Advertisement

CXCL5 Facilitates Melanoma Cell–Neutrophil Interaction and Lymph Node Metastasis

Open ArchivePublished:February 20, 2018DOI:https://doi.org/10.1016/j.jid.2018.01.035
      Chemokines influence tumor metastasis by targeting tumor, stromal, and hematopoietic cells. Characterizing the chemokine mRNA expression profile of human primary melanoma samples, we found CXCL5 significantly up-regulated in stage T4 primary melanomas when compared to thin melanomas (T1 stage). To characterize the role of CXCL5 in melanoma progression, we established a metastasizing murine xenograft model using CXCL5-overexpressing human melanoma cells. CXCL5 had no effect on melanoma proliferation in vitro and on primary tumor growth in vivo, but CXCL5-overexpressing tumors recruited high amounts of neutrophils and exhibited significantly increased lymphangiogenesis in our severe combined immune-deficient mouse model. Recruited neutrophils were found in close proximity to or within lymphatic vessels, often in direct contact with melanoma cells. Clinically, CXCL5-overexpressing melanomas had significantly increased lymph node metastases. We were able to translate these findings to human patient samples and found a positive correlation between CXCL5 expression, numbers of neutrophils in stage T4 primary melanoma, and the occurrence of subsequent locoregional metastasis.

      Abbreviation:

      SCID (severe combined immune-deficient)

      Introduction

      Chemokines, or chemotactic cytokines, are a family of more than 50 small peptides and 20 receptors. They primarily regulate trafficking and positioning of cells by activating the seven-transmembrane spanning G protein–coupled chemokine receptors. Depending on the position of the first two conserved cysteine residues in the extracellular N-terminal part, chemokines are classified as either CC, CXC, CX3C, or C chemokines (
      • Baggiolini M.
      Chemokines and leukocyte traffic.
      ,
      • Le Y.
      • Zhou Y.
      • Iribarren P.
      • Wang J.M.
      Chemokines and chemokine receptors: their manifold roles in homeostasis and disease.
      ). They usually bind to their specific signaling receptors, but in order to maintain chemokine gradients in tissues, scavenger receptors, or so-called atypical chemokine receptors, exist (
      • Ulvmar M.H.
      • Hub E.
      • Rot A.
      Atypical chemokine receptors.
      ). Besides their main function in immune cell trafficking in steady-state tissue homeostasis or inflammation, chemokines play an important role in cancer biology: primary tumor-derived chemokines are able to remodel the cellular composition of the local milieu by attraction of leukocytes, endothelial cells, or fibroblasts (
      • Balkwill F.
      Cancer and the chemokine network.
      ,
      • Chow M.T.
      • Luster A.D.
      Chemokines in cancer.
      ,
      • Zlotnik A.
      Chemokines and cancer.
      ). Additionally, CXC chemokines that carry the proangiogenic ELR-motif (glu-leu-arg) N-terminal to the first cysteine residue in their sequence exhibit proangiogenic functions (
      • Belperio J.A.
      • Keane M.P.
      • Ehlert L.E.
      • Arenberg D.A.
      • Burdick M.D.
      • Strieter R.M.
      CXC chemokines in angiogenesis.
      ). Chemokines alter the cellular composition and activate specific signaling pathways that affect epithelial–mesenchymal transition, activation of the vasculature, and break down of the extracellular matrix, events known to initiate metastasis (
      • Nagarsheth N.
      • Wicha M.S.
      • Zou W.
      Chemokines in the cancer microenvironment and their relevance in cancer immunotherapy.
      ). Single chemokines have been analyzed extensively with regard to their respective tumorigenic effects, but only a few data exist to understand the milieu-specific parallel and sequential crosstalk of different chemokines.
      Metastasis is a complex process of sequential events and starts with the dissemination of primary tumor cells into the local environment. Tumor cells invade into blood or lymphatic vessels and travel through the circulation before they cross the endothelial barrier at distant sites to form daughter clones in a new environment. Selection of metastatic sites does not happen randomly, chemokine/chemokine receptor axes do exist for guidance (
      • Chambers A.F.
      • Groom A.C.
      • MacDonald I.C.
      Dissemination and growth of cancer cells in metastatic sites.
      ,
      • Vanharanta S.
      • Massague J.
      Origins of metastatic traits.
      ). CXCL12/CXCR4 is the most prominent chemokine/chemokine receptor pair in cancer biology that guides CXCR4+ metastatic tumor cells to lungs, liver, or bone marrow, which are natural sources of CXCL12 (
      • Chatterjee S.
      • Behnam Azad B.
      • Nimmagadda S.
      The intricate role of CXCR4 in cancer.
      ). Another example is the CCR7/CCL19,CCL21 axis that guides tumor cells into the lymphatic vasculature and facilitates lymph node metastasis (
      • Kochetkova M.
      • Kumar S.
      • McColl S.R.
      Chemokine receptors CXCR4 and CCR7 promote metastasis by preventing anoikis in cancer cells.
      ,
      • Legler D.F.
      • Uetz-von Allmen E.
      • Hauser M.A.
      CCR7: roles in cancer cell dissemination, migration and metastasis formation.
      ). Tumor cells that express CCR9 are able to recognize CCL25 and, therefore, can be targeted to the small bowel. Interestingly, patients with CCR9-expressing melanomas have a >60% chance of small bowel metastases developing (
      • Amersi F.F.
      • Terando A.M.
      • Goto Y.
      • Scolyer R.A.
      • Thompson J.F.
      • Tran A.N.
      • et al.
      Activation of CCR9/CCL25 in cutaneous melanoma mediates preferential metastasis to the small intestine.
      ).
      Melanoma, a highly malignant skin tumor, primarily metastasizes via the lymphatic system to develop lymph node metastases (
      • Meier F.
      • Will S.
      • Ellwanger U.
      • Schlagenhauff B.
      • Schittek B.
      • Rassner G.
      • et al.
      Metastatic pathways and time courses in the orderly progression of cutaneous melanoma.
      ). Melanomas are able to induce lymphangiogenesis both in vivo and in vitro (
      • Shields J.D.
      • Borsetti M.
      • Rigby H.
      • Harper S.J.
      • Mortimer P.S.
      • Levick J.R.
      • et al.
      Lymphatic density and metastatic spread in human malignant melanoma.
      ), and the number of lymphatic vessels has been shown to correlate with the metastatic risk (
      • Pastushenko I.
      • Vermeulen P.B.
      • Carapeto F.J.
      • Van den Eynden G.
      • Rutten A.
      • Ara M.
      • et al.
      Blood microvessel density, lymphatic microvessel density and lymphatic invasion in predicting melanoma metastases: systematic review and meta-analysis.
      ). Successful lymphatic metastasis requires successful tumor cell migration into lymphatic vessels. Primary melanomas express a wide variety of chemokines and chemokine receptors that are able to induce angiogenesis and attract immune cells (
      • Richmond A.
      • Yang J.
      • Su Y.
      The good and the bad of chemokines/chemokine receptors in melanoma.
      ), but only a few of them, for example, CXCL8 (IL-8) (
      • Singh R.K.
      • Varney M.L.
      IL-8 expression in malignant melanoma: implications in growth and metastasis.
      ), CXCL1 (
      • Dhawan P.
      • Richmond A.
      Role of CXCL1 in tumorigenesis of melanoma.
      ), and CXCL10 (
      • Antonicelli F.
      • Lorin J.
      • Kurdykowski S.
      • Gangloff S.C.
      • Le Naour R.
      • Sallenave J.M.
      • et al.
      CXCL10 reduces melanoma proliferation and invasiveness in vitro and in vivo.
      ), have been studied in more detail. Chemokine-induced alterations in the cellular composition of the local tumor milieu influence tumor cell invasion and transmigration into the vasculature. Resident and circulating immune cells, foremost tumor-associated macrophages, tumor-associated neutrophils, or regulatory T cells, have been reported to suppress antitumor immunity and actively assist the metastatic process (
      • Kitamura T.
      • Qian B.Z.
      • Pollard J.W.
      Immune cell promotion of metastasis.
      ).
      In our study, we identified a pattern of 12 chemokines that are differentially regulated in melanocytic nevi and melanomas of different tumor stages. CXCL5, a potent neutrophil attractor, was up-regulated most strongly in thick melanomas. We also identified a lymphangiogenic potential of CXCL5 that ultimately led to increased lymph node metastasis.

