Circular RNA hsa_circ_0006117 Facilitates Pancreatic Cancer Progression by Regulating the miR-96-5p/KRAS/MAPK Signaling Pathway

Circular RNAs (circRNAs) play key roles in many malignant tumors, including pancreatic cancer (PC); however, whether circular RNA hsa_circ_0006117, a newly identified circRNA, has a role in PC has not been investigated. Here, in order to elucidate the role and potential molecular mechanisms of circRNAs, we utilized bioinformatic tolls to screen the differentially expressed circRNAs in PC. Subsequently, circular RNA hsa_circ_0006117 was identified as being highly expressed in PC tissues in a screen of two GEO datasets, which was further verified in PC cell lines and tissues. Then, its molecular characteristics were investigated using methods such as Sanger sequencing and fluorescence in situ hybridization (FISH). Functional experiments subsequently indicated that circular RNA hsa_circ_0006117 facilitated the malignant behaviors of PC cells, prompting that it plays an oncogenic role in PC. Moreover, we found that circular RNA hsa_circ_0006117 exerts its PC-promoting effects via activating the KRAS/mitogen-activated protein kinase (MAPK) signaling pathway. Through bioinformatics exploration and dual-luciferase reporter assays, miR-96-5p was identified as a downstream target of circular RNA hsa_circ_0006117. A series of assays confirmed that circular RNA hsa_circ_0006117 acted as a miR-96-5p sponge, thereby promoting the malignant features of PC in a miR-96-5p/KRAS axis-dependent manner. Taken together, our study indicated, for the first time, that the specifically highly expressed circular RNA hsa_circ_0006117 facilitates PC progression via the modulation of the miR-96-5p/KRAS/MAPK signaling pathway and might be a hopeful therapeutic target for PC.


Introduction
Pancreatic cancer (PC) is a leading cause of cancer-related deaths worldwide among all cancers with the lowest fiveyear overall survival rate (10%) [1]. Although surgical resection, neoadjuvant therapies, and comprehensive targeted treatment have improved the treatment options, early metastasis and invasion, combined with the lack of effective and precise targeted therapies, still limit the prognosis of PC patients [2][3][4]. Indeed, a severely poor prognosis with overall survival of just 10-16 months is usually related to patients with locally advanced or unresectable PC [5]. is highlights the urgency in elucidating the molecular regulatory network in PC and searching for new or more effective therapeutic targets for treating disease.
Increasing evidence suggests that noncoding RNAs have widely participated in PC development [6][7][8]. Circular RNAs (circRNAs) are a sort of noncoding RNA stemmed from the back-splicing of precursor messenger RNAs (mRNAs). ey have a stable covalent circular structure and are resistant to digestion by RNase R [9]. Compared with their linear precursors, circRNAs have functions that are independent of their host genes [10], such as serving as microRNA (miRNA) sponges or protein scaffolds, interacting with RNA-binding proteins, regulating alternative splicing or transcription, and generating pseudogenes [11,12]. Emerging shreds of evidence have revealed that circRNAs act as miRNA sponges to affect downstream targets and play roles in various cancers, including PC [13,14]. Nevertheless, the roles of circRNAs in PC remain poorly understood.
Members of the rat sarcoma (RAS) oncogene family, including Kirsten-RAS (KRAS), Harvey-RAS (HRAS), and neuroblastoma-RAS (NRAS), exhibit the highest mutation frequency in human cancers, with associated mutations being identified in approximately 30% of all cancers [15]. KRAS is a driver gene of many diseases and one of the most common and frequently mutated genes in PC [16,17]. KRAS oncogenic mutations lead to the continued activation of downstream molecules, and the KRAS/mitogen-activated protein kinase (MAPK) signaling pathway is strongly associated with the development of PC, both of which enhance the malignant potential of this cancer [18,19]. Meanwhile, KRAS is also involved in the regulation of noncoding RNAs in some cancers [20][21][22]. However, the association between circRNAs and KRAS in PC has not been explored, nor has the underlying regulatory relationship.
Here, circular RNA hsa_circ_0006117 with high expression in PC was reidentified from two circRNA microarrays from the Gene Expression Omnibus (GEO) database. Subsequent functional experiments illustrated that circular RNA hsa_circ_0006117 promoted the rapid development of PC cells. Moreover, our results suggested that circular RNA hsa_circ_0006117 could activate the MAPK signaling pathway by relieving the miR-96-5p-mediated posttranscriptional suppression of KRAS. Taken together, we found that circular RNA hsa_circ_0006117 adsorbed miR-96-5p and acted in an axis-dependent regulation of miR-96-5p/ KRAS/MAPK, thereby facilitating proliferation, migration, and invasion in PC. Circular RNA hsa_circ_0006117 has the potential to represent a promising target for PC therapy.

