Exosomal CTCF Confers Cisplatin Resistance in Osteosarcoma by Promoting Autophagy via the IGF2-AS/miR-579-3p/MSH6 Axis

Cancer-derived exosomes participate in carcinogenesis and progression of cancers, including metastasis and drug-resistance. Of note, CTCF has been suggested to induce drug resistance in various cancers. Herein, we aim to investigate the role of cisplatin- (CDDP-) resistant osteosarcoma- (OS-) derived exosomal CTCF in OS cell resistance to CDDP and its mechanistic basis. Differentially expressed transcription factors, long noncoding RNAs (lncRNAs), miRNAs, and genes in OS were retrieved using bioinformatics approaches. Exosomes were extracted from CDDP-resistant OS cells and then cocultured with parental OS cells, followed by lentiviral transduction to manipulate the expression of CTCF, IGF2-AS, miR-579-3p, and MSH6. We assessed the in vitro and in vivo effects on malignant phenotypes, autophagy, CDDP sensitivity, and tumor formation of OS cells. It was established that CTCF and IGF2-AS were highly expressed in CDDP-resistant OS cells, and the CDDP-resistant OS cell-derived exosomal CTCF enhanced IGF2-AS transcription. CDDP-resistant OS-derived exosomes transmitted CTCF to OS cells and increased CDDP resistance in OS cells by activating an autophagy-dependent pathway. Mechanistically, CTCF activated IGF2-AS transcription and IGF2-AS competitively bound to miR-579-3p to upregulate MSH6 expression. Additionally, the promoting function of exosomal CTCF-mediated IGF2-AS/miR-579-3p/MSH6 in OS cell resistance to CDDP was confirmed in vivo. Taken together, CDDP-resistant OS-derived exosomal CTCF enhanced resistance of OS cells to CDDP via activating the autophagy-dependent pathway, providing a potential therapeutic consideration for OS treatment.


Introduction
Osteosarcoma (OS) is the most common primary malignant bone tumor, mainly occurring in children and adolescents [1]. Tall stature, high birth weight, certain inherited cancer susceptibility syndromes, and common genetic variants are the main risk factors for OS [2]. Surgery combined with chemotherapy has greatly improved the prognosis of patients with OS, but the prognosis of recurrent or metastatic OS is a big challenge [3]. Cisplatin (CDDP) is regarded as a standard drug for the treatment of OS, but the resistance of OS to CDDP limits its effectiveness [4]. OS is a tumor believed to originate from bone-forming mesenchymal stem cells, and abnormal oncogene activation and tumor suppressor gene inactivation due to somatic mutations and epigenetic mechanisms play a key pathogenic role in OS [5]. In this regard, efforts are required to explore mutated epigenetic mechanisms involved in OS progression and OS resistance to CDDP for the development of OS treatment.
Exosomes are cell-derived extracellular vesicles which participate in intercellular communication and induce changes in cell behavior by transporting proteins, DNA, RNA, or lipids between cells and are concerned with tumorigenesis and drug resistance of OS [6]. CCCTC-binding factor (CTCF) is a highly conserved zinc finger protein that regulates high-order chromatin tissue and participates in various gene regulatory processes, which is often altered by hemizygous deletions or mutations in human cancers [7,8]. Although CTCF is essential for cell survival, inadequate haploid CTCF is related to tumor development and hypermethylation [9]. CTCF has been revealed as a critical transcriptional factor to mediate long noncoding RNAs (lncRNAs) to affect cancer progression [10,11]. Long noncoding RNA IGF2-AS is upregulated in gastric adenocarcinoma tissues and is associated with poor prognosis in gastric adenocarcinoma patients [12]. A previous report has pointed out the elevated IGF2 expression in OS after chemotherapy, which prompted us to examine its functional contribution to drug resistance [13]. It was generally known that lncRNAs modulate the expression of different micro-RNAs (miRNAs), and the binding of miRNAs to lncRNAs led to the increase in the expression of miRNA target genes [14]. Using bioinformatics software, we found that IGF2-AS could competitively bind to microRNA-579-3p (miR-579-3p) to upregulate MutS homolog 6 (MSH6) expression. miR-579-3p is strongly downregulated before and after resistance to targeted therapies in matched tumor samples from patients [15]. Further, it has been reported that miR-579-3p inhibits proliferation, invasion, and migration of squamous cell lung cancer cells [16]. MSH6 is one of the members of the MutS family which could heterodimerize with MSH2 to form a mismatch recognition complex to mediate DNA mismatch repair (MMR) in eukaryotes [17]. Silencing MSH6 combined with CDDP inhibited OS cell proliferation [18].
