eissa, 2013). In this regard, we found that autophagy defi-
ciency in myeloid cells reduced severity of EAE. Reduced
severity of EAE in myeloid-specific autophagy deficient mice
was not due to impairment of peripheral antigen-specific T cell
activation. These results are consistent with our recent findings
supporting dendritic cells (DCs) as a major antigen presenting
cells during EAE (Bhattacharya et al., 2014a). LysMCre-medi-
ated deletion predominantly affects macrophages and neutro-
phils, whereas DCs remain largely unaffected (Clausen et al.,
1999). Normal peripheral T cell activation may reflect unaltered
functions of DCs in these mice. Autophagy-deficient neutrophils
showed reduced secretion of MMP-9, a matrix metalloprotei-
nase involved in breakdown of blood-CNS barrier in EAE.
MMP-9 is considered a potential therapeutic target in multiple
sclerosis and clinical trials with MMP-9 inhibitors are currently
under way (Muroski et al., 2008). Reduced severity of EAE in
Atg7MD mice correlated with reduced permeability of blood-spi-
nal cord barrier and reduced CNS cellular infiltration. Although
our studies utilized well-characterized neutrophil-mediated
in vivo models, they do not exclude possible roles of other
myeloid cells in these models.
In summary, this study revealed an important role of auto-
phagy in neutrophils during inflammatory and autoimmune
diseases. Neutrophils constitute the major portion of circu-
lating leukocytes in humans and are considered the first line
of cellular defense and major inflammatory mediator in the
immune system. Neutrophilic secretion is required not only
during infection but also to maintain normal homeostasis.
Increased neutrophilic secretion has been implicated in the
pathogenesis of diseases such as chronic obstructive lung
disease and cystic fibrosis. Recent studies proposed alter-
ation of the autophagy pathway as a therapeutic approach
to multiple diseases. Our study uncovered autophagy as a
major regulator of neutrophil function and suggested the pos-
sibility that targeting the autophagy pathway might offer an
important approach to treat neutrophil-mediated inflammatory
processes.
EXPERIMENTAL PROCEDURES
Mice
C57BL/6 and LysM-Cre transgenic mice in a C57BL/6 background were pur-
chased from Jackson Laboratory. Atg7MD and Atg5MD mice have been previ-
ously described (Bonilla et al., 2013; Shin et al., 2013; Abdel Fattah et al.,
2015). All mice used were housed in BCM vivarium (biosafety level 2). All animal
protocols were approved by institutional boards of Baylor College of Medicine.
In Vivo Disease Models
EAE and contact CHS models were induced as described previously (Bhatta-
charya et al., 2014a). For EAE, active immunization was performed with
MOG35–55 peptide emulsified in Freund’s complete adjuvant, along with two
doses of pertussis toxin. Mice were monitored daily for weight changes and
clinical signs (Supplemental Experimental Procedures). CHS was induced by
sensitizing and challenging the mice with Oxazolone or DNFB. Ear thickness
was measured before and 24 hr after the challenge. Acute dermatitis was
induced by applying 20 mg of PMA in 20 ml of DMSO to one ear and DMSO
only (vehicle) to the other ear, and ear swelling was measured after 24 hr. Biolu-
minescence imaging was performed as previously described (Gross et al.,
2009). In brief, mice were anaesthetized by isoflurane and imaging was per-
formed in a Xenogen In Vivo Imaging System (IVIS), 10 min after intraperitoneal
