Microvesicles secreted by macrophages shuttle invasion-potentiating microRNAs into breast cancer cells.

Abstract

Tumor-associated macrophages (TAMs) are alternatively activated cells induced by interleukin-4 (IL-4)-releasing CD4+ T cells. TAMs promote breast cancer invasion and metastasis; however, the mechanisms underlying these interactions between macrophages and tumor cells that lead to cancer metastasis remain elusive. Previous studies have found microRNAs (miRNAs) circulating in the peripheral blood and have identified microvesicles, or exosomes, as mediators of cell-cell communication. Therefore, one alternative mechanism for the promotion of breast cancer cell invasion by TAMs may be through macrophage-secreted exosomes, which would deliver invasion-potentiating miRNAs to breast cancer cells.
We utilized a co-culture system with IL-4-activated macrophages and breast cancer cells to verify that miRNAs are transported from macrophages to breast cancer cells. The shuttling of fluorescently-labeled exogenous miRNAs from IL-4-activated macrophages to co-cultivated breast cancer cells without direct cell-cell contact was observed. miR-223, a miRNA specific for IL-4-activated macrophages, was detected within the exosomes released by macrophages and was significantly elevated in the co-cultivated SKBR3 and MDA-MB-231 cells. The invasiveness of the co-cultivated breast cancer cells decreased when the IL-4-activated macrophages were treated with a miR-223 antisense oligonucleotide (ASO) that would inhibit miR-223 expression. Furthermore, results from a functional assay revealed that miR-223 promoted the invasion of breast cancer cells via the Mef2c-β-catenin pathway.
We conclude that macrophages regulate the invasiveness of breast cancer cells through exosome-mediated delivery of oncogenic miRNAs. Our data provide insight into the mechanisms underlying the metastasis-promoting interactions between macrophages and breast cancer cells.

Full text

Page 1

RESEARCHOpen Access
Microvesicles secreted by macrophages shuttle
invasion-potentiating microRNAs into breast
cancer cells
Mei Yang1,2, Jingqi Chen1, Fang Su3, Bin Yu2, Fengxi Su1, Ling Lin1,4, Yujie Liu1, Jian-Dong Huang2*and
Erwei Song1*
Abstract
Background: Tumor-associated macrophages (TAMs) are alternatively activated cells induced by interleukin-4 (IL-4)-
releasing CD4+T cells. TAMs promote breast cancer invasion and metastasis; however, the mechanisms underlying
these interactions between macrophages and tumor cells that lead to cancer metastasis remain elusive. Previous
studies have found microRNAs (miRNAs) circulating in the peripheral blood and have identified microvesicles, or
exosomes, as mediators of cell-cell communication. Therefore, one alternative mechanism for the promotion of
breast cancer cell invasion by TAMs may be through macrophage-secreted exosomes, which would deliver