      Results

      Chemokine Expression Differs among Human Melanoma Tumor Stages

      In a severe combined immune-deficient (SCID) hairless outbred mouse xenotransplantation model, we identified a panel of 12 chemokines strongly expressed in primary tumors and lymph node metastases (Table 1). The relevance of this panel was confirmed by the presence of these chemokines in different primary melanoma cell lines, as well as in human fibroblasts and endothelial cells (Supplementary Figure S1 online). In samples of human nevi and cutaneous melanomas (T1 to T4), all chemokines were differentially expressed and displayed gradual expression changes according to the tumor stage (Figure 1a). CXCL5 was the most up-regulated chemokine between T1 and T4 melanomas. Interestingly, analysis of a human melanoma cell line isolated from a lymph node metastasis (DLN) showed a 60-fold increase in CXCL5 expression when compared to the parental cell line (MCMD1), which was engrafted into SCID mice (Figure 1b) (
      • Swoboda A.
      • Schanab O.
      • Tauber S.
      • Bilban M.
      • Berger W.
      • Petzelbauer P.
      • et al.
      MET expression in melanoma correlates with a lymphangiogenic phenotype.
      ). CXCL5 protein levels, as determined by immunohistochemistry on primary melanoma sections, also correlated with mRNA expression levels (data not shown).
      Table 1Chemokine expression profiling by Affymetrix Genome Array: strongest expressed chemokines in M24 primary tumors and lymph node metastases
      SkinLymph Node
      CCL6CCL5
      CCL8CCL6
      CCL21aCCL8
      CXCL1CCL19
      CXCL2CCL21a
      CXCL5CCL22
      CXCL14CXCL12
      CXCR13
      Figure 1
      Figure 1CXCL5 is strongest up-regulated in high risk human cutaneous melanoma. (a) Fold change (FC) expression in primary melanomas (T1, n = 18; T2, n = 21; T3, n = 19; T4, n = 21) and nevi (N, n = 11), normalized to normal skin (n = 8). P values shown for differences between T1 and T4 melanomas. (b) FC of CXCL5 mRNA levels of lymph node metastasis normalized to primary tumor. (c) CXCR2 mRNA expression (ΔCt) in T1 (n = 7), T2 (n = 7), T3 (n = 7) and T4 (n = 13) melanomas. (d) Left: Positive correlation of ΔCt CXCL5 to CXCR2 mRNA expression in T4 melanomas (n = 13). Right: FC of CXCR2 in CXCL5high (n = 6) compared to CXCL5low (n = 7) T4 melanomas. (e) Representative immunohistochemical stainings of CXCR2 in healthy skin and primary melanoma. Insert: IgG2a isotype control. Scale bar = 200 μm. Mean ± standard error of mean (c, d); one-way analysis of variance, Tukey post-hoc test or two-tailed t test (c, d). P < 0.05.
      To identify possible targets for CXCL5, we analyzed mRNA expression levels of CXCR2 and DARC in human primary melanoma samples; whereas CXCR2 is the signaling receptor, DARC is a non-signaling scavenger receptor. CXCR2 mRNA expression was similar in all tumor stages (T1 to T4; Figure 1c). Interestingly, CXCL5 mRNA expression in T4 melanomas varied much more than in other tumor stages, pointing toward the presence of two subpopulations. T4 melanomas could be separated into a CXCL5high- and a CXCL5low-expressing population. In these two populations, CXCR2 mRNA expression levels correlated positively with CXCL5 mRNA expression levels (Figure 1d). Using immunohistochemistry, CXCR2 was detected on melanoma cells and on keratinocytes (Figure 1e). In contrast, DARC mRNA expression decreased significantly with tumor thickness, however, it was still abundantly present on cutaneous endothelial cells of thick primary melanomas, as detected by immunohistochemistry (Supplementary Figure S2 online).

      CXCL5 Increases Lymph Node and Lung Metastasis In Vivo and In Vitro

      To address the role of CXCL5 in the progression of primary melanomas, two CXCL5 low-expressing human melanoma cell lines, M24met and A375, were stably transfected with human full-length CXCL5 (M24 CXCL5ox, A375 CXCL5ox) or empty vector (M24pBMN, A375pBMN) (Supplementary Figure S3 online). Expression profiles of chemokines and angiogenic factors remained unchanged in these cell lines after transfection (data not shown). CXCL5ox did not alter cell proliferation or wound closure kinetics in vitro (Figure 2a). To study in vivo effects of CXCL5 on metastatic behavior, M24 CXCL5ox or A375 CXCL5ox cells or their respective control cell lines were injected intracutaneously into the right flank of SCID hairless outbred mice. Tumors were excised at a volume of 300 mm3 without significant differences in primary tumor growth (Figure 2b). At this time point, CXCL5ox tumors led to detectable CXCL5 serum levels, whereas animals with control tumors had no detectable CXCL5 serum levels (Figure 2c). Spontaneous metastases were evaluated on tissue sections of locoregional lymph nodes 14 days (M24 tumors) or 21 days (A375 tumors) after primary tumor resection, stained with an anti-human VIMENTIN antibody (Figure 2d). In both cell lines, CXCL5ox led to a significant increase in lymph node metastasis (64% M24 CXCL5ox or 56% A375 CXCL5ox macrometastasis compared to 14.3% or 25% of macrometastases in the respective control groups) (Figure 2d). Neutrophil depletion with anti-LY6G antibody treatment led to a decrease in lymph node metastasis (36.4% M24 CXCL5ox anti-LY6G or 33.3% A375 CXCL5ox anti-LY6G macrometastasis compared to 0% or 7.7% of macrometastasis in the respective isotype-treated control groups) (Supplementary Figure S4 online). The average weight of M24 CXCL5ox lymph nodes was 191.9 ± 80.9 mg compared to 49.8 ± 21.3 mg in the M24ctrl group. A375 CXCL5ox lymph nodes had an average weight of 240.8 ± 126.0 mg compared to 80.5 ± 29.3 mg in the control group (Figure 2e).
      Figure 2
      Figure 2CXCL5 expression increases lymph node metastasis in a mouse xenograft model. (a) Left: Proliferation assay of M24 and A375 in vitro. Right: Representative pictures of an in vitro scratch assay for A375. (b) Growth curves of M24 and A375 xenografts in severe combined immune-deficient mice (n = 5 per group, two independent experiments). (c) Enzyme-linked immunosorbent assay of CXCL5 serum levels in mice, carrying either M24 control (n = 2) or M24 CXCL5ox (n = 2) primary tumors. (d) Frequencies of lymph node metastases in control and CXCL5ox mice. Representative staining of anti-human VIMENTIN-stained lymph node micrometastasis. Scale bar = 100 μm. (e) Weight (in milligrams) of tumor draining lymph nodes. M24 control, n = 14, M24 CXCL5ox, n = 14, A375 control, n = 12, A375 CXCL5ox, n = 14, three independent experiments (d, e). Mean ± standard error of mean (a, b, c, e); statistical analysis: two-tailed t test (c, e), χ2 test (d). P < 0.05.

