Targeting SPP1-orchestrated neutrophil extracellular traps-dominant pre-metastatic niche reduced HCC lung metastasis

Abstract

Background: The mechanisms by which tumor-derived factors remodel the microenvironment of target organs to facilitate cancer metastasis, especially organ-specific metastasis, remains obscure. Our previous studies have demonstrated that SPP1 plays a key role in promoting metastasis of hepatocellular carcinoma (HCC). However, the functional roles and mechanisms of tumor-derived SPP1 in shaping the pre-metastatic niche (PMN) and promoting lung-specific metastasis are unclear.

Methods: Orthotopic metastasis models, experimental metastasis models, CyTOF and flow cytometry were conducted to explore the function of SPP1 in shaping neutrophil-dominant PMN and promoting HCC lung metastasis. The main source of CXCL1 in lung tissues was investigated via fluorescence activated cell sorting and immunofluorescence staining. The expression of neutrophils and neutrophil extracellular traps (NETs) markers was detected in the lung metastatic lesions of HCC patients and mouse lung specimens. The therapeutic significance was explored via in vivo DNase I and CXCR2 inhibitor assays.

Results: SPP1 promoted HCC lung colonization and metastasis by modifying pulmonary PMN in various murine models, and plasma SPP1 levels were closely associated with lung metastasis in HCC patients. Mechanistically, SPP1 binded to CD44 on lung alveolar epithelial cells to produce CXCL1, thereby attracting and forming neutrophil-abundant PMN in the lung. The recruited neutrophils were activated by SPP1 and then formed NETs-dominant PMN to trap the disseminated tumor cells and promote metastatic colonization. Moreover, early intervention of SPP1-orchestrated PMN by co-targeting the CXCL1-CXCR2 axis and NETs formation could efficiently inhibit the lung metastasis of HCC.

Conclusions: Our study illustrates that HCC-lung host cell-neutrophil interactions play important roles in PMN formation and SPP1-induced HCC lung metastasis. Early intervention in SPP1-orchestrated PMN via CXCR2 inhibitor and DNase I is a potential therapeutic strategy to combat HCC lung metastasis.

Keywords: C-X-C motif chemokine ligand 1; Hepatocellular carcinoma; Lung epithelial cells; Pre-metastatic niche; SPP1.

Fig. 1

SPP1 is closely associated with the lung metastasis of HCC. A ELISA detection of plasma SPP1 concentration in samples of NLM (n = 18) and LM (n = 9). B qRT‒PCR analysis of SPP1 mRNA levels in tissues of NLM (n = 9) and LM (n = 6). The P values were calculated by Student’s t-test. C, D SPP1 protein levels in plasma (C) and pro-metastatic gene expression in the lungs (D) of mice at 7 days after SPP1 treatment (n = 6 per group). The P values were calculated by Student’s t-test. E Tissue-clearing technique showed early colonization of Hepa1-6^GFP in the lungs at 1 day after i.v. injected in PBS or SPP1 pre-treated mice (n = 3 per group). The P values were calculated by Student’s t-test. F The effect of SPP1 on HCC metastasis was evaluated in the SPP1-PMN model of Hepa1-6 (n = 7 per group) and H22 cell lines (n = 6 per group). The P values were calculated by two-tailed Fisher exact test (for Hepa1-6) and Student’s t-test (for H22), respectively. G–J Experimental schematics of modified orthotopic spontaneous metastasis model (G). Quantification analysis of SPP1 protein levels in plasma (H) and pro-metastatic gene expression in the lungs (I) of mice at 14 days after orthotopic liver transplantation (n = 5 per group). Shown were representative images of lung metastatic loci of Hepa1-6^vector/SPP1 in the SPP1-orchestrated spontaneous metastasis model (J, n = 5 per group). The P values were calculated by a student’s t-test (for H-I) and two-tailed Fisher exact test (for J), respectively. K I.v. injection of Hepa1-6^vector or Hepa1-6^SPP1 cells with matched CM-treatment for lung metastasis analysis. Representative images and the ratio of lung metastasis were shown (n = 7 per group). The P values were calculated by the two-tailed Fisher exact test. L The effect of anti-SPP1 neutralizing antibody on HCC metastasis was evaluated in the experimental metastasis model (n = 7 per group). The P values were calculated by the two-tailed Fisher exact test. For all tests, significance is determined with a 95% confidence interval (ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001)

