As is well known, numerous cancers have a poor prognosis because of early diagnosis difficult. A high proportion of patients have an advanced stage of cancer at diagnosis, indicating that tumors have spread to adjacent or distant organs, tissues, and lymph nodes, indicating that a poor prognosis. Consequently, it is indispensable to develop novel biomarkers that are reliable for predicting cancer patient diagnosis and prognosis.
UHRF1, also known as NP95 or ICBP90, is a nuclear protein gene associated with tumorigenesis. UHRF1 primary exists in the S phase of perinocytic phase,which is tightly associated with cell proliferation. UHRF1 can bind to the trans-CCAAT box in the promoter region of topoisomeraseⅡa to regulate the expression of this enzyme[23].
UHRF1 is widely considered to be a key modulator protein of methylated DNA and histone[24]. UHRF1 recruits DNA methyltransferase to newly synthesized DNA, which plays a key role in maintaining DNA methylation, and methylated modification is the key to transmitting epigenetic information during cell division[25]. UHRF 1 is overexpressed in a variety of different types of tumors, such as bladder cancer[26], colorectal cancer[27], lung adenocarcinoma[28] and intrahepatic cholangiocarcinoma[14]. UHRF1 can regulate the expression of multiple oncogenes or suppressor genes in tumors. Ectopic expression of UHRF1 can accelerate the proliferation or metabolism of cancer cells, and downregulation of UHRF1 can induce cell cycle arrest, DNA damage reaction and apoptosis.
Niinuma et al. indicated that the combination of UHRF1 depletion and histone deacetylase inhibition reactivated the silenced genes and strongly suppressed colorectal cancer cell proliferation[27]. Geminin is an inhibitor of DNA replication licensing and cell cycle. Chen et al.[29] study demonstrated that overexpressing UHRF1 was involved in the proliferation of vascular smooth cells by reducing inhibitory geminin protein levels to promote cell cycle and activating PI3K-Akt signaling. Kim et al.[30] investigated that UHRF1 is overexpressed in HPV-induced cervical cancers. They further detected the upregulated UHRF1 and the down-regulated thioredoxin interacting protein (TXNIP) in cervical cancer by Western blotting and immunohistochemistry. Kim's results indicate HPV might induce carcinogenesis through UHRF1-mediated methylation TXNIP promoter, decreasing the expression of tumor suppressor genes(TSG). Reports indicated that repression of regulator of G-protein signaling (RGS)2 increases cancer cell growth. Ying et al.[31] studies demonstrated that epigenetic repression of RGS2 by UHRF1 contributes to bladder cancer progression. In bladder cancer cells, UHRF1 inhibits regulator of RGS2 expression by increasing the methylation of CpG nucleotides of the RGS2 promoter. High UHRF1 expression of correlated with aberrant TGS2 promoter methylation in bladder tumors, which results in the loss of TGS2 expression.
Uncontrolled cell proliferation is the primary feature of malignant tumors in cell biology, which is closely connected with abnormal cell cycle regulation[32]. Knockdown of UHRF1 in A431 and Scl-1 attenuated cutaneous squamous cell carcinoma (cSCC) cell proliferation, migration, and invasion, leading to G2/M cell cycle arrest and apoptosis. Through a mouse xenograft model, Li et al. found UHRF1 deficiency ameliorated tumor growth[33]. Wan et al.[22] results showed that UHRF1 was overexpressed in almost all of the PCa cell lines. In addition, knockdown of UHRF1 expression by specific siRNA resulted in downregulation of UHRF1 expression, which inhibited the proliferation and migration of PCa cells and eventually induced apoptosis. Downregulation of UHRF1 by RNAi inhibited proliferation and clonogenic ability of medulloblastoma (MB) cell lines with cell cycle arrest in G1/G2-phase. Meanwhile, MB cells transfected with lentish UHRF1 showed increased p16 expression and location shift of CDK4[20]. P16and CDk4 loss stimulated cell proliferation, and accelerated cell cycle progression in vitro and in vivo[34,35]. Consequently, Zhang et al. indicated the hight expression of UHRF1 may promote cell proliferation in MB[20]. What's more, Tu et al.[28] observed knockdown of UHRF1 increase the G1 phase was while decrease the G2 phase was in lung adenocarcinoma(ADC)cell and induces cell apoptosis. Hence, the up-regulated UHRF1 in lung ADC is accompanied by the poor overall survival.