injection of luminol. Quantitation was done by software (Living Image In Vivo
Imaging Software, PerkinElmer).
Isolation of Neutrophils from Bone Marrow or Peritoneal Cavity
Single-cell suspension from bone marrow, after ACK lysis, was subjected to
gradient centrifugation in 60%/80% Percoll, and neutrophils were collected
from the interface (Halpert et al., 2011), washed in HBSS without Ca2+ and
Mg2+, and stained for flow cytometry or used for in vitro stimulation. Peritoneal
lavage was performed with 10 ml of ice-cold HBSS without Ca2+ and Mg2+ 4 hr
after injection of 1 ml thioglycollate, and neutrophils were isolated by Percoll
gradient.
BMN Stimulation and Assessing Degranulation
Isolated BMN were allowed to settle for 20–30 min at 37
C, 5% CO2 and then
pre-treated for 5 min with 10 mg/ml of cytochalasin B. BMN were then stimu-
lated with 1 mM fMLF, 1 mg/ml of LPS or PMA, or 10 ng/ml of GM-CSF for
various time points. In some experiments, neutrophils were pre-treated for
2 hr with 50 mM CQ and then cytochalasin B was added for 5 min, followed
by fMLF stimulation. Supernatant was collected by centrifugation and cells
were prepared for flow cytometry. MMP-9 zymography was performed using
Novex 10% Zymogram (Gelatin) Gel. For MPO detection, supernatant was
incubated with TMB reaction mixture (1 mg of 3,30 ,5,50 -tetramethylbenzidine
in 1 ml of DMSO and 9 ml of 0.1 M NaH2PO4 (pH 5.5) with 2 ml of fresh 30%
hydrogen peroxide) for 15 min (in dark) and OD615 was measured immediately.
For detection of b-glucuronidase, supernatant was incubated overnight with
3.15 mg/ml p-nitrophenyl-itroglucuronide in 0.17 M sodium acetate buffer
(pH 4.0), reaction was terminated by adding 0.4 M glycine (pH 10), and
OD405 was measured immediately.
Measurement of ROS Production
Cytochrome c reduction assay and OxyBURST Green H2HFF BSA assay
were performed to detect extracellular ROS production using 0.5 3 106 to 1 3
106 cells/well or 2 3 106 cells/well, respectively in a 96-well plate. Cytochrome
c or OxyBURST Green H2HFF BSA was added (final concentration of
1.5 mg/ml or 10 mg/ml, respectively) to the cells and baseline absorbance/fluo-
rescence was recorded. fMLF was then added (final concentration: 1 mM in
200 ml/well) and absorbance/fluorescence was monitored continuously at
550 nm (absorbance) or at 488 nm excitation and 530 nm emission (fluores-
cence). FcOxyburst assays were performed according to the manufacturer’s
instructions (Invitrogen). In brief, flow cytometric measurement was performed
using a total of 2 3 106 cells/tube. After baseline reading, FcOxyburst reagent
was added (125 mg/ml in a final volume of 400 ml), and data were collected contin-
uously with LSR Fortessa (Becton Dickinson). FACSDiva software was used for
data collection, and FlowJo software was used for analysis. Cell viability assays
were performed using lactate dehydrogenase (LDH) detection kit (Roche).
Statistical Analysis
Normally distributed data were analyzed by Student’s t test (Welch correction
was applied in case of significant difference in variance between the groups),
and non-parametric comparisons were performed by Mann-Whitney test. For
EAE data, curves were compared by ANOVA, daily clinical scores and weight
changes by Student’s t test. Paired t test was used for comparisons between
matched pairs (in vivo bioluminescence and degranulation of primary granules
after DPI treatment). Analyses were performed by GraphPad Prism 6.
1738 Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures
and five figures and can be found with this article online at http://dx.doi.org/
10.1016/j.celrep.2015.08.019.
ACKNOWLEDGMENTS
We thank Dr. Clifford Harding for DOBW cells. We acknowledge current and
former members of the N.T.E. lab for stimulating discussions and technical
assistance. This study was supported by the National Heart, Lung, and Blood
Institute and by the Cytometry and Cell Sorting Core at Baylor College of Med-
icine with funding from the NIH (AI036211, CA125123, and RR024574) and the
expert assistance of Joel M. Sederstrom.
Received: March 6, 2015
Revised: May 7, 2015
Accepted: August 6, 2015
Published: September 3, 2015
REFERENCES
Abdel Fattah, E., Bhattacharya, A., Herron, A., Safdar, Z., and Eissa, N.T.
(2015). Critical role for IL-18 in spontaneous lung inflammation caused by auto-
phagy deficiency. J. Immunol. 194, 5407–5416.
Aid, S., Silva, A.C., Candelario-Jalil, E., Choi, S.H., Rosenberg, G.A., and
Bosetti, F. (2010). Cyclooxygenase-1 and -2 differentially modulate lipopoly-
saccharide-induced blood-brain barrier disruption through matrix metallopro-
teinase activity. J. Cereb. Blood Flow Metab. 30, 370–380.
Amulic, B., Cazalet, C., Hayes, G.L., Metzler, K.D., and Zychlinsky, A. (2012).
Neutrophil function: from mechanisms to disease. Annu. Rev. Immunol. 30,
459–489.
Baehner, R.L., Karnovsky, M.J., and Karnovsky, M.L. (1969). Degranulation of
leukocytes in chronic granulomatous disease. J. Clin. Invest. 48, 187–192.
Bennett, J., Basivireddy, J., Kollar, A., Biron, K.E., Reickmann, P., Jefferies,
W.A., and McQuaid, S. (2010). Blood-brain barrier disruption and enhanced
vascular permeability in the multiple sclerosis model EAE. J. Neuroimmunol.
229, 180–191.
Bhattacharya, A., and Eissa, N.T. (2013). Autophagy and autoimmunity cross-
talks. Front. Immunol. 4, 88.
Bhattacharya, A., Parillon, X., Zeng, S., Han, S., and Eissa, N.T. (2014a). Defi-
ciency of autophagy in dendritic cells protects against experimental autoim-
mune encephalomyelitis. J. Biol. Chem. 289, 26525–26532.
Bhattacharya, A., Prakash, Y.S., and Eissa, N.T. (2014b). Secretory function of
autophagy in innate immune cells. Cell. Microbiol. 16, 1637–1645.
Bonilla, D.L., Bhattacharya, A., Sha, Y., Xu, Y., Xiang, Q., Kan, A., Jagannath,
C., Komatsu, M., and Eissa, N.T. (2013). Autophagy regulates phagocytosis by
modulating the expression of scavenger receptors. Immunity 39, 537–547.
Borregaard, N. (1997). Development of neutrophil granule diversity. Ann. N Y
Acad. Sci. 832, 62–68.
Cadwell, K., Liu, J.Y., Brown, S.L., Miyoshi, H., Loh, J., Lennerz, J.K., Kishi, C.,
Kc, W., Carrero, J.A., Hunt, S., et al. (2008). A key role for autophagy and the
autophagy gene Atg16l1 in mouse and human intestinal Paneth cells. Nature
456, 259–263.
Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz, R., and Förster, I. (1999).
Conditional gene targeting in macrophages and granulocytes using LysMcre
mice. Transgenic Res. 8, 265–277.
DeSelm, C.J., Miller, B.C., Zou, W., Beatty, W.L., van Meel, E., Takahata, Y.,
Klumperman, J., Tooze, S.A., Teitelbaum, S.L., and Virgin, H.W. (2011). Auto-
phagy proteins regulate the secretory component of osteoclastic bone resorp-
tion. Dev. Cell 21, 966–974.
Geering, B., and Simon, H.U. (2011). Peculiarities of cell death mechanisms in
neutrophils. Cell Death Differ. 18, 1457–1469.
Gold, S.B., Hanes, D.M., Stites, D.P., and Fudenberg, H.H. (1974). Abnormal
kinetics of degranulation in chronic granulomatous disease. N. Engl. J. Med.
