Circ-0012417/ miR-29c-5p Axis Stimulates Thyroid Cancer Growth

Background: There is an increasing tendency of the incidence rate of thyroid cancer. Despite the advance of widespread early diagnose and good prognosis of papillary thyroid cancer (PTC), some patients still suffered from the invasiveness and metastasis of thyroid cancer. More and more evidences demonstrated that circular RNAs (circRNAs) play important roles in tumorigenesis, tumor progression and metastasis. This study aims to identify oncogenic cirRNAs and explore its underlying molecular mechanisms in thyroid cancers. Methods: With circRNA chips in GEO database, we found that circ-0012417 was over-expressed in thyroid cancer. After silencing circ-0012417 or ectopic over-expressed circ-0012417, CCK-8 was used for detecting cell proliferation and transwell is used for examining metastasis and invasion abilities. Western Blot was used to detect changes of downstream proteins. Dual luciferase reporter assay was used to verify whether miR-29c-5p is the target of circ-0012417. Tumor xenografts in nude mice were used for further in vivo verications. Results: Circ-0012417 is overexpressed in thyroid cancer, and silencing circ-0012417 could inhibit cancer cell proliferation, metastasis and invasion. MiR-29c-5p may be the target of circ-0012417 and inhibiting miR-29c-5p could reverse this phenomenon. And further analysis of TCGA and GTEx databases showed that EGFL7 may be the target of miR-126 and a promising thyroid cancer prognosis indicator. Conclusion: Circ-0012417 promotes thyroid cancer progression through regulating miR-29c-5p /EGFL7 pathway. Our nding provided novel therapeutic targets for thyroid cancer treatment.


Background
Thyroid cancer is one of most common endocrine tumors and its incidence rate is increasing. Thyroid cancer is classi ed in 4 types: follicular thyroid cancer (FTC), papillary thyroid cancer (PTC), anaplastic thyroid cancer (ATC) and medullary thyroid cancer (MTC). While FTC and PTC accounts for around 90% thyroid cancer in clinical. Early diagnose, radical surgery and TSH inhibition therapy and I 131 is recommended for patients with tumor invasion and metastasis. Because of early invasion and metastasis, relapse and I 131 resistance, a subset of ATC patients suffered with only 6 months overall survival (OS). Besides, narrow therapy strategies limited bene ts for patients with signi cant invasion and metastasis. Circular RNAs (circRNAs), as the name suggests, are a novel group of covalently closed loop RNAs without 5' hat or polyadenylation at their 3' ends [1]. CircRNAs are mainly constituted by back-spliced exons [2]. MiRNAs are a group of short non-coding RNAs (16-29 nt) with post-transcriptional regulation functions. CircRNAs can function as "miRNA-sponges" to protect the target mRNAs from miRNA-induced cleavage. In fact, circRNA-miRNA-mRNA network regulation had been found in several type of cancer [3]. For example, circ-ITCH could sponge miR-17/miR-224 to inhibit bladder cancer progression [4]. Concretely, these circRNA-miRNA networks inhibit the expression of target genes and regulate tumorigenesis, tumor progression, invasion and metastasis. What's more, some circ-RNAs could also be reliable prognostic indicators. In advanced colon cancer, a four-circRNA-based classi er could be a good recurrence prognostic strati cation [5].
To uncover the roles of cirRNAs in thyroid cancer, we analyzed circRNA chips in GEO database and found that circ-0012417 was overexpressed in thyroid cancer. Thus, this study focused on demonstrating the functions and its molecular mechanisms of circ-0012417 in thyroid cancers.

Patient tissue samples and cell lines
This study was approved by the ethics committee of the First a liated hospital of Jinan University. All patients provided a written informed consent. Fresh-frozen cancer tissues and adjacent normal tissues from 30 patients with thyroid cancer were used for this study. Fresh-frozen tissues from 50 patients with benign thyroid nodules were used as control.
Thyroid cancer cell lines (K1, BHT-101, KHM-5M, B-CPAP) were all purchased from the Culture Collection of Chinese Academy of Sciences (Shanghai, China). All cells were preserved as per the manufacturer's instructions. Brie y, the cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum, 2 mM penicillin and streptomycin (Gibco BRL, NY, USA). All cell lines were cultured in humidi ed air supplemented with 5% CO2 at 37 °C.

