research depletion of tumor-associated macrophages ... · jor public health problem and the fifth...

3420

Published OnlineFirst June 22, 2010; DOI: 10.1158/1078-0432.CCR-09-2904

Cancer Therapy: Preclinical Clinical
Cancer
Research

Depletion of Tumor-Associated Macrophages Enhances theEffect of Sorafenib in Metastatic Liver Cancer Models byAntimetastatic and Antiangiogenic Effects

Wei Zhang1,2, Xiao-Dong Zhu1, Hui-Chuan Sun1, Yu-Quan Xiong1, Peng-Yuan Zhuang1,Hua-Xiang Xu1, Ling-Qun Kong1, Lu Wang1, Wei-Zhong Wu1, and Zhao-You Tang1

Abstract

Authors' AFudan UnInvasion, T2DepartmeInstitute an

Note: SuppResearch O

W. Zhang,

CorresponZhongshan200032, PeE-mail: zyta

doi: 10.115

©2010 Am

Clin Canc

Dow

Purpose: To investigate the role of macrophages in tumor progression under sorafenib treatment andto explore whether combination of drugs that deplete macrophages improved the antitumor effect ofsorafenib.Experimental Design: Tumor growth, lung metastasis, and tumor angiogenesis were observed in

HCCLM3-R and SMMC7721, two human hepatocellular carcinoma xenograft nude mouse models, whentreated with sorafenib (30 mg/kg daily, n = 6 per group) or a vehicle as control. Macrophage infiltrationwas measured in the peripheral blood and in sorafenib-treated tumor by immunohistochemistry andflow cytometry with F4/80 antibody and CD11b antibody. The effect of macrophage depletion on tumorangiogenesis and metastasis after sorafenib treatment, using two drug target macrophages, zoledronic acid(ZA) and clodrolip, was measured in the two models of hepatocellular carcinoma.Results: Although sorafenib significantly inhibited tumor growth and lung metastasis, it induced a

significant increase in peripheral recruitment and intratumoral infiltration of F4/80- and CD11b-positivecells, which was accompanied with elevation of colony-stimulating factor-1, stromal-derived factor 1α,and vascular endothelial growth factor in the tumor and elevation of plasma colony-stimulating factor-1 and mouse vascular endothelial growth factor in peripheral blood, suggesting the role of macrophagesin tumor progression under sorafenib treatment. Depletion of macrophages by clodrolip or ZA in com-bination with sorafenib significantly inhibited tumor progression, tumor angiogenesis, and lung metas-tasis compared with mice treated with sorafenib alone. ZA was more effective than clodrolip.Conclusions: Macrophages may have an important role in tumor progression under sorafenib treat-

ment. ZA is promising when combined with sorafenib to enhance its antitumor effect. Clin Cancer Res;

16(13); 3420–30. ©2010 AACR.

Sorafenib treatment increased overall survival and timeto progression of patients with advanced hepatocellularcarcinoma (HCC) by <3 months compared with patientswho received placebo (1, 2). Patients with lung metastasishad a poorer response to sorafenib treatment (3). Emerg-ing clinical evidence as well as preclinical findings revealedthat invasiveness and metastatic behavior of tumor cells

ffiliations: 1Liver Cancer Institute and Zhongshan Hospital,iversity, Key Laboratory for Carcinogenesis and Cancerhe Chinese Ministry of Education, Shanghai, P.R. China andnt of Hepatobiliary Surgery, Tianjin Medical University Cancerd Hospital, Tianjin, P.R. China

lementary data for this article are available at Clinical Cancernline (http://clincancerres.aacrjournals.org/).

X-D. Zhu and H-C. Sun contributed equally to this work.

ding Author: Zhao-You Tang, Liver Cancer Institute andHospital, Fudan University, 136 Yi Xue Yuan Road, Shanghai,ople's Republic of China. Phone and Fax: 86-21-64037181;[email protected].

8/1078-0432.CCR-09-2904

erican Association for Cancer Research.

er Res; 16(13) July 1, 2010

Researcon Augusclincancerres.aacrjournals.org nloaded from

are increased after vascular endothelial growth factor(VEGF)–targeted therapy (4–6), which may explainwhy, after a transitory growth delay of primary tumorand prolongation of progression-free survival, clinical re-sponses do not endure and tumors relapse as more inva-sive metastatic disease, thereby limiting the benefit foroverall survival (7).Tumor-associated macrophages facilitate tumor angio-

genesis, and extracellular matrix degradation and remodel-ing, and also promote tumor cell motility (8). Recentstudies reveal the direct communication between macro-phages and tumor cells that leads to the invasion andegress of tumor cells into blood vessels (9). Thus, macro-phages are at the center of the invasive microenvironmentand are important targets for cancer therapy (8). Further-more, macrophages are derived from myeloid cells, whichhave been reported as a primary cause of resistance to anti-VEGF therapy, and the elimination of these cells from thetumor microenvironment significantly restricts tumorgrowth in spontaneous and xenograft murine tumor mod-els (10). Moreover, a phase II study of sunitinib, another

h. t 23, 2021. © 2010 American Association for Cancer

http://clincancerres.aacrjournals.org/

Translational Relevance

Hepatocellular carcinoma (HCC) is currently a ma-jor public health problem and the fifth most commoncause of cancer-related death among men in the Unit-ed States. Sorafenib prolonged time to progression andoverall survival of HCC patients by only 3 months,and patients with metastasis had a poorer responseto sorafenib treatment. Thus, further improvement lieson the combination of sorafenib with other classes ofdrugs. Zoledronic acid (ZA) is now clinically availableas an antimetastatic drug that prevents and treats bonemetastases in patients with solid tumors by targetingosteoclasts, which are macrophages homing to thebone. In the current study, we used two HCC cell lineswith high and low metastatic potential, and found thatsorafenib induced macrophage recruitment and infil-tration to HCC xenograft. When combination with sor-afenib, ZA reduced the numbers of macrophagesinduced by sorafenib, decreased tumor angiogenesis,and, above all, further inhibited tumor progressionand reduced lung metastasis. This work might be ap-plied to clinical practice by combination of ZA withsorafenib for HCC patients.

