Knockout of MDA-9/Syntenin (SDCBP) expression in the microenvironment dampens tumor-supporting inflammation and inhibits melanoma metastasis

Cancer development and progression to metastasis is a complex process, which largely depends on bidirectional communication between tumor cells and their microenvironment. Melanoma differentiation associated gene-9 (mda-9, also known as Syntenin-1, SDCBP), a gene first cloned by our group, is robustly expressed in multiple cancers including melanoma and contributes to invasion and metastasis in a tumor cell-intrinsic manner. However, the role of MDA-9/Syntenin in the tumor cell-extrinsic microenvironment remains unclear even though MDA-9/Syntenin is ubiquitously expressed in most organs that are active metastatic sites for melanoma, e.g., lung, lymph node, brain, and liver. In this study, we explored the effect of environmental mda-9/syntenin expression on melanoma growth and metastasis using multiple immunocompetent animal models, syngeneic B16 xenograft and intravenous B16 mouse model and a genetically engineered mouse (GEM) model of melanoma. Host-deficient expression of mda-9/syntenin in mice negatively impacted on subcutaneously implanted B16 tumor growth and lung metastasis. Absence of MDA-9/Syntenin in the lung microenvironment suppressed tumor growth by modulating in situ Interleukin 17A (IL17A) expression and impaired the recruitment of myeloid derived suppressor cells (MDSCs) and Th17 cells as compared to genetically wild type animals. Additionally, loss of mda-9/syntenin expression in a spontaneous melanoma model (melanocyte-specific pten loss and BrafV600E mutation) significantly delayed tumor initiation and suppressed metastasis to the lymph nodes and lungs. The present study highlights a novel role of mda-9/syntenin in tumor-promoting inflammation and immune suppression. These observations along with other documented roles of MDA-9/Syntenin in cancer and metastasis support the potential relevance of MDA-9/Syntenin in the carcinogenic process and as a target for developing improved therapies by using either genetic or pharmacologic approaches to treat and prevent melanoma and other cancers.


INtrODUctION
Metastasis is a multifaceted process that initiates with the migration of tumor cells from a primary site and culminates in the formation of new secondary tumors in distant organs [1]. For successful survival and growth of metastatic cells in their new environment, disruption of homeostasis of the host organ is crucial [2]. Each organ contains multiple cell types organized in an ordered/structured architecture and numerous molecular/ biochemical processes are necessary to maintain physiological functions [1]. Through mechanisms that require further clarification, specific biochemical and molecular changes, either induced by invading cancer cells or occurring as spontaneous events, can disrupt this organization and facilitate tumor cell colonization and growth [3]. Numerous studies support the importance of cross talk between the resident cells (in target organs) and invading cells in mediating metastasis [3]. In principle, the homing organ can serve as a 'perfect ecosystem' that can provide nutritional support and protection from the immune system to facilitate tumor growth.
Accumulating evidence suggests that a local immune-suppressive and inflammatory microenvironment is a key element for tumor progression and invasion [34][35][36]. Myeloid derived suppressor cells (MDSCs), a heterogeneous population of cells of myeloid origin, have garnered attention due to their immune suppressive functions in a tumor bearing host [37][38][39]. These effects are elicited by suppressing effector T cells [37], converting naïve CD4 + T cells to regulatory T cells (Tregs) [40] and inhibiting T cell trafficking [41]. CD4 + Th17, a subset of CD4 + T cells, is an additional type of immune suppressive cell that also infiltrates tumors and correlates with tumor progression [42]. Interleukin 17A (IL-17A), a pro-inflammatory cytokine secreted by CD4 + Th17 cells, triggers tumor cells to produce interleukin 6 (IL-6), which in turn activates STAT3-dependent survival and angiogenesis [43]. Additionally, IL-17 production in the tumor microenvironment promotes infiltration of MDSCs to promote immune suppression and to amplify tumorpromoting inflammation [44]. The behavior of cancer cells is influenced to a great extent by various cytokines produced by resident immune or non-immune cells in the tumor microenvironment in response to invading tumor cells. In this study, we show that lack of mda-9/ syntenin expression in the host lung influences the local inflammatory network, indicated by the reduced level of pro-inflammatory cytokines such as IL-6 and IL-17A, as well as diminished accumulation of Th17 cells and MDSCs. This defect in tumor-supporting inflammation strongly suppresses tumor progression as evidenced by a delay and reduction in metastatic melanoma development.

