Immunoregulatory Monocyte Subset Promotes Metastasis Associated With Therapeutic Intervention for Primary Tumor

Systemic and local inflammation associated with therapeutic intervention of primary tumor occasionally promotes metastatic recurrence in mouse and human. However, it remains unclear what types of immune cells are involved in this process. Here, we found that the tissue-repair-promoting Ym1+Ly6Chi monocyte subset expanded as a result of systemic and local inflammation induced by intravenous injection of lipopolysaccharide or resection of primary tumor and promoted lung metastasis originating from circulating tumor cells (CTCs). Deletion of this subset suppressed metastasis induced by the inflammation. Furthermore, transfer of Ym1+Ly6Chi monocytes into naïve mice promoted lung metastasis in the mice. Ym1+Ly6Chi monocytes highly expressed matrix metalloproteinase-9 (MMP-9) and CXCR4. MMP-9 inhibitor and CXCR4 antagonist decreased Ym1+Ly6Chi-monocyte-promoted lung metastasis. These findings indicate that Ym1+Ly6Chi monocytes are therapeutic target cells for metastasis originating from CTCs associated with systemic and local inflammation. In addition, these findings provide a novel predictive cellular biomarker for metastatic recurrence after intervention for primary tumor.


INTRODUCTION
Systemic and local inflammation caused by cancer therapy is now recognized as an important risk factor for cancer recurrence. Surgical resection of primary tumor, chemotherapy, or radiation therapy can awake dormant cancer cells and induce metastatic outgrowth in distant organs through inflammation (1)(2)(3)(4)(5)(6)(7). In addition to these cancer treatments, it has also been reported that inflammation caused by bacterial infection and cigarette smokeexposure, promotes cancer dormancy escape and metastasis (8,9). Such immune cells as neutrophils, macrophages, and monocytes are involved in cancer recurrence caused by inflammation. Recently, neutrophils have received increased attention with regard to their role in promoting cancer progression and metastasis associated with inflammation. For instance, neutrophils were reported to play critical roles in promoting lung metastases mediated by producing proinflammatory cytokines (10). Neutrophil extracellular traps (NETs) awake dormant cancer cells through interaction with cancer cells. NETs also trap circulating tumor cells (CTCs) and lead to increased formation of metastasis (9,11,12). In addition to neutrophils, monocytes and macrophages were also reported to be involved in cancer recurrence. The depletion of CD11b + macrophages reduces lung metastasis of breast cancer cells (13). Vascular endothelial growth factor A (VEGFA)-secreting macrophages promote the extravasation of cancer cells and lung metastasis (14). It was also reported that monocytes recruited to metastasis site by the CCL2-CCR2 axis differentiate into macrophages and promote extravasation and survival of cancer cells (14,15). A recent report indicated that not only neutrophils, but also monocytes, awake dormant cancer cells (7). Considering these reports, the types of immune cells involved in cancer progression and metastasis presumably depend on the context of inflammation or experimental models. However, immune cells involved in actual cancerrelated events in patients are not well understood.
Blood monocytes play critical roles in inflammation as a component of mononuclear phagocyte system. In the steadystate conditions, monocytes consist of two or three subpopulations in mouse or human, respectively (16,17). Classical monocytes (Ly6C hi CCR2 + CX3CR1 − in mouse, CD14 + CD16 − in human) are recruited into an inflamed site in a CCR2-dependent manner, and act as inflammation-promoting immune cells (14,18,19). In contrast, non-classical monocytes (Ly6C low CCR2 − CX3CR1 + in mouse, CD14 dim CD16 + in human) are differentiated from Ly6C hi monocytes in an Nr4A-dependent manner, patrol the vasculature during homeostasis, and contribute to cancer immunosurveillance (20). Intermediate dim → + monocytes (CD14 + CD16 + ) in human have been suggested to be responsible for the proliferation and stimulation of T cells (21). These monocyte subsets have been considered to coordinately engage in various immune responses in tissue injury or cancer. Recently, however, emergency hematopoiesis including monopoiesis during inflammation or other immune responses has been extensively studied, and several reports have identified bone marrow (BM)-derived atypical novel monocyte subsets that are rarely observed in the steady-state condition. In mouse, inflammation induced by microbial stimulation gives arise to neutrophil-like Ly6C hi monocytes derived from granulocyte-macrophage progenitors (GMPs), but not MDPs (22,23). Ceacam1 + Msr + Ly6C low monocytes called segregated-nucleus-containing atypical monocytes (SatM) emerge in lung of bleomycin-treated mouse and are involved in fibrosis (24). Ly6C hi MHCII hi Sca-1 hi monocytes arise in BM of acute gastrointestinal infected mouse and are considered to regulate immune response via the production of prostaglandin E2 and IL-10 (25). These reports suggest the possibility that a novel inflammation-related subset of monocytes can modulate cancer progression and metastasis associated with inflammation. However, details of such a monocyte subpopulation remain unknown.
We previously reported that GMP-derived atypical Ly6C hi monocytes characterized by Ym1 expression (Ly6C hi Ym1 + monocytes) are produced in BM during the recovery phase of tissue injury. These monocytes share some characteristics with granulocytes and exhibit the immunoregulatory phenotype that contributes to tissue repair and regeneration (22). Here, we show that not neutrophils, but Ym1 + Ly6C hi monocytes contribute to promoting metastasis caused by inflammation associated with intervention for primary tumor. These findings demonstrate that the mechanisms of tissue repair are closely related to metastasis and provide a novel therapeutic target for the metastasis.