      CXCL5 Overexpression Leads to Influx of Neutrophils into Primary Tumors and Intratumoral Lymph Vessels

      CXCL5 is a known neutrophil chemoattractant (
      • Walz A.
      • Burgener R.
      • Car B.
      • Baggiolini M.
      • Kunkel S.L.
      • Strieter R.M.
      Structure and neutrophlil-activating properties of a novel inflammatory peptide (ENA-78) with homology to interleukin 8.
      ). We detected neutrophils in primary tumors and lymph nodes. In CXCL5ox tumors, neutrophils accounted for up to 2% of total tumor area, as assessed by immunohistochemistry with an anti-LY6G antibody. In control groups, <0.1% of total tumor area was LY6G-positive (Figure 3a). Neutrophils were not arbitrarily distributed; >50% of tumor-infiltrating neutrophils were located in close association with lymphatic vessels (Figure 3b). Importantly, we found neutrophils alone or attached to tumor cells in lymphatic vessels draining primary tumor sites and in the marginal sinus of lymph nodes (Figure 3c). Only locoregional lymph nodes did harbor detectable numbers of neutrophils, other lymph node locations were neutrophil negative, thus excluding a general neutrophil-recruiting effect by high CXCL5 serum levels in animals with CXCL5-overexpressing tumors (data not shown).
      Figure 3
      Figure 3Neutrophils support tumor cells in lymphatic traveling. (a) Neutrophils per tumor area (n = 5, two independent experiments). Representative immunohistochemical images of anti-LY6G–stained neutrophils in sections of CXCL5ox and control xenografts; Scale bar = 100 μm. (b) Distance: neutrophils to lymphatic vessels (LVs) in CXCL5ox mice (≤10, 11–20, 21–200 μm, n = 141, three tumor samples). Immunofluorescent image of a neutrophil in close proximity to a LV in a xenograft section; Scale bar = 20 μm. (c) Immunofluorescent images of primary xenografts and draining lymph node. Top: Neutrophils in close proximity to LVs (left); neutrophils attached to a melanoma cell inside a LV (right); Bottom: Melanoma cells alone and together with neutrophils inside LVs (left); lymph node sinus (white line) with infiltrating neutrophils and melanoma cells (right). Scale bar = 20 μm. Mean ± standard error of mean (a, b); two-tailed t test (a), one-way analysis of variance (b). P < 0.05.

      CXCL5 Supports Lymphangiogenesis and Tumor Cell Migration across Lymphatic Endothelial Cells

      Quantification of lymphatic vessels density by immunohistochemistry revealed a significant increase in LYVE-1–positive lymphatic vessels (CXCL5ox tumors: 7.9% of total tumor area vs. 4.7% in control groups) but no significant difference in CD31-positive blood vessel count (Figure 4a). This observation suggests a previously unknown lymphangiogenic effect of CXCL5. FACS analysis and immunofluorescent stainings revealed the expression of CXCR2 on human lymphatic endothelial cells in vitro (Figure 4b, Supplementary Figure S5a online). Additionally, human recombinant CXCL5 induced dose-dependently sprouts in immortalized human lymphatic endothelial cells in a three-dimensional spheroid sprouting assay. Recombinant human CXCL5 was able to induce cumulative sprout lengths at a similar extent as human recombinant VEGFC. Treatment with anti-CXCR2 blocking antibody in parallel to CXCL5 stimulation inhibited sprouting (Figure 4c, Supplementary Figure S5b). Lymphatic vessels in human skin showed expression of CXCR2, as determined by immunohistochemistry, therefore, CXCL5 could exert its pro-lymphangiogenic effect also in vivo (Supplementary Figure S5c). Because of the close association of neutrophils with tumor cells in lymphatic vessels, we explored the idea that neutrophils assist in tumor cell intravasation into lymphatic vessels. Indeed, when neutrophils were added to tumor cells, reverse transendothelial migration across lymphatic endothelial cells from the basal to the luminal side in vitro was increased by 3.5-fold compared to transmigration of tumor cells alone (Figure 4d).
      Figure 4
      Figure 4Lymphangiogenic potential of CXCL5. (a) Percentages of LYVE-1 and CD31-positive stained areas in sections of primary CXCL5ox and control tumors (n = 6, two independent experiments). Scale bar = 100 μm. (b) FACS histogram of lymphatic endothelial cells stained against CXCR2. One representative out of three independent experiments. (c) Cumulative length (×103 μm) of sprouts from immortalized human lymphatic endothelial cells spheroids stimulated with rCXCL5 ± anti-CXCR2, VEGFC, or bovine serum albumin. One representative out of three independent experiments. Representative photographs of sprouting spheroids stimulated with rCXCL5 or VEGFC. (d) Total numbers of A375 melanoma cells transmigrated across a fibronectin-coated lymphatic endothelial cell monolayer in the presence or absence of neutrophils. One representative out of two independent experiments. Statistical analysis by two-tailed t test (a, d); mean ± standard error of mean (a, d). P < 0.05.

      Relevance of CXCL5 in Human Melanoma

      Similarly, in human patient samples, CXCL5-expressing melanomas had a higher amount of infiltrating neutrophils when compared to primary melanomas with weak, or without, expression of CXCL5 (Figure 5a). In human melanoma samples, there was a positive correlation between the presence of ulceration and the number of neutrophils (Figure 5b). Higher levels of CXCL5 in primary melanoma samples correlated with increased likelihood of locoregional metastases (Figure 5c), identifying CXCL5 as a possible prognostic marker in primary human melanoma.
      Figure 5
      Figure 5CXCL5-expressing human primary melanomas have a higher risk for metastasis. (a) CD66b+ neutrophils per mm2 in CXCL5-positive (n = 27) or -negative (n = 13) histological sections of human melanoma samples. Representative images of anti-CD66b–stained neutrophils on human melanomas, classified as CXCL5high or CXCL5low. Scale bar = 100 μm. (b) CD66b+ Neutrophils per mm2 in histological sections of human melanoma samples graded in ulcerated (n = 27) or not ulcerated (n = 36). (c) Log2 fold-change expression of CXCL5 (mRNA) in thick human primary melanomas (tumor stage T3/T4) of patients with (n = 14) or without (n = 7) metastases. Statistical analysis by two-tailed t test and data are represented as box plot with individual data points. P < 0.05.

