Targeting FSTL1 Prevents Tumor Bone Metastasis and ... · Targeting FSTL1 Prevents Tumor Bone Metastasis and Consequent Immune Dysfunction Chie Kudo-Saito, Takafumi Fuwa, Kouichi

Microenvironment and Immunology

Targeting FSTL1 Prevents Tumor Bone Metastasis andConsequent Immune Dysfunction

Chie Kudo-Saito, Takafumi Fuwa, Kouichi Murakami, and Yutaka Kawakami

AbstractBonemetastasis greatly deteriorates the quality of life in patientswith cancer. Althoughmechanisms have been

widely investigated, the relationship between cancer bone metastasis and antitumor immunity in the host hasbeen much less studied. Here, we report a novel mechanism of bone metastasis mediated by FSTL1, a follistatin-like glycoprotein secreted by Snailþ tumor cells, whichmetastasize frequently to bone.We found that FSTL1 playsa dual role in bonemetastasis—in oneway bymediating tumor cell invasion andbone tropismbut also in a secondway by expanding a population of pluripotent mesenchymal stem-like CD45�ALCAMþ cells derived from bonemarrow. CD45�ALCAMþ cells induced bonemetastasis de novo, but they also generated CD8low T cells with weakCTL activity in the periphery, which also promoted bone metastasis in an indirect manner. RNA interference-mediated attenuation of FSTL1 in tumor cells prevented bone metastasis along with the parallel increase inALCAMþ cells andCD8low T cells. These effectswere accompanied byheightened antitumor immune responses invitro and in vivo. In clinical specimens of advanced breast cancer, ALCAMþ cells increasedwith FSTL1 positivity intumor tissues, but not in adjacent normal tissues, consistent with a causal connection between these molecules.Ourfindings define FSTL1 as an attractive candidate therapeutic target to prevent or treat bonemetastasis, whichremains a major challenge in patients with cancer. Cancer Res; 73(20); 6185–93. �2013 AACR.

IntroductionBone metastasis of tumor cells is frequently seen in patients

with cancer, particularly with breast cancers and prostatecancers, and greatly deteriorates the quality of life in patients,leading to poor prognosis (1). Chemokines and its receptors areone of the representative molecules regulating bone tropism oftumor cells. For examples, CXCR4þ tumor cells are attracted byCXCL12 secreted from stromal cells in bone marrow (2), andCXCL12maintains proliferation and survival of the tumor cells(3). Bone-derived TGFb and PDGF also promote tumor pro-gression in bone marrow (4). CCL2 is known as anotherchemokine critical for bone metastatic mechanism in variouscancers (5, 6). The excess of tumor growth in bone marrowresults in disruption of skeletal integrity, and causes abnormalosteogenesis or osteolysis in patients with cancer (1). One ofthe most influential molecules is RANKL, which is highlyexpressed in normal mesenchymal cells including osteoblastsand stromal cells, and metastatic tumor cells undergoingepithelial-to-mesenchymal transition (EMT; ref. 7).

A variety of therapeutics targeting these molecules hasbeen developed for treating bone metastasis (4). However,their subjects are mostly postmetastasis events, but notpremetastatic state. Elimination of the trigger for tumorseeding toward bone would be a higher priority of cancertherapy rather than treatment of tumor progression andosteo-imbalance in bone marrow. Moreover, although bonemetastasis mechanisms have been widely investigated, therelationship with antitumor immunity has been rarelyexplored, although bone marrow is an essential organ forhematopoiesis and immune responses (8). We previouslyshowed a close relationship between cancer metastasis andimmunosuppression focusing on an EMT-governing tran-scriptional factor Snail (9). Snail expression in tumor cellspromotes both tumor metastasis and induction of immu-noregulatory cells simultaneously. Some immunoregulatorymembers such as mesenchymal stem cells (MSC; ref. 10) andmyeloid-derived suppressor cells (MDSC; ref. 11) are origi-nated from bone marrow. To understand the interactionbetween bone metastasis and antitumor immunity mayprovide a new insight into bone metastasis mechanism. Inthis study, we attempted to elucidate a new mechanism ofcancer bone metastasis from the immunologic perspective.

Materials and MethodsCell lines and mice

Murine melanoma B16-B10 cells were kindly provided byCell Resource Center for Biomedical Research at TohokuUniversity in Japan. Human tumor cells including melanomaHS294T and pancreatic cancer Panc1 were purchased fromAmerican Type Culture Collection, and were authenticated by

Authors' Affiliation: Division of Cellular Signaling, Institute for AdvancedMedical Research, Keio University School of Medicine, Shinjuku-ku,Tokyo, Japan

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Authors: Chie Kudo-Saito, Division of Cellular Signaling,Institute for Advanced Medical Research, Keio University School of Med-icine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. Phone: 81-3-5363-3778; Fax: 81-3-5362-9259; E-mail: [email protected]; andYutakaKawakami, E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-1364

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 6185

on January 13, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 21, 2013; DOI: 10.1158/0008-5472.CAN-13-1364

http://cancerres.aacrjournals.org/

short tandem repeat profiling. Some clones were transfectedwith plasmid vector pcDNA3.1 encoding murine or humansnail gene as described before (9), and/or with lentiviral vectorencoding GFP gene (Biogenova) for in vivo study. Tumor cellswere cultured routinely in 10% FBS/Dulbecco's Modified EagleMedium (Invitrogen), and sometimes in 2% FBS/Opti-MEM(Invitrogen) for preparation of supernatant used for assays.Female C57BL/6 and C57BL/6-CAG-EGFP mice (designatedEGFPþ mice) were purchased from SLC, and were maintainedunder pathogen-free conditions until use according to theprotocols approved by the Animal Care and Use Committeeat Keio University School of Medicine (Tokyo, Japan).

Stimulation and characterization of bone marrow cellsBone marrow cells (BMC) were stimulated with 25%

tumor supernatant (1 � 105 cells/10 mL/3 days, 0.22 mmfiltration) or FSTL1 (5 ng/mL; Thermo Scientific) in 2% FBS/Opti-MEM for 7 to 10 days. CD45� BMCs were sorted fromthe culture using a MACS system with microbeads-conju-gated monoclonal antibody (mAb; Miltenyi Biotec), and weretested for MSC activities: differentiation activity into mes-enchymal lineages using Mesenchymal Stem Cell FunctionalIdentification Kit (R&D Systems), self-renewal activity using2% agarose medium, and immunoregulatory activity usingcoculture with splenic T cells in vitro, and coinjection withtumors in vivo. The details were described in SupplementaryData. For tracking CD45�ALCAMþ cell division, the sortedCD45�ALCAMþ cells were stained with red fluorescent dyePKH26 (Sigma) before assay.

