Hepato‐entrained B220+ CD11c+ NK1.1+ cells regulate pre‐metastatic niche formation in the lung

Abstract Primary tumours establish metastases by interfering with distinct organs. In pre‐metastatic organs, a tumour‐friendly microenvironment supports metastatic cells and is prepared by many factors including tissue resident cells, bone marrow‐derived cells and abundant fibrinogen depositions. However, other components are unclear. Here, we show that a third organ, originally regarded as a bystander, plays an important role in metastasis by directly affecting the pre‐metastatic soil. In our model system, the liver participated in lung metastasis as a leucocyte supplier. These liver‐derived leucocytes displayed liver‐like characteristics and, thus, were designated hepato‐entrained leucocytes (HepELs). HepELs had high expression levels of coagulation factor X (FX) and vitronectin (Vtn) and relocated to fibrinogen‐rich hyperpermeable regions in pre‐metastatic lungs; the cells then switched their expression from Vtn to thrombospondin, both of which were fibrinogen‐binding proteins. Cell surface marker analysis revealed that HepELs contained B220+ CD11c+ NK1.1+ cells. In addition, an injection of B220+ CD11c+ NK1.1+ cells successfully eliminated fibrinogen depositions in pre‐metastatic lungs via FX. Moreover, B220+ CD11c+ NK1.1+ cells demonstrated anti‐metastatic tumour ability with IFNγ induction. These findings indicate that liver‐primed B220+ CD11c+ NK1.1+ cells suppress lung metastasis.

In the manuscript by S. Hiratsuka, the authors have identified a rare NK cells (B220+CD11c+NKL1+) which are derived in the liver and in turn these cells are recruited to the pre-metastatic lung where these cells support the upregulation of interferon and convert fibronogen to vitronectin that together suppress lung metastasis. This work opens up a new avenue of research supportive of the systemic effects of cancer beyond the crosstalk between the primary tumor, bone marrow and pre-metastatic/metastatic lung. The work emphasizes the need to focus on new explorations to alter pre-metastatic sites preventing metastasis. Despite the extreme novelty of this work there are questions that remain.
It is unclear how tumor cells activate these HepELs. Are the HepELs only found in the liver and where in the liver are they found. It seems the original publication describing these cells describe these NK cells in lymph nodes as well. What does it mean to have Factor X in these specialized NK cells. Are there other coagulation factors packaged inside. Do they have a role in clotting. How are these cells recruited specifically in the lung. Once in the lung, is interferon release derived only from the HepELs or other cells contribute to this process. Interferon can promote the immune system to attack the tumors but also promote inflammation, which can support metastasis. The authors need to explain IFN's role further. How do the HepEls convert fibrinogen to vitronectin. Do these NK cells bind other ECM molecules. What molecular pathways are ongoing here. In most cases of the models used here, metastasis progresses. So when do these cells interfere with metastasis, a time course study (not just 72 hours) on these particular cells in the liver, lung, blood and bone marrow are warranted in these studies. When metastases progresses, do these cells remain or disappear. Do these cells get re-educated, or do other cells override the effects of the HepEls The knockdown and OE of Factor X in these particular cells will greatly enhance the novelty and function of these cells during metastasis.
In addition, it is very startling that the authors state that B16, LLC and EO771 implanted at primary sites do not promote lung metastasis. It is very common to get lungs mets from B16 and LLC tumors whihc were subcutaneously implanted and for EO771 implanted in the mammary gland to metastasize to the lungs. Furthermore, EO771 cells implanted in the spleen readily grow in the liver.
The authors should comment on the role of HepELs in the primary tumors.