Given the aforementioned evidence, we proposed a hypothesis that exosomes delivering CTCF could regulate the IGF2-AS/miR-579-3p/MSH6 axis to enhance resistance of OS cells to CDDP. 2.2. In Silico Prediction. The hTFtarget database was used to predict the upstream transcription factor of IGF2-AS. The promoter sequence of IGF2-AS was obtained from the UCSC database. The JASPAR database was applied to predict the binding sites of transcription factors and downstream promoters. The Vesiclepedia database was adopted to retrieve the enrichment of transcription factors in exosomes.

Materials and Methods
The Gene Expression Omnibus (GEO) database was used to obtain the OS-related GSE41445 expression profile. There were 57 samples in the GSE41445 expression profile, which were not related to this study. After deletion of these 57 samples, 3 normal samples and 3 OS samples were left. The limma package in R language was used for differential analysis. Top 300 genes with logFC > 1 were considered significantly highly expressed genes.
The GeneCards database was used to identify genes related to OS. Related genes were retrieved through the STRING database, followed by construction of an interaction network. The Cytoscape database was employed for drawing and calculating the core level. KEGG enrichment analysis was conducted through the KOBAS 3.0 database. The MEM database was applied to perform large-scale analysis to determine the coexpression relationship among genes, and the lncATLAS database was used to predict the subcellular location of lncRNA to determine its downstream regulatory mechanism. The StarBase database was adopted to predict the relationship of lncRNA-miRNA-gene and obtain binding sites.
The effects of exosomes carrying CTCF (Exo-CTCF) on autophagy and CDDP resistance of CDDP-resistant OS cells via regulating IGF2-AS expression were examined. After transduction with lentivirus for 24 h, MG63 cells were treated with 10 μg Exo-CTCF and 1 nM rapamycin (HY-10219, MCE) for 48 h. Cells were treated with Exo-CTCF alone (Exo-CTCF) or in the presence of lentiviral sh-NC (Exo-CTCF+sh-NC), lentiviral sh-IGF2-AS (Exo-CTCF +sh-IGF2-AS), or/and rapamycin (Exo-CTCF+sh-IGF2-AS +rapamycin). Those without any treatment functioned as the control. All above-mentioned cells were treated with 8 μM CDDP (HY-17394, MCE) for 48 h and then collected for subsequent testing. 2.8. Isolation and Identification of Exosomes. Differential centrifugation was used to separate exosomes from the cell culture supernatant of MG63 and MG63/CDDP. The collected supernatant was centrifuged at 300 × g for 10 min, 2000 × g for 10 min, and 10000 × g for 30 min. Exosomes were resuspended with phosphate PBS and filtered with a 0.22 μm filter to remove smaller cell debris. After being washed with PBS, exosomes were resuspended by ultracentrifugation at 110000 × g for 2 h to remove PBS. The supernatant was subjected to ultracentrifugation at 110000 × g for 2 h. The obtained precipitates were exosomes. The collected exosomes were resuspended in PBS and filtered with 0.22 μm to remove smaller cell debris. Exosomes were resuspended with PBS and ultracentrifuged at 110000 × g for 2 h to remove the PBS. Then, the second ultracentrifugation under the same conditions was performed. The precipitate was collected and stored at -80°C for later use. The above centrifugation was performed at 4°C. Exosomes were cultured at cell culture medium containing exosome-free serum (C38010050, Viva Cell, Shanghai, China). Transmission electron microscopy (H-7650, HITACHI, Tokyo, Japan) was applied to identify exosomes. Total 10 μL of exosomes was dropped on the parafilm, and the copper mesh Formvar film was placed face down on the suspension. Each exosome sample was prepared with 2-3 copper meshes. Then, 100 μL PBS was added to the parafilm, and the copper mesh (Formvar membrane facing down) was placed on a PBS droplet with tweezers for 5 min. The copper mesh was treated with 50 μL of 1% glutaraldehyde droplets for 5 min and treated in 100 μL of double distilled water for 2 min. The copper mesh was treated with a 50 μL drop of uranyl oxalate for 5 min. The copper mesh was then placed on a 50 μL drop of methylcellulose-UA for 10 min. The copper mesh was removed with a stainless steel ring, followed by gently absorbing the excess liquid on the filter paper, leaving a thin layer of methyl cellulose membrane. The dried copper mesh was placed under 100 keV to take electron micrographs.
Exosome size was evaluated by nanoparticle tracking analysis (NTA), which could automatically track and determine particle size based on the diffusion coefficient of the Brownian motion. Exosomes were resuspended and mixed in 1 mL PBS, and the filtered PBS was used as a control. Then, the diluted exosomes were injected into the NanoSight LM10 instrument to measure the particle size.
