Function and Potential Molecular Mechanism of High STIL Expression in Osteosarcoma


 BackgroundSCL/TAL1 interrupting locus (STIL) is associated with the progression of several tumors; however, the biological role of STIL in osteosarcoma remains poorly understood.MethodsIn this study, the clinical significance of STIL in osteosarcoma was analyzed by gene chip data recorded in public databases. STIL expression was silenced in osteosarcoma cell lines to observe the effects on proliferation, apoptosis, invasion, and migration. Differentially expressed genes (DEGs) in the osteosarcoma chip were analyzed using The Limma package, and STIL co-expressed genes were obtained via the Pearson correlation coefficient. The potential molecular mechanism of STIL in osteosarcoma was further explored by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.ResultsMeta-analysis revealed that STIL was highly expressed in osteosarcoma, and the standardized mean difference (SMD) was 1.52. STIL also had a good ability to distinguish osteosarcoma from non-osteosarcoma samples (area under the curve [AUC] = 0.96). After silencing STIL, osteosarcoma cell proliferation decreased, apoptosis increased, and the migratory and invasion ability decreased. A total of 294 STIL differentially co-expressed genes were screened, and a bioinformatics analysis found that differentially co-expressed genes were primarily enriched in the cell signaling pathways. The protein-protein interaction (PPI) network indicated that the hub differentially co-expressed genes of STIL were CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB.Conclusions﻿STIL is associated with osteosarcoma proliferation and invasion, and may represent a novel target for its diagnosis and treatment.


Abstract
Background SCL/TAL1 interrupting locus (STIL) is associated with the progression of several tumors; however, the biological role of STIL in osteosarcoma remains poorly understood.

Methods
In this study, the clinical signi cance of STIL in osteosarcoma was analyzed by gene chip data recorded in public databases. STIL expression was silenced in osteosarcoma cell lines to observe the effects on proliferation, apoptosis, invasion, and migration. Differentially expressed genes (DEGs) in the osteosarcoma chip were analyzed using The Limma package, and STIL co-expressed genes were obtained via the Pearson correlation coe cient. The potential molecular mechanism of STIL in osteosarcoma was further explored by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

Results
Meta-analysis revealed that STIL was highly expressed in osteosarcoma, and the standardized mean difference (SMD) was 1.52. STIL also had a good ability to distinguish osteosarcoma from non-osteosarcoma samples (area under the curve [AUC] = 0.96). After silencing STIL, osteosarcoma cell proliferation decreased, apoptosis increased, and the migratory and invasion ability decreased. A total of 294 STIL differentially co-expressed genes were screened, and a bioinformatics analysis found that differentially co-expressed genes were primarily enriched in the cell signaling pathways. The protein-protein interaction (PPI) network indicated that the hub differentially co-expressed genes of STIL were CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB.
Conclusions STIL is associated with osteosarcoma proliferation and invasion, and may represent a novel target for its diagnosis and treatment.
Background Among children and adolescents, osteosarcoma represents one of the most common type of malignant bone tumor, with an incidence of approximately 3 in 1 million people each year [1]. Osteosarcoma originates from the mesenchymal tissue, which is mainly located in the metaphysis of the long bones and is the most common around the knee joint [2]. Since osteosarcoma has the characteristics of rapid cell proliferation, strong invasive ability, and early metastasis [3,4], the patient ve-year survival rate with non-metastatic and metastatic osteosarcoma is still only 60% and 30%, respectively [5][6][7], despite advances in treatment techniques (e.g., surgery and chemotherapy).
Therefore, there is an urgent need to gain in-depth insight into the molecular mechanisms that result in the development of osteosarcoma and identify novel molecular therapeutic targets.
STIL is an important factor in cellular mitosis and centriole replication. In the cell cycle, STIL expression increases gradually during the early stages, peak in the middle stage, and rapidly decreases in the later stage [8]. Throughout this process, STIL interacts with a variety of proteins to ensure mitotic stability [9][10][11][12]. Thus, changes in STIL expression are associated with chromosome instability and even cancer. Some studies have demonstrated that STIL is overexpressed in a variety of cancers. Erez et al. reported that the high STIL expression in lung cancer tissue is both a marker of cell proliferation, as well as a marker of the metastatic potential of lung cancer [13]. Moreover, the study by Ouyang et al. found that high STIL expression in nasopharyngeal carcinoma may down-regulate the expression of ITGA2, Smad2, and JAK1, as well as promote the proliferation and invasion of nasopharyngeal carcinoma cells [14]. Despite these ndings, the clinical signi cance and biological function of STIL in osteosarcoma remains unknown. Therefore, we studied the biological function of STIL in osteosarcoma and investigated its underlying molecular mechanism.
In this study, the clinical characteristics of STIL in osteosarcoma were analyzed by integrating the gene chip data in public databases. The biological function of STIL was studied in a series of in vitro experiments. In addition, we obtained the differentially co-expressed genes related to STIL, and the potential molecular mechanism was analyzed using bioinformatics methods.

