Decreased lncRNA, TINCR, promotes growth of colorectal carcinoma through upregulating microRNA-31

Abnormal expression in terminal differentiation-induced noncoding RNA (TINCR), a long non-coding RNA (lncRNA), has been reported in different human cancers, including colorectal carcinoma (CRC). Moreover, the molecular mechanisms that underlie the effects of TINCR on CRC remain unclear. Here, by a set of bioinformatics studies, we found that microRNA-31 (miR-31), the oncogenic miRNA that robustly upregulates in CRC, was a sponge miRNA for TINCR. TINCR and miR-31 levels were inversely correlated in both CRC tissues and CRC cell lines. Luciferase reporter assay revealed a specific binding site on TINCR for miR-31. Suppression of TINCR promoted CRC cell growth and migration in vitro, while overexpression of TINCR inhibited CRC cell growth and migration in vitro. TINCR depletion increased tumor xenograft growth in vivo, while TINCR overexpression inhibited it. Together, our study suggests that re-expressing TINCR may suppress invasive outgrowth of CRC through miR-31.


INTRODUCTION
Colorectal carcinoma (CRC) accounts for about 90% of all colorectal carcinoma cases [1], and is one of the leading reasons for cancer-related mortality [2]. Surgical removal of the primary tumor, endoscopic noninvasive treatment, radiotherapy, chemotherapy and radiation have now been co-applied during therapy, which has significantly improved the overall patient survival [3]. However, clinical outcomes of patients with CRC are still unsatisfactory and need further improvement [4,5]. Therefore, better description and comprehension of the regulators of the aggressive manner of CRC are urgently required for developing novel therapeutic strategies in clinic.
Long non-coding RNAs (lncRNAs) are a group of endogenous and non-protein coding RNAs that are more than 200 nucleotides in length, and were first identified from sequencing and microarray analyses of the whole genome and transcriptome [6]. LncRNAs are functionally heterogeneous and abundant RNAs acting in all cellular compartments that can form complexes with DNA, RNA, and proteins. Accumulating evidence shows that lncRNAs are dysregulated in nearly all types of human cancer, and significantly influence diverse pathophysiological processes, including innate immunity, metabolism and carcinogenesis [7][8][9]. An increasing number of studies have indicated that long non-coding RNAs (lncRNAs) play important roles in tumorigenesis and tumor development [10][11][12]. Numerous lncRNAs are dysregulated in CRC, and are implicated in the genesis and development of CRC [13]. Accordingly, therapies that target lncRNAs might be attractive strategies for treating patients with CRC.
TINCR was initially discovered as a lncRNA required for induction of key differentiation genes in epidermal tissue, including genes mutated in human skin diseases characterized by disrupted epidermal barrier formation AGING [14,15]. Aberration in TINCR expression has been identified in multiple human cancer types, and its aberrant expression has been shown to have significant functions in cancer progression. TINCR has been reported involved in the tumorigenesis of many different human cancers [16][17][18][19][20], including CRC [21,22]. However, the mechanisms that underlie the effects of TINCR on CRC remain unclear.
Here, by a set of bioinformatics studies, we found that microRNA-31 (miR-31), the oncogenic miRNA that robustly upregulates in CRC, was a sponge miRNA for TINCR. TINCR and miR-31 levels were inversely correlated in both CRC tissues and CRC cell lines. Luciferase reporter assay revealed a specific binding site on TINCR for miR-31. Suppression of TINCR promoted CRC cell growth and migration in vitro, while overexpression of TINCR inhibited CRC cell growth and migration in vitro. TINCR depletion increased tumor xenograft growth in vivo, while TINCR overexpression inhibited it. Together, our study suggests that reexpressing TINCR may suppress invasive outgrowth of CRC through miR-31.

