tmpA13D TMP

A RT I C L E S
Macrophage-secreted granulin supports pancreatic

cancer metastasis by inducing liver fibrosis
Sebastian R. Nielsen1, Valeria Quaranta1, Andrea Linford1, Perpetua Emeagi1, Carolyn Rainer1,
Almudena Santos1, Lucy Ireland1, Takao Sakai2, Keiko Sakai2, Yong-Sam Kim3,4, Dannielle Engle5,6,
Fiona Campbell1, Daniel Palmer1, Jeong Heon Ko3,4, David A. Tuveson5,6,7, Emilio Hirsch8, Ainhoa Mielgo1
and Michael C. Schmid1,9
Pancreatic ductal adenocarcinoma (PDAC) is a devastating metastatic disease for which better therapies are urgently needed.
Macrophages enhance metastasis in many cancer types; however, the role of macrophages in PDAC liver metastasis remains poorly
understood. Here we found that PDAC liver metastasis critically depends on the early recruitment of granulin-secreting
inflammatory monocytes to the liver. Mechanistically, we demonstrate that granulin secretion by metastasis-associated
macrophages (MAMs) activates resident hepatic stellate cells (hStCs) into myofibroblasts that secrete periostin, resulting in a
fibrotic microenvironment that sustains metastatic tumour growth. Disruption of MAM recruitment or genetic depletion of granulin
reduced hStC activation and liver metastasis. Interestingly, we found that circulating monocytes and hepatic MAMs in PDAC
patients express high levels of granulin. These findings suggest that recruitment of granulin-expressing inflammatory monocytes
plays a key role in PDAC metastasis and may serve as a potential therapeutic target for PDAC liver metastasis.
Pancreatic cancer is among the most lethal cancers in part owing to
its aggressive metastatic nature1,2. Metastatic spreading is a multistage
process, which starts with the dissemination of cancer cells from the
primary tumour site and ends with clinically detectable metastatic
outgrowth at distant organs. Tumours can release large numbers
of cancer cells into the circulation, but only a small proportion of
these cells are able to successfully survive and colonize the hostile
environment at the distant metastatic site3,4. Thus, the successful
outgrowth of metastatic tumour cells in this new distant environment
is a severe rate-limiting step during metastasis. Emerging evidence
indicates that the colonization of a new organ by metastatic tumour
cells critically depends on the support of non-cancerous stromal
partners58. Multiple steps of the metastatic cascade are supported
by stromal partners, particularly macrophages and, in many cancers,
metastasis correlates with increased macrophages at the metastatic
site914. PDAC is characterized by its formation of ductal structures
and a rich stromal compartment containing mainly fibroblasts, stellate
cells and infiltrating immune cells. In response to the presence of
tumour cells, quiescent fibroblasts and stellate cells become activated
myofibroblasts that express alpha smooth muscle actin15 (SMA). In

non-pathological conditions, myofibroblasts are critical for wound
healing16. In cancer, the role of myofibroblasts is controversial at
present. Previous studies have suggested that myofibroblasts support
tumour growth and restrict the delivery of chemotherapeutic drugs
to the tumour1719. However, recent reports have shown that genetic
depletion of myofibroblasts results in tumour progression and
metastasis20,21. Most pancreatic cancer studies have focused on the role
of stromal partners at the primary site; however, the role of stromal
cells at the secondary metastatic site, which for PDAC is most often
the liver, remains poorly understood.
At the moment, the best treatment option for PDAC patients
is resection of the pancreatic tumour, but, unfortunately, by the
time PDAC patients are diagnosed, most (80%) present with
non-resectable metastatic cancer. Moreover, more than 60% of the
patients whose tumours are resected, relapse with distant hepatic
recurrence within the first 24 months after surgery22,23. Thus, a better
understanding of the mechanisms underlying the metastatic process in
pancreatic cancer is critical to improve treatment and patient outcome.
Department of Molecular and Clinical Cancer Medicine, University of Liverpool, Ashton Street, Liverpool L69 3GE, UK. 2Department of Molecular and Clinical
Pharmacology, University of Liverpool, Ashton Street, Liverpool L69 3GE, UK. 3Aging Intervention Research Center, KRIBB, 125 Gwahak-ro, Yuseong-gu,
Deajeon 305-806, Korea. 4Korea University of Science and Technology, 217 Gajeong-ro, Yuseong-gu, Daejeon 305-350, Korea. 5Cold Spring Harbor Laboratory, Cold
Spring Harbor, New York 11724, USA. 6Lustgarten Pancreatic Cancer Research Laboratory, Cold Spring Harbor, New York 11724, USA. 7Rubenstein Center for
Pancreatic Cancer Research, Memorial Sloan Kettering Cancer Center, New York, New York 10065, USA. 8Department of Molecular Biotechnology and Health
Sciences, Center for Molecular Biotechnology, University of Torino, Via Nizza 52, 10126 Turin, Italy.
9
Correspondence should be addressed to M.C.S. (e-mail: mschmid@liv.ac.uk)
Received 13 August 2015; accepted 11 March 2016; published online 18 April 2016; corrected online 20 May 2016; DOI: 10.1038/ncb3340
NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016
2016 Macmillan Publishers Limited. All rights reserved
549
A RT I C L E S
Liver metastasis
CD68
SMA
P < 0.001
P = 0.002
CD45
CK
250
DAPI/Cancer cells/SMA
P = 0.005
P = 0.01
P < 0.001
100
50
50
20
tdTomatoRed
F4/80
Merged + DAPI
tdTomatoRed
SMA
Merged + DAPI
Myofibroblasts
2
1
BM-derived
macrophages
tdTomatoRed+
Macrophages
P < 0.001
on
t
Da rol
Da y 5
y
12
Connective tissue
(percentage of area)
Day 5
P < 0.001
Chimaeric mice
WT + tdTomatoRed BM
Resident
macrophages (KCs)
tdTomatoRed
NS NS
10
f
5
on
P = 0.045
30
P < 0.001
B
MTS
40
20
0
LM
Healthy P = 0.003
liver
Liver
metastasis
60
P=
0.003
40
T
N
K
N
eu
M .
ac
.
SMA
Percentage of CD45+ cells
CD68
60
tr
Da ol
Da y 5
y
12
150
Cells per FOV
CD45
Cells per FOV
200
HL
Day 12
P < 0.001
80
SMA
CD11b
Day 5
Cytokeratin
Healthy liver
Day 12
Figure 1 Metastatic PDAC cells induce macrophage recruitment and

activation of myofibroblasts in the liver. (a) Identification of pan-cytokeratin
(CK)+ metastatic pancreatic cancer cells, haematopoietic cells (CD45+ ),
macrophages (CD68+ ) and myofibroblasts (SMA+ ) as predominant cell
types in the hepatic metastatic microenvironment of pancreatic cancer
by immunohistochemical analysis of human biopsies. Representative
micrographs and quantification of the data are shown (n = 5 PDAC
patients, n = 5 healthy subjects; five fields assessed per sample; mean
s.e.m.; two-tailed unpaired t-test). HL, healthy liver; LM, liver metastasis;
FOV, field of view. (b) Established metastatic nodules in the liver were
processed 12 days post intrasplenic implantation of 1 106 KPCs and
analysed by flow cytometry. Composition of intrametastatic leukocytes is
shown as a percentage of CD45+ cells using the following definitions:
B cells (CD45+ CD3neg B220+ ); T cells (CD45+ B220neg CD3+ ), NK cells
(CD45+ B220neg CD3neg NK1.1+ ), neutrophils (CD11b+ Ly6G+ F4/80neg ), MAMs
(CD11b+ F4/80+ ) (n = 4 healthy livers; n = 8 liver metastasis; data combine
two independent experiments; mean s.e.m.; two-tailed unpaired t-test).
(c) Representative immunofluorescence staining of myofibroblasts (SMA+ )
clustering around metastatic KPCluc/zsGreen (zsGreen) cancer cells in the liver

at 5 and 12 days after implantation. Histogram: quantification of myeloid
(CD11b+ ) and myofibroblast (SMA+ ) cell frequency in livers during the
course of metastasis formation. Nuclei were counterstained with DAPI
(n = 6 mice per time point; four fields assessed per sample; data combine
two independent experiments; mean s.e.m.; two-tailed unpaired t-test).
(d) Representative Massons trichrome staining (MTS) of tumour-bearing
livers at 5 and 12 days after implantation. Histogram: quantification of area
occupied by fibrotic stroma (n = 6 mice per time point; four fields assessed
per sample; data combine two independent experiments; mean s.e.m.;
two-tailed unpaired t-test). (e) Schematic of the generation of chimaeric mice
resulting in tdTomatoRed-positive BM-derived macrophages and non-labelled
resident Kupffer cells (KCs). (f) Chimaeric mice from e were intrasplenically
implanted with 1 106 Panc02 cells and livers were collected after day
12. Immunofluorescence analysis of BM-derived tdTomatoRed+ cells in
combination with F4/80 staining and SMA staining in tumour-bearing livers.
Nuclei were counterstained with DAPI (data are from six mice per condition;
one experiment). Scale bars, 100 m; NS, not significant.
RESULTS
Metastatic PDAC cells trigger macrophage recruitment and an
extensive stromal response in the liver
The most common route of PDAC metastasis is to the liver. To
understand whether and how stromal cells influence liver PDAC
metastasis, we first analysed liver biopsies from advanced metastatic
PDAC patients and healthy volunteers by immunohistochemical

and immunofluorescence techniques. We found that metastatic
tumour cells (cytokeratin+ ) are surrounded by an abnormal stromal
compartment rich in haematopoietic immune cells (CD45+ ),
macrophages (CD68+ ), myofibroblasts (SMA+ ; PDGFR+ ) and
connective tissue deposition (Fig. 1a and Supplementary Fig. 1ac).
550
A RT I C L E S
0.1
T
PI
(3
)K
0.0
T
PI
(3
)K
T
PI
(3
)K
10
Cells per FOV
DAPI/F4/80/SMA
80
SMA
P = 0.002
60
P < 0.001
40
20
(Days) 0
T
Clodronate treatment
NS
60
40
20
0
PL
CL
PI
(3
Analysis
9
0.00
)K
Intrasplenic
implantation
0.05
0.10
F4/80
PI(3)K /
WT
0.15
0.2
20
0.20
0.3
30
P < 0.001
T
PI
(3
)K
10
Metastatic lesion area (mm2)
20
0.4
0.25
30
40
Metastatic foci (per 100 mm2 liver)
c
P < 0.001
40
0.5
Monocytes
P = 0.006
T
PI
(3
)K
10
Viable cells (%)
Viable cells (%)
50
Macrophages
P = 0.012
Metastatic foci
(per 100 mm2 liver)
CD45
P = 0.024
Viable cells (%)
12
F4/80
SMA
Clodronate liposomes
100
P < 0.001
80
Cells per FOV
DAPI/F4/80/SMA
PBS liposomes
Macrophage depletion
60
P = 0.001
40
20
0
0.04
Metastatic lesion area (mm2)
Seeding
P < 0.001
0.03
0.02
0.01
0.00
PL
CL
PL
CL
Figure 2 Macrophages promote myofibroblast activation and metastatic

growth. (ad) Liver metastasis was induced by intrasplenic implantation
of 1 106 KPC cells. Entire livers were collected and analysed 12 days
later. (a) Haematopoietic cells (CD45+ ), MAMs (CD45+ CD11b+ F4/80+ ),
and IMs (CD45+ CD11b+ Ly6C+ F4/80neg Ly6Gneg ) from tumour-bearing livers
of WT versus PI(3)K/ mice were evaluated by flow cytometry (n = 6 mice
WT; n = 8 mice PI(3)K/ ; data combine two independent experiments;
individual data points, horizontal lines represent mean s.e.m.; two-tailed
unpaired t-test). (b,c) Quantification of metastatic frequency (b) and average
metastatic lesion size (c) in WT and PI(3)K/ mice by haematoxylin and
eosin-stained liver sections (n = 7 mice WT; n = 9 mice PI(3)K/ ; data
combine two independent experiments; mean s.e.m.; two-tailed unpaired
t-test). (d) Representative immunofluorescence staining and quantification of
MAM (F4/80+ ) and myofibroblast (SMA+ ) cell frequency in livers in WT and
PI(3)K/ . Nuclei were counterstained with DAPI (n = 6 mice WT; n = 8 mice
PI(3)K/ ; four fields assessed per sample; data combine two independent
experiments; mean s.e.m.; two-tailed unpaired t-test). (eh) Liver
metastasis was induced by intrasplenic implantation of 1 106 KPC cells.
Macrophages were depleted by clodronate liposome treatment after initial
colonization of the liver had occurred. (e) Schematic illustration of the
experiment. (f) Representative immunofluorescent staining and quantification
of MAM (F4/80+ ) and myofibroblast (SMA+ ) cell frequency in tumourbearing livers treated with liposomes containing PBS (PL) or clodronate (CL).
Nuclei were counterstained with DAPI (n = 4 mice per condition; five fields
assessed per sample; one experiment; mean s.e.m.; two-tailed unpaired
t-test). (g,h) Evaluation of metastatic frequency (g) and area covered by
metastatic cells (h) in tumour-bearing livers of mice treated with PL or CL
(n = 4 mice per condition; all metastatic nodules assessed from one section
per sample; one experiment; individual data and mean s.e.m.; two-tailed
unpaired t-test). Scale bars, 100 m; NS, not significant.
To further evaluate the type of immune cells accumulating at

the metastatic site, we intrasplenically injected KPC-derived cells
(FC1199), isolated from the genetically engineered mouse model
of PDAC (KrasG12D ; Trp53R172H ; Pdx1Cre mice)24. In this model,
tumour cells migrate to the liver through the portal circulation
(a common way of metastasis occurring in humans)25,26, and generate

metastases restricted to the liver (Supplementary Fig. 1d). We
analysed established metastatic tumours at day 12, by flow cytometry,
and found that the percentages of CD45+ immune cells, B220+ B
cells, CD3+ T cells, Nk1.1+ NK cells, CD11b+ Ly6G+ neutrophils,
551
A RT I C L E S
a
BM
macrophages
BM
macrophages
Control
WT MCM
Grn/ MCM
Grn/ MCM + rGrn
Acta2
400
200
0
WT MCM
0
WT MCM
Grn/ MCM
rGranulin
+
+
Grn/ MCM
rGranulin
0.8
0.6
0.4
0.2
+ +
+
60
40
20
0
0
+
80
P = 0.001
P = 0.021
P < 0.001
100
P < 0.001
600 P < 0.001
P < 0.001
P < 0.001
P = 0.004
P < 0.001
1.0
2-
c0
macrophages
Col1a1
Relative
invasion (%)
10
+
+ +
+
d Metastasis-associated
P < 0.001
P = 0.003
20
15
0
WT MCM
Grn/ MCM
rGranulin
Pa
n
Metastasis-associated
macrophages
Acta2
Fold change
mRNA levels
+
+
M
0
50
M
KP
0
C
Pa
-C
M
nc
02
-C
M
100
C
-C
150
KP
P < 0.001
200
Granulin protein (ng ml1)
P < 0.001
0
WT MCM
Grn/ MCM
rGranulin
250
Relative Grn mRNA levels

(normalized to Gapdh)
P < 0.001
P = 0.008
Relative
invasion (%)
Fold change
mRNA levels
Col1a1
P = 0.005
P = 0.013
P = 0.009
Figure 3 Granulin secreted by macrophages activates hepatic stellate cells.

