Netrin-1 Promotes the Immunosuppressive Activity of MDSCs in Colorectal Cancer

In recent years, the tumor microenvironment has become appreciated as a hotbed for providing protective conditions for tumor cell growth and helping acquire phenotypes for metastasis (1). Once the tumor niche is established, the tumor subverts beneficial anticancer immune effector cells into dysfunctional bystanders, or into immunosuppressive cells (2). Of note, myeloid-derived suppressor cells (MDSC) have immunosuppressive activity and promote resistance to effector immune cells. MDSCs represent a heterogeneous population that comprises myeloid cell progenitors and precursors of myeloid cells arrested in an immature state with an immunosuppressive function (3, 4), and MDSCs accumulate and expand in large numbers in the bone marrow, spleen, and tumor microenvironment during tumor development (2, 5). In tumor-bearing mice or patients with cancer, all subpopulations show a ability to suppress T-cell responses in an antigen-specific or nonantigen-specific manner (5). Moreover, the induction of MDSCs plays a crucial role as an immune escape mechanism at most tumor sites, and MDSCs actively promote tumor angiogenesis, maintain tumor cell stemness, and promote tumor invasion and metastasis (6). Because of their unique features, MDSCs are not present in steady-state conditions but are recruited and expanded during tumor progression. These suggest that MDSCs may be a cancer immunotherapy target without possible side effects. Understanding the molecular mechanisms that regulate the accumulation and function of MDSCs will allow for more precise targeted therapy.

Netrin-1 belongs to the laminin-related factor family and interacts with multiple receptors, including deleted in colorectal cancer (DCC) and its homolog receptor neogenin, UNC5 homolog family members (UNC5A-D in humans and UNC5H1–4 in rodents), Down syndrome cell adhesion molecule (DSCAM), the newly discovered alternative receptor adenosine receptor 2B (A2BR), and members of the integrin family (7, 8). Netrin-1 and its receptors, either as partners or opponents, are condition-dependent in some types of cancer (9–11). The UNC5 family and DCC are recognized as dependent receptors and induce tumor cell apoptosis in the absence of netrin-1 (11, 12). Thus, most tumors selectively depend on receptor loss or netrin-1 upregulation to provide tumor survival and invasion signals. In intestinal tissue, netrin-1 is heterogeneously expressed with gradient characteristics. Netrin-1 is more highly expressed at the proximal intestinal villus, which can cause intense cellular proliferation, rather than at the distal tip of the villus, which controls cell apoptosis. The ectopic expression of netrin-1 or receptor loss may be causes of intestinal tumorigenesis (13). Netrin-1 expression has a direct correlation with inflammation and cancer. Although receptor loss, such as DCC and the UNC5 family, has a high frequency of occurrence, rather than increasing the autocrine expression of netrin-1 in sporadic colorectal cancer, significantly elevated netrin-1 expression is a requirement for the development of colitis-associated colorectal cancer (14, 15). Hypoxia is an inducible factor of netrin-1 expression that is dependent on HIF1α. Netrin-1 interacts with A2BR on polymorphonuclear neutrophils (PMN) and then weakens hypoxia-induced inflammation (16). However, the effects of tumor-derived netrin-1 on MDSCs have not yet been elucidated.

In our research, we found that netrin-1 secreted by colorectal cancer cells plays an important role in regulating the immunosuppressive function of MDSCs. Netrin-1 can enhance the ability of MDSCs to inhibit T cells by upregulating a variety of effector molecules. In addition, we further explored the mechanism by which netrin-1 functions. Netrin-1 activates the internal cAMP/PKA/CREB signaling pathway through A2BR to enhance the immunosuppressive activity in MDSCs, thus providing a novel candidate target for therapeutic intervention.

The murine colon cancer lines MC38 and CT26 were obtained from the Cell Bank, Chinese Academy of Sciences in 2019. Cells were cultured for a maximum of 15 passages in RPMI1640 (Gibco, 21875–034) with 10% FBS (Gibco, 12484028) and 1% penicillin/streptomycin antibiotics (Gibco, 10378016) in a humidified atmosphere at 37°C with 5% CO₂. Cell lines were not authenticated. Both cell lines were routinely tested to ensure there was no Mycoplasma contamination with MycAway-Color One-Step Mycoplasma Detection Kit (40612ES25, Yeasen Biotechnology). MC38 cells were transfected with shRNA scramble and shRNA NTN1 (Genechem) according to the manufacturer's instructions. The transfected cells screened with 1 μg/mL puromycin (MCE, HY-B1743A) to generate stably transfected shRNA scramble and shRNA NTN1 cell lines, as stated below.

Fresh peripheral blood samples from 36 cases of patients with colorectal cancer and 36 cases of healthy control individuals with similar ages and sex ratios were collected at the Affiliated People's Hospital, Jiangsu University. Peripheral blood mononuclear cells (PBMC) were isolated by density-gradient centrifugation using Ficoll–Hypaque solution (Haoyang Biological Technology, LDS1075) and subsequently analyzed by flow cytometry as described below. Plasma samples were collected and stored at −80°C for netrin-1 measurement. Colorectal cancer tissues and adjacent colorectal tissues were collected from patients with colorectal cancer undergoing surgery and were stored at −80°C for RNA extraction or fixed in 4% paraformaldehyde and embedded in paraffin to prepare tissue sections for IHC analysis as stated in sections below. The protocol was approved by the Ethics Committee of the Affiliated People's Hospital of Jiangsu University. All subjects gave written informed consent in accordance with the Declaration of Helsinki.