invasion-potentiating miRNAs to breast cancer cells.
Results: We utilized a co-culture system with IL-4-activated macrophages and breast cancer cells to verify that
miRNAs are transported from macrophages to breast cancer cells. The shuttling of fluorescently-labeled exogenous
miRNAs from IL-4-activated macrophages to co-cultivated breast cancer cells without direct cell-cell contact was
observed. miR-223, a miRNA specific for IL-4-activated macrophages, was detected within the exosomes released
by macrophages and was significantly elevated in the co-cultivated SKBR3 and MDA-MB-231 cells. The invasiveness
of the co-cultivated breast cancer cells decreased when the IL-4-activated macrophages were treated with a miR-
223 antisense oligonucleotide (ASO) that would inhibit miR-223 expression. Furthermore, results from a functional
assay revealed that miR-223 promoted the invasion of breast cancer cells via the Mef2c-b-catenin pathway.
Conclusions: We conclude that macrophages regulate the invasiveness of breast cancer cells through exosome-
mediated delivery of oncogenic miRNAs. Our data provide insight into the mechanisms underlying the metastasis-
promoting interactions between macrophages and breast cancer cells.
Background
Breast cancer is the most common malignant tumor in
women, and distal metastasis of highly invasive breast
cancer cells is the major cause of death in these women.
Tumor-associated macrophages that infiltrate the breast
cancer stroma are the most notable migratory hemato-
poietic cell-type in the tumor microenvironment and
function to promote the invasiveness of breast cancer
cells [1-4]. Macrophages are heterogeneous cells that
respond differently to various stimulating signals and
display numerous phenotypes [5]. Fully polarized M1, or
classically activated, macrophages are stimulated by
microbial agents or pro-inflammatory factors, such as
lipopolysaccharides (LPS); whereas M2, or alternatively-
activated, macrophages are responding to anti-inflam-
matory molecules, such as interleukin-4 (IL-4). The M1
and M2 macrophage phenotypes represent the two
extremes of a broad range of macrophage functional
states. Among these functional states, M2 macrophages
activated by IL-4 have been associated with breast can-
cer invasion, metastasis and poor patient prognosis
[2,6-10].
Previous studies have shown that TAMs promote
breast cancer progression and metastasis by releasing a
* Correspondence: jdhuang@hku.hk; songerwei02@yahoo.com.cn
1Breast Tumor Center, Sun-Yat-Sen Memorial Hospital, Sun-Yat-Sen University,
Guangzhou, PR China
2Department of Biochemistry, University of Hong Kong, Pokfulam, Hong
Kong SAR, China
Full list of author information is available at the end of the article
Yang et al. Molecular Cancer 2011, 10:117
http://www.molecular-cancer.com/content/10/1/117
© 2011 Yang et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.

