LncRNA CASC19 Enhances the Radioresistance of Nasopharyngeal Carcinoma by Regulating the miR-340-3p/FKBP5 Axis

Radioresistance remains a serious obstacle encountered in the radiotherapy of nasopharyngeal carcinoma (NPC). Both mRNAs and non-coding RNAs (ncRNAs), including long ncRNA (lncRNA) and microRNA (miRNA), play essential roles in radiosensitivity. However, the comprehensive expression profiles and competing endogenous RNA (ceRNA) regulatory networks among lncRNAs, miRNAs, and mRNAs in NPC radioresistance are still bewildering. In this study, we performed an RNA-sequencing (RNA-seq) assay in the radioresistant NPC cells CNE2R and its parental cells CNE2 to identify the differentially expressed lncRNAs, miRNAs, and mRNAs. The ceRNA networks containing lncRNAs, miRNAs, and mRNAs were predicted on the basis of the Pearson correlation coefficients and authoritative miRanda databases. In accordance with bioinformatic analysis of the data of the tandem mass tag (TMT) assay of CNE2R and CNE2 cells and the gene chip assay of radioresistant NPC samples in pre- and post-radiotherapy, the radioresistance-related signaling network of lncRNA CASC19, miR-340-3p, and FKBP5 was screened and further verified using an RT-qPCR assay. CASC19 was positively associated with FKBP5 expression while negatively correlated with miR-340-3p, and the target binding sites of CASC19/miR-340-3p and miR-340-3p/FKBP5 were confirmed using a dual-luciferase reporter assay. Moreover, using an mRFP–GFP–LC3 maker, it was found that autophagy contributed to the radioresistance of NPC. MiR-340-3p inhibition or FKBP5 overexpression could rescue the suppression of autophagy and radioresistance induced by CASC19 knockdown in CNE2R cells. In conclusion, the CASC19/miR-340-3p/FKBP5 network may be instrumental in regulating NPC radioresistance by enhancing autophagy, which provides potential new therapeutic targets for NPC.


Introduction
Nasopharyngeal carcinoma (NPC), a unique type of head and neck cancer arising from the nasopharynx, has remarkable geographic and racial distribution worldwide. NPC has a high prevalence in Southeast Asia and southern China [1,2]. Due to its unique anatomical location and relatively high radiosensitivity, radiotherapy has become the first choice for the treatment of NPC [3]. However, approximately 30% of NPC patients eventually develop recurrence or/and distant metastasis due to radioresistance [4]. Therefore, it is crucial to further investigate the molecular mechanism underlying the residual NPC cells after irradiation to avoid radiotherapy failure.
At present, many non-coding RNAs (ncRNAs) have been identified and demonstrated to play vital roles in multiple physiological and pathological processes. Based on the size, these ncRNAs are subdivided into small ncRNAs (<200 nt, such as microRNAs) and long ncRNAs (lncRNAs) (>200 nt) [5]. Dysregulated lncRNAs are involved in tumorigenesis and

Expression Profiles of lncRNAs, miRNAs, and mRNAs in NPC Cells
We previously successfully established a radioresistant cell line (named CNE2R) from human NPC cells CNE2 [25]. In comparison with CNE2 cells, CNE2R cells had acquired radioresistance ( Figure 1A). Then, the RNA sequences of these two cell lines, including lncRNAs, miRNAs, and mRNAs, were acquired using the Illumina Hiseq2000 platform. It was found that there were 71 DElncRNAs (47 upregulation and 24 downregulation), 358 DEmiRNAs (166 upregulation and 192 downregulation), and 1155 DEmRNAs (645 upregulation and 510 downregulation) between CNE2R and CNE2 cells ( Figure 1B). Moreover, the differential expressed genes (DEGs) between CNE2 and CNE2R cells were analyzed using a TMT quantitative proteomic assay, and the volcano plot of these DEGs is shown in Figure 1C. 1B). Moreover, the differential expressed genes (DEGs) between CNE2 and CNE2R cells were analyzed using a TMT quantitative proteomic assay, and the volcano plot of these DEGs is shown in Figure 1C. Figure 1. RNA profile analyses of CNE2 and its paired radioresistance CNE2R cells: (A) colony formation assay was performed to evaluate the radiosensitivity of CNE2 and CNE2R cells. * p < 0.05 between indicated groups; (B) volcano plots of the DElncRNAs, DEmRNAs, and DEmiRNAs between CNE2 and CNE2R cells based on RNA−seq assay; (C) volcano plot of differentially expressed genes (DEGs) between CNE2 and CNE2R cells based on TMT assay.

