TM4SF1 promotes EMT and cancer stemness via Wnt/ β-catenin/SOX2 pathway in colorectal cancer

Background Transmembrane 4 L six family member 1 (TM4SF1) is upregulated in several epithelial cancers and closely associated with poor prognosis. However, the role of TM4SF1 and the potential mechanism in colorectal cancer (CRC) remains elusive. Methods The expression data were obtained via the Oncomine database, the Cancer Genome Atlas database(TCGA) and Gene Expression Omnibus (GEO) database, and immunohistochemistry (IHC), RT-qPCR and western blot analysis. The effect of TM4SF1 on CRC cells were investigated by Transwell assay, wound healing assay and sphere formation assay. A series of in vitro and in vivo experiments were conducted to reveal the mechanisms of TM4SF1 modulating EMT and cancer stemness in CRC. Results TM4SF1 was significantly overexpressed in CRC and positively correlated with the poor prognosis. Down-regulation of TM4SF1 inhibited cell migration, invasion and tumor sphere formation in SW480 and LoVo cells. Moreover, TM4SF1 silencing inhibited the epithelial to mesenchymal transition (EMT) mediated by transforming growth factor-β1(TGF-β1). Mechanistically, Gene set enrichment analysis (GSEA) and western blot analysis revealed that TM4SF1 modulated SOX2 expression in a Wnt/β-catenin activation-dependent manner. Furthermore, we found that knockdown of TM4SF1 suppressed the expression of c-Myc leading to decreased c-Myc binding to the SOX2 gene promoter. Finally, Depletion of TM4SF1 inhibited metastasis and tumor growth in a xenograft mouse model. Therefore, our studies substantiate a novel mechanism by which TM4SF1 maintains cancer cell stemness and EMT via the Wnt/β-catenin/c-Myc/SOX2 axis during the recurrence and metastasis of CRC. and DLD1were cultured in RPMI 1640 medium (Hyclone, USA). All of the cell lines were cultured in medium supplemented with 10% fetal bovine serum (FBS, Sciencecell). Primary antibodies against the proteins, TM4SF1, N-cadherin, Vimentin, SOX2, MMP9, CD133, Par3 and β-catenin were purchased from Abcam (USA), while E-cadherin, c-Myc, and ZO1 were from Santa Cruz (USA), and Smad2, CD44 and were from Sigma (USA). The second antibody, affinity-purified, biotinylated rabbit anti-rat IgG, was purchased from Sigma (USA).


Introduction
Colorectal cancer (CRC) is one of the most prevalent malignancy worldwide and remains the third leading cause of global cancer-related morbidity and mortality (1). Distant invasion and metastasis are responsible for up to 90% of CC-associated mortality. Although great improvements in clinical diagnosis and comprehensive therapy have partly prolonged the survival, the prognosis of these patients remains poor (2). Transmembrane 4 L6 family member 1 (TM4SF1) is a member of the tetraspanin-related L6 family (TM4SF) characterized by four highly conserved transmembrane domains, two extracellular loops and a small intracellular loop (3,4). It has been reported that the tetraspanin family has 39 members (TM4SF1, TM4SF4, TM4SF5, TM4SF18, TM4SF19 and TM4SF20), which shared certain tetraspanin functions including stabilizing cell signaling complexes and regulate cell proliferation, adhesion, motility, and involved in many biological processes. TM4SF1 was initially defined as a tumorassociated antigen in 1993. Immunostaining of nanopodia at both the light and electron microsopic levels localized TM4SF1 in a regularly spaced, banded pattern, forming TMED (TM4FS1-enriched domains) that anchor nanopodia to regulate cell movement (4,5). Studies have confirmed that TM4SF1 is highly expressed in various epithelial cancer cells, including pancreatic, liver, lung and bladder cancers (6,7). Andrew reported that TM4SF1 was also upregulated in endothelial cells lining angiogenic tumor blood vessels (8). They also found that TM4SF1 serves as a molecular organizer that is essential for the formation of nanopodia and the maturation of angiogenesis. A study conducted in mouse model suggested that TM4SF1 targeting antibody-drug could significantly decreased the vascular network and may be promising to restrain tumor growth and angiogenesis (9)(10)(11). Recently, most studies about TM4SF1 had mainly focused on that TM4SF1 functions as direct target of some miRNAs (miR-141, miR-9 miR-206) and its biology function on cancer cells (12)(13)(14). However, the molecular mechanisms of TM4SF1 in CRC remain elusive. Therefore, further investigation is warranted to identify the downstream target genes of TM4SF1 that involved in CRC development.
revealed that EMT plays an important role in the enrichment of cells with CSC properties, which was believed to be the origin of cancer progression. Therefore, all these findings may open a new approach for research in CRC. In this study, we found that TM4SF1 was significantly overexpressed in CRC and positively correlated with poor prognosis. In addition, TM4SF1 silencing suppressed CRC stemness and EMT, which were critical for CRC invasion and metastasis. Mechanistically, we found that TM4SF1 promotes cell metastasis and maintains the phenotypes of EMT and cancer stemness via Wnt/β-catenin/c-Myc/SOX2 pathway in CRC.

Materials And Methods Specimens and Immunohistochemistry (IHC)
The cancer tissues (T) and paired adjacent tissues (N) were obtained from the Union Hospital from July 2012 to April 2017. None of the patients had received any radio-chemotherapy before operation.
Immunohistochemistry staining was performed as described elsewhere (16), and specific antibodies were used as follows:TM4SF1, β-catenin (1:100 dilution, Abcam, USA) and SOX2, CD133 (1:80 dilution; Abcam, USA). The expression of TM4SF1 was evaluated according to the intensity of the staining (0, 1+, 2 + and 3+) and the percentage of positive cells, which were separated by 0 (0%), 1(1-25%), 2 (26-50%), 3 (51-75%) and 4 (76-100%). And, the result is considered by the staining index (SI), SI= (intensity score in 1) × (positive score in 2). SI < 3 was classified as low expression, while SI ≥ 4 was classified as high expression. Furthermore, the study was approved by the Human Subjects Protection qPCR And WB Real-time PCR was performed as described elsewhere (6). The primers of RT-PCR were designed in our lab and synthesized by Sangon (Shanghai, China Table 2). Western blot and band density analysis were conducted as described before 22 . The specific protein bands were normalized to GAPDH and Tubulin.

Transwell Assays And Wound Healing
The migration and invasion were examined by Boyden chamber assay (8 µm pore). CRC cells were resuspended with 200 µL FBS free medium (1 × 10 5 cells) and added to the top chamber (BD, USA).
Medium supplemented with 10% FBS was added to the lower chamber. After 24 hours, the cells were fixed, stained and the number of cells in six randomly selected fields were counted under the microscope. The invasion assay was similarly conducted with a modified Boyden Chamber whose upper chamber was coated with Matrigel (BD Bioscience, USA), and the rest of protocol performed in a similar manner as the migration assay.
Cells (10 × 10 4 ) were seeded in 12-well plates in low concentration of serum fresh growth medium (2%) and scratched with a 10 µL pipette tip after cells had covered the bottom of the plate. The scratch width was measured by inverted microscope in six randomly selected fields at 0 h, 24 h, which was used to determine the migrating ability of cancer cells. Data presented were repeated three times.
After incubated with DAPI (Biosharp Biotech, Hefei, China, 1:1000) for 5 min, the cells were observed under a fluorescence microscope within 4 hours.

ChIP Assays
We performed ChIP assays following the instruction of EZ-ChIP™ kit (Millipore, Billerica, MA). The following antibodies were used for immunoprecipitation: rabbit anti-c-Myc (Abcam, UK), rabbit anti-IgG (Abcam, USA). The ChIP DNA sample or 1% total Input (5 uL) were precipitated, washed, dried, resuspended in water. The enrichment of the specific amplified region was analyzed by qRT-PCR.

Tumor xenograft and metastasis in vivo
Male 4-week-old nude mice were purchased from Beijing HFK Bioscience Co. Ltd. (Beijing, China).
Stably transfected cells were injected into the flank of 4-weeks old male Balb/c nu/nu mice (20mice in each group). We measured the size of xenografts using calipers (calculated volume shortest diameter/2 × longest diameter) weekly. The subcutaneous tumors were harvested at 4-5weeks, weighed and measured for the diameter, and then were stained by immunohistochemistry. The metastasis models were induced by tail-vein injection as described in previous studies, as the procedure described elsewhere (16). The mice were anesthetized by chloral hydrate (4%, 0.2 ml/20 g) and sacrificed by cervical dislocation and then their lung tissues were collected. We counted the number of metastatic nodules on the surface of each lung under the microscope. The experiment was completed in the Surgery Experimental Animal Center of Tongji Medical College, and was approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology.

Statistical analysis
All of the experiments were independently performed at least three times, and all results are presented as mean ± standard deviation. Pearson's correlation coefficient analysis, the Kaplan-Meier plot, logistic regression analyses were performed to assess the covariates of TM4SF1 expression.
GraphPad Prism (version 6; La Jolla, CA, USA) and SPSS statistics software (version 23; IBM, Armonk, NY, USA) were used for statistical analysis. P ≤ 0.05 was deemed to be statistically significant.

Results
Overexpression of TM4SF1 is correlated with poor prognosis nlm.nih.gov/gds) which suggested that TM4SF1 was overexpressed in CRC (P < 0.01, Fig. 1a,b). The IHC examination revealed an obvious abundance of TM4SF1 protein in CRC tissues (Fig. 1c). In addition, overexpression of TM4SF1 was significantly correlated with tumor T stage and lymph node metastasis ( Fig. 1d, e, P < 0.01), but not with the age, gender, tumor size, or tumor differentiation (P > 0.05, Table 1). Moreover, Kaplan-Meier analysis indicated that CRC patients with elevated expression of TM4SF1 suffered from poor survival ( Fig. 1f, P < 0.01), which was consistent with the studies conducted by Sveen, Smith and Marisa in R2 genomic analysis ( Fig. 1g-i, P < 0.01, https://hgserver1.amc.nl/cgi-bin/r2/main.cgi). As shown in Fig. 1j, k TM4SF1 were significantly upregulated in CRC tissues compared with normal tissues (fold change = 1.59, P < 0.01). Then we found that the expression of TM4SF1 in CRC cell lines (HCT116, LoVo, RKO, SW480, DLD1) was higher than that in normal colon epithelial cells (NCM460 cells and FHC cells, Fig. 1l).

TM4SF1 promotes cell migration, invasion and EMT in CRC cells
Specific shRNAs (sh-Control, sh-TM4SF1#1/2) and TM4SF1 plasmids were transfected in SW480 and LoVo cells and the expression of TM4SF1 was confirmed with qPCR and WB (Fig. 2a, S1a). Then wound healing assay indicated that depletion of TM4SF1 significantly suppressed scratch wound healing and TM4SF1 overexpression enhanced the migration of CRC cells (Fig. 2b, S1b). Consistent with these results, Transwell Assay confirmed that knockdown of TM4SF1 inhibited the migration and invasion of SW480 and LoVo cells. In contrast, cells with TM4SF1 overexpression exhibited more aggressive potential in migration and invasion (Fig. 2c, S1c). The qPCR and WB showed that TM4SF1 knockdown increased the expression of E-cadherin, ZO1 and decreased the expressions of Vimentin, N-cadherin MMP9 (Fig. 2d), while TM4SF1 overexpression increased the expression of Vimentin, N-cadherin, βcatenin, and decreased the expressions of E-cadherin ( Figure S1e). As shown in Fig. 2e, sh-TM4SF1transfected SW480 cells presented an increased distribution of ZO-1 on the cell membrane and a decreased expression of Vimentin in the cytoplasm or nucleus. Interestingly, TM4SF1 silencing-SW480 cells showed a conversion from spindle-like mesenchymal phenotype with obvious characteristics of the interstitial cells to cobblestone-like shape, as observed under contrast microscope (Fig. 2f).
TM4SF1 is involved in the process of the EMT induced by TGF-β1.
To investigate whether TM4SF1 involved in the EMT induced by TGF-β1, SW480 and LoVo cell lines were treated with recombinant human TGF-β1 protein at different concentrations (0, 10, 20 ng/mL) for 48 h. The results showed that TGF-β1 significantly promoted the migration and invasion of CRC cells (Fig. 3a, b, S2a, b) and increased the expressions of TM4SF1, Smad2, Vimentin, N-cadherin, and MMP9, while decreased the expression of E-cadherin (Fig. 3c). Furthermore, we found that TM4SF1deficiency remarkably decreased the potential of cell migration and invasion induced by TGF-β1 ( Fig. 3a, b), and enhanced the expression of epithelial markers (E-cadherin), and suppressed the expression of mesenchyme marker (N-cadherin, Vimentin, and Smad2, MMP9) in both protein level and mRNA level as compared to that in the control group (Fig. 3d). These results suggested TM4SF1 is involved in the process of EMT induced by TGF-β1 in CRC cells.
TM4SF1 modulates the stemness properties and enhanced the sensitivity of fluorouracil.
Cancer stemness is one of the hotspot mechanisms leading to CRC development. To elucidate whether TM4SF1 is associated with stem cell-like properties of CRC cells, we performed sphere formation assay and found that knockdown of TM4SF1 expression significantly decreased the capability of sphere formation in both cell lines, and TM4SF1 overexpression could enhance the sphere formation (Fig. 3e, S1d). To further examine the function of TM4SF1 in cell resistance to chemotherapy, we treated the infected CRC cells with fluorouracil. CCK-8 assay revealed that the viability of sh-TM4SF1 transfected cells was remarkably lower than that of the control cells (Fig. 3f). In addition, we found that the IC50 value of TM4SF1 knockdown cells was 50 nmol/L, which is lower than that of the matched group. Consistently, silencing of TM4SF1 significantly decreased the expressions of stemness markers such as CD133, CD44, SOX2, and ALDHA1, while TM4SF1 overexpression increased the expression of CD133, SOX2 ( Figure S1e). These data indicated that TM4SF1 could regulate the stemness and sensitized the cells to fluorouracil.
TM4SF1 drove EMT and cancer stemness of CRC through Wnt/β-catenin signaling To investigate the molecular mechanism of TM4SF1 in EMT and cancer stemness, we performed RNA-Seq to identify the differentially expressed genes (DEGs) between TM4SF1-silencing CRC cells and control cells. Gene set enrichment analysis indicated that Wnt signaling pathway was one of the most impaired in TM4SF1-deficient CRC cells (Fig. 4a). The depleted cells showed decreased expression of Wnt/β-catenin target genes (c-Myc, Axin2, TCF7, MMP7) (Fig. 4b).
In addition, we found that TM4SF1 knockdown reduced levels of total β -catenin and the nuclear translocation of β-catenin (Fig. 4c).
Based on these observations, we evaluated whether activation of the Wnt signaling was able to reverse the suppression effects of TM4SF1 knockdown on EMT and stemness. Therefore, TM4SF1deficient cells were differentiated treated with or without LiCL (a GSK-3β inhibitor which activates the β-catenin-mediated Wnt signaling pathway). It showed that LiCL activated the Wnt/β-catenin signaling as well as upregulated the expression of Wnt/β-catenin target genes (c-Myc, Axin2, TCF7, MMP7) in the TM4SF1 deficiency cells (Fig. 4d, S3a). In addition, LiCL remarkably suppressed the expression Ecadherin, and enhanced the expression of MMP9, Vimentin, and CD133, and CD144 in both protein and mRNA level (Fig. 4e, S3b). Further, we found that activation of β-catenin rescued the EMT phenotype and diminished TM4SF1-deficiency mediated inhibitory effect of cell migration and invasion (Fig. 4f, g) and suppression of cells stemness (Fig. 4h). These results suggested that TM4SF1 may maintain the stemness and mesenchymal phenotype in a Wnt/β-catenin signaling-dependent manner.
TM4SF1 promotes EMT of CRC cells via Wnt/β-catenin/ SOX2 signaling pathway It has been revealed that SOX2 was over-expressed in many cancer stem progenitor cells and involved in chemotherapy resistance, metastasis and recurrence of cancer (23)(24)(25). Hence, we asked whether SOX2 was necessary for the TM4SF1-induced EMT and stemness of CRC cells. We investigated the expression of SOX2 (normal vs. cancer) in GEO database (https://www.ncbi.nlm.nih.gov/gds, GSE70880) which revealed that SOX2 was overexpressed in CRC(P < 0.05 Fig. 5a). Spearman's rank correlation analysis revealed that there was a strong positive correlation between the expressions of TM4SF1 and SOX2 in CRC tissues (r = 0.53, P < 0.05, Fig. 5b).
Further, we found that SOX2 was significantly overexpressed in CRC tissues compared to the paired normal tumor adjacent tissues (Fig. 5c, d). As shown in Fig. 5e, SOX2 was significantly downregulated in the sh-TM4SF1 SW480/LoVo cells, as well as the expression of Axin2, MMP7, and c-Myc. Then we constructed stable SOX2 over-expression vector in TM4SF1 deficient cells, which showed that overexpression of SOX2 significantly potentiated the migration and invasion of SW480 and LoVo cells ( Fig. 5f, g). And SOX2 overexpression significantly promoted the sphere formation in TM4SF1-deficient cells (Fig. 5h). As shown in Fig. 5i, the SW480-SOX2 cells presented a decreased distribution of ZO-1 on the cell membrane and an increased expression of Vimentin and SOX2 in the cytoplasm.
Furthermore, we found that up-regulation of SOX2 rescued the TM4SF1-deficiency mediated downregulation of N-cadherin, Vimentin, MMP9, CD133, CD44 and up-regulation of the expressions of Ecadherin (Fig. 5j, S3c). All these results indicated that SOX2 up-regulation abolished the suppression effect of sh-TM4SF1 on EMT and stemness. On this basis, we further investigated the effect of Wnt/βcatenin activation on the TM4SF1-mediated regulation of SOX2 expression. We treated SW480/LoVosh-TM4SF1 cells with LiCL as we have described previously. It showed that activation of Wnt signaling restored the expression of c-Myc, SOX2 reduced by sh-TM4SF1 (Fig. 6a). Further, we found SOX2deficiency could significantly downregulate the expression of SOX2 but not the expression of c-Myc in the LiCL treated cells (Fig. 6b).
c-Myc directly binds the SOX2 promoter and regulates SOX2 expression in CRC cells.
In the canonical Wnt/β-catenin signaling pathway, T-cell factor/Lymphoid enhancer factor (TCF/LEF) transcription factors activate the target genes by recruiting the β-catenin transcriptional co-activator and binding to the Wnt responsive DNA elements (26). But we cannot identify a consensus TCF/LEF binding site in the promoter region of SOX2, which suggested that the Wnt/β-catenin signaling may not modulate SOX2 directly and the target genes may mediate the Wnt/β-catenin/SOX2 signaling. It has been reported that c-Myc could activate SOX2 gene transcription by binding to the SOX2 promoter in lung cancer cells (27). To further demonstrate the role of c-Myc in the regulation of SOX2 expression in CRC cells, the SW480/LoVo cells were transiently transfected with si-c-Myc or si-Control.
The results revealed that c-Myc knockdown significantly decreased the expression of SOX2 (Fig. 6c).
And overexpression of c-Myc significantly increased the expression of SOX2 in mRNA and protein level, suggesting that SOX2 may exert a transcriptional regulation of SOX2 (Fig. 6d). We identified an essential binding site for c-Myc (5′-CACATG-3′) in the SOX2 promoter (Fig. 6e). Then we performed a ChIP assay using an anti-c-Myc antibody in SW480 cells to determine whether c-Myc could bind to the sequence of SOX2 promoter region. Cross-linked chromatin was prepared from SW480 cells and immunoprecipitation was performed using either the anti-c-Myc antibody or IgG, and the sequence containing the putative c-Myc binding site, was amplified. As shown in (Fig. 6f), it demonstrated that c-Myc could bind to the SOX2 promoter. Further, we found that SOX2 knockdown could restored the migration and sphere formation in the c-Myc-overexpression cells (Fig. 6g, h). And we found that SOX2 deficiency could decreased the expression of N-cadherin,CD133,CD44, and upregulated the expression of E-cadherin. These results further demonstrate that Wnt/β-catenin/ c-Myc regulates SOX2 expression in CRC cells (Fig. 6i).

TM4SF1 Promotes Tumorigenicity And Tumor Metastasis In Mice
To validate the function of endogenous TM4SF1 in vivo, we constructed stable transfected SW480 cells, and then subcutaneously injected them into BALB/c-nu mice and then measured tumor growth in the xenograft mice model weekly (Fig. 7a). As shown in Fig. 7b, mouse implanted tumor with TM4SF1-silenced cells(n = 30) were smaller in volume and lighter in weight compared to the control mice(n = 30). We also examined the expression of TM4SF1 SOX2 and β-catenin in the xenograft by Western blotting and IHC; the results revealed that the xenograft inoculated with TM4SF1 knockdown cells showed a significantly diminished expression of TM4SF1, SOX2, β-catenin, CD133 (Fig. 7c, d).
Next, the tail vein injection tumor metastasis model was constructed. We determined the metastatic nodules in lungs 6 weeks after the inoculation by H&E staining. Mice injected with TM4SF1 knockdown cells showed less or no tumor nodules in lungs when compared with the control group. And hematoxylin and eosin staining of lung samples are shown in (Fig. 7e). These results indicate that TM4SF1 is involved in tumorigenicity and tumor metastasis in vivo.

Discussion
Although recent advances in therapy have been applied to prolong survival rate, CRC remains one of most aggressive malignancies characterized by rapid tumor recurrence and early metastasis (28).
Recently, EMT and stemness has been widely accepted as the crucial biological process for driving tumor cells invasion and metastatic dissemination from the primary tumors. Growing evidence has implied that elevated expression of TM4SF1 is positively correlated with aggressive progression and occurrence in various epithelial malignant carcinomas. A wide range of studies have demonstrated that TM4SF1 exerts a promoting effect on cell proliferation, survival via JAK2/STAT3 signaling, PI3K/AKT/mTOR related signal pathways (6,7). A recent study suggested that TM4SF1 could regulate apoptosis and cell cycle via the PARγ-SIRT1 feedback loop in bladder cancer cells (29). And several studies reported that TM4SF1 functions as the direct target of some miRNAs (miR-141, miR-9 miR-203) to promote self-renewal, invasion and migration in esophageal and breast cancer cells (12,13).
However, the role and mechanism of TM4SF1 in CRC progression and metastasis remains largely unknown remain elusive. In this study, we found that TM4SF1 was significantly overexpressed and positively correlated with depth of invasion, T stage, lymph node metastasis, and distant metastasis in patients with CRC, which suggested that TM4SF1 may serve as a potential biomarker to predict metastasis and prognosis of CRC.
EMT is regarded as the prerequisite step for initial tumor cells to be motile and invasive leading to metastasis and recurrence in many cancers (30,31). Cancer stem cells possess characteristics of selfrenewal and multipotent differentiation potential which may lead to tumorigenicity, therapeutic resistance, relapse and metastasis (32). CSCs are able to resist chemotherapy through a series of self-defense lines. In essence, CSCs are the "roots" of aggressive tumors for which we currently have no effective treatments (30). Accumulating studies revealed that EMT plays an important role in the enrichment of cells with CSC properties and therapy resistance (33). Thus, targeting biochemical composition in EMT and cancer stemness have become a frontier of cancer therapy. Jia et al. reported that TM4SF1 promoted gemcitabine resistance of pancreatic cancer via downregulating ABCB1 and ABCC1 (6). A recent study revealed that TM4SF1 coupled with DDR1 was shown to promote the cancer stem cell traits and metastatic reactivation and by PKCα-dependent JAK2/Stat3 signaling in breast cancer (32). We found that TM4SF1 knockdown suppressed the migration, invasion and EMT induced by TGF-β1. In addition, we confirmed that TM4SF1 knockdown sensitized the cells to fluorouracil and suppressed the stemness of CSCs in CRC There have been abundant research results on the abnormal activation of Wnt/β-catenin pathway in human cancers (34). It is demonstrated that most CRCs are associated with aberrant Wnt/β-catenin signaling, whose activation could increase amount of β-catenin protein in the nucleus forms complexes with TCF/LEF to regulate target gene ex-pression (Axin2, SOX4, TCF7, c-Myc, MMP7) (35).
And the Wnt/β-catenin signaling is one of the most pivotal pathways contributing to EMT and selfrenewal of cancer CSCs during tumor metastasis (36). A recent study suggested that TM4SF1 could promote migration and metastasis by positively regulating the Wnt/β-catenin signaling in hepatocellular carcinoma. Consistent with this research, we found that TM4SF1 deficiency significantly suppressed the stemness and EMT associated invasion and metastasis of CRC cells by restrained the Wnt/β-catenin signal pathway. In addition, activation of β-catenin reversed the suppression effects of TM4SF1 silencing on stemness and EMT in vitro. All these results validated that TM4SF1 promotes the migration, invasion and oncosphere formation of CRC cells via Wnt/β-catenin signaling pathway.
SOX2 is one of the key members of the SOX family gene and plays critical role in maintenance of selfrenewal and pluripotency of embryonic stem cells. Studies have revealed that SOX2 is an oncogene known to be amplified and overexpressed in the carcinogenesis, metastasis and recurrence in many cancer types (37,38). Wang et al. reported that SOX2 enhanced the capability of cell migration and invasion and was associated with poor outcomes in LSCC. In addition, studies have highlighted the central roles of Wnt/β-catenin in EMT process and stemness maintaining by regulating SOX2 in cancer cells (39). A recent study revealed that TM4SF1 overexpression increases the expression of SOX2 and NANOG, sustains the manifestation of cancer stem cell traits, and drives metastatic reactivation in the lung and brain (32). Therefore, we speculated that TM4SF1-deficiency may suppress the proficiency of EMT and sphere formation by regulating the expression of SOX2. Consistent with the speculation, we found that SOX2 overexpression reversed the down-regulation of EMT or stemness hallmarks and attenuated the suppression of stemness induced by sh-TM4SF1. Further, we validated that TM4SF1 could regulate the expression of SOX2 via the Wnt/β-catenin signal pathway.
There is strong evidence showing that the activation of Wnt/β-catenin may result in the accumulation and nuclear translocation of β-catenin (40,41).  (46). Interestingly, we identified the noncanonical E box, 5'-CACATG-3', in the sequence of SOX2 promoter. Therefore, we performed ChIP assay and found that c-Myc binded to the SOX2 promoter, suggesting that c-Myc regulates SOX2 expression through recruitment to the promoter. In our study, we found that TM4SF1 modulates the Wnt/β-catenin signaling mediated regulation of Sox2 expression via c-Myc in CRC.

Conclusions
Our study revealed that TM4SF1 overexpression is closely correlated with the clinicpathological features of CRCs and predicts an unfavorable prognosis for CRC patients. It presents a novel mechanism by which TM4SF1 could facilitate cell migration, invasion, and maintain stemness via the Wnt/β-catenin/c-Myc/SOX2 axis in CRCs. Moreover, our findings provide evidence for the roles of crosstalk between SOX2 and β-catenin signaling in CRC development. Therefore, TM4SF1 may be a potential prognostic marker and target for metastatic prediction and individualized drug therapy which may prevent tumor metastasis and improve the prognosis of cancer patients.
Declarations of Science and Technology (Wuhan, China).

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.