Stabilization of IGF2BP1 by USP10 promotes breast cancer metastasis via CPT1A in an m6A-dependent manner

Metastasis leads to the vast majority of breast cancer mortality. Increasing evidence has shown that N6-methyladenosine (m6A) modification and its associated regulators play a pivotal role in breast cancer metastasis. Here, we showed that overexpression of the m6A reader IGF2BP1 was clinically correlated with metastasis in breast cancer patients. Moreover, IGF2BP1 promoted distant metastasis in vitro and in vivo. Mechanistically, we first identified USP10 as the IGF2BP1 deubiquitinase. USP10 can bind to, deubiquitinate, and stabilize IGF2BP1, resulting in its higher expression level in breast cancer. Furthermore, by MeRIP-seq and experimental verification, we found that IGF2BP1 directly recognized and bound to the m6A sites on CPT1A mRNA and enhanced its stability, which ultimately mediated IGF2BP1-induced breast cancer metastasis. In clinical samples, USP10 levels correlated with IGF2BP1 and CPT1A levels, and breast cancer patients with high levels of USP10, IGF2BP1, and CPT1A had the worst outcome. Therefore, these findings suggest that the USP10/IGF2BP1/CPT1A axis facilitates breast cancer metastasis, and this axis may be a promising prognostic biomarker and therapeutic target for breast cancer.

Ubiquitination, as a posttranslational modification, not only takes part in a large number of physiological events but also facilitates the origin and progression of cancer [28]. To date, the number of identified deubiquitinases (DUBs) is more than 100 [29]. By removing ubiquitin (Ub) from the substrate, DUBs can rescue the specific protein from being marked for degradation and maintain its protein level [28]. The deregulation of some DUBs is involved in the metastasis of BC [30][31][32]. Recently, the degradation of IGF2BPs has been reported to be related to the ubiquitin-proteasome system. However, the DUBs of IGF2BPs have not yet been discovered [33,34].
In our study, we report the oncogenic role of IGF2BP1 in BC metastasis and identified USP10 as the deubiquitinase of IGF2BP1 for the first time. Moreover, we identified CPT1A, which has been reported to endow BC cells with the potential for metastasis [35][36][37], as the target gene of IGF2BP1. Based on our clinical prognostic data, BC patients with high expression levels of USP10, IGF2BP1, and CPT1A had the worst outcome. Therefore, our study uncovers an essential role of the USP10/IGF2BP1/ CPT1A axis in the regulation of BC metastasis and provides a promising prognostic biomarker and therapeutic target for BC patients with metastasis.

Patients and clinical samples
Examination of the signatures on the informed consent forms was exempted following the rules approved by the Ethics Committees of Nanjing Drum Towel Hospital and the Affiliated Hospital of Nanjing University Medical School. The cases in which patients were simultaneously suffering from other malignant diseases had previously been ruled out. In this study, 80 paraffin-embedded tissues obtained from the Department of Pathology of Nanjing Drum Hospital after BC resection from 2010 to 2019 were used. Two pathologists individually confirmed the results of haematoxylin and eosin (HE) staining and immunohistochemical (IHC) staining. The ages of the patients ranged from 27 to 74 years. The latest follow-up date was December 21, 2019.

Cell lines and cell culture
Human HEK293T, MDA-MB-468 and HCC1806 cells were obtained from the Shanghai Institute of Biochemistry and Cell Biology. The corresponding results with MDA-MB-231 cells and metastatic derivative cells LM2, BM6, and 1833 were described in a previous study [38]. HCC1806 cells were cultured in 1640 medium supplemented with 10% foetal bovine serum (FBS) and penicillin-streptomycin solution (Biochannel, Nanjing, China). Other cell lines were cultured in DMEM supplemented with 10% foetal bovine serum (FBS) and penicillin-streptomycin solution (Biochannel, Nanjing, China). All cells were cultured in an incubator with 5% CO2 at 37 °C. Cells were stored at −80 °C using CELLSAVING (New Cell & Molecular Biotech, Suzhou, China).