TINCR and miR-31 levels inversely correlate in CRC tissues and cell lines
To explore the potential functions of TINCR in the development of CRC, its expression pattern as well as association with metastasis was investigated in 77 pairs of CRC tissues and peritumoral tissues (PTT) ( Table 2). Interestingly, the data of RT-qPCR revealed that TINCR was slightly but significantly reduced in CRC tissues, compared to PTT (Figure 2A, p=0.04). However, the increase in miR-31 levels in CRC tissues was dramatical and significant, compared to PTT ( Figure 2B, p<0.0001). In addition, all 77 cases were put together, which showed a significant inverse correlation (r=-0.3; p<0.0001, Figure  2C). Further measurements of TINCR ( Figure 2D) and miR-31 ( Figure 2E) expression were done in different human CRC cell lines, including HCT116, SW480, SW620, T84, HT15, LS174T, SNU-C1, Caco-2, LoVo and RKO, for selecting proper cell lines for mechanistic studies. Of note, the levels of TINCR and miR-31 in these 10 cell lines also exhibited an inverse correlation (r=-0.82, p<0.01, Figure 2F). Together, these data suggest a possible functional interaction between TINCR and miR-31.

TINCR is a molecular sponge to interact with miR-31 in CRC
Next, the binding (site) between TINCR and miR-31 was predicted by StarBase (http://starbase.sysu.edu.cn/) ( Figure 3A). We thus prepared luciferase reporter for wildtype (wt) TINCR and TINCR with a mutant at the miR-31 binding site (mut) ( Figure 3A). CaCO2 cells expressed relatively high miR-31 and low TINCR ( Figure 2D, 2E), and thus were selected as the cell line for depleting miR-31 or for overexpressing TINCR.
SW480 cells expressed relatively low miR-31 and high TINCR ( Figure 2D, 2E), and thus were selected as the cell line for overexpressing miR-31 or for depleting TINCR. First, transfection with as-miR-31 significantly reduced miR-31 levels in Caco-2 cells ( Figure 3B), while transfection with miR-31 significantly increased miR-31 levels in SW480 cells ( Figure 3C). LncRNA can serve as a molecular sponge to interact with miRNAs [29]. Thus, we performed a luciferase reporter assay to examine whether TINCR could interact with miR-31 in CRC cells. In SW480 cells, the transfection AGING of miR-31 markedly reduced the luciferase activity of TINCR-wt (p<0.05), whereas the luciferase activity of the TINCR-mut was unaffected upon miR-31 upregulation ( Figure 3D). Moreover, in Caco-2 cells, the transfection of as-miR-31 markedly increased the luciferase activity of TINCR-wt (p<0.05), whereas the luciferase activity of the TINCR-mut was unaffected upon miR-31 downregulation ( Figure 3E). Together, these data suggest that TINCR is a molecular sponge to interact with miR-31 in CRC.

Suppression of TINCR promotes CRC cell growth in vitro without affecting cell apoptosis
Then, we examined the effects of TINCR modification on CRC cell growth in vitro. First, the CCK-8 assay was performed to evaluate the influence of TINCR overexpression or depletion on CRC cell growth. The absorbance was significantly lower in TINCRtransfected Caco-2 cells ( Figure 5A), and significantly higher in shTINCR-transfected SW480 cells ( Figure  5B). However, neither overexpression of TINCR in AGING Caco-2 cells ( Figure 5C, 5D), nor depletion of TINCR in SW480 cells ( Figure 5E, 5F), appeared to significantly alter apoptosis. Thus, the effects of TINCR on CRC growth should be mediated through cell proliferation, rather than through cell apoptosis.

Suppression of TINCR promotes CRC cell metastasis in vitro
The migration of CRC cells in vitro was then examined in TINCR-modified CRC cells. The transwell migration assay showed that the overexpression of TINCR suppressed the migratory ability of Caco-2 cells, shown by representative images ( Figure 6A) and by quantification (p<0.05, Figure 6B). On the other hand, depletion of TINCR increased the migratory ability of SW480 cells, shown by representative images ( Figure 6C) and by quantification (p<0.05, Figure 6D). Thus, suppression of TINCR promotes CRC cell metastasis in vitro.

