CircLDLR Promotes Papillary Thyroid Carcinoma Tumorigenicity by Regulating miR-637/LMO4 Axis

Background Circular RNAs (circRNAs) have been reported to play important roles in the development and progression of papillary thyroid carcinoma (PTC). However, the function and molecular mechanism of circRNA low-density lipoprotein receptor (circLDLR) in the tumorigenesis of PTC remain unknown. Results In this study, circLDLR was found to be markedly upregulated in PTC tissues and cell lines, and knockdown of circLDLR inhibited PTC cell proliferation, migration, and invasion but induced apoptosis in vitro. Moreover, circLDLR acted as a sponge for miR-637, and miR-637 interference reversed the anticancer effects of circLDLR knockdown on PTC cells. LMO4 was verified to be a target of miR-637; LMO4 upregulation abolished miR-637 mediated inhibition of cell growth and metastasis in PTC. Additionally, circLDLR could indirectly modulate LMO4 via acting as a sponge of miR-637 in PTC cells. Besides that, xenograft analysis showed that circLDLR knockdown suppressed tumor growth in vivo via regulating LMO4 and miR-637. Conclusion Taken together, these results demonstrated that circLDLR promoted PTC tumorigenesis through miR-637/LMO4 axis, which may provide a novel insight into the understanding of PTC tumorigenesis and be useful in developing potential targets for PTC treatment.


Introduction
Thyroid cancer is the most common type of human malignancies in endocrine system and is histologically categorized into papillary, follicular, medullary, anaplastic, and poorly differentiated thyroid cancer [1,2], among which, papillary thyroid carcinoma (PTC) is the most common type of thyroid carcinomas with a rising prevalence at an average annual rate of nearly 4% in recent years [1,2]. Early PTC has relatively good survival, while the 5-year survival rate of PTC with advanced stage is only 59% [3]. Importantly, although great advance in the clinical treatment of PTC, including surgery and/or physiotherapy or chemotherapy, recurrence and metastasis still occurred with high rates [4]. Thus, it is of great significance to further identify the pathogenesis of PTC to develop novel effective treatments for PTC. Circular RNAs (circRNAs) are one of stable and highly conservative noncoding RNA molecules with a covalently closed loop lacking the 5 ′ -end cap and the 3 ′ -end poly A tail [5,6]. Increasing studies have shown the involvement of cir-cRNAs in the tumorigenesis of many types of malignancy, such as gastric cancer [7], bladder cancer [8], hepatocellular carcinoma [9], and so on. Besides, previous studies also revealed that abnormal expression of circRNAs is drawn into the initiation and progression of PTC through by affecting the malignant behaviors of cancer cells [10,11]. Circular RNA low-density lipoprotein receptor (circLDLR, hsa_circ_ 0003892) is a derived from LDLR gene with the length of 544 bp; it locates at chr19: 11230767-11238761. According to the analysis of GSE93522 dataset, the expression of circLDLR was found to be aberrantly upregulated in PTC. However, the biological functions of circLDLR in the progression of PTC remain unknown.
MicroRNAs (miRNAs) are one of small noncoding RNAs and can modulate specific gene expression programs by regulating posttranslational processes [9]. miRNAs have been documented to be involved in a variety of biological functions in cancers, thus affecting tumor progression [12].
Recently, Yuan et al. demonstrated that lncRNA HOTTIP promoted PTC progression by inducing cell malignant biological behavior via targeting miR-637 [13]. However, the action and molecular mechanism of miR-637 in PTC progression remain vague. Additionally, it is well known that miRNAs commonly exert their functions through targeting 3 ′ -UTR of their target genes [14]. LIM-only protein 4 (LMO4) is a member of LIM domain only proteins (LMOs), which belongs to a subfamily of LIM-containing proteins [15]. Previous study showed LMO4 was increased in PTC and promoted cell growth, migration, and invasion via cir-cBACH2/miR-139-5p/LMO4 axis [16]. Thus, it is not clear whether LMO4 is involved in PTC tumorigenesis as a target of miR-637.
In this study, we investigated the function of circLDLR in PTC carcinogenesis in vitro and in vivo and evaluated whether circLDLR exerted it biological functions via circLDLR/miR-637/LMO4 in regulating cell biological behaviors of PTC.