Cell Lines
The murine melanoma cell line, B16F10 (Riken Cell Bank, Ibaraki, Japan), was maintained in RPMI1640 medium supplemented with 10% fetal bovine serum (FBS) and 100 units/mL of penicillin-streptomycin at 37°C in a humidified incubator with 5% CO 2 .

X-Ray Irradiation
B16 cells (1 × 10 6 cells) were implanted subcutaneously into the back of WT mice. Seven to eight days after implantation, mice under anesthesia were immobilized in a customized harness that allowed the implanted tumor to be exposed, whereas the remainder of the body was shielded by 3.5 cm of lead. Mice were irradiated in a Faxitron CP-160 irradiator (Faxitron Xray Corporation).

Quantitative RT-PCR (qRT-PCR)
For the analysis of mRNA levels in Ym1 + -or Ym1 -Mo, sorted Ym1 + -or Ym1 -Mo RNA was extracted and converted into cDNA, and qRT-PCR was performed on the cDNA as above. Expression levels were normalized to 18s rRNA. The following primer sequences were used for each gene:

RNA-Sequencing
Sorted cells were lysed and their total RNAs were extracted with RNeasy Mini kit (QIAGEN). Five hundred picograms of total RNA was subjected to DNA library preparation for RNA sequencing analysis using SMART-Seq v4 Ultra Low Input RNA Kit (TAKARA) and Nextera XT DNA Library Preparation Kit (Illumina). Sequencing was performed on a NextSeq 500 sequencer (Illumina) in the 75-bp single-end read mode. Data with the fragments per kilobase of exon per million reads (FPKM) were used for further analysis after mapping of the sequence reads. PCA analysis of RNA-sequencing was performed using AltAnalyze. The R package limma was used to identify differentially expressed genes. For PCA analysis, RNA-seq data in BM naïve monocytes, Ym1 + Ly6C hi monocytes, and Ym1 -Ly6C hi monocytes were retrieved from the Gene Expression Omnibus database (accession number GSE118032) (22).

Invasion Assay
Ly6C hi monocytes (2 or 5 × 10 6 /mL) sorted from LPS-treated WTor Ym1-Venus mice were incubated in serum-free medium (Advanced-RPMI1640, Thermo Fischer Scientific) at 37°C in 5% CO 2 for 24 h. The culture supernatant was centrifugated (10,000 ×g, 30 min at 4°C). The supernatants were collected for the invasion assay. Human melanoma cell line A375 (ATCC-CRL-1619, 1 × 10 6 / mL) was suspended in serum-free medium (RPMI1640, Wako) and added into the upper chamber of a 24-well Transwell chamber that had been coated with Matrigel (Corning, BioCoat 354480) in the presence or absence of monocyte-culture supernatant. The lower chamber contained RPMI1640 containing 0.1% FBS as a chemoattractant. Assays were carried out at 37°C in 5% CO 2 for 24 h. At the end of the incubation, the non-invading cells on the upper surface of the filter were mechanically removed. The invading cells that migrated through the Matrigel and the 8-µm pore membrane, were fixed with 4% paraformaldehyde/PBS for 5 min, and stained with 0.1% crystal violet (WAKO) for 20 min. The proportion of invading cells was calculated using BZ-X710 software (Keyence).