Phenotype of mda-9/syntenin knockout (mda-9 -/-) mice
As reported previously by Tamura et al. [28] using syntenin-1-deficient mice, we did not observe any phenotypic differences (Supplementary Figure S1), embryonic lethality, developmental abnormalities, or adult sterility in our mda-9 -/mice. Additionally, body weight changes up-to six months and various organ weights at 3 months were similar between mda-9 -/and wild type (WT) C57BL/6 mice (data not shown). We confirmed in our mda-9 -/mice the absence of MDA-9/Syntenin protein in multiple tissues (lung shown) from mice homozygous mda-9-deficient mice are less supportive for murine B16-derived xenograft tumor growth Our previous studies firmly established the relationship between mda-9/syntenin expression and melanoma metastasis [8,45]. In the present study, we evaluated mda-9/syntenin as a host factor and defined whether host expression could influence tumor growth when B16 cells were implanted subcutaneously in mda-9 -/mice and their WT littermates (C57BL/6). B16 cells abundantly express MDA-9/syntenin (data not shown, [46]. Overall, B16 cells grew faster in WT mice compared with mda-9 -/mice. During the time course of the experiment (from day 3 to day 24) delayed tumor growth was observed in the mda-9 -/vs. mda-9 WT mice ( Figure 1B). The tumor volumes from WT mice were ~2fold greater than tumor volumes from mda-9 -/mice at day 15 post-inoculation. By day 18 and 24, ~2.5-fold greater tumor volumes were found in WT mice as compared to mda-9 -/mice. The average tumor weight was ~40% less in the mda-9 -/group than in the WT group ( Figure  1C). In total, these results suggest that mda-9/syntenin deficiency in the microenvironment negatively impacted tumor (melanoma) growth. H & E sections from WT mice indicated substantial pigmented cells in tumors, which were less apparent in the mda-9 -/group ( Figure 1D). Oncotarget 46851 www.impactjournals.com/oncotarget mda-9/syntenin deficiency in the lungs of mice modulates B16 lung nodule growth The lungs are the most common site for melanoma metastases [47]. Injection of B16 cells through the lateral tail vein results in pulmonary metastasis in C57BL/6 animals. To define a potential impact of mda-9/syntenin deficiency in the host lung on the development of metastatic nodules, B16 cells were injected into the tail vein of mda-9 -/and WT mice. Lung metastases were evaluated by gross morphology of the lungs (Figure 2A), determining the number of metastatic nodules on the lung surface ( Figure 2B), and lung morphology (presence of tumor and lung structure) ( Figure 2C) on days 15 and 21 [48]. Overall, the WT group manifested enhanced pathology compared with mda-9 -/mice, as indicated by an increased number of pigmented melanoma nodules in the lungs at each time point (Figure 2A). Visible lung nodules were first apparent between day 3 to day 6 after B16 injection in WT mice, while nodules only became apparent at day 6 in the mda-9 -/mice, and dramatically fewer nodules were present at this time point (mda-9 -/-: 0.33 ± 0.57 vs. WT: 9.33 ± 5.5). Similarly, at later time points (day 9, 15 and 21) the difference in lung nodule numbers was also significantly different between these two groups (Figure 2A and 2B). Additionally, the size of nodules was comparatively larger at all later time points (day 15 and 21, Figure 2C) in WT mice.
In a second experiment, mda-9/syntenin expression was downregulated in B16 cells using shRNA for mda-9/ syntenin (shmda-9), and these and control cells (shcon) were injected via tail vein into both WT and mda-9 -/mice. After 6 days, mice were sacrificed and nodules the lungs of WT and mda-9 -/mice. A cohort of 15-age (6 weeks) and sex (male) matched wild type (C57BL/6) and mda-9 -/mice were inoculated with B16 cells (1 X 10 5 ) by intravenous injection to generate experimental lung metastases. Every 3 rd day, 3 mice from each group were sacrificed and lungs were collected, fixed with formalin and examined for nodules. Representative photographs of lungs with tumor metastases are presented. B. Graphical representation of average surface nodule number/mouse from each time point is presented. C. Photomicrographs of gross hematoxylin/eosin-stained lung sections are shown. D. In a separate study, a cohort of 6 animals from each mouse type (WT and mda-9 -/-) was inoculated with shcon or shmda-9 transfected B16 cells through tail-vein injection. After 6 days, lungs from each group were analyzed for lung nodule development. Representative photographs and the average number of nodules/mouse from each experimental group are shown in the upper and lower panel, respectively. E. Representative photomicrographs of gross lung hematoxylin/eosin stained lung sections are shown.