CircRNAs expression pro le analysis
The gene expression pro les (which one) of TC were reviewed from the Gene Expression Omnibus database (GEO, http://www.ncbi. nlm.nih.gov/geo). The online analysis tool GEO2R (http://www.ncbi.nlm.nih.gov/geo/geo2r/) was utilized to assess the differentially expressed genes (DEGs) in this study. The adjusted P values were then calculated using the Benjamini and Hochberg false discovery rate method to correct for the occurrence of false positive results.

RNase R resistance analysis of circRNAs
Hsa-circ-0012417 from BHT-101 and B-CPAP was treated with RNase ((4 U/mg, Epicenter) and incubated for 30 minutes at 37 °C. The treated RNAs were then reverse transcribed using speci c primers. Next, quantitative real-time PCR (qRT-PCR) was used to determine the expression of has-circ-0012417 in the TC cell lines.
RNA extraction and quantitative real-time PCR (RT-PCR) assay Total RNA was extracted from the thyroid cancer tissues/cell lines and adjacent normal tissues/cell lines using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). After assuring RNA quanti cation and quality using NanoDro2000c (Thermo Scienti c, Waltham, USA), 2 µg of total RNA were reverse transcribed to cDNA using the BestarTM qPCR RT kit (#2220, DBI Bioscience, China). Quantitative polymerase chain reaction (RT-PCR) was then done using the BestarTM qPCR MasterMix (#2043, DBI Bioscience,China).

Lentivirus constructs and cell transfection
For in vivo studies, circRNA0012417-knockdown cells (si-circRNA0012417) and negative control cells (NC) were utilized. Lentivirus particles were bought from Hanbio Biotechnology (Shanghai, China). 48 hours after transfection, stable-transfection cells were established by puromycin selection (2 µg/ml) applied for 10 days. LV003-has-circ-0012417 overexpression and negative cells (LV003 vector control) were established in the same way.
The miR-29c, and the miR-miR-29c inhibitor were acquired from GenePharma (Shanghai, China). B-CPAP and BHT101 cells were seeded in 6-well plates and then transfected the next day using Lipofectamine3000 (Invitrogen, CA, USA) as per the manufacturer's instructions. Cells were harvested at 48 hours post transfection followed by con rmation of transfection e ciency using qRT-PCR.

Cell proliferation assay
The cell proliferation assay was performed using Cell Counting Kit-8 (CCK-8, Dojindo, Osaka, Japan). B-CPAP and BHT101 cells in different groups were seeded into 96-well plates with serum-free DMEM medium. Each group was duplicated with three independent wells. Cell proliferation was observed at 24, 48 and 72 hours. Prior to observation, each well was treated with 10 µl CCK-8 reagent and then cells were incubated at 37℃ for another 3 hours before measuring absorbance at 450 nm utilizing a spectrophotometer. All experiments were done in triplicate.
Transwell migration assay and invasion assay B-CPAP and BHT101 cells (5 × 104) in different groups were resuspended in 200 µl serum-free medium and seeded into the upper chamber of 24-well plates (Corning, New York, NY, USA) with (invasion) or without (migration) Matrigel (BD Biosciences, New York, NY, USA), while 600 µl medium containing 20% FBS were supplied at the lower chamber as chemoattractant. After incubation at 5% CO2 /37 °C for 20 hours, cells were xed with 4% paraformaldehyde for 30 min and then stained with 0.1% crystal violet for 30 min. The number of cells that migrated or invaded was determined in ve randomly selected elds using an inverted microscope.
Cell cycle assay Cells (2 × 10 6 ) were collected and xed with 70% ethanol at 4 °C overnight, and 100 µl RNase A (Keygen Biotech, Nanjing, China) were then added, followed by incubation at 37 °C for 30 min. Cells were then stained with 400 µl propidium iodide (Keygen Biotech) for 30 min. A FACS Calibur ow cytometer (Becton Dickinson) was applied to evaluate cells at 488 nm, and ModFit LT software (Verity Software House) for analysis. All assays were repeated at least three times.
Luciferase activity was measured at 48 hours after transfection by a Dual Luciferase Reporter Assay System (Promega, Madison, WI, USA) as per the manufacturer's instructions.

