DNA of neutrophil extracellular traps promotes cancer metastasis via CCDC25

Abstract

Neutrophil extracellular traps (NETs), which consist of chromatin DNA filaments coated with granule proteins, are released by neutrophils to trap microorganisms1,2,3. Recent studies have suggested that the DNA component of NETs (NET-DNA) is associated with cancer metastasis in mouse models4,5,6. However, the functional role and clinical importance of NET-DNA in metastasis in patients with cancer remain unclear. Here we show that NETs are abundant in the liver metastases of patients with breast and colon cancers, and that serum NETs can predict the occurrence of liver metastases in patients with early-stage breast cancer. NET-DNA acts as a chemotactic factor to attract cancer cells, rather than merely acting as a ‘trap’ for them; in several mouse models, NETs in the liver or lungs were found to attract cancer cells to form distant metastases. We identify the transmembrane protein CCDC25 as a NET-DNA receptor on cancer cells that senses extracellular DNA and subsequently activates the ILK–β-parvin pathway to enhance cell motility. NET-mediated metastasis is abrogated in CCDC25-knockout cells. Clinically, we show that the expression of CCDC25 on primary cancer cells is closely associated with a poor prognosis for patients. Overall, we describe a transmembrane DNA receptor that mediates NET-dependent metastasis, and suggest that targeting CCDC25 could be an appealing therapeutic strategy for the prevention of cancer metastasis.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: NETs in pre-metastatic livers promote cancer metastasis.
Fig. 2: NET-DNA binds to CCDC25 on tumour cells and facilitates their distant metastases.
Fig. 3: The transmembrane protein CCDC25 interacts with NET-DNA at its N terminus.
Fig. 4: CCDC25 interacts with ILK at its C terminus and signals through the ILK–β-parvin cascade.

Data availability

Source data are provided with this paper. All other data are available from the corresponding author upon reasonable request.

References

  1. 1.

    Brinkmann, V. et al. Neutrophil extracellular traps kill bacteria. Science 303, 1532–1535 (2004).

    ADS  CAS  Article  Google Scholar 

  2. 2.

    McDonald, B., Urrutia, R., Yipp, B. G., Jenne, C. N. & Kubes, P. Intravascular neutrophil extracellular traps capture bacteria from the bloodstream during sepsis. Cell Host Microbe 12, 324–333 (2012).

    CAS  Article  Google Scholar 

  3. 3.

    Fuchs, T. A. et al. Novel cell death program leads to neutrophil extracellular traps. J. Cell Biol. 176, 231–241 (2007).

    CAS  Article  Google Scholar 

  4. 4.

    Tohme, S. et al. Neutrophil extracellular traps promote the development and progression of liver metastases after surgical stress. Cancer Res. 76, 1367–1380 (2016).

    CAS  Article  Google Scholar 

  5. 5.

    Cools-Lartigue, J. et al. Neutrophil extracellular traps sequester circulating tumor cells and promote metastasis. J. Clin. Invest. 123, 3446–3458 (2013).

    CAS  Article  Google Scholar 

  6. 6.

    Cedervall, J., Zhang, Y. & Olsson, A. K. Tumor-induced NETosis as a risk factor for metastasis and organ failure. Cancer Res. 76, 4311–4315 (2016).

    CAS  Article  Google Scholar 

  7. 7.

    Albrengues, J. et al. Neutrophil extracellular traps produced during inflammation awaken dormant cancer cells in mice. Science 361, eaao4227 (2018).

    Article  Google Scholar 

  8. 8.

    Park, J. et al. Cancer cells induce metastasis-supporting neutrophil extracellular DNA traps. Sci. Transl. Med. 8, 361ra138 (2016).

    Article  Google Scholar 

  9. 9.

    Kessenbrock, K. et al. Netting neutrophils in autoimmune small-vessel vasculitis. Nat. Med. 15, 623–625 (2009).

    CAS  Article  Google Scholar 

  10. 10.

    Yu, F. et al. let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 131, 1109–1123 (2007).

    CAS  Article  Google Scholar 

  11. 11.

    Li, P. et al. PAD4 is essential for antibacterial innate immunity mediated by neutrophil extracellular traps. J. Exp. Med. 207, 1853–1862 (2010).

    CAS  Article  Google Scholar 

  12. 12.

    van der Windt, D. J. et al. Neutrophil extracellular traps promote inflammation and development of hepatocellular carcinoma in nonalcoholic steatohepatitis. Hepatology 68, 1347–1360 (2018).

    Article  Google Scholar 

  13. 13.

    Noubouossie, D. F. et al. In vitro activation of coagulation by human neutrophil DNA and histone proteins but not neutrophil extracellular traps. Blood 129, 1021–1029 (2017).

    CAS  Article  Google Scholar 

  14. 14.

    Warnatsch, A., Ioannou, M., Wang, Q. & Papayannopoulos, V. Neutrophil extracellular traps license macrophages for cytokine production in atherosclerosis. Science 349, 316–320 (2015).

    ADS  CAS  Article  Google Scholar 

  15. 15.

    Yang, H. et al. New insights into neutrophil extracellular traps: mechanisms of formation and role in inflammation. Front. Immunol. 7, 302 (2016).

    PubMed  PubMed Central  Google Scholar 

  16. 16.

    Lood, C. et al. Neutrophil extracellular traps enriched in oxidized mitochondrial DNA are interferogenic and contribute to lupus-like disease. Nat. Med. 22, 146–153 (2016).

