Tumor-derived exosomal miR-425-5p and miR-135b-3p enhance colorectal cancer progression through immune suppression and vascular permeability promotion

Abstract

BACKGROUND

Colorectal cancer (CRC) is a leading cause of cancer-related morbidity and mortality globally. Exosomal microRNAs (miRNAs) are known to modulate tumor progression by influencing immune responses and vascular dynamics. However, the roles of specific exosomal miRNAs, such as miR-425-5p and miR-135b-3p, in CRC remain unclear.

AIM

To explore the specific roles and underlying mechanisms of exosomal miR-425-5p and miR-135b-3p in CRC progression.

METHODS

Differentially expressed miRNAs were identified through microarray analysis of exosomes isolated from CRC tissues and adjacent normal mucosa. Functional roles of miR-425-5p and miR-135b-3p were evaluated in vitro using macrophage polarization, T cell differentiation, and vascular permeability assays, as well as in vivo tumor formation and metastasis experiments in nude mice. Validation experiments were performed using CRC cell lines (HCT116 and SW620).

RESULTS

Exosomal miR-425-5p and miR-135b-3p were significantly upregulated in CRC compared to normal tissues. Functional studies revealed that miR-425-5p promotes macrophage M2-like polarization and suppresses T cell proinflammatory responses, while miR-135b-3p enhances vascular permeability and angiogenesis. Inhibition of these miRNAs in CRC cell-derived exosomes significantly suppressed tumor growth and metastasis in nude mice, reprogramming the tumor microenvironment toward reduced angiogenesis and enhanced immune activation. Combined inhibition of both miRNAs resulted in the most pronounced effects.

CONCLUSION

Exosomal miR-425-5p and miR-135b-3p drive CRC progression by promoting immune suppression and vascular permeability. Their inhibition offers a promising strategy for modulating the tumor microenvironment and limiting CRC metastasis.

Key Words: Colorectal cancer; Exosomes; MiR-425-5p; MiR-135b-3p; Immune modulation; Vascular permeability

Core Tip: This study identifies exosomal miR-425-5p and miR-135b-3p as critical regulators of colorectal cancer (CRC) progression. These microRNAs (miRNAs) are significantly upregulated in CRC-derived exosomes and modulate the tumor microenvironment through distinct mechanisms. MiR-425-5p promotes immune suppression by inducing macrophage M2 polarization and skewing T cell differentiation toward pro-tumor subsets, while miR-135b-3p enhances vascular permeability and angiogenesis. Inhibition of these miRNAs in preclinical models significantly reduces tumor growth and metastasis, suggesting they represent promising therapeutic targets for CRC. The findings provide novel insights into exosomal miRNA-mediated immune modulation and vascular remodeling in CRC progression.

INTRODUCTION

Colorectal cancer (CRC) is one of the most prevalent malignancies worldwide, with high rates of morbidity and mortality[1,2]. Despite advances in diagnosis and treatment, the mechanisms underlying CRC progression and metastasis remain incompletely understood[3,4]. Exosomes, small extracellular vesicles secreted by various cell types, have emerged as critical mediators of intercellular communication within the tumor microenvironment[5]. Notably, exosomal microRNAs (miRNAs) are increasingly recognized as pivotal regulators of cancer biology, influencing processes such as immune modulation, angiogenesis and metastasis[6-8]. Several studies have reported specific exosomal miRNAs linked to macrophage polarization and T cell differentiation to promote CRC progression, such as miR-934, miR-106b-5p, miR-203a-3p, miR-106a-5p, miR-1246 in macrophage polarization[9-13], and miR-149-3p, miR-208b, miR-34a in T cell differentiation[14-16]. However, the current understanding of exosomal miRNAs in CRC remains fragmented, particularly regarding their roles in immune regulation and vascular permeability.

Among these miRNAs, miR-425-5p and miR-135b-3p have been implicated in the progression of several cancers[17-22]. Previous studies have shown that miR-425-5p is involved in modulating immune responses and promoting tumor growth in hepatocellular carcinoma and pancreatic carcinoma[23,24]. In CRC, miR-425-5p is found upregulated in patients’ serum and tumor tissues[25-27], and is close related to liver metastasis via M2 macrophage polarization[28,29]. However, its precise mechanistic contribution to immune regulation in CRC remains unclear. Similarly, miR-135b-3p has been associated with enhanced angiogenesis and epithelial-mesenchymal transition in breast cancer and adenocarcinoma[19-21], but its role in CRC, particularly as components of tumor-derived exosomes, remains unexplored.

In this study, we systematically address these gaps by investigating the roles of exosomal miR-425-5p and miR-135b-3p in CRC progression. Through microarray analysis of tumor tissue exosomes, we identified these miRNAs as highly upregulated in CRC. Functional enrichment analyses linked miR-425-5p to immune cell infiltration and miR-135b-3p to vascular permeability, highlighting novel regulatory functions in CRC progression. To further explore these roles, we examined their effects on immune modulation, angiogenesis and tumor growth using both in vitro and in vivo models.

Our findings demonstrate that exosomal miR-425-5p promotes macrophage polarization toward an immunosuppressive phenotype and skews T cell differentiation toward pro-tumor subsets, while miR-135b-3p enhances vascular permeability and angiogenesis. Furthermore, in vivo inhibition of these miRNAs suppresses tumor growth and metastasis, reprogramming the tumor microenvironment to a less aggressive state. By systematically elucidating the roles of these two miRNAs in immune regulation and vascular permeability - two underexplored aspects of exosomal miRNA function in CRC - our study provides novel insights into the tumor microenvironment and suggests potential therapeutic strategies targeting exosomal miRNAs in CRC.