Macrophage Depletion Improves the Effect of Sorafenib


inhibitor of multiple tyrosine kinase inhibitors whosespectrum of activity overlaps that of sorafenib, found thata high level of plasma stroma-derived factor-1α (SDF-1α)was associated with rapid progression or high mortality inHCC patients receiving sunitinib treatment (11). BecauseSDF-1α is a chemokine for progenitor cells (12), it is pos-sible that monocytes/macrophages may take part in tumorprogression. It is still unclear, however, whether sorafenibtreatment induces increased macrophage infiltration intumors and whether the antitumor effect of sorafenibcould be enhanced by the elimination of macrophages.Clodronate-containing liposome (clodrolip), in which

the bisphosphonate clodronate is encapsulated into lipo-some, has been shown to significantly reduce the num-ber of infiltrating mononuclear macrophages in tumors(13–15) and peripheral blood (14, 16), and to inhibitsurvival of macrophages in vitro (14). It was reported thatclodrolip depletes macrophages by 80.5% (16) to 93%(13), but it is not clinically available. Zoledronic acid(ZA) is a highly charged hydrophilic molecule that doesnot readily cross the plasma cell membrane (17); itreaches pharmacologically active concentrations only incells that exhibit marked fluid-phase endocytosis, suchas osteoclasts and macrophages (18). ZA has been usedto treat bone metastasis and reduce skeletal tumor bur-den in many types of cancers by inhibiting the activityof osteoclasts. ZA is well tolerated, either alone or com-bined with chemotherapy (19), but the combination ofZA with targeted therapies has not been investigated yet.It is reported that ZA has direct antitumor effect (20, 21),

www.aacrjournals.org

Researcon Augusclincancerres.aacrjournals.org Downloaded from

but only in tumors such as breast (20) or cervicalcarcinoma (21) in which macrophage infiltration isprominent, or in bone in which an extremely higher con-centration of ZA can be achieved (20). Further, the con-centration of ZA at which it presents a direct inhibitingeffect on a tumor cell is 12.5- to 50-fold higher than thatin a patient or animal study. In tumors without signifi-cant macrophage infiltration, ZA has no antitumor effect(17, 22). It is evident that ZA depletes tumor-associatedmacrophages by reverting their polarization from M2 toM1 phenotype (23) and by suppressing their expressionof matrix metalloproteinase-9 (24, 25).The present study used orthotropic nude mouse model

with two human HCC cell lines to investigate the role ofmacrophages in tumor progression under sorafenib treat-ment. Using two drugs that specifically target macrophages,clodrolip and ZA, we investigated the combination effectsof these drugs with sorafenib on tumor growth, tumor an-giogenesis and metastasis.

Materials and Methods

Animals, cell lines, and treatmentsMale athymic BALB/c nu/nu mice (5-wk-old) were ob-

tained from the Shanghai Institute of Materia Medica,Chinese Academy of Science. All mice were bred in lam-inar flow cabinets under specific pathogen-free condi-tions. The experimental protocol was approved by theShanghai Medical Experimental Animal Care Committee.Two HCC cell lines were used in this study. SMMC7721,which is of low metastatic capacity and was establishedby the Shanghai Institute of Cell Biology, Chinese Acad-emy of Sciences, Shanghai, and HCCLM3, a human HCCcell line with high metastatic potential that originatedfrom MHCC97 (26). Stable red fluorescent protein–expressing HCCLM3 (LM3-R) cells by infection with len-tivirus containing full-length cDNA of red fluorescentprotein were established by our institute (27). The mu-rine monocyte/macrophage cell line RAW 264.7 cells(mouse leukemic monocyte macrophage cell line) wasestablished by the Shanghai Institute of Cell Biology,Chinese Academy of Sciences, Shanghai. The LM3-R,SMMC7721, and RAW 264.7 cells were maintained at37°C with a 5% CO2 in air atmosphere in DMEM sup-plemented with 10% (volume/volume) heat-inactivatedfetal bovine serum and antibiotics (100 U/mL penicillin,100 mg/mL streptomycin).Clodrolip and PBS-containing liposome were gifts of

Roche Diagnostics GmbH. ZA was purchased fromNovartis.Sorafenib was prepared as previously described (28) andformulated at 4-fold (4×) of the highest dose in a cremo-phor EL/ethanol (50:50) solution. This 4× stock solutionwas prepared fresh daily. Final dosing solutions wereprepared on the day of use by dilution to 1× with endo-toxin-free distilled water (Life Technologies) and mixed byvortexing immediately before dosing. Sorafenib was final-ly prepared to a 1× solution with cremophor EL/ethanol/water (12.5:12.5:75, the vehicle solution).

Clin Cancer Res; 16(13) July 1, 2010 3421



Zhang et al.

3422


Treatment and grouping of nude mice bearingorthotopic HCC tumorsSeven days after orthotopic implantation of tumor cells

into the liver, the mice were randomly assigned to receivethe following treatments (n = 6 for each group): (a) a dailyoral dose of vehicle solution and a daily i.p. injection ofPBS-containing liposome (control group), (b) a daily oraldose of sorafenib at 30 mg/kg (sorafenib group), (c) i.p.injection of clodrolip at 100 μg/kg; thrice weekly (clodro-lip group), (d) a daily oral dose of sorafenib at 30 mg/kgplus an i.p. injection of clodrolip at 100 μg/kg thrice week-ly (sorafenib plus clodrolip group), (e) a daily s.c. injec-tion of ZA at 100 μg/kg (ZA group), or (f) a daily oraldose of sorafenib at 30 mg/kg plus a daily s.c. injectionof ZA at 100 μg/kg (sorafenib plus ZA group). All the micewere treated for 5 weeks.