Functional analysis of tumor cellsTumor cells (5 � 104) were assessed for Matrigel invasion

(6 hours) and chemotaxis (4 hours) to CCL2 (R&D Systems)or CXCL12 (R&D Systems) using a Transwell chamber with amembrane (pore size, 8 mm; BD Biosciences). FSTL1 in thesupernatant fluids was measured using the ELISA kit (R&DSystems) according to the manufacturer's instructions. Forknockdown of fstl1 or snail expression, the specific siRNAs orthe scrambled oligonucleotides as a control (3 mg; Invitrogen)were complexed with jetPEI (PolyPlus), and were thentransfected into tumor cells. The transfection efficacy wasvalidated by reverse transcriptase (RT)-PCR 1 to 2 days aftertransfection. Several kinds of siRNAs targeting on a differentsequence position of the genes were initially used, and asiRNA having the highest knockdown efficiency was mainlyused. In a setting, tumor cells were stimulated with FSTL1(5 ng/mL) in combination with/without CCL2 (5 ng/mL) for3 days before assay.

Flow-cytometric analysisAfter Fc blocking, cells were stained with immunofluo-

rescence (Fluorescein isothiocyanate, phycoerythrin, orCyChrome)-conjugated mAbs specific for mouse and/orhuman ALCAM (eBioscience), CCR2 (R&D Systems), CCR5,CD146 (Abcam), CD271, CD45, CD8, CXCR4, DIP2A (Santa-Cruz), Foxp3, PDGFRa (R&D Systems), RANK (R&D Sys-tems), RANKL (Avnova), tetramer for gp70 (a tumor antigenexpressed in B16-F10; MBL), or the appropriate isotype

control antibodies. The immunofluorescence-conjugatedsecondary antibodies were used if necessary. Antibodiesexcept designated ones were purchased from BD Pharmin-gen. For intracellular staining, cells were treated with Cyto-fix/Cytoperm solution (BD). The immunofluorescence wasanalyzed and compared with the isotype controls by Cell-quest software using a FACSCalibur cytometer (BD).

Analysis of clinical tissuesParaffin-embedded tumor tissue sections obtained from

patients with stage III breast cancer (n ¼ 26) and normalmammary tissue sections (n ¼ 6) were purchased (HumanCancer Tissue Array; MBL). To analyze mRNA expression,semiquantitative RT-PCR was conducted using AmpdirectPlus (Shimadzu) and the specific paired primers for humanfstl1, alcam, and gapdh (Supplementary Data) as describedbefore (9). The digital images of the bands were quantifiedusing NIH ImageJ software, and the signal intensity wasnormalized to gapdh expression as a control (normal tissues,< 0.1). To analyze protein expression, immunohistochemicalanalysis was conducted with anti-FSTL1 mAb (Abcam) andanti-ALCAM mAb (LSBio) according to a general protocol(Supplementary Data). As immunofluorescence intensity, pixelcounts were automatically measured at two fields under thesame 40, 6-diamidino-2-phenylindole level using a LSM700Laser Scanning Microscope (Carl Zeiss), and the average datawere plotted in graphs (FSTL1, < 66, ALCAM, < 233 in normaltissues).

FSTL1 blocking therapy in vivoTo evaluate siRNA efficacy on tumor dissemination from

the primary tumor site, and on tumor metastatic noduleformation, tumor cells were implanted into mice bothsubcutaneously (5 � 105) to prepare easy accessible sitesfor siRNA injection, and intravenously (1 � 105) to makevisible bone metastasis for a short term while F10-snailþ

tumor-implanted mice were living. Seven days later, poly-ethyleneimine-complexed siRNAs (5 mg) were injected intothe subcutaneous tumors (n ¼ 3–5/group), and the trans-fection efficacy was validated by RT-PCR or Western blot-ting 1 to 2 days after injection. Tumor volume (0.5 � length� width2, mm3) was measured 1 to 2 times a week. Tendays after siRNA injection, the subcutaneous tumors, bonemarrow, and spleens were analyzed by flow cytometry, andsplenic CD8þ T cells were tested for IFNg production (24hours) and tumor killing (4 hours) as described before (9).

Statistical analysisSignificant differences (P < 0.05) were evaluated using the

unpaired two-tailed Student t test. The significance was con-firmed by the nonparametric Mann–Whitney U test using thecumulative data of the repeated experiments, when the num-ber of n was small in one experiment. Mouse survival wasanalyzed by Kaplan–Meier method and ranked according tothe Mantel–Cox log-rank test. Correlation between tumorbone metastasis and ALCAMþ cells in mice, or between FSTL1andALCAMexpressions in clinical tissueswas evaluated by thenonparametric Spearman rank test.

Kudo-Saito et al.

Cancer Res; 73(20) October 15, 2013 Cancer Research6186




ResultsSnailþ tumor cells generate CD45�ALCAMþ

mesenchymal stem cells in bone marrowWe previously establishedmurine snail-transduced B16-F10

melanoma (designated F10-snailþ) having typical EMT fea-tures (9). When the F10-snailþ tumor cells were implantedboth subcutaneously and intravenously in C57BL/6 mice,

tumor metastatic dissemination was more severely observedin various tissues such as lymph node, lung, and bone marrow,as compared with that of the mice implanted with B16-F10tumors transduced with empty vectors (designated F10-mock;Supplementary Fig. S1A), although the subcutaneous tumorgrowth was slower than F10-mock growth as shown before(12). Particularly, there are many mice having black bones due

CD146

F10-mock: 6.01

F10-snail+: 46.31

ALCAM

F10-mock: 8.01

F10-snail+: 44.30

CD271

F10-mock: 20.22

F10-snail+: 70.84

CCR5

F10-mock: 48.92

F10-snail+: 80.89

PDGFRαF10-mock: 25.56

F10-snail+: 66.47

DIP2A

F10-mock: 31.03

F10-snail+: 58.10

Intensity

D

No

su

p.

F10-m

ock

F1

0-s

na

il+

2.66 9.30 25.39

Ad

ipo

cyte

s

Oste

ob

lasts

No sup. F10-mock F10-snail+

Sp

he

res

SS

C

CD45−

No sup. F10-mock F10-snail+ A

No

. o

f ce

lls (

x1

05)

Bulk 0.8

0.6

0.4

0.2

0.0 N

o s

up

.