Page 2

variety of cytokines that regulate the survival and inva-
siveness of tumor cells and stimulate tumor angiogen-
esis [11,12]. More recent data have demonstrated that
macrophages are able to produce microvesicles, also
known as exosomes, which shuttle proteins or micro-
RNAs (miRNAs) into adjacent cells within the microen-
vironment [13,14]. Exosomes are derived from
multivesicular endosomes that fuse with the plasma
membrane and are shed into the extracellular space.
These particles range in size from 50 to 100 nm. A wide
variety of cells may release exosomes, but their contents
vary depending on the cell-type of origin and its activa-
tion status [15]. One topic of considerable interest is
that exosomes contain miRNAs that mediate intercellu-
lar communication [16-19]. miRNAs are short, non-cod-
ing RNAsthatregulate
complementary mRNAs [20]. The shuttling of these
molecules between cells aids in regulating the biology of
target cells. miR-223 is specific for alternatively-activated
M2 macrophages induced by IL-4 and is associated with
the regulation of human granulopoiesis [21].
In the present study, we demonstrate that exogenous
miRNAs transfected into IL-4-activated M2 macro-
phages can be shuttled into co-cultivated breast cancer
cells in the absence of direct cell-cell contact with the
macrophages. Exosomes containing miR-223 were
released by M2 cells and were then internalized by co-
cultivated breast cancer cells that did not express this
miRNA. The exosome-shuttled miR-223 promoted the
invasiveness of breast cancer cells in vitro. This process
of invasion could be inhibited by transfecting miR-223
antisense oligonucleotides (ASO) into the tumor cells.
Our study provides evidence for the delivery of inva-
sion-potentiating miR-223 by IL-4-activated macro-
phages to breast cancer cells via exosomes and may
highlight a novel communication mechanism between
TAMs and cancer cells.
theexpressionof
Methods
Isolation and activation of human monocyte-derived
macrophages
Institutional approval from the local research ethical
committees (Internal Review and the Ethics Boards of
the Sun-Yat-Sen Memorial Hospital, Sun-Yat-Sen Uni-
versity) was obtained prior to conducting the study.
Human mononuclear cells were isolated from the per-
ipheral blood of healthy donors by Ficoll density gradi-
ent centrifugation at 450 × g for 25 min at room
temperature. The mononuclear cells were washed three
times with PBS and plated at a density of 5 × 106per
well in 24-well plates and incubated for 1.5 h in DMEM
alone. Subsequently, non-adherent cells were washed
away with warm Hank’s solution, and the adherent
monocytes were cultured in DMEM containing 10%
fetal bovine serum (FBS). Media was changed every 3
days, and the resulting monocyte-derived macrophages
(MDMs) were activated by adding IL-4 (45 ng/ml) to
the culture medium for 3 days.
Cell cultures and co-cultivation
HEK-293T cells and the breast cancer cell lines, SKBR3
and MDA-MB-231, were cultured in DMEM containing
10% FBS. Co-cultivation of macrophages and breast can-
cer cells was performed in 24-well Boyden chambers
(Corning, cat. no. 3413). Macrophages were seeded on
the 0.4 μM inserts, which are permeable to supernatants
but not to cellular components. Breast cancer cells were
seeded in the lower chambers and grown for the indi-
cated periods of time.
Microarray
MicroRNA expression profiles were generated as
described previously [22], and hybridization using Disco-
vArray miRNA arrays was performed according to the
manufacturer’s instructions.
Quantitative reverse transcription-PCR
Quantitative real-time reverse transcription-PCR (qRT-
PCR) for miR-223 was conducted using Real-Time PCR
Universal Reagent (GenePharma Co., Ltd.) and the MX-
3000P Real-Time PCR machine (Stratagene). All reac-
tions were performed in a 20 μL reaction volume in tri-
plicate. The primers used for miR-223 and U6 snRNA
are as follows: miR-223 forward, 5’CATTGTCAGTT
TGTCAAATACC3’; miR-223 reverse, 5’CCATGAGA-
GATCCCTAGCG3’ and U6 snRNA forward, 5’ATTG-
GAACGATACAGAGAAGAT3’; U6 snRNA reverse,
5’GGAACGCTTCACGAATTT3’. PCR amplification
consisted of an initial denaturation step at 95°C for 3
min, followed by 40 cycles of 95°C for 30 s, 62°C for 40
s, and 72°C for 30 s. Standard curves were generated,
and the relative amount of miR-223 was normalized to
the amount of U6 snRNA (2- ΔCt).
Luciferase reporter plasmid construction
The pMIR-REPORT miRNA expression reporter (firefly
luciferase reporter plasmid; Ambion) was used for plas-
mid construction. The 3’-UTR of Mef2c, containing two
miR-223 target sequences, miR-223 complementary
sequence (TGGGGTATTTGACAAACTGACA) and lin-
4 complementary sequence (CTAGTCACACTT-
GAGGTCTCAGGGA), were separately cloned into the
3’-UTR of the pMIR-REPORT plasmid according to the
manufacturer’s instructions (Ambion).
Invasion assay
Invasion was measured by assessing the cell migration
rate through an artificial basement membrane in a
Yang et al. Molecular Cancer 2011, 10:117
http://www.molecular-cancer.com/content/10/1/117
Page 2 of 13

