COAD is the fourth most deadly cancer in the world, with a mortality rate of approximately 50% every year worldwide[2]. Surgical treatment and chemotherapy of early COAD can significantly improve the survival time of patients, but advanced COAD with chemotherapy and radiotherapy often have drug resistance and liver metastasis. With the in-depth research on the tumor microenvironment, CSCs, a type of cell subset, in COAD have been found to play an important role in the drug resistance of patients and the regulating immune cells in the tumor microenvironment[7, 8]. Tumor with high stemness has the characteristics of CSCs, which inhibit T cells and NK cell activation and finally leads to a poor outcome of immunotherapy[21–24]. But the in-depth mechanism of COAD stemness related genes resisting immunotherapy is not clear. In this study, we focus on identifying the genes related the COAD stemness and evaluating the correlation between COAD stemness-related genes and the immune cells in tumor environment.
We identified that as normal colon tissue developed into tumor tissue and deteriorates into lymph node metastasis status, mRNAsi scores become higher. Tumors with high stemness are more likely to migrate to other organs and lead to disease progression in patients[14], which is consistent with the increased mRNAsi scores in COAD. The result indicated that the stemness of COAD might be related to COAD progression and metastasis, but COAD patients with different mRNAsi scores had no significant difference in survival rate.
WGCNA analysis was applied to hierarchical cluster. the modules related the COAD stemness, and the brown and green modules were defined as key modules. Functional annotations of genes from the brown module mainly focused on contractile fiber, myofibril, actin binding and cGMP-PKG signaling pathway, which is down-expressed in COAD and closely related to COAD metastasis and recurrence[25, 26]. Besides, functional analysis of key genes from the green module was only enriched in cell cycle. The genes from the green module were highly positively correlated with mRNAsi scores, which might remain the stemness of COAD. The PPI network and co-expression diagrams indicated that genes in the same module had a high co-expression and interaction correlation with each other. The genes in green module had a highly positive relation with COAD stemness, and they were up-expressed in the spheres of HCT116. Hence, PLK4, TTK, CHEK1, KIF18A and BUB1 were identified as COAD stemness-related genes from the green module.
The functions of COAD stemness-related genes were mainly enriched in the cell cycle, which explained the phenomenon that CSCs promote COAD in a highly self-renewing state[27]. The Polo-like kinases (PLKs) play a critical role in centrosome replication and cell division and checkpoint regulation of mitosis, and Plk4 over-expression or activity can cause chromosomal instability and mitotic errors which would promote colon cancer development [28, 29]. TTK is an essential kinase to regulate chromosome alignment and error correction, which is positively correlated with the worse prognosis in several cancer[30, 31]. High expression level of TTK in colon cancer might regulate the functions of spindle assembly checkpoint and mitochondria to enhance the survival of aneuploid cancer[32]. BUB1 mitotic checkpoint serine/threonine kinase (BUB1) could regulate the chromosome segregation and was reported over-expressing in many cancers and was also positively correlated with poor prognosis[33, 34]. Bub1 insufficiency accelerates the rate at which whole chromosomes are lost or gained in ApcMin/+ mice, thereby perhaps accelerating Apc LOH and driving colonic tumorigenesis[35]. KIF18A, a member of the kinesin superfamily of molecular motor proteins, is a microtubule depolymerase and plays an important role in regulating the stability of microtubule plus ends in the mitotic spindle, and colon cancer cells with high expression of KIF18A show stronger migration[36].
Previous studies have found that colon CSCs inhibit immune cells activity in the tumor microenvironment and led to the poor outcome of immunotherapy[37–39]. we used the ESTIMATE algorithm to infer immune score and explored the correlation between immune score and mRNAsi score and COAD stemness-related genes[40]. The immune scores were found to be negatively correlated with the mRNAsi scores and COAD stemness-related genes. This would suggest that high expression of COAD stemness-related genes prevents the immunity of patients, which might cause the failure of the clinical effectiveness of immunotherapy.
Subsequently, we analyzed the association between TIICs and COAD stemness-related genes, and high TTK in COAD might lead to less infiltration of memory resting CD4 + T cells and more infiltration of M1 Macrophages. CD4 + T-cell activation is required for effective CD8 T-cell responses, and the higher infiltration of resting CD4 + T cells might lead to low CD8 T cells responses and poor immunotherapy[41]. This suggested that COAD with high stemness promoted poor outcome of immunotherapy by affecting the inactivation of memory CD4 + T cells. Interestingly, in COAD with high expression of TTK, the infiltration of M1 Macrophages was higher. Macrophages as a key driver of tumor inflammation, contribute to nurturing cancer stem cells and modulate CSC formation and maintenance by transferring into the M1 phenotype [42, 43]. This evidence suggests that inactivation of CD4 + T cells and low infiltration of M1 Macrophages might be detrimental to cancer stemness related gene expression and immune escape. Overall survival analysis had elucidated significant roles for high M1 Macrophages in predicting prognosis for COAD patients, the COAD patients with low M1 Macrophages infiltration were considered to have a better prognosis. The result above demonstrated that the stemness of COAD was related to infiltrating immune cell in the tumor microenvironment.
In conclusion, PLK4, TTK, CHEK1, KIF18A and BUB1 were positively correlated with stemness of COAD, and TTK might be associated with CD4 T cells activation and M1 Macrophages infiltration in the tumor microenvironment. Also, the increased infiltration of memory activated CD4 + T cells and M1 Macrophages might further contribute to formation and immune escape of COAD with high stemness. TTK may have the potential as a prognostic biomarker and adjuvant target to promote immunotherapy sensitivity to improve outcomes in COAD. But conclusion derived from our study was mainly based on bioinformatic analysis, but it was not enough to reflect the gene expression in COAD with high stemness.