F10- m

ock

F1

0-s

na

il+

CD45

B

T-c

ell

pro

life

ratio

n

(OD

450–65

5 n

m)

Cyto

kin

es (

ng

/mL

) C

Snail+

+aAL

10.86

Mock

7.16

Snail+

+Cont

20.59

SS

C

CD45


+ aALCAM

2.0

1.5

1.0

0.5

0.0

E

IFNγ 1.0

0.8

0.6

0.4

0.2

0.0

No

BM

Cs

No

su

p.

F10-m

ock

F1

0-s

na

il+

+ a

AL

CA

M

No

BM

Cs

No

su

p.

F10- m

ock

F1

0-s

na

il+

+ a

AL

CA

M

No

BM

Cs

No

su

p.

F10-m

ock

F1

0-s

na

il+

+ a

AL

CA

M

No

BM

Cs

No

su

p.

F10-m

ock

F10-s

nail+

+ a

AL

CA

M

TNFα IL-10

��

��

��

BMCs Treated with

0.0 0.5 1.0 1.5

5.0 4.0 3.0 2.0 1.0 0.0

BMCs Treated with BMCs Treated with BMCs Treated with

(CD45− cells)

FABP4+

Osteocalcin+

* *

* *

* *

*

12.25 27.44 53.95

PKH26-labeled CD45-ALCAM+ cells

No. of CD45− cells (x105)


Figure 1. Snailþ tumor cells induce CD45�ALCAMþMSCs. BMCs were stimulated with supernatant of murine melanoma B16-F10 cells transfected with snail(F10-snailþ) or empty vector (F10-mock) for 7 to 10 days and were analyzed for CD45� cells by flow cytometry (A). The cultured BMCs are shown inphoto (scale bar, 50mm). The number ofCD45�BMCs in the culturewas calculated (n¼3,mean�SD). The sortedCD45�BMCswere tested for differentiationinto FABP4þ adipocytes and osteocalcinþ osteoblasts (B), sphere formation (B), and immunosuppression (C). Scale bar, 200 mm. In C, the sortedCD45� BMCs were added to the T-cell proliferation system (1:10) with/without anti-ALCAM neutralizing mAb (5 days), and cytokines in the supernatantfluids were measured (n ¼ 3, mean � SD). D, the CD45� BMCs were further analyzed for MSC marker expression by flow cytometry (thin lines, F10-mock–treated BMCs; thick lines, F10-snailþ-treated BMCs). PKH26-labeled CD45�ALCAMþ BMCs proliferated in response to F10-snailþ supernatant,and the CD45� BMC increase was suppressed by anti-ALCAM neutralizing mAb (E). The number of BMCs was calculated (n ¼ 3, mean � SD). �, P < 0.002versus F10-mock–treated BMCs. ^^, P < 0.002 versus control IgG. Data in each panel are representative of three independent experiments. PDGFRa,platelet-derived growth factor receptor a; IL, interleukin.

Role of FSTL1 in Cancer Bone Metastasis

www.aacrjournals.org Cancer Res; 73(20) October 15, 2013 6187




to melanoma metastasis in the subcutaneous/intravenoustumor model (41/48 ¼ 85.4%; F10-mock, 1/12 ¼ 8.3%), andthe F10-snailþ tumor cells spontaneously disseminated tobone rather than other tissues even after only subcutaneousimplantation (Supplementary Fig. S1B). These results suggestthat Snail regulates bone tropism of B16-F10 tumor cells.

To examine how Snailþ tumor cells would affect bonemarrow microenvironment, we stimulated BMCs with F10-snailþ supernatant for 7 to 10 days. Many colony-forming cellswere observed in the culture, and CD45� cells significantlyincreased as compared with the culture with F10-mock super-natant (P ¼ 0.0002; Fig. 1A). Indeed, in addition to tolerogenicdendritic and regulatory T cells (Treg; but not MDSCs)reported before (9), CD45� cells also increased in the F10-snailþ-implanted mice (Supplementary Fig. S1C). CD45�

BMCs sorted from the culture with F10-snailþ supernatantmore frequently differentiated into mesenchymal lineagessuch as adipocytes and osteoblasts, and formed large sphere

colonies as compared with the CD45� BMCs treated with noneor F10-mock supernatant (Fig. 1B). This indicates that F10-snailþ-induced CD45� BMCs are MSCs having pluripotencyand self-renewal activity. The Snailþ tumor-induced MSCs(designated sMSCs) significantly suppressed splenic T-cellproliferation and cytokine production, indicating immunosup-pressive activity (P < 0.002 vs. F10-mock–treated BMCs; Fig.1C). These results suggest that Snailþ tumor cells expandMSCsin bone marrow microenvironment.

The sMSCs highly expressed someMSCmarkers reported inhuman MSCs (13), particularly ALCAM and CD146, as com-pared with the F10-mock–treated BMCs (Fig. 1D). However,the CD45�CD146þ subpopulation hardly differentiated intoosteoblasts (Supplementary Fig. S2). In contrast, the CD45�

ALCAMþ subpopulation showed high differentiation activity(Supplementary Fig. S2), proliferated in response to F10-snailþ

supernatant (Fig. 1E). ALCAM blocking with the specific mAbsignificantly suppressed the sMSC increase (P¼ 0.001; Fig. 1E),

17.25 12.83 8.65 10.11 17.96 32.83

Tetr

am

er

(CD3+ cells in spleen) CD8

B

A 1,500

1,000

500

0

i.v.

Tu

mo

r vo

lum

e (

mm

3)

BMCs Treated with

Tu

mo

r o

nly

No

su

p.

F10-m

ock

F1

0-s

na

il+

C

100 101 102 103 104

FL3-Height100 101 102 103 104

FL3-Height100 101 102 103 104

FL3-Height

2.67 0.34 0.17

EGFP+ BMCs in the tumors

100 101 102 103 104

FL1-Height

2.99 100 101 102 103 104

FL1-Height100 101 102 103 104

FL1-Height

0.06 0.02

GFP+ Tumor cells in the BM


EGFP+ BMCs Treated with

ALC

AM

/tum

or

F10-mock–treated BMCs

F10-snail+-treated BMCs

Tumor (s.c.) BM

Tumor (s.c.) BM

100

80

60

40

20

0 0 5 10 15 20 25

No. of ALCAM+ in BM

0 10 20 30 40

No. of ALCAM+ in tumor

100

80

60

40

20

0

R 2 = 0.609

P < 0.05

R2 = 0.533

P < 0.05

AL

CA

M/tum

or

No

. o

f tu

mo

r ce

lls in

BM

2,500

2,000

1,500

1,000

500

0

Tum

or

(mm

3)

2.0

1.5

1.0

0.5

0.0

BMCs Treated with

IFN

γ fr

om

CD

8+

(ng

/mL

)

Tumor only F10-mock F10-snail+

Tumor + BMCs treated with

Tu

mo

r o

nly

No

su

p.