Female BALB/c and C57BL/6 mice, ages 6 to 8 weeks old, were purchased from the Laboratory Animal Center of Jiangsu University and kept under specific pathogen-free conditions. All experiments with mice were performed with the consent of the Institutional Animal Care and Use Committee of Jiangsu University (No. UJS-IACUC-AP-20190305019) and complied with the Guide for the Care and Use of Laboratory Animals published by the US NIH.

MC38 and CT26 (2 × 10⁶) cells were subcutaneously implanted into the right flanks of female C57BL/6 and BALB/c mice, respectively. Mice were euthanized on Day 14, 21, and 28 after tumor injection; tumor tissues, draining lymph nodes (dLN), and spleens were then harvested and analyzed by flow cytometry as indicated in the “Flow cytometry” section.

For some in vivo experiments, shRNA NTN1 and shRNA scramble transfected MC38 (2 × 10⁶) cells were injected into the right flanks of female C57BL/6. Tumor progression was monitored every 2 days. The tumor volume was measured every day by Vernier calipers and calculated using the following formula: tumor volume (mm³) = (length × width²)/2. Mice were euthanized on Day 18 after tumor injection, and tumor tissues, dLNs, and spleens were harvested and analyzed by flow cytometry.

MC38 (1 × 10⁶) cells were subcutaneously implanted into the right flanks of female C57BL/6. On days 10 and 14 after tumor injection, mice were treated with intraperitoneal injection of 100 μg monoclonal anti-DR5 (BioXCell, MD5–1) to deplete the endogenous MDSCs. Armenian hamster IgG isotype (BioXCell, BE0091) was used as a control. In MDSC transfer experiments, mice were divided into four groups: vehicle (PBS) group, MDSC group, netrin-1-treated MDSC (NTN-1 MDSC) group, and PSB1115 prior to netrin-1-treated MDSC (PSB+NTN-1 MDSC) group. Sorted MDSCs were cultured as mentioned below, and 2 × 10⁶ pretreated MDSCs per mouse were suspended with 100 μL PBS and administrated into tumor tissues on days 16 and 22 after tumor injection. Tumor progression was monitored every 2 days. The tumor volume was measured every day by Vernier calipers and calculated using the following formula: tumor volume (mm³) = (length × width²)/2. Mice were euthanized on Day 28 after tumor injection, tumor tissues, dLNs, and spleens were harvested and analyzed by flow cytometry.

Tumor tissues were minced into 1 to 2 mm³ and digested in RPMI1640 supplemented with 5% FBS, 0.5 mg/mL collagenase type V (Sigma-Aldrich, C9263), 0.2 mg/mL hyaluronidase (Sigma-Aldrich, H1115000), and 0.015 mg/mL DNase I (Sigma-Aldrich, DN25) at 37°C for 1 hour and filtered by a 70 μm cell strainer (Corning, 352350). Red blood cells were then lysed by ACK lysis buffer (Gibco, A1049201) to prepare single-cell suspensions. Spleen and dLNs were grinded in RPMI1640 and filtered by 70 μm cell strainer for further experiments.

To isolate MDSCs, mice bearing subcutaneous tumors were sacrificed after tumor injection for 28 days. Spleens were homogenized, and splenocytes passed through a 70 μm strainer and washed with PBS (Biosharp, BL302A). Red blood cells were then lysed by ACK lysis buffer and a Myeloid-Derived Suppressor Cell Isolation Kit (Miltenyi Biotec, 130–094–538) was used to isolate MDSCs according to the manufacturer's instructions. In this process, cells were suspended in and washed with PBE buffer containing PBS, 2 mmol/L EDTA (Sigma-Aldrich, E9884), and 0.5% FBS. After the magnetic separation, the isolated MDSCs were maintained in RPMI1640 supplemented with 10% FBS and 1% penicillin /streptomycin (referred to as complete medium) for further studies. The MDSC population was defined by the expression of the cell surface antigens CD11b and Gr-1. The purity of the isolated cell populations was determined by flow cytometry after labeling with PE/Cy7-conjugated CD11b (BioLegend, 101216) and PE-conjugated Gr-1 (BioLegend, 108408), and the frequency of CD11b⁺Gr-1⁺ cells was >90%. Murine CD4⁺ T cells and CD8⁺ T cells were isolated from the spleen of C57BL/6 mice using mouse CD4 microbeads (Miltenyi Biotec, 130–117–043) and mouse CD8 microbeads (Miltenyi Biotec, 130–116–478). The sorted purity of CD4⁺ T and CD8⁺ T cells was more than 90% by flow cytometry.

Murine MC38 cells were engineered to contain netrin-1 knockdown and a luciferase construct to facilitate in vivo bioluminescence imaging. For this purpose, netrin-1 knockdown was achieved by using lentivirus-delivered shRNA (Genechem). The lentivirus contained shRNA targeting the Netrin-1 gene (NTN1), and the component order was hU6-MCS-CBh-gcGFP-IRES-puromycin. The following sequences were used:

shRNA scramble (CON313): CCGGTTCTCCGAACGTGTCACGTCTCGAGA-CGTCACGTTCGGAGGCGGAGAATTTTTG

NTN1 shRNA (9116511): CCGGGTGGAAGTTCACCGTGAACATCTCGAG

ATGTTCACGGTGAACTTCCACTTTTTG

MC38 cells were suspended and plated into 24-well plates and allowed to reach a confluency of 20% to 30%. Then, the medium was replaced with fresh culture medium containing an appropriate number of viral particles and 1× HistransG A for 8 hours. The transfection mix was then replaced with fresh culture medium. Successful shRNA transduction was identified by fluorescence microscopy or flow cytometry using the GFP tag. Cells successfully infected with lentivirus were further selected with 1 μg/mL puromycin. After puromycin selection of luciferase‐positive cell populations, the efficacy of shRNA transduction was determined by Western blot analysis, qRT‒PCR, and flow cytometry, as indicated below.

