Anti-tumor effect of miR-1291 in colorectal cancer

Background: MiR-1291 has an anti-tumor effect in carcinoma of kidney, esophagus, pancreas, and prostate. However, it’s role in colorectal cancer (CRC) has not been elucidated. Methods: In this study, we explored the effect of miR-1291 in CRC cells (HCT116, DLD-1, and HT29) in vitro, and performed a tumor growth inhibitory assay in a mouse therapeutic model using DLD-1 cells. Flow cytometric analysis and Western blotting were performed to examine a role of miR-1291 in cell cycle regulation. We performed luciferase reporter assay to verify the interaction between DCLK1 and miR-1291 in HCT116 cells. Cancer stemness was evaluated by identifying the expression of BMI1 and CD133, as well as performing sphere formation assay. Results: We found that miR-1291 signicantly suppressed the proliferation, invasion, cell mobility, and colony formation capability of CRC cell lines. MiR-1291 caused altered expression of the cell cycle-regulatory proteins, representatively CDK inhibitors p21 WAF1/CIP1 and p27 KIP1 or CDK4. Moreover, intravenous administration of miR-1291 loaded on the super carbonate apatite delivery system signicantly inhibited a tumor growth in the DLD-1 xenograft mouse model. A luciferase reporter assay showed that miR-1291 directly bound the 3’ UTR sequence of DCLK1 and suppressed its expression at both the mRNA and protein levels in HCT116 expressing the DCLK1 protein. In addition, miR-1291 suppressed cancer stem cell (CSC) markers BMI1 and CD133 as well as sphere formation ability in HCT116 cells. Conclusions: Taken together, these ndings indicate that miR-1291 has an anti-tumor effect by modulating multiple functions, including cell cycle, invasiveness, and cancer stemness. Our data suggest that miR-1291 could be a promising nucleic acid medicine against CRC.

Results: We found that miR-1291 signi cantly suppressed the proliferation, invasion, cell mobility, and colony formation capability of CRC cell lines. MiR-1291 caused altered expression of the cell cycleregulatory proteins, representatively CDK inhibitors p21 WAF1/CIP1 and p27 KIP1 or CDK4. Moreover, intravenous administration of miR-1291 loaded on the super carbonate apatite delivery system signi cantly inhibited a tumor growth in the DLD-1 xenograft mouse model. A luciferase reporter assay showed that miR-1291 directly bound the 3' UTR sequence of DCLK1 and suppressed its expression at both the mRNA and protein levels in HCT116 expressing the DCLK1 protein. In addition, miR-1291 suppressed cancer stem cell (CSC) markers BMI1 and CD133 as well as sphere formation ability in HCT116 cells.
Conclusions: Taken together, these ndings indicate that miR-1291 has an anti-tumor effect by modulating multiple functions, including cell cycle, invasiveness, and cancer stemness. Our data suggest that miR-1291 could be a promising nucleic acid medicine against CRC.

Background
Colorectal cancer (CRC) is the third most widespread cancer and the second most deadly cancer in the world. Approximately 1.8 million new cases of CRC and 881,000 related deaths were estimated in 2018 [1,2]. In the past few decades, improved treatment options have become available, including surgery, radiotherapy, chemotherapy, and molecular-targeted therapy for advanced CRC [3][4][5][6]. However, the 5-year survival rate of CRC is <65% due to cancer relapse [2].
MicroRNAs (miRNAs) are short (18-25 nucleotides) internally originated non-coding RNAs that mainly bind to the 3'-untranslated region (3' UTR) of target mRNAs, contributing to mRNA cleavage or translational suppression [7,8]. MiRNAs play an important role in many biological progresses, including tumor growth, apoptosis, invasion, and survival [9], which are closely related to oncogenesis and tumor progression. Recent studies have shown that the pathological mechanisms underlying CRC depend on a variety of signaling pathways, including Wnt/β-catenin, EGFR, TGF-β, TP53, and epithelial-tomesenchymal transition, and miRNAs play a pivotal role in regulating these pathways [10][11][12]. For example, miR-4689 has an anti-tumor effect on mutant KRAS CRC by inhibiting the EGFR pathway [13]. In addition, miR-34a can inhibit cell proliferation and increase the expression of p21 WAF1/CIP1 in HCT116 and RKO colorectal cancer cells [14].
Through in silico analysis and in vitro selection using doublecortin-like kinase 1 (DCLK1) as a target molecule ( Supplementary Fig. S1a, b), we focused on miR-1291. DCLK1 is over-expressed in subsets of cancers and has an oncogenic function [15,16]. In the Caki-2 renal cancer cell line, knocking down DCLK1 contributes to inhibition of cell proliferation, invasion, and wound-healing [17]. In 2013, DCLK1 was reported as a marker distinguishing tumor stem cells from intestinal normal stem cells [18]. Accumulating evidence supports the involvement of DCLK1 in the stemness of CRC [19,20].
Cancer stem cells (CSCs) have distinct capability for self-renewal, unlimited proliferation, reduced capacity to undergo apoptosis, and multi-potential differentiation [21,22]. Conventional anticancer drugs and radiotherapy may reduce tumor bulk, but CSCs can still survive, as they confer resistance to these therapies and cause distant metastases and recurrence in various cancers [23][24][25]. Thus, the development of CSC-targeted therapy may be an effective approach for overcoming the shortage of current therapies and completely cure CRC patients [26][27][28]. Several CSC markers for CRC have been found, including CD133, BMI1, and LGR5 [24,[29][30].
In recent years, miR-1291 has been demonstrated to have anti-tumor effects in carcinoma of the kidney, esophagus, pancreas, and prostate [31][32][33][34]. However, to the best of our knowledge, miR-1291 in CRC has not been reported. Therefore, we investigated the anti-tumor effect of miR-1291 in CRC cells in an effort to improve our understanding of the potential mechanisms of miR-1291 in CRC.

