Timing Is Critical for an Effective Anti-Metastatic
Immunotherapy: The Decisive Role of IFNc/STAT1-
Mediated Activation of Autophagy
Jun Yan1, Zi-Yan Wang1, Hong-Zhen Yang1, Han-Zhi Liu1, Su Mi1, Xiao-Xi Lv1, Xiao-Ming Fu1, Hui-Min
Yan1, Xiao-Wei Zhang1, Qi-Min Zhan2, Zhuo-Wei Hu1*
1Molecular Immunology and Pharmacology Laboratory, State Key Laboratory of Bioactive Substances and Functions of Natural Medicines, Institute of Materia Medica,
Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China, 2State Key Laboratory of Molecular Oncology, Cancer Institute and Cancer
Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
Background: Immunotherapy is often recommended as an adjuvant treatment to reduce the chance of cancer recurrence
or metastasis. Interestingly, timing is very important for a successful immunotherapy against metastasis, although the
precise mechanism is still unknown.
Methods and Findings: Using a mouse model of melanoma metastasis induced by intravenous injection of B16-F10 cells,
we investigated the mechanism responsible for the diverse efficacy of the prophylactic or therapeutic TLR4 and TLR9
agonist complex against metastasis. We found that the activation of TLR4 and TLR9 prevented, but did not reverse,
metastasis because the potency of this combination was neither sufficient to overcome the tumor cell-educated immune
tolerance nor to induce efficacious autophagy in tumor cells. The prophylactic application of the complex promoted
antimetastatic immunity, leading to the autophagy-associated death of melanoma cells via IFNc/STAT1 activation and
attenuated tumor metastasis. IFNc neutralization reversed the prophylactic benefit induced by the complex by suppressing
STAT1 activation and attenuating autophagy in mice. However, the therapeutic application of the complex did not suppress
metastasis because the complex could not reverse tumor cell-induced STAT3 activation and neither activate IFNc/STAT1
signaling and autophagy. Suppressing STAT3 activation with the JAK/STAT antagonist AG490 restored the antimetastatic
effect of the TLR4/9 agonist complex. Activation of autophagy after tumor inoculation by using rapamycin, with or without
the TLR4/9 agonist complex, could suppress metastasis.
Conclusion and Significance: Our studies suggest that activation of IFNc/STAT1 signaling and induction of autophagy are
critical for an efficacious anti-metastatic immunotherapy and that autophagy activators may overcome the timing barrier for
immunotherapy against metastasis.
Citation: Yan J, Wang Z-Y, Yang H-Z, Liu H-Z, Mi S, et al. (2011) Timing Is Critical for an Effective Anti-Metastatic Immunotherapy: The Decisive Role of IFNc/STAT1-
Mediated Activation of Autophagy. PLoS ONE 6(9): e24705. doi:10.1371/journal.pone.0024705
Editor: Francesco Dieli, University of Palermo, Italy
Received May 8, 2011; Accepted August 16, 2011; Published September 1 , 2011
Copyright: ? 2011 Yan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by grants from the National Major Basic Research Program of China (2006CB503808, http://www.973.gov.cn/English/Index.
aspx), Creation of Major New Drugs (2009ZX09301-003-13; 2009ZX09301-003-9-1, http://program.most.gov.cn/), National Nature Scientific Foundation
(30973557,http://www.nsfc.gov.cn), and Ph.D. Programs Foundation of Ministry of Education of China (20070023035,http://www.cutech.edu.cn). The funders had
no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
The main cause of cancer mortality is disseminated disease,
rather than the primary tumor . Conventional treatments, such
as surgery, radiotherapy and chemotherapy, have little effect on
metastasis and recurrence, especially if a large proportion of the
tumor has already metastasized at the time of diagnosis. Thus,
metastasis remains the most formidable challenge in cancer
therapy. Metastasis is determined by the interaction between the
tumor cells and the host tissue microenvironment . Immuno-
therapy is particularly well suited to eliminate residual tumor cells,
especially quiescent and cancer stem cells because immunotherapy
manipulates the microenvironment to induce cancer immunity,
thus eradicating metastatic tumor cells .
Numerous anticancer immunotherapeutic strategies have been
developed, including active immunization (i.e., cancer vaccines
and adjuvants), passive immunization (i.e., adoptive cell immuno-
therapy), and antibodies and small molecular inhibitors that
modulate the tumor microenvironment . However, the clinical
results of immune-based strategies for treating human cancer have
not met expectations. This limited success is largely attributed to
the immune tolerance observed in cancer patients . Indeed,
during tumor progression, increased immunosuppressive factors
and immune evasion protect the host from the induction of an
efficacious anti-cancer response by immunotherapy . Further-
more, the timing for immunotherapy is another key factor for
determining the outcome of the therapy; however, the mechanism
underlying this remains unclear.
PLoS ONE | www.plosone.org1September 2011 | Volume 6 | Issue 9 | e24705
Toll-like receptors (TLRs) are a family of conserved pattern-
recognition receptors (PRRs) that mediate the inflammatory
response by detecting conserved motifs of pathogen- or damage-
associated molecular patterns (PAMPs or DAMPs) . Both
developed and emerging TLR agonists for cancer treatment act as
stand-alone therapies or in combination with various agents [8,9].
However, anticancer responses are often not achieved under
physiological conditions , and numerous TLR-based immu-
notherapy strategies for cancer treatment eventually fail . The
clinical impact of these studies is highlighted by the recent failure
of the Stage III clinical trial of CpG 7909 in non-small cell lung
cancer . Both the TLR4 agonist Escherichia coli lipopolysac-
charide (EC-LPS) and the TLR9 agonist CpG oligodeoxynucleo-
tide (CpG ODN) are immunostimulants and can induce a potent
Th1-type immune response in vivo. Moreover, TLR4 acts in
synergy with TLR9 in the induction of IL-12p70 in mouse
dendrite cells (DCs) [13,14]. We therefore designed an immuno-
therapeutic regimen consisting of EC-LPS plus CpG ODN to
assess the effect of this potent immunotherapy regimen in a
metastatic mouse model of B16 melanoma cells. Despite an
optimal synergistic combination of EC-LPS plus CpG ODN with
a similar dose and frequency, only prophylactic (versus therapeu-
tic) administration of this complex attenuated metastasis, indicat-
ing that effective antimetastatic immunotherapy depends critically
on administration timing. We further investigated what mecha-
nism(s) was responsible for the different efficacy resulting from the
timing of the complex’s delivery. Our study indicated that
perturbation of signal transducers and activators of transcription
1/3 (STAT1/3) and autophagy induction accounted for the
complex’s distinctive efficacy against metastasis. Our study may
provide guidance in designing rational immunotherapeutic
strategies for patients with advanced cancers.
Materials and Methods
E. Coli 0111: B4 LPS (Ultra-Pure) was purchased from
InvivoGen. CpG ODN 1826 (59- TCC ATG ACG TTC CTG
ACG TT-39, phosphorothioate-modified) and CpG 1826 control
(59- TCC ATG AGC TTC CTG AGC TT-39, phosphorothioate-
modified) came from Beijing SBS Corporation. FITC-, PE- or PE-
Cy5-conjugated anti-CD3, CD4, CD8, CD25, F4/80, CD206,
IgG2b and IgG2a mAb, IFNc, IL-12p70, IL-4, IL-10 and TGF-
b1 ELISA kits were purchased from eBioscience (San Diego, CA).
Anti-STAT1, p-STAT3 (Ser727), STAT3, suppressor of cytokine
signaling 3 (SOCS3), proliferating cell nuclear antigen (PCNA),
phosphoinositide 3-kinases 85a (PI3K85a), PI3K110a and Actin
antibody were obtained from Santa Cruz Biotechnology, Inc.
(Santa Cruz, CA). Anti-p-STAT1 (Tyr701), SOCS1, cleaved
caspase-3, beclin 1, histone H3, mammalian target of rapamycin
(mTOR), p-mTOR, immunity-related GTPase family M member
1 (IRGM1), glycogen synthase kinase 3 (GSK3), p-GSK3, AKT
and p-AKT antibodies were purchased from Cell-Signaling
Technology Inc. (Danvers, MA). Anti-LC3B antibody was from
Abcam Inc. (Cambridge, UK). Anti-IFNc antibody (clone
XMG1.2) and isotype-matched IgG1 were from BD Biosciences
Pharmingen (San Diego, CA). Rapamycin and anti-P62 antibody
were from Sigma-Aldrich, Inc. (St. Louis, MO).