CeRNA Networks were Constructed on the Basis of Screened mRNAs and CASC19
Our previous study demonstrated that CASC19 contributed to the poor prognosis and radioresistance of NPC [25]. To know the potential targets of CASC19, we established CASC19-related ceRNA networks based on the above DElncRNAs, DEmiRNAs, DEmR-NAs, and DEGs. As a result, six mRNAs (FKBP5, CBX5, CTSD, DHTKD1, FKBP10, and HSP90B1) and four miRNAs (miR-340-3p, miR-3065-3p, miR-3179, and miR-3614-3p) regulated by CACS19 were included in the ceRNA networks, shown in a Sankey diagram ( Figure 2A). Meanwhile, gene chip analysis gave the heat map of the distribution of DEGs Figure 1. RNA profile analyses of CNE2 and its paired radioresistance CNE2R cells: (A) colony formation assay was performed to evaluate the radiosensitivity of CNE2 and CNE2R cells. * p < 0.05 between indicated groups; (B) volcano plots of the DElncRNAs, DEmRNAs, and DEmiRNAs between CNE2 and CNE2R cells based on RNA−seq assay; (C) volcano plot of differentially expressed genes (DEGs) between CNE2 and CNE2R cells based on TMT assay.

CeRNA Networks were Constructed on the Basis of Screened mRNAs and CASC19
Our previous study demonstrated that CASC19 contributed to the poor prognosis and radioresistance of NPC [25]. To know the potential targets of CASC19, we established CASC19related ceRNA networks based on the above DElncRNAs, DEmiRNAs, DEmRNAs, and DEGs. As a result, six mRNAs (FKBP5, CBX5, CTSD, DHTKD1, FKBP10, and HSP90B1) and four miRNAs (miR-340-3p, miR-3065-3p, miR-3179, and miR-3614-3p) regulated by CACS19 were included in the ceRNA networks, shown in a Sankey diagram ( Figure 2A). Meanwhile, gene chip analysis gave the heat map of the distribution of DEGs between the tumor samples of three NPC radioresistant patients before and after radiotherapy ( Figure 2B), and a total of 292 aberrantly expressed genes were obtained. Only one commonly high-expressed gene, FKBP5, was identified in the cross-section among the multi-omic analyses of the above RNA-seq, TMT, and gene chip ( Figure 2C). Further analyses using GSE12452 and GSE13597 datasets in the GEO database demonstrated that the expression of FKBP5 was upregulated in NPC tissue in comparison with normal tissue ( Figure 2D). Analysis of the GEPIA database also demonstrated that the expression of the FKBP5 gene was obviously upregulated in 519 HNSC (head and neck squamous cell carcinoma) tissues compared with that in 44 normal tissues ( Figure 2E). Therefore, the high expression of FKBP5 was probably an important prognostic marker of NPC radioresistance. Moreover, it was reported that miR-340 could be decreased by ionizing radiation (IR) [26]. Hence, to demonstrate the feasibility and value of the constructed ceRNA networks, we chose the CASC19/miR-340-3p/FKBP5 axis for further validation.

Expressions of CASC19, FKBP5 mRNA and miR-340-3p in NPC Cells
We then examined the difference in the expression levels of CASC19, FKBP5 mRNA, and miR-340-3p in radiosensitive and radioresistant NPC cell lines. It was confirmed that CASC19 and FKBP5 were significantly highly expressed ( Figure 3A,B), and miR-340-3p was lowly expressed in CNE2R cells, compared with CNE2 cells (Figure 3C), indicating that CASC19 and FKBP5 had a negative correlation with miR-340-3p. Western blot assay also demonstrated that the FKBP5 protein level in CNE2R cells was higher than that in CNE2 cells ( Figure 3D). Moreover, the expressions of these ceRNAs could be altered via irradiation in a dose-dependent manner. The RT-PCR assay showed that the expressions of CASC19 and FKBP5 generally increased with irradiation dose ( Figure 3E,F), but the expression of miR-340-3p decreased with irradiation dose ( Figure 3G).