siRNA targeting A2BR and corresponding negative control (siNC; siN0000001–1-5) were designed and manufactured by RiboBio. The targeting specific sequences were listed as follows: siA2BR1: CCACAAGATCATCTCCAGA; siA2BR2: ACTACTTTCTGGTATCCCT; siA2BR3: GCTACATGGTGTATGGCAA. MDSCs were sorted, as above mentioned, and plated into 24-well plates at an 80% confluency for 1 hour. siRNAs were transfected at 50 nmol/L into MDSCs using Lipofectamine 3000 (Invitrogen, L3000150) at 37°C for 24 hours or cocultured with MC38 at 1:2 ratio for 24 hours. The mRNA and protein expression levels of A2BR in MDSCs were assessed by Western blot analysis and qRT-PCR. The Arg1 activity and ROS production were measured as indicated below.

For acquiring MC38 conditioned medium (TCCM), cell density of MC38 cells was reached into 60% confluency, fresh medium was added to replace the culture medium and was collected 24 hours. Collected TCCM was filtered by 0.22 μm sterile syringe filter (Millipore, SLGSR33SS) and used to stimulate MDSCs later.

MDSCs were seeded in 24-well plates (1.5 × 10⁶ MDSCs/well) and were treated with recombinant Netrin-1 protein (R&D Systems, 1109-N1) at a dose of 200 ng/mL for 24 hours. Cells were harvested to detect the immunosuppressive factors, including Arg1, ROS, PD-L1, and supernatants were collected and used for NO, IL10, TGFβ, and PGE2 detection as mentioned below. Otherwise, netrin-1-treated MDSCs were harvested at 0, 5, 15, 30, and 60 minutes for CREB phosphorylation detection by Western blot analysis. For inhibiting A2BR activity, the A2BR antagonist PSB1115 (Tocris Bioscience, 2009) was added at 10 μmol/L for 30 minutes before netrin-1 treatment for 24 hours. DMSO plus netrin-1-treated MDSCs served as control. In addition, MDSCs were pretreated with 10 μmol/L PSB1115 for 30 minutes and then cocultured with MC38 cells at 1:2 ratio in complete medium for 24 hours. To identify the effect of cAMP and PKA on CREB phosphorylation and Arg1 activity, MDSCs were incubated with 20 μmol/L PKA inhibitor H89 (Selleck, S1582) or 10 μmol/L adenylyl cyclase agonist Forskolin (Beyotime, S1612) for 30 minutes and 24 hours, respectively. For netrin-1 blocking experiments, MDSCs were incubated in TCCM with 10 μg/mL of neutralizing antibody to netrin-1 (R&D Systems, AF1109) or isotype IgG control (R&D Systems, AB-108-C) for 24 hours. In addition, MDSCs were cocultured with shRNA scramble or shRNA NTN1 transfected MC38 at 1:2 ratio and incubated in complete medium for 24 hours.

For the proliferation assay, MDSCs were treated with recombinant netrin-1 and PSB1115 or transfected with siRNA as described above 1 day in advance prior to coculture with T cells. CFSE (5 μmol/L, Invitrogen, C34554) labeling was used to measure T-cell proliferation as described previously (17). CFSE labeled T cells were cocultured with pretreated MDSCs at a 1:1 ratio in round-bottomed, 96-well plates in the presence of 10 μg/mL purified anti-mouse CD3 (BioLegend, 100253) and 5 μg/mL purified anti-mouse CD28 (BioLegend, 102121) for 72 hours. T cells without anti-CD3 and anti-CD28 were used as a negative controls. T-cell proliferation was determined by CFSE dye dilution by flow cytometry as indicated below.

Total RNA was extracted from cultured MDSCs or MC38 and CT-26 with TRizol (Invitrogen, 15596018). One microgram of RNA were reverse-transcribed into cDNA using the PrimeScript RT Reagent Kit (Takara, RR037A) according to the manufacturer's instructions. The mRNA expression of target genes was quantified using TB Green premix (Takara, RR420A) and was conducted on C1000 Thermal Cycler instrument with CFX96 detection systems (Bio-Rad). Each reaction was performed in duplicate, and changes in relative gene expression normalized to β-actin expression was determined using the comparative threshold cycle method. Primer sequences are shown in Table 1. The efficiency and specificity of all primers were confirmed using the Basic Local Alignment Search Tool of the National Center for Biotechnology Information (http://blast.ncbi.nlm.nih.gov).

Total proteins from cultured MDSCs, MC38, and CT26, and human colorectal tissues were extracted using RIPA lysis buffer (CoWin Bioscience, CW2333) supplemented with protease inhibitor cocktail (CoWin Bioscience, CW2200) and phosphatase inhibitor cocktail (CoWin Bioscience, CW2383). Twenty-five micrograms of proteins samples were separated on 10% SDS-PAGE polyacrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, IPVH00010). The membranes were first probed with the indicated primary antibodies at 4°C overnight, followed by a horseradish peroxidase (HRP)-linked secondary antibody for 1 hour at room temperature. Protein detection was conducted using enhanced chemiluminescence (ECL) substrate (Millipore, WBULS0500), and the signal intensities were quantified by an ImageQuant LAS4000 chemiluminescence gel imaging and analysis system (GE Healthcare). Monoclonal rabbit anti-netrin-1 (ab126729, 1:1,000), monoclonal rabbit anti-CREB (ab32515, 1:1,000), monoclonal anti-CREB (phosphoS133; ab32096, 1:1,000), and HRP-conjugated goat anti-rabbit IgG (ab205718, 1:10,000) were purchased from Abcam. Monoclonal rabbit anti-β-actin (AC038, 1:100,000), monoclonal rabbit anti-Arg1 (A4923, 1:5,000), and polyclonal rabbit anti-A2BR (A1953, 1:2,000) were purchased from Abclonal.

