Screenplays & Play

Granulocyte-colony stimulating factor promotes proliferation, migration and invasion in glioma cells

Granulocyte-colony stimulating factor promotes proliferation, migration and invasion in glioma cells
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Granulocyte-colony stimulating factor promoteslung metastasis through mobilization of Ly6G+Ly6C+ granulocytes Marcin Kowanetz, Xiumin Wu, John Lee, Martha Tan, Thijs Hagenbeek, Xueping Qu, Lanlan Yu, Jed Ross,Nina Korsisaari, Tim Cao, Hani Bou-Reslan, Dara Kallop, Robby Weimer, Mary J. C. Ludlam, Joshua S. Kaminker,Zora Modrusan, Nicholas van Bruggen, Franklin V. Peale, Richard Carano, Y. Gloria Meng, and Napoleone Ferrara 1 Genentech Inc., South San Francisco, CA 94080This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected in 2006.Contributed by Napoleone Ferrara, October 21, 2010 (sent for review September 26, 2010) Primingofthe organ-speci 󿬁 c premetastaticsitesisthoughttobeanimportant yet incompletely understood step during metastasis. Inthis study, we show that the metastatic tumors we examinedoverexpress granulocyte-colony stimulating factor (G-CSF), whichexpands and mobilizes Ly6G+Ly6C+ granulocytes and facilitatestheir subsequent homing at distant organs even before the arrivalof tumor cells. Moreover, G-CSF – mobilized Ly6G+Ly6C+ cells pro-duce the Bv8 protein, which has been implicated in angiogenesisandmobilizationofmyeloidcells.Anti – G-CSForanti-Bv8antibodiessigni 󿬁 cantly reduced lung metastasis. Transplantation of  Bv8   nullfetallivercellsintolethallyirradiatedhostsalsoreducedmetastasis.We identi 󿬁 ed an unexpected role for Bv8: the ability to stimulatetumorcellmigrationthroughactivationofoneoftheBv8receptors,prokineticin receptor (PKR)-1. Finally, we show that administrationofrecombinantG-CSFissuf 󿬁 cienttoincreasethenumbersofLy6G+Ly6C+cellsinorgan-speci 󿬁 cmetastaticsitesandresultsinenhancedmetastatic ability of several tumors. breast cancer  |  myeloid  |  CSF3  |  prokineticin 2 M etastasis is a major cause of death from solid tumors. Tometastasize, tumor cells need to degrade and invade theextracellular matrix, intravasate, be carried through blood orlymphatic vessels, extravasate at the secondary site, and  󿬁 nally,establish secondary tumors (1). In addition, recent evidencessuggest that, at least in some circumstances, tumors are able tomodify the distant microenvironment before arrival of metastatictumor cells to create the so-called premetastatic niche (2). Thisability of tumors to affect distant tissues is expected to enablecancer cells to target speci 󿬁 c organs in which they can initiatesecondary tumor growth and supports Paget ’ s seed and soil hy-pothesis. Bone marrow-derived cells (BMDCs) are thought to bemajorplayersintheseprocesses(3,4).Althoughseveralmoleculeshave been implicated (3 – 6), the mechanisms of tumor-dependentBMDC mobilization and the precise identity and signi 󿬁 cance of these cells in metastasis are incompletely understood.VEGFR-1, a tyrosine kinase receptor that binds VEGF-A,VEGF-B, and placenta growth factor (7, 8), has been implicatedas one of the key regulators of BMDC mobilization and premeta-static priming owing to its expression in a population of hemato-poietic progenitor cells and the ability of anti – VEGFR-1 antibodiesto reduce metastasis (3). However, more recent studies have ques-tioned this conclusion and reported that anti – VEGFR-1 treatmenthas no effect in clinically relevant models of metastasis, raising thepossibility that alternative pathways mediate tissue priming for me-tastasis (9). Therefore, key challenges are further de 󿬁 ning the mo-lecular and cellular changes occurring in the premetastatic tissuesandidentifyingthefactorsinitiatingsuchpremetastaticenvironment.In the present study, we analyzed several metastatic and non-metastatic breast cancer models for their ability to trigger BMDCmobilization. This analysis led to identi 󿬁 cation of Ly6G+Ly6C+myeloid cells as a major cell type that accumulates in premetas-tatic tissues and facilitates colonization by cancer cells and sub-sequent metastasis. We also identi 󿬁 ed tumor-derived granulocyte-colony stimulating factor (G-CSF) as a key initiator and regulatorof these processes. Results Metastatic Tumors Induce Gene Expression Changes in PremetastaticLungs.  