Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Omental macrophages secrete chemokine ligands that promote ovarian cancer colonization of the omentum via CCR1

Omental macrophages secrete chemokine ligands that promote ovarian cancer colonization of the... ARTICLE https://doi.org/10.1038/s42003-020-01246-z OPEN Omental macrophages secrete chemokine ligands that promote ovarian cancer colonization of the omentum via CCR1 1 1 2 1 3 4 Venkatesh Krishnan , Supreeti Tallapragada , Bruce Schaar , Kalika Kamat , Anita M. Chanana , Yue Zhang , 5 6 7 8 1,9 1,10 Sonia Patel , Vinita Parkash , Carrie Rinker-Schaeffer , Ann K. Folkins , Erinn B. Rankin & Oliver Dorigo The omentum is the most common site of ovarian cancer metastasis. Immune cell clusters called milky spots are found throughout the omentum. It is however unknown if these immune cells contribute to ovarian cancer metastasis. Here we report that omental macro- phages promote the migration and colonization of ovarian cancer cells to the omentum through the secretion of chemokine ligands that interact with chemokine receptor 1 (CCR1). We found that depletion of macrophages reduces ovarian cancer colonization of the omentum. RNA-sequencing of macrophages isolated from mouse omentum and mesenteric adipose tissue revealed a specific enrichment of chemokine ligand CCL6 in omental mac- rophages. CCL6 and the human homolog CCL23 were both necessary and sufficient to promote ovarian cancer migration by activating ERK1/2 and PI3K pathways. Importantly, inhibition of CCR1 reduced ovarian cancer colonization. These findings demonstrate a critical mechanism of omental macrophage induced colonization by ovarian cancer cells via CCR1 signaling. Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, Stanford School of Medicine, Stanford Cancer Institute, Stanford, CA, USA. 2 3 Genome Technology Center, Stanford, CA, USA. Department of Obstetrics and Gynecology, David Geffen School of Medicine, University of California, Los 4 5 Angeles, CA, USA. Department of Physiology and Chemistry, Genentech, San Francisco, CA, USA. Department of Thoracic Head and Neck Medical Oncology, Division of Cancer Medicine, University of Texas Health Science Center at Houston, Houston, TX, USA. Department of Pathology, Yale School of Medicine and Yale School of Public Health, New Haven, CT, USA. Department of Surgery, Division of Urology, University of Chicago, Chicago, IL, USA. 8 9 Department of Pathology, Stanford School of Medicine, Stanford, CA, USA. Department of Radiation Oncology, Stanford School of Medicine, Stanford, CA, 10 ✉ USA. Stanford Women’s Cancer Center, Stanford, CA, USA. email: odorigo@stanford.edu COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 1 1234567890():,; ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z varian cancer is the seventh most common cancer in To investigate the role of tissue-resident macrophages in women with 300,000 new cases and 185,000 deaths metastatic colonization to the omentum, we first verified the Oworldwide in 2018 . In 2020, an estimated 21,750 new presence of macrophages in milky spots in naive human and cases of ovarian cancer will be diagnosed in the USA and 13,940 murine omentum (Fig. 1b, c). Interestingly, in omenta from women will die from the disease . Ovarian cancer originates in ovarian cancer patients without gross disease, microscopic the ovary or fallopian tube and metastasizes throughout the metastases were restricted to milky spots and were not observed peritoneal cavity . Metastatic disease, rather than the primary in the adipose tissue at this early stage (Fig. 1b). Similarly, tumor, is the cause of ovarian cancer-related deaths . However, microscopic metastases in murine omentum were localized and the fundamental mechanisms that support intraperitoneal confined to the milky spots at early timepoints (24 h) after metastasis are not well understood. intraperitoneal (i.p.) injection of ID8 cells (Fig. 1c). The omentum is a preferred metastatic site for ovarian cancer We next studied the dynamics of intraperitoneal metastasis in cells. The majority of women diagnosed with high grade serous our murine ovarian cancer model. ID8 murine ovarian cancer 5–7 ovarian cancer present with omental metastases . In experi- cells were injected i.p. into C57BL/6 mice and the anatomic mental models of ovarian cancer, ovarian cancer cells metastasize distribution of metastatic colonization was followed over time. to the omentum within 1–6 h of injection . The omentum is an Interestingly, the omentum was the first site of macroscopic adipose tissue that contains secondary lymphoid structures with a metastases within one week of injection and remained the only variety of cell types, including adipocytes, blood vessels, and site of peritoneal metastasis for 6 weeks after which metastases 9–15 clusters of leukocytes known as milky spots . Milky spots appeared at other sites (Fig. 1d). This is similar to observations in contain macrophages, B, T, and NK cells, which rapidly efflux patients with ovarian cancer who frequently present with omental 10,16 into the peritoneum during pathogenic challenge . While metastasis upon diagnosis but may not have other sites of 5,6 various cell types, including neutrophils, mesothelial cells, and significant disease burden . Milky spots increased in number adipocytes, have been shown to support ovarian cancer metastasis and size after cancer cell colonization at week 1 post injection. to the omentum, little is known regarding the role of tissue- The number of immune cells, in particular macrophages, was 7,15,17 resident immune cells in omental metastasis . found to increase by week 2. We observed continued expansion of Macrophages enhance metastasis in several cancer models. In tumor from milky spots into the adipose tissue of the omentum breast cancer, macrophages facilitate the early hematogenous following early colonization (Fig. 1e). These findings suggest that dissemination of cancer cells from the primary tumor by altering the immune cell-rich milky spots are sites of early metastasis of vascular permeability and facilitating migration into the blood- ovarian cancer cells within the omentum. stream . Chemokines enhance the interactions between cancer cells and macrophages in a chemokine cascade that promotes Macrophages promote ovarian cancer colonization of the breast cancer metastasis . In ovarian cancer, macrophage-driven omentum. Previous studies have demonstrated in both murine inflammation can promote intraperitoneal metastasis and ascites (ID8) and human ovarian cancer models (CaOV3, HeyA8, or production . Specific subsets of omental macrophages might be SKOV3ip.1) that colonization of the omentum was independent involved in promoting metastasis to the omentum . of B-, T-, and NK- cells . To determine the role of macrophages In this study, we describe the role of omental macrophage- in omental colonization, we depleted macrophages in C57BL/6 derived chemokine ligands in promoting ovarian cancer cell and athymic nude mice using clodronate liposomes prior to i.p. colonization of the omentum. Gene expression profiling of injection of mouse ID8 or human SKOV3ip.1 ovarian cancer mouse omental macrophages compared to mesenteric adipose cells, respectively (Fig. 2a). Strikingly, we observed an 11-fold macrophages shows distinct gene expression patterns with reduction in the ovarian cancer burden in the omentum of significant upregulation of the chemokine ligand CCL6 prior to macrophage depleted mice at 7 days following i.p. injection of and during colonization of the omentum by ovarian cancer tumor cells (Fig. 2b). Interestingly, when macrophages were cells. CCL6 induces the migration of ovarian cancer cells as an depleted using clodronate liposome three days after injection of important step in colonization via the chemokine receptor SKOV3ip.1 cells, we failed to observe a decrease in omental CCR1. Furthermore, inactivation of CCR1 in mouse ovarian colonization. (Supplementary Fig. 1). These findings support our cancer cells blocks CLL6 induced migration and abolishes their hypothesis that macrophages are crucial for the initiation of ability to colonize the omentum. In ovarian cancer patients, the ovarian cancer colonization of the omentum. expression of CCR1 is associated with a poorer prognosis. In addition, CCL23, the human homolog of CCL6, is expressed on macrophages isolated from human omentum. These observa- Omental macrophage-derived conditioned media promote tions validate the importance of CCR1 and its ligand CCL23 in migration of ovarian cancer cells. Our findings above indicate an human ovarian cancer. important role for omental macrophages in metastatic coloniza- tion of ovarian cancer. Therefore, we investigated whether omental macrophages can induce migration of ovarian cancer Results cells as an important aspect of metastatic colonization. For this Omentum as a niche for ovarian cancer metastasis. The purpose, CD11b cells were isolated from normal mouse omenta omental vasculature is characterized by the presence of numerous and cultured to generate conditioned media (Supplementary branching blood vessels ending in glomerulus-like capillary beds Fig. 2a). CD11b cells isolated from murine gonadal fat were near the periphery of the omentum. Immune cells aggregate used for comparison since our prior studies had demonstrated a around and within these capillary beds to form milky spots. The lack of early metastasis to intraperitoneal adipose outside of the + 21 + milky spots contain clusters of monocytes (CD11b ), B cells, omentum . Conditioned media generated by CD11b cells iso- macrophages, CD11c cells, and T cells all under a gap in the lated from mouse omenta significantly increased migration of ID8 mesothelial lining (Fig. 1a). The endothelial lining of the capil- ovarian cancer cells (Supplementary Fig. 2b). Compared to con- laries and the overlying mesothelium are adapted to facilitate the ditioned media from omental CD11b cells, conditioned media transmigration of immune cells. Additional structural elements from gonadal fat CD11b cells had only minor effects on ovarian are provided by, fibroblasts, mesenchymal cells, collagen, and cancer cell migration. To validate these findings in human tissue, elastic fibers (Fig. 1a). we isolated macrophages from human omentum collected from 2 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE ab CD68 H&E CD45 Unstimulated Naive Mouse Omentum Ovarian Cancer Cells Unstimulated/Naive Milky Spot CD45 H&E F4/80 Stimulated/Cancer-Colonized Mouse Omentum Adipocyte Adipocyte Tumor-Associated Macrophages Tumor-Associated Macrophages ID8 in C57BL/6 - 24hpi ID8 in C57BL/6 - 24hpi Dendritic Cell (CD11c+ cells) Dendritic Cell (CD11c+ cells) Fibroblasts Fibroblasts Ovarian Cancer Cell Ovarian Cancer Cell B-Cell B-Cell Mesothelial Cell Mesothelial Cell T-Cell T-Cell Extracellular Matrix Extracellular Matrix Monocytes Monocytes Hematogenous Circulation Hematogenous Circulation b Lymphatic Circulation Lymphatic Circulation M Macrophages acrophages Stimulated/Cancer-Colonized Milky Spot de Naive F4/80 Omentum Peritoneum Week 1 Diaphgragm Liver Week 2 Week 4 1.0 0.5 Naive Week 1 Week 2 Week 4 0.0 12 34 56 789 Weeks Post Injection Fig. 1 Omentum as a niche for ovarian cancer metastasis. a Representation of a naïve unstimulated milky spot and ovarian cancer metastasis activated milky spots within the mouse omentum. The number of macrophages associated with the expanding cancer-immune cells regions increase and cancer cells grow past milky spots into the central adipose region, eventually replacing it. b, c Omental milky spots can be detected in human and mouse samples. IHC of naive omental tissue sections stained for H&E, CD45 and macrophages (mouse F4/80 and human CD68). Uninvolved milky spot in human (b; inset a, b) and mouse (c; inset a). Microscopic ovarian metastases were identified within milky spots in human omenta (b; inset c), and mouse (c; inset b); scale bar = 100 µm. d C57BL/6 mice were injected with ID8 cancer cells and overt (>1 mm) peritoneal metastases over a period of 9 weeks were quantitated for metastatic spread within the abdomen (n = 6). e Representative digital scans of whole mouse omental sections stained for F4/80. Milky spots at week 0 (naive un-injected omentum) and week 1, week 2, week 3, week 4, post injection of ID8 cells. (Bottom) ×40 magnification of milky spot regions at each time point. (Scale bar = 1×, 1000 μM; 40×, 50 μM). + + + individuals with benign disease (CD45 CD14 CD68 cells; RNA-sequencing of omental macrophage reveals distinct gene Supplementary Fig. 3b). Similar to the effects shown in murine expression patterns. We next investigated the mechanisms that CD11b cells, all human macrophage conditioned media sig- regulate the omental macrophage-induced colonization of the nificantly enhanced the migration of human SKOV3.ip1 cells omentum. For this purpose, changes in gene expression were (Supplementary Fig. 2d). These findings suggest that omental studied in macrophages during the omental colonization process. + + + macrophages secrete chemoattractants into the tumor micro- Macrophages (CD45 /CD11b /F4/80 ) were isolated from mouse environment that enhance migration and, therefore, the meta- omenta prior to injection of ID8 cells, and 24 hours and seven days static potential of ovarian cancer cells. post-injection (dpi). RNA sequencing analysis was performed on COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 3 Avg. No. of Overt Peritoneal Metastases Human Omentum Mouse Omentum ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z Liposome ID8 or SKOV3ip.1 Liposome Liposome Euthanize & Treatment (1 x 10 cells) Treatment Treatment FFPE Tissues Day 0 Day 1 Day 3 Day 5 Day 7 ID8 in C57BL/6 SKOV3ip.1 in Athymic Nude CD45 F4/80 Pan-CK CD45 F4/80 Pan-CK ** PBS Clodronate *** PBS Clodronate cd e Ccl6 10 Ccl9 8 Ccl24 6 Ccl8 MY:0h Ccl25 MY:24h Ccl3 MY:7d Ccl2 CCL6 Ccl4 OM:0h Ccl7 OM:24h Ccl11 Ccl17 OM:7d Ccl12 Ccl27a Ccl5 PC1: 86% Variance Ccl22 Ccl20 Ccl1 My0h My24h My7d OM0h OM24h OM7d 0 200 400 600 800 1000 Average Expression (FPKM) in Omentum fg h i Mice CCL6 ELISA Human CCL23 ELISA 1.0 0.20 0.8 0.15 0.6 0.10 5000 0.4 800 0.05 0.2 400 0.0 0.00 0 Omentum Gonadal Uterine Mesenteric RAW mOM-M 1 mOM-M 2 PBMC hOM-M 1 hOM-M 2 Fat Fat Fat j k C57BL/6 Athymic Nude Human CCL6; F4/80; DAPI CCL6; F4/80; DAPI CCL23; CD68; DAPI macrophages at the respective time points (Supplementary Fig. 3a). macrophages before injection of ID8 cells (1283 genes were dif- RNA isolated from mesenteric adipose tissue macrophages ferentially expressed between omentum and mesentery derived was used for comparison. Hierarchical clustering and principal- macrophages). Second, omental macrophages showed a significant component analysis (PCA) of 12756 genes expressed by omental change in gene expression after the injection of ID8 cells, while macrophages during intraperitoneal metastasis resulted in various there was no change in the gene signature from mesenteric adipose important observations. First, naive omental macrophages showed tissue-derived macrophages. (Fig. 2c). Genes expressed in omental a gene signature distinct from mesenteric adipose tissue-derived macrophages at 24 h and 7 dpi were significantly enriched for cell 4 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio hOm-M hOm-M hOm-M PBMC OC156 OC157 OC158 OC159 PBS Clodronate Liposome CCL6 (1/ΔCT) PC2: 8% Variance Normal Omentum CCL23 (1/ΔCT) Clodronate Liposome PBS CCL6 (pg/mL) Log Fold Change (Omentum/Mesentery) 2 024 6 CCL23 (pg/mL) % of SKOV3ip.1 Cells Per Omentum % of ID8 cells per Omentum COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Fig. 2 Ovarian cancer cell colonization of the omentum is macrophage dependent. a Experimental design of clodronate liposome treatment for 7 dpi of ID8/SKOV3ip.1 in vivo metastasis assay. b Representative omental images of IHC for lymphocytes (CD45), macrophages (F4/80) and cancer cells (pan- CK). Quantification of DAB (pan-CK) staining area for each model is shown on the right (*p < 0.05; n = 3). (c–e) c Principal-component analysis (PCA) of genes expressed in the omental (OM) and mesenteric (MY) adipose tissue macrophages. Each data point represents macrophages derived from three omenta (n = 3). d Heat map of all chemokines expressed in (c). e Plot of chemokine genes highly upregulated in naive omental macrophages compared to naive mesenteric macrophages (timepoint zero). f Gene expression analysis of CCL6 on mouse naive omental macrophages isolated from C57BL/6 mice (n = 2) by q-RT PCR. RAW 264.7 cells were used as positive control. g Gene expression analysis of CCL23 on human omental macrophages isolated from patient with benign disease (n= 2). PBMCs were used as positive control. h ELISA of CCL6 levels in conditioned media generated from weight matched naive omentum and peritoneal adipose tissues (n = 3). i ELISA of CCL23 levels on conditioned media generated from benign patient samples. Omental macrophages (hOM-Mϕ1, 2, and 3) and whole-omental-tissue (OC157, 158, and 159) each represent a different benign sample. Conditioned media derived from PBMCs were used as control. j Immunofluorescence staining for co-localization of CCL6 (red) and F4/80 (green) and nuclei (blue, DAPI) on naive omental tissue from C57BL/6 and athymic nude mice. k Similarly, CCL23 (red) and CD68 (green) on the human omentum derived from benign patient samples. (scale bar = 20 µm). adhesion, immune response, cytokine-cytokine receptor interac- Phospho-proteomic profiling of CCL6 treated ID8 cells and CCL23 tion, and chemotaxis (Supplementary Fig. 3c). Among the che- treated SKOV3ip.1 cells showed increased levels of pERK1/2 in both mokines, CCL6 was found to be the gene with the highest fold cell lines and increased