Tumor tissues, dLNs, spleens, cultured MDSCs, MC38, and human PBMCs single-cell suspensions were prepared as indicated above, and 1 × 10⁶ cells were stained with the indicated antibodies for 30 minutes in PBS under dark and cold conditions. After washing with PBS, the samples were counted and analyzed by flow cytometry. For mouse MDSCs, as well as subgroup analysis, cells were incubated with relevant fluorochrome-conjugated anti-mouse CD45 (100712), anti-mouse/human CD11b (101216), anti-mouse Gr-1 (108408), Ly6G (127608), and Ly6C (128006) antibodies. All antibodies used were obtained from BioLegend. Anti-mouse PD-L1 (BioLegend, 124314) was used to detect PD-L1 intensity on MDSCs. For MDSCs analysis in human PBMCs, cells were incubated with anti-CD11b (BioLegend, 101216), anti-CD33 (BioLegend, 366616), and anti-HLA-DR (BioLegend, 327014). To detect the infiltration of cytotoxic T lymphocytes (CTL) and T helper 1 (Th1) cells in tumor tissues, spleens, and dLNs, 2 × 10⁶ cells were resuspended and stimulated in complete medium with 1 μg/mL ionomycin (Sigma-Aldrich, I3909), 2 ng/mL monensin (eBioscience, 00–4505–51), and 50 ng/mL PMA (Sigma-Aldrich, P1585) for 5 hours. Cells were collected and washed with PBS and stained with monoclonal anti-mouse CD8a (BioLegend, 100712) or anti-mouse CD4 (BioLegend, 100422). For intracellular staining, cells were fixed and permeabilized with intracellular fixation/permeabilization buffer (eBioscience, 00–8222–49/00–8333–56) and stained with monoclonal anti-mouse IFNγ (eBioscience, 12–7311–82). Rat IgG1 κ PE (eBioscience, 12–4301–82) were used as isotype control. For Treg staining, 2 × 10⁶ cells were resuspended and stained with monoclonal anti-CD4 and anti-CD25 (eBioscience, 45–0251–82) for surface antigens. Cells were then permeabilized with a Foxp3/Transcription Factor Staining Buffer Kit (eBioscience, 00–5523–00) and incubated with a monoclonal anti-Foxp3 (eBioscience, 12–4776–42). Rat IgG2a κ PE were used as isotype control (eBioscience, 12–4321–80). Moreover, shRNA scramble and shRNA NTN1 MC38 cells were permeabilized and incubated with monoclonal anti-Ki67 (eBioscience, 48–5698–82). Cells were resuspended in 200 μL of PBS for subsequent flow cytometric analysis. Data were acquired using a flow cytometer (Beckman Coulter, CytoFLEX) and processed using FlowJo (Tree Star, Inc.).

The oxidation-sensitive dye 2,7-dichlorodihydrofluorescein diacetate (H2DCFDA, Invitrogen, D399) was used to measure reactive oxygen species (ROS) production by MDSCs. MDSCs were treated as mentioned above, harvested and washed with PBS, incubated with or without 30 ng/mL PMA in the presence of 2.5 μmol/L dichlorofluorescein in 1 mL PBS for 1 hour at 37°C, washed with PBS, and analyzed using flow cytometry as stated above.

Total nitric oxide (NO) produced by MDSCs was determined with a Griess Reagent Kit (Promega, G2930). After netrin-1 treatment for 24 hours, the supernatant of MDSCs was collected. Fifty microliters of supernatant samples were added into 96-well plates in duplicate or triplicate, and then carried out according to the procedures provided by the manufacturer. The absorbances were measured between 520 and 550 nm by a Multi-Mode Reader (Bio-Tek, Cytation 5) and analyzed by Curve Expert 3.2.

Arginase activity was measured in cell lysates by a QuantiChrom Arginase Assay Kit (Bioassay Systems, DARG-100). MDSCs were treated as mentioned above, 1 × 10⁶ cells were harvested per sample and washed with PBS. The samples were centrifuged at 1,000 × g at 4°C for 10 minutes. The cell pellets were lysed for 10 minutes in 100 μL RIPA lysis buffer. The lysates were then centrifuged at 14,000 × g at 4°C for 10 minutes, and 80 μL supernatant was used for arginase assays according to the manufacturer's instructions. All samples need to prepare a blank control and read the optical density (OD) at 430 nm by Cytation 5 and calculate Arg1 activity (U/L): (OD_sample − OD_blank)/(OD_standard − OD_water) × 10.4.

Freshly isolated MDSCs (1 × 10⁷) were exposed to the indicated concentrations of recombinant Netrin-1 or PSB1115 over 60 minutes at 37°C. Cells were lysed with ice-cold hypotonic water containing 0.1 mol/L HCl, centrifuged at 1,000 × g for 10 minutes for acquiring supernatant. The supernatants were diluted at 1:2 ratio and cAMP concentrations were determined with cAMP ELISA Kit (Cayman Chemical, 581001). The absorbances were measured at 410 nm by Cytation 5 and analyzed by Curve Expert 3.2.