Other comments: Figure 1B, Lymph nodes should be considered 1E, it appears that FX and CD45 colocalize predominantly, and do not merge that much, comment Figure 2: 2B, it is very strange that the laser technology also photoconverts rare cells within a region, it should be a patch for the entire laser area. Comment The authors ignore their own data which show that CD4+CD45+ cells also increase in response to TCM and they also have Factor X associated with them. Comment The authors need to comments on other works on NK cells in particular (such as Andy Moeller and colleagues) that promote the progression of the premetastatic niche.  Figure 4, no Y axis labels in D Figure 5, the authors should have isolated these unique cells from the lung itself where they are recruited and not just the liver where they exit with time. This is a very interesting study on a timely and impactful issue. The authors of this manuscript study the potential role of a "third organ" (in this case the liver) in the control of the premetastatic niche in the lung. These authors show that "liver-educated" leukocytes that they name as HepELs and characterized by B220+CD11c+NK1.1+ surface markers, are critical anti-metastatic cells in the lung. In the revised manuscript, Fig  Comments: Are the HepELs only found in the liver and where in the liver are they found.
Answer: We searched HepELs in various organs in tumour-bearing mice. As shown in Appendix Fig S7, HepELs were prominently found in the lung and liver in no-tumour mouse but in the lung, liver, peripheral blood, and tumour tissue in tumour-bearing mouse. Further qPCR analysis revealed that the cells in the liver in 3mm-tumour-bearing mouse displayed remarkably high levels of FX expression. In 10mm-tumour bearing mouse, HepELs isolated from the lung, liver, and tumour tissue are also showed high expression levels of FX. Our immunohistochemical data showed that liver HepELs in tumour-bearing mouse were observed in a diffusely-scattered pattern.
Comments: It seems the original publication describing these cells describe these NK cells in lymph nodes as well. What does it mean to have Factor X in these specialized NK cells.
Answer: First, CD45 leukocytes in the liver and lung expressed FX and but not CD45 leukocytes in the lymph node. Remarkably, the FX expression levels of CD45 leukocytes in the liver and lung were enhanced in the tumour-bearing state (revised Fig.   1B). Next, B220 + CD11c + NK1.1 + cells in various organs such as lung, liver, peripheral blood, bone marrow, lymph node and the primary tumour were investigated. We collected samples 2, 7, and 14 days after the tumour cell implantation; their approximate tumour sizes were 0 mm (2 days), 3 mm (7 days), and 10 mm (14 days) in diameter, respectively. The sorted cells were used for the qPCR analyses (Appendix Fig. S7, upper) We also added the functional analyses data to show FX-dependent coagulation in HepELs as described later.
We added this data shown in revised Appendix Fig. S7 and sentences as below.
B220 + CD11c + NK1.1 + cells in various organs such as lung, liver, peripheral blood, bone marrow, lymph node and the primary tumour were investigated. We collected samples 2, 7, and 14 days after the tumour cell implantation; their approximate tumour sizes were 0 mm (2 days), 3 mm (7 days), and 10 mm (14 days) in diameter, respectively. Among them, the FX expression levels in B220 + CD11c + NK1.1 + cells isolated from the liver of 3mm tumour-bearing mice were remarkably high (Appendix Fig. S7, upper panel). We would like to note that the FX expressions in the cells derived from the lung and tumour tissues in 10mm tumour-bearing mice were also observed. (Appendix Fig. S7, upper panel).
Comments: Are there other coagulation factors packaged inside. Do they have a role in clotting.
Answer: Based on the array data, coagulation factor 5 and factor 13 as well as factor 10 were packed in peripheral blood leukocytes in tumour-bearing mouse (For reviewer   Table). Among them, coagulation factor 10 was upregulated in tomour-beaing mice. To determine the coagulation activity, we calculated prothrombin time (PT) using the B220 + CD11c + NK1.1 + cells derived from the liver and lung in tumour-bearing mice. We concluded that B220 + CD11c + NK1.1 + cells play a role in clotting because addition of these cells in plasma reduced PT. (data statement was shown in the responses for FX-OE-HepELs).
We added sentences in the revised text as below.