2.9. Uptake of Exosomes. An equal volume of 2 μM PHK67 dye (MINI67, Sigma-Aldrich, St. Louis, MO, USA) was added to the extracted exosomes. After 5 min of culture at room temperature, 10 mL of exosome medium was added into exosomes to stop staining. Exosomes were centrifuged at 400 × g for 10 min, and the supernatant was discarded. Then, 10 mL of exosomes was added and centrifuged again at 400 × g for 5 min. The supernatant was discarded, and the precipitate was obtained after which was PHK67labeled exosomes. A special cell slide was placed on the top of the petri dish, and MG63 cells were cultured in the petri dish. When cell density reached 50%, cells were added to exosomes labeled with PHK67 and coincubated at room temperature for 24 h. The slide was taken out and washed slowly with PBS 3 times. Cells were soaked in 4% paraformaldehyde at room temperature for 30 min, permeabilized with 2% Triton X-100 for 15 min, and blocked with 2% BSA for 45 min. The blocking solution was discarded. After staining with DAPI (2 μg/mL, C1005, Beyotime, Shanghai, China), the slides were mounted. Fluorescence intensity was detected by using an upright fluorescence microscope.

Chromatin Immunoprecipitation (ChIP).
To detect the enrichment of CTCF in the promoter region of IGF2-AS, the EZ-Magna ChIP TMA kit (Merck Millipore, Billerica, MA) was used for ChIP experiments. The OS cells in the logarithmic growth phase were cross-linked and cultured with 1% formaldehyde for 10 min. The cells were treated with the protease inhibitor and then sonicated until 200-1000 bp chromatin fragments were obtained. In each group, 100 μL of supernatant (DNA fragment) was added to 900 μL ChIP dilution buffer, 20 μL of 50 × PIC, and 60 μL Protein A Agarose/Salmon Sperm DN to invert and mix at 4°C for 1 h. After centrifugation, the supernatant was taken out, and 20 μL was used as input. In the experimental group, the supernatant was mixed with 2 μg of CTCF antibody (ab128873, rabbit, Abcam Inc.). In the NC group, 2 μg of anti-IgG (ab172730, rabbit, Abcam Inc.) was added, and 60 μL of Protein A Agarose/Salmon was added to each tube. Sperm DNA was inverted at 4°C for 2 h. After centrifugation, the precipitate was washed with 1 mL of low-salt buffer, high-salt buffer, LiCl solution, and TE, respectively. Each tube was eluted twice with 250 mL ChIP Wash Buffer. Then, 20 mL of 5 M NaCl was used to reverse cross-link. The DNA was recovered, and the promoter sequence of IGF2-AS in the complex (F: 5 ′ -AGAATTCAGGGGCCCCATCC-3 ′ , R: 5 ′ -CTGCATCTGCACTCAGACGG-3 ′ ) was conducted with quantification.
2.11. Dual-Luciferase Reporter Gene Assay. The IGF2-AS sequence containing the CTCF binding site and mutation binding site was cloned into a pGL3-Promoter luciferase reporter vector (E1751, Promega, Madison, WI, USA) to construct IGF2-AS-wild type (WT) and IGF2-AS-mutant type (MUT) reporter vectors. HEK-293T cells were cotransfected with oe-NC or oe-CTCF and the above-mentioned reporter vectors using the Lipofectamine 2000 kit (11668019, Invitrogen, Carlsbad, CA). After 24 h of transfection, the supernatant of HEK-293T cells was collected. The luciferase activity was measured using the dual-luciferase reporter analysis system (E1910, Promega, Madison, WI, USA). The luciferase activity was directly measured by the ratio of firefly luciferase to Renilla luciferase. The same method was used to detect the binding ability of IGF2-AS with miR-579-3p and miR-579-3p with MSH6.
2.12. RNA Immunoprecipitation (RIP). The RIP assay was performed according to the instructions of the EZ-Magna RIP kit (17-701, Millipore, Billerica, MA, USA). Cells were lysed with RIP lysis buffer. After resuspending magnetic beads in RIP Wash Buffer, cells were probed with the AGO2 antibody (ab186733, 1 : 30, Abcam Inc.) or antirabbit IgG (ab6721, 1 : 30, Abcam Inc.) for 30 min. Cells were incubated with the lysed cell supernatant at room temperature overnight at 4°C. After a short centrifugation, the supernatant was discarded. Cells were added with RIP Wash Buffer and washed 6 times. The RNA in the magnetic bead-antibody complex was extracted. The expression of IGF2-AS and miR-579-3p in the complex was analyzed by RT-qPCR. All experimental steps were performed according to the instruction manual.

