Purpose: Local breast cancer relapse after breast-saving surgery and radiotherapy is associated with increased risk of distant metastasis formation. The mechanisms involved remain largely elusive. We used the well-characterized 4T1 syngeneic, orthotopic breast cancer model to identify novel mechanisms of postradiation metastasis.
Experimental Design: 4T1 cells were injected in 20 Gy preirradiated mammary tissue to mimic postradiation relapses, or in nonirradiated mammary tissue, as control, of immunocompetent BALB/c mice. Molecular, biochemical, cellular, histologic analyses, adoptive cell transfer, genetic, and pharmacologic interventions were carried out.
Results: Tumors growing in preirradiated mammary tissue had reduced angiogenesis and were more hypoxic, invasive, and metastatic to lung and lymph nodes compared with control tumors. Increased metastasis involved the mobilization of CD11b+c-Kit+Ly6GhighLy6Clow(Gr1+) myeloid cells through the HIF1-dependent expression of Kit ligand (KitL) by hypoxic tumor cells. KitL-mobilized myeloid cells homed to primary tumors and premetastatic lungs, to give rise to CD11b+c-Kit− cells. Pharmacologic inhibition of HIF1, silencing of KitL expression in tumor cells, and inhibition of c-Kit with an anti-c-Kit–blocking antibody or with a tyrosine kinase inhibitor prevented the mobilization of CD11b+c-Kit+ cells and attenuated metastasis. C-Kit inhibition was also effective in reducing mobilization of CD11b+c-Kit+ cells and inhibiting lung metastasis after irradiation of established tumors.
Conclusions: Our work defines KitL/c-Kit as a previously unidentified axis critically involved in promoting metastasis of 4T1 tumors growing in preirradiated mammary tissue. Pharmacologic inhibition of this axis represents a potential therapeutic strategy to prevent metastasis in breast cancer patients with local relapses after radiotherapy. Clin Cancer Res; 18(16); 4365–74. ©2012 AACR.
Patients with local breast cancer relapse after conservative breast cancer surgery and radiotherapy have an increased risk of metastasis formation. To date there is no effective therapy to prevent progression to metastasis in these patients. Using a well-characterized model of murine breast cancer, we identified the KitL/c-Kit axis as being critically involved in promoting metastasis of tumors growing in preirradiated mammary tissue. The mechanisms involve KitL-dependent mobilization of metastasis-promoting CD11b+c-Kit+Ly6GhighLy6Clow(Gr1+) myeloid cells. These findings have 2 immediate translational implications. First, circulating CD11b+c-Kit+ cells might be used as biomarkers to identify patients at risk for postradiation metastasis. Second, inhibition of c-Kit could be an attractive approach to improve efficacy of radiotherapy or to prevent metastasis in breast cancer patients with local relapses after radiotherapy. Translational studies aimed at validating these results in patients are warranted.
Adjuvant radiotherapy provides survival advantages compared with surgery alone, and it is nowadays standard treatment in the management of several cancers, including breast cancer (1). Despite progress in the delivery mode, locoregional postradiotherapy relapses still occur in a fraction of treated patients. Relapses occurring within a preirradiated area are associated with an increased risk of local invasion, metastasis formation, and poor prognosis compared with relapses occurring outside of the irradiated area, including in breast (2) and head and neck cancers (3). Experimental evidence supports these clinical observations. In murine xenograft models, tumors developing within preirradiated beds are more invasive and metastatic compared with tumors growing outside irradiated beds, a condition also referred to as tumor bed effect (4, 5). One common feature of the tumor bed effect is the appearance of hypoxia (6–8), which is likely due to the inhibition of angiogenesis by ionizing radiations (9, 10). Tumor hypoxia is associated with increased invasiveness, greater risk of metastasis formation, and shorter disease-free survival in different human tumors, including soft tissue sarcoma (11), head and neck (12), cervical (13), and breast cancers (14). In spite of its clinical relevance, the cellular and molecular mechanisms underlying the tumor bed effect are still not fully elucidated. Using human melanoma cells grafted in preirradiated, immunosuppressed mice, Rofstad and colleagues showed that interleukin-8 and the receptor of urokinase-type plasminogen activator are important mediators of the tumor bed effect in this model (8, 15). We have previously shown that the matricellular protein cysteine rich protein 61 (CYR61) and integrin αVβ5 cooperate to promote invasion and metastasis of squamous cell carcinoma and colorectal adenocarcinoma growing subcutaneously in preirradiated, immunosuppressed mice (7).