MATERIALS AND METHODS

Patient samples and cell lines

CRC and adjacent normal tissue specimens were obtained from patients who underwent surgical procedures at Zhongshan City People’s Hospital Cancer Center. Diagnoses were confirmed independently by at least two pathologists, adhering to the Chinese Joint Committee on Cancer guidelines. The study was approved by the Ethics Committee of Zhongshan City People’s Hospital (Approval No. 2025-002), with all participants providing informed consent.

Human CRC cell lines (HCT116, SW620, SW480, HT-29), normal colon epithelial cell line (FHC), macrophage cell line (THP-1), primary naive CD4⁺ T cells, and human umbilical vein endothelial cells (HUVECs) were obtained from American Type Culture Collection (Rockville, MD, United States). Authentication and culture were conducted in specific media supplemented with 100 g/L foetal bovine serum (FBS) (Gibco, NY, United States). HCT116 and HT-29 were cultured in McCoy’s 5A medium, SW620 and SW480 in Leibovitz L-15, FHC in DMEM: F-12, and THP-1 in RPMI-1640 (Gibco, NY, United States). Primary naive CD4⁺ T cells were maintained in RPMI-1640 with 100 g/L HAS (Med Chem Express, NJ, United States), while HUVECs were grown in vascular cell basal medium with an endothelial cell growth kit-BBE (American Type Culture Collection, Rockville, MD, United States). All cells were incubated at 37 °C in a humidified 50 mL/L CO₂ atmosphere.

Exosome isolation and characterization

Exosomes were isolated and characterized as previously described[30]. In brief, exosomes were extracted using an exo Easy Maxi Kit (Qiagen, Germany) according to the manufacturer’s instructions. Following isolation, exosomes were characterized using transmission electron microscopy to confirm morphology, nanoparticle tracking analysis for size distribution, and western blotting for exosomal markers, including tumor susceptibility gene 101 protein (TSG101) and CD81. The purity of exosome preparations was validated by assessing the absence of calnexin, an endoplasmic reticulum marker. The primary antibodies were anti-TSG101 (28283-1-AP), anti-CD81 (66866-1-Ig) and anti-calnexin (10427-2-AP) (all from Proteintech, IL, United States).

Microarray analysis

Total RNA from CRC patient-derived exosomes was extracted using the All-in-One microRNA extraction kit (GeneCopoeia, Rockville, MD, United States) per the manufacturer’s instructions. RNA concentration was determined using a Qubit 3.0 Fluorometer (Invitrogen, Carlsbad, CA, United States), and integrity was assessed via an Agilent 2100 Bioanalyzer (Applied Biosystems, Foster City, CA, United States).

Small RNA libraries were constructed using 100 ng total RNA input with the VAHTS Small RNA Library Prep Kit (Vazyme Biotech, Nanjing, China) compatible with Illumina platforms. The workflow involved sequential enzymatic processing, beginning with adapter ligation targeting small RNA species (miRNA, small interfering RNA, PIWI-interacting RNA). Reverse transcription was then performed to generate cDNA templates, followed by polymerase chain reaction (PCR) enrichment and purification of amplified libraries. Quality control was conducted via Agilent 2100 Bioanalyzer profiling using a DNA 1000 chip, with library quantification achieved through KAPA Biosystems’ library quantification assay (Kapa Biosystems, Wilmington, MA, United States).

Libraries were diluted, equimolarly pooled, and clustered before undergoing single-end sequencing (50 bp read length). Bioinformatics analysis, including differential expression and Kyoto Encyclopedia of Genes and Genomes enrichment analysis, was performed to identify significant molecular changes and related biological pathways.

Real-time quantitative PCR analysis

Total RNA from exosomes was isolated as described in the microarray analysis section, while total RNA from cells was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, United States) following the manufacturer’s protocol. cDNA synthesis was carried out with either the miRNA first-strand cDNA Synthesis Kit (Vazyme, Nanjing, China) or MonScriptRTIII Super Mix (Monad, China).

PCRs were performed on a CFX96 Touch^™ real-time PCR system (Bio-Rad, Hercules, CA, United States) using miRNA Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) or MonAmp ChemoHS qPCR Mix (Monad, China). U6 and GAPDH served as internal controls for exosomes and cells respectively. The relative gene abundance was calculated via the 2^-ΔΔCT formula, and the primer sequences are listed in Table 1.

Table 1 Primer sequence for real-time quantitative polymerase chain reaction.