The mouse melanoma cell line B16-F10 (B16 cells, CRL-6475)
was cultured in RPMI 1640 (Invitrogen Corporation, Carlsbad,
CA) supplemented with 2 g/L Na2CO3, 100 units/ml penicillin,
50 mg/ml streptomycin, and 10% FBS at 37uC in 5% CO2.
Female C57BL/6 mice were purchased from Vital River Lab
Animal Technology, Co. Ltd (Beijing, China) and maintained
under standard conditions in an animal facility at the Institute of
Materia Medica. Animal care and experimentation were con-
ducted in accordance with guidelines of the Institutional
Committee for the Ethics of Animal Care and Treatment in
Biomedical Research of Chinese Academy of Medical Sciences
and Peking Union Medical College (Permit NO. 002802). All mice
used in these studies were between 5 and 6 weeks of age. The tail
veins of the mice were injected with 56105B16-F10 cells that were
resuspended in 200 ml PBS, as previously described . The mice
were euthanized with excessive anesthesia at 14 days, and whole
lungs were isolated to calibrate the lung weight index (wet lung
weight (mg) per body weight (g)). An anatomical microscopic
metastasis quantitation was performed by counting the metastasis
nodes with different diameters on the surface of all lobes of lungs
for each mouse by two observers in a blinded fashion. The lungs
were then fixed with 4% paraformaldehyde to prepare for
To determine the impact of immunotherapy on tumor cell
metastasis, mice were injected i.p. with TLR4 agonist EC-LPS
(12.5 mg/kg) plus TLR9 agonist CpG ODN (0.25 mg/kg) or PBS
plus CpG control (0.25 mg/kg) every three days over the seven
days before (Days 7, 4, 1) or after (Days 1, 4, 7) tumor cell
inoculation . Sham and B16 bearing mice received PBS
vehicle alone. To determine the role of STAT1 activation in the
protective effect of prophylactic treatment, IFNc-neutralizing or
isotype-matched antibody (100 mg/mouse) was injected intrave-
nously with TLR4/9 agonist complex before tumor inoculation.
Alternatively, human recombinant IFNc (16106U/kg) alone was
applied to the mice once a day before tumor cell inoculation .
To determine the role of autophagy in metastasis, rapamycin (i.p.,
10 mg/kg/day) was applied with or without the TLR4/TLR9
agonist complex after tumor cell inoculation . To determine
the role of STAT3, the mice were treated with the TLR4/TLR9
complex with or without the JAK2/STAT3 antagonist AG490
(i.p., 30 mg/kg/day) after tumor cell inoculation .
To analyze protein expression, the right lung lobes were lysed in
RIPA buffer (50 mM Tris-HCl, pH 7.4, 1% NP-40, 0.25% Na-
deoxycholate, 150 mM NaCl, 1 mM EDTA) supplemented with
50 mM NaF, 20 mM b-glycerophosphate and a complete
protease inhibitor cocktail (Roche). Cytoplasmic and nuclear
fractions were prepared as described previously. Protein
concentrations were determined by bicinchoninic acid reagent.
Proteins were separated by sodium dodecyl sulfate polyacrylamide
gel electropheresis (SDS-PAGE) (12%, 10%, or 8% acrylamide)
and transferred to polyvinylidene fluoride (PVDF) membranes.
Membranes were incubated overnight at 4uC with primary
antibodies. After washing, the blots were incubated with the
appropriate HRP-conjugated secondary antibody and processed
to detect electrochemiluminescence signals (Amersham Bioscienc-
es). The signal intensity was determined with the Gel-proH
Analyzer (Gel-Pro Plus Version 6.0, Media Cybernetics).
Hematoxylin and eosin (H&E) and immunofluorescence
The left lower lobe of the lung was isolated, fixed, paraffin
embedded and coronally sliced into 4-mm thicknesses. The tissue
sections were stained with H&E. Protocols for immunofluores-
cence staining for LC3II and LAMP1 have been described
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org2 September 2011 | Volume 6 | Issue 9 | e24705
previously . The apoptosis of lung tissues was detected with
terminal deoxynucleotidyl transferase (TdT) nick-end labeling
(TUNEL) using the DeadEndTM Fluorometric TUNEL System
(Promega, WI). The tissues were observed with a Leica SP2
confocal microscope (Leica Microsystems, PA) equipped with the
appropriate filter system. The images were analyzed with Leica
confocal software. The autophagosomes were evaluated using the
coexpression of LC3 and LAMP 1. Autophagy-associated cell
death was determined with both LC3 and TUNEL .
Electron microscopy (EM)
The left lower lung lobe was cut into 1-mm3slices and fixed in
2.5% glutaraldehyde in cacodylate buffer (0.1 M, pH 7.2) and in
1% osmium tetroxide (OsO4) in cacodylate buffer. After fixation,
the slices were dehydrated and embedded in Spurr’s resin. For
light microscopy, semithin sections (2 mm) were stained with
toluidine blue. Ultrathin sections (70 nm) were cut, and structural
analysis was performed using an Olympus CM-12 Transmission
Electron Microscope operated at 80 KV. To quantify autophagic
structures, digital images were acquired. The number of
autophagosomes and autolysosomes in the cell cytoplasm was
determined per microscope area, and four separate thin sections
were analyzed for each mouse.
The lungs were perfused with PBS through the right ventricle,
and then whole lungs were harvested and dissected into
approximately 1-mm pieces. Single-cell suspensions were prepared
with 2 ml of dispase containing collagenase (2 mg/ml) and DNase
(50 mg/ml) for 30 min . After RBC lysis, the digested lungs
were resuspended in PBS and sequentially filtered through 70-mm
filters, and each single-cell suspension was divided into four parts
to analyze the number of CTL, Th, M1, M2 and Treg cells in the
lung. The cells were incubated with saturated concentrations of
FITC-, PE- and/or PE-cy5-conjugated mAb against CD3, CD4,
CD8, CD25, Foxp3, CD11b, CD11c, MHC I, MHC II, F4/80 or
CD206. Isotype-matched mAbs were used in the control samples.
CD4+and CD8+cells were gated from CD3+cells. CD4+CD25+
Tregs were gated from Foxp3+cells. M1 and M2 cells were gated
from CD11b+cells. Data were analyzed using CellQuest software
(Becton Dickinson, San Jose, CA).
ELISA for cytokines in lung tissue
The right lung lobes were lysed in PBS supplemented with
complete protease inhibitor cocktail. Lung tissue homogenate was
diluted with lysis buffer to a final protein concentration of 500 mg
per ml. The expression of IFNc, IL-12p70, IL-4, IL-10 and TGF-
b1 in lung tissue homogenates was detected using ELISA kits
according to the manufacturer’s instructions.
Differences between groups were assessed by ANOVA. Survival
curves were compared by the log-rank test. The results were
presented as the mean 6 standard error (S.E.). P values ,0.05
were considered statistically significant.
Timing determines the efficacy of the TLR4/9 agonist
complex against metastasis
To investigate the optimal timing for initiating anticancer
immunotherapy with the TLR4 agonist EC-LPS plus the TLR9
agonist CpG, mice were injected i.v. with B16-F10 melanoma
cells, and the TLR4/TLR9 agonist complex was injected i.p.
either before (prophylactic) or after (therapeutic) tumor cell
inoculation every three days for three doses. Control mice were
treated with PBS or the TLR4/TLR9 agonist complex without
B16 cell inoculation. The PBS-treated mice inoculated with B16-
F10 cells formed a great number of macroscopic pulmonary
metastases two weeks after tumor cell inoculation. The first animal
deaths occurred on the 23rd day, and all animals had died by the
34th day after tumor cell inoculation (Fig. 1A, 1B, 1C).