Kit ligand (KitL), also known as stem cell factor or Steele factor, is a cell-surface protein existing in 2 alternatively spliced isoforms (16). One isoform contains a cleavage site that allows protease-mediated shedding from the cell surface as homodimeric soluble KitL, whereas the other one cannot be released and remains associated to the cell surface (17). KitL binds to the tyrosine kinase receptor c-Kit (18). During development, KitL and c-Kit play critical roles in guiding cell migration, in particular, at sites of hematopoiesis, in the central nervous system, in the gut, and in the skin (melanogenesis; ref. 19). C-Kit is downregulated in adult tissues, except in hematopoietic stem/progenitor cells in the bone marrow, in melanocytes, and in mast cells. Accordingly, after development the KitL/c-Kit axis is essential for the maintenance of hematopoiesis and for mast cell survival and function in peripheral tissues (20). In cancer c-Kit acts as oncogene in several tumors, in particular gastrointestinal stromal tumors (GIST), mastocytosis, and melanoma (20), through activating mutations in the extracellular or intracellular domain (18) or through an autocrine KitL/c-Kit loop (21). The KitL also activates tissue-resident mast cells to generate a tumor-promoting angiogenic microenvironment (22, 23). A role of the KitL/c-Kit axis in metastasis formation, however, has remained unexplored.
In this work, we investigated cellular and molecular mechanisms underlying the tumor bed effect in breast cancer by using a model mimicking local relapse after radiotherapy. Presented results identify the KitL/c-Kit axis as a previously unsuspected mediator of metastasis in breast cancer.
Materials and Methods
Antibodies, reagents, and cell lines
Biotinylated rat anti-CD31 (MEC 13.3), fluorescein isothiocyanate (FITC)-conjugated anti-CD11b (M1/70), PE-conjugated anti-c-Kit (2B8), APC-conjugated anti-CXCR4 (2B11), PE-conjugated rat IgG2b isotype control (A95-1), and CD16/CD32 Fc-Block (2.4G2) were purchased from BD Biosciences. PE-conjugated anti-CCR5 (HM-CCR5), unconjugated and PercP-conjugated anti-F4/80 (BM8), PercP-conjugated anti-Sca1 (D7), Pacific blue–conjugated anti-CD31 (390), unconjugated and Pacific blue–conjugated anti-CD11b (M1/70), APC-conjugated anti-CD45 (30-F11), FITC-conjugated anti-CD11c (N418), FITC-conjugated Harmenian Hamster IgG2 isotype control (HTK888), Alexa Fluor 647–conjugated anti-Ly-6G, PerCP-conjugated anti-Ly-6C, PE-conjugated hamster IgG (HTK888), PercP-conjugated rat IgG2a (RTK2758), and Pacific blue–conjugated rat IgG2a (RTK2758) isotype controls were purchased from Biolegend. Pacific blue–conjugated anti-CD45 (30-F11), APC-conjugated anti-Gr-1 (RB6-8C5), APC-conjugated anti-CD123 (5B11), APC-conjugated rat IgG2a isotype control, and APC-conjugated rat IgG2b isotype control were purchased from eBioscience. APC-conjugated anti-VEGF-R1 (141522) was purchased from R&D. LIVE/DEAD fixable near-IR dead cell stain kit and 4′,6-diamidino-2-phenylindole were purchased from Invitrogen. The hydroxyprobe-1 kit for detection of tissue hypoxia was obtained from HPI Inc.. 4T1 cell line was generously provided by Dr. Fred R. Miller (Michigan Cancer Foundation, Detroit, MI; ref. 24). HEPA-1 C1C7 and HEPA-1 C4 cells were obtained from Dr. Isabelle Desbaillet-Hakimi (CHUV, Lausanne, Switzerland). For all experiments, cells were grown in Dulbecco's modified Eagle's medium high glucose supplemented with 10% fetal calf serum and 1% Penicillin/Streptomycin (all purchased from Invitrogen). For hypoxic treatment, cells were placed into a hypoxic chamber at 0.1% O2. HIF1-inhibitor NSC-134754 (Developmental Therapeutics Program, NCI/NIH) was used at 1 μmol/L.