Expressions of CASC19, FKBP5 mRNA and miR-340-3p in NPC Cells
We then examined the difference in the expression levels of CASC19, FKBP5 mRNA, and miR-340-3p in radiosensitive and radioresistant NPC cell lines. It was confirmed that CASC19 and FKBP5 were significantly highly expressed ( Figure 3A,B), and miR-340-3p was lowly expressed in CNE2R cells, compared with CNE2 cells (Figure 3C), indicating that CASC19 and FKBP5 had a negative correlation with miR-340-3p. Western blot assay also demonstrated that the FKBP5 protein level in CNE2R cells was higher than that in CNE2 cells ( Figure 3D). Moreover, the expressions of these ceRNAs could be altered via irradiation in a dose-dependent manner. The RT-PCR assay showed that the expressions of CASC19 and FKBP5 generally increased with irradiation dose ( Figure 3E,F), but the expression of miR-340-3p decreased with irradiation dose ( Figure 3G).

CASC19 Negatively Regulated miR-340-3p to Promote Autophagy and Enhance Radioresistance
To know the relationship between the functions of CASC19 and miR-340-3p in radioresistance, we interfered in CNE2R cells with the CASC19 smart silencer (siCASC19),

CASC19 Negatively Regulated miR-340-3p to Promote Autophagy and Enhance Radioresistance
To know the relationship between the functions of CASC19 and miR-340-3p in radioresistance, we interfered in CNE2R cells with the CASC19 smart silencer (siCASC19), miR-340-3p mimics, miR-340-3p inhibitor, and their negative controls with high efficiencies ( Figure 4A-C). It was found that the expression of CASC19 was significantly downregulated by miR-340-3p mimics ( Figure 4D), and the transfection of CNE2R cells with siCASC19 decreased cell survival after irradiation, which was rescued by the miR-340-3p inhibitor ( Figure 4E). Therefore, CASC19 might promote the radioresistance of NPC cells by regulating miR-340-3p. The bioinformatic analysis predicted that miR-340-3p targeted the 3 UTRs of CASC19 with complementary binding sites ( Figure 4F). Then, we constructed luciferase reporter plasmids containing wild-type (WT) CASC19 or mutant-type (MUT) CASC19 and co-transfected them with miR-340-3p mimics into CNE2R cells. It was found that the luciferase activity of the reporter was reduced under the situation of MUT-CASC19 but had no alteration under WT-CASC19 ( Figure 4F), demonstrating that CASC19 serves as a sponge for miR-340-3p.   Furthermore, to determine whether CASC19 contributes to the autophagy-regulated radioresistance of NPC cells by negatively regulating miR-340-3p, we transfected siCASC19 and/or miR-340-3p inhibitor together with an autophagy indicator Ad-mRFP-GFP-LC3 into radioresistant CNE2R cells. As a result, siCASC19 reduced the number of autophagic LC3 spots in CNE2R cells, which was also rescued by the miR-340-3p inhibitor ( Figure 4G). In addition, the ratio of LC3II/I was significantly reduced, and the expression of P62 was increased in CNE2R cells by siCASC19 transfection ( Figure 4H), but these alterations did not occur when the cells were transfected with siCASC19 together with the miR-340-3p inhibitor. Accordingly, the functions of CASC19 in radioresistance and autophagy relied on its role of acting as the sponge of miR-340-3p.

CASC19 Positively Regulated the Expression of FKBP5 through miR-340-3p
It was found that the mRNA and protein expression levels of FKBP5 were significantly decreased in CNE2R cells after transfection with miR-340-3p mimics ( Figure 5A,B). Using the bioinformatic prediction tools of miRanda, RNAhybrid, and TargetScan databases, it was demonstrated that miR-340-3p could target the 3 UTR of FKBP5 mRNA with complementary binding sites, which was validated with the luciferase reporter gene assay ( Figure 5C). Further investigation showed that the downregulation of CASC19 positively affected FKBP5 expression in CNE2R cells ( Figure 5D), and the expression of CASC19 was decreased after knocking down FKBP5 ( Figure 5E). To verify the FKBP5-dependent radioresistance function of CASC19, we transfected CNE2R cells with siCASC19 or its scrambled vector together with the FKBP5-overexpression vector or its corresponding empty vector. Our results showed that the overexpression of FKBP5 increased the ratio of LC3 II/I but decreased the expression of P62 ( Figure 5F). Taken together, these findings demonstrated that CASC19 contributed to the radioresistance of NPC cells via the CASC19/miR-340-3p/FKBP5 regulatory axis.

