Cten promotes epithelial‐mesenchymal transition through the post‐transcriptional stabilization of Snail

Cten promotes cell migration however the knowledge of underlying signalling pathways is sparse. We have shown that Cten downregulates E‐cadherin, a feature of epithelial to mesenchymal transition (EMT). This prompted us to investigate whether Cten further contributed to EMT processes to regulate cell motility. The regulation of Snail by Cten was investigated following overexpression, knockdown (by RNA‐interference) or knockout of Cten in HCT116, Caco‐2 and SW620 colorectal cancer (CRC) cell lines. Subsequently, the cycloheximide (CHX) pulse chase assay was used to investigate changes in Snail protein stability and the functional relevance of Cten‐Snail signalling was investigated. Snail was identified as a downstream target of Cten signalling using multiple approaches of Cten expression manipulation. Furthermore, this activity was mediated through the SH2 domain of Cten. The CHX assay confirmed that Cten was regulating Snail at a post transcriptional level and this was through the prevention of Snail degradation. Cell migration, invasion and colony formation efficiency were increased following forced expression of GFP‐Cten but subsequently lost when Snail was knocked down, demonstrating a functional Cten‐Snail signalling axis. In conclusion, we have described a novel Cten‐Snail signaling pathway that contributes to cell motility in CRC, mediated by the stabilization of Snail protein. This finding potentially furthers the understanding of EMT regulatory networks in cancer metastasis.

epithelial cell lines, Cten was shown to regulate cell migration by a "tensin switch" mechanism whereby the upregulation of Cten was associated with a decrease in Tensin 3 expression. Downstream, differential binding to DLC1 induced cell motility through RhoA-ROCK signaling. 2,8 In CRC, and possibly cancer tissue in general, it is likely that alternative signaling mechanisms exist as firstly, we have found no evidence of a tensin switch in CRC. 9 Secondly, we found that DLC1 expression is often lost in CRC, suggesting that DLC1 is not a major player in Cten signaling. Despite the mechanistic differences, there is a common functional activity of Cten in stimulating cell migration and invasion, consequently alternative signaling mechanisms must be present. 10,11 EMT is a process whereby epithelial cells acquire a mesenchymal phenotype, enhancing cell migration. This is a process native to physiological events such as wound healing and embryogenesis and it is also likely to play a role in metastasis. [12][13][14] The loss of E-cadherin at adherens junctions causes disruption of cell-to-cell adhesion thereby allowing invasion and migration away from the primary tumor.
However this is also accompanied by additional molecular changes which induce cell motility. [15][16][17][18] Snail is a transcription factor central to the regulation of EMT through the downregulation of E-cadherin but is also known to regulate other genes associated with EMT. [12][13][14] Interestingly, we have previously shown that Cten represses E-cadherin and therefore may induce EMT. 10 Considering the data that both we and others have published, we hypothesized that Cten may induce cell motility through the regulation of additional biomarkers of EMT. Here, we demonstrate that Cten regulates the protein stability of Snail which is mediated via Cten's SH2 domain. Furthermore, we show that Cten signals through Snail to increase cell migration and invasion in CRC cell lines.

| Cell culture
HCT116, Caco-2, and SW620 colorectal cell lines, a kind gift from Prof Ian Tomlinson, were cultured in DMEM (GlutaMAX™ supplement, Thermo Fisher Scientific, Carlsbad, CA) supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were maintained at 37°C in a 5% CO 2 atmosphere. Cell line identities were verified by high resolution melting (HRM) mutation analysis as described previously. 19 2.2 | Cten deletion in SW620 CRISPR/Cas9 gene technology allows editing of the genome of cell lines. We reasoned that by knocking out both copies of Cten in SW620, we would be able to carry out both gene knockdown and gene over expression experiments in isogenic cell lines. A plasmid construct expressing GFP-tagged Cas9 and guide RNA targeting exon 3 of Cten (CCGCCAGATCAAGGTGCCACGA) (Sigma, St Louis, MO) was transfected into SW620 cells. A total of 10 µg CRISPR-Cas9 construct was transfected with 10 µL Lipofectamine according to the manufacturer's instructions. Forty-eight hours post transfection, GFP expressing cells were sorted into 96-well plates using the Astrios Cell Sorter (Beckman Coulter, High Wycombe, UK) and single cells were expanded to form clonal cell lines. Genomic DNA (gDNA) was extracted from the resulting clones using the Genelute Mammalian Genomic DNA Extraction Kit (Sigma) and the region around the CRISPR target site amplified (Supplementary Table S1). Clones were screened for mutation by HRM and those revealing a shift in melting temperature from the wild-type cell line were selected. The amplicons were cloned into a TOPO vector using the TOPO TA cloning kit (Thermo Fisher Scientific) and recombinants purified using the Genelute Plasmid Miniprep Kit (Sigma) according to the manufacturer's protocol. In order to obtain a homozygous knockout, a clone showing mutation of one allele was then expanded and put through another round of gene editing to mutate the second allele.

| Site-directed mutagenesis
We have recently shown that the conserved arginine at position 474 is essential for the functioning of the SH2 domain in Cten (manuscript in preparation). In brief, the Phusion site directed mutagenesis kit (Thermo Fisher Scientific) was used to convert the arginine to an alanine residue (GFP-Cten R474A ) according to the manufacturer's protocol. The presence of the mutation was confirmed by sequencing.

| Western blot
Cell lysates were obtained using RIPA buffer (Thermo Fisher Scientific) supplemented with phosphatase and protease inhibitor (Thermo Fisher Scientific). A total of 50 µg protein was heated at 95°C for 5 min then cooled on ice for 5 min. The protein was fractionated on a 4-12% NUPAGE Bis-Tris gel with NUPAGE MOPS running buffer (Thermo Fisher Scientific) using the NUPAGE gel electrophoresis system (Thermo Fisher Scientific). Proteins were transferred onto a PVDF membrane (GE Life Science, Chicago, IL) using the Trans Blot semi-dry transfer system (Biorad, Hercules, CA). Following blocking in 5% milk 0.01% tween PBS or 5% BSA 0.01% tween TBS, membranes were incubated with optimally diluted primary antibodies; Cten 1:10 000 (Sigma) and Actin 1:50 000 (Sigma) diluted in 5% milk 0.01% tween and PBS, Snail 1:1000 (Cell Signaling, Danvers, MA) diluted in 5% BSA 0.01% tween PBS overnight. Following washing, membranes were incubated with the appropriate anti-mouse or antirabbit secondary antibody 1:10 000 (Sigma) for 1 h at room temperature. The ECL prime detection kit (GE Life Sciences) was used for protein band visualization using X-ray film (GE Life Sciences) or the C-DiGiT Blot Scanner (LI-COR, Lincoln, NE).

| Cycloheximide chase assay
CHX inhibits translation allowing the rates of protein degradation to be evaluated. HCT116 cells were transfected with either GFP-Cten or GFP-EV and 24 h post transfection were treated with 100 μg/mL CHX (Sigma). Protein lysates were collected following 0, 1, 2, or 4 h exposure to CHX. Western blotting was performed as described above.

| Co-immunoprecipitation
Cell lysates were pre-cleared by incubating with 20 μL Protein A/G agarose beads (Thermo Fisher Scientific) at 4°C with rotation for 30 min. Lysates were centrifuged at 4°C at 13 000 rpm for 5 min and the supernatant retained. A total of 2 μg Cten antibody was added to 500 μg pre-cleared lysate and incubated rotating overnight at 4°C. A total of 30 μL Protein A/G beads were added to the IP reactions and left rotating overnight at 4°C. Separately, 500 μg of pre-cleared lysate (without antibody) was also incubated with Protein A/G beads as a negative control. The beads were pelleted by centrifugation at 13 000 rpm at 4°C for 5 min and washed twice in ice cold PBS. Beads were re-suspended in 10 μL NUPAGE loading Buffer (Thermo Fisher Scientific) and heated at 95°for 5 min, kept on ice for 5 min, and centrifuged for 2 m at 13 000 rpm before loading onto an SDS gel for Western blot analysis. A total of 50 μg lysate was loaded for the input.  Table S1). The run cycle comprised 95°C for 2 min, 40× (95°C for 3 s, annealing temperature for 30 s) and a melt

| PrestoBlue proliferation assay
PrestoBlue (Thermo Fisher Scientific) was used as an indirect method to measure the total number of live cells. A total of 1 × 10 5 cells were seeded in a 24-well plate and allowed to adhere for 3 h.

| Transwell cell migration and invasion assays
The Transwell system (Corning, Corning, NY) was used to assess changes in cell migration. The Transwell inserts (6.5 mm diameter; 8 μm pore size) were incubated in DMEM at 37°C for 1 h prior to use. Following this, 600 μL of DMEM (20% FBS) was added to the outer wells of the Transwell plate and the Transwell inserts placed inside. A total of 1 × 10 5 cells in DMEM (10% FBS) were seeded onto the Transwell insert. The plate was incubated at 37°C for 24 h.
Following this, the cells that had migrated through to the bottom of the outside well, using the higher FBS concentration as a chemoattractant, were manually counted. Triplicate wells were seeded for each experimental condition. The Transwell invasion assay was performed according to this protocol with the exception that 2 × 10 5 cells were seeded onto a Transwell insert coated in Basement membrane extract (3 mg/mL, Corning) and cells allowed to migrate for 48 h prior to counting.

| Colony formation assay
Colony formation in soft agar was used to assess anchorage independent cell growth. A total of 1 mL 1% agar layer (Sigma) containing DMEM was plated in 6-well plates. Overlaying this was 1 mL 0.7% agarose layer (Sigma) containing 2500 cells in DMEM.
Plates were incubated at 37°C for 21 days and fed with 0.5 mL of DMEM. Following this, the plates were stained with 0.005% crystal violet 4% formaldehyde for 1 h. The number of colonies of approximately >50 cells in size were manually counted and the colony formation efficiency determined (number of colonies counted/number of cells seeded × 100).

| Statistical analysis
Statistical analysis was performed using GraphPad Prism (v6). The Shapiro-Wilk test was used to test for normality. The unpaired T-test or ANOVA statistical tests were applied for experiments with two or more than two treatment groups, respectively.  Table S3). Functional evaluation of SW620 ΔCten revealed that, consistent with previously published data, loss of Cten was associated with a reduction in both cell migration and invasion (Fig. 1C,D). 7 The creation of isogenic cell lines with both the presence and absence of the full length Cten gene provides a suitable a model to study Cten biology.

| Cten is a positive regulator of Snail expression
Since we found little evidence of DLC1 expression in CRC ( Supplementary   Fig. S1), we investigated alternative Cten signaling mechanisms in this tumor type. We have shown that Cten is a regulator of E-cadherin expression, which suggests it may be involved in the regulation of EMT. 10 We investigated whether Cten may also regulate Snail expression in CRC.
The colorectal cell line HCT116, which expresses very low endogenous levels of Cten, was transfected with either GFP-Cten or GFP-EV expression constructs and the changes in Snail protein expression investigated by Western blot ( Fig. 2A and supplementary Fig. S2). Over

| Cten regulates Snail in an SH2 dependent manner
Cten contains an SH2 and a PTB domain at its C-terminus. SH2 domains partake in signal transduction events via tyrosine phosphorylation and this domain in Cten and other Tensin family members has previously been shown to be critical for its activity. 20,21 We sought to determine whether the SH2 domain of Cten was required for the upregulation of Snail expression. HCT116 cells were transfected with GFP-Cten, GFP-Cten R474A , or GFP-EV expression constructs and the changes in Snail protein expression were assessed by Western blot (Fig. 2B and supplementary Fig. S2).

| Cten increases Snail protein stability
To further investigate the mechanism of Snail upregulation by Cten we next performed qRT-PCR to determine whether this was occurring at a transcriptional or post-transcriptional level. Cten was over expressed in HCT116 and knocked down in SW620 cells. In both experiments, there was no change in Snail mRNA expression compared to the control (Fig. 3A,B). This suggests that the regulation of Snail by Cten is occurring at a post-transcriptional level.
Expression may be regulated post-transcriptionally either by increased protein synthesis or reduced protein degradation. The CHX chase assay was used to determine whether Cten stabilized Snail protein preventing its degradation (Fig. 3C). HCT116 cells following transfection of GFP-Cten or the empty vector control were treated with 100 μg/mL CHX to inhibit protein synthesis. This allowed for the tracking of protein degradation by Western blot. Our data showed that when Cten was present, Snail protein degradation was markedly delayed. In the cells transfected with GFP-Cten, Snail protein expression was still highly expressed 2 h after treatment whereas, in the control cells, Snail protein had mostly been degraded at 1 h.
Cten has been shown to form a complex with β-catenin in the nucleus. 22 As both Cten and Snail translocate between the nucleus and cytoplasm, it was hypothesized that they could form a physical complex and once in complex, Cten could prevent the degradation of Snail protein. To investigate protein binding interactions, a co-immunoprecipitation experiment was performed however, this revealed that Snail and Cten proteins did not bind to each other using this assay (Fig. 3D). Together, these results show that Cten regulates Snail protein stability but as they do not form a physical complex, this is probably due to signaling downstream of Cten mediated by the SH2 domain.

| The regulation of Snail by Cten is functionally active
Having shown that Cten regulates Snail, we next wanted to investigate whether this interaction was functionally relevant. Both Cten and Snail regulate cell invasion and migration and since Cten regulates Snail expression, it is possible that Cten may regulate these activities through Snail signaling. We have previously shown that Cten has no effect on cell proliferation 10 but before performing further assays, it was necessary to determine whether Snail had any effect on cell number. Snail was knocked down using siRNA in HCT116 and following this, the PrestoBlue assay was performed to assess cell proliferation (Fig. 4A,B). The PrestoBlue assay showed no change in activity when Snail was knocked down compared to the control. This

| DISCUSSION
EMT is considered to play a critical role in cancer metastasis endowing a cell with greater migratory capabilities as well as properties of "stemness." Although progress has been made in recent years to try and elucidate the underlying signaling mechanisms, there is still a lot about the regulation of this process that remains unknown. We have previously shown that Cten is a regulator of EMT and that it stimulates cell motility in tumor cells. 10 To our knowledge, this is the first time that it has been shown that Cten may also mediate EMT through the positive regulation of Snail expression and that this regulation is achieved through enhancing the stability of Snail protein (summarized in Fig. 6).
Using multiple approaches to modulate Cten expression, we have shown that any induced changes were followed by similar changes in Snail is a transcription factor which requires tight regulation to ensure | 2607 appropriate expression of downstream targets. It is often regulated at the protein level to ensure signals can be promptly switched off upon external stimuli and this fits with our data. 23 The exact mechanism by which this stabilization occurs is unclear. We were unable to demonstrate the formation of a complex between Cten and Snail and thus an alternative explanation must be sought.
Snail nuclear localization is essential for its transcriptional activity.
The phosphorylation of Snail regulates its export from the nucleus and subsequent degradation via the ubiquitin proteasome pathway. 23 It would be of interest to determine whether Cten signaling is involved in this process. Also of interest, is to determine whether Cten regulates the expression of Snail downstream targets. Cten has been shown to regulate E-cadherin, a known target of Snail signaling, however, this regulation was at a post-transcriptional level and since Snail is a transcriptional regulator of E-cadherin, it is unlikely that these signal in the same pathway. 10 Snail controls the transcriptional activity of a number of other genes involved in EMT in addition to genes involved in other cellular processes which could possibly also be targets of Cten. 24 We are confident that Cten can be added to the list of genes which can regulate Snail and that this relationship is important to the functional activity of Cten. We over expressed Cten while at the same time knocking down Snail to create a situation where Cten was present but there was no Snail present. This resulted in an abrogation of the effect of Cten on cell migration, cell invasion and colony formation.
These data confirm that the induction of Snail by Cten is not just a bystander phenomenon. We have previously found that Cten upregulates ILK and FAK to increase cell motility and both of these proteins are now known to play a role in EMT. 6,11 It would be of interest to determine whether ILK or FAK are signaling intermediates in the Cten-Snail pathway. The role of EMT in cancer metastasis has been well documented however most of this work has been performed in vitro. These studies were performed in cell lines and validation of these experiments in animal models is required to confirm that these effects also occur in vivo. There is accumulating in vitro evidence for the contribution of EMT to cancer metastasis however, the occurrence and relevance of EMT in vivo is debated. In mouse models of pancreatic and lung cancer, although EMT contributed to chemoresistance, it was not required by metastasising cells. 25,26 However, EMT is complex and at present is not fully understood and consequently these mouse models used may not fully recapitulate EMT processes in human cancer.
In conclusion, we have uncovered a novel mechanism of Snail regulation in CRC which increases cell motility and colony formation.
Knowledge of mechanisms regulating cell migration may help to identify novel markers for therapeutic targeting of cancer metastasis.