F1

0-m

ock

F1

0-s

na

il+

s.c. (mixed)

D 0.0 0.5 1.0 1.5

TGFβ TNFα TSP1 CCL2 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5

No sup.

F10-mock

F10-snail+

(ng/mL)

Supernatant of the BMCs treated with

0 50 100 150 200 0 20 40 60 80

No sup. Control Ig aTNFα aCCL2

Invasion (no. of cells) No. of tumor cells in BM sM

SC

su

p.

No BMCs

Control Ig

aCCL2��

sM

SC

s

* *

*

*

*

*

Figure 2. The Snailþ tumor-induced MSCs induce both tumor progression and immune dysfunction via generation of impaired CD8low T cells. A, mice weresubcutaneously (s.c.) implanted with F10-mock tumor cells (3 � 105) and 7 days later, received intravenous (i.v.) injection with the 10-day–culturedCD45�BMCs (3� 105) obtained from EGFPþmice (n¼ 5 per experiment). Eighteen days later, tumor size was measured (mean� SD), and tumor-infiltratingcells were analyzed for the injected EGFPþ BMCs by flow cytometry. B, mice were subcutaneously injected with the mixture of CD45� BMCs (3 � 105) andGFPþ F10-mock tumor cells (3 � 105). Twenty-five days later, tumor size was measured (n ¼ 5, mean � SD), and BMCs and SPCs were analyzedfor the injected GFPþ tumor cells or CD8þgp70-tetramerþ T cells by flow cytometry (n ¼ 3, pooled). CD8low T cells are shown at the top left in dot plot.IFNg-producible activity of the splenic CD8þ/low T cells was evaluated (n ¼ 3, mean � SD). C, the subcutaneous tumors and bone marrow (BM) tissueswere harvested from the mice and were immunocytochemically analyzed for ALCAMþ cells and GFPþ tumor cells. The number of positive cells wasmicroscopically counted at triplicate sites, and the average data were plotted in graphs (n ¼ 15). Scale bar, 50 mm. D, some EMT inducers in the CD45�

BMC-cultured supernatant (7 days) were measured (n ¼ 3, mean � SD). To evaluate the EMT-inducible activity, F10-mock tumor cells were stimulatedwith supernatant of F10-snailþ-treated BMCs (sMSCs) with/without anti-TNFa or anti-CCL2 neutralizing mAb for 3 days and were tested by Matrigelinvasion assay (n ¼ 3, mean � SD). Also, mAbs were intratumorally injected into the 7-day subcutaneous tumors mixed with sMSCs (3 � 105) andGFPþF10-mock cells (3�105), and 18days later, the number ofGFPþ tumor cells in bonemarrowwasmicroscopically counted (n¼5,mean�SD). �,P<0.01versus BMCs treated with none or F10-mock. ^^, P < 0.01 versus control Ig. Data in each panel are representative of three independent experiments.

Kudo-Saito et al.





and significantly improved T-cell responses (P < 0.002; Fig. 1C).These data suggest that ALCAM is partly but critically respon-sible for the sMSC properties. ALCAM is a member ofthe immunoglobulin (IgG) superfamily, and binds to CD6-expressed T cells with high affinity, and to ALCAM-expressed

various cells with low affinity (14). The ALCAM–ALCAMhomophilic binding is strengthen by ALCAM clustering onthe cell surface, and has been recently reported as a criticalmolecular interaction for proliferation and colonization ofcancer stem cells (14). However, the functional role of ALCAMin MSCs remains unclear. Probably, sMSC expansion may bemediated by this homophilic interaction via ALCAM cluster-ing, and T-cell suppression may be mediated by the hetero-philic interaction.

The sMSCs induce both tumor bone metastasis andimmune dysfunction via generation of impairedCD8low T cells

We next conducted in vivo study using GFPþ tumor cells orEGFPþ BMCs for cell tracking after injection in mice. WhenEGFPþ sMSCs were intravenously injected in mice on day 7after F10-mock subcutaneous implantation, the sMSCs mainlyinfiltrated in the tumors, and tumor growth was significantlypromoted (P¼ 0.002 vs. F10-mock–treated BMCs; Fig. 2A). Tosee how the infiltrating sMSCs would affect tumor microen-vironment more clearly, the sMSCs weremixed with F10-mocktumor cells, and injected subcutaneously in mice. The tumorsgrew more aggressively (P ¼ 0.013 vs. F10-mock–treatedBMCs), and CD8low T cells predominantly increased in thetumors (Supplementary Fig. S3A). CD8low T cells also increasedin spleen that was hypertrophic with abundant adipose tissueshaving increased CD45� cells including the injected sMSCs(Fig. 2B and Supplementary Fig. S3B). The CD8þ/low T cellssorted from spleen hardly produced IFNg in response to atumor antigen gp70 peptide, although weakly reacted to gp70tetramer (Fig. 2B). It is inferred that such CTL dysfunctionwould be induced by the sMSCs infiltrating in spleen. In thesMSC-injected mice, tumor cells increased in bone marrow(Fig. 2B and Supplementary Fig. S3C), and the significantcorrelation was seen between the number of tumor cellsand ALCAMþ cells in bone marrow and the primary tumorsite (P < 0.05; Fig. 2C). This suggests that sMSCs could inducebone metastasis de novo followed by further increase ofALCAMþ sMSCs. Indeed, stimulation with the sMSC super-natant significantly enhanced tumor invasion in vitro (P ¼0.001), and bone metastasis in vivo (P ¼ 0.006; Fig. 2D). Wefound that the sMSCs significantly more produced CCL2among soluble factors tested as EMT-inducible molecules (P¼ 0.004 vs. F10-mock–treated BMCs; Fig. 2D). CCL2 blockingwith the specific mAb significantly inhibited the sMSC-induced tumor invasion in vitro (P ¼ 0.002 vs. control IgG)and bone metastasis in vivo (P ¼ 0.006; Fig. 2D). Theseresults suggest that CCL2 is, at least in part, involved in thesMSC-induced bone metastasis mechanism. There may be amolecule that directly regulates this mechanism more sig-nificantly. The further study is needed. Tumor metastasis-promoting activity of MSCs has been already shown usingxenograft model implanted with human breast cancer cellsand human bone marrow-derived MSCs, and CCL5 has beenshown as a molecule essential for the activity (15). However,the sMSCs produced no CCL5, and no lung metastaticnodules were observed in the coinjected mice in our study(data not shown).