Fig. 2

SPP1 promotes HCC lung metastasis in a neutrophil-dependent manner. A TSNE plot projection of single cells from PBS-PMN and SPP1-PMN, colored by 23 clusters. The lungs were harvested after i.p. injecting SPP1 for 7 days (related to Fig. 1D) (n = 4 per group). B The quantification of the distinct immune cell subtypes in PBS- and SPP1-PMN group. C, D IHC staining of neutrophils (Ly6G⁺) in the lung tissues (C). To verify, the number of neutrophils is counted in five random fields and took up the average per sample (D, n = 5 per group). The P values were calculated by a student’s t-test. Scale bar, 100 μm. E, F FC analysis of neutrophils (CD45⁺CD11b⁺Ly6G⁺) in the BALF and MPO activity in the lungs from SPP1-PMN model (n = 5 per group). The P values were calculated by a student’s t-test. G, H FC analysis of immune cells in lung tissues from the orthotopic model of Hepa1-6^vector/SPP1 mice (n = 4 per group). Lungs were harvested at 14 days after orthotopic inoculation. Percentages of neutrophils (CD45⁺CD11b⁺Ly6G⁺), macrophages (CD45⁺CD11b⁺F4/80⁺), NK cells (CD45⁺ NK1.1⁺), CD4 + T cells (CD45⁺CD3⁺CD4⁺CD8⁻), CD8 + T cells (CD45⁺CD3⁺CD4⁻CD8⁺), B cells (CD45⁺B220⁺) were shown. The P values were calculated by a student’s t-test. I, J Representative IHC images and quantification of Ly6G staining and in the lungs of Hepa1-6 orthotopic implantation model and anti-SPP1 treatment model (n = 5 per group). Scale bar, 100 μm. The P values were calculated by a student’s t-test. K, L Representative H&E images and quantification of lung metastasis loci of Hepa1-6^SPP1 (K) or H22 (L) in SPP1-associated experimental metastasis model after depleting neutrophils by anti-Ly6G antibody. For all tests, significance is determined with a 95% confidence interval (ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001)

Fig. 3

SPP1 promotes neutrophil recruitment by stimulating lung epithelial cells to increase CXCL1 secretion. A CM from different groups was placed in the lower chambers. Primary neutrophils were isolated and then placed in the upper chambers and their migration into the lower chambers was counted. The P values were calculated by one-way ANOVA. B qRT-PCR showed the expression of neutrophil-related chemokines and cytokines in lung tissue in SPP1-PMN (upper panel, related to Fig. 1C) and in Hepa1-6^vector/SPP1-PMN (lower panel). The P values were calculated by Student’s t-test. C qRT‒PCR analysis of Cxcl1 expression in epithelial cells, neutrophils, macrophages, fibroblasts, T cells, and B cells purified from the lung tissues in the Hepa1-6^vector/SPP1-PMN model. D Representative images (right panel) and quantitative analysis of SPP1-his distribution in different tissues of mice with i.p. administration of SPP1-his. Scale bar, 50 μm. E, F IF staining of SPP1-his and lung epithelial cell (SFTPC⁺) colocalization (E) and SPP1-positive rates (F) in different cells of mice with i.p. administration of SPP1-his (related to Figure S5C, Supporting Information). Scale bar, 20 μm. G Representative IF image of SPP1-his internalization by MLE-12 cells. Scale bar, 10 μm. H qRT‒PCR showed the neutrophil-related chemokines and cytokines in MLE-12 cells treated with various concentrations of SPP1 (upper panel) or cocultured with Hepa1-6^vector/SPP1 (lower panel). The P values were calculated by Student’s t-test. I Migrated cell numbers of murine neutrophils recruited by CM of MLE-12 cells, SPP1-pretreated MLE-12 cells or MLE-12 cells treated with 1 µg/ml SPP1 plus CXCR2 inhibitor (SB225002, 40 nM) or by MLE-12 CM plus 1 µg/ml SPP1. The P values were calculated by one-way ANOVA. J IHC images and quantification of neutrophils in the Hepa1-6-PMN models following CXCR2 inhibitor treatment (n = 6 per group). The P values were calculated by Student’s t-test. Scale bar, 100 μm. For all tests, significance was determined with a 95% confidence interval (ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001)

Fig. 4

SPP1 increases CXCL1 expression in MLE-12 cells through the CD44/STAT3 axis. A Heatmap of the SPP1 receptor genes detected by RNA-seq in the lung tissue after i.p. administration of SPP1 for 7 days. B–E Immunoblotting of CXCL1 expression in MLE-12 shCD44 (B) cells, MLE-12 shITGAV (C) cells, MLE-12 shITGB1 (D) cells, MLE-12 shITGA5 (E) cells treated with SPP1 for 48 h. F–H Immunoblot (F), ELISA (G) and qRT-PCR (H) showed CXCL1 expression in MLE-12 shCD44 cells treated with SPP1 or MLE-12 cells treated with SPP1 and/or integrin inhibitor (MK-0429, 20 nM). The P values were calculated by a student’s t-test. I Representative IF images of SPP1-his and CD44 colocalization in MLE-12 shNC and MLE-12 shCD44 cells. Scale bar, 10 μm. J Relative migrated cell numbers of murine neutrophils recruited by CM of CD44-, ITGB1- or ITGAV-knockdown MLE-12 cells treated with SPP1. The P values were calculated by a student’s t-test. K Changes in p-STAT3, STAT3, and CXCL1 in MLE-12 cells under SPP1 stimulation. L qRT‒PCR (upper panel) or immune blot (lower panel) analysis of CXCL1 expression in MLE-12 cells treated with SPP1, Stattic (20 μM), and SPP1 plus Stattic. The P values were calculated by one-way ANOVA. M Immunoblotting of CXCL1, STAT3, p-STAT3 in lung tissues of mice at 7 days after PBS or SPP1 treatment. N–Q Overlapping of genes resulting from the comparison of upregulated genes in mouse lung tissue in SPP1-PMN mice model and in SPP1-pretreated MLE-12 cells (N). qRT-PCR (O) and ELISA (P) showed the SPP1 expression in SPP1-pretreated MLE-12 cells. The pattern diagram (Q) of MLE-12 self-amplified SPP1 expression under exogenous SPP1 stimulation. The P values were calculated by a student’s t-test. For all tests, significance was determined with a 95% confidence interval (ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001)