Clinical Specimens.
Tumor tissues and para-tumor samples were collected from 45 PTC patients in Wuhan No.1.Hospital and then immediately stored at -80°C until RNA isolation. The screening of included PTC patients was shown in Fig. S1. Follow-up was conducted regularly every 3 months in the first 2 years after surgery and reduced to every 6 months from the third year. The last follow-up was performed in April 2019. We had obtained the agreement of the Ethics Committee of Wuhan No.1.Hospital, and written informed consent was collected form all patients before this study.

Quantitative Real-Time Polymerase Chain Reaction
(qRT-PCR). The isolation of total RNA was performed with the help of Trizol reagent (Invitrogen). A PARIS Kit (Invitrogen) was used to define the subcellular localization of circLDLR according to the manufacturer's instructions. Complementary DNA (cDNA) generation was conducted using a High-Capacity cDNA Reverse Transcription Kit (Qiagen, Valencia, CA, USA), and then, qPCR was implemented by a miScript SYBR Green PCR Kit (Qiagen). The relative expression was detected using the 2 -ΔΔCt method with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or U6 small nuclear B noncoding RNA (U6) as an internal control. The PCR reaction program started at 95°C for 2 min, 40 cycles for 95°C for 10 s followed by 60°C for 30 s. The same experiment was repeated three times, and the average was taken. The primer sequences are as follows: circLDLR: 2.4. RNase R Treatment. The RNA (2 μg) was interacted with or without RNase R (3 U/mg, Qiagen) at 37°C for 20 min. Then, the resulting RNA was purified using an RNeasy MinElute Cleanup Kit (Qiagen) and subjected to qRT-PCR analysis. The results represent as the average of three independent replicates. For colony formation assay, transfected cells (5000/well) with RPMI-1640 medium were seeded in 6-well plates, and the medium was replaced with new medium every 3 days. Finally, the typical images were photographed, and the number of visible colonies (≥50 cells) was counted after 14 days of incubation. The experiment was repeated three times. 2.13. Statistical Analysis. Data of at least three experiments were shown as mean values with standard deviation (SD). All statistical analyses were performed using the GraphPad Prism 7 software. Statistical differences between groups were analyzed using Student's t-test, one-way or two-way analysis of variance (ANOVA) followed by Tukey's test as appropriate. The correlation analysis was carried out with Pearson correlation analysis. P value < 0.05 ( * ), P value < 0.01 ( * * ), or P value < 0.001 ( * * * ) was considered as statistically significant.

CircLDLR Expression in PTC and Its Correlation with
Overall Survival. The expression of circLDLR in 45 PTC tissues and matched normal tissues was detected, and qRT-PCR analysis displayed that circLDLR was remarkably elevated in tumor tissues relative to those in normal tissues (Figure 1(a)). Then, the correlations of circLDLR expression and special clinicopathological parameters of PTC were analyzed; it was proved that higher circLDLR expression was correlated with advanced TNM stages, tumor size, and lymph node metastasis ( Table 1, P < 0:05). Besides that, patients with PTC were divided into two groups depending on the median level of circLDLR expression, and we found that patients in the high circLDLR group had a significantly shorter overall survival than those in the low circLDLR group (Figure 1(b)). Then, the expression of circLDLR in cells was measured; as expected, circLDLR also was higher in PTC cell lines (TPC-1 and IHH-4) than those in normal Nthy-ori 3-1 cell lines (Figure 1(c)). Afterwards, the stability of circLDLR was investigated. Total RNA from proliferating TPC-1 and IHH-4 cells was treated with RNase R; by contrast with linear LDLR mRNA, circLDLR could resistant to 3 Disease Markers the degradation by RNase R, indicating circLDLR stably functioned as a typical circRNA (Figure 1(d)). Besides that, qRT-PCR indicated that circLDLR was distributed mainly in the cytoplasm (Figure 1(e)). These data confirmed that circLDLR expression was elevated in PTC, and high circLDLR predicted poor overall survival in patients with PTC.