Source of public databases
The gene microarray data were acquired from the ArrayExpress and Gene Expression Omnibus (GEO) databases, respectively. The following keywords were used to search the databases: "osteosarcoma" OR "osteosarcomas" AND "Homo sapiens". The inclusion criteria of the data were as follows: 1) the experimental group and control group consisted of the osteosarcoma group and non-osteosarcoma group; 2) the patient did not receive radiotherapy, chemotherapy, or additional adjuvant therapy prior to the patient's tissue was extracted; and 3) the number of samples in the control group and experimental group is ≥ 3. In addition, GSE21257 contains clinical data that can be used to analyze the clinical signi cance of STIL in osteosarcoma.

Western blot (WB) analysis
The cell lysates were prepared using RIPA cleavage buffer and phenylmethanesulfonyl uoride uoride (Beyotime Biotechnology, China), to extract the protein from the cells after full cleavage. The protein was fractionated by SDS-polyacrylamide gel electrophoresis, electroblotted on 0.45 µm polyvinylidene di uoride membranes (Millipore, USA), blocked with 5% skimmed milk powder at 37 °C for 1 h, washed three times with 1 × PBS-T (1000 mL 1×PBS + 0.5 mL Tween-20), and incubated with the primary antibody overnight at 4 °C. After three washes, a secondary antibody was added to the membrane and incubated for 2 h, and the membrane was washed ve times. Finally, an ECL kit (Beyotime Biotechnology, China) was used for detection. Anti-STIL and anti-GAPDH antibodies were purchased from Proteintech. A goat anti-mouse IgG (H + L) -HRP secondary antibody was purchased from Jackson ImmunoResearch.

Proliferation assay
A Cell Counting Assay Kit-8 (CCK-8, Dojindo Laboratories, Japan) was used to analyze the transfected U-2 OS and HOS cells. Each type of cell was transferred into 96-well plates at 5000 cells/well, and CCK8 was added to each well of the plate and incubated for 24 h, 48 h, or 72 h, respectively. The optical density was measured using an enzyme labeling instrument at a wavelength of 450 nm.

Transwell migration and invasion assays
The invasion and migratory ability of the cells was evaluated using the Transwell method. To determine the migratory ability, the transfected osteosarcoma cells were suspended in serum-free medium (cell density: 2×10 5 /mL) and added to a Transwell chamber (8.0 µµ; Becton, Dickinson and Company, USA). Next, 500 µΛ of complete medium was added to the lower chamber, and the plate was incubated for 48 h, xed in paraformaldehyde, and washed with a crystal violet staining solution at room temperature for 20 min. The unbound crystal violet was washed away and the number of cells that had migrated were counted. The invasion ability involved the Transwell chamber to be covered with Matrigel (Becton, Dickinson and Company, USA) followed by steps similar to that used to assess the cell migration ability.
Acquisition of STIL differentially co-expressed genes in osteosarcoma The DEGs in each gene chip was calculated using the Limma software package, with LogFC ≥ 1 and p < 0.05 as the standards. Genes that appeared in at least two studies were selected as DEGs. In addition, the correlation between STIL and other genes in each gene chip was assessed. Using Pearson's r ≥ 0.5 and p < 0.05 as the screening criteria, as well as genes that appeared in more than two studies were selected as the STIL co-expressed genes. The genes obtained from the intersection of DEGs and co-expressed genes were considered to be STIL differentially co-expressed genes in osteosarcoma.