Inflammatory and immunosuppressive cells are reduced in the lungs of mda-9 -/mice
Since we observed a lower number of metastatic nodules in the lungs of mda-9 -/mice as compared with their WT counterparts (Figure 2), we hypothesized that mda-9/syntenin expression in the tumor microenvironment might result in an alteration in tumor-supporting inflammation in mda-9/syntenin-deficient mice. To test this assumption, we first analyzed the expansion of MDSCs in the lungs of tumor-bearing WT and mda-9 -/mice using flow cytometry. Following the establishment of metastatic B16 tumors, lungs from WT mice contained significantly higher numbers of both monocytic MDSCs (M-MDSCs, CD11b + Ly6C high Ly6G -) and granulocytic MDSCs (G-MDSCs, CD11b + Ly6C low Ly6G + ) than those from mda-9 -/mice ( Figure 3A and 3B). The differences were significant for M-MDSCs, a prominent subtype for melanoma [49], at all time points analyzed in this study (day 3, 9 and 15). It is worth noting that at day 21, the accumulation of MDSCs (both subtypes) in mda-9 -/mice were not significantly different from the wild type (Supplementary Figure S2). It is possible that the lack of mda-9 expression in the microenvironment has a more significant impact on host-mediated anti-tumor activity at early stage of tumor progression and over time the tumor cells in the niche overcome the immune suppression and their growth is expanded. Consistent with this hypothesis, we observed nodule sizes (representing increased tumor cell number) were dramatically larger on day 21 compared with day 15 (Figure 2A) in mda-9 -/mice.
Interestingly, we observed several pro-inflammatory cytokines such as IL-17A, IL-6 and TNF-α that were downregulated (at an mRNA level) in mda-9-deficient naïve lungs, suggesting a potential impairment of protumorigenic responses to invading tumor cells ( Figure  4A). To determine the consequences of absence of mda-9 in the lungs of mice, we next examined the infiltration of CD4 + Th17 populations in the lungs of WT and mda-9 -/mice in response to intravenous injected B16 cells. Flow cytometry analysis revealed lower numbers of CD4 + Th17 cells in the lung tumors from mda-9 -/mice as compared to WT mice only at days 9 and 15 (total accumulation was gradually decreases as tumor progression occurred in both groups). Again, consistent with the MDSCs data presented in Supplementary Figure S2, CD4 + Th17 populations at day 21 were similar between WT and mda-9 -/mice. To evaluate potential changes in the expression of the pro-inflammatory cytokine IL-17A in response to B16 cell colonization, qPCR was performed with RNA isolated from mouse lungs. At day 3, lower levels of IL- Oncotarget 46854 www.impactjournals.com/oncotarget 17A mRNA were evident in mda-9 -/mice which was comparable to that of the naïve lungs (compare Figure  4A vs Figure 4C). At this time point, we did not find any difference in CD4 + Th17 cells (data not shown). A significant downregulation of IL-17A mRNA ( Figure 4C) correlated with the CD4 + Th17 cells at day 9 ( Figure 4B and Figure 4C). However, at day 15, we did not observe any differences of IL-17A mRNA in mda-9 -/and WT groups. We did not explore the molecular mechanism in detail, but it is possible that other resident cells of the tumor microenvironment such as macrophages or T cells might produce IL-17A as a consequence of the presence of localized tumor cells. Overexpression of IL-17A might then facilitate the suppression of the host-mediated immune surveillance and allow the tumor cells to grow in mda-9 -/mice, as evident in day 15 or later time points ( Figure 2B). Further systematic studies are required to explore these phenomena, which are currently a focus in our laboratory.