Mo

ck

Sn

ail+

A

Sn

ail+

B

FS

TL

1 (

ng

/mL

) 80

60

40

20

0

A

Co

ntr

ol

siF

ST

L1

#1

siF

ST

L1

#2

siS

na

il #

1

siS

na

il #

2

Control siFSTL1

18.1136.12

10.3228.01

Intensity

CC

R2

CX

CR

4

300

200

100

0

No

. o

f ce

lls

F1

0-m

ock

None

Co

ntr

ol

siF

ST

L1 #

1

siF

ST

L1 #

2

siS

na

il #

1

siS

na

il #

2

F10-snail+

100

80

60

40

20

0

% P

ositiv

e c

ells

CCR2 CXCR4��

��

��

��

RANKL RANK

No

ne

FS

TL

1

CC

L2

Co

mb

.

No

ne

FS

TL

1

CC

L2

Co

mb

.

C

D

FS

TL

1 (

ng

/mL

)

15

10

5

0

B

Mo

ck

Sn

ail+

A

Sn

ail+

B

Mo

ck

Sn

ail+

A

Sn

ail+

B

B16-F10 HS294T Panc1

100

80

60

40

20

0

300

200

100

0

No

. o

f ce

lls

No

ne

FS

TL

1

CC

L2

Co

mb

.

No

. o

f ce

lls 200

100

0

B16-F10

Human tumors

**

**

**

*

*

**

**

**

*

*

*

*

0 20 40 60 80 100

ControlsiFSTL1Control

siFSTL1Control

siFSTL1

Chemotaxis (no. of cells)

None

CCL2

CXCL12*

*

Figure 3. FSTL1 regulates bone tropism of Snailþ tumor cells. A, increaseof FSTL1 in the 3-day–cultured supernatant of snail-transfected tumorcells (murine melanoma B16-F10, human melanoma HS294T, andhuman pancreatic cancer Panc1; n¼ 3, mean � SD). �, P < 0.001 versusmock. B, transfectionwith siRNAs specific for fstl1 or snail into F10-snailþ

cells inhibitedFSTL1production. �,P<0.001 versus control siRNA.C, thesiRNA-fstl1 transfection also inhibited cell invasion (n ¼ 3, mean � SD),bone metastasis-associated molecule expressions, and chemotacticactivity to the ligands. �, P < 0.03 versus control siRNA. D, FSTL1stimulation enhanced tumor invasion (n ¼ 3, mean � SD) and bonemetastasis-associated molecule expressions of F10-mock cells (n ¼ 3,mean�SD). �,P < 0.002 versus none,^^,P < 0.01 versus FSTL1 alone.Data in each panel are representative of three independent experiments.






FSTL1 is a key molecule governing Snail-induced bonemetastasis

To identify the specific molecule regulating the Snail-induced bone metastasis, we reviewed literature focusing onmolecules that associate with "bone"-related diseases such asarthritis, leukemia, and osteosarcoma, and about 20 candi-dates were chosen for analysis of mRNA and protein expres-sions. FSTL1 was the molecule that significantly and widelyupregulated in murine and human Snailþ tumor cells (Fig. 3Aand Supplementary Fig. S4). FSTL1 is a follistatin-like glyco-protein that binds to DIP2A receptor (16, 17). FSTL1 is orig-inally known to regulate organ tissue formation in embryos(18), and its increase has been reported in some diseases suchas rheumatoid arthritis (19, 20) and osteosarcoma (21). How-ever, the functional role in cancer and metastasis has neverbeen shown. Transfectionwith fstl1-specific siRNAs into Snailþ

tumor cells significantly inhibited tumor invasion (P < 0.004),bone metastasis-associated molecule expressions such asCCR2 and CXCR4, and migration to the ligand such as CCL2and CXCL12 (Fig. 3B and C and Supplementary Fig. S5), inaddition to the typical reversal changes of EMT (e.g., increasedadhesion, and decreased CD44 expression; data not shown).When Snail�/low tumor cells were stimulatedwith FSTL1, thesemolecular expressions and tumor invasion were significantlyenhanced (P < 0.002; Fig. 3D). This suggests that FSTL1 is a

critical effector molecule for the high metastatic activity ofSnailþ tumor cells. CCL2 combination synergistically elevatedCXCR4 andRANKexpressions (Fig. 3D). It is inferred that CCL2produced from the sMSCs and Snailþ tumor cells (12) maycollaborate for promotion of bone metastasis. More interest-ingly, CD8low T-cell–inducible sMSC-like cells increased whenBMCs were stimulated with FSTL1 (Fig. 4). The sMSCs highlyexpressed DIP2A (Fig. 1D). These results suggest that FSTL1 isresponsible for both tumor metastasis and immune dysfunc-tion in cancer bone metastasis mechanism.

We validated the FSTL1 effect in human system. Whenhuman peripheral blood mononuclear cell (PBMC) werestimulated with FSTL1 for 7 days, CD45�/lowALCAMþ cellsincreased in the culture (Supplementary Fig. S5D). When thesorted CD45�/lowALCAMþ cells were coinjected subcutane-ously with human breast cancer MDA231 cells into inimmunodeficient mice, tumor bone metastasis was caused(Supplementary Fig. S5D), although the subcutaneous tumorgrowth was not enhanced (data not shown). ALCAMþ

MSC increase may be more closely associated with tumorbone metastasis rather than T-cell impairment, and T-cellimpairment may be required for further tumor progression.We analyzed tumor tissues of patients with stage III breastcancer possibly having bone metastasis and deterioratedimmunity immunohistologically, because bone metastasis

A

100 101 102 103 104

FL3-Height

6.02

F10-snail+ supernatant

siFSTL1

100 101 102 103 104

FL3-Height

20.50(sMSC)

Control

100 101 102 103 104

FL3-Height

9.75(MSC)

100 101 102 103 104

FL3-Height

12.18

100 101 102 103 104

FL3-Height

18.30(fMSC)

bFGFFSTL1

CD45

SS

C

Control siFSTL1

Ad

ipo

cyte

sO

ste

ob

lasts

FSTL1

F10-snail+ supernatantB

Sp

he

res

None

100 101 102 103 104

FL3-Height

1.9310.83

100 101 102 103 104

FL3-Height100 101 102 103 104

FL3-Height

Foxp3

7.47 2.025.91 25.86

CD8

None

CControl Ig Anti-ALCAM

Coculture with fMSC

2.0

1.5

1.0

0.5

0.0

OD

45

0–

65

5 n

m

IFN

γ (n

g/m

L)