pAkt in SKOV3ip.1 cells (Supplementary upregulation in naive omental macrophages compared to naive Fig. 4d, e). pERK1/2 and pAkt regulate migration in part by 25–28 mesenteric macrophages (timepoint zero). In addition, we phosphorylating LIMK1 and p-Cofilin (Fig. 3f and Supple- observed a significant increase of CCL6 expression in omental but mentary Fig. 4f–i). We confirmed the upregulation of p-cofilin by not mesenteric macrophages during colonization (Fig. 2d, e). Since immunofluorescence in both ID8 and SKOV3ip.1 as early as 10 min 22–24 CCL6 is expressed primarily by macrophages ,wehypothe- post treatment with CCL6 and CCL23, respectively (Fig. 3g, h). We sized that macrophage-derived CCL6 might promote early omental hypothesized that phosphorylationofERK1/2byCCL6and CCL23 metastasis by interaction with its receptor CCR1 expressed on regulates MYO9A, which modulates downstream p-LIMK1 and p- ovarian cancer cells. Cofilininthe migrationpathway (Fig. 3i). In vitro migration of ID8 We validated the expression of CCL6 and CCL23 in flow-sorted and SKOV3ip.1 cells toward CCL6 or CCL23 was inhibited by mouse or human omental macrophages, respectively by qRT-PCR pharmacologic targeting of MEK (trametinib) and PI3K pathway (Fig. 2f, g). Conditioned media generated from naive mouse (BEZ235) respectively (Fig. 3j–l).In addition,inhibition ofMyosin-9 omenta contained high levels of secreted CCL6 (Fig. 2h). High (blebbistatin) resulted in a two-fold decreased migration of both ID8 levels of CCL23 were also detected in conditioned media from and SKOV3ip.1 cells toward CCL6 and CCL23, respectively (Fig. 3m, whole human omentum without tumor and human omentum n). The concentrations of BEZ235, trametinib and blebbistatin used derived macrophages (Fig. 2i). Immunofluorescence staining of in the migration assays did not affect the viability of ovarian cancer tissue sections further confirmed the presence of CCL6-expressing cells (Supplementary Fig. 4j–n). Our findings suggest that CCL6 or F4/80 macrophages in mouse omenta and CCL23-expressing CCL23 induce migration by activating the ERK and PI3-Kinase CD68 macrophages in human omenta (Fig. 2j, k). pathways resulting in enhanced downstream signaling via MYO9A and p-Cofilin. CCL6 and CCL23 promote ovarian cancer migration.We Genetic and pharmacological inhibition of CCR1 in human hypothesized that CCL6 and CCL23 are important mediators of and mouse ovarian cancer cells reduces migration toward colonization in our mouse model and in human omentum CCL6/CCL23. CCL6 and CCL23 signal through the CCR1 che- respectively. Both chemokine ligands can induce migration which mokine receptor. To determine the role of CCR1 in macrophage- is a crucial aspect of colonization. Therefore, we investigated the induced omental colonization, we first verified the expression of ability of CCL6 and CCL23 to enhance migration of mouse and CCR1 in ID8 and SKOV3ip.1 cells (Supplementary Fig. 5a). CCR1 human ovarian cancer cell lines. Both murine ID8 and human was expressed in ID8 and SKOV3ip.1 derived tumor tissue SKOV3ip.1 cell lines showed a concentration-dependent migra- (Fig. 4a, b). The functional importance of CCR1 in CCL6 and tion toward CCL6 and CCL23 (Fig. 3a, b). This effect was not due CCL23 induced migration was examined in ID8 and to increased cell proliferation as neither ID8 nor SKOV3ip.1 cells SKOV3ip.1 cells with deletion of CCR1 by CRISPR-Cas9. Deletion cultured in serum-free medium in the presence of 100 ng/mL of CCR1 was verified by sequencing the targeted locus and by CCL6 or 200 ng/mL CCL23 showed differences in cell prolifera- qRT-PCR for CCR1 expression (Supplementary Fig. 5b–e). tion (Fig. 3c). Antibody-mediated neutralization of CCL6 abro- Induction of migration by CCL6 was completely abrogated in gated the migration of ID8 cells toward mouse omentum- ID8-CCR1-null cells. (Fig. 4c). Similarly, CCR1-deleted SKO- conditioned media (Fig. 3d). Similarly, induction of SKOV3ip.1 V3ip.1 cells failed to respond to CCL23 induced migration migration toward conditioned media from human omentum was (Fig. 4d). No differences in tumor cell proliferation or viability blocked by CCL23 neutralizing antibody (Fig. 3e). were observed in the CCR1-null cell lines relative to the parental To better understand the cellular mechanisms of CCL6 and cells (Supplementary Fig. 5f, g). We next utilized a small molecule CCL23 mediated migration, we performed microarray gene expres- inhibitor of CCR1 (UCB35625) to investigate the therapeutic sion analysis on CCL6 treated ID8 cells. We found that CCL6 potential of CCR1 blockade in reducing ovarian cancer migration. induced upregulation of 98 genes and downregulation of 108 genes. CCR1 inhibition by UCB35625 blocked CCL6 and CCL23 Upregulated genes were clustered in pathways that regulate cell induced migration of ID8 and SKOV3ip.1, respectively (Fig. 4e, f). migration, such as integrin-mediated cell adhesion, MAPK signaling pathway, and signaling by RhoGTPases (Supplementary Table 1). Among important regulators of migration, Myosin 9A (Myo9A), CCR1 knock-down in ovarian cancer cells reduces colonization LIMK1, and Cofilin were significantly upregulated in CCL6 treated of the omentum. We next studied if CCR1 is important in ID8 cells compared to untreated cells (Supplementary Fig. 4a–c). omental colonization by ovarian cancer cells in vivo. mRuby- COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 5 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ab c ID8 MTS assay 16 hours SKOV3ip.1 MTS assay 16 hours 1.50 1.50 250 250 1.25 1.25 **** 200 200 1.00 1.00 ** **** 150 150 *** ** ** * &%,-.#/0$*+%01#"+.2)%1#3!/4!"3#$#"5#%&6%7/#4&8"9#'*:;("<=)#*>+?@=-&#*:-.#>"A:%0*'"+' ** 0.75 0.75 ** 100 100 0.50 0.50 50 50 0.25 0.25 0 0 0.00 0.00 SF 1 102550 100 200 SF 1 10 25 50 100 200 SF 100ng 200ng 10%FBS SF 100ng 200ng 10%FBS Concentration of CCL23 (ng/mL) Concentration of CCL6 (ng/mL) Concentration of CCL6 (ng/mL) Concentration of CCL23 (ng/mL) SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL de f ID8 + 200ng/mL CCL6 SKOV3ip.1 + 200ng/mL CCL23 No Treatment 30 minutes No Treatment 360 minutes SF 250 SF pERK1/ERK2 (p44/p42) – 44/42kDa mOM CM hOM CM 200 Total ERK1/ERK2 – 44/42kDa 200 **** IgG IgG pAKT (Ser473) – (60kDa) **** NAb 150 NAb **** 150 **** **** Total AKT– 60kDa 10% FBS 10% FBS 100 **** **** **** pLIMK1(T508) – 72kDa pCofilin (Ser3) – 19kDa Total Cofilin – 19kDa hOM CM - + + + + + - mOM CM - + + + + + - IgG (µg/mL) - - 2.5 - 5 - - Alpha tubulin– 51kDa IgG (µg/mL) - - 2.5 - 5 - - NAb (µg/mL) - - - 2.5 - 5 - NAb (µg/mL) - - - 2.5 - 5 - No Treatment 10 minutes 30 minutes 360 minutes phospho-cofilin; actin; DAPI No Treatment 10 minutes 30 minutes 360 minutes phospho-cofilin; actin; DAPI CCL6 or CCL23 ij CCR1 ** ** ERK1/2 PI3K/AKT Myo 9a/b LIMK1 200ng/mL 100nM BEZ235 SF +200ng/mL CCL23 CCL23 Cofilin 1 Chemotaxis k l *** 100 *** **** **** ** ** **** 50 **** SF 200ng/mL CCL23 2.5 M Trametinib 5 M Trametinib 10 M Trametinib SF 200ng/mL 2.5uM Trametinib 5uM Trametinib 10uM Trametinib CCL6 +200ng/mL CCL6 +200ng/mL CCL6 +200ng/mL CCL6 +200ng/mL CCL23 +200ng/mL CCL23 +200ng/mL CCL23 mn 150 *** ** 100 **** 100 **** **** **** 50 50 0 0 SF 200ng/mL CCL6 25uM Bleb + 50uM Bleb + 75uM Bleb + SF 200ng/mL CCL23 25uM Bleb + 50uM Bleb + 75uM Bleb + 200ng/mL CCL6 200ng/mL CCL6 200ng/mL CCL6 200ng/mL CCL23 200ng/mL CCL23 200ng/mL CCL23 labeled ID8 parental cells and ID8-CCR1-null cells were injected to ID8-CRISPR-control cells with a three-fold reduction at 24 h i.p. into C57BL/6 mice. At 24 hours post-injection (hpi) and and a seven-fold reduction at 7 days post injection (Fig. 5b–d and seven dpi, omenta were harvested and assessed for metastasis by Supplementary Fig. 6a–f). histology and flow cytometry (Fig. 5a). Numerous foci of ID8 We then determined whether CCR1 is associated with a more parental (control) cells were observed within the milky spots of aggressive in vivo progression of tumor growth and its impact on the omentum 24 hpi and seven dpi. At both time points, omental survival. ID8-parental, ID8-CRISPR-Control, or ID8-CCR1-null tissue contained significantly less ID8-CCR1-null cells compared cells were injected i.p. into mice. Mice injected with ID8-CCR1- 6 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio #Migrated ID8 cells/200X Field # Migrated ID8 cells/ 200X Field #Migrated ID8 cells/200X Field ID8 + CCL6 (200 ng/mL) # Migrated ID8 cells/200X Field SKOV3ip.1 + CCL23 (200 ng/mL) # Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/ 200X Field # Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/200XField Absorbance at 595nm Absorbance at 595nm COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Fig. 3 Ovarian cancer cells show enhanced migration toward CCL6/CCL23 by activating Myosin9 via the Cofilin signaling in vitro. a, b Transwell migration assay of ID8/SKOV3ip.1 cells toward varying concentrations of CCL6 and CCL23 with representative crystal violet staining. Statistical significance (*p < 0.05) was determined for all conditions versus the serum-free media control by ordinary one-way ANOVA analysis. c Quantitation of cell proliferation of ID8/SKOV3ip.1 cells in the presence of CCL6/CCL23 (100 ng/mL–200 ng/mL) by MTS assay. 10% FBS was used as the positive control. d, e Transwell migration assays with ID8 cells in the presence of anti-CCL6 (2.5 μg/mL and 5 μg/mL; n = 3) toward mouse omental conditioned media and SKOV3ip.1 cells toward human omental conditioned media in the presence of anti-CCL23 (2.5 μg/mL–5 μg/mL; n = 3). Statistical significance was determined by two-way ANOVA analysis for each condition corresponding to the serum-free (SF) control (*p < 0.05). f Western blot analysis of ID8 treated with CCL6 (200 ng/mL) and SKOV3ip.1 treated with CCL23 (200 ng/mL) (n = 3). g, h Immunofluorescence staining for phospho-cofilin (red), actin cytoskeleton (green), and nuclei (blue, DAPI) on ID8 and SKOV3ip.1 ovarian cancer cells treated with CCL6 or CCL23, respectively, at different time points (n = 3); (scale bar = 20 µm). i Schematic representation of the CCR1-CCL6/CCL23 signaling pathway leading to increased migratory potential of the ovarian cancer cells. j Transwell migration assay of SKOV3ip.1 toward CCL23 treated with BEZ235 (PI3K Inhibitor) (**p < 0.008). k, l Transwell migration assay of ID8 toward CCL6 and SKOV3ip.1 cells toward CCL23 treated with trametinib (MEK inhibitor). m, n Transwell migration assay of ID8 toward CCL6 and SKOV3ip.1 cells toward CCL23 treated with blebbistatin (myosin inhibitor). Statistical significance was determined by two-way ANOVA analysis with Tukey’s multiple comparison t-test with each condition compared to the 200 ng/mL CCL23 treatment condition. null cells had a significantly slower progression of tumor growth on studies in immune-deficient mice . Our current study and longer survival compared to mice injected with ID8-CRISPR- demonstrates that omentum resident macrophages promote early Control or ID8 parental cells (Fig. 5e). After 9 weeks, all mice omental metastatic colonization. Gene expression profile of (n = 5/group) injected with ID8-parental cells had developed macrophages derived from the omentum demonstrated a tissue- ascites and extensive peritoneal metastases (Fig. 5f). In contrast, specific pattern compared to other intra-abdominal adipose tis- none of the mice injected with ID8-CCR1-null cells had sues. Macrophages in other organs including the lung or the liver developed ascites or other signs of disease at this time point. are likewise characterized by specific gene signatures which The median overall survival (OS) for mice injected with ID8- convey important and specific functions in their respective tissue CCR1-null cells was 156 days and, therefore, significantly longer microenvironments . Based on our data, omental macrophages compared to both ID8-parental (OS: 64 days) and ID8-CRISPR- secreted chemokine ligands support early metastasis of ovarian Control (OS: 97 days) (Fig. 5e). cancer cells providing a possible explanation for the frequent The data generated in the ID8 tumor model suggests that involvement of the omentum in ovarian cancer patients. deletion of CCR1 expression resulted in extended survival Chemokine ligands can exert a variety of effects on cellular pro- 31,32 in vivo. To assess the levels of CCR1 in clinical samples, IHC of liferation, angiogenesis, and metastasis . CCL6 and its functional a tissue microarray showed that CCR1 was highly expressed in human homolog CCL23 are predominantly expressed by macro- epithelial ovarian cancers across different subtypes, including phages, eosinophils, and neutrophils . The expression of these serous, endometrioid, and clear cell histologies. Interestingly, chemokine ligands can be upregulated during cancer, peritonitis, and 34–36 borderline ovarian tumors, which typically have better prog- chronic rhinosinusitis . CCL6 is a known chemoattractant for nosis compared to invasive ovarian cancer, did not show any macrophages, and to a lesser extent, B cells, T-helper cells, and 37,38 significant expression of CCR1 (Fig. 5g). Based on gene eosinophils . It is able to promote innate immune response by NK expression data from a cohort of ovarian cancer datasets, cell activation in animal models of peritonitis .IL-4, aknown sti- including TCGA , patients with high CCR1 expression had a mulant for an immune-suppressive-M2-macrophage subtype, indu- shortened disease-free survival compared to patients with low ces secretion of CCL6 by macrophages . Similarly, CCL23 is IL-4- CCR1 expressing tumors. CCR1 expression levels in stage I and inducible in a STAT6-dependent manner . In a mouse model of II ovarian cancer patients showed a significant difference in lung cancer, overexpression of CCL6 accelerated tumor growth, and median progression free survival of 37 months with high CCR1 increased metastatic spread .Similarly,higherlevels of CCL23 and compared to 96 months with low CCR1 expression (HR 2.13; its corresponding receptor CCR1 correlated with reduced metastasis- p = 0.032) (Fig. 5h, i). These findings lend further support to free survival of patients with breast cancer . the important effect of CCR1 on the oncological outcome of The omental and intraperitoneal microenvironment, including ovarian cancer patients. ascites, contains a variety of chemokine ligands. CCL8 (MCP2), CCL13 (MCP4), CCL14 (HCC1), CCL15 (MIP5), and CCL16 (HCC4) bind to CCR1 albeit with different affinity and could, Discussion therefore, potentially promote colonization . However, our The current study, to our knowledge, identifies a novel mecha- experiments in omentum-macrophage-conditioned media nism of ovarian cancer colonization of the omentum mediated underscore the importance of murine CCL6 and human CCL23. by chemokine ligands secreted by omentum resident macro- Neutralization of CCL6 and CCL23 in these complex environ- phages. We demonstrate that macrophage secreted CCL6 or ments with a large variety of other chemokine ligands completely CCL23 promotes colonization of ovarian cancer cells to abrogated the migration of ovarian cancer cells. We hence pro- omental milky spots. Targeting of CCR1 in ovarian cancer pose that the effects in vivo are mainly mediated by CCL6 and cells leads to a significant reductioninmigration andomental CCL23. Conversely, other chemokine receptors may play a role in metastasis demonstrating the importance of CCR1 in ovarian promoting ovarian cancer cell metastasis to the omentum. CCL6, cancer cell metastasis. for example, can bind to CCR3 which has been implicated in The role of macrophages or other immune-effector cells in metastasis . Neither ID8 nor SKOV3ip.1 cells used in our study ovarian cancer metastasis is still not well understood. We have expressed CCR3, and we, therefore, excluded its role in the previously demonstrated that ovarian cancer cells metastasize observed reduction of metastasis in CCR1 deleted cells. However, initially to milky spots in the omentum but not other intra- CCR3 is expressed in human ovarian cancer and might, therefore, abdominal adipose tissues including the mesentery that lack 21 be another important receptor to promote metastasis in milky spots . This pattern of early omental colonization was patients . independent of the presence of B cells, T cells or NK cells based COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 7 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z a b Mouse Omental Metastasis Human Omental Metastasis CCR1 CCR1 CCR1 TROMA EpCAM EpCAM CCR1 TROMA DAPI CCR1 EpCAM DAPI DAPI CCR1 EpCAM DAPI DAPI DAPI SF c d SF 1 ng/mL 1ng/mL 10 ng/mL 10ng/mL 250 25 ng/mL 25ng/mL 50 ng/mL 50ng/mL 100 ng/mL 100ng/mL 200 **** **** **** **** 200ng/mL 200 ng/mL 10% FBS 10% FBS 150 **** **** **** **** *** ** 0 0 SKOV3ip.1 CRISPR CTRL SKOV3ip.1 CCR1 null ID8 CRISPR CTRL ID8 CCR1 null ID8 CRISPR CTRL SKOV3ip.1 CRISPR CTRL SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SKOV3ip.1 CCR1 null ID8 CCR1 null SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS e f SF SF 100 ng/mL CCL6 100 ng/mL CCL23 200 ng/mL CCL6 **** 200 ng/mL CCL23 **** *** *** 50 50 0 0 No Treatment UCB35625 No Treatment UCB35625 No treatment UCB35625 No treatment UCB35625 SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL Fig. 4 Genomic deletion and pharmacological inhibition of CCR1 in human and mouse ovarian cancer cells inhibits their migratory capacity toward CCL6/CCL23 in vitro. a Immunofluorescence for CCR1 (red), cytokeratin (green), and nuclei (blue, DAPI) on ID8 cells metastasized to omentum in C57BL/6 and athymic nude mice (scale bar = 20 µm). b Immunofluorescence for CCR1 (red), EPCAM (green), and nuclei (blue, DAPI) on omenta derived from patients with high-grade serous cancer (scale bar = 20 µm). c, d Transwell migration assay of ID8 and SKOV3ip.1 