PRL-3 up-regulates exosomal ITGαvβ5 expression to promote liver pre-metastatic niche formation and colon cancer liver metastasis

Abstract

Liver pre-metastatic niches (PMN) formation is a pivotal process in colorectal cancer liver metastasis (CLM). Phosphatase of regenerating liver-3 (PRL-3) has been demonstrated as a key factor in promoting CRC progression (e.g., therapeutic resistance and metastasis), but its role in liver PMN formation remains unknown. Using mouse models and CRC patient samples, we herein reveal that high PRL-3 expression in CRC cells could enhance the recruitment of myeloid-derived suppressor cells (MDSCs) into the liver and impair the hepatic infiltration of CD8⁺ T cells, thereby promoting the liver PMN formation and CLM. Mechanistically, high PRL-3 expression could activate the Src/STAT3 signaling pathway in CRC cells and thus up-regulate integrin αvβ5 (ITGαvβ5) expression in their secreted exosomes, which could specifically target F4/80⁺ macrophages in the liver to activate the P38/STAT1 signaling pathway. With this activation of P38/STAT1 pathway, the secretion of C-X-C motif chemokine ligand 12 (CXCL12) from F4/80⁺ macrophages is significantly improved, which could enhance the recruitment of MDSCs into the liver and impair the hepatic infiltration of CD8⁺ T cells, ultimately leading to the liver PMN formation and CLM. Taken together, our findings not only uncover the important role of PRL-3 in CLM via promoting the liver PMN formation, but also provide the evidence for the treatment of CLM by targeting PRL-3.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40164-025-00733-5.

Introduction

Colorectal cancer (CRC) is a common gastrointestinal malignancy and appropriately 30% of patients would develop liver metastasis with 5-year survival rate less than 20% [1, 2]. Pre-metastatic niches (PMN), majorly characterized by the infiltration of bone marrow-derived progenitor cells, has been demonstrated as the foundation of metastasis [3], as PMN can facilitate the survival and colonization of tumor cells in target organs by enhancing the evasion of immune surveillance, remodeling extracellular matrix (ECM), and modulating angiogenesis [4]. For example, as a type of immature bone marrow-derived cells, myeloid-derived suppressor cells (MDSCs) have been identified as the major cellular components of PMN in the lung prior to the lung metastasis of breast cancer [5, 6]. Therefore, uncovering the key factors contributing to the liver PMN formation and elucidating their regulatory mechanisms could not only provide new insight into understanding of CRC liver metastasis (CLM), but also facilitate the development of novel strategies for CRC treatment.

Phosphatase of regenerating liver-3 (PRL-3) belongs to the protein tyrosine phosphatase family [7]. In the past decade, numerous researches have demonstrated the important role of PRL-3 in promoting CRC progression (e.g., therapeutic resistance and metastasis) via multiple mechanisms, including the activation of PI3K/AKT pathway in CRC cells to evade stress-induced apoptosis and promotion of epithelial-mesenchymal transition (EMT) [8–11]. Moreover, emerging evidences have revealed that high PRL-3 expression could remodel the tumor microenvironment via enhancing the secretion of inflammatory cytokines (e.g., interlukin-8, IL-8) and improving the infiltration of immunosuppressive cells (e.g., tumor-associated macrophages, TAMs) to promote CRC progression [12, 13]. In spite of these pioneered findings, it remains elusive whether PRL-3-induced CLM is based on the regulation of liver PMN formation as PMN is the foundation of metastasis. Although several studies have shown that tumor-derived exosomes including CRC-derived exosomes could regulate PMN formation in the liver and promote CLM [14–16], their relationship with PRL-3 and the underlying regulatory mechanisms remain to be elucidated.

We herein, for the first time, uncovered the important role of PRL-3 in promoting the liver PMN formation and CLM. Mechanically, high PRL-3 expression could activate the classic Src/STAT3 signaling pathway in CRC cells to up-regulate integrin αvβ5 (ITGαvβ5) expression in their secreted exosomes, which could specifically target F4/80⁺ macrophages in the liver to activate the P38/STAT1 signaling pathway, leading to the enhanced secretion of C-X-C motif chemokine ligand 12 (CXCL12) from macrophages to recruit MDSCs for the formation of liver PMN and ultimately induce CLM (Fig. 1). Our findings not only uncover the new biological function of PRL-3, but also provide a novel strategy for the treatment of CLM by targeting PRL-3.

Fig. 1.

Schematic illustration of the mechanism by which PRL-3 promotes the liver PMN formation and CLM. High PRL-3 expression could activate the Src/STAT3 signaling pathway in CRC cells to up-regulate TGαvβ5 expression in their secreted exosomes, which could specifically target F4/80⁺ macrophages in the liver and activate their P38/STAT1 signaling pathway, leading to the enhanced secretion of CXCL12 from hepatic F4/80⁺ macrophages for recruitment of MDSCs into the liver to induce the liver PMN formation and ultimate CLM

Methods

Patients and clinical samples

Tumor samples of 338 colon adenocarcinoma patients who underwent R0 resection without neoadjuvant chemotherapy between 2014 and 2018 were collected from the Department of Gastrointestinal Surgery, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University. Among these patients, there are 30 patients at stage I, 115 patients at stage II, 118 patients stage III, and 75 patients stage IV with liver metastasis. Serum samples were collected from 21 healthy volunteers and 47 colon adenocarcinoma patients. Pathological diagnosis, Ki67, and other biomarkers were verified independently by two pathologists. All samples were collected with informed consent from the donors according to the International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS). The study was approved by the Institutional Review Board (IRB) of Sun Yat-Sen Memorial Hospital.

Cell culture

Human CRC cell lines (LoVo, HT116, HT29, and SW480) were purchased from the Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences. Mouse CRC cell lines (MC38 and CT26) were obtained from the American Type Culture Collection (ATCC). The cells were incubated in culture medium containing 10% fetal bovine serum (FBS) and cultured at 37 °C in a humidified atmosphere containing 5% CO₂.

PRL-3 knockdown and over-expression

CT26 and HT116 cells were respectively seeded in 6-well plates (1 × 10⁵ per well) and incubated in 2 mL of culture medium containing 10% FBS overnight. Subsequently, the mixture of lipofectamine 3000 and siRNA targeting PRL-3 (Table S1) was added according to the transfection reagent protocols. After 12 h incubation, the medium was removed and the cells were further incubated in fresh medium for another 48 h. Thereafter, the cells were digested by trypsin and the total RNA was extracted using Trizol (Invitrogen, USA) for quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis of PRL-3 mRNA expression. The total protein was also extracted using lysis buffer supplemented with protease inhibitor cocktail and phenylmethanesulfonyl fluoride (PMSF) for western blot analysis of PRL-3 protein expression. For the stable PRL-3 silencing, the cells were transfected with the shRNA targeting PRL-3 (Table S2) enveloped by lentivirus according to the method described above. To up-regulate PRL-3 expression, MC38 cells were incubated with the PRL-3-expressing plasmid (Table S2) enveloped by lentivirus for 12 h. Subsequently, the medium was removed and the cells were further incubated in fresh medium. After 48 h incubation, the total RNA and protein were extracted from the cells (denoted MC38-PRL-3) for qRT-PCR and western blot analysis of PRL-3 expression.

PRL-3 knockout

CRISPR-Cas9 gene editing technology was used to knock out PRL-9 in CT 26 cells. In brief, CT26 cells were seeded in 6-well plates (1 × 10⁵ per well) and incubated in 2 mL of culture medium containing 10% FBS overnight. Subsequently, the cells were transfected with pSB-CRISPR plasmids and SB100X transposase according to method described above. The single-guide RNA (sgRNA) targeting PRL-3 includes two sequences, sgPRL-3#1 (TTTGACGATGGGGCGCCCCC) and sgPRL-3#2 (GAAGGATGGCATCACCGTTG). The efficiency of PRL-3 knockout was examined by qRT-PCR and western blot.