To investigate changes triggered by primary tumors in lungsbefore the arrival of metastatic tumor cells, we initially used the4T1-related lines of mouse breast carcinoma as a model (10, 11).These cells provide a phenotypic spectrum ranging from non-metastatic cells (67NR and 168FARN) to cells able to complete allsteps of metastasis (4TO7, 66c14, and 4T1) (Fig. S1  A ). We per-formedcDNAmicroarraycomparingtotallungsfrommicewithoutany tumors (naïve) with those from mice bearing nonmetastatic(67NR) ormetastatic (4T1) breast carcinomas in the premetastaticphase (Fig. 1  A  and  SI Text ,  De  󿬁  ning the Premetastatic Lungs ). Ouranalysis identi 󿬁 ed 260 genes speci 󿬁 cally up-regulated and 274genes down-regulated more than twofold in lungs of mice bearing4T1 tumors relative to lungs of naïve mice or mice bearing non-metastatic 67NR tumors (Fig. 1  A ). We decided to focus on  Bv8 ,because it was one of the top up-regulated genes (Fig. S1  D ). Bv8 isa secreted protein that has been previously characterized asa proangiogenic factor (12), an inducer of growth and mobilizationof hematopoietic cells (13), and a neuromodulator (14, 15).Quantitative (q)RT-PCR analysis of   Bv8  expression in lung tissuescon 󿬁 rmed the microarray results (Fig. S1  E ). Moreover, we founda strong correlation between high Bv8 expression and metastaticpotential in multiple tumor models examined. Increased Bv8 levels were measured in the premetastatic lungs of mice bearing mouse66c14, 4TO7, and 4T1 (Fig. 1  B  and Fig. S2  A ) as well as humanMDA-MB-231 (Fig. S2  B ) breast carcinomas. We also detectedelevated Bv8 levels in the lungs of mice bearing Lewis Lung Car-cinoma (LLC) (Fig. S2  B ). Furthermore, high  Bv8  expression was found in premetastatic lungs of 8-wk-old polyoma virus middleT antigen under control of mouse mammary tumor virus promoter(MMTV-PyMT) transgenic mice (Fig. S2  B ). Premetastatic stage inMMTV-PyMT mice was con 󿬁 rmed by histology (Fig. S2 C ). Bv8-Expressing Ly6G+Ly6C+ Granulocytes Are a Major Component ofthe Premetastatic Lung Microenvironment.  Cd11b+Gr1+ myeloidsare known to be a major source of Bv8 (16, 17). Furthermore, they  Author contributions: M.K. and N.F. designed research; M.K., X.W., J.L., M.T., T.H., X.Q.,L.Y., J.R., N.K., T.C., H.B.-R., and D.K. performed research; L.Y. contributed new reagents/ analytic tools; M.K., T.C., H.B.-R., D.K., R.W., M.J.C.L., J.S.K., Z.M., N.v.B., F.V.P., R.C., andY.G.M. analyzed data; and M.K. and N.F. wrote the paper.The authors declare no con 󿬂 ict of interest.Freely available online through the PNAS open access option. 1 To whom correspondence should be addressed. E-mail: article contains supporting information online at 21248 – 21255  |  PNAS  |  December 14, 2010  |  vol. 107  |  no. 50  have been implicated in several important processes in tumor bi-ology (18 – 21). Therefore, we hypothesized that metastatic tumorsare able to modify the lung microenvironment, at least in part,throughmobilizationofCd11b+Gr1+cellsfromthebonemarrow(BM) and subsequent homing in the lungs. In agreement with thishypothesis, we observed increased numbers of Cd11b+Gr1+ cellsin 󿬁 ltrating lungs of mice bearing the metastatic tumors 4TO7,66c14, 4T1, LLC, MDA-MB-231, and MMTV-PyMT, whereaslungsfrom micebearing nonmetastatictumors67NRdid not showany increase in Cd11b+Gr1+ cells (Fig. 1 C  and Fig. S2  D ). Inaddition, we detected only very low numbers of these cells inorgans that are not typically a target for metastasis in the 4T1model, such as kidney (Fig. S2  E ).To further de 󿬁 ne the composition of the premetastatic mi-croenvironment, we performed FACS analysis to characterizeCd11b+Gr1+ cells and identify subpopulations accumulating inthe premetastatic lungs (Fig. 1  D ). Because the anti-Gr1 antibody recognize two antigens, Ly6G and Ly6C, these cells represent aheterogeneous population that includes granulocytes (expressingboth Ly6G and Ly6C), monocytes (expressing Ly6C but notLy6G), macrophages, dendritic cells, and myeloid suppressorcells (20, 22). We observed a marked enrichment ( 󿬁  ve- to six-fold) in Ly6C+Ly6G+ granulocytes and a modest accumulationof Ly6C+Ly6G- monocytes and F4/80+ macrophages