    CAS  Article  Google Scholar 

  17. 17.

    Gehrke, N. et al. Oxidative damage of DNA confers resistance to cytosolic nuclease TREX1 degradation and potentiates STING-dependent immune sensing. Immunity 39, 482–495 (2013).

    CAS  Article  Google Scholar 

  18. 18.

    Berlow, R. B., Dyson, H. J. & Wright, P. E. Hypersensitive termination of the hypoxic response by a disordered protein switch. Nature 543, 447–451 (2017).

    ADS  CAS  Article  Google Scholar 

  19. 19.

    Chen, F. et al. Extracellular vesicle-packaged HIF-1α-stabilizing lncRNA from tumour-associated macrophages regulates aerobic glycolysis of breast cancer cells. Nat. Cell Biol. 21, 498–510 (2019).

    CAS  Article  Google Scholar 

  20. 20.

    Schuberth, C. & Buchberger, A. Membrane-bound Ubx2 recruits Cdc48 to ubiquitin ligases and their substrates to ensure efficient ER-associated protein degradation. Nat. Cell Biol. 7, 999–1006 (2005).

    CAS  Article  Google Scholar 

  21. 21.

    Yipp, B. G. et al. Infection-induced NETosis is a dynamic process involving neutrophil multitasking in vivo. Nat. Med. 18, 1386–1393 (2012).

    CAS  Article  Google Scholar 

  22. 22.

    Nie, M. et al. Neutrophil extracellular traps induced by IL8 promote diffuse large B-cell lymphoma progression via the TLR9 signaling. Clin. Cancer Res. 25, 1867–1879 (2018).

    Article  Google Scholar 

  23. 23.

    Su, S. et al. CD10+GPR77+ cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness. Cell 172, 841–856.e16 (2018).

    CAS  Article  Google Scholar 

  24. 24.

    Baer, C. et al. Suppression of microRNA activity amplifies IFN-γ-induced macrophage activation and promotes anti-tumour immunity. Nat. Cell Biol. 18, 790–802 (2016).

    CAS  Article  Google Scholar 

  25. 25.

    Wang, C. Y. et al. ZIP8 is an iron and zinc transporter whose cell-surface expression is up-regulated by cellular iron loading. J. Biol. Chem. 287, 34032–34043 (2012).

    CAS  Article  Google Scholar 

  26. 26.

    Roberts, T. L. et al. HIN-200 proteins regulate caspase activation in response to foreign cytoplasmic DNA. Science 323, 1057–1060 (2009).

    ADS  CAS  Article  Google Scholar 

  27. 27.

    Zhang, Z. et al. The helicase DDX41 senses intracellular DNA mediated by the adaptor STING in dendritic cells. Nat. Immunol. 12, 959–965 (2011); correction 13, 196 (2012).

    CAS  Article  Google Scholar 

  28. 28.

    Abe, T. et al. STING recognition of cytoplasmic DNA instigates cellular defense. Mol. Cell 50, 5–15 (2013).

    CAS  Article  Google Scholar 

  29. 29.

    Xu, R. H. et al. NANOG is a direct target of TGFβ/activin-mediated SMAD signaling in human ESCs. Cell Stem Cell 3, 196–206 (2008).

    CAS  Article  Google Scholar 

  30. 30.

    Zhang, Z. et al. DDX1, DDX21, and DHX36 helicases form a complex with the adaptor molecule TRIF to sense dsRNA in dendritic cells. Immunity 34, 866–878 (2011).

    CAS  Article  Google Scholar 

  31. 31.

    Jensen, R. B., Carreira, A. & Kowalczykowski, S. C. Purified human BRCA2 stimulates RAD51-mediated recombination. Nature 467, 678–683 (2010).

    ADS  CAS  Article  Google Scholar 

  32. 32.

    Lu, H., Pannicke, U., Schwarz, K. & Lieber, M. R. Length-dependent binding of human XLF to DNA and stimulation of XRCC4·DNA ligase IV activity. J. Biol. Chem. 282, 11155–11162 (2007).

    CAS  Article  Google Scholar 

  33. 33.

    Yamamoto, Y. & Sakisaka, T. Molecular machinery for insertion of tail-anchored membrane proteins into the endoplasmic reticulum membrane in mammalian cells. Mol. Cell 48, 387–397 (2012).

    CAS  Article  Google Scholar 

  34. 34.

    Hotz, M. & Nelson, W. J. Pumilio-dependent localization of mRNAs at the cell front coordinates multiple pathways required for chemotaxis. Nat. Commun. 8, 1366 (2017).

    ADS  Article  Google Scholar 

  35. 35.

    Zhu, Y., Wang, G. Z., Cingöz, O. & Goff, S. P. NP220 mediates silencing of unintegrated retroviral DNA. Nature 564, 278–282 (2018).

    ADS  CAS  Article  Google Scholar 

  36. 36.

    Li, M. et al. The histone modification reader ZCWPW1 is required for meiosis prophase I in male but not in female mice. Sci. Adv. 5, eaax1101 (2019).

    ADS  Article  Google Scholar 

  37. 37.

    Tvorogov, D. et al. Effective suppression of vascular network formation by combination of antibodies blocking VEGFR ligand binding and receptor dimerization. Cancer Cell 18, 630–640 (2010).

    CAS  Article  Google Scholar 

  38. 38.