Statistics
Data were analyzed either by analysis of variance (ANOVA) followed by multiple comparison, or by the t-test with Prism (GraphPad Software, CA). P values < 0.05 were considered significant.

Ly6C hi Monocytes, but Not Neutrophils, Promote Lung Metastasis Accelerated by Systemic Inflammation
Inflammation is one of the most important factors that promote cancer metastasis (2,(7)(8)(9)(10)(11)(28)(29)(30). The metastasis cascade involves multiple processes, including invasion of cancer cells into adjacent tissue, intravasation, survival in blood circulation, extravasation of circulating tumor cells (CTCs), and subsequent outgrowth at distant sites (31). Among these steps, the outgrowth of CTCs at distant sites was proven to be enhanced by systemic inflammation in an experimental metastasis model (8,10,11,29,30,32). However, the precise mechanisms of inflammationinduced metastasis originating from CTCs remain unknown. To explore these mechanisms, we first compared some forms of systemic inflammation induced by different Toll-like receptor (TLR) ligands from the perspective of promoting metastasis in an experimental metastasis model ( Figure 1A). Mice were treated with different TLR ligands, followed by intravenous (i.v.) injection of B16 melanoma cells. Consistent with previous reports (8,10,30), the systemic injection of lipopolysaccharide (LPS) via tail vein promoted the formation of metastatic foci of B16 melanoma cells originating from CTCs in lung ( Figures 1B,  C). The mRNA expression levels of premelanosome protein (Pmel) and dopachrome tautomerase (Dct) genes, both of which are highly expressed in B16 melanoma cells (33), were significantly elevated in the lungs of LPS-treated mice ( Figure  1D), indicating outgrowth of B16 melanoma cells in the lungs. On the other hand, CpG-ODN or Poly (I:C) had negligible effects on metastasis ( Figures 1B-D), indicating that systemicinflammation-induced enhancement of metastasis depends on the mode of inflammation.
It was reported that neutrophils and monocytes are involved in metastasis under inflammatory conditions (7,(9)(10)(11). In fact, both neutrophils and monocytes accumulated in lung in the early phase (Day1 to 2 after systemic injection of LPS) of inflammation (Supplemental Figures 1A, B). Therefore, we focused on the role of neutrophils and monocytes in the inflammation-induced promotion of metastasis. The depletion of neutrophils by anti-Ly6G monoclonal antibody (mAb) injection had no effects on lung metastasis ( Figures 1E, F, and Supplemental Figure 2). On the other hand, anti-Gr-1 mAb, which depletes both neutrophils and Ly6C hi monocytes, but not Ly6C low monocytes, suppressed the metastatic formation ( Figure 1G and Supplemental Figure 2), suggesting that Ly6C hi monocytes, but not neutrophils, contributed to promoting metastasis induced by systemic injection of LPS. We previously reported that BM and peripheral blood monocytes highly expressed CD204, a class A scavenger receptor (26), and that both Ly6C hi and Ly6C low monocytes but not neutrophils were specifically deleted in peripheral blood by diphtheria toxin (DT) injection in CD204-DTR mice (26) (Supplemental Figure 3). In these mice, the number of metastatic foci was decreased by DT injection ( Figures 1H-J), indicating that monocytes are responsible for the promotion of lung metastasis. The injection of anti-Gr-1 mAb did not increase inflammatory cytokine production induced by LPS ( Figure 1K), indicating that the suppression of metastasis by anti-Gr-1 mAb is not attributed to the suppression of inflammatory cytokine production.
Albrengues et al. recently reported that neutrophil extracellular traps (NETs) play a critical role in the awakening of dormant cancer cells and the growth of metastatic lesions in lung, when mice were injected intranasally (i.n.) with LPS (9). Thus, we compared i.n. and i.v. routes of LPS administration in terms of NET formation in lung. As was previously reported, the i.n. injection of LPS induced the citrullination of histone H3, a specific marker for NET formation in lung. On the other hand, the injection of LPS via tail vein never induced NET formation in lung ( Figures 1P,  Q). These results strongly suggest that NET formation is not attributed to inflammation-induced promotion of metastasis in the case of systemic injection of LPS.
To further confirm the role of Ly6C hi monocytes in systemicinflammation-induced metastasis originating from CTCs, we purified Ly6C hi monocytes from either naïve or LPS-injected mice and injected intravenously these monocytes into naïve mice. After that, we injected cancer cells ( Figure 1L). As shown in Figures 1M-O, Ly6C hi monocytes from LPSinjected mice (LPS Mo) facilitated the formation of metastatic foci in lungs, whereas Ly6C hi monocytes from naïve mice (Naïve Mo) did not. We counted the number of transferred monocytes and B16 cells in lung soon after injection of these cells. However, there was no significant difference in the cell number of these cells in the lungs between Naïve Mo-transferred-and LPS Motransferred mice (Supplemental Figure 4), suggesting functional difference between Naïve Mo and LPS Mo in lung. Taken together, the systemic injection of LPS provides Ly6C hi monocytes with the ability to promote metastasis.