Detection of metastasis by fluorescent microscopeand H&E stainingTumors were excised and measured by largest (a) and

smallest (b) diameters to calculate tumor volume byV = ab2/2. The lungs were also excised, and red fluorescentprotein–positive metastatic foci were imaged by fluores-cent imaging (stereomicroscope: Leica MZ6; illumination:Leica L5FL; C-mount: 0.63/1.25; charge-coupled device:DFC 300FX), and integrated absorbance was quantitatedby Image-Pro Plus software (Media Cybernetics, Inc.) aspreviously described (27). After examination of theimages, orthotopic tumors and lungs were fixed with10% buffered formalin and were paraffin embedded. Serialsections were cut at 5 μm for histologic study. Sections oflung were stained with H&E to pathologically determinethe presence of lung metastasis.

Immunohistochemical staining for CD31, F4/80,and CD11bTo assess the distribution of CD31 and F4/80, we stained

sections with rat anti-CD31 (1:200; BD Pharmingen), ratanti-F4/80 antibody (1:50; Serotec), and rat anti-CD11bantibody (1:50; Serotec). Rabbit anti-rat IgG/horseradishperoxidase (1:100; Serotec) or rabbit anti-rat FITC anti-body (1:100; Santa Cruz) were applied as secondary anti-bodies. Staining was evaluated by two independentobservers. For negative controls, primary antibodies werereplaced by PBS. For quantitation of mean vessel densityin sections stained for CD31, five fields at ×100 magni-fication in the “hotspot” were captured for each tumor,and microvessel density (MVD) was quantitated asCD31-positive area/total area. Quantitation of F4/80-or CD11b-positive cells in an HCC graft was calculatedas F4/80- or CD11b-positive area/total area. The datawere expressed as mean ± SEM.

Flow cytometry analysis of number of F4/80-positive cellsMouse whole blood was obtained with heparin used as

the anticoagulation agent, centrifuged to separate plasma,and stored at −80°C until tested. Blood cells of the lower

Clin Cancer Res; 16(13) July 1, 2010


layer were processed with lysing buffer (BD Pharmingen)for 20 minutes at room temperature and were then centri-fuged at 1,500 rpm for 5 minutes. Cells were then incubatedwith primary rat anti-F4/80 antibody (1:10; Serotec) or ratanti-CD11b (1:10; Serotec) for 1 hour at 37°C in PBS con-taining 0.5% (w/v) bovine serum albumin. After beingwashed with PBS, the cells were incubated with goat anti-rat IgG -FITC as a second antibody for 1 hour at 37°C.Negative controls were obtained without adding primaryantibody. FITC-labeled cells were enumerated with aFACSCalibur cytometer (BD Biosciences), and cell countswere expressed as % grated (percentage of F4/80-positivecells or CD11b-positive cells to mononuclear cells in theperipheral blood).

ELISA of plasma proteinsLevels of plasma VEGF of mouse origin (mouse VEGF)

and colony-stimulating factor-1 (CSF-1) of human originwere analyzed by ELISA with the use of QuantikineELISA kits (R&D Systems). All analyses were carried outin duplicate.

Quantitative real-time PCR analysisTotal RNA was extracted from cell lines and frozen tu-

mor specimens (n = 6 for each group) using Trizol Reagent(Invitrogen). Total RNA (2 μg) was reverse transcribedusing a PrimeScript RT reagents kit (TaKaRa). Reverse tran-scription-PCR was performed before quantitative real-timePCR. Messenger RNA expressions were determined by real-time PCR using SYBR Premix Ex Taq II (TaKaRa). PCR am-plification cycles were programmed for 10 seconds at 95°C,followed by 40 cycles of 95°C for 5 seconds and 60°C for30 seconds. Data were collected after each annealing step.Glyceraldehyde-3-phosphate dehydrogenase was used asan endogenous control to normalize for differences in theamount of total RNA in each sample. Relative expression ofgenes was calculated and expressed as 2−ΔCt, as previouslydescribed (25), and fold changes were calculated by nor-malization mRNA of the sorafenib-treated samples to thatof control samples. The primers used for the amplificationof human genes were as follows: VEGF, forward 5′-GGGAGCGGCTGACATTAT-3′ and reverse 5′-TGTTTCATC-CACCACCAA-3′; CSF-1, forward 5′-GGAGTGGACACCTG-CAGTCT-3′ and reverse 5′-TGTGCAGGGGCTGCTCACCA-3′;SDF-1α, forward 5′-ATGAACGCCAAGGTCGTGGTC-3′and reverse 5′-CTTGTTTAAAGCTTTCTCCAGGTACT-3′;and glyceraldehyde-3-phosphate dehydrogenase, forward5-GGAGTCAACGGATTTGGT-3 and reverse 5′-GTGATGG-GATTTCCATTGAT-3′.

Western blotting analysisTumor protein was extracted and the protein concentra-

tion was determined with the BCA protein assay (Pierce),and equal amounts of protein were subjected to 12% SDS-PAGE. After gel electrophoresis, the proteins were trans-ferred to the polyvinylidene difluoride membranes(Immobilon PVDF; Millipore). The membranes wereblocked for 1 hour at room temperature in 5% nonfat

Clinical Cancer Research





dry milk in TBS containing 0.05% Tween 20, followedby overnight incubation at 4°C with primary antibodies.The membranes were then incubated with horseradishperoxidase–labeled secondary antibody (Chemicon) for1 hour at room temperature. Peroxidase activity was de-tected through chemiluminescence (SuperSignal WestFemto luminol substrate and peroxide buffer; Pierce).Primary antibodies used include anti-CSF-1 (Santa CruzBiotechnology), anti-SDF-1α (Santa Cruz Biotechno-logy), anti-VEGF (Santa Cruz Biotechnology), and anti–glyceraldehyde-3-phosphate dehydrogenase (KangchengTechnology).