with genomic deletion of CCR1 toward CCL6 and CCL23. e, f Transwell migration assay of ID8 and SKOV3ip.1 toward CCL6 and CCL23 with pharmacological inhibition of CCR1 using a small molecule antagonist UCB35625 (100 nM). Data shown represent mean and s.d of technical replicates (n = 3) for each condition. Statistical significance was determined by ordinary two-way ANOVA analysis comparing each condition to serum free (SF) media (**p < 0.005; ****p < 0.0001). A role for CCR1 in cancer metastasis has been suggested in that enhanced metastasis . CCR1 activation enhanced the other studies, but not in the context of ovarian cancer or medi- interaction between “metastasis associated macrophages” (MAM) ated by omental macrophage-derived CCL6/CCL23. CCR1 and cancer cells in part through integrin α4 resulting in increased expression was found to increase in colon cancer cells during extravasation of cancer cells and metastasis . In a study by Zhu metastasis to the liver . In a separate study, CCR1 sustained liver et al., osteopontin (OPN) was found to upregulate CCR1 expre- metastasis by promoting local recruitment of bone marrow (BM)- ssion in hepato-cellular carcinomas. CCR1 knockdown resulted derived cells that secrete the MMP9 and MMP2 metalloprotei- in reduction of migration, invasion and pulmonary metastasis 45 48 nases required for tissue invasion . Activation of EGFR signaling induced by OPN in vitro and in vivo . via CCR1 has been shown to contribute to breast cancer invasion CCR1 promotes monocytic tissue infiltration and plays a major and metastasis . The CCR1 mediated interaction between mac- role in various other disease processes, including autoimmune rophages and breast cancer cells induced a chemokine cascade diseases like rheumatoid arthritis. Clinical trials targeting CCR1 8 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio C57BL/6 # Migrated ID8 cells/200X Field # Migrated ID8 cells/200X Field Athymic Nude # Migrated SKOV3ip.1 cells/200X Field # Migrated SKOV3ip.1 cells/200X Field Human Omentum COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Euthanize Euthanize (n=9/group), (n=9/group), ID8-mRuby Flow cytometry Flow cytometry (1x10 Cells; i.p.) & FFPE Tissues & FFPE Tissues Day 7 Day 0 Day 1 b d 24 Hours Post Intraperitoneal Injection of ID8 in C57BL/6 DAB Quantitation Flow Cytometry Quantitation Naive Omentum ID8 Parental ID8 CCR1 null 24 Hours Post Injection ** ID8 Parental ID8 CCR1 null ID8 Parental ID8 CCR1 null 7 Days Post Intraperitoneal Injection of ID8 in C57BL/6 DAB Quantitation Flow Cytometry Quantitation Naive Omentum ID8 Parental ID8 CCR1 null 7 Days Post Injection 15 40 0 0 ID8 Parental ID8 CCR1 null ID8 Parental ID8 CCR1 null e g ID8-Parental ID8-CRISPR-CTRL **** p<0.0001 **** **** **** ID8-CCR1-null BL vs. CC BL vs. EC *** BL vs. SC 0 50 100 150 200 Days elapsed ID8 CCR1 null ID8 Parental ID8 CRISPR CTRL Histotype N=1104 N=99 All Stages Stages 1 and 2 Time (months) Time (months) for the treatment of rheumatoid arthritis and multiple sclerosis study, and safety and tolerability of CCR1 receptor antagonists in have so far shown that the targeting agents used are well tolerated human clinical trials provide a rationale for using CCR1 but have limited clinical efficacy . Targeting CCR1 in cancer antagonists as a new therapeutic target in ovarian and other patients to reduce metastasis or induce changes in the tumor cancers. microenvironment might lead to anti-tumor immune responses Various other cells types, such as neutrophils, mesothelial cells due to reduction of monocytic infiltration. The findings in our and adipocytes have been described to regulate ovarian cancer COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 9 BL (N=13) SC (N=82) EC (N=66) CC (N=30) Diaphragm Peritoneal Wall Mesentery Omentum Percent survival Progression Free Survival % of DAB Staining per Omentum % of DAB Staining per Omentum CCR1 Expression (%) Progression Free Survival % of ID8-mRuby Cells % of ID8-mRuby Cells ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z Fig. 5 Colonization of the omentum is disrupted by CCR1 knock-down in the cancer cells. a Schematic of experimental design for the 24hpi and 7dpi of ID8 in vivo metastasis assay. b Representative sections of omenta collected at 24 hpi and 7 dpi of stained for CK8/18 by IHC (scale bar = 100 µm). c Quantification of DAB staining area as an indicator of ID8 cancer cell burden in omental tissues. d Quantitation of flow cytometry analysis of ID8-mRUBY cells colonizing omentum at 24 hpi and 7 dpi (*p < 0.05). e Kaplan–Meier curve showing percent-survival after i.p injection of ID8- parental, CRISPR- Control, or CCR1-null cells (**p < 0.001) (n = 6/group). f Representative H&E stained peritoneal tissues at experimental end point. H&E images of ID8- CCR1-null group are from mice euthanized at 97 days, the experimental end point of the ID8-CRISPR-Control group. (scale bar = 100 µm). g Examples of 3 patient cores of endometrioid histology from a tissue microarray stained for CCR1 and IHC quantitation of CCR1 levels in a tissue microarray with different ovarian cancer histologies (****p < 0.0001). h, i Kaplan–Meier survival analysis (log-rank test) of CCR1 expression on progression free survival of serous ovarian cancer patients in h all stages and i stages I and II. 15,17,50–53 metastasis . Neutrophil Extracellular Traps (NET) for- expressing cells. All cells were maintained under standard tissue culture conditions (i.e., in a humidified incubator at 37 °C supplemented with 5% CO ). mation provides a premetastatic omental niche conducive for implantation of ovarian cancer cells. Pharmacological blockade of NET formation decreased omental colonization by ovarian cancer Patient samples. Omenta were obtained from patients under approved IRB protocol at the Dept. of Obstetrics and Gynecology, Stanford Hospital. Informed cells . A unique subset of omental tissue-resident macrophages consent was obtained from each patient prior to sample collection. Omental spe- + + (CD163 Tim4 ) can promote a premetastatic niche in the cimens with HGSC and omenta derived from patients with benign disease were omentum . Other studies have focused on the attachment of obtained and processed for histological analysis and in vitro assays as ovarian cancer cells to mesothelial cells that outline the peritoneal described below. cavity. This process is regulated by multiple adhesion molecules, 50,53 proteases and extracellular matrix components . Fibronectin Mice. All mice were housed, maintained, and euthanized according to APLAC receptors like α5β1‐integrins, for example, are expressed on the protocol and under the supervision of the Stanford Animal Resource Center. 6–8 weeks, female, inbred C57BL/6 (C57BL/6NHsd; immunocompetent) and nude surface of ovarian cancer cells and mediate the adhesion through (Athymic Nude-Foxn1nu; T-cell deficient) mice were obtained from Harlan myosin‐mediated traction forces . Furthermore, TGF‐β secreted Laboratories (Indianapolis, IN). All animal studies were performed in accordance by cancer cells activates a RAC1/SMAD‐mediated signaling with Stanford APLAC approved protocols. All mice were randomly assigned to pathway in mesothelial cells, which results in transcriptional appropriate treatment groups. upregulation of the fibronectin gene and induces an EMT‐like phenotype in mesothelial cells . The adhesion of cancer cells to In vivo experimental metastasis assays. Exponentially growing untagged or mRuby 3,51 tagged-ID8 parental, ID8-CCR1-null cells, ID8-CRISPR-control cells were trypsi- the mesothelium is followed by mesothelial clearance . In this nized and prepared as a single-cell suspension at a concentration of 2 × 10 cells/ process, mesothelial cells are displaced by ovarian cancer spher- mL in ice cold PBS. Animals were injected intraperitoneally with 500 μL of the cell oids followed by invasion of cancer cells into the deeper layer of suspension (1 × 10 cells). For all experiments, 500 μL of PBS was injected as a peritoneal tissues . Specifically, hypoxic signaling increased negative control in a parallel group of control mice. At the experimental endpoint (24 h or 7 dpi) of each assay, mice were euthanized in accordance with APLAC expression of lysyl oxidase (LOX) in mesothelial and ovarian protocol via CO asphyxiation and a secondary method of cervical dislocation was cancer cells to promote collagen crosslinking and tumor cell performed. For the time course experiment, grossly visible overt peritoneal invasion . Further, during ovarian cancer progression, omental metastases that were greater than 1 mm, from each mouse, were counted. Tissues adipocytes provide fatty acids as an energy source to ovarian were then harvested, processed, sectioned, and stained as described in “Methods”. cancer cells . Macrophage depletion using clodronate liposomes. For both pre-cancer and post- Based on our current work and the published literature, we cancer injection experiments, peritoneal macrophages were depleted by injection of propose a multi-step model for omental colonization. Mesothelial clodronate-containing liposomes (clodronateliposome.org). In vivo depletion of cells promote ovarian cancer cell adhesion to the omentum. macrophages was assessed for several dosages by cytology performed on omental Subsequent migration and homing to milky spots is induced by dissociated cells and by flow cytometry with F4/80 antibody. A dose of 100 μLof liposomes diluted in PBS was selected and administered i.p. to mice every alternate omental resident macrophages via chemokine ligands. The day. Control mice were administered either PBS as a control for unstimulated interaction between macrophages and cancer cells creates an macrophages or saline liposomes to control for any nonspecific effects of liposome inflammatory environment that recruits other immune-effector administration. Macrophage depletion was maintained during the experimental cells, including neutrophils. The formation of NETs further period. enhances colonization of ovarian cancer cells. Adipocytes then provide the energy required for the continuous growth of tumor Isolation of murine and human omental macrophages. Omenta excised from cells. Further studies are needed to explore this complex process C57BL/6 mice (n = 9) and processed as three separate groups (n = 3 per group). A and identify opportunities for therapeutic interventions. modified protocol of the adipose tissue dissociation kit (130–105–808, Miltenyi Biotech) was used to obtain a single cell suspension of the mouse omenta. Human omental samples were processed using the human tumor dissociation kit (130-095- 929, Miltenyi Biotech) to dissociate soft tumors into a single cell suspension. The Methods cells were ACK lysed and re-suspended in PBS containing 1% FBS for flow cyto- Cell lines. Human SKOV3.ip1 ovarian cancer cell lines were cultured and main- metry applications to analyze and sort for mouse or human omental macrophages. tained at 37 °C in Dulbecco’s Modified Eagle Medium (DMEM) (10-013-CV, Corning Cellgro) supplemented with 10% (v/v) fetal bovine serum (FBS) (S11150, Atlanta Biologicals, Lawrenceville, GA, USA) and 1% penicillin-streptomycin (Pen Antibodies and flow cytometry. The isolated mouse and human omental cells Strep) (v/v) (15240, Invitrogen). Similarly, the murine ID8 ovarian cancer cell line were re-suspended in PBS supplemented with 1% FBS. The mouse panel designed was cultured in DMEM supplemented with 4% FBS and 1% Penicillin Strep and 1% to sort for murine omental macrophages included CD45-FITC (clone-30-F11, Insulin-Transferrin-Selenium (25-800-CR, Corning). RAW 264.7 cells and human eBioscience), CD11b-APC (clone-M1/70, eBioscience), F4/80-PE (clone-BM8, peripheral blood mononuclear cells (hPBMCs) were cultured in 10% RPMI and eBioscience). Similarly, the human panel designed to sort for human omental 10% FBS and 1% penicillin-streptomycin]. ID8 or SKOV3ip.1 ovarian cancer cells macrophages included CD45-FITC (clone-HI30, eBiosciences), CD14-PE (clone- that stably express mRuby (ID8-mRuby/SKOV3ip.1-mRuby) were constructed by 61D3, eBioscience), CD68-PECy7 (Clone-eBioY1/82A, eBioscience). In both the lentiviral delivery of pLVX-mRuby expression vector (Stanford genomics facility). mouse and human flow panel, aqua amine (L34957, Fisher Scientific) was used as The mRuby tagged expression vector along with the packaging vector master mix the Live/Dead stain and Compensation Beads (01-1111, eBioscience) were used for PMDL(PVSV-G) and PSPAX2 lentiviral expression plasmid was transfected into compensation controls. Labeled cells were sorted by flow cytometry on a BD HEK293T cells to generate the viral conditioned medium. Fluorescence-activated InfluxFlow Sorter or analysis was performed on BD LSR II at Stanford shared cell sorting using a BD FACS Aria II system (BD Biosciences, San Jose, CA) at the FACS facility. For the RNA-seq analysis, CD45+CD11b+F4/80+ cells were sorted Stanford Flow Cytometry Core Facility was used to select for high mRuby- into buffer RLT and were processed immediately for RNA isolation. 10 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE RNA sequencing and analyses. For each sample in the whole transcriptome sample in triplicate. Cells were incubated for a period of 16 h overnight in a 37 °C sequencing library, 43–66 million 75-basepair paired-end reads were acquired from incubator (5% CO ). After incubation, the top and the bottom chambers were the sequencer. Read quality was determined with FastQC 0.11.4. The reads were washed with PBS and cells were fixed with 4% paraformaldehyde for 10 min. After aligned to the mouse reference genome (NCBI37/mm9) using STAR 2.4.2a, with washing with PBS, the cells were stained with 0.05% Crystal Violet for 40 min. Both splice junctions defined by the GTF file (UCSC). On average, 93% of reads were the chambers were then washed with tap water and the cells on the top of the insert aligned to the reference genome, and 87% of reads uniquely aligned to the reference were gently scraped with a cotton swab. The positive-staining cells adherent to the genome. Gene expression in raw count was determined by STAR with quant mode, bottom of the transwell, were examined under the microscope. Data are shown as GeneCounts setting. Cufflinks 2.2.1 was used to normalize gene expression in mean ± s.d. from three independent experiments. fragments per kilobase per million aligned reads (FPKM). Further, differential expression between the conditions was evaluated using DEseq2 1.12.4. The fold Protein isolation and western blotting. Protein was extracted from ID8 and change between each sample is calculated using DEseq2. The genes that were either SKOV3ip.1 cells treated with CCL6/CCL23 at different time points by RIPA lysis upregulated 2-fold or downregulated 2-fold were selected for gene set enrichment extraction method (pH 7.4) containing Halt protease inhibitor cocktail (#1861278, analysis using DAVID (https://david.ncifcrf.gov/) with gene set annotation GO BP Thermofisher Scientific) and EDTA (0.5%) under agitation at 4 °C for 30 min. A and KEGG database. The top 10 gene sets are only used for visualization. bicinchoninic acid assay (Pierce, Rockford, IL, USA) was used to determine total protein concentration. Equal amounts of protein were separated on (4–12%) SDS- PAGE and transferred to nitrocellulose membranes (Sigma, St Louis, MO, USA). CRISPR/Cas9 genome editing. Human (AACTTGTAGTCGATCCAGAA) and murine (GAACACTAGAGAATACAGG) CCR1 sgRNA sequences were identified Membranes were blocked in 5% low-fat dry milk in TBS-T for 1-h at room tem- perature and then incubated overnight at 4 °C. Dilutions for the primary antibodies using the Broad Institute Genetic Perturbation Platform Web Portal’s sgRNA design tool . Control non-targeting sequences for human (ACGGAGGCTAA were used according to the manufacturer’s recommendation. The primary anti- bodies used for this study are listed as follows: anti-pERK1/2 (p44/42) (#9101S, Cell GCGTCGCAA) and mouse (GCGAGGTATTCGGCTCCGCG) were taken from the non-targeting sequences in the human and mouse GeCKOv2 libraries . Oli- Signaling Technologies),anti-total ERK1/2 (#9102S, Cell Signaling Technologies), anti-pAKT (Ser473) (#4058S, Cell Signaling Technologies),anti-total AKT(#9272S, goduplexes were cloned into pLentiCRISPRv2 (a gift from Feng Zhang; Addgene Cell Signaling Technologies), anti-pLIMK (#ab194798,Abcam), anti-p-Cofilin (Ser3) plasmid 52961 ) and transformed to Stbl3 bacteria. pLentiCRISPRv2 vectors (#3313L, Cell Signaling Technologies), anti-Cofilin (#5175S, Cell Signaling Tech- containing the sgRNA guides were transfected with pMD2.G and pSPAX2 (gifts nologies), anti-alpha tubulin (#3873S, Cell Signaling Technologies). Following day, from Didier Trono; Addgene plasmids 12259 and 12260) to 293T cells using Fugene (Roche) in order to produce lentiviral vectors. After 24 h of culture, after three washes for 10 min in TBS-T, the membrane was incubated with IRDye- 680 nm conjugated-goat anti-rabbit (#925-32211, LICOR) and IRDye-800nm con- supernatants were collected, filtered through a 0.45 μm filter and added to monolayers of either ID8 (mouse) or SKOV3ip.1 (human) cell lines. Transduced jugated-donkey anti-mouse (#926-32212, LICOR) secondary antibodies at 1:5000 for 1-h at room temperature. Membranes were washed for 10 min in TBS-T before ID8/SKOV3ip.1 cells were then selected in medium containing 0.6/0.4 µg/mL puromycin for 1 week and individual clones were selected by limiting dilution and being imaged by the LICOR Odyssey Clx imaging system. The intensity of the protein bands was semi-quantified using the Image studio software (Windows then expanded in puromycin containing media. The targeted genomic regions were amplified by PCR (mouse CCR1 forward: 5′- application