    Shibue, T., Brooks, M. W. & Weinberg, R. A. An integrin-linked machinery of cytoskeletal regulation that enables experimental tumor initiation and metastatic colonization. Cancer Cell 24, 481–498 (2013).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

This work was supported by grants from the National Key Research and Development Program of China (2016YFC1302300, 2017YFA0106300), the Natural Science Foundation of China (81621004, 81490750, 81720108029, 81930081, 91942309, 81672614, 81802643, 81902688, 81902699, 81972465, 81802645, 91940305), Guangdong Science and Technology Department (2017B030314026), Clinical Innovation Research Program of Guangzhou Regenerative Medicine and Health Guangdong Laboratory (2018GZR0201004), Guangzhou Science Technology and Innovation Commission (201803040015, 201508020008, 201508020249), Natural Science Foundation of Guangdong Province (2016A030306023, 2017A030313878, 2018A030310085, 2019A1515011485), Pearl River S&T Nova Program of Guangzhou (201710010083), Tip-top Scientific and Technical Innovative Youth Talents of Guangdong Special Support Program (2016TQ03R553) and the China Postdoctoral Science Foundation (BX20190396, 2019M663270, 2018M640868, 2019TQ0367, 2019M660228). This research is partly supported by Fountain-Valley Life Sciences Fund of University of Chinese Academy of Sciences Education Foundation.

Author information

Affiliations

Authors

Contributions

L.Y., Q.L., S.S. and E.S. conceived the ideas, designed the experiments and wrote the manuscript. L.Y. performed most of the experiments and analysed the data. The bio-layer interferometry assays were reproduced independently by X.Z. The flow cytometry assays and three-dimensional coculture assays were performed by L.Y. and X.L. and were reproduced by J. Li. The adhesion assays and Boyden chamber assays were performed by L.Y. and B.Z. and reproduced by Y.X. Chemotaxis experiments were performed by J.C. and reproduced by X.C. and J. Liu. Immunoblotting assays were reproduced independently by H.L. Q.L. and E.S. provided samples from patients for clinical data analysis. D.H. and F.C. performed the NET quantification independently. All authors contributed to the revision of the manuscript.

Corresponding authors

Correspondence to Shicheng Su or Erwei Song.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Ilaria Malanchi and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 NETs are predominantly presented in liver metastases of breast cancer.

a, Representative images of haematoxylin and eosin (H&E) staining (first column) and immunofluorescence staining for myeloperoxidase (red), citrullinated histone H3 (green) and DAPI (blue) (subsequent columns) in human primary breast cancer (n = 461) and metastases (Met) in liver (n = 20), lung (n = 23), brain (n = 7) or bone (n = 33). b, NET quantification was performed by immunofluorescence staining using Imaris 9.0 Microscopy Image Analysis Software. The first column indicates MPO, H3cit and DAPI staining and the second column indicates H3cit staining in the same tissue section. Columns A and B show the results of analysis using the Imaris 9.0 Software. Column A indicates the H3cit-positive signal area, and column B shows the percentage of H3cit areas in the whole section. c, Correlation between serum MPO–DNA and plasma MPO–DNA levels in breast cancer samples (n = 72, the Pearson’s correlation coefficient R value and the P value are shown). d, Plasma and serum levels of MPO–DNA in patients with breast cancer with (n = 14) or without (n = 58) distant organ metastases. Data are mean ± s.e.m., **P = 0.0052 (plasma) and 0.0035 (serum), calculated using two-tailed Student’s t-test. e, Kaplan–Meier survival curves for patients with breast cancer with low (n = 135) and high (n = 136) serum MPO–DNA levels. The significance was assessed using a two-sided long-rank test. f, Receiver operator characteristic (ROC) curves to predict liver, lung, bone or brain metastases from serum MPO–DNA levels. n = 271, AUC, area under curve. Source data

Extended Data Fig. 2 NETs promote liver metastases.

a, b, Mouse 4T1 (a) and human MDA-MB-231 (b) breast cancer cells were injected into mammary fat pads of BALB/c mice (a) and NOD/SCID mice (b), respectively. At various time points (0, 15 and 40 days) after tumour inoculation, the mice were killed and examined for NET infiltration and tumour metastases in the liver and lungs. Representative images of H&E staining and immunofluorescence staining for H3cit and MPO to denote NET infiltration in the liver (left) and the lung (right) are shown, white arrows indicate NETs. n = 5 per group. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test; ****P < 0.0001; for a, ns = 0.5985, *P = 0.0422, **P = 0.0072; for b, ***P = 0.0003, ns = 0.7300, **P = 0.0016 (day 0) and 0.0063 (day 15) compared with day 40. c, MDA-MB-231 breast cancer cells were injected into mammary fat pads of NOD/SCID mice, and the tumour tissues, liver tissues and plasma were collected at different time points after tumour inoculation. The dynamics of NET expression in the primary tumours and the liver, the plasma MPO–DNA levels and the expression of liver HPRT1 mRNA relative to mouse Gapdh expression were shown (n = 3 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ****P < 0.0001; for liver NETs group: *P = 0.0478, **P = 0.0064; for plasma MPO–DNA group: **P = 0.0014; for liver Met group: *P = 0.0454, compared with d7 group. d, Representative images of immunofluorescence co-staining for H3cit with CK and epithelial cell adhesion molecule (EpCAM) for tumour cells, platelet-derived growth factor receptor beta (PDGFRβ) for pericytes, α-smooth muscle actin (α-SMA) for stromal cells and CD31 for endothelial cells in the metastatic liver tissues of NOD/SCID mice intrasplenically injected with MDA-MB-231 cells. White arrows indicate the areas that are shown in the higher-magnification images in the top-right corners. Scale bars, 50 μm. n = 3 biologically independent experiments. e, Representative images and quantification of NETs in the LPS-induced neutrophils isolated from PAD4+/+ and PAD4−/− mice. Scale bars, 10 μm. n = 6 biologically independent animals. Data are mean ± s.d. ****P < 0.0001, assessed using a two-tailed Student’s t-test. f, g, Representative images (left) and quantification (right) of liver NET formation (f) and liver metastases (g) of luciferase-E0771 tumour cells injected into the spleens (1 × 106 cells per mouse of wild-type and DNase I-treated female C57BL/6 mice (n = 6 mice per group). Data are mean ± s.d. *P = 0.0122, ****P < 0.0001 assessed using a two-tailed Student’s t-test. Source data