      Discussion

      Melanomas metastasize predominantly via the lymphatic system. A positive correlation between lymphangiogenesis and metastases has been established (
      • Shields J.D.
      • Borsetti M.
      • Rigby H.
      • Harper S.J.
      • Mortimer P.S.
      • Levick J.R.
      • et al.
      Lymphatic density and metastatic spread in human malignant melanoma.
      ). It is currently thought that melanomas metastasize to the locoregional lymph node basin after active tumor cell migration into dermal lymphatic vessels, followed by passive flow to the lymph nodes. Several published mouse melanoma models show different mechanisms that support the formation of lymph node metastasis: CCL19 and CCL21, chemokines constitutively expressed by lymphatic endothelial cells, guide CCR7-positive melanoma cells toward the lymph node (
      • Wiley H.E.
      • Gonzalez E.B.
      • Maki W.
      • Wu M.T.
      • Hwang S.T.
      Expression of CC chemokine receptor-7 and regional lymph node metastasis of B16 murine melanoma.
      ). CCR8 expression showed to be a requirement for melanoma cell entry into the CCL1-expressing lymph node sinus (
      • Das S.
      • Sarrou E.
      • Podgrabinska S.
      • Cassella M.
      • Mungamuri S.K.
      • Feirt N.
      • et al.
      Tumor cell entry into the lymph node is controlled by CCL1 chemokine expressed by lymph node lymphatic sinuses.
      ). In addition, expression of α4 integrin increases the interaction between tumor cells and lymphatic endothelial cells and thereby affects the outcome in lymph node metastasis (
      • Rebhun R.B.
      • Cheng H.
      • Gershenwald J.E.
      • Fan D.
      • Fidler I.J.
      • Langley R.R.
      Constitutive expression of the β4 integrin correlates with tumorigenicity and lymph node metastasis of the B16 murine melanoma.
      ). So far, CXCL5 has been associated with various human cancers, but not directly linked to lymph node metastasis. CXCL5 acts proangiogenic in human non-small cell lung cancer and pancreatic cancer (
      • Arenberg D.A.
      • Keane M.P.
      • DiGiovine B.
      • Kunkel S.L.
      • Morris S.B.
      • Xue Y.Y.
      • et al.
      Epithelial-neutrophil activating peptide (ENA-78) is an important angiogenic factor in non-small cell lung cancer.
      ,
      • Li A.
      • King J.
      • Moro A.
      • Sugi M.D.
      • Dawson D.W.
      • Kaplan J.
      • et al.
      Overexpression of CXCL5 is associated with poor survival in patients with pancreatic cancer.
      ). In renal carcinoma, gastric cancer, and nasopharyngeal carcinomas, CXCL5 serum levels have predictive relevance for prognosis and progression (
      • Park J.Y.
      • Park K.H.
      • Bang S.
      • Kim M.H.
      • Lee J.E.
      • Gang J.
      • et al.
      CXCL5 overexpression is associated with late stage gastric cancer.
      ,
      • Zhang H.
      • Xia W.
      • Lu X.
      • Sun R.
      • Wang L.
      • Zheng L.
      • et al.
      A novel statistical prognostic score model that includes serum CXCL5 levels and clinical classification predicts risk of disease progression and survival of nasopharyngeal carcinoma patients.
      ). In addition, in intrahepatic cholangiocarcinoma and gastric cancer, CXCL5 serum levels and development of metastasis correlated positively (
      • Park J.Y.
      • Park K.H.
      • Bang S.
      • Kim M.H.
      • Lee J.E.
      • Gang J.
      • et al.
      CXCL5 overexpression is associated with late stage gastric cancer.
      ,
      • Zhou S.L.
      • Dai Z.
      • Zhou Z.J.
      • Chen Q.
      • Wang Z.
      • Xiao Y.S.
      • et al.
      CXCL5 contributes to tumor metastasis and recurrence of intrahepatic cholangiocarcinoma by recruiting infiltrative intratumoral neutrophils.
      ).
      Our data point to an even more complex situation, mediated and/or initiated by CXCL5. We identified a dual role of this chemokine in the mediation of lymphatic metastasis. First, CXCL5 acts directly mitogenic on lymphatic endothelial cells in vitro and, secondly, overexpression of tumor-derived CXCL5 increases the number of intra- and peritumoral neutrophils. Interestingly, although we were able to describe a pro-lymphangiogenic effect of CXCL5 in our model, we were not able to recapitulate the reported proangiogenic effect of CXCL5. This could be explained by an already high pro-angiogenic drive caused by other angiogenic factors (VEGFA, PDGF, as reported previously (
      • Homsi J.
      • Daud A.I.
      Spectrum of activity and mechanism of action of VEGF/PDGF inhibitors.
      ), which could not be further increased by CXCL5. Nevertheless, at least in our hand, recombinant CXCL5 stimulation of human endothelial cells (human umbilical vein endothelial cells) failed to induce angiogenesis (data not shown).
      Neutrophils were found in direct contact with melanoma cells in and around lymphatic vessels and they increased the number of reverse transmigrating melanoma cells throughout lymphatic endothelial cells in vitro. In melanoma, UV-induced neutrophils have been shown to stimulate angiogenesis and enhance melanoma migration toward blood endothelial cells (
      • Bald T.
      • Quast T.
      • Landsberg J.
      • Rogava M.
      • Glodde N.
      • Lopez-Ramos D.
      • et al.
      Ultraviolet-radiation-induced inflammation promotes angiotropism and metastasis in melanoma.
      ). During infections, neutrophils utilize lymphatics rather than blood vessels to migrate from skin sites of infections into draining lymph nodes. The fast re-localization from the inflammation to draining lymph nodes allows neutrophils to modulate the adaptive response of the lymph node prior to the activation of other innate immune cells (
      • Chtanova T.
      • Schaeffer M.
      • Han S.J.
      • van Dooren G.G.
      • Nollmann M.
      • Herzmark P.
      • et al.
      Dynamics of neutrophil migration in lymph nodes during infection.
      ,
      • Hampton H.R.
      • Bailey J.
      • Tomura M.
      • Brink R.
      • Chtanova T.
      Microbe-dependent lymphatic migration of neutrophils modulates lymphocyte proliferation in lymph nodes.
      ). Our data suggest that, also during metastasis, neutrophils use the tumor draining lymphatic vessels to the locoregional lymph nodes in order to facilitate tumor cell travel and seeding. The role of neutrophils in tumorigenesis is still controversial, different opposing roles have been identified (
      • Chao T.
      • Furth E.E.
      • Vonderheide R.H.
      CXCR2-dependent accumulation of tumor-associated neutrophils regulates T-cell immunity in pancreatic ductal adenocarcinoma.
      ,
      • Granot Z.
      • Henke E.
      • Comen E.A.
      • King T.A.
      • Norton L.
      • Benezra R.
      Tumor entrained neutrophils inhibit seeding in the premetastatic lung.
      ,
      • Lopez-Lago M.A.
      • Posner S.
      • Thodima V.J.
      • Molina A.M.
      • Motzer R.J.
      • Chaganti R.S.
      Neutrophil chemokines secreted by tumor cells mount a lung antimetastatic response during renal cell carcinoma progression.
      ,
      • Spicer J.D.
      • McDonald B.
      • Cools-Lartigue J.J.
      • Chow S.C.
      • Giannias B.
      • Kubes P.
      • et al.
      Neutrophils promote liver metastasis via Mac-1-mediated interactions with circulating tumor cells.
      ). Reduction of CXCL5 expression in squamous cell carcinoma and bladder cancer T24 cells decreased proliferation and impaired migration of tumor cells (
      • Miyazaki H.
      • Patel V.
      • Wang H.
      • Edmunds R.K.
      • Gutkind J.S.
      • Yeudall W.A.
      Down-regulation of CXCL5 inhibits squamous carcinogenesis.
      ,
      • Zheng J.
      • Zhu X.
      • Zhang J.
      CXCL5 knockdown expression inhibits human bladder cancer T24 cells proliferation and migration.
      ). Neutrophils, together with elevated acute-phase proteins (e.g., CRP) or hypoalbuminemia are signs of a systemic inflammatory reaction that may predict poor outcomes in cancer patients (
      • Jamieson N.B.
      • Glen P.
      • McMillan D.C.
      • McKay C.J.
      • Foulis A.K.
      • Carter R.
      • et al.
      Systemic inflammatory response predicts outcome in patients undergoing resection for ductal adenocarcinoma head of pancreas.
      ,
      • Kennelly R.P.
      • Murphy B.
      • Larkin J.O.
      • Mehigan B.J.
      • McCormick P.H.
      Activated systemic inflammatory response at diagnosis reduces lymph node count in colonic carcinoma.
      ,
      • Wang J.
      • Kalhor N.
      • Hu J.
      • Wang B.
      • Chu H.
      • Zhang B.
      • et al.
      Pretreatment neutrophil to lymphocyte ratio is associated with poor survival in patients with stage I-III non-small cell lung cancer.
      ). A high absolute neutrophil count or an increased neutrophil to lymphocyte ratio was linked to poor overall survival of cancer patients in phase 1 clinical trials. An increased neutrophil to lymphocyte ratio has been regarded to mirror sustained angiogenesis and proliferation of tumor cells (
      • Kumar R.
      • Geuna E.
      • Michalarea V.
      • Guardascione M.
      • Naumann U.
      • Lorente D.
      • et al.
      The neutrophil-lymphocyte ratio and its utilisation for the management of cancer patients in early clinical trials.
      ).
      In summary, we identified a closed communication circuit among melanoma cells, neutrophils, and lymphatic endothelial cells, regulated by CXCL5. This interaction controls and facilitates the transfer of tumor cells from primary tumor sites into locoregional lymph nodes. The relevance of these findings was established in murine melanoma models, but importantly, the same regulatory circuit was also found in human melanoma patients under normal expression levels of CXCL5. Currently, the search for serum-detectable biomarkers has been strongly intensified because of the need for patient stratification with regard to the newly available treatment strategies (e.g., immune therapy). We also demonstrate that locally situated, primary tumor-derived factors might serve as relevant biomarkers, although further research clearly is needed.

