CRC samples (n=471) and control samples (n=41) were downloaded from the TCGA-Colon Adenocarcinoma (COAD) database and MARCH9 expression was assessed by individual cancer stage (stage 1-stage 4), histological subtype (adenocarcinoma and mucinous adenocarcinoma), and nodal metastasis status (N0-N2).

For cell lines, the human colonic cancer cell lines, HCT116 and RKO, were chosen and purchased from ATCC. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone) that was supplemented with 10% fetal bovine serum (FBS, Hyclone) and 1% penicillin-streptomycin solution (Solarbio), and incubated in 5% CO₂ (PHcbi, MCO-18AC). Lipofectamine RNA-iMax reagent (Invitrogen) was used for cell transfection. The sequences of siMARCH9#1 (5’-GCAGTGGAAGGTCCTAAATTA-3’), siMARCH9#2 (5’-GTCCAGATTGCTGCCATAGTT-3’) and siCtrl(5’-UUCUCCGAACGUGUCACGU-3’) were synthesized from GenePharma (Shanghai). The VigoFect repeat (Vigorous Biotechnology) was used for MARCH9 overexpression.

The total RNAs of the HCT116 and RKO cells were isolated using Trizol Reagent. (Ambion, Life Technologies). ReverTra Ace qPCR RT Master Mix with gDNA Remover (Toyobo) was used for reverse transcription and 2* SYBR Green qPCR Master Mix (Low ROX) (Servicebio) was used for real-time PCR. The following primers were utilized: MARCH9-F: 5’-CTCCTCTGTCTACCGCATCTT-3’; MARCH9-R: 5’-TCTCCTCCTATGTCCTTGGTCT-3’; GAPDH-F: 5’-GACTCATGACCACAGTCCATGC-3’; and GAPDH-R: 5’-AGAGGCAGGGATGATGTTCTG-3’.

Whole-cell lysates were prepared in RIPA lysis buffer (Beyotime) containing a protease cocktail (Beyotime). BCA Protein Assay Kit (Thermo) was used to evaluate total protein concentrations. Proteins were separated with 10% SDS-PAGE and transferred to a nitrocellulose membrane, and the membrane was blocked with 0.5% bovine serum albumin (BSA) for 1 h at 25°C. Anti-MARCH9 (proteintech, 1:1000), anti-CYLD (Abcam 1:1000), anti-p-p65 (1:1000), and anti-vinculin (ABclonal, 1:1000) antibodies were added to the membranes and incubated at 4°C overnight, separately. After washing three times with TBST, secondary antibodies were added and the membrane was incubated for 1 h at 25°C. IRDye@800CW reagent (LI-COR) and Odyssey CLx imager (LI-COR) were used to visualize the protein bands.

Two CRC cell lines containing about 300 cells each were planted into 6 cm tissue culture plates and 4 mL DMEM was added. The cells were cultured for 3 weeks to form colonies. The cells were then fixed in 100% methanol for 20 min and stained with 0.1% crystal violet solution (Solarbio) for 10 min. The positive colonies were photographed and quantitated using a digital camera.

Transfected HCT116 and RKO cells were seeded in the upper chamber of a Transwell plate in DMEM containing 10% FBS. DMEM containing 20% FBS was added to the lower chamber. An 8.0 μm pore membrane was inserted to allow the cells to migrate, and the plate was incubated at 37°C for 24 h. The migrated cells were fixed, stained with crystal violet, and ten random fields of each well were counted.

Cell apoptosis in transfected cells was analyzed using a FITC-Annexin V apoptosis kit (Yeasen, Shanghai). In brief, cultured cells were suspended, centrifuged, washed with PBS, and resuspended in cold Binding Buffer. FITC-Annexin V reagent (1.25 μL) was added to the cells and apoptotic cells were assessed using a A00-1-1122 flow cytometer (Beckman Kurt Biotechnology (Suzhou) Co., Ltd).

For cell cycle analysis, the transfected cells were fixed with 70% cold ethanol (700 μL) at 4°C for 4 h, washed three times with PBS, and stained with the fluorescent dye propidium iodide (PI; Sigma) at 37°C for 30 min in the dark. The ratio of cells in various cell cycle phases was assessed using a Beckman flow cytometer.

In brief, a total of 10⁴ siCtrl, siMARCH9#1 and siMARCH9#2 HCT116 and RKO cells were seeded into 96-well plates. 1.5 hours later, the activity of caspase 3/7 was measured by using Caspase-Glo reagent (Promega), according to the manufacturers’ protocols.

Female BALB/c nude mice that were 6 weeks old (Beijing vital-river Co.td) were raised in a pathogen-free facility and divided into a shMARCH9 group (n=5) and a shCtrl group (n=5). The mice were subcutaneously injected with shMARCH9#1 or shCtrl transfected HCT116 cells (4 × 10⁶ cells) in the right axilla. After 45 days, the mice were anesthetized and sacrificed, and the tumors were excised, weighed, and photographed. Tumor tissues were used to detect the expression of MARCH9 by western blotting.

All the experimental results was used to analyze in GraphPad Prism 7 software and data were expressed as the mean ± SD. The Student’s t-test was used to compare differences between the groups. Statistical significance was considered at p < 0.05.

We firstly explored the clinical significance of MARCH9 in CRC. Based on TCGA data, MARCH9 expression was significantly higher in CRC than adjacent normal tissue samples ( Figure 1A ). MARCH9 expression was also much higher in CRC stages 1-4 ( Figure 1B ), adenocarcinoma and mucinous adenocarcinoma ( Figure 1C ), and N0-N2 samples ( Figure 1D ).

Next, we investigated the role of MARCH9 in CRC by using loss-of-function assay. As shown by RT-qPCR and Western blot analysis, both siMARCH9#1 and siMARCH9#2 could effectively knock down MARCH9 expression in two CRC cell lines ( Figures 2A, B ). The CCK-8 assay was used to investigate whether MARCH9 knockdown influenced CRC cell proliferation. CRC cell lines transfected with siMARCH9#1 and siMARCH9#2 had significantly lower viability than the siCtrl group ( Figure 2C ). Knockdown of MARCH9 also inhibited the colony formation and migration of CRC cells ( Figures 2D, E ). We also checked the effect of MARCH9 on the proliferation of colorectal normal cells. The results showed that knockdown of MARCH9 suppressed the proliferation of FHC cells ( Supplementary Figure 1 ). Therefore, MARCH9 expression is critical for the proliferation and growth of colorectal normal and CRC cells.

To confirm the function of MARC9 in CRC, we then performed gain-of-function experiments. As shown by RT-qPCR and Western blot analysis, endogenous MARCH9 expression was greatly upregulated after MARCH9 overexpression in two CRC cell lines ( Figures 3A, B ). This resulted in an increase in cell viability and proliferation ( Figures 3C, D ). MARCH9 overexpression also enhanced CRC cell migration, supporting an oncogenic role for this protein in CRC. These results confirmed the oncogenic function of MARCH9 in CRC ( Figure 3E ).

We then explore the role of MARCH9 in CRC cell apoptosis and cell cycle progression. MARCH9 knockdown significantly promoted apoptosis in HCT116 and RKO cells as shown by flow cytometry ( Figures 4A, B ). Furthermore, downregulation of MARCH9 enhanced the caspase 3/7 activity in both cell lines ( Figure S2 ), suggesting the underlying mechanisms how MARCH9 regulates apoptosis. There was a higher number of cells in the G0/G1 phase of the cell cycle and a lower number in the S and G2/M phases in two CRC cell lines ( Figures 4C, D ). These results indicated that MARCH9 downregulation induced cell apoptosis and blocked the cell cycle in the G0/G1 phase.

Cylindromatosis (CYLD) is an important tumor suppressor gene that inhibits the development of various cancers through inactivation of NFκB (p65) (25, 26). A wide range of studies have shown that the expression of CYLD could be suppressed at transcription level by different non-coding RNAs (26–28), whereas whether CYLD abundance was repressed at post-transcription manner in colorectal cancer should be determined. Based on western blot results, we showed that knockdown of MARCH9 induced CYLD protein expression and reduced p65 phosphorylation in two CRC cell lines ( Figure 5A ), while overexpression of MARCH9 had the opposite effect on CYLD expression and p65 phosphorylation ( Figure 5B ). Total protein expression of p65 did not change after knocking down or overexpressing MARCH9 ( Figure 5 ), indicating that MARCH9 regulates p65 phosphorylation but not expression levels. These results suggest that MARCH9 suppresses the expression of CYLD at post-transcription manner, leading to activation of p65.

To explore the role of p65 activity regulated by MARCH9 in CRC, we treated Ctrl and MARCH9 CRC cells with a specific p65 inhibitor, named Caffeic Acid Phenethyl Ester (CAFE). While MARCH9 overexpression promoted CRC cell viability ( Figure 6A ), increased the number of cell colonies ( Figure 6B ), and increased cell migration ( Figure 6C ), co-expression of CAFE weakened these tumor-promoting effects. By contrast, the inhibitory effect of CAFÉ on the growth and migration of Ctrl cells was smaller than that of MARCH9 overexpressing cells ( Figures 6A-C ). These results suggest that MARCH9 may accelerate CRC progression by regulating p65.

Lastly, we studied the in vivo function of MARCH9 by implanting shCtrl and shMARCH9 CRC into the BALB/C nude mice. The shMARCH9#1 group of BALB/C nude mice had smaller tumors with significantly lower weight than the shCtrl group ( Figures 7A, B ). Western blot analysis revealed lower expression of MARCH9 and higher expression of CYLD in the xenograft tumor tissues from the shMARCH9#1 group than tissues from the siCtrl group ( Figure 7C ). In addition, MARCH9 knockdown suppressed the phosphorylation of p65, while had no effect on the total protein expression of p65 in the tumor tissues ( Figure 7C ). Collectively, these data suggest that MARCH9 knockdown suppressed HCT116 cell tumorigenicity.