Inhibition of FKBP5 Increased Radiosensitivity by Inhibiting Autophagy
It has been reported that FKBP5 can promote autophagy and increase the radioresistance of melanoma [27]. To know the regulation function of FKBP5 in the radiosensitivity of NPC, we transfected CNE2R and CNE2 cells with siFKBP5 and its negative control with a high transfection efficiency ( Figure 6A,B). After siFKBP5 interference, the survival fractions of irradiated NPC cells were significantly decreased, which was considerably profound in CNE2R cells ( Figure 6C). To determine whether FKBP5 contributes to the autophagyregulated radioresistance of NPC cells, we transfected siFKBP5 and Ad-mRFP-GFP-LC3 into radioresistant CNE2R cells and found that the number of autophagic LC3 spots in the siFKBP5 group sharply decreased in comparison with the negative control of siRNA ( Figure 6D). In line with this finding, the suppression of FKBP5 also decreased the ratio of LC3II/I in CNE2R cells ( Figure 6E). These results demonstrated that siFKBP5 enhanced the radiosensitivity of NPC cells by inhibiting autophagy.

Inhibition of FKBP5 Increased Radiosensitivity by Inhibiting Autophagy
It has been reported that FKBP5 can promote autophagy and increase the radioresistance of melanoma [27]. To know the regulation function of FKBP5 in the radiosensitivity of NPC, we transfected CNE2R and CNE2 cells with siFKBP5 and its negative control with a high transfection efficiency ( Figure 6A,B). After siFKBP5 interference, the sur- ably profound in CNE2R cells ( Figure 6C). To determine whether FKBP5 contributes to the autophagy-regulated radioresistance of NPC cells, we transfected siFKBP5 and Ad-mRFP-GFP-LC3 into radioresistant CNE2R cells and found that the number of autophagic LC3 spots in the siFKBP5 group sharply decreased in comparison with the negative control of siRNA ( Figure 6D). In line with this finding, the suppression of FKBP5 also decreased the ratio of LC3II/I in CNE2R cells ( Figure 6E). These results demonstrated that siFKBP5 enhanced the radiosensitivity of NPC cells by inhibiting autophagy.

