{"paper_id":"ee33b162-e3f9-4ca6-a3e3-516d9362ad7d","body_text":"Metastasis is associated with advanced tumors and is the main cause of cancer-related death. It occurs in all malignant tumors and in certain benign diseases, such as endometriosis and tuberous sclerosis complex. 1 , 2  It is an integrated process that involves several biological mechanisms, such as digestion of the basement membrane and extracellular matrix (ECM), separation from the primary tumor, penetration into a blood vessel through transendothelial migration, and proliferation at distant organs. 3 , 4  Invasion and migration play a crucial role in this multistep process and are central features of cancer cells. They can contribute to metastasis individually or in combination. 5 , 6  For instance, invasion triggers metastasis by providing space and direction for cancer cell migration by digesting the basement membrane and the ECM. Since they are oriented to a blood vessel or lymph vessel, cancer cells can penetrate them by changing their morphology. 5  Since invasion and migration function simultaneously at critical steps of metastasis, they should be investigated together to obtain further insight.\nMany genes and pathways are involved in invasion, migration, and metastasis; specifically, matrix metalloprotease 2 ( MMP2 ) and Ras-related C3 botulinum toxin substrate 1 ( RAC1 ) play a pivotal role in invasion and migration. 7 , 8 \n MMP2  is a member of the MMP family that functions in ECM degradation.  MMP2  can degrade the basement membrane by targeting its most abundant component, type IV collagen. 9  Similarly,  RAC1  plays an important role in migration. 10  It is a member of the  RAC  subfamily and promotes F-actin polymerization to reorganize the cytoskeleton, which functions in migration. 11 , 12 \n MMP2  and  RAC1  activity is negatively regulated by  WNK2 , a member of the WNK (with no K = lysine) family of protein kinases. 13 , 14  Outside of conventional gene mutations, epigenetic mechanisms play an important role in the regulation of  WNK2  expression. 15\nChromobox protein homolog 8 ( CBX8 ), together with  CBX2 ,  CBX4 ,  CBX6 , and  CBX7 , forms the CBX family, which is the central part of the polycomb repressive complex 1 (PRC1). 16  PRC1 and PRC2 can epigenetically regulate genes by methylating histones, such as di- or tri-methylation of lysine 27 on histone H3 (H3K27me2/me3), which results in chromatin remodeling and silencing of genes. 17 , 18  However,  CBX8  can also regulate gene expression in a PRC1-independent manner. 19 , 20  Recent studies show that increased expression of  CBX8  is associated with many types of cancer, though these findings require additional characterization and functional analyses. 21 ,  22 ,  23  Here, we explored the functional mechanism of  CBX8  in cancer to determine its association with metastasis.\n\nTo explore the expression pattern of  CBX8  in cancer, we performed a literature search on The Cancer Genome Atlas database and compared  CBX8  expression levels between tumor tissues and normal tissues. The results showed that  CBX8  was expressed at low levels in most normal tissues except the testis, pituitary, and fibroblasts. However,  CBX8  was overexpressed in most cancers compared with corresponding normal tissues ( Figure 1 A). Considering the  CBX8  expression profile and clinical implications, we further examined its expression in brain tumors (data published in Tang et al. 24 ), breast cancer, and lung cancer. In each group,  CBX8  expression was significantly higher in tumor tissues than in normal tissues ( Figure S1 ), and glioma patients with higher  CBX8  expression displayed significant poorer overall survival than those with lower  CBX8  expression ( Figure 1 D).  