Dot blot assay
This experiment followed the protocol in the bio-protocol database (https://bio-protocol.org/ e2095). Briefly, mRNA isolated from the indicated cells was further denatured and then spotted onto a Hybond-N+ membrane and crosslinked by UV Crosslinker. After blocking the membrane, an anti-m6A antibody (1:1000; Abcam, USA) was incubated with the membrane overnight at 4 °C. Then, the membrane was further incubated with an HRP-conjugated anti-mouse antibody. Finally, images were recorded with a CCD camera (Tanon, Shanghai, China). Methylene blue (MB) was used to interact with mRNA and also used for normalization.

Mouse model of BC lung metastasis
Female BALB/c nude mice (4-5 weeks, 18-20 g) were purchased from GemPharmatech (Nanjing, China) and raised in a specific pathogen-free (SPF) experimental animal room. Treated cells (1×10 6 /100 μl PBS) were injected through the tail vein of the nude mice. Pulmonary metastasis was evaluated by bioluminescence imaging at 4 or 8 weeks. Then, the mice were sacrificed, and the lung tissues were imaged and fixed in 4% paraformaldehyde for further analyses.

RNA isolation and quantitative real-time PCR (qRT-PCR)
Total RNA was isolated from cultured cells using TRIzol reagent (Invitrogen), and reverse transcription was carried out by HiScript II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme) following the protocol provided by the manufacturer. The primers involved in the qRT-PCR experiments are listed in Supplementary Table 2. qRT-PCR was performed with 2 × ChamQ SYBR qPCR Green Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China). The results were calculated by the 2 -ΔΔCt formula and normalized to the GAPDH levels.

RNA immunoprecipitation (RIP)
A MagnaRIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, MA, USA) was used following the manufacturer's instructions. The cell lysates were incubated with beads coated with 5 µg of control mouse IgG or anti-m6A antibody (Abcam, MA, USA) with gentle rotation at 4 °C overnight. Then, specifically captured products were obtained for the detection of CPT1A expression by qRT-PCR. The primers involved in the qRT-PCR experiments are listed in Supplementary Table 2.
Haematoxylin and eosin (HE) and immunohistochemical (IHC) staining HE and IHC staining followed the standard protocols described in a previous study [39]. The results of IHC staining were evaluated considering both the staining intensity and proportion of tumour cells. The intensity was evaluated as follows: 0, negative; 1, weak; 2, moderate; and 3 strong. The proportion score was determined as follows: 0, less than 5%; 1, 5% to 25%; 2, 26% to 50%; 3, 51% to 75%; and 4, greater than 75%. The final IHC staining index was calculated by multiplying the two numbers (the final values ranged from 0 to 12).

Immunofluorescence (IF) assay
The indicated cells were seeded in a 24-well plate at a density of 2 × 10 4 cells per well. After 24 h, the cells were fixed with methanol for 15 min at room temperature and treated with 0.3% Triton X-100 for 15 min. After blocking for 30 min at room temperature, the indicated antibodies were added to the plate separately for incubation overnight at 4 °C. After being washed with PBS 3 times, the plate was further incubated with the secondary antibodies for 2 h at room temperature in the darkroom. DAPI was used to stain the nuclei. Finally, the cells were observed and imaged by fluorescence microscopy (LEICA DMi8, Germany).

Western blotting assay
Western blotting assays were performed according to standard protocols. Briefly, RIPA lysis buffer (P0013B, Beyotime) supplemented with protease inhibitor cocktail (P1005, Beyotime) and PMSF (ST507-10 ml, Beyotime) was used to lyse cells for 30 min at 4 °C. After centrifugation, the supernatants were collected. The protein concentration was also determined before western blotting. After resolving the proteins by SDS-polyacrylamide electrophoresis, the proteins were transferred to PVDF membranes. After the blotted membranes were blocked for 1 h, they were incubated with the indicated primary antibody overnight at 4 °C and then incubated with HRP-conjugated secondary antibodies for 1 h. Finally, the results were recorded by a CCD camera (Tanon, Shanghai, China) by ECL chemiluminescence (Millipore, Billerica, MA, USA). All antibodies are listed in Supplementary Table 3.