Ym1 + Ly6C hi Monocyte Subset Plays a Vital Role in Lung Metastasis
We next sought to reveal the properties of Ly6C hi monocytes in mice treated with LPS. We previously identified a subpopulation of Ly6C hi monocytes that are characterized by a high expression of Ym1 (22). Ym1 + Ly6C hi monocytes greatly expanded in BM during the recovery phase of systemic inflammation induced by LPS administration or tissue injury. These monocytes infiltrating into an injured site exhibited immunoregulatory and tissuereparative phenotypes. These findings of the roles of Ym1 + Ly6C hi monocytes in tissue repair prompted us to speculate that Ym1 + Ly6C hi monocytes could play roles in systemic inflammation-induced metastasis. We first monitored the accumulation of Ym1 + Ly6C hi monocytes in lung after systemic inflammation by using Ym1-Venus mice. As expected, when Ym1-Venus mice were injected with LPS, a large number of Ym1 + Ly6C hi monocytes were accumulated in the lungs (Figures 2A-C). Intriguingly, a small number of Ym1 + Ly6C hi monocytes were found in the lungs of mice injected with either CpG-ODN or Poly(I:C) (Figures 2A-C), both of which had no effects on metastasis formation in lung ( Figures 1B-D). An increase in the number of Ym1 + Ly6C hi monocytes was also observed in mice injected with both LPS and cancer cells (Supplemental Figures 5A, B), suggesting the role of Ym1 + Ly6C hi monocytes in lung metastasis.
We previously generated Ym1-DTR mice in which Ym1expressing cells were deleted by DT injection (Supplemental Figure 6). As shown in Figures 2D-F, the transient deletion of Ym1-positive cells on Days 1 and 4 significantly suppressed lung metastasis induced by LPS injection. To further reveal the role of Ym1 + Ly6C hi monocytes in promoting metastasis originating from CTCs, we purified Ym1 + Ly6C hi or Ym1 -Ly6C hi monocytes from BM of LPS-treated Ym1-Venus mice and injected those cells into naïve mice. After that, we injected cancer cells (Figures 2G, H). The injection of Ym1 + Ly6C hi monocytes resulted in a large number of metastatic foci in lung compared with the injection of Ym1 -Ly6C hi monocytes ( Figures 2I, J). These results clearly indicate that Ym1 + Ly6C hi monocytes have the ability to promote lung metastasis.