Mouse model, irradiation, and drug treatment
Adult (5–7 weeks of age) BALB/c female mice (Charles River Laboratories) were used as host animals for grafted tumors. BALB/c mice expressing GFP under the α-actin promoter were generously provided by Dr. S. Swain (Trudeau Institute, Saranac Lake, NY). Primary tumors were initiated by the injection of 4T1 tumor cells (5 × 104 cells/mouse) into the right fourth mammary gland in 50 μL of 1:5 mixture of Matrigel (BD Biosciences) and PBS. Before injection, the fourth mammary gland was locally irradiated with a single 20 Gy dose by using an X-ray unit (PHILIPS, RT 250, Germany), operated at a 125 kV, 20 mA, with a 2-mm Al filter. Drugs were administered as following: Clodrolip: 2 mg/20 g mouse body weight as initial dose, followed by 1 mg for the subsequent doses injected intraperitoneally every fourth day starting from day 6; NSC-134754: 5 mg/kg dose, injected intraperitoneally daily from day 5; ACK2 (Biolegend): 50 μg dose, injected intraperitoneally 4 times every third day from day 10. Nilotinib (AMN107-AA; kindly provided by Novartis): 20 mg/kg dose, administered daily by gavage, from day 7. All animal experiments have been subjected to control and authorization by the cantonal veterinary service (Vaud and Fribourg). Tumor volume and lung metastases were assessed as previously described (7). For the quantification of lymphatic and liver metastases, axillary lymph nodes and liver sections were stained with hematoxylin and eosin (HE) and assessed for the presence of metastatic foci. For the irradiation of established tumors, 4T1 tumors were initiated as above, and when they were palpable (day 8 after tumor initiation), they were locally irradiated with the same settings (20 Gy, single dose). ACK2 treatment was carried out as above.
For flow cytometry on blood circulating cells, 50 μL of fresh blood were collected from the tail vein in 3 μL of 0.5 mol/L EDTA. Bone marrow–resident cells were collected from femoral bones and filtered to form single-cell suspensions. For flow cytometry analyses on tumors and lungs, mice were perfused with PBS by intracardiac injection. Mammary tissue, tumors, and lungs were excised, mechanically disrupted, enzymatically digested, and then filtered to obtain single-cell suspensions. Staining and acquisition were done as described (25). Samples were acquired with a FACS LSR II, FACScalibur (Becton-Dickinson) or MACS Quant Analyzer Miltenyi and data analyzed using FCS Express Version 3 (De Novo Software) or FlowJo (Tree Star, Inc.).
c-Kit+ cell depletion and fluorescence-activated cell sorting
Peripheral blood mononuclear cells (PBMC) were isolated using the Ficoll procedure. To assess the effect of circulating c-Kit+ cells on metastasis, 10 × 106 PBMCs were injected at day 8 and 12 either directly or after depletion of c-Kit+ cells using the EasySep Magnet procedure according to the manufacturer's instructions (Stemcell Technologies). For in vivo cell tracking, PBMCs were sorted using FACS Aria (Becton-Dickinson). The purity of the sorted samples was assessed by direct flow cytometry reanalysis and viability of the cells estimated with Trypan blue staining. A total of 2 × 106 c-Kit+ cells were then injected intravenously in 4T1-bearing mice.
Statistical comparisons were carried out by a 2-tailed Student t test or one-way ANOVA with Bonferroni posttest using Prism 5.0 GraphPad Software. Results were considered to be significant with P < 0.05. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Additional methods are available as Supplementary Material.
Tumors growing in preirradiated mammary tissue have decreased vascular density, are more hypoxic, invasive, and metastatic
To investigate the effect of mammary tissue irradiation on breast cancer progression, we preirradiated the fourth mammary gland of BALB/c mice with 20 Gy X-ray single dose before implanting 4T1 tumor cells (24). Whereas in current clinical practice, adjuvant radiotherapy in breast cancer is delivered in fractionated doses, in this model it was not possible to carry out multiple irradiations because of technical and ethical issues. Nevertheless, the single X-ray dose was chosen to correspond to the cumulative dose of approximately 60 Gy delivered to breast cancer patients during fractionated therapy (26). Mammary tissue preirradiation had no significant effect on 4T1 primary tumor growth (Supplementary Fig. S1A). Tumors growing in a preirradiated mammary tissue showed reduced microvascular density (MVD), increased hypoxia, and necrosis (Fig. 1A), consistent with radiation-induced inhibition of angiogenesis (9, 10). They were also more invasive into the surrounding muscle and fat tissues (Fig. 1B) and more metastatic to ipsilateral axillary lymph nodes (Fig. 1C), lungs (Fig. 1D), and liver (Supplementary Fig. S1B).