Acquisition of Gene Expression and Identification of Differentially Expressed circRNAs (DECs).
Using the keywords "circular RNA," "pancreatic cancer" or "pancreatic ductal adenocarcinoma," the expression profile of circular RNAs in PC were searched from the GEO (https://www.ncbi. nlm.nih.gov/geo/) database. e microarray datasets GSE69362 and GSE79634 were included and downloaded for screening potential DECs. After identifying DECs using the R "limma" package, Venn diagram analysis (http://bioinfor matics.psb.ugent.be/webtools/Venn/) was performed to overlap and focus candidate DECs.

RT-qPCR, Agarose Gel Electrophoresis, and Sanger
Sequencing. We used an RNA-easy Isolation Reagent (Vazyme, Nanjing, China) to extract RNA based on its protocol. Real-time quantitative reverse transcription PCR (RT-qPCR) was performed using cDNA Synthesis Kit (Vazyme) and qPCR Probe Kit (Vazyme). Relative mRNA expression was analyzed with the 2 −ΔΔCT method and standardized to those of the appropriate internal references. All primers are displayed in Table 1. Divergent primers were used for PCR amplification of reverse-transcribed cDNA and 2% agarose gel was selected for agarose gel electrophoresis. en, agarose gel was used for Sanger sequencing with the assistance of Cloud-Seq Biotech (Shanghai, China).

RNase R Assay and Actinomycin D Treatment.
Total RNA (2 μg) was incubated with or without 3 U/μg RNase R (Epicentre, Madison, WI, USA) at 37°C for 30 minutes. en, RT-qPCR was used for analyzing the abundance of circular RNA hsa_circ_0006117 and its parent gene protein tyrosine phosphatase receptor type A (PTPRA). Meanwhile, PC cells were treated with 2 μg/mL actinomycin D (Sigma-Aldrich, St Louis, MO, USA) or dimethyl sulfoxide (DMSO) (Sigma-Aldrich) for 12 hours. After harvesting the cells, the stability of circular RNA hsa_circ_0006117 and linear RNA PTPRA was tested by RT-qPCR.

Separation of Cytoplasmic and Nuclear Fractions.
All subcellular RNA components of PC cells were isolated with a Cytoplasmic & Nuclear RNA Purification Kit (Norgen Biotek, orold, ON, Canada) based on the protocol. While the cytoplasmic/nuclear RNA ratio was assessed by RT-qPCR, U6 was used as a positive reference for nuclear RNA and GAPDH as a positive reference for cytoplasmic RNA.

Fluorescence In Situ Hybridization (FISH). A FISH Kit
(RiboBio, Guangzhou, China) was applied to visualize the subcellular localization of circRNA. PC cells were sowed in 24-well plates and reproduced until 60%-70% confluence. After fixation in 4% paraformaldehyde and permeabilization with TritonX-100, the cells were hatched with a prehybridization buffer for 30 min at 37°C and then with a Cy3-labeled circular RNA hsa_circ_0006117 probe (RiboBio) in hybridization buffer overnight at 37°C. Representative pictures were obtained using a fluorescence microscope (Olympus, Tokyo, Japan) at ×400 magnification.