Extended Data Fig. 3 8-OHdG-enriched NETs are predominantly detected in liver metastases of colon cancer.

a, Representative images of H&E (first column) and immunofluorescence (subsequent columns) staining for MPO (red), H3cit (green) and DAPI (blue) in human primary colon cancer and metastases (Met) in the liver, lung, bone or brain. b, NETs infiltrated in primary colon cancer tissues (n = 130) and in liver (n = 16), lung (n = 12), bone (n = 3) and brain (n = 5) metastases. Data are mean ± s.e.m., two-sided one-way ANOVA with Tukey test, ****P < 0.0001, *P = 0.0359, ns > 0.9999 (bone met.) and = 0.9710 (brain met.) compared with primary tumour. Met, metastases. c, d, Representative images of confocal microscopy (c) and quantification (d) of NETosis, denoted by H3cit and MPO immunofluorescence staining in the liver tissues at various time points (0, 10 and 20 days) following intrasplenic injection of HCT116 colon cancer cells. n = 5 mice per time point. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, ****P < 0.0001. e, Representative scanning electron microscopy images of normal or PMA-stimulated neutrophils (NETs) and cell-free NETs isolated from PMA-stimulated neutrophils (cell-free NETs). f, Representative images of 8-OHdG staining in the NETs (top) produced by PMA-stimulated neutrophils or normal neutrophils (bottom). g, 8-OHdG levels in the genomic and NET-DNA of human neutrophils, determined by 8-OHdG ELISA assays (n = 6 biologically independent samples). Data are mean ± s.d. ***P = 0.0001 calculated using a two-tailed Student’s t-test. h, Silver staining of His–TREX1 expressed and purified from Escherichia coli. i, j, Agarose gel analysis of the genomic DNA and NET-DNA from neutrophils incubated with increasing concentrations of recombinant TREX1 protein (i) or with 100 ng ml−1 of recombinant TREX1 protein for increasing time periods (j). k, Dynamics of the levels of genomic DNA and NET-DNA treated with recombinant TREX1 at 100 ng ml−1, quantified from the agarose gels in j. l, Adhesion and migration assays for MDA-MB-231 cells stimulated with 5 μg ml−1 neutrophil DNA or 5 μg ml−1 NETs in the presence or the absence of DNase I. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, ****P < 0.0001; for migration assays: *P = 0.0390, **P = 0.0011 and ns = 0.6578; for adhesion assay: *P = 0.0469 and ns = 0.2841. m, Migration assays for MDA-MB-231 cells in Boyden chambers. NET-DNA at increasing concentrations (0–5 μg ml−1) or pretreated with DNase I was added to the culture media in the lower chambers. n = 5 biologically independent experiments, Data are mean ± s.d. two-sided one-way ANOVA with Tukey test, ****P < 0.0001, ***P = 0.0003, ns = 0.6978 (1 μg ml−1) and 0.9372 (5 μg ml−1 + DNase I) compared with the untreated cells. Scale bars, 100 μm. n, MDA-MB-231 cells were randomly attached to the seeding chamber in PBS. The media in the left chamber was replaced with media containing 5 μg ml−1 NET-DNA. Tracks of individual cells are shown as coloured lines (left). The spider plot (right) demonstrates tracks of the bulk cells. Data in e, f, hk, n were representative of three biologically independent experiments. Source data

Extended Data Fig. 4 CCDC25 binds to NET-DNA.