Transwell assay
The invasion and migration abilities of the cells were evaluated using Transwell chambers with 8 µm pore filters. After 48 h of transfection, cells (3 × 10 4 /100 µl) were seeded on the upper chambers coated with or without 70 µl of Matrigel (BD Biosciences) in serum-free DMEM. DMEM containing 10% FBS was added to the lower chambers. After incubation for 24 h at 37 °C, the cells in the chambers were collected and fixed with methanol or 4% paraformaldehyde. After removing the nonmigrating or noninvading cells with cotton swabs, the chambers were stained with crystal violet solution for 20 min. The results were recorded at ×200 magnification under a microscope. All assays were repeated three times in duplicate.

RNA-seq and MeRIP-seq
RNA-seq and MeRIP-seq were performed by Gene Denovo (Guangzhou, China). Briefly, after extracting total RNA from LM2 cells with stable IGF2BP1 knockdown and the corresponding control cells, RNA was immunoprecipitated with an anti-m6A antibody and further used to construct RNA libraries, which were sequenced using Illumina NovaSeq TM 6000 by Gene Denovo Biotechnology Co. (Guang Zhou, China).

Statistical analysis
A paired t test or Wilcoxon signed-rank test was used to evaluate the mRNA levels of the major m6A-related enzymes in 112 breast cancer samples and the paired normal tissues from TCGA database. The differences in IGF2BP1 IHC staining indices in 11 primary breast tumours and the paired normal breast tissues were assessed by the Wilcoxon signed-rank test (grouped). Kaplan-Meier survival curves were analysed by the log-rank test. The correlations between the IHC staining indices of USP10 and IGF2BP1 were analysed by Spearman's test. Student's t test or conventional one-way ANOVA was used to identify statistically significant differences (P < 0.05) between the treated and control groups. GraphPad PRISM 8.0 was used to analyse all the data (GraphPad Software, La Jolla, CA, USA). All experiments were repeated three times. Data are shown as the mean ± SD.

Elevated IGF2BP1 expression resulted in poor prognosis of BC
To identify the roles of m6A regulators in BC, we first screened the expression of key m6A writers (METTL3, METTL14, and WTAP), erasers (ALKBH5 and FTO), and readers (YTHDC1, YTHDC2, YTHDF1, YTHDF3, IGF2BP1, IGF2BP2, and IGF2BP3) in BC samples and the paired normal breast tissues using The Cancer Genome Atlas (TCGA) data. The mRNA expression levels of YTHDC2, YTHDF1, YTHDF3, IGF2BP1, and IGF2BP3 were significantly increased in tumour tissues compared with paired normal breast tissues, while the METTL14, WTAP, FTO, and IGF2BP2 expression levels were significantly decreased in tumour tissues ( Figure 1A). Correspondingly, the mRNA expression levels of METTL3, ALKBH5, and YTHDC1 showed no significant difference between the paired normal and BC tissues ( Figure S1A). In addition, we examined all of these genes of interest in the BC tissues and paired controls in GEO datasets (GSE22820 and GSE86374) ( Figure  1B). By taking the intersection of these results, we found that the mRNA expression levels of IGF2BP1, WTAP and FTO showed significant differences between the paired normal and BC tissues in all three datasets ( Figure 1C). Among these three genes, only IGF2BP1 mRNA expression was associated with poor overall survival (OS) and disease-free survival (DFS) of BC patients based on TCGA datasets ( Figure 1D and Figure S1C).
Furthermore, to explore the function of m6A modification in the metastasis of BC, we first detected the RNA m6A levels in MDA-MB-231, LM2 (MDA-MB-231 lung metastatic derivative), BM6 (MDA-MB-231 bone metastatic derivative), and 1833 (MDA-MB-231 brain metastatic derivative) cells. The RNA m6A levels were markedly increased in LM2, BM6, and 1833 cells compared with that in MDA-MB-231 cells as determined by dot blot assay ( Figure 1E). Then, we examined the mRNA and protein expression levels of the key m6A writers (METTL3, METTL14, and WTAP), erasers (ALKBH5 and FTO), and readers (YTHDC1, YTHDC2, YTHDF1, YTHDF3, IGF2BP1, IGF2BP2, and IGF2BP3) in the four cell lines. Interestingly, the protein expression of IGF2BP1 was significantly higher in LM2, BM6, and 1833 cells than in MDA-MB-231 cells, but this phenomenon did not occur at the mRNA level ( Figure  1F and Figure S1B).
We further confirmed this finding in our clinical samples and found elevated levels of the IGF2BP1 protein in BC tissues compared with paired normal breast tissues by immunohistochemical (IHC) staining (n=11; Figure 1G). To further evaluate the prognostic effect of IGF2BP1, we also assessed IGF2BP1 expression by IHC staining in 80 BC tissue samples. BC patients with higher IGF2BP1 expression had shorter OS and DFS ( Figure 1H and Figure S1D). Moreover, primary BC tissues from patients with distant metastasis had higher expression of IGF2BP1 than those without distant metastases (n=7; Figure 1I). These data suggest that IGF2BP1 might play a crucial role in the malignant process of BC, especially in its metastasis.