Fig. 5

SPP1 induces NETs formation in PMN. A KEGG enrichment bubble diagrams of upregulated genes in SPP1-stimulated neutrophils. B, C Representative IF images of NETs formation (B) from mice neutrophils with different administrations. NETs are stained with Sytoxgreen (green), and DNA is stained with Hoechst (blue). To quantify, NETs extension from five random views is shown (C). The P values were calculated by one-way ANOVA. Scale bar, 100 μm. D Representative IF images and quantified NETs percentage from human neutrophils treated with PBS or rhSPP1. The P values were calculated by Student’s t-test. Scale bar, 100 μm. E IF images of DIL-labeled Hepa1-6 cells trapped within NETs. F, G Quantified NETs extension in lungs of SPP1-PMN at day 7 (F, related to Fig. 1C) and in lungs of Hepa1-6 orthotopic model at day 14 (G, related to Fig. 1G). The P values were calculated by Student’s t-test. H Representative images of adhered DIL-labeled Hepa1-6 cells in SPP1-PMN with the treatment of GSK484 or DNase I, the lungs were harvested 1 day after i.v. injected Hepa1-6 cells. To quantify, tumor cells were counted in five random views for each mouse and took an average count (n = 8 per group). The P values were calculated by one-way ANOVA. Scare bar, 100 μm. I, J The effect of DNase I or GSK484 on experimental metastasis model of Hepa1-6^SPP1. Representative images of H3cit + and MPO + signals in lung tissues (I) and lung metastasis lesions (J). The P values were calculated by one-way ANOVA. Scale bar, 100 μm. K The effect of DNase I or GSK484 on SPP1-PMN model of H22. The lungs were harvested at 14 days after i.v. injected (n = 5 per group). The P values were calculated by one-way ANOVA. For all tests, significance was determined with a 95% confidence interval (ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001)

Fig. 6

CXCR2 inhibitor plus DNase I treatment inhibits HCC lung metastasis. A IHC staining of SPP1 in HCC lung metastasis lesions from Huashan cohort (left). Matched IF staining of CXCL1 + and SFTPC + cells in lung tissues (middle), and IF staining of MPO + and H3cit + cells in metastatic tumor lesions (right). Scare bar, 25 μm. B–E Quantification analysis of CXCL1 (B), percent of CXCL1 + SFTPC + cells (C), NETs formation (H3cit⁺, D), and Neutrophils (MPO⁺ E) in SPP1-high/-low groups (n = 6 per group). The P values were calculated by Student’s t-test. Scare bar, 25 μm. F Correlation analysis between serum MPO-DNA and SPP1 concentration in HCC samples (n = 21, Pearson correlation). G Experimental setup. Briefly, GFP-labeled Hepa1-6 (Hepa1-6^GFP) cells were i.v. injected into C57BL/6 mice bearing orthotopic Hepa1-6 (unlabeled) at their PMN stage. For the early and late treatment groups, DNase I and CXCR2 inhibitor-SB225002 were given at day 0 and day 14, respectively. At the endpoint, the lung metastatic burden of Hepa1-6^SPP1−GFP was detected by BLI (n = 6 per group). H Tumor metastasis burden was observed and quantified by fluorescence imaging. Representative lung images (left) and quantified results (right) were shown, respectively. Data were means ± SD. The P values were calculated by one-way ANOVA. I, J Quantification analysis (I) and HE images (J) of lung metastasis lesions. The P values were calculated by Fisher’s exact test, one-tailed. K The immunostaining of MPO + and H3cit + and the signal intensity were quantified and analyzed, and representative images of indicated staining of lungs were shown. The P values were calculated by one-way ANOVA. Scale bar, 50 μm. For all tests, significance was determined with a 95% confidence interval (ns, P > 0.05; **P < 0.01; ****P < 0.0001)

Fig. 7

Schematic diagram. In HCC, primary tumor-derived SPP1 stimulated lung epithelial cells to secrete more CXCL1 through SPP1/CD44/STAT3 axis. Elevated CXCL1 levels in the lung microenvironment promoted neutrophil infiltration and NETs formation to facilitate HCC lung metastasis