a, Schematic of NET-DNA pull-down assays. b, Mass spectrometry analysis identified CCDC25 as the cytoplasmic membrane protein from MDA-MB-231 cells pulled down by the biotinylated NET-DNA. c, Immunoblotting of the membrane proteins of MDA-MB-231 cells pulled down by biotinylated NET-DNA and detected by an anti-CCDC25 antibody. d, EMSA demonstrated NET-DNA binding to CCDC25 super-shifted by an anti-CCDC25 antibody. Membrane proteins of MDA-MB-231 cells and the biotinylated NET-DNA were incubated with or without the antibody against CCDC25, IgG (negative control), or 20-fold excess of unbiotinylated NET-DNA. e, The binding of NET-DNA to the membrane proteins of MDA-MB-231 cells transduced with a control sgRNA or two CCDC25-sgRNAs was evaluated by EMSA. f, EMSA reveals the interaction of biotinylated NET-DNA with increasing concentrations of CCDC25. The protein–DNA complex is denoted by a red asterisk. g, Binding kinetics of CCDC25 and NET-DNA generated from the above EMSA assays in f. n = 3. Data are mean ± s.d. h, Purified NETs were coupled to magnetic beads, treated with Proteinase K (left) and DNase I (right), and incubated with His–CCDC25. The interaction of NETs and CCDC25 was evaluated by the precipitation of NETs–beads and blotted with anti-His antibody. His–CCDC25 mixed with beads without DNA served as a negative control (empty beads). The digestion efficiency of the protein and DNA components of NETs by Proteinase K and DNase I was confirmed by immunoblotting for H3cit and agarose gel analysis for DNA. i, j, Three different biotinylated heterologous 90-bp DNA duplexes with random sequences were either irradiated with UV-C light or were not irradiated. i, The relative 8-OHdG content in the DNA was determined by ELISA. n = 3. Data are mean ± s.d., **P = 0.0059 as determined by a two-tailed Student’s t-test. j, DNA pull-down assay for His–CCDC25. The resultant CCDC25 was detected by anti-His western blot analysis. His–CCDC25 mixed with beads without DNA served as a negative control (empty beads). k, Representative bio-layer interferometry showing CCDC25 binding to 8-OHdG-enriched DNA (left) and non-8-OHdG-enriched DNA (Right). The coloured lines show the data for five different concentrations of CCDC25 as indicated. Data in bf, hk are representative of three biologically independent experiments. Source data

Extended Data Fig. 5 NETs promote tumour metastasis via CCDC25.

a, Western blots for CCDC25 expression in MDA-MB-231 cells that were untreated (UT), transfected with a control sgRNA (ctrl-sg) or with one of two CCDC25 sgRNAs (sgCCDC251 and sgCCDC252). b, MDA-MB-231 cells untreated (UT) or transfected with a control sgRNA or one of two CCDC25 sgRNAs were added into culture plates coated with fibronectin or NETs, and cell adhesion was evaluated. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, ****P < 0.0001. c, MDA-MB-231 cells transduced with either a control sgRNA or one of two CCDC25-sgRNAs were treated with 5 μg ml−1 NET-DNA or left untreated (UT), and stained with phalloidin (F-actin, green) and DAPI (nuclei, blue). Scale bars, 20 μm. FLPs, filopodium-like protrusions. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, **P = 0.0011, ns >0.9999. d, MDA-MB-231 cells were transduced with a control sgRNA or with CCDC25-sgRNA, and monitored for migration in a chemotaxis chamber with NET-DNA on the left. Red and blue lines demonstrate the migration tracks of control and CCDC25-knockout tumour cells, respectively. e, MDA-MB-231 cells transduced with a control sgRNA or with one of two sgRNAs for CCDC25 were treated with 5 μg ml−1 NET-DNA, 50 ng ml−1 IGF, 50 ng ml−1 MIF or 50 ng ml−1 EGF or were left untreated. Cell proliferation was assessed by the Cell Counting Kit-8 (CCK-8) assay. n = 3 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ****P < 0.0001. f, MDA-MB-231 cells transduced with a control sgRNA or with one of two sgRNAs for CCDC25 were treated with 5 μg ml−1 NET-DNA or 5 μg ml−1 apoptotic DNA or were left untreated, and cell migration and adhesion were evaluated. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ****P < 0.0001. g, 8-OHdG levels in NET-DNA and apoptotic DNA, determined by 8-OHdG ELISA assays, n = 6 biologically independent experiments. Data are mean ± s.d. ****P < 0.0001 as calculated by a two-tailed Student’s t-test. h, Luciferase-MDA-MB-231 cells transduced with a control or CCDC25-targeting sgRNA were intravenously injected into NOD/SCID mice that were pretreated with LPS or untreated; representative images and quantification of lung metastases in mice with the indicated treatments are shown (n = 6 mice per group). Two-sided one-way ANOVA with Tukey test, *P = 0.0132, ns = 0.9958. i, MCF-7 cells that were untreated (MCF-7 WT), or transduced with negative control (MCF-7 NC) or with CCDC25-overexpression vectors (MCF-7 OE) were added into the culture plates coated with fibronectin or NETs, and cell adhesion was evaluated. n = 6 biologically independent experiment. Data are mean ± s.d. Two-sided one-way ANOVA with Tukey test. ns = 0.4764 (fibronectin group) and 0.9744 (NETs group), ****P < 0.0001. j, Migration tracks of the MCF-7 cells transduced with negative control (CCDC25 NC) or with CCDC25-overexpression vectors (CCDC25 OE) in a chemotaxis chamber containing culture media with 5 μg ml−1 NETs. Red and blue lines demonstrate the tracks of control and CCDC25-overexpressed tumour cells, respectively. k, Luciferase-MCF-7 cells with (OE MCF-7) or without (NC MCF-7) CCDC25 overexpression were intravenously injected into NOD/SCID mice that were pretreated with LPS or were untreated. Representative images and quantification of lung metastases in the mice with indicated treatments are shown (n = 6 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ns = 0.9989, *P = 0.0170. l, Representative images and quantification of liver metastases in NOD/SCID mice that were intrasplenically injected with luciferase MCF-7 cells with (MCF-7 OE) or without (MCF-7 NC) CCDC25 overexpression, n = 6 mice per group). Data are mean ± s.d., **P = 0.0052 as calculated using a two-tailed Student’s t-test. m, n, HCT116 cells transduced with a control shRNA or one of two CCDC25-shRNAs were treated with 5 μg ml−1 NET-DNA or were left untreated, and were stained with phalloidin (F-actin, red) and DAPI (nuclei, blue). Scale bars, 10 μm. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. **P = 0.0026, ***P = 0.0008, ns = 0.9817. o, HCT116 cells were transduced with a control shRNA (shluc) or with CCDC25-shRNA, and monitored for migration in a chemotaxis chamber with NET-DNA on the left. Red and blue lines show the migration tracks of control and CCDC25-knockdown tumour cells, respectively. p, Luciferase-HCT116 cells transduced with a control shRNA or with CCDC25-targeting shRNA were intravenously injected into NOD/SCID mice pretreated with LPS or untreated. Representative images and quantification for lung metastases in mice with the indicated treatments are shown (n = 6 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. *P = 0.0435, ns = 0.9922. q, NOD/SCID mice were intrasplenically injected with luciferase-HCT116 cells, which were transduced with a control or with one of two CCDC25-targeting shRNAs. Representative images and quantification of liver metastases with indicated treatments were shown (n = 6 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. *P = 0.0117 (sh1 versus shluc) and 0.0142 (sh2 versus shluc), ns = 0.9948. Data in a, d, j, o are representative of three independent experiments. Source data

