Colon cancer is one of the most common malignant tumors in the world, with high morbidity and mortality. Although systematic medical treatment and surgical technique have made a great progress, many patients with recurrence and metastasis of colon cancer still have poor prognosis. CSCs are a small group of hidden tumor cells with the potential for self-renewal and multi-differentiation. Nowadays, accumulated studies have shown the action of CSCs in tumor initiation, progression and malignization, and therapeutic resistance. However, it is still a challenge to identify the therapeutic modalities targeting CSCs. Herein, we used WGCNA to screen out key genes associated with COAD stemness based on mRNAsi.
Firstly, we found COAD tissue exhibited higher stemness than normal tissue, consistent with other reports in bladder cancer, gastric cancer, lung cancer, and breast cancer. Patients with higher mRNAsi had better survivals, which was similar to a previous report, but contrary to what we expected. After eliminating the effect of tumor purity, we found no statistically significant difference in survival curves between the two groups. Differences in treatment options might result in similar survival rates between the two groups. Next, we constructed a coexpression network to analyze the expression of DEGs, and divided DEGs into different gene modules. Green module had the highest positive correlation with mRNAsi, brown and red modules had negative correlations with mRNAsi. Hence, key genes affecting the COAD stemness should be screened from green module, where 27 key genes were obtained based on GS and MM. These genes were upregulated in COAD tissue. There were high coexpression relationships among them at the protein and transcription levels.
Next, KEGG and GO analyses showed that key genes were involved in organelle fission and cell cycle, DNA replication, cellular senescence pathways. The enriched signaling pathway included human T−cell leukemia virus 1 infection and viral carcinogenesis, indicating that these genes may have a role in the origin of cancer and tumorigenesis. Previous studies have shown a high prevalence of JCV infection in colon cancer, which induces chromosomal instability and provoke metastasis. HPV infection may represent an important factor in the transformation of precancerous lesions to neoplastic phenotypes in colon cancer. P53 signaling pathway plays an important role in metabolism, aging, apoptosis, proliferation, cell cycle regulation, and inhibition of tumor expression[21-24]. P53 mutations play a key role in adenoma-carcinoma transformation, chemoresistance, and poorer prognosis in colon cancer[25-29]. These findings suggest that key genes may play important roles in tumorigenesis, progression and drug resistance. Finally, the expression profile of each key gene was verified in multiple cancer tissues by external data from GEO, GEPIA, and Oncomine.
KEGG and GO analyses suggested key genes could regulate tumor stemness by affecting cell cycle. BUB1, BUB1B, and MAD2L1 played a central role in mitosis, they could regulate the spindle checkpoint activity, and delay the onset of anaphase ensuring proper chromosome alignment and segregation[30-34]. BUB1 mutations could cause chromosomal instability, DNA damage, and increase the risk of colon cancer[33, 35, 36]. BUB1, BUB1B, and MAD2L1 were highly expressed in many other cancer types[11, 37-40], which played important roles in maintaining CSCs potential [36, 41-43] and predicted a poor prognosis[38, 44, 45]. The protein encoded by SPC25 was involved in kinetochore-microtubule interaction and spindle checkpoint function. Previous studies have shown that SPC25 upregulation increases CSCs properties and predicted poor prognosis in lung adenocarcinoma, breast cancer, hepatocellular carcinoma and head and neck cancer [46-50].In this study, we innovatively found that SPC25 was highly related to COAD stem cell properties. No studies have found the role of SPC25 in colon cancer.
CCNA2 bound and activated CDK1, thus induced both cell cycle G1/S and G2/M transition[51-53]. CCNA2 and CDK1 were key players in regulating cell cycle progression and mitosis, both of them were upregulated in colon cancer[54, 55]. CENPA is the centromere-specific histone-H3-like variant, which specify the equal chromosome segregation and plays a central role in centromere structure and function[56-58]. CENPA was overexpressed at protein and transcriptional levels in colorectal cancer tissues, which could be related to aneuploidy.
NCAPG, and NCAPH mediate mature chromosome condensation and stabilization during mitosis and meiosis[60-63]. Both of them were overexpressed in colorectal cancers [40, 64]. Furthermore, study showed that NCAPH promoted colon cancer cells proliferation, and NCAPH knockdown motivated G2/M cell cycle arrest and phase cell apoptosis. Some studies showed that the upregulation of NCAPG and NCAPH was associated with proliferation, progression, resistance and poor prognosis in other cancers[65-71]. Some studies showed SUV39H2 was associated with stem cell self-renewal and chemosensitivity[72-74].
DNA2, ORC6, MCM10, and RFC4 are essential for the initiation of the DNA replication, chromosome maintenance and cell proliferation in eukaryotic cells[75-81]. ORC6 was implicated in colorectal cancer initiation and progression. Reduction of ORC6 expression could increase the sensitivity of colon cancer cells to cisplatin and 5-fluorouracil. High levels of RFC4 were in association with differentiation and poor prognosis in colorectal cancer, the loss of RFC4 suppressed cell proliferation and promoted the cessation of the S-phase cell cycle arrest. These four genes were high expressed and related to prognosis in other cancers, and could be potential therapeutic targets[45, 84-88]. Experimental research is required to demonstrate their role in CSCs.
The differential expression of CHEK1, XRCC2, DDIAS and NEIL3 were related to DNA damage and repair[89-92]. XRCC2 polymorphisms and its protein were associated with susceptibility of colorectal cancer[93, 94]. CHK1 inhibitors are promising in chemo- and radio- sensitization. XRCC2 single nucleotide polymorphisms may influence the risk and survival of breast cancer. NEIL3 were highly expressed in tumor tissues and cells with high proliferative potential.
KIF18A and KIF23 are members of the kinesin superfamily of microtubule-associated molecular motors, which are related to chromosome separation, intracellular transport, mitotic spindle formation and cytokinesis. Aberrant expressions of KIF18A and KIF23 were related to metastasis and poor prognosis in colorectal cancer[98, 99]. TTK could regulate tumor proliferation and differentiation and predict prognosis in colon cancer. PLK4 expression plays a key role in centrosome duplication for cell division. The upregulation of PLK4 may induce carcinogenesis and metastasis by regulating the Wnt/β‑catenin signaling pathway, which may be a therapeutic target.
Studies found that upregulation of MTFR2 promoted proliferation, invasion, migration through switching glucose metabolism in breast cancer and oral squamous carcinoma[102, 103]. Upregulated expression of MTFR2 predicted a poor prognosis[104-106]. Activation of MELK was significantly associated with prolonging survival and promoting proliferation of CSCs in many organs. MELK expression was correlated with resistance of radiation and chemotherapy in colorectal cancer, as well as survival of patients with multiple cancer types[109-111].The clinical trials and basic experiments for cancers with MELK inhibitor are ongoing. The overexpression of NUP107 increased resistance to oxidative damage in cervical cancer. Previous studies have indicated that DEPDC1B induces tumor cell proliferation, migration, invasion and metastasis through various mechanisms[113-116]. PNPT1 encodes a polyribonucleotide nucleotidyltransferase, mutations in PNPT1 inhibit RNA from importing into mitochondria and result in respiratory-chain deficiency. There no study has showed the role of PNPT1 in tumorigenesis and development.