Colorectal cancer-derived extracellular vesicles induce liver premetastatic immunosuppressive niche formation to promote tumor early liver metastasis

Abstract

To date, the effect of colorectal cancer (CRC)-derived extracellular vesicles (EVs) on liver pre-metastatic niche (PMN) remain incompletely understood. 1 To investigate the role of CRC-derived EVs in the remodeling of the liver PMN, we isolated EVs from CT26 cell culture supernatant. The characteristics of EVs (the morphology, size, and markers) were identified by transmission electron microscopy (TEM), Nanoparticle Tracking Analysis (NTA), and western blotting (Supplementary Fig. S1a–e ). Then, BALB/c mice with an intact liver immune status were pretreated with CRC-derived EVs and phosphate-buffered saline (PBS) for one month, and a liver metastasis model was established via spleen injection of tumor cells to determine whether EVs influence liver metastasis (Fig. 1a ). Two hours after the operation, in vivo imaging indicated that the animal model was successfully established, and the tumor fluorescence intensity in the liver was consistent between the two groups (Supplementary Fig. S2a, b ). Amazingly, the liver tumor fluorescence intensity in the EVs group was significantly stronger than that in the PBS group after 24 h (Fig. 1b ; Supplementary Fig. S2c, d ). The number of liver tumor nodules in the EVs group was also significantly higher than that in the PBS group on day 4 (Fig. 1c ; Supplementary Fig. S2e, f ). The efficacy of EVs in promoting tumor liver metastasis was consistent regardless of whether the spleen was being preserved when the animal model was established (Supplementary Fig. S3a–e ). Fig. 1 CRC-derived EVs initiate a sequential event to remodel the liver premetastatic niche and promote tumor liver metastasis. a Schematic study overview. b Left: Representative bioluminescence imaging showed that, compared with the PBS group ( n = 10), the liver tumor fluorescence intensity was significantly stronger in the EVs group ( n = 8) at 24 h after operation. Right: Quantifying tumor fluorescence intensity further confirmed that the tumor burden was heavier in the EVs group. The detail of the animal experiment was also shown in Fig. S2. c Left: Representative liver images showed more metastatic nodules on the liver surface in the EVs group. Right: Quantifying liver tumor nodules confirmed that the mean number of liver tumor nodules was significantly higher in the EVs group (13.33 vs. 52.43, p = 0.009). d Left: Representative image demonstrated that the CD11b + Gr1 + cells population screened by flow cytometry in the liver PMN after CRC-derived EVs education increased ( n = 3 in the PBS group, n = 3 in the EVs group). Right: Quantitative analysis confirmed that the mean proportion of the CD11b + Gr1 + cells subset in the CD45 + lymphocyte population increased significantly in the EVs group (7.09% vs. 14.17%, p = 0.012). In rescue assay, after CRC-derived exsomal TGF-β1 knockout and HSCs-derived CXCL12 depletion, the mean proportion of CD11b + Gr1 + cells subset was 8.86% in the EVs with CXCL12 depletion group and 7.92% in the EVs TGF-β1 KO group, respectively. Compared with the PBS group, there was no significant difference among the three group in Fig S4 h , i . e Left: Representative image demonstrated that, in the liver PMN, the expression of NKG2D in NK cells was downregulated after CRC-derived EVs education. Right: Quantitative analysis confirmed that compared with the PBS group ( n = 3), NKG2D expression level in NK cells was significantly downregulated in the EVs group ( n = 3) (8.5% vs. 1.15%, p < 0.05). f Left: Representative liver images showed the metastatic nodules on the liver surface in the control group ( n = 5), PBS group ( n = 5), and EVs group ( n = 6). Right: The mean number of liver tumor nodules was 16 in the control group, 5.6 in the PBS group, and 8.2 in the EVs group. Statistical analysis showed that, for tumor burden, there were no significant differences between the PBS group and the EVs group after the CD11b + Gr1 + cells subset depletion (5.6 vs. 8.2, p = 0.608). The tumor nodule number in the PBS group and the EVs group tended to decrease compared with that in the control group. g Left: Representative liver images showed the metastatic nodules on the liver surface in the control group ( n = 5), PBS group ( n = 6), and EVs group ( n = 6). Right: The mean number of liver tumor nodules was 7 in the control group, 44.67 in the PBS group, and 50 in the EVs group. Statistical analysis showed that, for tumor burden, there were no significant differences between the PBS group and the EVs group after CD45 + GM1 + cells subset depletion (44.67 vs. 50, p = 0.805). Tumor nodule number was significantly higher in the PBS group and the EVs group than in the control group. These results demonstrated that the regulated NK cells in the liver PMN remodeled by EVs have an important role in promoting tumor liver metastasis. h Representative imaging showed that DID dye–labeled CRC-derived EVs mainly accumulated in the liver at 2 and 24 h after injection. i Immunofluorescence confocal analysis demonstrated that DID + EVs were mainly absorbed by Desmin + HSCs. j Left: Immunofluorescence confocal analysis demonstrated that HSCs were massively activated and significantly increased their expression of α-SMA and fibronectin after CRC-derived EVs education for 4 weeks. Right: Statistical analysis further confirmed that α-SMA and fibronectin intensity was significantly increased after CRC-derived EVs education for 4 weeks. k Representative imaging showed that quiescent HSCs absorbed EVs in vitro. l Left: A co-culture assay showed that tumor migration was enhanced after HSCs were co-cultured with exogenous Tgf-β1 and CRC-derived EVs and transformed into CAFs. Right: Statistical analysis further confirmed that after co-culture with exogenous Tgf-β1 and CRC-derived EVs, CAFs enhanced tumor migration, and the effect was abolished by Tgf-β1 inhibitor. m The volcano map showed that chemokines such as CXCL 12 were highly expressed. n ELISA showed that CXCL12 was highly expressed in serum after mice were educated with CRC-derived EVs for 4 weeks. o ELISA confirmed that EVs and exogenous Tgf-β1 enhanced the expression of CXCL12 in the cell supernatant of CAFs, and the expression of CXCL12 was inhibited by Tgf-β1 inhibitor. p Left: After the mice were educated with Tgf-β1–knockout EVs for 4 weeks, a CRC liver metastasis model was established to confirm the effect of Tgf-β1 knockout. Representative liver images showed liver tumor fluorescence intensity in the PBS group ( n = 6) and the EVs Tgf-β1 group ( n = 6). Right, There was no significant difference in tumor burden between the PBS group ( n = 6) and EVs Tgf-β1 group ( n = 6) at 24 h after operation. The detail of the animal experiment was also shown in Fig. S13. q The expression level of exosomal TGFB1 in plasma was significantly higher in CRC patients with synchronous liver metastasis ( n = 18) than in those without metastasis ( n = 20). r Survival analysis suggested that patients with higher TGFB1 expression levels had worse overall survival. s The working model in this study ( https://biorender.com/ ). c , f , g scale bar: 1 cm; I , j , l , scale bar: 100 μm; k scale bar: 50 μm; GI gastrointestinal tract, Inh inhibitor, CM culture medium, HSCs hepatic stellate cells. Data are shown as means ± SD. ‘ns’ means no significance, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 Full size image

Topics & Concepts