Fluorescence In Situ Hybridization (FISH).
RiboTM lncRNA FISH Probe Mix (Red) (C10920, Ruibo Bio, China) was used to detect the subcellular colocalization of IGF2-AS and miR-579-3p. Cells were seeded in a 24-well plate. When confluence reached 60-70%, cells were fixed in 4% formaldehyde for 10 min. The membrane was ruptured with 0.5% TritonX-100 for 5 min. Cells in each well were added with 200 μL of prehybridization solution and incubated at 37°C for 30 min, followed by incubating with Cy3-labeled IGF2-AS and FITC-labeled miR-579-3p cell probes, overnight at 37°C. The cells were stained with DAPI (C1005, Beyotime Biotechnology, Shanghai, China). Finally, the cells were observed and photographed in 5 different fields of view under a fluorescence microscope (Olympus, Tokyo, Japan). The nude mice were randomly inoculated with lentiviral sh-NC-transduced MG63 cells or lentiviral sh-MSH6transduced MG63 cells under the armpit and injected with 100 μL PBS or 10 μg MG63/CDDP cell-derived exosomes via the tail vein (n = 6). All above mice were injected with CDDP. Exosomes were configured in sterile PBS and injected into the tail vein every 3 days (10 μg exosomes in 100 μL PBS was injected each time).

Transmission Electron
When the tumor grew to 50 mm 3 , nude mice were injected intraperitoneally with CDDP (3 mg/kg) twice a week. Tumor volume was measured with a Vernier caliper every 1 week. The tumor volume calculation formula was W = 1/2 * a * b 2 (a represented the long diameter, b represented the short diameter). After 4 weeks, the nude mice were sacrificed to remove the xenograft tumor tissues.
2.17. Statistical Analysis. Data statistical analyses were processed using SPSS 21.0 statistical software (IBM Corp. Armonk, NY). Measurement data were expressed as mean ± standard deviation. Data between two groups were compared using the paired t-test or unpaired t-test. Comparisons among multiple groups were conducted by one-way analysis of variance (ANOVA), followed by Tukey's post hoc test. Two-way analysis of variance was applied for cell viability comparison among multiple groups, and repeated measures analysis of variance for tumor data comparison. A value of p < 0:05 indicated significant difference.  Figure 1), which indicated that the CDDPresistant OS cells were successfully constructed. We found that IGF2-AS was significantly highly expressed in CDDPresistant OS cells through RT-qPCR detection results (Figure 1(a)).

Exosomes Derived from OS Cells
According to the prediction of the hTFtarget database, 70 upstream transcription factors of IGF2-AS were obtained. Among them, CTCF targeted IGF2-AS in 24 types of tissues (Supplementary Table 2), which was the transcription factor targeting the most types of tissues. This result indicated that CTCF might be the key upstream gene of IGF2-AS. Subsequently, an analysis through the Vesiclepedia database found that CTCF was enriched in tumor-derived exosomes. Therefore, we speculated that exosomes from tumor cells delivered CTCF to activate IGF2-AS.
Exosomes were isolated from the supernatant of MG63 and MG63/CDDP cell culture medium, and the morphology of exosomes was observed by TEM (Supplementary Figure 2A). The isolated exosomes had a typical exosome-like morphology, with basically uniform round or oval membranous vesicles. NTA showed that the diameter of exosomes ranged from 30 nm to 150 nm (Supplementary Figure 2B). Western blot results showed that the surface of exosomes significantly expressed the surface markers CD63 and CD9 of exosomes but did not express calnexin (Supplementary Figure 2C). The above results confirmed that extracted exosomes were successfully derived from OS cells.
In addition, the results of RT-qPCR and Western blot analysis showed that the expression of CTCF in CDDPresistant OS cells was higher than that in CDDP-sensitive OS cells (Figure 1(b)). In addition, the expression of CTCF in MG63/CDDP-Exo was higher than that in MG63-Exo (Figure 1(c)). After cocultivating PKH67-(green) labeled MG63/CDDP-Exo with MG63 cells for 24 h, observation under a fluorescence microscope revealed that MG63 cells presented obvious PKH67 staining, which confirmed that MG63 cells had successfully taken up exosomes (Figure 1(d)). RT-qPCR results showed that the expression of CTCF and IGF2-AS after coculture with MG63/CDDP-Exo was higher than that in the control (Figure 1(e)).
Moreover, oe-CTCF was transduced in MG63 cells, and the RT-qPCR results showed that the expression of CTCF and IGF2-AS upon the oe-CTCF treatment was elevated (Figure 1(f)). ChIP experiment found that CTCF could directly bind to the promoter region of IGF2-AS (Figure 1(g)). The dual-luciferase reporter gene assay showed that the cell luminescence intensity was increased after overexpression of CTCF, but the cell luminescence intensity did not change significantly after the site mutation (Figure 1(h)).