Extended Data Fig. 6 NETs promote tumour metastasis via CCDC25.

a, Primary breast cancer cells transduced with a control sgRNA or with one of two sgRNAs for CCDC25 were treated with 5 μg ml−1 NET-DNA or were untreated, and cell migration, adhesion and cytoskeleton remodelling were evaluated. n = 5 biologically independent experiments. Data are mean ± s.d. two-sided one-way ANOVA with Tukey test. b, NOD/SCID mice were intrasplenically injected with luciferase-primary breast cancer cells transduced with a control sgRNA or with one of two sgRNAs for CCDC25. Representative images (left) and quantification (right) of liver metastases after the indicated treatments are shown (n = 5 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, **P = 0.0040 (sg1 versus ctrl-sg) and 0.0018 (sg2 versus ctrl-sg), ns = 0.8865. c, Representative images (left) and quantification (right) of lung metastases in wild-type (WT) PyMT mice and in PyMT;CCDC25−/− (PyMT-KO) mice pretreated with LPS or untreated (n = 5 mice per group). Data are mean ± s.d., two-sided one-way ANOVA with Tukey test, *P = 0.0244, ns >0.9999. d, Tumour burden in wild-type PyMT mice and PyMT;CCDC25−/− mice pretreated with LPS or untreated. n = 5 mice per group. Data are mean ± s.d., two-sided one-way ANOVA, ns = 0.9934. e, Representative images (left) and quantification (right) of liver metastases in C57BL/6 mice intrasplenically injected with tumour cells derived from the wild-type PyMT mice and PyMT-KO mice (n = 5 mice per group). Data are mean ± s.d., *P = 0.0215 determined by a two-tailed Student’s t-test. f, Validation of the polyclonal CCDC25 antibody. Indicated cell lysates from MDA-MB-231 cells transduced with a control sgRNA or with one of two sgRNAs for CCDC25 were subjected to western blot analysis probing with the polyclonal CCDC25 antibody. g. Representative bio-layer interferometry data of polyclonal antibody binding to recombinant protein CCDC25. The coloured lines show the data for five different concentrations of recombinant CCDC25 as indicated. h, Inhibitory effects of a polyclonal CCDC25 blocking antibody (5 μg ml−1) on NET-induced migration, adhesion and cytoskeleton arrangement of MDA-MB-231 cells. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. i, Representative images and quantification of liver metastases of NOD/SCID mice intrasplenically injected with luciferase-MDA-MB-231 cells, which were treated with IgG as a control or with a CCDC25 antibody. n = 6 mice per group. Data are mean ± s.d., **P = 0.0042 as calculated by a two-tailed Student’s t-test. Data in f, g are representative of three independent experiments. Source data

Extended Data Fig. 7 CCDC25 is associated with poor prognosis in multiple malignant tumours.

a, CCDC25 expression in multiple cancer types in the Human Protein Atlas database. b, c, Representative immunohistochemical staining images (left) of CCDC25 expression in primary breast cancer (n = 202) and liver metastasis (n = 17) (b) and in primary colon cancer (n = 134) and liver metastasis (n = 16) (c). Scale bars, 50 μm. Lines within the violin plots (right) mark the 25th, 50th and 75th percentiles. *P = 0.0300, **P = 0.0056 as calculated by a two-sided Mann–Whitney U-test. d, Representative immunofluorescence co-staining images of CCDC25 with CK for tumour cells, CD31 for endothelial cells or α-SMA for stromal cells in human primary breast cancer. n = 5. e, Representative immunofluorescence staining images of CCDC25 in human primary breast cancer. The areas marked by the white boxes are shown magnified in the insets in the top right. n = 6. f, Representative immunofluorescence staining images for CCDC25 and H3cit in the liver metastases of patients with breast cancer. Insets as in e. n = 5. g, Representative immunohistochemical images for low and high CCDC25 expression in human primary breast cancer. Scale bars, 200 μm. Blue and red arrows indicate cancer cells and non-malignant cells, respectively. n = 573 in the low-CCDC25 group and n = 268 in the high-CCDC25 group. h, Kaplan–Meier survival curves for patients with colon cancer with high (n = 39) and low (n = 95) CCDC25 expression in the primary tumours. Comparisons are performed using a two-sided log rank test. i, Kaplan–Meier curves showing the overall survival of patients with breast cancer with high and low CCDC25 expression in The Cancer Genome Atlas (TCGA) breast cancer online database (n = 1,100); overall survival curves (n = 1,107) and recurrence-free survival curves (n = 909) of patients with lung cancer with high and low CCDC25 expression in the TCGA lung cancer online database; and overall survival curves of patients with myeloma with high and low CCDC25 expression in the Mulligan Myeloma online database (n = 264). The optimal survival cut points were determined by X-Tile statistical software. Comparisons were performed using a two-sided log-rank test. Source data