Overexpression of TINCR inhibits CRC growth in vivo
In vivo xenograft experiments were conducted to analyze the role of TINCR on tumor growth in vivo.
Caco-2 cells transduced with lentivirus carrying TINCR or scramble with luciferase, and SW480 cells AGING transduced with lentivirus carrying shTINCR or scramble with luciferase, were subcutaneously injected into the NOD/SCID mice. The formation of xenograft-derived tumor was measured every week. After 8 weeks, the quantification of tumor formation in living animals was performed using bioluminescence detection system. The increase in TINCR expression inhibited Caco-2 tumor cell growth in vivo, shown by quantification (p<0.05, Figure   7A), and by representative images ( Figure 7B). While the decrease in TINCR expression enhanced SW480 tumor cell growth in vivo, shown by quantification (p<0.05, Figure 7A), and by representative images ( Figure 7B). These data were consistent with the weekly measurement of tumor volume ( Figure 7C-7D). Collectively, the overexpression of TINCR inhibits CRC growth through suppression of miR-31. AGING

DISCUSSION
An increasing number of studies have demonstrated the important regulatory roles of lncRNAs in carcinogenesis and cancer progression [30][31][32]. A variety of lncRNAs are aberrantly expressed in CRC, and play dispensable roles in regulating a wide range of biological activities, such as cell proliferation, cycle, apoptosis, metastasis, and epithelial-mesenchymal transition [6]. Therefore, identification of the detailed roles of lncRNAs in CRC pathogenesis might facilitate the development of effective targets for the treatment of CRC patients.
TINCR expression is also reduced in prostate [16] and lung [17] cancers. For example, prostate cancer patients with low TINCR expression exhibit a poorer prognosis than those with high TINCR expression [16]. In contrast, TINCR is upregulated in hepatocellular carcinoma, and high TINCR expression is significantly correlated with tumor size, tumor differentiation, TNM stage, and vascular invasion [18]. Hepatocellular carcinoma patients with high TINCR expression have shorter disease-free survival and overall survival than those with low TINCR expression [18]. Increased expression of TINCR is also observed in breast [19] and gastric [20] cancers.

AGING
Previous studies have shown that insufficient TINCR expression promotes proliferation, metastasis through activating EpCAM cleavage in CRCs [21,22]. Here, we found that TINCR is a sponge for miR-31. Given that miR-31 was shown to target ARID1A to further transactivate EpCAM and subseqently enhance the oncogenicity and stemness of head and neck squamous cell carcinoma [33], it is highly possible that miR-31 is the mediator for the TINCR-regulated EpCAM, reported by Zhang et al. [21]. In a recent study, TINCR was however found to increase in CRC and seemed to work as a sponge for tumor suppressor miR-7 [34]. To figure out the possible reason for this discrepancy, we tried to surf TCGA database. However, our approach failed due to poor availability of the raw data from these studies, which prevented a deep analysis on the relationship between TINCR and interested miRNAs.
On the other hand, the patients' information provided in the two studies did not show significant difference in the selection of tumor stage and malignancy. Thus, we think that the only reason for this discrepancy of TINCR levels may be due to regional difference in the participated cases, as people from these two regions had major difference in appetite, which could affect the delicate molecular characteristics of the CRCs.
Here, we provided a set of strong evidence to demonstrate the interaction between TINCR and miR-31. First, we detected an obvious inverse correlation between TINCR and miR-31 in not only the clinical CRC specimens, but also in selected commonly used CRC lines. Second, bioinformatic studies demonstrated the binding site for miR-31 on TINCR, which was proved to be functional in a luciferase reporter assay. Third, overexpressing TINCR in CRCs decreased miR-31, while depleting TINCR increased miR-31.
Next, we found that suppression of TINCR promoted CRC cell growth and migration in vitro, without affecting cell apoptosis. Overexpression of TINCR inhibited CRC cell growth and migration in vitro, without affecting cell apoptosis. TINCR depletion increased tumor growth in vivo, while TINCR overexpression inhibited tumor growth in vivo. The function of TINCR in CRC appeared to be tumorsuppressive, which is different from how it functioned in gastric cancer, breast cancer and hepatocellular carcinoma. These likely inconsistent results in different cancers may result from the difference in expression of the interactive partners of TINCR in different cancers, e.g. certain miRNAs functioning as sponge miRNAs or certain genes and their mRNAs [7][8][9].
The effects of TINCR on CRC growth appeared not to be attributable to cell apoptosis. Since cell growth is a summary of cell proliferation and cell death, it is expected that increase in CRC cell growth by TINCR depletion could be primarily resulting from increases in cell proliferation. Of note, the depletion of TINCR on increases in cell metastasis could help to detach cell junctions and increase the intercellular spaces to promote cell proliferation [35]. Thus, the effect of TINCR on metastasis may be more important than its direct effect on cell proliferation in CRC.
The miR-31 is a well-known oncogenic miRNA in many cancers [23][24][25][26][27][28], including CRC [36]. It has multiple determined target genes and the net effect is tumor promoting [23][24][25][26][27][28]36]. Hence, in the current study, we did not look at its downstream genes, since it is not the key question here. In summary, our study showed that TINCR is an endogenous sponge of miR-31, while the decrease in TINCR expression increases the miR-31 expression, thereby enhancing tumor growth in CRC. Our research provides novel data regarding the mechanisms underlying CRC pathogenesis, and may be helpful in identifying promising novel targets for the treatment of CRC.