291, 332–337.
Gross, S., Gammon, S.T., Moss, B.L., Rauch, D., Harding, J., Heinecke, J.W.,
Ratner, L., and Piwnica-Worms, D. (2009). Bioluminescence imaging of mye-
loperoxidase activity in vivo. Nat. Med. 15, 455–461.
Halpert, M.M., Thomas, K.A., King, R.G., and Justement, L.B. (2011). TLT2 po-
tentiates neutrophil antibacterial activity and chemotaxis in response to G pro-
tein-coupled receptor-mediated signaling. J. Immunol. 187, 2346–2355.
Harding, C.V., Collins, D.S., Kanagawa, O., and Unanue, E.R. (1991). Liposome-
encapsulated antigens engender lysosomal processing for class II MHC pre-
sentation and cytosolic processing for class I presentation. J. Immunol. 147,
2860–2863.
Jesaitis, A.J., Buescher, E.S., Harrison, D., Quinn, M.T., Parkos, C.A., Livesey,
S., and Linner, J. (1990). Ultrastructural localization of cytochrome b in the
membranes of resting and phagocytosing human granulocytes. J. Clin. Invest.
85, 821–835.
Kumar, V., and Sharma, A. (2010). Neutrophils: Cinderella of innate immune
system. Int. Immunopharmacol. 10, 1325–1334.
Levine, B., Mizushima, N., and Virgin, H.W. (2011). Autophagy in immunity and
inflammation. Nature 469, 323–335.
McColl, S.R., Staykova, M.A., Wozniak, A., Fordham, S., Bruce, J., and Willen-
borg, D.O. (1998). Treatment with anti-granulocyte antibodies inhibits the
effector phase of experimental autoimmune encephalomyelitis. J. Immunol.
161, 6421–6426.
Muroski, M.E., Roycik, M.D., Newcomer, R.G., Van den Steen, P.E., Opdenak-
ker, G., Monroe, H.R., Sahab, Z.J., and Sang, Q.X. (2008). Matrix metallopro-
Cell Reports 12, 1731–1739, September 22, 2015 ª2015 The Authors 1739
teinase-9/gelatinase B is a putative therapeutic target of chronic obstructive
pulmonary disease and multiple sclerosis. Curr. Pharm. Biotechnol. 9, 34–46.
Opdenakker, G., Van den Steen, P.E., Dubois, B., Nelissen, I., Van Coillie, E.,
Masure, S., Proost, P., and Van Damme, J. (2001). Gelatinase B functions as
regulator and effector in leukocyte biology. J. Leukoc. Biol. 69, 851–859.
Pak, V., Budikhina, A., Pashenkov, M., and Pinegin, B. (2007). Neutrophil ac-
tivity in chronic granulomatous disease. Adv. Exp. Med. Biol. 601, 69–74.
Patel, K.K., Miyoshi, H., Beatty, W.L., Head, R.D., Malvin, N.P., Cadwell, K.,
Guan, J.L., Saitoh, T., Akira, S., Seglen, P.O., et al. (2013). Autophagy proteins
control goblet cell function by potentiating reactive oxygen species produc-
tion. EMBO J. 32, 3130–3144.
Rumble, J.M., Huber, A.K., Krishnamoorthy, G., Srinivasan, A., Giles, D.A.,
Zhang, X., Wang, L., and Segal, B.M. (2015). Neutrophil-related factors as bio-
markers in EAE and MS. J. Exp. Med. 212, 23–35.
Shin, J.N., Fattah, E.A., Bhattacharya, A., Ko, S., and Eissa, N.T. (2013). Inflam-
masome activation by altered proteostasis. J. Biol. Chem. 288, 35886–35895.
Steinbach, K., Piedavent, M., Bauer, S., Neumann, J.T., and Friese, M.A.
(2013). Neutrophils amplify autoimmune central nervous system infiltrates by
maturing local APCs. J. Immunol. 191, 4531–4539.
Ushio, H., Ueno, T., Kojima, Y., Komatsu, M., Tanaka, S., Yamamoto, A., Ichi-
mura, Y., Ezaki, J., Nishida, K., Komazawa-Sakon, S., et al. (2011). Crucial role
for autophagy in degranulation of mast cells. J. Allergy Clin. Immunol. 127,
1267–1276.e6.
Vocanson, M., Hennino, A., Rozières, A., Poyet, G., and Nicolas, J.F. (2009).
Effector and regulatory mechanisms in allergic contact dermatitis. Allergy
64, 1699–1714.
Cell Reports
Supplemental Information