IGF2BP1 promoted BC metastasis in vitro and in vivo
To determine the role of IGF2BP1 in BC metastasis, we first established stable IGF2BP1 knockdown BC cells (LM2 and 1833) (Figure 2A and Figure S2A). The results showed that knockdown of IGF2BP1 suppressed the migration and invasion ability of LM2 and 1833 cells ( Figure 2B and Figure  S2B-C). In addition, we constructed IGF2BP1-overexpressing BC cells using the parental MDA-MB-231 cell line ( Figure 2C). The results showed that ectopic expression of IGF2BP1 significantly promoted the migration and invasion ability of MDA-MB-231 cells ( Figure 2D).
We further explored the prometastatic effect of IGF2BP1 in vivo. LM2-luciferase cells with IGF2BP1 knockdown and the corresponding control cells were injected into nude mice through the tail vein. Intriguingly, downregulation of IGF2BP1 significantly suppressed lung metastasis compared with that in the control group after 4 weeks, as indicated by bioluminescence imaging and haematoxylin and eosin (HE) staining of pulmonary metastases ( Figure  2E-F). These data suggest that IGF2BP1 may act as an oncogene that promotes BC metastasis.

The USP10/IGF2BP1 axis promoted BC metastasis in vitro and in vivo
The expression of USP10 was significantly upregulated in BC tissues compared with paired normal breast tissues based on TCGA data ( Figure  4A). Moreover, higher expression of USP10 was associated with a poor prognosis of BC patients ( Figure 4B). We performed IHC staining of 80 BC tissues and confirmed that higher USP10 expression was associated with shorter OS and DFS in BC patients ( Figure 4C-D and Figure S4A). As shown in Figure S4B, there was no significant relationship between the mRNA expression of USP10 and IGF2BP1. However, the protein level of USP10 were significantly correlated with that of IGF2BP1 in BC samples (n=80, p<0.001, Figure 4E).
Subsequently, knocking down USP10 dramatically suppressed the migration and invasion of LM2 and 1833 cells ( Figure 4F and Figure S4C-D). The USP10 inhibitor Spautin-1 also inhibited the migration of LM2 and 1833 cells ( Figure 4G). Moreover, overexpressing IGF2BP1 markedly rescued the inhibition of migration caused by USP10 knockdown in LM2 and 1833 cells ( Figure 4H and Figure S4E-F). Conversely, suppressing IGF2BP1 significantly blocked USP10 overexpression-induced migration in MDA-MB-231 cells ( Figure 4I). Next, BC lung metastasis mouse models were constructed, and the number and size of lung metastatic lesions were analysed by bioluminescence imaging and HE staining of the pulmonary metastases at 8 weeks after injection. The results indicated that IGF2BP1 could block BC metastasis caused by USP10 overexpression in vivo ( Figure 4J-K). These results suggest a pivotal role of the USP10/IGF2BP1 axis in BC metastasis.