Enrichment analysis of STIL differentially co-expressed genes and PPI analysis
The network tool, DAVID (https://david.ncifcrf.gov/), was used to perform an enrichment analysis of the GO and KEGG pathways for the obtained differentially co-expressed genes. In addition, the differentially co-expressed genes were used to construct a PPI network through the STRING (https://string-db.org/) database, and Cytoscape 3.8.0 was used to visualize the PPI network with a score of 0.9 or higher. Statistical analysis SPSS23.0 statistical software was used to perform an independent sample t-test, and the results were displayed as the mean ± standard deviation. SMD and summarized receiver operating characteristic (SROC) were calculated using STIL expression by STATA15.0. The Q test of the χ 2 test and I 2 were used to detect heterogeneity. When p > 0.05 and I 2 < 50%, there was considered to be no heterogeneity, and the use of a xed effects model was employed. Alternatively, a continuous variable meta-analysis was performed in conjunction with a random effects model. The overall results were subjected to a sensitivity analysis to determine the effects of each gene chip. Publication bias was evaluated using Begg's and Egger's tests. In addition, the Kaplan-Meier method was used for survival analysis, and the difference in the survival rate was calculated with a log-rank test. Statistical signi cance was considered at p < 0.05.

STIL expression in osteosarcoma
Based on the inclusion and exclusion criteria, we screened 13 osteosarcoma chips ( Table 1). The STIL expression data were extracted from the 13 gene chips. As shown in Fig. 1, STIL was highly expressed in the E-MEXP-3628, GSE14359, GSE126209, GSE33383, GSE36001, GSE39262, GSE42352, and GSE68591 chips. There was no difference in expression in the GSE11414, GSE12865, GSE87624, GSE99671, and GSE19276 chips. In addition, in the meta-analysis of 13 chips, p < 0.05, and an I 2 value of 66.9% were considered to be heterogeneous. Therefore, the random effects model was used to conduct a variable continuous meta-analysis. STIL expression in 376 osteosarcoma samples was higher than that in 92 normal samples, and the total SMD integrated by the random effects model was 1.52 (95% con dence interval [CI]: 0.98-2.05) (Fig. 2a). The sensitivity analysis suggested that the exclusion of any chip data had little effect on the total effect (Fig. 2b). The results of the Begg's test showed that p = 0.180 had no publication bias, while the Egger's test showed an absence of publication bias (Fig. 2c).

Effect of STIL on osteosarcoma cell proliferation and apoptosis
Of the osteoblast cell lines and four osteosarcoma cell lines, it was found that the up-regulation of STIL mRNA was the most signi cant in the U-2 OS and HOS cell lines (Fig. 4a). To study the effect of STIL on osteosarcoma cells, we inhibited STIL expression in the two cell lines with the highest expression levels (U-2 OS and HOS). The WB results showed that the knockdown effect at the siRNA-2 site was the most signi cant (Fig. 4b). The CCK8 results revealed that a knockdown of STIL could signi cantly inhibit the proliferation of osteosarcoma cells (Fig. 4c). In addition, the apoptosis experiments revealed that compared with the control group, there was an increase in the level of apoptotic osteosarcoma cells after knocking down STIL, which was statistically signi cant in U-2 OS cells, but not in HOS cells (Fig. 4d).

Effect of STIL on osteosarcoma cell migration and invasion
The Transwell assay results revealed that osteosarcoma cell migration was signi cantly inhibited after STIL was silenced (Fig. 5a). In addition, the invasive ability was tested and there was a signi cantly weakened invasive ability of STIL compared to that of the control group after STIL was silenced in the two cell lines (Fig. 5b).

