Tn antigen promotes human colorectal cancer metastasis via H‐Ras mediated epithelial‐mesenchymal transition activation

Abstract Tn antigen is a truncated O‐glycan, frequently detected in colorectal cancer (CRC), but its precise role in CRC metastasis is not well addressed. Here we investigated the effects of Core 1 β3Gal‐T specific molecular chaperone (Cosmc) deletion‐mediated Tn antigen exposure on CRC metastasis and its underlying mechanism. We first used CRISPR/Cas9 technology to knockout Cosmc, which is required for normal O‐glycosylation, and thereby obtained Tn‐positive CRC cells. We then investigated the biological consequences of Tn antigen expression in CRC. The results showed that Tn‐positive cells exhibited an enhanced metastatic capability both in vitro and in vivo. A further analysis indicated that Tn antigen expression induced typical activation of epithelial‐mesenchymal transition (EMT). Mechanistically, we found that H‐Ras, which is known to drive EMT, was markedly up‐regulated in Tn‐positive cells, whereas knockdown of H‐Ras suppressed Tn antigen induced activation of EMT. Furthermore, we confirmed that LS174T cells (Tn‐positive) transfected with wild‐type Cosmc, thus expressing no Tn antigen, had down‐regulation of H‐Ras expression and subsequent inhibition of EMT process. In addition, analysis of 438 samples in TCGA cohort demonstrated that Cosmc expression was reversely correlated with H‐Ras, underscoring the significance of Tn antigen‐H‐Ras signalling in CRC patients. These data demonstrated that Cosmc deletion‐mediated Tn antigen exposure promotes CRC metastasis, which is possibly mediated by H‐Ras‐induced EMT activation.

Mucin type O-glycosylation is initiated by addition of N-acetyl galactosamine (GalNAc) to either serine or threonine, thus forming a biosynthetic intermediate O-linked structure called Tn antigen. 8,9 Tsynthase (core 1 β1,3-galactosyltransferase), a key enzyme in the Oglycosylation pathway, converts Tn antigen to more complex O-glycans that eventually modify many secreted and transmembrane glycoproteins. 10,11 Of note, the expression and activity of T-synthase require an endoplasmic reticulum (ER)-resident molecular chaperone Cosmc. [12][13][14] Dysfunction in T-synthase or Cosmc leads to the expression of Tn antigen, which is indeed a characteristic of abnormal Oglycosylation. 15,16 It has been reported that Tn antigen is detected in almost 90% of human colon cancer whereas it is rarely detected in normal tissue. 15,17 Tn antigen expression is correlated with tumour invasion and poor prognosis in CRC patients, and it is proposed as a potential target for immunotherapy. 6,18 Nevertheless, it remains elusive whether Tn antigen may play a causative role in colorectal cancer development and metastasis.
In this study, we forcedly induced Tn antigen expression in both CRC cell lines (HCT116, SW480) through knockout of the Cosmc chaperone. Both in vitro and in vivo experiments showed that Tn antigen directly promoted cancer invasion and metastasis of cells or tumours, suggesting a metastasis-promoting role for Tn antigen.
Moreover, we found that Tn antigen activated the EMT process, which was responsible for the observed oncogenic alterations, by up-regulating the expression of H-Ras.

| Flow cytometry
Cells were harvested from the culture flask and resuspended in 100 μL PBS (1 × 10 6 /mL). The mouse anti-Tn mAb (CA3638, clone 12A8-C7-F5, 10 μg/mL, kindly provided by Dr. Tongzhong Ju of the Emory University School of Medicine in Altanta, USA) was used to detect the expression of Tn antigen. After incubation at 4°C for 1 hour, the cells were incubated with PE-labelled goat antimouse IgM (BD, 562033) for 60 minutes at 4°C. Then the cells were washed twice and then analysed using the flow cytometer (Canto II; BD Bioscience, Franklin Lakes, NJ, USA).

| Cell migration and invasion assays
After being starved for 24 hours, cells (2 × 10 5 ) with serum-free medium were seeded into the upper chamber of transwell plate (BD Bioscience, 8 μm pore size) pre-coated with or without Matrigel (BD Bioscience). A 500 μL 10% serum-containing medium was added to the lower chamber. After being cultured for 24 hours or 48 hours, the migrated or invaded cells were counted under the microscope after fixation with 4% paraformaldehyde and staining with 0.1% crystal violet.

| Establishment of transplantable metastatic murine models
The male BALB/c nude mice aged 6 weeks were purchased from Charles River Laboratories (Beijing, China) and maintained under specific pathogen-free conditions. All animal experiments were performed under the guidelines of Institutional Animal Care and Use Committee at Capital Medical University (Beijing, China) and Medical Research Center of Beijing Chao-Yang Hospital.
We performed two transplantable metastatic murine models to explore the role of Tn antigen in the metastasis of CRC in vivo. For intrasplenic injection mouse models, 6-week-old BALB/c nude mice were anesthetized with chloral hydrate (400 mg/kg) and then incised in the left upper lateral abdomen for 1 cm. The prepared single cell suspension of HCT116 cells with or without Tn antigen expression (2 × 10 6 per mouse) were injected into the spleen of each mouse (n = 6 for each group) respectively. After 8 weeks, the mice were killed and the livers and lungs were excised and fixed with formalin for H&E staining.
For orthotopic implantation models, 2 × 10 6 HCT116Tn+ cells or HCT116Tn-cells were first injected subcutaneously into 6-week-old BALB/c nude mice to attain subcutaneous xenografts that were subsequently minced into 2 mm 3 pieces and subserosally transplanted into the cecum of other mice (n = 5 for each group). When cachexia occurred, the mice were killed. The xenografts and macroscopically visible metastatic lymph nodes were isolated and embedded in paraffin for histological examinations and immunohistochemistry staining.

| Knockdown of H-Ras with shRNA
The plasmid containing short hairpin RNA (shRNA) against H-Ras

| Re-expression of Cosmc in LS174T cells
The GV367-EGFP-Cosmc lentiviral particles were produced by Shanghai Genechem Co. Ltd (Shanghai, China) . LS174T cells were transfected with the lentiviral particles with or without a Cosmc gene insert using polybrene (Genechem). The transfected cells were selected using puromycin (2 μg/mL) after 2 days and then pooled for further studies.  Table S1.
Relative changes in mRNA were normalized with GAPDH and calculated using 2 -▵▵CT methods.

| Western blotting and antibodies
Cells and tissues from xenografts were lysed with RIPA lysis buffer (Solarbio, Shanghai, China) and the protein concentration was determined by BCA assay kit (Thermo Fischer, Waltham, MA, USA).
Equal amount of denatured protein was electrophoresed on 10% SDS-PAGE and transferred onto PVDF membranes (Millipore, Bedford, MA, USA). The membranes were blocked with 5% defatted milk and then probed with primary antibodies overnight at 4°C. After incubation of specific HRP-conjugated secondary antibodies, signals were visualized by ECL kit (Millipore) according to the manufacturer's instruction.

| TCGA colon cancer dataset
To depict the association between Cosmc and H-Ras in CRC patients, the RNA-seq data of the 438 cases from TCGA colon cancer dataset were obtained from the cBioPortal for Cancer Genomics (http://www.cbioportal.org/).

| Statistical analysis
Each experiment was repeated three times independently to verify reproducibility. Data analysis was performed using the SPSS 22.0 statistical software. The data from in vitro experiments were presented as the mean ± SD. The differences of survival time between the mice bearing Tn-positive cells and the mice bearing Tn-negative cells in the orthotopic models was evaluated using a log-rank test.
The none-parametric Mann-Whitney U test was used when the metastasis rate between two groups was compared. Student's t test was used for the statistical analysis. P < 0.05 was considered statistically significant.

| Cosmc deficiency induces the expression of Tn antigen in human CRC cell lines
To explore the roles of Tn antigen, we first used CRISPR/Cas9 technology to knockout the gene encoding Cosmc, which is specifically

| Tn antigen promotes invasive and metastatic properties of colorectal cancer
We conducted the transwell assays (matrigel-uncoated for migration We also established an orthotopic implantation murine model, which ideally mimics the in vivo invasion and metastasis of cancer, to investigate the effects of Tn antigen on CRC metastasis. 19,20 Although there was no liver or lung metastasis detected in the mice undergoing orthotopic transplantation, we still observed that the mice implanted with Tn-positive cells presented with more lymph node metastases compared with those implanted with Tn-negative cells (5/5, 100% vs 3/5, 60%) ( Figure 2F and Figure S1). Besides, Tn antigen led to a significantly reduced survival time in the mice bearing Tn-positive xenotransplants (P = 0.047) ( Figure 2G and Table S2).
Collectively, Tn antigen expression significantly enhanced the in vivo metastasis of CRC (9 of 11 vs 4 of 11, P = 0.034, Table 1), and resulted in a worse prognosis.  Figure 3D). Together, these data indicated that Tn antigen expression was able to activate the EMT process in CRC.

| Tn antigen targets H-Ras to activate the EMT process
We further investigated the mechanisms by which Tn antigen activates EMT. Accumulating evidence shows that Ras signalling activates EMT during cancer metastasis. 24,25 We reasoned whether Tn antigen has an influence on Ras family proteins (H-Ras, K-Ras and N-Ras), and consequently, activates EMT. We found that H-Ras, but not K-ras or N-ras, was prominently up-regulated in both Tn-positive CRC cells as compared with that in Tn-negative cells ( Figure 4A and  Tn-1/6 0/6 3/5 4/11 a Distant metastases to specific organs or regional lymph nodes occurred in both mouse models. bThere was a significantly increased incidence of metastasis in Tn-positive cells implanted mice. (9 of 11 vs 4 of 11, P* < 0.05).

| Re-expression of Cosmc in LS174T cells (Tnpositive) reduces H-Ras expression and subsequent activation of EMT
To further confirm the above results, we included another colorectal

| Determination of Tn antigen-H-Ras signalling in patients
As these experiments indicated that Tn antigen expression was associated with H-Ras mediated EMT, we further investigated 438 cases in the TCGA cohort to delineate the relationship between Tn antigen and H-Ras in patients. Because expression of Tn antigen was induced by Cosmc deficiency, we performed a correlation analysis between Cosmc and H-Ras using the transcriptome data of colon cancer samples in TCGA. As it can be seen, Cosmc expression levels showed a significant reverse correlation with H-Ras levels (r = −0.29, P < 0.0001, Figure 6A) in the colon cancer tissues. The expression levels of H-Ras in Cosmc-low patients were significantly higher compared with the Cosmc-high patients (P < 0.0001, Figure 6B). All these data were consisted with the above in vitro and in vivo results and highlighted the significance of Tn antigen-H-Ras signalling in patients.

| DISCUSSION
In this study, we investigated the pathological role of Tn antigen in colorectal cancer metastasis using in vitro and in vivo experiments.
We used precise gene editing to knockout the Cosmc chaperone to  34 Hofmann et al demonstrated that forced expression of Tn antigen in pancreatic cancer cells leads to increased migration and decreased proliferation and apoptosis. 35 Lin et al also found that Tn expression in oral squamous cell carcinoma enhances the invasive potential of tumour cells through up-regulating EGFR phosphorylation and activity. 36 Conversely, Bergstrom and colleagues unexpectedly found that Tn antigen is not involved with cancer progression in a murine model of colorectal cancer; instead intestinal inflammation is a major mechanism responsible for tumour development. 37 Song et al even reported that Tn antigen expression suppressed breast cancer development in mice via impairment of MUC1 expression, which is usually highly expressed in various cancers. 38 All of these findings suggest that the role of Tn antigen in cancer progression and metastasis needs to be further understood.
In the process of O-glycosylation, the Tn antigen is normally Our results clearly showed that both CRC cells expressing Tn antigen enhanced the capabilities of cell migration and invasion in vitro. More importantly, two types of transplantation metastatic mouse models (intrasplenic injection model and orthotopical implantation model 42,43 ) confirmed that tumour metastasis was more evident in Tn-positive group compared with the control mice expressing no Tn antigen. These findings provide direct evidence that Tn antigen promotes CRC metastasis, which may be considered as a target for immunotherapy in future.
We next asked how Tn antigen affects cancer metastasis in CRC.
Because of tumour metastasis is a complicated and multi-step course in which EMT has been considered to be associated with the increased invasive and metastatic properties, [44][45][46] we assumed that activation of EMT may play an important role in Tn antigen-induced metastasis. We found that Tn antigen significantly activated the EMT pathway, demonstrated by a reduced expression of epithelial cell marker such as E-cadherin, and an enhanced expression of mesenchymal cell markers with both in vitro and in vivo assays. We thereby concluded that Tn antigen promotes cancer metastasis through activation of the EMT pathway. We further explored the mechanism as to how Tn antigen activates EMT.
More recently, it has been reported that Ras/MAPK signalling pathway is responsible for activation of EMT, and is associated with tumourigenesis. 24,47,48 Suppression of Ras signalling by Rnd1 inhibited the following signalling cascades and consequently deregulated the activation of EMT. 49  In summary, our data showed that Tn antigen expression, a hallmark of abnormal O-glycosylation, may contribute to colon cancer metastasis. Tn antigen may promote activation of the EMT process by up-regulating the expression of H-Ras, as summarized in our model ( Figure 6C). Our study underscored the importance of Tn antigen in CRC progression and metastasis, and suggested that anti-Tn antigen may hold a great promise for tumour immunotherapy.