Page 3

modified Boyden chamber (Corning, cat. no. 3422). The
membrane consisted of polycarbonate (8 μm pore dia-
meter) and was coated on ice with Matrigel (BD Bios-
ciences)diluted in serum-free
resuspended in DMEM were seeded into the upper well
of the chamber (100 μl), while the lower well was filled
to the top (approximately 600 μl) with complete med-
ium (DMEM containing 10% FBS). Cells were incubated
for 4-8 h. The cells in the upper well that did not
migrate were scraped off, and the cells that migrated
onto the outer side of the upper well membrane were
stained with crystal violet. Invading cells were observed
under a microscope and counted for statistical analysis.
DMEM. Cells
Transfection of miRNA mimics and miRNA ASO
MicroRNA mimics and inhibitors (miRNA antisense oli-
gonucleotides (ASO)) were purchased from GenePharma
Co., Ltd. The sequences are provided in Table 1. In
vitro transfection of miRNAs (miR-223, lin-4 and miR-
NC) and miRNA ASO (miR-223 ASO and miR-NC
ASO) were performed using X-tremeGENE siRNA
Transfection Reagent according to the manufacturer’s
instructions (Roche, cat. no. 04 476 093 001).
Luciferase assays
To determine whether miRNAs were shuttled from
macrophages to breast cancer cells, the levels of miR-
NAs in the target cancer cells were assessed using luci-
ferase assays. Briefly, the pMIR-REPORT vector with
either the miR-223 or the lin-4 complementary
sequence in its 3’-UTR (firefly luciferase reporter vector,
Ambion) were transfected into SKBR3 cells using Lipo-
fectamine2000 (Invitrogen) according to the manufac-
turer’s instructions. A Renilla luciferase vector
(Promega) was co-transfected as an internal control for
normalization. After transfection, cells were washed to
remove untransfected plasmids or miRNAs. SKBR3 cells
were then co-cultivated in Boyden chambers with
macrophages activated with IL-4 or transfected with lin-
4 mimics, as described above, for 24-72 h. pMIR-
REPORT and Renilla luciferase activities were assayed
using the Dual-Luciferase assay kit (Promega). pMIR-
REPORT luciferase activity was normalized to the
Renilla luciferase activity. To determine whether Mef2c
mRNA is a target of miR-223, the pMIR-REPORT vec-
tor containing the 3’-UTR of Mef2c was co-transfected
with miR-223 into HEK-293T cells. After 24-48 h of
incubation, the cells were lysed, and luciferase activity
was detected as described above.
Shuttling assays for fluorescently-labeled miRNA
To further visualize the shuttling of miRNAs, Cy3-
labeled miRNAs (GenePharma) were transfected into
macrophages, as described above. Macrophages were
washed to remove the residual transfection reagent 24-h
after transfection. Macrophages carrying Cy3-miRNA
were then placed onto transwell inserts, and SKBR3
cells were seeded in the lower wells of Boyden cham-
bers. After incubation for 24-48 h, SKBR3 cells were
collected for fluorescence microscopy and flow cyto-
metric analyses.
Western blots
Cells were lysed with RIPA lysis buffer and protease inhi-
bitors (Sigma-Aldrich). Nuclear protein was collected
according to previously described protocols [23]. A total
of 20 μg of protein per sample was separated on SDS-
PAGE gels and transferred onto nitrocellulose membranes.
Membranes were blocked and incubated with antibodies
against b-catenin (1:1000, Cell Signaling Technology,
#9587) or Mef2c (1:1000, Cell Signaling Technology,
#9792) overnight at 4°C. Primary antibody incubation was
followed by incubation with HRP-conjugated secondary
antibodies (Amersham Pharmacia). HRP signals were then
visualized by enhanced chemiluminescence (Pierce).
Confocal microscopy
Cells prepared on coverslips were fixed in 4% PFA, trea-
ted with 0.3% Triton X-100, blocked with 5% BSA and
incubated with an anti-b-catenin antibody (1:200, Cell
Signaling Technology, #9587) overnight at 4°C. After
being washed with PBS, cells were incubated with a
FITC-conjugated secondary antibody for 1 h and coun-
terstained with PI prior to inspection under a confocal
microscope.
Immunofluorescence
SKBR3 breast cancer cells were co-cultured with Cy3-
preloaded macrophages for 24-48 h. After co-culture,
both macrophages and SKBR3 were fixed in 4% PFA,
treated with 0.1% Triton X-100, blocked in 3% BSA and
incubated with an anti-CD68 antibody (1:200, Dako) for
2 h at room temperature. After being washed, cells were
incubated with an Alexa Fluor 488-conjugated second-
ary antibody, and then counterstained with DAPI before
inspection under fluorescence microscope.
Immunohistochemistry
All tumor samples of invasive breast cancer were
obtained from female patients at the No. 2 Affiliated
Table 1 The sequences of microRNA mimics and
inhibitors
hsa-miR-223 UGUCAGUUUGUCAAAUACCCCA
miR-NC (negative control) UUCUCCGAACGUGUCACGUTT
hsa-miR-223 inhibitor (ASO)UGGGGUAUUUGACAAACUGACA
miR-NC ASOCAGUACUUUUGUGUAGUACAA
Yang et al. Molecular Cancer 2011, 10:117
http://www.molecular-cancer.com/content/10/1/117
Page 3 of 13