No

ne

MS

C

sM

SC

fMS

C

CD8+ cocultured with

No

ne

MS

C

sM

SC

fMS

C

CD8+ cocultured with

Control siFSTL1 FSTL1

F10-snail+ supernatant

DTumor cells + BMCs treated with

Tumor ALCAM

0.8

0.6

0.4

0.2

0.0

FABP4+

Osteocalcin+

* * **

Figure 4. FSTL1 generates the CD8low T-cell-inducible sMSCs. BMCs were treated with stimulants for 10 days and were analyzed by flow cytometry (A).The sorted CD45� BMCs treated with none (MSC), F10-snailþ supernatant (sMSCs), or FSTL1 (fMSCs) were tested for differentiation into FABP4þ

adipocytes and osteocalcinþ osteoblasts (B), sphere formation (B), and immunosuppression (C). Scale bar, 200 mm. In C, the CD45� BMCs were addedto the CD8þ T-cell proliferation system (1:10) with/without anti-ALCAM neutralizing mAb (5 days), and IFNg in the supernatant fluids were measured(n¼3,mean�SD; �,P<0.003 vs.MSC).CD8expression of the coculturedCD8þTcellswas alsoanalyzedby flowcytometry. TheCD45�BMCs (3�105)weremixedwithGFPþF10-mock cells (3�105) andwere subcutaneously injected intomice. D, fourweeks later, theBMCsof themicewere immunocytochemicallyanalyzed for ALCAMþ cells and GFPþ tumor cells. Scale bar, 50 mm. Data in each panel are representative of three independent experiments.

Kudo-Saito et al.





frequently occurs in patients with breast cancer (1). Colo-nizing ALCAMþ cells significantly increased in FSTL1þ

tumor tissues, but not normal tissues, and the significantcorrelation between FSTL1 and ALCAM increase was seen(P < 0.01; Fig. 5). This suggests a causal connection betweenthese molecules in tumor microenvironment. To clarify theusefulness of these molecules for diagnosis and prognosis ofpatients with cancer, the further studies are needed usingother cancers and bone marrow samples of patients atvarying stages.

FSTL1 blockade inhibits bone metastasis and immunedysfunction in vivoTo evaluate the antitumor efficacy of FSTL1 blocking,

F10-snailþ cells were implanted both subcutaneously andintravenously in mice, and polyethyleneimine-complexedsiRNA-fstl1 was injected into the subcutaneous tumors on day7. Tumor growth (P¼ 0.012; Fig. 6A) and bone metastasis (P¼

FSTL1(P = 0.009)

ALCAM(P = 0.008)

FS

TL

1 m

RN

A e

xp

ressio

n 0.8

0.6

0.4

0.2

0

0 0.2 0.4 0.6 0.8 1.0

ALCAM mRNA expression

Normal Cancer Normal Cancer

y = 0.5923x + 0.0415

R2 = 0.6393, P < 0.01

FSTL1(P = 0.020)

ALCAM(P = 0.025)

FS

TL

1 p

ixe

l co

un

ts (

x1

03) 3

2

1

00 1 2 3 4

ALCAM pixel counts (x103)

Normal Cancer Normal Cancer

Pix

el co

un

ts (

x1

03)

y = 0.6701x + 236.86

R2 = 0.7417, P < 0.01

BA0.8

0.6

0.4

0.2

0

6

4

2

0

Re

lative

exp

ressio

n

C

Hig

h

Low High

Lo

w

FSTL1

AL

CA

M

Normal

Tumor

Partial enlargement

Stroma

Figure 5. ALCAMþ cells correlatively increase in FSTL1þ tumor tissues ofpatients with advanced breast cancer. Tumor tissues obtained frompatients with stage III breast cancers (n ¼ 26) and normal mammarytissues (n ¼ 6) were analyzed by semiquantitative RT-PCR (A) andimmunohistochemically using anti-FSTL1 mAb and anti-ALCAM mAb(B and C). Bar graph indicates mean � SD. C, representative photos(scale bar, 100 mm). The partial enlarged photo shows a tumor sitepositive for FSTL1 and colonizing cells strongly positive for ALCAM.ALCAM staining was only faintly seen in stroma.

Contr

ol

siS

na

il #

1

siS

na

il #

2

siF

ST

L1 #

1

siF

ST

L1 #

2

F10-m

ock

15

10

5

0

GF

P+

Tu

mo

r ce

lls (

%)

0 7 10 14 17

2,000

1,500

1,000

500

0

Tum

or

volu

me (

mm

3)

F

Sp

ecific

lysis

(%

)

25

20

15

10

5

0

IFN

γ (n

g/m

L)

Contr

ol

siS

na

il

siF

ST

L1

F10-m

ock

Na

ive

F10-snail+

1.5

1.0

0.5

0.0

A

F10-mock

F10-snail+

Control siSnail siFSTL1

Contr

ol

siS

na

il

siF

ST

L1

F10-m

ock

Na

ive

F10-snail+

Days after implantation

D

40

20

0

AL

CA

M+

ce

lls (

%) Tumor BMCs SPCs

Contr

ol

siF

ST

L1

siS

na

il

Contr

ol

siF

ST

L1

siS

na

il

Contr

ol

siF

ST

L1

siS

na

il

B

2.86 8.83 16.51 5.80 4.61 10.91 3.60 13.31

CD8

1.49 3.73 6.23 3.77 0.67 2.23 0.92 2.73

Tetr

am

er

EF

oxp3

100

80

60

40

20

0

0 7 14 21 28 Days after implantation

Su

rviv

al (%

)

F10-snail+

C siS siF M

S

F

A

**

* **

* *

**

C

Figure 6. FSTL1 blocking inhibits tumor bone metastasis and antitumorimmune dysfunction in vivo. F10-snailþ or F10-mock tumor cells wereimplanted both subcutaneously (5� 105) and intravenously (1� 105) intomice, and 7 days later, siRNAwas injected into the subcutaneous tumors(n ¼ 5–10 per experiment). Knockdown efficacy on day 2 was confirmedby Western blotting (S, Snail; F, FSTL1; A, Actin; A). In the mice havingF10-snailþ tumors injected with siRNA-fstl1 (closed triangles, siF) orsiRNA-snail (closed squares, siS), the subcutaneous tumor growth(n¼ 10, mean� SD; A) and bone metastasis (n¼ 9, mean� SD; B) weresuppressed, and the mouse survival was prolonged by (n¼ 10; C). Opencircles (M), F10-mock tumors. Closed circles (C), F10-snailþ tumorsinjected with control siRNA. Ten days after siRNA injection, thesubcutaneous tumors, bone marrow, and spleens of the mice wereanalyzed for ALCAMþ cells as MSCs (n ¼ 6, mean � SD; D). SPCs werefurther analyzed for CD8þgp70-tetramerþ T cells and CD8þFoxp3þ Tcells by flow cytometry (E), and the sorted CD8þ T cells were tested forF10-mock tumor killing (ET ratio ¼ 50:1; n ¼ 3, mean � SD; F) and IFNgproduction (n¼3,mean�SD; F). �,P<0.02 versus control siRNA.Data ineach panel are representative of three independent experiments.