(a) Quantification of SMA (Acta2), and collagen 1a (Col1a) mRNA levels
in primary hStCs stimulated with isogenic macrophage-conditioned media
(MCM) generated from WT or granulin-deficient (Grn/ ) macrophages, and
MCM generated from Grn/ macrophages in the presence of recombinant
granulin (rGranulin) as determined by qPCR (n = 3 independent experiments;
mean s.e.m.; two-tailed unpaired t-test). (b) Quantification of hStC
invasion towards MCM generated from WT or Grn/ macrophages and
MCM generated from Grn/ macrophages in the presence of rGranulin
(n = 3 independent experiments; mean s.e.m.; two-tailed unpaired t-test).
(c,d) The same as in a,b, but MCM was generated using in vivo-derived
MAMs. Therefore, liver metastasis was induced by intrasplenic implantation
of 1 106 KPC cells into isogenic WT + WT BM and WT + Grn/ BM

chimaeric mice (n = 3 independent experiments; mean s.e.m.; two-tailed
unpaired t-test). (e) Representative immunofluorescence images of d showing
Vybrant Dil (Em565)-labelled murine hStCs showing invasion towards MCM
generated from WT and Grn/ MAMs and MCM generated from Grn/
MAMs in the presence of recombinant granulin (rGranulin) (data are from
three independent experiments). (f,g) Quantification of Grn mRNA levels
by qPCR (f) and granulin secretion by ELISA (g) in primary unstimulated
(M0) macrophages and macrophages stimulated with isogenic conditioned
media (CM) generated from murine KPC and Panc02 tumour cells (n = 3
independent experiments; mean s.e.m.; two-tailed unpaired t-test).
Scale bar, 100 m.
and F4/80+ macrophages were increased in tumour-bearing livers

compared with tumour free livers (Supplementary Fig. 2ac).
However, among the CD45+ immune cells, metastasis-associated
macrophages (MAMs; CD11b+ F4/80+ Ly6Gneg CCR2+ ) were the
most predominant cell population (Fig. 1b, Supplementary Figs 2d
and 3a,b). Interestingly, metastatic tumour cells initially induced
a rapid accumulation of CD11b+ F4/80neg Ly6C+ LY6Gneg CCR2+
inflammatory monocytes (IMs) followed by an accumulation of
MAMs (Supplementary Fig. 3a). In contrast, accumulation of
SMA+ myofibroblasts and stromal expansion were detected only
in established experimental and spontaneous metastatic lesions
(Fig. 1c,d and Supplementary Fig. 3bg).
Macrophages in the liver can be broadly categorized into two
classes: embryonically derived tissue-resident macrophages known
as Kupffer cells (KCs), and infiltrating macrophages derived from
IMs that originate from the bone marrow2730 (BM). To investigate
whether the increased number of MAMs we observe is due to an
expansion of resident KCs or to the recruitment of BM-derived cells,

we generated BM chimaeras by engrafting tdTomato red (tdTomato)
BM into irradiated wild-type mice (WT + tdTomato+ BM) or
WT BM into irradiated WT mice (WT + WT BM). As KCs in
the liver are radio-resistant31, they remain of host origin and are
therefore F4/80+ tdTomatoneg , whereas BM-derived macrophages are
of donor origin and F4/80+ tdTomato+ (Fig. 1e). After confirming
successful BM reconstitution (Supplementary Fig. 3h), chimaeric
mice were intrasplenically implanted with the murine PDAC cell
line Panc0232. We found that MAMs were exclusively tdTomato+ ,
and, thus, derived from BM (Fig. 1f and Supplementary Fig. 3i).
Conversely, in adjacent normal liver tissue, F4/80+ macrophages
remained tdTomatoneg , consistent with resident F4/80+ KCs (Fig. 1f).
In addition, SMA+ metastasis-associated myofibroblasts remained
tdTomatoneg , suggesting that these cells are not of BM origin, but
are locally activated mesenchymal cells, such as resident hepatic
stellate cells (hStCs) or fibroblasts (Fig. 1f and Supplementary Fig. 3j).
552
15
Liver metastasis
KPC
CD68
CD68
Granulin
80
P < 0.001
P < 0.001
Cells per FOV
0
Granulin
MAMs
Stroma (immunecell-depleted)
Tumour cells
P < 0.001
60
40
20
P < 0.001
HL
LM
60
40
Healthy liver
Liver metastasis
Human
CD68
Granulin
20
0
150
Tgfb
Grn
Cells per FOV
2.0
P = 0.001
1.5
1.0
Granulin
Relative mRNA levels

Healthy liver
80
10
Granulin protein (ng ml1)
MAMs
Stroma (immunecell-depleted)
Tumour cells
CD68
Fold change Grn mRNA levels
A RT I C L E S
0.5
P < 0.001
100
50
0
KC
MAM
HL
LM
Figure 4 Granulin is highly expressed in hepatic metastatic lesions

and metastasis-associated macrophages are the main source of granulin
secretion. (a,b) Quantification of Grn mRNA levels (a) and granulin protein
levels (b) in intrametastatic pancreatic cancer cells, immune-cell-depleted
stromal cells (zsGreenneg CD45neg ), and MAMs (CD45+ F4/80+ ) isolated by
fluorescence-activated cell sorting from established tumour-bearing livers 12
days after intrasplenic implantation of 1 106 KPCluc/zsGreen cancer cells
(a, data are from three pooled mice; one experiment; b, n = 3 independent
experiments; mean s.e.m.; two-tailed unpaired t-test). (c) Quantification
of Grn and Tgfb mRNA levels in tissue resident macrophages (KCs) and
MAMs sorted from established metastatic lesions from chimaeric WT mice
harbouring tdTomatoRed+ BM as described in Fig. 1g (data are from
three pooled mice; one experiment). (d) Spontaneous metastatic hepatic

tumours derived from KPC mice were isolated and analysed. Representative
images of immunohistochemistry staining for MAMs (CD68+ ) and granulin
expression on serial tissue sections from metastatic lesions and healthy
liver and quantification of the data (n = 5 control mice, n = 5 KPC
mice, five fields assessed per sample; mean s.e.m.; two-tailed unpaired
t-test). HL, healthy liver; LM, liver metastasis. (e) Representative images of
immunohistochemistry staining for MAMs (CD68+ ) and granulin expression
on serial tissue sections from human metastatic PDAC lesions and healthy
liver and quantification of the data (n = 5 healthy subjects, n = 5 PDAC
samples; five fields assessed per sample; mean s.e.m.; two-tailed unpaired
t-test). Scale bars, 100 m; asterisks indicate matched tissue areas.
Together, our findings demonstrate that: in PDAC liver metastasis,

MAMs originate from IMs; MAMs represent the predominant
immune cell population in metastatic liver lesions; and MAM
accumulation in the liver precedes myofibroblast activation.
that depletion of PI(3)K will ablate IM trafficking to the metastatic

site and consequently abolish MAM accumulation in the liver. Indeed,
macrophage and IM numbers were significantly reduced in metastatic
livers of PI(3)K/ mice compared with WT mice as measured
by flow cytometry analysis (Fig. 2a). Importantly, PI(3)K depletion
markedly reduced metastatic frequency, average metastatic lesion
size, and SMA+ myofibroblast numbers, in both KPC (Fig. 2bd
and Supplementary Fig. 4a) and Panc02 (Supplementary Fig. 4be)
liver-metastasis-bearing mice. Together, these results suggest that
inhibition of BM-derived macrophage recruitment to the liver
prevents pancreatic cancer metastasis to the liver and is accompanied
by a decrease in myofibroblast activation.
As most (80%) PDAC patients present with liver metastases
at the time of diagnosis, or relapse with hepatic recurrence after
surgical removal of primary pancreatic tumours1, we next focused our
Macrophages promote myofibroblast activation and metastatic

growth
To determine the functional role of MAMs in pancreatic cancer
metastasis, we next intrasplenically implanted KPC cells into isogenic
PI(3)K/ mice (p110/ )33 and control WT mice. PI(3)K is
mainly expressed in the haematopoietic compartment, and PI(3)K
knockout mice have a defect in monocyte recruitment in response
to inflammatory and tumour-derived signals34,35. As we previously
found that MAMs are IM-derived macrophages that are actively
recruited from the BM to the metastatic liver (Fig. 1f), we reasoned
553
A RT I C L E S
a WT
Grn/
DAPI/F4/80/SMA
WT
F4/80
SMA
NS
e
WT
Grn/
50
Ac
C t a2
ol
1a
1
Fn
W
G T
rn
/
100
Grn/ BM
0
80
60
40
20
Ki67
WT BM
0.5
P < 0.001
80
Caspase 3
WT BM
60
0.4
0.3
0.1
40
Grn/ BM
0.05
Grn/ BM
20
NS
2
1
0
/
M2
Figure 5 Granulin depletion prevents myofibroblast activation and PDAC

metastasis. (al) Liver metastasis was induced by intrasplenic implantation
of 5 105 KPC cells into WT and granulin-deficient (Grn/ ;) mice
(ae) and chimaeric WT + WT BM and WT + Grn/ BM mice (fl).
Entire livers were collected and analysed 12 days later. (a) Representative
images of haematoxylin and eosin staining of liver sections (data are
from 6 WT and 7 Grn/ mice; one experiment). (b) Metastatic frequency
(n = 6 WT mice, n = 7 Grn/ mice; all metastatic nodules assessed
from one section per sample; one experiment; individual data points,
horizontal lines represent mean s.e.m.; two-tailed unpaired t-test).
(c) Metastatic area (n = 6 WT mice, n = 7 Grn/ mice; all metastatic
nodules assessed from one section per sample; one experiment; mean
s.e.m.; two-tailed unpaired t-test). (d) Representative immunofluorescent
staining and quantification of MAM (F4/80+ ) and myofibroblast (SMA+ )
cell frequency in tumour-bearing livers. Nuclei were counterstained with
DAPI (n = 6 WT mice, n = 7 Grn/ mice; four fields assessed per sample;
one experiment; mean s.e.m.; two-tailed unpaired t-test). (e) qPCR
analysis of multiple hStC activation markers in intrametastatic myofibroblasts
isolated from established metastatic lesions (data are from three pooled
mice per condition; one experiment). (f) Metastatic frequency (n = 5 WT
+ WT BM mice, n = 6 WT + Grn/ BM mice; all metastatic nodules
554
W
G TB
rn M
12
Ifn
g
Ar
g
Re 1
tn
C la
hi
3l
3
Il-
M1
BM
BM
Relative mRNA levels

0.6
Caspase-3+ cells per FOV
WT BM
Grn/ BM
Percentage of Ki67+ tumour cells

W
G TB
rn M
rn
W
T
BM
WT BM
Grn/ BM
Ac
C ta2
ol
1a
1
Fn
0.2
P < 0.001
Fold change
(relative to Grn/ BM)
F4/80
SMA
NS
DAPI/F4/80/SMA
WT BM
0.4
W
G T
rn BM
W
G T
rn BM
/
BM
BM
10
P = 0.002
0.6
BM
20
h
Metastatic lesion area
(mm2)
Metastatic foci
(per 100 mm2 liver)
30
10
W
G T
rn
/
NS
40
20
0.1
Fold change
(relative to Grn/)
0.2
Grn/
0.3
40
30
rn
10
P = 0.043
W
T
20
0.4

(mm2)
Metastatic foci
(per 100 mm2 liver)
NS
30
Cells per FOV
Cells per FOV
P < 0.001
assessed from one section per sample; one experiment; individual data
points, horizontal lines represent mean s.e.m.; two-tailed unpaired
t-test). (g) Metastatic area (n = 5 WT + WT BM mice, n = 6 WT +
Grn/ BM mice; all metastatic nodules assessed from one section per
sample; one experiment; mean s.e.m.; two-tailed unpaired t-test). (h)
Representative immunofluorescence staining and quantification of MAM
(F4/80+ ) and myofibroblast (SMA+ ) cell frequency in tumour-bearing livers.
Nuclei were counterstained with DAPI (n = 5 WT + WT BM mice, n = 6
WT + Grn/ BM mice; four fields assessed per sample; one experiment;
mean s.e.m.; two-tailed unpaired t-test). (i) qPCR analysis of multiple
hStC activation markers in intrametastatic myofibroblasts isolated from
established metastatic lesions (data are from five pooled mice per condition;
one experiment). (j) qPCR analysis of multiple M2- and M1-macrophageassociated genes in MAMs isolated from metastatic tumours developed in
WT + WT BM and WT + Grn/ BM mice (data are from six pooled mice
per condition; one experiment). (k,l) Representative immunohistochemical
staining and quantification of Ki67+ tumour cell frequency (k) and cleaved
caspase 3+ cell numbers (l) in metastatic livers (n = 5 WT + WT BM
mice, n = 6 WT + Grn/ BM mice; five fields assessed per sample; one
experiment; mean s.e.m.; two-tailed unpaired t-test). Scale bars, 100 m;
NS, not significant.
A RT I C L E S
a
ECM organization
ECM disassembly
Periostin
3
Fold change
Platelet activation
Wound healing
Haemostasis
Collagen metabolic process
Blood vessel development
20
30
40
50
log (fold change)
Mf CM
KPC
Control
Anti-POSTN
KPC
Panc02
Healthy liver
Panc1
40
5
20
0
Mf CM:
Anti-POSTN:
20
0
+
+
+
Periostin IHC
KPC
Colonies per well
60
10
+
+
P < 0.001
25
20
15
10
5
0
HL LM
P = 0.023
P = 0.006
P < 0.001
60
80
15
P = 0.004
P = 0.002
P < 0.001
40
Liver metastasis
Periostin IHC
10
Human PDAC
ANXA2
GREM1
P4HB
THBS1
LUM
FBN1
DCN
SPARC
CTSD
CCDC80
COL5A1
BGN
PPIB
EFEMP1
POSTN
HSPG2
EFEMP2
BMP1
TIMP1
COL4A2
MFAP2
PLEC
CTGF
LTBP1
COL5A2
TIMP2
ADAM9
COL6A1
LAMB1
LAMC1
LAMA1
ACTN1
PXDN
COL3A1
CYR61
COL1A2
COL6A3
COL12A1
LTBP4
SERPINE1
GSN
COL4A1
COL1A1
LOXL2
CST3
MMP2
COL4A5
FBLN1
VTN
CD44
SERPINF2
ECM1
NID2
SERPINH1
MMP1
FN
A2M
TGFBI
Axon development
P < 0.001
4
3
2
1
0
HL LM
+
+
Figure 6 Myofibroblast-secreted periostin enhances pancreatic cancer cell