Deletion of mda-9/syntenin in BRaf V600E /Pten fl/fl / Tyr-cre ER2 mice delays tumor onset and inhibits formation of distant metastases
In the BRaf V600E /Pten fl/fl /Tyr-cre ER2 mouse model, tamoxifen-mediated activation of cre results in activation of expression of the Braf oncogene and knockout of the Pten tumor suppressor gene specifically in melanocytes [50]. In the absence of tamoxifen, most of these mice display no discernible phenotype. However, topical or systemic administration of 4-hydroxytamoxifen (4-HT) results in the rapid development of melanomas that recapitulate the key pathophysiological features of human melanoma, including metastases in lymph nodes and lungs, with animals requiring euthanasia within 3-6 weeks. To comprehend the role of mda-9/syntenin in melanoma progression, we crossed mda-9 -/mice with the BRaf V600E / Pten fl/fl /Tyr-cre ER2 animals to develop mice that lack MDA-9/Syntenin expression. Systemic administration of 4-hydroxytamoxifen (4-HT) to parental BRaf V600E /Pten fl/fl / Tyr-cre ER2 mice induced pigmented lesions within 7 to 10 days ( Figure 5A). However, deletion of mda-9/syntenin (BRaf V600E /Pten fl/fl /Tyr-cre ER2 /mda-9 -/-) resulted in delayed tumor formation ( Figure 5A). Additionally, we also observed less metastasis to regional lymph nodes and lungs in BRaf V600E /Pten fl/fl /Tyr-cre ER2 /mda-9 -/-( Figure 5B), which were assessed pathologically (presence of pigmented cells or lung morphology). These results provide further documentation of the role of mda-9/syntenin expression in the microenvironment in regulating primary tumor and metastasis formation in an autochthonous transgenic mouse model of melanoma containing an intact immune system.