0.0002;Fig. 6B) were significantly suppressed, and the mouse survival(P < 0.0001; Fig. 6C) was significantly prolonged as comparedwith those of the control siRNA group. In the siRNA-fstl1–injectedmice, increase of ALCAMþ cells (Fig. 6D) andCD8low Tcells (Fig. 6E) was not seen, and tumor-specific CD8þ T-cellresponses were significantly elevated (P < 0.002; Fig. 6F). ThesiRNA-fstl1 efficacy was higher than that of siRNA-snail, par-ticularly on induction of ALCAMþ sMSCs and CD8low T cells.These results suggest that FSTL1 knockdown prevents not onlytumor metastasis, but also immune dysfunction throughALCAMþ sMSC decrease, and therefore antitumor CTLs areinduced rightly for eliminating tumor cells from the mice.FSTL1 blockade may be a promising strategy for treatingcancer bone metastasis of patients.

DiscussionThe relationship between cancer bone metastasis and anti-

tumor immunity in the host has been rarely investigated. Thisstudy revealed a novelmechanism, which is governed by FSTL1produced from Snailþ tumor cells undergoing EMT. FSTL1plays a dual role in cancer bone metastasis, in one way bymediating tumor invasion and bone tropism, and in a secondway by expanding pluripotent and multifunctional CD45�

ALCAMþ MSC-like cells, named as sMSCs. The sMSCs arepossibly derived fromBMCs, and disseminate all over the body.Tumor-infiltrating sMSCs induce bone metastasis de novofollowed by further increase of sMSCs in bone marrow. ThesMSCs are able to generate CD8low T cells with almost no CTLactivities, leading to dysfunction of antitumor immuneresponses. The number of ALCAMþ cells is correlated withthe amount of tumor bone metastasis in a murine model, andALCAMþ cells correlatively increase in FSTL1þ tumor tissuesof patients with advanced breast cancer, who may have bonemetastasis and deteriorated immune status frequently. FSTL1blocking in vivo significantly prevents both tumor bonemetas-tasis and increase of ALCAMþ cells and CD8low T cells, andsignificantly ameliorates induction of antitumor immuneresponses in the host. Thus, FSTL1 is a prominent trigger todrive cancer bone metastasis accompanied by immune dys-function (Supplementary Fig. S6).

Although FSTL1 increase in some tumor cell lines has beenshown previously (21), the functional role has not been inves-tigated in cancer. We clarified that FSTL1 critically regulatescancer bone metastasis. FSTL1 would be a potential target fortreating bone metastasis in patients. FSTL1 combination withother established drugs would treat bone metastasis moreeffectively. It has been shown that bone marrow-derived MSCsexpress FSTL1 (22). Also, someALCAMþ cells expressed FSTL1in the clinical tumor tissues tested in our study. FSTL1may be auseful marker for "mesenchymal" cells regardless of normal orcancerous cells.

CCL2 is one of the molecules involved in bone metastaticmechanism (5, 6). Surely, CCL2 significantly increased in Snailþ

tumor cells, and siRNA-ccl2 injection significantly suppressedSnailþ tumor growth and dissemination/metastasis as shownbefore (12). However, the siRNA-ccl2 injection hardly sup-pressed CD45�ALCAMþ MSC expansion in the treated mice,

and the siRNA-ccl2 efficacy was lower than siRNA-fstl1 efficacyon bone metastasis and CD8low T-cell induction (data notshown). This suggest that FSTL1 blocking is more effective intreating cancer, at least in association with bone metastasis,because CD45�ALCAMþ MSCs are the upstream cells capableof inducing various immunoregulatory cells such as tolero-genic dendritic and Treg cells.

In contrast with human MSCs, murine MSCs have not beenfully characterized yet. AsmurineMSCs, nonstimulated adher-ent CD45� BMCs obtained from na€�ve mice are generally usedafter long-term culture to obtain the adequate number ofMSCs (13). However, we found that murine MSCs, which werephenotypically and functionally similar to human MSCs (13),increased inmice having SnailþFSTL1þ tumors in vivo, and areexpanded by tumor-derived FSTL1 for a short term in vitro.FSTL1 was selectively expressed in tumor cells of patients,and stimulation with FSTL1 induced sMSC-like CD45�/low

ALCAMþ cells in human PBMCs. A specific circumstanceunder cancer would lead to the finding of the sMSCs.

Some reports have shown induction of immunosuppressiveT cells byMSCs. However, CD8 reduction was not shown in thereport of Foxp3þ Treg induction (23), and Foxp3 expressionwas not shown in the report of CD4low or CD8low T-cellinduction (24). Our study additionally showed that the sMSCsgenerate CD8low T cells partly including CD8lowFoxp3þ cells.CD8 reduction in T cells may be mediated by zinc-fingerprotein MAZR (25) and/or cKrox (26), which are negativeregulators of CD8 expression in thymocyte lineage differenti-ation. The signaling pathway of ALCAM, which is possiblyrequired for CD8low T-cell induction, may be of interest forelucidation of this mechanism.

Taken together, FSTL1 is a critical determinant governingcancer bone metastasis accompanied by immune dysfunction,and may be an attractive target for treating cancer via repro-gramming of antitumor immune responses in patients.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: C. Kudo-SaitoDevelopment of methodology: C. Kudo-SaitoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Kudo-Saito, K. MurakamiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Kudo-SaitoWriting, review, and/or revision of the manuscript: C. Kudo-SaitoAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): C. Kudo-Saito, T. FuwaStudy supervision: C. Kudo-Saito, Y. Kawakami

Grant SupportThis work was supported by Grant-in-Aid for Scientific Research from

Japan Society for the Promotion of Science (19590405, 21590445, 18591484, and23240128), The Sagawa Foundation for Promotion of Cancer Research, and KeioGijuku Academic Development Funds.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact.

Received May 10, 2013; revised July 11, 2013; accepted July 31, 2013;published OnlineFirst August 21, 2013.

Kudo-Saito et al.