growth. (a) Top gene ontology (GO) functions of secreted proteins enriched
in human myofibroblasts (Mf) following in vitro education with macrophageconditioned media (CM). (b) Fold change rank of identified proteins
associated with ECM organization. Data were obtained from one experiment
assessing two biologically independent samples (green, periostin). (c) Colonyformation assay of primary murine KPC cells, murine Panc02, and human
Panc1 cells in the presence or absence of Mf CM and periostin-neutralizing
antibody (anti-periostin). Representative images reflecting colonies and
single cells are shown for KPC cells (n = 3 independent experiments;

mean s.e.m.; two-tailed unpaired t-test). (d) Immunohistochemical (IHC)
staining of periostin in human liver biopsies (n = 5 healthy subjects,
n = 5 PDAC; five fields assessed per sample; mean s.e.m.; two-tailed
unpaired t-test) and in spontaneous metastatic liver tumours collected
from KPC mice (n = 5 mice per condition; five fields assessed per sample,
mean s.e.m.; two-tailed unpaired t-test). Representative micrographs
and quantification of the data. HL, healthy liver; LM, liver metastasis.
Scale bars, 100 m.
studies on investigating whether MAMs are required for supporting

the metastatic growth of already disseminated cancer cells. To address
this question, we chemically depleted MAMs in vivo, using clodronate
liposomes36 (CLs). CL treatment was started at day 3 post intrasplenic
injection of KPC cells, a time point in which initial seeding of the
liver by KPC cells has already occurred (Fig. 2e and Supplementary
Fig. 5a,b). As expected, in response to CL treatment, MAM numbers
were significantly reduced (Fig. 2f and Supplementary Fig. 5c).
Interestingly, we could barely detect any SMA+ myofibroblasts in
metastatic livers of CL-treated mice (Fig. 2f and Supplementary
Fig. 5c). The ablation of MAMs and the prevention of SMA+
myofibroblast accumulation was not cell line specific because similar
results were observed using Panc02 cells (Supplementary Fig. 5d,e).
Although the metastatic frequency was only modestly affected
by CL treatment in both models, the size of the lesion area
covered by metastatic cells was significantly reduced in response
to macrophage depletion (Fig. 2g,h and Supplementary Fig. 5f,g).
Together, these results suggest that depletion of MAMs prevents the
activation of myofibroblasts and impairs the progression of metastatic

lesions even after initial colonization of the metastatic site by
cancer cells.
Granulin secreted by macrophages triggers myofibroblast
activation
We next sought to understand how MAMs regulate myofibroblast
activation at the metastatic site. In this respect, we found that
macrophage conditioned media (CM) acts as a strong activator
of quiescent primary fibroblasts in culture and enhances their
invasion and proliferation (Supplementary Fig. 6ac). To identify the
macrophage-derived factors responsible for myofibroblast activation,
we next exposed human THP-1 macrophages to the pancreatic cancer
cell line Panc1 CM in vitro, and performed a secretome analysis.
We identified several extracellular matrix (ECM) proteases associated
with macrophage function3739 (Supplementary Tables 1 and 2).
Among the most highly secreted proteins we identified granulin
(Grn), a secreted glycoprotein with a relative molecular mass of
555
A RT I C L E S
0
WT MCM
Grn/ MCM
rGranulin
DAPI/Periostin
+
+
P < 0.001
WT + WT BM
1.0
P = 0.001
1.0
0.5
0
WT MCM
Grn/ MCM
rGranulin
+
+
MTS
P < 0.001
P < 0.001
WT + WT BM
10
P < 0.001
0.6
0.4
WT + Grn/ BM
0.2
Connective tissue
6
4
2
co
T
W
G L
G rn
rn /
/
BM
PI ntro
(3
)K l
co
T
P < 0.001
P < 0.001
PI ntro
(3
)K l
WT + Grn/ BM
Periostin/DAPI (a.u.)
P < 0.001
0.8
BM
+
+
1.5
G L
G rn
rn /
/
BM
P = 0.01
2.0
Fold change
Postn mRNA levels
MAM
P = 0.008
10 P < 0.001
Fold change
Postn mRNA levels
Fold change
Postn mRNA levels
3 P = 0.003
0
WT MCM
Grn/ MCM
rGranulin
MAM
P = 0.001
W
G T
rn
W /
G TB
rn M
BM macrophage
P = 0.004
Periostin (ng ml1)
Figure 7 Macrophage-derived granulin induces periostin expression by

hepatic stellate cells in vitro and in vivo. (a) Evaluation of periostin
(Postn) mRNA expression levels in primary hStCs following stimulation with
CM collected from BM WT or Grn/ macrophages in the presence or
absence of recombinant granulin (rGranulin) (n = 3 independent experiments;
mean s.e.m.; two-tailed unpaired t-test). (b,c) Evaluation of Postn mRNA
levels by qPCR (b) and periostin protein levels by ELISA (c) in primary
hStCs following stimulation with CM collected from in vivo-derived WT
or Grn/ MAMs, in the presence or absence of recombinant granulin
(rGranulin) (n = 3 independent experiments; mean s.e.m.; two-tailed
unpaired t-test). (d,e) Evaluation of periostin deposition and fibrotic stroma
formation in metastatic livers of control WT mice, WT mice treated with
clodronate liposomes (CL), PI(3)K/ , Grn/ , and WT + Grn/ BM mice

12 days after intrasplenic implantation of KPC cells. (d) Representative
immunofluorescence staining and quantification of periostin deposition.
Nuclei were counterstained with DAPI. (e) Representative Massons
trichrome staining (MTS) and quantification of area occupied by fibrotic
stroma (n = 4 mice per condition; four fields assessed per sample; data
combine five independent experiments; mean s.e.m.; two-tailed unpaired
t-test). (f) Quantification of Postn mRNA expression levels by qPCR in
intrametastatic myofibroblasts isolated from tumour-bearing livers of WT
and Grn/ mice, and chimaeric WT + WT BM and WT + Grn/ BM
mice (data are from six pooled mice per condition; one experiment).
Scale bars, 100 m.
approximately 70,000 that has previously been shown to mediate

wound healing, by stimulating fibroblast migration40, and to induce
fibrosis in breast cancer41. As the main source of myofibroblasts in
the liver is activated resident hStCs (Supplementary Fig. 3j)16, we
next isolated primary hStCs from mice and stimulated them with
CM generated from primary murine WT BM-derived macrophages
or Grn/ (granulin deficient) murine BM macrophages (BMMs). We
found that CM generated from WT, but not from Grn/ , BMMs
efficiently activated isogenic hStCs and promoted their migration
(Fig. 3a,b). Importantly, addition of recombinant granulin to Grn/
BMM CM was sufficient to restore hStC activation and migration
(Fig. 3a,b). Next, we isolated MAMs from tumour-bearing livers
derived from chimaeric WT + WT BM and WT + Grn/ BM
mice, and prepared MAM CM. We confirmed that only CM from
WT MAMs efficiently induced hStC activation and migration,
whereas CM from Grn/ MAM was unable to induce activation or

migration of hStCs (Fig. 3ce). In addition, we found that granulin
expression and secretion is increased in tumour-educated (Fig. 3f,g)
and alternatively (M2-like) activated macrophages (Supplementary
Fig. 6d,e). Together, these data indicate that cancer-cell-derived factors
induce the secretion of granulin in macrophages and that MAMs
activate resident hStCs through granulin.
556
Granulin is highly expressed in pancreatic cancer metastatic

lesions of mice and humans and MAMs are the main source of
granulin in vivo.
In vivo, we found that within the metastatic tumour microenvironment, granulin is highly expressed in MAMs, and that
MAMs are the main source of granulin secretion in metastatic
lesions (Fig. 4a,b). When we analysed chimaeric mice harbouring
A RT I C L E S
IMs
RMs
P = 0.005
Periostin
Fold change
Grn mRNA levels
(relative to healthy)
105
CD14
104
103
6
4
2
0
103 104
CD16
Promoting
metastatic
growth
Granulin
105
HSs
Failing
metastatic
growth
Macrophage
function
PDAC patients
d
Percentage of monocytes
Healthy subjects
PDAC patients
80
60
NS
40
20
Fold change
Grn mRNA levels
(relative to healthy)
3
P = 0.016
100
Inhibition of recruitment or
depletion of granulin
2
Inflammatory monocytes
Metastasis-associated macrophage
Myofibroblasts
Pancreatic cancer cells

0
IMs
RMs
Healthy
KPC
Figure 8 Metastatic PDAC patients have increased circulating inflammatory

monocytes that express high levels of granulin. (a) Peripheral mononuclear
cells were isolated from healthy subjects and metastatic PDAC patients.
Representative dot plot of inflammatory monocytes (IMs; CD14hi CD16neg )
and resident monocytes (RMs; CD14dim CD16hi ) after gating for the
CD45+ CD3neg B220neg CD19neg Sytoxneg cell population (data are from 6
different PDAC patients and 6 different healthy subjects). (b) Quantification
of a. Percentage of monocyte populations in healthy subjects and metastatic
PDAC patients (n = 6 different healthy; n = 6 different PDAC samples;
individual data points, horizontal lines represent mean s.e.m.; twotailed unpaired t-test). NS, not significant. (c) Quantification of Grn mRNA
levels in IMs isolated from metastatic PDAC patients and healthy subjects
(HSs) as described in a,b (n = 3 different healthy, n = 4 different PDAC
samples; mean s.e.m.; two-tailed unpaired t-test). (d) Quantification of
Grn mRNA levels in circulating IMs sorted from KPC mice with pathological
confirmed liver metastasis or tumour-free littermates (data are from four
pooled mice per condition; one experiment). (e) Schematic depicting the
role of macrophage-derived granulin in activation of hStCs and in PDAC
liver metastasis. Inflammatory monocytes are recruited to the liver by
metastatic pancreatic tumour cells through a PI(3)K-dependent mechanism.
Once in the metastatic tissue, differentiated macrophages stimulate the
activation and recruitment of resident hStCs through granulin secretion,
resulting in excessive accumulation of myofibroblasts. Granulin-induced
myofibroblasts release high levels of the ECM protein periostin, thereby
enhancing survival and growth of metastatic pancreatic cancer cells in a
hostile environment. Interruption of this sequence, by either preventing
macrophage accumulation or by abolishing granulin expression in recruited
macrophages, limits metastatic growth of pancreatic cancer cells.
tdTomatoRed BM, we found that BM-derived MAMs, but not resident

KCs, express high levels of granulin (Grn) (Fig. 4c). Surprisingly,
expression levels of Tgfb, a common activator of hStCs42, were low in
MAMs and KCs (Fig. 4c).
Finally, we further confirmed that granulin is highly expressed
in the stroma of spontaneous hepatic metastatic lesions of human
and murine tissues24,43 (Fig. 4d,e and Supplementary Fig. 6f). Taken
together, our findings indicate that granulin is highly expressed by
MAMs and that MAMs are the main source of granulin in PDAC
liver metastasis.
hepatic metastatic tumour burden (Fig. 5a,c). Whereas depletion of

granulin did not reduce the total number of MAMs (Fig. 5d), nor their
polarization at the metastatic site (Supplementary Fig. 7a), granulin
deficiency abolished the accumulation of activated myofibroblasts
in metastatic liver lesions (Fig. 5d). Quantitative gene expression
analysis of myofibroblast activation markers (Acta2, Fn, Col1a1)16
further confirmed that lack of granulin prevents the activation of
intrametastatic hStCs into myofibroblasts in vivo (Fig. 5e).
Next, we characterized metastatic pancreatic tumour growth in
WT mice transplanted with BM from Grn/ or WT animals. We
found that depletion of granulin in the haematopoietic compartment
was sufficient to suppress metastatic growth of KPC cells (Fig. 5f,g)
and to prevent myofibroblast activation in vivo (Fig. 5h,i), while
MAM numbers and polarization remained unchanged (Fig. 5j and
Supplementary Fig. 7b). Interestingly, ablation of granulin in the
haematopoietic compartment markedly reduced the number of
proliferating (Ki67+ ) metastatic pancreatic tumour cells (Fig. 5k),
but apoptotic (cleaved caspase 3+ ) tumour cell numbers remained
unchanged (Fig. 5l). Taken together, these results suggest that granulin
secretion by MAMs is required for hStCs activation and for metastatic
growth of pancreatic tumours in the liver.
Depletion of granulin prevents liver fibrosis and PDAC