DIscUssION
The primary direction of research on MDA-9/ Syntenin (SDCBP) in cancer over the past two decades focused on defining its role in the context of tumor cells from different organ sites (7-11, 30,31,33,45,52,53) Although expression is significantly lower in normal vs. transformed cells, the MDA-9/Syntenin protein is ubiquitously expressed in the mouse (in both developing and adult stages [28,51]) and in adult human organs, e.g., skin, liver, gall bladder, colon (www. Proteinatlas.org). Tamura et al. first demonstrated that knockout of MDA-9 (syntenin-1; SDCBP) in mice was not embryonic lethal and mice that developed did not show obvious abnormalities when grown in pathogen-free conditions [28]. The mda-9 -/mouse model used in the present study was independently generated and characterized, corroborating the prior results that deletion of mda-9/syntenin in all tissues of the mouse is not lethal. In agreement with Tamura et al. [28], we also hypothesize that the lack of mda-9/syntenin in this animal is probably compensated by another variant of Syntenin, such as Syntenin-2, thereby preventing this knockout from being lethal.
MDA-9/Syntenin functions as a scaffold protein that interacts with a multitude of partner protein(s) and is involved in the generation or stabilization of active complexes, which play quintessential roles in multiple physiological activities [52,53]. In this way mda-9 can elicit various effects depending on its location in a cell and its binding partners. As indicated, previous studies have focused primarily on the role of mda-9 in cancer cells without defining the role of this molecule in normal tissue and the microenvironment. We hypothesized that lack of mda-9/syntenin expression might impact significantly inflammatory cytokine(s) in the microenvironment through STAT-3/NF-κB pathways. This positive loop and accumulation of proinflammatory cytokines then recruit inflammatory and immunosuppressive cells (e.g., Th17 cells and MDSCs) to support and facilitate tumor outgrowth. We hypothesize that MDA-9/Syntenin plays a critical role in this early induction phase of the metastatic process by promoting an inflammatory and immunosuppressive environment, possibly via the STAT-3/NF-κB pathways. Absence of MDA-9/Syntenin negatively impacts on the inflammatory cascade and blocks the mobilization of inflammatory and immunosuppressive cells, which, as a consequence, leads to reduced tumor cell colonization and/or growth.
Oncotarget 46856 www.impactjournals.com/oncotarget only when cells were challenged with exogenous insults, such as those resulting from implantation or injection of tumor cells. To validate this hypothesis, we examined the tumorigenic/metastatic behavior of B16 cancer cells when placed in contact with mda-9/syntenin deficient tissue in an mda-9 -/mouse. As predicted, tumor growth and metastasis was suppressed in mda-9 -/mice as compared with WT animals. Migration and seeding of cancer cells in the lung after tail vein injection depends on the interactions of tumor cells and the resident host cells, i.e., the microenvironment [54]. Many studies have shown that the microenvironment has the capacity to eliminate, minimize or enhance tumorigenesis [55,56] depending on its composition and the stromal cell proportions or their activation states [34,57].
Our study reveals that ablation of mda-9/syntenin impairs tumor-promoting inflammation and tumorassociated immunosuppression, characterized by fewer inflammatory and immunosuppressive cells, e.g., MDSCs, Th17 in the lungs with metastases. Malignant progression is often associated with inflammation and immunosuppression in the tumor sites [58]. The potential tumor-promoting role of MDSCs has been well documented in studies involving multiple animal models (reviewed by Quail et al, [55,59] and clinical samples [60]. Pro-inflammatory IL-17A-producing Th17 cells, are often associated with tumors (reviewed Guery et al, [61]) with both good and bad prognoses. The tumor-promoting mechanism of Th17 cells is dependent on the production of IL-17A, a pro-inflammatory cytokine, that is known to promote tumor angiogenesis [62,63], stimulate tumor cells to produce IL-6 [43], and in some murine models recruit MDSCs within tumors [44]. Additionally, Th17 cells may also exert immunosuppressive functions via direct inhibition of tumor-specific CD8 + T effector cells [64], and development of IL-17 + Foxp3 + regulatory T cells [65]. The recruitment and expansion of various immune cells at least partially