These results showed that 4T1 breast tumors growing in a preirradiated mammary tissue undergo a tumor bed effect that recapitulates clinically relevant features of breast cancers locally relapsing after radiotherapy.
Tumors growing in preirradiated mammary tissue recruit metastasis-promoting CD11b+ cells
In tumor growing in preirradiated beds, we observed a significant increase in the recruitment of CD11b+ myeloid cells, and of 2 subpopulations thereof, CD11b+F4/80+ and CD11b+Gr1+ (Fig. 2A and Supplementary Fig. S2A). Irradiation of the mammary tissue without tumor implantation did not induce recruitment of CD11b+ myeloid cells (Fig. 2A). Enhanced recruitment occurred in the tumor periphery as shown by F4/80 staining (Fig. 2B). Treatment with clodrolip to deplete phagocytic myeloid cells (27) did not affect primary tumor growth (Supplementary Fig. S2B), but significantly reduced the number of CD11b+ and F4/80+ cells in the tumor (Fig. 2C, Supplementary Fig. S2C, left panel, and Supplementary Fig. S2D), MVD (Supplementary Fig. S2C, right panel, and Supplementary Fig. S2D), and lung metastasis formation (Fig. 2D and Supplementary Fig. S2E).
These results showed that mammary tissue irradiation enhances recruitment of CD11b+ cells contributing to lung metastasis formation.
HIF1-dependent kit ligand expression in hypoxic tumor cells promotes metastasis
Hypoxic tumors can mobilize bone marrow–derived cells by releasing soluble factors (28). We therefore screened for cytokines induced by hypoxia in 4T1 cells and potentially involved in CD11b+ cell mobilization. In addition to VEGF-A and MCP-1, KitL was among the cytokines significantly induced by hypoxia (24 hours, 0.1% O2) in 4T1 cells in vitro at both the mRNA and protein levels (Fig. 3A and Supplementary Fig. S3A). Because bone marrow–derived cells expressing the KitL receptor c-Kit (20) are present in the premetastatic niche (29), but the putative role of KitL/c-Kit axis in metastasis formation has not been assessed yet, we decided to focus on KitL. VEGF-A expression was monitored as a control for effective hypoxia and activity of the HIF1 inhibitor NSC-134754 (ref. 30; Supplementary Fig. S3A and S3B). Inhibition of HIF1 in 4T1 cells by NSC-134754 and genetic deficiency of the β subunit of the HIF complex in HEPA1 cells prevented hypoxia-induced upregulation of KitL mRNA and protein (Fig. 3A, Supplementary Fig. S3B and S3C). 4T1 tumors growing within a preirradiated bed had increased KitL protein (Fig. 3B) and mRNA (Supplementary Fig. S4A) levels, and this increase was blunted by NSC-134754 treatment (Fig. 3B). Mammary tissue irradiation alone did not induce KitL expression (Supplementary Fig. S4B). Plasma levels of KitL were higher in preirradiated tumor-bearing mice compared with controls (Fig. 3C) showing systemic release of KitL (31). In contrast, CXCL12, a chemokine previously reported to promote CD11b+ cell recruitment within irradiated tumors (32), was not induced in this model (Supplementary Fig. S4C). Silencing of KitL expression in 4T1 cells through lentiviral mediated expression of KitL-specific short-hairpin (sh) RNAs (Supplementary Fig. S4D and S4E) suppressed lung metastasis formation of 4T1 tumors growing in preirradiated mammary tissue (Fig. 3D), whereas it did not affect 4T1 tumor cell growth in vitro or in vivo (Supplementary Fig. S4F and S4G).
These results showed that KitL expression is induced by hypoxia in a HIF1-dependent manner, and that KitL is required for increased metastatic spreading, but not primary growth, of 4T1 tumors implanted in preirradiated mammary tissue.