in thepremetastatic lungs of 4T1 tumor-bearing mice (Fig. 1  D ). We didnot observe any signi 󿬁 cant changes in the frequency of eosino-phils (SiglecF+ cells) or dendritic cells in response to the pri-mary tumor (Fig. 1  D ). Moreover, we detected the previously described VEGFR1+CD117+ hematopoietic progenitor cells(HPCs) (3), although their frequency was much lower than that of Ly6G+Ly6C+ cells (2.7% for HPC vs. 27% for Ly6G+Ly6C+cells)(Fig.1  D ).Wethenstainedsectionsofpremetastaticlungswithanti-Ly6G antibody to visualize Ly6G+Ly6C+ cells. In agreement with the FACS data, we observed increased accumulation of thesecellsinlungsderivedfrommicebearingmetastatictumors(Fig.S2  F  and  G ). Among Cd11b+Gr1+ cells isolated from lungs, BM, orspleen of mice bearing 4T1 tumors, only Ly6G+Ly6C+ cellsstrongly expressed Bv8 (Fig. 1  E  and Fig. S2  H  ). However, wedetected only marginal  Bv8  transcripts levels in Ly6G-Ly6C+ andLy6G-Ly6C- cells, suggesting that the primary tumor secretes fac-tors that speci 󿬁 cally up-regulate Bv8 expression in Ly6G+Ly6C+granulocytes (Fig. 1  E ). Furthermore, we detected Bv8-positive cellsonly in lungs isolated from mice bearing tumors that are able tometastasizebothinthepremetastaticandmetastaticphases(Fig.S2  I   and  J  ). Thus, Bv8-expressing Ly6G+Ly6C+ cells seem to bea major component of the premetastatic microenvironment. Tumor-Derived G-CSF Initiates a Premetastatic Environment in Dis-tant Organs.  To determine which cytokine(s) secreted by tumorcells might be responsible for mobilizing Ly6G+Ly6C+ cells, wemeasured plasma and tumor levels of several factors previously implicated in mobilization of myeloid cells: VEGF-A, placentagrowth factor (PlGF), stromal-derived factor (SDF)1 α , macro-phage colony stimulating factor (M-CSF), GM-CSF, and G-CSF(23) (Fig. 2  A  and  B ). Among these, only G-CSF fully correlated with the ability of tumors to metastasize (Fig. 2  A and  B ). We alsofoundthatG-CSFisreleasedbymetastatictumorcellsinvitro(Fig.S3  A ). G-CSF is a major regulator of granulopoiesis produced by a variety of cell types (24) and plays a key role in neutrophil mo-bilization from the BM (25). Another attractive feature of G-CSFas a potential regulator of a process taking place at a distant sitefrom the tumor of srcin is its endocrine mode of action (24). Re-cently,G-CSFhasbeenalsoimplicatedintheacutemobilizationof endothelialprogenitorsinducedbyvasculardisruptingagents(26).Interestingly, in mice bearing human MDA-MB-231 tumors, wedetected increased tumor and plasma levels of both human and B CA 67NR -4 40 Naive Non-metastatic tumorMetastatic tumorExpressionNaive/Non-metastatic Metastatic 024681012141618Naïve 67NR 66c14 4T1    L  u  n  g   C   d   1   1   b  +   G  r   1  +   f  r  e  q  u  e  n  c  y   [   %   ] Non-metastaticMetastatic ** 110100100010000PBS 67NR 66c14 4T1    L  u  n  g   B  v   8   [  p  g   /  m  g   ] Non-metastaticMetastatic ** 4T1 Lung tissue in the pre-metastatic phase D 05101520253035Granulocytes(Ly6G+Ly6C+)Monocytes(Ly6C+Ly6G-)Macrophages (F4/80+)Eosinophils (SiglecF+)HPC(CD117+VEGFR1+)    %   o   f   l   i  v  e  c  e   l   l  s Naive4T1 ** Pre-metastatic lungs E 0 50 100 150 200 250 300 350   r  e   l  a   t   i  v  e  e  x  p  r  e  s  s   i  o  n   t  o   l  u  n  g   L  y   6   G  -   L  y   6   C  - Naive4T1 Lung Bone MarrowSpleenLy6G-Ly6C+Ly6G+Ly6C+Ly6G-Ly6C-Ly6G-Ly6C+Ly6G+Ly6C+Ly6G-Ly6C-Ly6G-Ly6C+Ly6G+Ly6C+Ly6G-Ly6C-Bv8 RNA in isolated cells *** Fig. 1.  Bv8   is strongly up-regulated in premeta-static lungs of mice bearing metastatic tumors. (  A )Design and results of the microarray study com-paring gene expression in lungs from BALB/c naïvemice and mice bearing nonmetastatic tumors 67NRand metastatic 4T1 tumors. Lungs from differentexperimental groups are color-coded (gray, fromnaïve mice; blue, from 67NR-bearing mice; orange,from 4T1-bearing mice). Note that the metastaticgene expression pro 󿬁 le is clearly separated fromboth naïve and nonmetastatic pro 󿬁 les. Each pro 󿬁 lecolumn represents one individual mouse. ( B ) Bv8protein concentrations in the premetastatic lungsof mice bearing various tumors ( n  = 3 per group).