Page 4

Hospital, Sun-Yat-Sen University. All patient samples
were collected with informed consent according to the
Internal Review and the Ethics Boards of the Sun-Yat-
Sen Memorial Hospital, Sun-Yat-Sen University. Sam-
ples were fixed, paraffin embedded and sectioned into
5- μM slices. Macrophages were visualized by immuno-
histochemistry staining using an anti-CD68 (1:200,
Dako) antibody. Bound primary antibody was detected
by using a horseradish peroxidase conjugated secondary
antibody, which was then developed in DAB solution
(Dako). Pictures were taken under a light microscope.
Exosome purification and labeling
The same number of IL-4-activated or unactivated
macrophages were cultured in exosome-free medium
(DMEM containing 10% FBS that was depleted of exo-
somes by ultra-centrifuging at 100,000 × g for 18 h).
Conditioned media were collected after 3-5 days of
incubation. Exosomes were purified by differential cen-
trifugation. Briefly, the conditioned media were centri-
fuged at 500 × g for 30 min and 16,500 × g for 20 min
to eliminate cells and cellular debris, respectively. Super-
natants were filtered through 0.22 μm filters. Exosomes
were pelleted by ultracentrifugation at 120,000 × g for
180 min, washed in PBS, pelleted again and resuspended
in PBS. Exosome preparations were stained with CM-
DiI (CellTracker, C7000), a fluorescent dye that labels
the plasma membrane, according to the manufacturer’s
instructions. Next, exosomes were diluted in complete
medium (DMEM with 10% FBS) and were added into
the cell cultures. At the indicated time points, cells were
examined under a confocal microscope and analyzed
using flow cytometry.
RNase treatment of Exosomes
The culture of unactivated and IL-4-activated macro-
phages and the process through which exosomes were
collected are described above. Exosomes were treated
with RNase according to previously described protocols
[24]. Briefly, separated exosomes were incubated with
RNase A at a final concentration of 100 U/ml, with or
without 1% Triton X-100, at room temperature for 30
min. Exosomes were washed with PBS to remove resi-
dual RNase and Triton X-100. Exosomes were incubated
with breast cancer cells prior to performing invasion
assays.
Electronic microscopy
Exosome preparations were mixed with equal quantities
of freshly prepared 4% paraformaldehyde for 20 min.
Samples were washed in water, pelleted by ultracentrifu-
gation and then fixed for 5 min in 1% glutaraldehyde.
After this process, exosomes were re-suspended in
water, and 5 μl of the samples were loaded onto carbon-
coated formvar grids. Exosomes were stained for 10 min
with saturated aqueous uranyl and examined using an
electron microscope (Hitachi S-4800 FEG SEM).
Statistical analyses
All data are expressed as mean ± SD. Statistical analyses
were performed using paired Student’s t-tests.
Results
Co-cultivation with IL-4-activated macrophages elevates
miR-223 levels in breast cancer cells
Because TAMs located in the stroma of breast cancers
are primarily M2 macrophages activated by IL-4-pro-
ducing CD4+T cells [25], we mimicked this TAM-
populated microenvironment by co-cultivating SKBR3
breast cancer cells with IL-4-activated MDMs in a
Boyden chamber, which prevents direct cell-cell con-
tact. To optimize a physiologically-relevant cell num-
ber ratio for the co-culture experiments, we quantified
the amount of macrophage infiltration present in
patient samples that were pathologically diagnosed as
invasive breast cancer. As shown in Additional file 1
Figure S1(A), a large number of macrophages (CD68-
positive) infiltrated breast tumors, especially in the
tumor-associated stromal border, where many invasive
tumor cells were also located. Because previous studies
suggested that macrophages produce exosomes, which
shuttle proteins or microRNAs (miRNAs) into adjacent
cells within the microenvironment, we focused on the
neighboring tumor cells and macrophages. We calcu-
lated the cell ratio based upon the neighboring tumor
cell and macrophage populations in a local context as
indicated in Additional file 1 Figure S1(B). The ratio of
tumor cells to their adjacent macrophages ranged from
1:1 to 1:7. Subsequently, we tested the effects of
macrophage to breast cancer cell ratios, ranging from
1:1 to 5:1, in the co-culture system. The effects of
macrophages on breast cancer cells were observed at
ratios starting from 1:1. First, we screened for miRNAs
that were differentially expressed between macrophages
and breast cancer cells using a DiscovArray miRNA
microarray. All the microarray data were listed in
Additional file 2 Table S1. We found three microRNAs
that were abundantly expressed in macrophages but
not in SKBR3 or MDA-MB-231 breast cancer cells
(Figure 1A). Using qRT-PCR, we confirmed that miR-
223 was overexpressed in IL-4-activated MDMs but
was not highly expressed in either SKBR3 or MDA-
MB-231 breast cancer cells (Figure 1B). Additionally,
miR-223 is involved in cancer progression [26,27];
therefore, we focused on miR-223 expression levels in
breast cancer cells after co-cultivation with IL-4-acti-
vated MDMs. We seeded breast cancer cells and
macrophages in co-culture Boyden chambers, as
Yang et al. Molecular Cancer 2011, 10:117
http://www.molecular-cancer.com/content/10/1/117
Page 4 of 13