Prophylactic administration of the TLR4/TLR9 agonist complex
increased the animals’ survival rate (40% survival on the 34th day),
prolonged the survival time (13% of mice remained alive on the
45th day), and decreased the number of metastatic nodules (6465
vs. 257618 nodules/lung, p,0.001), compared with the PBS
treatment (Fig. 1A, 1B, 1C). However, therapeutic administration
of the complex neither suppressed metastasis (280617 nodules/
lung) nor enhanced animal survival (all died on the 33rd day).
Thus, prophylactic, but not therapeutic, administration of the
TLR4/9 agonist complex attenuated the pulmonary metastasis of
B16 melanoma cells.
Many therapies suppress tumor progression by inducing
programmed cell death and/or by inhibiting tumor cell prolifer-
ation . We thus examined the markers of apoptosis and
proliferation in the lung tissue. Two weeks after the final injection
of the TLR4/9 agonist complex, the expression of activated
caspase-3 and PCNA in the lung tissue of the mice treated with the
immune complex was similar to that in the mice treated with PBS
in the absence of tumor cell inoculation (Fig. 1D). Prophylactic
administration with the TLR4/9 agonist complex induced an
increase in the expression of activated caspase-3 (0.7860.09 vs.
0.1160.03, p,0.05) and a decrease in PCNA expression (8.262.1
vs. 27.662.0, p,0.05), compared to PBS administration in the
lung tissues. However, therapeutic application of the TLR4/
TLR9 agonist complexneither
(0.1560.04) nor attenuated tumor cell proliferation (33.1611)
(Fig. 1D). In fact, the therapeutic application of the TLR4/TLR9
agonist complex suppressed caspase-3 activity (0.4760.16 vs.
1.6460.33, p,0.01) compared to the mice treated with PBS in the
early stage of metastasis (Supplementary Fig. S1A). Therefore, two
different timing regimens of the TLR4/9 agonist complex had
different efficacies against metastasis due to their different
capacities for regulating apoptosis and proliferation.
Prophylactic or therapeutic application of the TLR4/TLR9
agonist complex differentially regulates the inflammatory
milieu in the lung of B16 bearing mice
To determine the influence of the complex on the immune
system in control animals, mice were injected with PBS or the
TLR4/9 agonist complex, and immune responses in lung tissue
were examined at 2 weeks after final injection of the complex. We
found that the lung-infiltrating immune cells and the expression of
cytokines in the mice treated with the complex were similar to
those in the mice treated with PBS in the absence of tumor cell
inoculation (Fig. 2A, B). We then examined the infiltration of
immune cells and the expression of cytokines in the lung tissues
after tumor cell inoculation. An immunosuppressive microenvi-
ronment was formed in the lung tissues of the PBS-treated B16-
bearing mice, with suppressed infiltration or secretion of
CD3+CD8+T cells, CD3+CD4+T cells, M1 cells, IFNc, and IL-
12p70 and increased infiltration or secretion of M2 cells, Treg
cells, IL-4, IL-10, and TGF-b (Fig. 2A, 2B). Prophylactic
intervention induced antitumor immunity in the lung tissues,
including enhanced infiltration or secretion of CD3+CD4+T cells
(6.0560.12% vs. 3.7460.73%, p,0.05), M1 cells (18.3161.30%
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org3September 2011 | Volume 6 | Issue 9 | e24705
vs. 12.6860.91%, p,0.05), IFNc (212620 vs. 8762 pg/mg
protein, p,0.001), and IL-12p70 (8069 vs. 2062 pg/mg protein,
p,0.001) and reduced infiltration or expression of M2 cells
(7.9160.89% vs. 13.3760.95%, p,0.05), Treg cells (5.0460.33%
vs. 24.2064.35%, p,0.01), IL-4 (66611 vs. 10069 pg/mg
protein, p,0.05), IL-10 (5165 vs. 141619 pg/mg protein,
p,0.05), and TGF-b1 (0.7660.13 vs. 1.9060.23 ng/mg protein,
p,0.01) compared to PBS administration (Fig. 2A, 2B). However,
therapeutic intervention failed to increase the infiltration or
(93622 pg/mg protein), and IL-12p70 (4267 pg/mg protein) or
attenuate the infiltration or expression of M2 cells (12.2361.0%),
IL-4 (78612 pg/mg protein), and IL-10 (107613 pg/mg protein).
Therapeuticintervention increased the infiltration of M1
cells (18.1660.79% vs. 12.6860.91%, p,0.01) and decreased
the infiltration or expression of Treg cells (5.8661.50%
1.9060.23 ng/mg protein, p,0.05) in the lung tissue (Fig. 2A, 2B).
To compare the immune response directly regulated by the
TLR4/9 agonist complex alone or by tumor cells alone in the lung
tissue, the mice injected with B16 cells or PBS were treated with or
without the complex for three doses. In the second day after final
injection of the complex, the mice were sacrificed and the lung-
infiltrating immune cells were analyzed by flow cytometry. The
mice treated with the complex without B16 cells increased the
infiltration of MHCIhighDCs, MHCIIhighDCs, CD3+CD8+T
cells, and M1 cells and decreased the infiltration of M2 cells and
Treg cells in the lung tissues as compared with the PBS-treated
control mice (Fig. S2A–F). Compared to the mice treated with the
complex with B16 cell inoculation, the mice treated with the
complex alone resulted in the increased infiltration of MHCIhigh
DCs, MHCIIhighDCs, and M1 cells in the lung tissues by
p,0.05) andTGF-b1 (0.9060.11vs.
Figure 1. Prophylactic, but not therapeutic, administration of the complex attenuates pulmonary metastasis of B16 melanoma
cells. C57BL/6 mice were injected i.v. with B16-F10 melanoma cells (56105/mouse) and administered with the TLR4 agonist EC-LPS (12.5 mg/kg) plus
the TLR9 agonist CpG ODN (0.25 mg/kg) i.p. every 3rdday for 3 doses before (prophylactic group) or after (therapeutic group) tumor cell inoculation.
Control animals were treated with PBS or EC-LPS plus CpG ODN with an identical dosage and frequency as indicated in prophylactic group. The mice
were humanely sacrificed, and their lungs were excised 14 days after tumor cell inoculation. Externally visible melanoma nodules on the lung surface
were counted using stereo microscopy. (A) Kaplan-Meier graph representing the cumulative survival of mice in the indicated treatment groups. The
data were analyzed using Kaplan-Meier survival analysis (n=15 per group). (B) The metastatic nodules were counted and data presented as the mean
6 S.E. (n=15). (C) Data are representative lung samples (upper panel) and representative H&E staining of lung sections (below panel) (magnification:
1006). (D) The expression of cleaved caspase-3 and PCNA was detected by western blot (left panel) and corresponding quantification (right panel) in
lung tissues 14 days after inoculation. Data are presented as the mean 6 S.E. (n=5 mice per group).
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org4September 2011 | Volume 6 | Issue 9 | e24705
3.560.51-, 3.860.53-, and 2.060.34- fold, respectively (Fig. S2A,
S2B, S2D). However, the mice treated with the complex with B16
cell inoculation decreased the infiltration of CD11C+MHCIhigh
DCs and CD11C+MHCIIhighDCs, but did not change the
infiltration of CTL and M1 cells in the lung tissues as compared
with the mice treated with PBS with B16 cell inoculation. In the
lung tissues from the mice treated with the complex with B16 cell
inoculation, the percentage of M2 cells was increased compared
with those from the mice treated with PBS with B16 cell
inoculation. These data proved that the application of the
TLR4/9 complex without B16 cells activates both innate and
adaptive immunity by regulating DC maturation and M1
polarization in the lung. When the TLR4/TLR9 agonist complex
is applied after tumor cell inoculation, it is unable to reverse the
immunosuppressive tissue environment induced by tumor cells.