Ym1 + Ly6C hi Monocytes Express Metastasis-Related Genes
To elucidate the mechanisms underlying the promotion of metastasis by Ym1 + Ly6C h i monocytes, we sought to characterize the Ym1-Venus + Ly6C hi monocyte subpopulation that accumulated in lung during systemic inflammation. Flow cytometry analysis revealed that the Ym1 + Ly6C hi monocyte subpopulation expressed the same levels of several monocyte surface markers as the Ym1 -Ly6C hi monocyte subpopulation ( Figure 3A). Next, we globally compared the mRNA expression profiles of Ym1 + Ly6C hi monocytes and Ym1 -Ly6C hi monocytes from lungs of LPS-treated mice by RNA sequencing analysis. PCA analysis demonstrated that the two monocyte subsets exhibited obvious differences in gene expression after infiltrating into lung ( Figure 3B). In addition, the gene expression of lung monocyte subsets clearly differed from previously reported that of BM monocyte subsets (22) ( Figure 3B).
While Chi3l3 (Ym1-coding gene), known as a marker of M2 macrophages (34), is highly expressed in Ym1 + Ly6C hi monocytes, Ym1 + Ly6C hi monocytes did not show higher expression of any other M2 genes (Supplemental Figure 7). Interestingly, the expression of several metastasis-related genes (35) such as Mmps, Vegf, Cox2, and Il1b was enhanced in Ym1 + Ly6C hi monocytes, while expression levels of inflammatory cytokines except Il1b were not different between two subsets ( Figures 3C, D and, Supplemental Figure 7). In addition to these genes, Lcn2, which is reported to enhance matrix metalloproteinase-9 (MMP-9) activity by stabilizing MMP-9 (36), was also expressed in higher levels in Ym1 + Ly6C hi monocytes. The high expression of those genes in Ym1 + Ly6C hi monocytes was also confirmed by PCR analysis ( Figure 3E).

MMP-9 Is Essential for Ly6C hi Monocyte-Promoting Lung Metastasis
MMP-9 plays an important role in the invasion and metastasis of cancer cells (37)(38)(39)(40). Thus, we next sought to examine the roles of MMP-9 in metastasis promotion by Ym1 + Ly6C hi monocytes. We first examined the protein levels of MMP-9 in the culture supernatant of purified Ly6C hi monocytes using gelatin zymography. The higher protein levels of proMMP-9, latent form of MMP-9 was observed in the culture supernatant of LPS Mo compared with Naïve Mo ( Figure 4A). We also confirmed that Ym1 + Ly6C hi monocytes in LPS Mo showed higher protein levels of proMMP-9 than Ym1 -Ly6C hi monocytes in LPS Mo ( Figure 4A). We next demonstrated whether the culture supernatant of LPS Mo promoted cancer cell invasion in vitro using the Matrigel invasion assay. The culture supernatant of LPS Mo induced a significant increase in cancer cell invasion compared with that of Naïve Mo ( Figure 4B). We next demonstrated the in vivo contribution of Ym1 + Ly6C hi monocyte derived-MMP-9 to metastasis. The sequential injection of MMP-9 inhibitor suppressed LPS-promoted lung metastasis (Supplemental Figure 8). However, previous report suggested that MMP-9 induces not only the awakening of dormant cancer cells in lung through extracellular matrix remodeling (ECM) but also outgrowth of tumor growth in the late phase of metastasis (9). Thus, to inhibit in vivo MMP-9 enzymatic activity only at early time points, WT mice were injected with MMP-9 inhibitor on Days -1, 0, and 1 only ( Figure 4C). Treatment of WT mice with the inhibitor at these time points also suppressed the number of metastatic foci (Figures 4D, E). The metastasis-promoting effects of injecting LPS Mo were also canceled by the early injection of the inhibitor (Figures 4F, G). We further tried to demonstrate that Ym1 + Ly6C hi monocyte-derived MMP-9 is responsible for the progression of metastasis by using lipocalin 2 (Lcn2) -deficient mice (27). As described above, Lcn2 is reported to be responsible for stability of MMP-9 (36). In fact, the protein levels of MMP-9 in Lcn2 -/-LPS Mo were lower than those in Lcn2 +/-LPS Mo ( Figure  4H). The injection of Lcn2 -/-LPS Mo resulted in the reduced number of metastatic foci in lung compared with the case of Lcn2 +/-LPS Mo injection ( Figure 4I). Taken together, these results indicate that Ym1 + Ly6C hi monocyte derived-MMP-9 has a strong impact on the promotion of lung metastasis.