define the inflammatory status of the tumor microenvironment [66]. Although the precise mechanisms require further clarification, our results indicate that the basal levels of two major pro-inflammatory cytokines IL-6 and IL-17A are downregulated in the lungs from mda-9 -/mice. IL-17A is predominantly produced by immune cells, including T cells, macrophages, dendritic cells (DC), natural killer cells, natural killer T cells, and γδ-T cells [67]. Tissue resident cells such as fibroblasts also modulate the local inflammatory conditions by producing various cytokines and chemokines [57]. Although context dependent, STAT-3 activation has been reported to be involved in IL-17 regulation [68,69]. Considering that we have previously shown that MDA-9/Syntenin can activate STAT-3 in cancer cells [33], we postulate that lack of mda-9/syntenin in our mda-9 -/mice may abrogate STAT-3-mediated IL-17 production. Precisely how MDA-9/Syntenin regulates IL-17 expression is currently an area of active investigation.
Although somewhat controversial, IL-17 is often considered to exert its' tumor-promoting effects in the context of inflammation [55]. More precisely, when IL-17 is exogenously produced in the microenvironment, tumor growth is enhanced as documented in studies with IL-17 -/mice [43,44,70] or intra-tumoral injection of IL-17 siRNA [71]. Using B16 cells, Wang et al [43] demonstrated that IL-17A induced IL-6 production by tumor cells, which in turn enhanced tumor growth in a STAT-3-dependent manner. Additionally, this study also revealed strong CD + 8 T cell accumulation and IFN -γ production in IL-17 -/mice. Our mda-9 -/mouse model, which has a lower level of IL-17 mRNA compared to WT mice, also displayed reduced B16 tumor growth and lung metastases. It is possible that the lower expression of IL-17 in mda-9 -/mouse lungs impairs infiltration of MDSCs, an inflammatory cell population with immunosuppressive activity. Although we did not see complete regression of tumors (at the primary or metastatic sites), kinetics of tumor development and pathological state of tumorbearing lungs correlated with the accumulation of these inflammatory and immunosuppressive cells. Finally, our hypothesis that mda-9/syntenin expression in stroma is also important in tumor and metastatic development is validated in a spontaneous melanoma metastasis model. Additional studies will be needed to further understand the cross-talk of tumor cells (either expressing MDA-9/ Syntenin or not) within an mda-9/syntenin deficient tumor microenvironment, which will provide further insight into the multiple levels of involvement and functions of this gene in this complex process.
The present study is the first demonstration of the relevance of MDA-9/Syntenin expression in regulating inflammation and immunosuppression in the microenvironment thereby impacting tumor growth and metastasis. A schematic model of the role of MDA-9/ Syntenin in the tumor cell and microenvironment on metastasis is shown in Figure 6. Taking into account our previous research (7-11, 31,45,52,53) the present study also supports the newer concept that MDA-9/ Syntenin is a key protein product influencing metastasis by regulating both tumor cells and the microenvironment. Accordingly, this protein is an attractive therapeutic target for inhibiting metastasis. In principle, mda-9/syntenin could be targeted for inhibition genetically, using shRNA or siRNA delivered by nanoparticles or other approaches. For this strategy to be successful it would be necessary for the inhibitory RNAs to be delivered to metastatic cells in diverse and sometimes difficult to access regions of the body and to efficiently reduce expression to a level that would be therapeutically efficacious. Since a key mechanism by which MDA-9/Syntenin regulates cellular phenotypes is through protein-protein interactions through its PDZ domains, identifying small molecule inhibitors that specifically target the PDZ domains thereby disrupting the interactions and altering down stream signaling critical Oncotarget 46857 www.impactjournals.com/oncotarget for activity would be of significant value. We are currently exploring both of these strategies to capitalize on the potential importance of mda-9/syntenin as a regulator of cancer invasion and metastasis in melanoma and other cancers. Finally, the mda-9 -/mouse models we have created will pave the way in comprehending the role of the microenvironment in the development and therapy of melanoma and potentially additional cancers.

