Identification of Driver Genes and Key Pathways of Osteosarcoma Shows K-7174 as a Novel Anti-osteosarcoma Drug by inhibiting VCAM1

Background Osteosarcoma is one of the leading causes in cancer-related death of children and adolescents. However current standard therapeutic strategy, surgery combined with neoadjuvant chemotherapy is very limited in effects. As big data mining and analysis using bioinformatics method has been applied in the diagnosis and treatment of many cancers, we want to use bioinformatic combined with experimental assays to found new molecular targets and test new drug for osteosarcoma. Methods The gene chip of osteosarcoma samples constructed by Richter GH et al were downloaded from the Gene Expression Omnibus (GEO) database, Gene Ontology analysis (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed on differential expression genes which screened by bioinformatics methods. Protein-protein interaction network was constructed by suing STRING database to found hub gene. Combined with pertinent literature, genes of interest and corresponding drug was selected. Series of experiments were performed on the osteosarcoma cell lines in vitro, involved cell viability test, colony formation assay, migratory and invasive tests, western blot as well. Results A total of 1069 DEGs were obtained from data, including 375 up-regulated genes and 694 down-regulated genes. Differentially expressed genes mainly involve biological processes such as cellular immune function, such as interferon-gamma-mediated signaling pathway and antigen processing and presentation of peptide. Among top 20-ranked degree hug gene evaluated by PPI network, vascular cell adhesion molecule-1(VCAM1) was picked out. An VCAM1 inhibitor K-7174 was treated in U2OS and MG63 cell lines. In vitro experiments have shown that K-7174 can inhibit the proliferation, migration and invasion of osteosarcoma, the protein expression of VCAM1was also decreased by K-7174. VCAM1 could be a potential target for osteosarcoma and K-7174 promises to be a therapeutic drug after more nuanced evaluation in animal and clinical trials.


Introduction
Osteosarcoma which originated from primitive mesenchymal cells is a kind of primary malignant tumor. Most patients diagnosed with osteosarcoma are children and adolescents under 20 years old. Despite its low morbidity, the mortality rate and incidence of metastasis of osteosarcoma are extremely high [1]. In the past few decades, surgeries were the main methods for the treatment of osteosarcoma, however the overall outcome is poor with high disability rate. With new progress in medical technology, surgery combined with neoadjuvant chemotherapy has become the standard treatment [2]. Chemotherapy drugs mainly include adriamycin, methotrexate, ifosfamide and cisplatin. Since the combination of surgery and neoadjuvant chemotherapy, the 5-year survival rate for OS patients has been improved [3]. However, osteosarcoma prognosis is still suboptimal, distant metastasis especially lung metastasis is common in early stage and patients are prone to relapse after treatment [4]. The long-term use of chemotherapy drugs is often accompanied by drug toxicity also cause adverse effects on multiple organs in patients. Therefore, it is very important to identify new treatment methods of osteosarcoma.
In recent years, with the development and application of bioinformatics and microarray technology, Bioinformatics data-mining of gene has been applied in the diagnosis and treatment of many diseases [5]. And it is possible for researchers to better understand the mechanism of osteosarcoma at molecular level. Many bioinformatics analysis techniques based on genomics, transcriptomics and proteomics have revealed the high genetic heterogeneity of osteosarcoma [6]. Many genes, such as EGR1, CXCL10 and MYC have been found to be associated to the progression and metastasis of osteosarcoma [7]. Since database updated and new data uploaded, there has been a continual increase in the number of new targets gene available for osteosarcoma treatment, which suggests that additional genetic factors remain to be found.
In this study, bioinformatics method was used to further investigate OS microarray data.
A total of 1069 differential expression genes (DEGs)between metastatic human osteosarcoma tissue and non-metastatic human osteosarcoma tissue were selected by screening. Gene Ontology (GO) analysis and pathway enrichment were performed for functional and pathway analysis of DEGs, and protein-protein interaction (PPI) network was also constructed for DEGs. Finally, hub genes with top 20 ranked degree in the PPI network were picked out. Among top 20-ranked hub genes, there were various gene that had been known to development, progression and relapse of osteosarcoma, such as TP53 and MYC. We notice vascular cell adhesion molecule-1 (VCAM1) which has been reported closely associated with distant metastasis and poor prognosis of patients with breast and colorectal cancer. However, VCAM1 was not fully understood in tumorigenesis and metastasis of osteosarcoma. And it was still unknown if VCAM1 could be a potential therapeutic target of osteosarcoma. So, aiming at VCMA1 we used K-7174 as potential drug in subsequent in-vitro experiments. In osteosarcoma cell lines U2OS and MG63, we found that k-7174 could inhibit the proliferation, migration and invasion of osteosarcoma cells by CCK-8, colony formation, scratch and Transwell assay. Further western blot confirmed that VCAM1 could indeed inhibit the expression of VCAM1 in osteosarcoma cell. Our results showed that k-7174 could effectively inhibit the progression and metastasis of osteosarcoma in vitro, and VCAM1 could become a potential treatment target for osteosarcoma.