Activation of the transcription factors STAT1/STAT3 is crucial
in determining whether inflammation in the tumor microenviron-
ment promotes or inhibits cancer development [15,23]. Because
the prophylactic or therapeutic application of the TLR4/TLR9
agonist complex differentially regulated the expression of Th1
cytokines IFNc and IL-12p70 or Treg cytokine IL-10 (Fig. 2B),
which has been coupled with the activation of JAK-STAT1 or
STAT3 signaling cascade [23,24], we examined whether different
timing regimens of the TLR4/9 agonist complex differentially
Figure 2. Prophylactic or therapeutic application of the complex differently regulates the inflammatory milieu and STAT1
activation. Mice were treated as indicated in the legend of Figure 1 and sacrificed 14 days after B16 melanoma cell inoculation. Lung single-cell
suspensions were prepared as indicated in the Methods. (A) CD3+CD8+T cells, CD3+CD4+T cells, Foxp3+CD4+CD25+Treg cells, CD11b+F4/80+CD2062
M1 and CD11b+F4/80+CD206+M2 macrophages were determined by flow cytometry. Data are the mean 6 S.E. (n=5). (B) The levels of antitumor
cytokines IFNc and IL-12p70 and suppressive factors IL-4, IL-10, TGF-b were detected in lung homogenates from mice using ELISA kits. Data are the
mean 6 S.E. (n=5). (C) The expression of STAT1 and STAT3 singling molecules in the lung tissue. The lungs were excised, and the cytoplasmic and
nuclear fractions were extracted as described in the Methods. The expression of p-STAT1, STAT1, p-STAT3, STAT3, and histone H3 in nucleic extracts
and SOCS1, SOCS3, and b-actin in the cytoplasm were detected with Western blotting. Left panel is representative western blots and right panels are
summary results. Data are presented as the mean 6 S.E. of five mice per group.
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org5 September 2011 | Volume 6 | Issue 9 | e24705
regulated the balance of STAT1/3 activity. As shown in Fig. 2C,
the phosphorylation or expression of STAT3 and SOCS3
increased, while the phosphorylation or expression of STAT1
and SOCS1 decreased in the lung tissues of the PBS-treated B16-
bearing mice as compared to those of PBS-treated control mice.
Prophylactic intervention reversed tumor-suppressed phosphory-
lation or expression of STAT1 (1.8260.20 vs. 0.4660.10,
p,0.001) and SOCS1 and suppressed the tumor-induced
3.7461.07, p,0.05) and SOCS3 in the lung tissues. However,
therapeutic intervention could not reverse the tumor cell-induced
(2.6561.07) in the lung tissues. Perturbation of the STAT1/3
balance induced by the different timing regimens of TLR4/9
agonist complex application directed cytokine/growth factor
signals from apoptotic to proliferative or from cancer immuno-
surveillance to cancer immunoediting.
of STAT3 (0.7260.24vs.
and STAT3 activation
Prophylactic, but not therapeutic, application of the
TLR4/TLR9 agonist complex activates autophagy in the
melanoma cells of metastatic nodes
Autophagy plays many roles as an immunological effector, such
as mediating TLR- and Th1 cytokine-induced responses .
Previous studies have shown that IRGM1 plays a crucial role in
host resistance to a variety of intracellular pathogens by promoting
phagolysosome maturation and autophagy. Its expression is
induced by the IFNc/STAT1 signal [26,27]. We found that the
expression levels of IRGM1, LC3B-II, and beclin-1 in the lung of
the prophylactically treated B16-bearing mice were markedly
increased compared to those in the therapeutically treated and the
PBS-treated B16-bearing mice (Fig. 3A). Furthermore, the P62
level was significantly elevated in the lung tissues of therapeutically
treated and PBS-treated B16-bearing mice, whereas it was
decreased in the lungs of the prophylactically treated B16-bearing
mice (Fig. 3A). These data suggest that prophylactic, but not
therapeutic, administration of the immune complex activates
autophagy in the lungs. To determine where autophagy occurred
in the lung sections, autolysosomes or autophagosomes were
detected using a confocal microscope and anti-LC3B and anti-
LAMP1 antibodies. In the lungs from PBS-treated and therapeu-
tically treated B16-bearing mice, autolysosomes (red and green
foci) only occurred at the perimeter of metastasis nodes but not
within the nodes (Fig. 3B). However, in the lung tissue from the
prophylactically treated mice, autolysosomes were located both at
the perimeter and at the center of the nodes (Fig. 3B). Therefore,
the number of autolysosomes in metastatic nodes was markedly
increased after prophylactic treatment. Meanwhile, what about
the changes of autophagic activity in metastatic tumor cells after
indicated treatments? p62 is targeted for lysosomal degradation
during autophagy, and the expression levels of p62 inversely
correlate with autophagic activity . The accumulation of p62
in the lung tissues was examined by confocal microscope. We
found that the accumulation of p62 only appeared in metastatic
nodes of B16 melanoma cells but not in normal lung tissues,
suggesting autophagic activity in melanoma cells is lower than that
in normal cells. Furthermore, prophylactic treatment reduced the
accumulation of p62 in melanoma cells (Fig. 3B). These data
suggest that prophylactic, but not therapeutic, administration of
the immune complex activates autophagy in the melanoma cells.
Because we observed that the prophylactic application of the
complex promotes cell death (Fig. 1D), we investigated whether
cell death depended on complex-activated autophagy .
Electron microscopic analysis of melanoma cells in the lung
revealed that melanoma cells in the prophylactically treated mice
(but not in the therapeutically treated or PBS-treated B16-bearing
mice) exhibited a pronounced vacuolization in the cytoplasm and
displayed signs of apoptosis (chromatin margination) (Fig. 3C).
Consistently, the number of cells with LC3 dots and TUNEL-
positive nuclei in the metastatic nodules was markedly enhanced in
the prophylactically treated B16-bearing mice (5.360.8% vs.
0.560.3%, p,0.01), but not in the therapeutically treated ones
(0.960.3% vs. 0.560.3%, p.0.05) (Fig. 3D). Approximate 70% of
TUNEL-positive cells in metastatic nodes were accompanied with
LC3 dots in the lung sections from prophylactically treated B16-
bearing mice. Moreover, we found that LC3BII and beclin-1
expression and the number of autolysosomes were increased, but
cleaved caspase-3 expression was not changed on Day 3 after
tumor cell inoculation in the prophylactically treated B16-bearing
mice (Fig. S1A-S1C), suggesting that the activation of autophagy
preceded apoptosis and that prophylactic administration of the
TLR4/9 agonist complex promotes melanoma cell death by
stimulating autophagy-associated cell death.
PI3K/Akt/mTOR signaling negatively regulates autophagy
. We investigated whether the differential regulation of PI3K/
Akt/mTOR signaling was responsible for the different efficacy of
two timing regimens against metastasis. PI3K/Akt/mTOR
signaling was activated in the lung tissue from PBS-treated B16-
bearing mice, as indicated by the enhanced expression or
phosphorylation of PI3K (p110a) (1.8160.23), PI3K (p85a)
(3.9761.0), AKT (2.8360.47), GSK3 (2.4360.45), and mTOR
(2.2960.48) (Fig. S3). However, prophylactic intervention caused
a significant reduction in the expression or phosphorylation of
(1.1860.14 vs. 2.0860.21, p.0.05), GSK3b (1.3660.12 vs. 4.85
61.46, p,0.05) and mTOR (1.2260.21 vs. 2.1360.50, p.0.05)
compared to therapeutic intervention (Fig. S3). These results
indicate that the prophylactic but not therapeutic administration
of the TLR4/9 agonist complex reverses tumor cell-induced
activation of the PI3K/AKT/mTOR signaling.
Neutralization of IFNc reverses the antimetastatic role of
the TLR4/TLR9 agonist complex
To determine whether the activation of IFNc-STAT1 signaling
and autophagy was responsible for the antimetastatic effects
produced by the prophylactic administration of the TLR4/9
agonist complex, we examined the antimetastatic role of IFNc
alone and IFNc-neutralizing antibody plus the TLR4/9 agonist
complex treatment. We found that the prophylactic application of
IFNc reduced the number of metastatic nodules by 47616% and
suppressed the phosphorylation or expression of PCNA and P62
while enhancing the phosphorylation or expression of activated
caspase-3, LC3BII, beclin-1, and STAT1 as compared to PBS
administration in B16-bearing mice (Fig. 4A, 4B, 4D, S4).