CBX8  expression was higher in breast cancer and lung cancer cell lines than in non-cancer cell lines ( Figures 1 B and 1C). Taken together, these data indicated that  CBX8  has a potential role in tumorigenesis. Figure 1 CBX8  Expression Is Upregulated in a Variety of Cancers (A)  CBX8  shows higher mRNA-seq expression profile comparing multiple human cancer tissues (in red color) versus normal tissues (in blue) as well as gray (missing TCGA normal sample type). The figure was generated from the COH Bioinformatics Core using Broad Institute TCGA Genome Data Analysis Center (2016): Analysis-ready standardized TCGA data from Broad GDAC Firehose 2016_01_28 run; Broad Institute of MIT and Harvard. Dataset ( https://doi.org/10.7908/C11G0KM9 ). (B and C) Comparison of  CBX8  expression in breast cancer, lung cancer, and primary breast to brain metastatic cancer cells via (B) western blotting, and (C) qRT-PCR analysis. ∗p < 0.05. (D) Kaplan-Meier overall survival analysis shows that glioma patients with higher  CBX8  expression (n = 131) displayed significantly poorer overall survival than those with lower  CBX8  expression (n = 135; p = 1.94e−7) using Glioma Affymetrix microarray cohorts.\nCBX8  Expression Is Upregulated in a Variety of Cancers\n(A)  CBX8  shows higher mRNA-seq expression profile comparing multiple human cancer tissues (in red color) versus normal tissues (in blue) as well as gray (missing TCGA normal sample type). The figure was generated from the COH Bioinformatics Core using Broad Institute TCGA Genome Data Analysis Center (2016): Analysis-ready standardized TCGA data from Broad GDAC Firehose 2016_01_28 run; Broad Institute of MIT and Harvard. Dataset ( https://doi.org/10.7908/C11G0KM9 ). (B and C) Comparison of  CBX8  expression in breast cancer, lung cancer, and primary breast to brain metastatic cancer cells via (B) western blotting, and (C) qRT-PCR analysis. ∗p < 0.05. (D) Kaplan-Meier overall survival analysis shows that glioma patients with higher  CBX8  expression (n = 131) displayed significantly poorer overall survival than those with lower  CBX8  expression (n = 135; p = 1.94e−7) using Glioma Affymetrix microarray cohorts.\nCBX8  knockdown and overexpressing cell lines were generated in glioblastoma (U-251 MG), breast cancer (MDA-MB-231), and lung cancer (A549), and  CBX8  expression levels were validated by quantitative reverse-transcriptase PCR (qRT-PCR) and western blotting ( Figure S2 ). Further, the Transwell invasion assay was performed in  CBX8 -modulated U-251 MG, MDA-MB-231, and A549 cell lines. In U-251 MG and MDA-MB-231 cells, invasion was significantly higher in the  CBX8  overexpression group than in the control and silencing groups (p < 0.05). An increasing trend of invasion was found between the control group and the  CBX8  silencing group, although this was not statistically significant. In A549 cells, invasion increased with increasing  CBX8  expression, although the difference was not statistically significant. This may be attributed to the high standard deviation ( Figure 2 B). Consistent with the results of the Transwell invasion assay,  CBX8  expression was positively associated with migration, as shown in the Transwell migration assay ( Figure 2 A) and scratch assay ( Figures 2 C and  S3 ). Statistically significant differences were observed in the overexpression group versus the control group and the overexpression group versus the silencing group in the three cell lines in the migration assay. The scratch assay was performed to measure the mobility of the cells after modulation of  CBX8  expression. In U-251 MG and MDA-MB-231 cells, the mobility was significantly increased in the overexpression group compared with the control and silencing groups. In A549 cells, an increasing trend in mobility was observed in correlation with increasing  CBX8 . The results thus showed that  CBX8  expression was positively associated with invasion and migration. The effect of  CBX8  modulation on proliferative capacity of MDA-MB-231 cells was evaluated with colony-formation assay and indicated  CBX8  promoted proliferation ( Figure 2 D). Figure 2 Effects of  CBX8  Modulation on the Invasion and Migration of U-251 MG, MDA231-MB, and A549 Cells via Transwell Assay A total of 8,000 cells/well for migration and 50,000 cells/well for invasion in no serum medium were seeded in the upper chamber, respectively. After 8 h incubation for migration or 20 h for invasion, cells passing through the lower chamber membrane were counted, and corresponding images (10×) were captured in at least three random fields. Representative images and quantification of cell numbers are presented here. Overexpression (OE) of  CBX8  significantly increased migration and invasion capacity of U-251 MG, MDA231-MB, and A549 cells, whereas silencing (knockdown [KD])  CBX8  decreased migration and invasion. (A) Transwell migration assay. (B) Transwell invasion assay. ∗p < 0.05. (C) Scratch assay images of U-251 MG, MCF-7, and A549 with  CBX8  modulation (KD, OE) were analyzed based on the online platform WimScratch compared to corresponding wild-type (WT) cells. (D) Effect of  CBX8  modulation on proliferative capacity of MDA-MB-231 cells with  CBX8  silencing (KD) or OE. Images of representative wells from colony-formation assay and corresponding graph are shown. Colonies were stained with 0.1% crystal violet. Data are representative of three independent experiments performed in triplicate. ∗p < 0.05.\nEffects of  CBX8  Modulation on the Invasion and Migration of U-251 MG, MDA231-MB, and A549 Cells via Transwell Assay\nA total of 8,000 cells/well for migration and 50,000 cells/well for invasion in no serum medium were seeded in the upper chamber, respectively. After 8 h incubation for migration or 20 h for invasion, cells passing through the lower chamber membrane were counted, and corresponding images (10×) were captured in at least three random fields. Representative images and quantification of cell numbers are presented here. Overexpression (OE) of  CBX8  significantly increased migration and invasion capacity of U-251 MG, MDA231-MB, and A549 cells, whereas silencing (knockdown [KD])  CBX8  decreased migration and invasion. (A) Transwell migration assay. (B) Transwell invasion assay. ∗p < 0.05. (C) Scratch assay images of U-251 MG, MCF-7, and A549 with  CBX8  modulation (KD, OE) were analyzed based on the online platform WimScratch compared to corresponding wild-type (WT) cells. (D) Effect of  CBX8  modulation on proliferative capacity of MDA-MB-231 cells with  CBX8  silencing (KD) or OE. Images of representative wells from colony-formation assay and corresponding graph are shown. Colonies were stained with 0.1% crystal violet. Data are representative of three independent experiments performed in triplicate. ∗p < 0.05.\nWe next generated an animal model to evaluate the functional role of  CBX8  in metastasis. The luciferase gene was transfected into  CBX8 -modulated cell lines (U-251 MG, MDA-MB-231, and A549) and injected into NOD scid gamma (NSG) mice through the tail vein. Luciferase imaging was performed 28–35 days after injection to monitor the location and growth of cells in the lung ( Figures 3 A and 3B). The results showed that luciferase count and density were increased in relation to  CBX8  upregulation, especially in the  CBX8  overexpression group. Hematoxylin and eosin (H&E) staining showed that both the tumor number and the area of the tumor locus were increased in association with  CBX8  upregulation ( Figures 3 C, 3D, and  S3 ). Figure 3 Effects of  CBX8  Modulation on Tumor Metastasis via Tail Vein Model (A and B) Representative bioluminescent images (A) and bioluminescent density (B) of mice (n = 8) injected with  CBX8 -modulated (KD, OE) U-251 MG, MDA-MB-231, and A549 cells as well as corresponding WT controls. ∗p < 0.05. (C) Representative images of lung tissues of NSG mice injected with  CBX8 -modulating MDA-MD-231 cells. (D) H&E staining of lung tissues from mice injected with  CBX8 -modulated MDA-MB-231 and A549 cells as well as corresponding vector controls.\nEffects of  CBX8  Modulation on Tumor Metastasis via Tail Vein Model\n(A and B) Representative bioluminescent images (A) and bioluminescent density (B) of mice (n = 8) injected with  CBX8 -modulated (KD, OE) U-251 MG, MDA-MB-231, and A549 cells as well as corresponding WT controls. ∗p < 0.05. (C) Representative images of lung tissues of NSG mice injected with  CBX8 -modulating MDA-MD-231 cells. (D) H&E staining of lung tissues from mice injected with  CBX8 -modulated MDA-MB-231 and A549 cells as well as corresponding vector controls.\nChromatin immunoprecipitation sequencing (ChIP-seq) data indicate that  CBX8  can bind to  WNK2 . 25  A systematic literature search showed that  WNK2  is a potential target of  CBX8 . To determine whether reduced  WNK2  expression was associated with increased  CBX8  expression,  CBX8  and  WNK2  expressions were analyzed in U-251 MG, MDA-MB-231, and A549 cells under  CBX8  modulation. The results showed that  WNK2  had a negative relationship with  CBX8 , indicating that  CBX8  may suppress  WNK2  ( Figure 4 A). To further address the association between  CBX8  and  WNK2 , ChIP was performed in  CBX8 -overexpressing cell lines (U-251 MG) ( Figures 4 B and 4C). The results indicated that  CBX8  can directly interact with the promoter of  WNK2 , which is consistent with the  CBX8  suppressing function. To examine the effect of  WNK2  on  MMP2  and  RAC1 ,  WNK2  was knocked down and  MMP2  and  RAC1  activities were analyzed. The results confirmed that  WNK2  negatively regulated  MMP2  and  RAC1  ( Figures S4 A–S4C and S4E). Figure 4 WNK2  Is a Target of  CBX8 . (A)  WNK2  Expression Is Directly Impacted under  CBX8  Modulation in Different Tumor Types (A) qRT-PCR indicates that  WNK2  is significantly upregulated in  CBX8  silencing (KD) tumor cells, while there is striking downregulation in U-251 MG and MDA-MB-231 cells with  CBX8  OE. ∗p < 0.05. (B) Gel images of PCR products and qRT-PCR graph. (C) From ChIP assay using  CBX8  OE U-251 MG cells. With the positive input control,  CBX8  promoter sequence located between primers was detected in 11  CBX8  pull-down products compared to the negative IgG control.\nWNK2  Is a Target of  CBX8 . (A)  WNK2  Expression Is Directly Impacted under  CBX8  Modulation in Different Tumor Types\n(A) qRT-PCR indicates that  WNK2  is significantly upregulated in  CBX8  silencing (KD) tumor cells, while there is striking downregulation in U-251 MG and MDA-MB-231 cells with  CBX8  OE. ∗p < 0.05. (B) Gel images of PCR products and qRT-PCR graph. (C) From ChIP assay using  CBX8  OE U-251 MG cells. With the positive input control,  CBX8  promoter sequence located between primers was detected in 11  CBX8  pull-down products compared to the negative IgG control.\nTo explore the involvement of  CBX8  in invasion and migration, we performed a literature search and tested potential genes associated with these processes. The results showed that  MMP2  and  RAC1  changed considerably in response to the modulation of  CBX8  expression. The  MMP2  expression pattern was evaluated by qRT-PCR in all  CBX8 -modulated cell lines. The results showed that  MMP2  expression and activity increased in association with  CBX8  upregulation ( Figures 5 A, 5B, and  S4 F). Similarly,  RAC1  activity increased in association with  CBX8  upregulation, though a mild change was observed in the  RAC1  expression pattern ( Figure 5 C, 5D, and  S4 F). Figure 5 CBX8  Modulation (KD, OE) Has a Positive Relation with the Expression and Activity of  MMP2  and  RAC1 (A–D) qRT-PCR (A and C) and protein activity assay (B and D) suggest that increased  CBX8  induces the expression level and protein activity of MMP2 in U-251 MG, MDA-MB-231, and A549 cells, whereas knockdown of  CBX8  has a suppressive effect. ∗p < 0.05.\nCBX8  Modulation (KD, OE) Has a Positive Relation with the Expression and Activity of  MMP2  and  RAC1\n(A–D) qRT-PCR (A and C) and protein activity assay (B and D) suggest that increased  CBX8  induces the expression level and protein activity of MMP2 in U-251 MG, MDA-MB-231, and A549 cells, whereas knockdown of  CBX8  has a suppressive effect. ∗p < 0.05.\n\nAs the leading cause of death, metastasis has been extensively reported in the literature. However, the metastatic process is complex, and the underlying mechanism remains elusive. Invasion and migration are critical biological properties of cancer cells and integrated processes that are difficult to separate from each other. In addition to other properties such as colonization, the cooperation between invasion and migration is crucial to promote metastasis. Recently, epigenetic mechanisms, involving  CBX8 , were shown to play an important role in metastasis. 26 , 27\nCBX8  acts as an oncogene and is involved in tumor progression. 21 ,  22 ,  23  Alterations in  CBX8  are observed in many cancers, such as glioblastoma, colorectal cancer (CRC), breast cancer, leukemia, and hepatocellular carcinoma. 22 ,  23 ,  24 , 28 , 29  Specifically,  CBX8  promotes proliferation, differentiation, invasion, and metastasis, promoting breast tumorigenesis and increasing proliferation in CRC. 22 , 24  However, the role of  CBX8  in metastasis is context dependent. Yuan et al. 30  reported that  CBX8  promotes invasiveness in bladder cancer. However, in esophageal squamous cell carcinoma (ESCC) and CRC,  CBX8  suppresses metastasis by inhibiting Snail and p53. 24 , 31  Most studies found CBX8 acted as an oncogene in cancers. In one type of cancer, CBX8 may even have a function of promoting proliferation and inhibiting metastasis. 24 , 31  CBX8 has a function through various pathways in different cancers. However, the mechanism is not reported clearly in each type of cancer. Here, we explored the role of  CBX8  in glioblastoma, breast cancer, and lung cancer and found that the efficiency for promoting metastasis differs among cancers. The effect of  CBX8  on metastasis was more significant in glioblastoma and breast cancer than in lung cancer. Additionally, though sharing a positive relationship with both,  CBX8  had a stronger effect on invasion than migration.\nCBX8  functions through PRC1-dependent and PRC1-independent pathways. 17 ,  18 ,  19 ,  20  In the PRC1-dependent pathway,  CBX8  suppresses genes by assembling the PRC1 complex with other proteins. However, an increasing number of studies showed that  CBX8  can function in a PRC1-independent manner, such as  CBX8  binding to Snail in ESCC, to p53 in CRC, and to p16Ink4a in CRC. 20 , 31  Here, we showed that  CBX8  can suppress  WNK2  expression by binding to the promoter of  WNK2 .\nWNK2 is a kinase that negatively regulates invasion and migration by suppressing  MMP2  and  RAC1  expression and activity. 13 , 14  However, the low activity of WNK2 in cancer is associated with low levels of expression rather than changes in gene sequence, indicating that an epigenetic mechanism is involved in the regulation of  WNK2 . 15 \n CBX8  mutations are rarely reported, whereas overexpression of  CBX8  is widely found in various cancers. 22 ,  23 ,  24 , 28 , 29  Overexpressed  CBX8  may function in cancer by suppressing antioncogenes. Here, we showed that  CBX8  suppressed  WNK2  expression by binding to the  WNK2  promoter, increasing  MMP2  and  RAC1  expression and activity.\nAs a member of the MMP family,  MMP2  plays a central role in malignant tumors and predicts a poor prognosis.  MMP2  promotes invasion by promoting ECM degradation to provide space and direction for cell movement and digestion of the basement membrane to maintain tissue organization. 9 \n RAC1  mainly regulates cancer cell migration. Both expression and activity of  MMP2  and  RAC1  are negatively regulated by  WNK2 . 13 , 14  After validating the negative relationship between  CBX8  and  WNK2 , we found that  MMP2  expression and activity were increased in association with  CBX8  overexpression. Both invasion and migration increased, indicating that  CBX8  can suppress  WNK2  and promote invasion and migration by increasing the expression and activity of  MMP2  and  RAC1 .\nCollectively, the results indicated that  CBX8  was upregulated in glioblastoma, breast cancer, and lung cancer. Overexpression of  CBX8  can promote cancer cell invasion and migration, leading to metastasis. This process is dependent on the  CBX8 -mediated suppression of  WNK2 , which results in increased expression and activity of  MMP2  and  RAC1 .\n\nU-251 MG, MDA-MB-231, and A549 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). U-251 MG and MDA-MB-231 cells were grown in Dulbecco’s modified Eagle medium (DMEM); A549 cells were grown in F-12K Nutrient Mixture, Kaighn’s Mod (FK-12). All media contained 10% fetal bovine serum (FBS) (Omega Scientific) and 1% penicillin-streptomycin.  CBX8 -overexpressing and -silencing plasmids were purchased from OriGene (overexpression: RC203417L1; knockdown: TF317361). Plasmids with a scramble sequence were used as the silencing group control, and those with a flag tag sequence were used as the overexpressing group control (generated by Gang Li). Cells were transfected with 4 mg of each plasmid using Lipofectamine 2000 (Invitrogen) following the manufacturer’s instruction. Overexpressing groups were selected with G418 and silencing groups were selected with puromycin ( Table 1 ). Table 1 Antibiotic Concentrations Used for Selection after Transfection Cell Type G418 (overexpressing) Puromycin (silencing) U251MG 1,000 μg/mL 2 μg/mL MDA-MB-231 1,200 μg/mL 2 μg/mL A549 1,000 μg/mL 2 μg/mL\nAntibiotic Concentrations Used for Selection after Transfection\nQuantitative PCR (qPCR) was performed to analyze  CBX8 ,  WNK2 , and  MMP2  expression in different cell lines. After extraction of total RNA using the Trizol Reagent (Ambion), 500 ng of RNA template was mixed with 5 × PrimeScript RT Master Mix (Bio-Rad) in RNase-Free dH 2 O. The mixture was maintained at 37°C for 15 min for reverse transcription, followed by heat inactivation at 85°C for 5 s.\nFor qPCR, 1 μL of cDNA template was added to a master mix consisting of SYBR Premix EX TaqII (Bio-Rad), forward and reverse primers ( Table 2 ), and ROX reference dye (TaKaRa Bio) to form a 20 μL reaction mixture. Duplicates of each sample were made for each run. The reaction was performed using the 7900HT Fast Real-Time PCR System (Applied Biosystems); the thermal cycling conditions were 30 s at 95°C, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. The internal loading control was glyceraldehyde-3-phosphate dehydrogenase (GAPDH); the relative  CBX8 ,  WNK2 , and  MMP2  expression was calculated with the 2-ΔCt formula, where ΔCt = CtCBX − CtGAPDH. A list of primers used is included in  Table 2 . Table 2 Primers for qRT-PCR Gene Forward Primer Chain Reverse Primer Chain CBX8 5′-GTGAAATGGAAGGGATG-3′ 5′-GTTTTGGGCTTG GGTC-3′ WNK2 5′-GTGCACGATCCTGAAATC-3′ 5′-CAGTTTCTTGGG GTCTTCC-3′ MMP2 5′-CTTGACCCATGCATTCTC-3′ 5′-CATCCCAATGAC CTCATC-3′ RAC1 5′-GCTTTTGCGGAGATTTTGA-3′ 5′-CCCGTGACACTTT CATTCCT-3′ GAPDH 5′-GGTAGGGAGTTCGAGACCAG-3′ 5′-TCAACGCAGTTC AGTTAGGC-3′\nPrimers for qRT-PCR\nProtein lysates were generated using radioimmunoprecipitation assay (RIPA) buffer (Cell Signaling Technology) with a protease inhibitor cocktail (Sigma-Aldrich). After incubation for 30 min on ice, the protein lysates were centrifuged at 17,000 rpm for 10 min. Protein concentration was quantified using the bicinchoninic acid method (BCA) (Thermo Fisher Scientific); the optical density (OD) of reaction mixtures was measured at 570 nm, and protein quantities were obtained by referring to the bovine serum albumin standard curve. The samples (20 μg for each lane) paralleled with a protein ladder (10–170 kDa; Thermo Fisher Scientific) were resolved in a 10% SDS-PAGE gel; the proteins were then transferred to polyvinylidene fluoride (PVDF) membranes (GE Healthcare Life Sciences, Vienna, Austria) using a wet electroblotting system (Bio-Rad Laboratories, Pittsburgh, PA, USA). For immunoblotting, the membranes were first blocked with 5% dry fat-free milk for 1 h at room temperature and washed twice with Tris-buffered saline containing 0.1% Tween 20 (TBST; Affymetrix). The membranes were incubated for 2 h at room temperature with different primary antibodies, including anti-CBX8 (Santa Cruz) and anti-GAPDH (Cell Signaling Technology). This was followed by washing twice with TBST and blotting with the corresponding horseradish peroxidase (HRP)-conjugated secondary antibodies (anti-mouse or anti-rabbit; 1:10,000) for 1 h. After the incubation, membranes were washed three times with TBST to remove any non-specific binding of secondary antibody. Chemiluminescent detection of signals was performed with HRP substrates (Merck Millipore), and the signal was visualized in X-ray films.\nWound healing assay was carried out to determine the cell protrusion and migration ability of tumor cells. U-251 MG, MDA-MB-231, and A549 cells were grown in their respective media and then seeded in 6-well plates and cultured until reaching 100% confluency as a monolayer. A new 20 μL pipette tip was used to make three parallel straight scratches gently and slowly with a marker on the bottom of the dish. The resulting gap distance was considered equivalent to the outer diameter of the end of the tip. Cells were washed with phosphate-buffered saline (PBS) three times and cultured in medium without FBS to inhibit proliferation. Images were acquired using phase contrast and a 10× objective. The wound was measured at 0, 12, 24, 48, and 72 h. Images were analyzed with the online platform WimScratch. Migration was determined by comparing relative gap areas.\nThe MTT assay was utilized to determine cell viability. Cells were grown in 96-well plates at a density of 5 × 10 4  cells/well and cultured in medium with or without FBS for 72 h. Then, 20 μL MTT (5 mg/mL) was added to each well and incubated at 37°C for 4 h. DMSO (150 μL) was added to solubilize the formazan crystals. The plate was then wrapped in foil and placed on an orbital shaker for 15 min. The amount of formazan salt was determined by measuring the OD at 490 nm using a Bio-Rad 680 microplate reader (Bio-Rad Laboratories, Hercules, CA, USA). Cell viability was expressed as the percent OD value of each group relative to that of control cells.\nCell migration and invasion were assessed using the Millipore 24-well Millicell Chamber with pore size 8 mm (Millipore). For the migration assay, 8,000 cells in DMEM without FBS were added to the upper chamber of the insert. For the invasion assay, 50,000 cells in DMEM without FBS were added to the upper chamber pre-coated with Matrigel (Sigma). DMEM with 20% FBS was placed in the lower chamber. Cells were incubated for 8 h for the migration assay and 20 h for the invasion assay. Then, cells were fixed and stained with 0.1% crystal violet, and non-invading cells were removed with cotton swabs. The number of cells on the lower surface of the chamber membrane was counted under a microscope with a 20× objective in five random fields.\nCells were seeded in a 6-well plate at 80%–90% confluence for 24 h without antibiotics. A total of 2,500 ng DNA was diluted with 5.0 μL RNAi. Small interfering RNAs (siRNAs) were from Dharmacon (designated siRNA #1) and Sigma (designated siRNA #2). Cells in all conditions designated as “Control” were transfected with a pool of siRNAs that do not target human genes using Lipofectamine 2000 in 100 μL Opti-MEM Medium without serum, using the manufacturer’s protocol. Cells were harvested 24 h after transfection. Pools of at least three siRNAs were used to dilute potential off-target effects.\n8- to 10-week-old NSG mice (Jackson Laboratory) were used. The mice were maintained under specific pathogen-free (SPF) conditions in an animal facility and given a pelleted regular rodent diet and water. Tumor cells were counted and diluted to 5 × 10 6 /mL. Prior to injection, mice were weighed and warmed for 10 min in a commercially available warming box to dilate the veins. Then, mice were lightly anesthetized by xylene and ketamine. Approximately 10 6  cells (200 μL) were injected into the tail vein (n = 8). After 6–8 weeks, luciferase imaging was performed to evaluate lung metastasis. Then, mice were euthanized, weighed, and lung tissues were harvested. H&E staining was performed to assess metastasis. All experiments involving mice were performed according to Tianjin Medical University Cancer Institute and Hospital animal guidelines. To generate tumor cell lines stably expressing the luciferase gene, cells were transfected with pGL4.51 plasmids (Promega, WI, USA), followed by G418 selection and confirmation with the presence of luciferin according to manufacture instructions.\nFormaldehyde (1%) was directly added to cell media. The cross-linking reaction was terminated by adding 0.125 M glycine. Cells were pelleted after centrifugation, followed by preparation of nuclear lysates using Magna ChIP protein G Kit (Millipore, Billerica, MA, USA). Cells were resuspended with 1 mL per 5 × 10 7  cells of swelling buffer. Nuclei were pelleted by microfuge at 2,500 × g for 5 min at 4°C and were resuspended in 1 mL per 1 × 10 8  cells of prepared Nuclei Lysis Solution and incubated on ice for 10 min. Samples were sonicated, then 1 μL of 100 mM phenylmethanesulfonylfluoride (PMSF) was added per 100 μL of blocked Staph A cells. Approximately 1 × 10 7  cells were used for each immunoprecipitation (IP). An immunoglobulin G (IgG) negative control sample was also included. Primary antibodies (1 μg each) were added to each sample and incubated on a rotating platform at 4°C overnight. Secondary antibody (1 μg) was added and incubated for an additional hour at 4°C. The antibody/protein/DNA complexes were eluted by adding 50 μL of IP elution buffer at room temperature, followed by shaking and centrifuging at 14,000 rpm for 3 min at room temperature. Supernatant was removed, then 4 μL of 5 M NaCl (0.2 M NaCl final) was added to each IP sample tube. All samples were incubated at 67°C overnight. Then, 1 μL of 10 mg/mL RNase A was added to each sample and incubated for 30 min at 37°C. ChIP samples were used for PCR reaction.\nPCR was performed with primers ( Table S1 ): binding site: 5′-TGGGCAACATGGTGAAA-3′ and 5′-AAGCAATCCTCCCACCTCA-3′.\nRAC1  activity was measured using colorimetric G-LISA assays (Cytoskeleton, Denver, CO, USA) according to the manufacturer’s instructions. Briefly, cells were lysed with 1 mg/mL GL36. The lysate was incubated in a RAC1-GTP affinity plate for 30 min on a shaker at 200 rpm. The plate was washed with running buffer, then  RAC1  activity was analyzed on the Molecular Devices SpectraMax 250 at 490 nm.\nMMP2 activity was measured using MMP Activity Assay Kit (Fluorometric - Green) (Abcam, USA) according to the manufacturer’s instructions. Briefly, cells were lysed and incubated with 2 mM 4-aminophenylmercuric acetate (APMA) working solution in 96-well plate per assay. Then, 50 μL of MMP Green Substrate working solution was added to the sample and control wells of the assay plate. Fluorescence intensity was assessed with a fluorescence plate reader at Excitation Wavelength (Ex)/Emission Wavelength (Em) = 490/525 nm.\n\nY.J., Y.W., and C.Z. conducted the experiment and collected data. Y.J. and M.Y.C. contributed to the conception of the study and wrote the paper.\n\nThe authors declare no competing interests.","source_license":"CC-BY-4.0","license_restricted":false}