Gene	Primer sequence (5’-3’)
U6	Forward: CTCGCTTCGGCAGCACA
	Reverse: AACGCTTCACGAATTTGCGT
MiR-425-5p	Forward: AAUGACACGAUCACUCCCGUUGA
	Reverse: AGTGCAGGGTCCGAGGTATT
MiR-135b-3p	Forward: AUGUAGGGCUAAAAGCCAUGGG
	Reverse: AGTGCAGGGTCCGAGGTATT
GAPDH	Forward: AATGACCCCTTCATTGAC
	Reverse: TCCACGACGTACTCAGCGC
IL-1β	Forward: AACCTCTTCGAGGCACAAGG
	Reverse: AGATTCGTAGCTGGATGCCG
IL-6	Forward: CCTTCGGTCCAGTTGCCTTCT
	Reverse: TCTGAGGTGCCCATGCTACA
TNF-α	Forward: CTGAGACTCTCCTGTGCAGC
	Reverse: ACTCCCTGAAGAGACGGTGA
IL-4	Forward: CTTTGCTGCCTCCAAGAACA
	Reverse: GTTCCTGTCGAGCCGTTTCA

IL-1β: Interleukin 1 beta; IL-6: Interleukin 6; TNF-α: Tumor necrotic factor alpha; IL-4: Interleukin 4.

Western blot analysis

Cells were lysed using RIPA buffer, and total protein concentration was measured via a BCA Protein Assay Kit (Pierce, Waltham, MA, United States). Equal amounts (20 μg) of protein were separated on a 120 g/L sodium-dodecyl sulfate gel electrophoresis gel (80 V, 40 minutes) and transferred to polyvinylidene fluoride membranes. Membranes were blocked with 100 g/L nonfat milk, incubated overnight at 4 °C with primary antibodies, and then with secondary antibodies for 1 hour at room temperature. Signals were detected using the iBright 1500 system (Thermo Fisher, Waltham, MA, United States), and band intensities were quantified with ImageJ. Antibodies: Anti-zonula occludens-1 (anti-ZO-1) (21773-1-AP), anti-occludin (27260-1-AP), anti-claudin-5 (29767-1-AP) (all from Proteintech, IL, United States).

In vitro functional assays

Transfection: MiR-425-5p andmiR-135b-3p inhibitors, along with negative control (NC) recombinant (Med Chem Express, NJ, United States), were transfected into CRC cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, United States). Cells were cultured to 85% confluence, and exosomes were isolated using exosome concentration solution at 4 °C. Exosomes were then incubated with THP-1 cells, primary naïve T cells, or HUVECs for further analysis.

Flowcytometry: THP-1 cells were stained with CD86-FITC, inducible nitric oxide synthase (iNOS)-Alexa Fluor 488, CD163-PE, and CD206-APC. CD4⁺ T cells were labeled with CD4-FITC, interferon gamma (IFN-γ)-PE, interleukin 4 (IL4)-PE, IL17A-PE, CD25-FITC, forkhead box protein P3 (Foxp3)-PE, and Zombie NIR^™ for live/dead discrimination. Tumor cells from mouse models were dissociated using Miltenyi Biotec’s dissociation kit and GentleMACS^™, then stained with CD4-FITC, IFNγ-PE, IL4-PE, IL17A-PE, CD25-APC, Foxp3-PE, and 7-AAD. Flowcytometry was performed on a CytoFLEX S (Beckman Coulter, Brea, CA, United States), with data analyzed via Kaluza software. Antibodies are listed in Table 2.

Table 2 Antibodies used in flow cytometry.

Antibody	Catalog number	Manufacture
THP-1 cells
Anti-CD86-FITC	557343	BD Biosciences, NJ, United States
Anti-iNOS	ab283655	Abcam, United Kingdom
Alexa Fluor® 488	ab236553	Abcam, United Kingdom
Anti-CD163-PE	560933	BD Biosciences, NJ, United States
Anti-CD206-APC	561763	BD Biosciences, NJ, United States
CD4+ T cells
Anti-CD4-FITC	561005	BD Biosciences, NJ, United States
Anti-IFN-γ-PE	561056	BD Biosciences, NJ, United States
Anti-IL4-PE	554485	BD Biosciences, NJ, United States
Anti-IL17A-PE	560438	BD Biosciences, NJ, United States
Anti-CD25-FITC	560990	BD Biosciences, NJ, United States
Anti-Foxp3-PE	560082	BD Biosciences, NJ, United States
Mouse tumor cells
Anti-CD4-FITC	553046	BD Biosciences, NJ, United States
Anti-IFNγ-PE	562020	BD Biosciences, NJ, United States
Anti-IL4-PE	562044	BD Biosciences, NJ, United States
Anti-IL17A-PE	561020	BD Biosciences, NJ, United States
Anti-CD25-APC	561048	BD Biosciences, NJ, United States
Anti-Foxp3-PE	563101	BD Biosciences, NJ, United States

CD86: T-lymphocyte activation antigen CD86; iNOS: Inducible nitric oxide synthase; CD163: Scavenger receptor cysteine-rich type 1 protein M130; CD206: Macrophage mannose receptor 1; CD4: T-cell surface glycoprotein CD4; IFN-γ: Interferon gamma; IL-4: Interleukin 4; CD25: Interleukin-2 receptor subunit alpha; FoxP3: Forkhead box protein P3.

Immunofluorescence staining: Cell slices or 4-μm tumor sections were blocked with 10 g/L BSA (Sigma - Aldrich, Saint Louis, MI, United States). They were then incubated overnight at 4 °C with primary antibodies: Anti-CD86 (13395-1-AP), anti-CD206 (18704-1-AP), anti-ZO-1 (21773-1-AP), anti-occludin (27260-1-AP), and anti-claudin-5 (29767-1-AP) (all from Proteintech, IL, United States). After washing with phosphate buffered saline, samples were incubated for 2 hours at room temperature with fluorescence-conjugated secondary antibodies (Abcam, United Kingdom). Nuclei were stained with DAPI (Beyotime, China), and images were taken using a ZEISS LSM 900 confocal microscope.