Extended Data Fig. 8 CCDC25 is a transmembrane protein and its N terminus interacts with NET-DNA.

a, Transmembrane helix prediction for CCDC25. One transmembrane helix was predicted by ProtScale (https://web.expasy.org/protscale/). One confidently predicted helix (score above 0) spans residues from around 60 to 80. The N terminus (about 40 residues) and the C terminus (about 40 residues) of CCDC25 are predicted to reside at the external or cytosolic sides of the cytoplasmic membrane owing to their hydrophilicity. b, Membrane pellets of MDA-MB-231 lysates were resuspended in the lysis buffer or buffer containing a high salt concentration (3 M NaCl), 5 M urea, 0.2 M Na2CO3 (alkaline) or 0.1% Triton-X100 after centrifugation at 20,000g. The resulting lysates with the indicated treatments were separated into membrane pellets (P) and supernatants (S) by centrifugation. Immunoblotting was performed using anti-CCDC25 antibody, anti-CXCR4 antibody (positive control) or anti-ATP5b antibody (negative control). c, Schematics of the different CCDC25 variants. d, EMSA showing the binding of wild-type or mutant CCDC25 with NET-DNA. eg, MDA-MB-231 cells were transduced with control sgRNA (Ctrl-sg) or sgRNA for CCDC25 alone (sgCCDC25) or along with ectopic expression of full-length wild-type CCDC25 (sgCCDC25+WT) or the CCDC25 AA21–25 mutant (sgCCDC25+Mutant(21–25)). e, The expression of indicated proteins was determined by western blot. f, Filopodium-like protrusions of the cells with the indicated treatments were stained with phalloidin (F-actin, red) and DAPI (nuclei, blue). Scale bars, 10 μm. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ****P < 0.0001, ***P = 0.0007 and ns >0.9999. g, Migration tracks for the tumour cells with the indicated treatments in a chemotaxis chamber containing culture media with 5 μg ml−1 NET-DNA. Ctrl-sg, grey; sgCCDC25, blue; sgCCDC25+WT, green; sgCCDC25+Mutant(21-25), orange. h, CCDC25-knockout MDA-MB-231 cells with or without the ectopic expression of full-length wild-type CCDC25 or the CCDC25 AA21–25 mutant (M1) were injected into the spleen of NOD/SCID mice. The quantification of liver metastases is shown. n = 5 mice per group. Data are mean ± s.d., one-way ANOVA with Tukey test. **P = 0.0023 (sgCCDC25 versus ctrl-sg) and 0.0013 (sgCCDC25 + M1 versus ctrl-sg), ns = 0.9758. Data in b, d, e, g are representative of three biologically independent experiments. Source data

Extended Data Fig. 9 CCDC25 interacts with ILK.

a, Cytosolic extracts from the HeLa cells transfected with EGFP-His-tagged-CCDC25 with (+) or without (−) NETs treatment were immunoprecipitated using anti-His antibody. Bound proteins were eluted and visualized by silver staining. A precipitated protein band of 55 kDa was submitted for mass spectrometry. b, The full amino-acid sequence of human ILK. The sequences in yellow are the tryptic peptides identified by liquid chromatography–mass spectrometry. c, Mass spectrometry analysis of the two peptides highlighted in b. d, Confocal microscopy showing the colocalization of CCDC25 with ILK in the HeLa cells transfected with His-tagged full-length CCDC25. Quantification was performed using Leica Confocal Software (fourth row). Scale bars, 10 μm. e, Immunoblotting of CCDC25 (top) or ILK (bottom) in the lysates (input) or immunoprecipitates (IgG or anti-ILK) of MDA-MB-231 cells stimulated with (+) or without (−) NETs treatment. Data in a, d, e were representative of three biologically independent experiments.

Extended Data Fig. 10 CCDC25 interacts with ILK at its C terminus and signals through the ILK–β-parvin cascade.