Western Blot.
Dissolving PC cells with RIPA lysis buffer (Boster, Wuhan, China), 10% SDS-PAGE was used to segregate proteins and then transferred them onto PVDF membranes (Millipore, MA, USA). After blocking with 5% milk (total protein) or 5% BSA (phosphorylated protein), the membranes were hatched with the corresponding primary antibodies and an HRP-conjugated secondary antibody. Immunoreactive protein bands were exposed with an ECL reagent (Boster) and quantified by Image Lab Software. e following commercially antibodies were used in this experiment: KRAS, GAPDH, phosphorylated-MEK1/2 (P-MEK1/2), MEK1/2, phosphorylated-ERK1/2 (P-ERK1/2), and ERK (GAPDH and KRAS from ABclonal Technology, Wuhan, China; others from Cell Signaling Technology, Beverly, MA, USA).  e migration rate was normalized using the 0 h scratch area. At the same time, transwell assays was used to estimate the migratory and invasive capabilities of PC cells. e upper chamber was pretreated or not with 60 μL Matrigel (Matrigel BD Biosciences, NY, USA) before transfection of 200 μL serum-free PC cells (5 × 10 4 cells/well). And 600 μL of 20% FBS medium was placed in the lower chamber. Migrating or invading cells were counted after 24-28 h of incubation. Representative pictures were captured with an inverted microscope (Olympus).

Statistical
Analysis. SPSS version 26.0 (SPSS IBM, Armonk, NY, USA) was applied for statistical analysis. Independent sample Student's t-tests were used for comparisons between two groups, whereas one-way ANOVA was employed for comparisons of more than two groups. Pearson's correlation curve analysis was applied for assessing the correlations between different indicators. All tests were two-tailed, and a P value of <0.05 suggested statistical significance.

Identification and Characterization of Circular RNA hsa_circ_0006117.
We identified potential DECs from GSE69362 and GSE79634. Venn diagram analysis showed that the overlapping circRNAs (circular RNA hsa_-circ_0006117 & circular RNA hsa_circ_0029634) in two datasets were both elevated in PC tissues (Figure 1(a) and Supplementary figures S1A and S1B). We further found that circular RNA hsa_circ_0006117 (Figure 1(b)) was differentially expressed between PC and paracancerous samples in each data set. en, we concentrated on exploring circular RNA hsa_circ_0006117. RT-qPCR analysis indicated that circular RNA hsa_circ_0006117 (Figure 1(c)) was significantly upregulated in 20 pairs of PC tissues. Meanwhile, circular RNA hsa_circ_0006117 expression was upregulated in PC cells compared with that in HPDE cells (Figure 1(d)). In terms of the annotation in circBase (http://www.circbase.org/), we found that circular RNA hsa_circ_0006117 is derived from 5′ of exon 8 to 3′ of exon 9 of PTPRA (Chr20: 2944917-2945848) by back-splicing ( Figure 1(e), upper panel). us, we designed divergent and convergent primers to amplify the back-spliced (circular RNA hsa_circ_0006117) and linear products (PTPRA), respectively. Circular RNA hsa_circ_0006117 could only be amplified from cDNA obtained from MIA PaCa-2 cells using divergent primers (Figure 1(e) (lower panel), Figure 1(f)). Next, Sanger sequencing of PCR amplicons validated the back-splicing site of circular RNA hsa_-circ_0006117 ( Figure 1(g)). After that, the stability of circular RNA hsa_circ_0006117 in PC cells treated with RNase R or actinomycin D was analyzed. RT-qPCR results revealed that circular RNA hsa_circ_0006117 had strong tolerance to RNase R digestion, whereas the linear form of PTPRA was rapidly degraded (Figure 1(h)). A stability assay using actinomycin D treatment also demonstrated that circular RNA hsa_circ_0006117 was substantially more stable than linear PTPRA in PC cells (Supplementary figure S1C). Furthermore, separation of cytoplasmic and nuclear fractions (Figure 1(i)) and FISH (Figure 1(j)) suggested that the subcellular localization of circular RNA hsa_circ_0006117 was mainly in the cytoplasm.