1.5 × 10⁵ MC38 and CT-26 were cultured for 24 hours and collected the supernatant to assess soluble netrin-1 using the Mouse Netrin-1 ELISA Kit (RayBiotech, ELM-NTN1–1). MDSCs were treated as indicated above for 24 hours, the supernatants were collected to detect TGFβ, IL10, and PGE2 production. ELISA kits for TGFβ (MultiSciences, EK981), IL10 (MultiSciences, EK210), and PGE2 (Cayman Chemical, 514010) were used according to the manufacturer's protocol. IL2 and IFNγ in the coculture system of the T-cell proliferation assay were measured using mouse ELISA kits (MultiSciences, EK202, and EK280) according to the protocol provided by the manufacturer. Plasma from patients with colorectal cancer and healthy control individuals were unfrozen and diluted by ELISA buffer at a 1:3 dilution ratio. Using Human Netrin-1 ELISA Kit (CUSABIO, CSB-E11899h) to detect netrin-1 concentrations in diluted plasma. All absorbance was read at 450 nm by Cytation 5 and analyzed by Curve Expert 3.2.

MC38 and CT26 were plated onto coverslips (NEST) and fixed with 4% paraformaldehyde. The cells were washed three times with PBS for 5 minutes each. The cells were permeabilized for 20 minutes with 0.1% Triton X-100, and the wash procedure was repeated. Cells were incubated with 5% BSA (Solarbio Life Science, A8020) in PBS for 30 minutes to block nonspecific binding and were incubated with anti-mouse Netrin-1 (Abcam, ab126729,1:200) or polyclonal anti-mouse A2BR (Abclonal, A1953, 1:200), in PBS with 1% BSA at 4°C overnight and washed three times for 5 minutes each with PBS. The cells were then incubated with anti-rabbit IgG-FITC (Invitrogen, 65–6111, 1:200) for 60 minutes at 37°C, counterstained with DAPI (Invitrogen, D1306, 1:1,000) in PBS for 10 minutes, and washed three times for 5 minutes each with PBS. Samples were mounted in appropriate antifade mounting medium (Invitrogen, P36961) and visualized with a Delta Vision (GE Healthcare).

The paraffin-embedded colorectal cancer tissues and adjacent tissues from patients with colorectal cancer were prepared for IHC tests. The paraffin-embedded tissues were cut into 5 μm sections and then sequentially deparaffinized to water, antigen retrieved, and 5% BSA blocking. Processed sections were incubated with anti-netrin-1 (CUSABIO, CSB-PA016127LA01HU), performed using a biotinylated anti-rabbit IgG and streptavidin-HRP from SABC Three-Step Kit (Boster Biotechnology, SA1020), followed by colorimetric detection using DAB Substrate Kit (Boster Biotechnology, AR1022). Tissues were observed and representative images were taken using an Olympus BX41 microscope. Brown cytoplasmic staining represented the positive netrin-1 expression.

The GEPIA database (http://gepia2021.cancer-pku.cn/) consists of RNA sequencing expression data of 9,736 tumors and 8,587 normal samples derived from the TCGA and Genotype-Tissue Expression (GTEx) databases (18). The correlation of netrin-1 expression level and MDSC markers (CD33,ITGAM) in tissues were processed and acquired using GEPIA 2021.

Statistical analysis was achieved using GraphPad Prism Software. Data are presented as the means ± SEM and were considered to be statistically significant when the P value was equal to or less than 0.05. Two groups of data were compared using unpaired two-tailed Student t tests. ANOVA was used to compare multiple groups. Two-way ANOVA with Bonferroni correction was used to compare the growth curves of multiple groups. The association between MDSCs and netrin-1 was analyzed by Spearman Rank Correlation.

All data generated in this study are available within the article and its supplementary data files.

To select colorectal cancer cell lines with higher netrin-1 expression, we compared the expression of netrin-1 in two well-established mouse colorectal cancer cell lines, MC38 and CT26 (Fig. 1). The MC38 cell line is derived from a grade III adenocarcinoma chemically induced in C57BL/6 female mice and is a transplantable mouse tumor model (19). The characteristics of CT26 cells are consistent with undifferentiated/refractory human colorectal cancer and have been shown to be modestly immunogenic (20). In addition, MC38 and CT26 cells represent microsatellite instability (MSI) and microsatellite-stable (MSS) colorectal cancer, respectively (21, 22). Compared with CT26 cells, MC38 cells had more Netrin-1 mRNA expression (Fig. 1A). Western blot analysis and immunofluorescence results showed that netrin-1 protein was more highly expressed in MC38 cells and was mainly located in the cytoplasm (Fig. 1B and C). To further determine the netrin-1 secretion ability, we used ELISAs to detect soluble netrin-1 in the culture supernatant of MC38 and CT26 cells, and MC38 cells released more netrin-1 (Fig. 1D). These results showed that the MC38 colon cancer cell line had abundant netrin-1 expression and a potent ability to release netrin-1 into the extracellular environment.

