Background
Colorectal cancer (CRC) is a very common cancer in the Western World, and in spite of improvements in surgery, chemotherapy and screening, it ranks in the second place regarding cancer-related deaths in this part of the world [
1]. The development of CRC occurs in a stepwise manner, developing from a benign preneoplastic lesion to a more metastatic disease that has a poor survival rate (11%), which is motivated by a series of genetic and epigenetic alterations [
2]. The escape of immune surveillance of tumor cells is a pitfall that cannot be overlooked, and the aim of immunotherapy is to recruit immune cells and to remove tumor cells by stopping tumor cells from escaping from immune surveillance [
3].
Interestingly, tumor- and immune cells-released extracellular vesicles (EVs) exert functional roles in immune processes, involving immune cell priming and activation, and immune escape under both local and systemic contexts [
4]. EVs are cells-derived particles ranging from 30 to 1000 nm in size, enclosed within a phospholipid bilayer [
5]. EVs from mesenchymal stem cells (MSCs) harbor a healing effect, reverting the malignant phenotype of CRC cells [
6]. The roles of EVs between intercellular communication is due to their capability of transferring proteins, lipids and nucleic acids, thus manipulating many physiological and pathological functions in recipient and parent cells [
7]. Because of their high abundance and their function as mediators of gene expression, microRNAs (miRNAs), small non-coding RNAs with 19–24 nucleotides, have been identified as potential markers in several cancer types, including CRC [
8]. A quantity of EV-miRNAs is linked to development or dismal overall survival of CRC [
9]. For instance, the plasma level of miR-30d-5p shuttled by EVs was enhanced in patients with metastatic CRC [
10]. Here in our study, miRNA-based microarray revealed that miR-222 was one of the most remarkably upregulated miRNAs in CRC cells co-cultured with MSC-derived EVs. miR-222 is located on the human chromosome Xp11.3 and plays significant parts in the modulation of a broad spectrum of cancers [
11]. In the context of CRC, miR-222 overexpression contributed to promoted cell migration and invasion [
12]. Additionally, exosomal miR-222-3p was significantly enhanced in patients with lymph node metastasis, indicating its potential predictive roles in papillary thyroid cancer [
13]. However, the relevance of miR-222 from MSC-EVs to the progression, especially to the immune escape of CRC remains unclear. Therefore, we examined if miR-222 derived from MSC-EVs is involved in immune evasion in CRC and the potential mechanism of action.
Methods
Ethics approval
The study was performed as per the Declaration of Helsinki and ratified by the Ethics Committee of The First Hospital of Jilin University. The patient enrolled for MSC extraction signed an informed consent before enrollment. The animal experimental protocol was approved by the Committee on the Ethics of Animal Experiments of The First Hospital of Jilin University. All animal procedures were performed in line with the Guide for the Care and Use of Laboratory Animals issued by the National Institutes of Health (Bethesda, MA, USA).
Human sample
Fresh tumor tissues were collected from a patient with CRC admitted to The First Hospital of Jilin University on September 13, 2019. The patient was 53 years old and free of a history of other chronic diseases or any other cancers, and had not received radiotherapy or chemotherapy prior to treatment at The First Hospital of Jilin University. The patient was diagnosed with CRC at stage II without lymph node metastasis by colonoscopy and tissue biopsy. We surgically removed the CRC tissues and gave adjuvant chemotherapy to the patient after the surgery. The prognosis of this patient is now good. The excised tissue samples were soaked in 95% ethanol to prevent contamination, and then washed in phosphate-buffered saline (PBS) containing 1% penicillin/streptomycin. The samples were cut into thin slices and detached in type IV collagenase (Thermo Fisher Scientific Inc., Waltham, MA, USA) for 3 h at 37 °C. After detachment, the tissues were rinsed with PBS and passed through a 70 μm sieve (Corning Glass Works, Corning, N.Y., USA). We centrifuged the filtrate and cultured the cells in erythrocyte lysis buffer (Sigma-Aldrich Chemical Company) to eliminate erythrocytes. Cells were grown in low-glucose DMEM (Thermo Fisher) containing 10% fetal bovine serum (FBS) and penicillin/streptomycin at 37 °C, 5% CO2 in a cell culture incubator, with the medium renewed every 2 days. The passages were performed at appropriate times. After the fourth passage, MSCs were observed under the microscope for identification and used in subsequent experiments.
Isolation and characterization of EVs
The cell culture medium was firstly subjected to continuous centrifugation at 300 g for 5 min, at 1200 g for 20 min, and at 10,000 g for 30 min to discard cells and cell debris, followed by a centrifugation at an ultra-high speed of 100,000 g for 60 min at 19 °C using a Sorvall WX Ultra series centrifuge in an F50L-2461.5 rotor (Thermo Fisher). The resulting precipitate was washed with PBS and ultracentrifuged again at 100,000 g for 1 h, and the obtained vesicles were resuspended in Roswell Park Memorial Institute-1640 medium for identification.
EVs were identified by the nanoparticle tracking analysis (NTA) system (NTA 3.2 Dev Build 3.2.16, Malvern Panalytical Ltd., UK). The Brownian motion of the EVs was irradiated by a laser beam and recorded by a camera. NTA was converted by the Stokes-Einstein equation to the size distribution of EVs, which was measured in triplicate. For transmission electron microscopy (TEM), the EVs were fixed with 2% paraformaldehyde and loaded onto carbon-coated copper grids. The grids were placed on 2% gelatin at 37 °C for 20 min and rinsed with 0.15 M glycine in PBS. The morphology of EVs was viewed under a Philips CM120 TEM (Philips Research, Eindhoven, The Netherlands). The expression of EVs-specific markers, tumor susceptibility gene 101 (TSG101) and CD81, was measured by Western blot. Additional file
1 (Supplementary Table S1) presents related antibodies.
Cell culture and treatment
The cell lines SW480 (CCL-228), HCT116 (CCL-247), 293 T (CRL-3216) were from the American Typical Culture Collection (Manassas, VA, USA). NCM460 cells (MZ-0658) were from Mingzhoubio (Ningbo, Zhejiang, China). To avoid cell contamination, STR genotyping was performed on all cells (including primary MSCs) during the first week of cell culture to confirm cell purity. Mycoplasma contamination was detected in the cells by isolation culture method, and then identified every 2 months during the experiment. The results showed no cross-contamination or mycoplasma contamination in the cells. CRC cells were incubated in complete DMEM containing 10% FBS (Thermo Fisher), 100 mg/mL penicillin and 10 mg/mL streptomycin (Thermo Fisher) for 48 h. The extracted 100 μg/mL EVs were added to the medium and incubated for 24 h. The miR-222 inhibitor, small interfering RNA (si) targeting ATF3 (si-ATF3) and AKT1 (si-AKT1), and their respective negative controls (miR-222 control and NC) were generated by GenePharma (Shanghai, China). The Cy3-labeled miRNA mimic was produced by GE Dharmacon. miR-222 inhibitor, si-ATF3, and si-AKT1 were transfected into EVs-treated CRC cells using Lipofectamine 2000 (Thermo Fisher) according to instructions. Cells were incubated until stable after transient transfection.
Immunofluorescence staining
After transfection with Cy3-miR-222 mimic, MSCs (1 × 106 cells/well) were co-cultured with CRC cells at a 1:1 ratio using Transwell plates (0.4 mm polycarbonate filter, Corning) for 12 h. CRC cells were placed in the basolateral chamber, and MSCs in the apical chamber. The cells were then fixed with 4% paraformaldehyde at 4 °C for 15 min, incubated with 0.5% Triton-100 X for 20 min, and sealed with anti-fluorescence quenching sealant Vectashield (Vector Laboratories Inc., Burlingame, CA, USA). The presence of Cy3 red fluorescence in CRC cells was observed by fluorescence microscopy (BX63, Olympus Optical Co., Ltd., Tokyo, Japan).
5-Ethynyl-2′-deoxyuridine (EdU) labelling
The cells in logarithmic growth period were plated at 1.6 × 105 cells/well and cultivated in 96-well plates for 2 d. EdU (50 mM, Cell-Light EdU Apollo 488 kit, Guangzhou RiboBio Co., Ltd., Guangzhou, Guangdong, China) was supplemented to the cells for a 4-h incubation at 37 °C. The cells were treated with 4% formaldehyde solution for 15 min, with 0.5% Triton X-100 for permeabilization, with 100 mL Apollo Mix for about 30 min at ambient temperature and stained in 100 mL Hoechst33342 staining solution for 30 min before being viewed under a fluorescence microscope (BX63, Olympus). To measure the proportion of EdU-positive cells (red), the EdU-positivity rate was calculated by Image-Pro Plus 6.0 software (Media Cybernetics, Bethesda, MD, USA).
Microarray analysis
Gene expression analysis of SW480 and HCT116 cells was conducted before and after EV treatment. RNA was isolated from cells using TRIzol reagent (Thermo Fisher) and reversely transcribed into complementary DNA (cDNA) using a Superscript reverse transcriptase kit (Transgene Biotech, Beijing, China). cDNA was hybridized with Human miRNA Expression Microarray V4.0 (Arraystar Inc., Rockville, MD, USA) and GeneChip™ Human Gene 1.0 ST Array (Thermo Fisher). Gene expression data were obtained after incubating the hybridization microarray with DNA in an incubator for 24 h with a GeneChip™ Scanner 3000 7G system (Thermo Fisher), and the resulting data were analyzed by R-project. Affy (Bioconductor) was used for normalization and quality control of expression data, and Pheapmap (Bioconductor) was used to screen differentially expressed genes at |Log2FoldChange| > 1, p < 0.01 and to plot the heatmap.
Reverse transcriptase quantitative PCR (RT-qPCR)
After isolation using TRIzol reagent (Thermo Fisher), the RNA was reversely transcribed into cDNA with the help of a Superscript reverse transcriptase kit (Transgene) according to standard instructions. qPCR was conducted on an ABI7300 real-time PCR system (ABI7300, USA) using the Super SYBR Green kit (Transgene). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as an endogenous control for ATF3 and AKT1, and U6 as an endogenous control for miR-222. PCR primers were synthesized by Sangon (Shanghai, China). The specific primer sequences are as follows: miR-222: forward, 5′-CCCTCAGTGGCTCAGTAG-3′, reverse, 5′-CCACCAGAGACCCAGTAG-3′; ATF3: forward, 5′-CTCTGCGCTGGAATCAGTCA-3′, reverse, 5′-CCTCGGCTTTTGTGATGGA-3′; AKT1: forward, 5′-TCCTCCTCAAGAATGATGGCA-3′, reverse, 5′-GTGCGTTCGATGACAGTGGT-3′; U6: forward, 5′-CTCGCTTCGGCAGCACA-3′, reverse, 5′-AACGCTTCACGAATTTGCGT-3′; GAPDH: forward, 5′-AGTGGCAAAGTGGAGATT-3′, reverse: 5′-GTGGAGTCATACTGGAACA-3′.
Transwell assay
Transwell migration and invasion analyses were performed in 24-well Transwell chambers (0.8 μm pore size, Millipore Corp, Billerica, MA, USA). Transfected CRC cells were trypsinized with 0.25% trypsin and resuspended in FBS-free DMEM. Cells were added to the apical chamber of the Transwell, and DMEM plus 10% FBS was supplemented to the basolateral chamber. Cells were removed after 2 d of incubation. After fixation with 4% paraformaldehyde for 30 min at ambient temperature, cells on the subsurface of the membrane were stained with 0.1% crystal violet overnight. Cells from the bottom of the apical chamber were counted by a microscope (Axiolab 5, Carl Zeiss, Oberkochen, Germany) in four randomly selected areas. For invasion experiments, BD Matrigel (1:8, Corning) was used for the apical chamber coating at a ratio of 30 μL per well.
Flow cytometry
Apoptosis of CRC cells was evaluated using flow cytometry and the Annexin-V-FLUOS staining kit (Roche Diagnostics, Co., Ltd., Rotkreuz, Switzerland). Cells at the logarithmic growth stage were collected at 48 h post-transfection and resuspended in binding buffer. Apoptotic cells were stained with Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) for 15 min at ambient temperature with light avoidance. Fluorescence signals were collected by FACSCanto (BD Bioscience, San Jose, CA, USA) and analyzed using FlowJo 8.7.1 software (Ashland, OR, USA).
The surface markers of MSCs were also identified by flow cytometry. The cells were detached with 0.25% trypsin once at an 80% cell confluence, and centrifuged to prepare a single cell suspension. After the cell concentration was adjusted to 1 × 106 cells/mL, 100 μL single cell suspension was reacted with 20 μL human monoclonal antibodies to CD73, CD90, CD105, CD14, CD19, and CD45 (Beyotime Biotechnology Co., Ltd., Shanghai, China) for 30 min, followed by treatment with FITC-labeled secondary antibody (Roche) for 30 min. Fluorescence signals were collected by FACSCanto (BD Bioscience) and analyzed by FlowJo 8.7.1 software.
Tumor xenografts in nude mice
Thirty 4-week-old female BALB/c nude mice (15 ± 2.52 g) were from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The transfected CRC cells (miR-222 control, miR-222 inhibitor, miR-222 inhibitor + NC, miR-222 inhibitor + si-ATF3, si-ATF3 + NC, and si-ATF3 + si-AKT1) were injected subcutaneously into mice at 1 × 107 cells per mouse (n = 5), and the mouse tumor volume was detected every 7-d using a vernier caliper. Tumor volume changes in nude mice were recorded based on the formula tumor = L × W2/2. After 28 d, mice were euthanized by intraperitoneal injection of 1% sodium pentobarbital at 150 mg/kg, and tumors were resected and photographed with weights measured. Following euthanasia, animal death was confirmed by observing the lack of heartbeat, respiratory arrest, pupil dilation, and lack of nerve reflex.
Immunohistochemistry
Mouse tumor tissues were routinely embedded, de-waxed and hydrated. The tissues were treated with 3% H
2O
2 for 15 min at room temperature, followed by another 15-min treatment at room temperature with normal goat serum sealant (Beijing Solabio Life Sciences Co., Ltd., Beijing, China). Afterwards, the tissues were probed with CD3 monoclonal antibody at 4 °C overnight, then with the secondary antibody at 37 °C for 15 min. The diaminobenzidine reaction was carried out by an incubation with 40 μL horseradish-labeled Streptomyces ovalbumin working solution (Solarbio) for 15 min, and then the sections were dehydrated and sealed after a 30-s hematoxylin counter-staining. Microscopy (Axiolab 5, Zeiss) was utilized to observe and count CD3-positive cells in sections. See Additional file
1 (Supplementary Table S1) for antibody information.
In situ hybridization
The paraffin-embedded tissue sections were dewaxed, hydrated and subjected to a 10-min treatment with 15 mg/mL proteinase K (Exiqon, Denmark) at 37 °C. Double-digoxigenin was utilized to label the miR-222 probe, and tissue sections were immersed in 50 μL hybridization solution and hybridized with the probe (500 ng/mL) at 37 °C for 18 h. The sections were washed with sodium citrate in saline solution, incubated in a sealant containing 2% goat serum (Solarbio) for 4 h at room temperature. Finally, the sections were stained with Nuclear Fast Red (Sangon) for 1 min and visualized by Aperio Scanscope Virtua (Aperio Scanscope FLGL, Aperio) to determine the miR-222 expression.
3’untranslated region (3’UTR) luciferase reporter assays
The target genes of miR-222 were predicted using StarBase (
http://starbase.sysu.edu.cn/index.php). The fact that ATF3 is a direct target of miR-222 was confirmed using a dual-luciferase reporter gene assay. The wild-type (WT) sequences (GenePharma) with miR-222 binding sites and mutant (MT) sequences in the mRNA 3’UTR were synthesized artificially, and the pmiR-RB-REPORT-ATF3–3’UTR plasmid (RiboBio) was treated with restriction enzymes. The target gene fragments were synthesized and inserted into the pmiR-RB-REPORT vector with miR-222 mimic, respectively. HEK293T cells were collected at 48 h post-transfection, and fluorescence intensity was measured using a luciferase assay kit (Beyotime).
Chromatin immunoprecipitation (ChIP)
The binding sites between ATF3 to AKT1 were predicted using JASPAR (
http://jaspar.binf.ku.dk/). Cells were incubated in 37% formaldehyde at 37 °C for 10 min, collected and added to sodium dodecyl sulfate (SDS) lysis buffer and protease inhibitor complex (Thermo Fisher) for ultrasonic fragmentation. A Pierce Agarose ChIP Kit (Thermo Fisher) was applied. The cells were centrifuged at 10,000 g for 10 min at 4 °C, and reacted with 900 μL ChIP dilution buffer, 20 μL 50x protease inhibitor cocktail and 60 μL ProteinA agarose at 4 °C for 60 min. The mixture was then probed with antibodies at 4 °C, and the precipitated complexes were washed with 60 μL ProteinA agarose and 250 μL eluent. The supernatant was de-crosslinked by adding 20 μL NaCl (5 M) and eluted again by adding 500 μL eluent. After another 1-h incubation with 1 μL RNaseA at 37 °C, DNA fragments were recovered for PCR analysis. Antibody information is exhibited in the Additional file
1 (Supplementary Table S1).
Western blot
Cells or EVs were homogenized in radio-immunoprecipitation assay lysis buffer (Roche) supplemented with protease inhibitors. Protein concentration measurement was carried out using a bicinchoninic acid protein assay kit (Beyotime). The cultured cells were harvested after an 800-g centrifugation at 4 °C for 5 min, ice-bathed with 5x lysis solution for 10 min, followed by a 10-min centrifugation at 12,000 g and 4 °C. The supernatant was separated by SDS-polyacrylamide gel electrophoresis and trans-blotted to polyvinylidene fluoride membranes (Millipore). The membranes were sealed with 5% skimmed milk powder and probed with primary antibodies for 16 h, followed by the incubation with secondary antibodies for 2 h at 37 °C. See the Additional file
1 (Supplementary Table S1) for all antibody information. Optical density (OD) value measurement was conducted using ImageJ software (version 1.8.0; NIH).
Statistical analysis
All quantitative results were obtained from triplicate assays and analyzed using GraphPad Prism 6 (GraphPad, San Diego, CA, USA). Data were exhibited as the mean ± standard deviation (SD). The significant difference (p < 0.05) between groups was determined by the independent t test and one-way or two-way analysis of variance (ANOVA) among multiple groups, along with Tukey’s post hoc test.
Discussion
Distinctive characteristics of MSCs make them highly promising in the cell-based therapy of cancers, and MSCs have the potency to suppress the immune system and support tumor cells to escape from immune responses [
14]. On the other hand, tumor-derived EVs have been implicated in different events, including angiogenesis, chemoresistance as well as immune evasion, and their role has also been well-established in biological pathways involved in CRC initiation and progression [
15]. In this study, we identified a novel regulatory mechanism expediting immune escape of CRC cells. We found that MSC produce EVs containing miR-222 to potentiate CRC cell malignant phenotype. miR-222 targets ATF3 in CRC cells and promotes immune escape of CRC cells by activating the AKT pathway.
The first finding of this study was that MSC-EVs encourage CRC cells to grow and proliferate. Ramírez-Ricardo et al. proposed that circulating EVs from patients with breast cancer could elevate migration and invasion of breast cancer cells [
16]. In addition, tumor-derived EVs may contribute to favored tumor aggressiveness through both direct and indirect manners and are involved in tumor immune escape [
17]. After that, we conducted microarray analysis to detect differential changes in miRNAs, followed by experimental validation. The upregulation of miR-222 in CRC cells was revealed as a result of MSC-EVs transfer. Further functional experiments were carried out in CRC cells with miR-222 knockdown. In addition to hampering malignant aggressiveness in vitro, miR-222 was also found to reduce tumor growth and immune escape in vivo. Consistent with our findings, endoplasmic reticulum stress-evoked exosomal miR-27a-3p promoted immune escape in breast cancer [
18]. As regards to the function of miR-222, increased miR-222-3p expression was associated with metastasis and a poor prognosis in renal clear cell carcinoma [
19]. In addition, miR-222 was enriched in retinoblastoma tissues and cells, which facilitated resistance of retinoblastoma cells to vincristine, a chemotherapeutic agent [
20]. Under the context of CRC, HCT116 and SW480 cells illustrated repressed invasion and migration abilities and enhanced apoptosis in response to miR-222-3p inhibitor [
21]. During the malignant transformation of normal colorectal epithelial cells, CCR5, FasL and HLA-E expression elevated remarkably, whereas Fas expression reduced [
22]. By contrast, we observed that miR-222 inhibitor diminished HLA-A, CCR5, FasL and HLA-E expression, while restored Fas expression, implying that that the immune escape was prevented. Likewise, Cojo et al. believed that upregulation of hsa-miR-222 could protect against apoptosis in HIV-infected CD4
+ T cells [
23].
In this study, we also noted that miR-222 had a binding relationship with ATF3, and ATF3 was not only a target of miR-222, but also downregulated in CRC cells after MSC-EVs treatment. Interestingly, overexpression of ATF3 was linked to good survival rates in CRC patients, and silencing of ATF3 promoted proliferation, migration, and clonogenic growth of CRC cells [
24]. Consistently, our rescue experiments disclosed that downregulation of ATF3 reversed the inhibitory effects of miR-222 knockdown on CRC cell growth and immune escape. Kim et al. reported that ATF3, a member of the ATF/CREB family of transcription factors, is tightly related to apoptosis in CRC cells with the involvement of many signaling components, including AKT [
25]. Moreover, the depletion of ATF3 promoted activation of the AKT signaling, evidenced by higher extent of AKT phosphorylation, to accelerate prostate cancer development [
26]. Our bioinformatics analysis and ChIP assays provided evidence that ATF3 directly bound to and shared a negative correlation with AKT1 in CRC cells. Further western blot analyses established that miR-222 inhibitor resulted in the AKT pathway deficit, while si-ATF3 led to the AKT pathway activation. Also, high-mobility group A1 expedited uveal melanoma progression via the PI3K/AKT pathway and oncogenic miR-222 [
27]. In the same vein, tumor-secreted exosomal miR-222 facilitated pancreatic cancer progression by potentiating the AKT pathway [
28]. Additional rescue experiments in our study indicated that AKT1 knockdown reversed the supporting role of si-ATF3 in tumor growth and immune escape in CRC.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.