VIO.02), with normalization of each protein against alpha tubulin. TTCCACTGCTTCAGGCTCTT-3′ and reverse: 5′-TCATTTGTTCTGTTCCC CTCTT-3′ and human CCR1 forward: 5′-TACCAGCCCAAAGAGGTTCA-3′ and Phospho-protein array. Protein lysates isolated from ID8 and SKOV3ip.1 cells reverse: 5′-TGACACGACCACAGAGTTTGA-3′) and analyzed using TIDE-seq . stimulated with CCL6 or CCL23 at 30 min time points were processed and assayed To characterize the specific lesion in each cell line, PCR products were Topo cloned using the proteome-profiler human phospho-kinase array (#ARY003B, R&D Sys- (Life Technologies) and 10 colonies from each isolated cell line was sequenced. tems) according to manufactures protocol. Membranes were incubated with pro- tein lysates of ID8 (500 μg) or SKOV3ip.1 (250 μg) for each condition, respectively. The membranes were imaged by Chemiluminescence by using LICOR Odyssey Clx Cell proliferation assays. For cell proliferation assays, cells were seeded in 96-well imaging system. Protein bands were semi-quantified using the Image studio soft- plates (1.5 × 10 cells/well). At the indicated time points, cell proliferation was ware (Windows application VIO.02), using normalization controls as per the detected by MTS assay (Promega, G3530) according to the manufacturer’s manufacture’s protocol. instructions. Data are shown as mean ± s.d. from three independent experiments. Immunohistochemistry. Slides were first deparaffinized with xylene treatment for Real-time RT–PCR. Total RNA was isolated with the RNeasy kit (Qiagen) and 60 min, 30 min, and 10 min three times each. Slides were then rehydrated, and reverse-transcribed into cDNA using the Superscript Strand III cDNA SuperMix antigen retrieval was performed in 1× citrate buffer (pH 6.0) for 5 min in a con- (Invitrogen). Real-time PCR was performed using SSO Fast Evergreen Supermix ventional steamer. Slides were incubated in a blocking buffer solution (Protein block (Biorad) on an Applied Biosystems 7500 Fast Real-Time PCR System at Functional serum-free ready-to-use, DAKO) for 60 min at room temperature. After standard Genomics Facility at Stanford. Expression of the target genes were normalized to slide preparation as described above, slides were incubated with primary antibody the housekeeping gene 18S. Mouse RAW 264.7 cells and human PBMC were used overnight and subsequent secondary antibodies (Supplementary Table 1). After as a positive control for the all the gene expression studies. washing, slides were either developed for immunohistochemistry or incubated with The target gene-specific primers were obtained from Integrated DNA Technologies DAPI and imaged using an EVOS Fluorescence microscope. Level of CCR1 on (IDT). The primer sequences used in this study are shown in Supplementary Table 2. patient tissue core samples on a tissue microarray was determined by HALO image All real-time RT-PCRs were performed in triplicates and the relative mRNA −ΔCT analysis software area quantification algorithm (Indica Labs, Corrales, NM). expression of each target gene was determined by using the formula 2 (C ,cycle threshold) where ΔC = C (target gene)− C (18S). The comparative expression T T T −ΔΔCT level of each target gene between different samples was 2 . Statistics and reproducibility. All grouped data are presented as mean ± s.e.m. Significance between groups was analyzed by one-way ANOVA or Student t-test using GraphPad Prism. KM plotter (https://kmplot.com/analysis/) was used to Generation of tissue and cell conditioned media. Three freshly excised mouse analyze the prognosis of CCR1 in ovarian cancer. The analysis was conducted for omenta were pooled and processed to produce a single cell suspension with all serous samples with an ‘auto-select best cutoff’ feature. The data with statistical minimal cell loss. Gonadal fat was similarly processed in parallel. Macrophages analyses were performed with biological replicates on different samples of the same (CD11b cells) were isolated from stromal and vascular cells (SVC) by positive cell type on different days, yielding similar results. The number of biological selection using CD11b microbeads (Miltenyi Biotech) (Supplementary Fig. 2). To replicates is included in the figure legends. Human CCL23 ELISA data shown in assess the functional phenotype, 1 × 10 cells were plated and allowed to condition Fig. 2i is plotted as separate data points to show inter-patient variability. serum-free media for 24 h after which the media were used in standard transwell migration assay. Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Transwell migration assays Toward chemokines CCL6/CCL23. For cell migration assays, 1.5 × 10 (ID8) or 3 × 10 (SKOV3ip.1) cells in serum-free media were plated in the top chamber of 8.0- Data availability μm 12-well plate transwell insert (12-well format, 8-μm pore; BD PharMingen). Raw data and analyzed data for the RNA-sequencing and microarray gene expression CCL6 (250-06, Peprotech) CCL23 (300-29, Peprotech), or omental conditioned analysis can be accessed from Gene Expression Omnibus using the accession number medium was added to the bottom well of the plate (35-3503, BD Falcon) as GSE153685 and GSE153790, respectively. Source data are available in Supplementary chemoattractant. Anti-CCL6 (MAB000487, R&D systems) and anti-CCL23 Data 1. All other data are available from authors on reasonable request. (MAB371, R&D systems) antibodies were added to omentum-conditioned media in neutralization experiments. For pharmacological Inhibition, CCR1 inhibitor— UCB35625 (2757, Tocris), PI3K inhibitor—BEZ235 (S1009, Selleckchem), MEK Received: 2 March 2020; Accepted: 6 August 2020; inhibitor—trametinib (S2673, Selleckchem) and myosin inhibitor—blebbistatin (1760, Tocris) were used for the migration assays. Analysis was performed on each COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 11 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z References 32. Yi, F., Jaffe, R. & Prochownik, E. V. The CCL6 chemokine is differentially 1. Bray, F. et al. Global cancer statistics 2018: GLOBOCAN estimates of regulated by c-Myc and L-Myc, and promotes tumorigenesis and metastasis. incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer Cancer Res. 63, 2923–2932 (2003). J. Clin. 68, 394–424 (2018). 33. Heng, T. S. & Painter, M. W., Immunological Genome Project, C. The 2. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2020. CA Cancer J. Immunological Genome Project: networks of gene expression in immune cells. Clin. 70,7–30 (2020). Nat. Immunol. 9, 1091–1094 (2008). 3. Yeung, T. L. et al. Cellular and molecular processes in ovarian cancer 34. LaFleur, A. M., Lukacs, N. W., Kunkel, S. L. & Matsukawa, A. Role of CC metastasis. A review in the theme: cell and molecular processes in cancer chemokine CCL6/C10 as a monocyte chemoattractant in a murine acute metastasis. Am. J. Physiol. Cell Physiol. 309, C444–C456 (2015). peritonitis. Mediators Inflamm. 13, 349–355 (2004). 4. Lambert, A. W., Pattabiraman, D. R. & Weinberg, R. A. Emerging biological 35. Nagarsheth, N., Wicha, M. S. & Zou, W. Chemokines in the cancer principles of metastasis. Cell 168, 670–691 (2017). microenvironment and their relevance in cancer immunotherapy. Nat. Rev. 5. Gerber, S. A. et al. Preferential attachment of peritoneal tumor metastases to Immunol. 17, 559–572 (2017). omental immune aggregates and possible role of a unique vascular 36. Poposki, J. A. et al. Increased expression of the chemokine CCL23 in microenvironment in metastatic survival and growth. Am. J. Pathol. 169, eosinophilic chronic rhinosinusitis with nasal polyps. J. Allergy Clin. Immunol. 1739–1752 (2006). 128,73–81 e74 (2011). 6. Pearce, O. M. T. et al. Deconstruction of a metastatic tumor 37. Asensio, V. C. et al. C10 is a novel chemokine expressed in experimental microenvironment reveals a common matrix response in human cancers. inflammatory demyelinating disorders that promotes recruitment of macrophages Cancer Disco. 8, 304–319 (2018). to the central nervous system. Am.J.Pathol. 154,1181–1191 (1999). 7. Nieman, K. M. et al. Adipocytes promote ovarian cancer metastasis and 38. Berger, M. S. et al. The chemokine C10: immunological and functional provide energy for rapid tumor growth. Nat. Med. 17, 1498–1503 (2011). analysis of the sequence encoded by the novel second exon. Cytokine 8, 8. Khan, S., Taylor, J. L. & Rinker-Schaeffer, C. W. Disrupting ovarian cancer 439–447 (1996). metastatic colonization: insights from metastasis suppressor studies. J. Oncol. 39. Coelho, A. L. et al. The chemokine CCL6 promotes innate immunity via 2010, 286925 (2010). immune cell activation and recruitment. J. Immunol. 179, 5474–5482 (2007). 9. Meza-Perez, S. & Randall, T. D. Immunological functions of the omentum. 40. Orlofsky, A., Lin, E. Y. & Prystowsky, M. B. Selective induction of the beta Trends Immunol. 38, 526–536 (2017). chemokine C10 by IL-4 in mouse macrophages. J. Immunol. 152, 5084–5091 10. Krist, L. F. et al. Cellular composition of milky spots in the human greater (1994). omentum: an immunochemical and ultrastructural study. Anat. Rec. 241, 41. Novak, H. et al. CCL23 expression is induced by IL-4 in a STAT6-dependent 163–174 (1995). fashion. J. Immunol. 178, 4335–4341 (2007). 11. Liu, J., Geng, X. & Li, Y. Milky spots: omental functional units and hotbeds for 42. Yan, H. H. et al. CCL9 induced by TGFbeta signaling in myeloid cells peritoneal cancer metastasis. Tumour Biol. 37, 5715–5726 (2016). enhances tumor cell survival in the premetastatic organ. Cancer Res. 75, 12. Mebius, R. E. Lymphoid organs for peritoneal cavity immune response: milky 5283–5298 (2015). spots. Immunity 30, 670–672 (2009). 43. Long, H. et al. Autocrine CCL5 signaling promotes invasion and migration of 13. Rangel-Moreno, J. et al. Omental milky spots develop in the absence of CD133+ ovarian cancer stem-like cells via NF-kappaB-mediated MMP-9 lymphoid tissue-inducer cells and support B and T cell responses to peritoneal upregulation. Stem Cells 30, 2309–2319 (2012). antigens. Immunity 30, 731–743 (2009). 44. Akram, I. G., Georges, R., Hielscher, T., Adwan, H. & Berger, M. R. The 14. Cleypool, C. G. J., Schurink, B., van der Horst, D. E. M. & Bleys, R. chemokines CCR1 and CCRL2 have a role in colorectal cancer liver Sympathetic nerve tissue in milky spots of the human greater omentum. J metastasis. Tumour Biol. 37, 2461–2471 (2016). Anat. 236, 156–164 (2019). 45. Hirai, H. et al. CCR1-mediated accumulation of myeloid cells in the liver 15. Lee, W. et al. Neutrophils facilitate ovarian cancer premetastatic niche microenvironment promoting mouse colon cancer metastasis. Clin. Exp. formation in the omentum. J. Exp. Med. 216, 176–194 (2019). Metastasis 31, 977–989 (2014). 16. Shimotsuma, M., Takahashi, T., Kawata, M. & Dux, K. Cellular subsets of the 46. Shin, S. Y. et al. C-C motif chemokine receptor 1 (CCR1) is a target of the milky spots in the human greater omentum. Cell Tissue Res. 264, 599–601 (1991). EGF-AKT-mTOR-STAT3 signaling axis in breast cancer cells. Oncotarget 8, 17. Etzerodt, A., et al. Tissue-resident macrophages in omentum promote 94591–94605 (2017). metastatic spread of ovarian cancer. J. Exp. Med. 217, e20191869 (2020). 47. Qian, B. Z. et al. CCL2 recruits inflammatory monocytes to facilitate breast- 18. Linde, N. et al. Macrophages orchestrate breast cancer early dissemination and tumour metastasis. Nature 475, 222–225 (2011). metastasis. Nat. Commun. 9, 21 (2018). 48. Zhu, Y. et al. C-C chemokine receptor type 1 mediates osteopontin-promoted 19. Kitamura, T. et al. CCL2-induced chemokine cascade promotes breast cancer metastasis in hepatocellular carcinoma. Cancer Sci. 109, 710–723 (2018). metastasis by enhancing retention of metastasis-associated macrophages. J. 49. Gladue, R. P., Brown, M. F. & Zwillich, S. H. CCR1 antagonists: what have we Exp. Med 212, 1043–1059 (2015). learned from clinical trials. Curr. Top. Med. Chem. 10, 1268–1277 (2010). 20. Robinson-Smith, T. M. et al. Macrophages mediate inflammation-enhanced 50. Kenny, H. A. et al. Mesothelial cells promote early ovarian cancer metastasis metastasis of ovarian tumors in mice. Cancer Res. 67, 5708–5716 (2007). through fibronectin secretion. J. Clin. Invest. 124, 4614–4628 (2014). 21. Clark, R. et al. Milky spots promote ovarian cancer metastatic colonization of 51. Kenny, H. A., Kaur, S., Coussens, L. M. & Lengyel, E. The initial steps of peritoneal adipose in experimental models. Am. J. Pathol. 183,576–591 (2013). ovarian cancer cell metastasis are mediated by MMP-2 cleavage of vitronectin 22. Orlofsky, A., Berger, M. S. & Prystowsky, M. B. Novel expression pattern of a and fibronectin. J. Clin. Invest. 118, 1367–1379 (2008). new member of the MIP-1 family of cytokine-like genes. Cell Regul. 2, 52. Natarajan, S. et al. Collagen remodeling in the hypoxic tumor-mesothelial 403–412 (1991). niche promotes ovarian cancer metastasis. Cancer Res. 79, 2271–2284 (2019). 23. Shay, T. & Kang, J. Immunological Genome Project and systems immunology. 53. Niedbala, M. J., Crickard, K. & Bernacki, R. J. Interactions of human ovarian Trends Immunol. 34, 602–609 (2013). tumor cells with human mesothelial cells grown on extracellular matrix. An 24. Ma, B. et al. The C10/CCL6 chemokine and CCR1 play critical roles in the in vitro model system for studying tumor cell adhesion and invasion. Exp. Cell pathogenesis of IL-13-induced inflammation and remodeling. J. Immunol. Res. 160, 499–513 (1985). 172, 1872–1881 (2004). 54. Iwanicki, M. P. et al. Ovarian cancer spheroids use myosin-generated force to 25. Klemke, M., Kramer, E., Konstandin, M. H., Wabnitz, G. H. & Samstag, Y. An clear the mesothelium. Cancer Disco. 1, 144–157 (2011). MEK-cofilin signalling module controls migration of human T cells in 3D but 55. Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize not 2D environments. EMBO J. 29, 2915–2929 (2010). off-target effects of CRISPR-Cas9. Nat. Biotechnol. 34,184–191 (2016). 26. Makowska, K. A., Hughes, R. E., White, K. J., Wells, C. M. & Peckham, M. 56. Sanjana, N. E., Shalem, O. & Zhang, F. Improved vectors and genome-wide Specific myosins control actin organization, cell morphology, and migration in libraries for CRISPR screening. Nat. Methods 11, 783–784 (2014). prostate cancer cells. Cell Rep. 13, 2118–2125 (2015). 57. Brinkman, E. K., Chen, T., Amendola, M. & van Steensel, B. Easy quantitative 27. Wang, W., Eddy, R. & Condeelis, J. The cofilin pathway in breast cancer assessment of genome editing by sequence trace decomposition. Nucleic Acids invasion and metastasis. Nat. Rev. Cancer 7, 429–440 (2007). Res. 42, e168 (2014). 28. Yamaguchi, H. & Condeelis, J. Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochim Biophys. Acta 1773, 642–652 (2007). 29. Gyorffy, B., Lanczky, A. & Szallasi, Z. Implementing an online tool for genome- Acknowledgements wide validation of survival-associated biomarkers in ovarian-cancer using We thank Drs. Roger Pederson (Stanford) and Mickey Hu (Panorama Research Institute) microarray data from 1287 patients. Endocr. Relat. Cancer 19, 197–208 (2012). for critical reading of the paper. This work was supported by the Mary Lake Polan 30. Gosselin et al. Environment drives selection and function of enhancers Gynecologic Oncology Endowment for Research (O.D.), the Vivian Scott Fellowship in controlling tissue-specific macrophage identities. Cell 159, 1327–1340 (2014). Gynecologic Oncology (O.D.), the Dean Pizzo Stanford Cancer Center Research Award 31. Sarvaiya, P. J., Guo, D., Ulasov, I., Gabikian, P. & Lesniak, M. S. Chemokines (O.D.), the Ovarian Cancer Research Fund Alliance—Ann and Sol Schreiber Mentored in tumor progression and metastasis. Oncotarget 4, 2171–2185 (2013). Investigator Award (V.K.), and generous donation of Tell Every Amazing Lady About 12 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Ovarian Cancer Louisa McGregor Ovarian Cancer Foundation, also known as T.E.A.L. Reprints and permission information is available at http://www.nature.com/reprints (V.K.), Stanford Medical Scholars Fellowship Program (A.M.C.). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author contributions V.K. and O.D. designed experiments; B.S. designed CRISPR/Cas9 experiments; V.K., S.T., K.K., and A.M.C. carried out and analyzed experiments; V.K. and S.P. analyzed histology Open Access This article is licensed under a Creative Commons specimens; Y.Z. analyzed RNA-seq data set; V.P. and A.F.K. provided clinical samples Attribution 4.0 International License, which permits use, sharing, from early and late stage of ovarian cancer disease; V.K. prepared human tissue samples; adaptation, distribution and reproduction in any medium or format, as long as you give V.K. prepared the figures wrote the initial draft of the paper; all authors edited the paper; appropriate credit to the original author(s) and the source, provide a link to the Creative C.R.S., E.B.R., and O.D. provided supervision. Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless Competing interests indicated otherwise in a credit line to the material. If material is not included in the The authors declare no competing interests. article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from Additional information the copyright holder. To view a copy of this license, visit http://creativecommons.org/ Supplementary information is available for this paper at https://doi.org/10.1038/s42003- licenses/by/4.0/. 020-01246-z. © The Author(s) 2020 Correspondence and requests for materials should be addressed to O.D. COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 13 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Communications Biology Springer Journals

Omental macrophages secrete chemokine ligands that promote ovarian cancer colonization of the omentum via CCR1

Loading next page...