Ym1 + Ly6C hi Monocytes Contribute to the Promotion of Lung Metastasis Induced by Tumor Resection
It was reported that the resection of primary tumor triggers a high frequency of tumor-dormancy escape and metastatic relapse in cancer (41)(42)(43). In mouse, inflammation associated with surgery triggered the outgrowth of distinct tumors or promoted metastasis (2,7). Thus, we first sought to examine whether Ym1 + Ly6C hi monocytes are involved in the formation of metastatic foci after resection of tumors. B16 cells were inoculated subcutaneously. Subcutaneous tumors were removed by resection followed by i.v. injection of B16 cells 24 h after the resection. As shown in Figures 5A, B, the proportion of Ym1 + Ly6C hi monocytes, but not total Ly6C hi monocytes, was drastically increased within 2 days after tumor resection. In addition to resection of primary tumor, it has already been reported that radiation exposure to primary tumor promotes cancer metastasis in mouse (44,45). As expected, an increase in Ym1 + Ly6C hi monocytes was also observed in tumor-bearing mice treated with irradiation therapy ( Figure 5C). Furthermore, we found that resection of the tumors promoted metastasis originating from CTCs ( Figure 5D). We then tried to identify the immune cells responsible for the promotion of metastasis induced by resection. As shown in Figure 5E, anti-Ly6G mAb had no effects on metastasis. On the other hand, anti-Gr-1 mAb suppressed the metastatic formation. Taken together, we concluded that Ym1 + Ly6C hi monocytes contributed to the promotion of metastasis induced by tumor resection or radiation exposure to tumor.

Inhibition of CXCR4 Signaling Reduces Lung Metastasis by Ym1 + Ly6C hi Monocytes
It is presumably critical for Ym1 + Ly6C hi monocytes to accumulate in situ for the promotion of lung metastasis. Chong et al. reported that the lung accumulation of Ly6C hi monocytes is dependent on the CXCR4-CXCL12 signaling axis in the LPS-induced inflammation state (46). In fact, CXCR4 expression was observed on Ly6C hi monocytes, but not B16 cells ( Figure 6A). AMD3100, a CXCR4 antagonist, inhibited the accumulation of Ly6C hi monocytes, but not neutrophils ( Figure 6B) in lung associated with systemic inflammation. We then examined whether AMD3100 suppressed lung metastasis promoted by the surgical resection of primary tumor resection. The promotion of lung metastasis was inhibited by the treatment with AMD3100 ( Figures 6C, D). These findings suggest that CXCR4 is a novel therapeutic target for controlling lung metastasis associated with surgical intervention for cancer by inhibiting the accumulation of Ym + Ly6C hi monocytes in potential metastatic organs.