Cell-based assays

Tube formation assay: Matrigel (Corning, NY, United States) was polymerized in 24-well plates at 37 °C for 30 minutes. Treated HUVECs were seeded onto the matrigel and incubated at 37 °C in 50 mL/L CO₂. Tube formation was observed at 12 hours under a microscope and quantified by counting the tube structures.

Transwell migration assay: Exosome-treated HUVECs in serum-free medium were seeded into transwell inserts (8-μm pore, BD Biosciences, NJ, United States). The lower chamber contained medium with 100 g/L FBS. After 12 hours, migrated cells on the membrane’s lower surface were stained with hematoxylin and counted under a light microscope in four random fields (200 ×).

Permeability assay: Rhodamine B isothiocyanate-dextran with a molecular weight around 70000 (Sigma, MI, United States) was introduced to into the upper compartment of transwell filters (0.4-μm pore, BD Biosciences, NJ, United States) seeded with HUVECs (10⁵ cells/well). The HUVECs were co-cultured with CRC cells for 3 days. After incubation for 30 minutes, lower chamber fluorescence was quantified at 544 nm excitation/590 nm emission.

Aortic ring assay: Thoracic aortas from 8-12-week-old Sprague Dawley rats were sectioned (1 mm), placed on matrigel-coated wells, and incubated at 37 °C for 30 minutes. Vascular cell basal medium with 100 g/L FBS, vascular endothelial growth factor (VEGF) (10 ng/mL), and heparin (25 μg/mL) was added. After 24 hours, aortic rings were incubated with conditioned medium or exosomes from CRC cells transfected with miR-425-5p inhibitor or NC. Sprouts were quantified on day 5.

In vivo tumor formation assays

Female Balb/c-nude mice (4-6 weeks, Guangdong Zhiyuan Biomedical Technology Co., Ltd., China) were housed pathogen-free under a 12-hour light/dark cycle. Experimental procedures were approved by the Laboratory Animal Ethics Committee of Lai’an Technology (Approval No. G2024123).

Mice were subcutaneously injected with CRC cells (2 × 10⁶/mouse) transfected with miR-425-5p inhibitor, miR-135b-3p inhibitor, both inhibitors, or NC. After 42 days, mice were exposure to excess CO₂ to ensure unconsciousness, and tumor, liver, and lung tissues were collected immediately. Tumor volume, weight, and gross size were recorded, and metastatic nodules in the liver/lungs were counted.

Tumors were fixed informal in, embedded in paraffin, and sectioned (4 μm). Hematoxylin and eosin staining (Beyotime, China) was performed. For immunohistochemistry and immunofluorescence, primary antibodies included anti-CD34 (14486-1-AP), anti-CD206 (60143-1-Ig), and anti-iNOS (22226-1-AP) (all from Proteintech, IL, United States). Tumor-infiltrating T cells were analyzed as described. Cytokine levels [IL-1β, IL-6, tumor necrosis factor (TNF)-α] in tumor lysates were measured using QuantiCyto^® ELISA kits (EMC001b, EMC004 from NeoBioscience, D721217 from Sangon Biotech, China).

Statistical analysis

Data are presented as mean ± SD. Statistical analysis was performed using GraphPad Prism 8.2 (GraphPad Software, United States) with unpaired t test or one-way ANOVA, followed by Dunnett’s test for multiple comparisons. Statistical significance was considered as ^aP < 0.05, ^bP < 0.01, ^cP < 0.001, ^dP < 0.0001.

RESULTS

Exosomal miR-425-5p and miR-135b-3p are upregulated in CRC patients and cell lines

To explore the dysregulation of exosomal miRNAs, we extracted exosomes from tumor tissues and adjacent normal mucosa tissues of CRC patients, and then performed microarray (fold change > 1.5, false discovery rate < 0.01) (Figure 1A). Compared with normal tissues, there were 30 miRNAs with abnormal expression in CRC tumor tissue exosomes. Twenty of them were upregulated and 10 were downregulated (Figure 1B). We chose two upregulated miRNAs (miR-425-5p and miR-135b-3p) for further investigation. Real-time quantitative PCR analyses confirmed their up-regulation in exosomes from CRC cell-lines (SW480, HT-29, SW620, HCT116) compared to normal cells (FHC) (Figure 1C). Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses of miR-425-5p and miR-135b-3p focus on NOD-like receptor signaling pathway and VEGF signaling pathway respectively (Figure 1D). We suspect that abnormally expressed miR-425-5p may be associated with immunity and miR-135b-3p may be associated with vascular permeability.

Figure 1 Exosomal miR-425-5p and miR-135b-3p are upregulated in colorectal cancer patients and cell lines. A: Heatmap; B: Volcano plot of differentially expressed tissue exosomal microRNAs (miRNAs) between tumors tissues and normal adjacent tissues of colorectal cancer patients; C: Real-time quantitative polymerase chain reaction validation of miR-425-5p and miR-135b-3p from exosomes of colorectal cancer cell lines (SW480, HT-29, SW620, HCT116) and normal cells (FHC); D: Kyoto Encyclopedia of Genes and Genomes enrichment analyses of miR-425-5p and miR-135b-3p; E: Transmission electron microscopy images; F: Nanoparticle tracking analysis of exosomes isolated from HCT116 cells; G: Western blot analysis of exosomes isolated from HCT116 cells. ^aP < 0.05 vs FHC-exo (n = 3 per group). NC-exo: Negative control-exosome; TSG101: Tumor susceptibility gene 101 protein.