metastatic growth
To investigate whether granulin is required for PDAC metastatic
progression in vivo, we next inoculated Grn/ and WT mice
with isogenic KPC cells through intrasplenic implantation. We
observed that lack of granulin expression did not alter metastatic
frequency because WT and Grn/ mice showed similar numbers
of metastatic nodules (Fig. 5a,b). However, the area covered by
metastatic tumour cells in Grn/ mice was significantly smaller
compared with WT, resulting in an overall significant decrease of
557
A RT I C L E S
Granulin induces periostin expression in hStCs thereby allowing
tumour growth
Next, we sought to gain a better understanding of how MAM-induced
myofibroblast activation promotes pancreatic cancer cell growth.
To address this question, we stimulated human fibroblasts with
macrophage CM and performed a mass spectrometry quantitative
analysis of the secretome from fibroblasts exposed to macrophage
CM compared with unexposed fibroblasts. We found that macrophage
CM induces fibroblast secretion of proteins associated with ECM
remodelling (Fig. 6a and Supplementary Table 3), and in particular
the secretion of the ECM component periostin (Fig. 6b). Periostin
has been reported to enhance metastatic growth of breast and
colon cancer cells through activation of Wnt and v 3 Akt/PKB
signalling pathways, respectively8,44. Thus, we next tested whether
periostin expression by activated myofibroblasts is necessary to
promote pancreatic cancer cell survival and growth. We found
that myofibroblast CM markedly promoted colony formation and
proliferation of PDAC cancer cells, and that addition of a periostinneutralizing antibody completely abolished these effects (Fig. 6c
and Supplementary Fig. 7c). We confirmed that periostin acts in a
paracrine myofibroblasttumour cell loop because we did not detect
any periostin expression in PDAC cells (Supplementary Fig. 7d).
Moreover, we found that periostin was markedly upregulated in
spontaneous hepatic metastatic lesions of human and murine tissues
compared with healthy control livers (Fig. 6d and Supplementary
Fig. 7e). Importantly, the induction of periostin expression in
primary hStCs was strictly dependent on granulin, because CM
generated from Grn/ BMM (Fig. 7a) and MAMs (Fig. 7b,c) was
unable to induce periostin expression in hStCs. To examine whether
inhibition of myofibroblast activation by MAM-derived granulin
also affects periostin expression and stroma expansion in vivo, we
analysed metastatic liver tissue sections of WT, PI(3)K/ , Grn/ ,
WT + Grn/ BM, and WT + CL-treated mice. In all models,
hepatic periostin expression levels and connective tissue deposition
were found to be significantly reduced compared with metastasisbearing WT control animals (Fig. 7d,e and Supplementary Fig. 7f,g).
In agreement with these findings, we found that intrametastatic
myofibroblasts isolated from Grn/ and WT + Grn/ BM mice
showed an approximately threefold reduction in periostin expression
relative to control myofibroblasts isolated from WT tumour lesions
(Fig. 7f). Taken together, these findings indicate that MAMs induce the
expression of ECM proteins in myofibroblasts, particularly periostin,
in a granulin-dependent manner which then supports metastatic
pancreatic tumour growth.
Although less frequently, PDAC can also sometimes metastasize
to the lung1. Similar to what we observed in the liver, granulin
was markedly expressed in the stromal compartment of pulmonary
metastatic lesions, especially in areas rich in CD68+ macrophages
(Supplementary Fig. 8a). Depletion of granulin in the haematopoietic
compartment (WT + Grn/ BM) was sufficient to reduce pulmonary
metastatic growth of PDAC cells, and to prevent myofibroblast
activation, fibrotic stromal expansion, and periostin deposition
in vivo, while MAM numbers remained unchanged (Supplementary
Fig. 8be). To further explore whether this mechanism is specific
for the metastatic site, we next analysed primary PDAC tumours for
granulin and periostin expression. Interestingly, we found that at the
558
primary tumour site granulin is mainly expressed by tumour cells,

and not by macrophages, while periostin was strongly expressed in the
surrounding stromal compartment (Supplementary Fig. 8f).
Taken together, these data suggest that granulin-expressing
macrophages play a critical role at the metastatic site, but not at the
primary tumour site.
Circulating IMs from metastatic PDAC patients express granulin
As granulin-expressing MAMs originate from the BM, we next
investigated whether circulating IMs express granulin. Therefore, we
collected fresh blood samples from PDAC patients with metastases,
but before therapeutic intervention, and from healthy control
subjects, and purified IMs (CD14hi CD16neg ) and resident monocytes
(CD14dim CD16hi ) by fluorescence-activating cell sorting45. We found
that the percentage of IMs isolated from metastatic PDAC patients
was significantly higher compared with healthy subjects, while no
differences were observed in the percentage of resident monocytes
(Fig. 8a,b and Supplementary Fig. 8g). Importantly, IMs isolated from
PDAC patients and from metastatic KPC mice exhibited abnormally
elevated granulin expression levels compared with healthy control
subjects (Fig. 8c,d).
DISCUSSION
In this study, we aimed to gain a better understanding of the
role of macrophages in pancreatic cancer metastasis, with the hope
to improve therapies for this devastating disease. In this respect,
we found that MAMs support pancreatic cancer metastasis by
secreting granulin, which consequently promotes the activation of
resident hStCs into SMA+ myofibroblasts that secrete high levels of
periostin (Fig. 8e). Collectively, our studies provide a comprehensive
functional analysis of stromatumour interaction in PDAC metastasis,
and support the rationale for the development of therapeutic
approaches targeting stromal secreted pro-metastatic factors such
as granulin and periostin to disrupt the interrelationship between
MAMs, myofibroblasts and cancer cells. In addition, we observed
an increase of granulin-expressing IMs in blood from metastatic
PDAC patients and metastatic KPC mice compared with healthy
subjects. This observation suggests that granulin could also potentially
be used as a predictive biomarker of PDAC metastasis. However,
to assess the potential use of granulin as a predictive biomarker of
PDAC metastasis, granulin expression levels in IMs derived from
inflammatory, non-cancerous diseases, such as pancreatitis, would be
worth exploring in future analysis.
The fact that MAMs isolated from metastatic lesions secrete high
levels of granulin, but resident macrophages (KCs) do not (Fig. 4c),
indicates that the recruitment of monocyte-derived macrophages is a
critical step during PDAC metastasis. We and others have previously
shown that cancer cells release chemotactic factors, including CCL2,
SDF-1 and mCSF1, to attract pro-tumorigenic monocytes to the
tumour, in a PI(3)K-dependent manner11,35,4648. Consistent with
this, we found that prevention of macrophage recruitment resulted
in decreased hStCs activation, reduced periostin levels and decreased
metastasis to the liver.
Using an unbiased mass spectrometry secretome analysis of cancereducated macrophages, we identified the glycoprotein granulin.
Granulin has previously been associated with fibroblast activation
A RT I C L E S
and migration during wound healing40. In addition, in breast
cancer, granulin expression correlates with increased fibrosis and
poor survival5,41. In agreement with a role for granulin in fibrosis,
we found that macrophage-secreted granulin plays a critical role
in PDAC metastasis by activating resident hStCs and stimulating
the secretion of periostin. The precise signalling pathway by
which granulin activates hStCs and primary fibroblasts remains
unknown, as the cognate cell-surface receptor to which granulin
binds is still controversial4951. Interestingly, depletion of granulin
in the haematopoietic compartment did not change macrophage
recruitment to the liver, macrophage polarization, or CD8 + T cell
infiltration (Fig. 5 and Supplementary Fig. 7).
In agreement with ref. 52, we found that macrophages play a
key role in PDAC metastasis. Whereas ref. 52 describes a role for
resident hepatic macrophages in the establishment of a pre-metastatic
niche that facilitates initial tumour cells seeding, our work focuses
on understanding the subsequent step of metastatic growth. In this
respect, we found that once tumour cells have reached the metastatic
site, their survival and outgrowth capacity critically depends on the
further recruitment of IMs that secrete granulin.
Unfortunately, current imaging approaches are unable to detect
micro-metastases and by the time pancreatic cancer patients are
diagnosed, micro-metastatic spreading has already occurred in most
cases1. Our findings suggest that recruitment of IMs that express
granulin plays a key role in pancreatic cancer metastasis and may
serve both as a prognostic marker, and a potential target for PDAC
liver metastasis.