Exosome isolation and identification

To isolate exosomes from serum, peripheral blood samples were centrifuged at 1000 ×g for 20 min at 4 °C to obtain the serum. After adding thrombin, the serum was diluted with an equal volume of phosphate buffered saline (PBS) solution and centrifuged (2000 ×g) at 4 °C for 30 min. The supernatant was then transferred to an ultracentrifuge tube and centrifuged (12,000 ×g) at 4 °C for 45 min. Thereafter, the supernatant was collected and further centrifuged (110,000 ×g) at 4 °C for 2 h. The obtained precipitate was suspended in PBS solution and filtered through a 0.22 μm filter to obtain the exosomes. The specific markers (Alix, Tsg101, CD9, CD81, RSP11, and calnexin) were examined to identify the collected exosomes. The size of exosomes was examined by nanoparticle tracking analysis (NTA) system and their morphology was observed by transmission electron microscopy (TEM).

For the isolation of exosomes derived from MC38-PRL-3 or CT26 cells, the cells were incubated in a 150 cm² culture flask and the medium was replaced with fresh medium when the confluence reaches ~ 90%. After 24 h incubation, the culture medium was collected and centrifuged at 4 °C (1000 ×g) for 10 min. Subsequently, the supernatant was collected and the ExoQuick-TC exosome precipitation kit (System Biosciences, USA) was used to isolate the exosomes. The obtained exosomes were suspended in PBS solution and then filtered through a 0.22 μm filter to remove large extracellular vesicles. The exosomes were identified by examining the specific markers (Alix, Tsg101, CD9, CD81, RSP11, and calnexin). Their size and morphology were examined by NTA and TEM, respectively.

High-throughput proteomics

The exosomes in the serum of colon adenocarcinoma patients with high PRL-3 (n = 3) and low PRL-3 expression (n = 3) were isolated according to method described above. The obtained exosomes were lysed with strong RIPA buffer (Beyotime, Wuhan, China) and 200 µg of exosomal protein was used for proteomics analysis. Before the analysis, the protein was precipitated in acetonitrile and desalted on an SPE column. Subsequently, the desalted samples were dried and resuspended for proteomics analysis using the Data independent acquisition (DIA) method, in which all ions within a selected m/z range are fragmented and analyzed in a second stage of tandem mass spectrometry. The final protein expression profile was profiled by Proteome Discoverer Software (ver. 2.4) containing a complex proteomic standard. Differential protein expression was screened using the criteria of log2FC ≥ 1 and FDR < 0.05. Gene set enrichment analysis (GSEA) was employed to analyze the biological functions of differentially expressed proteins, including biological processes (BPs) and molecular functions (MFs).

Animals

C57BL/6 female mice (6–8 weeks old) were purchased from Guangdong Yaokang Biotechnology Co., Ltd. and bred under specific pathogen-free (SPF) conditions. All in vivo studies were performed by a protocol (#AEP20220056) approved by the Institutional Animal Care and Use Committee at Sun Yat-Sen Memorial Hospital.

Influence of PRL-3 on the liver PMN formation and CLM

The influence of PRL-3 on the liver PMN formation and CLM was evaluated using an orthotopic CRC tumor-bearing mouse model. In brief, healthy C57BL/6 mice were anesthetized with 2% isoflurane and the abdominal cavity was opened to expose the cecum. Subsequently, 1 × 10⁶ luciferase (Luc)-expressing MC38 (denoted MC38-Luc) cells suspended in 50 µL of matrix gel were injected into the cecum and the abdominal cavity was closed with surgical grapes. At 2, 3, and 5 weeks post the injection, the tumor growth and liver metastasis were monitored by using bioluminescence imaging. Prior to imaging, D-luciferin was intraperitoneally injected at a dose of 150 mg/kg and the mice were viewed using an IVIS Lumina III imaging system (Perkin-Elmer, USA). After the imaging, the mice were sacrificed and the livers were collected for histopathological examination of liver metastasis. Additionally, the livers were homogenized into single cell suspension, which was subjected to examine the liver PMN formation via flow cytometry analysis of the infiltration of MDSCs and CD8⁺ T cells into the liver and qRT-PCR analysis of the expression of metastasis-associated inflammatory factors.

Influence of CRC-derived exosomes on the liver PMN formation and CLM

The exosomes derived MC38-PRL-3 or CT26 cells were isolated according to the method described above. Subsequently, the obtained exosomes were suspended in PBS solution and intravenously injected into healthy C57BL/6 mice at a dose of 0.3 mg/kg every two days (n = 12). After consecutive injection for 2 weeks, half of the mice (n = 6) were sacrificed and the livers were collected for the detection of liver PMN formation via flow cytometry analysis of the infiltration of MDSCs. The rest of mice (n = 6) were anesthetized with 2% isoflurane and the abdominal cavity was opened to expose the spleen. Subsequently, 1 × 10⁶ MC38-Luc cells suspended in 50 µL of matrix gel were injected into the spleen and the abdominal cavity was closed with surgical grapes. At 21 days post the cell injection, the mice were viewed using an IVIS Lumina III imaging system to examine the liver metastasis according to the method described above. After the imaging, the mice were sacrificed and the livers were collected for histopathological examination of liver metastasis and flow cytometry analysis of the infiltration of MDSCs and CD8⁺ T cells into the liver.

Establishment of F4/80⁺ macrophages and MDSCs

Macrophages and MDSCs were established by incubating mouse bone marrow-derived monocytes (BMDMs) in the conditioned medium. Briefly, femurs of healthy C57BL/6 mice were collected under aseptic conditions and the bone marrow was flushed out with PBS solution. After filtration through 70 μm membrane, the BMDMs were collected via centrifugation at 1500 rpm. Subsequently, the obtained BMDMs were suspended in culture medium containing 10% FBS and 20 ng/mL of macrophage colony-stimulating factor (M-CSF) and then transferred to a 75 cm² culture flask. After incubation for 5–7 days, flow cytometry was used to examine the specific markers of F4/80⁺ macrophages. For the establishment of MDSCs, the isolated BMDMs were cultured in the medium containing 10% FBS and 20 ng/mL of granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin 6 (IL-6). After incubation for 3–5 days, flow cytometry was used to examine the specific markers of MDSCs.

Transwell assay

Transwell assay was conducted to evaluate the influence of CRC-derived exosomes on the ability of F4/80⁺ macrophages to recruit MDSCs. In brief, MDSCs were seeded in the upper chamber of a 6-well Boyden Chamber device with 0.4 μm pore size (Corning, USA) at a density of 2 × 10⁵ cells/mL, whereas F4/80⁺ macrophages pre-treated with the exosomes derived from MC38-PRL-3 or CT26 cells were seeded in the lower chamber at a density of 2 × 10⁵ cells/mL. After 24 h incubation, MDSCs attached to the bottom of the upper chamber were counted and imaged under a microscope.

Transcriptome sequencing and data analysis

F4/80⁺ macrophages were seeded in 6-well plates (1 × 10⁵ per well) and incubated in 2 mL of culture medium containing 10% FBS overnight. Thereafter, the exosomes derived from MC38-PRL-3 cells were added and the cells were further incubated for another 48 h. Subsequently, the cells were digested by trypsin and the total RNA was extracted using Trizol for RNA-sequencing analysis (IGE Biotechnology, Guangzhou, China). The differentially expressed genes (DEGs) and downstream enrichment analysis were performed using the company’s online interaction analysis system.