References1. Weilbaecher KN, Guise TA, McCauley LK. Cancer to bone: a fatal

attraction. Nat Rev Cancer 2011;11:411–25.2. Wang J, Loberg R, Taichman RS. The pivotal role of CXCL12 (SDF-1)/

CXCR4 axis in bone metastasis. Cancer Metastasis Rev 2006;25:573–87.

3. Burger JA, Kipps TJ. CXCR4: a key receptor in the crosstalk betweentumor cells and their microenvironment. Blood 2006;107:1761–7.

4. Lipton A. Future treatment of bone metastases. Clin Cancer Res2006;12:6305s–08s.

5. Lu Y, Chen Q, Corey E, Xie W, Fan J, Mizokami A, et al. Activation ofMCP-1/CCR2 axis promotes prostate cancer growth in bone. Clin ExpMetastasis 2009;26:161–9.

6. Cai Z,ChenQ,ChenJ, LuY, XiaoG,WuZ, et al.Monocyte chemotacticprotein 1 promotes lung cancer-induced bone resorptive lesions invivo. Neoplasia 2009;11:228–36.

7. Zhau HE, Odero-Marah V, Lue HW, Nomura T, Wang R, Chu G, et al.Epithelial to mesenchymal transition (EMT) in human prostate cancer:lessons learned from ARCaP model. Clin Exp Metastasis 2008;25:601–10.

8. Feuerer M, Beckhove P, Garbi N, Mahnke Y, Limmer A, Hommel M,et al. Bone marrow as a priming site for T-cell responses to blood-borne antigen. Nat Med 2003;9:1151–7.

9. Kudo-Saito C, ShirakoH, Takeuchi T, Kawakami Y. Cancermetastasisis accelerated through immunosuppression during Snail-induced EMTof cancer cells. Cancer Cell 2009;15:195–206.

10. Uccelli A, Moretta L, Pistoia V. Mesenchymal stem cells in health anddisease. Nat Rev Immunol 2008;8:726–36.

11. Bianchi G, BorgonovoG, Pistoia V, Raffaghello L. Immunosuppressivecells and tumourmicroenvironment: focus onmesenchymal stemcellsand myeloid derived suppressor cells. Histol Histopathol 2011;26:941–51.

12. Kudo-Saito C, Shirako H, Ohike M, Tsukamoto N, Kawakami Y. CCL2is critical for immunosuppression to promote cancer metastasis. ClinExp Metastasis 2013;30:393–405.

13. Beyer Nardi N, da SilvaMeirelles L.Mesenchymal stem cells: isolation,in vitro expansion and characterization. Handb Exp Pharmacol2006:249–82.

14. Weidle UH, Eggle D, Klostermann S, Swart GW. ALCAM/CD166:cancer-related issues. Cancer Genomics Proteomics 2010;7:231–43.

15. Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW, et al.Mesenchymal stem cells within tumour stroma promote breast cancermetastasis. Nature 2007;449:557–63.

16. Tanaka M, Murakami K, Ozaki S, Imura Y, Tong XP, Watanabe T,et al. DIP2 disco-interacting protein 2 homolog A (Drosophila) is acandidate receptor for follistatin-related protein/follistatin-like 1–analysis of their binding with TGF-beta superfamily proteins. Febs J2010;277:4278–89.

17. Ouchi N, Asaumi Y, Ohashi K, Higuchi A, Sono-Romanelli S, OshimaY,et al. DIP2A functions as a FSTL1 receptor. J Biol Chem 2010;285:7127–34.

18. Adams D, Larman B, Oxburgh L. Developmental expression ofmouse Follistatin-like 1 (Fstl1): dynamic regulation during organo-genesis of the kidney and lung. Gene Expr Patterns 2007;7:491–500.

19. Miyamae T,Marinov AD, SowdersD,Wilson DC, Devlin J, Boudreau R,et al. Follistatin-like protein-1 is a novel proinflammatory molecule.J Immunol 2006;177:4758–62.

20. Li D, Wang Y, Xu N, Wei Q, Wu M, Li X, et al. Follistatin-like protein 1 iselevated in systemic autoimmune diseases and correlated with dis-ease activity in patients with rheumatoid arthritis. Arthritis Res Ther2011;13:R17.

21. Hambrock HO, Kaufmann B, Muller S, Hanisch FG, Nose K,Paulsson M, et al. Structural characterization of TSC-36/Flik:analysis of two charge isoforms. J Biol Chem 2004;279:11727–35.

22. Genovese JA, Spadaccio C, Rivello HG, Toyoda Y, Patel AN. Electro-stimulated bone marrow human mesenchymal stem cells producefollistatin. Cytotherapy 2009;11:448–56.

23. PrevostoC, Zancolli M, Canevali P, Zocchi MR, Poggi A. Generation ofCD4þ or CD8þ regulatory T cells upon mesenchymal stem cell-lymphocyte interaction. Haematologica 2007;92:881–8.

24. Giuliani M, Fleury M, Vernochet A, Ketroussi F, Clay D, Azzarone B,et al. Long-lasting inhibitory effects of fetal liver mesenchymalstem cells on T-lymphocyte proliferation. PLoS ONE 2011;6:e19988.

25. Bilic I, Koesters C, Unger B, Sekimata M, Hertweck A, Maschek R,et al. Negative regulation of CD8 expression via Cd8 enhancer-mediated recruitment of the zinc finger protein MAZR. Nat Immunol2006;7:392–400.

26. Jenkinson SR, Intlekofer AM, Sun G, Feigenbaum L, Reiner SL,Bosselut R. Expression of the transcription factor cKroxin peripheral CD8 T cells reveals substantial postthymic plasticityin CD4-CD8 lineage differentiation. J Exp Med 2007;204:267–72.






2013;73:6185-6193. Published OnlineFirst August 21, 2013.Cancer Res Chie Kudo-Saito, Takafumi Fuwa, Kouichi Murakami, et al. Immune DysfunctionTargeting FSTL1 Prevents Tumor Bone Metastasis and Consequent

Updated version

10.1158/0008-5472.CAN-13-1364doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/08/23/0008-5472.CAN-13-1364.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/20/6185.full#ref-list-1

This article cites 25 articles, 8 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/20/6185.full#related-urls

This article has been cited by 4 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/73/20/6185To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-13-1364

http://cancerres.aacrjournals.org/content/suppl/2013/08/23/0008-5472.CAN-13-1364.DC1

http://cancerres.aacrjournals.org/content/73/20/6185.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/20/6185.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/20/6185


Documents

Targeting FSTL1 Prevents Tumor Bone Metastasis and ... · Targeting FSTL1 Prevents Tumor Bone Metastasis and Consequent Immune Dysfunction Chie Kudo-Saito, Takafumi Fuwa, Kouichi