METHODS
Methods and any associated references are available in the online
version of the paper.
Note: Supplementary Information is available in the online version of the paper
ACKNOWLEDGEMENTS
We thank the flow cytometry and cell sorting facility and the animal facility at the
University of Liverpool for provision of equipment and technical assistance. We
are grateful to J. Y. Kim and J. S. Yoo at the Korea Basic Science Institute, Mass
Spectrometry Research Centre, for their assistance. We acknowledge the Liverpool
Tissue Bank for provision of tissue samples. We thank A. Taylor and P. Murray,
University of Liverpool, for technical help with lentiviral particle infection. We
thank L. Young, CRUK Cambridge Research Institute, for assistance with animal
models. We also thank the patients and their families, as well as the healthy blood
donors who contributed tissue samples and blood donations to these studies. These
studies were supported by grants from the Medical Research Council (grant number
MR/L000512/1) and the Pancreatic Cancer Research Fund (M.C.S.), a National
Institute for Health Research Biomedical Research Unit funding scheme through
a NIHR Pancreas BRU/Cancer Research UK PhD fellowship (S.R.N.), North West
Cancer Research (M.C.S. and V.Q.), and a Sir Henry Dale Fellowship jointly funded
by the Wellcome Trust and the Royal Society (A.M., grant number 102521/Z/13/Z).
AUTHOR CONTRIBUTIONS
S.R.N. designed and performed most of the experiments, analysed and interpreted
the data, and contributed to the preparation of the manuscript. V.Q. performed qPCR
experiments, immunofluorescence staining and immunoblotting. A.L. performed
human immune cell analysis and immunohistochemistry. V.Q., P.E., A.S. and
L.I. helped with in vivo experiments. C.R. performed flow cytometry and cell
sorting. A.S. performed immunohistochemistry. T.S. and K.S. provided primary
and immortalized hStCs. Y.-S.K. and J.H.K. performed proteomic analysis. D.E.
and D.A.T. provided primary murine KPC-derived pancreatic cancer cells. F.C.
helped with the analysis and interpretation of tumour biopsies. D.P. provided patient
samples. E.H. provided the PI(3)K/ mouse colony. A.M. provided conceptual
advice, designed experiments, and wrote the manuscript. M.C.S. conceived and
supervised the project, interpreted data, and wrote the manuscript. All authors
critically analysed and approved the manuscript.
COMPETING FINANCIAL INTERESTS
The authors declare no competing financial interests.
Published online at http://dx.doi.org/10.1038/ncb3340
Reprints and permissions information is available online at www.nature.com/reprints
1. Ryan, D. P., Hong, T. S. & Bardeesy, N. Pancreatic adenocarcinoma. N. Engl. J. Med.
371, 21402141 (2014).
2. Von Hoff, D. D. et al. Increased survival in pancreatic cancer with nab-paclitaxel plus
gemcitabine. N. Engl. J. Med. 369, 16911703 (2013).
3. Vanharanta, S. & Massague, J. Origins of metastatic traits. Cancer Cell 24,
410421 (2013).
4. Nguyen, D. X., Bos, P. D. & Massague, J. Metastasis: from dissemination to organspecific colonization. Nat. Rev. Cancer 9, 274284 (2009).
5. McAllister, S. S. & Weinberg, R. A. The tumour-induced systemic environment
as a critical regulator of cancer progression and metastasis. Nat. Cell Biol. 16,
717727 (2014).
6. Quail, D. F. & Joyce, J. A. Microenvironmental regulation of tumor progression and
metastasis. Nat. Med. 19, 14231437 (2013).
7. Erler, J. T. et al. Hypoxia-induced lysyl oxidase is a critical mediator of bone marrow
cell recruitment to form the premetastatic niche. Cancer Cell 15, 3544 (2009).
8. Malanchi, I. et al. Interactions between cancer stem cells and their niche govern
metastatic colonization. Nature 481, 8589 (2012).
9. Kitamura, T., Qian, B. Z. & Pollard, J. W. Immune cell promotion of metastasis. Nat.
Rev. Immunol. 15, 7386 (2015).
10. Sevenich, L. et al. Analysis of tumour- and stroma-supplied proteolytic networks
reveals a brain-metastasis-promoting role for cathepsin S. Nat. Cell Biol. 16,
876888 (2014).
11. Qian, B. Z. et al. CCL2 recruits inflammatory monocytes to facilitate breast-tumour
metastasis. Nature 475, 222225 (2011).
12. Qian, B. Z. & Pollard, J. W. Macrophage diversity enhances tumor progression and
metastasis. Cell 141, 3951 (2010).
13. Kitamura, T. et al. CCL2-induced chemokine cascade promotes breast cancer
metastasis by enhancing retention of metastasis-associated macrophages. J. Exp.
Med. 212, 10431059 (2015).
14. Kim, S. et al. Carcinoma-produced factors activate myeloid cells through TLR2 to
stimulate metastasis. Nature 457, 102106 (2009).
15. Ohlund, D., Elyada, E. & Tuveson, D. Fibroblast heterogeneity in the cancer wound.
J. Exp. Med. 211, 15031523 (2014).
16. Pellicoro, A., Ramachandran, P., Iredale, J. P. & Fallowfield, J. A. Liver fibrosis and
repair: immune regulation of wound healing in a solid organ. Nat. Rev. Immunol. 14,
181194 (2014).
17. Olive, K. P. et al. Inhibition of Hedgehog signaling enhances delivery of chemotherapy
in a mouse model of pancreatic cancer. Science 324, 14571461 (2009).
18. Sherman, M. H. et al. Vitamin d receptor-mediated stromal reprogramming
suppresses pancreatitis and enhances pancreatic cancer therapy. Cell 159,
8093 (2014).
19. Erez, N., Truitt, M., Olson, P., Arron, S. T. & Hanahan, D. Cancer-associated
fibroblasts are activated in incipient neoplasia to orchestrate tumor-promoting
inflammation in an NF-B-dependent manner. Cancer Cell 17, 135147 (2010).
20. Ozdemir, B. C. et al. Depletion of carcinoma-associated fibroblasts and fibrosis
induces immunosuppression and accelerates pancreas cancer with reduced survival.
Cancer Cell 25, 719734 (2014).
21. Rhim, A. D. et al. Stromal elements act to restrain, rather than support, pancreatic
ductal adenocarcinoma. Cancer Cell 25, 735747 (2014).
22. Neoptolemos, J. P. A randomized trial of chemoradiotherapy and chemotherapy after
resection of pancreatic cancer. N. Engl. J. Med. 350, 12001210 (2004).
23. Sperti, C., Pasquali, C., Piccoli, A. & Pedrazzoli, S. Recurrence after resection for
ductal adenocarcinoma of the pancreas. World J. Surg. 21, 195200 (1997).
24. Hingorani, S. R. et al. Trp53R172H and KrasG12D cooperate to promote
chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in
mice. Cancer Cell 7, 469483 (2005).
25. Lim, C. et al. Tumour progression and liver regenerationinsights from animal models.
Nat. Rev. Gastroenterol. Hepatol. 10, 452462 (2013).
26. Little, E. C. et al. Novel immunocompetent murine models representing advanced
local and metastatic pancreatic cancer. J. Surg. Res. 176, 359366 (2012).
27. Geissmann, F. et al. Development of monocytes, macrophages, and dendritic cells.
Science 327, 656661 (2010).
28. Hashimoto, D. et al. Tissue-resident macrophages self-maintain locally throughout
adult life with minimal contribution from circulating monocytes. Immunity 38,
792804 (2013).
29. Yona, S. et al. Fate mapping reveals origins and dynamics of monocytes and tissue
macrophages under homeostasis. Immunity 38, 7991 (2013).
30. Schulz, C. et al. A lineage of myeloid cells independent of Myb and hematopoietic
stem cells. Science 336, 8690 (2012).
31. Klein, I. et al. Kupffer cell heterogeneity: functional properties of bone marrow
derived and sessile hepatic macrophages. Blood 110, 40774085 (2007).
559
A RT I C L E S
32. Corbett, T. H. et al. Induction and chemotherapeutic response of two transplantable
ductal adenocarcinomas of the pancreas in C57BL/6 mice. Cancer Res. 44,
717726 (1984).
33. Hirsch, E. et al. Central role for G protein-coupled phosphoinositide 3-kinase in
inflammation. Science 287, 10491053 (2000).
34. Camps, M. et al. Blockade of PI3K suppresses joint inflammation and damage in
mouse models of rheumatoid arthritis. Nat. Med. 11, 936943 (2005).
35. Schmid, M. C. et al. Receptor tyrosine kinases and TLR/IL1Rs unexpectedly activate
myeloid cell PI3k, a single convergent point promoting tumor inflammation and
progression. Cancer Cell 19, 715727 (2011).
36. van Rooijen, N. & van Nieuwmegen, R. Elimination of phagocytic cells in the
spleen after intravenous injection of liposome-encapsulated dichloromethylene
diphosphonate. An enzyme-histochemical study. Cell Tissue Res. 238,
355358 (1984).
37. Shree, T. et al. Macrophages and cathepsin proteases blunt chemotherapeutic
response in breast cancer. Genes Dev. 25, 24652479 (2011).
38. Coussens, L. M., Tinkle, C. L., Hanahan, D. & Werb, Z. MMP-9 supplied by bone
marrow-derived cells contributes to skin carcinogenesis. Cell 103, 481490 (2000).
39. Hildenbrand, R., Wolf, G., Bohme, B., Bleyl, U. & Steinborn, A. Urokinase
plasminogen activator receptor (CD87) expression of tumor-associated macrophages
in ductal carcinoma in situ, breast cancer, and resident macrophages of normal breast
tissue. J. Leukoc. Biol. 66, 4049 (1999).
40. He, Z., Ong, C. H., Halper, J. & Bateman, A. Progranulin is a mediator of the wound
response. Nat. Med. 9, 225229 (2003).
41. Elkabets, M. et al. Human tumors instigate granulin-expressing hematopoietic cells
that promote malignancy by activating stromal fibroblasts in mice. J. Clin. Invest.
121, 784799 (2011).
560
42. Hellerbrand, C., Stefanovic, B., Giordano, F., Burchardt, E. R. & Brenner, D. A. The
role of TGF1 in initiating hepatic stellate cell activation in vivo. J. Hepatol. 30,
7787 (1999).
43. Gidekel Friedlander, S. Y. et al. Context-dependent transformation of adult pancreatic
cells by oncogenic K-Ras. Cancer Cell 16, 379389 (2009).
44. Bao, S. et al. Periostin potently promotes metastatic growth of colon cancer by
augmenting cell survival via the Akt/PKB pathway. Cancer Cell 5, 329339 (2004).
45. Gordon, S. & Taylor, P. R. Monocyte and macrophage heterogeneity. Nat. Rev.
Immunol. 5, 953964 (2005).
46. DeNardo, D. G. et al. Leukocyte complexity predicts breast cancer survival
and functionally regulates response to chemotherapy. Cancer Discov. 1,
5467 (2011).
47. Pyonteck, S. M. et al. CSF-1R inhibition alters macrophage polarization and blocks
glioma progression. Nat. Med. 19, 12641272 (2013).
48. Schmid, M. C. et al. Combined blockade of integrin-41 plus cytokines SDF1 or IL-1 potently inhibits tumor inflammation and growth. Cancer Res. 71,
69656975 (2011).
49. Etemadi, N., Webb, A., Bankovacki, A., Silke, J. & Nachbur, U. Progranulin does
not inhibit TNF and lymphotoxin- signalling through TNF receptor 1. Immunol. Cell
Biol. 91, 661664 (2013).
50. Tang, W. et al. The growth factor progranulin binds to TNF receptors and is therapeutic
against inflammatory arthritis in mice. Science 332, 478484 (2011).
51. Chen, X. et al. Progranulin does not bind tumor necrosis factor (TNF) receptors and is
not a direct regulator of TNF-dependent signaling or bioactivity in immune or neuronal
cells. J. Neurosci. 33, 92029213 (2013).
52. Costa-Silva, B. et al. Pancreatic cancer exosomes initiate pre-metastatic niche
formation in the liver. Nat. Cell Biol. 17, 816826 (2015).
METHODS
DOI: 10.1038/ncb3340
METHODS
Cells. Murine pancreatic cancer cells KPC FC1199, referred to as KPC, were
generated in the Tuveson laboratory (Cold Spring Harbor Laboratory, New York,
USA) isolated from PDA tumour tissues obtained from KrasG12D/+ ; p53R172H/+ ; Pdx1
Cre mice of a pure C57BL/6 background as described previously with minor
modifications24. For some experiments, KPCluc/zsGreen cells were generated by using
pHIV LuczsGreen (gift from B. Welm, University of Utah, USA, Addgene plasmid
no. 39196) lentiviral particle infection. Infected cells were selected for high zsGreen
expression levels using a flow cytometry cell sorter (ARIA, BD).
Primary murine hepatic stellate cells were isolated from murine livers of 1012week-old female C57BL/6 mice according to previously reported protocols53. The
murine immortalized hepatic stellate cell line from livers of C57BL/6 mice was
provided by T. Sakais laboratory, University of Liverpool, UK53. Primary murine
dermal fibroblasts were generated as previously described19. All fibroblasts were used
at an early passage (passage 24).
Human primary pancreatic fibroblasts were acquired from Vitro Biopharma,
human peripheral blood monocytic cell line THP-1 and human pancreatic
cancer cell line PANC1 were acquired from ATCC, and murine C57BL/6 Panc02
pancreatic ductal carcinoma cells were obtained from the NCI DCTD Tumor
Repository, NIH.
All cells were cultured in DMEM + 10% FBS and supplemented with 100 units of
penicillin, 100 g ml1 streptomycin and 2.5 g ml1 amphotericin B, except human
primary macrophages and THP-1 cells, which were cultured in RPMI + 10% FBS
and supplemented with penicillin/streptomycin and amphotericin B.
Primary murine macrophages were generated by flushing the bone marrow
from the femur and tibia of C57BL/6 mice followed by incubation for five days
in DMEM containing 10% FBS and 10 ng ml1 murine M-CSF (Peprotech).
For some experiments, differentiated macrophages were polarized into an M1,
M2 or tumour-educated phenotype using INF (20 ng ml1 , Peprotech) and
LPS (100 ng ml1 , Sigma-Aldrich); IL-4 (20 ng ml1 , Peprotech); or tumourconditioned media, respectively. Primary human macrophages were generated
by purifying CD14+ monocytes from human total blood followed by incubation
for five days in RPMI containing 10% FBS and 50 ng ml1 recombinant human
M-CSF (Peprotech). THP-1 cells were incubated with 50 nM phorbol 12myristate 13-acetate for 72 h in RPMI + 10% FBS to generate THP-1-derived
macrophages. For some experiments, differentiated primary macrophages
or THP-1-derived human macrophages were polarized into an M1, M2 or
tumour-educated phenotype using INF (50 ng ml1 , Peprotech) and LPS
(10 ng ml1 , Sigma-Aldrich); IL-4 (40 ng ml1 , Peprotech); or tumour-conditioned
media, respectively.
All cells were routinely tested negative for the presence of mycoplasma
contamination. None of the cell lines used in this manuscript is listed in the ICLAC
and NCBI Biosample database of misidentified cell lines. Cell lines have been directly
purchased from ATTC and NCI DTCP respectively, as described above.
Mice. We obtained transgenic mice lacking PI(3)-kinase expression (p110/ )
on the C57BL/6 background33 from E. Hirsch, Institute of Molecular Biotechnology
Center, University of Torino, Italy. Grn/ mice (B6(Cg)Grntm1.1Aidi /J) and
tdTomatoRed mice (B6.129(Cg)Gt(ROSA)26Sortm4(ACTBtdTomato,EGFP)Luo /J) both on
the C57BL/6 genetic background were purchased from The Jackson Laboratory.
C57BL/6 mice were purchased from Charles River. KrasG12D/+ ; p53R172H/+ ; Pdx1Cre
mice24 were purchased from CRUK, Cambridge Research Institute, Cambridge.
All animal experiments were performed in accordance with current UK
legislation under an approved project licence PPL 40/3725 (M. Schmid). Mice
were housed under specific pathogen-free conditions at the Biomedical Science
Unit at the University of Liverpool. In general, for animal studies the group size
was calculated by power analysis using a significance level kept at 5% and the
power at 80% (according to approved corresponding Home Office Project License
Application). For tumour studies, female animals aged 68 weeks were used. Animals
were randomly assigned to experimental groups. The investigators were not blinded
to allocation during experiments and outcome assessments.
Bone marrow transplantation. Bone marrow transplantation was performed by
reconstituting the bone marrow of lethally irradiated (10 Gy) female, 6-week-old
C57BL6 mice by tail vein injection of 5 106 total bone marrow cells isolated
from tdTomatoRed mice or Grn/ mice35. After 4 weeks, successful engraftment
of tdTomatoRed bone marrow was assessed by flow cytometry. WT recipient mice
reconstituted with tdTomatoRed bone marrow showed exclusively tdTomatoRed+
haematopoietic cells, whereas WT recipient mice reconstituted with unlabelled WT
bone marrow did not show any tdTomatoRed+ signal. After 4 weeks, engraftment of
Grn/ bone marrow was assessed by genomic DNA PCR according to The Jackson
Laboratory protocol on peripheral blood cells from fully recovered bone-marrowtransplanted mice. After confirmation of successful bone marrow reconstitution,
mice were enrolled in tumour studies.
Metastasis studies. Experimental liver metastasis was performed by implanting