KitL mobilizes CD11b+c-Kit+ cells
To ascertain cells responding to KitL, we set up to identify c-Kit-expressing cells in tumor-bearing mice. C-Kit expression was undetectable on CD45− and CD45+ cells recovered from primary tumors, whereas it was detectable at low frequency on bone marrow cells and circulating CD45+ myeloid cells (Fig. 4A and Supplementary Fig. S5A). C-Kit+ cells were virtually undetectable in the blood of tumor-free mice (Fig. 4B). Circulating c-Kit+ cells in preirradiated, tumor-bearing mice expressed CD11b and Gr1 but were negative for F4/80, CCR5, CXCR4, VEGF-R1, Sca1, and CD123 (a marker for mast cells; ref. 33) expression (Supplementary Fig. S5B). Their morphology was consistent with young, immature myeloid cells (Supplementary Fig. S5C). Circulating c-Kit+ cells were unable to form hematopoietic colonies, in contrast to bone marrow–derived c-Kit+ cells (Supplementary Fig. S5D). Further phenotypical analysis revealed that circulating c-Kit+CD11b+ cells were Ly6Ghigh and Ly6Clow and CD11c negative (Supplementary Fig. S6A and S6B). Tumor-infiltrating CD11b+ cells were predominantly Ly6Ghigh (Supplementary Fig. S6C). The Ly6GhighLy6Clow phenotype is indicative of CD11b+ granulocytic myeloid-derived suppressor cells (MDSC; ref. 34), consistent with the well-documented ability of 4T1 tumors to expand and mobilize MDSCs (35, 36). Tumors implanted in preirradiated mammary tissue, but not mammary tissue irradiation alone, significantly enhanced the frequency of circulating CD11b+c-Kit+ cells, compared with control tumors (Fig. 4B), and KitL silencing in 4T1 tumor cells or systemic administration of NSC-134754 significantly reduced it (Fig. 4C).
Mobilized CD11b+c-Kit+ cells home to tumors and premetastatic lungs in a KitL-dependent manner
To monitor the fate of circulating c-Kit+ cells, we isolated circulating GFP+CD11b+c-Kit+ cells from tumor-bearing GFP-BALB/c mice (more than 95% enriched, not shown) and injected them into recipient BALB/c mice bearing 4T1 tumors implanted in preirradiated mammary tissue. Six hours after intravenous injection, GFP+CD11b+ cells were detectable in the blood and in tumor tissue; however, only approximately 25% and none of the transferred cells retained c-Kit expression, respectively (Fig. 5A). KitL silencing in 4T1 tumors significantly reduced radiation-induced recruitment of CD11b+ cells to primary tumors (Fig. 5B). In a second adoptive transfer experiment, we showed that GFP+CD11b+c-Kit+ cells also home to lungs of BALB/c recipient mice bearing tumors growing in preirradiated mammary tissue, with only approximately 25% of them retaining c-Kit expression (Fig. 5C). C-Kit+ cells were also present in lungs of tumor-bearing mice 15 days after tumor implantation, and their frequency increased in mice bearing tumors growing in preirradiated beds compared with nonirradiated controls. This increase was blunted by KitL silencing in tumor cells (Fig. 5D).
These results indicated that tumor-derived KitL mobilizes CD11b+c-Kit+ cells to home to primary tumors and premetastatic lungs, but once recruited they rapidly lose c-Kit expression.
Mobilized CD11b+c-Kit+ cells promote lung metastasis
To obtain evidence whether tumor-mobilized c-Kit+ myeloid cells promote lung metastasis, we isolated PBMCs from preirradiated tumor-bearing mice and injected either the total PBMCs or PBMCs depleted of c-Kit+ cells (Supplementary Fig. S7A) into nonirradiated tumor-bearing mice and monitored lung metastasis formation. Transfer of total PBMCs increased lung metastasis by about 5-fold, whereas transfer of c-Kit+ cell–depleted PBMCs resulted in a significantly reduced increase in lung metastasis (Fig. 6A). No effects on primary tumor growth were observed (Supplementary Fig. S7B).