Expression and clinical characteristics of STIL
The data of 53 osteosarcoma cases were extracted from the GSE21257 tissue microarray and their clinical features were analyzed (Table 2). A signi cant difference was observed in the level of STIL expression between the group aged < 16 years old and that aged ≥ 16 years old (p < 0.010). A signi cant difference was also observed between the metastasis and non-metastasis groups (p < 0.007), whereas there was no statistical signi cance based on gender. At the same time, the survival analysis in Fig. 6a shows that the prognosis of high STIL expression was lower than that of low STIL expression, and the ve-year survival rates were 0.44 and 0.727, respectively (p < 0.05). GO and KEGG analysis of STIL differentially co-expressed genes A total of 2425 DEGs and 1663 STIL co-expressed genes were screened from the public database, and there were 294 differentially co-expressed genes identi ed in the overlapping regions (Fig. 6b). The GO enrichment analysis of STIL differentially co-expressed genes (Fig. 7a -c) revealed that the biological process (BP) was primarily enriched in the areas of cellular and mitotic nuclear division, as well as sister chromatid cohesion.
The nucleoplasm, kinetochore, and spindle exhibited the primary enrichment of cellular component (CC). Protein binding, ATP binding, and single-stranded DNA-dependent ATPase activity were mainly enriched in molecular function (MF). The KEGG analysis indicated that the primary enrichment of the STIL co-expressed genes was in the areas of cell cycle, mismatch repair, and oocyte meiosis (Fig. 7d). Tables 3 and 4

PPI network analysis
The STIL differentially co-expressed genes were analyzed by PPI network. The rst six genes (CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB) were selected as hub differentially co-expressed genes by degree analysis in Cytoscape (Fig. 8). Therefore, these six hub differentially coexpressed genes were considered to be the key genes that interact with STIL.
Validation of hub differentially co-expressed genes based on 13 gene chips Using the 13 gene chips, we found that six hub differentially co-expressed genes (CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB) were highly expressed in osteosarcoma (Fig. 9). In addition, according to the SROS curve results, these genes also exhibit the ability to distinguish osteosarcoma from non-osteosarcoma samples (Fig. 10).