( C  ) FACS analysis of Cd11b+Gr1+ cells in the pre-metastatic lungs of BALB/c mice bearing varioustumors. Asterisk indicates signi 󿬁 cant differencerelative to the nonmetastatic group. ( D ) FACSanalysis of different cell populations in premeta-static lungs from mice bearing 4T1 tumors 2 wkafter tumor inoculation. ( E  ) Bv8 expression inLy6GLy6C cell subpopulations isolated from lungsof naïve or 4T1 tumor-bearing mice. In  D  and  E  ,asterisk indicates signi 󿬁 cant difference relative tonaïve group. Graphs present means  ±  SEM. Kowanetz et al. PNAS  |  December 14, 2010  |  vol. 107  |  no. 50  |  21249      M     E     D     I     C     A     L     S     C     I     E     N     C     E     S     I     N     A     U     G     U     R     A     L     A     R     T     I     C     L     E  mouse G-CSF, suggesting that host cells in 󿬁 ltrating the tumor arealso a signi 󿬁 cant source of G-CSF (Fig. 2  A ). Administration of anti – G-CSF antibody to mice bearing mouse 4T1, 66c14, orMMTV-PyMT tumors completely abolished the accumulation of Ly6G+Ly6C+cellsintheperipheralbloodandlungs(Fig.2 C andFigs. S4  A  – C and S5). It also reduced lung Bv8 levels (Fig. S6  A  –  D ). In contrast, treatment with anti – GM-CSF, anti – M-CSF, oranti-VEGF neutralizing antibodies did not have any signi 󿬁 canteffect on the number of Ly6G+Ly6C+ cells in any of the exam-inedtissues(Fig.2 C andFigs.S4andS5  A and  B ).Theneutralizingactivityoftheseantibodieswascon 󿬁 rmedininvitrobioassays(Fig.S7). Also, none of these treatments reduced Bv8 concentrations inthepremetastaticlungs(Fig.S6  A and  B ).Treatmentwithanti-Bv8antibodyalsoreducedmobilizationofLy6G+Ly6C+cellsintothepremetastatic tissues in mice bearing 4T1, 66c14, and MMTV-PyMT tumors (Fig. 2 C  and Figs. S4 and S5). To de 󿬁 ne the tissuespeci 󿬁 city of Ly6G+Ly6C+ cells mobilization, we examined thepresence of Ly6G+ cells in various tissues during the premeta-static phase (Fig. 2 C  and Fig. S4  A ). In mice bearing 4T1 tumors, we could detect signi 󿬁 cant increases in the numbers of these cellsnotonlyinthelungbutalsointheliverandspleen,organsinwhichmetastasis is known to occur, although at later time points than inlungs (11, 27) (Fig. S4  A ). In tissues in which 4T1 tumors do not orrarely metastasize,we did not detect any Ly6G+ cells (muscle andbrain) or their accumulation was minimal (kidney). Neutralization of G-CSF, Bv8, or Ly6G+Ly6C+ Cells Reduces Metastasis. We orthotopically inoculated 66c14, 4T1, MDA-MB-231, orMMTV-PyMT tumors cells in the mammary fat pad of mice. Anti-Bv8 or anti – G-CSF treatment resulted in signi 󿬁 cant re-duction of lung metastasis in all models tested (Fig. 3  A  –  D  andFig. S8 G ), whereas it had little effect on primary tumor growth    N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1 SDF1a PlGF VEGF-A M-CSF GM-CSF G-CSF   p   l  a  s  m  a  c  o  n  c .   [  p  g   /  m   l   ] 050010001500200025003000350040004500    N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1   N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1   N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1   N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1 Non-metastaticMetastatic ACB PRE-METASTATIC PHASE    N  a   ï  v  e   6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   t   1 *** Naïve(SCID/bg) MDA-MB-231 050100150200250300350   p  g   /  m   l ** 02004006008001000    N  a   ï  v  e   I   S   O  a  n   t   i  -   G  -   C   S   F  a  n   t   i  -   B  v   8  a  n   t   i  -   G   M  -   C   S   F  a  n   t   i  -   M  -   C   S   F  a  n   t   i  -   V   E   G   F  a  n   t   i  -   P   l   G   F LUNGSLIVERKIDNEYBRAINMUSCLE    L  y   6   G  +  c  e   l   l  s   /  m  m     2 4T1    N  a   ï  v  e   I   S   O  a  n   t   i  -   G  -   C   S   F  a  n   t   i  -   B  v   8  a  n   t   i  -   G   M  -   C   S   F  a  n   t   i  -   M  -   C   S   F  a  n   t   i  -   V   E   G   F  a  n   t   i  -   P   l   G   F 4T1    N  a   ï  v  e   I   S   O  a  n   t   i  -   G  -   C   S   F  a  n   t   i  -   B  v   8  a  n   t   i  -   G   M  -   C   S   F  a  n   t   i  -   M  -   C   S   F  a  n   t   i  -   V   E   G   F  a  n   t   i  -   P   l   G   F 4T1    N  a   ï  v  e   4   T   1   N  a   ï  v  e   4   T   1 ************* ** Ly6G+ cells in the pre-metastatic tissues   m   G  -   C   S   F   h   G  -   C   S   