Circular RNA hsa_circ_0006117 Facilitated the Proliferation, Migration, and Invasion of PC Cells.
A series of loss-offunction experiments were applied to illustrate the function of circular RNA hsa_circ_0006117 in PC. For this, we designed siRNAs targeting the unique back-splicing site of circular RNA hsa_circ_0006117 (Supplementary figure S2) and used RT-qPCR to confirm that transfection of these siRNAs reduced the expression of circular RNA hsa_-circ_0006117, but not that of PTPRA (Figure 2(a)). Because circRNA-si#1 and circRNA-si#2 elicited the best results, they were used for subsequent experiments. e proliferation capacity was suppressed by the downregulation of circular RNA hsa_circ_0006117 in PC cells, which was confirmed by CCK-8 (Figure 2(b)) and colony formation (Figures 2(c) and 2(d)) assays. Moreover, both wound healing (Figure 2(e)) and transwell (Figures 2(f ) and 2(g)) assays indicated that downregulating circular RNA hsa_-circ_0006117 markedly weakened the invasiveness and migratory capability of PC cells.

Circular RNA hsa_circ_0006117 Maintained the Malignant Characteristics of PC via Activating the MAPK Signaling
Pathway. Studies have suggested that exon-derived circR-NAs mainly exert their functions in tumor development by sponging miRNAs [12]. To explore whether circular RNA hsa_circ_0006117 functions as a miRNA sponge, we used bioinformatics databases, such as circBANK, miRDB, Tar-getScan, and miRTarBase, to predict the targets of circular RNA hsa_circ_0006117 within the competitive endogenous RNA network associated with this circRNA. Meanwhile, we found that the MAPK and RAS signaling pathways were the most likely downstream targets of circular RNA hsa_-circ_0006117 in KEGG enrichment analysis (Figure 3(a)). Subsequently, combined with gene expression profile from e Cancer Genome Atlas (TCGA) database, we focused on genes involved in the MAPK signaling pathway and identified highly expressed KRAS, GRB2, IGF2BP2, and RAP1A in PC tissues as the most significantly enriched genes, implicating them as potential targets of circular RNA hsa_-circ_0006117 (Figure 3(b) and Supplementary figures S3A-S3C). In addition, we found that KRAS reduced overall survival in patients with PC based on data from the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia2.cancer-pku.cn/#index) (Figure 3(c)).
erefore, we further constructed shRNAs using the sequence of circRNA-si#1 and circRNA-si#2 and then transfected them into PC cells for follow-up experiments. Interestingly, the mRNA (Figure 3    e circular RNA hsa_circ_0006117 probe was labeled with Cy3 (red) and the nucleus was counterstained with DAPI (blue). Representative images were obtained at ×400 magnification (bars: 20 μm). All values were shown as means ± SD, ns: not significant, * P < 0.05, and * * P < 0.001. 6 Journal of Oncology   Transwell migration (f ) and invasion (g) assays in circular RNA hsa_circ_0006117silenced PC cells were applied to evaluate the invasiveness and migration capabilities of PC, respectively. Representative images were obtained at ×200 magnification (bars: 50 μm). All values were shown as means ± SD, * P < 0.05, and * * P < 0.001. protein (Figure 3(e)) expression of KRAS were both downregulated in PC cells transfected with shRNA targeting circular RNA hsa_circ_0006117. Moreover, a direct relationship between the expression of circular RNA hsa_-circ_0006117 and KRAS in PC tissue samples was revealed by Pearson's correlation analysis (Figure 3(f )). It is well known that KRAS contributes to the upregulation of the MAPK signaling pathway, we hypothesized that circular RNA hsa_circ_0006117 was also involved in the influence on the MAPK signaling pathway. We verified that the protein levels of phosphorylated mitogen-activated extracellular signal-regulated kinase 1/2 (P-MEK1/2) and phosphorylated extracellular signal-regulated kinases 1/2 (P-ERK1/2) were lower in circular RNA hsa_circ_0006117-silenced PC cells than in the negative controls, whereas the total MEK1/2 and ERK 1/2 protein level remained unchanged (Figure 3(g)).