      Materials and Methods

      Generation of Stable Cell Lines

      A375 and M24met human melanoma cells were infected with virus-containing supernatants produced from Phoenix cells (provided by H. Stockinger, Medical University Vienna, Austria) containing either retroviral plasmid pBMN-I-GFP (gift from Michael Detmar, ETH Zurich, Switzerland) or pBMN-I-GFP with the human CXCL5 gene (NM_002994.3, gift from Andrew Yeudall, Augusta University, GA). GFP-positive cells were selected by FACS.

      Cell Culture

      Human melanoma cell line A375 (LGC, Teddington, UK) and the stably transfected A375 were cultivated in DMEM, supplemented with 10% fetal calf serum and 2 mM l-glutamine (Life Technologies, Carlsbad, CA). Patient-derived melanoma cell lines (gift from Walter Berger, Medical University Vienna [
      • Pirker C.
      • Holzmann K.
      • Spiegl-Kreinecker S.
      • Elbling L.
      • Thallinger C.
      • Pehamberger H.
      • et al.
      Chromosomal imbalances in primary and metastatic melanomas: over-representation of essential telomerase genes.
      ]), M24met melanoma cells (kindly provided by RA Reisfeld [
      • Mueller B.M.
      • Romerdahl C.A.
      • Trent J.M.
      • Reisfeld R.A.
      Suppression of spontaneous melanoma metastasis in scid mice with an antibody to the epidermal growth factor receptor.
      ]), and the stably transfected M24met were grown in RPMI-1640 (Life Technologies) with identical supplements. Lymphatic endothelial cells, isolated from human foreskin (
      • Gröger M.
      • Loewe R.
      • Holnthoner W.
      • Embacher R.
      • Pillinger M.
      • Herron G.S.
      • et al.
      IL-3 Induces expression of lymphatic markers Prox-1 and Podoplanin in human endothelial cells.
      ), were lentivirally infected with the human telomerase gene to extend their lifespan (immortalized human lymphatic endothelial cells) and cultivated in Iscove’s modified Dulbecco’s medium (Lonza, Basel, Switzerland), supplemented with 10% fetal calf serum, 2 mM l-glutamine (Life Technologies), and 0.4% endothelial cell growth supplement/heparin (PromoCell, Heidelberg, Germany). Human umbilical vein endothelial cells were isolated from human umbilical cords and propagated as described previously (
      • Loewe R.
      • Holnthoner W.
      • Gröger M.
      • Pillinger M.
      • Gruber F.
      • Mechtcheriakova D.
      • et al.
      Dimethylfumarate inhibits TNF-induced nuclear entry of NF-kB/p65 in human endothelial cells.
      ). The use of human umbilical cords for the isolation of human umbilical vein endothelial cells has been approved by the Ethics Commission of the Medical University of Vienna.

      In Vitro Cell Proliferation and Wounding Assay

      Cell proliferation was determined with the Picogreen kit (Life Technologies) according to manufacturer’s instructions. For wounding assays, cell lines were plated in triplicate and cultivated until confluency. The monolayer was disrupted using a 200-μl pipet tip and scratches were documented every 24 hours using a Zeiss AxioCam ICc3 and Zeiss Axio Vision Rel.4.8 camera (Zeiss, Jena, Germany).

      SCID Hairless Outbred Xenotransplantation Experiments

      Animal procedures were approved by the animal care and use committee of the Medical University of Vienna and carried out according to the Association for Assessment and Accreditation of Laboratory Animal Care guidelines and the Guide for the Care and Use of Laboratory Animals (
      National Institutes of Health
      Guide for the Care and Use of Laboratory Animals (publication no. 86–23).
      ). M24met and A375 (2 × 106 cells) were injected intracutaneously into the right flank of 6- to 8-week-old pathogen-free, female SCID hairless outbred mice (Charles River, Sulzfeld, Germany) (
      • Loewe R.
      • Valero T.
      • Kremling S.
      • Pratscher B.
      • Kunstfeld R.
      • Pehamberger H.
      • et al.
      Dimethylfumarate impairs melanoma growth and metastasis.
      ). Tumor sizes were measured using a caliper and calculated using the following formula: V = (π/6) × (length) × (width)2. Tumors were excised after reaching a volume of 300 mm3 and defects were sutured. Ten days (M24met) or 21 days (A375) after tumor resection, animals were sacrificed and lymph nodes were removed. Tumors and lymph nodes were fixed in 4% paraformaldehyde and embedded in paraffin. Lymph nodes were completely cut in 4-μm sections and stained with anti-human VIMENTIN Clone V9 (M0725; Dako, Glostrup, Denmark). Micrometastasis is considered preservation of the lymph node capsule and presence of vital lymph node tissue. Macrometastasis is considered no preservation of the lymph node capsule. For depletion experiments, animals were intraperitoneally treated with 100 μg rat IgG2a, κ isotype, or rat anti-mouse Ly6G clone 1A8 (#BE0075-1; BioXCell, West Lebanon, NH) 24 hours prior to CXCL5ox tumor cell injection. Antibody treatment was repeated every 48 hours until tumors were resected at a volume of 300 mm3. The effectiveness of depletion was monitored by regular FACS analysis (Cytoflex, Beckman Coulter, Brea, CA) of peripheral blood and histological stainings of primary tumors.