F 0 100 200 300 400 500 600 700 800 900 1000    6   7   N   R   1   6   8   F   A   R   N   4   T   O   7   6   6  c   1   4   4   T   1   M   M   T   V  -   P  y   M   T  p  g   /  m  g **** MDA-MB-231 ** G-CSF in tumor lysates    M   M   T   V  -   P  y   M   T  m   h   M   D   A  -   M   B  -   2   3   1   M   M   T   V  -   P  y   M   T  m   h   M   D   A  -   M   B  -   2   3   1   M   M   T   V  -   P  y   M   T  m   h   M   D   A  -   M   B  -   2   3   1   M   M   T   V  -   P  y   M   T   M   D   A  -   M   B  -   2   3   1  m   h   M   M   T   V  -   P  y   M   T   M   D   A  -   M   B  -   2   3   1  m   h **    M   M   T   V  -   P  y   M   T * mG-CSFhG-CSF Fig. 2.  Increased G-CSF and Bv8 levels are associated with a metastatic phenotype. (  A ) Plasma levels of various cytokines in the premetastatic phase in micebearing nonmetastatic or metastatic tumors. In samples from mice with MDA-MB-231 tumors, m indicates mouse, and h is human. ( B ) Levels of G-CSF inwhole-tumor extracts isolated from mice in the premetastatic phase. ( C  ) Numbers of Ly6G+ cells (per millimeter squared) in various organs during thepremetastatic phase in mice bearing 4T1 tumors. Asterisk indicates signi 󿬁 cant difference relative to naïve group; two asterisks indicate signi 󿬁 cant differencerelative to isotype control (ISO) group ( C  ). Data shown are means  ±  SEM. 21250  | Kowanetz et al.  ( SI Text ,  Anti-Bv8 Anti  – G-CSF Antibodies Inhibit Metastasis ).Depending on the model, anti-Bv8 and anti – G-CSF treatmentsresulted in 30 – 60% and 40 – 70% inhibition, respectively.When4T1or66c14tumorcellswereorthotopicallyinoculatedinmicelackingBv8inbonemarrowcells(BMCs)(Fig.3  E andFig.S8  D and  E ), metastasiswas reduced to a degree comparablewith thatobserved in response to anti-Bv8 treatment, providing a genetic validation of the  󿬁 ndings obtained with antibodies.To assess the role of Ly6G+Ly6C+ cells in the development of metastasis, mice bearing orthotopic 66c14 tumors were treated with anti-Ly6G antibody (clone 1A8 speci 󿬁 cally depleting neu-trophils) (28) or anti-Gr1 antibody (clone RB6-8C5 depletingboth monocytes and neutrophils) (Fig. 3  F  ). Treatment with eachantibody signi 󿬁 cantly reduced the number of lung metastases toa degree comparable with anti-Bv8 or anti – G-CSF antibody. Theeffects of anti-Ly6G treatment were almost equivalent to those of the anti-Gr1 antibody, indicating that Ly6G+Ly6C+ neutrophilsare largely responsible for the metastasis in the tested models. Pretreatment with Recombinant G-CSF Induces Metastasis.  Weasked whether G-CSF administration is suf  󿬁 cient to initiate a premeta-static microenvironment. We analyzed the frequency of Ly6G+Ly6C+ and Cd11b+Gr1+ cells in mice treated daily with recom-binant(r)G-CSF for5 consecutive d(Fig. 4  A and  B ).Asexpected,delivery of rG-CSF induced mobilization of Cd11b+Gr1+ andLy6G+Ly6C+ cells, because we detected marked increases inthesecellsinBM,peripheralblood,and spleen.Wealsomeasureda signi 󿬁 cant increase in the frequency of Cd11b+Gr1+ and Ly6G+Ly6C+ cells in the lungs (Fig. 4  A  and  B ), coincident with in-creased expression of Bv8 (Fig. 4 C ). In addition, we observed anincreased frequency of Cd11b+Gr1+ cells (and enhanced Bv8expression) in the liver but the kidneys (Fig. 4 C ). To determine whether these effects may result in enhanced metastatic potential,miceweretreatedwithrG-CSFfor5consecutivedbeforeandafterinjection of tumor cells through the tail vein. Despite the fact thatthisroutebypassesseveralofthestepsoccurringduringmetastasis,it enabled us to focus on the potential role of G-CSF and the G-CSF – initiated premetastatic microenvironment at the  󿬁 nal stagesof metastasis, such as extravasation, survival, and tumor growth atthe distant organs. We injected mice with the metastatic cell lines66c14, 4T1, and B16F10 or nonmetastatic cell line 67NR. rG-CSFadministration increased the numbers of lung metastases in miceinjected with B16F10, 4T1, or 6614c cells (Fig. 4  D ). This was fol-lowed by increases in lung mass, which were in agreement with thenumber of visible metastases (Fig. S9  A ). Remarkably, the non-metastaticcellline67NRexhibitedmetastaticbehaviorinthelungsafter pretreatment with G-CSF (Fig. 4  E ). Authenticity of thetumors in mice pretreated with G-CSF was con 󿬁 rmed by histo-logical analysis of H&E-stained sections (Fig. S9  B –  D ). G-CSF – Induced Metastasis Depends on Bv8-Expressing Ly6G+Ly6C+Cells.  