Western blotting
Total protein was extracted from cells using RIPA lysis buffer (Keygen Biotech). The lysate protein was separated by 10% SDS-PAGE and electrophoretically transferred onto polyvinylidene di uoride (PVDF) membranes (Millipore, Bedford, MA, USA). After blocking with 5% non-fat milk in TBST for 2 h, proteins were then incubated with primary and secondary antibodies. Signal detection was visualized by using enhanced chemiluminescence. Finally, relative concentration was calculated using Quantity One software

Tumor formation in nude mice
The validated thyroid cancer cell line hsa-circ-0012417 (B-CPAP Circ) and its negative control strain (B-CPAP NC) were expanded and cultured. 1 × 107 cells was injected subcutaneously into 10 BALB/c-nu/nu female nude mice weighing between 18-22 g (4-5 weeks old). 14 days after transfection, the length and width of the tumor were measured once a week for 5 weeks. After 5 weeks, the mice were sacri ced by cervical dislocation. The tumor length, width and weight were all measured and photographed.

Statistical Analysis
Statistical analysis was done using SPSS 18.0 software (SPSS, Chicago). T test was used to analyze the relationship between the indicators of each group. P value less than 0.05 was deemed as statistically signi cant.

Hsa-circ-0012417 is overexpressed in thyroid cancer
According to circRNA chips of thyroid cancer in GEO database, we focused on 4 dysregulated RNAs: hsacirc-0012417, hsa-circ-0000690, hsa-circ-0006357, hsa-circ-0007694. We then measured in 14 pairs of human thyroid cancer tissues and adjacent normal tissues, and we found that only has-circ-0012417 was signi cantly elevated in cancer tissues compared with respective normal tissues (p < 0.05) (Fig. 1A). We then tried amplify circ-0012417 using human complementary DNA (cDNA) and genomic DNA (gDNA) as templates, and we got the circ-0012417 fragment from cDNA templates, but we failed to obtain circ-0012417 from gDNA (Fig. 1B). This result indicating that the intact circ-0012417 fragment contains different splicing sites and it could be generated by back-splicing. Sanger sequencing was further performed to verify the sequence information and the conjugation site information of hsa-circ-0012417 was showed (Fig. 1C). These results demonstrated that circ-0012417 was identi ed as a novel circRNA, which is overexpressed in thyroid cancers.

Silence of circ-0012417 inhibits proliferation in thyroid cancer cells
Since circ-0012417 was higher-expression in thyroid cancer tissues, we next test thyroid cancer functions in thyroid cancer cells. Firstly, we determined circ-0012417 expression in four different thyroid carcinoma cell lines: KHM-5M, K1, BHT101 and B-CPAP and found that hsa-circ-0012417 was relatively high expressed in KHM-5M, BHT101 and B-CPAP than in K1 ( Fig. 2A). Then we constructed two si-circ0012417-1 and si-circ0012417-2 which could successfully inhibit circ-0012417 levels (Fig. 2B).
Finally, we detected that the proliferation of B-CPAP and BHT101 was inhibited when transfected with si-circ0012417 ( Fig. 2C) (p < 0.01 and p < 0.05, respectively).

Silence of circ-0012417 suppresses the metastasis of thyroid cancer cells
We then explore roles of has-circ-0012417 in biological behaviors of thyroid cancer. When transfected with si-circ0012417, the migration ability of B-CPAP and BHT101 cells were signi cantly decreased to 41.02% and 37.50% (Fig. 3A). Cell invasion decreased to 33.33% in the B-CPAP (P < 0.01), and to 56.91% in the BHT101 cell line respectively (P < 0.01) (Fig. 3B). Transfection of si-circ-0012417 also resulted in a signi cant decrease in the wound healing ability of B-CPAP and BHT101 cells (Fig. 3C). In conclusion, si-circ0012417 inhibit proliferation, migration, invasion and wound healing capacity of thyroid cancer cells.
Exogenous circ-0012417 stimulate he proliferation and metastasis of thyroid cancer cells We then explore gain of functions experiments of circ-0012417 in B-CPAP and BHT101. Over-expressed circ-0012417 induced increased cell proliferation in B-CPAP and BHT101 (Fig. 4A&B). As expected, with exogenous circ-0012417, the migration and invasion of B-CPAP and BHT101 cells increased signi cantly (P < 0.01 and P < 0.01 in B-CPAP, P < 0.001 and P < 0.001 in BHT101, respectively) (Fig. 4C&D). Furthermore, we wondered whether circ-0012417 regulated Wnt/β-catenin signaling. Both western Blot and Real-time PCR results showed that β-catenin, c-myc and cyclin D1 were up-regulated by circ-0012417 (Fig. 5). Previous researches had proved that β-catenin activation regulate cell cycle and cell adhesion [6,7], c-myc and cyclin-D1 could also promote cell cycle progression [8,9]. With these results, circ0012417 may regulate cell cycle to stimulate cell proliferation by upregulating β-catenin, c-myc and cyclin D1 proteins.
Silencing hsa-circ-0012417 inhibits tumor growth in vivo Following these in vitro studies, we further build xenograft models with B-CPAP cell line. Consistent with previous results, compared with NC group, silencing hsa-circ-0012417 signi cantly inhibited tumor growth by measurement of tumor size once a week (Fig. 7A). At the end of the animal experiment, tumor weight in si-circ-0012417 group is obviously decreased than NC treatment (Fig. 7B&C). Following transfection with si-circ0012417, we observe a signi cant decrease in tumor volume and weight in vivo, supporting the hypothesis that hsa-circ-0012417 is a tumor growth factor.