Discussion
Radioresistance is a serious obstacle encountered in NPC treatment. During the past few decades, besides mRNAs, an increasing number of ncRNAs, including miRNA and lncRNA, have been demonstrated to play an essential role in the radioresistance of tumors. Actually, some pivotal mRNAs and ncRNAs have been reported to be involved in NPC radioresistance [28,29]. Moreover, several lines of evidence have shown that some ncRNAs can form ceRNA networks to competitively regulate the expression of critical coding genes [18]. Thus far, there have been many reports on the whole transcriptome expression profiles and related ceRNA regulatory networks in carcinogenesis and tumor chemoresistance [30,31]. However, there are few studies in the literature about the expression profiles of ncRNAs and related-ceRNA regulatory networks for the radioresistance of NPC. Therefore, clarifying the ceRNA regulatory network can improve the understanding of the mechanism of NPC radioresistance.
CASC19 is located on the 8q24 region of the chromosome and is referred to as an "oasis" of non-coding RNA [32]. Previous studies demonstrated that CASC19 was involved in clear cell renal cell carcinoma (ccRCC) [33], non-small cell lung cancer (NSCLC) [34], glioma [35], and pancreatic cancer [36]. Moreover, CASC19 could accelerate chondrocyte apoptosis and proinflammatory cytokine production and exacerbate osteoarthritis development by regulating the miR-152-3p/DDX6 axis [37]. Evidence suggested that the aberrantly expressed CASC19 was associated with chemotherapeutic drugs [38]. Notably, our previous study revealed that CASC19 contributed to NPC radioresistance [25]. As the prior function of CASC19 was mainly through the indirect adsorption of miRNAs, which affected its downstream target genes [33][34][35][36][37], we analyzed the RNA-seq of CNE2R and its parental CNE2 cells. Based on the analyses of the expression profiles of lncRNA, miRNA, and mRNA (Supplementary Tables S1-S3), as well as the TMT data of NPC cells (CNE2R vs. CNE2) (Supplementary Table S4) and the gene chip data of clinical NPC samples (Supplementary  Table S5), we proposed that a novel ceRNA network of CASC19/miR-340-3p/FKBP5 contributed to the radioresistance of NPC. Currently, miR-340 is widely regarded as a tumor suppressor in a variety of cancers. For example, miR-340 suppressed cell proliferation, promoted cell apoptosis, and inhibited cell migration and invasion in gastric cancer [39,40]. In ovarian cancer, miR-340 suppressed cell metastasis by targeting FHL2 [41]. Moreover, miR-340 could target BMI1, and the ectopic upregulation of BMI1 partially reversed the effect of miR-340 on the growth and migration of colorectal cancer [42]. In addition, several other studies found that miR-340 also acted as a tumor suppressor in glioblastoma [43,44], breast cancer [45], and non-small cell lung cancer [46]. In particular, the radiation-regulated miR-340/429/IL-4/β-catenin/Stat6 signaling axis effectively enhanced tumor progression and metastasis of human carcinomas [26]. Our study demonstrated that CASC19 could act as the sponge of miRNAs and it downregulated the intracellular level of miR-340-3p and thus influenced cell autophagy and radioresistance.
As a downstream target gene of miR-340-3p, FKBP5 was identified using the bioinformatic analyses of RNA-seq, TMT, and gene chip assays of cells and clinic samples. It was found that FKBP5 could mediate the effect of the CASC19/miR-340-3p axis on autophagy and the radioresistance of NPC cells. Previous studies have shown that FKBP5, also known as FK506 binding protein 5, is a member of the immunoaffinity protein family and a modulator of the Hsp90 chaperone protein and glucocorticoid receptor [47]. Several lines of evidence have revealed its functions in cancer biology [48]. Particularly, FKBP5 played an essential role in the resistance of melanoma to radiotherapy [27]. On the other hand, FKBP5 has also been proven to be a potent inducer of autophagy [49][50][51]. For example, FKBP5 could bind to BECN1, a key regulator of autophagy, and change its phosphorylation and protein expression levels, thereby promoting autophagy. In agreement with these findings, our data also disclosed that FKBP5 contributed to radioresistance by promoting autophagy in NPC cells. FKBP5 was also confirmed to be the target gene of miR-340-3p. Furthermore, we identified that CASC19 and FKBP5 presented a positive correlation, indicating that CASC19 increased radioresistance by enhancing autophagy through the miR-340-3p/FKBP5 axis.
Taken together, this study presented the comprehensive analysis of the lncRNA, miRNA, and mRNA expression profiles between radioresistance and radiosensitive NPC cells and built a novel ceRNA regulatory network, CASC19/miR-340-3p/FKBP5, as a key player in modulating the radiosensitivity of NPC cells by promoting autophagy (Figure 7). These results improved the knowledge of the mechanism of NPC radioresistance and might provide potential molecular targets for increasing the efficacy of NPC radiotherapy.
lines of evidence have revealed its functions in cancer biology [48]. Particularly, FKBP5 played an essential role in the resistance of melanoma to radiotherapy [27]. On the other hand, FKBP5 has also been proven to be a potent inducer of autophagy [49][50][51]. For example, FKBP5 could bind to BECN1, a key regulator of autophagy, and change its phosphorylation and protein expression levels, thereby promoting autophagy. In agreement with these findings, our data also disclosed that FKBP5 contributed to radioresistance by promoting autophagy in NPC cells. FKBP5 was also confirmed to be the target gene of miR-340-3p. Furthermore, we identified that CASC19 and FKBP5 presented a positive correlation, indicating that CASC19 increased radioresistance by enhancing autophagy through the miR-340-3p/FKBP5 axis.
Taken together, this study presented the comprehensive analysis of the lncRNA, miRNA, and mRNA expression profiles between radioresistance and radiosensitive NPC cells and built a novel ceRNA regulatory network, CASC19/miR-340-3p/FKBP5, as a key player in modulating the radiosensitivity of NPC cells by promoting autophagy (Figure  7). These results improved the knowledge of the mechanism of NPC radioresistance and might provide potential molecular targets for increasing the efficacy of NPC radiotherapy.