DISCUSSION
In this study, we demonstrated that Ym1 + Ly6C hi monocytes, but not neutrophils, promote inflammation-induced lung metastasis associated with intervention for primary tumor. Inhibition of the accumulation of Ym1 + Ly6C hi monocytes in lung or inhibition of MMP9 reduced lung metastasis, suggesting that Ym1 + Ly6C hi monocytes are a therapeutic target for the metastasis. Recent reports highlighted the critical roles of NETs in metastasis associated with inflammation in mouse. For instance, intranasal injection of LPS triggers marked neutrophil recruitment, detects numerous NET formation, and promotes lung metastasis (9). The induction of peritonitis results in NET formation in liver and facilitates liver metastasis (11). The inhibition of NET formation with DNase and neutrophil elastase inhibitor suppresses lung and liver metastases (11). In these experimental models, inflammation is elicited by local infection or tissue injury. However, such local inflammation at a distant site from the primary tumor dose not seems to occur in tumor-bearing patients. In this study, we demonstrated that a novel monocyte subset, but not neutrophils, play a critical role in the progression of metastasis associated with inflammation caused by primary tumor resection or irradiation. Although neutrophils were recruited into lung in this type of inflammation, NET formation hardly occurred at this site.
These results indicate that the context of inflammation determines the types of immune cells primarily responsible for promoting metastasis, and suggest that therapeutic target cells for metastasis prevention need to be carefully selected according to the actual situation of cancer patients. Recent reports have shown that a functionally distinct monocyte subset is differentiated in BM in response to certain inflammatory stimuli (16,23,25). This monocyte subset is differentiated from granulocyte-monocyte progenitors (GMPs) and shares some characteristics with granulocytes. In line with this concept, we showed in our previous report that immunoregulatory Ym1 + Ly6C hi monocytes were generated from GMPs in BM during the recovery phase of tissue injury and contributed to inflammatory response associated with tissue repair (22). It is well known that tissue repair and wound healing after tissue injury consist of multiple processes including regeneration of parenchymal cells, ECM remodeling and angiogenesis (47). These processes also contribute to tumor progression and metastasis (32). For example, the high expression of wound-response gene increased the risk of metastasis in human (48). Increased MMPs, which are critical molecules for ECM remodeling in injured tissue, are correlated with low overall survival rate in cancer patients (49). In fact, lung metastasis of B16F10 was decreased in MMP9-deficient mouse (40). Regarding the regeneration of blood vessels in wound tissue, the hypoxic condition is detected by hypoxia inducible factor alpha (HIF1a), which induces the production of vascular endothelial growth factor (VEGF) to promote angiogenesis (50). The same mechanisms also apply in cancer tissue, and neutralizing antibody against VEGF or VEGF receptor was reported to inhibit lung metastasis (51,52). In this respect, the tissue repair process shares common features with cancer progression and metastasis. This study, together with our previous report, reveals a link between tissue repair and cancer progression from the perspective of cell population. In the future, the relationship between these monocytes and primary tumor progression should be investigated.
Patients who undergo resection of primary tumors face the risk of metastatic recurrence that peaks sharply 12 to 18 months after surgery (41)(42)(43). Although the cause of early metastatic relapse has been debated, a recent report has indicated that systemic inflammation induced by resection triggers the outgrowth of distant dormant tumors in mouse, implicating that Ly6C hi monocytes are essential effector cells for the induction of the outgrowth (7). However, it is unclear what monocyte subset is involved in tumor progression induced by resection. In the present study, we showed that immunoregulatory Ym1 + Ly6C hi monocytes promoted lung metastasis of CTCs in an MMP-9-and CXCR4dependent manner. MMP-9 secreted by Ym1 + Ly6C hi monocytes may degrade extracellular matrix and promote infiltration of cancer cells into metastatic tissues.
These results suggest that the risk of metastatic recurrence after resection can be reduced by developing a therapeutic method targeting immunoregulatory Ym1 + Ly6C hi monocytes.
The prediction and prevention of metastasis is a vital clinical task in cancer treatment. A previous report suggested that the high expression of wound response signature in tumors is a predictor of poor patient survival and increased risk of metastasis in human (48). In this study, we observed the rapid increase of Ym1 + Ly6C hi monocyte numbers in peripheral blood in the early stage of lung metastasis. Given that the emergence of Ym1 + Ly6C hi monocytes is implicated in the initial step of wound healing or tissue repair, monitoring of these monocytes in peripheral blood after surgical intervention for primary tumor may be a useful predictive cellular biomarker for metastasis. For this purpose, the human counterpart of mouse Ym1 + Ly6C hi monocytes should be identified. A recent study that employed single-cell RNA sequencing uncovered a four-monocyte population in healthy human peripheral blood (53). Furthermore, in cancer patients, monocytic myeloid-derived suppressor cells have emerged as the major negative regulators of immune responses (54)(55)(56). In any case, estimation of the emergence of novel and atypical monocyte subsets associated with inflammation is important for the development of therapeutic strategies for metastasis.

DATA AVAILABILITY STATEMENT
The RNA-seq data has been deposited to the GEO -accession number is GSE174199.

ETHICS STATEMENT
The animal study was reviewed and approved by Tokyo University of Pharmacy and Life Sciences Animal Use Committee.