e Circular RNA hsa_circ_0006117-Mediated Malignant Progression of PC Was Dependent on KRAS.
To verify whether circular RNA hsa_circ_0006117 facilitates PC progression through regulating KRAS, we overexpressed KRAS in circular RNA hsa_circ_0006117-silenced PC cells to rescue the inhibitory effects of circRNA-sh#2. Proliferative capacity assays revealed that ectopically expressed KRAS could partially rescue the effects induced by circular RNA hsa_circ_0006117 silencing and facilitated the growth of PC cells (Figures 4(a) and 4(b)). Similarly, wound healing (Figure 4(c)) and transwell (Figures 4(d) and 4(e)) assays demonstrated that KRAS overexpression partially rescued the circular RNA hsa_circ_0006117 knockdown-induced effects and reinforced the migration and invasion potential of PC cells. Furthermore, in PC cells where circular RNA hsa_-circ_0006117 was knocked down, KRAS transfection partially restored the expression of P-MEK1/2 and P-ERK1/2 when the total MEK1/2 and ERK1/2 remained constant (Figure 4(f)).

e Circular RNA hsa_circ_0006117-Dependent KRAS Regulation in PC Progression Was Mediated by miR-96-5p.
e above studies have shown that circular RNA hsa_-circ_0006117 may play PC-promoting effects by sponging   (g) e expression of P-MEK1/2, MEK1/2, P-ERK1/2, and ERK1/2 were detected in circular RNA hsa_circ_0006117-silenced PC cells. All values were shown as means ± SD, * P < 0.05, and * * P < 0.001. 8 Journal of Oncology  miRNAs. Hence, we used bioinformatic analysis to explore which miRNAs can bind both circular RNA hsa_-circ_0006117 and KRAS, and miR-96-5p was identified as a possible candidate (Figure 5(a)). RT-qPCR indicated that miR-96-5p was lower expressed in PC cells than in HPDE cells ( Figure 5(b)). On the contrary, miR-96-5p content in PC cells transfected with circRNA-sh#1 and circRNA-sh#2 increased ( Figure 5(c)), whereas the content of KRAS in PC cells transfected with miR-96-5p mimics reduced ( Figures 5(d), and 5(e)). Furthermore, the data presented in Figure 5(f ) showed that miR-96-5p content was inversely associated with the content of both circular RNA hsa_circ_0006117 and KRAS. ese results suggested that circular RNA hsa_circ_0006117 and KRAS may share miR-96-5p binding sites. We subsequently generated pmiR-RB-Report ™ constructs containing either circular RNA hsa_circ_0006117 or KRAS 3′UTR sequences and confirmed them by sequencing (Supplementary figures S4A-S4D). As expected, the luciferase intensity of PC cells cotransfected with miR-96-5p mimics and circRNA-WT was significantly weaker than that cotransfected with miR-96-5p mimics and circRNA-MUT, prompting that miR-96-5p and circular RNA hsa_circ_0006117 have complementary binding sequences (Figures 5(g) and 5(h)). Meanwhile, the same complementary sequences were also present in the KRAS 3′UTR (the predicted binding sites are displayed in Figure 5(i)). Subsequently, the results demonstrated that PC cells cotransfected with the miR-96-5p mimics KRAS-WT, but not KRAS-MUT, displayed reduced luciferase activity, suggesting that miR-96-5p interacted with KRAS by binding to sequences in its 3′UTR (Figure 5(j)).