Production of mouse VEGF by macrophage induced byconditioned medium from tumor cellsThe LM3-R and SMMC7721 cell lines were seeded in six-

well plates, and were allowed to grow to 60% confluenceand then washed twice with serum-free DMEM, followedby treatment with serum-free DMEM containing 5 μmol/Lsorafenib or same volume of vehicle for 24 hours. Then,DMEM was removed and fresh serum-free DMEM wasadded into the well. Another 24 hours later, the condi-tioned medium was collected and added to RAW 264.7cells (mouse leukemic monocyte macrophage cell line, es-tablished by the Shanghai Institute of Cell Biology, Chi-nese Academy of Sciences, Shanghai), which were seededin six-well plates. After incubation for 24 hours, the medi-um was collected and centrifuged to remove cellular de-bris, and the supernatants were frozen at −80°C untilassayed for mouse VEGF by ELISA. The number of RAWcells was also counted to normalize the mouse VEGF pro-tein expression in ELISA. All experiments were done intriplicate.

Statistical analysisData were analyzed by the computer program SPSS

15.0 (SPSS, Inc.) by means of an unpaired two-tailedStudent's t test. Continuous data were expressed as mean ±SEM. Results were considered statistically significant at aP value of <0.05.

Results

Clodrolip and ZA inhibit tumor growth and lungmetastasis when combined with sorafenibAfter 5 weeks of treatment, sorafenib-treated mice had a

smaller tumor volume compared with that of mice in thecontrol group in the LM3-R model (1.03 ± 0.05 cm3 versus3.37 ± 0.23 cm3, P < 0.001) and in the SMMC7721 model(1.02 ± 0.1cm3 versus 1.86 ± 0.3 cm3, P < 0.05). Com-pared with sorafenib-treated mice, mice treated with acombination of sorafenib and clodrolip reduced tumorvolume by 31% (P < 0.01) in LM3-R model and 23%(P < 0.05) in SMMC7721 model, and combination ofZA with sorafenib reduced tumor volume by 67%(P < 0.01) in the LM3-R model and 39% (P < 0.001) inthe SMMC7721 model (Fig. 1A and C). In both models,although clodrolip alone did not inhibit tumor growth,



it significantly enhanced the antitumor effects of sorafe-nib (Fig. 1A). Mice treated with ZA alone had a signifi-cantly smaller tumor volume compared with that ofcontrols (Fig. 1A and C), which may not rule out me-chanisms other than the macrophage depletion effectof ZA, such as an antiangiogenic effect (29, 30) or inhi-bition of matrix metalloproteinase 9 expression bymacrophages other than by reducing the number ofmacrophages (31).Lung metastasis was also significantly inhibited by

sorafenib in the LM3-R model (integrated absorbance:58.0 ± 3.65 versus 167.7 ± 15.80 per lung, P < 0.001;Fig. 1B), and the combination of clodrolip or ZA withsorafenib further inhibited lung metastasis comparedwith sorafenib alone (integrated absorbance: 36.17 ±6.36 and 3.67 ± 2.28 versus 58.0 ± 3.65, P < 0.05 andP < 0.001, respectively). In the SMMC7721 model, lungmetastases were found in two of six mice in the controlgroup; no lung metastasis was found in the sorafenibgroup or the combination group.Weight loss was similar in mice treated by sorafenib or

ZA, but clodrolip, used either alone or combined withsorafenib, was associated with significant loss of bodyweight in LM3-R model (Fig. 1D). However, no signi-ficant body weight loss was found in mice bearingSMMC7721 tumors.

Clodrolip and ZA reduce F4/80- or CD11b-positivecells in peripheral blood and macrophageinfiltration in sorafenib-treated tumorsWe examined macrophage infiltration in tumors in re-

sponse to sorafenib treatment in both models by immu-nohistologic staining of F4/80 and CD11b antibodies.Histologic studies revealed prominent infiltration of F4/80- or CD11b-positive cells in tumors (F4/80-positivestaining area, 1.427% ± 0.47% versus 0.067% ± 0.019%,P < 0.05; Fig. 2C; CD11b-positive staining area, 5.0% ±0.81% versus 1.06% ± 0.24%, P < 0.01; Fig. 2D), whereasmacrophage infiltration was scarce in tumors of the con-trol group. In the SMMC7721 model, significant increasedmacrophage infiltration in tumor was also found(P < 0.05, respectively; Fig. 2E and F). Macrophage infiltra-tion was significantly suppressed by the combination ofsorafenib with clodrolip or ZA, but not in the clodrolip-or ZA-treated tumors (Fig. 2A-D).We also examined the number of monocytes/macro-

phages in peripheral blood by flow cytometry with F4/80 antibody and CD11b antibody (32). In the LM3-Rmodel, sorafenib treatment increased the number of F4/80- and CD11b-positive cells in peripheral blood com-pared with that of controls (16.09% ± 1.29% versus8.37% ± 0.88% for F4/80-positive cells and 15.68% ±2.14% versus 11.43% ± 1.62% for CD11b-positive cells,respectively, P < 0.05). Mice treated with sorafenib withclodrolip or ZA had significantly fewer F4/80- or CD11b-positive cells in the peripheral blood than mice treatedwith sorafenib alone. However, mice treated with clodro-lip alone or ZA alone did not significantly decrease the




Zhang et al.