1 106 KPC cells or 1 106 Panc02, respectively, unless otherwise stated, in
20 l PBS into the spleen of immunocompetent isogenic C57BL/6 mice using a
Hamilton 29 G Syringe as previously described32,52. Experimental lung metastasis
was performed by injecting 5 105 KPC cells in 200 l PBS into the tail vein of
immunocompetent isogenic C57BL/6 mice using an insulin syringe. At the indicated
time points, mice were euthanized and metastatic tumour burden was assessed
by quantifying the frequency and size of metastatic lesions in haematoxylin and
eosin-stained paraffin-embedded liver or lung sections by microscopy using ZEN
imaging software.
Clodronate liposome treatment. Macrophages were depleted in vivo by
intraperitoneal injections of 200 l PBS liposomes (Control) or clodronate
liposomes (5 mg of clodronate per 1 ml) from ClodronateLiposomes.com54.
Flow cytometry. Single-cell suspensions from murine livers were prepared by
mechanical and enzymatic disruption in Hanks balanced salt solution (HBSS) with
1 mg ml1 collagenase P (Roche) as previously described55 with some modifications.
Briefly, cells in suspension were centrifuged for 5 min at 1,200 r.p.m., resuspended in
HBSS and filtered through a 500 m polypropylene mesh (Spectrum Laboratories).
The cell suspension was resuspended in 1 ml 0.05% trypsin and incubated at 37 C
for 5 min. Cells were filtered through a 70 m cell strainer and resuspended in PBS
+ 5% BSA. Cells were blocked for 10 min on ice with FC Block (BD Pharmingen,
Clone 2.4G2) and then stained with Sytox viability marker (Life Technologies)
and fluorochrome-conjugated antibodies (Supplementary Table 5). Flow cytometry
was performed on a FACSCanto II (BD Biosciences) and FACS performed on a
FACSAria (BD Biosciences). For FACS experiments, cells were sorted directly into
RLT buffer + -mercaptoethanol according to the manufacturers instructions for
RNA isolation.
Magnetic bead isolation of cells. Samples for magnetic bead isolation were prepared
from livers as described above for preparation of flow cytometry samples. Then
samples were stained and PDGFR+ or F4/80+ cells were isolated according to the
manufacturers instructions (Miltenyi).
Conditioned medium preparation. Conditioned medium from all cells were
generated according to previous reports56. Briefly, the medium was removed from
70% confluent cells and the cells were washed three times with PBS before
addition of serum-free medium. Cells were incubated for 1824 h in serum-free
medium and then collected and filtered through 0.45 m filters before use or stored
at 20 C.
Proteomics. To analyse the secretome of cancer-educated macrophages, THP-1derived human macrophages were stimulated for 24 h with conditioned medium
from the human pancreatic cancer cell line PANC1. Cells were then washed
three times with PBS and incubated for 24 h in serum-free DMEM. After 24 h of
incubation, over 100 ml of cell culture supernatant was collected and filtered through
0.45 m filters for mass spectrometry. The samples were spiked with 10 g of human
serum albumin (Sigma-Aldrich) as an internal reference and all proteins in each
sample were concentrated using MWCO 3kDa Amicon Ultra Centrifugal Filtration
Tubes (Millipore) according to the manufacturers instructions. The protein mixtures
were reduced, alkylated, and tryptic-digested as described previously57. Each sample
was individually labelled with a TMT kit (Thermo Scientific), and then aqueous
hydroxylamine solution (5% w/v) was added to quench the reaction. After being
combined, labelled peptide samples were vacuum dried, and then dissolved in
0.1% formic acid for LCMS/MS analysis. The labelled samples were analysed
using a two-dimensional LCMS/MS system consisting of a nanoACQUITY
UltraPerformance LC System (Waters) and an LTQ Orbitrap Elite mass spectrometer
(Thermo Scientific) equipped with a nano-electrospray source57. Briefly, 5 l aliquots
of samples were resolved on a strong cation exchange (5 m, 3 cm) column,
trapped on a C18 trap column, and then separated on a 200 mm microcapillary
C18 column (particle size 3 m) packed into 100 m silica tubing. During the
chromatographic separation, the mass spectrometer was operated in a datadependent mode at the voltage of 2 kV with acquisition parameters as follows: five
data-dependent CIDHCD dual MS/MS scans per full scan; CID scans acquired
with two-microscan averaging; full scans and HCD scans acquired at resolution
60,000 and 15,000 respectively, with two-microscan averaging; 35% normalized
collision energy (NCE) in CID and 45% NCE in HCD; 1 Da isolation window.
Previously fragmented ions were excluded for 60 s. Each selected parent ion was
first fragmented by CID and then by HCD. MS/MS spectra were analysed against
the Uniprot human database (released 18 March 2013). ProLucid58 was used to
identify the peptides with a precursor mass error of 25 ppm and a fragment ion
mass error of 600 ppm. TMT modification (+229.1629), carbamidomethylation
at cysteine, and oxidation at methionine were considered. For better peptide
NATURE CELL BIOLOGY
METHODS
DOI: 10.1038/ncb3340
identifications and quantifications, six reporter ions were extracted from the HCD
spectrum within 0.02 mass tolerance using home-made software and inserted
into the corresponding CID spectrum with the same precursor ions59. The output
data files were filtered and sorted to compose the protein list using DTASelect60
with 2 peptides assignments for a protein identification and a false positive
rate 0.01.
For quantitative analysis, Census in IP2 pipeline (Integrated Proteomics) was
used. Protein abundance was calculated from the average of a reporter ions
intensities from all constituent peptides from a protein61. Two different TMT
reagents were used to control the quality in quantification, and data with 30%
variation were excluded. Relative protein abundance was calculated from the ratios
of reporter ion intensities.
To identify secreted proteins in the THP-1 macrophages, the full list of identified
proteins was analysed using the gene ontology consortium web page for the GO
term Extracellular Vesicular Exosome (Gene ontology: tool for the unification
of biology62) and then referenced to the Uniprot database (www.uniprot.org)
for confirmation.
Dermal fibroblast/hepatic stellate cell activation. Primary PDGFR+ murine
dermal fibroblasts or hepatic stellate cells were isolated as previously described19,53.
On the day before stimulation, dermal fibroblasts were washed three times
with PBS and then incubated overnight in DMEM + 1%FBS. The next
morning, dermal fibroblasts were stimulated with macrophage-conditioned medium
(supplemented with 1% FBS) or DMEM + 1%FBS as a control and incubated
for 24 h.
Murine hepatic stellate cells were washed three times with PBS and then
incubated overnight in DMEM + 2%FBS. The next morning, hepatic stellate cells
were stimulated with macrophage-conditioned medium (supplemented with 2%
FBS) from WT or Grn/ macrophages or DMEM + 2%FBS as a control and
incubated for 24 h. To rescue the phenotype from Grn/ macrophage-conditioned
medium, Grn/ MCM were supplemented with recombinant murine progranulin
(R&D Systems, 2557-PG-050) at 1 g ml1 for activation experiments.
ELISA. Hepatic stellate cells were stimulated with macrophage-conditioned medium
(supplemented with 2% FBS) from WT or Grn/ macrophages or DMEM + 2%FBS
as a control and incubated for 24 h. After 24 h, the hepatic stellate cells were washed
three times with pre-warmed PBS and then incubated for 48 h in serum- and
phenol red-free DMEM. After 48 h, the medium was collected and used for ELISA
to measure periostin according to the manufacturers instructions (R&D Systems,
MOSF20). For murine macrophages, cells were stimulated with KPC- or Panc02conditioned medium for 24 h, then washed three times with PBS and incubated
in serum- and phenol red-free DMEM. The conditioned medium was collected
after 48 h and used for ELISA to measure granulin according to the manufacturers
instructions (LS Bio, LS-F284)
Transwell migration assay. Transwell inserts with 8 m pore size (Corning) were
coated with 50 l growth factor-reduced Matrigel (BD Biosciences) and allowed to
solidify. Macrophages were seeded in DMEM + 10%FBS in a 24-well plate. Growth
medium was removed, cells were washed three times with PBS, and DMEM + 1%
FBS was added. Next, fibroblasts were added in DMEM + 1% FBS into the inserts
and allowed to migrate through the insert membrane for 48 h. The inserts were
removed, and the inside was swabbed thoroughly with cotton swabs and then fixed
in methanol containing 0.05% crystal violet. The number of migrated fibroblasts was
counted by bright field microscopy. For Transwell migration of hepatic stellate cells,
the hepatic stellate cells were labelled with Vybrant Dil (Life Technology, emission
565) according to the manufacturers protocol before addition to the inserts. Relative
migration was evaluated by the amount of pixels from each condition relative to the
pixels from the control using Zen Software.
Proliferation assay. Human primary pancreatic fibroblasts were seeded at a
concentration of 2,000 cells per well in 96-well plates in DMEM + 10% FBS.
Cells were washed three times with PBS and then stimulated with conditioned
medium (supplemented with +1% FBS) prepared from primary human or THP-1
macrophages or DMEM + 1% FBS as a control.
Cell viability was determined by measuring the conversion of water-soluble 3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to an insoluble
formazan. The concentration of formazan was determined by optical density
at 595 nm.
KPC cells were seeded at a concentration of 2,000 cells per well in 96-well
plates in DMEM + 10% FBS. Cells were washed three times with PBS and
then stimulated with serum-free DMEM as a control or hepatic stellate cellconditioned medium with or without a neutralizing antibody against periostin (R&D
Systems, AF2955). Cell viability was determined as described by measuring after 48
and 72 h.
Multicellular colony formation assay. Tumour cells were strained through a 70 m

cell filter to ensure a single-cell suspension. The cells were then embedded at a
concentration of 2,000 cells per well in a 0.3% agar mix consisting of either DMEM
+ 5%FBS (control) or myofibroblast-conditioned medium + 5%FBS. A neutralizing
anti-periostin antibody (murine: R&D Systems, AF2955; human: Abcam14041)
was incubated for 10 min with the myofibroblast-conditioned medium before
embedding into the agar matrix. A layer of control medium or myofibroblastconditioned medium was added on top of the agar matrix with or without the antiperiostin antibody. The colonies were fixed and stained with 0.05% crystal violet in
methanol/PBS and the number of colonies per well was counted at day 14 after cell
seeding by bright-field microscopy.
RTqPCR. Cells were lysed in RLT buffer +-mercaptoethanol according to
the manufacturers instructions for RNA isolation. Total RNA purification was
performed with the RNeasy kit and cDNA was generated using QuantiTect Reverse
Transcription Kit (Qiagen) according to the manufacturers instructions. Total
RNA (500 ng) was used to generate cDNA. RTqPCR assays were performed with
LightCycler 480 SYBR Green I Master Mix (Roche) on a LightCycler 480 (Roche).
Relative expression levels were normalized to Gapdh expression according to the
formula 2(CT for gene of interestCT Gapdh) . Fold increase in expression levels was calculated
by the comparative CT method 211CT (ref. 63).
Primers are listed in Supplementary Table 4.
Immunofluorescence. For human tissue sections, antigen retrieval was carried
out in high-pH buffer using the Dako PT-Link System. Tissue sections were
blocked in PBS + 0.1% bovine serum albumin and 8% normal goat serum
prior incubation with primary antibodies. Tissue sections were stained with
secondary antibodies including DAPI (Life Technologies) and mounted using Dako
Fluorescent Mounting Medium.
Murine liver tissues were embedded in Optimal Cutting Temperature (OCT)
medium and stored at 80 C. Tissue sections were fixed in ice-cold acetone,
permeabilized with PBS + 0.1% Triton X-100, blocked with PBS + 8% normal goat
serum, and incubated with primary antibodies. Next, tissue sections were washed in
PBS, stained with secondary antibodies including DAPI (Life Technologies, 1:100)
and mounted using Dako Fluorescent Mounting Medium. Goat anti-rat or goat antirabbit secondary antibodies conjugated to AlexaFluor488 and AlexaFluor594 were
used (Abcam, 1:500).
Livers from experiments involving ZsGreen/Luciferase-transfected cells were
fixed using a sucrose gradient method to preserve the zsGreen fluorescence64. All
tissue sections were imaged using an Axio Observer Light Microscope with the
Apotome.2 (Zeiss) and quantified using the Zen Software (Zeiss). Antibodies are
listed in Supplementary Table 5.
Immunohistochemical analysis. Deparaffinization and antigen retrieval was
performed using an automated DAKO PT-link. Paraffin-embedded human and
mouse PDA tumours were immunostained using the DAKO envision+ systemHRP. Tissue sections were incubated with primary antibodies followed by HRPconjugated secondary antibody (from DAKO envision kit). Staining was developed
using diamino-benzidine and counterstained with haematoxylin. Antibodies are
listed in Supplementary Table 5.
Connective tissue staining (Massons trichrome staining). Massons trichrome
staining was performed according to the manufacturers instructions (Abcam
ab150686).
Human studies. Human studies using blood samples were approved by the National
Research Ethnics (NRES) Service Committee North WestGreater Manchester
North 08/H1011/36 and NRES Committee North WestCheshire 11/NW0083 and
REC15/NW/0477. Human blood samples were obtained from advanced PDAC
patients with liver metastasis (all pathologically confirmed) or from control healthy
subjects. All individuals provided informed consent for blood donation on approved
institutional protocols. Blood was collected in sodium heparin (NaH) tubes and
immediately processed for mononuclear cell isolation using gradient centrifugation
(Histopaque 1077, Sigma-Aldrich) according to the manufacturers protocol.
Mononuclear cells were resuspended in PBS + 2% BSA and blocked using
human TrueStain FcX (Biolegend) following by staining with antibodies and
the viability marker SYTOX Green (Life Technologies). Flow cytometry analysis
was performed on a FACSCanto II (BD Biosciences) and flow-cytometrybased cell sorting was performed on a FACSAria (BD Biosciences). For cell
sorting, cells were sorted directly into RLT buffer +-mercaptoethanol according
to the manufacturers instructions for RNA isolation (see qPCR section for
further details).
Paraffin-embedded human tissue sections from control healthy subjects and
advanced PDAC patients with liver metastasis were obtained from the Liverpool
NATURE CELL BIOLOGY
METHODS
DOI: 10.1038/ncb3340
Tissue Bank, University of Liverpool, UK or approved by NRES Committee North
West Cheshire REC15/NW/0477. All samples were pathologically confirmed.
Statistics and reproducibility. The level of significance was determined by a twotailed unpaired Students t-test and analysed with GraphPad Prism5 software.
P < 0.05 was considered significant.
Immunohistochemical and immunofluorescence analyses of human healthy and
metastatic liver biopsies (Figs 1a, 4e, 6d and Supplementary Fig. 1ac) are shown
from n = 5 healthy subjects and n = 5 metastatic patients, and human primary
PDAC tumour (Supplementary Fig. 8f) n = 4. Images were acquired and analysed
using five different visual fields from each tissue section. For immunohistochemical
and immunofluorescence analysis of murine tissue sections, the following numbers
of different visual fields from each tissue section were analysed: Supplementary
Fig. 4e (2 fields); Supplementary Fig. 3d (3 fields); Figs 1c,d,f, 2d, 5d,h and 7d,e
(4 fields); Figs 2f, 4d, 5j,k and 6d, and Supplementary Figs 5e and 8a,ce (5
fields), Supplementary Figs 6f and 7e (8 fields). For Figs 2b,c,g,h and 5b,c,f,g,
and Supplementary Figs 4c,d, 5f,g and 8b all metastatic nodules were counted
and measured (size) within one liver section. Numbers of different samples used
to acquire images are reported in the corresponding figure legends. Transwell
migration assays (Fig. 3b,d) are shown as the mean from n = 3 independent
experiments. For each experiment, 10 different visual fields per condition were
counted. Colony formation assays (Fig. 6c) are shown as the mean from n = 3
independent experiments; total number of colonies per cell type per experiment was
counted. qPCR analysis (Fig. 3a,c,f and Supplementary Figs 6d,e and 7d) are shown
as the mean from n = 3 independent experiments. qPCR analyses of in vivo-derived
murine cells are shown as the mean from three technical replicates whereby samples
were pooled from several animals from each condition: Figs 4a,c and 5e (three pooled
mice); Fig. 8d (four pooled mice), Fig. 5i (five pooled mice), Figs 5l and 7f (six pooled
mice); qPCR analyses of in vivo-derived human cells are shown as mean from n = 3
healthy and n = 4 PDAC patients (Fig. 8c). ELISA analyses (Figs 3g, 4b and 7c) are
shown as the mean from n = 3 independent experiments, whereby two (Figs 4b and
7c) and three (Fig. 3g) technical replicates per condition per experiment were used.
Access to public repository for mass spectrometry data. Mass Spec data have been
placed in the public repository MassIV database, accession ID MSV000079491 with
the title THP_1 and human fibroblast secretomes.
Direct link: ftp://massive.ucsd.edu/MSV000079491.
53. Moriya, K. et al. A fibronectin-independent mechanism of collagen fibrillogenesis in
adult liver remodeling. Gastroenterology 140, 16531663 (2011).
54. Van Rooijen, N. & Sanders, A. Liposome mediated depletion of macrophages:
mechanism of action, preparation of liposomes and applications. J. Immunol.
Methods 174, 8393 (1994).
55. McAllister, F. et al. Oncogenic Kras activates a hematopoietic-to-epithelial IL-17
signaling axis in preinvasive pancreatic neoplasia. Cancer Cell 25, 621637 (2014).
56. Kaplan, R. N. et al. VEGFR1-positive haematopoietic bone marrow progenitors initiate
the pre-metastatic niche. Nature 438, 820827 (2005).
57. Washburn, M. P., Wolters, D. & Yates, J. R. 3rd Large-scale analysis of the yeast
proteome by multidimensional protein identification technology. Nat. Biotechnol. 19,
242247 (2001).
58. Carvalho, P. C. et al. YADA: a tool for taking the most out of high-resolution spectra.
Bioinformatics 25, 27342736 (2009).
59. Li, Z. et al. Systematic comparison of label-free, metabolic labeling, and isobaric
chemical labeling for quantitative proteomics on LTQ Orbitrap Velos. J. Proteome
Res. 11, 15821590 (2012).
60. Tabb, D. L., McDonald, W. H. & Yates, J. R. 3rd DTASelect and Contrast: tools
for assembling and comparing protein identifications from shotgun proteomics.
J. Proteome Res. 1, 2126 (2002).
61. Raso, C. et al. Characterization of breast cancer interstitial fluids by TmT labeling,
LTQ-Orbitrap Velos mass spectrometry, and pathway analysis. J. Proteome Res. 11,
31993210 (2012).
62. Ashburner, M. et al. Gene ontology: tool for the unification of biology. The Gene
Ontology Consortium. Nat. Genet. 25, 2529 (2000).
63. Schmittgen, T. D. & Livak, K. J. Analyzing real-time PCR data by the comparative
C(T) method. Nat. Protoc. 3, 11011108 (2008).
64. Bleriot, C. et al. Liver-resident macrophage necroptosis orchestrates type 1
microbicidal inflammation and type-2-mediated tissue repair during bacterial
infection. Immunity 42, 145158 (2015).
NATURE CELL BIOLOGY
Corrigendum: Macrophage-secreted granulin supports

pancreatic cancer metastasis by inducing liver fibrosis
Sebastian R. Nielsen, Valeria Quaranta, Andrea Linford, Perpetua Emeagi, Carolyn Rainer, Almudena Santos,
Lucy Ireland, Takao Sakai, Keiko Sakai, Yong-Sam Kim, Dannielle Engle, Fiona Campbell, Daniel Palmer,
Jeong Heon Ko, David A. Tuveson, Emilio Hirsch, Ainhoa Mielgo and Michael C. Schmid
Nature Cell Biology 18, 549560 (2016); published online 18 April 2016; corrected after print 20 May 2016
In the version of this Article originally published, in the last sentence of the Acknowledgements section, Pancreas Biomedical Research Unit
should have read National Institute for Health Research Biomedical Research Unit funding scheme through a NIHR Pancreas BRU. This has been
corrected in all online versions of the Article.
NATURE CELL BIOLOGY