The immunosuppressive mechanism of MDSCs involves many aspects, including crosstalk with tumor cells, which affects the release of soluble immunomodulatory factors and the interaction between ligands and receptors. Thus, we tried to determine whether abundant netrin-1 release into the tumor microenvironment could affect MDSC populations. We observed differences in the proportions of MDSCs in the two murine colorectal cancer xenograft models. During the progression of the tumor xenografts, the proportion of MC38-derived MDSCs was significantly higher than that of CT26-derived MDSCs in the tumor and spleen (Supplementary Figs. S1A and S1B). In addition, the arginase (Arg1) activity and ROS level of MC38-derived MDSCs were also significantly increased (Supplementary Figs. S1C–S1E), which indicates that netrin-1 likely participates in MDSCs infiltration and immunosuppressive function.

To gain insights into the effects of netrin-1 on the immunosuppressive function of MDSCs, we sorted splenic MDSCs from MC38 tumor-bearing mice for further investigation and analyzed cell purity using flow cytometry. As shown in Fig. 2A, the sorted MDSC purity was higher than 90%. The endogenous Netrin-1 mRNA expression in MDSCs was significantly lower than that in MC38 cells, which suggests that netrin-1 is mainly secreted from tumor cells (Supplementary Fig. S2A). We then used recombinant Netrin-1 to treat MDSCs in vitro. The proliferation of activated CD4⁺ and CD8⁺ T cells was inhibited when CD4⁺ and CD8⁺ T cells were cocultured with netrin-1-treated MDSCs (Fig. 2B and C). Consistently, the soluble IL2 and IFNγ produced by CD4⁺ and CD8⁺ T cells were decreased in netrin-1-treated MDSC group (Fig. 2D). Moreover, the classic effector molecules of MDSCs, such as Arg1 activity and expression as well as ROS production, were increased after netrin-1 treatment, whereas NO production and inhibitory PD-L1 intensity were constant compared with those in the control group (Fig. 2E-I). In addition, some immunosuppressive cytokines, such as IL10 and TGFβ, were significantly induced in cultures containing netrin-1 (Supplementary Figs. S2B and S2C). PGE2 production in MDSCs were not affected after netrin-1 treatment (Supplementary Fig. S2D). These findings collectively indicate that netrin-1 treatment uniquely enhances the suppressive activity of MDSCs in vitro.

To further assess whether tumor-derived netrin-1 has comparable effects on MDSCs, we used filtered TCCM with netrin-1 neutralizing antibody or isotype IgG to treat MDSCs. Because of netrin-1 activity blockade, the suppressive function of activated MDSCs was limited (Fig. 2J and K). In addition, netrin-1 was knocked down using lentivirus-delivered shRNA constructs. The mRNA and protein expression of netrin-1 in these cells was confirmed by qRT-PCR, Western blot analysis, and flow cytometry (Supplementary Figs. S3A–S3D). When cocultured with netrin-1 knockdown MC38 cells in vitro, the immunosuppressive activity of MDSCs was also decreased (Supplementary Figs. S4A–S4D). Together, these data suggest that tumor-derived netrin-1 is a potential agonist for immunosuppressive functions in MDSCs.

Netrin-1 functions as a ligand interacting with a receptor, and the different effects depend on the receptor to which it binds. However, whether netrin-1 can directly target MDSCs via surface receptors in the tumor microenvironment remains unknown. We next evaluated the involvement of netrin-1 signaling events. First, we identified netrin-1 receptors on the surface of MDSCs and found that, among eight reported netrin-1 receptors, only the adenosine A2B receptor (A2BR) was highly expressed on MDSCs (Fig. 3A). Compared with CD11b⁺Gr-1⁺ cells in the spleen from naïve, nontumor-bearing mice, the MDSCs in the spleen from MC38-bearing mice had more A2BR expression (Fig. 3B–D). In contrast, PSB1115 treatment abolished the effects of treatment with recombinant netrin-1 on the suppression of T-cell proliferation (Fig. 3E and F). The effector molecules IL2 and IFNγ released by T cells were increased with PSB1115 treatment (Fig. 3G). The upregulated Arg1 activity, Arg1 expression, and ROS production mediated by netrin-1 were inhibited by PSB1115 (Fig. 3H–J). In addition, secreted IL10 and TGFβ also declined with antagonist treatment (Supplementary Figs. S5A and S5B). When PSB1115-pretreated MDSCs were cocultured with MC38 cells in vitro, they partially blocked the immunosuppressive activity of MDSCs (Fig. 3K and L). In addition, we knocked down the A2BR on MDSCs using three effective A2BR siRNA (Supplementary Fig. S6A) to confirm the immunosuppressive function in vitro. We found that even in the presence of exogenous netrin-1, Arg1 expression, and activity were decreased (Supplementary Figs. S6B–S6C). Immunosuppressive ROS production was also reduced after A2BR knockdown (Supplementary Fig. S6D). These experimental data suggest that netrin-1 targets A2B receptors and may play important roles in netrin-1-mediated immunosuppression of MDSCs in our culture system.

A previous study reports that A2BR is involved in netrin-1-mediated axon outgrowth and may cause cAMP activation in the neural system (23). Netrin-1 also activates A2BR and enhances intracellular cAMP levels, dampening the inflammatory response (24). Thus, we measured cAMP concentrations in netrin-1-treated MDSCs. The results showed that intracellular cAMP concentrations were increased through netrin-1 stimulation (Fig. 4A). Studies with MDSCs pretreated with the A2BR antagonist PSB1115 abrogated netrin-1-induced cAMP production, which suggests that A2BR signaling is required for netrin-1–mediated cAMP synthesis (Fig. 4A). Moreover, activated CREB transcription factor signaling is one key cascade pathway for the induction of immunomodulatory molecules in MDSCs (25, 26). We determined CREB activation in netrin-1-treated MDSCs and found that netrin-1 could rapidly enhance CREB phosphorylation within 60 minutes (Fig. 4B). To further confirm that the downstream PKA was a key protein that affected the immunosuppressive activity of MDSCs, we treated MDSCs in vitro with the PKA selective inhibitor H89, which competitively binds with ATP (27). MDSCs pre-exposed to H89 completely reversed netrin-1-induced Arg1 activity, whereas cells preexposed to the adenylyl cyclase agonist forskolin exhibited a higher level of Arg1 activity than those treated with netrin-1 alone (Fig. 4C). In addition, PSB1115 and H89 inhibited netrin-1-induced CREB phosphorylation and forskolin promoted phosphorylated CREB (Fig. 4D). These studies indicated the involvement of the cAMP/PKA/CREB pathway in netrin-1-induced MDSCs to create a protumor microenvironment.