Discussion
Thyroid cancer is originated from thyroid follicular cells or parafollicular cells and its morbidity is increasing worldwide in recent years [11]. The 5-year survival of patients with papillary thyroid microcarcinoma was around 99% [11]. But there're still a part of patients who suffered thyroid cancer metastasis and recurrence. CircRNAs are a group of endogenous noncoding RNAs (ncRNAs), and characterized by a closed covalently loop with back-spliced exons [12]. CircRNAs are dysregulated in many solid tumors like hepatocellular carcinoma [13,14], gastric carcinoma [15,16] and colorectal cancer [17,18]. CircRNAs usually sponge miRNA to modulate enrgy metabolism, angiogenesis, invasion and metastasis in carcinoma. Recent researches had found that in thyroid cancer, circRNA_10217, circRNA_NEK6 and circ_0067934 could promote tumor growth [19][20][21], while circ-ITCH could suppress tumor progression [22].In this manuscript, we rst screened thyroid cancer related circRNA chips results from GEO database. Then hsa-circ-0012417 was upregulated in thyroid tumor via pPCR in 14 pairs of thyroid tumor tissues and para-carcinoma tissue (Fig. 1A). DNA gel and sanger sequencing supported that hsa-circ-0012417 is a circular RNA (Fig. 1B-D). Trough gain of functions and loss of functions experiments, circ-001241 could stimulate cell division and promote cell proliferation, metastasis and invasion ( Fig. 2-5). Researches had demonstrated that activation of WNT/β-catenin pathway take parts in the occurrence and progression of thyroid carcinoma [23], and aberrant WNT/β-catenin pathway expression is related with the survival of patients with thyroid carcinoma [24]. As expected, we found that silencing circ-0012417 could inhibit WNT/β-catenin. Further we found that inhibiting circ-0012417 could stimulate miR-29c-5p expression, and inhibiting miR-29c-5p could rescue decreased migration and invasion induced by silencing hsa-circ-001241, while dual luciferase report assay demonstrated that circ-0012417 may directly target miR-29c-5p (Fig. 6). And in vivo study showed that silencing hsa-circ-001241 could inhibit tumor growth signi cantly (Fig. 7). Finally, with TCGA and GEO database, we found that EGFL7 and LEMD1, the possible target of hsa-circ-001241, could be possible prognostic predictors in thyroid cancer.
CircRNAs are new class of non-coding RNAs and are mostly generated by back-splicing of exons of protein-coding genes. Like microRNAs and long non-coding RNAs, it is well-established that circRNAs exerts important physiological functions, especially in cancers [25]. Given that circRNAs are highly stable and their expression pattern highly depends on cell speci city, circRNAs could be considered as the novel biomarker for cancer diagnosis [26]. In present study, we identi ed that hsa-circ-0012417 is highly enriched in human thyroid cancer tissues and it was required for thyroid cancer cell proliferation and invasion. Although, the speci city and accuracy of applying hsa-circ-0012417 in thyroid cancer diagnosis needed to further evaluation, our nding suggested that hsa-circ-0012417 could be considered as a novel biomarker candidate for thyroid cancers.