KCTD12 Regulates Colorectal Cancer Cell Stemness through the ERK Pathway

Targeting cancer stem cells (CSCs) in colorectal cancer (CRC) remains a difficult problem, as the regulation of CSCs in CRC is poorly understood. Here we demonstrated that KCTD12, potassium channel tetramerization domain containing 12, is down-regulated in the CSC-like cells of CRC. The silencing of endogenous KCTD12 and the overexpression of ectopic KCTD12 dramatically enhances and represses CRC cell stemness, respectively, as assessed in vitro and in vivo using a colony formation assay, a spheroid formation assay and a xenograft tumor model. Mechanistically, KCTD12 suppresses CRC cell stemness markers, such as CD44, CD133 and CD29, by inhibiting the ERK pathway, as the ERK1/2 inhibitor U0126 abolishes the increase in expression of CRC cell stemness markers induced by the down-regulation of KCTD12. Indeed, a decreased level of KCTD12 is detected in CRC tissues compared with their adjacent normal tissues and is an independent prognostic factor for poor overall and disease free survival in patients with CRC (p = 0.007). Taken together, this report reveals that KCTD12 is a novel regulator of CRC cell stemness and may serve as a novel prognostic marker and therapeutic target for patients with CRC.


KCTD12 is involved in the self-renewal ability of CRC cells in vitro and in the tumorigenesis of CRC cells in vivo.
We further explored the functions of KCTD12 in the self-renewal and tumorigenesis of CRC cells. First, as shown in Fig. 3A by the colony formation assay, the knockdown of KCTD12 in HT29 cells  significantly enhanced the cells' colony formation capacity, whereas the overexpression of KCTD12 in both DLD1 and HCT116 cells reduced this capacity (Fig. 3B). However, the alteration of KCTD12 in these cells did not affect their proliferation (Fig. 3C). These results indicate that KCTD12 is involved in the self-renewal ability of CRC cells in vitro. Second, as shown in Fig. 4, the knockdown of KCTD12 in HT29 cells promoted, whereas the overexpression of KCTD12 in DLD1 cells inhibited, tumor growth in nude mice, as measured by tumor volumes and weights. These results suggest that KCTD12 plays a crucial role in CRC tumorigenesis in vivo.
Silencing of KCTD12 enhances the drug resistance of CRC cells. As chemoresistance is another characteristic of cancer stem cells, we then sought to test whether there is a correlation between KCTD12 level and drug resistance in CRC cells. KCTD12 is a prognostic biomarker for the treatment of GIST patients with imatinib mesylate 16 , and imatinib is a potential therapeutic strategy for patients with CRC 17 . We therefore chose to use both 5-FU, a commonly used therapeutic for patients with CRC, and imatinib mesylate to test our hypothesis. As shown in Fig. 5A, HT29 cells with KCTD12 knockdown displayed enhanced viability in the presence of varying concentrations of imatinib and 5-FU. Consistently, decreases of 20% and 60% in apoptosis rates were detected in the KCTD12 knockdown HT29 cells treated with 100 μ M imatinib and 10 μ g/ml 5-FU, respectively (Fig. 5B), results that were further supported by the cleavage of both PARP and procaspase 3 (Fig. 5C). Given that chemoresistance affords CSCs drug-exclusion properties, the side population (SP) cells using Hoechst-33342 dye were examined in KCTD12 knockdown HT29 cells. The results indicated that the down-regulation of KCTD12 dramatically increased the SP abundance (Fig. 5D). Collectively, the results suggest that the down-regulation of KCTD12 enhances the resistance of CRC cells to both imatinib and 5-FU.
KCTD12 regulates CRC cell stemness via the ERK pathway. Given that KCTD12 acts as a component of the GABA B complex, downstream of which is the ERK pathway, we sought to determine whether the ERK pathway is involved in the KCTD12-mediated regulation of CRC cell stemness. As shown in Fig. 6A,B, phosph-ERK1/2 levels were dramatically increased in HT29 cells with silenced KCTD12 and decreased in DLD1 cells with overexpressed KCTD12. Moreover, U0126, an ERK 1/2 inhibitor, abrogated the increases in CD44, CD133 and CD29 levels in HT29 cells induced by the knockdown of KCTD12 (Fig. 6C). Likewise, the inhibition of the ERK pathway by U0126 reduced the sizes of spheres in KCTD12 knockdown HT29 cells (Fig. 6D). These results indicate that KCTD12 regulates CRC cell stemness via the ERK pathway.   Low KCTD12 expression indicates a poor prognosis of patients with CRC. Finally, we analyzed the clinical relevance of KCTD12 in CRC samples. As shown in Fig. 7A, the protein level of KCTD12 was significantly higher in normal tissues than in CRC tumor tissues. Immunohistochemical (IHC) staining showed that KCTD12 was localized to the cytoplasm of CRC cells (Fig. 7B) and that its expression was significantly lower in CRC tissues compared with their adjacent normal tissues (p < 0.0001, Fig. 7C). To investigate the correlation between KCTD12 expression and the clinicopathological features of patients with CRC, the 157 patient samples were divided into two groups (high and low KCTD12) based on IHC density scores. The chi-square test revealed that the KCTD12 level was strongly related to the clinical stage (p = 0.027), tumor size (p = 0.021) and vital status (p = 0.003) ( Table 1). As shown in Fig. 1D, Kaplan-Meier survival curves and the log-rank test showed that the KCTD12 expression level was significantly correlated with overall survival (OS) and disease free survival (DFS) of patients with CRC (p = 0.001). The univariate and multivariate analyses revealed that the KCTD12 expression   (Table 2). To examine the correlation between the KCTD12 expression level and CRC clinical stages, Stage I and Stage II cases were grouped together in our analysis due to a limited sample number. The results showed that the expression level of KCTD12 in stage I and II cases is higher than that in stage III and stage IV cases (p < 0.05, Fig. 7E). Taken together, these results indicate that KCTD12 may serve as a prognostic indicator for patients with CRC.

Discussion
In this report, the down-regulation of KCTD12 is detected in colorectal CSC-like cells, and a low level of KCTD12 is associated with a poor prognosis of patients with CRC. Functionally, KCTD12 regulates CRC cell stemness characteristics, such as self-renewal, tumorigenesis and drug resistance, through the ERK pathway. This is the first report to reveal that KCTD12 regulates CRC cell stemness through the ERK pathway.
CRC stem cells have been proven to initiate tumorigenesis and recurrence of CRC during progression of the disease, although how to target CRC stem cells remains a problem in this field 6 . Although multiple molecules and signaling pathways have been shown to associate with the stemness of CRC cells, the regulatory mechanisms of CRC stem cells are not yet completely elucidated. In this report, we demonstrated that KCTD12 regulates CRC cell stemness, as the silencing and the overexpression of KCTD12 promotes and inhibits, respectively, the stemness characteristics of CRC cells, such as self-renewal, tumorigenesis and drug resistance in vitro and in vivo, most likely via an ERK-dependent mechanism. Consistently, a lower level of KCTD12 is a negative independent prognostic factor for patients with CRC. This information may be useful for patients with CRC, especially for those with a low level of KCTD12. Given that MAPK/ERK pathway activation has been linked to the regulation of CRC pathogenesis, progression and chemoresistance 18 , the combination of an ERK inhibitor, such as U0126, with imatinib and/or 5-FU would be beneficial to the treatment of certain CRC patients, as this combination would increase drug sensitivities.
CSCs represent a small subpopulation of cancer cells that are capable of maintaining self-renewal, survival and chemoresistance. To settle this problem in the context of traditional chemotherapeutics, many studies have shown that the combination of chemoresistance-related proteins and agents, such as the ABC transporter 19 , the multi-pass membrane protein SLC6A6 20 and microRNAs 21 , with conventional chemotherapeutic drugs improve the efficacy of CRC treatment. The new chemotherapeutic combination (5-FU, irinotecan, and oxaliplatin) and monoclonal antibodies against the epidermal growth factor receptor prevented the metastasis of CRC and improved the survival of patients compared with what has been observed over the past decades 22 . Elevated Lgr5 levels promoted spheroid resistance to 5-FU and oxaliplatins 23 . It is clear that the discovery of more molecules regulating CSCs will improve therapeutic efficacy.
KCTD8, 12 and 16 are identified as components of the auxiliary GABA B receptor subunits 24 , but only KCTD12 generates a desensitizing GABA B receptor response 8,25 and suppresses the proliferation of GIST cells via interference with GABA B signaling 26 , which has been reported to play important roles in several human cancers, including CRC 27,28 . This notion is also supported by our finding that KCTD12 may be an important regulator of CRC cell stemness via the ERK pathway, a downstream component of GABA B receptor signaling. Although not currently understood, we will further investigate how KCTD12 modulates the ERK pathway, and we speculate that the GABA B signaling pathway may also be crucial for cell survival, tumorigenicity, and drug resistance in some cancers, including CRC, as previously shown in the literature [29][30][31] .
In summary, we have determined that KCTD12 plays an important role in the tumorigenesis of CRC progression via activation of the ERK signaling pathway and could serve as a useful biomarker for the prognosis of patients with CRC.

Materials and Methods
Human colorectal cancer clinical specimens. 157 1 and included the stages of the patients' tumors: stage Ι (N = 12), stage II (N = 38), stage III (N = 72) and stage IV (N = 35). Ten paired CRC tissue specimens and the adjacent normal tissues were collected and stored at − 80 °C immediately after surgery for the western blotting assay. All patients provided written informed consent for research purposes, according to guidelines approved by the institutional Review Board of Ethics at the Sun Yat-Sen University Cancer Center. The experimental protocols of all experiments involving human were approved by the ethical committee of Sun Yat-Sen University Cancer Center and performed in accordance with approved guidelines and regulations.
Immunohistochemistry and immunoblotting. The immunohistochemical staining for KCTD12 was performed in 157 paraffin-embedded primary CRC tissues as follows: the paraffin-embedded specimens were deparaffinized and stained with anti-KCTD12 antibody overnight. The KCTD12 expression was detected by the secondary-rabbit HRP-conjugated antibody and a DAB chromogen kit. The IHC scores for KCTD12 in CRC tissues were calculated by two independent pathologists and were composed of the score for the percentage of positively stained tumor cells and the grade of the staining intensity. The percentages of positively stained tumor cells were scored according to the following rules: 0, no positive tumor cells; 100, < 25%; 200, 25%-50%; 300, > 75%. The staining intensities were divided into no staining, weak staining, moderate staining and strong staining. The expression of KCTD12 in adjacent carcinoma and malignant tissues were compared in accordance with the above rules. Protein extractions form the CRC cells and tissues were subjected to SDS-PAGE, followed by transfer onto PVDF membranes. The membranes were incubated with the antibodies anti-KCTD12, anti-CD44, anti-CD29, and anti-CD133 (Proteintech, Chicago, IL, USA). The Hsp70 and GAPDH antibodies (SantaCruz, Santa Cruz, CA) were used as controls.
Cell proliferation assays. For the proliferation assay, HT29, DLD1 and HCT116 cells were seeded at a density of 1.5 × 10 3 cells per well in 96 well microplates and cultured for six days. For cell viability analysis, the density of cells was 5 × 10 3 cells per well in 96-well microplates. One day following seeding, cells were stained with 3-(4,5-dimethyl-2thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) dye (0.5 mg/ml, Sigma, Saint Louis, MI) for 4 h at 37 °C, the culture medium was removed, and cells were dissolved in dimethyl sulphoxide (DMSO, Sigma, Saint Louis, MI). The absorbance was measured with a multifunctional microplate reader at 490 nm. The absorbance was normalized to the absorbance at the first day and calculated. Each experiment was performed in triplicate.

RNA Isolation and qRT-PCR.
Total RNA was extracted from cells using the RNA extraction kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The RNA concentration was measured with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Rockland, DE, USA). Reverse transcription and qRT-PCR was performed with the SYBR GreenER TM two step qPCR kit (Invitrogen, Paisley, UK).

Stable cell lines.
To establish stable cell lines with KCTD12 knockdown or overexpression, we selected two effective sgRNA sequences for KCTD12 knockdown with CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) technology 32 . Stable cell lines with overexpression of KCTD12 were established based on Pbabe-retrovirus vectors. 293T cells were co-transfected with expression vectors and virus skeleton vectors. Viruses were collected after incubation for 24 h and 48 h and were then used to infect HT29, DLD1 and HCT116 cells. The stable cell lines were then selected with 0.5 μ g/ml puromycin and isolated after10 days incubation.
Flow cytometry analysis. For the apoptosis assay, the cells were harvested and washed with cold PBS, and then the cells were stained first with Annexin V (Nanjing kaiji Bio-Tek Corporation, Jiangsu, China) for 20 min at 4 °C in the dark and second with PI solution (50 μ g/ml). Cells were then analyzed using a CytomicsTM FC 500 instrument (Beckman Coulter, USA). The results were analyzed and displayed with CXP software and Flow J software.
For the assay of CD molecules expression, the cells (5 × 10 5 ) were collected and washed twice with PBS. The cells were stained with 2 μ g anti-CD44 and anti-CD133 antibodies or respective control IgG in the dark and analyzed with the equipment. The analyses were performed with Flow J software. Each experiment was performed in triplicate.
Colony formation assays. Cells were seeded in 6-well plates at a density of 5 × 10 2 cells per well and cultured for 12 days. The colonies were washed once with PBS and fixed with methyl alcohol for 30 min at room temperature. The colonies were stained with 1% crystal violet for 2 min and were counted.
Sphere formation assays. Five thousand cells per well were seeded in ultra-low attachment 6-well plates (Corning, Tweksburg, MA) and incubated in DMEM/F12 (1:1) supplemented with B27 (Invitrogen, Life Technologies Inc. Grand Island, NY), 25 ng/ml fibroblast growth factor-basic (bFGF, Sigma, St, Louis, MO) and 20 ng/ml epidermal growth factor (EGF, Sigma, St, Louis, MO). The number of spheres with a diameter of >50 μ m was quantified by Image J software.
In vivo tumorigenicity experiments. Male BALB/c nude mice (4 week old, 16-18 g) were randomly divided into 3 groups (n = 7/group) for the KCTD12 knockdown experiment and into 2 groups (n = 5/group) for the KCTD12 overexpression experiment. For tumor cell implantation, the cells with KCTD12 knockdown or KCTD12 overexpression (1.5 × 10 6 ) suspended in 100 μ l PBS were injected into the armpits of mice. The length, width and thickness of tumors were examined every two days, and the weights of tumors were calculated at the end of the experiment. All experiments were performed in accordance with the Institutional Animal Care and Scientific RepoRts | 6:20460 | DOI: 10.1038/srep20460 Use Committee of Sun Yat-sen University. All experimental protocol involving mice were approved by the ethical committee of Sun Yat-Sen University Cancer Center and performed in accordance with approved guidelines and regulations.
The inhibitor for ERK1/2 and cell lines. HT29 cells were treated with 30 μ M U0126 to inhibit the activity of the ERK1/2 signaling pathway of an equivalent concentration of DMSO as a control. The colorectal cancer cell lines HT-29, HCT116, and DLD-1 and the embryonic kidney cell line 293T were purchased from American Type Culture Collection.
Cell viability assays. For cell viability analysis after treatment with imatinib and 5-Fluorouracil (5-FU), cells were plated in 96-well microplates at a density of 5 × 10 3 cells per well and cultured overnight; This was followed by the addition of increasing concentrations of drugs and incubation for 24 h or 48 h, and then cell viability was determined by the MTT assay. For cell apoptosis analysis, cells were seeded in 6-well plates at a density of 5 × 10 5 cells per well and treated with 100 μ M imatinib for 24 h or with 10 μ g/ml 5-FU for 48 h. The apoptosis rates were detected with the Annexin V/PI kit according to the manufacturer's instructions.

SP cells assay.
The effects of KCTD12 on the SP cells fraction were evaluated using KCTD12 knockdown HT29 cells; 7 × 10 5 cells were collected and divided into two groups. One group was pretreated with 50 μ g/ml verapamil for 15 min at 37 °C, and then both groups were incubated with the DNA binding dye Hoechest33342 at a final concentration of 0.1 μ g/ml for 90 min at 37 °C with gentle agitation every 15 min. Statistical analysis. Statistical analyses were performed with the SPSS version 16.0 software (version 16.0, SPSS Inc., Chicago, IL, USA) and the GraphPad PRISM software (GraphPad Software Inc., San Diego, CA). The correlations between KCTD12 expression and OS and DFS were analyzed with Kaplan-Meier Survival and the log rank test. The relationship between KCTD12 expression and clinicopathological features of CRC cancers was determined by the Pearson Chi-Square test. For multivariate statistical analysis, a Cox regression model was used. Data were analyzed using Student's t-test or one/two way ANOV methods and represented as the means ± SEM; p < 0.05 was considered statistically significant.