a, Phosphorylation of ILK substrates (AKT and GSK3β) was analysed. Whole-cell lysates of MDA-MB-231 cells transduced with two CCDC25-targeting or control sgRNAs with or without NETs stimulation at 5 μg ml−1 were subjected to immunoblotting with the indicated antibodies. b, c, GTP bound or total RAC1 and CDC42 were examined in the lysates of MDA-MB-231 cells transduced with ILK sgRNAs (b) or with PARVB shRNAs (c) and stimulated with or without NETs. d, MDA-MB-231 cells transduced with a control shRNA or with one of two PARVB-shRNAs were treated with or without 5 μg ml−1 NETs. The representative images of filopodium-like protrusions of the cells (stained with phalloidin (F-actin, green)) are shown on the left, and the quantification is shown on the right. Scale bars, 20 μm. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. ***P = 0.0001(UT versus UT + NETs) and 0.0007 (UT versus shluc + NETs), ns = 0.9402 (UT versus sh1 + NETs) and 0.7870 (UT versus sh2 + NETs). e, Migration tracks of the MDA-MB-231 cells transduced with a control sgRNA or ILK sgRNA in a chemotaxis chamber containing culture media with 5 μg ml−1 NET-DNA. Red and blue lines denote the tracks of control and ILK-knockout tumour cells, respectively. f, Migration tracks of the MDA-MB-231 cells transduced with a control shRNA or with PARVB-shRNA in a chemotaxis chamber containing culture media with 5 μg ml−1 NET-DNA. Red and blue lines denote the tracks of control and β-parvin-knockdown tumour cells. g, h, MDA-MB-231 cells transduced with a control sgRNA or two sgRNAs for ILK were treated with or without 5 μg ml−1 NET-DNA. g, Left, representative images of immunofluorescence staining for ki67 (green) and F-actin (red) in MDA-MB-231 cells in a 3D culture system. Scale bars, 20 μm. Right, quantification of the ki67-positive tumour cells. n = 5 biologically independent experiments. Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. **P = 0.0050 (UT versus sgILK1) and 0.0055 (UT versus sgILK2), *P = 0.0328, ns >0.9999 (sgILK1 versus sgILK1 + NETs) and 0.9990 (sgILK2 versus sgILK2 + NETs). h, Cell proliferation was assessed by the CCK-8 assay in a 2D culture system. n = 3 biologically independent experiments, Data are mean ± s.d., two-sided one-way ANOVA with Tukey test. *P = 0.0105, ****P < 0.0001, ns >0.9999 (ILK sgRNA1 versus ILK sgRNA1 + NETs) and = 0.9969 (ILK sgRNA2 versus ILK sgRNA2 + NETs). i, MDA-MB-231 cells that were untreated or transduced with a control or with one of two ILK sgRNAs were intrasplenically injected into NOD/SCID mice, and liver metastatic nodules were counted 30 days after injection. n = 6 mice per group. Data are mean ± s.d., ns >0.9999 and **P = 0.0021 (UT versus sgILK1) and 0.0016 (UT versus sgILK2), determined by a two-sided one-way ANOVA with Tukey test. j, Schematics highlighting the major findings of this study. Data in ac, e, f are representative of three independent experiments. Source data

Supplementary information

Supplementary Figures

This file contains uncropped blots for figures 2–4 and extended data figures 3-6 and 8-10.

Reporting Summary

Video 1: MDA-MB-231 cells efficiently migrated towards higher NET-DNA gradient in µ-slide chemotactic chamber.

MDA-MB-231 cells were randomly attached to the seeding chamber in PBS. The left well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 3n.

Video 2: Knocking out CCDC25 reduced the chemotactic response of cancer cells towards NET-DNA.

MDA-MB-231 cells transduced with CCDC25-sgRNA (stained with CFSE, Green) or a control sgRNA were randomly attached to the seeding chamber in PBS. The left well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 5d.

Video 3: Enforced expression of CCDC25 in MCF-7 cells promoted directional migration towards NET-DNA.

MCF-7 cells transduced with empty vectors (stained with CFSE, Green) or CCDC25-overexpression vectors were randomly attached to the seeding chamber in PBS. The down well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 5j.

Video 4: Knocking down CCDC25 reduced the chemotactic response of cancer cells towards NET-DNA.

HCT116 cells transduced with CCDC25-shRNA (stained with CFSE, Green) or a control shRNA were randomly attached to the seeding chamber in PBS. The left well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 5o.

Video 5: Ectopic expression of wild-type CCDC25, but not its mutant at AA21-25, successfully rescued the NET-DNA-induced directional chemotaxis abolished by the knockout of CCDC25.

MDA-MB-231 cells were transduced with control sgRNA (Ctrl-sg) or sgRNA for CCDC25 alone (sgCCDC25) or along with ectopic expression of full-length CCDC25 (sgCCDC25+ WT) or mutant (21-25) CCDC25 (sgCCDC25+Mutant (21-25)) and randomly attached to the seeding chamber with 5 µg/ml NET-DNA on the left. Migration tracks of the ctrl-sg, sgCCDC25(stained with Hoechst 33342, Blue), sgCCDC25+WT (stained with CFSE, Green), sgCCDC25+Mutant (21-25) (stained with celltracker Orange CMTMR Dye, Orange) cancer cells were shown in Extended Data Fig. 8g.

Video 6: Knocking out ILK reduced the chemotactic response of cancer cells towards NET-DNA.

MDA-MB-231 cells transduced with ILK-sgRNA (stained with CFSE, Green) or a control sgRNA were randomly attached to the seeding chamber in PBS. The left well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 10e.

Video 7: Knocking down β-Parvin reduced the chemotactic response of cancer cells towards NET-DNA.

MDA-MB-231 cells transduced with β-Parvin-shRNA (stained with CFSE, Green) or a control shRNA were randomly attached to the seeding chamber in PBS. The right well was filled with media containing 5 µg/ml NET-DNA. Migration tracks of the cells were shown in Extended Data Fig. 10f.

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Yang, L., Liu, Q., Zhang, X. et al. DNA of neutrophil extracellular traps promotes cancer metastasis via CCDC25. Nature 583, 133–138 (2020). https://doi.org/10.1038/s41586-020-2394-6

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.