Discussion
PC displays highly invasive and metastatic characteristics [23,24]. Advances in the comprehensive systemic treatment of PC have not resulted in improvements in the prognosis of this cancer [25]. Numerous pieces of evidence have suggested that circRNAs have inspired the development of PC. A study by Li et al. [26] revealed that circ-IARS secreted by PC cells affected endothelial monolayer permeability, thereby facilitating PC invasion and metastasis. In addition,  another study [14] suggested that circ-PDE8A could facilitate the invasiveness capability of PC cells through the stimulation of the MET/ERK or AKT pathways, and exosomal circ-PDE8A has been related to the PC process and the prognosis of PC patients. In contrast, reduced circRNA expression has also been linked with the inhibition of PC progression. For example, Kong et al. [6] revealed that circNFIB1 regulates the miR-486-5p/PIK3R1 axis and further suppresses lymphatic metastasis in PC. Analogously, circular RNA hsa_circ_0006117 is reported to be low expressed and identified as a tumor-inhibiting factor in nonsmall cell lung carcinomas [27] and bladder cancer [28] recently. Surprisingly, our analysis results suggested that circular RNA hsa_circ_0006117 was highly expressed in PC, which indicated that it may have tissue specific and play a cancer-promoting role in PC. After all, the tissue specificity of circRNAs is one of their most striking features. However, current studies have not found a cancer-promoting role of circular RNA hsa_circ_0006117. We reported for the first time that circular RNA hsa_circ_0006117 is upregulated and played a promotive role in PC. ese clues provided novel insights for further exploring the role of circular RNA hsa_circ_0006117. Furthermore, our results confirmed the PC tissue specificity of circular RNA hsa_circ_0006117 expression and revealed its independent biological function. e KRAS/MAPK signaling pathway is strongly related to the growth and survival of cancer cells [29,30]. is pathway is also involved in resistance to chemotherapy, autophagy, and metabolic reprogramming, all of which contribute to the malignant capacity of PC [19,31]. In addition, KRAS promotes PC development through the regulation of noncoding RNA or nucleotide synthesis, while KRAS is a well-known activator that mediates the phosphorylation/activation of the MAPK signaling pathway [32,33]. Here, our results also support this idea. However, whether there is an association between KRAS and circRNAs (e, f ) e invasive and migratory capabilities of PC cells were investigated by transwell migration (e) and invasion (f ) assay, respectively. Representative images were obtained at ×200 magnification (bars: 50 μm). (g) Protein expression of key genes in the MAPK signaling pathway in circular RNA hsa_circ_0006117-silenced and (or) miR-96-5p-knockdown PC cells. All values were shown as means ± SD, * P < 0.05, and * * P < 0.001.
in PC has not been investigated. In our research, we first demonstrated that circular RNA hsa_circ_0006117 could upregulate the expression of KRAS via the competitive absorption of miRNA. And we found the cancer-promoting role of miR-96-5p in PC, which is consistent with previous work by others [34]. Intriguingly, previous studies [17,35] have identified KRAS as a candidate for genetic therapy for PC treatment. Further, a recently published study suggested that an inhibitor targeting mutated KRAS represented an effective treatment for some types of tumors [36]. Besides, therapies targeting MEK/ERK are considered to be promising treatments to slow the progression of PC [31]. Our study indicated that circular RNA hsa_circ_0006117 could stimulate the MAPK signaling pathway via up-regulating the phosphorylation of MEK/ERK and accelerate PC progression in a KRAS-dependent manner. Combined, we believed that circular RNA hsa_circ_0006117 may be a promising pharmacological target, and exploring the functions of KRAS-associated circRNAs may be of value for clinical application. Accumulating evidence has indicated that circRNAs can play both oncogenic and tumor-suppressor roles via sponging miRNAs [37,38]. For instance, Luo et al. [39] uncovered that circCCDC9 modulates the miR-6792-3p/CAV1 axis, thereby suppressing the development of gastric cancer. Another study revealed [40] that hsa_circ_001783 adsorbed miR-200c-3p to accelerate the malignant behavior of breast cancer. In our study, we also discovered that circular RNA hsa_circ_0006117 plays an oncogenic role in PC by sponging a miRNA (miR-96-5p). Notably, in similar research, most studies first identify miRNAs and then explore the associated downstream circRNA regulatory network. However, we first focused on the highly expressed genes in potential signaling pathways of PC and then searched for downstream targets and further determined the circRNA-related regulatory network, which led to miR-96-5p/KRAS and was identified as the downstream target of circular RNA hsa_circ_0006117 in PC. is provides a novel idea for the investigation of the ceRNA mechanism.
Of course, it is undeniable that not all ceRNA mechanisms can be successfully verified by this approach. Combined, all of our work indicated that circular RNA hsa_circ_0006117 promotes PC progression in a manner that is dependent on the downregulation of miR-96-5p.

Conclusion
In summary, we have identified circular RNA hsa_-circ_0006117 as a specifically highly expressed circRNA in PC. We further found that circular RNA hsa_circ_0006117 facilitates the malignant behaviors of PC through regulating the miR-96-5p/KRAS/MAPK signaling pathway (Figure 7). ese results suggest that circular RNA hsa_circ_0006117 may contribute to the potential therapeutic target of PC.

Data Availability
e data used to support the findings of this study are included within the article and supplementary information files.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this study.  14 Journal of Oncology