      Enzyme-Linked Immunosorbent Assay

      Mouse serum was analyzed for human CXCL5 using the Quantikine ELISA Kit (Abcam, Cambridge, MA) according to the manufacturer’s instructions.

      Chemokine Profile of Primary Tumors and Lymph Node Metastases

      We used previously published microarray data, deposited in the Gene Expression Omnibus database GSE26656 (Samples: GSM656300, GSM656301, GSM656302, GSM656303, GSM656304) and GSE30688 (Samples: GSM761206, GSM761207, GSM761208).

      RNA Extraction and Real-Time PCR

      RNA from cell culture or paraffin embedded tumor patient samples (approved by the institutional ethics committee) was extracted using the RNeasy Mini kit (Qiagen, Valencia, CA) or RNeasy FFPE kit (Qiagen). RNA was transcribed into cDNA with RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas, Vienna, Austria). Applied Biosystems Taqmans were used for detection of CCL5 (Hs00174575_m1), CCL8 (Hs99999026_m1), CCL19 (Hs00171149_m1), CCL21 (Hs00171076_m1), CCL22 (Hs99999075_m1), CCL23 (Hs00270756_m1), CCL27 (Hs00171157_m1), CXCL2 (Hs00236966_m1), CXCL5 (Hs00171085_m1), CXCL6 (Hs00237017_m1), CXCL12 (Hs00171022_m1), CXCL13 (Hs00757930_m1), CXCL14 (Hs01557413_m1), CXCR2 (Hs01011557_m1), DARC (Hs01011079_s1). Raw data were normalized to β2-microglobulin (cat# 4333766-1006021; Applied Biosystems, Foster City, CA). Reactions were carried out on Applied Biosystems StepOne Plus cycler and analyzed with StepOne Software, version 2.1. Analysis of relative gene expression was calculated according to the ΔΔCt method (
      • Livak K.J.
      • Schmittgen T.D.
      Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method.
      ).

      Three-Dimensional In Vitro Spheroid Sprouting Assay

      Spheroids of immortalized human lymphatic endothelial cells were produced using the hanging drop method as described previously (
      • Laib A.M.
      • Bartol A.
      • Alajati A.
      • Korff T.
      • Weber H.
      • Augustin H.G.
      Spheroid-based human endothelial cell microvessel formation in vivo.
      ). Spheroids were embedded into a collagen/methocel matrix and stimulated with recombinant human CXCL5 (#254-XB; R&D Minneapolis, MN) with or without mouse anti-human CXCR2 blocking antibody Clone 48311 (#MAB331; R&D), supernatants from VEGFC-overexpressing cells as positive control or Iscove’s modified Dulbecco’s medium/3% bovine serum albumin as negative control. After 6 hours, pictures were taken and sprouts were counted by two independent observers.

      Immunohistochemistry, Immunofluorescence, and Flow Cytometry

      Primary antibodies

      The following primary antibodies were used: mouse anti-human VIMENTIN Clone V9 (#M0725; Dako, Vienna, Austria), rabbit anti-mouse LYVE-1 (#DP3513; Acris, Herford, Germany), rabbit anti-mouse CD31 (#Rb-10333; Neomarkers, Fremont, CA), mouse anti-human CXCL5 Clone 33160 (#MAB254; R&D), rat anti-mouse LY6G Clone 1A8 (#551459; BD Pharmingen, San Diego, CA), mouse anti-human CD66b Clone G10F5 (#555723; BD Pharmingen), mouse anti-human CXCR2 Clone ab24963 (#ab24963; Abcam), and goat anti-human Prox-1 (#AF2727; R&D).
      Paraffin sections were deparaffinized followed by antigen retrieval by boiling in citrate buffer (Dako). Sections were blocked with phosphate-buffered saline with Tween 20/3% bovine serum albumin (Sigma Aldrich, St. Louis, MO) and incubated overnight with first antibody. Secondary antibodies were applied for 1 hour and counterstained with either hematoxylin or DAPI (Thermo Scientific, Vienna, Austria). Immunohistochemistry included: biotinylated secondary antibodies (Vector Laboratories, Burlingame, CA) and horseradish peroxidase development system (AEC; Dako, Vienna, Austria). Immunofluorescence included: Alexa546 goat anti-rabbit, Alexa546 donkey anti-goat, Alexa488 donkey anti-rat, and Alexa633 goat anti-mouse secondary antibodies (Invitrogen, Vienna, Austria), or Alexa488 donkey anti-mouse (Jackson, West Grove, PA). All immunofluorescent images were acquired using confocal laser scanning microscopy (LSM780; Carl Zeiss). Clinical records were retrieved and assessed for the occurrence of locoregional metastases and correlated to the absence or presence of CXCL5 in tumor sections. Observation period was between 6 and 11 years, mean observation period was 8.1 ± 2.1 years. The study was approved by the institutional ethics committee (ethics committee of the Medical University Vienna) and according to the Helsinki declaration. Written informed consent was obtained.

      Image Analysis

      Tumor infiltrating neutrophils or intratumoral blood/lymphatic vessels in xenografts were stained immunohistochemically with anti-LY6G, anti-LYVE-1, or anti-CD31 antibodies and quantified by positive pixel count in relation to total tumor pixel count using Image Scope software, version 12.3.0.5056 (Leica, Wetzlar, Germany). Numbers of anti-LY6G–positive tumor-infiltrating neutrophils in human tumors were counted with Strataquest Analysis software (TissueGnostics, Vienna, Austria). The distance of neutrophils to lymphatic vessels in xenograft sections was evaluated by immunofluorescent labeling with anti-LY6G, anti-LYVE-1, and DAPI using ZEN software (Zeiss) and classified into three categories (≤10, 11–20, and 21–200 μm).

      Reverse Transmigration

      Lymphatic endothelial cells (5 × 105 cells) were seeded on the undersurface of fibronectin (Sigma Aldrich)-coated 3-μm Transwells (Merck Millipore, Darmstadt, Germany) in EBM-2 (Lonza) supplemented with 10% fetal calf serum and EGM-2MV Single Quots supplement mix (Lonza). Cells were allowed to adhere for 1 hour; Transwells were turned and placed back into the wells. After a confluent monolayer had formed, lymphatic endothelial cells were stimulated with TNF-α (10  ng/ml) and IFN-γ (2.5 U/ml). A375 cells (1.75 × 106 cells) alone or in combination with neutrophils (1.75 × 106 cells) were added into the upper well. Neutrophils were isolated from healthy volunteers using density gradient centrifugation (Histopaque1119, Histopaque1077; Sigma Aldrich). Migration was performed in EBM-2 medium (+0.5% fetal calf serum in the bottom chamber) for 4 hours at 37°C in a humidified CO2-containing atmosphere. Transmigrated cells were stained with anti-human VIMENTIN antibody and analyzed on a FACS Aria using Diva software (BD, Vienna, Austria).