Page 5

described in Figure 2A. Interestingly, breast cancer
cells co-cultured with IL-4-activated macrophages
exhibited a profound increase in cellular miR-223
levels (2.7 ± 0.5- and 3.3 ± 0.98-fold in SKBR3 and
MDA-MB-231 cells, respectively) relative to cells that
were not co-cultured (blank) or were co-cultured with
unactivated macrophages (Figure 1C). To evaluate the
function of elevated miR-223 in breast cancer cells, we
used a luciferase reporter gene containing a sequence
complementary to miR-223 in its 3’-UTR. Co-culturing
with IL-4-activated macrophages reduced luciferase
reporter activity in SKBR3 breast cancer cells (Figure
1D and Additional file 3 Figure S2), which suggests
that the elevated miR-223 levels observed in breast
cancer cells are capable of silencing target gene expres-
sion. Furthermore, direct transfection of miR-223
mimics, but not a scrambled negative control miRNA
(miR-NC), also suppressed the reporter gene activity in
SKBR3 cells (Figure 1D). These data suggest that co-
culturing breast cancer cells with IL-4-activated
macrophages increases the level of functional miR-223
in breast cancer cells.
miRNAs released by macrophages are shuttled into
breast cancer cells
To determine whether miRNAs released by IL-4-acti-
vated macrophages are shuttled into co-cultured breast
cancer cells, we transfected macrophages with either
Cy3-labeled miR-223 or non-mammalian lin-4 miRNA
prior to co-culture with SKBR3 cells. Co-culture was
performed in a 24-well Boyden chamber with a 0.4- μm
insert (Figure 2A). Fluorescence microscopy analysis
indicated the presence of Cy3-miRNA in SKBR3 cells,
with approximately 15 positive cells per field of view,
when co-cultured with macrophages transfected with
Cy3-labeled miR-223. Fluorescence was not detected in
cells that were not co-cultured or that were co-cultured
with macrophages transfected with unlabeled miR-223
Figure 1 miR-223 was up-regulated in breast cancer cell lines co-cultured with IL-4-activated macrophages. (A) miRNA profiling in
macrophages and the breast cancer cell line SKBR3 using the DiscovArray miRNA array. Three of the miRNAs that were present at high levels in
macrophages but absent in SKBR3 cells are shown. (Un-Mac, unactivated macrophages; IL4-Mac, IL-4-activated macrophages). (B) Quantitative
real-time PCR (qRT-PCR) confirmed a high expression level of hsa-miR-223 in macrophages but not in SKBR3 or MDA-MB-231 cells. * p < 0.05. (C)
Co-culture of breast cancer cells with macrophages increased miR-223 levels in breast cancer cells. Breast cancer cell lines (SKBR3 and MDA-MB-
231) were cultured alone (blank) or co-cultured with unactivated or IL-4-activated macrophages without direct cell-cell contact. miR-223
expression in SKBR3 and MDA-MB-231 cells was detected using qRT-PCR. Relative levels of miR-223 expression normalised to U6 rRNA levels are
presented. * p < 0.05. (D) Functional assay of miR-223 in breast cancer cells. miR-223-targeting luciferase reporter containing a miR-223
complementary sequence within the 3’-UTR of the luciferase reporter gene was constructed. SKBR3 cells were transfected with the reporter gene
and cultured alone (blank) or co-cultured with IL-4-activated macrophages. As controls, SKBR3 cells were co-transfected with the reporter gene
and miR-NC, miR-660 or miR-223. Relative luciferase activities (normalised to Renilla luciferase activity) are presented. * p < 0.05.
Yang et al. Molecular Cancer 2011, 10:117
http://www.molecular-cancer.com/content/10/1/117
Page 5 of 13