S U P P L E M E N TA R Y I N F O R M AT I O N
DOI: 10.1038/ncb3340
Cytokeratin
CD45
Liver metastasis (LM)
400
P=0.005
Cells / FOV
300
200
DAPI/CD68/CK19
SMA
CD68
PDGFR
CK19
DAPI/CD68
/PDGFR
Healthy liver (HL)
CD68
Liver metastasis (LM)
MTS
Connective tissue
(% area)
Healthy liver (HL)
50
40
30
20
10
0
P<0.001
P<0.001
HL
LM
40 P<0.001
30
20
10
0
HL LM
DAPI/zsGreen
HE
Experimental metastasis model
Healthy
Day 5
Day 12
Figure S1 - Schmid
Supplementary Figure 1 Morphological characterisation of metastatic PDAC

lesions in human and mouse. (a) Lower magnification of representative
micrographs shown in Fig. 1a. Identification of pan-cytokeratin (CK)+
metastatic pancreatic cancer cells, hematopoietic cells (CD45+),
macrophages (CD68+) and myofibroblasts (SMA+) as predominant
cell types at the hepatic metastatic microenvironment of pancreatic
cancer by immunohistochemical analysis of human biopsies (data are
from 5 PDAC patients and 5 healthy subjects; five fields assessed per
sample). HL= healthy liver, LM = liver metastasis. (b) Identification of
cytokeratin (CK) 19+ metastatic pancreatic cancer cells, tumour associated
macrophages (CD68+) and myofibroblasts (PDGFR+) as predominant cell
types at the hepatic metastatic microenvironment of pancreatic cancer
by immunofluorescence analysis of human biopsies. Representative
micrographs are shown and quantification of data (n= 5 PDAC, n = 5
healthy subjects; five fields assessed per sample; mean s.e.m; two-tailed
unpaired t-test). HL= healthy liver, LM=liver metastasis. (c) Representative
Massons trichrome staining and quantification of area occupied by fibrotic

stroma in human biopsies (n= 5 PDAC patients, n = 5 healthy subjects;
five fields assessed per sample; mean s.e.m; two-tailed unpaired t-test).
HL= healthy liver, LM=liver metastasis. (d) Experimental metastasis model
by intrasplenic implantation of 1x106 KrasG12D;Trp53R172H;Pdx1-Cre
(KPC) mice derived pancreatic cancer cells expressing a luciferase/zsGreen
lentiviral reporter plasmid (KPCluc/zsGreen). Liver tissues were isolated at
day 5 and day 12 post implantation. (Upper panel) Liver tissue sectioned
stained by Hematoxylin and Eosin (HE) showing initial micrometastatic
lesions of disseminated cancer cells at day 5 post implantation followed
by the generation of an excessive stromal microenvironment surrounding
disseminated cancer cells at day 12 post implantation. (Lower panel)
Representative immunofluorescence staining of disseminated KPCluc/zsGreen
(zsGreen) cells. Data are from 6 mice per time point; four fields assessed
per sample; data combine two independent experiments. Scale bars = a,
200 m; b, c, d, 100m.
WWW.NATURE.COM/NATURECELLBIOLOGY
1
Supplementary Figure 2 FACS-based quantification of immune cell

infiltrating the metastatic site in PDAC. Experimental metastasis
model by intrasplenic implantation of 1x106 KPC-derived pancreatic
cancer cells. After 12 days, liver tissues were isolated, enzymatically
digested and resulting single cell suspensions were stained for flow
cytometry analysis. Nave livers were used as controls (healthy).
(a) Representative flow cytometry dot plots showing gating
strategy used to quantify intrametastatic B cells (CD45+B220+),
T cell (CD45+CD3+), NK cells (CD45+NK1.1+CD3negB220neg),
Neutrophils (CD45+CD11b+F4/80negLy6G+Ly6C+, inflammatory
monocytes (CD45+CD11b+F4/80negLy6GnegLy6C+) and macrophages
(CD45+CD11b+F4/80+). Only viable cells (SYTOXneg) were used (data are
from 5 healthy livers or 8 liver metastasis; repeated two times with similar
results). (b) Quantification of CD45+ hematopoietic cells detected in
healthy control livers (Ctrl) and metastatic tumour bearing livers (LM) (n =
5 Ctrl mice; n = 8 LM mice; data combine two independent experiments;
individual data points, horizontal lines represent mean s.e.m; two-tailed
unpaired t-test). (c) Percentage of intrametastatic immune cells among
total viable cells gated according to (a) in healthy control livers (Ctrl) and
metastatic tumour bearing livers (LM) (individual data points, horizontal
lines represent mean s.e.m; two-tailed unpaired t-test). Data shown
combine two independent experiments; total mice: n = 5 Ctrl mice, n = 6
LM mice (B220, NK1.1, CD3); n = 5 Ctrl mice; n = 9 LM mice (Ly6G, Ly6C,
F4/80). (d) Representative flow cytometry dot plots showing gating strategy
to analyse CCR2 (CD192) expression levels. Intrametastatic inflammatory
monocytes (CD45+CD11b+F4/80negLy6GnegLy6C+) and macrophages
(CD45+CD11b+F4/80+) express CCR2 (data are from 5 Ctrl mice or 8 LM
mice; repeated two times with similar results). ns, not significant.
Supplementary Figure 3 Liver PDAC metastasis induces the recruitment of

monocyte-derived macrophages, followed by the activation of resident hStCs.
(a, b) Experimental metastasis model by intrasplenic implantation of 1x106
KPC cells. (a) Intrametastatic CD11b+ cells showing a marked expansion of
the macrophage population (F4/80+, blue) in established macro-metastatic
lesions (day 12, D12), while inflammatory monocytes (Ly6C+Ly6Gneg, red)
predominantly increased during early micro-metastatic spreading (day 5,
D5). Ctrl= 5 mice; Day 5 = 2 mice, Day 12 = 5 mice; data combine two
independent experiments. (b) Representative immunofluorescence staining
showing increased CD11b+ cells (upper) in micro-metastatic livers, followed
by excessive accumulation of aSMA+ myofibroblasts in livers with macrometastatic lesions. Macrophages (lower) are equally distributed in healthy livers
(Kupffer cells) and show increase clustering in tumour bearing livers. Data are
from 6 mice per time point; two independent experiments. (c) Representative
Massons trichrome staining (MTS) of healthy control liver indicating absence
of fibrotic stroma (lack of blue colour). Data are from 6 mice; two independent
experiments. (d) Statistical comparison showing a positive correlation (Pearson)
between increased numbers of aSMA+ myofibroblasts and area occupied by
metastatic cancer cells in tumour bearing murine livers. solid line = best fit,
dashed lines, 95% confidence intervals. Total n = 20 mice; four independent

experiments. (e - g) Primary tumours and spontaneous metastatic hepatic
tumours derived KPC mice. (e) Representative HE images showing the
presence of an excessive stromal compartment at both sites (data are from 5
mice per condition, four fields assessed per sample). (f) Representative images
of liver tumour sections stained for MAMs (F4/80+), pancreatic cancer cells
(CK19+), or myofibroblasts (aSMA+) showing the presence of an excessive
metastatic microenvironment mainly consisting of MAMs and myofibroblasts
surrounding the tumour cells (data are from 5 mice per condition, four fields
assessed per sample). (g) Flow cytometry analysis showing a marked expansion
of the macrophage population (F4/80+) in metastatic livers (ML) compared to
healthy livers (HL) (n = 5 mice per condition, data combine five independent
experiments; mean s.e.m.; two-tailed unpaired t-test). (h - j) Chimeric WT +
tdTomatoRed BM (WT+tdTR BM) mice. (h) Successful BM reconstitution was
confirmed by flow cytometry analysing total circulating CD11b+Gr1+ cells. (i,
j) Representative images showing co-localization of tdTomatoRed signal with
F4/80+ MAMs (i), but not with aSMA+ myofibroblasts in the metastatic lesion
(below white line). Data are from 6 mice per condition; one experiment. Scale
bars, b, c, f, 100m; e, 50m; i, j, 25m.
3
30
20
10
0
60
40
20
0.1
/-
PI
3K
-
T
W
3K
/-
0.2
WT
PI3K-/-
Panc02
1.5
P=0.018
1.0
0.5
0.0
W
PI T
3K
/-
DAPI/SMA
0.3
0.0
PI
P=0.008
/-
40
0.4
P=0.019
3K
-
Mets / liver section
50
80
Panc02
SMA+ cells / area (a.u.)
60
P<0.001
Panc02
PI
Panc02
Lesion size (mm 2)
b
F4/80+ (% of CD45+)
PI3K-/-
WT
Figure S4 - Schmid
Supplementary Figure 4 The recruitment of monocyte derived

macrophages is necessary for efficient pancreatic cancer metastasis.
1x106 KPC-derived cells (a) or 1x106 Panc02 cells (b - e) were
intrasplenically injected into age and sex matched WT and PI3Kg-/- (-/-)
mice. After 12 days, total livers were harvested and analysed by flow
cytometry, immunohistochemistry, and immune fluorescence based
methods. (a) Representative HE staining of liver tissue sections showing a
marked decreased size of metastatic tumours in livers of PI3Kg deficient
(-/-) mice (data are from 7 WT and 9 PI3Kg-/- mice; data combine two
independent experiments). (b) Percentage of intrametastatic macrophages
among CD45+ cells quantified by flow cytometry. PI3Kg deficiency (-/-)
results in a marked reduction of MAMs compared to WT (n = 6 mice
per condition; one experiment; individual data points, horizontal lines

represent mean s.e.m; two-tailed unpaired t-test). (c, d) Quantification
of metastatic frequency (c) and average metastatic lesion size (d) in WT
and PI3Kg knockout mice (-/-) by hematoxylin and eosin (HE) stained
liver sections (n = 6 mice per condition; all metastatic nodules assessed
from one section per sample; one experiment; c, individual data points,
horizontal lines represent mean s.e.m; d, mean s.e.m; two-tailed
unpaired t-test). (e) Representative immunofluorescence staining and
quantification of myofibroblasts (aSMA+) cell frequency in livers in WT
and PI3Kg knockout mice (-/-). Nuclei were counterstained with DAPI (n
= 6 mice per condition; two fields assessed per sample; one experiment;
mean s.e.m; two-tailed unpaired t-test).
In vivo bioluminescence imaging

2 days post intra-splenic implantation
In vivo bioluminescence imaging

9
1.0x10
Total Flux
(Photons/Second)
5.0
4.0
3.0 x10
2.0
1.0
Control
KPC
luc/zsGreen
1.0x10
1.0x10
1.0x10
Colour Scale
Min = 1.16e7
Max = 5.80e8
1.0x10
Ctrl
Day 2
KPC
PBS Liposomes
20
Clodronate Liposomes
P=0.042
20
15
SSC
Panc02
25
15
10
10
5
0
F4/80
1.5
Fold change
(normalised to PBS)
PBS Liposomes
5
PL
CL
PL CL
P=0.049
FSC
1.0
PDGFR
F4/80
SMA
60
P=0.003
40
20
0
PL CL
15
Metastatic foci
(per 100mm2 liver)
100 P=0.01
80
Positive cells / FOV
PL CL
PBS Liposomes
P=0.038
10
5
0