 
/lp/springer-journals/omental-macrophages-secrete-chemokine-ligands-that-promote-ovarian-vAW3Q5Tg0R

References (71)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
eISSN
2399-3642
DOI
10.1038/s42003-020-01246-z
Publisher site
See Article on Publisher Site

Abstract

ARTICLE https://doi.org/10.1038/s42003-020-01246-z OPEN Omental macrophages secrete chemokine ligands that promote ovarian cancer colonization of the omentum via CCR1 1 1 2 1 3 4 Venkatesh Krishnan , Supreeti Tallapragada , Bruce Schaar , Kalika Kamat , Anita M. Chanana , Yue Zhang , 5 6 7 8 1,9 1,10 Sonia Patel , Vinita Parkash , Carrie Rinker-Schaeffer , Ann K. Folkins , Erinn B. Rankin & Oliver Dorigo The omentum is the most common site of ovarian cancer metastasis. Immune cell clusters called milky spots are found throughout the omentum. It is however unknown if these immune cells contribute to ovarian cancer metastasis. Here we report that omental macro- phages promote the migration and colonization of ovarian cancer cells to the omentum through the secretion of chemokine ligands that interact with chemokine receptor 1 (CCR1). We found that depletion of macrophages reduces ovarian cancer colonization of the omentum. RNA-sequencing of macrophages isolated from mouse omentum and mesenteric adipose tissue revealed a specific enrichment of chemokine ligand CCL6 in omental mac- rophages. CCL6 and the human homolog CCL23 were both necessary and sufficient to promote ovarian cancer migration by activating ERK1/2 and PI3K pathways. Importantly, inhibition of CCR1 reduced ovarian cancer colonization. These findings demonstrate a critical mechanism of omental macrophage induced colonization by ovarian cancer cells via CCR1 signaling. Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, Stanford School of Medicine, Stanford Cancer Institute, Stanford, CA, USA. 2 3 Genome Technology Center, Stanford, CA, USA. Department of Obstetrics and Gynecology, David Geffen School of Medicine, University of California, Los 4 5 Angeles, CA, USA. Department of Physiology and Chemistry, Genentech, San Francisco, CA, USA. Department of Thoracic Head and Neck Medical Oncology, Division of Cancer Medicine, University of Texas Health Science Center at Houston, Houston, TX, USA. Department of Pathology, Yale School of Medicine and Yale School of Public Health, New Haven, CT, USA. Department of Surgery, Division of Urology, University of Chicago, Chicago, IL, USA. 8 9 Department of Pathology, Stanford School of Medicine, Stanford, CA, USA. Department of Radiation Oncology, Stanford School of Medicine, Stanford, CA, 10 ✉ USA. Stanford Women’s Cancer Center, Stanford, CA, USA. email: odorigo@stanford.edu COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 1 1234567890():,; ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z varian cancer is the seventh most common cancer in To investigate the role of tissue-resident macrophages in women with 300,000 new cases and 185,000 deaths metastatic colonization to the omentum, we first verified the Oworldwide in 2018 . In 2020, an estimated 21,750 new presence of macrophages in milky spots in naive human and cases of ovarian cancer will be diagnosed in the USA and 13,940 murine omentum (Fig. 1b, c). Interestingly, in omenta from women will die from the disease . Ovarian cancer originates in ovarian cancer patients without gross disease, microscopic the ovary or fallopian tube and metastasizes throughout the metastases were restricted to milky spots and were not observed peritoneal cavity . Metastatic disease, rather than the primary in the adipose tissue at this early stage (Fig. 1b). Similarly, tumor, is the cause of ovarian cancer-related deaths . However, microscopic metastases in murine omentum were localized and the fundamental mechanisms that support intraperitoneal confined to the milky spots at early timepoints (24 h) after metastasis are not well understood. intraperitoneal (i.p.) injection of ID8 cells (Fig. 1c). The omentum is a preferred metastatic site for ovarian cancer We next studied the dynamics of intraperitoneal metastasis in cells. The majority of women diagnosed with high grade serous our murine ovarian cancer model. ID8 murine ovarian cancer 5–7 ovarian cancer present with omental metastases . In experi- cells were injected i.p. into C57BL/6 mice and the anatomic mental models of ovarian cancer, ovarian cancer cells metastasize distribution of metastatic colonization was followed over time. to the omentum within 1–6 h of injection . The omentum is an Interestingly, the omentum was the first site of macroscopic adipose tissue that contains secondary lymphoid structures with a metastases within one week of injection and remained the only variety of cell types, including adipocytes, blood vessels, and site of peritoneal metastasis for 6 weeks after which metastases 9–15 clusters of leukocytes known as milky spots . Milky spots appeared at other sites (Fig. 1d). This is similar to observations in contain macrophages, B, T, and NK cells, which rapidly efflux patients with ovarian cancer who frequently present with omental 10,16 into the peritoneum during pathogenic challenge . While metastasis upon diagnosis but may not have other sites of 5,6 various cell types, including neutrophils, mesothelial cells, and significant disease burden . Milky spots increased in number adipocytes, have been shown to support ovarian cancer metastasis and size after cancer cell colonization at week 1 post injection. to the omentum, little is known regarding the role of tissue- The number of immune cells, in particular macrophages, was 7,15,17 resident immune cells in omental metastasis . found to increase by week 2. We observed continued expansion of Macrophages enhance metastasis in several cancer models. In tumor from milky spots into the adipose tissue of the omentum breast cancer, macrophages facilitate the early hematogenous following early colonization (Fig. 1e). These findings suggest that dissemination of cancer cells from the primary tumor by altering the immune cell-rich milky spots are sites of early metastasis of vascular permeability and facilitating migration into the blood- ovarian cancer cells within the omentum. stream . Chemokines enhance the interactions between cancer cells and macrophages in a chemokine cascade that promotes Macrophages promote ovarian cancer colonization of the breast cancer metastasis . In ovarian cancer, macrophage-driven omentum. Previous studies have demonstrated in both murine inflammation can promote intraperitoneal metastasis and ascites (ID8) and human ovarian cancer models (CaOV3, HeyA8, or production . Specific subsets of omental macrophages might be SKOV3ip.1) that colonization of the omentum was independent involved in promoting metastasis to the omentum . of B-, T-, and NK- cells . To determine the role of macrophages In this study, we describe the role of omental macrophage- in omental colonization, we depleted macrophages in C57BL/6 derived chemokine ligands in promoting ovarian cancer cell and athymic nude mice using clodronate liposomes prior to i.p. colonization of the omentum. Gene expression profiling of injection of mouse ID8 or human SKOV3ip.1 ovarian cancer mouse omental macrophages compared to mesenteric adipose cells, respectively (Fig. 2a). Strikingly, we observed an 11-fold macrophages shows distinct gene expression patterns with reduction in the ovarian cancer burden in the omentum of significant upregulation of the chemokine ligand CCL6 prior to macrophage depleted mice at 7 days following i.p. injection of and during colonization of the omentum by ovarian cancer tumor cells (Fig. 2b). Interestingly, when macrophages were cells. CCL6 induces the migration of ovarian cancer cells as an depleted using clodronate liposome three days after injection of important step in colonization via the chemokine receptor SKOV3ip.1 cells, we failed to observe a decrease in omental CCR1. Furthermore, inactivation of CCR1 in mouse ovarian colonization. (Supplementary Fig. 1). These findings support our cancer cells blocks CLL6 induced migration and abolishes their hypothesis that macrophages are crucial for the initiation of ability to colonize the omentum. In ovarian cancer patients, the ovarian cancer colonization of the omentum. expression of CCR1 is associated with a poorer prognosis. In addition, CCL23, the human homolog of CCL6, is expressed on macrophages isolated from human omentum. These observa- Omental macrophage-derived conditioned media promote tions validate the importance of CCR1 and its ligand CCL23 in migration of ovarian cancer cells. Our findings above indicate an human ovarian cancer. important role for omental macrophages in metastatic coloniza- tion of ovarian cancer. Therefore, we investigated whether omental macrophages can induce migration of ovarian cancer Results cells as an important aspect of metastatic colonization. For this Omentum as a niche for ovarian cancer metastasis. The purpose, CD11b cells were isolated from normal mouse omenta omental vasculature is characterized by the presence of numerous and cultured to generate conditioned media (Supplementary branching blood vessels ending in glomerulus-like capillary beds Fig. 2a). CD11b cells isolated from murine gonadal fat were near the periphery of the omentum. Immune cells aggregate used for comparison since our prior studies had demonstrated a around and within these capillary beds to form milky spots. The lack of early metastasis to intraperitoneal adipose outside of the + 21 + milky spots contain clusters of monocytes (CD11b ), B cells, omentum . Conditioned media generated by CD11b cells iso- macrophages, CD11c cells, and T cells all under a gap in the lated from mouse omenta significantly increased migration of ID8 mesothelial lining (Fig. 1a). The endothelial lining of the capil- ovarian cancer cells (Supplementary Fig. 2b). Compared to con- laries and the overlying mesothelium are adapted to facilitate the ditioned media from omental CD11b cells, conditioned media transmigration of immune cells. Additional structural elements from gonadal fat CD11b cells had only minor effects on ovarian are provided by, fibroblasts, mesenchymal cells, collagen, and cancer cell migration. To validate these findings in human tissue, elastic fibers (Fig. 1a). we isolated macrophages from human omentum collected from 2 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE ab CD68 H&E CD45 Unstimulated Naive Mouse Omentum Ovarian Cancer Cells Unstimulated/Naive Milky Spot CD45 H&E F4/80 Stimulated/Cancer-Colonized Mouse Omentum Adipocyte Adipocyte Tumor-Associated Macrophages Tumor-Associated Macrophages ID8 in C57BL/6 - 24hpi ID8 in C57BL/6 - 24hpi Dendritic Cell (CD11c+ cells) Dendritic Cell (CD11c+ cells) Fibroblasts Fibroblasts Ovarian Cancer Cell Ovarian Cancer Cell B-Cell B-Cell Mesothelial Cell Mesothelial Cell T-Cell T-Cell Extracellular Matrix Extracellular Matrix Monocytes Monocytes Hematogenous Circulation Hematogenous Circulation b Lymphatic Circulation Lymphatic Circulation M Macrophages acrophages Stimulated/Cancer-Colonized Milky Spot de Naive F4/80 Omentum Peritoneum Week 1 Diaphgragm Liver Week 2 Week 4 1.0 0.5 Naive Week 1 Week 2 Week 4 0.0 12 34 56 789 Weeks Post Injection Fig. 1 Omentum as a niche for ovarian cancer metastasis. a Representation of a naïve unstimulated milky spot and ovarian cancer metastasis activated milky spots within the mouse omentum. The number of macrophages associated with the expanding cancer-immune cells regions increase and cancer cells grow past milky spots into the central adipose region, eventually replacing it. b, c Omental milky spots can be detected in human and mouse samples. IHC of naive omental tissue sections stained for H&E, CD45 and macrophages (mouse F4/80 and human CD68). Uninvolved milky spot in human (b; inset a, b) and mouse (c; inset a). Microscopic ovarian metastases were identified within milky spots in human omenta (b; inset c), and mouse (c; inset b); scale bar = 100 µm. d C57BL/6 mice were injected with ID8 cancer cells and overt (>1 mm) peritoneal metastases over a period of 9 weeks were quantitated for metastatic spread within the abdomen (n = 6). e Representative digital scans of whole mouse omental sections stained for F4/80. Milky spots at week 0 (naive un-injected omentum) and week 1, week 2, week 3, week 4, post injection of ID8 cells. (Bottom) ×40 magnification of milky spot regions at each time point. (Scale bar = 1×, 1000 μM; 40×, 50 μM). + + + individuals with benign disease (CD45 CD14 CD68 cells; RNA-sequencing of omental macrophage reveals distinct gene Supplementary Fig. 3b). Similar to the effects shown in murine expression patterns. We next investigated the mechanisms that CD11b cells, all human macrophage conditioned media sig- regulate the omental macrophage-induced colonization of the nificantly enhanced the migration of human SKOV3.ip1 cells omentum. For this purpose, changes in gene expression were (Supplementary Fig. 2d). These findings suggest that omental studied in macrophages during the omental colonization process. + + + macrophages secrete chemoattractants into the tumor micro- Macrophages (CD45 /CD11b /F4/80 ) were isolated from mouse environment that enhance migration and, therefore, the meta- omenta prior to injection of ID8 cells, and 24 hours and seven days static potential of ovarian cancer cells. post-injection (dpi). RNA sequencing analysis was performed on COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 3 Avg. No. of Overt Peritoneal Metastases Human Omentum Mouse Omentum ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z Liposome ID8 or SKOV3ip.1 Liposome Liposome Euthanize & Treatment (1 x 10 cells) Treatment Treatment FFPE Tissues Day 0 Day 1 Day 3 Day 5 Day 7 ID8 in C57BL/6 SKOV3ip.1 in Athymic Nude CD45 F4/80 Pan-CK CD45 F4/80 Pan-CK ** PBS Clodronate *** PBS Clodronate cd e Ccl6 10 Ccl9 8 Ccl24 6 Ccl8 MY:0h Ccl25 MY:24h Ccl3 MY:7d Ccl2 CCL6 Ccl4 OM:0h Ccl7 OM:24h Ccl11 Ccl17 OM:7d Ccl12 Ccl27a Ccl5 PC1: 86% Variance Ccl22 Ccl20 Ccl1 My0h My24h My7d OM0h OM24h OM7d 0 200 400 600 800 1000 Average Expression (FPKM) in Omentum fg h i Mice CCL6 ELISA Human CCL23 ELISA 1.0 0.20 0.8 0.15 0.6 0.10 5000 0.4 800 0.05 0.2 400 0.0 0.00 0 Omentum Gonadal Uterine Mesenteric RAW mOM-M 1 mOM-M 2 PBMC hOM-M 1 hOM-M 2 Fat Fat Fat j k C57BL/6 Athymic Nude Human CCL6; F4/80; DAPI CCL6; F4/80; DAPI CCL23; CD68; DAPI macrophages at the respective time points (Supplementary Fig. 3a). macrophages before injection of ID8 cells (1283 genes were dif- RNA isolated from mesenteric adipose tissue macrophages ferentially expressed between omentum and mesentery derived was used for comparison. Hierarchical clustering and principal- macrophages). Second, omental macrophages showed a significant component analysis (PCA) of 12756 genes expressed by omental change in gene expression after the injection of ID8 cells, while macrophages during intraperitoneal metastasis resulted in various there was no change in the gene signature from mesenteric adipose important observations. First, naive omental macrophages showed tissue-derived macrophages. (Fig. 2c). Genes expressed in omental a gene signature distinct from mesenteric adipose tissue-derived macrophages at 24 h and 7 dpi were significantly enriched for cell 4 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio hOm-M hOm-M hOm-M PBMC OC156 OC157 OC158 OC159 PBS Clodronate Liposome CCL6 (1/ΔCT) PC2: 8% Variance Normal Omentum CCL23 (1/ΔCT) Clodronate Liposome PBS CCL6 (pg/mL) Log Fold Change (Omentum/Mesentery) 2 024 6 CCL23 (pg/mL) % of SKOV3ip.1 Cells Per Omentum % of ID8 cells per Omentum COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Fig. 2 Ovarian cancer cell colonization of the omentum is macrophage dependent. a Experimental design of clodronate liposome treatment for 7 dpi of ID8/SKOV3ip.1 in vivo metastasis assay. b Representative omental images of IHC for lymphocytes (CD45), macrophages (F4/80) and cancer cells (pan- CK). Quantification of DAB (pan-CK) staining area for each model is shown on the right (*p < 0.05; n = 3). (c–e) c Principal-component analysis (PCA) of genes expressed in the omental (OM) and mesenteric (MY) adipose tissue macrophages. Each data point represents macrophages derived from three omenta (n = 3). d Heat map of all chemokines expressed in (c). e Plot of chemokine genes highly upregulated in naive omental macrophages compared to naive mesenteric macrophages (timepoint zero). f Gene expression analysis of CCL6 on mouse naive omental macrophages isolated from C57BL/6 mice (n = 2) by q-RT PCR. RAW 264.7 cells were used as positive control. g Gene expression analysis of CCL23 on human omental macrophages isolated from patient with benign disease (n= 2). PBMCs were used as positive control. h ELISA of CCL6 levels in conditioned media generated from weight matched naive omentum and peritoneal adipose tissues (n = 3). i ELISA of CCL23 levels on conditioned media generated from benign patient samples. Omental macrophages (hOM-Mϕ1, 2, and 3) and whole-omental-tissue (OC157, 158, and 159) each represent a different benign sample. Conditioned media derived from PBMCs were used as control. j Immunofluorescence staining for co-localization of CCL6 (red) and F4/80 (green) and nuclei (blue, DAPI) on naive omental tissue from C57BL/6 and athymic nude mice. k Similarly, CCL23 (red) and CD68 (green) on the human omentum derived from benign patient samples. (scale bar = 20 µm). adhesion, immune response, cytokine-cytokine receptor interac- Phospho-proteomic profiling of CCL6 treated ID8 cells and CCL23 tion, and chemotaxis (Supplementary Fig. 3c). Among the che- treated SKOV3ip.1 cells showed increased levels of pERK1/2 in both mokines, CCL6 was found to be the gene with the highest fold cell lines and increased pAkt in