Nearly all myeloma patients eventually relapse, even those who experienced a complete response to previous treatment[17]. Therefore, clarifying underlying mechanisms that result in disease recrudesce may turn on new therapeutic approaches for MM. In the study, we firstly constructed a co-expression network of 10826 genes by the WGCNA algorithm based on GSE82307. Through analyzing module significance, module membership, and gene significance, 68 genes were recognized as representative hub genes in the blue hub module. Then performing GO enrichment and KEGG pathway analysis and constructing a PPI network, we comprehended the biological characteristics of hub genes and the relationship among proteins expressed by hub genes. Furthermore, a total of 46 DEGs including 16 down-regulated genes and 30 up-regulated genes were screened in 33 MM patients. At last, 14 cross genes BIRC5, CDCA5, CENPE, RRM2, TTK, FOXM1, HJURP, CEP55, PBK, UHRF1, SKA1, NUF2, CDC20, TACC3 of hub genes obtained by WGCNA and DEGs were filtered out as key genes. Survival analysis indicated patients with a lower expression of CDC20, CDCA5, CENPE, CEP55, FOXM1, HJURP, NUF2, RRM2, SKA1, TTK showed significantly both longer OS and EFS. Thus, the ten key genes might be potential biomarkers for the diagnosis and treatment of MM. The three genes CDCA5, CEP55, HJURP are usually overexpressed in several cancers and strongly correlated with the poor prognosis of patients. Notably, there is no relevant research on the effects of the three genes on MM so far. Apart from that, the remaining seven genes have been studied to varying degrees in MM.
CDCA5 encodes sororin which is a regulatory protein during mitosis of eukaryotic cells. Sororin is pivotal for embryo development, maintenance of cohesion between sister chromatids, and correct chromosomal separation[18]. Sororin gene CDCA5 has been associated with the development and progression of several types of human cancers. In hepatocellular carcinoma, knockdown of CDCA5 by lentivirus-mediated shRNA resulted in cell cycle arrest at the G2/M phase and cell apoptosis [19]. Moreover, higher expression of CDCA5 in hepatocellular carcinoma was associated with increased tumor size, microvascular infiltration, as well as poor prognosis of OS and DFS[20, 21]. Likewise, CDCA5 knockdown could inhibit the growth of esophageal squamous cell carcinoma cells. And the expression of CDCA5 promoted tumor cell proliferation, distant metastasis, drug resistance. The high expression of CDCA5 was related to tumor progression and shorter OS in esophageal squamous cell carcinoma[22]. Guanghou Fu et al. found that overexpression of CDCA5 in UMUC3 cells(bladder cancer cell line) could hasten cell proliferation and hinder cell apoptosis[23]. In an additional experiment, they also found overexpression of CDCA5 in bladder cancer cells activated PI3K/AKT pathway to promote cell proliferation and inhibited mitochondrial-mediated cell apoptosis through reducing expression of downstream proteins, such as cleaved caspase-3, caspase-9, and cleaved PARP. The results of our study demonstrated that CDCA5 was up-regulated in relapsed MM. And a higher expression of CDCA5 heralded a poor prognosis.
CEP55 is a centrosome- and midbody-associated protein and a key regulator of cytokinesis. Over-expressed CEP55 promotes proliferation, survival, and metastasis of multiple cancer cells at several levels[24]. Some characteristic components of CEP55 are also present in cell cycle and proliferation genes whose overexpression is also detrimental to clinical outcome[24]. CEP55 has been studied in the context of cancers including breast cancer, hepatocellular carcinoma, osteosarcoma et al. In breast cancer, the high CEP55 expression blocked cancer cell apoptosis and contributed to chemotherapy resistance, associated with genomic instability. For breast tumors exhibiting high CEP55 expression, blocking MEK1/2-PLK1 signaling reduced outgrowth of tumors[25]. In hepatocellular carcinoma, CEP55 promoted activation of the JAK2/STAT3 signaling pathway and induced expression of matrix metalloproteinases via the same signaling, which led to disease deterioration. Moreover, CEP55 was highly expressed in hepatocellular carcinoma cells, which heralded a poor prognosis[26]. In osteosarcoma, CEP55 was significantly overexpressed in tumor tissues, which was associated with tumor size, tumor metastasis, and reduced OS. Meanwhile, knockdown of CEP55 inhibited cell proliferation, invasion, and migration[27]. In the data set GSE82307, CEP55 was identified as one of the key genes related to relapsed MM, and the ROC curve indicated that CEP55 was a sign of poor prognosis for MM.
HJURP is a histone chaperone of nucleosomes that plays a dual role in coordinating the recruitment of CENP-A and CENP-C. HJURP assists in the assembly of functional centromeres and mediates chromosome separation and cell division[28]. The dysfunction of HJURP has been confirmed in kind of cancers. HJURP was highly expressed in liver cancer cells and tissues, which was a sign of poor prognosis. The high expression of HJURP promoted the proliferation of liver cancer cells by down-regulating p21 through MAPK/ERK1/2 and AKT/GSK3β pathways[29]. Besides, the up-regulation of HJURP enhanced migration and invasion of liver cancer cells by interacting with sphingosine kinase 1[30]. In glioblastoma, HJURP knockout caused severe clonal ability and survival damage of glioblastoma cell lines. At the same time, these cell lines became more sensitive to radiation therapy after HJURP was reduced[31]. Moreover, HJURP was highly expressed in cells and tissues of pancreatic ductal cell carcinoma. HJURP may contribute to tumor cell growth, migration, invasion, and metastasis via the MDM2/p53 signaling pathway in pancreatic ductal cell carcinoma[32].
In order to further verify the carcinogenic ability and molecular roles of CDCA5, CEP55, HJURP in MM, we will continue to conduct in vivo, in vitro, and clinical research. The remaining seven genes CDC20, FOXM1, RRM2, TTK, CENPE, SKA1, NUF2 have been researched in MM to different depths. In the high-risk group of MM patients, the expression of CDC20 was significantly increased, accompanied by a shorter os and enrichment of genes related to proliferation. Besides, CDC20 knockout in myeloma cell lines could weaken cell apoptosis[33]. In MM patients showing high FOXM1 expression, both their DFS and OS were reduced. In vivo and in vitro experiments, down-regulation of FOXM1 inhibited the growth of myeloma cells, while up-regulation of FOXM1 resulted in the opposite[34]. RRM2 was significantly up-regulated. In terms of possible mechanisms, RRM2 participated in the proliferation and apoptosis of MM cells through the Wnt/β-catenin signaling pathway[35]. TTK was a kinase related to the prognosis of MM, and its expression was associated with the high-risk group of MM. Besides, TTK inhibitors could significantly undermine the viability and proliferation capacity of myeloma cell lines and could be combined with Melphalan or IMiD as a new treatment strategy[36]. CENPE and SKA1 were upregulated in side population cells of MM, encoding for proteins associated with cell cycle and mitosis[37]. At last, during the culture of the fresh bone marrow of myeloma, down-regulation of NUF2 expression regulating mitosis and transcription was observed, which could reveal ensuing proliferation arrest[38]. However, the three genes CENPE, SKA1, NUF2 are initially found in MM. And their roles remain to be fully elucidated.
In summary, we adopted the WGCNA analysis to determine ten novel biomarkers CDCA5, CEP55, HJURP, CDC20, FOXM1, RRM2, TTK, CENPE, SKA1, NUF2 for recurrence and prognosis of MM. The results are beneficial to shed light on molecular mechanisms of progression and filter out targeted agents.