Consistently, IFNc treatment enhanced the number of cells with
LC3 dots and TUNEL-positive nuclei in metastatic nodes
(3.060.6% vs. 0.560.2%) (Fig. 4C). However, blocking the IFNc
produced by the TLR4/9 agonist complex with an IFNc-
neutralizing antibody almost doubled the number of metastatic
nodules compared to PBS administration (425687 vs. 234640
nodules/lung, p,0.01) (Fig. 4A, 4B). Indeed, blocking IFNc
suppressed apoptosis and autophagy-associated cell death and
significantly promoted proliferation, as indicated by the attenuated
expression of activated caspase-3, LC3BII, and beclin-1, by
decreased the percentage of LC3B positive, LC3B-TUNEL
positive,and TUNELpositivecells, and bytheenhanced expression
of PCNA and accumulation of p62 (Fig. 4C, 4D). Moreover,
the prophylactic application of TLR4/TLR9 complex-activated
STAT1 was blocked by the IFNc-neutralizing antibody (Fig. S4).
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org6September 2011 | Volume 6 | Issue 9 | e24705
Figure 3. Prophylactic, but not therapeutic, application of TLR4/TLR9 agonist complex activates autophagy in melanoma cells. Mice
were treated as indicated in the legend of Figure 1 and sacrificed 14 days after B16 melanoma cell inoculation. The lungs were obtained, and tissue
extracts or lung sections were prepared for indicated analysis. (A) The expression of IRGM1, LC3B-II/LC3B-I, beclin-1, and p62 in the lung tissues were
detected with Western blot analysis. The representative immune blots are shown in left panel, and statistical results are shown in the right panel.
Data are the mean 6 S.E. (n=5). (B) Upper: representative immunofluorescence microphotograph of LC3B and LAMP-1. Lung sections were stained
for LC3 (red) and LAMP-1 (green). Arrows point to LC3- and Lamp1- positive cells. Scale bar: 30 mm. Bottom: representative immunofluorescence
microphotograph of p62 accumulation. Lung sections were stained for p62 (green) and DAPI (blue). Green areas indicate accumulation of p62 in
melanoma cells. Scale bar: 30 mm. (C) Analysis of autophagy in lung sections by transmission electron micrograph (TEM). Autophagosomes and
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org7 September 2011 | Volume 6 | Issue 9 | e24705
However, therapeutic application of IFNc or IFNc plus the
complex had no antimetastatic effect on B16-bearing mice (data
not shown). These data suggest whether or not the IFNc/STAT1
signaling and autophagy are activated is critical for the antimeta-
static efficacy produced by prophylactic application of the TLR4/
TLR9 agonist complex.
Autophagy activation by rapamycin after tumor
inoculation suppresses tumor metastasis
To verify that the absence of autophagy activation might be
responsible for the complex’s failure to elicit an antimetastatic
effect after tumor inoculation, rapamycin was administered with or
without the TLR4/TLR9 agonist complex after tumor inocula-
tion. Rapamycin is an autophagy activator targeting mTOR. We
found that rapamycin, with or without the TLR4/TLR9 agonist
complex, markedly decreased the number of tumor metastatic
nodes and enhanced the phosphorylation or expression of STAT1,
IRGM1, cleaved caspase-3, and LC3BII, while suppressing the
phosphorylation or expression of STAT3, PCNA, and P62
compared to PBS (Fig. 5A, 5B). Compared to rapamycin alone,
the TLR4/TLR9 agonist complex plus rapamycin did not
produce a more potent antimetastatic efficacy (80612 vs. 37610
nodules/lung, p,0.05) but even partially restrained the antimeta-
static activity of rapamycin by suppressing the expression of
IRGM1 (2.6560.48 vs. 4.3860.51, p,0.05) and LC3BII
(1.8660.25 vs. 3.3760.76, p,0.05), and augmenting the phos-
phorylation of STAT3 (0.5660.14 vs. 0.2260.04, p,0.05) and the
expression of P62 (0.2960.14 vs. 0.1060.03, p,0.05) in the lung
tissues, and by enhancing the accumulation of p62 in metastatic
nodes of lung sections (Fig. 5A, 5B, 5C). These data indicate that
autophagy is a critical defense mechanism against metastasis
independent of immunotherapy.
Inhibiting STAT3 by AG490 induces anti-tumor activity via activation of
STAT1 and autophagy.
Activated STAT3 can suppress STAT1 activity directly or by
inducing inhibitory molecules, such as SOCS . To assess
whether STAT3 activation restrained the TLR4/TLR9 agonist
complex-induced STAT1 activation and autophagy-associated
tumor cell death, AG490, a selective JAK/STAT inhibitor, was
administered with or without the complex after tumor inoculation.
Mice treated with AG490 alone showed an antimetastatic effect
with decreased lung metastatic nodes (189622 vs. 278629
nodules/lung, p,0.05), STAT3 suppression, STAT1 activation
and IRGM1 expression when compared to the PBS-treated B16-
bearing mice (Fig. 6A, 6B). However, the administration of the
TLR4/TLR9 complex plus AG490 resulted in a further reduction
of metastatic nodules (98612 vs. 189622 nodules/lung, p,0.05)
with the activation of caspase-3 (2.8460.6 vs. 1.4260.27, p,0.05)
and autophagy in the lungs (Fig. 6A, 6B). Additionally, the mice
treated with the TLR4/TLR9 agonist complex plus AG490
showed a higher level of STAT3 suppression and IRGM1
expression compared to the mice treated with or without the
TLR4/TLR9 complex (Fig. 6B). These data indicate that the
inhibition of STAT3 reverses the suppressed STAT1 activity and
autophagy caused by tumor cells, which produces anti-metastatic
efficacy (Fig. 7).
Despite significant advances in cancer immunology and
immunotherapy, clinical investigations have had marginal success
[6,31]. The reasons underlying the relatively low clinical responses
to immunotherapy in cancer patients include 1) suboptimal
synergistic combinations of immunotherapeutic agents and 2)
delayed timing for administering the immunotherapeutic agents.
Regarding the first reason, recent studies indicate that there is not
only insufficient antitumor immunity, but also too many
immunosuppressive factors existing in the tumor environment
[6,32]. Thus, the optimal synergistic combinations of immuno-
therapy should include components that can enhance the
antitumor capability and components that can eliminate the
tumor-promoting factors from the tumor environment .
Regarding the second reason, immunotherapy should be applied
as early as possible, rather than at a later stage of the disease or
after other treatments have failed in the clinical trial. For instance,
beginning immunotherapy a day or two before surgery can boost
the immune system and block its suppression by psychological and
physiological stress .
In current study, we assessed the efficacy of an immunother-
apeutic regimen consisting of the TLR4 agonist EC-LPS plus the
TLR9 agonist CpG ODN against tumor metastasis. TLR agonists
have been shown to be Myd88-associated TLR (TLR2, TLR5,
TLR7 and TLR9) agonists and TRIF-coupled TLR (TLR3 and
TLR4) agonists that can act in synergy to induce high levels of
proinflammatory cytokines when applied simultaneously .
Furthermore, TLR agonists acting in synergy showed an increased
and sustained capacity to prime Th1 responses [14,34]. It has been
established that Th1 responses are crucial for protection against
tumor development and progression. Our data show that
triggering TLR4 and TLR9 simultaneously with LPS plus CpG
before tumor inoculation inhibits tumor metastasis significantly,
whereas triggering either TLR4 or TLR9 has no effect on
metastasis (data not shown). However, the potent immunothera-
peutic complex can only prevent disease and is unable to
therapeutically suppress metastasis, similar to the failures of
immunotherapy seen in patients with late-stage cancer [10,35],
suggesting that timing is crucial for efficacious anticancer
We found that the prophylactic or therapeutic application of the
TLR4/TLR9 agonist complex differentially regulated Th1
responses and subsequent tumor cell death by activating IFNc/
STAT1 signaling (in the case of prophylactic treatment) or by
activating STAT3 (in the case of therapeutic treatment), which is
responsible for the different efficacy against tumor metastasis.