      Conflict of Interest

      The authors state no conflict of interest.

      Acknowledgments

      We thank Tamara Prusa of the animal care facility, Monika Weiss, Ingrid Fae, Claudia Kokesch, and Karin Neumüller for technical assistance. Special thanks to Sabine Rauscher of the imaging core facility for her continuous help. Supported by L’Oréal foundation For Women in Science, OeNB Anniversary Fund grants 13672, 15148 and Austrian Science Fund (FWF) P24022-B21.

      Supplementary Material

      References

        • Amersi F.F.
        • Terando A.M.
        • Goto Y.
        • Scolyer R.A.
        • Thompson J.F.
        • Tran A.N.
        • et al.
        Activation of CCR9/CCL25 in cutaneous melanoma mediates preferential metastasis to the small intestine.
        Clin Cancer Res. 2008; 14: 638-645
        • Antonicelli F.
        • Lorin J.
        • Kurdykowski S.
        • Gangloff S.C.
        • Le Naour R.
        • Sallenave J.M.
        • et al.
        CXCL10 reduces melanoma proliferation and invasiveness in vitro and in vivo.
        Br J Dermatol. 2011; 164: 720-728
        • Arenberg D.A.
        • Keane M.P.
        • DiGiovine B.
        • Kunkel S.L.
        • Morris S.B.
        • Xue Y.Y.
        • et al.
        Epithelial-neutrophil activating peptide (ENA-78) is an important angiogenic factor in non-small cell lung cancer.
        J Clin Invest. 1998; 102: 465-472
        • Baggiolini M.
        Chemokines and leukocyte traffic.
        Nature. 1998; 392: 565-568
        • Bald T.
        • Quast T.
        • Landsberg J.
        • Rogava M.
        • Glodde N.
        • Lopez-Ramos D.
        • et al.
        Ultraviolet-radiation-induced inflammation promotes angiotropism and metastasis in melanoma.
        Nature. 2014; 507: 109-113
        • Balkwill F.
        Cancer and the chemokine network.
        Nat Rev Cancer. 2004; 4: 540-550
        • Belperio J.A.
        • Keane M.P.
        • Ehlert L.E.
        • Arenberg D.A.
        • Burdick M.D.
        • Strieter R.M.
        CXC chemokines in angiogenesis.
        J Leukoc Biol. 2000; 68: 1-8
        • Chambers A.F.
        • Groom A.C.
        • MacDonald I.C.
        Dissemination and growth of cancer cells in metastatic sites.
        Nat. Rev Cancer. 2002; 2: 563-572
        • Chao T.
        • Furth E.E.
        • Vonderheide R.H.
        CXCR2-dependent accumulation of tumor-associated neutrophils regulates T-cell immunity in pancreatic ductal adenocarcinoma.
        Cancer Immunol Res. 2016; 4 (96–82)
        • Chatterjee S.
        • Behnam Azad B.
        • Nimmagadda S.
        The intricate role of CXCR4 in cancer.
        Adv Cancer Res. 2014; 124: 31-82
        • Chow M.T.
        • Luster A.D.
        Chemokines in cancer.
        Cancer Immunol Res. 2014; 2: 1125-1131
        • Chtanova T.
        • Schaeffer M.
        • Han S.J.
        • van Dooren G.G.
        • Nollmann M.
        • Herzmark P.
        • et al.
        Dynamics of neutrophil migration in lymph nodes during infection.
        Immunity. 2008; 29: 487-496
        • Das S.
        • Sarrou E.
        • Podgrabinska S.
        • Cassella M.
        • Mungamuri S.K.
        • Feirt N.
        • et al.
        Tumor cell entry into the lymph node is controlled by CCL1 chemokine expressed by lymph node lymphatic sinuses.
        J Exp Med. 2013; 210: 1509-1528
        • Dhawan P.
        • Richmond A.
        Role of CXCL1 in tumorigenesis of melanoma.
        J Leuko Biol. 2002; 27: 9-18
        • Granot Z.
        • Henke E.
        • Comen E.A.
        • King T.A.
        • Norton L.
        • Benezra R.
        Tumor entrained neutrophils inhibit seeding in the premetastatic lung.
        Cancer Cell. 2011; 20: 300-314
        • Gröger M.
        • Loewe R.
        • Holnthoner W.
        • Embacher R.
        • Pillinger M.
        • Herron G.S.
        • et al.
        IL-3 Induces expression of lymphatic markers Prox-1 and Podoplanin in human endothelial cells.
        J Immunol. 2004; 173: 7161-7169
        • Hampton H.R.
        • Bailey J.
        • Tomura M.
        • Brink R.
        • Chtanova T.
        Microbe-dependent lymphatic migration of neutrophils modulates lymphocyte proliferation in lymph nodes.
        Nat Commun. 2015; 6: 7139
        • Homsi J.
        • Daud A.I.
        Spectrum of activity and mechanism of action of VEGF/PDGF inhibitors.
        Cancer Control. 2007; 14
        • Jamieson N.B.
        • Glen P.
        • McMillan D.C.
        • McKay C.J.
        • Foulis A.K.
        • Carter R.
        • et al.
        Systemic inflammatory response predicts outcome in patients undergoing resection for ductal adenocarcinoma head of pancreas.
        Br J Cancer. 2005; 92: 21-23
        • Kennelly R.P.
        • Murphy B.
        • Larkin J.O.
        • Mehigan B.J.
        • McCormick P.H.
        Activated systemic inflammatory response at diagnosis reduces lymph node count in colonic carcinoma.
        World J Gastrointest Oncol. 2016; 8: 62-68
        • Kitamura T.
        • Qian B.Z.
        • Pollard J.W.
        Immune cell promotion of metastasis.
        Nat Rev Immunol. 2015; 15: 73-86
        • Kochetkova M.
        • Kumar S.
        • McColl S.R.
        Chemokine receptors CXCR4 and CCR7 promote metastasis by preventing anoikis in cancer cells.
        Cell Death Differ. 2009; 16: 664-673
        • Kumar R.
        • Geuna E.
        • Michalarea V.
        • Guardascione M.
        • Naumann U.
        • Lorente D.
        • et al.
        The neutrophil-lymphocyte ratio and its utilisation for the management of cancer patients in early clinical trials.
        Br J Cancer. 2015; 112: 1157-1165
        • Laib A.M.
        • Bartol A.
        • Alajati A.
        • Korff T.
        • Weber H.
        • Augustin H.G.
        Spheroid-based human endothelial cell microvessel formation in vivo.
        Nat Protocols. 2009; 4: 1202-1215
        • Le Y.
        • Zhou Y.
        • Iribarren P.
        • Wang J.M.
        Chemokines and chemokine receptors: their manifold roles in homeostasis and disease.
        Cell Mol Immunol. 2004; 1: 95-104
        • Legler D.F.
        • Uetz-von Allmen E.
        • Hauser M.A.
        CCR7: roles in cancer cell dissemination, migration and metastasis formation.
        Int J Biochem Cell Biol. 2014; 54: 78-82
        • Li A.
        • King J.
        • Moro A.
        • Sugi M.D.
        • Dawson D.W.
        • Kaplan J.
        • et al.
        Overexpression of CXCL5 is associated with poor survival in patients with pancreatic cancer.
        Am J Pathol. 2011; 178: 1340-1349
        • Livak K.J.
