Bone marrow micro environment contains cellular components,extracellular matrix and soluble components that form a heterogeneous networks and play an important role in the occurrence and development of myeloma.And WGCNA is an suitable tool to identify the causes and mechanism of myeloma recurrence by identifying the weighted co-expression relationships among genes and then clustering them into highly synergistic modules.Genes with strong relationships between clinic trait and modules were selected for KEGG and GO enrichment analysis, followed by TATFs-gene regulatory network.We found that three modules play different roles in the relapse of myeloma. As the study progressed, we found 19 kinds of cytokine/cytokine receptors upregulated in relapsed myeloma patients in brown module(Supplementary Table4) ,with CC13 ,CCL8, CCL23, CCL16, CCL24, IL18 thought to be regulated by TATFs(CEBPA and ELF5 ). CCL18 CCL2 CCL8 CXCL12 were related to infiltration of M2 Macrophages.
CCL18 secreted by M2 macrophages promotes tumor cell migration and invasion in gallbladder cancer [32]. CCL2 promotes macrophages-associated chemoresistance in multiple myeloma [33]. CCL8 promotes postpartum breast cancer by recruiting M2 macrophages [34]. A study showed that inhibition of M2 polarization can overcome myeloma resistance to lenalidomide by reducing CXCL12 expression[35]. However, the relationship between CCL18 and CCL8, M2 macrophages and myeloma recurrence has not been studied yet.
A large number of cytokine such as CXCL-12, IL-6, IL-8 and IL-17 have been elucidated and they construct a pro-inflammatory and permissive environment for the survival of myeloma cells[36]. Some mechanism of myeloma are finely elucidated,but the role of cytokine are still remaining unclear in the relapse of myeloma,and we provide dysregulated cytokine and their relationship with CEBPA and ELF5,suggesting that receptors for chemokine are an important point,as pharmacological inhibitors of individual receptors may have limited efficacy due to the redundancy,as well as a tendency toward compensatory increased expression of other chemokine family members[37].So it's critical to discover that key gene regulating cytokines and blocking them may simultaneously cause the turnover of a large number of chemokine,thus overcoming drug resistance and allowing cells to promote apoptosis and prevent myeloma cell metastasis. We therefore hypothesize that CEBPA and ELF5 may be the potential target for reversing cytokine.
In the CENPA-based regulatory network,we provide another possible pathogenesis of myeloma recurrence. In the yellow module,high expression of CENPA in relapsed myeloma was positively related to E2F2,suggesting that E2F2 may be associated with poor prognosis, after all, its expression in breast cancer correlated with lower recurrence-free survival[38]. BIRC5/survivin: A multitasking protein with dual roles in promoting cell proliferation and preventing apoptosis. Overexpression of AURKA lead to drug resistance in gastric cancer by up-regulating the gene expression of the anti-apoptotic protein survivin[39].
Furthermore, over-expression of survivin in the presence of anti-myeloma drugs was found to increase the viability of MM cells[40]. SKA1:knockdown of SKA1 shows a significant reduction in proliferation of bladder cancer cells[41]. DIAPH3:Highly expressed DIAPH3 promoted proliferation, non-anchored growth and invasion of pancreatic cancer cells[42]. Genes above are upregulated in yellow module,revealing that CENPA and other genes may play a key role in cell cycle,which lead to myeloma recurrence and promote tumor proliferation and drug resistance. The high level of genes expression in the yellow module has been rarely studied in myeloma,which provide a new direction for our research in myeloma patients.
In sky blue module:BST1 is found in relapsed multiple myeloma patients at a expression foldchange level of 4. BST1 is a analogue of CD38, which has similar functions such as hydrolysis NAD+ to adenosin (ADO). BST1 produce adenosine diphosphate ribose(ADPR) which is subsequently converted to AMP and eventually to ADO ,causing immunosuppressive functions, and is also growth factor for osteoblasts and osteoclasts. Interestingly BST1/CD157 was undetectable in myeloma cell line, suggesting that upregulated BST1 is a key cause of relapse in patients[43].The association between BST1 and drug resistance in myeloma cells have not been reported, and the high level of BST1 expression in our study suggests that it may contribute to the survival and immune escape of myeloma cells,leading to the recurrence of the disease.
Two of our screens from the hub genes were used to determine prognosis.TPX2 and PRAM1 showed independent predictive power in predicting EFS in myeloma, and both sets of genes have been rarely reported in myeloma. TPX2 is a spindle assembly factor required for mitotic spindles,and normal assembly,and its inhibition leat to a significant increase in mitotic index and metaphase myeloma cells.Therefore, it may be feasible to obtain inhibitors from it to suppress myeloma proliferation cells[44]. Furthermore,high levels of TPX2 indicated lower survival rate and promoted cancer cell migration and invasion in Non-Small Cell Lung Cancer[45].PRAM1 had never been reported in research of myeloma,but the high expression of PRAM1 is a sign for favorable prognosis in the CN-AML[46].This is consistent with our study that TPX2 is associated with high recurrence rates and PRAM1 gene expression is a protective factor in patients with myeloma.
The present study has several strengths and limitations. Bioinformatic tools have been used in our study to provide a general but comprehensive understanding of the mechanism of myeloma recurrence at the transcriptome level, and we provide some potential area of study worthy of in-depth study to overcome drug resistance and metastasis. A series of pivotal genes and response group pathways were identified to elucidate the molecular mechanisms of myeloma recurrence and to identify potential therapeutic targets.However,this study dose not profound enough and more experiments are needed to validate the function of moleculars in the regulatory networks of multiple myeloma patients.