Influence of F4/80⁺ macrophage depletion and CXCL12 neutralization on CLM

Clodronate liposome suspension (Yeasen Biotechnology, Shanghai, China) was intravenously injected into healthy C57BL/6 mice (n = 3) at a dose of 200 µL per mouse. At 48 h post the injection, the mice were sacrificed and the livers were collected for the determination of hepatic macrophage depletion using immunofluorescence (IF). Subsequently, healthy C57BL/6 mice with hepatic macrophage depletion were randomly divided into two groups (n = 6) and then intravenously injected with the exosomes derived from MC38-PRL-3 or CT26 cells at a dose of 0.3 mg/kg every two days. After three consecutive injections, the mice were anesthetized with 2% isoflurane and the abdominal cavity was opened to expose the spleen. Subsequently, 1 × 10⁶ MC38-Luc cells suspended in 50 µL of matrix gel were injected into the spleen and the abdominal cavity was closed with surgical grapes. At 21 days post the cell injection, the mice were viewed using an IVIS Lumina III imaging system to examine the liver metastasis. After the imaging, the mice were sacrificed and the livers were collected for histopathological examination of liver metastasis and flow cytometry analysis of the infiltration of MDSCs and CD8⁺ T cells into the liver.

To evaluate the influence of CXCL12 neutralization on CLM, healthy C57BL/6 mice were randomly divided into two groups (n = 6) and then received the intravenous injection of the exosomes derived from MC38-PRL-3 or CT26 cells at a dose of 0.3 mg/kg every two days. After three consecutive injections, the mice were anesthetized with 2% isoflurane and the abdominal cavity was opened to expose the spleen. Subsequently, 1 × 10⁶ MC38-Luc cells suspended in 50 µL of matrix gel were injected into the spleen and the abdominal cavity was closed with surgical grapes. Subsequently, CXCL12 monoclonal antibody was intraperitoneally injected at a dose of 10 µg per mouse every two days. After consecutive injection for 14 days, the mice were viewed using an IVIS Lumina III imaging system to detect the liver metastasis. Similarly, the mice were sacrificed after imaging and the livers were collected for histopathological examination of liver metastasis and flow cytometry analysis of the hepatic infiltration of MDSCs and CD8⁺ T cells.

Statistical analysis

The in vitro data were presented as the mean ± S.D of three independent experiments. All statistical analyses were performed using SPSS 16.0 statistical software package and Graphpad Prism 9.0. Unpaired two-sided Student’s t-test and one-way ANOVA were used to compare the cell and animal experiments with different treatments, and post hoc tests were used to analyze difference between groups. Spearman’s rank correlation analysis was used to analyze the correlation between PRL-3 and the clinical characteristics of patients. Kaplan-Meier curve and log-rank test were used to compare overall survival (OS) and disease-free survival (DFS) in different patient groups. A Cox proportional hazards regression model was used to evaluate the correlations between various clinical features of patients with their prognosis. In all cases, *P < 0.05, **P < 0.01, and ***P < 0.001.

Results

High PRL-3 expression promotes the liver PMN formation and CLM via enhancing the recruitment of MDSCs

To evaluate the ability of PRL-3 to regulate the liver PMN formation and CLM, we up-regulated PRL-3 expression in murine CRC cell line MC38 (i.e., MC38-PRL-3, Fig. 2A, Fig. S1A) and injected into the cecum of C57BL/6 mice (Fig. 2B). Tumor growth was monitored using IVIS imaging system and the cecum, liver, and lung of mice were respectively harvested to assess the metastatic burden at 2, 3, and 5 weeks post the cell injection. As shown in Fig. 2C-E, two mice could develop liver metastasis at 5 weeks post the injection of MC38-PRL-3 cells. We further used qRT-PCR and flow cytometry to examine the expression of metastasis-associated inflammatory factors and the infiltration of MDSCs into the liver. As displayed in Fig. 2F, the expression of metastasis-associated inflammatory factors including S100A8/A9, MMP9, and Bv8 dramatically increases in the liver at 2 and 3 weeks post the injection of MC38-PRL-3 cells. Moreover, the results of flow cytometry analysis show that the infiltration of MDSCs into the liver is significantly improved at 3 weeks post the cell injection (Fig. 2G). All these results clearly indicate that high PRL-3 expression in CRC cells could indeed promote the liver PMN formation by enhancing the infiltration of MDSCs and the time frame of liver PMN formation is appropriately 3 weeks post injection of MC38-PRL-3 cells in the ceca.

Fig. 2.

High PRL-3 expression promotes the liver PMN formation and CLM via enhancing the recruitment of MDSCs. (A) Western blot analysis of PRL-3 expression in MC38-PRL-3 and MC38-VT cells. (B) Schematic illustration of the establishment of orthotopic CRC tumor via the cecal injection of MC38-PRL-3 or MC38-VT cells. (C) IVIS images of orthotopic CRC tumor-bearing mice and the photograph of their liver, lung, and colon at 2, 3, and 5 weeks post the cecal injection of MC38-PRL-3 or MC38-VT cells. Red arrows indicate the orthotopic tumors. (D) HE staining of the liver and lung and IHC staining of PRL-3 expression in the colon collected from the orthotopic CRC tumor-bearing mice shown in (C). Round circles indicate the tumor tissues in the colon and liver tissue. (E) Incidence of liver and lung metastasis in the orthotopic CRC tumor-bearing mice based on the results shown in (C) and (D). (F) qRT-PCR analysis of the expression of inflammatory factors in the liver of orthotopic CRC tumor-bearing mice shown in (C). (G) Flow cytometry analysis of CD45⁺CD11b⁺Gr-1⁺ cells (i.e., MDSCs) and the statistical results of the proportion of MDSCs in CD45⁺ cells in the liver of orthotopic CRC tumor-bearing mice shown in (C). (H) Schematic illustration of the evaluation of liver PMN formation and CLM in a mouse model educated with the exosomes derived from MC38-PRL-3 or MC38-VT cells followed by splenic injection of MC38-Luc cells. (I) Flow cytometry analysis of the infiltration of MDSCs and CD8⁺ T cells into the liver of mice (n = 6) educated with the exosomes derived from MC38-PRL-3 or MC38-VT cells for 14 days. (J) IVIS images of tumor-bearing mice and the photograph of their liver and spleen collected from the mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells. Red arrows indicate the metastatic tumors in the liver. (K) HE staining of the liver collected from the mice (n = 6) shown in (J) and the statistical results of CLM. Round circles indicate the metastasis tumor in the liver. (L, M) IF analysis of metastatic tumors, MDSCs, F4/80⁺ macrophages, and CD8⁺ T cells in the liver collected from the mice (n = 6) shown in (J). *P < 0.05, **P < 0.01