Methods
Cell lines and cell culture Human CRC cell lines DLD-1, HT29, HCT116 were purchased from the American Type Culture Collection (Rockville, MD, USA). These cell lines were authenticated by morphological inspection, short tandem repeat pro ling, and mycoplasma testing. DLD-1, HT29 cells were cultured in RPMI 1640 medium, and HCT116 cells were cultured in Dulbecco's modi ed Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were cultured in the humidi ed incubator at 37°C and 5% CO 2 . All cells were passaged every 2 or 3 days.

Clinical tissue samples
Paired clinical tissue specimens (normal mucosa and colorectal cancer tissue) were collected from 20 patients who had surgery at Osaka University Hospital between 2016 and 2017. All tissue specimens were stored at -80 °C until RNA extraction. All patients gave written informed consent, in accordance with the guidelines approved by the Institutional Research Board of the institute. This study was conducted under the supervision of the Ethics Board of Osaka University Hospital.

RNA isolation
Total RNA was collected from cultured cells using TRIzol Reagent (Thermo Fisher Scienti c) followed by phenol-chloroform extraction and ethanol precipitation. And miRNA was collected from tissue specimens and cultured cells using the miRNeasy kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Total RNA concentration and purity were measured using a NanoDrop one spectrophotometer (Thermo Fisher Scienti c), at 260 and 280 nm (A 260/280 ) wavelengths.
Quantitative real-time PCR analysis of messenger RNA A High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scienti c) was used to synthesize the complementary DNA from 2.5 μg of total RNA according to the manufacturer's instructions. The quantitative real-time PCR (qRT-PCR) for DCLK1, BMI1 and CD133 RNA were performed using was ampli ed using oligonucleotide primers and the LightCycler 480 Real-Time PCR system (Roche, Basel, Switzerland). The ampli cation products were detected using the THUNDERBIRD SYBR qPCR Mix (TOYOBO, Osaka, Japan), and the level of target gene expression was calculated. The qRT-PCR conditions were 95°C for 30 sec; 40 cycles of 95°C for 10 sec, and 60°C for 10 sec, 72°C for 30 sec. The expression of the target gene was normalized to endogenous GAPDH expression. Relative expression was quanti ed by the 2 -ΔΔCt method. The PCR primers are listed in Supplementary Table S1. Proliferation assay Cells were seeded in 96-well plates at a density of 4000-8000 per well and were transfected with miR-NC or miR-1291 at a nal concentration of 30 nM the second day after seeding. Twenty-four, 48, and 72 hours after transfection,10 µl of Cell Counting Kit-8 (DOJINDO Molecular technologies, Inc., Kumamoto, Japan) was added to each well, and the 96-well plates were shaded for 2 hours. After that, the absorbance was detected by Multiskan Go (Thermo Fisher Scienti c), and the subtraction difference of absorbance between wavelength of 630 nm and 450 nm was used to determine cell number.