Autophagy Is Required
for Neutrophil-Mediated Inflammation
Abhisek Bhattacharya, Qin Wei, Jin Na Shin, Elmoataz Abdel Fattah, Diana L. Bonilla,
Qian Xiang, and N. Tony Eissa
Supplemental Data

Figure S1 (Related to figure 1 & 3). Deficiency of Autophagy reduced

degranulation, but did not alter global neutrophil functions. (A-C) BMN from control
or Atg7M∆ mice were incubated with FITC-zymosan A particles for 15 min at 37°C. Flow
cytometric analysis was performed to evaluate phagocytosis of zymosan by BMN (A)
and surface levels of CD11b (B). Supernatants were collected and subjected to
gelatinase zymography (C). Densitometry analysis and a representative zymogram are
shown. Each lane or symbol represents one mouse. N=8-10/group. (D) Transwell
migration assays were performed for 1 hr using BMN and medium (Med) with or without
1 µM fMLF or 10 nM GM-CSF in the lower chamber. N=3/ group. Data are mean ±
SEM; *p <0.05, **p<0.01. MFI, mean fluorescence intensity. NS: Not significant.

  1  
Figure S2 (Related to figure 3). No gross abnormalities in numbers and
percentages of circulating blood cells in Atg7M∆ mice. Blood was collected by
cardiac puncture and analysis was performed by automated analyzer. Total numbers of
red blood cells (RBC), white blood cells (WBC) and platelets are shown (A). Total (B)
and differential (C) counts of circulating blood cells are shown. Data are mean ± SEM.
Each symbol represents one mouse, *p<0.05. LUC, large unstained cells.  

  2  
Figure S3 (Related to figure 4). Characterization of permeability during EAE. (A)
Spinal cord lysates from naïve or mice during EAE were subjected to gelatinase
zymography. Representative image is shown, each lane represents one mouse. N=7-9/
group. (B) Sixteen days after EAE induction (at the peak of EAE), mice were injected
intraperitoneally with Evan’s blue dye for 4 hr and organs were collected and weighed.
Dye was extracted by formamide and quantified by colorimetric assay. Each symbol
represents one mouse. Data are mean ± SEM. NS, not significant.  

  3  
   4  
Figure S4 (Related to figure 4). Myeloid specific deletion of Atg7 did not produce
any gross defect in T cell compartments (A) Single cell suspensions from spleens
were prepared at the peak (day 16) or remission (day 26) phase of EAE and incubated
with the indicated concentration of MOG peptide for 72 hr. Proliferation was measured
by incorporating tritiated thymidine for the last 18 hr of the assay. N=6-7/ group. (B)
Naïve peritoneal macrophages were collected, purified by adherence, stimulated with
LPS (1µg/ ml) for 40 hr and incubated for 24 hr with the indicated concentration of
ovalbumin and DOBW hybridoma cells. Supernatant was collected and assayed for IL-2
by ELISA. N=8-10/ group. (C-F) Spleen and thymus were collected from naïve control or
Atg7M∆ mice. Single cell suspensions were prepared and stained for different T cell
markers followed by intracellular staining for Foxp3. Total cell count (C), percent of
CD3+ cells (D), different T cell compartments in thymus (E) and Treg cells in thymus
and spleen (F) were enumerated. (G) MACS-purified CD4 T cells were stained with
CFSE and stimulated for 5 days with plate bound anti-CD3, anti-CD3 and soluble anti-
CD28 or 2 µg/ml of concanavalin A. Proliferation was measured by flow cytometric
estimation of CFSE dilution. N=3-8/ group. (H-I) Spleens were collected from naïve
control or Atg7M∆ mice. Single cell suspensions were prepared and transwell migration
assays were performed in response to various concentrations of sphingosine-1-
phosphate in the lower chambers of the transwell. T cell populations were identified in
both input cells and migrating cells using flow cytometry. Data are expressed for CD4
(H) and CD8 (I), as chemotactic index (fold increase over medium only). (J) Single cells
suspensions from spleens of control and Atg7-deficient mice were stained with green
and red cell trackers respectively, mixed in equal number and then injected into control
and Atg7-defiicent mice. Mice were sacrificed after 18-20 hr, blood and spleen were
collected and flow cytometry was performed to identify stained cells. N=3/group. Data
are mean ± SEM.  