Our microarray data indicates that coagulation factor 5 and factor 13 as well as factor 10 (FX) were packed in peripheral blood leukocytes in tumour-bearing mouse (Appendix Table S1 : GSE76506). Among them, FX was upregulated in tumour-beaing mice. We set up a coagulation assay system to measure prothrombin time (PT) of B220 + CD11c + NK1.1 + cells derived from tumour-bearing mice. To determine PT, we recorded absorbance at 671 nm after mixing HemosIL RecombiPlasTin with samples.
Comment: How are these cells recruited specifically in the lung.
Answer: We consider that HepELs leaving the liver go into the circulatory system, and are trapped in fibrinogen-enriched niche in the lung by an interaction with fibrinogen binding molecules (such as Vtn or TSP, shown in Fig 5) expressed in the cell. A neutralizing anti-Vtn Ab significantly blocked the binding of liver HepELs to fibrinogen coated plate ( Fig 7D). In addition, we confirmed that a neutralizing anti-Vtn and anti-TSP Ab blocked the binding of lung HepELs to a fibrinogen coated plate ( Fig   7E). We also tested the binding ability of HepELs to other ECM components as a reviewer requested later (detailed response is shown later) Comments: Once in the lung, is interferon release derived only from the HepELs or other cells contribute to this process. Interferon can promote the immune system to attack the tumours but also promote inflammation which can support metastasis. The We added the following sentences in the section of discussion.
It has been reported that IFN-γ has dual opposite roles as anti-metastatic immune response and promotion of metastatic ability of tumour cells via activated nuclear factor Our data present that HepELs were able to attach to collagen I or FN although their affinities were not as high as that to fibrinogen (shown in revised Fig.7E). Moreover, the HepEL-Fibronectin/Collagen I interactions were Vtn/TSP independent.
We added sentences in the revised text as below.
We focused on TSP as a ligand molecule of fibrinogen because TSP was upregulated in lung HepELs but not in liver HepELs (Appendix Table S2). Our results exhibit that anti-Vtn Ab inhibited binding of lung HepELs to a fibrinogen coated plate. Similarly, neutralizing anti-TSP Ab blocked the binding of lung HepELs ( Fig 7E). Then, we examined binding abilities of HepELs to other ECM such as collagen I and fibronectin.(FN) Our data present that HepELs were able to attach to collagen I or FN although their affinities were not as high as that to fibrinogen (Fig.7E). Moreover, the HepEL-FN/collagen I interactions were Vtn/TSP independent.
Comment: What molecular pathways are ongoing here. In most cases of the models used here, metastasis progresses. So when do these cells interfere with metastasis, a time course study (not just 72 hours) on these particular cells in the liver, lung, blood and bone marrow are warranted in these studies. When metastases progresses, do these cells remain or disappear.
Answer: Thank you for the reviewer's comment. As shown above, we found that CCL2 and CXCL1 stimulated FX in HepELs (Revised Fig. 7C). In addition, We investigated involvement of any of transcription factors peculiar to liver by using siRNA transfection or electroporation technique. Our data, shown in "For Reviewer data 2", suggest that FX expression in HepELs are regulated by Foxa1, Cebp-α and Rela, because knockdown of each one of the three transcription factors achieved 35-50% reduction of FX expression. We added a sentence as shown below in discussion in the revised manuscript.
Deciphering molecular mechanisms responsible for the FX induction in HepELs in a liver environment is ongoing. Based on our data, we speculate that FX expression in HepELs is regulated by multiple transcription factors such as Foxa1, Cebp-α and Rela,(unpublished data).
We analyzed populations of B220 + CD11c + NK1.1 + cells in tumour-bearing mouse liver, lung, bone marrow, peripheral blood, lymph node, and tumour tissues to find out that the triple positive cells were observed in tumour-bearing liver, blood and tumour tissues, as stated above. In addition, the sorted cells showed remarkable upregulation of FX in the liver, lung and tumour tissues in the presence of a primary tumour (revised Appendix Fig. S7). However, the number of HepELs gradually decrease during tumour progression (please compare HepELs in 3mm-vs 10mm-size tumour-bearing mice).