After validating the important role in promoting the liver PMN formation, we next investigate how PRL-3 exerts this biological function. It is known that tumor-derived exosomes are widely involved in the inter-microenvironment communication, including the regulation of PMN formation in target organs [15]. To examine whether the PRL-3-induced liver PMN formation is mediated by exosomes derived from CRC cells, we pre-educated C57BL/6 mice with the exosomes derived from MC38-PRL-3 cells (Fig. S1B and 1 C) for 2 weeks and then intrasplenically injected MC38-Luc cells to observe the liver PMN formation at 3 weeks post the cell injection (Fig. 2H). It could be found that the exosome pre-education could significantly enhance the hepatic recruitment and infiltration of MDSCs and reduce the hepatic infiltration of CD8⁺ T cells (Fig. 2I). However, there is no obvious change in the infiltration of natural killer (NK) cells and regulatory T cells (Tregs) into the liver (Fig. S2). These results highlight the importance of CRC cell-derived exosomes on the liver PMN formation by enhancing the hepatic infiltration of MDSCs and inhibiting the hepatic infiltration of CD8⁺ T cells. To further prove this conclusion, we isolated the exosomes derived from CT26 cells, a murine CRC cell line with high PRL-3 expression (Fig. S3A-D), and examined the influence of these exosomes on the liver PMN formation. Similarly, these exosomes could also dramatically enhance the recruitment and infiltration of MDSCs into the liver (Fig. S3E) and inhibit the hepatic infiltration of CD8⁺ T cells (Fig. S3F). Due to this significant contribution to the liver PMN formation, higher level of liver metastasis of MC38-Luc cells could be observed in the mice pre-educated with the exosomes derived from MC38-PRL-3 or CT26 cells, as demonstrated by stronger liver metastatic burden (Fig. 2J and K, Fig. S3G and 3 H), more hepatic infiltration of MDSCs (Fig. 2L, Fig. S3I and 3 K), and less hepatic infiltration of CD8⁺ T cells (Fig. 2M, Fig. S3J and 3 L). Taken together, all these results clearly reveal that CRC cells with high PRL-3 expression could secrete specific exosomes to enhance the liver PMN formation and promote CLM.

PRL-3 up-regulates ITGαvβ5 expression in CRC-derived exosomes to promote the liver PMN formation and CLM

Having confirmed the enhanced liver PMN formation by exosomes derived from CRC cells with high PRL-3 expression, we next investigated the intrinsic regulatory mechanism. To this end, we collected the surgically resected tumor samples and peripheral blood from 6 clone cancer patients, among which 3 patients developed postoperative liver metastasis. PRL-3 expression in tumor tissues of patients and isolated exosomes from their peripheral serum was examined by immunohistochemistry (IHC) and proteomic analysis (Fig. 3A, Fig. S4A and 4B), respectively. It could be found that the patients with liver metastasis show much higher PRL-3 expression in their tumor tissues compared to the patients without metastasis (Fig. 3B). The results of proteomic analysis show that there are 2678 differentially expressed proteins in the isolated exosomes from the peripheral serum in these two group patients (Fig. 3C). GeneSet Enrichment Analysis (GSEA) of these differentially expressed proteins indicates that biological process involved in symbiotic interaction and cell-matrix adhesion are positively regulated in the exosomes of patients with liver metastasis (Fig. 3D). Moreover, these two positively regulated pathways share 11 upregulated proteins (i.e., SERPINB9, ITGβ5, CFHR5, ANPEP, TRIM62, GRK2, VPS18, CXCL6, SNW1, PPID, and ZC3H12A), among which ITGβ5 is particularly noted (Fig. 3E) as it plays a crucial role in promoting tumor progression by regulating tumor cell adhesion to the matrix, remodeling extracellular matrix, and improving angiogenesis [17]. More importantly, recent researches have demonstrated the significant contribution of ITGβ5 in tumor-derived exosomes to the liver metastasis of various cancer types (e.g., colon and pancreatic cancer) via the formation of functional dimer ITGαvβ5 with ITGαv [18]. Therefore, we speculate that PRL-3 may up-regulate ITGαvβ5 expression in the exosomes derived from CRC cells to enhance the liver PMN formation and CLM. To validate this speculation, we examined ITGαvβ5 expression in the exosomes isolated from the peripheral serum of healthy volunteers (n = 21) and colon cancer patients (n = 47). As expected, colon cancer patients show much higher expression of ITGαvβ5 (Fig. 3F and G) in their serum exosomes compared to the healthy volunteers. We also examined ITGαvβ5 expression in the serum exosomes of colon cancer patients with different PRL-3 expression levels and the results indicate the positive correlation of PRL-3 expression with ITGαvβ5 (Fig. 3H and I, Fig. S4C and 4D). In addition, significantly increased ITGαvβ5 expression could be also observed in the serum exosomes of IV stage colon cancer patients with liver metastasis compared to the I-III stage patients (Fig. 3J and K). Further receiver operating characteristic (ROC) curve analysis also indicates the good specificity and sensitivity of ITGαvβ5 in the serum exosomes of colon cancer patients to CLM (Fig. 3L and M). Moreover, the results of univariate and multivariate analysis show that the serum exosomal ITGαvβ5 could be an independent prognostic risk factor for liver metastasis of CRC patients (Table S3), implying that the serum exosomal ITGαvβ5 could be used as a prognostic markers for CRC patients.

Fig. 3.

High PRL-3 expression is positively correlated with ITGαvβ5 expression in the serum exosomes and liver metastasis of CRC patients. (A) Schematic illustration of the detection of PRL-3 expression in tumor tissues and serum exosomes of colon cancer patients. (B) IHC analysis of PRL-3 expression in the tumor tissues of colon cancer patients (n = 6). (C) Heatmap of differentially expressed proteins in the serum exosomes isolated from colon cancer patients with high (n = 3) and low (n = 3) PRL-3 expression. (D) GSEA analysis of the differentially expressed proteins showing the positive regulation of biological processes associated with system interconnections and cell-matrix adhesion in the serum exosomes isolated from colon cancer patients with high PRL-3 expression. (E) Heatmap of the same differentially expressed proteins including ITGβ5 enriched in the biological processes associated with system interconnections and cell-matrix adhesion. (F, G) Level of ITGβ5 (F) and ITGαv (G) in the serum exosomes isolated from healthy volunteers (n = 21) and colon cancer patients (n = 47). (H, I) Level of ITGβ5 (H) and ITGαv (I) in the serum exosomes isolated from colon cancer patients with high (n = 27) and low (n = 20) PRL-3 expression. (J, K) Level of ITGβ5 (J) and ITGαv (K) in the serum exosomes isolated from colon cancer patients at stage I-III (n = 34) and IV with CLM (n = 13). (L, M) ROC curve showing the sensitivity and specificity of ITGβ5 (L) and ITGαv (M) to CLM of colon cancer patients (n = 47). *P < 0.05; **P < 0.01; ***P < 0.001

To further prove that PRL-3 promotes the liver PMN formation and CLM by up-regulating the exosomal ITGαvβ5 expression, we first examined the expression PRL-3 and ITGαvβ5 in various CRC cell lines and demonstrated the positive correlation between PRL-3 and ITGαvβ5 expression (Fig. S5A, 5B, S6A, and 6B). Based on this result, we further isolated the exosomes derived from MC38 and CT 26 cells and then examined their ITGαvβ5 expression. It could be found that up-regulating PRL-3 expression in MC38 cells could dramatically enhance ITGαvβ5 expression in their secreted exosomes (Fig. 4A, Fig. S7A). However, silencing PRL-3 expression in CT26 cells could significantly suppress ITGαvβ5 expression in their secreted exosomes (Fig. 4B, Fig. S7B). Inspired by these encouraging results, we next pre-educated mice with the exosomes derived from CT26 cells with stable ITGβ5 silencing by shRNA (denoted CT26-shITGB5 Exo) for 2 weeks and then intrasplenically injected MC38-Luc cells to evaluate the influence of these exosomes on the liver PMN formation (Fig. 4C, Fig. S7C). As shown in Fig. 4D and E, the impaired hepatic infiltration of MDSCs but enhanced hepatic infiltration of CD8⁺ T cells clearly indicates the suppressed liver PMN formation. More importantly, the liver metastasis ability of MC38-Luc cells is significantly impaired in the mice pre-educated with the CT26-shITGB5 Exo, as demonstrated by the weak liver metastatic burden (Fig. 4F) and less metastatic nodes in the liver (Fig. 4G). In addition, the results of flow cytometry (Fig. 4H and I) and IF staining (Fig. 4J and K) also indicate the less infiltration of MDSCs and more infiltration of CD8⁺ T cells into the liver metastatic tumor tissues of mice pre-educated with the CT26-shITGB5 Exo. These results demonstrate that PRL-3 could up-regulate ITGαvβ5 expression in CRC cells, which could be secreted into their exosomes to promote the liver PMN formation and CLM by enhancing the hepatic infiltration of MDSCs and impairing the hepatic infiltration of CD8⁺ T cells.

