Exosomal circ-0100519 promotes breast cancer progression via inducing M2 macrophage polarisation by USP7/NRF2 axis

Abstract

Background: Breast cancer (BC) is one of the most prevalent malignant tumours that threatens women health worldwide. It has been reported that circular RNAs (circRNAs) play an important role in regulating tumour progression and tumour microenvironment (TME) remodelling.

Methods: Differentially expression characteristics and immune correlations of circRNAs in BC were verified using high-throughput sequencing and bioinformatic analysis. Exosomes were characterised by nanoparticle transmission electron microscopy and tracking analysis. The biological function of circ-0100519 in BC development was demonstrated both in vitro and in vivo. Western blotting, RNA pull-down, RNA immunoprecipitation, flow cytometry, and luciferase reporter were conducted to investigate the underlying mechanism.

Results: Circ-0100519 was significant abundant in BC tumour tissues and related to poor prognosis. It can be encapsulated into secreted exosomes, thereby promoting BC cell invasion and metastasis via inducing M2-like macrophages polarisation.Mechanistically, circ-0100519 acted as a scaffold to enhance the interaction between the deubiquitinating enzyme ubiquitin-specific protease 7 (USP7) and nuclear factor-like 2 (NRF2) in macrophages, inducing the USP7-mediated deubiquitination of NRF2. Additionally, HIF-1α could function as an upstream effector to enhance circ-0100519 transcription.

Conclusions: Our study revealed that exosomal circ-0100519 is a potential biomarker for BC diagnosis and prognosis, and the HIF-1α inhibitor PX-478 may provide a therapeutic target for BC.

Keywords: HIF‐1α; NRF2; USP7; breast cancer; circ‐0100519; exosomes; macrophages.

FIGURE 1

The identification of circ‐0100519 and its characteristics as a BC biomarker. (A) The heatmap for immune cell score displayed the distribution of immune score expression in several clinical characteristics (PAM50, stage, immune score, and age were shown), together with the abundance of circRNA. (B) Dot plot showing the connection between the top 16 circRNAs and several immune cells. The size of the dot reflected the significance of the correlation (blue represents negative; red represents positive). (C) The GASE pathway analysis of circ‐0100519. (D) The expression levels of circ‐0100519 in various PAM50 subtypes (n = 196 independent replicates). (E, F) The expression levels of circ‐0100519 in BC tumour tissues and matched adjacent tissues verified by qRT‐PCR (n = 60 independent replicates) and ISH (n = 10 independent replicates). Scale bar = 100 µm. (G, H) Overall survival and disease‐free survival of circ‐0100519 low and circ‐0100519 high groups in BC patients were analysed by Kaplan–Meier curves and log‐rank tests. (I) Risk factors associated with poor prognosis of 196 BC patients were assessed by multivariate cox regression analysis. (J) Circ‐0100519 in serum exosomes of normal people (n = 10 independent replicates) and BC patients (n = 20 independent replicates). (K) Correlation of circ‐0100519 expression between BC tumours and serum exosomes. (L) The expression levels of circ‐0100519 in MCF‐10A, BC cells and human monocyte THP‐1 cells were detected by qRT‐PCR. (M) The closed‐loop structure of circ‐0100519 was revealed using Sanger sequencing. A representative data set is displayed as mean ± SEM values. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 2

Exosomal circ‐0100519 promotes BC cell proliferation and metastasis in vitro. (A) CCK‐8 assays were used to assess the viabilities of BC cells after treated with shNC or shcirc‐0100519. (B) Schematic diagram showing preactivated THP‐1 cocultured with BT‐549 or T‐47D. (C, D) Colony formation and EdU assays (scale bar = 50 µm) were used to detect the proliferative potential of BC cells. (E) The viabilities of BC cells were showed by CCK‐8 assays. (F) Analysis of the effect of downregulated circ‐0100519 expression on BC cell apoptosis. (G) Transwell migration assays were used to identify migration ability of BC cells. Scale bar = 1 mm. (H) FCM was used to assess CD206 expression in THP‐1. (C–H: BT‐549 or T‐47D cells were treated with shNC or shcirc‐0100519 before coculturing with THP‐1.) (I) TEM and NTA were used to identify exosomes obtained from the BT‐549 and T‐47D cell supernatants. The peak diameters of BT‐549‐Exo and T‐47D‐Exo were 80 and 95 nm, respectively. Scale bar = 50 nm. (J) Exosomes’ biomarker analysis using Western blot. The circ‐0100519 expression levels in exosomes obtained from the BT‐549 and T‐47D cell supernatants. The circ‐0100519 expression levels in macrophages treated with indicated treatments. (K) Confocal microscopy was used to verify macrophages can engulf exosomes and circ‐0100519. Scale bar = 10 µm. A representative data set is displayed as mean ± SEM values of three independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 3

