Osteosarcoma is challenging cancer with high chemotherapeutic resistance and metastatic incidence. Because the prognosis of osteosarcoma is still poor, identifying prognostic genes involved in the pathogenesis of metastatic osteosarcoma and constructing a predictive model of survival are of crucial importance for management policy. In this study, we used the TARGET datasets to identify DEGs between metastatic osteosarcoma and non-metastatic osteosarcoma. In total, 214 genes were consistently expressed differentially in metastatic osteosarcoma (114 up-regulated and 100 down-regulated). Twenty-six significantly enriched functions were obtained in GO function and KEGG pathway enrichment analyses, such as cell-cell signaling, inflammatory response, cytokine activity, immune response, and cytokine-cytokine receptor interaction. A PPI network was constructed to investigate the interrelationship of the DEGs. Survival analysis and univariate Cox analyses were used to find out candidate genes for multivariate analyses. Fifteen genes were screened from the survival analysis and univariate Cox analyses. Multivariate analyses demonstrated that five genes were independent prognostic indicators for osteosarcoma, specifically three were up-regulated (MUC17, MYC, TAC4) and two were down-regulated (HERC5, OR8G5). After verification, we found that this five-genes signature was not affected by clinical traits such as age at diagnosis, gender, metastasis, and relapse. Not only did the five genes have a high potential of distinguishing the patient with a high risk of metastasis from osteosarcoma patients, but it also could be regarded as a signature to predict the prognosis of osteosarcoma patients.
The MYC proto-oncogene, a bHLH transcription factor (MYC), also known as MRTL, MYCC, c-Myc, and bHLHe39, is a proto-oncogene and encodes a nuclear phosphoprotein that plays a role in cell cycle progression, apoptosis, and cellular transformation. Amplification of this gene is frequently observed in numerous human cancers and also has been reported in osteosarcoma. MYC gene up-regulation in osteosarcoma has been associated with poor prognosis of osteosarcoma patients [14–15]. HOTTIP up-regulates MYC, and the positive feedback loop formed by HOTTIP and MYC promotes osteosarcoma cell migration and invasion [16]. Overexpression of miR-34a up-regulates MYC expression and accelerates osteosarcoma cell apoptosis induced by cisplatin. MYC was necessary for osteosarcoma apoptosis induced by the combination of miR-34a and cisplatin [17]. Recombinant adenovirus encoding antisense c-myc (Ad-Asc-myc) increases the sensitivity of osteosarcoma cells to cisplatin in vitro and induced apoptosis [18].
HECT and RLD domain containing E3 ubiquitin protein ligase 5 (HERC5), also known as CEB1 and CEBP1, is a member of the HERC family of ubiquitin ligases and encodes a protein with a HECT domain and five RCC1 repeats. Zhu et al reported that HERC5 was negative correlated with SOX18 expressions in osteosarcoma cell, and SOX18 was overexpressed in OS and promoted migration and invasion of ostesarcoma [19]. But them did not elaborate further on the role of HERC5 in osteosarcoma.
None of the other three prognostic genes (MUC17, TAC4, OR8G5) have been reported in osteosarcoma, but MUC17 and TAC4 have been shown to play an important role in many cancers. MUC17 was more highly expressed in gastric cancer, and inhibited the progression of gastric cancer through a MYH9-p53-RhoA regulatory feedback loop to limited imflammatory responses [20]. TAC4 was a member of the tachykinin family of neurotransmitter-encoding genes. Tachykinin proteins are cleaved into small, secreted peptides that activate members of a family of receptor proteins. Hemokinin-1 (HK-1), the newest tachykinin encoded by the TAC4 gene, promoted migration of melanoma cells [21]. Studies on OR8G5 are still infrequent at present, and there are no relevant studies on tumor field.
Our study indicates that the five prognostic genes play a potentially important role in the pathogenesis of metastatic osteosarcoma and may serve as target biomarkers for the prognosis of osteosarcoma. But four of them have not been researched in osteosarcoma. Further experiments are needed to validate and refine their function and signaling pathways. Besides, there are still several limitations that cannot be ignored in our study. First, the number of eligible cases was not particularly sufficient and some meaningful clinical data were incomplete, so we had to remove some cases, which may cause relevant bias. Second, lack of independent large-scale biological tissue samples or cell lines to verify the results obtained in this study and completely demonstrate the underlying mechanism of five genes in osteosarcoma. Our findings will be more reliable if another independent dataset could be used for validation.
Osteosarcoma is a complex regulatory network, multigene as biomarkers achieved higher specificity and sensitivity compared with single-gene. The present study identifies five target genes related to the metastasis of osteosarcoma. This five-genes signature can be used to predict the prognosis of osteosarcoma patients and may become new potential therapeutic targets for treatment of osteosarcoma.