Matrigel invasion assay
Cells were seeded in BD BioCoat Matrigel Invasion Chambers (BD Biosciences, San Jose, CA, USA) at a density of 50,000-100,000 cells per chamber. The cells were transfected with the miRNAs at a nal concentration of 50 nM. After 48-72 h of transfection, invaded cells were xed with 10% formalin and then stained with hematoxylin for counting.
Wound healing assay Cells were seeded in ibidi culture 2-well inserts (ibidi, Gräfel ng, Germany) in 24-well plates at a con uent density. The inserts were removed after 24 hours to create wounds. The miRNAs were transfected at a nal concentration of 30 nM. At 0-48 hours after transfection, the areas of the wounds were measured using ImageJ software.

Colony formation assay
Cells were seeded in a 6-well plate at a density of 1 × 10 5 cells per well, incubated overnight, and then transfected with miR-NC or miR-1291 at a nal concentration of 30 nM for 8 hours and then reseeded in 6-well plates at a density of 500 cells per well. After 10 days, the cells were xed with methanol and stained by crystal violet for counting.

Statistical analysis
Data are presented as the mean ± SEM. Statistical analyses were performed using GraphPad Prism 5 (San Diego, CA, USA) and Microsoft Excel. The statistical differences between the miR-NC and miR-1291 groups were analyzed by student's t-test (two-tailed). The expression levels of miRNAs in normal and cancer colorectal tissues were analyzed using the Wilcoxon signed-rank test. In vivo tumor growth was analyzed with one-way ANOVA followed by Bonferroni's multiple comparisons test. P<0.05 was considered signi cant.

Expression of miR-1291 in CRC cells and clinical tissue specimens
We evaluated the expression of miR-1291 in CRC cell lines and non-tumor human HEK293 cells by qRT-PCR ( Supplementary Fig. S2a). We then compared the expression level of miR-1291 in normal mucosa and CRC cancer tissues from 20 paired clinical tissue specimens. Although the tumor tissues tended to express generally low miR-1291 expression, statistical difference was not noted between normal colonic mucosa and CRC tissues (Supplementary Fig. S2b).

MiR-1291 inhibited cell proliferation
To con rm the anti-tumor effect of miR-1291 in CRC cell growth, we detected the absorbance of cells transfected with miR-NC or miR-1291 as determined by Cell Counting Kit-8. The miR-1291 transfection group presented a signi cant low absorbance compared to the miR-NC transfection group in DLD-1, HT29, and HCT116 cells after 48 and 72 hours of transfection (Fig. 1b). MiR-1291 signi cantly suppressed the proliferation ability of the three cell lines.

MiR-1291 inhibited cell invasion
MiR-NC or miR-1291 was transfected into CRC cells to evaluate the effect on invasion ability. The cells invading through Matrigel were stained with hematoxylin 48 hours for DLD-1 or 72 hours for HT29 and HCT116 after transfection and then counted. The invasion ability of miR-1291-transfected cells was signi cantly inhibited compared to miR-NC transfected cells in the three cell lines tested (Fig. 1c).

MiR-1291 inhibited cell migration ability
We evaluated the effect of miR-1291 on the wound-healing ability of CRC cells. The wound area of the cells was measured at the same location every 24 hours after miR transfection. The migration ability of DLD-1, HT29, and HCT116 cells was signi cantly suppressed in the miR-1291 group compared to miR-NC group either at 24 hours or 48 hours or both (Fig. 2a).

MiR-1291 inhibited colony formation ability
To con rm the function of miR-1291 in cell colony formation ability, we observed and evaluated the colony formation of the CRC cells which were transfected with miR-NC or miR-1291. The colonies were stained with crystal violet 10 days after transfection, and we counted the number of the colonies. MiR-1291 signi cantly inhibited the colony formation potential compared to miR-NC in DLD-1, HT29, and HCT116 cells (Fig. 2b).