The above results indicated that CTCF and IGF2-AS were highly expressed in CDDP-resistant OS cells, and the CDDP-resistant OS cell-derived exosomes carrying CTCF could promote the transcriptional expression of IGF2-AS in OS cells.

OS-Derived Exosomal CTCF Activates the Autophagy
Signaling Pathway via Upregulating IGF2-AS to Enhance the CDDP Resistance of OS Cells. We further moved to test the influence of exosomal CTCF on CDDP resistance and the involvement of IGF2-AS. RT-qPCR was applied to detect the silencing efficiency of lentiviral sh-IGF2-AS in MG63 cells. It was noted that lentiviral sh-IGF2-AS-1 and sh-IGF2-AS-2 could knock down the expression of IGF2-AS in MG63 cells, among which lentiviral sh-IGF2-AS-2 had better silencing effect and was used in subsequent experiments (Figure 2(a)).
Lentiviral sh-IGF2-AS was transduced in parental cells after coculture with exosomes derived from CDDPresistant OS cells. We found that the expression of CTCF  7 Journal of Oncology and IGF2-AS was elevated in the parental cells after coculture with Exo-CTCF. The upregulated expression of IGF2-AS caused by Exo-CTCF was reversed by additional sh-IGF2-AS treatment, but there was no significant difference in the expression of CTCF (Figure 2(b)).
Subsequently, MG63 cells were treated with different concentrations of CDDP. CCK-8 and flow cytometry results showed that the cell viability and IC50 were increased upon the coculture with Exo-CTCF, while cell apoptosis was diminished. Compared with the coculture with Exo-CTCF alone, the cell viability and IC50 value were diminished in cells with Exo-CTCF and sh-IGF2-AS treatment, and cell apoptosis was increased (Figures 2(c) and 2(d)). The above results indicated that exosomes derived from CDDPresistant OS cells delivered CTCF to increase the expression of IGF2-AS, thereby enhancing the CDDP resistance of OS cells.
Ultrastructural analysis by TEM showed that, after 48 h of intervention with 8 μM CDDP, the number of autophagosomes in MG63/CDDP cells was higher than that in MG63 cells (Figure 2(e)). Western blot analysis showed that compared with untreated MG63 cells, the expression of p62 in MG63/CDDP cells was reduced, and the expression of Beclin1 and the ratio of LC3-II/LC3-I were increased. This result indicated that the autophagy was increased in CDDP-resistant OS cells (Figure 2(f)).
MG63 cells were subjected to lentiviral transduction and autophagy agonist rapamycin (1 nM, HY-10219, MCE) intervention to further study the interaction among Exo-CTCF mediated-IGF2-AS, autophagy, and CDDP resistance in OS cells. TEM ultrastructural analysis showed that after 48 h of CDDP intervention, autophagosomes were increased in MG63 cells treated with Exo-CTCF. After downregulating the expression of IGF2-AS, the regulatory effect of Exo-CTCF on autophagosomes in cells was inhibited. The num-ber of autophagosomes in cells was further increased after rapamycin treatment (Figure 2(g)).
In addition, the expression of p62 in cells cocultured with Exo-CTCF was diminished, and the expression of Beclin1 and the ratio of LC3-II/LC3-I were increased. In comparison with the Exo-CTCF alone, Exo-CTCF and sh-IGF2-AS treatment led to increased expression of p62 and reduced expression of Beclin1 and the ratio of LC3-II/LC3-I. The effects of Exo-CTCF and sh-IGF2-AS treatment were abolished by additional rapamycin treatment (Figure 2(h)). The CCK-8 and flow cytometry showed that cell viability and IC50 value were increased and cell apoptosis was diminished after rapamycin intervention (Figures 2(i) and 2(j)).
The above results indicated that exosomes from CDDPresistant OS cells transmit CTCF to upregulate the expression of IGF2-AS and activate the autophagy signaling pathway, thereby enhancing the CDDP resistance of OS cells.