Exosomal circ‐0100519 promotes BC development via altering the polarisation of M2‐like macrophages in vivo. (A) The mouse tumourigenicity model's workflow. (B) The Kaplan–Meier curve was provided. (C) Photographs of tumours collected from 4T1‐bearing Balb/c mice (n = 6 per group). (D) The curve graph exhibited the tumour volume measured at different time‐points (n = 6 per group). (E) Ki67 IHC staining was applied to tumour sections (n = 6 for each group). Scale bar = 200 µm. (F, G) FCM was utilised to evaluate CD206 expression in Balb/c mouse tumours. (H) Immunofluorescent images of mouse tumour sections using F4/80 and CD206 antibodies. Scale bar = 50 µm. (I) Bioluminescence images of lung metastasis model injected with shNC or shcirc‐0100519 exosomes via tail vein injection. (J) Lung metastatic nodules were showed by H&E staining. Scale bar = 200 µm. A representative data set is displayed as mean ± SEM values of three to six independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 4

circ‐0100519 suppresses NRF2 ubiquitination via interacting with USP7 and NRF2. (A) RNA pull‐down employed the biotinylated circ‐0100519 probe to identify the interaction of circ‐0100519 with USP7 or NRF2 in macrophage. RIP assay further confirmed the interaction between circ‐0100519 and USP7/NRF2 with USP7 antibody/NRF2 antibody. (B) Colocalisation of circ‐0100519 (RED) with USP7 proteins (GREEN) and NRF2 proteins (PURPLE). Scale bar = 5 µm. (C) Co‐IP assays verified the binding association of circ‐0100519 with USP7 and NRF2 in THP‐1 cells. (D) Exogenous protein interactions were identified in THP‐1 cells. (E) Expression of circ‐0100519 and NRF2 RNA levels in THP‐1 transfected with USP7 or Vector. (F, G) Expression of NRF2 and USP7 RNA levels in THP‐1 transfected with circ‐0100519 or pcDNA. (H) The expression of NRF2 and Keap1 in protein level after overexpression or knockdown of circ‐0100519 in THP‐1. (I) The NRF2 protein level in THP‐1 with circ‐0100519 overexpression (using proteasome inhibitor MG132). (J) The USP7 and NRF2 protein level in specified time point following treatment with cycloheximide (CHX, 10 µg/mL) in transfected THP‐1 cells. (K) IP assays showed the ubiquitination modification level of NRF2 in THP‐1 cells with indicated treatments. (L) NRF2 protein and ubiquitination levels in THP‐1 transfected with shcirc‐0100519 or shNC were analysed using Western blot, either in the presence or absence of the proteasome inhibitor MG132. (M) Ubiquitination modification levels of NRF2 in THP‐1 cells after circ‐0100519 overexpression or knockdown. A representative data set is displayed as mean ± SEM values of three independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 5

circ‐0100519 suppresses NRF2 ubiquitination via interacting with USP7 and NRF2. (A) 3D structure was employed to predict the precise binding positions for the interaction of circ‐0100519, USP7 and NRF2. (B) RNA pull‐down employing sequentially deleted circ‐0100519 fragments showed the binding region of circ‐0100519 with USP7. (C) The structure of circ‐0100519 fragment (nt 1−160) predicted by RNAfold. (D) Deletion of hA, hB, hC and hD (ΔhA+ΔhB+ΔhC+ΔhD) abolished the binding of circ‐0100519 fragment (nt 1−160) with USP7. (E) RNA pull‐down employing sequentially deleted circ‐0100519 fragments showed the binding region of circ‐0100519 with NRF2. (F) The structure of circ‐0100519 fragment (nt 450−550) predicted by RNAfold. (G) Double deletion of hE and hF (ΔhE + ΔhF) abolished the binding of circ‐0100519 fragment (nt 450−550) with NRF2. (H) RIP assays were performed employing anti‐USP7 or anti‐NRF2 antibodies in THP‐1 cells by agarose gel electrophoresis analysis. (I) RNA pull‐down assays were used to verify the interaction between circ‐0100519‐WT or circ‐0100519‐MUT and USP7/NRF2. (J) IP assays showed the ubiquitination modification level of NRF2 in THP‐1 cells with indicated treatments. (K) Schematic illustration of NRF2, displaying the wild‐type and truncations of NRF2. (L) The truncated NRF2 mutant harbouring 83−333aa retained the binding ability with circ‐0100519. Full length or the NRF2 fragments was transfected into THP‐1 cells for circ‐0100519 RNA‐pull down assay. A representative data set is displayed as mean ± SEM values of three independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 6