These findings are consistent with reports that STAT1 and
STAT3 play opposite roles in cancer immunity  and that
IFNc/STAT1 activation is important in TLR agonist-induced
cellular inflammation . Although the precise mechanism is
required further investigation, tumor cell-induced STAT3 activa-
tion may largely be responsible for the suppression of IFNc/
STAT1 signaling and Th1 responses in mice treated with the
TLR4/9 agonist complex after tumor cell inoculation. We and
others have previously shown that the constitutive activation of
STAT3 in melanoma cells determines the development of tumor
autolysosomes were counted in 10 melanoma cells in every lung section. Data are the mean 6 S.E (n=4 mice/group). Typical autophagosomes and
autolysosomes (asterisk) and melanin granules (arrowheads) are indicated. N indicates nuclei of the cell. Scale bar: 500 nm. (D) Co-localization of LC3B
immunostaining (red) and TUNEL (green) in the tumor nodes was detected in lung tissue sections. Left three panels are representative images.
Arrows point to cells with LC3 dots and TUNEL-positive nuclei. Scale bar: 15 mm. Right panel is bar graph showing the percentage of cells with TUNEL-
positive nuclei or with TUNEL positive nuclei and LC3 dots relative to the total number of cells in each section. Twelve images from each lung
specimen were counted. Data are presented as the mean 6 S.E (n=6 mice per group).
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org8 September 2011 | Volume 6 | Issue 9 | e24705
Figure 4. IFNc neutralization reverses the protective role of TLR4/TLR9 agonist complex against tumor metastasis. C57BL/6 mice were
injected with B16-F10 melanoma cells (56105) and humanely sacrificed 14 days after tumor cell inoculation. The mice were intraperitoneally treated
with the TLR4/TLR9 complex (dosage and frequency as in Fig. 1) with or without an IFNc-neutralizing antibody (100 mg/mouse) or human
recombinant IFNc (16106U/kg once a day) before tumor cell inoculation. Control animals were treated with PBS. Externally visible metastases on the
lung surface were counted, as described in the legend of Figure 1. (A) Representative lung samples and representative pulmonary H&E staining
(magnification: 1006). (B) Metastatic nodules were summarized. Data are the mean 6 S.E. (n=10). (C) Effects of targeting IFNc on apoptosis or
autophagy-associated cell death were evaluated by confocal analysis of LC3 immunostaining (red) and TUNEL (green) in metastatic nodes of lung
tissue sections. Left panels are representative immunofluorescence microphotographs of LC3B and TUNEL. Arrows point to cells with LC3 dots and
TUNEL-positive nuclei. Scale bar: 15 mm. Right panel is a bar graph to show the percentage of cells with TUNEL-positive nuclei or with TUNEL positive
nuclei and LC3 dots relative to the total number of cells in each section. Twelve images from each lung specimen were counted. Data are the mean 6
S.E (n=6 mice/group). (D) The expression of apoptosis and autophagy-related proteins in lung tissues as indicated was detected by Western blots.
Representative immune blots are shown in the left panel, and the statistical results are shown in the right panel. Data are presented as the mean 6
S.E. (n=5 mice per group).
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org9 September 2011 | Volume 6 | Issue 9 | e24705
immune tolerance and tumor progression [15,36]. Furthermore,
STAT3 can be induced directly and rapidly by TLR4 and TLR9
agonists . For the reciprocal regulation of STAT1/3 activity,
STAT3 inhibition by JAK/STAT antagonist AG490 may allow
STAT1 activation and the expression of antitumor cytokines to
suppress tumor metastasis. In fact, therapeutic administration of
the TLR4/9 agonist complex plus AG490 is able to suppress the
STAT3 activity, and the anti-metastatic efficacy is thus augmented
or restored compared to the AG490 or TLR4/9 complex
treatment alone (Fig. 7).
The role of autophagy in tumorigenesis and metastasis remains
controversial because autophagy either promotes cell death or cell
survival [38,39]. However, the induction of autophagy- associated
cell death has been identified as a crucial tumor-suppressing
Figure 5. Augmentation of autophagy with rapamycin protects against tumor metastasis. C57BL/6 mice were injected with B16-F10
melanoma cells (56105) and were humanely sacrificed 14 days after inoculation of tumor cells. Mice were intraperitoneally administered with the
TLR4/TLR9 agonist complex (dosage and frequency as in the legend of Fig. 1) with or without rapamycin (10 mg/kg, once a day) after tumor cell
inoculation. (A) Rapamycin protects against metastasis. Metastatic nodules were counted and data are presented as the mean 6 S.E. (n=10). (B)
Rapamycin activates autophagy and regulates STAT1/3 signaling. The expression of STAT1/3 singling and autophagy-related molecules in the lung
tissue was detected by Western blot. The lungs were excised and the cytoplasmic and nuclear fractions were extracted as described in the Methods.
The expression of p-STAT1, STAT1, p-STAT3, STAT3, and histone H3 in nucleic extracts and IRGM1, LC3B, cleaved caspase-3, P62, PCNA, and b-actin in
the cytoplasm were detected with Western blotting. Left panel is representative western blots and right panels are summary results. Data are
presented as the mean 6 S.E. of 5 mice per group. (C) Rapamycin treatment decreases p62 accumulation in the lung tissue. Representative
immunofluorescence microphotographs were presented to show p62 accumulation in melanoma cells. Lung sections were stained for p62 (green)
and DAPI (blue). Scale bar: 30 mm.
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org10 September 2011 | Volume 6 | Issue 9 | e24705
mechanism. Our results now clearly demonstrate that the
autophagy-associated cell death is involved in the mechanism by
which the prophylactic application of the TLR4/9 agonist
complex promotes B16 melanoma cell apoptosis. Notably, the
autophagy process is commonly regulated by the cytokines and
transcription factors in tumor microenvironment or by the tumor
itself . Our studies indicate that autophagy activation in tumor
cells from the mice treated prophylactically with the TLR4/9
agonist complex is associated with the elevated levels of IFNc
expression and STAT1 phosphorylation. In contrast, IFNc/
STAT1 signaling and autophagy are not activated in tumor cells
from the lungs of therapeutically treated mice. Indeed, IFNc
neutralization alone suppressed STAT1 activation and autophagy
in the lung tissues from the prophylactically treated mice, which
resulted in a deprivation of the TLR4/9 agonist complex-induced
antimetastatic effect. Through reversing the activated STAT3 by
AG490, the suppressed STAT1 activity and autophagic activity
were restored, which led to an antimetastatic effect in mice treated
therapeutically with the TLR4/9 complex. Furthermore, rapa-
mycin, which induces autophagy by inhibiting mTOR kinase,
enhances STAT1 activity in the lungs of B16-bearing mice and
produces a potent anti-metastatic action. These data suggest that
IFNc/STAT1-activated autophagy is critical for the anti-meta-
static role of the TLR4/9 agonist complex. Consistent with our
findings, Li et al found that suppressing STAT1 phosphorylation
by fludarabine or by silencing the expression of STAT1 inhibits
the expression of LC3BI/II and decreases the number of
autophagosomes induced by IFN-c in primary human macro-
phages . However, Chang et al. reported that embryonic
fibroblasts from autophagy-deficient mice are resistant to IFNc-
induced STAT1 activation . Thus, STAT1 can interact
positively with autophagy although price mechanism requires to
be identified (Fig. 7).
On the other hand, our studies indicate that therapeutic
treatment of mice with the TLR4/9 agonist complex after
inoculation of B16F10 melanoma cells can not reverses tumor
cell-induced STAT3 activation, IL-10 expression, and autophagy
suppression in the lung tissues. Similarly to the IFNc/STAT1
signaling, STAT3 and IL-10 can form a positively regulatory loop
to promote tumor progression and metastasis through sustaining
immunosuppressive environment in tumor tissue . Van Grol
et al recently reported that IL-10 suppressed autophagy induced
Figure 6. Therapeutic application of TLR4/TLR9 agonist complex and AG490 act synergistically to attenuate metastasis. C57BL/6
mice were injected with B16-F10 melanoma cells (56105) and were humanely sacrificed 14 days after tumor cell inoculation. The mice were
intraperitoneally injected with the TLR4/TLR9 agonist complex (dosage and frequency stated in the legend of Fig. 1) with or without AG490 (30 mg/
kg, once a day) after tumor cell inoculation. (A) Metastatic nodules were counted and summarized, and the data are the mean 6 S.E. (n=10). (B) The
expression of STAT1/3 singling and autophagy-related molecules in the lung tissue. The lungs were excised and the cytoplasmic and nuclear fractions
were extracted as described in the Methods. The expression of p-STAT1, STAT1, p-STAT3, STAT3, and histone H3 in nucleic extracts and IRGM1, LC3B,
cleaved caspase-3, P62, PCNA, and b-actin in the cytoplasm were detected with Western blotting. Left panel is representative western blots and right
panels are summary results. Data are presented as the mean 6 S.E. of 5 mice per group.