IGF2-AS May Competitively Bind to miR-579-3p and
Upregulates the Expression of MSH6. The focus of this study shifted to find the downstream genes of IGF2-AS. The top 300 upregulated genes were retrieved through differential analysis on microarray GSE41445 (Figure 3(a)). The top 400 genes related to OS were obtained from the GeneCards database. Totally, 15 important and significantly upregulated genes were obtained by drawing the Venn diagram based on the intersection of the list of upregulated and OS-related genes (Figure 3(b)). Among them, only TIMP1 [19], MSH6 [20], and FGF2 [21] have been reported in the previous literature to be related to the CDDP resistance of OS cells. The interaction genes of the three genes were predicted using the STRING database, followed by constructing an interaction network. It was found that the core degree of MSH6 was 5, while the core degrees of TIMP1 and FGF2 were both 2 (Figure 3(c)). Therefore,   9 Journal of Oncology we speculated that MSH6 was one of the most important genes related to OS chemoresistance.
The box plot confirmed that MSH6 was highly expressed in OS (Figure 3(d)). The STRING database was applied again to predict the 10 interaction genes of MSH6 (Figure 3(e)). KOBAS 3.0 was used to perform KEGG enrichment analysis on MSH6 and the interaction genes. It was found that the top 3 KEGG pathways with significant MSH6 enrichment were mismatch repair, CDDP drug resistance, and colorectal cancer (Figure 3(f)). The coexpression prediction of the MEM database found that IGF2-AS and MSH6 were significantly coexpressed (Figure 3(g)). Most of the samples in the significantly coexpressed probes were shown in red, which indicated that their expression was mainly positively correlated.
The lncATLAS database predicted that the subcellular localization of IGF2-AS was in the cytoplasm (Figure 3(h)). IGF2-AS may target MSH6 through miRNAs. Using the StarBase database, it was confirmed that there were 16 and 93 targeted binding miRNAs for IGF2-AS and MSH6, respectively. The intersection of four important miRNAs was miR-4731-5p and miR-579-3p, miR-664b-3p, and miR-3126-5p (Figure 3(i)). Studies have found that the expression of miR-579 was low in OS. Increasing expression of miR-579 could inhibit the development of OS [22]. The StarBase database obtained the binding sites between IGF2-AS and miR-579-3p and between miR-579-3p and MSH6 (Figure 3(j)).
Based on the above results, we speculated that IGF2-AS may competitively bind to miR-579-3p to upregulate the expression of MSH6, thereby enhancing the CDDP resistance of OS cells.

IGF2-AS Competitively Binds to miR-579-3p to
Upregulate the Expression of MSH6. In vitro cell experiments were used to verify the regulatory interaction among IGF2-AS, miR-579-3p, and MSH6. The data of RT-qPCR and Western blot analysis showed that the expression of miR-579-3p was diminished in CDDP-resistant OS cells, and the expression of MSH6 was increased (Figure 4(a)). The     Journal of Oncology dual-luciferase reporter gene assay showed that the miR-579-3p mimic could significantly reduce the luciferase activity of the IGF2-AS-WT plasmid but did not affect the luciferase activity of the IGF2-AS-MUT plasmid, suggesting that IGF2-AS may directly interact with miR-579-3p (Figure 4(b)).
In addition, the FISH assay found that IGF2-AS and miR-579-3p were coexpressed in the cytoplasm (Figure 4(c)). The results of the RIP assay showed that the expressions of IGF2-AS and miR-579-3p in cells treated with AGO2 were higher than those in the cells treated with IgG. This result indicated that there was a direct binding relationship between IGF2-AS and miR-579-3p (Figure 4(d)). IGF2-AS was overexpressed or silenced in MG63 cells, and the RT-qPCR results confirmed that after overexpression of IGF2-AS, the expression of miR-579-3p was diminished, and after silencing the expression of IGF2-AS, the expression of miR-579-3p was increased (Figure 4(e)). The above results indicated that miR-579-3p could be sponged by IGF2-AS.
The dual-luciferase reporter gene assay was used to verify the targeting of miR-579-3p to MSH6. It was found that the miR-579-3p mimic could reduce the luciferase activity of the MSH6-WT plasmid but did not affect MSH6-MUT luciferase activity of the plasmid, which suggested that miR-579-3p may directly interact with MSH6 (Figure 4(f)). Further, overexpression of miR-579-3p reduced the expression of MSH6, and the depleted miR-579-3p elevated the expression of MSH6 (Figure 4(g)). The above results validated that miR-579-3p could target MSH6 and downregulate the expression of MSH6.
In vitro cell experiments were conducted to verify the interaction among IGF2-AS, miR-579-3p, and MSH6. Data of RT-qPCR and Western blot analysis showed that compared with cells treated with oe-NC and NC mimic, the expressions of IGF2-AS and MSH6 were increased, and the expression of miR-579-3p was reduced in cells treated with oe-IGF2-AS-WT and NC mimic. The expression of IGF2-AS was increased in cells with oe-IGF2-AS-MUT and NC mimic treatment, and there was no significant difference in the expression of miR-579-3p and MSH6. In contrast to cells treated with oe-IGF2-AS-WT and NC mimic, the expressions of IGF2-AS and miR-579-3p in cells treated with oe-IGF2-AS-WT and miR-579-3p mimic were increased, while the expression of MSH6 was diminished (Figure 4(h)).