Protocol approval, patients and mice
The current study and animal work were approved by the Ethics Committee of Zhongshan Hospital of Fudan University and carried out in accordance with the Declaration of Helsinki. Written informed consent was provided by all the enrolled patients before their participation into the research. In total, 77 pairs of CRC tissues and adjacent normal tissues were collected from patients at Zhongshan Hospital of Fudan University. Immediately after surgical resection, all these tissue specimens were snap frozen in liquid nitrogen and then stored at −80°C until further use. CRC patients who had been treated with chemotherapy or radiotherapy prior to surgical resection were excluded from the research.

Bioinformatics analysis and luciferase reporter assay
The binding (site) between TINCR and miR-31 was predicted by miRcode (http://www.mircode.org/ mircode) and StarBase (http://starbase.sysu.edu.cn/). The fragments of TINCR containing the predicted wildtype (wt) and mutant (mut) miR-31-binding sites were cloned into pmirGLO reporter vectors (Promega Corporation, Madison, WI, USA) to generate the TINCR-wt and TINCR-mut plasmids, respectively. For the reporter assay, cells were seeded into 24-well plates one day before transfection. The generated luciferase reporter plasmids, along with the miR-31 or antisense for miR-31 or scramble, were transfected into cells using Lipofectamine 2000. The transfected cells were collected after 48 h of transfection and subjected to the dual luciferase reporter assay (Promega) for the measurement of the luciferase activity. Firefly luciferase activity was normalized to the level of Renilla luciferase activity.

Transfection and transduction
Small hairpin interfering RNAs (shRNA) against TINCR (shTINCR), TINCR and nontargeting control shRNA (scramble) used a pcDNA3.1-CMV-GFP plasmid as backbone (Clontech, Mountain View, CA, USA). Cells were plated into six-well plates at a density of 5 × 10 5 cells per well. Cell transfection was conducted with Lipofectamine 2000 (Invitrogen, Carlsbad, USA) according to the protocol recommended by the manufacturer. For in vivo xenograft experiments, plasmids that co-express luciferase and TINCR/shTINCR/ scramble were used to prepare lentiviral vectors. Seeded HEK293T cells (ATCC) were co-transfected with 5 µg of prepared plasmids and 5 µg each of packaging plasmids (REV, pMDL and VSV-G) by Lipofectamine-2000. The supernatant was removed 48 hours after transfection and filtered through the 0.45 µm syringe filter, after which the lentivirus in supernatant was further processed, isolated and titrated. For in vitro transduction, a multiplicity of infection (MOI) of 100 was used and the incubation time was 48 hours to allow completeness of viral infection.