To further investigate the effects of exosomal miR-425-5p on tumor cell-induced immune responses and exosomal miR-135b-3p on tumor cell-induced vascular permeability, we extracted exosomes from the most upregulated HCT116 cells transfected with miR-425-5p inhibitor or miR-135b-3p inhibitor. Characterization of HCT116-derived exosomes were carried out by transmission electron microscopy (Figure 1E) and nanoparticle tracking analysis (Figure 1F). In addition, western blot analysis confirmed the expression of specific exosome markers (TSG101 and CD81) and the absence of the endoplasmic reticulum resident protein calnexin, which serves as a negative control for exosome purity (Figure 1G).

Exosomal miR-425-5p inhibition promotes macrophage M1-like polarization and proinflammatory T cell differentiation

To investigate the effects of exosomal miR-425-5p on immune cells, exosomes isolated from HTC116 cells transfected with miR-425-5p inhibitor or NC were cultured with human macrophage cell line THP-1. Flow cytometry analysis showed an increased proportion of CD86⁺iNOS⁺ M1 macrophages [CD86⁺: P < 0.001, 95% confidence interval (CI): 10.66-15.26; iNOS⁺: P < 0.001, 95%CI: 13.65-16.09] and a decreased proportion of CD163⁺CD206⁺ M2 macrophages (CD163⁺: P < 0.001, 95%CI: -83.48 to -72.77; iNOS⁺: P < 0.001, 95%CI: -15.40 to -8.995) when THP-1 cultured with miR-425-5p inhibitor exosomes compared to NC (Figure 2A). Immunofluorescent results confirmed the changes, as evidenced by upregulation of M1 marker CD86 and downregulation of M2 marker CD206 when miR-425-5p was depleted (CD86: P < 0.05, 95%CI: 4036-17670; CD206: P < 0.05, 95%CI: -10519 to -1032) (Figure 2B). These results indicate that exosomal miR-425-5p from CRC cells affects macrophages polarization and depletion of it promotes toward a M2-like phenotype.

Figure 2 Exosomal miR-425-5p inhibition promotes macrophage M1-like polarization and proinflammatory T cell differentiation. A: Flow cytometry analysis of CD86⁺iNOS⁺ and CD163⁺CD206⁺ cells ratio in THP-1 cells cultured with exosomes isolated from HTC116 cells transfected with negative control and miR-425-5p inhibitor; B: Immunofluorescent staining of CD86 (red) and CD206 (green) in THP-1 cells with DAPI nuclei (blue); C: Real-time quantitative polymerase chain reaction analysis of cytokine expression in CD4⁺ T cells cultured with exosomes; D and E: Flow cytometry analysis of CD4⁺ T cell subsets. ^aP < 0.05, ^cP < 0.001 vs negative control-exosome (n = 3 per group). NC-exo: Negative control-exosome; CD4: T-cell surface glycoprotein CD4; CD25: Interleukin-2 receptor subunit alpha; CD86: T-lymphocyte activation antigen CD86; CD163: Scavenger receptor cysteine-rich type 1 protein M130; CD206: Macrophage mannose receptor 1; FoxP3: Forkhead box protein P3; IFN-γ: Interferon gamma; IL-1β: Interleukin 1 beta; IL-4: Interleukin 4; IL-6: Interleukin 6; iNOS: Inducible nitric oxide synthase; Th: T helper cells; TNF-α: Tumor necrotic factor alpha; Treg: Regulatory T cells.

To further explore the role of exosomal miR-425-5p in T cell differentiation, primary naïve CD4⁺ T cells were cultured with exosomes isolated from HCT116 cells transfected with miR-425-5p inhibitor or NC. Real-time quantitative PCR analysis of T cell cytokines revealed significant changes in pro- and anti-inflammatory cytokine expression. In the miR-425-5p inhibitor group, proinflammatory cytokines IL-1β, IL-6, and TNF-α were significantly increased (IL-1β: P < 0.05, 95%CI: 0.3545-6.125; IL-6: P < 0.05, 95%CI: 0.5998-4.264; TNF-α: P < 0.05, 95%CI: 0.4767-4.433), while the anti-inflammatory cytokine IL-4 was markedly decreased (IL-4: P < 0.05, 95%CI: -2.439 to -0.4751) (Figure 2C). Flow cytometry analysis of CD4⁺ T cell subsets showed pronounced shifts in differentiation patterns. The proportion of CD4⁺IFNγ⁺ T-helper 1 (Th1) cells was significantly increased, whereas CD4⁺IL-4⁺ Th2 cells were decreased in the miR-425-5p inhibitor group (Th1/Th2: P < 0.001, 95%CI: 0.5486-0.5894) (Figure 2D and E). Additionally, CD4⁺IL-17A⁺ Th17 cells were elevated, and the proportion of immunosuppressive CD4⁺CD25⁺Foxp3⁺ regulatory T cells (Tregs) was reduced compared to the NC group (Th17/Treg: P < 0.001, 95%CI: 0.3680-0.4026) (Figure 2D and E). These results collectively indicate that exosomal miR-425-5p from CRC cells modulates T cell differentiation, and its depletion skews differentiation toward a proinflammatory Th1 and Th17 phenotype while reducing anti-inflammatory Th2 and immunosuppressive Treg subsets. To validate the effects of exosomal miR-425-5p on immune modulation, similar experiments were conducted using SW620 cells. The results, which corroborate the findings from HCT116 cells, are presented in Supplementary Figures 1 and 2, further supporting the role of miR-425-5p in regulating macrophage polarization and T cell differentiation.