Fig. 4.

PRL-3 up-regulates ITGαvβ5 expression in CRC-derived exosomes to promote the liver PMN formation and CLM. (A, B) Western blot analysis of ITGαvβ5 expression in MC38-PRL-3 cells and CT26 cells with PRL-3 silencing and ITGαvβ5 expression in their secreted exosomes. (C) Schematic illustration of the evaluation of liver PMN formation and CLM in a mouse model educated with the exosomes derived from CT26 cells with ITGβ5 silencing followed by splenic injection of MC38-Luc cells. (D, E) Flow cytometry analysis of the infiltration of MDSCs (D) and CD8⁺ T cells (E) into the liver of mice (n = 6) educated with the exosomes derived from CT26 cells with ITGβ5 silencing for 14 days. (F) IVIS images of the tumor-bearing mice and photograph of the liver and spleen collected from the mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells. (G) HE staining of the liver collected from the mice (n = 6) shown in (F). Round circles indicate the metastasis tumor in the liver. (H, I) Flow cytometry analysis of the infiltration of MDSCs (H) and CD8⁺ T cells (I) into the liver of mice (n = 6) shown in (F). (J, K) IF analysis of metastatic tumors, MDSCs, F4/80⁺ macrophages, and CD8⁺ T cells in the liver of mice (n = 6) shown in (F). *P < 0.05; **P < 0.01; ***P < 0.001

PRL-3 up-regulates ITGαvβ5 expression by activating the Src/STAT3 pathway

After understanding the role of PRL-3 in promoting CLM via up-regulating ITGαvβ5 expression in CRC cells and their secreted exosomes, we next investigate the inherent mechanism how PRL-3 regulates ITGαvβ5 expression. To this end, we employed protein-protein interaction (PPI) analysis (https://cn.string-db.org/) to predict the possible mechanism, and the results show that PRL-3 may regulate ITGαvβ5 expression by activating the classic Src/STAT3 signaling pathway (Fig. 5A). To validate this prediction, we examined the activity of Src/STAT3 signaling pathway in various CRC cell lines (e.g., CT26, MC38, and HCT116). Indeed, high ITGαvβ5 expression and elevated phosphorylation level of Src and STAT3 could be observed in CRC cells with high PRL-3 expression (Fig. 5B-D, Fig. S8A-C). Based on this result, we further treated CT26 and MC38-PRL-3 cells with Src inhibitor PP2 and then examined ITGαvβ5 expression. As expected, PP2 treatment could significantly suppress ITGαvβ5 expression (Fig. 5E and F, Fig. S8D and 8E). The similar tendency could be also observed in the cells treated with STAT3 inhibitor cryptotanshinone (Fig. 5G, Fig. S8F) or STAT3 siRNA (Fig. S5I and 6 J). Moreover, the treatment with PP2, cryptotanshinone, or STAT3 siRNA could also down-regulate PRL-3 expression in CRC cells. This result is consistent with the previous report that PRL-3 and Src/STAT3 pathway could form positive feedback loop to regulate the downstream ITGαvβ5 expression in tumor cells [7].

Fig. 5.

PRL-3 up-regulates ITGαvβ5 expression by activating the Src/STAT3 pathway. (A) STRING analysis of protein-protein interaction showing the possible regulation of ITGαvβ5 by PRL-3 through the Src/STAT3 signaling pathway. (B) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in MC38-PRL-3 cells. (C, D) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in CT26 (C) and HCT116 (D) cells with PRL-3 silencing. (E, F) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in CT26 (E) and MC38-PRL-3 (F) cells treated with the Src inhibitor PP2. (G) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in CT26 cells treated with the STAT3 inhibitor cryptotanshinone. (H) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in CT26 cells with PRL-3 gene knockout by CRISPR-Cas9. (I) Western blot analysis of ITGαvβ5 expression and the activity of Src/STAT3 signaling pathway in CT26 cells with PRL-3 gene knockout followed by rescue with the plasmid expressing PRL-3

It is known that the enzyme activity of PRL-3 mainly depends on the Cys104 conserved catalytic residue in its PTP domain. Therefore, we constructed a mutation site at Cys104 and found that PRL-3^C104S mutant could dramatically block the activity of Src/STAT3 pathway and thus inhibit ITGαvβ5 expression (Fig. S5D and 6D). Furthermore, we also knocked out PRL-3 expression in CT26 cells and then conducted a rescue experiment with the wild type PRL-3 or PRL-3^C104S mutant type. It could be found that the wild type PRL-3 could enhance the phosphorylation of Src and STAT3 and up-regulate ITGαvβ5 expression (Fig. 5H and I, Fig. S8G and 8 H), indicating that the enzyme activity of PRL-3 is mainly based on the Cys104 conserved catalytic residue. The similar tendency could be also found in the exosomes isolated from CT26 and other CRC cells with high PRL-3 expression, in which PRL-3 could up-regulate ITGαvβ5 expression in the exosomes by activating the Src/STAT3 pathway in CRC cells (Fig. S5C-H and S6C-I) and this pathway activation mainly depends on the Cys104 conserved catalytic residue of PRL-3.

CRC-derived ITGαvβ5⁺ exosomes specifically target F4/80⁺ macrophages to enhance the recruitment of MDSCs and promote CLM

Having elucidated the regulation of ITGαvβ5 expression in both CRC cells and their secreted exosomes by activating the Src/STAT3 pathway, we next explore the mechanism by which these CRC-derived ITGαvβ5⁺ exosomes enhance the recruitment and infiltration of MDSCs into the liver. To this end, we performed the pre-education experiment by labeling the exosomes derived from MC38 and CT26 cells with the fluorescent dye PKH26 and then intravenously injecting into C57BL/6 mice (Fig. 6A). The fluorescent dye PKH26 is a classic membrane-labeling dye that can facilitate the monitoring of exosomes in the liver. As displayed in Fig. 6B, compared to hepatic stellate cells, F4/80⁺ macrophages in the liver show much higher internalization of the exosomes derived from CRC cells with high PRL-3 expression. If silencing PRL3 expression in CRC cells, the internalization of these exosomes by F4/80⁺ macrophages is significantly suppressed (Fig. S9A and 9B). With these results, we further investigate the inherent reason for the specific internalization of these CRC-derived ITGαvβ5⁺ exosomes by F4/80⁺ macrophages. As the ligand of ITGαvβ5, fibronectin has been reported to highly express in the liver, particularly on F4/80⁺ macrophages, which could enhance the accumulation of ITGαvβ5⁺ exosomes in the liver and promote their specific internalization by F4/80⁺ macrophages [18]. Therefore, we employed flow cytometry to examine fibronectin expression in the mouse liver. Indeed, F4/80⁺ macrophages show much stronger fibronectin expression than that of hepatic stellate cells and hepatocytes (Fig. S10). Due to this high fibronectin expression, F4/80⁺ macrophages present much stronger ability to internalize the exosomes derived from MC38-PRL-3 than that of hepatic stellate cell LX2 and hepatocyte LO2 (Fig. 6C). However, this high exosome internalization is dramatically impaired when pre-treating F4/80⁺ macrophages with fibronectin antibody (Fig. 6D), indicating that the specific recognition between the exosomal ITGαvβ5 and highly expressed fibronectin on F4/80⁺ macrophages is main driven force for the targeted internalization of CRC-derived ITGαvβ5⁺ exosomes. Based on this result, we next established F4/80⁺ macrophages and MDSCs using mouse bone marrow cells (Fig. S9C), and then investigated the influence of CRC-derived ITGαvβ5⁺ exosomes on the ability of macrophages to recruit MDSCs in a Transwell system (Fig. 6E). Because F4/80⁺ macrophages could specifically uptake the exosomes derived from MC38-PRL-3 and CT26 cells (Fig. 6F), the exosome treatment could dramatically enhance their ability to recruit MDSCs (Fig. 6G, Fig. S9D). However, there is no apparent influence on the ability of F4/80⁺ macrophages to recruit MDSCs when treated these macrophages with the exosomes derived from CT26 cells with PRL-3 or ITGβ5 silencing (Fig. 6G, Fig. S9D-F).