NRF2 is a functional downstream mediator of circ‐0100519 and USP7. (A–C) Colony formation, EdU assays (scale bar = 50 µm), and Transwell migration assays (scale bar = 1 mm) were used to detect the proliferative and migration potential of BC cells. (THP‐1 cells transfected with Vector or NRF2 plasmids were cocultured with exosomes isolated from circ‐0100519 knockdown tumour cells.) (D) The Kaplan–Meier curve was provided. (E) Photographs of tumours collected from 4T1‐bearing Balb/c mice with indicated treatment (n = 6 per group) (F) The curve graph exhibited the tumour volume measured at different time‐points (n = 6 per group). (G) Tumour sections were stained with TUNEL. (H) FCM was utilised to evaluate CD206 expression in Balb/c mouse tumours. (I) Immunofluorescent images of mouse tumour sections using circ‐0100519, NRF2, F4/80 and CD206 antibodies. Scale bar = 25 µm. (J) Bioluminescence images of lung metastasis model with indicated treatments. (K) Lung metastatic nodules were showed by H&E staining. Scale bar = 200 µm. (L) Western blotting showed NRF2 protein levels in BMDMs or TAMs from WT and NRF2^M‐KO mice. (M) Kaplan–Meier overall survival curves of NRF2^M‐KO+shNC and NRF2^M‐KO+shcirc‐0100519 mice with indicated treatment. (N) Immunofluorescent images of mouse tumour sections using F4/80 and CD206 antibodies. Scale bar = 50 µm. (O) Multicolour immunofluorescence using circ‐0100519, CK19, and CD206 antibodies in patient tumour sections (n = 20 pairs). Scale bar = 50 µm. (P) Patient tumour sections were stained with Ki67 IHC. Scale bar = 100 µm. A representative data set is displayed as mean ± SEM values of three to six independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 7

HIF‐1α serves as an upstream effector to enhance circ‐0100519 transcription. (A) Relationship between the expression of circ‐0100519 and HIF‐1α according to 196 tissues sequencing. (B) Potential HIF‐1α‐binding site in the genomic region adjacent to the TSS of EPSTI1. (C) Putative or mutant HIF‐1α binding sequences are shown by red characters in the binding areas. Lucifer activity of reporter vectors with the EPSTI1 promoter in BT‐549 cells. (D) HIF‐1α binding to the EPSTI1 promoter in BT‐549 cells was examined using CHIP. Utilised as negative controls were two circ‐0100519 promoter regions that were not anticipated to be bound by HIF‐1α. Data (3 independent biological replicates) are shown as mean ± SD. (E) Relative expression levels of circ‐0100519 after treatment with the HIF‐1α inhibitor PX‐478 assessed by qRT‐PCR. (F) Colony formation assays were used to detect the BC cells' proliferative potential. (G) FCM was used to assess CD206 expression in THP‐1. (F, G: BT‐549 cells were pretreated with 30 µM PX‐478 or PBS for 24 h and meanwhile transfected with circ‐0100519 or pcDNA, and then cocultured with THP‐1 cells. BT‐549 cells were pretreated with 30 µM PX‐478 or PBS for 24 h and cocultured with THP‐1 cells transfected with NRF2 plasmid or Vector) (H, I) Western blotting and qRT‐PCR in THP‐1 cells after pretreated BT‐549 cells with 30 µM PX‐478 or PBS for 24 h and cocultured with THP‐1 cells. (J) Photographs of tumours collected from 4T1‐bearing Balb/c mice with PX‐478 treatment. A representative data set is displayed as mean ± SEM values of three to five independent replicates. ns, not significant, *p < .05, **p < .01, ***p < .001, ****p < .0001.