The above results indicated that IGF2-AS could competitively bind to miR-579-3p to upregulate the expression of MSH6 in OS cells.
Ultrastructural analysis of TEM showed that, after CDDP treatment for 48 h, overexpression of IGF2-AS elevated autophagosomes in cells, but it was abolished by silencing of MSH6 ( Figure 5(c)). Western blot analysis results manifested that oe-IGF2-AS treatment could decrease the expression of p62 and increase the expression of Beclin1 and the ratio of LC3-II/LC3-I, which were revoked by additional sh-MSH6 treatment ( Figure 5(d)). The CCK-8 and flow cytometry showed that cell viability and IC50 value were increased after overexpression of IGF2-AS, and cell apoptosis was diminished, but these effects caused by oe-IGF2-AS could be reversed by sh-MSH6 (Figures 5(e) and 5(f)).
The above results indicated that the IGF2-AS/miR-579-3p/MSH6 axis could activate the autophagy signaling pathway, thereby increasing the resistance of OS cells to CDDP.

Exosomal CTCF Activates the Autophagy Signaling
Pathway by Regulating the IGF2-AS/miR-579-3p/MSH6 Axis and Increases the Resistance of OS Cells to CDDP. Furthermore, exosomes were extracted from MG63/CDDP cells. MG63 cells were cocultured with extracted exosomes and then transduced with lentiviral sh-MSH6. RT-qPCR and Western blot analysis presented that coculture with Exo-CTCF could elevate expression of CTCF, IGF2-AS, and MSH6 and reduce the expression of miR-579-3p, but the effect on MSH6 expression was reversed by sh-MSH6 treatment, accompanied by no significant change on expressions of CTCF, IGF2-AS, and miR-579-3p ( Figure 6(a)).
In addition, the ultrastructural analysis of TEM exhibited that Exo-CTCF increased a number of autophagosomes in MG63 cells after CDDP intervention for 48 h, while silencing MSH6 expression reversed the effect of Exo-CTCF caused on the number of autophagosomes in MG63 cells (Figure 6(b)). Western blot analysis results confirmed that Exo-CTCF could diminish p62 level in and elevate expressions of Beclin1 and the ratio of LC3-II/LC3-I, which could be revoked by silencing of MSH6 (Figure 6(c)). The CCK-8 and flow cytometry showed that cell viability and IC50 value were increased and apoptosis was diminished after coculture with Exo-CTCF, but sh-MSH6 could reverse this effect caused by Exo-CTCF (Figures 6(d) and 6(e)).
The above results demonstrated that exosomes from CDDP-resistant OS cells delivered CTCF into OS cells, which activated the autophagy signaling pathway via

CDDP-Resistant OS Cell-Derived Exosomal CTCF Increases Tumor Formation and CDDP Resistance in Nude
Mice Bearing Xenografts of OS Cells. Next, we investigated the effect of CDDP-resistant OS-Exo-CTCF on the IGF2-AS/miR-579-3p/MSH6 axis by subcutaneous xenograft tumors in nude mice. It was noted that compared with mice with PBS treatment, the weight and volume of tumors in response to CDDP treatment were reduced. Exo-CTCF treatment could promote weight and volume of the tumor, but its effect could be abolished by sh-MSH6 treatment (Figures 7(a)-7(c)).
In addition, the results of RT-qPCR and Western blot analysis presented that the weight and volume of tumors from mice treated with those with CDDP alone were increased in comparison with those with PBS treatment alone. The increased expressions of CTCF, IGF2-AS, and MSH6 and diminished expression of miR-579-3p were caused by Exo-CTCF, and it was confirmed that sh-MSH6 could revoke the impact on MSH6, but it causes no significant effects on expressions of CTCF, IGF2-AS, and miR-579-3p (Figure 7(d)).
The above results validated that CTCF delivered by exosomes derived from CDDP-resistant OS cells could increase the tumorigenesis and CDDP resistance of OS cells in nude mice.