Exosomal miR-135b-3p enhances vascular permeability and angiogenesis in endothelial cells

To explore the role of exosomal miR-135b-3p in vascular permeability, exosomes isolated from HCT116 cells transfected with miR-135b-3p inhibitor or NC were cultured with HUVECs. Microscopic observation revealed a reduction in angiogenesis in the inhibitor group compared to NC (branch points: P < 0.05, 95%CI: -47.92 to -18.08) (Figure 3A). Cell invasion assays showed a slight but consistent decrease in invasive capacity in the inhibitor group (invaded cells: P < 0.01, 95%CI: -279.8 to -75.51) (Figure 3B). Similarly, the rhodamine B permeability assay indicated a modest decrease in permeability in HUVECs treated with miR-135b-3p inhibitor exosomes (P < 0.01, 95%CI: -6.134 to -2.512) (Figure 3C). In the aortic ring assay, the number of microvascular sprouts was significantly reduced in the inhibitor group (P < 0.05, 95%CI: -119.3 to -48.71) (Figure 3D). ZO-1, occludin, and claudin-5 are tight junction proteins and lack of them leads to increased vascular permeability[31]. Western blot analysis demonstrated increased expression of them in the inhibitor group, indicating enhanced endothelial barrier integrity (ZO-1: P < 0.05, 95%CI: 1.206-1.374; occludin: P < 0.05, 95%CI: 0.9003-1.106; claudin-5: P < 0.05, 95%CI: 0.8941-1.106) (Figure 3E). Immunofluorescence analysis further confirmed the upregulation of ZO-1, occludin and claudin-5 in HUVECs treated with miR-135b-3p inhibitor exosomes (claudin-5: P < 0.05, 95%CI: 528.5-2945; occludin: P < 0.05, 95%CI: 239.5-2161; ZO-1: P < 0.05, 95%CI: 420.6-2121) (Figure 3F). These results suggest that exosomal miR-135b-3p from CRC cells contributes to increased vascular permeability and angiogenesis. Similar experiments conducted using SW620 cells yielded comparable results, further supporting the role of miR-135b-3p in vascular permeability. These findings are presented in Supplementary Figure 3.

Figure 3 Exosomal miR-135b-3p enhances vascular permeability and angiogenesis in endothelial cells. A: Representative images and branch points of microscopic observation of angiogenesis in human umbilical vein endothelial cells (HUVECs) cultured with exosomes from HCT116 cells transfected with miR-135b-3p inhibitor or negative control; B: Representative images and invaded cells of cell invasion assay; C: Permeability of the HUVEC monolayers to rhodamine–dextran (70 kDa) after exposure to exosomes for 72 hours; D: Representative images of the aortic ring and numbers of microvascular sprouts; E: Western blot analysis of tight junction proteins zonula occludens-1, occludin, and claudin-5 in HUVECs treated with exosomes; GAPDH as an internal control; F: Immunofluorescent staining of tight junction proteins zonula occludens-1, occludin, and claudin-5 (green) in HUVECs treated with DAPI nuclei (blue). ^aP < 0.05, ^bP < 0.01 vs negative control-exosome (n = 3 per group). NC-exo: Negative control-exosome; ZO-1: Zonula occludens-1.

In vivo roles of exosomal miR-425-5p and miR-135b-3p in tumor growth, metastasis and immune microenvironment modulations

Inhibition of exosomal miR-425-5p and miR-135b-3p significantly suppressed tumor growth, metastasis, and altered the tumor microenvironment in vivo. Compared to the model group, tumor volume, weight, and gross size were markedly reduced in the miR-425-5p inhibitor group (gross tumor volume: P < 0.05, 95%CI: -4.327 to -3.651; tumor weight: P < 0.05, 95%CI: -0.4184 to -0.3523), the miR-135b-3p inhibitor group (gross tumor volume: P < 0.01, 95%CI: -3.431 to -2.755; tumor weight: P < 0.01, 95%CI: -0.3317 to -0.2656), and most notably in the combined inhibition group (gross tumor volume: P < 0.01, 95%CI: -5.846 to -5.170; tumor weight: P < 0.01, 95%CI: -0.5737 to -0.5076) (Figure 4A). Metastatic burden, as indicated by the number of liver and lung tumors, was also significantly decreased in the miR-425-5p inhibitor group (liver tumor: P < 0.05, 95%CI: -3.253 to 0.5865; lung tumor: P < 0.05, 95%CI: -5.358 to -2.642), the miR-135b-3p inhibitor group (liver tumor: P < 0.01, 95%CI: -5.253 to -1.414; lung tumor: P < 0.01, 95%CI: -7.691 to -4.976), with the combined inhibition group demonstrating the lowest metastasis rates (liver tumor: P < 0.001, 95%CI: -6.586 to -2.747; lung tumor: P < 0.001, 95%CI: -9.691 to -6.976) (Figure 4B). Histological analyses revealed increased immune cell infiltration and reduced angiogenesis in the inhibitor groups (Figure 4C). Immunoblotting of CD34 showed reduction in the miR-425-5p inhibitor group (P < 0.05, 95%CI: -5.024 to -2.309), the miR-135b-3p inhibitor group (P < 0.01, 95%CI: -6.358 to -3.642), and the combined inhibition group (P < 0.001, 95%CI: -9.024 to -6.309), further confirming the anti-angiogenic effects (Figure 4D).