(mm2)
0.5
PL CL
0.15
P=0.026
0.10
0.05
0.0
PL CL
Figure S5 - Schmid
Supplementary Figure 5 Disseminated PDAC cells depend on MAMs to sustain

growth after colonization of the metastatic site. (a, b) In vivo bioluminescence
imaging to monitor metastatic colonization of the liver (a) Representative
image showing radiance of a mouse two days post intrasplenically implantation
of 1x 106 PDAC cancer cells (KPCluc/zsGreen) and tumour free control mouse.
Main signal detected originates from the liver area, while some residual signal
is measured from the injection site, the spleen. (b) Quantification of radiance
(total flux) measure two days post implantation confirming colonization of
the liver by implanted cancer cells has occurred (data are from 2 control or
10 KPCluc/zsGreen mice; one experiment, individual data points, horizontal
lines represent mean). (c) Representative flow cytometry dot plots and
quantification showing a marked reduction of macrophages (F4/80+) and
myofibroblasts (PDGFR+) in KPC cell induced tumour bearing livers of mice
treated with Clodronate Liposomes (CL) compared to control mice treated with
PBS Liposomes (PL) (n = 3 mice per condition; one experiment; individual
data points, horizontal lines represent mean s.e.m; two-tailed unpaired
t-test). (d) Percentage of macrophages (F4/80+) among viable cells in Panc02

cell induced tumour bearing livers of mice treated with Clodronate Liposomes
(CL, data from 2 mice) compared to control mice treated with PBS Liposomes
(PL, data from 2 mice) (one experiment; individual data points, horizontal
lines represent mean). (e) Representative immunofluorescence staining and
quantification of MAMs (F4/80+) and myofibroblasts (aSMA+) cell frequency in
Panc02 cell induced liver tumours of mice treated with liposomes containing
control PBS (PL) or clodronate (CL). Nuclei were counterstained with DAPI (n
= 4 mice per condition; five fields assessed per sample; one experiment; mean
s.e.m; two-tailed unpaired t-test). (f, g) Evaluation of metastatic frequency
(f) and lesion area covered by metastatic cells (g) in Panc02 tumour bearding
livers of mice treated with liposomes containing control PBS (PL) or clodronate
(CL). Nuclei were counterstained with DAPI (n = 5 mice per condition; all
metastatic nodules assessed from one section per sample; one experiment;
unpaired t-test). Scale bars, 100m.
5
b
Relative invaded
cell number (fold)
3
2
1
P<0.001
2.5
P<0.001
2.0
1.5
1.0
0.5
0.2
Ctrl THP-1 M
CM CM
0.5
0
1.5
P<0.001
P=0.016
ns
1.0
0.5
0.0
LP
LP
S+
IF
nc IL
02 -4
KP C
C M
C
M
P=0.022
Murine
Pa
P=0.001
P<0.001
Day 0
Day 3
Day 5
P<0.001
*
tro
1.0
P<0.001
0.4
S+ M
IF 0
N
Pa
nc IL1 4
C
M
Granulin
relative mRNA levels
(normalised to Gapdh)
Human
1.5
P<0.001
C
on
Ac
ta
C 2
ol
1a
1
Fn
PdxCre-ERT KP
Healthy liver
Liver Metastasis
PdxCre-ERT KP
CD68
Granulin
60
Granulin
Cells/FOV
CD68
0.6
Granulin
relative mRNA levels
(normalised to Gapdh)
Fold change
mRNA levels
Human
Mouse
Absorbance (595 nm)
Control
Macrophage CM
40
20
Supplementary Figure 6 High levels of granulin are specifically expressed

by tumour educated macrophages in vitro and in vivo. (a) Quantification
of aSMA (Acta2) and collagen 1a (Col1a) mRNA levels in primary murine
dermal fibroblasts stimulated with isogenic macrophage conditioned
media (CM) as determined by qPCR. Bar graph show fold up regulation
Figure S6
Schmid
compared to unstimulated
and -are
displayed as mean (data are from a
single experiment, repeated four times with similar results; Supplementary
Table 6). (b) Quantification of human (green) and mouse (red) myofibroblast
that migrated towards human and mouse macrophages, respectively, in a
matrigel-coated transwell assay (n = 4 independent experiments; mean
s.e.m.; two-tailed unpaired t-test) (c) Quantification of human myofibroblast
proliferation in the presence of control media, THP-1 derived macrophage
CM, or primary macrophage CM (n = 4 independent experiments; mean
s.e.m.; two-tailed unpaired t-test). (d) Quantification of granulin mRNA
expression levels by qPCR in human primary unstimulated macrophages
(M0) or primary macrophages stimulated with either interleukin (IL) -4
HL
LM
(M2-like phenotype), interferon (IFN) g / LPS (M1-like phenotype) or

educated with tumour conditioned media generated from human Panc1
cells (n = 3 independent experiments mean s.e.m.; two-tailed unpaired
t-test). (e) Quantification of granulin mRNA expression levels by qPCR
in murine primary macrophages stimulated with either interleukin (IL)
-4 (M2-like phenotype), interferon (IFN) g / LPS (M1-like phenotype) or
educated with tumour conditioned media generated from murine KPC
and Panc02 PDAC cancer cells (n = 3 independent experiments; mean
s.e.m; two-tailed unpaired t-test). (f) Metastatic hepatic tumours
derived from the spontaneous mouse pancreatic cancer model Pdx1CreERT;KrasG12D;Trp53R172H; (PdxCre-ERT KP mice) were isolated and
morphometrically analysed. Representative images of immunohistochemistry
staining for MAMs (CD68+) and granulin expression on serial tissue sections
from metastatic lesions and healthy liver and quantification of the data (n
= 1 mouse per condition; eight fields assessed per sample; bars represent
means). ns, not significant.
Absorbance (595 nm)
4
2
BM
-/-
BM
0.4
D
3
D
2
D
0
P<0.001
0.3
0.2
0.1
PdxCreERT KP
Healthy liver
Liver metastasis
o.
CD8+ T cells
(% of CD45+)
0.5
br
10
1.0
Fi
20
P<0.001
Periostin
relative mRNA
30
ns
P<0.001
1.5
KP
8
P=0.017
BM
-/rn
BM
WT Grn-/-
P<0.001
rn
20
40
P<0.001
2.0
60
40
CD8+ T cells
(% of CD45+)
ns
WT Grn-/-
Control
hStC-CM
hStC-CM + Anti-Periostin
BM
80
BM
Total macrophages
(% of CD45+)
WT Grn-/-
-/-
10
10
20
rn
20
ns
30
ns
10
30
40
ns
40
CD206+ macrophages
(% of total M)
Total macrophages
(% of CD45+)
CD206+ macrophages
(% of total M)
Periostin
Periostin IHC
(% area)
15
10
5
H
L
LM
Control
Clodronate
PI3K-/-
Grn-/-
Control
Clodronate
PI3K-/-
Grn-/-
DAPI/Periostin
MTS
Figure S7 - Schmid
Supplementary Figure 7 Depletion of granulin does not affect the

recruitment of macrophage to the metastatic site, their activation, or
intrametastatic effector T cell numbers, but markedly reduces stromal
expansion. 5x105 KPC-derived cells were intrasplenically injected into WT
and Grn-/- mice (a), and chimeric WT+ WT BM and WT + Grn-/- BM mice (b).
After 12 days, total livers were harvested and analysed my flow cytometry.
(a) Quantification of intrametastatic total MAMs (F4/80+), CD206+ MAMs,
and CD8+ effector T cells by flow cytometry isolated from WT and Grn/- mice (n = 6 mice per condition; one experiment; individual data points,
horizontal lines represent mean s.e.m; two-tailed unpaired t-test). (b)
Quantification of intrametastatic total MAMs (F4/80+), CD206+ MAMs,
and CD8+ effector T cells by flow cytometry isolated from WT + WT BM
and WT + Grn-/- BM mice mice (n = 6 mice per condition; one experiment;
unpaired t-test). (c) Quantification of murine KPC cancer cell proliferation
in the presence or absence of Mf CM and periostin neutralising antibody
(anti-Periostin) (n = 4 independent experiments; mean +/- SEM; twotailed unpaired t-test). (d) Quantification of periostin expression by qPCR
in primary activated murine myofibroblasts and KPC cells. Periostin
expression was undetectable in KPC cells while myofibroblasts expressed
high levels of periostin (n = 3 independent experiments; mean s.e.m).
(e) Immunohistochemical analysis of periostin in healthy liver (HL) and
spontaneous metastatic livers (ML) collected from Pdx1-CreERT KP mice,
respectively. Representative micrographs and quantification of the data
are shown (n = 1 mouse per condition; eight fields assessed per sample;
bars represent means). (f, g) Representative images of the evaluation of
periostin deposition and fibrotic stroma formation (MTS) in metastatic livers
of control WT, WT mice treated with clodronate liposomes, PI3Kg-/-, Grn-/-,
and WT + Grn-/- BM mice 12 days after intra-splenic implantation of KPC
cells (data are from 4 mice per condition, four fields assessed per sample;
data combine five independent experiments). Scale bars = 100mm. NS, not
significant.
7
10
Periostin
BM
0.5
0
T
W
Human PDAC
Primary Tumour
2.0 P<0.001
Granulin
1.5
1.0
0.5
BM
-/-
rn
BM
Periostin
Connective tissue
(% area)
WT BM
1.0
rn
BM
-/-
rn
T
W
MTS
Grn-/- BM
1.5
BM
10
Grn-/- BM
20
P=0.016
2.0
BM
30
0.1
-/-
P<0.001
0.2
Periostin IHC (% area)
ns
50
40
W
T
G BM
rn
-/BM
0.3 P=0.004
W
T
B
rn M
-/BM
20
F4/80
SMA
Cells / FOV
Metastatic foci
Grn-/- BM
DAPI/F4/80/SMA
WT BM
Grn-/- BM
ns
30
WT BM
Hematoxylin/Eosin
CD68
Granulin
WT BM
Metastatic lesion size

(mm 2 )
Counts
Inflammatory monocytes (IM)

Resident monocytes (RM)
CCR2
Figure S8 - Schmid
Supplementary Figure 8 Depletion of granulin in the hematopoietic

compartment reduces pulmonary metastasis and myofibroblast activation
in the lung. (a - e) 5x105 KPC-derived cells were intravenously injected into
age and sex matched chimeric WT + WT BM and WT + Grn-/- BM mice.
After 12 days, total lungs were harvested and analysed. (a) Representative
images of immunohistochemistry staining for MAMs (CD68+) and granulin
expression on serial tissue sections from lung metastatic lesions and
quantification of the data (data are from 5 WT + WT BM mice or 4 WT
+ Grn-/- BM mice; four fields assessed per sample; one experiment). (b)
Representative images and quantification of WT BM and Grn-/- BM derived
lung tissue stained with HE showing a marked reduction of area covered
by metastatic cancer cells in Grn-/- BM mice compared to WT BM mice
(mean s.e.m.; two-tailed unpaired t-test), while frequency of metastatic
lesions (metastatic foci) remained unchanged (n = 5 WT + WT BM mice,
n = 4 WT + Grn-/- BM mice; all metastatic nodules assessed from one
section per sample; one experiment; individual data points, horizontal line
represents mean s.e.m.; two-tailed unpaired t-test). (c) Representative
immunofluorescence staining and quantification of MAMs (F4/80+) and
myofibroblasts (aSMA+) cell frequency in tumour bearing lungs of WT +
WT BM and WT + Grn-/- BM mice. Nuclei were counterstained with DAPI

(n = 5 WT + WT BM mice, n = 4 WT + Grn-/- BM mice; five fields assessed
per sample; one experiment; mean s.e.m.; two-tailed unpaired t-test).
(d) Representative immunofluorescence staining and quantification of
periostin expression in tumour bearing lungs of WT + WT BM and WT +
Grn-/- BM mice. Nuclei were counterstained with DAPI (n = 5 WT + WT
BM mice, n = 4 WT + Grn-/- BM mice; five fields assessed per sample; one
experiment; mean s.e.m.; two-tailed unpaired t-test). (e) Representative
Massons trichrome staining and quantification of area occupied by fibrotic
stroma in tumour bearing lungs of WT + WT BM and WT + Grn-/- BM mice
(n = 5 WT + WT BM mice, n = 4 WT + Grn-/- BM mice; five fields assessed
per sample; one experiment; mean s.e.m.; two-tailed unpaired t-test).
(f) Immunohistochemical analysis of periostin and granulin expression in
human primary PDAC tumours. Representative micrographs are shown (data
are from 4 different patients per condition). (g) IM from blood samples from
healthy subjects and metastatic PDAC patients expressed high levels of the
chemokine receptor CCR2 (CD192). Representative histogram shown from
6 different samples per condition. Scale bars a, b, c, e, f, 100 m; d, e,
200 m; inset, 20 m.
Supplementary Tables
Supplementary Table 1: Identified proteins in tumour educated macrophage secretome.
Top 20 identified proteins secreted by tumour educated macrophages are listed according to their abundance.
Supplementary Table 2. Secretome macrophages.
Complete list of secreted proteins associated with GO term Extracellular Vesicular Exosome identified by mass spectrometry in human macrophages
educated with Panc1 conditioned media.
Supplementary Table 3. Secretome myofibroblasts.
Complete list of secreted proteins associated with GO term Extracellular Matrix Organistion identified by mass spectrometry as differentially expressed in
human fibroblasts educated with primary macrophage conditioned media compared to unstimulated fibroblasts.
Supplementary Table 4: Primers.
List of primers used for qPCR analysis in this study.
Supplementary Table 5: Antibodies.
List of antibodies used in this study.
Supplementary Table 6: Statistics source data.
Activation assays primary dermal fibroblasts.
9

tmpA13D TMP

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

tmpA13D TMP

Enviado por

Direitos autorais:

Formatos disponíveis

A RT I C L E S

Macrophage-secreted granulin supports pancreatic

myofibroblasts that express alpha smooth muscle actin15 (SMA). In

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Percentage of CD45+ cells

Cells per FOV

Cells per FOV

Figure 1 Metastatic PDAC cells induce macrophage recruitment and

clustering around metastatic KPCluc/zsGreen (zsGreen) cancer cells in the liver

PDAC patients and healthy volunteers by immunohistochemical

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Cells per FOV

Metastatic lesion area (mm2)

Metastatic foci (per 100 mm2 liver)

Viable cells (%)

Viable cells (%)

Figure 2 Macrophages promote myofibroblast activation and metastatic

To further evaluate the type of immune cells accumulating at

(a common way of metastasis occurring in humans)25,26, and generate

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Grn/ MCM + rGrn

600 P < 0.001

Granulin protein (ng ml1)

Relative Grn mRNA levels

Figure 3 Granulin secreted by macrophages activates hepatic stellate cells.

of 1 106 KPC cells into isogenic WT + WT BM and WT + Grn/ BM

and F4/80+ macrophages were increased in tumour-bearing livers

expansion of resident KCs or to the recruitment of BM-derived cells,

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Cells per FOV

Cells per FOV

Relative mRNA levels

Granulin protein (ng ml1)

Fold change Grn mRNA levels

Figure 4 Granulin is highly expressed in hepatic metastatic lesions

three pooled mice; one experiment). (d) Spontaneous metastatic hepatic

Together, our findings demonstrate that: in PDAC liver metastasis,

that depletion of PI(3)K will ablate IM trafficking to the metastatic

Macrophages promote myofibroblast activation and metastatic

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Figure 5 Granulin depletion prevents myofibroblast activation and PDAC

Relative mRNA levels

Caspase-3+ cells per FOV

Percentage of Ki67+ tumour cells

Metastatic lesion area

Cells per FOV

Cells per FOV

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

log (fold change)

Colonies per well

Figure 6 Myofibroblast-secreted periostin enhances pancreatic cancer cell

single cells are shown for KPC cells (n = 3 independent experiments;

studies on investigating whether MAMs are required for supporting

activation of myofibroblasts and impairs the progression of metastatic

NATURE CELL BIOLOGY VOLUME 18 | NUMBER 5 | MAY 2016

2016 Macmillan Publishers Limited. All rights reserved

Periostin (ng ml1)