CircLDLR Knockdown Suppresses Cell Malignant
Phenotypes in PTC. To explore the function of circLDLR in PTC, TPC-1 and IHH-4 cells were transfected with si-circLDLR or si-NC; then, qRT-PCR analysis showed that circLDLR expression was significantly reduced by si-circLDLR as expected (Figure 2(a)). Subsequently, CCK-8 assay indicated that circLDLR knockdown inhibited the proliferation of TPC-1 and IHH-4 cells (Figure 2(b)); similarly, colony formation analysis also revealed that circLDLR silencing decreased the number of colonies formed in TPC-1 and IHH-4 cells (Figure 2(c)). Conversely, the apoptosis of TPC-1 and IHH-4 cells was induced by circLDLR knockdown (Figure 2(d)). Meanwhile, transwell assay suggested that the invasion ability of TPC-1 and IHH-4 cells was suppressed by circLDLR downregulation (Figure 2 [17]. Thus, whether circLDLR enhanced the malignant biological behaviors of PTC cell by sponging miRNAs was explored. According to prediction of CircInteractome, seven miRNAs with high score were selected, and we found that circLDLR knockdown significantly led to an increase of miR-637 expression in TPC-1 cells (Fig. S2). Thus, we hypothesized that miR-637 might be a target of circLDLR. The binding sites of miR-637 on circLDLR are shown in Figure 3(a). Then, the dual luciferase reporter assay   7 Disease Markers exhibited that miR-637 overexpression dramatically reduced the luciferase activity of the circLDLR-wt group in TPC-1 and IHH-4 cells but not the circLDLR-mut group (Figure 3(b)). In the meanwhile, we discovered that circLDLR and miR-637 pulled down by Anti-Ago2 was predominantly enriched in the Ago2 overexpression group compared with the control Anti-IgG group in TPC-1 and IHH-4 cells (Figure 3(c)). These results confirmed that circLDLR targeted miR-637 in PTC. Additionally, qRT-PCR analysis indicated that miR-637 expression was inhibited by circLDLR overexpression but was elevated by circLDLR downregulation in TPC-1 and IHH-4 cells (Figure 3(d)). Subsequently, the expression of miR-637 in PTC was detected, and miR-637 was decreased in PTC tissues (Figure 3(e)) and was negatively correlated with circLDLR ( Figure 3(f)). Also, miR-637 had a low expression in TPC-1 and IHH-4 cells compared with the normal Nthyori 3-1 cell lines (Figure 3(g)). Therefore, these results indicated that miR-637 was a target of circLDLR and might be associated with the development of PTC.