Microarray Data
The gene expression profiles of osteosarcoma named GSE73166 were downloaded

Statistical analysis
GraphPad Prism 8.0 was used to make statistical analysis and draw statistical chart The data were analyzed using the GraphPad Prism 8.0 program (GraphPad Software, Inc., San Diego, CA). An independent-samples T test was selected to analyze data. Data are expressed as the mean ± standard deviation of three repeated experiments. P-values just below 0.05 was considered to be statistically significant.

Differential expression analysis for gene expression.
A total of 1069 differential expression genes or DEGs were identified, of which 375 genes were upregulated and 694 genes were downregulated. The volcano map of differential genes is shown in Figure 1A

Functional and pathway enrichment analysis of DEGs
All DEGs were uploaded to the DAVID online tool for further GO functional analysis and KEGG pathway enrichment analysis. AS shown in Table 2, Figure 1C and D, Figure 2A, B, C and D, the results showed that upregulated DEGs were mainly associated with cellular immune function, such as interferon-gamma-mediated signaling pathway and antigen processing and presentation of peptide. Furthermore, the down-regulated genes were also associated inflammation reaction and immune response, including ECM-receptor interaction and regulation of inflammatory response.
GSEA results also reveal DEGs mainly enriched in Wnt and Notch signal pathway.   Figure 1B. MCODE analysis were further performed, top 3 significant modules were selected as shown in Figure 3 and Table 3.     We analyzed the effect of K-7174 on the invasion ability of U2OS and MG63 cell lines by Transwell assay. As shown in Figure 6A,

The expression of VCAM1 was inhibited by K-7174 in osteosarcoma cell lines
As K-7174 was reported to be an inhibitor of VCAM1, but it was still unknown whether the expression of VCAM1 in osteosarcoma cell lines could be decreased by K-7174.
Western blot assay was performed to gain further insight. As shown in Figure 7A,  and blots were included in supplement materials.

