The transcription factors E2F family were reported in the regulation of cell cycle and apoptosis. Some of them could activate the quiescent cells from G phase to S phase and they might be the indispensable and determant factor for the survival of CRC cells, as E2F knockdown slows down the GⅠ/S transition and the prolifration rate of human intestinal epithelial cells . In spite of this, survival analysis of CRC patients with bioinformatic tools has not been widely reported so far. In this study, we integrated several online resourses about the expression of E2F factors, and evaluated their prognostic values in OS and DFS.
Among all these E2F family, E2F1 is the most extensively investigated member of E2F family in CRC[23–25]. A research reported suppressive function of E2F1 showed that the absence of E2F1 gene lead to an increased incidence and progression of tumorigenesis induced by PTEN loss. While others found that E2F1 was overexpressed in CRC patients and E2F1 could promote the aggressiveness of human colorectal cancer by activating the ribonucleotide reductase small subunit M2. In this study, we disclosed that the mRNA expression of E2F1 was significantly higher in both COAD and READ based on ONCOMINE and TCGA datasets. Generally speaking, E2F2 was regarded as the “activator” by acting on downstream proteins to promote proliferation and cell growth. In the Table 1, we found that the expression of E2F2 were significantly lower in CRC tissues, which is in accord with Xanthoulis’s report that E2F2 was expressed in 41cases at low levels. However, the transcriptional levels of E2F2 were significantly higher both in COAD and READ. It might be explained by Johnson that the ability of some E2F family members to behave as both oncogene and tumor suppressor gene can be reconciled by putting E2F into context. However, the OS and DFS were not associated with the differential E2F1 or E2F2 expression levels after the follow up of 150 months. Besides, neither the clinical stages of COAD and READ patients were not correlated with E2F1 or E2F2 expression.
The overexpression of E2F3 in CRC tissues and cell lines has been substantiated by RT-qPCR. Both in vitro and in vivo experiments, Zhang demonstrated that high levels of miR-34a expression could inhibit the expression of direct target gene E2F3, and then inhibit the growth of CRC cells. CircPRMT5 is an upstream regulator of E2F3 and it could promote the proliferation of CRC by inducing the expression of E2F3. Comparing with other E2F members, E2F4 has been reported in a lower extent. In Garneau’s report, they found that E2F4 is required for cell cycle progression of not only normal intestinal crypt cells but also CRC cells. In an immunohistochemical study of 100 cases, Xanthoulis found that E2F4 expressed nuclear immunopositivity in all cases. In our study, the expressions of E2F3 were significantly higher in CRC tissues than normal tissues. While the data about the expressions of E2F4 was absent. Secondly, the expressions of E2F3 was correlated with the clinical stages in COAD patients, implying the E2F3 could be used as the reference index of tumor staging for this disease. Besides, we found that the higher E2F3 and E2F4 expression were significantly related with poor OS in COAD patients but not with DFS. So, these two factors could be applied to evaluate the prognosis of COAD patients.
E2F5 has been widely reported with upregulation in several tumors and played a crucial part in the cell proliferation, cell migration and invasion of these diseases[31, 32]. Yu found that the E2F5 with higher expression significantly increased CRC cells proliferation and could also reverse the inhibition of CRC cells proliferation induced by upstream SNHG6 silencing. Similarly, the facilitating effect of E2F5 on CRC cells progression could be suppressed by upstream miR-34a. The higher expression of E2F6 was reported in gastric and breast cancers[35, 36]. However, there is no study about the expression and prognostic role of E2F6 in CRC patients as far as we know. The E2F7 was significantly higher in CRC cells as reported by Liu’s study, and the decreased E2F7 expression by all-transretinoic acid could reverse the progression of CRC cells. The low-frequent missense variant in E2F7 was reported with significant association with CRC risk, indicating the important role of E2F7 in development of this tumor. Similarly, the expressions of E2F8 in CRC patients were significantly higher, and these CRC cells with E2F8 knockdown showed lower growth rates. In this study, the expressions of E2F5, E2F6, E2F7 and E2F8 were significantly higher in CRC tissues than normal tissues. But the expression levels of these four E2F members were neither related with the clinical stages nor the OS and DFS of COAD and READ patients.
In general, we thoroughly explored the expression and survival evaluation of these E2F members in CRC patients,and found a new path with bioinformatic tools to analyze the complexity of CRC. We found that E2F1/2/3/5/7/8 were significantly higher in CRC patients than that in normal samples and E2F3 could be used as the biomarker for clinical staging in COAD patients. Besides, the lower E2F3 and E2F4 expression predicted higher OS and DFS rate.