Discussion
OS is one of the most common primary bone malignancies, and the occurrence of chemotherapy resistance is the main reason for the high incidence of OS patients [23]. CDDP is one of the most widely used agents in the treatment of OS, but the drug resistance of OS to CDDP is a serious problem [24]. Herein, novel therapies tailored with the aid of molecular biomarkers are urgently needed to prevent resistance of OS cells to CDDP. In the present study, we elucidated the regulatory role of CTCF encapsulated by CDDP-resistant OS cell-derived exosomes in CDDP resistance of OS cells via regulation of the IGF2-AS/miR-579-3p/MSH6 axis. In this study, the OS cell line MG63 was continuously exposed to CDDP to establish CDDP-resistant cell lines (MG63/CDDP). Our findings unraveled higher expression of CTCF in MG63/CDDP cells than in CDDP-sensitive cell lines. Findings obtained from a study confirmed that CTCF is upregulated in colorectal cancer cells and elevated CTCF is able to promote drug resistance in colorectal cancer [25]. Sun et al. also pointed out that knockdown of CTCF can suppress cell progression and chemoresistance in head and neck squamous cell carcinoma [26]. EGR1 and CTCF were found to be involved in the progression of OS cells by interacting with chemokines and their receptors [27]. Similar to these results, our results validated that overexpressed CTCF conferred enhanced CDDP resistance in OS cells. We then validated that CTCF was also overexpressed in exosomes derived from MG63/CDDP cells. It has been reported that CDDP-resistant OS cell-derived exosomes contribute to enhancing chemoresistance of OS cells to CDDP, which could function as a promising biomarker and therapeutic strategy against OS chemoresistance by carrying therapeutic components [28]. As a consequence, it could be concluded that OS cell-derived exosomes transmitted CTCF into OS cells to enhance resistance of OS cells to CDDP. CTCF has been reported to activate downstream gene expression in cancer [29]. Our bioinformatics analysis and tests substantiated that CTCF could bind to the promoter of IGF2-AS as a transcription factor. IGF2-AS has been confirmed to be upregulated and exhibit tumor-promoting properties in various cancers such as hepatocellular carcinoma [30] and breast cancer [31]. Our finding manifested that IGF2-AS was also highly expressed in MG63/CDDP cells. Further evidence revealed that CDDP-resistant cell-derived exosomal CTCF upregulated IGF2-AS to accelerate the autophagydependent pathway so as to confer elevated CDDP resistance in OS cells, as companied by elevated autophagosomes and increased Beclin1 expression and the ratio of LC3-II/ LC3-I in MG63/CDDP cells. LC3 and Beclin1 are wellknown autophagy-related genes and increased the LC3-II/ LC3-I ratio, and Beclin1 expression is the hallmark of autophagy [32]. Autophagy is a self-degrading system that is common in the treatment of multidrug resistance in cancers and enables to prevent cancer cells from chemotherapeutic drugs [33]. It has been well documented that autophagy exerts key roles in increasing CDDP resistance in OS cells [34]. By contrary, inhibited autophagy is capable of boosting sensitivity of OS cells to CDDP, along with the reduced Beclin1 level and LC3-II/LC3-I ratio [35,36]. In recent years, studies have found that lncRNA-mediated dysregulation of targets and pathways is considered to be the key cause of chemotherapy resistance [37,38]. lncRNAs acting as competitive endogenous RNAs (ceRNAs) have been confirmed through sponging miRNAs to change the expression of their target genes in tumorigenesis [39]. Overexpressed IGF2-AS is confirmed in gastric adenocarcinoma tissues which exhibits a great functional role in promoting cancer cell progression by sponging miR-503 as a ceRNA to regulate SHOX2 expression [12]. Additionally, silencing of IGF2-AS downregulates CREB1 expression through sponging miR-195 to retard tumorigenesis of gastric cancer [40]. Consistent with these findings, the present study validated that IGF2-AS could upregulate MSH6 expression through sponging miR-579-3p to boost resistance of OS cells to CDDP. There is a growing body of evidence supporting the effective role of miR-579-3p played in overcoming drug resistance in cancer treatment [15,41]. More importantly, miR-579 is underexpressed in OS, and the improvement of miR-579 expression significantly inhibits the development of OS [22]. Moreover, it has been established that MSH6 overexpression promotes the progression of OS [20]. The combined effect of MSH6 gene silencing and CDDP can effectively inhibit the proliferation of OS cells and promote apoptosis [18]. Furthermore, the promoting function of exosomal CTCF-mediated IGF2-AS through the miR-579-3p/ MSH6 axis in OS cell resistance to CDDP was confirmed in tumor-bearing nude mice.
Taken together, the data acquired in this study led to a conclusion that CDDP-resistant OS cell-derived exosomal CTCF activated the autophagy signaling pathway through the IGF2-AS/miR-579-3p/MSH6 axis to enhance OS cell resistance to CDDP (Figure 8). Our findings deepen our understanding of the resistance of OS cells to CDDP and provide a theoretical basis for development of novel targeted therapies to overcome chemotherapy resistance in OS.