To de 󿬁 ne the role of Bv8-expressing Ly6G+Ly6C+ cellsin metastasis, mice were treated with rG-CSF together with anti-Bv8, anti-Ly6G, or anti-Gr1 neutralizing antibody. Anti-Bv8treatment reduced the increase in Cd11b+Gr1+ cells in thelungs (Fig. S9  E ). However, anti-Gr1 or anti-Ly6G antibodiescompletely abolished the G-CSF effect on Cd11b+Gr1+ cells. As expected, anti – G-CSF antibody treatment prevented the G-CSF effects (Fig. S9  E ). We next injected MDA-MB-231, 66c14,or 4T1 cells through the tail vein in mice that had been pre-treated with G-CSF in the presence or absence of anti-Bv8 an-tibody. Anti-Bv8 antibody treatment signi 󿬁 cantly reduced G-CSF – induced lung metastases of 66c14 (Fig. 4  F   and Fig. S9  F  ),4T1 cells (Fig. S9 G ), or MDA-MB-231 (Fig. S9  H  ). Treatment with an anti-Gr1 or anti-Ly6G antibody also markedly inhibitedthe prometastatic effects of G-CSF (Fig. 4  F  ), further supportingthe hypothesis that Bv8-expressing Ly6G+Ly6C+ cells are pre-dominantly responsible for the G-CSF – induced premetastaticpriming and metastasis. G-CSF Does Not Enhance Metastasis in Mice Treated with CytotoxicChemotherapy.  We treated mice bearing orthopic 66c14 or 4T1tumors with taxotere in the presence or absence of G-CSF asdescribed in  SI Materials and Methods . Taxotere treatment ef-fectively reduced primary tumor growth (Fig. S10  A  and  B ) andnumbers of lung metastases (Fig. S10 C ). Acute administration of recombinant G-CSF did not have any effect on the primary tu-mor and metastasis nor did it antagonize taxotere effects, whereas it reduced taxotere-induced neutropenia (Fig. S10  D and  E ). Similarly, taxotere reduced metastasis in mice pretreated with G-CSF and injected with tumor cells through the tail vein(Fig. S10  F  ). Interestingly, taxotere did not abolish the G-CS – induced accumulation of myeloid cells, suggesting that most of its antimetastatic activity was caused by its direct antisurvivaleffect on the tumor cells (Fig. S10 G ). G-CSF Induces Expression of Several Prometastatic Molecules. Microarray analysis (Fig. 1  A ) indicated that, in addition to Bv8,the premetastatic lungs are enriched in a number of factors thathave been previously shown to promote metastasis, includingMMP-9, S100A8, and S100A9 (Fig. S1  D ). Although MMP-9 hasbeenreportedtopromoteinvasion(29)andsurvival(30)oftumorcells in the lung environment, S100A8 and S100A9 mediate re-cruitment of myeloid and tumor cells in the lungs (4). To in- vestigate whether the G-CSF/Bv8 axis promotes metastasis, atleast in part, through regulation of these molecules, we measuredtheir expression levels in lungs of mice pretreated with rG-CSF.We found that all three molecules are signi 󿬁 cantly up-regulated(Fig. S11  A ) after G-CSF treatment. Anti-Bv8 signi 󿬁 cantly re-duced the G-CSF – induced expression of MMP-9, S100A8, andS100A9(Fig.S11  A ).Wethenaskedwhetherincreasedexpressionof these factors is directly linked to the Ly6G+Ly6C+ cells ac-cumulating in the premetastatic lungs. We isolated several cellsubpopulations from naïve or premetastatic lungs and sub-sequently measured the expression of Bv8, MMP-9, S100A8, andS100A9(Fig.S11  B and  D ).Interestingly,onlyLy6G+Ly6C+andCd11b+Gr1+ cells sorted from lungs of mice pretreated withG-CSF or mice bearing 4T1 tumors expressed signi 󿬁 cant levels of MMP-9, S100A8, and S100A9 along with Bv8 (Fig. S11  B  –  D ). Wedid not detect signi 󿬁 cant levels of these molecules in Ly6G-Ly6C+or Ly6G-Ly6C- cells or in Cd11b+Gr1- and Cd11b-Gr1- cells. Also, expression of both S100 proteins was markedly induced inLy6G+Ly6C+ cells isolated from mice pretreated with G-CSF ormice bearing 4T1 tumors (Fig. S11  B  and  D ). These  󿬁 ndingssuggest that G-CSF can increase local concentrations of MMP-9,S100A8,andS100A9throughmobilizationofLy6G+Ly6C+cellsintothelungsandbyinducingexpressionofthesemoleculesinthemobilized cells. Bv8 Directly Stimulates Tumor Cell Migration and Metastasis.  