Altered expression of cell cycle components by treatment of miR-1291
Cells were examined for the change in cell cycle regulatory protein expression after transfection of miR-1291 in the standard medium supplemented with FBS. In DLD-1 cells, the expression of CDK inhibitors p21 WAF1/CIP1 and p27 KIP1 were up-regulated and CDK4 and CDC25A level decreased at 48 hours after transfection with treatment of miR-1291 (Fig. 3a). In HT29 cells, the expression of p21 WAF1/CIP1 and p27 KIP1 protein increased at 48 hours compared to miR-NC group (Fig. 3a). In HCT116 cells, transfection of miR-1291 up-regulated the expression of p21 WAF1/CIP1 and p27 KIP1 , down-regulated the expression of CDK4, compared to miR-NC cells (Fig. 3a). We then examined the cell cycle distribution after serum starvation ( Supplementary Fig. S3a, b; Time points: 0, 12, 24, and 24 hours). Twelve hours refed with FBS, cells in G1 phase were signi cantly increased in miR-1291 treated cultures of DLD-1 cells (60.81% vs 67.89%, Fig. 3b). However, cell population in each cell cycle phase was not apparently changed in HT29 cells throughout the time points examined (Fig. 3b, Supplementary Fig. S3b). MiR-1291 increased the percentage of cells in G2/M phase in HCT116 cells compared to miR-NC after 24 hours of addition of serum to the starved cells (26.35% vs 33.15%, Fig. 3b).

Anti-tumor effects of miR-1291 in vivo
We used a DLD-1 tumor xenograft mouse model to verify the anti-tumor effect of miR-1291 in vivo. Compared to miR-NC group and no treatment group, systemic administration of miR-1291 on super carbonate apatite signi cantly inhibited the growth of tumor (Fig. 4a). No obvious body weight loss of mice was observed among the three groups (Fig. 4b).
MiR-1291 directly targeted DCLK1 We performed luciferase reporter assays to con rm whether miR-1291 directly bind to DCLK1. In HCT116 cells, co-transfection with miRNA-1291 signi cantly inhibited the luciferase activity of wild type of DCLK1-3′ UTR reporter vector, compared to miR-NC (**P <0.01). On the other hand, no signi cant difference in luciferase activity was noted between miR-1291 group and miR-NC group in use of 2nucleotide mutated type (Mut) or 3-nucleotide deleted type (Del) DCLK1-3′ UTR reporter vector (Fig. 5a, b). Transfection with miR-1291 signi cantly suppressed the expression of DCLK1 at both the mRNA (Fig. 5c) and protein levels (Fig. 5d, Supplementary Fig. S4) compared to miR-NC. In miR-1291 treated group, the protein expression of DCLK1 was reduced to 30.6% of that treated with miR-NC (Fig. 5d). These ndings indicate that DCLK1 is a direct target of miR-1291 in HCT116.

MiR-1291 suppressed stem-like properties in HCT116 cells
Our previous study showed that HCT116, but not DLD-1 and HT29 expressed DCLK1 [16], therefore, we assessed stemness of HCT116 cells after treatment of miR-1291. We demonstrated that other CSC markers in addition to DCLK1, including BMI1 and CD133, were signi cantly down-regulated by miR-1291 overexpression at mRNA level (Fig. 6a, b). MiR-1291 treatment also down-regulated the protein level of BMI1 by Western blot analysis (Fig. 6c). Moreover, the ratio of CD133 positive cells decreased with treatment of miR-1291 by ow cytometric analysis (Fig. 6d). Furthermore, miR-1291 treatment signi cantly decreased the ability of sphere formation in HCT116 cells (Fig. 6e). These ndings suggest that miR-1291 may be involved in the regulation of stem cell properties through direct inhibition of DCLK1 in HCT116.