Fig. 6.

CRC-derived ITGαvβ5⁺ exosomes specifically target F4/80⁺ macrophages to enhance the recruitment of MDSCs and promote CLM. (A) Schematic illustration of labeling the exosomes derived from CT26 or MC38-PRL-3 cells with fluorescent dye PKH26 and then intravenously injecting into healthy mice (n = 6) every two days. (B) IF analysis of the localization of PKH26-labeled exosomes in the liver of mice (n = 6) at 7 days post the first injection of exosomes. Macrophages were stained with F4/80 (green fluorescence) while hepatic stellate cells were stained with α-SMA (pink fluorescence). (C) Fluorescence images of F4/80⁺ macrophages, LX2, and LO2 cells incubated with the PKH26-labeled exosomes derived from MC38-PRL-3 cells for 24 h. (D) Fluorescence images of F4/80⁺ macrophages treated with fibronectin antibody (2.5 ng/mL) for 2 h followed by incubated with the PKH26-labeled exosomes derived from MC38-PRL-3 cells for 24 h. (E) Schematic illustration of the differentiation of BMDMs into MDSCs and F4/80⁺ macrophages for examining the recruitment of MDSCs by F4/80⁺ macrophages in a Boyden Chamber device. (F) Fluorescence images of F4/80⁺ macrophages incubated with the PKH26-labeled exosomes derived from MC38-PRL-3 cells or CT26 cells with PRL-3 silencing for 24 h. (G) Migration of MDSCs co-cultured with F4/80⁺ macrophages treated with the exosomes derived from MC38-PRL-3 cells or CT26 cells with PRL-3 silencing for 24 h. (H) Schematic illustration of the evaluation of CLM in a mouse model with macrophage depletion by clodronate liposomes that educated with the exosomes derived from MC38-PRL-3 or CT26 cells followed by splenic injection of MC38-Luc cells. (I) IVIS images of the tumor-bearing mice and photograph of the liver and spleen collected from the mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells. Red arrows indicate the metastatic tumors in the liver. (J, K) Flow cytometry (J) and IF analysis (K) of the infiltration of MDSCs into the liver of mice (n = 6) shown in (I). (L, M) Flow cytometry (L) and IF analysis (M) of the infiltration of CD8⁺ T cells into the liver of mice (n = 6) shown in (I). *P < 0.05; **P < 0.01; ***P < 0.001

To further validate the pivotal role of F4/80⁺ macrophages in promoting the recruitment of MDSCs, we depleted F4/80⁺ macrophages from the mouse liver using clodronate liposomes, and then pre-educated the mice with the exosomes derived from CT26 cells followed by intrasplenical injection of MC38-Luc cells (Fig. 6H). As displayed in Fig. 6I, the macrophage depletion could dramatically impair the liver metastasis of mice pre-educated with the exosomes. The similar tendency could be also observed in the mice pre-educated with the exosomes derived from MC38-PRL-3 cells (Fig. S9G). In addition, the results of flow cytometry and IF analysis show an impaired hepatic infiltration of MDSCs (Fig. 6J and K, Fig. S9H and 9I) and an enhanced hepatic infiltration of CD8⁺ T cells (Fig. 6L and M, Fig. S9J and 9 K). All these results clearly reveal that F4/80⁺ macrophages in the liver are the main target cells of CRC-derived ITGαvβ5⁺ exosomes and they could enhance the ability of F4/80⁺ macrophages to recruit MDSCs and thus promote the liver PMN formation and CLM.

CRC-derived ITGαvβ5⁺ exosomes up-regulate CXCL12 in F4/80⁺ macrophages by activating the P38/STAT1 pathway

To explore the underlying mechanism by which the CRC-derived ITGαvβ5⁺ exosomes enhance the ability of F4/80⁺ macrophages to recruit MDSCs, we co-cultured F4/80⁺ macrophages with these exosomes and analyzed their differentially expressed genes (DEGs). Compared to the exosomes derived from MC38-VT cells, around 350 genes are upregulated in F4/80⁺ macrophages treated with the exosomes derived from MC38-PRL-3 cells (Fig. 7A and B). Among these DEGs, CXCL12 is particularly noted as it plays a crucial role in regulating the extracellular matrix and chemotaxis. Further detecting the supernatant of F4/80⁺ macrophages indicates that the CRC-derived ITGαvβ5⁺ exosomes could indeed enhance the secretion of CXCL12 from F4/80⁺ macrophages (Fig. 7C and D). Moreover, when neutralizing CXCL12 with monoclonal antibody, the ability of F4/80⁺ macrophages to recruit MDSCs is significantly impaired (Fig. 7E), implying that CXCL12 is the key factor enhancing the ability of F4/80⁺ macrophages to recruit MDSCs.

Fig. 7.

CRC-derived ITGαvβ5⁺ exosomes up-regulate CXCL12 in F4/80 + macrophages by activating the P38/STAT1 pathway. (A, B) Volcano plot (A) and heatmap (B) of DEGs in F4/80⁺ macrophages incubated with the exosomes derived from MC38 and MC38-PRL-3 cells for 24 h. (C, D) Level of CXCL12 in the supernatant of F4/80⁺ macrophages incubated with the exosomes derived from MC38-PRL-3 cells (C) and CT26 cells with PRL-3 silencing (D) for 24 h. (E) Migration of MDSCs co-cultured with F4/80⁺ macrophages incubated with the exosomes derived from CT26 or MC38-PRL-3 cells for 24 h followed by CXCL12 monoclonal antibody. (F) Western blot analysis of CXCL12 expression and the activity of P38/STAT1 signaling pathway in F4/80⁺ macrophages after 24 h incubation with the exosomes derived from MC38-PRL-3 cells or CT26 cells with PRL-3 silencing. (G) Migration of MDSCs co-cultured with F4/80⁺ macrophages incubated with the exosomes derived from MC38-PRL-3 cells for 24 h followed by P38 inhibitor SB203580 or STAT1 inhibitor NSC 118,218. (H) Western blot analysis of CXCL12 expression and the activity of P38/STAT1 signaling pathway in F4/80⁺ macrophages treated with the formula shown in (G). (I) Schematic illustration of the evaluation of CLM in a mouse model educated with the exosomes derived from MC38-PRL-3 cells followed by splenic injection of MC38-Luc cells and then intraperitoneal injection of CXCL12 monoclonal antibody. (J) IVIS images of tumor-bearing mice and photograph of the liver and spleen of mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells. (K, L) HE staining of the liver of mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells (K) and the statistical results of CLM (L). Round circles indicate the metastatic tumor in the liver. (M, N) Flow cytometry of the infiltration of MDSCs (M) and CD8⁺ T cells (N) into the liver of mice (n = 6) at 21 days post the splenic injection of MC38-Luc cells. *P < 0.05; **P < 0.01; ***P < 0.001