Comments: Do these cells get re-educated, or do other cells override the effects of the HepEls Answer: The HepELs, obtained from primary tumour-stimulated mice and cultured overnight, showed low level of FX expression (please see NoCM column in Fig 7B).
We were able to regain the FX expression in the cells by using Liver-CM in the culture media (please see LiCM column). Thus, we consider that the HepELs can be re-educated in terms of FX expression. CD4 + cells may have an ability to override the effects of HepELs because they have FX expression in tumour-bearing mice. However, our animal study showed these cells supported metastasis (For reviewer data 1). Thus, it is very difficult to conclude that other cells do override the effects of the HepELs.
Comment: The knockdown and OE of Factor X in these particular cells will greatly enhance the novelty and function of these cells during metastasis.
Answer: The HepELs prepared from FX-knockdown mice (95% knockdown mouse) have very low levels of FX so that the cells can be used as FX knockdown cells. Fig. 8B presents FX knockdown increased in our tumour cell homing assay. Because FX transgenic mouse is not available, we tried to obtain FX-OE equivalent cells in a different way. We succeeded to prepare FX-OE equivalent cells by introducing recombinant FX conjugated with polyethylene glycol (PEG)-lipid. We tried to obtain FX-overexpressed (OE) HepELs and succeeded to obtain those cells. We first characterized our FX built HepELs in clotting assay to measure prothrombin time (PT).
FX-OE HepELs displayed shorter PT in the assay. (please see the revised text shown below). Then, the FX-OE HepELs were applied in the metastasis assay. The number of metastatic rhodamin-labeled tumour cells was decreased in tumour-bearing lungs after injection of lung HepELs and FX-OE HepELs enhanced the inhibitory activity of HepELs ( Fig 8C).
We added sentences as below and revised Fig. 8C.
Comments: In addition, it is very startling that the authors state that B16, LLC and EO771 implanted at primary sites do not promote lung metastasis. It is very common to get lungs mets from B16 and LLC tumours which were subcutaneously implanted and for EO771 implanted in the mammary gland to metastasize to the lungs. Furthermore, EO771 cells implanted in the spleen readily grow in the liver.
Answer: We would like to thank for the reviewer's attention. We would like to emphasize that in our assay system we have never observed macro-and micrometastasis in the pre-metastatic organs. Lung metastasis is observed only after the primary tumour resection, when B16, LLC, or E0771 are subcutaneously implanted.
Because these cells readily metastasize when they are implanted in the mammary fad pad or injected into the tail vein, there is no doubt that location of the primary tumour is one of the most important factors. We also would like to add that 3LL cells, a subline of LLC, relatively easily accomplish lung metastasis even in the case of subcutaneous implantation. We clearly stated this point with references in the revised text and methods section as shown below.

Text
In this study, a key point of our pre-metastatic model system is that spontaneous metastasis from the primary site was observed only after the primary tumour resection, although an intravenous injection of these cells easily attained lung metastasis Answer: In our in vitro study, the HepELs derived from TCM-primed mouse attacked tumour cells in vitro (Fig. 4C-E), suggesting that HepELs potentially have anti-metastatic activity. However, given the fact that the HepELs population in tumor tissue decreased slightly in the primary tumour growth (Appendix Fig. S7), the effect of HepELs is limited.
We added sentences as shown below in discussion of the revised text.
The B220 + CD11c + NK1.1 + HepELs derived from TCM-primed mouse attacked tumour cells in vitro (Fig. 4C-E). Moreover, they were found in the primary tumor, and they became FX + in the later stage of tumour (Appendix Fig. S7). These data suggest that HepELs potentially have anti-metastatic activity. However, given the fact that the HepELs population in tumor tissue decreased slightly in the primary tumour growth (Appendix Fig. S7), the effect of HepELs against the primary tumour is limited.