3424


number of F4/80- or CD11b-positive cells in peripheralblood compared with that of control (Fig. 3A and B). Inthe SMMC7721 model, sorafenib treatment also increasedF4/80- and CD11b-positive cell recruitment in peripheralblood compared with that of controls (3.8 ± 0.48% versus2.51 ± 0.29% for F4/80-positive cells and 15.8 ± 0.5%versus 10.6 ± 0.57% for CD11b-positive cells, P < 0.05,respectively). Sorafenib-induced recruitment of F4/80-positive cells to peripheral blood were depleted by thecombination with clodrolip or ZA (2.56 ± 0.19% and1.9 ± 0.2% versus 3.8 ± 0.48%, P < 0.05, respectively),whereas sorafenib-induced CD11b-positive cells (15.8 ±0.5% versus 12.2 ± 1.66%, P < 0.05) were depleted bythe combination of sorafenib with ZA.



CSF-1, SDF-1α, and VEGF are elevated aftersorafenib treatmentCSF-1, SDF-1α, and VEGF are the important chemo-

kines for macrophages (33, 34). We assessed the expres-sion of three molecules in tumors by Western blottingand/or quantitative real-time PCR, and in the plasma byELISA in the LM3-R and SMMC7721 model. We foundthat CSF-1 was elevated by 5.1-fold in the LM3-R tumorand 2.2-fold in the SMMC7721 tumor, and SDF-1α waselevated by 2.1-fold and 1.5-fold in both tumor models,respectively. VEGF mRNA was elevated by 1.65-fold inSMMC7721 tumor. Although the VEGF mRNA was notsignificantly increased in the LM3-R tumor, the VEGF pro-tein was significantly elevated in this model (Fig. 4A-C).

Fig. 1. ZA and clodrolip enhanced tumor growth inhibition and reduced lung metastasis. A, sorafenib significantly inhibited tumor growth, and clodrolip andZA enhanced the tumor inhibitory effect of sorafenib. B, sorafenib significantly inhibited lung metastasis, and clodrolip and ZA further reduced lungmetastasis in sorafenib-treated mice. IOD, integrated absorbance. C, tumor volume was significantly reduced by the combination of sorafenib with clodrolipor ZA in the orthotopic model of the SMMC7721 cell line. D, clodrolip alone or combined with sorafenib was associated with significant loss of bodyweight in the LM3-R model (*, P < 0.05; **, P < 0.01; ***, P < 0.001).






The increase of CSF-1, SDF-1α, and VEGF was increasedmacrophage infiltration in the tumor and increased mac-rophage recruitment in the peripheral blood. As CSF-1was reported to be one of the most important chemo-kines to recruit macrophages and was elevated to a muchgreater extent than SDF-1α and VEGF, we further as-sessed CSF-1 level in the plasma by ELISA and foundthat plasma CSF-1 concentration was elevated by 2.5-fold in sorafenib-treated mice in the LM3-R model(P < 0.05) and 1.55-fold (P < 0.05) in the SMMC7721model (Fig. 4D and E), when compared with that of thenontreated mice.

ZA and clodrolip inhibit tumor angiogenesis insorafenib-treated miceSorafenib significantly inhibited tumor angiogenesis as

indicated by reduced MVD in the LM3-R model (1.28% ±0.15% versus 3.77% ± 0.44%, P < 0.001), and sorafenibcombined with clodrolip or ZA further inhibited tumorangiogenesis (0.69% ± 0.08% or 0.3% ± 0.07% versus



1.28% ± 0.15%, P < 0.01 or P < 0.001, respectively), sug-gesting that macrophages play a role in tumor angiogen-esis (Fig. 5A and B). Mouse plasma VEGF was elevatedafter sorafenib treatment (218.2 ± 18.2 pg/mL versus129.4 ± 5.4 pg/mL, P < 0.01) and was reduced by ZAor clodrolip (123.0 ± 2.9 pg/mL or 93 ± 21.8 pg/mL ver-sus 218.2 ± 18.2 pg/mL, P < 0.01 or P < 0.001, respec-tively; Fig. 5C). Mouse plasma VEGF treated withclodrolip alone did not differ significantly from that ofthe control group (112.7 ± 6.0 pg/mL versus 129.4 ±5.4 pg/mL, respectively, P > 0.05; Fig. 5C).Because macrophages are one of the major sources of

VEGF, we treated RAW 264.7 cells with conditioned mediafrom sorafenib-treated cells or nontreated cells, and exam-ined the mouse VEGF in the supernatants of macrophages.We found that after being treated by the conditioned me-dia from sorafenib-treated LM3-R and SMMC7721 cells,macrophage secreted a higher level of mouse VEGF, com-pared with that treated with the conditioned media ofnontreated tumor cells (Fig. 5D).

Fig. 2. F4/80- and CD11b-positive cells in the tumor tissue. A and B, immunohistochemistry analysis of tumors stained with F4/80-specific antibody (A)and CD11b-specific antibody (B). Sorafenib increased intratumoral macrophages, whereas clodrolip and ZA reduced macrophage infiltration. C and D,percent of F4/80-positive cells (C) and CD11b-positive cells (D) in LM3-R tumor. E and F, sorafenib also induced significant increase in the infiltration ofF4/80- and CD11b- positive cells in SMMC7721 tumor. Original magnification, ×100. *, P < 0.05; **, P < 0.01.




Zhang et al.