IGF2BP1 recognized the m6A modification on CPT1A mRNA to maintain mRNA stability
To explore the molecular mechanism by which IGF2BP1 promoted BC metastasis, we performed methylated RNA immunoprecipitation sequencing (MeRIP-seq) and RNA sequencing (RNA-seq) in LM2 cells with stable IGF2BP1 knockdown and the corresponding control cells. The m6A peaks identified by MeRIP-seq were mainly distributed in the coding sequence (CDS), 3' untranslated region (3'UTR), start codon, and stop codon ( Figure 5A). m6A occurs mostly in the RRACH (R=G or A, H=A, C, or U) consensus sequence [20,43]. Our MeRIP-seq results revealed that the GGACT motif was highly enriched in the immunopurified RNA in both control LM2 cells and IGF2BP1 knockdown LM2 cells and the abundance of m6A was decreased in the mRNA of IGF2BP1 knockdown LM2 cells ( Figure 5B). Furthermore, MeRIP-seq revealed that 6120 transcripts had fewer m6A peaks in IGF2BP1 knockdown cells than in control cells (fold change >1.5). Moreover, RNA-seq revealed that 300 transcripts were markedly decreased in IGF2BP1 knockdown cells compared with control cells (fold change >1.5) ( Figure 5C and Figure S5A-B). Integrating the MeRIP-seq and RNA-seq results, 89 transcripts overlapped ( Figure 5C). Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis showed that these transcripts mainly functioned in the Hedgehog signaling pathway and PPAR signaling pathway ( Figure 5D). Then, the expression of six candidate genes (CCND2, MEGF8, GLI3, CPT1A, ANGPTL4, PLIN4) in these two pathways was further verified in IGF2BP1deficient LM2/1833 cells and IGF2BP1-overexpressing MDA-MB-231 cells by qRT-PCR ( Figure 5F-G and Figure S5C). We found that the expression of CPT1A and GLI3 was consistently regulated by IGF2BP1 in these cells. Additionally, the m6A abundance in CPT1A and GLI3 mRNA was significantly decreased upon IGF2BP1 knockdown, as shown by the MeRIP-qPCR assay ( Figure 5H and Figure S5D). Overexpression of IGF2BP1 also enriched the m6A abundance in CPT1A mRNA ( Figure 5H). The Integrative Genomics Viewer (IGV) also indicated that the m6A levels on the exons of CPT1A mRNA were decreased in IGF2BP1 knockdown LM2 cells compared with control cells ( Figure 5E). Moreover, the protein level of CPT1A was decreased in IGF2BP1 knockdown LM2/1833 cells but increased in IGF2BP1-overexpressing MDA-MB-231 cells compared with the corresponding control cells ( Figure 5J-K). IGF2BP1 is recognized to stabilize target mRNA transcripts with m6A motifs [23]. We then treated IGF2BP1 knockdown LM2 cells and control cells with actinomycin D, an inhibitor of transcription [44], which showed that the mRNA level of CPT1A was less stable in IGF2BP1 knockdown LM2/1833 cells ( Figure 5I). However, we found that knockdown of IGF2BP1 reduced m6A modification of GLI3, but did not affect GLI3 mRNA stability, indicating that IGF2BP1 may not completely depend on m6A modification to regulate GLI3 expression ( Figure  S5D-E). Together, our findings indicated that m6A modification maintains stable CPT1A expression in an IGF2BP1-dependent manner in metastatic BC cells.