Accumulated studies suggest that the abnormal regulation of DNA and histone methylation by UHRF1 is tightly associated with the uncontrolled cell proliferation. UHRF1 was overexpressed in ESCC tissues and was an independent prognostic factor for ESCC patients[18]. In ESCC cells, knockdown of UHRF1 caused global DNA hypomethylation, inhibited cell proliferation, induced cell cycle arrest at the G2/M phase, accompanied by activation of Wee1 and DNA damage response pathway, and eventually induced apoptosis[18]. The UHRF1-DNMT1 axis plays a key role in DNA maintenance methylation in mammals. Accumulative studies demonstrate that DNMT1 is broadly overexpressed in cancers, which contributes to cancer cell proliferation and tumorigenesis[36]. Li[37] provided evidence that multiple transcription factors including E2F1 and SP1 mediate the activation of UHRF1-DNMT1 axis by the activated MEK/ERK pathway, which encourages cancer cells to proliferate.
Cancer metabolism plays an underlying role in driving cancer cell phenotype during tumor progression and promoting cancer aggressiveness[38]. UHRF1 plays a key role in cell metabolism by regulating glycolysis[39]. Studies have been reported that UHRF1 is markedly elevated in various cancer cells, suggesting that UHRF1 may affect cell proliferation by regulating cell metabolism[40,41]. Hu et al.[42] demonstrated that silencing UHRF1 significantly inhibited aerobic glycolysis, cell proliferation and tumor growth in pancreatic cancer cells. Furthermore, by knocking down UHRF1 in pancreatic cancer cells, decreased hypoxia inducible factor (HIF)1α levels、HIF1α targeted glycolytic genes and inhibit Sirtuin expression , which could negatively regulate aerobic glycolysis and suppress cell proliferation and tumor growth[42].
UHRF1 could function as an E3 ubiquitin ligase and has ubiquitination function. The ubiquitination modification of histones can also influence the methylation process of deoxyribonucleic acid. The investigators found that the ubiquitination of UHRF1 at the histoneH3lysine23 (H3K23) site in histones coupled the methylation of DNA to the DNA replication process[43]. The RING domain of UHRF1 can be monoubiquitinated to modify H3K23 in the S-phase of the cell cycle, which is essential for target recruitment of DNMT1 to UHRF1[43]. As previously described that the UHRF1-DNMT1 axis plays a key role in the maintenance of DNA methylation. Zhou et al.[46]data showed that UHRF1 down-regulated expressed in osteogenic differentiation, but overexpressed in osteosarcoma. Moreover, UHRF1 has been shown to exist in nuclear chromatin, bind to transcription-related proteins, and mediate protein degradation through ubiquitination, regulate DNA repair and cell cycle and promote proliferation of osteosarcoma. The researchers detected that UHRF1 plays an epigenetic regulatory role in tumor angiogenesis[47]. Achour et al.[48]confirmed that UHRF1/DNMT1 complex was associated with vascular endothelial growth factor (VEGF) in Jurkat cell line and immortalized human vascular smooth muscle fine cell line (HVTS-SM1), and verified that UHRF1 and DNMT1 exist in the same macromolecular complex in cells. Concurrently,Down-regulation of UHRF1 and DNMT1 can increase the expression level of p16INK4A and decrease the expression level of VEGF. However, p16INK4A is a negative regulator of VEGF. Therefore, It was concluded that UHRF1/DNMT1 complex could promote the regeneration of tumor angiogenesis by down-regulating p16INK4A to enhance the expression of VEGF gene. Wang et al.[49] proved that UHRF1 could negatively regulate the expression of p16INK4A in colorectal cancer tissues. The expression level of p16INK4A in colorectal cancer cell lines DLD1 and SW620 was significantly elevated after transfecting with Lenti-ShuHRF1 to knockdown UHRF1. It was inferred that UHRF1 may promote tumor angiogenesis by mediating P16INK4A-VEGF pathway. Wu et al.[50] determined the effects of UHRF1 overexpression on osteosarcoma cell proliferation, migration, invasion, angiogenesis, and metastasis. They further found that UHRF1 overexpression induced angiogenesis by suppressing AMPK activation and Semaphorin 3E (SEMA3E) expression.
Recent studies have been also reported that the overexpression of UHRF1 can induce tumor cell proliferation and metastasis through mediating certain signaling pathways. UHRF1 could inhibit the role of AMPK activity in cells by recruiting phosphatase PP2A to dephosphorylate AMPK T-172[39]. AMPK has been reported to inhibit mTOR signaling. Hence, the upregulated UHRF1 could reduced AMPK activity, and relieve the inhibition of AMPK in the mTOR signaling pathway, thus promoting the growth of hepatocellular carcinoma cells [45,46]. Zhu et al.[14] found that UHRF1 is overexpressed in intrahepatic cholangiocarcinoma(ICC) tissues. The downregulated UHRF1 attenuated the transition of the G1/S cell cycle and then suppressed cell proliferation and tumor growth. Further, Zhu et al.[14] showed that Mir-124-3p, an upstream regulator of UHRF1 expression, inhibited UHRF1 expression. The elevated Mir-124-3p suppressed cell proliferation and led to the arrest of the cell cycle. Together, these findings demonstrate that UHRF1, regulated by miR-124-3p, acts as a tumor promoter by promoting cell proliferation in ICC. Kuang et al.[8] suggested that the overexpression of UHRF1 promoted the metastasis of papillary thyroid cancer cells, and the inhibition of UHRF1 decreased the metastasis of anaplastic thyroid cancer cells, and the potential signaling pathway might be that UHRF1 directly bound and activated the transcription factor c-Jun/AP-1 in the nucleus, increasing the transcription of IL-6 and MIF.