Figure 4 Inhibition of exosomal miR-425-5p and mir-135b-3p reduces tumor growth, metastasis, and angiogenesis in vivo. A: Body weight, tumor volume, tumor weight and gross size of tumors from mice treated with HCT116 cells; B: Number of liver and lung tumors; C: Representative images of hematoxylin and eosin staining in tumor tissues; D: Immunohistochemical staining of CD34 in tumor tissues. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001 vs model (n = 3 per group). CD34: Hematopoietic progenitor cell antigen CD34.

Further analyses showed a pronounced shift in the immune microenvironment. Macrophages exhibited a proinflammatory polarization, with increased M1 marker (iNOS) expression and decreased M2 marker (CD206) expression shown by immunofluorescence staining (Figure 5A). Specifically, compared to the model group, iNOS expression was significantly increased in the miR-425-5p inhibitor group (P < 0.01, 95%CI: 4100-7726), miR-135b-3p inhibitor group (P < 0.05, 95%CI: 1642-5268), and combined inhibition group (P < 0.001, 95%CI: 7462-11088). CD206 expression was significantly decreased in the miR-425-5p inhibitor group (P < 0.01, 95%CI: -9169 to -2203), miR-135b-3p inhibitor group (P < 0.05, 95%CI: -6469 to 496.2), and combined inhibition group (P < 0.001, 95%CI: -17503 to -10537).

Figure 5 Exosomal miR-425-5p and mir-135b-3p inhibition modulates the immune microenvironment in tumor tissues. A: Immunofluorescence staining of M1 marker inducible nitric oxide synthase and M2 marker CD206 in tumor tissues; B: Flow cytometry analysis of tumor-infiltrating T cells, including CD4⁺IFNγ⁺ T helper 1 (Th1), CD4⁺IL-4⁺ Th2, CD4⁺IL-17A⁺ Th17, and CD4⁺CD25⁺Foxp3⁺ regulatory T cell; C: ELISA results of interleukin-1β, interleukin-6, and tumor necrotic factor alpha in tumor tissues. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001 vs model (n = 3-5 per group). CD206: Macrophage mannose receptor 1; iNOS: Inducible nitric oxide synthase; Th: T helper cells; Treg: Regulatory T cells; IL-1β: Interleukin 1 beta; IL-6: Interleukin 6; TNF-α: Tumor necrotic factor alpha.

Flow cytometry of tumor-infiltrating T cells indicated increased Th1 (CD4⁺IFNγ⁺) and Th17 (CD4⁺IL-17A⁺) populations, along with reduced Th2 (CD4⁺IL-4⁺) and Treg (CD4⁺CD25⁺Foxp3⁺) populations (Figure 5B). Compared to the model group, Th1/Th2 ratios were significantly increased in the miR-425-5p inhibitor group (P < 0.001, 95%CI: 1.157-2.075), miR-135b-3p inhibitor group (P < 0.05, 95%CI: 0.1080-1.026), and combined inhibition group (P < 0.001, 95%CI: 5.949-6.867). Th17/Treg ratios showed increases in the miR-425-5p inhibitor group (P < 0.001, 95%CI: 1.325-1.787), miR-135b-3p inhibitor group (P < 0.001, 95%CI: 0.5619-1.025), and combined inhibition group (P < 0.001, 95%CI: 11.33-11.79).

Correspondingly, ELISA results demonstrated elevated levels of proinflammatory cytokines IL-1β, IL-6 and TNF-α in the inhibitor groups, with the highest levels in the combined inhibitor group (Figure 5C). Compared to the model group, IL-1β levels were slightly decreased in the miR-425-5p inhibitor group (P < 0.05, 95%CI: -19.79 to -6.363), and significantly increased in the miR-135b-3p inhibitor group (P < 0.05, 95%CI: 36.93-50.36), and combined inhibition group (P < 0.01, 95%CI: 108.7-122.2). IL-6 levels showed significant increases in the miR-425-5p inhibitor group (P < 0.05, 95%CI: 83.94-242.5), miR-135b-3p inhibitor group (P < 0.05, 95%CI: 70.66-229.2), and combined inhibition group (P < 0.01, 95%CI: 162.9-321.4). TNF-α levels were significantly elevated in all inhibitor groups (miR-425-5p inhibitor: P < 0.05, 95%CI: 53.76-125.3; miR-135b-3p inhibitor: P < 0.05, 95%CI: 11.00-82.51; combined inhibition: P < 0.01, 95%CI: 99.32-170.8).

To validate these findings, the same in vivo experiments were performed using SW620 cells. The results were consistent with those observed in HCT116 cells, further supporting the roles of exosomal miR-425-5p and miR-135b-3p in promoting tumor growth, angiogenesis, and immune evasion. These validation results are presented in Supplementary Figures 4 and 5. These findings underscore that the inhibition of miR-425-5p and miR-135b-3p can effectively reprogram the tumor microenvironment to a less aggressive phenotype.