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org11 September 2011 | Volume 6 | Issue 9 | e24705
by rapamycin via activation of STAT3 partially  while Park
et al report that IL-10 inhibits autophagy in macrophages via
activation of PI3K pathway. Thus, activation of IL-10/
STAT3 can impair autophagy induction and decrease autophagy-
associated tumor cell death. Interestingly, we observe that direct
activation of autophagy causes a significant inhibition of STAT3
activity although we do not know what mechanism responsible for
this regulation at this time. These studies indicate that IFNc/
STAT1 signaling plays a critical role in autophagy induction by
TLR agonists, and IL-10/STAT3 signaling serves as a negative
modulator of autophagy in response to tumor cells (Fig. 7).
In addition to the balance of STAT1/3 signaling, PI3K
signaling may be an alternative mechanism responsible for the
different antimetastatic roles produced by preventive or therapeu-
tic administration of the TLR4/9 agonist complex. Autophagy
acts as an immunological effector for TLR activation and Th1
cytokines . However, the PI3K pathway functions as a
negative regulator of TLR responses to downregulate autophagy
in cancer cells . The present study indicates that prophylactic
application of the TLR4/9 agonist complex decreases the
expression of PI3K p85 and p110 subunits, suppresses the
activation of AKT and the AKT downstream target GSK-3b,
which results in mTOR inactivation and autophagy activation.
However, the activation of PI3K/AKT/mTOR signaling by B16-
F10 cells cannot be suppressed with a therapeutic application of
the TLR4/9 agonist complex, which even further augments the
activated PI3K-AKT-mTOR signaling and reverses partially the
rapamycin-induced attenuation of mTOR . These findings
indicate that the differentiated regulation of PI3K/AKT/mTOR
signaling is also responsible for the different efficacies caused by
prophylactic or therapeutic administration of the TLR4/9 agonist
In summary, we have identified a mechanism underlying the
failure of an immunotherapeutic protocol against tumor progres-
sion and metastasis; the tumor cells activate STAT3 to hijack host
immune cells to protect the IFNc/STAT1 signaling from
activation and subsequently protect tumor cells from autophagy-
associated cell death (Fig. 7). Moreover, we demonstrate that
autophagy is a suppressive mechanism of metastasis and is
regulated by the tumor microenvironment. Our studies not only
suggest that administration timing is crucial for an efficacious
cancer immunotherapy but also indicate a novel strategy to induce
an effective antimetastatic response. Immunotherapy plus an anti-
inflammatory agent (AG490) or autophagy activator (rapamycin)
may be a rational immunotherapy against tumor progression and
agonist complex suppressed apoptosis and autophagy
in the metastatic cells. The mice were sacrificed on the 3rdday
after B16 melanoma cell inoculation. Lung tissue extracts were
prepared as described in the Methods. (A) Western blot analysis
and the corresponding quantification of cleaved caspase-3 and
PCNA in lung tissues 3 days after tumor cell inoculation. Data are
mean 6 S.E. (n=5). (B) Western blot analysis and the
corresponding quantification of the autophagy-related proteins
LC3B-II/LC3B-I, beclin-1 and p62 in lung tissues. Data are mean
6 S.E. (n=5). (C) Representative immunofluorescence micropho-
tograph of LC3 and LAMP-1. Lung tissue sections were stained
for LC3 (red) and LAMP-1 (green). Arrows point to LC3- and
Lamp1- positive cells. Scale bar: 30 mm.
Therapeutic application of the TLR4/9
fails to reverse the tumor cells-induced suppressive
immune responses. Mice were i.v. injected with B16
melanoma cells (56105/mouse) (10 mice/group) or with equal
volume of PBS (10 mice/group). Five of B16-bearing mice and five
of PBS-treated mice were administered with the TLR4/9 agonist
complex for three doses as indicated in the legend of Figure 1. The
mice were sacrificed on the second day after the final dose of the
complex administration. The lung single-cell suspensions were
prepared as indicated in the Methods. CD11c+MHC Ihighcells (A),
CD11c+MHC IIhighcells (B), CD3+CD8+T cells (C), CD11b+F4/
80+CD2062M1 macrophages (D), CD11b+F4/80+CD206+M2
macrophages (E) and Foxp3+CD4+CD25+Treg cells (F) were
detected with flow cytometry. The data are represented as the
mean percentage of positive cells 6 S.E. (n=6).
Therapeutic administration of the complex
TLR4/9 agonist complex differentially regulated PI3K-
AKT-mTOR signaling. Mice were treated as indicated in the
legend of Figure 1 and sacrificed on the 14th day after B16
melanoma cell inoculation. Lung tissue extracts were prepared as
described in the Methods. Lung tissue extracts were prepared as
described in the Methods. The expression of PI3K110a, PI3K85a,
p-AKT, AKT, p-GSK3, GSK3, p-mTOR and mTOR was
detected by western blot. Data are mean 6 S.E. (n=5).
Prophylactic or therapeutic application of the
scriptional activity. Mice were treated as indicated in the
Neutralization of IFNc regulated STAT tran-
Figure 7. Mechanisms account for the different efficacies of
prophylactic or therapeutic TLR4/9 agonist complex against
metastasis. Prophylactic application of the TLR4/9 agonist complex
results in the activation of IFNc/STAT1 signaling which stimulates
autophagy and autophagy-associated tumor cell death. Neutralization
of IFNc inhibits STAT1 and results in the attenuation of autophagy.
Therapeutic application of the TLR4/9 agonist complex can not
suppress metastasis because STAT3 is activated and expression of IL-
10 is enhanced by tumor cells, which reciprocally results in the
suppression of IFNc/STAT1 signaling, autophagy, and subsequent
tumor cell death. Inhibition of STAT3 activation by AG490 (or IL-10
antagonist) or activation of autophagy by rapamycin may suppress
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org12 September 2011 | Volume 6 | Issue 9 | e24705
legend of Figure 1 and sacrificed on the 14th day after B16 Download full-text
melanoma cell inoculation. Lung tissue extracts were prepared as
described in the Methods. The expression of p-STAT1, STAT1,
p-STAT3, STAT3 and histone H3 in the nucleus and SOCS1,
SOCS3 and b-actin in the cytoplasm were detected by western
blot. The left panel represents immune blots, and the right panel is
the related statistical results. Data are mean 6 S.E. (n=5).
Conceived and designed the experiments: JY ZWH. Performed the
experiments: JY ZYW HZY HZL SM XXL XMF HMY XWZ. Analyzed
the data: JY. Contributed reagents/materials/analysis tools: JY QMZ.
Wrote the paper: JY ZWH.
1. Kopfstein L, Christofori G (2006) Metastasis: cell-autonomous mechanisms
versus contributions by the tumor microenvironment. Cell Mol Life Sci 63:
2. de Visser KE, Eichten A, Coussens LM (2006) Paradoxical roles of the immune
system during cancer development. Nat Rev Cancer 6: 24–37.
3. Blattman JN, Greenberg PD (2004) Cancer immunotherapy: a treatment for the
masses. Science 305: 200–205.
4. Ruttinger D, Winter H, van den Engel NK, Hatz R, Jauch KW, et al. (2010)
Immunotherapy of cancer: key findings and commentary on the third Tegernsee
conference. Oncologist 15: 112–118.