SKOV3ip.1 cells (Supplementary upregulation in naive omental macrophages compared to naive Fig. 4d, e). pERK1/2 and pAkt regulate migration in part by 25–28 mesenteric macrophages (timepoint zero). In addition, we phosphorylating LIMK1 and p-Cofilin (Fig. 3f and Supple- observed a significant increase of CCL6 expression in omental but mentary Fig. 4f–i). We confirmed the upregulation of p-cofilin by not mesenteric macrophages during colonization (Fig. 2d, e). Since immunofluorescence in both ID8 and SKOV3ip.1 as early as 10 min 22–24 CCL6 is expressed primarily by macrophages ,wehypothe- post treatment with CCL6 and CCL23, respectively (Fig. 3g, h). We sized that macrophage-derived CCL6 might promote early omental hypothesized that phosphorylationofERK1/2byCCL6and CCL23 metastasis by interaction with its receptor CCR1 expressed on regulates MYO9A, which modulates downstream p-LIMK1 and p- ovarian cancer cells. Cofilininthe migrationpathway (Fig. 3i). In vitro migration of ID8 We validated the expression of CCL6 and CCL23 in flow-sorted and SKOV3ip.1 cells toward CCL6 or CCL23 was inhibited by mouse or human omental macrophages, respectively by qRT-PCR pharmacologic targeting of MEK (trametinib) and PI3K pathway (Fig. 2f, g). Conditioned media generated from naive mouse (BEZ235) respectively (Fig. 3j–l).In addition,inhibition ofMyosin-9 omenta contained high levels of secreted CCL6 (Fig. 2h). High (blebbistatin) resulted in a two-fold decreased migration of both ID8 levels of CCL23 were also detected in conditioned media from and SKOV3ip.1 cells toward CCL6 and CCL23, respectively (Fig. 3m, whole human omentum without tumor and human omentum n). The concentrations of BEZ235, trametinib and blebbistatin used derived macrophages (Fig. 2i). Immunofluorescence staining of in the migration assays did not affect the viability of ovarian cancer tissue sections further confirmed the presence of CCL6-expressing cells (Supplementary Fig. 4j–n). Our findings suggest that CCL6 or F4/80 macrophages in mouse omenta and CCL23-expressing CCL23 induce migration by activating the ERK and PI3-Kinase CD68 macrophages in human omenta (Fig. 2j, k). pathways resulting in enhanced downstream signaling via MYO9A and p-Cofilin. CCL6 and CCL23 promote ovarian cancer migration.We Genetic and pharmacological inhibition of CCR1 in human hypothesized that CCL6 and CCL23 are important mediators of and mouse ovarian cancer cells reduces migration toward colonization in our mouse model and in human omentum CCL6/CCL23. CCL6 and CCL23 signal through the CCR1 che- respectively. Both chemokine ligands can induce migration which mokine receptor. To determine the role of CCR1 in macrophage- is a crucial aspect of colonization. Therefore, we investigated the induced omental colonization, we first verified the expression of ability of CCL6 and CCL23 to enhance migration of mouse and CCR1 in ID8 and SKOV3ip.1 cells (Supplementary Fig. 5a). CCR1 human ovarian cancer cell lines. Both murine ID8 and human was expressed in ID8 and SKOV3ip.1 derived tumor tissue SKOV3ip.1 cell lines showed a concentration-dependent migra- (Fig. 4a, b). The functional importance of CCR1 in CCL6 and tion toward CCL6 and CCL23 (Fig. 3a, b). This effect was not due CCL23 induced migration was examined in ID8 and to increased cell proliferation as neither ID8 nor SKOV3ip.1 cells SKOV3ip.1 cells with deletion of CCR1 by CRISPR-Cas9. Deletion cultured in serum-free medium in the presence of 100 ng/mL of CCR1 was verified by sequencing the targeted locus and by CCL6 or 200 ng/mL CCL23 showed differences in cell prolifera- qRT-PCR for CCR1 expression (Supplementary Fig. 5b–e). tion (Fig. 3c). Antibody-mediated neutralization of CCL6 abro- Induction of migration by CCL6 was completely abrogated in gated the migration of ID8 cells toward mouse omentum- ID8-CCR1-null cells. (Fig. 4c). Similarly, CCR1-deleted SKO- conditioned media (Fig. 3d). Similarly, induction of SKOV3ip.1 V3ip.1 cells failed to respond to CCL23 induced migration migration toward conditioned media from human omentum was (Fig. 4d). No differences in tumor cell proliferation or viability blocked by CCL23 neutralizing antibody (Fig. 3e). were observed in the CCR1-null cell lines relative to the parental To better understand the cellular mechanisms of CCL6 and cells (Supplementary Fig. 5f, g). We next utilized a small molecule CCL23 mediated migration, we performed microarray gene expres- inhibitor of CCR1 (UCB35625) to investigate the therapeutic sion analysis on CCL6 treated ID8 cells. We found that CCL6 potential of CCR1 blockade in reducing ovarian cancer migration. induced upregulation of 98 genes and downregulation of 108 genes. CCR1 inhibition by UCB35625 blocked CCL6 and CCL23 Upregulated genes were clustered in pathways that regulate cell induced migration of ID8 and SKOV3ip.1, respectively (Fig. 4e, f). migration, such as integrin-mediated cell adhesion, MAPK signaling pathway, and signaling by RhoGTPases (Supplementary Table 1). Among important regulators of migration, Myosin 9A (Myo9A), CCR1 knock-down in ovarian cancer cells reduces colonization LIMK1, and Cofilin were significantly upregulated in CCL6 treated of the omentum. We next studied if CCR1 is important in ID8 cells compared to untreated cells (Supplementary Fig. 4a–c). omental colonization by ovarian cancer cells in vivo. mRuby- COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 5 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ab c ID8 MTS assay 16 hours SKOV3ip.1 MTS assay 16 hours 1.50 1.50 250 250 1.25 1.25 **** 200 200 1.00 1.00 ** **** 150 150 *** ** ** * &%,-.#/0$*+%01#"+.2)%1#3!/4!"3#$#"5#%&6%7/#4&8"9#'*:;("<=)#*>+?@=-&#*:-.#>"A:%0*'"+' ** 0.75 0.75 ** 100 100 0.50 0.50 50 50 0.25 0.25 0 0 0.00 0.00 SF 1 102550 100 200 SF 1 10 25 50 100 200 SF 100ng 200ng 10%FBS SF 100ng 200ng 10%FBS Concentration of CCL23 (ng/mL) Concentration of CCL6 (ng/mL) Concentration of CCL6 (ng/mL) Concentration of CCL23 (ng/mL) SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL de f ID8 + 200ng/mL CCL6 SKOV3ip.1 + 200ng/mL CCL23 No Treatment 30 minutes No Treatment 360 minutes SF 250 SF pERK1/ERK2 (p44/p42) – 44/42kDa mOM CM hOM CM 200 Total ERK1/ERK2 – 44/42kDa 200 **** IgG IgG pAKT (Ser473) – (60kDa) **** NAb 150 NAb **** 150 **** **** Total AKT– 60kDa 10% FBS 10% FBS 100 **** **** **** pLIMK1(T508) – 72kDa pCofilin (Ser3) – 19kDa Total Cofilin – 19kDa hOM CM - + + + + + - mOM CM - + + + + + - IgG (µg/mL) - - 2.5 - 5 - - Alpha tubulin– 51kDa IgG (µg/mL) - - 2.5 - 5 - - NAb (µg/mL) - - - 2.5 - 5 - NAb (µg/mL) - - - 2.5 - 5 - No Treatment 10 minutes 30 minutes 360 minutes phospho-cofilin; actin; DAPI No Treatment 10 minutes 30 minutes 360 minutes phospho-cofilin; actin; DAPI CCL6 or CCL23 ij CCR1 ** ** ERK1/2 PI3K/AKT Myo 9a/b LIMK1 200ng/mL 100nM BEZ235 SF +200ng/mL CCL23 CCL23 Cofilin 1 Chemotaxis k l *** 100 *** **** **** ** ** **** 50 **** SF 200ng/mL CCL23 2.5 M Trametinib 5 M Trametinib 10 M Trametinib SF 200ng/mL 2.5uM Trametinib 5uM Trametinib 10uM Trametinib CCL6 +200ng/mL CCL6 +200ng/mL CCL6 +200ng/mL CCL6 +200ng/mL CCL23 +200ng/mL CCL23 +200ng/mL CCL23 mn 150 *** ** 100 **** 100 **** **** **** 50 50 0 0 SF 200ng/mL CCL6 25uM Bleb + 50uM Bleb + 75uM Bleb + SF 200ng/mL CCL23 25uM Bleb + 50uM Bleb + 75uM Bleb + 200ng/mL CCL6 200ng/mL CCL6 200ng/mL CCL6 200ng/mL CCL23 200ng/mL CCL23 200ng/mL CCL23 labeled ID8 parental cells and ID8-CCR1-null cells were injected to ID8-CRISPR-control cells with a three-fold reduction at 24 h i.p. into C57BL/6 mice. At 24 hours post-injection (hpi) and and a seven-fold reduction at 7 days post injection (Fig. 5b–d and seven dpi, omenta were harvested and assessed for metastasis by Supplementary Fig. 6a–f). histology and flow cytometry (Fig. 5a). Numerous foci of ID8 We then determined whether CCR1 is associated with a more parental (control) cells were observed within the milky spots of aggressive in vivo progression of tumor growth and its impact on the omentum 24 hpi and seven dpi. At both time points, omental survival. ID8-parental, ID8-CRISPR-Control, or ID8-CCR1-null tissue contained significantly less ID8-CCR1-null cells compared cells were injected i.p. into mice. Mice injected with ID8-CCR1- 6 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio #Migrated ID8 cells/200X Field # Migrated ID8 cells/ 200X Field #Migrated ID8 cells/200X Field ID8 + CCL6 (200 ng/mL) # Migrated ID8 cells/200X Field SKOV3ip.1 + CCL23 (200 ng/mL) # Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/ 200X Field # Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/200X Field #Migrated SKOV3ip.1 cells/200XField Absorbance at 595nm Absorbance at 595nm COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Fig. 3 Ovarian cancer cells show enhanced migration toward CCL6/CCL23 by activating Myosin9 via the Cofilin signaling in vitro. a, b Transwell migration assay of ID8/SKOV3ip.1 cells toward varying concentrations of CCL6 and CCL23 with representative crystal violet staining. Statistical significance (*p < 0.05) was determined for all conditions versus the serum-free media control by ordinary one-way ANOVA analysis. c Quantitation of cell proliferation of ID8/SKOV3ip.1 cells in the presence of CCL6/CCL23 (100 ng/mL–200 ng/mL) by MTS assay. 10% FBS was used as the positive control. d, e Transwell migration assays with ID8 cells in the presence of anti-CCL6 (2.5 μg/mL and 5 μg/mL; n = 3) toward mouse omental conditioned media and SKOV3ip.1 cells toward human omental conditioned media in the presence of anti-CCL23 (2.5 μg/mL–5 μg/mL; n = 3). Statistical significance was determined by two-way ANOVA analysis for each condition corresponding to the serum-free (SF) control (*p < 0.05). f Western blot analysis of ID8 treated with CCL6 (200 ng/mL) and SKOV3ip.1 treated with CCL23 (200 ng/mL) (n = 3). g, h Immunofluorescence staining for phospho-cofilin (red), actin cytoskeleton (green), and nuclei (blue, DAPI) on ID8 and SKOV3ip.1 ovarian cancer cells treated with CCL6 or CCL23, respectively, at different time points (n = 3); (scale bar = 20 µm). i Schematic representation of the CCR1-CCL6/CCL23 signaling pathway leading to increased migratory potential of the ovarian cancer cells. j Transwell migration assay of SKOV3ip.1 toward CCL23 treated with BEZ235 (PI3K Inhibitor) (**p < 0.008). k, l Transwell migration assay of ID8 toward CCL6 and SKOV3ip.1 cells toward CCL23 treated with trametinib (MEK inhibitor). m, n Transwell migration assay of ID8 toward CCL6 and SKOV3ip.1 cells toward CCL23 treated with blebbistatin (myosin inhibitor). Statistical significance was determined by two-way ANOVA analysis with Tukey’s multiple comparison t-test with each condition compared to the 200 ng/mL CCL23 treatment condition. null cells had a significantly slower progression of tumor growth on studies in immune-deficient mice . Our current study and longer survival compared to mice injected with ID8-CRISPR- demonstrates that omentum resident macrophages promote early Control or ID8 parental cells (Fig. 5e). After 9 weeks, all mice omental metastatic colonization. Gene expression profile of (n = 5/group) injected with ID8-parental cells had developed macrophages derived from the omentum demonstrated a tissue- ascites and extensive peritoneal metastases (Fig. 5f). In contrast, specific pattern compared to other intra-abdominal adipose tis- none of the mice injected with ID8-CCR1-null cells had sues. Macrophages in other organs including the lung or the liver developed ascites or other signs of disease at this time point. are likewise characterized by specific gene signatures which The median overall survival (OS) for mice injected with ID8- convey important and specific functions in their respective tissue CCR1-null cells was 156 days and, therefore, significantly longer microenvironments . Based on our data, omental macrophages compared to both ID8-parental (OS: 64 days) and ID8-CRISPR- secreted chemokine ligands support early metastasis of ovarian Control (OS: 97 days) (Fig. 5e). cancer cells providing a possible explanation for the frequent The data generated in the ID8 tumor model suggests that involvement of the omentum in ovarian cancer patients. deletion of CCR1 expression resulted in extended survival Chemokine ligands can exert a variety of effects on cellular pro- 31,32 in vivo. To assess the levels of CCR1 in clinical samples, IHC of liferation, angiogenesis, and metastasis . CCL6 and its functional a tissue microarray showed that CCR1 was highly expressed in human homolog CCL23 are predominantly expressed by macro- epithelial ovarian cancers across different subtypes, including phages, eosinophils, and neutrophils . The expression of these serous, endometrioid, and clear cell histologies. Interestingly, chemokine ligands can be upregulated during cancer, peritonitis, and 34–36 borderline ovarian tumors, which typically have better prog- chronic rhinosinusitis . CCL6 is a known chemoattractant for nosis compared to invasive ovarian cancer, did not show any macrophages, and to a lesser extent, B cells, T-helper cells, and 37,38 significant expression of CCR1 (Fig. 5g). Based on gene eosinophils . It is able to promote innate immune response by NK expression data from a cohort of ovarian cancer datasets, cell activation in animal models of peritonitis .IL-4, aknown sti- including TCGA , patients with high CCR1 expression had a mulant for an immune-suppressive-M2-macrophage subtype, indu- shortened disease-free survival compared to patients with low ces secretion of CCL6 by macrophages . Similarly, CCL23 is IL-4- CCR1 expressing tumors. CCR1 expression levels in stage I and inducible in a STAT6-dependent manner . In a mouse model of II ovarian cancer patients showed a significant difference in lung cancer, overexpression of CCL6 accelerated tumor growth, and median progression free survival of 37 months with high CCR1 increased metastatic spread .Similarly,higherlevels of CCL23 and compared to 96 months with low CCR1 expression (HR 2.13; its corresponding receptor CCR1 correlated with reduced metastasis- p = 0.032) (Fig. 5h, i). These findings lend further support to free survival of patients with breast cancer . the important effect of CCR1 on the oncological outcome of The omental and intraperitoneal microenvironment, including ovarian cancer patients. ascites, contains a variety of chemokine ligands. CCL8 (MCP2), CCL13 (MCP4), CCL14 (HCC1), CCL15 (MIP5), and CCL16 (HCC4) bind to CCR1 albeit with different affinity and could, Discussion therefore, potentially promote colonization . However, our The current study, to our knowledge, identifies a novel mecha- experiments in omentum-macrophage-conditioned media nism of ovarian cancer colonization of the omentum mediated underscore the importance of murine CCL6 and human CCL23. by chemokine ligands secreted by omentum resident macro- Neutralization of CCL6 and CCL23 in these complex environ- phages. We demonstrate that macrophage secreted CCL6 or ments with a large variety of other chemokine ligands completely CCL23 promotes colonization of ovarian cancer cells to abrogated the migration of ovarian cancer cells. We hence pro- omental milky spots. Targeting of CCR1 in ovarian cancer pose that the effects in vivo are mainly mediated by CCL6 and cells leads to a significant reductioninmigration andomental CCL23. Conversely, other chemokine receptors may play a role in metastasis demonstrating the importance of CCR1 in ovarian promoting ovarian cancer cell metastasis to the omentum. CCL6, cancer cell metastasis. for example, can bind to CCR3 which has been implicated in The role of macrophages or other immune-effector cells in metastasis . Neither ID8 nor SKOV3ip.1 cells used in our study ovarian cancer metastasis is still not well understood. We have expressed CCR3, and we, therefore, excluded its role in the previously demonstrated that ovarian cancer cells metastasize observed reduction of metastasis in CCR1 deleted cells. However, initially to milky spots in the omentum but not other intra- CCR3 is expressed in human ovarian cancer and might, therefore, abdominal adipose tissues including the mesentery that lack 21 be another important receptor to promote metastasis in milky spots . This pattern of early omental colonization was patients . independent of the presence of B cells, T cells or NK cells based COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 7 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z a b Mouse Omental Metastasis Human Omental Metastasis CCR1 CCR1 CCR1 TROMA EpCAM EpCAM CCR1 TROMA DAPI CCR1 EpCAM DAPI DAPI CCR1 EpCAM DAPI DAPI DAPI SF c d SF 1 ng/mL 1ng/mL 10 ng/mL 10ng/mL 250 25 ng/mL 25ng/mL 50 ng/mL 50ng/mL 100 ng/mL 100ng/mL 200 **** **** **** **** 200ng/mL 200 ng/mL 10% FBS 10% FBS 150 **** **** **** **** *** ** 0 0 SKOV3ip.1 CRISPR CTRL SKOV3ip.1 CCR1 null ID8 CRISPR CTRL ID8 CCR1 null ID8 CRISPR CTRL SKOV3ip.1 CRISPR CTRL SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SKOV3ip.1 CCR1 null ID8 CCR1 null SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS SF 1 ng/mL 10 ng/mL 25 ng/mL 50 ng/mL 100 ng/mL 200 ng/mL 10% FBS e f SF SF 100 ng/mL CCL6 100 ng/mL CCL23 200 ng/mL CCL6 **** 200 ng/mL CCL23 **** *** *** 50 50 0 0 No Treatment UCB35625 No Treatment UCB35625 No treatment UCB35625 No treatment UCB35625 SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL SF 100 ng/mL 200 ng/mL Fig. 4 Genomic deletion and pharmacological inhibition of CCR1 in human and mouse ovarian cancer cells inhibits their migratory capacity toward CCL6/CCL23 in vitro. a Immunofluorescence for CCR1 (red), cytokeratin (green), and nuclei (blue, DAPI) on ID8 cells metastasized to omentum in C57BL/6 and athymic nude mice (scale bar = 20 µm). b Immunofluorescence for CCR1 (red), EPCAM (green), and nuclei (blue, DAPI) on omenta derived from patients with high-grade