mda-9/syntenin knockout mice
To investigate the physiological role of mda-9/ syntenin in metastasis development we generated a line of global mda-9/syntenin knockout mice (mda-9 -/-), starting with a line of mda-9 "knockout first" mice (Sdcbp tm1a(KOMP) Wtsi ) obtained from the Knockout Mouse Project (KOMP) Repository [72], which were bred first to β-actin-FLPe mice (JAX stock #005703) to generate the conditional knockout (CKO) allele in which exon 3 is floxed. These mice were then bred to EIIa-cre mice (JAX stock #003724) to delete exon 3, which results in a translational frameshift, leaving only 17 residues of the 299-residue MDA-9/ Syntenin protein. We confirmed the absence of MDA-9/ Syntenin protein in multiple tissues (lung shown) from mice homozygous for the knockout allele as well as the original "knockout first" allele, and retention of expression in mice with the floxed allele ( Figure 1A).

Compound Braf V600E /Pten fl/fl /Tyr-cre ER2/ mda-9 -/mice
The BRaf V600E /Pten fl/fl /Tyr-cre ER2 tamoxifen-inducible genetically engineered mouse model of metastatic melanoma (51) accurately recapitulates the full process of tumor initiation, progression and metastasis. This model was crossed with mda-9 -/mice to develop BRaf V600E /Pten fl/ fl /Tyr-cre ER2 /mda-9 -/mice. Production of the BRaf V600E / Pten fl/fl /Tyr-cre ER2 and BRaf V600E /Pten fl/fl /Tyr-cre ER2 /mda-9 -/mice were particularly challenging because of the large number of genetically modified alleles being bred into each mouse. To facilitate efficient production of these mice, we generated sub-colonies of optimal breeders. For example, to generate BRaf V600E /Pten fl/fl /Tyr-cre ER2 mice of the desired genotype (heterozygous for a modified Braf allele that expresses WT Braf until cre is expressed, and is then converted to the mutant BRaf V600E allele; homozygous for a Pten floxed allele, and hemizygous for the Tyr-cre ER2 transgene), we generated sub-colonies of: i) BRaf V600E / Pten fl/fl /Tyr-cre ER2 and ii) Braf fl/fl / Pten fl/fl (cre-) mice. When mice of these two genotypes were interbred, 50% of the offspring were of the experimental genotype, while the remaining 50% were cre-controls. Similarly, we also generated the following parental sub-colonies to develop the mda-9 -/mice: iii) Braf WT/WT / Pten fl/fl / Tyr-cre ER2 / mda-9 -/and iv) Braf fl/fl / Pten fl/fl / mda-9 -/-(cre-) mice.

Western blotting analysis
Freshly isolated organs (e.g., Lung) were minced with sterile scalpels and incubated with cell lysis buffer for 30 minutes on ice followed by homogenization using a glass homogenizer. All steps were performed on ice. Protein lysates were collected after centrifugation. For western blotting, we used MDA-9/syntenin (1:50, Cat.

RNA extraction, cDNA preparation, and real-time Pcr
Total RNA was extracted from different mouse organs as indicated using QIAGEN miRNAeasy Mini Kit (QIAGEN). After confirming the quality of RNA, ABI cDNA Synthesis Kit (Applied Biosystems, Carlsbad, CA) was used to prepare cDNA. Real time PCR was performed using an ABI ViiA7 fast real-time PCR system and TaqMan gene expression assays according to the manufacturer's protocol. All relevant probes (mouse specific Taqman probe for IL-17A, IL-6, TNF-α, GAPDH) were purchased from Applied Biosystems (Carlsbad, CA).

Xenograft and experimental metastasis
For xenograft experiments, 2 X 10 5 B16 cells were subcutaneously injected in the right dorsal flank of mice as previously described [73]. Tumor volume was measured twice weekly with a caliper and calculated using formulas π/6 × larger diameter × (smaller diameter) 2 [74]. Experimental lung metastasis was performed as previously described [73]. Briefly, B16 cells (1 X 10 5 cells per mouse) suspended in 100 µL PBS (without Ca 2+ and Mg 2+ ) were injected intravenously to generate pulmonary metastases. Based on the experimental design, mice were sacrificed at different time points.

Routine pathological staining
Tumors or organs were fixed in formalin. Samples were then embedded in paraffin, sectioned, and haematoxylin/eosin stained by the Virginia Commonwealth University Anatomic Pathology Laboratory.

Statistical analysis
Data are shown as the average ± S.D. Unpaired twotailed Student t-tests were conducted for calculating the "p" value between two groups. A "p" value <0.05 was considered statistically significant.

cONFLIcts OF INtErEst
The authors declare no conflict of interest.