These results indicated that mobilized CD11b+c-Kit+ cells significantly contributed to promote lung metastasis.
c-Kit inhibition reduces mobilization of CD11b+c-Kit+ cells and attenuates lung metastases
To test whether inhibition of c-Kit impinges on lung metastasis formation, we treated preirradiated, tumor-bearing mice with the anti-c-Kit blocking antibody ACK2 (37). ACK2 treatment suppressed radiation-induced mobilization of c-Kit+ cells, accumulation of CD11b+ cells in primary tumors (Fig. 6B), and attenuated lung metastasis formation by approximately 50% (Fig. 6C, top panel). Treatment with nilotinib (AMN107-AA), a tyrosine kinase inhibitor that effectively blocks c-Kit activity (38), suppressed lung metastasis formation to a similar extent (Fig. 6C, bottom panel). ACK2 or nilotinib treatments did not impinge on primary tumor growth (Supplementary Fig. S7C and S7D). To assess whether this mechanism may also apply to irradiated tumors, we irradiated established 4T1 tumors. Irradiation significantly delayed tumor growth and increased lung metastasis (Supplementary Fig. S8A and S8B). Treatment with ACK2 reduced the frequency of circulating CD11b+c-Kit+ cells and the number of lung metastases (Supplementary Fig. S8C and S8D) without impinging on tumor growth (Supplementary Fig. S8E).
These results showed that c-Kit activity is critically involved in mediating the mobilization of CD11b+c-Kit+ cells and in promoting lung metastasis formation of 4T1 tumors growing in preirradiated mammary tissue or after tumor irradiation.
In spite of their clinical relevance, the cellular and molecular mechanisms mediating metastasis of tumors locally relapsing after radiotherapy are still largely elusive. In this work while using the well-characterized 4T1 murine breast cancer model, we discovered a novel mechanism promoting metastasis of breast cancer growing in a preirradiated bed.
The main novel finding is the identification of the KitL/c-Kit axis as a crucial element of this mechanism. We show that tumors growing in a preirradiated bed are poorly vascularized and hypoxic, and that hypoxia induces HIF1-dependent KitL expression by tumor cells. Released KitL mobilizes CD11b+c-Kit+Ly6GhighLy6Clow (Gr1+) myeloid cells, which recruit to primary tumors and to lungs and facilitate lung metastasis formation. Expression of KitL in perinecrotic tumor regions was reported in human cancers, including breast cancer, and correlates with progression (39). The tumor-promoting effect of KitL was considered due to KitL stimulating c-Kit+ tumor cells or c-Kit+ endothelial cells and mast cells, resulting in enhanced cell growth (40), angiogenesis (41), and immunosuppression (42).
The second main finding is the identification of CD11b+c-Kit+Ly6GhighLy6Clow (and F4/80−CCR5−CXCR4−Flt1−CD123−) cells as novel CD11b+ subpopulation with metastasis-promoting activity. Recruited CD11b+c-Kit+ cells rapidly lose c-Kit from their surface, to give rise to CD11b+c-Kit− cells, which may be the actual cells promoting metastasis. This is consistent with the enriched presence of CD11b+ cells at the invasive tumor front and the observed reduced metastasis formation following depletion by clodrolip. CD11b+ cells are known to promote tumor growth, invasion, and metastasis through multiple mechanisms, including angiogenesis (43), vasculogenesis (10, 32), immunosuppression (44), lymphangiogenesis (45), and tumor cell motility (46). Taken together, our results identify KitL as a factor inducing the mobilization and recruitment to tumors and lungs of CD11b+ myeloid cells, in addition to already identified factors, such as VEGF-A (47), CXCL12 (48), CXCL5 (46), and PlGF (49). We are currently investigating the mechanisms by which CD11b+c-Kit+ cells (and CD11b+ c-Kit− cells derived thereof) promote metastasis. The reduced MVD induced by clodrolip treatments in tumors growing in a preirradiated bed suggests that these cells possess vessel-forming properties, possibly by vasculogenesis, as recently reported (32). The Ly6GhighLy6Clow phenotype consistent with a granulocytic MDSC population (34) also raises the possibility that immunosuppressive effects might contribute to promote metastasis.