Discussion
In this study, 13 gene chips were collected and it was found that STIL expression was signi cantly increased in osteosarcoma. Highly expressed STIL was associated with the ability to distinguish osteosarcoma from non-osteosarcoma samples, and the patients with high STIL expression are associated with a poor prognosis. Additionally, the in vitro experiments revealed that the silencing of STIL could signi cantly block proliferation, induce apoptosis, and reduce migration and invasion in osteosarcoma cells. Therefore, STIL plays a role of proto-oncogene in osteosarcoma and has important signi cance in the occurrence and development of osteosarcoma. Together, these ndings provide new evidence that STIL may represent a new biomarker in osteosarcoma.
STIL was rst isolated from T cell chromosomes from T cell acute lymphoblastic leukemia [15]. In earlier studies, it was found that a deletion of the STIL locus in mice and zebra sh was embryonically lethal [16,17]. Moreover, STIL mutations in humans can lead to primary hereditary microcephaly and even cancer [18][19][20]. This is the rst study to report the overexpression of STIL in osteosarcoma, and demonstrate that silencing STIL inhibits cell proliferation, promotes cell apoptosis, and suppresses invasion and migration capabilities. Kasai et al. found that STIL was able to regulate the Hh signaling pathway through interacting with Sufu and Gli1 to affect cell proliferation [11]. Wu et al. found STIL to be highly expressed in prostate cancer and could regulate the growth of prostate cancer cells through the MAPK/ERK, PI3K/Akt, and AMPK signaling pathways [21]. Furthermore, Rabinowicz et al. reported that the targeted inhibition of highly expressed STIL could signi cantly improve the e cacy of DNA-damaging drugs for the treatment of ovarian cancer, and suggested that STIL might be a novel therapeutic target [22]. In recent studies, Wang et al. attenuated the IGF-1/PI3K/AKT pathway by knocking out STIL in gastric cancer, which inhibited cellular proliferation and reduced clone formation ability [23]. These results are consistent with our results and provide evidence for the role of STIL as a protooncogene.
To further clarify the potential molecular mechanism of STIL in osteosarcoma, we performed a GO and KEGG enrichment analysis on the STIL differentially co-expressed genes, and they were found to be signi cantly enriched in the cell cycle pathway. The PPI network results also showed that STIL and the hub co-expressed genes were proteins involved in the cell cycle. The above results indicate that STIL and the differentially coexpressed genes may affect the mitosis and cell proliferation of osteosarcoma cells through cell cycle signaling pathways. The cell cycle is known to have an important function in cellular growth and proliferation. Prior studies have reported that miR-671-5p and miR-299-5p target cell cycle regulation and mediate osteosarcoma proliferation [24,25]. Zhang et al. found that Ludartin induces apoptosis and cell cycle arrest at the G2/M checkpoint through the elevated expression of p21WAF1 in osteosarcoma cells [26]. Cell cycle pathways have also been established to play an important role in other tumors [27][28][29]. As an important factor in the process of mitosis, STIL may participate in the progression of osteosarcoma by regulating the cell cycle.
In the PPI network, we selected six genes (CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB) as the core STIL differentially co-expressed genes in osteosarcoma. CDK1 is a gene related to the cell cycle, and its abnormal expression leads to the development of tumors [30]. In the study conducted by Huang et al, microRNA-199a-3p, as a tumor suppressor gene was found to exhibit low expression in osteosarcoma and may interact with highly expressed CDK1 in the development of osteosarcoma [31]. CCNA2 and CCNB2 are members of cyclin family, which are critical for both cellular proliferation and apoptosis. Shekhar et al. found that CCNA2 is the common target of miR-449a and miR-424 in osteosarcoma, which inhibits tumor progression by inhibiting CCNA2 expression [32]. In another study, silencing CDC6 reduced CCNA2 expression and suppressed osteosarcoma cell proliferation and invasion [33]. CDC20 is a gene that regulates the cell cycle, and is reported to be involved in osteosarcoma development by analyzing the gene chip data [34]. Moreover, apcin blocks osteosarcoma cell growth and invasion by reducing the level of CDC20 expression, indicating that CDC20 may represent a potential therapeutic target [35,36]. In a recent study, CDC20 was found to exhibit high levels of expression in osteosarcoma cisplatin-resistant cell lines, which enhanced the sensitivity of drug-resistant cell lines to cisplatin by knocking out CDC20 [37]. Moreover, studies using bioinformatics analyses have found that RFC4 may interact with BUB1, which may function to promote osteosarcoma occurrence and development of [38]. AURKB is a serine/threonine kinase that has been proposed to stimulate the invasion and proliferation of osteosarcoma through PTK2/PI3K/AKt/NF-κB signaling pathway and VCP [39,40]. Thus, AURKB inhibitors may also provide a new option for the treatment of osteosarcoma [41]. In our study, we found that these six genes are highly expressed in osteosarcoma. Therefore, these ndings suggest that STIL may play a role in osteosarcoma progression through regulation of expression of these identi ed genes.
Our study is associated with certain limitations: 1) there is a large heterogeneity in our analysis (67%). Although we attempted to identify the source of heterogeneity through a sensitivity analysis, the results showed that no particular research to be the source of the heterogeneity; and 2) although we have demonstrated that STIL functions as a proto-oncogene in osteosarcoma, its potential molecular mechanism requires further veri cation both in vivo and in vitro.

Conclusion
In general, we con rmed that STIL plays a role as a proto-oncogene, is highly expressed in osteosarcoma, and is involved in osteosarcoma proliferation and invasion. Moreover, the bioinformatics analysis showed that STIL may participate in the biological function of osteosarcoma by regulating CDK1, CCNB2, CDC20, CCNA2, BUB1, and AURKB. Therefore, STIL may represent a potential biomarker for the diagnosis and     PPI analysis of the STIL differentially co-expressed genes in osteosarcoma. Six Hub genes were selected in the PPI network of STIL differentially co-expressed genes. PPI: protein-protein interaction.