5. Jemal A, Siegel R, Ward E, Hao Y, Xu J, et al. (2008) Cancer statistics, 2008.
CA Cancer J Clin 58: 71–96.
6. Lasaro MO, Ertl HC (2010) Targeting inhibitory pathways in cancer
immunotherapy. Curr Opin Immunol 22: 385–390.
7. Rubartelli A, Lotze MT (2007) Inside, outside, upside down: damage-associated
molecular-pattern molecules (DAMPs) and redox. Trends Immunol 28:
8. Romagne F (2007) Current and future drugs targeting one class of innate
immunity receptors: the Toll-like receptors. Drug Discov Today 12: 80–87.
9. Kanzler H, Barrat FJ, Hessel EM, Coffman RL (2007) Therapeutic targeting of
innate immunity with Toll-like receptor agonists and antagonists. Nat Med 13:
10. Fournier P, Schirrmacher V (2009) Randomized clinical studies of anti-tumor
vaccination: state of the art in 2008. Expert Rev Vaccines 8: 51–66.
11. Klinman D, Shirota H, Tross D, Sato T, Klaschik S (2008) Synthetic
oligonucleotides as modulators of inflammation. J Leukoc Biol 84: 958–964.
12. Schmidt C (2007) Clinical setbacks for toll-like receptor 9 agonists in cancer. Nat
Biotechnol 25: 825–826.
13. Theiner G, Rossner S, Dalpke A, Bode K, Berger T, et al. (2008) TLR9
cooperates with TLR4 to increase IL-12 release by murine dendritic cells. Mol
Immunol 45: 244–252.
14. Napolitani G, Rinaldi A, Bertoni F, Sallusto F, Lanzavecchia A (2005) Selected
Toll-like receptor agonist combinations synergistically trigger a T helper type 1-
polarizing program in dendritic cells. Nat Immunol 6: 769–776.
15. Yang HZ, Cui B, Liu HZ, Mi S, Yan J, et al. (2009) Blocking TLR2 activity
attenuates pulmonary metastases of tumor. PLoS One 4: e6520.
16. Liu YY, Cai WF, Yang HZ, Cui B, Chen ZR, et al. (2008) Bacillus Calmette-
Guerin and TLR4 agonist prevent cardiovascular hypertrophy and fibrosis by
regulating immune microenvironment. J Immunol 180: 7349–7357.
17. Noguchi M, Imaizumi K, Kawabe T, Wakayama H, Horio Y, et al. (2001)
Induction of antitumor immunity by transduction of CD40 ligand gene and
interferon-gamma gene into lung cancer. Cancer Gene Ther 8: 421–429.
18. Hahnel PS, Thaler S, Antunes E, Huber C, Theobald M, et al. (2008) Targeting
AKT signaling sensitizes cancer to cellular immunotherapy. Cancer Res 68:
19. Burdelya L, Catlett-Falcone R, Levitzki A, Cheng F, Mora LB, et al. (2002)
Combination therapy with AG-490 and interleukin 12 achieves greater
antitumor effects than either agent alone. Mol Cancer Ther 1: 893–899.
20. Ho HH, Ivashkiv LB (2006) Role of STAT3 in type I interferon responses.
Negative regulation of STAT1-dependent inflammatory gene activation. J Biol
Chem 281: 14111–14118.
21. Salazar M, Carracedo A, Salanueva IJ, Hernandez-Tiedra S, Lorente M, et al.
(2009) Cannabinoid action induces autophagy-mediated cell death through
stimulation of ER stress in human glioma cells. J Clin Invest 119: 1359–1372.
22. Kim HS, Lee MS (2007) STAT1 as a key modulator of cell death. Cell Signal
23. Yu H, Pardoll D, Jove R (2009) STATs in cancer inflammation and immunity: a
leading role for STAT3. Nat Rev Cancer 9: 798–809.
24. Chang YP, Tsai CC, Huang WC, Wang CY, Chen CL, et al. (2010) Autophagy
facilitates IFN-gamma-induced Jak2-STAT1 activation and cellular inflamma-
tion. J Biol Chem 285: 28715–28722.
25. Delgado MA, Deretic V (2009) Toll-like receptors in control of immunological
autophagy. Cell Death Differ 16: 976–983.
26. Singh SB, Davis AS, Taylor GA, Deretic V (2006) Human IRGM induces
autophagy to eliminate intracellular mycobacteria. Science 313: 1438–1441.
27. Bafica A, Feng CG, Santiago HC, Aliberti J, Cheever A, et al. (2007) The IFN-
inducible GTPase LRG47 (Irgm1) negatively regulates TLR4-triggered proin-
flammatory cytokine production and prevents endotoxemia. J Immunol 179:
28. Mizushima N, Yoshimori T, Levine B (2010) Methods in mammalian autophagy
research. Cell 140: 313–326.
29. Jin S, White E (2007) Role of autophagy in cancer: management of metabolic
stress. Autophagy 3: 28–31.
30. Cully M, You H, Levine AJ, Mak TW (2006) Beyond PTEN mutations: the
PI3K pathway as an integrator of multiple inputs during tumorigenesis. Nat Rev
Cancer 6: 184–192.
31. Umansky V, Malyguine A, Shurin M (2009) New perspectives in cancer
immunotherapy and immunomonitoring. Future Oncol 5: 941–944.
32. Curiel TJ (2007) Tregs and rethinking cancer immunotherapy. J Clin Invest 117:
33. Avraham R, Benish M, Inbar S, Bartal I, Rosenne E, et al. (2010) Synergism
between immunostimulation and prevention of surgery-induced immune
suppression: an approach to reduce post-operative tumor progression. Brain
Behav Immun 24: 952–958.
34. Raman VS, Bhatia A, Picone A, Whittle J, Bailor HR, et al. (2010) Applying
TLR synergy in immunotherapy: implications in cutaneous leishmaniasis.
J Immunol 185: 1701–1710.
35. Rosenberg SA, Yang JC, Restifo NP (2004) Cancer immunotherapy: moving
beyond current vaccines. Nat Med 10: 909–915.
36. Molavi O, Ma Z, Hamdy S, Lai R, Lavasanifar A, et al. (2008) Synergistic
antitumor effects of CpG oligodeoxynucleotide and STAT3 inhibitory agent JSI-
124 in a mouse melanoma tumor model. Immunol Cell Biol 86: 506–514.
37. Kortylewski M, Kujawski M, Herrmann A, Yang C, Wang L, et al. (2009) Toll-
like receptor 9 activation of signal transducer and activator of transcription 3
constrains its agonist-based immunotherapy. Cancer Res 69: 2497–2505.
38. Marx J (2006) Autophagy: is it cancer’s friend or foe? Science 312: 1160–1161.
39. Kenific CM, Thorburn A, Debnath J (2010) Autophagy and metastasis: another
double-edged sword. Curr Opin Cell Biol 22: 241–245.
40. Li JC, Au KY, Fang JW, Yim HC, Chow KH, et al. (2011) HIV-1 trans-
activator protein dysregulates IFN-gamma signaling and contributes to the
suppression of autophagy induction. AIDS 25: 15–25.
41. Van Grol J, Subauste C, Andrade RM, Fujinaga K, Nelson J, et al. (2011) HIV-
1 inhibits autophagy in bystander macrophage/monocytic cells through Src-Akt
and STAT3. PLoS One 5: e11733.
42. Park HJ, Lee SJ, Kim SH, Han J, Bae J, et al. (2011) IL-10 inhibits the starvation
induced autophagy in macrophages via class I phosphatidylinositol 3-kinase
(PI3K) pathway. Mol Immunol 48: 720–727.
43. Deretic V (2009) Multiple regulatory and effector roles of autophagy in
immunity. Curr Opin Immunol 21: 53–62.
44. Sinnberg T, Lasithiotakis K, Niessner H, Schittek B, Flaherty KT, et al. (2009)
Inhibition of PI3K-AKT-mTOR signaling sensitizes melanoma cells to cisplatin
and temozolomide. J Invest Dermatol 129: 1500–1515.
IFNc/STAT1-Activated Autophagy and Metastasis
PLoS ONE | www.plosone.org13 September 2011 | Volume 6 | Issue 9 | e24705