  5  
Figure. S5 (Related to figure 4). Neutrophil depletion abolished the difference in
severity of EAE between control and Atg7M∆ mice. EAE was induced in mice. After
the onset of EAE, isotype control or neutrophil-depleting (1A8) antibodies were injected
intraperitoneally every alternate day starting from day 10 (A). Daily clinical scores (B),
cumulative scores (C) and weight (D) were monitored. N=4-8/ group, data represent
mean ± SEM. *p <0.05, **p<0.01, NS, not-significant  

  6  
Supplemental experimental procedures

EAE Induction and in vivo treatments  

EAE was induced as described (Bhattacharya A et al., 2014a). Briefly, eight to twelve

week-old mice were actively immunized with 100 µl of 1 mg/ml MOG35-55 peptide (EZ

Biolab) emulsified in Freund’s complete adjuvant (5 mg/ ml of Mycobacterium

tuberculosis extract H37Ra in incomplete Freund’s adjuvant). At and 48 hr after

immunization, all mice received intraperitoneal injections of 500 ng of pertussis toxin

(List Biological Laboratories). Mice were monitored daily for weight changes and clinical

symptoms, which were scored as follows: 0: no overt abnormalities, 1: limp tail or hind

limb weakness, 2: limp tail and hind limb weakness, 3: partial paralysis of hind limbs, 4:

complete hind limb paralysis, 5: moribund or death from EAE. At grade 5, mice were

sacrificed for humane reasons. Disease incidence was marked when an animal showed

clinical signs for at least two consecutive days. Neutrophil depletion experiments were

done 8-10 days after immunization for EAE. Mice were injected intraperitoneally every

other day for 10 days (5 doses) with either 500 µg of 1A8 antibody or 250 µg of the 2A3

isotype control (BioXcell). Investigators scoring the mice for all EAE experiments were

blinded to mice genotype.  

Induction of Contact hypersensitivity

Contact hypersensitivity was induced as described (Bhattacharya A et al., 2014a).

Briefly, mice were sensitized (day 0) by applying 50 µl of 5% (wt/vol) Oxazolone (4-

Ethoxymethylene-2-phenyl-2-oxazolin-5-one) in absolute ethanol, or by applying, on day

0 and day 1, 50 µl of 0.5% DNFB in acetone: olive oil (4:1) to the shaved abdomen.

  7  
Mice were challenged on day 5 by applying 20 µl of 0.2% DNFB on one ear (10 µl on

both sides) or on day 6 by applying 20 µl of 1% Oxazolone on one ear and solvent

alone to the other ear. For mice with ear tags, the tagged ear was always used for

vehicle treatment. Ear thickness was measured before and 24 hours after the challenge

with a micrometer (Mitutoyo) and ear swelling was calculated by using the formula:  

(Challenged ear after- Challenged ear before)- (Vehicle treated ear after-Vehicle treated ear

before).  