3426


Discussion

The present study showed that sorafenib treatment in-duced a significant increase in the recruitment of F4/80-and CD11b-positive cells in blood and the infiltration ofthese cells in two orthotopic HCC xenografts. These re-sults were accompanied with an elevated plasma mouseVEGF level, suggesting that macrophages may play a role



in the progression of a tumor under sorafenib treatment.When combined with sorafenib, clodrolip and ZA, twodrugs that deplete macrophages, significantly reduced in-tratumoral macrophage infiltration as well as decreasedmacrophage recruitment in peripheral blood comparedwith that of mice treated with sorafenib alone. In addi-tion, sorafenib-induced elevation of mouse VEGF was al-so decreased by clodrolip and ZA, which led to the

h. t 23, 2021. © 2010 American

Fig. 3. Recruitment of F4/80-and CD11b-positive cells inperipheral blood. A, percentage ofF4/80- and CD11b-positive cellsin peripheral blood were measuredby flow cytometry; the flowcytometry data are representativeof five to six samples in eachgroup. B, columns, mean of five tosix samples in each group; bars,SEM. *, P < 0.05; **, P < 0.01.


Association for Cancer




enhanced inhibition of tumor growth, tumor angiogene-sis, and lung metastasis.In the present study, clodrolip alone did not inhibit tu-

mor growth in sorafenib- nontreated mice but did presenta synergistic inhibitory effect on tumor growth when com-bined with sorafenib, indicating the importance of sorafe-nib-induced macrophages in tumor progression undersorafenib treatment. Macrophages can be attracted intohypoxic, perinecrotic areas along a trail of necrotic debrisemanating from such areas (8). Although T cell–mediatedcellular immune mechanisms are absent in nude mice, ithas been shown that the presence of macrophages can beactivated (35). We found that sorafenib caused severe hyp-oxia and necrosis in the treated tumor, as indicated by pi-monidazole and hypoxia-inducible factor 1α staining (seeSupplementary Figure). Hypoxia-inducible factor 1α hasbeen found to induce the recruitment of bone marrow–derived vascular modulatory cells, including F4/80- and



CD11b-positive cells, to regulate tumor angiogenesis andinvasion (36). In the present study, CSF-1, SDF-1α, andVEGF, all chemokines for macrophages (37, 38), wereelevated in sorafenib-treated tumor. Plasma CSF-1 alsoincreased after sorafenib treatment. CSF-1 can activatemacrophages, thereby contributing to macrophage-tumorinteraction, and may finally lead to tumor progression (9).It was also reported that CSF-1 has a direct effect on tumorgrowth, so we cannot rule out the possibility that CSF-1may directly participate in tumor progression under sora-fenib treatment. However, plasma CSF-1 levels were un-changed or even slightly elevated (not significantly) afterthe combination of either clodrolip or ZA with sorafenib,indicating that the effect of clodrolip or ZA was by target-ing macrophages rather than CSF-1.Upon activation by cytokines produced by cancer cells,

macrophages can release a vast diversity of growth factors,cytokines, and inflammatory mediators; many of these

Fig. 4. CSF-1, SDF-1α, and VEGFwere elevated in sorafenib-treatedtumor. A to C, protein level ofCSF-1, SDF-1α, and VEGF byWestern blotting of LM3-R tumor(A), and mRNA level of M-CSF,SDF-1, and VEGF of sorafenib-treated tumor by quantitative PCRwere normalized to control (B and C).GAPDH, glyceraldehyde-3-phosphate dehydrogenase.CSF-1 was elevated by 5.1-fold inthe LM3-R tumor and 2.2-fold inthe SMMC7721 tumor. SDF-1αwas also elevated by 2.1-fold and1.5-fold in both tumor models,respectively. VEGF was elevatedby 1.65-fold in the SMMC7721tumor but not elevated in the LM3-Rtumor. D and E, plasma CSF-1level were also significantlyelevated in sorafenib-treated micecompared with control in theLM3-R tumor (D) and SMMC7721tumor (E), and the combination ofclodrolip and ZA significantlyreduced plasma VEGF in bothmodels and also reduced plasmaCSF-1 in the SMMC7721 model.*, P < 0.05; **, P < 0.01;***P < 0.001.




Zhang et al.

3428


factors are key agents in cancer metastasis (39). Macro-phage-produced VEGF regulates malignant progressionthrough stimulation of tumor angiogenesis and tumor cellinvasion (34). In a mouse model of cervical cancer, inhi-bition of matrix metalloproteinase 9 in macrophagesblocked the release of VEGF and thereby inhibited angio-genesis and tumor growth (21). In the present study,sorafenib treatment increased plasma mouse VEGF con-centration, which was accompanied with an increasednumber of F4/80- and CD11b-positive cells in the periph-eral blood and increased macrophage infiltration in the tu-mor. Because the plasma level of mouse VEGF in the micetreated with clodrolip or ZA alone did not differ from that



of control mice, it is highly possible that the decrease ofsorafenib-induced plasma mouse VEGF associated withmacrophage depletion and MVD decrease, which is consis-tant with the results from in vitro experiments in LM3-Rand SMMC7721 cell lines.The present study showed that the combination of sor-

afenib with either clodrolip or ZA had a better antitumorefficacy than sorafenib alone in two orthotopic HCC mod-els, and the main effect of intratumoral-infiltrated macro-phage was tumor promoting. However, we do agree thatthis conclusion needs to be validated in an immunity-competent mouse model. Clodrolip was toxic to mice inthe present study and still is not clinically available. ZA has

Fig. 5. Clodrolip and ZA reduced tumor angiogenesis of sorafenib-treated tumor. A, immunohistochemistry analysis of tumors stained with anti-CD31(BD Pharmingen). B, quantitation of tumor vasculature indicated that clodrolip and ZA inhibited angiogenesis in combination with sorafenib. MVD of CD31 ineach section was measured in five images from each treatment (n = 6). Bars, MVD in each treatment. Original magnification, ×100. **, P < 0.01; ***, P < 0.001.C, combination with clodrolip or ZA was associated with decreased plasma VEGF of mouse origin (mouse VEGF). D, in vitro assessment of mouse VEGFin the supernatant of macrophage culture. Microphages were treated by conditionedmedium of sorafenib-treated or nontreated LM3-R and SMMC7721 cells.






been widely used in the prevention of bone metastasis andreduction of skeletal tumor burden in many kinds of can-cer, mainly by the inhibition of osteoclasts activity; ZA iswell tolerated, either alone or combined with chemother-apy (19). We believe that ZA may also have other mecha-nism of action in addition to macrophage depletion;however, the antimacrophage effect is one of the major ef-fects of ZA, as shown by our data. Another reason why weused ZA in this study is that ZA is clinical available to fur-ther improve the effect of sorafenib in clinical practice.