Cell Culture and Irradiation
Human NPC cell line CNE2 was purchased from Shanghai Cell Bank (Shanghai, China). The radioresistant CNE2R cell line was previously established from its parental CNE2 cells by fractionated irradiation with a total dose of 60 Gy [25]. These cells were cultured with RPMI-1640 medium (Gibco, Hangzhou, China) supplemented with 10% fetal bovine serum (Gibco Invitrogen, Grand Island, NY, USA), 1% penicillin-streptomycin (Gibco) and maintained in a humidified atmosphere of 5% CO2 at 37°C. NPC cells were irradiated with different doses using an X-ray irradiator (12 mA, 2 mm aluminum filtration; X-RAD 320, Precision X-Ray Inc., North Branford, CT, USA) at a dose rate of 0.883 Gy/min.

Cell Culture and Irradiation
Human NPC cell line CNE2 was purchased from Shanghai Cell Bank (Shanghai, China). The radioresistant CNE2R cell line was previously established from its parental CNE2 cells by fractionated irradiation with a total dose of 60 Gy [25]. These cells were cultured with RPMI-1640 medium (Gibco, Hangzhou, China) supplemented with 10% fetal bovine serum (Gibco Invitrogen, Grand Island, NY, USA), 1% penicillin-streptomycin (Gibco) and maintained in a humidified atmosphere of 5% CO 2 at 37 • C. NPC cells were irradiated with different doses using an X-ray irradiator (12 mA, 2 mm aluminum filtration; X-RAD 320, Precision X-Ray Inc., North Branford, CT, USA) at a dose rate of 0.883 Gy/min.

Colony Formation Assay
The radiosensitivities of NPC cell lines were assessed using a colony formation assay. Logarithmic growing cells were irradiated with different doses (2, 4, 6, and 8 Gy) of γ-rays (137-Cs, Gammacell-40, MDS Nordion, Ottawa, ON, Canada) at room temperature, cultured for about 10 days to form colonies, and then fixed and stained with crystal violet. Colonies consisting of more than 50 cells were counted for survival fraction. The cell survival curve (SF) was fitted with the single-hit multitarget model as SF = 1 − (1 − exp (−k × D)) N .

RNA Isolation for Sequencing Analysis
The total cellular RNA was extracted using TRIzol reagent (Invitrogen, San Diego, CA, USA) according to the manufacturer's instructions. DElncRNAs, DEmRNAs, and DEmiRNAs between the indicated cell lines were assessed using the DEGseq R package as previously described [29]. The transcript with Pval < 0.01 was set as the threshold for significant differences. Sequencing libraries for miRNAs were generated using NEBNext ® Multiplex Small RNA Library Prep Set for Illumina ® (New England Biolabs, Ipswich, MA, USA) following the manufacturer's recommendation. All sequencing programs and analyses were performed by Novogene Company (Beijing, China).

TMT Quantitative Proteomic Analysis
Total proteins were extracted from CNE2 and CNE2R cells, and their concentrations were detected with a BCA kit according to the manufacturer's instructions. A quantity of 0.2 mg of protein from each sample was used for TMT analysis. A high-resolution mass spectrometer Q Exactive plus (Thermo Fisher Scientific) was used to perform the quantitative proteomics analysis. The MS/MS data were performed with the Maxquant search engine (v.1.5.2.8).

Gene Chip Detection of Tumor Tissues of NPC Patients
Formalin-fixed paraffin-embedded (FFPE) NPC samples of three radioresistant patients pre-and post-radiotherapy were applied for gene chip detection (Bohao Biotechnology Co., Shanghai, China). According to the Declaration of Helsinki, all samples were obtained after ethical approval with the informed consent of patients receiving treatment at the Nanfang Hospital of Southern Medical University (Guangzhou, China).

Construction of CASC19-Associated ceRNA Network
To construct a radioresistance-related lncRNA-associated ceRNA network, we first performed RNA-seq analysis to screen the DEmiRNAs that may be involved in radioresistance [30]. Secondly, we constructed the lncRNA-miRNA-mRNA interaction network to predict the 'sponge' function of lncRNA. According to the expression levels of lncRNAs, miRNAs, and mRNAs, the Pearson correlation coefficients between miRNA and mRNA as well as miRNA and lncRNA were calculated, where the negative correlation coefficient > 0.85 was selected for further analyses using the miRanda database to obtain the ceRNA network. Moreover, applying the above TMT data and the gene chip data, we utilized the paired dysregulated lncRNA, miRNA, and mRNA expression profiles to construct the CASC19-associated ceRNA network involved in radioresistance using the R and Cytoscape software.