serous cancer (scale bar = 20 µm). c, d Transwell migration assay of ID8 and SKOV3ip.1 with genomic deletion of CCR1 toward CCL6 and CCL23. e, f Transwell migration assay of ID8 and SKOV3ip.1 toward CCL6 and CCL23 with pharmacological inhibition of CCR1 using a small molecule antagonist UCB35625 (100 nM). Data shown represent mean and s.d of technical replicates (n = 3) for each condition. Statistical significance was determined by ordinary two-way ANOVA analysis comparing each condition to serum free (SF) media (**p < 0.005; ****p < 0.0001). A role for CCR1 in cancer metastasis has been suggested in that enhanced metastasis . CCR1 activation enhanced the other studies, but not in the context of ovarian cancer or medi- interaction between “metastasis associated macrophages” (MAM) ated by omental macrophage-derived CCL6/CCL23. CCR1 and cancer cells in part through integrin α4 resulting in increased expression was found to increase in colon cancer cells during extravasation of cancer cells and metastasis . In a study by Zhu metastasis to the liver . In a separate study, CCR1 sustained liver et al., osteopontin (OPN) was found to upregulate CCR1 expre- metastasis by promoting local recruitment of bone marrow (BM)- ssion in hepato-cellular carcinomas. CCR1 knockdown resulted derived cells that secrete the MMP9 and MMP2 metalloprotei- in reduction of migration, invasion and pulmonary metastasis 45 48 nases required for tissue invasion . Activation of EGFR signaling induced by OPN in vitro and in vivo . via CCR1 has been shown to contribute to breast cancer invasion CCR1 promotes monocytic tissue infiltration and plays a major and metastasis . The CCR1 mediated interaction between mac- role in various other disease processes, including autoimmune rophages and breast cancer cells induced a chemokine cascade diseases like rheumatoid arthritis. Clinical trials targeting CCR1 8 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio C57BL/6 # Migrated ID8 cells/200X Field # Migrated ID8 cells/200X Field Athymic Nude # Migrated SKOV3ip.1 cells/200X Field # Migrated SKOV3ip.1 cells/200X Field Human Omentum COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Euthanize Euthanize (n=9/group), (n=9/group), ID8-mRuby Flow cytometry Flow cytometry (1x10 Cells; i.p.) & FFPE Tissues & FFPE Tissues Day 7 Day 0 Day 1 b d 24 Hours Post Intraperitoneal Injection of ID8 in C57BL/6 DAB Quantitation Flow Cytometry Quantitation Naive Omentum ID8 Parental ID8 CCR1 null 24 Hours Post Injection ** ID8 Parental ID8 CCR1 null ID8 Parental ID8 CCR1 null 7 Days Post Intraperitoneal Injection of ID8 in C57BL/6 DAB Quantitation Flow Cytometry Quantitation Naive Omentum ID8 Parental ID8 CCR1 null 7 Days Post Injection 15 40 0 0 ID8 Parental ID8 CCR1 null ID8 Parental ID8 CCR1 null e g ID8-Parental ID8-CRISPR-CTRL **** p<0.0001 **** **** **** ID8-CCR1-null BL vs. CC BL vs. EC *** BL vs. SC 0 50 100 150 200 Days elapsed ID8 CCR1 null ID8 Parental ID8 CRISPR CTRL Histotype N=1104 N=99 All Stages Stages 1 and 2 Time (months) Time (months) for the treatment of rheumatoid arthritis and multiple sclerosis study, and safety and tolerability of CCR1 receptor antagonists in have so far shown that the targeting agents used are well tolerated human clinical trials provide a rationale for using CCR1 but have limited clinical efficacy . Targeting CCR1 in cancer antagonists as a new therapeutic target in ovarian and other patients to reduce metastasis or induce changes in the tumor cancers. microenvironment might lead to anti-tumor immune responses Various other cells types, such as neutrophils, mesothelial cells due to reduction of monocytic infiltration. The findings in our and adipocytes have been described to regulate ovarian cancer COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 9 BL (N=13) SC (N=82) EC (N=66) CC (N=30) Diaphragm Peritoneal Wall Mesentery Omentum Percent survival Progression Free Survival % of DAB Staining per Omentum % of DAB Staining per Omentum CCR1 Expression (%) Progression Free Survival % of ID8-mRuby Cells % of ID8-mRuby Cells ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z Fig. 5 Colonization of the omentum is disrupted by CCR1 knock-down in the cancer cells. a Schematic of experimental design for the 24hpi and 7dpi of ID8 in vivo metastasis assay. b Representative sections of omenta collected at 24 hpi and 7 dpi of stained for CK8/18 by IHC (scale bar = 100 µm). c Quantification of DAB staining area as an indicator of ID8 cancer cell burden in omental tissues. d Quantitation of flow cytometry analysis of ID8-mRUBY cells colonizing omentum at 24 hpi and 7 dpi (*p < 0.05). e Kaplan–Meier curve showing percent-survival after i.p injection of ID8- parental, CRISPR- Control, or CCR1-null cells (**p < 0.001) (n = 6/group). f Representative H&E stained peritoneal tissues at experimental end point. H&E images of ID8- CCR1-null group are from mice euthanized at 97 days, the experimental end point of the ID8-CRISPR-Control group. (scale bar = 100 µm). g Examples of 3 patient cores of endometrioid histology from a tissue microarray stained for CCR1 and IHC quantitation of CCR1 levels in a tissue microarray with different ovarian cancer histologies (****p < 0.0001). h, i Kaplan–Meier survival analysis (log-rank test) of CCR1 expression on progression free survival of serous ovarian cancer patients in h all stages and i stages I and II. 15,17,50–53 metastasis . Neutrophil Extracellular Traps (NET) for- expressing cells. All cells were maintained under standard tissue culture conditions (i.e., in a humidified incubator at 37 °C supplemented with 5% CO ). mation provides a premetastatic omental niche conducive for implantation of ovarian cancer cells. Pharmacological blockade of NET formation decreased omental colonization by ovarian cancer Patient samples. Omenta were obtained from patients under approved IRB protocol at the Dept. of Obstetrics and Gynecology, Stanford Hospital. Informed cells . A unique subset of omental tissue-resident macrophages consent was obtained from each patient prior to sample collection. Omental spe- + + (CD163 Tim4 ) can promote a premetastatic niche in the cimens with HGSC and omenta derived from patients with benign disease were omentum . Other studies have focused on the attachment of obtained and processed for histological analysis and in vitro assays as ovarian cancer cells to mesothelial cells that outline the peritoneal described below. cavity. This process is regulated by multiple adhesion molecules, 50,53 proteases and extracellular matrix components . Fibronectin Mice. All mice were housed, maintained, and euthanized according to APLAC receptors like α5β1‐integrins, for example, are expressed on the protocol and under the supervision of the Stanford Animal Resource Center. 6–8 weeks, female, inbred C57BL/6 (C57BL/6NHsd; immunocompetent) and nude surface of ovarian cancer cells and mediate the adhesion through (Athymic Nude-Foxn1nu; T-cell deficient) mice were obtained from Harlan myosin‐mediated traction forces . Furthermore, TGF‐β secreted Laboratories (Indianapolis, IN). All animal studies were performed in accordance by cancer cells activates a RAC1/SMAD‐mediated signaling with Stanford APLAC approved protocols. All mice were randomly assigned to pathway in mesothelial cells, which results in transcriptional appropriate treatment groups. upregulation of the fibronectin gene and induces an EMT‐like phenotype in mesothelial cells . The adhesion of cancer cells to In vivo experimental metastasis assays. Exponentially growing untagged or mRuby 3,51 tagged-ID8 parental, ID8-CCR1-null cells, ID8-CRISPR-control cells were trypsi- the mesothelium is followed by mesothelial clearance . In this nized and prepared as a single-cell suspension at a concentration of 2 × 10 cells/ process, mesothelial cells are displaced by ovarian cancer spher- mL in ice cold PBS. Animals were injected intraperitoneally with 500 μL of the cell oids followed by invasion of cancer cells into the deeper layer of suspension (1 × 10 cells). For all experiments, 500 μL of PBS was injected as a peritoneal tissues . Specifically, hypoxic signaling increased negative control in a parallel group of control mice. At the experimental endpoint (24 h or 7 dpi) of each assay, mice were euthanized in accordance with APLAC expression of lysyl oxidase (LOX) in mesothelial and ovarian protocol via CO asphyxiation and a secondary method of cervical dislocation was cancer cells to promote collagen crosslinking and tumor cell performed. For the time course experiment, grossly visible overt peritoneal invasion . Further, during ovarian cancer progression, omental metastases that were greater than 1 mm, from each mouse, were counted. Tissues adipocytes provide fatty acids as an energy source to ovarian were then harvested, processed, sectioned, and stained as described in “Methods”. cancer cells . Macrophage depletion using clodronate liposomes. For both pre-cancer and post- Based on our current work and the published literature, we cancer injection experiments, peritoneal macrophages were depleted by injection of propose a multi-step model for omental colonization. Mesothelial clodronate-containing liposomes (clodronateliposome.org). In vivo depletion of cells promote ovarian cancer cell adhesion to the omentum. macrophages was assessed for several dosages by cytology performed on omental Subsequent migration and homing to milky spots is induced by dissociated cells and by flow cytometry with F4/80 antibody. A dose of 100 μLof liposomes diluted in PBS was selected and administered i.p. to mice every alternate omental resident macrophages via chemokine ligands. The day. Control mice were administered either PBS as a control for unstimulated interaction between macrophages and cancer cells creates an macrophages or saline liposomes to control for any nonspecific effects of liposome inflammatory environment that recruits other immune-effector administration. Macrophage depletion was maintained during the experimental cells, including neutrophils. The formation of NETs further period. enhances colonization of ovarian cancer cells. Adipocytes then provide the energy required for the continuous growth of tumor Isolation of murine and human omental macrophages. Omenta excised from cells. Further studies are needed to explore this complex process C57BL/6 mice (n = 9) and processed as three separate groups (n = 3 per group). A and identify opportunities for therapeutic interventions. modified protocol of the adipose tissue dissociation kit (130–105–808, Miltenyi Biotech) was used to obtain a single cell suspension of the mouse omenta. Human omental samples were processed using the human tumor dissociation kit (130-095- 929, Miltenyi Biotech) to dissociate soft tumors into a single cell suspension. The Methods cells were ACK lysed and re-suspended in PBS containing 1% FBS for flow cyto- Cell lines. Human SKOV3.ip1 ovarian cancer cell lines were cultured and main- metry applications to analyze and sort for mouse or human omental macrophages. tained at 37 °C in Dulbecco’s Modified Eagle Medium (DMEM) (10-013-CV, Corning Cellgro) supplemented with 10% (v/v) fetal bovine serum (FBS) (S11150, Atlanta Biologicals, Lawrenceville, GA, USA) and 1% penicillin-streptomycin (Pen Antibodies and flow cytometry. The isolated mouse and human omental cells Strep) (v/v) (15240, Invitrogen). Similarly, the murine ID8 ovarian cancer cell line were re-suspended in PBS supplemented with 1% FBS. The mouse panel designed was cultured in DMEM supplemented with 4% FBS and 1% Penicillin Strep and 1% to sort for murine omental macrophages included CD45-FITC (clone-30-F11, Insulin-Transferrin-Selenium (25-800-CR, Corning). RAW 264.7 cells and human eBioscience), CD11b-APC (clone-M1/70, eBioscience), F4/80-PE (clone-BM8, peripheral blood mononuclear cells (hPBMCs) were cultured in 10% RPMI and eBioscience). Similarly, the human panel designed to sort for human omental 10% FBS and 1% penicillin-streptomycin]. ID8 or SKOV3ip.1 ovarian cancer cells macrophages included CD45-FITC (clone-HI30, eBiosciences), CD14-PE (clone- that stably express mRuby (ID8-mRuby/SKOV3ip.1-mRuby) were constructed by 61D3, eBioscience), CD68-PECy7 (Clone-eBioY1/82A, eBioscience). In both the lentiviral delivery of pLVX-mRuby expression vector (Stanford genomics facility). mouse and human flow panel, aqua amine (L34957, Fisher Scientific) was used as The mRuby tagged expression vector along with the packaging vector master mix the Live/Dead stain and Compensation Beads (01-1111, eBioscience) were used for PMDL(PVSV-G) and PSPAX2 lentiviral expression plasmid was transfected into compensation controls. Labeled cells were sorted by flow cytometry on a BD HEK293T cells to generate the viral conditioned medium. Fluorescence-activated InfluxFlow Sorter or analysis was performed on BD LSR II at Stanford shared cell sorting using a BD FACS Aria II system (BD Biosciences, San Jose, CA) at the FACS facility. For the RNA-seq analysis, CD45+CD11b+F4/80+ cells were sorted Stanford Flow Cytometry Core Facility was used to select for high mRuby- into buffer RLT and were processed immediately for RNA isolation. 10 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE RNA sequencing and analyses. For each sample in the whole transcriptome sample in triplicate. Cells were incubated for a period of 16 h overnight in a 37 °C sequencing library, 43–66 million 75-basepair paired-end reads were acquired from incubator (5% CO ). After incubation, the top and the bottom chambers were the sequencer. Read quality was determined with FastQC 0.11.4. The reads were washed with PBS and cells were fixed with 4% paraformaldehyde for 10 min. After aligned to the mouse reference genome (NCBI37/mm9) using STAR 2.4.2a, with washing with PBS, the cells were stained with 0.05% Crystal Violet for 40 min. Both splice junctions defined by the GTF file (UCSC). On average, 93% of reads were the chambers were then washed with tap water and the cells on the top of the insert aligned to the reference genome, and 87% of reads uniquely aligned to the reference were gently scraped with a cotton swab. The positive-staining cells adherent to the genome. Gene expression in raw count was determined by STAR with quant mode, bottom of the transwell, were examined under the microscope. Data are shown as GeneCounts setting. Cufflinks 2.2.1 was used to normalize gene expression in mean ± s.d. from three independent experiments. fragments per kilobase per million aligned reads (FPKM). Further, differential expression between the conditions was evaluated using DEseq2 1.12.4. The fold Protein isolation and western blotting. Protein was extracted from ID8 and change between each sample is calculated using DEseq2. The genes that were either SKOV3ip.1 cells treated with CCL6/CCL23 at different time points by RIPA lysis upregulated 2-fold or downregulated 2-fold were selected for gene set enrichment extraction method (pH 7.4) containing Halt protease inhibitor cocktail (#1861278, analysis using DAVID (https://david.ncifcrf.gov/) with gene set annotation GO BP Thermofisher Scientific) and EDTA (0.5%) under agitation at 4 °C for 30 min. A and KEGG database. The top 10 gene sets are only used for visualization. bicinchoninic acid assay (Pierce, Rockford, IL, USA) was used to determine total protein concentration. Equal amounts of protein were separated on (4–12%) SDS- PAGE and transferred to nitrocellulose membranes (Sigma, St Louis, MO, USA). CRISPR/Cas9 genome editing. Human (AACTTGTAGTCGATCCAGAA) and murine (GAACACTAGAGAATACAGG) CCR1 sgRNA sequences were identified Membranes were blocked in 5% low-fat dry milk in TBS-T for 1-h at room tem- perature and then incubated overnight at 4 °C. Dilutions for the primary antibodies using the Broad Institute Genetic Perturbation Platform Web Portal’s sgRNA design tool . Control non-targeting sequences for human (ACGGAGGCTAA were used according to the manufacturer’s recommendation. The primary anti- bodies used for this study are listed as follows: anti-pERK1/2 (p44/42) (#9101S, Cell GCGTCGCAA) and mouse (GCGAGGTATTCGGCTCCGCG) were taken from the non-targeting sequences in the human and mouse GeCKOv2 libraries . Oli- Signaling Technologies),anti-total ERK1/2 (#9102S, Cell Signaling Technologies), anti-pAKT (Ser473) (#4058S, Cell Signaling Technologies),anti-total AKT(#9272S, goduplexes were cloned into pLentiCRISPRv2 (a gift from Feng Zhang; Addgene Cell Signaling Technologies), anti-pLIMK (#ab194798,Abcam), anti-p-Cofilin (Ser3) plasmid 52961 ) and transformed to Stbl3 bacteria. pLentiCRISPRv2 vectors (#3313L, Cell Signaling Technologies), anti-Cofilin (#5175S, Cell Signaling Tech- containing the sgRNA guides were transfected with pMD2.G and pSPAX2 (gifts nologies), anti-alpha tubulin (#3873S, Cell Signaling Technologies). Following day, from Didier Trono; Addgene plasmids 12259 and 12260) to 293T cells using Fugene (Roche) in order to produce lentiviral vectors. After 24 h of culture, after three washes for 10 min in TBS-T, the membrane was incubated with IRDye- 680 nm conjugated-goat anti-rabbit (#925-32211, LICOR) and IRDye-800nm con- supernatants were collected, filtered through a 0.45 μm filter and added to monolayers of either ID8 (mouse) or SKOV3ip.1 (human) cell lines. Transduced jugated-donkey anti-mouse (#926-32212, LICOR) secondary antibodies at 1:5000 for 1-h at room temperature. Membranes were washed for 10 min in TBS-T before ID8/SKOV3ip.1 cells were then selected in medium containing 0.6/0.4 µg/mL puromycin for 1 week and individual clones were selected by limiting dilution and being imaged by the LICOR Odyssey Clx imaging system. The intensity of the protein bands was semi-quantified using the Image studio software (Windows then expanded in puromycin containing media. The targeted