Conclusions

Although sorafenib significantly inhibited tumor growthand metastasis, it induced the elevation of F4/80- andCD11b-positive cells in peripheral blood and increasedthe infiltration of intratumoral macrophages, suggestingthat macrophages have a role in the progression of tumorunder sorafenib treatment. With the depletion of macro-



phages by clodrolip or ZA, tumor progression, tumor an-giogenesis, and metastasis were significantly inhibited. ZAis clinical available and is promising to be combined withsorafenib for HCC patients.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Grant Support

National Natural Science Foundation of China and the Research GrantsCouncil (no. 30731160, 30872505), and China National “211” Project forHigher Education, and National Key Scientific Projects: 2008ZX10002-019,021.

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 toindicate this fact.

Received 11/03/2009; revised 03/17/2010; accepted 05/05/2010;published OnlineFirst 06/22/2010.

References
1. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepa-
tocellular carcinoma. N Engl J Med 2008;359:378–90.2. Cheng AL, Kang YK, Chen Z, et al. Efficacy and safety of sorafenib in

patients in the Asia-Pacific region with advanced hepatocellular car-cinoma: a phase III randomised, double-blind, placebo-controlledtrial. Lancet Oncol 2009;10:25–34.

3. Yau T, Chan P, Ng KK, et al. Phase 2 open-label study of single-agent sorafenib in treating advanced hepatocellular carcinoma in ahepatitis B-endemic Asian population: presence of lung metastasispredicts poor response. Cancer 2009;115:428–36.

4. Loges S, Mazzone M, Hohensinner P, Carmeliet P. Silencing or fuel-ing metastasis with VEGF inhibitors: antiangiogenesis revisited.Cancer Cell 2009;15:167–70.

5. Ebos JM, Lee CR, Cruz-Munoz W, Bjarnason GA, Christensen JG,Kerbel RS. Accelerated metastasis after short-term treatment witha potent inhibitor of tumor angiogenesis. Cancer Cell 2009;15:232–9.

6. Paez-Ribes M, Allen E, Hudock J, et al. Antiangiogenic therapy elicitsmalignant progression of tumors to increased local invasion and dis-tant metastasis. Cancer Cell 2009;15:220–31.

7. Kerbel RS. Tumor angiogenesis. N Engl J Med 2008;358:2039–49.8. Condeelis J, Pollard JW. Macrophages: obligate partners for tumor

cell migration, invasion, and metastasis. Cell 2006;124:263–6.9. Goswami S, Sahai E, Wyckoff JB, et al. Macrophages promote the

invasion of breast carcinoma cells via a colony-stimulating factor-1/epidermal growth factor paracrine loop. Cancer Res 2005;65:5278–83.

10. Pulaski HL, Spahlinger G, Silva IA, et al. Identifying alemtuzumab asan anti-myeloid cell antiangiogenic therapy for the treatment of ovar-ian cancer. J Transl Med 2009;7:49.

11. Zhu AX, Sahani DV, Duda DG, et al. Efficacy, safety, and potentialbiomarkers of sunitinib monotherapy in advanced hepatocellular car-cinoma: a phase II study. J Clin Oncol 2009;27:3027–35.

12. Sweeney EA, Lortat-Jacob H, Priestley GV, Nakamoto B,Papayannopoulou T. Sulfated polysaccharides increase plasmalevels of SDF-1 in monkeys and mice: involvement in mobilizationof stem/progenitor cells. Blood 2002;99:44–51.

13. Zeisberger SM, Odermatt B, Marty C, Zehnder-Fjallman AH,Ballmer-Hofer K, Schwendener RA. Clodronate-liposome-mediateddepletion of tumour-associated macrophages: a new and highlyeffective antiangiogenic therapy approach. Br J Cancer 2006;95:272–81.

14. Hiraoka K, Zenmyo M, Watari K, et al. Inhibition of bone and muscle

metastases of lung cancer cells by a decrease in the number ofmonocytes/macrophages. Cancer Sci 2008;99:1595–602.

15. Gazzaniga S, Bravo AI, Guglielmotti A, et al. Targeting tumor-associated macrophages and inhibition of MCP-1 reduce angiogen-esis and tumor growth in a human melanoma xenograft. J InvestDermatol 2007;127:2031–41.

16. Nakao S, Kuwano T, Tsutsumi-Miyahara C, et al. Infiltration of COX-2-expressing macrophages is a prerequisite for IL-1 β-induced neo-vascularization and tumor growth. J Clin Invest 2005;115:2979–91.

17. Ottewell PD, Monkkonen H, Jones M, Lefley DV, Coleman RE, HolenI. Antitumor effects of doxorubicin followed by zoledronic acid in amouse model of breast cancer. J Natl Cancer Inst 2008;100:1167–78.

18. Thompson K, Rogers MJ, Coxon FP, Crockett JC. Cytosolic entry ofbisphosphonate drugs requires acidification of vesicles after fluid-phase endocytosis. Mol Pharmacol 2006;69:1624–32.

19. Aapro M, Abrahamsson PA, Body JJ, et al. Guidance on the use ofbisphosphonates in solid tumours: recommendations of an interna-tional expert panel. Ann Oncol 2008;19:420–32.