In vivo assessment of permeability  

EAE was induced in C57Bl/6 mice and permeability was measured as described

(Bennett J et al., 2010) with minor modifications. Briefly, 0.5% Evans blue dye was

injected intraperitoneally (50-100 mg/ kg body weight) at the peak of the disease (~16

days following immunization). For LPS-induced breakdown of blood-brain barrier, LPS

was injected intraperitoneally (5 mg/kg body weight) at 0 and 48 hr. Evan’s blue dye

was injected 16-20 hr after the second LPS injection. Four hours after Evan’s blue

injection, mice were deeply anaesthetized, perfused through the left ventricle with 30 ml

cold PBS and brain, spinal cord, spleen, a part of the liver (medial lobe) and left kidney

were collected. Organs were weighed and dye was extracted with 1 ml of formamide for

5 days at 37°C. The amount of dye was computed from a standard curve of Evans blue

by spectrophotometric measurement (OD650).  

In vivo competitive homing assay  

Single cell suspension was prepared from spleen mechanically and stained with

CellTrackers Green or Red (Molecular Probes). Equal numbers of labeled cells from

autophagy-deficient and littermate control mice were mixed and injected in mice retro-

  8  
orbitally. Mice were sacrificed after 18-20 hr, spleens were collected, single cell

suspension prepared and flow cytometry was performed to identify red and green cells.  

Isolation of CNS infiltrating lymphocytes  

Mice were sacrificed at the peak of EAE (~16 days after immunization) and

mononuclear cells from brain and spinal cord were isolated by gradient centrifugation

using 37/70% Percoll gradient (GE Healthcare) after digestion with collagenase and

DNase I. Cells were counted and stained for flow cytometric analysis after Fc blocking.

HL-60 culture and neutrophil differentiation  

HL-60 cells were obtained from ATCC. Neutrophil differentiation was performed by

culturing HL-60 cells for 5-7 days in complete RPMI-1640 medium supplemented with

1.3% DMSO. Cells were stimulated with cytochalasin B (10 µg/ ml) and 100 nM fMLF,

and assays for degranulation were performed as described above. For pharmacological

inhibition of class III PI3K pathway, differentiated HL-60 (neutHL-60) cells were treated

with 3-methyladenine (3-MA, 5 mM), LY-294002 (50 µM) or wortmanin (800 nM) for 2 hr

and then stimulated with cytochalasin B and fMLF.  

Culture of macrophages derived from bone marrow or spleen  

Macrophages from bone marrow or spleen were prepared as previously described

(Bonilla et al., 2013). Briefly, single cell suspension from bone marrow was treated with

ACK lysis buffer and cells were cultured for 6-7 days in presence of 10% conditioned

medium from L929 cells. Half of the medium was discarded and fresh medium added

after 3-4 days.  

Zymosan phagocytosis  

  9  
FITC-zymosan A particles (Invitrogen) were added to BMN at a ratio of 1:5-1:10.

Synchronous binding was performed by centrifuging the plates at 100g for 2 min at 4°C.

After washing with ice cold PBS, phagocytosis was initiated by adding pre-warmed

complete RPMI-1640. Plates were incubated for 15 min at 37°C and supernatant was

collected for zymography. Cells were treated with Trypsin-EDTA to remove zymosan

particles adherent to the surface, washed and then stained for flow cytometry.

Complete Blood Count  

Blood was collected into EDTA-coated tubes (BD Microtainer) by cardiac puncture from

deeply anaesthetized mice and complete and differential blood counts were performed

by automated blood analyzer (Advia 120).  

Single cell suspension and flow cytometry  

Single cell suspension from spleen, thymus and lymph nodes were prepared by

mechanical disruption over cell strainer (BD Falcon) followed by RBC lysis using ACK

lysis buffer (Quality Biological, Inc). Intracellular staining for Foxp3 was performed using

Cytofix/Cytoperm Plus (BD) kit according to manufacturer’s instructions. In all neutrophil

experiments for CD11b upregulation, analysis was performed on Ly6G+CD11b+ cells.