To further investigate the potential effects of tumor-derived netrin-1 on the immunosuppressive function of MDSCs in vivo, we constructed shRNA NTN1 MC38 and shRNA scramble MC38 cells and transplanted the cells into mice to dynamically monitor tumor growth. As expected, tumor progression of shRNA NTN1 MC38 tumors in mice were significantly delayed compared with that in the shRNA scramble group (Fig. 5A). Tumor burden was clearly alleviated in the shRNA NTN1 group (Fig. 5B and C). Interestingly, the proliferation rate and Ki67 expression of MC38 cells were not affected by netrin-1 knockdown in vitro (Supplementary Figs. S7A and S7B). MDSC infiltration in tumors was slightly decreased, and both PMN-MDSCs and M-MDSCs were decreased in the tumor tissues of the shRNA NTN1 group (Fig. 5D and E). The proportions of MDSCs were decreased while MDSC subsets remained constant in the spleen of shRNA NTN1 mice (Fig. 5F and G). Immunosuppressive components, such as ROS and Arg1, were decreased in MDSCs from shRNA NTN1 tumor bearing mice (Fig. 5H and I). Moreover, we found that netrin-1 knockdown led to a decreased proportion of immunosuppressive Tregs (Supplementary Fig. S8A), while enhancing the infiltration and activation of Th1 cells and cytotoxic CD8⁺ T cells (Supplementary Figs. S8B and S8C), suggesting that netrin-1 knockdown in tumors could boost the antitumor immune responses.

To further rule out the effect of netrin-1 on tumor cells or surrounding stromal cells in tumor microenvironment, we consumed endogenous MDSCs by intraperitoneal injection of anti-DR5 antibody, the MDSCs proportions were reduced to half of original (Supplementary Fig. S9A). We then administrated differently treated MDSCs into tumor tissues after MDSCs depletion (Supplementary Fig. S9B). Tumor growth was accelerated by netrin-1–treated MDSC transfer (Supplementary Fig. S7C). Effector T-cell infiltration in tumor tissues and spleen were decreased when transferring netrin-1–treated MDSCs, while increased T cells infiltration was observed in the PSB1115 pre-incubated MDSC group. However, there were no significance changes in effector T cells from dLNs with MDSCs administration (Supplementary Figs. S9D–S9F). These data showed the direct effects of netrin-1 on MDSCs. Collectively, these results suggest that targeting netrin-1 in tumor cells can impair the infiltration and immunosuppressive function of MDSCs, while restoring antitumor immunity in MC38 tumor-bearing mice.

To determine the correlation between netrin-1 expression and MDSC proportion in patients with colorectal cancer, we analyzed the proportion of CD11b⁺CD33⁺HLA-DR⁻ MDSCs in patient PBMCs and measured the netrin-1 concentration in plasma from patients with colorectal cancer. The MDSC proportion in PBMCs from patients with colorectal cancer was higher than that in healthy controls, which further indicated expansion and accumulation of MDSCs in peripheral blood from patients with colorectal cancer (Fig. 6A). Netrin-1 levels in plasma from patients with colorectal cancer were also increased compared with healthy controls (Fig. 6B). We also found a positive correlation between MDSC proportions and netrin-1 levels in patients with colorectal cancer (Fig. 6C). Consistently, GEPIA database analysis showed that the expression of Netrin-1 positively correlated with CD33 and ITGAM (encodes CD11b), markers of human MDSCs (Fig. 6D). Furthermore, we confirmed netrin-1 expression in colorectal cancer tissue. Western blot analysis and IHC images showed that netrin-1 was higher in colorectal cancer tissue compared with adjacent colorectal tissue (Fig. 6E and F). Taken together, these results indicate that netrin-1 accumulates, and has high-expression, in colorectal cancer tissues, suggesting that netrin-1 is partially responsible for reducing the accumulation and immunosuppressive activity of MDSCs in the tumor microenvironment and thus enhancing antitumor immunity.

Here, we identified the tumor cell-derived netrin-1 ligand and its cognate MDSC adenosine receptor A2BR as a ligand–receptor signaling axis regulating MDSC immunosuppressive function and tumor burden in vivo. This netrin-1/A2BR axis activated cAMP-PKA signaling and increased immunosuppressive molecules, such as Arg1 and ROS, which led to functional MDSCs enhancement. The netrin-1-A2BR axis represents a novel intercellular autonomous signaling network by which netrin-1 produced by cells in tumor niches binds to MDSCs to fine-tune MDSCs dynamics, in particular immunosuppressive activity (Fig. 7).