20. Peng H, Sohara Y, Moats RA, et al. The activity of zoledronic Acid onneuroblastoma bonemetastasis involves inhibition of osteoclasts andtumor cell survival and proliferation. Cancer Res 2007;67:9346–55.

21. Giraudo E, Inoue M, Hanahan D. An amino-bisphosphonate targetsMMP-9-expressing macrophages and angiogenesis to impair cervi-cal carcinogenesis. J Clin Invest 2004;114:623–33.

22. Soltau J, Zirrgiebel U, Esser N, et al. Antitumoral and antiangiogenicefficacy of bisphosphonates in vitro and in a murine RENCA model.Anticancer Res 2008;28:933–41.

23. CosciaM,Quaglino E, IezziM, et al. Zoledronic acid repolarizes tumor-associated macrophages and inhibits mammary carcinogenesisby targeting the mevalonate pathway. J Cell Mol Med. 2009 doi:10.1111/j.1582-4934.2009.00926.x.

24. Tsagozis P, Eriksson F, Pisa P. Zoledronic acid modulates anti-tumoral responses of prostate cancer-tumor associated macro-phages. Cancer Immunol Immunother 2008;57:1451–9.

25. Melani C, Sangaletti S, Barazzetta FM, Werb Z, Colombo MP.Amino-biphosphonate-mediated MMP-9 inhibition breaks thetumor-bone marrow axis responsible for myeloid-derived suppressorcell expansion and macrophage infiltration in tumor stroma. CancerRes 2007;67:11438–46.

26. Tian J, Tang ZY, Ye SL, et al. New human hepatocellular carcinoma(HCC) cell line with highly metastatic potential (MHCC97) and its ex-pressions of the factors associated with metastasis. Br J Cancer1999;81:814–21.




Zhang et al.

3430


27. Yang BW, Liang Y, Xia JL, et al. Biological characteristics of fluores-cent protein-expressing human hepatocellular carcinoma xenograftmodel in nude mice. Eur J Gastroenterol Hepatol 2008;20:1077–84.

28. Chang YS, Adnane J, Trail PA, et al. Sorafenib (BAY 43-9006) inhibitstumor growth and vascularization and induces tumor apoptosis andhypoxia in RCC xenograft models. Cancer Chemother Pharmacol2007;59:561–74.

29. Santini D, Vincenzi B, Galluzzo S, et al. Repeated intermittent low-dose therapy with zoledronic acid induces an early, sustained, andlong-lasting decrease of peripheral vascular endothelial growth fac-tor levels in cancer patients. Clin Cancer Res 2007;13:4482–6.

30. Daubine F, Le Gall C, Gasser J, Green J, Clezardin P. Antitumor ef-fects of clinical dosing regimens of bisphosphonates in experimentalbreast cancer bone metastasis. J Natl Cancer Inst 2007;99:322–30.

31. Ahn GO, Brown JM. Matrix metalloproteinase-9 is required for tumorvasculogenesis but not for angiogenesis: role of bone marrow-derived myelomonocytic cells. Cancer Cell 2008;13:193–205.

32. Kim SJ, Kim JS, Papadopoulos J, et al. Circulating monocytesexpressing CD31: implications for acute and chronic angiogenesis.Am J Pathol 2009;174:1972–80.



33. Sica A, Schioppa T, Mantovani A, Allavena P. Tumour-associatedmacrophages are a distinct M2 polarised population promotingtumour progression: potential targets of anti-cancer therapy. Eur JCancer 2006;42:717–27.

34. Lin EY, Li JF, Bricard G, et al. Vascular endothelial growth factorrestores delayed tumor progression in tumors depleted of macro-phages. Mol Oncol 2007;1:288–302.

35. Sharp AK, Colston MJ. The regulation of macrophage activity in con-genitally athymic mice. Eur J Immunol 1984;14:102–5.

36. Du R, Lu KV, Petritsch C, et al. HIF1α induces the recruitment ofbone marrow-derived vascular modulatory cells to regulate tumorangiogenesis and invasion. Cancer Cell 2008;13:206–20.

37. Grunewald M, Avraham I, Dor Y, et al. VEGF-induced adult neovas-cularization: recruitment, retention, and role of accessory cells. Cell2006;124:175–89.

38. Petit I, Jin D, Rafii S. The SDF-1-CXCR4 signaling pathway: a molec-ular hub modulating neo-angiogenesis. Trends Immunol 2007;28:299–307.

39. Siveen KS, Kuttan G. Role of macrophages in tumour progression.Immunol Lett 2009;123:97–102.




2010;16:3420-3430. Published OnlineFirst June 22, 2010.Clin Cancer Res Wei Zhang, Xiao-Dong Zhu, Hui-Chuan Sun, et al. Antimetastatic and Antiangiogenic EffectsEffect of Sorafenib in Metastatic Liver Cancer Models by Depletion of Tumor-Associated Macrophages Enhances the

Updated version

10.1158/1078-0432.CCR-09-2904doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2010/08/25/1078-0432.CCR-09-2904.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/16/13/3420.full#ref-list-1

This article cites 38 articles, 8 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/16/13/3420.full#related-urls

This article has been cited by 8 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] at

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/16/13/3420To request permission to re-use all or part of this article, use this link

Research. on August 23, 2021. © 2010 American Association for Cancerclincancerres.aacrjournals.org Downloaded from


http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-09-2904

http://clincancerres.aacrjournals.org/content/suppl/2010/08/25/1078-0432.CCR-09-2904.DC1

http://clincancerres.aacrjournals.org/content/16/13/3420.full#ref-list-1

http://clincancerres.aacrjournals.org/content/16/13/3420.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/16/13/3420


research depletion of tumor-associated macrophages ... · jor public health problem and the fifth...

Documents