To investigate how the CRC-derived ITGαvβ5⁺ exosomes regulate the expression of CXCL12 in F4/80⁺ macrophages, Gene Ontology analysis was used to analyze the DEGs in the macrophages treated with the exosomes derived MC38-PRL-3 cells. The result shows that the DEGs could be particularly enriched in the PI3K-AKT and JAK-STAT signaling pathways (Fig. S11A). Previous studies have reported that the P38/STAT1 pathway is widely involved in activating the PI3K-AKT and JAK-STAT pathways [19, 20]. Therefore, we examined the activity of P38/STAT1 pathway in F4/80⁺ macrophages using the exosomes derived from MC38-PRL-3 cells. Indeed, these exosomes could activate the P38/STAT1 pathway in F4/80⁺ macrophages (Fig. 7F, Fig. S12A and 12B). Based on these findings, we examined CXCL12 expression in F4/80⁺ macrophages and evaluated their ability to recruit MDSCs after the treatment with the exosomes derived from MC38-PRL-3 cells followed by P38 or STAT1 inhibitor. It could be found that the addition of P38 or STAT1 inhibitor could dramatically impair the ability of F4/80⁺ macrophages to recruit MDSCs (Fig. 7G, Fig. S11B) and inhibit their CXCL12 expression (Fig. 7H, Fig. S11C, 12 C, and 12D). With this result, we further investigated the role of CXCL12 in promoting CLM in vivo. As shown in Fig. 7I-L, MC38-Luc cells show an impaired ability to metastasize in the liver of mice pre-educated with their secreted exosomes following by CXCL12 antibody treatment. Further flow cytometry analysis shows that CXCL12 antibody treatment could significantly impair the hepatic filtration of MDSCs and enhance the hepatic infiltration of CD8⁺ T cells (Fig. 7M and N). Moreover, because blocking the interaction between the exosomal ITGαvβ5 and fibronectin on F4/80⁺ macrophages could significantly suppress the exosome internalization by F4/80⁺ macrophages (Fig. 6D), concurrent blockage of CXCL12 and ITGαvβ5 with their antibodies could further synergistically suppress the liver metastasis of MC38-Luc cells (Fig. S13). We also examined the level of CXCL12 in peripheral blood of mice in cecum orthotopic tumor model or exosome pre-education model, and the results indicate that the level of CXCL12 is dramatically improved in the peripheral blood of mice pre-educated with the exosomes derived from MC38-PRL-3 or CT26 cells (Fig. S11D-F). If silencing ITGβ5 expression in these CRC cells or depleting F4/80⁺ macrophages from the mice, the level of CXCL12 decreases significantly in the peripheral blood (Fig. S11G-I). Base on the above results, it could be concluded that the CRC-derived ITGαvβ5⁺ exosomes could up-regulate CXCL12 expression in F4/80⁺ macrophages in the liver, which could be secreted into the liver microenvironment to enhance the hepatic infiltration of MDSCs and impair the hepatic infiltration of CD8⁺ T cells to promote CLM.

High PRL-3 and ITGαvβ5 expression is associated with the liver metastasis and poor prognosis of CRC patients

We finally evaluated the influence of PRL-3 and ITGαvβ5 expression on the liver metastasis and prognosis of CRC patients. Both the NCBI database and our clinical data indicate that ITGαvβ5 is highly expressed in tumor tissues of CRC patients compared to normal colon tissues and liver tissues (Fig. S14A-C). We further examined the expression of PRL-3 and ITGαvβ5 in the tumor tissues of colon patients at different stages (n = 338). As displayed in Fig. 8A, compared to the patients at stage I and II, the patients at stage III and IV show much higher expression of PRL-3 and ITGαvβ5 in their tumor tissues. Moreover, the expression of PRL-3 and ITGαvβ5 is further enhanced in liver metastatic tumors (Fig. 8B-D). In addition, Spearman’s rank correlation analysis reveals a positive correlation between PRL-3 and ITGαvβ5 expression (Fig. 8E and F).

Fig. 8.

High PRL-3 and ITGαvβ5 expression is associated with the liver metastasis and poor prognosis of CRC patients. (A-D) IHC staining (A) and immunoreactive score of PRL-3 and ITGαvβ5 (B-D) in the tumor tissues of colon cancer patients (n = 338). (E, F) Spearman’s rank analysis of the correlation between PRL-3 and ITGαvβ5 expression shown in (B-D). (G, H) ROC curve analysis of the specificity and sensitivity of PRL-3 and ITGαvβ5 to CLM (G) and TNM stage (H) of colon cancer patients (n = 338). (I, J) OS (I) and DFS (J) of colon cancer patients with high (n = 182) and low (n = 156) ITGβ5 expression in their tumor tissues. ***P < 0.001

In terms of survival prognosis, a noteworthy finding from the ROC curve analysis indicates that both PRL-3 and ITGαvβ5 show a high specificity and sensitivity for TNM stage and liver metastasis (Fig. 8G and H), but a moderate sensitivity for the differentiation and lymph node metastasis (Fig. S14D and 14E). We further divided the colon patients into two groups based on their PRL-3 and ITGαvβ5 expression levels, and then analyzed their clinical features and prognosis. It can be found that the expression of PRL-3 and ITGαvβ5 in both groups is associated with vascular invasion, pTNM stage, and liver metastasis of patients (Table S4). In addition, the patients with low expression of PRL-3 and ITGαvβ5 show much better OS (Fig. 8I, Fig. S14F and 14G) and DFS (Fig. 8J, Fig. S14H and14I). The result of univariate regression analysis also reveals that both PRL-3 and ITGαvβ5 are the risk factors for postoperative OS and DFS of CRC patients (Table S5). In multivariate regression analysis, PRL-3 expression could be an independent risk factor for postoperative OS while ITGαvβ5 could be an independent risk factor for postoperative DFS of CRC patients (Table S6). Taken together, high PRL-3 and ITGαvβ5 expression in the tumor tissues of CRC patients is closely associated with inferior long-term survival outcomes and heightened risk of liver metastasis.

Discussion

CRC recurrence and the occurrence of metastasis, primarily targeting the liver, are pivotal determinants of prognosis and are a leading cause of mortality [2]. The complex interplay between cancer cells and the liver microenvironment facilitates the establishment of liver PMN, essential for tumor colonization and progression. Understanding the molecular mechanisms that underpin the formation of PMN is crucial for developing effective therapeutic strategies to combat metastatic disease. It has been found in breast cancer, pancreatic cancer and CRC that the regulation of PMN in target organs promotes tumor metastasis [21–23]. Studies have shown that bone marrow-derived progenitor cells are the main regulatory cells promoting the PMN formation, and MDSCs are the main components, which are mainly related to the inhibition of tumor immunity and the promotion of microenvironment angiogenesis [24, 25]. PRL-3 is a member of the tyrosine phosphatase protein family. A mutation at position 104 (C104S) inhibits tumor invasion [26]. Our previous research has demonstrated that PRL-3 can regulate the invasion and liver metastasis of colon cancer by modulating the colon cancer microenvironment [12, 27]. In this study, our findings suggest that PRL-3 enhances the production of ITGαvβ5 in colon cancer cell-derived exosomes, thereby mediated the impact of colon cancer on hepatic MDSCs recruitment, facilitating PMN formation and promoting liver metastasis. These insights provide a foundational understanding of how PRL-3 mediates immune evasion in metastatic colon cancer and suggest novel avenues for targeted therapy.