Other comments:  Comment: 2B, it is very strange that the laser technology also photoconverts rare cells within a region, it should be a patch for the entire laser area.
Answer: We used deep UV lamp with a 436 nm bandpass filter as light source.
Resulting blue light was transmitted by an optical fiber which allows us to irradiate a particular region (a circle of 10-12 mm in diameter and 100-150 m in depth) in the liver. Figure 2B data were taken deep inside the liver so that the photoconversion did not occur in the liver stromal cells. We have observed the photoconversion in the surface liver cells (data not presented in this paper).
Comment: The authors ignore their own data which show that CD4+CD45+ cells also increase in response to TCM and they also have Factor X associated with them.
Answer: Thank you for your comment. Our data show that CD4 + T cells, which expressed FX, promoted metastasis (For reviewer data1). These results indicate that some activated immune cells might support metastasis. We added the statement in the discussion of revised manuscript.
Comment: The authors need to comments on other works on NK cells in particular (such as Andy Moeller and colleagues) that promote the progression of the premetastatic niche.
Answer: Thank you for your comment. We added the following sentence in the revised manuscript.
In addition, NK cell-type cytotoxic capacities of CD3 -NK1.1 + cells was reduced by In this study, all the photo-converted cells are expected to receive light exposure less than 5 min, because they are moving in the liver during the light exposure. This indicates that the photo-converted cells hold mixture of KikGR-green and kikGR-red.
Given the estimated light exposure time for each cell, vast majority was small amount of kikGR-red and large amount of kikGR-green. The cells locate slightly more upward than pure KikGR-green cells in the dot plot. In order to detect the cells holding small amount of kikGR-red, we set a region shown in Fig 3B. We also would like to note that we empirically confirmed that cells appeared in the region only after the light exposure.
Comment: Figure 4, no Y axis labels in D Answer: Thank you for your comment. We added label in Fig. 4D Comment: Figure 5, the authors should have isolated these unique cells from the lung itself where they are recruited and not just the liver where they exit with time.
Answer: Thank you for the reviewer's comment. We added data that B220 + CD11c + NK1.1 + cells from tumour-bearing mouse lung (lung HepELs) attached to fibrinogen via TSP (revised Fig. 7E), and suppressed metastatic tumour cell homing in the lungs (revised Fig. 8C).  Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are now globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending final editorial amendments.

For reviewer
Please submit your revised manuscript within two weeks. I look forward to seeing a revised form of your manuscript as soon as possible.
I look forward to reading a new revised version of your manuscript as soon as possible.
***** Reviewer's comments ***** Referee #3 (Remarks for Author): The revised version of the manuscript has improved greatly. Do the data meet the assumptions of the tests (e.g., normal distribution)? Describe any methods used to assess it.
Is there an estimate of variation within each group of data?
Is the variance similar between the groups that are being statistically compared?
Yes. The name of the test is described in method section (p28).
Yes, based upon experience. GraphPad Prism was used for the stastical analysis.

YOU MUST COMPLETE ALL CELLS WITH A PINK BACKGROUND ê
Sample size was not chosen to ensure adequate power to detect a pre--specified effect size. Sample size were chosen based on the similar studies from literatures. All sample sizes were described in the figure legends.
No statistical method was used to predetermine sample size(p28). Sample size were chosen based on the similar studies from literatures. All sample sizes were described in the figure legends.
Samples or animals were not excluded from the analysis.
Randomization was not used (p28). Animals were grouped based upon their genotype and littermates (p19).
The investigators were not blinded allocation during experiments and outcome assessment (p28).

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

Captions
The data shown in figures should satisfy the following conditions: Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).

B--Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner.
a statement of how many times the experiment shown was independently replicated in the laboratory.
Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects. This checklist is used to ensure good reporting standards and to improve the reproducibility of published results. These guidelines are consistent with the Principles and Guidelines for Reporting Preclinical Research issued by the NIH in 2014. Please follow the journal's authorship guidelines in preparing your manuscript.