genomic regions were amplified by PCR (mouse CCR1 forward: 5′- application VIO.02), with normalization of each protein against alpha tubulin. TTCCACTGCTTCAGGCTCTT-3′ and reverse: 5′-TCATTTGTTCTGTTCCC CTCTT-3′ and human CCR1 forward: 5′-TACCAGCCCAAAGAGGTTCA-3′ and Phospho-protein array. Protein lysates isolated from ID8 and SKOV3ip.1 cells reverse: 5′-TGACACGACCACAGAGTTTGA-3′) and analyzed using TIDE-seq . stimulated with CCL6 or CCL23 at 30 min time points were processed and assayed To characterize the specific lesion in each cell line, PCR products were Topo cloned using the proteome-profiler human phospho-kinase array (#ARY003B, R&D Sys- (Life Technologies) and 10 colonies from each isolated cell line was sequenced. tems) according to manufactures protocol. Membranes were incubated with pro- tein lysates of ID8 (500 μg) or SKOV3ip.1 (250 μg) for each condition, respectively. The membranes were imaged by Chemiluminescence by using LICOR Odyssey Clx Cell proliferation assays. For cell proliferation assays, cells were seeded in 96-well imaging system. Protein bands were semi-quantified using the Image studio soft- plates (1.5 × 10 cells/well). At the indicated time points, cell proliferation was ware (Windows application VIO.02), using normalization controls as per the detected by MTS assay (Promega, G3530) according to the manufacturer’s manufacture’s protocol. instructions. Data are shown as mean ± s.d. from three independent experiments. Immunohistochemistry. Slides were first deparaffinized with xylene treatment for Real-time RT–PCR. Total RNA was isolated with the RNeasy kit (Qiagen) and 60 min, 30 min, and 10 min three times each. Slides were then rehydrated, and reverse-transcribed into cDNA using the Superscript Strand III cDNA SuperMix antigen retrieval was performed in 1× citrate buffer (pH 6.0) for 5 min in a con- (Invitrogen). Real-time PCR was performed using SSO Fast Evergreen Supermix ventional steamer. Slides were incubated in a blocking buffer solution (Protein block (Biorad) on an Applied Biosystems 7500 Fast Real-Time PCR System at Functional serum-free ready-to-use, DAKO) for 60 min at room temperature. After standard Genomics Facility at Stanford. Expression of the target genes were normalized to slide preparation as described above, slides were incubated with primary antibody the housekeeping gene 18S. Mouse RAW 264.7 cells and human PBMC were used overnight and subsequent secondary antibodies (Supplementary Table 1). After as a positive control for the all the gene expression studies. washing, slides were either developed for immunohistochemistry or incubated with The target gene-specific primers were obtained from Integrated DNA Technologies DAPI and imaged using an EVOS Fluorescence microscope. Level of CCR1 on (IDT). The primer sequences used in this study are shown in Supplementary Table 2. patient tissue core samples on a tissue microarray was determined by HALO image All real-time RT-PCRs were performed in triplicates and the relative mRNA −ΔCT analysis software area quantification algorithm (Indica Labs, Corrales, NM). expression of each target gene was determined by using the formula 2 (C ,cycle threshold) where ΔC = C (target gene)− C (18S). The comparative expression T T T −ΔΔCT level of each target gene between different samples was 2 . Statistics and reproducibility. All grouped data are presented as mean ± s.e.m. Significance between groups was analyzed by one-way ANOVA or Student t-test using GraphPad Prism. KM plotter (https://kmplot.com/analysis/) was used to Generation of tissue and cell conditioned media. Three freshly excised mouse analyze the prognosis of CCR1 in ovarian cancer. The analysis was conducted for omenta were pooled and processed to produce a single cell suspension with all serous samples with an ‘auto-select best cutoff’ feature. The data with statistical minimal cell loss. Gonadal fat was similarly processed in parallel. Macrophages analyses were performed with biological replicates on different samples of the same (CD11b cells) were isolated from stromal and vascular cells (SVC) by positive cell type on different days, yielding similar results. The number of biological selection using CD11b microbeads (Miltenyi Biotech) (Supplementary Fig. 2). To replicates is included in the figure legends. Human CCL23 ELISA data shown in assess the functional phenotype, 1 × 10 cells were plated and allowed to condition Fig. 2i is plotted as separate data points to show inter-patient variability. serum-free media for 24 h after which the media were used in standard transwell migration assay. Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Transwell migration assays Toward chemokines CCL6/CCL23. For cell migration assays, 1.5 × 10 (ID8) or 3 × 10 (SKOV3ip.1) cells in serum-free media were plated in the top chamber of 8.0- Data availability μm 12-well plate transwell insert (12-well format, 8-μm pore; BD PharMingen). Raw data and analyzed data for the RNA-sequencing and microarray gene expression CCL6 (250-06, Peprotech) CCL23 (300-29, Peprotech), or omental conditioned analysis can be accessed from Gene Expression Omnibus using the accession number medium was added to the bottom well of the plate (35-3503, BD Falcon) as GSE153685 and GSE153790, respectively. Source data are available in Supplementary chemoattractant. Anti-CCL6 (MAB000487, R&D systems) and anti-CCL23 Data 1. All other data are available from authors on reasonable request. (MAB371, R&D systems) antibodies were added to omentum-conditioned media in neutralization experiments. For pharmacological Inhibition, CCR1 inhibitor— UCB35625 (2757, Tocris), PI3K inhibitor—BEZ235 (S1009, Selleckchem), MEK Received: 2 March 2020; Accepted: 6 August 2020; inhibitor—trametinib (S2673, Selleckchem) and myosin inhibitor—blebbistatin (1760, Tocris) were used for the migration assays. Analysis was performed on each COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 11 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z References 32. Yi, F., Jaffe, R. & Prochownik, E. V. The CCL6 chemokine is differentially 1. Bray, F. et al. Global cancer statistics 2018: GLOBOCAN estimates of regulated by c-Myc and L-Myc, and promotes tumorigenesis and metastasis. incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer Cancer Res. 63, 2923–2932 (2003). J. Clin. 68, 394–424 (2018). 33. Heng, T. S. & Painter, M. W., Immunological Genome Project, C. The 2. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2020. CA Cancer J. Immunological Genome Project: networks of gene expression in immune cells. Clin. 70,7–30 (2020). Nat. Immunol. 9, 1091–1094 (2008). 3. Yeung, T. L. et al. Cellular and molecular processes in ovarian cancer 34. LaFleur, A. M., Lukacs, N. W., Kunkel, S. L. & Matsukawa, A. Role of CC metastasis. A review in the theme: cell and molecular processes in cancer chemokine CCL6/C10 as a monocyte chemoattractant in a murine acute metastasis. Am. J. Physiol. Cell Physiol. 309, C444–C456 (2015). peritonitis. Mediators Inflamm. 13, 349–355 (2004). 4. Lambert, A. W., Pattabiraman, D. R. & Weinberg, R. A. Emerging biological 35. Nagarsheth, N., Wicha, M. S. & Zou, W. Chemokines in the cancer principles of metastasis. Cell 168, 670–691 (2017). microenvironment and their relevance in cancer immunotherapy. Nat. Rev. 5. Gerber, S. A. et al. Preferential attachment of peritoneal tumor metastases to Immunol. 17, 559–572 (2017). omental immune aggregates and possible role of a unique vascular 36. Poposki, J. A. et al. Increased expression of the chemokine CCL23 in microenvironment in metastatic survival and growth. Am. J. Pathol. 169, eosinophilic chronic rhinosinusitis with nasal polyps. J. Allergy Clin. Immunol. 1739–1752 (2006). 128,73–81 e74 (2011). 6. Pearce, O. M. T. et al. Deconstruction of a metastatic tumor 37. Asensio, V. C. et al. C10 is a novel chemokine expressed in experimental microenvironment reveals a common matrix response in human cancers. inflammatory demyelinating disorders that promotes recruitment of macrophages Cancer Disco. 8, 304–319 (2018). to the central nervous system. Am.J.Pathol. 154,1181–1191 (1999). 7. Nieman, K. M. et al. Adipocytes promote ovarian cancer metastasis and 38. Berger, M. S. et al. The chemokine C10: immunological and functional provide energy for rapid tumor growth. Nat. Med. 17, 1498–1503 (2011). analysis of the sequence encoded by the novel second exon. Cytokine 8, 8. Khan, S., Taylor, J. L. & Rinker-Schaeffer, C. W. Disrupting ovarian cancer 439–447 (1996). metastatic colonization: insights from metastasis suppressor studies. J. Oncol. 39. Coelho, A. L. et al. The chemokine CCL6 promotes innate immunity via 2010, 286925 (2010). immune cell activation and recruitment. J. Immunol. 179, 5474–5482 (2007). 9. Meza-Perez, S. & Randall, T. D. Immunological functions of the omentum. 40. Orlofsky, A., Lin, E. Y. & Prystowsky, M. B. Selective induction of the beta Trends Immunol. 38, 526–536 (2017). chemokine C10 by IL-4 in mouse macrophages. J. Immunol. 152, 5084–5091 10. Krist, L. F. et al. Cellular composition of milky spots in the human greater (1994). omentum: an immunochemical and ultrastructural study. Anat. Rec. 241, 41. Novak, H. et al. CCL23 expression is induced by IL-4 in a STAT6-dependent 163–174 (1995). fashion. J. Immunol. 178, 4335–4341 (2007). 11. Liu, J., Geng, X. & Li, Y. Milky spots: omental functional units and hotbeds for 42. Yan, H. H. et al. CCL9 induced by TGFbeta signaling in myeloid cells peritoneal cancer metastasis. Tumour Biol. 37, 5715–5726 (2016). enhances tumor cell survival in the premetastatic organ. Cancer Res. 75, 12. Mebius, R. E. Lymphoid organs for peritoneal cavity immune response: milky 5283–5298 (2015). spots. Immunity 30, 670–672 (2009). 43. Long, H. et al. Autocrine CCL5 signaling promotes invasion and migration of 13. Rangel-Moreno, J. et al. Omental milky spots develop in the absence of CD133+ ovarian cancer stem-like cells via NF-kappaB-mediated MMP-9 lymphoid tissue-inducer cells and support B and T cell responses to peritoneal upregulation. Stem Cells 30, 2309–2319 (2012). antigens. Immunity 30, 731–743 (2009). 44. Akram, I. G., Georges, R., Hielscher, T., Adwan, H. & Berger, M. R. The 14. Cleypool, C. G. J., Schurink, B., van der Horst, D. E. M. & Bleys, R. chemokines CCR1 and CCRL2 have a role in colorectal cancer liver Sympathetic nerve tissue in milky spots of the human greater omentum. J metastasis. Tumour Biol. 37, 2461–2471 (2016). Anat. 236, 156–164 (2019). 45. Hirai, H. et al. CCR1-mediated accumulation of myeloid cells in the liver 15. Lee, W. et al. Neutrophils facilitate ovarian cancer premetastatic niche microenvironment promoting mouse colon cancer metastasis. Clin. Exp. formation in the omentum. J. Exp. Med. 216, 176–194 (2019). Metastasis 31, 977–989 (2014). 16. Shimotsuma, M., Takahashi, T., Kawata, M. & Dux, K. Cellular subsets of the 46. Shin, S. Y. et al. C-C motif chemokine receptor 1 (CCR1) is a target of the milky spots in the human greater omentum. Cell Tissue Res. 264, 599–601 (1991). EGF-AKT-mTOR-STAT3 signaling axis in breast cancer cells. Oncotarget 8, 17. Etzerodt, A., et al. Tissue-resident macrophages in omentum promote 94591–94605 (2017). metastatic spread of ovarian cancer. J. Exp. Med. 217, e20191869 (2020). 47. Qian, B. Z. et al. CCL2 recruits inflammatory monocytes to facilitate breast- 18. Linde, N. et al. Macrophages orchestrate breast cancer early dissemination and tumour metastasis. Nature 475, 222–225 (2011). metastasis. Nat. Commun. 9, 21 (2018). 48. Zhu, Y. et al. C-C chemokine receptor type 1 mediates osteopontin-promoted 19. Kitamura, T. et al. CCL2-induced chemokine cascade promotes breast cancer metastasis in hepatocellular carcinoma. Cancer Sci. 109, 710–723 (2018). metastasis by enhancing retention of metastasis-associated macrophages. J. 49. Gladue, R. P., Brown, M. F. & Zwillich, S. H. CCR1 antagonists: what have we Exp. Med 212, 1043–1059 (2015). learned from clinical trials. Curr. Top. Med. Chem. 10, 1268–1277 (2010). 20. Robinson-Smith, T. M. et al. Macrophages mediate inflammation-enhanced 50. Kenny, H. A. et al. Mesothelial cells promote early ovarian cancer metastasis metastasis of ovarian tumors in mice. Cancer Res. 67, 5708–5716 (2007). through fibronectin secretion. J. Clin. Invest. 124, 4614–4628 (2014). 21. Clark, R. et al. Milky spots promote ovarian cancer metastatic colonization of 51. Kenny, H. A., Kaur, S., Coussens, L. M. & Lengyel, E. The initial steps of peritoneal adipose in experimental models. Am. J. Pathol. 183,576–591 (2013). ovarian cancer cell metastasis are mediated by MMP-2 cleavage of vitronectin 22. Orlofsky, A., Berger, M. S. & Prystowsky, M. B. Novel expression pattern of a and fibronectin. J. Clin. Invest. 118, 1367–1379 (2008). new member of the MIP-1 family of cytokine-like genes. Cell Regul. 2, 52. Natarajan, S. et al. Collagen remodeling in the hypoxic tumor-mesothelial 403–412 (1991). niche promotes ovarian cancer metastasis. Cancer Res. 79, 2271–2284 (2019). 23. Shay, T. & Kang, J. Immunological Genome Project and systems immunology. 53. Niedbala, M. J., Crickard, K. & Bernacki, R. J. Interactions of human ovarian Trends Immunol. 34, 602–609 (2013). tumor cells with human mesothelial cells grown on extracellular matrix. An 24. Ma, B. et al. The C10/CCL6 chemokine and CCR1 play critical roles in the in vitro model system for studying tumor cell adhesion and invasion. Exp. Cell pathogenesis of IL-13-induced inflammation and remodeling. J. Immunol. Res. 160, 499–513 (1985). 172, 1872–1881 (2004). 54. Iwanicki, M. P. et al. Ovarian cancer spheroids use myosin-generated force to 25. Klemke, M., Kramer, E., Konstandin, M. H., Wabnitz, G. H. & Samstag, Y. An clear the mesothelium. Cancer Disco. 1, 144–157 (2011). MEK-cofilin signalling module controls migration of human T cells in 3D but 55. Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize not 2D environments. EMBO J. 29, 2915–2929 (2010). off-target effects of CRISPR-Cas9. Nat. Biotechnol. 34,184–191 (2016). 26. Makowska, K. A., Hughes, R. E., White, K. J., Wells, C. M. & Peckham, M. 56. Sanjana, N. E., Shalem, O. & Zhang, F. Improved vectors and genome-wide Specific myosins control actin organization, cell morphology, and migration in libraries for CRISPR screening. Nat. Methods 11, 783–784 (2014). prostate cancer cells. Cell Rep. 13, 2118–2125 (2015). 57. Brinkman, E. K., Chen, T., Amendola, M. & van Steensel, B. Easy quantitative 27. Wang, W., Eddy, R. & Condeelis, J. The cofilin pathway in breast cancer assessment of genome editing by sequence trace decomposition. Nucleic Acids invasion and metastasis. Nat. Rev. Cancer 7, 429–440 (2007). Res. 42, e168 (2014). 28. Yamaguchi, H. & Condeelis, J. Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochim Biophys. Acta 1773, 642–652 (2007). 29. Gyorffy, B., Lanczky, A. & Szallasi, Z. Implementing an online tool for genome- Acknowledgements wide validation of survival-associated biomarkers in ovarian-cancer using We thank Drs. Roger Pederson (Stanford) and Mickey Hu (Panorama Research Institute) microarray data from 1287 patients. Endocr. Relat. Cancer 19, 197–208 (2012). for critical reading of the paper. This work was supported by the Mary Lake Polan 30. Gosselin et al. Environment drives selection and function of enhancers Gynecologic Oncology Endowment for Research (O.D.), the Vivian Scott Fellowship in controlling tissue-specific macrophage identities. Cell 159, 1327–1340 (2014). Gynecologic Oncology (O.D.), the Dean Pizzo Stanford Cancer Center Research Award 31. Sarvaiya, P. J., Guo, D., Ulasov, I., Gabikian, P. & Lesniak, M. S. Chemokines (O.D.), the Ovarian Cancer Research Fund Alliance—Ann and Sol Schreiber Mentored in tumor progression and metastasis. Oncotarget 4, 2171–2185 (2013). Investigator Award (V.K.), and generous donation of Tell Every Amazing Lady About 12 COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01246-z ARTICLE Ovarian Cancer Louisa McGregor Ovarian Cancer Foundation, also known as T.E.A.L. Reprints and permission information is available at http://www.nature.com/reprints (V.K.), Stanford Medical Scholars Fellowship Program (A.M.C.). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author contributions V.K. and O.D. designed experiments; B.S. designed CRISPR/Cas9 experiments; V.K., S.T., K.K., and A.M.C. carried out and analyzed experiments; V.K. and S.P. analyzed histology Open Access This article is licensed under a Creative Commons specimens; Y.Z. analyzed RNA-seq data set; V.P. and A.F.K. provided clinical samples Attribution 4.0 International License, which permits use, sharing, from early and late stage of ovarian cancer disease; V.K. prepared human tissue samples; adaptation, distribution and reproduction in any medium or format, as long as you give V.K. prepared the figures wrote the initial draft of the paper; all authors edited the paper; appropriate credit to the original author(s) and the source, provide a link to the Creative C.R.S., E.B.R., and O.D. provided supervision. Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless Competing interests indicated otherwise in a credit line to the material. If material is not included in the The authors declare no competing interests. article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from Additional information the copyright holder. To view a copy of this license, visit http://creativecommons.org/ Supplementary information is available for this paper at https://doi.org/10.1038/s42003- licenses/by/4.0/. 020-01246-z. © The Author(s) 2020 Correspondence and requests for materials should be addressed to O.D. COMMUNICATIONS BIOLOGY | (2020) 3:524 | https://doi.org/10.1038/s42003-020-01246-z | www.nature.com/commsbio 13

Journal

Communications BiologySpringer Journals

Published: Sep 22, 2020

There are no references for this article.