The essential process of cancer metastasis consist of cell motility, migration and invasion. Tumor metastasis begins with tumor migration and invasion through endothelial barriers, a process known as EMT that is characterized by loss of cell-cell adhesion and increased cell motility[51]. An essential hallmark of EMT is the loss of ecadherin expression, which has been identified in multitudinous malignancies and is associated with increased metastatic potential[32,52]. Accumulated studies suggest that UHRF1 plays an important role in promoting EMT in tumor cells. Liu et al.[53] datas demonstrated that UHRF1 promotes osteosarcoma cell invasion by downregulating the expression of eccadherin and increasing EMT in an Rb1-dependent manner. It has been found that the deficiency of UHRF1 triggered the upregulation of CXCR4 in hepatocellular carcinoma, activated AKT and JNK to increase the expression and secretion of IL-6. Furthermore, IL-6 readily activated the JAK/STAT3/Snail signaling axis, which subsequently contributed to EMT[54].
Furthermore, the overexpression of UHRF1 has been reported to be associated with inhibition of apoptosis in multiple malignancies[6,55-58]. Kim et al.[57] presented experimental evidences documenting that UHRF1 downmodulation can enhances antitumor effects of retinoblastoma (RB) cells to histone deacetylase (HDAC) inhibitors by augmenting oxidative stress-mediated apoptosis via downregulation of glutathione S-transferase α4(GSTA4) and thioredoxin-2(TXN2). Knockdown of UHRF1 expression can enhance caspase-3 expression and activation and promote apoptosis to inhibit intestinal tumogenesis[58]. Zhang et al.[59] demonstrated that Overexpression of UHRF1 down-regulated UbcH8 by reducing the expression of UBE2L6, and thereby inhibited apoptosis in cervical cancer cells. Thus, UHRF1 has a vital role in inhibiting cellular apoptosis. Nf-κb is an important nuclear transcription factor. Nf-kappab proteins have been reported to be modulators of the inflammatory, proliferative, and apoptotic responses of a cell to a very large number of different stimuli[60]. The transcription factor NF-κB could inhibit apoptosis and to induce drug resistance in cancer cells[61]. Hence, NF-κB has become a molecular target of some antitumor drugs[62]. Significantly, it has been reported that the overexpression of UHRF1 activates PI3K/NF-κB signaling pathway by inhibiting the mRNA expression level of PI3K/NF-κB signaling pathway mediators, which accelerates the development of colon cancer[63]. Boukhari[64] investigated that activation of CD47 in human astrocytoma cell lines U87, up-regulated the expression of UHRF1 with subsequent activated NF-κB-dependent mechanism, thereby promoting cell proliferation.
Up to now, there is no meta-analysis to evaluate of the relation between the expression of UHRF1 and he prognostic value of UHRF1 in patients with tumors. Based on the results of meta-analysis, we have found that the high expression levels UHRF1 were associated with poor OS and DFS. The results of GEPIA analysis showed UHRF1 was overexpressed in multitudinous malignant tumors, and the higher the expression level of UHRF1 in tumor tissues, the worse the prognosis of cancer patients. Similarly, the KMplot analysis results showed that the OS was significantly shortened and the DFS rate was significantly reduced in patients with malignant tumors with increased UHRF1 expression. Furthermore, the overexpression of UHRF1 protein is more common in cancer patients with poor clinicopathological parameters.These results suggested that the expression of UHRF1 protein can be deemed as a valuable prognostic marker in patients with malignant tumors.
Noteworthy is the fact that this present study had some limitations. First of all, only English language reports have been considered, so we might have missed important studies published in other languages. All but one of the included studies were from Britain and the rest were from China. Thus, the results may only reflect the clinical characteristics of cancer patients in china. Second, in consideration of the relatively small sample sizes in the literature, it is still imperative to assess DFS and PFS via increasing the sample size in specific tumors. Potential publication bias is very likely to exist, despite the lack of evidence from our statistical tests. In the end, most of the data from the articles are evaluated the UHRF1 overexpression in immunohistochemistry, however, one[21] of them did not perform by immunohistochemistry. This may result in inconsistencies when we perform meta-analysis. Therefore, it is necessary to increase the number of highquality studies that contain a large number of samples to avoid the various factors in the compound. Also, the pooled analysis of some clinicopathological characteristics hinted exist heterogeneity.