Discussion
MicroRNA is emerging as a next generation cancer treatment [41][42][43][44]. Studies showed that miR-1291 had anti-tumor effects in multiple cancers including renal cancer, esophagus cancer, pancreatic cancer, and prostate cancer [31][32][33][34]. However, to the best of our knowledge, there is no report of miR-1291 in CRC which is one of the widespread cancers in the world. In this study, we clearly demonstrated that miR-1291 exhibited anti-tumor effects in HCT116, DLD-1, and HT29 CRC cells in terms of cell proliferation, invasion, cell mobility, colony-forming abilities, and cell cycle regulation. Moreover, intravenous administration of miR-1291 loaded on the super apatite delivery system signi cantly inhibited in vivo tumor growth compared with miR-NC treated group.
In this study we started to select miRNAs using doublecortin-like kinase 1 (DCLK1) as a target [18,45] as well as Notch and Wnt signaling ( Supplementary Fig. S1). DCLK1 belongs to the protein kinase superfamily and the doublecortin family, and is over-expressed in several human malignancies, including colorectal, pancreas, kidney, and prostate cancer [46][47][48][49]. Excision of DCLK1-positive CSCs results in regression of the intestinal tumor without apparent impairment of normal tissue, which indicates that DCLK1 may be a novel target for CSC-targeted therapy [18]. Screening with ODC degron-transduced cells revealed that miR-1291 inhibited both CSC and non-CSC. Although degron (+) cells were more resistant to miR-1291 treatment compared with degron (-) cells, its growth inhibitory effect was even stronger than that given by a putative Onco-miR, miR-34a. Targeting both CSC and non-CSC by miR-1291 could be attributed to the feature of miRNA that can bind to multiple molecules.
We recently reported that sh DCLK1 clones, in which DCLK1 expression was silenced by short hairpin DCLK1 RNA, exhibited decreased cell growth, invasion, migration abilities and EMT in HCT116 [16]. Other investigators also showed that DCLK1 level was tightly associated with spheroid formation in HCT116, which is a hallmark for CSC [50,51]. In this study, we veri ed that miR-1291 directly bound to the 3' UTR of the DCLK1 mRNA sequence, leading to decreased expression of DCLK1 at both the mRNA and protein levels. In addition to DCLK1, miR-1291 lowered BMI1 and CD133 expression, which are also representative CSC markers in CRC. Moreover, we con rmed that miR-1291 inhibited sphere formation ability of HCT116 cells. These results support the notion that miR-1291 suppresses the cancer stemness through direct targeting DCLK1, suggesting that miR-1291 may be a novel CSC-targeted therapeutic strategy for CRC.
On the other hand, DLD-1 and HT29 cells which did not express DCLK1 [16] also exerted the potent tumor inhibitory effects, suggesting that certain other mechanism should be operating in these CRC cells and it may suppress non-stem cells. With this regard, studies reported that miR-1291 regulates SLC2A1/GLUT1 in A498 and 786-O renal cancer cells [31], MUC1 in human esophagus cancer EC9706 and EC-1 cells [32], the forkhead box protein A2-anterior gradient 2 (FOXA2-AGR2) pathway in PANC-1 pancreatic cancer cells [33]. Furthermore, miR-1291 has been shown to inhibit cell growth and tumorigenesis in prostate cancer by binding to MED1 [34]. These molecules as well as other targets should be further explored.
One of the major effects observed by miR-1291 treatment was drastic change in the cell cycle components. Western blot analysis showed increase in p21 WAF1/CIP1 and p27 KIP1 in the three CRC cell lines at 48 hours. These CDK inhibitors bind to and block the G1-S accelerators Cyclin D1-CDK4/6 complex and Cyclin E1-CDK2 complex, leading to restraint from the G1 to S phase [52][53][54]. We also found that CDK4 decreased in DLD-1 and HCT116. Under the condition supplemented with FBS we could not observe obvious change in the cell cycle distribution between miR-NC and miR-1291 treatment. Time course study after serum starvation revealed the delay in G1-S transition at 12 hours in DLD-1 and the delay in G2-M transition at 24 hours in HCT116. Collectively these results suggest that miR-1291 could cause dysregulation of cell cycle control even though its effectual action point and timing may differ by cell types. In terms of the association of cell cycle regulators with DCLK1, Chandrakesan et al. reported that DCLK1-positive cells had higher expression of p21 WAF1/CIP1 and p27 KIP1 and maintained quiescence in normal small intestine [55]. However, the current study demonstrated that miR-1291 inhibited DCLK1 and up-regulated p21 WAF1/CIP1 and p27 KIP1 which may cause G1 phase arrest. When we examined expression of these CDK inhibitors using the sh DCLK1 clones, p27 KIP1 level increased ( Supplementary   Fig. S5a, b). These ndings may partially help to explain an opposite role of DCLK1 in the cell cycle between normal and cancer cells.

Conclusions
In conclusion, miR-1291 presented a strong anti-tumor effect on CRC cells and played an important role in both delaying the cell cycle and suppressing cancer stemness. To the best of our knowledge, this is the rst study that demonstrates the function of miR-1291 in CRC. Considering the anti-tumor effect of miR-1291 in the broad range of cancer type [31][32][33][34], this microRNA may be one of the candidates for the next generation nucleic acid medicine using the practical DDS systems [11,12,35,43,44].

Declarations
Ethics approval and consent to participate: This study was performed in accordance with the Declaration of Helsinki and the study was approved by the Ethics Board of Osaka University (approval No. 13377-5; Osaka, Japan). Consent for publication: All patients gave written informed consent, in accordance with the guidelines approved by the Institutional Research Board of the institute.
Availability of data and materials: None.