        • Schmittgen T.D.
        Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method.
        Methods. 2001; 25: 402-408
        • Loewe R.
        • Holnthoner W.
        • Gröger M.
        • Pillinger M.
        • Gruber F.
        • Mechtcheriakova D.
        • et al.
        Dimethylfumarate inhibits TNF-induced nuclear entry of NF-kB/p65 in human endothelial cells.
        J Immunol. 2002; 168: 4781-4787
        • Loewe R.
        • Valero T.
        • Kremling S.
        • Pratscher B.
        • Kunstfeld R.
        • Pehamberger H.
        • et al.
        Dimethylfumarate impairs melanoma growth and metastasis.
        Cancer Res. 2006; 66: 11888-11896
        • Lopez-Lago M.A.
        • Posner S.
        • Thodima V.J.
        • Molina A.M.
        • Motzer R.J.
        • Chaganti R.S.
        Neutrophil chemokines secreted by tumor cells mount a lung antimetastatic response during renal cell carcinoma progression.
        Oncogene. 2013; 32: 1752-1760
        • Meier F.
        • Will S.
        • Ellwanger U.
        • Schlagenhauff B.
        • Schittek B.
        • Rassner G.
        • et al.
        Metastatic pathways and time courses in the orderly progression of cutaneous melanoma.
        Br J Dermatol. 2002; 147: 62-70
        • Miyazaki H.
        • Patel V.
        • Wang H.
        • Edmunds R.K.
        • Gutkind J.S.
        • Yeudall W.A.
        Down-regulation of CXCL5 inhibits squamous carcinogenesis.
        Cancer Res. 2006; 66: 4279-4284
        • Mueller B.M.
        • Romerdahl C.A.
        • Trent J.M.
        • Reisfeld R.A.
        Suppression of spontaneous melanoma metastasis in scid mice with an antibody to the epidermal growth factor receptor.
        Cancer Res. 1991; 51: 2193-2198
        • Nagarsheth N.
        • Wicha M.S.
        • Zou W.
        Chemokines in the cancer microenvironment and their relevance in cancer immunotherapy.
        Nat Rev Immunol. 2017; 17: 559-572
        • National Institutes of Health
        Guide for the Care and Use of Laboratory Animals (publication no. 86–23).
        National Institutes of Health, Bethesda, MD1985
        • Park J.Y.
        • Park K.H.
        • Bang S.
        • Kim M.H.
        • Lee J.E.
        • Gang J.
        • et al.
        CXCL5 overexpression is associated with late stage gastric cancer.
        J Cancer Res Clin Oncol. 2007; 133: 835-840
        • Pastushenko I.
        • Vermeulen P.B.
        • Carapeto F.J.
        • Van den Eynden G.
        • Rutten A.
        • Ara M.
        • et al.
        Blood microvessel density, lymphatic microvessel density and lymphatic invasion in predicting melanoma metastases: systematic review and meta-analysis.
        Br J Dermatol. 2014; 170: 66-77
        • Pirker C.
        • Holzmann K.
        • Spiegl-Kreinecker S.
        • Elbling L.
        • Thallinger C.
        • Pehamberger H.
        • et al.
        Chromosomal imbalances in primary and metastatic melanomas: over-representation of essential telomerase genes.
        Melanoma Res. 2003; 13: 483-492
        • Rebhun R.B.
        • Cheng H.
        • Gershenwald J.E.
        • Fan D.
        • Fidler I.J.
        • Langley R.R.
        Constitutive expression of the β4 integrin correlates with tumorigenicity and lymph node metastasis of the B16 murine melanoma.
        Neoplasia. 2010; 12: 173-182
        • Richmond A.
        • Yang J.
        • Su Y.
        The good and the bad of chemokines/chemokine receptors in melanoma.
        Pigment Cell Melanoma Res. 2009; 22: 175-186
        • Shields J.D.
        • Borsetti M.
        • Rigby H.
        • Harper S.J.
        • Mortimer P.S.
        • Levick J.R.
        • et al.
        Lymphatic density and metastatic spread in human malignant melanoma.
        Br J Cancer. 2004; 90: 693-700
        • Singh R.K.
        • Varney M.L.
        IL-8 expression in malignant melanoma: implications in growth and metastasis.
        Histol Histopathol. 2000; 15: 843-849
        • Spicer J.D.
        • McDonald B.
        • Cools-Lartigue J.J.
        • Chow S.C.
        • Giannias B.
        • Kubes P.
        • et al.
        Neutrophils promote liver metastasis via Mac-1-mediated interactions with circulating tumor cells.
        Cancer Res. 2012; 72: 3919-3927
        • Swoboda A.
        • Schanab O.
        • Tauber S.
        • Bilban M.
        • Berger W.
        • Petzelbauer P.
        • et al.
        MET expression in melanoma correlates with a lymphangiogenic phenotype.
        Hum Mol Genet. 2012; 21: 3387-3396
        • Ulvmar M.H.
        • Hub E.
        • Rot A.
        Atypical chemokine receptors.
        Exp Cell Res. 2011; 317: 556-568
        • Vanharanta S.
        • Massague J.
        Origins of metastatic traits.
        Cancer Cell. 2013; 24: 410-421
        • Walz A.
        • Burgener R.
        • Car B.
        • Baggiolini M.
        • Kunkel S.L.
        • Strieter R.M.
        Structure and neutrophlil-activating properties of a novel inflammatory peptide (ENA-78) with homology to interleukin 8.
        J Exp Med. 1991; 174: 1355-1362
        • Wang J.
        • Kalhor N.
        • Hu J.
        • Wang B.
        • Chu H.
        • Zhang B.
        • et al.
        Pretreatment neutrophil to lymphocyte ratio is associated with poor survival in patients with stage I-III non-small cell lung cancer.
        PLoS One. 2016; 11: e0163397
        • Wiley H.E.
        • Gonzalez E.B.
        • Maki W.
        • Wu M.T.
        • Hwang S.T.
        Expression of CC chemokine receptor-7 and regional lymph node metastasis of B16 murine melanoma.
        J Natl Cancer Inst. 2001; 93: 1638-1643
        • Zhang H.
        • Xia W.
        • Lu X.
        • Sun R.
        • Wang L.
        • Zheng L.
        • et al.
        A novel statistical prognostic score model that includes serum CXCL5 levels and clinical classification predicts risk of disease progression and survival of nasopharyngeal carcinoma patients.
        PloS One. 2013; 8: e57830
        • Zheng J.
        • Zhu X.
        • Zhang J.
        CXCL5 knockdown expression inhibits human bladder cancer T24 cells proliferation and migration.
        Biochem Biophys Res Commun. 2014; 446: 18-24
        • Zhou S.L.
        • Dai Z.
        • Zhou Z.J.
        • Chen Q.
        • Wang Z.
        • Xiao Y.S.
        • et al.
        CXCL5 contributes to tumor metastasis and recurrence of intrahepatic cholangiocarcinoma by recruiting infiltrative intratumoral neutrophils.
        Carcinogenesis. 2014; 35: 597-605
        • Zlotnik A.
        Chemokines and cancer.
        Int J Cancer. 2006; 119: 2026-2029