Abnormal myelopoiesis is a common phenomenon of tumor progression that results in the production of immunosuppressive myeloid cells, including MDSCs. MDSCs are considered the “queen bee” of the tumor microenvironment and supply immunosuppressive protection for tumors, promoting avoidance of immunosurveillance from the patient's immune system and immunotherapy (2). To date, MDSCs have become a promising target for cancer treatment due to their abundance in the tumor microenvironment, which represents a major obstacle to cancer immunotherapy. In fact, four strategies targeting MDSCs, including MDSC depletion, MDSC migration blockade, MDSC immunosuppressive function inhibition, and MDSC differentiation have been widely tested in preclinical and clinical studies (28, 29). However, the known functional crosstalk between MDSCs and tumors represents only the tip of the iceberg and seeking new MDSC targets for tumor therapy is becoming urgent and necessary.

Netrin-1 was first discovered to be a chemotropic cue secreted by the spinal cord, such as the floor plate, to guide the growth of commissural axons, and netrin-1 plays chemotactic or chemorepulsive roles dependent on the receptor expression profiles on target cells in the development of the nervous system (30). Outside neuromodulation, other important functions of netrin-1 and its receptors are gradually being revealed, such as cell apoptosis and survival, angiogenesis, and tumorigenesis (10, 31). The elevated netrin-1 autocrine activity in highly metastatic breast primary tumors has been considered an acquired selective advantage for tumor cells escaping netrin-1-dependent, receptor interaction-induced apoptosis (32). In colorectal cancer, previous studies have shown high netrin-1 expression in human colorectal tissues, which associates with a high inflammatory switch state and NF-κB activation (15, 33). Other studies also demonstrated selective netrin-1 expression in multiple human colorectal cancer cell lines, such as HCT8 and HCT116 (12, 34). However, whether tumor-derived netrin-1 affects stromal cells, such as MDSC function and accumulation, has not been elucidated. We found that netrin-1 was highly expressed in the murine colon adenocarcinoma cell line MC38 and enhanced the immunosuppressive function of MDSCs in vitro. Furthermore, netrin-1 blockade in TCCM treatment could inhibit Arg1 activity and ROS production in MDSCs. Consistently, in vivo experiments further showed that tumor MDSCs and splenic MDSCs exhibited poor suppressive molecule production, and MDSC infiltration reduction was observed in netrin-1 knockdown tumor-bearing mice.

A2BR is well known as one subtype of adenosine receptors (A1R, A2AR, A2BR, and A3R), which are G protein-coupled purinergic type 1 receptors (GPCR) that are regulated by extracellular adenosine (35). A previous study demonstrated that A2BR is involved in MDSC accumulation and expansion in tumor tissues (36). Apart from sensing adenosine, A2BR has also been shown to be a putative receptor of netrin-1, which is involved in the inflammatory response (37, 38). In our research, we found that netrin-1 interacts with A2BR and enhances the immunosuppressive activity of MDSCs. As expected, blocking A2BR activity with the A2BR antagonist PSB1115 or knockdown A2BR with siRNA in MDSCs reversed the effects of netrin-1 on MDSC-mediated T-cell suppression. Moreover, Arg1 activity and ROS production were decreased after A2BR blockade. A2BR, as an important GPCR, plays crucial roles in cAMP generation and downstream PKA/CREB activation. We found that netrin-1 treatment significantly enhanced the cAMP abundance in MDSCs, whereas PSB1115-pretreated MDSCs prevented the high cAMP production induced by netrin-1. A previous report shows that PGE2 and its receptor agonist also induce MDSC development by driving the cAMP/PKA/CREB pathway (39, 40). In our study, netrin-1 treatment did not affect PGE2 production in MDSCs, which suggests that netrin-1-induced cAMP generation is independent of PGE2 levels.

Circulating netrin-1 levels may contribute to the detection of cancers, such as colorectal cancer, advanced non–small cell lung cancer, and advanced gastric cancer (41–43). In addition, the level of serum netrin-1 positively correlates with clinical stage, prognosis, and recurrence of multiple tumors. Consistent with these published data, plasma netrin-1 levels were significantly increased in patients with colorectal cancer. There was a positive correlation between plasma netrin-1 levels and MDSC proportions in PBMCs from patients with colorectal cancer. However, the netrin-1 expression in patients with colorectal cancer remains controversial in several reports (41, 44). The possible reason may be associated with the relative expression of netrin-1 and receptors and DNA methylation in the tumor microenvironment. In our study, we found netrin-1 had high expression in patients with colorectal cancer, which correlated with MDSCs. In future, clinical evidence of netrin-1 expression and MDSCs in colorectal cancer tissues need to be collected and investigated.

Together, these findings provide novel insight into how tumor cells affect the immunosuppressive function of MDSCs by releasing netrin-1 and its interaction with A2BR, which could be a potential target for the treatment of tumors.

No author disclosures were reported.

X. Xia: Data curation, formal analysis, investigation, methodology, writing–original draft. Z. Mao: Resources, formal analysis, writing–review and editing. W. Wang: Formal analysis, investigation, methodology. J. Ma: Resources, formal analysis, funding acquisition. J. Tian: Resources, formal analysis, project administration. S. Wang: Conceptualization, supervision, funding acquisition, project administration, writing–review and editing. K. Yin: Conceptualization, formal analysis, project administration, writing–review and editing.

This work was supported by the Natural Science Foundation of Jiangsu (Grant No. BK20190242, BE2022779), Postgraduate Research & Practice Innovation Program of Jiangsu Province (Grant No. KYCX21_3403), Research Project of the Jiangsu Commission of Health (Grant No. K2019019), and Jiangsu Provincial Medical Key Discipline Cultivation Unit (Grant No. JSDW202241).

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.