These results have 2 immediate translational implications. First, they identify circulating CD11b+c-Kit+ cells as candidate biomarkers to stratify patients at risk for postradiation metastasis. We have attempted to validate this hypothesis using archive material, but the limited number of cases identified and restrictive local ethical regulations hindered this approach. Preliminary analyses in humans indicate that CD11b+c-Kit+ are present at low frequency in the peripheral blood of healthy volunteers and at higher frequency in patients with metastatic breast cancer.
Second, they point to the KitL/c-Kit axis as a potential therapeutic target to prevent metastasis of postradiation breast cancer relapses. C-Kit inhibition by tyrosine kinase inhibitors (TKI) is already used to treat cancers with uncontrolled c-Kit activity (ref. 50; due to c-Kilt overexpression, mutational activation, or KitL overexpression), including GIST (51) and melanoma (52). Here, stable silencing of KitL in 4T1 tumor cells, selective c-Kit inhibition by the ACK2 mAb, and treatment with nilotinib, a clinically approved TKI that effectively block c-Kit activation (38), all reduced metastasis of tumors growing in a preirradiated mammary tissue. C-Kit inhibition also reduced lung metastasis formation of 4T1 tumor that were treated with radiotherapy. On the basis of these findings, inhibition of c-Kit should be explored as a possible adjuvant approach in patients with local recurrences after radiotherapy and at increased risk for metastatic progression. It is important to note that these findings are based on one breast cancer cell line only, 4T1. This cell line is known to induce an unusually strong mobilization of MDSCs, which may not occur in all breast cancer models. It will also be important to ascertain whether in human breast cancer this mechanism is common to all breast cancer types, or whether it is specific for a particular breast cancer subtype. To this purpose, we are interrogating additional experimental breast cancer models and initiating a clinical study to monitor the frequency of circulating CD11b+c-Kit+ cells in breast cancer patients, before, during, and after surgery/radiotherapy. Monitoring CD11b+c-Kit+ cells in the circulation might also be a simple strategy to identify patients with an active KitL/c-Kit axis that may benefit from c-Kit inhibition.
Taken together we have shown that hypoxic breast tumors growing in a preirradiated environment mobilize CD11b+c-Kit+Ly6GhighLy6Clow granulocytic MDSCs through HIF1-mediated KitL release. These observations extend our understanding of the mechanisms of metastasis of tumors growing in a preirradiated bed by implicating for the first time the KitL/c-Kit axis in this effect (Fig. 6D). Translational studies aimed at validating these results in patients are warranted.
Disclosure of Potential Conflicts of Interest
C. Rüegg has ownership interest (including patents) in Diagnoplex SA. No potential conflicts of interest were disclosed by the other authors.
Conception and design: F. Kuonen, C. Rüegg
Development of methodology: F. Kuonen, J. Laurent, R. Schwendener, R.-O. Mirimanoff, C. Rüegg
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): F. Kuonen, J. Laurent, C. Secondini, E. Faes-van't Hull, G. Bieler, C. Rüegg
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): F. Kuonen, J. Laurent, C. Secondini, G. Lorusso, G. Bieler, S. Andrejevic-Blant, C. Rüegg
Writing, review, and/or revision of the manuscript: F. Kuonen, J. Laurent, C. Secondini, G. Lorusso, G.-C. Alghisi, S. Andrejevic-Blant, R.-O. Mirimanoff, C. Rüegg
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): F. Kuonen, J. Laurent, G. Bieler, G.-C. Alghisi, R. Schwendener
Study supervision: J. Laurent, C. Rüegg
The authors thank Drs. F.R. Miller, S. Swain, L. Naldini, R. Iggo, and A. Follenzi for providing cells, mice, or lentiviral vectors; Drs. C. Bourquin and T. Herbst for assistance with FACS analysis and discussion; and Dr. P. Manley for discussion and providing nilotinib.
This work was supported by the European Union under the auspices of the FP7 collaborative project TuMIC, contract no. HEALTH-F2-2008-201662 (to C. Rüegg); The Molecular Oncology Program of the National Center of Competence in Research (NCCR), a research instrument of the Swiss National Science Foundation (to C. Rüegg); a collaborative research project (CCRP) of Oncosuisse (OCS-01812-12-2005; to C. Rüegg); a Swiss National Science Foundation grant (31003A_135738; to C. Rüegg); The ISREC Foundation (to C. Rüegg); and a MD-PhD fellowship from Oncosuisse (to C. Rüegg).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.