MiR-637 Interference Reverses the Antitumor Effects
Induced by si-circLDLR in PTC Cells. We then investigated whether circLDLR promoted the progression of PTC by interacting with miR-637. TPC-1 and IHH-4 cells were transfected with si-NC, si-circLDLR, si-circLDLR + anti-NC, or si-circLDLR + anti-miR-637, and qRT-PCR analysis showed that miR-637 inhibition attenuated si-circLDLRinduced elevation of miR-637 in TPC-1 and IHH-4 cells (Figure 4(a)), suggesting the success of transfection. After that, rescue assay was conducted, and results showed miR-637 interference reversed circLDLR knockdownmediated impairments of the proliferation (Figures 4(b) and 4(c)), invasion (Figure 4(e)) and migration (Figure 4(f)), and promotion of apoptosis (Figure 4(d)) in TPC-1 and IHH-4 cells. Accordingly, western blot analysis also confirmed miR-637 inhibition could partly abate si-circLDLR-mediated migration and invasion inhibition and apoptosis enhancement, evidenced by the change in C-cas-pase3, MMP2, and MMP9 protein levels in TPC-1 and IHH-4 cells (Figures 4(g) and 4(h)). In all, we verified that 3.5. LMO4 Is a Target of miR-637 and circLDLR Regulates LMO4 via miR-637 in PTC Cells. The related target genes of miR-637 were further predicted. According to the prediction of Targetscan program, LMO4 was found that might be a target of miR-760 with putative binding sites ( Figure 5(a)). Then, a reduction of luciferase activity in TPC-1 and IHH-4 cells cotransfected with LMO4-wt and miR-637 mimic was detected ( Figure 5(b)). Moreover, RIP assay displayed that LMO4 and miR-637 expression was notably enriched in Ago2 immunoprecipitates relative to the control IgG immunoprecipitates ( Figure 5(c)). These data revealed that LMO4 was a target of miR-637. Afterwards, we discovered that LMO4 expression was decreased by miR-637 overexpression, and miR-637 overexpression reversed circLDLRinduced increase of LMO4 level in TPC-1 and IHH-4 cells ( Figure 5(d)). Thus, a circLDLR/miR-637/LMO4 axis in PTC was identified. After that, the expression of LMO4 was analyzed; we found that LMO4 expression was upregulated in PTC tumor tissues at mRNA and protein levels ( Figures 5(e) and 5(f)). Besides that, LMO4 was positively correlated with circLDLR ( Figure 5(g)), while negatively   3.6. MiR-637 Suppresses Cell Malignant Phenotypes in PTC Cells by Targeting LMO4. Given the axis of circLDLR/miR-637/LMO4 in PTC cells, we further studied whether LMO4 was involved in the action of circLDLR/miR-637 axis in PTC progression. TPC-1 and IHH-4 cells were transfected with NC, miR-637, miR-637 + vector, or miR-637 + LMO4, and then qRT-PCR analysis showed that LMO4 overexpression rescued miR-637 mimic-induced decrease of LMO4 in TPC-1 and IHH-4 cells (Figure 6(a)). Then, functional experiments were conducted. Results exhibited that miR-637 re-expression impaired the proliferation (Figures 6(b) and 6(c)), invasion ( Figure 6(e)), and migration ( Figure 6(f)) but induced apoptosis (Figure 6(d)) in TPC-1 and IHH-4 cells. Besides that, miR-637 overexpression led to the increase of C-caspase3 and decrease of MMP2 and MMP9 in TPC-1 and IHH-4 cells (Figures 6(g) and 6(h)). Thus, miR-637 inhibited PTC progression. However, rescue assay showed that the antitumor functions of miR-637 were significantly reversed by LMO4 overexpression in TPC-1 and IHH-4 cells (Figures 6(b) and 6(h)). Altogether, miR-637 suppressed PTC progression partly by targeting LMO4.  (Figure 7(d)) in xenograft tumors. Furthermore, western bolt analysis showed that PCNA was decreased, but C-caspase3 was increased in the xenograft tumors of the sh-circLDLR group compared with the sh-NC group, furthering revealing circLDLR silencing suppressed tumor growth in vivo. Collectively, circLDLR silencing hindered tumor growth in vivo via regulating miR-637 and LMO4.

Discussion
PTC is the most frequent endocrine malignancy and generally has a good prognosis. Nevertheless, accumulating evidence has exhibited that early stage extradural invasion, lymph node metastasis, distant metastases, and recurrence occasionally occur in some cases, which lead to a poor prognosis, even death in PTC patients [14,18]. Up to date, increasing researches have reported the implication of cir-cRNAs in the development and progression of cancers through modulating diverse cellular processes [16,19,20]. In PTC, some circRNAs have also been indicated to be