Anti-Bv8 treatment largely prevented the effects of G-CSF on thedevelopment of lung metastasis (Fig. 4  F   and Fig. S9  F   –  H  ).However, the same treatment only partially inhibited G-CSF – induced mobilization and homing of Cd11b+Gr1+ cells into thepremetastatic lungs (Fig. S9  E ). This  󿬁 nding suggests that addi-tional mechanisms are involved in the prometastatic activity of Bv8. We tested the hypothesis that Bv8 might have direct effectson tumor cells. We measured the expression levels of Bv8receptors (PKR-1/GPR73 and PKR-2/GPR73L1) in both non-metastatic and metastatic tumor cells. We detected signi 󿬁 cantexpression of   PKR-1  only in metastatic tumor cell lines (4TO7,66c14, 4T1, B16F10, and LLC as well as MDA-MB-231), whereas the nonmetastatic cell lines (67NR and 168FARN)exhibited much lower or undetectable levels of   PKR-1  (Fig. 5  A ).We were unable to detect  PKR-2  in any of the cell lines tested,except in LLC (Fig. S12  A ). Also,  G-CSFR  was undetectable inthese cells (Fig. S12  B ), arguing against the possibility that G-CSFmay have direct effects on cancer cells. To determine whetherPKR-1 is functional, we stimulated tumor cells with Bv8 and thenmeasured levels of phosphorylated ERK1/2 (Fig. S12 C ). Wedetected signi 󿬁 cant ERK1/2 activation in response to Bv8 only in Kowanetz et al. PNAS  |  December 14, 2010  |  vol. 107  |  no. 50  |  21251      M     E     D     I     C     A     L     S     C     I     E     N     C     E     S     I     N     A     U     G     U     R     A     L     A     R     T     I     C     L     E  the metastatic cell lines (Fig. S12 C ). Bv8 has been shown topromote migration of myeloid cells (31). Therefore, we consid-ered the possibility that Bv8 might also promote migration of metastatic tumor cells. Indeed, Bv8 induced migration of suchcells in a dose-dependent manner (Fig. 5  B ). In contrast, non-metastatic cell lines did not exhibit enhanced migration in re-sponse to Bv8. To con 󿬁 rm that the G-CSF – induced premeta-static microenvironment facilitates migration of tumor cells in vivo, we injected  󿬂 uorescently labeled (with CellTracker) 4T1 or66c14 cells into BALB/c nude mice pretreated with G-CSF in thepresence or absence of anti-Bv8, anti-Gr1, or anti – G-CSF anti-bodies, and we assessed the numbers of tumor cells that wereable to extravasate and seed the lung tissue. CellTracker wasstable in cells for more than 48 h after the labeling (Fig. S12  D ).We detected signi 󿬁 cant increases in the number of tumor cells inthe lungs of mice pretreated with G-CSF compared with controls(Fig. 5  C  and  D  and Fig. S12  E ). This effect was signi 󿬁 cantly reduced by administration of anti-Bv8, anti-Gr1, or anti – G-CSFantibodies (Fig. 5  C  and  D  and Fig. S12  E ). Asnoted(Fig.4  E ),nonmetastatic67NRcellsexhibitmetastaticpropertieswheni.v.injectedinmicepretreatedwithG-CSF,likely because of the substantially increased expression of MMP9 andother prometastatic molecules in the G-CSF – primed lungs. Wespeculated that metastasis of 67NR cells might be enhanced by overexpression of PKR-1. Indeed, overexpression of PKR-1 in67NR cells, to a level comparable with the expression in 66c14cells (Fig. S12  F  ), resulted in signi 󿬁 cant enhancement of 67NRmetastasis in the G-CSF pretreated mice (Fig. 5  E  and Fig. S12  G and  H  ). Conversely, PKR-1 – de 󿬁 cient 66c14 cells (Fig. S12  I  ) hadsigni 󿬁 cantly reduced numbers of G-CSF – induced metastasiscompared with cells expressing control shRNA (Fig. S12  J  ). Tumor Cells Colonizing the Lungs Express Elevated Levels of  G-CSF  and  PKR-1 .  We next asked whether cells capable of colonizing thelungs have a high expression of G-CSF or PKR-1. To accomplishthis, we performed in vivo clonal selection for MDA-MB-231cells that either reside at the primary tumor or are able to gen-erate metastasis in the lungs (Fig. S12  K  ). A similar methodology has been previously used to identify gene sets that mediate breastcancer metastasis to the lung and other organs (32, 33). Geneexpression pro 󿬁 ling revealed that MDA-MB-231 cells isolatedfrom lungs express much higher levels of both  G-CSF   and  PKR-1 along with  GM-CSF  ,  MMP-9 , and the previously identi 󿬁 ed  Cxcl1 and  MMP-1  (32) compared with cells isolated from the primary tumor or to parental cells (Fig. 5  F  ). The up-regulation of   