DISCUSSION

Our study identifies and characterizes the roles of exosomal miR-425-5p and miR-135b-3p in promoting CRC progression through immune modulation and vascular dynamics. Our findings reveal that both miRNAs are markedly enriched in CRC tumor-derived exosomes and synergistically orchestrate a tumor-permissive microenvironment. Inhibition of these miRNAs suppresses tumor growth, reduces metastasis, and shifts the tumor microenvironment to a less aggressive state. These results highlight their potential as therapeutic targets and biomarkers in CRC.

Exosomal miR-425-5p was found to be of crucial significance in immune modulation, promoting M2 macrophage polarization and skewing T cell differentiation toward immunosuppressive subsets. This finding is in line with Wang et al’s study, which demonstrated the role of exosomal miR-425-5p in M2 polarization in CRC[29]. Our research not only corroborates their results but also extends understanding of miR-425-5p’s involvement in T cell differentiation during CRC progression. Previously, the role of miR-425-5p in T cell modulation had not been reported. One study reported a positive correlation between miR-425-5p and Tregs in pancreatic cancer tissues[24], while another indicated a positive relationship between miR-425-5p and IL-23 in pancreatic cancer[32]. Given that IL-23 is involved in the survival and expansion of Th17 cells[33,34], our study shows that miR-425-5p modulates T cell differentiation towards immunosuppressive Tregs and Th2 subsets instead of Th17 and Th1 subsets, providing new perspectives on its function in CRC. This modulation likely occurs through the regulation of the NOD-like receptor signaling pathway, which has been implicated in immune responses[35,36]. Functional enrichment analyses implicate the NOD-like receptor pathway in miR-425-5p-mediated immune evasion, warranting further mechanistic validation. These findings position miR-425-5p as a central hub for tumor-driven immune suppression.

In parallel, miR-135b-3p appears to facilitate tumor metastasis by enhancing vascular permeability and angiogenesis. Its inhibition leads to the upregulation of tight junction proteins such as ZO-1, occludin, and claudin-5, thereby restoring endothelial barrier integrity. These results are consistent with previous studies in breast cancer, where miR-135b-3p has been implicated in promoting angiogenesis[21]. Our findings offer novel insights into its function in CRC by linking it to vascular permeability and endothelial barrier dysfunction. The observed effects may involve VEGF pathway modulation, given its established role in vascular permeability and miRNAs are known to post-transcriptionally regulate its expression[37,38]. Although further experiments are required to validate the direct molecular targets and signaling cascades involved, our data provide initial evidence that these miRNAs are key modulators of the tumor microenvironment.

The in vivo experiments underscore the synergistic effects of targeting both miRNAs. Combined inhibition of miR-425-5p and miR-135b-3p resulted in the most pronounced reduction in tumor growth and metastasis, along with significant changes in immune and vascular dynamics. These findings suggest that combinatorial targeting of these miRNAs could provide a more effective therapeutic strategy compared to single-agent approaches.

Our study also highlights the potential clinical utility of these miRNAs as biomarkers. Their elevated levels in CRC-derived exosomes, compared to normal tissues, suggest their use in non-invasive diagnostic or prognostic applications. The result of miR-425-5p aligns with previous studies in CRC tumors or serums[25-28,39], and we recommend miR-135b-3p as a new biomarker for CRC. Furthermore, their roles in immune suppression and vascular permeability make them attractive targets for combination therapies, such as immune checkpoint inhibitors or anti-angiogenic agents.

Despite these promising findings, there are limitations to our study. One limitation of our study is the relatively limited number of patient samples, which may affect the generalizability of the findings. Although our cohort includes well-characterized CRC specimens and the results were corroborated in multiple CRC cell lines and in vivo models, we recognize that expanding the sample size - especially including patients from various clinical stages and with different molecular subtypes - would further strengthen the conclusions. Future studies are planned to validate these results in larger and more diverse cohorts.

The use of cell-line-derived exosomes may not fully recapitulate patient tumor heterogeneity. Additionally, while functional enrichment analyses provided insights into potential pathways, direct validation of these mechanisms remains to be explored. While our study provides compelling evidence that exosomal miR-425-5p and miR-135b-3p play critical roles in modulating immune responses and vascular permeability in CRC, we recognize that further experimental validation of the underlying molecular mechanisms is necessary. Specifically, additional experiments such as gene knockout/overexpression studies, dual-luciferase reporter assays, and immunoprecipitation experiments would be valuable to confirm the involvement of the NOD-like receptor and VEGF signaling pathways, as well as to identify direct downstream targets of these miRNAs. Due to constraints in time and resources, these experiments were not included in the present study. However, we consider this an important avenue for future research, which will help to more fully elucidate the molecular mechanisms by which these exosomal miRNAs contribute to CRC progression. Expanding this work to include patient-derived exosomes and pathway-specific experiments will further strengthen the translational relevance of these findings.

CONCLUSION

This study establishes exosomal miR-425-5p and miR-135b-3p as key mediators of CRC progression, shaping immune evasion and vascular dysfunction. Their dual targeting offers a promising strategy to disrupt tumor-microenvironment crosstalk, paving the way for innovative therapeutic interventions in CRC.