Discussion
As a highly malignant primary bone tumor, osteosarcoma is characterized by immature osteoid tissue formed from tumor cells, osteosarcoma may occur in any bone of the body, although the femur and tibia are the most common location [8]. The poor prognosis and high mortality of osteosarcoma mainly due to early local diffusion and distant metastases of tumor. Local diffusion lead amputation to the only curative modality for surgery and cause disability and lung is the main target organ of osteosarcoma metastasis, 80% of patients with advanced osteosarcoma have lung metastasis and respiratory failure caused by that is the main cause of death of osteosarcoma patients [9]. Although the diagnosis and treatment of osteosarcoma have improved significantly in the past few decades, new treatment options like neoadjuvant chemotherapy and radiotherapy are constantly being explored in clinical practice, more than half of osteosarcoma patients still do not benefit from current treatments [10].
Therefore, the development of new target for therapy and effective drugs is urgent for both patients and medical researchers.
The exact etiology of osteosarcoma remains unclear, both genetic and environmental factors reported to contribute osteosarcoma development. Long term exposure to radiation and alkylating agents can also promote the development of osteosarcoma [11].
Studies have shown that the pathogenesis of osteosarcoma is closely related to genes and genetic factors [12]. Therefore, clarifying the pathogenesis of OS from the molecular level is a challenge but also an opportunity. With its rapid development, gene chip is widely used in the diagnosis, treatment and prognosis evaluation of cancers [13].
As a new interdisciplinary subject in the field of life science, bioinformatics could deeply mine the microarray data of gene expression profile. One the one hand, it provides the possible molecular mechanism basis for the occurrence and development of diseases, on the other hand, it also provides feasible ideas for experimental research.
Thus, based on the intersection of biology and information technology, we designed and implemented this study.
After tumor tissue from non-metastatic metastatic group compared with tumor tissue from metastatic group, A total of 1069 differential expression genes were identified, including 375 upregulated genes and 694 downregulated genes. The subsequent GO analysis revealed that DEGs mainly involved inflammatory-related and immunerelated biological processes, such as interferon-gamma-mediated signaling pathway, antigen processing and presentation of peptide, and regulation of inflammatory response. Both immunity and inflammation process constitute the basic characteristic of tumor microenvironment, although their relationship is still vague [14]. It is the speculated immune response is stronger in the early stage of tumor development, while inflammation response is stronger in the later stage. In the osteosarcoma, tumor cells not only control the recruitment of immune cells such as T-lymphocytes, mast cell and macrophages, but also regulate aggregation of inflammatory related factors like interleukin, interferon and transforming growth factor [15]. The dynamic state of both cell and molecules eventually establish a tolerant environment available for development, drug resistant and metastasis of osteosarcoma. In previous study, some agents which affect immune function has been used in osteosarcoma patients such as Anti-GD2 antibody and Anti-IGF-1R, some clinical effects were also observed [16,17].
Furthermore, KEGG and GSEA analysis were performed, the result showed DEGs were highly enriched in phagosome, antigen processing and presentation, wnt signaling pathway and ECM-receptor interaction. Both the processing and presentation of antigen and formation of phagosome are the process of cellular immunity [18]. Wnt signaling pathway and ECM-receptor interaction which was considered to be related to epithelial mesenchymal transition and tumor metastasis, has also been observed that the enrichment in various tumors like breast cancer and glioma [19,20]. the loss expression of TP53 and eventually lead to occurrence of osteosarcoma [21].
Animal models with a mutation in TP53 developing osteosarcoma also demonstrate that Tp53 essentiality [22]. Myc was reported to be related in OS development and metastasis by activating ERK pathways [23]. High expression of Myc increase cell proliferation, migration and clonogenicity in in osteosarcoma cell lines, and mutation of Myc gene were observed in more than 10% clinical cases of osteosarcoma [24]. While only only a few studies focused VCAM1 in osteosarcoma. VCAM1 is a 90-kDa glycoprotein which mainly expressed in the surface of endothelial cells [25]. Some cytokines, such as IFN and TNF could stimulate the expression of VCAM1 in other cells, fibroblasts, myoblasts and cancer cells were found express VCAM1 under inflammation and chronic conditions [26]. Six immunoglobulin (Ig)-like domains, one transmembrane domain and one cytoplasmic domain comprised the human VCAM1,and α4β1 integrin play mainly role in ligand binding with VCAM1. During inflammatory responses, ligands binding to VCAM-1 and stimulating a series of inner cell affair, subsequently rolling and firm adhesion between endothelium and leukocyte are regulated [27]. Besides inflammation, VCAM-1 has also reported to be involved in tumor angiogenesis and metastasis. In breast cancer, micro-vessel density are relevant to serum VCAM-1 [28]. In another study, VCAM-1 knockdown helped reduce tube formation ability of HUVEC that could be a potential angiogenic target [29]. Meanwhile， high VCAM-1 expression was also considered to be a hint for metastasis in various cancer cells. A comparative gene expression profile analysis of breast cancer cell suggest a lung metastatic activity with VCAM1 expression, and patients with early lymph node metastasis exhibited overexpression of VCAM1 in a study of colorectal cancer [30,31]. These related studies suggest that VCAM1 might be target of tumor metastasis.
In previous study, upregulation of VCAM-1 was observed to increase the migration of osteosarcoma cell [32]. Further study showed that CXCL13/CXCR5 axis activation facilitates the production of VCAM-1 via NF-kb pathway [33]. region, and was used as a potential anti-inflammatory drug [34]. In another study, K-7174 was also found to help recover the production of erythropoietin inhibited by IL-1 in animal model [35]. So far, the only research related the application of K-7174 in cancer treatment was from myeloma. K-7174 exhibits anti-myeloma effects by inhibiting class I histone deacetylases transcriptional activities, and the IC50 of K-7174 varied within 5μM to 20 μM in different myeloma cell lines [36]. Besides antiproliferation effect, anti-migration and anti-invasion effect of K-7174 were detected by us. Results of scratch assay and transwell showed both migration ability and invasive capacity of osteosarcoma cells were markedly suppressed by K-7174.Further more treatment of deletion K-7174 did reduce the protein expression level of VCAM-1 in osteosarcoma cells. Although, K-7174 has not been reported used to inhibit the osteosarcoma metastasis, inhibition of VCAM-1 by a natural compound called DHTI had been proved effective in anti-metastasis of osteosarcoma cell lines [37]. AS a VCAM1 inhibitor, K-7174 does exist potential for preventing tumor metastasis.

Conclusion
In this study, bioinformatics method was used to mine the osteosarcoma microarray data, and relevant DEGs was identified. Biological process, molecular function and signal pathway involved in the DEGs were also analyzed. The hub gene was obtained by functional enrichment. The gene of interest, VCAM1 was picked out from top 20ranked degree hub gene. In vitro experiments on osteosarcoma cell lines showed that an inhibitor of VCAM1, k-7174 could inhibit the proliferation, migration and invasion of tumor cells, reducing the protein expression of VCAM1 as well. Our study suggested that VCAM1 could be used as a target in treatment of osteosarcoma, and k-7174 was a potential drug. Of course, more detailed experiments including animal studies and clinical trials need to be performed to further validate our idea.