The following antibodies were used: CD8-APC (17-0081-81), F4/80APC-eFluor780 (47-

4801-82), all from eBioscience; CD11b-PerCP-Cy5.5 (550993), CD11b-APC-Cy7

(561039), CD4-APC (553051), CD4-PE (553048), Ly6G-PE (551461), CD8-FITC

(553030), CD3e-PE-Cy7 (561100) and Foxp3-PE (560414), all from BD Biosciences.

For F-actin reorganization studies, BMN were stimulated with 1 µM fMLF, fixed with ice

cold 4% paraformaldehyde, stained with AlexaFluor488-phalloidin for intracellular F-

actin and analyzed by flow cytometry. Flow cytometry was performed with LSR Fortessa

  10  
(Becton Dickinson), FACSDiva software was used for data collection and FlowJo

software was used for analysis.  

Transwell migration  

After 3-hr serum-starvation at 37°C, 106 splenocytes were placed on the top well of a 5-

µm polycarbonate insert (Costar) in 100 µl of medium and the bottom wells were set up

with 600 µl medium alone or with various concentrations of Sphingosine-1-phosphate.

After 3-hr incubation at 37°C, cells from the bottom wells were recovered, counted and

stained along with input splenocytes by flow cytometry. Migration was calculated as

percentage of number of input cells and chemotactic index was presented with respect

to the control well with medium only. Migration experiments for splenocytes were

conducted in serum-free RPMI 1640 medium supplemented with 0.5% BSA (fatty-acid

depleted, Sigma).

In transwell migration assay for neutrophils, complete RPMI 1640 medium and 3-µm

polycarbonate insert (Costar) were used with chemoattractants in the bottom wells.

After 1 hr at 37°C, media from the lower chamber were collected, 300 µl of HBSS (Ca2+

and Mg2+ free) containing 5 mM EDTA was added to the wells and the plate was

incubated on ice for 30 minutes. Wells were rinsed with 100 µl HBSS-EDTA and total

volume was made to 1 ml. Cells were counted using Countess automated cell counter

(Invitrogen).

T cell proliferation  

Antigen-specific T cell proliferation was measured by radioactive thymidine

incorporation. Single cell suspensions from spleen or lymph nodes were incubated with

the indicated concentration of MOG35-55 peptide for 3 days and 1 µCi/well of tritiated

  11  
thymidine was incorporated for the last 18 hours and measured by a scintillation

counter. For non-specific stimulation, CFSE dilution (Invitrogen) was used to measure

proliferation of total splenocytes or MACS-purified (Miltenyi) CD4 T cells as per

manufacturer’s instructions. Anti-CD3 and anti-CD28 (both from eBioscience) and

concanavalin A (Sigma) were used for stimulation.  

Antigen presentation assay  

Naïve peritoneal exudate cells were collected with ice cold medium and plated at 3X105

cells/ well in a 96-well plate, macrophages were isolated by adherence and stimulated

with LPS (1µg/ ml) for 40 hr. Antigen presentation to DOBW cells were performed with

Ovalbumin protein (Sigma) as described (Harding et al., 1991).

Immunoblotting  

Cell lysis was performed on ice using RIPA buffer with protease inhibitor cocktail (BD).

Lysates were collected after centrifugation, protein concentrations measured by BCA

protein assay kit (Pierce), boiled in Laemmli sample buffer for 5 minutes at 95°C and

resolved by SDS-PAGE. Following transfer onto a nitrocellulose membrane by Mini

Trans-Blot Cell (Bio-Rad), membranes were incubated with antibodies and imaging was

performed using an Odyssey Infrared Imaging System (LI-COR Bioscience). The

following antibodies were used: anti-ATG7 (600-401-487, Rockland Immunochemicals)

anti--actin (AM43020, Ambion), anti-p62 (03-GP62-C-1, American Research Products)

and anti-lactoferrin (ab77705, abcam).  

  12