Extraction of total RNA and reverse transcriptionquantitative polymerase chain reaction (RT-qPCR)
The extraction of total RNA was performed with a highpurity total RNA extraction kit (BioTeke Ltd., Beijing, China). Total RNA was reverse transcribed using a miScript Reverse Transcription kit (Qiagen GmbH, Hilden, Germany), according to the manufacturer's protocol. The cDNA was then used for the detection of miR-31 expression using a miScript SYBR Green PCR kit (Qiagen GmbH). the synthesis of cDNA was conducted using a PrimeScript first-strand cDNA synthesis kit (Takara Biotechnology Co., Ltd., Dalian, China); the cDNA was then subjected to qPCR using a SYBR Premix ExTaq kit (Takara Biotechnology Co.). The expression of miR-31 was normalized to glyceraldehyde phosphate dehydrogenase (GAPDH). Relative gene expression was calculated using the 2 -ΔΔCq method.

Cell counting kit-8 (CCK-8) assay
The transfected cells were collected after 24 h of incubation and suspended in complete culture medium. A total of 100μl of each suspension containing 2000 cells was seeded into 96-well plates. Cell proliferation was evaluated using the CCK-8 assay (Dojindo, Kumamoto, Japan) at four time points (0, 1, 2, and 3 days after incubation). For this assay, 10 μl of CCK-8 solution was added to the cells, followed by incubation at 37°C for an additional 2 h. The absorbance of the samples at a wavelength of 450 nm was measured using the VarioskanTM LUX microplate reader (Thermo Fisher Scientific).

Analysis of apoptosis by flow cytometry
The apoptosis rate was determined using an Annexin V fluorescein isothiocyanate (FITC) apoptosis detection kit (Biolegend, San Diego, CA, USA), referring to the protocols recommended by the manufacturer. After 48 h of culture, the transfected cells were collected and washed thrice with ice-cold phosphate buffer solution (PBS; Gibco; Thermo Fisher Scientific, Inc.). The transfected cells were then double-stained with 5 µl of Annexin V and 5 µL of PI diluted in 100 µl of binding buffer. Following incubation for 30 min in the dark, flow cytometry (FACScan; BD Biosciences, Bedford, MA, USA) was performed for the determination of the apoptotic condition of cells.

Transwell migration assay
At 48 h post-transfection, the cells were washed thrice with PBS and suspended in FBS-free DMEM. In total, 200 µL of cell suspension containing 5 × 10 4 transfected cells was plated into the upper compartments of transwell inserts (8 μM pore size, Costar, Cambridge, MA, USA). The bottom compartments were covered with 500 µL of DMEM containing 20% FBS; this medium acts as a chemoattractant. Following incubation for 24 h, the nonmigrating cells in the upper compartment were gently removed with a cotton swab, whereas the migrating cells were fixed in 4% paraformaldehyde and stained with 0.5% crystal violet. The migration of the cells was assessed by counting the average number of migrated cells in six randomly selected fields of each insert under an IX83 inverted microscope (Olympus Corporation, Tokyo, Japan).

In vivo xenograft experiments
Caco-2 cells transduced with lentivirus carrying TINCR or scramble with luciferase, and SW480 cells transduced with lentivirus carrying shTINCR or scramble with luciferase, were subcutaneously injected into the NOD/SCID mice (Shanghai SLAC Laboratory Animal Co. Ltd., Shanghai, China). The formation of xenograftderived tumor was measured every week. After 8 weeks, the quantification of tumor formation in living animals was performed using bioluminescence detection system (IVIS imaging system, Xenogen Corp., Alameda, CA, USA), 10 minutes after intraperitoneal injection of luciferin at 150 mg/kg body weight. The acquisition time was set to 1 minute and the binning value was 10.

Statistical analysis
The statistical analysis was performed with the GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). All results were expressed as the mean ± standard deviation (SD) from at least 5 independent experiments. All values represent the mean ± standard deviation (SD). The multivariate analysis was performed between TINCR and different clinicopathological characteristics in patients with CRC. The comparison between two groups was examined using the two-tailed Student's t-test, while one-way analysis of variance (ANOVA) followed by a Dunnett's post hoc test was used to determine the differences among multiple groups. The correlation between the expression of TINCR and miR-31 was evaluated by Spearman's correlation analysis. P values < 0.05 were deemed statistically significant.