LINC01615 activates ZEB2 through competitively binding with miR-3653-3p to promote the carcinogenesis of colon cancer cells

ABSTRACT

As a newly discovered cancer-related molecule, we explored the unreported mechanism of LINC01615 intervention in colon cancer.

LINC01615 expression in clinical samples and cells were detected. Effects of LINC01615 silencing/overexpression on the malignant development of colon cancer cells were analyzed through cell function experiments. Changes at the level of molecular biology were detected by quantitative real-time polymerase chain reaction and Western blot. Bioinformatics analysis and dual luciferase reporter assay were involved in the display and verification of targeted binding sequences. The rescue tests and correlation analysis examined the relationship among LINC01615, miR-3653-3p and zinc finger E-box binding homeobox 2 (ZEB2) in colon cancer cells. The xenograft experiment and immunohistochemistry were performed to verify these results.

TCGA suggested that LINC01615 was high-expressed in colon cancer, as verified in clinical and cell samples, and patients with LINC01615 overexpression suffered from a poor prognosis. Silent LINC01615 blocked the malignant development of colon cancer cells through regulating related genes expressions, while overexpressed LINC01615 had the opposite effect. LINC01615, which was targeted by miR-3653-3p, partially offset the inhibitory effect of miR-3653-3p on colon cancer cells. The downstream target gene ZEB2 of miR-3653-3p was high-expressed in colon cancer. MiR-3653-3p was negatively correlated with LINC01615 or ZEB2, while LINC01615 was positively correlated with ZEB2. Therefore, LINC01615 induced ZEB2 up-regulation, while miR-3653-3p reduced ZEB2 level. The results of in vivo studies were consistent with cell experiments.

LINC01615 competitively binds with miR-3653-3p to regulate ZEB2 and promote canceration of colon cancer cells.

Introduction

Colon cancer is a common malignant tumor of the digestive tract, which can occur anywhere from the cecum to the rectum [1], accompanied by the incidence only secondary to lung cancer and liver cancer worldwide [2]. Epidemiological studies have found that men have a greater possibility to suffer from colon cancer than women [3]. In the past 30 years, people’s living standards have been continuously improved with altered dietary structure. Most countries, including China, have witnessed the increasing incidence of colon cancer year by year [4], which has seriously endangered people’s lives and health.

At present, the treatment of colon cancer mainly resorts to surgery, chemotherapy and radiotherapy [5], among which surgery enjoys the longest history and the most obvious effects. For colon cancer patients with advanced metastasis or postoperative recurrence, chemical drug is the main treatment method [6]. However, this treatment method can not only kill the proliferating tumor cells, but also impair the function of normal cells, and has certain toxic side effects on the human body. Long-term application of the chemical drug is easy to cause gastrointestinal toxicity, bone marrow suppression, and toxic side effects of liver and kidney [7,8]. In the past few decades, the rapid development of molecular biology has enriched the theories of carcinogenic mechanisms and drug treatment research. A new generation of anti-tumor drugs and molecular-targeted drugs has come into being. Non-coding RNAs (ncRNAs) are one of the key molecules.

In 2012, the ENCyclopedia of DNA Elements research initiated by the US National Human Genome Research Institute (NHGRI) officially stated [9]: About 98% of the human genome is non-coding nucleic acid sequences, which are regulated by ncRNAs. The subsequent research continued to prove that ncRNAs are involved in the regulation of almost all physiological or pathological processes, such as embryonic development, cell proliferation, differentiation, apoptosis, infection and immune response, and about 90% of mutations, which are related to malignant tumors, cardiovascular diseases, neurological diseases, metabolic diseases, etc., are located in non-coding regions [10]. Long non-coding RNA (lncRNA) is an important functional ncRNA. A study has indicated that lncRNA occupies a vital part in many life activities such as dose compensation effect, epigenetic regulation, cell cycle regulation and cell differentiation regulation [11]. Many studies have revealed that abnormally expressed lncRNA affects the progression of colon cancer and interferes with the biological functions of colon cancer cells [12,13]. LINC01615 was discovered late, so there are fewer related reports. Ji Dong [14] screened the differentially expressed lncRNAs in hepatocellular carcinoma and found that LINC01615 has a strong correlation with extracellular matrix, which may further affect the metastasis of hepatocellular carcinoma. It was reported that LINC01615 is not only an optimal diagnostic lncRNA biomarker, but also a prognostic lncRNA biomarker of head and neck squamous cell carcinoma [15]. A recent study has shown that LINC01615 is highly expressed in colorectal cancer cells, and interacts with miR-491-5p to regulate the proliferation, migration and invasion of colorectal cancer cells [16]. We searched The Cancer Genome Atlas (TCGA) database and discovered that LINC01615 expression is abnormally increased in colon cancer. In order to find more specific molecular markers, we verified whether LINC01615 has an impact on the basic functions of colon cancer cells and clarified the corresponding molecular mechanism.

Method

1. Ethics statement

The clinical sample collection obtained the consent of the hospital (Blindedfor peer review) and the patient in written form. A total of 50 colon cancer patients (from January 2014 to December 2015) were enrolled. These patients have been diagnosed with the disease through pathological testing in this hospital. Colon cancer tissues and adjacent normal tissues were obtained surgically. All washed tissues were stored in a − 80°C refrigerator for subsequent experiments. In addition, we also tracked the survival of patients.

According to the requirements of animal ethics protection, we obtained the research permit and ordered 20 Balb/c nude mice (male, 6 weeks old) from Guangdong Medical Laboratory Animal Center (GDMLAC) for model construction. The breeding environment of nude mice was always maintained at 22°C with 55% humidity and 12 hours (h) of circulating light. We regularly checked and supplemented the feed and drinking water for the nude mice to ensure that the animals can eat and drink freely.

2. Bioinformatics analysis

Using the existing data from TCGA (https://portal.gdc.cancer.gov/) database to drill down the key genes of cancer is one of the commonly used methods in cancer research. We searched for the differential expression of LINC01615 between COAD (Colon adenocarcinoma) patients (num(T) = 275) and normal people (num(N) = 349) included in the TCGA. We also predicted the targeted miRNA of LINC01615 and its downstream genes through LncBase Predicted v.2 (http://carolina.imis.athena-innovation.gr/index.php?r=site%2Ftools) and TargetScan 7.2 (http://www.targetscan.org), respectively.

3. Cell culture

In order to ensure cell stability and experimental reliability, we purchased cells and cell culture media for in vitro researches from American Type Culture Collection (ATCC): CCD-18Co (CRL-1459; DMEM Medium, 30–2003), LS174T (CL-188; DMEM Medium), LOVO (CCL-229; ATCC-formulated F-12 K Medium, 30–2004), HCT116 (CCL-247; OptiMEM 1 Reduced Serum Medium, 31,985), HT29 (HTB-38; McCoy’s 5a Medium, 30–2007), and SW-620 (CCL-227; Leibovitz’s L-15 Medium, 30–2008). Different cells were inoculated in complete medium containing 10% fetal bovine serum suitable for growth. The BOVOGEN FBS (Fetal Bovine Serum, SFBS-X) used was uniformly provided by Wolcavi Beijing Biotechnology Co., Ltd. (China). The QP-50 carbon dioxide incubator (BIOBASE, China) maintained at 37°C with 5% CO₂.

4. Transfection

According to the experimental design, the researchers constructed overexpressed LINC01615 and silent LINC01615 (siLINC01615, silent sequence 5ʹ-AGGAGAGATTAAGACAGAAATGA-3ʹ), and purchased miR-3653-3p mimic (M; miR10018073-1-5), inhibitor (I; miR20018073-1-5) and their negative controls (MC, miR1N0000001-1-5; IC, miR2N0000001-1-5) from RIBOBIO (Guangzhou, China). Compared with the previous study, we chose the liposome transfection method: using the Lipo3000 Transfection Reagent kit (GK20006) modified by GLPBIO (USA), we transfected siLINC01615, LINC01615 overexpression plasmid, miR-3653-3p mimic, inhibitor and their respective negative controls into LOVO and HT29 cells. The cell transfection rate was detected through quantitative real-time polymerase chain reaction (qRT-PCR) method after 48 h.

5. Dual luciferase reporter assay

The dual luciferase reporter assay was used to check the reliability of the binding site by constructing the reporter plasmid of the target gene and analyzing the luciferase activity. Based on the results of bioinformatics research, we commissioned Wuhan genecreate company (China) to construct wild-type (WT) and mutant (MUT) reporter plasmids of LINC01615 or ZEB2 (zinc finger E-box binding homeobox 2). The reporter plasmids were co-transfected with miR-3653-3p mimic or miR-3653-3p mimic control into LOVO or HT29 cells according to grouping. 48 h later, we used the Dual-Lucy Assay Kit (D0010-100 T) produced by Solarbio (China) to continue incubating the transfected cells. Finally, the luciferase activity in LOVO and HT29 cells needed to be detected by GloMax 20/20 Luminometer (Promega, USA).

6. QRT-PCR

Total RNA in LOVO cells, HT29 cells, and collected colon cancer tissues and adjacent normal tissues was isolated under the action of TRIzol reagent (15,596–026, Invitrogen, USA). The concentration of RNA could be measured with the aid of METTLER TOLEDO LockPath™ Spectrophotometer (USA). According to the concentration calculated, the corresponding amount of RNA was added to the cDNA reaction system (Code No. 6110A, TaKaRa, Japan). Based on the requirements of the SYBR Green method, cDNA, primers, SYBR green reagent (A25742, ThermoFisher, USA) and DEPC water were combined into a 20 μl qRT-PCR system in a ratio of 1:1:5:3. The 96-well plate was placed into the StepOnePlus Real-Time PCR Systems (Applied Biosystems, USA) in line with the following settings: pre-denaturation at 95°C for 10 minutes (min), 40 cycles of denaturation at 95°C for 15 seconds (s), and annealing at 58°C for 1 min. The mRNA expression of the detection gene was calculated by 2^−ΔΔCT method [17]. The primers synthesized by Shanghai Sangon Company were listed in Table 1. It was worth noting that β-actin, GAPDH and U6 were internal references.

Table 1.

Primers for qRT-PCR

Genen	Forward primer (5ʹ-3ʹ)	Reverse primer (5ʹ-3ʹ)
LINC01615	TTGTACGAGCCAAGGCAGTG	GTGCTGTCTGCTCAACCAGT
p53	TGACACGCTTCCCTGGATTG	ACCATCGCTATCTGAGCAGC
p16	TTAGACACCTGGGGCTTGTG	AGGACCTTCGGTGACTGATGA
c-myc	AGGGAGATCCGGAGCGAATA	AGGGAGATCCGGAGCGAATA
Bcl-2	GAACTGGGGGAGGATTGTGG	ACTTCACTTGTGGCCCAGAT
cyclin D1	AGCTGTGCATCTACACCGAC	CTTCTGCTGGAAACATGCCG
miR-3653-3p	GGACTTTTAGGGGCAGCTGT	CGGCAATTGCACTGGATACG
Bax	CAGAGGCGGGGGATGATTG	GAGCTAGGGTCAGAGGGTCA
VEGF	GTCCTGGAGCGTGTACGTTG	CTTCCGGGCTCGGTGATTTA
E-cadherin	CAGGCCTCCGTTTCTGGAAT	GGTGTATACAGCCTCCCACG
N-cadherin	GTGCATGAAGGACAGCCTCT	CCGTGGCTGTGTTTGAAAGG
Vimentin	TCCGCACATTCGAGCAAAGA	ATTCAAGTCTCAGCGGGCTC
β-actin	CTCGCCTTTGCCGATCC	TCTCCATGTCGTCCCAGTTG
U6	CTCGCTTCGGCAGCACA	AACGCTTCACGAATTTGCGT
GAPDH	TGTGGGCATCAATGGATTTGG	ACACCATGTATTCCGGGTCAAT

7. Cell Counting Kit-8 (CCK-8)

The suspension of LOVO cell or HT29 cell at a concentration of 1 × 10⁵ cells/ml was transferred to a 96-well plate (500 μl/well) for incubation. We set two different time points, i.e. 24 h and 48 h, for detecting cell viability. After incubation, 10 μl of CCK-8 solution (GK10001, GLPBIO, USA) that could bind to mitochondria in the cells was added to the cells. Post 4-h binding, the MR-96A microplate reader (mindray, China) was used to detect the absorbance (OD) of LOVO or HT29 cells at 450 nm.

8. Tube formation assay

Human Umbilical Vein Endothelial Cells (HUVECs, CL-0122, Procell, China) possessed the ability to form the inner wall of blood vessels. Therefore, we co-cultured LOVO or HT29 cells with HUVECs for 48 h to observe the effects of LINC01615 and miR-3653-3p on the angiogenesis ability of HUVECs. The 96-well plate was inoculated with 50 μl Matrigel (354,234, BD, USA). It’s worth noting that Matrigel purchased was solid and needed to be melted into liquid first. HUVECs at a concentration of 1 × 10⁵ cells/ml were seeded in the gel-containing wells. After the 24 h routine incubation, the field of view of the DYS-40 microscope (DIANYING, China) was magnified 100 times to observe the angiogenesis in the well.

9. Cell migration and invasion experiments

The Transwell chamber procured from Corning could not only analyze the invasion ability of cells, but also observe cell migration changes. The main materials used in the experiment were provided by Corning: Transwell System (3464) and Matrigel (356,234). In the migration experiment, the upper layer where the cell seeded needn’t to be covered with Matrigel, and the lower layer needed to be filled with complete medium (600 µl). Differently, the invasion experiment required the cells to be seeded on the upper layer of the Matrigel-covered chamber. After 48 h of routine culture, the cells in the lower layer were collected and fixed. Giemsa stain (48,900, Sigma-Aldrich, Germany) was used to stain for 30 min at room temperature. The changes in cell migration and invasion were observed with a microscope at magnification of 250 times.

10. Western blot

This part of the experiment referred to the experimental method of Zhu et al. [17]. The reagents in the ProteoPrep® Total Extraction Sample Kit could be used to fully extract the total protein in tissue homogenate or colon cancer cells. Similarly, the protein concentration was detected by using BCA kit (P0012S, Beyotime, China). The protein was stabilized by the high temperature denaturation method and transferred onto the membrane carrier by Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). There were many types of this membrane carrier. We singled out the widely used Roche PVDF Western blotting Membranes (3,010,040,001, Switzerland). Bovine serum albumin (BSA) blocking solution (5%) treatment for 2 h effectively impeded the loss of the protein on the membrane during operation (room temperature). The antigen-antibody immune response was divided into two parts: on the first day, it was bound to the specific antigen (primary antibody) overnight (4°C); on the second day, it was bound to the secondary antibody that recognized the primary antibody for 2 h (room temperature). Finally, the step of chemiluminescence was the key to observe and analyze protein expression, requiring the participation of ECL luminescent fluid and Gel Doc™ XR+ system (BIORAD, USA). The final results were expressed by analyzing the gray value of the band on the film. The primary antibodies (Abcam, UK) used in the experiment were Anti-p53 antibody (1:500, ab26, 53 kDa), Anti-Bax antibody (1:1000, ab32503, 21 kDa), Anti-Bcl-2 antibody (1:1000, ab59348, 26 kDa), Anti-VEGF antibody (1:2000, ab1316, 21 kDa), Anti-E-cadherin antibody (1:10,000, ab40772, 97 kDa), Anti-Vimentin antibody (1:1000, ab92547, 54 kDa), Anti-N-cadherin antibody (1:500, ab18203, 130 kDa), and Anti-β-actin antibody (1:5000, ab8226, 42 kDa). The secondary antibodies (Abcam, UK) used in the experiment were Rabbit Anti-Mouse antibody (1:10,000, ab6728) and Goat Anti-Rabbit antibody (1:10,000, ab6721).

11. β-galactosidase staining assay

Cell senescence was detected by β-galactosidase staining Kit (C0602, Beyotime, Nanjing, China) according to the manufacturer’s instructions. In brief, LOVO or HT29 cells were fixed with fixative solution for 15 min at room temperature, and stained with the staining solution provided with the kit. The stained cells were observed and photographed under the microscope.

1. 12. The transplantation tumor inoculation experiment

The experiment process was strictly followed the principles of animal protection. After 5 days (d) of rearing, nude mice were randomly divided into siLINC01615 group (n = 10) and small interfering RNA negative control group (siNC, n = 10) according to their body weight. Colon cancer cells transfected with siLINC01615 or siNC were adjusted to the cell suspension at a concentration of 5 × 10⁶ cells/ml. The cell suspension was inoculated in a total volume of 0.1 ml into the unilateral armpit of nude mice according to the experimental requirements. At the same time, in order to ensure the timeliness of cell transfection, we supplemented the injection every 4 d.

The feeding process needed the monitor of volume changes of the transplanted tumor. We measured the tumor volume (mm³) on the 5th, 10th, 15th, 20th, 25th, and 30th d after the injection. Sigma brand (USA) pentobarbital sodium powder (P3761-25 G) was formulated into a 0.5% anesthesia solution with saline. On the 30th d, the nude mice were intraperitoneally injected (dose 100 mg/kg) with anesthesia solution and sacrificed. We took out the entire transplanted tumor tissues from the armpits of the sacrificed nude mice, and recorded the appearance and weight.

13. Immunohistochemistry

For the convenience of preservation, a part of the removed transplanted tumor tissues were made into paraffin-embedded sections. The paraffin was washed off, and then the sectioned tissues were immersed in Triton X-100 solution (P1080) and antigen retrieval reagent (C1032) provided by Solarbio. The processing time and temperature were in accordance with the reagent instructions. At room temperature, a blocked goat serum stock solution (SL038, Solarbio, China) was used to cover the tissue on the slice for 1 h. The following operation principle was similar to Western blot. Anti-ZEB2 antibody (ab230561) and Goat Anti-Rabbit antibody (ab205718), acting as primary and secondary antibodies, respectively, were used to incubate with the tissues for 24 h and 2 h, with the incubation temperature of the primary antibody being set at 4°C, and the secondary antibody being set at room temperature. The color-developed sections were observed by DYS-40 biological microscope (DIANYING, China) in 5 different fields (magnification 200 ×).

14. Statistical analysis

All the data in the study were obtained through SPSS 22.0 (IBM, USA) analysis and statistics. For different situations, we used the following different statistical methods: one-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test was used for the difference comparison between different groups; paired t-test was applied to compare the differential expressions of LINC01615, miR-3653-3p and ZEB2 in colon cancer tissues and adjacent normal tissues; and Pearson’s Correlation Coefficient was used in the correlation analyses of LINC01615, miR-3653-3p and ZEB2. Once SPSS reached a P < 0.05 result, we considered the difference to be statistically significant.

Result

1. The expression of LINC01615 in colon cancer and its prognostic influence

TCGA data showed that LINC01615 presented an abnormally increasing trend in COAD patients (Figure 1a). After verifying through clinical tissues and cells, we found that LINC01615 was evidently high-expressed in colon cancer tissues and cells (Figure 1b and 1d, p<0.05). By tracking the prognosis of patients, we found that patients with high expression of LINC01615 had lower 5-year survival rate than those with low expression of LINC01615 (Figure 1c, p = 0.0087). This signified that the aberrantly expressed LINC01615 is related to the prognosis of colon cancer patients. In addition, the relationship between LINC01615 expression and clinicopathological characteristics in colon cancer patients was analyzed, with the result indicating that LINC01615 was obviously correlated to tumor, node, and metastasis (TNM) stage and lymphatic metastasis (Table 2).

Figure 1.

The expression of LINC01615 in colon cancer and its prognostic influence.

(A) TCGA (The Cancer Genome Atlas, https://portal.gdc.cancer.gov/) database provided differential expression data of LINC01615 in COAD (Colon adenocarcinoma) patients and normal people. num (T) = 275, num (N) = 349. (B) The expression of LINC01615 in adjacent normal tissues and colon cancer tissues was compared by qRT-PCR (quantitative real-time polymerase chain reaction). GAPDH was an internal reference. Paired t-test was used to compare the differential expression of LINC01615 in colon cancer tissues and adjacent normal tissues. (C) 5-year survival rate analysis of patients with high or low LINC01615 expression. (D) The expression of LINC01615 in normal intestinal epithelial cells (CCD-18Co) and colon cancer cells (LS174T, LOVO, HCT116, HT29, SW-620) was compared by qRT-PCR. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups. GAPDH was an internal reference. All experiments were repeated three times to average. *p < 0.05, **p < 0.01, ***p < 0.001 vs. CCD-18Co

Table 2.

Relationship between LINC01615 expression and clinicopathological characteristics in colon cancer patients

Characteristics	LINC01615 expression		p value
	High expression(n = 25)	Low expression(n = 25)
Age(years)
≦60	12	11	0.777
> 60	13	14
Sex
Male	11	15	0.258
Female	14	10
Histological differentiation
Well	10	8	0.778
Moderate	9	9
Poor	6	8
TNM stage
I+ II	7	16	0.011
III+IV	18	9
Lymphatic metastasis
Negative	5	15	0.004
Positive	20	10

*Statistically significant(p < 0.05).

2. LINC01615 silencing blocked the further malignant development of colon cancer cells, while overexpressed LINC01615 generated the opposite effects.

We up-regulated or down-regulated the expression level of LINC01615 in colon cancer cells by exogenous means (Figure 2a-B, p < 0.001). In the detection of the basic functions of cancer cells (Figure 2c-f and Figure 3a-d), LINC01615 silencing was proved to successfully weaken the viability, migration and invasion of LOVO and HT29 cells, and reduce the angiogenesis ability of HUVECs, while LINC01615 overexpression brought completely opposite effects to colon cancer cells and HUVECs (p < 0.05). Based on the changes in these physiological functions, we detected the mRNA expressions of related genes (Figure 4a-b), revealing that the levels of p53 and p16 were promoted and those of C-myc, Bcl-2 and Cyclin D1 were inhibited after LINC01615 was silenced (Figure 4a, p < 0.001). As before, the overexpression of LINC01615 produced the opposite regulatory effect on these genes (Figure 4b, p < 0.001). In addition, LINC01615 silencing promoted the senescence of LOVO cells, while LINC01615 overexpression inhibited the senescence of HT29 cells (Figure 4c-d).

Figure 2.

LINC01615 silencing weakened the viability of colon cancer cells and the angiogenesis ability of HUVEC, while overexpressing LINC01615 did the opposite.

(A-B) The transfection rate of silenced/overexpressed LINC01615 was tested by qRT-PCR. GAPDH was an internal reference. (C-D) The effect of silenced/overexpressed LINC01615 on cell viability was tested by CCK-8 (Cell Counting Kit-8) assay. (E-F) The effect of silenced/overexpressed LINC01615 on the angiogenesis ability of HUVEC (Human Umbilical Vein Endothelial Cells) was tested by tube formation experiment. The magnification was 100 × . All experiments were repeated three times to average. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups (Figure A-F). *p < 0.05, ***p < 0.001 vs. siNC (negative control of small interfering RNA); ^#p < 0.05, ^##p< 0.01, ^###p < 0.001 vs. NC.

Figure 3.

LINC01615 silencing impeded the migration and invasion of colon cancer cells, while the overexpression of LINC01615 generated the opposite effect.

(A-B) The effect of silenced/overexpressed LINC01615 on cell migration was tested by the Transwell assay. The magnification was 250 × . (C-D) The effect of silenced/overexpressed LINC01615 on cell invasion was detected by the Transwell assay. The magnification was 250 × . All experiments were repeated three times to average. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups (Figure A-D). ***p < 0.001 vs. siNC;   ^###p < 0.001 vs. NC

Figure 4.

LINC01615 silencing promoted the expressions of p53 and p16 and cell senescence, and inhibited the expressions of C-myc, Bcl-2 and Cyclin D.

(A-B) The effect of silenced/overexpressed LINC01615 on genes related to cell physiological function was measured by qRT-PCR. β-actin was an internal reference. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups. (C-D) β-galactosidase staining assay was performed in colon cancer cells transfected with LINC01615 and siLINC01615. ***p < 0.001 vs. siNC; ^##p < 0.01, ^###p < 0.001 vs. NC

3. MiR-3653-3p, which was targeted by LINC01615, was lowly expressed in colon cancer tissues.

LncBase provided the binding sequences of LINC01615 and miR-3653-3p (Figure 5a). Based on these sequences, we verified their binding relationship through dual luciferase reporter assay. Figure 5b-c fully illustrated this point that the detection activity in the LINC01615-WT + miR-3653-3p M group was overtly lower than that in LINC01615-WT + miR-3653-3p MC group (p < 0.001). Thus, we tested the expression differences of miR-3653-3p in clinical samples. The results uncovered that miR-3653-3p was expressed oppositely to LINC01615, and its expression in colon cancer tissues was considerably suppressed (Figure 5d, p < 0.001). Moreover, siLINC01615 up-regulated the expression of miR-3653-3p, while overexpressed LINC01615 exerted a down-regulating effect (Figure 5e-F, p < 0.001).

Figure 5.

MiR-3653-3p, which was targeted by LINC01615, was low-expressed in colon cancer tissues.

(A) LncBase Predicted v.2 (http://carolina.imis.athena-innovation.gr/index.php?r=site%2Ftools) provided the binding sequence of LINC01615 and miR-3653-3p. (B-C) The dual luciferase reporter assay was used to verify the binding sequence of LINC01615 and miR-3653-3p. Unpaired t-test was used to compare the difference between different groups. (D) The expression of miR-3653-3p in adjacent normal tissues and colon cancer tissues was compared by qRT-PCR. U6 was an internal reference. Paired t-test was used to compare the differential expression of miR-3653-3p in colon cancer tissues and adjacent normal tissues. (E-F) The effect of silenced/overexpressed LINC01615 on miR-3653-3p was tested by qRT-PCR system. U6 was an internal reference. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups. All experiments were repeated three times to average. ^{^^^}p < 0.001 vs. MC (miR-3653-3p mimic control); ***p < 0.001 vs. siNC; ^###p < 0.001 vs. NC.

4. The role of miR-3653-3p in regulating the malignant development of colon cancer cells was partially offset by LINC01615.

After miR-3653-3p inhibitor/mimic was transfected into cells, the expression of miR-3653-3p also underwent notable down-regulation/up-regulation (Figure 6a-B, p < 0.001). The basic function tests of cancer cells (Figure 6c-f and Figure 7a-d) unraveled that miR-3653-3p inhibitor facilitated the viability, migration and invasion of cancer cells, and strengthened the angiogenesis ability of HUVECs. On the contrary, miR-3653-3p mimic showed strong cancer cell inhibition effects (p < 0.05). More importantly, siLINC01615 and overexpressed LINC01615 correspondingly neutralized the regulatory effects of miR-3653-3p inhibitor and mimic in colon cancer cells (Figure 6c-f and Figure 7a-D, p < 0.05).

Figure 6.

The effects of miR-3653-3p on the viability of colon cancer cells and the angiogenesis of HUVEC were partially offset by LINC01615.

(A-B) The transfection rate of miR-3653-3p inhibitor/mimic was detected by qRT-PCR. U6 was an internal reference. (C-D) The effects of LINC01615 and miR-3653-3p on cell viability were tested by CCK-8 experiment. (E-F) The effects of LINC01615 and miR-3653-3p on the angiogenesis ability of HUVEC were measured by tube formation experiment. The magnification was 100 × . All experiments were repeated three times to average. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups (Figure A-F). *p < 0.05, **p < 0.01, ***p < 0.001 vs. IC (miR-3653-3p inhibitor control) + siNC; ^#p < 0.05, ^###p < 0.001 vs. I + siNC; ^{^}p < 0.05, ^{^^}p < 0.01 vs. IC + siLINC01615; ^&p < 0.05, ^&&p < 0.01, ^&&&p < 0.001 vs. MC (miR-3653-3p mimic control) + NC; ^Δp<0.05, ^ΔΔp<0.01, ^ΔΔΔp<0.001 vs. M + NC; ^▲p < 0.05, ^▲▲▲p < 0.001 vs. MC + LINC01615.

Figure 7.

The effects of miR-3653-3p on the migration and invasion of colon cancer cells were partially offset by LINC01615.

(A-B) The effects of LINC01615 and miR-3653-3p on cell migration were tested by the Transwell assay. The magnification was 250 × . (C-D) The effects of LINC01615 and miR-3653-3p on cell invasion were detected by the Transwell assay. The magnification was 250 × . All experiments were repeated three times to average. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups (Figure A-D). ***p < 0.001 vs. IC + siNC; ^###p < 0.001 vs. I + siNC; ^{^^^}p < 0.001 vs. IC + siLINC01615; ^&&p < 0.01, ^&&&p < 0.001 vs. MC + NC; ^ΔΔp<0.01, ^ΔΔΔp<0.001 vs. M + NC; ^▲▲▲p < 0.001 vs. MC + LINC01615.

The mRNA and protein expressions of related genes also produced similar regulatory effects (Figure 8a-F, p < 0.05): miR-3653-3p inhibitor down-regulated the expressions of p53, Bax and E-cadherin, but up-regulated Bcl-2, VEGF, Vimentin and N-cadherin expressions; miR-3653-3p mimic had the complete opposite effect; siLINC01615 or overexpressed LINC01615 partially offset the regulation of miR-3653-3p inhibitor/mimic. This part fully demonstrated that LINC01615 could competitively bind to miR-3653-3p.

Figure 8.

The role of miR-3653-3p in regulating the expressions of genes related to physiological functions of cancer cells was partially offset by LINC01615.

(A-B and D-E) The effects of LINC01615 and miR-3653-3p on genes related to cell physiological functions were detected by qRT-PCR. (C and F) The effects of LINC01615 and miR-3653-3p on genes related to cell physiological functions were detected by Western blot. β-actin was an internal reference. All experiments were repeated three times to average. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups (Figure A-F). *p < 0.05, **p < 0.01, ***p < 0.001 vs. IC + siNC; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs. I + siNC; ^{^}p < 0.05, ^{^^}p < 0.01, ^{^^^}p < 0.001 vs. IC + siLINC01615; ^&&&p < 0.001 vs. MC + NC; ^ΔΔΔp<0.001 vs. M+ NC; ^▲p < 0.05, ^▲▲p < 0.01, ^▲▲▲p < 0.001 vs. MC + LINC01615.

5. ZEB2 was a downstream target gene of miR-3653-3p and was highly expressed in colon cancer.

TargetScan provided the binding sequences of miR-3653-3p and ZEB2 (Figure 9a). The following verification experiments were carried out based on these sequences (Figure 9b-c). The declined activity in the ZEB2-WT group was sufficient to prove the combination of the two (p < 0.001). The following cell expression test results manifested that the expression of ZEB2 was promoted by miR-3653-3p inhibitor,but was inhibited by miR-3653-3p mimic (Figure 9d-E, p < 0.01). Of course, these adjustments were partially offset by siLINC01615 or overexpressed LINC01615 (Figure 9d-E, p < 0.001).

Figure 9.

ZEB2 was a downstream target gene of miR-3653-3p, which was highly expressed in colon cancer.

(A) TargetScan 7.2 (http://www.targetscan.org) predicted the binding sequence of miR-3653-3p and ZEB2 (zinc finger E-box binding homeobox 2). (B-C) The dual luciferase reporter experiment was used to verify the binding sequence of miR-3653-3p and ZEB2. Unpaired t-test was used to compare the difference between different groups. ^{^^^}p < 0.001 vs. MC. (D-E) Effects of LINC01615 and miR-3653-3p on ZEB2 expression in cells were tested by qRT-PCR. β-actin was an internal control. One-way ANOVA followed by Dunnett’s post hoc test was used for the difference comparison between different groups. **p < 0.01, ***p < 0.001 vs. IC + siNC; ^###p < 0.001 vs. I + siNC; ^{^^^}p < 0.001 vs. IC + siLINC01615; ^&&&p < 0.001 vs. MC + NC; ^ΔΔΔp<0.001 vs. M + NC; ^▲▲▲p < 0.001 vs. MC + LINC01615. (F) The expression of ZEB2 in adjacent normal tissues and colon cancer tissues was compared by qRT-PCR. β-actin was an internal reference. Paired t-test was used to compare the differential expression of ZEB2 in colon cancer tissues and adjacent normal tissues. (G-I) The correlation of LINC01615, miR-3653-3p and ZEB2 in colon cancer tissues was obtained by the Pearson’s correlation coefficient analysis method. Pearson’s Correlation Coefficient was used in the correlation analysis of LINC01615, miR-3653-3p and ZEB2. All experiments were repeated three times to average.

Clinical sample inspection and correlation analysis unearthed that the expression of ZEB2 in colon cancer tissues was abnormally increased (figure 9f, p < 0.001), and was negatively correlated with miR-3653-3p expression (Figure 9h, p < 0.001, r = −0.7055), yet was positively correlated with LINC01615 (Figure 9i, p < 0.001, r = 0.4895). There was no doubt that LINC01615 and miR-3653-3p had a negative regulatory relationship in colon cancer (Figure 9g, p < 0.001, r = −0.7328).

6. Silencing of LINC01615 inhibited the growth of transplanted tumor and ZEB2 expression, whereas activating miR-3653-3p.

In the in vivo study, LINC01615 silencing conspicuously impeded the growth of transplanted tumors (volume and weight, Figure 10A-E, p < 0.001). The results on the expressions of ZEB2 and miR-3653-3p were also consistent with the results of in vitro studies. In detail, silenced LINC01615 inhibited the positive expression of ZEB2 (Figure 10F), while activating miR-3653-3p in tissues (Figure 10G-H, p < 0.001).

Figure 10.

Silencing of LINC01615 inhibited the growth and ZEB2 expression of transplanted tumor, whereas activating miR-3653-3p.

(A-B) Silencing of LINC01615 suppressed the volume of transplanted tumor. (C) The appearance and morphological changes of the transplanted tumor were recorded. (D-E) Silencing of LINC01615 suppressed the weight of the transplanted tumor. (F) Immunohistochemical staining was used to analyze the effect of siLINC01615 on the positive expression of ZEB2. Unpaired t-test was used to compare the difference between different groups (Figure A-B, D-E, G-H). The magnification was 200 × . (G-H) QRT-PCR was used to analyze the effect of siLINC01615 on the expression of miR-3653-3p. β-actin was an internal reference. All experiments were repeated three times to average. ***p < 0.001 vs. siNC.

Discussion

At present, the specific molecular biological mechanism of the occurrence and development of colon cancer has been yet to be fully annotated. The continuous development of ncRNA research has proposed new directions for elucidating the pathogenesis of colon cancer. In this study, we artificially regulated LINC01615 to be overexpressed or silenced in colon cancer cells, and observed the effects of LINC01615 on the physiological functions of colon cancer cells. Obviously, LINC01615-stimulated cancer cells augmented the rate of canceration. Extensive researches on lncRNA have enabled scientists to gradually clarify several ways in which lncRNA participates in cell regulation [18]. On the one hand, lncRNA combines with a single protein or protein complex and simultaneously recognizes DNA or RNA sequences, thereby helping protein complexes to regulate specific sites. On the other hand, lncRNA acts as a molecular scaffold and makes a profound impact upon the formation of protein complexes. The last and most widely used method by researchers is the competing endogenous RNAs (ceRNA) pathway. It means that lncRNA can act as a ceRNA and regulate gene expression by competitively binding to miRNA. Therefore, we retrieved and obtained miR-3653-3p that bound to LINC01615. Limited studies on LINC01615 have reported that LINC01615 potentially affects the extracellular matrix and further impacts on the metastasis of hepatocellular carcinoma [19]. In addition, LINC01615 is reported as an oncogene in colon cancer, and overexpression of LINC01615 boosts the proliferation, migration and invasion of colorectal cancer cells [16]. In this study, we also authenticated that LINC01615 functioned as an oncogene, and activated ZEB2 through competitively binding with miR-3653-3p to promote the carcinogenesis of colon cancer cells.

MiR-3653 is located on human chromosome 22. A research by Preetjote Gill et al. showed that miR-3653 could increase the risk of pancreatic neuroendocrine tumor metastasis [20]. Chen et al. put forward that the down-regulation of miR-3653 in gliomas often foreboded a poorer survival prognosis [21]. As a newly discovered miRNA, miR-3653-3p has only been reported to be differentially expressed in monozygotic twins, and may have the role of identifying monozygotic twins [22]. We have confirmed that the expression of miR-3653-3p in colon cancer is inhibited, and the up-regulation prevents further canceration of colon cancer cells. Aiming at verifying the relationship between LINC01615 and miR-3653-3p, we confirmed, for the first time, that LINC01615 can competitively bind to miR-3653-3p, inhibit the expression of miR-3653-3p, and offset the tumor-suppressive effects of miR-3653-3p.

In the next screening of downstream target gene, we found that miR-3653-3p can bind to the cancer-promoting factor ZEB2. ZEB2 is encoded by the ZFHXIB gene on the 2q22 chromosome, and belongs to the E-box binding zinc finger protein family of zinc finger structural transcription factors [23]. ZEB2 binds to the 5ʹ-CACCT sequence to act as a transcription factor and a promoter of downstream genes, thereby exerting a transcriptional inhibitory effect [23,24]. Many studies have pinpointed that ZEB2 plays a central role in the occurrence and development of tumors, such as gastric cancer, ovarian cancer, pancreatic cancer, and liver cancer [25]. In these tumors, ZEB2 not only promotes the proliferation of tumor cells, but also accelerates tumor cell invasion and metastasis. We confirmed the binding of miR-3653-3p and ZEB2 and found that ZEB2 can be activated by the positively correlated LINC01615and partially silenced by the negatively correlated miR-3653-3p. Studies have disclosed that ZEB2 is involved in the regulation of tumor cell cycle and the process of epithelial-mesenchymal transition (EMT) and promotes cell proliferation. ZEB2 can directly bind with the cyclin promoter sequence to inhibit its expression and cause the cell to block the G1 phase [26]. Besides, ZEB2 can combine with the E-cadherin promoter, thus suppressing the transcription of E-cadherin and inducing the transformation process of EMT, so as to enhance the invasion and metastasis of cells [27]. The evidence above makes our results more reliable.

The occurrence of colorectal cancer is a complex process incorporating multiple genes and steps. For example, once the balance between proto-oncogenes and tumor suppressor genes is broken, the function of cell apoptosis would be restricted, and telomerase would be activated [28]. The dysregulation of proto-oncogenes and tumor suppressor genes has always been the key to activating the cancer response. Colon cancer will form when colon cancer-related proto-oncogenes that control cell proliferation are continuously or excessively expressed, and tumor suppressor genes are not expressed or inactivated [29]. Therefore, we examined the regulation of LINC01615 on the expressions of proto-oncogene (C-myc) and tumor suppressor genes (p53 and p16). It was found that LINC01615 effectively inhibited the expression of p53 and p16, and up-regulation the activity of C-myc.

Of course, tumor metastasis and recurrence are inseparable from the regulation of EMT. EMT is the biological process of phenotypic transformation from polarized epithelial cells to mesenchymal cells [30]. After the EMT, epithelial cells undergo a short-term structural change, their polarity is lost, their contact with surrounding cells and extracellular matrix is reduced, and they obtain stronger migration and motility capabilities. In the meantime, the epithelial phenotype, such as the expression of E-cadherin, is gradually lost, and the characteristic phenotype of mesenchymal cells is obtained, such as the up-regulation of Vimentin and N-cadherin expressions [31,32]. In this study, LINC01615 silencing stimulated the up-regulation of E-cadherin, but inhibited the expressions of Vimentin and N-cadherin, and meanwhile offset the reverse regulation of miR-3653-3p inhibitor. It verified that the combination of miR-3653-3p and its ceRNA LINC01615 prevented the EMT process in colon cancer cells.

Additionally, with in-depth researches on tumors, scientists have realized that the growth, invasion and migration of malignant tumors are closely related to angiogenesis [33,34]. In malignant tumors, angiogenesis is dense and growing rapidly. On the one hand, the blood vessels in the tumor tissues can fully provide enough nutrients for tumor cell formation. On the other hand, due to the immature tumor blood vessels, the tumor cells can easily metastasize to the blood tract. As a key growth-promoting factor for angiogenesis, VEGF can activate host vascular endothelial cells, promote cell division and proliferation, and form tumor neovascularization [35,36]. This explains our experimental results that LINC01615 up-regulated the expression of VEGF, thereby promoting the angiogenesis ability of HUVEC.

In summary, we proved that LINC01615, as the ceRNA of miR-3653-3p, competitively activated the ZEB2 gene, thus facilitating the migration and invasion of colon cancer cells. LINC01615 and miR-3653-3p may become markers for early clinical diagnosis of colon cancer and targets for drug action.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Availability of Data and Materials

The analyzed data sets generated during the study are available from the corresponding author on reasonable request.

Disclosure statement

No potential conflict of interest was reported by the author(s).