IGF2BP1 promoted BC metastasis via the upregulation of CPT1A
To further characterize the function of CPT1A in BC metastasis, we first downregulated CPT1A with two different specific siRNAs and confirmed the knockdown efficiency by western blotting ( Figure 6A and Figure S6A). In addition, MDA-MB-231 cells stably overexpressing CPT1A were constructed ( Figure 6B). We found that knockdown of CPT1A dramatically suppressed the migration and invasion abilities of LM2 and 1833 cells, while CPT1A overexpression promoted the migration and invasion abilities of MDA-MB-231 cells ( Figure 6C-D and Figure S6B). Furthermore, ectopic expression of CPT1A rescued the decreased migration ability of IGF2BP1 knockdown LM2 and 1833 cells ( Figure  6E&G and Figure S6C). Moreover, downregulating CPT1A by specific siRNAs blocked the enhanced migration abilities observed by overexpression of IGF2BP1 ( Figure 6F&H). Thus, our data suggest that IGF2BP1 promoted BC metastasis by upregulating CPT1A expression.
Moreover, high expression of CPT1A mRNA also predicted poor OS of patients according to TCGA data ( Figure 6I). Furthermore, IHC staining in 80 clinical BC samples confirmed that patients with higher CPT1A protein expression had shorter OS and DFS ( Figure 6J-K and Figure S6D).
To further confirm the role of the USP10/ IGF2BP1/CPT1A axis in BC metastasis, we analysed the IHC staining images of USP10, IGF2BP1, and CPT1A in our BC samples. As shown in Figure 7A, the USP10 levels distinctly correlated with the IGF2BP1 and CPT1A levels in BC tissues. Moreover, BC patients with high levels of USP10, IGF2BP1, and CPT1A had the worst outcomes compared with those with low levels of ang or all of these indicators. ( Figure 7B). These data suggest that the USP10/ IGF2BP1/CPT1A axis may boost human BC progression and is correlated with BC patient survival.