GM-CSF   in MDA-MB-231 cells metastasizing to the lungs is unlikethe other models that we tested, suggesting that the role of GM-CSF is model-dependent. We did not detect increased expres-sion of   PKR-2 ,  M-CSF  ,  SDF1 α  ,  VEGF-A , or  PlGF  , suggestingthat, to metastasize, cancer cells selectively increase expressionof a narrow group of genes, which include  G-CSF   and  PKR-1 . Discussion Cd11b+Gr1+ and other myeloid cell types have been shown tofacilitate tumor growth in a number of studies (18 – 20, 34). Im-portantly, their human counterparts have been found to beoverproduced in cancer patients (35, 36). Cd11b+Gr1+ cellsrepresent a heterogeneous cell population comprised of neu-trophils, macrophages, and dendritic cells. They have been shownto promote invasion and metastasis through increased productionof matrix metalloproteinases (MMPs) and TGF- β 1 (37, 38) andhave been also implicated in suppression of T cell-mediatedresponses, hence the denomination of myeloid-derived suppres-sor cells (MDSC) (22, 23). However, we have no evidence thatimmunosuppression plays a role in the effects that we describedhere,sinceinhibitingmobilizationorfunctionofmyeloidcellshadsimilar effects in immuno-competent and immuno-de 󿬁 cient mice.Ourdataindicatethattumor-secretedG-CSFexpandsandmobi-lizes a subset of Cd11b+Gr1+ cells, Ly6G+Ly6C+ granulocytes,from BM and also induces Bv8 expression (Fig. 5 G ). Bv8, in turn,functions as a chemoattractant that enhances mobilization of BM-derived Ly6G+Ly6C+ granulocytes and facilitates their homingintothelungbeforearrivaloftumorcells.Aftertheyareinthelungs,G-CSF – mobilized Ly6G+Ly6C+ cells may serve as a major sourceof Bv8, MMP9, S100A8, and S100A9. MMP-9 has been shown toenhance invasion and metastasis in lungs (29, 30). S100A8 andS100A9 proteins have been shown to be key components of thepremetastaticnicheandtomediatemetastasisthroughmobilizationofmyeloidcellsandcancercellstolungs(4,39,40).Therefore,Ly6G+Ly6C+ cells mobilized by G-CSF create a protumsrcenic mi-croenvironment that supports extravasation, survival, and growth of secondarytumorsatdistantorgans.Interestingly,TNF α ,VEGF,andTGF β 1 have also been implicated in the regulation of S100A8 and D EF 0510152025303540 ISO anti-Bv8 anti-Ly6G anti-Gr1 anti-G-CSF **** NS 66c14n=10   m  e   t  a  s   t  a  s  e  s   /   l  u  n  g A 020406080100120 ISO anti-Bv8(2B9+3F1)anti-G-CSF anti-Bv8+ anti-G-CSF    P  e  r  c  e  n   t  a  g  e  o   f   l  u  n  g  m  e   t  a  s   t  a  s  e  s   (   I   S   O  =   1   0   0   %   ) 4T1 *** NS 0102030405060 ISO anti-Bv8 ISO anti-Bv8 4T1 Bv8 WTBv8 KO *** NS NS   m  e   t  a  s   t  a  s  e  s   /   l  u  n  g ISO anti-Bv82D3anti-mG-CSF+anti-hG-CSF    C   K  -   1   8  +  c  o   l  o  n   i  e  s   /   l  u  n  g  s  e  c   t   i  o  n  MDA-MB-231n=10 0510152025 ** C 4T166c14 ISO anti-Bv8 anti-G-CSF anti-Bv8 + anti-G-CSFISO anti-Bv8 anti-G-CSF anti-Bv8 + anti-G-CSF B 0510152025303540 ISO anti-Bv8 anti-G-CSF anti-Bv8+ anti-G-CSF 66c14 ** *   m  e   t  a  s   t  a  s  e  s   /   l  u  n  g Fig. 3.  Neutralization of G-CSF or Bv8 inhibits metastasis. (  A ) Number ofmetastases in lungs of mice bearing 4T1 tumors and treated with indicatedantibodies ( n  = 10) for 5.5 wk after tumor inoculation. ( B ) Number of me-tastases per lung in mice bearing 66c14 tumors 6 wk after tumor inoculation.Tumors were implanted, and treatment was performed as in  A  ( n  = 10). ( C  )Representative images of lung sections from mice bearing 4T1 or 66c14tumors and treated with the indicated antibodies. Sections correspond to  A (4T1 tumors) or  B  (66c14 tumors). Arrowheads indicate metastases. ( D ) Av-erage numbers of CK-18 – positive tumor colonies per lung section of SCID/bgmice bearing MDA-MB-231-X1.1 tumors and treated with indicated anti-bodies for 7 wk. ( E  ) Number of lung metastases in mice transplanted witheither Bv8 WT (white bars) or Bv8 KO (gray bars) fetal liver cells and treatedwith indicated antibodies. ( F  ) Numbers of lung metastases in mice bearing66c14 tumors and treated with indicated antibody. Asterisk indicates sig-ni 󿬁 cant difference relative to ISO group. Data shown are means  ±  SEM. 21252  | Kowanetz et al.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks