Forkhead box protein 3 (FOXP3) is implicated in tumor progression and prognosis in various types of tumor cells. We have recently reported that FOXP3 inhibited proliferation of gastric cancer (GC) cells through activating the apoptotic signaling pathway. In this study, we found that over-expression of FOXP3 inhibited GC cell migration, invasion and proliferation. Then, the label-free quantitative proteomic approach was employed to further investigating the down-stream proteins regulated by FOXP3, resulting in a total of 3,978 proteins quantified, including 186 significantly changed proteins. Caveolin-1 (CAV1), as a main constituent protein of caveolae, was one of those changed proteins up-regulated in FOXP3-overexpressed GC cells, moreover, it was assigned as one of the node proteins in the protein-protein interaction network and the key protein involved in focal adhesion pathway by bioinformatics analysis. Further biological experiments confirmed that FOXP3 directly bound to the promoter regions of CAV1 to positively regulate CAV1 transcription in GC cells. In summary, our study suggested that FOXP3 can be considered as a tumor suppressor in GC via positively regulating CAV1 through transcriptional activation, and this FOXP3-CAV1 transcriptional regulation axis may play an important role in inhibiting invasion and metastasis of GC cells. Data are available via ProteomeXchange under identifier PXD007725.
Gastric cancer (GC) is a common malignant tumor worldwide1. Although the incidence of GC is declining in most of developed countries, it is still one of the most common causes of cancer-related death in many Asian countries. In China, gastric cancer is the third and second cause of death for both men and women and the main cancers in rural areas. With an incidence of 42.92/105 and a mortality of 29.93/105 2, gastric cancer is still a great threat to people’s health in China.
The gene of FOXP3 is located on the short arm of the X chromosome at Xp.11.233. FOXP3 is a transcription factor that is necessary for induction of the immunosuppressive functions of regulatory T cells (Tregs), and it is a well-known marker of Tregs4. Recently, FOXP3 is reported to be expressed in various kinds of tumor cell including colorectal cancer5, melanoma6, non-small cell lung cancer7 and many other cancer cell lines8. Suh et al. found that tumoral FOXP3 expression is associated with favorable clinicopathological variables in gastric adenocarcinoma, and FOXP3 is associated with the Hippo pathway proteins, Lats2 and YAP expression9. Hao et al. demonstrated that FOXP3 was dominantly expressed in GC cells, and FOXP3 can act as a negative regulator of NF-κB activity to play a tumor suppressor role by reducing GC cell metastasis10.
In our previous studies, we have demonstrated that FOXP3 suppressed the GC cells proliferation by induction of GC cells apoptosis, and positive histological staining of FOXP3 in GC cells indicated better outcome11,12. The experimental results of FOXP3 in our previous studies were consistent with other findings in breast cancer and gastric cancer9,13. However, there is a huge difference for clinic treatment between early gastric cancer and metastatic gastric cancer. Hence, our study aimed to investigate the effects of FOXP3 on GC cell adhesion, migration and invasion and the mechanisms behind.
Label-free proteomic analysis is a recently emerged, efficient, powerful, and cost-effective approach for comparing multiple samples from different cells/tissues. In our previous studies, label-free proteomic analysis was applied to analyzing clinic and animal samples, such as endometrial tissues in proliferative and receptive phases14, omental adipose tissues from patients with gestational diabetes mellitus15, colonic biopsies from endoscopy16, and mouse corneal tissues17. Proteomic strategies were also employed to investigating the function of FOXP3. Rudra et al. purified FOXP3 complexes and explored their composition with mass spectrometry-based proteomics and identified 361 associated proteins, ~30% of which are transcription-related18. Kubach et al. reported a 2-dimensional gel electrophoresis (2D PAGE)-based proteomic work, and galectin-10 was identified as a novel marker and essential for CD4+ CD25+ FOXP3+ regulatory T cells anergy and suppressive function19.
In this study, overexpression of FOXP3 significantly inhibited cell migration, invasion and proliferation. Then label-free proteomic experiments were performed to analyze the FOXP3-overexpressed AGS cells and vector cells, resulting in a total of 3978 proteins identified, of which 186 proteins were significantly changed (fold change >1.5, students’ t test p value <0.01) between these two types of cells. The expression of CAV1 was significantly increased in AGS cells and played a central role in the protein-protein interaction networks constructed by these significantly changed proteins. Moreover, the results from KEGG pathway analyses enlightened the significant changes in the focal adhesion pathway, and CAV1 played an important biological role in this pathway. ChIP-PCR and luciferase assay were applied for biological validation, confirming that CAV1 was a direct transcription target of FOXP3. These results indicated that CAV1 may play an important role as the downstream transcriptional target of FOXP3 in regulation the adhesion, migration and invasion functions of GC cells.
Human GC cells (AGS) were obtained from the Cell Culture Center of Institute of Biochemistry and Cell Biolgy, Chinese Academy of Science (Shanghai, China). The AGS cells were maintained in Rosewell Park Memorial Institute (RPMI) 1640 medium (Hyclone, USA) supplemented with 10% fetal bovine serum (Invitrogen, USA) at 37 °C in a humidified atmosphere of 5% CO2.
For cell transfection, AGS cells were grown to 70 to 80% confluence in six well plates. Transfection was performed using lentivirus with a multiplicity of infection (MOI) of 50 (Shanghai GenePharma Co.,Ltd. China). The lentiviral vector containing FOXP3-specific gene was cotransfected with packaging vector pVSV-G into 293 T cells. The supernatant which contains virus was collected and added onto AGS cells 48 h after cotransfection. Cells stably expressing FOXP3 were selected using 2.5 μg/mL puromycin. Overexpression of FOXP3 was determined by using western blot to quantify the level of FOXP3 protein. The FOXP3-overexpressing cells were named AF and the vector-transfected cells were named ANC.
1 × 107 of cells were washed with ice-cold PBS three times, then lysed in SDT buffer containing 4% SDS (m/w), 100 mM DTT, 100 mM Tris, pH 7.6. The lysate was incubated at 95 °C for 5 min and then centrifuged at 15,000 g for 10 min, and the supernatant was used for the proteomics sample preparation. 100 μg of protein sample was digested using FASP method as previously described20. Each sample was mixed with 200 μL of UA buffer (8 M urea, 0.1 M Tris–HCl, pH = 8.5), loaded on a 10k Microcon filtration device (EMD Millipore, Billerica, MA, USA) and centrifuged at 14,000 g for 40 min at 20 °C. The concentrates were diluted in the device with 200 μL of UA solution and centrifuged again under the same conditions. The concentrate was then mixed with 100 μL of 50 mM indole acetic acid (IAA) in UA buffer and incubated for an additional 40 min at room temperature in darkness. After this, the IAA was removed by centrifugation at 14,000 g for 20 min. Next, the sample was diluted with 200 μL of UA buffer and centrifuged twice. Subsequently, 200 μL of 50 mM NH4HCO3 was added and the sample was centrifuged at 14,000 g for 40 min; this step was repeated twice. Finally, 100 μL of 5 0 mM NH4HCO3 and trypsin (1:50, enzyme to protein) were added to the sample, which was then incubated at 37 °C for 16 h. The tryptic peptide mixtures were collected for further analysis.
The reverse phase high performance liquid chromatography (RP-HPLC) separation was achieved on the EASY-nLC1000 HPLC system (Thermo Fisher Scientific,Grand Island, NY, USA) using a self-packed column (75 μm × 150 mm; 3 μm ReproSil-Pur C18 beads, 120 Å, Dr. Maisch GmbH, Ammerbuch, Germany) at a flow rate of 300 nL/min using 240 min gradients. The MS data were acquired on an Orbitrap Elite Hybrid Ion Trap-Orbitrap Mass Spectrometer (Thermo Fisher scientific, Grand Island, NY, USA) platform in a data dependent strategy, selecting the top 20 CID fragmentation events based on the precursor abundance in the survey scan (350–1600 m/z). For accurate mass measurements, the lock-mass option was employed21. The full mass was scanned in the Orbitrap analyzer with R = 60,000 (defined at m/z 400), and the subsequent MS/MS analyses were performed with R = 15,000.
The MS data were analyzed using the software MaxQuant22 (https://maxquant.org/, version 220.127.116.11). Carbamidomethyl (C) was set as a fixed modification, and oxidation (M, +15.99492 Da) was set as a variable modification. Proteins were identified by searching the MS and MS/MS data for the peptides against a decoy version of the International Protein Index (IPI) human database (version 3.87, 91,491 protein sequences; European Bioinformatics Institute). Trypsin/P was selected as the digestive enzyme with two potential missed cleavages. Protein abundance was calculated according to the normalized spectral protein intensity (LFQ intensity). The MS proteomics data have been deposited into the ProteomeXchangeConsortium23 via the PRIDE partner repository with the dataset identifier PXD007725.
Further analysis of the MaxQuant output protein quantitation data was performed using the Perseus software (version 18.104.22.168)24. First, hits to the reverse database, contaminants were eliminated. Then the LFQ intensities were logarithmized, imputed with random numbers from a normal distribution (width = 0.3, shift = 1.8), and sent to Student’s t-test analysis. Hierarchical clustering analysis was performed with the pheatmap package, which is based on the open-source statistical language R25, using Euclidean distance as the distance metric and complete method as the agglomeration method. A strict criterion including protein fold change >= 1.5 and p value of t-test analysis <= 0.01 was used for significant protein screening. The significant proteins were searched against the STRING database (version 10)26 for protein-protein interactions. STRING defines a metric called “confidence score” to define interaction confidence; we fetched all interactions which had a confidence score ≥ 0.7 (high confidence). The resulting interaction network had 64 nodes and 75 interactions. The network was visualized with Cytoscape software27 and protein regulation information was indicated as node color. Pathways enriched with significantly changing proteins were determined using a homemade pathway mapping tool based on the KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway database28. The enrichment within a given pathway was assessed by the hypergeometric distribution.
Total RNA was isolated from transfected and non-transfected cells using a commercial RNA isolation kit (TaKaRa Biotechnology, Japan) according to the manufacturer’s instructions. RNA (2 μL) from each sample was reverse transcribed to cDNA. The total volume was 10 μL. The complementary DNA reverse transcribed from the RNA was amplified by a Taq polymerase using the following primers for the FOXP3 cDNA: 5′-AAGCAGCACTACATTGACCTGAAA-3′ (forward) and 5′-GGTCTCCCCAAGCATCACTC-3′ (reverse). CAV1: 5′-GGGCATTTACTTCGCCATT-3′ (forward) and 5′-TGGAATAGACACGGCTGATG-3′ (reverse); GAPDH was also amplified as a control, using the following primers: 5′-GCACCGTCAAGGCTGAGAAC-3′ (forward) and 5′-TGGTGAAGACGCCAGTGGA-3′ (reverse). The amplification reaction was initiated by denaturing DNA at 95 °C for 5 min, followed by 30 cycles of template denaturing at 94 °C for 1 min, primer annealing at 60 °C for 1 min, and primer extension at 72 °C for 1 min. The qRT-PCR was applied using an ABI PRISM 7500 System (Perkin-Elmer, USA).
Protein expression in AGS cells was validated by western blotting. Cultured cells were collected and extracted using SDT buffer. The proteins were separated by molecular weight using 10% SDS-PAGE and then transferred to 0.22 μm polyvinyldifluoride membrane (PVDF) membrane (EMD Millipore, Billerica, MA, USA) at 100 V for 120 min. The blots were blocked with 5% BSA in Tris-buffered saline-Tween 20 (TBST buffer), followed by incubation with primary antibodies against GAPDH (Sigma aldrich, Saint Louis, MI), FOXP3(Abcam, Cambridge, United Kingdom) and CAV1 (Cell Signaling Technology, Danvers, MA, USA) overnight. The blots were washed three times with TBST and were incubated with the corresponding horseradish peroxidase conjugated secondary antibody (Santa cruz, USA) for 1 h. The signals were detected using an enhanced chemiluminescence (ECL) solution (Thermo Fisher scientific, Grand Island, NY, USA), and were captured by a gel imaging system (Amersham Imager 600; GE Healthcare, Waukesha, WI, USA). GAPDH protein was used as a loading control to normalize the Western blot data, and the relative values were calculated by the intensity of FOXP3 or CAV1 divided by the intensity of the loading control.
The motility of AGS cells was assessed using a wound-healing assay. AF cells and ANC cells were seeded in six well plates at a density of 105 cells/well in complete RPMI 1640. After incubation overnight to yield confluent monolayers for wounding, and washing the cells twice with PBS, 200 μL pipette tips were used to make wounds (4–5 parallel scratches/plate); subsequently, the cells were incubated with serum-free 1640 medium. Photographs were taken immediately (0 h) and 24 h after wounding from five randomly selected fields in each well. The distance migrated by the cell monolayer to close the wounded area during this time period was measured using ImageJ.
AF cells and ANC cells were seeded at an original density of 200 cells per 6 mm dish in a complete RPMI 1640 medium. After incubation at 37 °C for 7 days, adherent cells were washed twice and fixed with 4% paraformaldehyde for 30 min. They were then stained with methyl violet and washed twice with double distilled water….