Discussion
The decisive role of m6A modification in cancer progression has been widely reported in recent years [19,45]. m6A modification is dynamically regulated by its writers, erasers, and readers, which are involved in the progression of cancer metastasis [13,16,46], including BC [15,[47][48][49][50]. Here, we found that the m6A reader IGF2BP1 was upregulated in BC tissues, especially BC metastatic lesions, and was associated with a poor prognosis of BC. Functionally, we first identified that the deubiquitinase USP10 could bind to and stabilize the IGF2BP1 protein, resulting in its high expression level in BC. Furthermore, IGF2BP1 directly bound to and recognized the m6A modification on CPT1A mRNA to maintain its stability, which had been reported to endow BC cells with the potential for metastasis [35,37]. Thus, IGF2BP1 is capable of promoting BC distant metastasis ( Figure 7C).  IGF2BP1 is regarded as a classic RNA-binding protein involved in the regulation of gene expression [51,52]. Recent evidence has confirmed that IGF2BP1 can function as an m6A reader to recognize m6A sites and enhance targeted mRNA stability [23]. Recently, IGF2BP1 was reported to promote BC lung metastasis by cooperating with METTL3 and FTO and stabilizing the keratin 7 (KRT7)-AS/KRT7 mRNA duplex by recognizing m6A at A877 on KRT7-AS [27]. Here, we first discovered that IGF2BP1 was upregulated in BC tissues, especially in BC patients with distant metastasis. Moreover, the IGF2BP1 protein level, but not its mRNA level, was consistently elevated in the lung, bone, and brain BC metastatic derivative cell lines compared with the parental cell line, suggesting that IGF2BP1 plays an extensive role in the distant metastasis of BC.
The discrepancies of IGF2BP1 function in breast cancers need to be noted. It has been reported that IGF2BP1 functions as a tumor suppressor [53,54]. However, recent studies from different research groups demonstrated the oncogenic roles of IGF2BP1 in breast cancer, maintaining the glycolysis and stemness of BCSCs through IGF2BP1/c-Myc axis [55,56], promoting proliferation of breast cancer cells [57] and contributing the progression of breast cancer [58]. Based on GEO and TCGA datasets, Zhong et al. found that IGF2BP1 was an independent prognostic factor of breast cancer, and higher expression level of IGF2BP1 was associated with shorter overall survival of breast cancer patients [59], which was consistent with our results. In our study, we found the oncogenic role of IGF2BP1 in TNBC cell lines, and we also showed that high expression of IGF2BP1 was clinically correlated with metastasis in breast cancer patients. Therefore, due to the complexity of the pathological types of breast cancer, more studies are warranted.  The IGF2BP1 protein level has been reported to be widely upregulated in several aggressive cancers [27,[60][61][62][63][64]. However, the modulatory mechanisms underlying the abnormally elevated protein levels of IGF2BP1 in cancers are poorly understood. Previous studies have shown that posttranslational modification of some m6A regulators, such as ubiquitination [65], phosphorylation [66], sumoylation [67,68], and lactylation [69,70], could influence their biological function [33,65,71]. For example, F-box and WD repeat domain-containing 7 (FBW7), an E3-ubiquitin ligase, ubiquitinates YTHDF2 and suppresses tumour progression by antagonizing the tumour-promoting effect of YTHDF2 in ovarian cancer [72]. The ubiquitination of IGF2BP1/2/3 was first reported in their interaction with the E3-ubiquitin ligase tripartite motif-containing protein 25 (TRIM25) in non-small cell lung cancer (NSCLC) [33]. In addition, F-box/ SPRY domain-containing protein 1 (FBXO45), an E3-ubiquitin ligase, was found to promote IGF2BP1 ubiquitination and attenuate hepatocellular carcinoma (HCC) progression [73]. However, the deubiquitinase of IGF2BP1 has not yet been reported. Our study first identified USP10 as the deubiquitinating enzyme of IGF2BP1, which maintained a high level of IGF2BP1 expression in BC. Moreover, we highlighted that USP10 was an independent prognostic marker that predicts the outcomes of BC patients.
Metabolic reprogramming is a hallmark of malignant tumours [74]. Fatty acid oxidation (FAO) is of great importance in tumour progression because it increases the production of ATP and NADPH and enables cancer cells to survive under metabolic stress [75,76]. In our previous study on oesophageal cancer (ESCA), the m6A reader HNRNPA2B1 upregulated the expression of ATP citrate lyase (ACLY) and acetyl-CoA carboxylase (ACC1), two fatty acid synthetic enzymes, and promoted ESCA progression and metastasis [77]. In this study, by MeRIP and experimental verification, we first found that the FAO rate-limiting enzyme CPT1A [75,[78][79][80] was the downstream target of IGF2BP1. In colorectal cancer (CRC), CPT1A has been shown to strengthen resistance to anoikis and eliminate reactive oxygen species (ROS), thus promoting CRC cell metastatic capacity [79]. Here, we found that CPT1A mRNA could be stabilized by IGF2BP1 via recognition of its m6A modification, which led to its high level in BC and mediated IGF2BP1-induced distant metastasis in BC.

Conclusion
Our work revealed that the USP10/IGF2BP1/ CPT1A axis promoted BC metastasis. In clinical samples, the USP10 levels were correlated with the IGF2BP1 and CPT1A levels in BC tissues, and BC patients with high levels of USP10, IGF2BP1, and CPT1A had the worst outcome. Therefore, the USP10/IGF2BP1/CPT1A axis might be a potential predictor and therapeutic target for BC.

Ethics statement
The use of human breast cancer tissues and the waiver of patient consent in this study were approved by the Clinical Research Review Board of Nanjing Drum Tower Hospital. The study was conducted according to the principles expressed in the Declaration of Helsinki. The animal studies were reviewed and approved by the Ethics Committee for Animal Studies at The First Affiliated Hospital of Nanjing Medical University.

Author contributions
JJ S, QY Z, X Y and JH Y contributed equally to this work. JJ S and QY Z performed the experiments; X Y and JH Y analyzed the data; YX Y, BJ W, YP F, and HM Z provided the clinical samples; SQ G, C C, and Y Z constructed lung metastasis models in vivo; HY W analyzed the pathological tissues; JJ S and Q W wrote the paper; WJ Z, SY W, Q W, GF X, YZ Y, ZD W, and B W commented on the study and revised the paper; WJ Z, SY W, Q W and GF X designed and supervised the research.

Data availability
Data are available upon reasonable request.