PMN in the target organ plays a pivotal role in subsequent tumor cell colonization and proliferation, mainly due to its effects on microenvironment remodeling, such as angiogenesis, matrix remodeling, metabolic changes, and alterations in the immune microenvironment [15, 28]. PMN formation appears to result from the combined actions of tumor-derived secreted factors (TDSFs), exosomes, and bone marrow-derived cells (BMDCs) [4], with MDSCs, a subgroup of BMDCs, playing a significant role in this process [24, 29]. Therefore, investigating the mechanisms by which MDSCs are recruited may elucidate the process of PMN formation. Primary tumor-derived exosomes are implicated in the regulation of PMN formation [4]. For example, exosomes containing ITGBL1 secreted by CRC cells activated hepatic stellate cells to secrete IL-6 and IL-8, which promoted PMN formation [30]. In this study, we employed the exosome pre-education and liver metastasis model in C57BL/6 mice to reveal that the exosomes derived from CRC cells with high PRL-3 expression could enhance the infiltration of MDSCs and impair the infiltration of CD8⁺ T cells in the liver tissues, thus inducing liver PMN formation and promoting CLM. We analyzed the expression of PRL-3 in colon cancer tissues and the proteomics of peripheral serum exosomes by selecting patients who received radical surgery. Through postoperative survival follow-up, it was found that only patients with high PRL-3 levels developed liver metastasis. At the same time, peripheral serum exosome proteomic results showed that a variety of exosome proteins were involved in the regulation of microenvironment communication and cell matrix adhesion, one of which was ITGβ5, which only formed a dimer ITGαvβ5 with ITGαv and found to specifically regulate tumor liver metastasis.

Previous studies have revealed that exosomes carrying different types of integrins participate in organ-specific metastasis, including ITGα6β1 in the lungs and ITGαvβ5 in the liver [18]. ITGαvβ5 has been found in previous studies to play the role of adhesion receptors, activate tumor signaling pathways, and participate in tumor invasion and metastasis [31, 32]. At the same time, it was found that ITGαvβ5 in tumor cell-derived exosomes can reshape tumor microenvironment and promote tumor progression [33]. In this study, we showed that PRL-3 upregulated ITGαvβ5 levels in colon cancer cell-derived exosomes, recruiting MDSCs to the liver, which suppressed CD8⁺ T-cell infiltration, thereby promoting PMN formation in the liver and CLM. On the mechanism of ITGαvβ5 regulation, using protein-PPI predictions, we demonstrated that PRL-3 controls ITGαvβ5 expression in CRC cells by activating the Src/STAT3 pathway. At the same time, through inhibitors of Src/STAT3 pathway, CRISPR-Cas9-mediaed knockout and rescue experiments, we found that PRL-3 could activate the Src/STAT3 pathway through its 104th Cys residue and form a feedback loop with it, participating in the regulation of ITGαvβ5 expression. The exploration of other signaling pathways that interact with the Src/STAT3 axis could yield further insights into the complex regulatory networks that govern cancer cell behavior and immune interactions in the tumor microenvironment [34]. However, the underlying mechanism mediating the regulation of exosome production still needs further research in the future.

Although we have found that ITGαvβ5 is involved in regulating the liver PMN formation, the mechanism by which exosomal ITGαvβ5 regulates the infiltration of MDSCs in target organs is still unclear. As the largest digestive organ, the liver is rich in mesenchymal cells, including F4/80⁺ macrophages (including Kupffer cells), stellate cells, etc. Previous studies have reported the role of kupffer cells and hepatic stellate cells in regulating the formation of PMN [16, 35]. Kupffer cells, the primary resident stromal cells in the liver, are also known as hepatic or F4/80⁺ macrophages. They play a dual role in the microenvironment, exerting antitumor activity on the one hand and facilitating tumor metastasis. In the former, F4/80⁺ macrophages participate in phagocytosis and release oxygen metabolic products [36]. In the latter, they induce the expression of cell adhesion molecules in liver sinusoidal endothelial cells, facilitating the attachment of disseminated tumor cells to the liver [37]. Research has shown that F4/80⁺ macrophages take up exosome ITGαvβ5, leading to the secretion of the inflammatory factor S100 [18]. However, it is unclear whether this process regulates the recruitment of hepatic MDSC. In this study, it was found that F4/80⁺ macrophages in liver were the main target cells, rather than hepatic stellate cells for uptake of exosomes through in vivo tracing analysis of exosomes. We subsequently confirmed that F4/80⁺ macrophages recruit MDSCs significantly after taking up exosome ITGαvβ5. This occurs through the activation of the P38/STAT1 pathway and an increase in CXCL12 chemokine expression, which promotes MDSC migration. The current understanding of the mechanism underlying the regulation of MDSC infiltration is that chemokines CCL2, CXCL5, and CXCL12 act as primary inducers [38]. The significance of CXCL12 in promoting MDSC recruitment and modulating T-cell activity further suggests that targeting this pathway could enhance anti-tumor immunity and improve therapeutic outcomes in patients with CLM. Furthermore, PRL-3 and ITGαvβ5 high expression correlated with worse OS and DFS outcomes, and exosomal ITGαvβ5 in peripheral serum exhibited good specificity and sensitivity for the detection of liver metastasis of colon cancer. These suggest that PRL-3 and ITGαvβ5 may serve as a biomarker for assessing the risk of metastasis in colon cancer patients, which can guide treatment decisions and patient management.

In this study, further studies are needed on how PRL-3 is involved in regulating the production of exosomal ITGαvβ5, so as to further discover the specific mechanism by which PRL-3 could regulate the microenvironment of colon cancer. At the same time, in terms of clinical relevance, it is necessary to prospectively study the regulation of exosomal ITGαvβ5 levels in peripheral blood of patients with colon cancer on liver PMN formation and the sensitivity and specificity of liver metastasis judgment. Future studies will focus on the regulatory mechanisms of MDSCs on the formation of PMN, including immunosuppression, neutrophil extracellular trap formation, and ECM remodeling.

Conclusions

In summary, we have uncovered the important function of PRL-3 in promoting the liver PMN formation and CLM. Molecular mechanism study reveals that high PRL-3 expression could activate the Src/STAT3 pathway in CRC cells and thus up-regulate the expression of ITGαvβ5 in their secreted exosomes, which could specifically target F4/80⁺ macrophages in the liver to activate their P38/STAT1 signaling pathway. Due to this activation of P38/STAT1 pathway, the expression and secretion of CXCL12 is significantly improved, which could enhance the recruitment of MDSCs into the liver and impair the hepatic infiltration of CD8⁺ T cells, ultimately inducing the liver PMN formation and CLM. The findings in this work provide a solid evidence for the prevention and treatment of CLM by targeting PRL-3.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Q.L., H.X., Y.Z X., X.X., and Z.C. conceived the ideas and designed the experiments. Q.L., H.X., and Y.Z X. conducted the experiments, analyzed the data, and wrote the paper. L.L., Y.Q., and Y.Z. performed the experiments. H.X., L.L., and X.H. performed the experiments and analyzed the data of the revised manuscript. W.L. provided the technical and material support. L.L., W.L., and J.W. collected the clinical samples and information. Z.C., W.L., X.X., and J.H. reviewed and revised the manuscript. All the authors have read and approved the final manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (81871981, 82272650, 82272638), Natural Science Foundation of Guangdong Province (2023A1515010404, 2022A1515012348, 2023A1515010457), and Guangzhou Science and Technology Planning Project (202201010825).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All HCC tumor samples were collected with the informed consent of the patients in accordance with the International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS). The study was approved by the Institutional Review Board (IRB) of Sun Yat-Sen Memorial Hospital. All animal experiments were performed by a protocol approved by the Institutional Animal Care and Use Committee at Sun Yat-Sen Memorial Hospital.

Competing interests

The authors declare no competing interests.