Real-Time PCR Assay of RNA
After the extraction of total RNA from NPC cells using a TRIzol reagent, the cDNA was synthesized from 1 µg of the total RNA using a PrimeScript RT Reagent Kit with gDNA Eraser (Takara Biotechnology, Co., Ltd., Dalian, China), and then it was amplified using real-time qPCR with Ultra SYBR Mixture (Low ROX) (CoWin Biosciences, Beijing, China) on an MX3000P platform with the internal control of housekeeping gene β-actin. MiR-340-3p was detected using a Mir-X miRNA qRT-PCR TB Green Kit (Takara Biotechnology, Co., Ltd., Dalian, China) with U6 RNA as an internal control. All primers were synthesized by Sangon (Shanghai, China) with the below sequences. For the CASC19 gene, the forward primer was 5 -TTT AGC CTG CAT AGG ACC CTC-3 , and the reverse primer was 5 -GTC TGG TCA AAT TAC AAT CAG TTG G-3 . For the β-actin gene, the forward primer was 5 -CAT GTA CGT TGC TAT CCA GGC-3 , and the reverse primer was 5 -CTC CTT AAT GTC ACG CAC GAT-3 . For miR-340-3p, the forward primer was 5 -CGC GTC CGT CTC AGT TAC TTT ATA GC-3 . For U6, the forward primer was 5 -TGG AAC GCT TCA CGA ATT TGC G-3 , and the reverse primer was 5 -GGA ACG ATA CAG AGA AGA TTA G-3 . For the FKBP5 gene, the forward primer was 5 -TAT GGC TCG GCT GGC AGT CTC-3 , and the reverse primer was 5 -CCC TCT CCT TTC CGT TTG GTT CTC-3 .

Western Blotting Assay
Total proteins were extracted from NPC cells with an ice-cold RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China), isolated by SDS-PAGE, and transferred onto 0.22 µm PVDF membrane (Immobilon-P, Millipore Corporation). The membrane was blocked for 2 h with 5% nonfat milk and then incubated with primary antibodies overnight at 4 • C. After washing with 0.05% Tris-buffered saline/Tween (TBST) well, the membrane was incubated with HRP-conjugated secondary antibodies for 2 h. Then, the proteins were visualized with an ECL Chemiluminescence Kit (Millipore, St. Louis, MO, USA) according to the manufacturer's instructions, and the protein band images were analyzed with the Bio-Rad ChemiDoc XRS system.

Autophagic Flux Assay
Briefly, 3 × 10 5 cells per well were incubated in a 24-well plate overnight. Then, cells were transfected with mRFP-GFP-LC3 double-labeled adenovirus (Hanbio, Shanghai, China) to label autophagosomes. After 2 h of transfection, the cells were washed twice with pre-cooled PBS and stained with DAPI. Intracellular autophagy was observed using a high-content imaging system (ImageXpress Micro 4, Molecular Devices, San Francisco, CA, USA). Double labeling of LC3 (green) and mRFP (red) immunofluorescence corresponds to the changes in the autophagic flux. After merging, the yellow puncta in the cell image symbolize the autophagosomes.

Dual-Luciferase Reporter Assay
The CASC19 luciferase reporter vector harboring miR-340-3p binding sequence (CASC19-WT) and site-directed mutation of targeted sequence (CASC19-MUT), and the FKBP5 3untranslated region (3 -UTR) luciferase reporter vector containing miR-340-3p targeted sites (FKBP5-WT) and site-directed mutation of targeted sequence (FKBP5-MUT) were constructed according to manufacturer's instructions (Genomeditech, Shanghai, China). To confirm the targeting relationship between CASC19 and miR-340-3p, the cells under 50% confluence were transfected with the CASC19-WT or CASC19-MUT vector together with miR-340-3p mimics or miR-NC mimics. To detect whether FKBP5 is a direct target of miR-340-3p, the FKBP5-WT or FKBP5-MUT vector was transfected into cells together with miR-340-3p mimics or miR-NC mimics. After 24 h transfection, cell lysates were prepared, and the luciferase activity was measured using a dual-luciferase reporter kit (Promega, Madison, WI, USA) following the manufacturer's instructions.

Statistical Analysis
Data are presented as means ± SD of at least three independent experiments. Data were analyzed with Student's t-test or one-way analysis of variance (ANOVA) using SPSS 19.0 software (SPSS, Chicago, IL, USA). p < 0.05 indicated a significant difference between the indicated samples.