PCBP1 depletion promotes tumorigenesis through attenuation of p27Kip1 mRNA stability and translation

Abstract

BackgroundPoly C Binding Protein 1 (PCBP1) is an RNA-binding protein that binds and regulates translational activity of subsets of cellular mRNAs. Depletion of PCBP1 is implicated in various carcinomas, but the underlying mechanism in tumorigenesis remains elusive.MethodsWe performed a transcriptome-wide screen to identify novel bounding mRNA of PCBP1. The bind regions between PCBP1 with target mRNA were investigated by using point mutation and luciferase assay. Cell proliferation, cell cycle, tumorigenesis and cell apoptosis were also evaluated in ovary and colon cancer cell lines. The mechanism that PCBP1 affects p27 was analyzed by mRNA stability and ribosome profiling assays. We analyzed PCBP1 and p27 expression in ovary, colon and renal tumor samples and adjacent non-tumor tissues using RT-PCR, Western Blotting and immunohistochemistry. The prognostic significance of PCBP1 and p27 also analyzed using online databases.ResultsWe identified cell cycle inhibitor p27Kip1 (p27) as a novel PCBP1-bound transcript. We then demonstrated that binding of PCBP1 to p27 3’UTR via its KH1 domain mainly stabilizes p27 mRNA, while enhances its translation to fuel p27 expression, prior to p27 protein degradation. The upregulated p27 consequently inhibits cell proliferation, cell cycle progression and tumorigenesis, whereas promotes cell apoptosis under paclitaxel treatment. Conversely, knockdown of PCBP1 in turn compromises p27 mRNA stability, leading to lower p27 level and tumorigenesis in vivo. Moreover, forced depletion of p27 counteracts the tumor suppressive ability of PCBP1 in the same PCBP1 over-expressing cells. Physiologically, we showed that decreases of both p27 mRNA and its protein expressions are well correlated to PCBP1 depletion in ovary, colon and renal tumor samples, independent of the p27 ubiquitin ligase Skp2 level. Correlation of PCBP1 with p27 is also found in the tamoxifen, doxorubincin and lapatinib resistant breast cancer cells of GEO database.ConclusionOur results thereby indicate that loss of PCBP1 expression firstly attenuates p27 expression at post-transcriptional level, and subsequently promotes carcinogenesis. PCBP1 could be used as a diagnostic marker to cancer patients.Electronic supplementary materialThe online version of this article (10.1186/s13046-018-0840-1) contains supplementary material, which is available to authorized users.

hnRNP E1 at the crossroads of translational regulation of epithelial-mesenchymal transition

“…The function of the PCBP1 gene encoding the hnRNP E1 protein was first demonstrated through characterization of its role as a negative regulator of alternative splicing. Since then, many additional roles have also been discovered, and most of them appear to have critical participation in the maintenance of cell phenotype integrity [34,39,67] . At the molecular level, the ability of hnRNP E1 to specifically bind to mRNA species often leads to a direct or indirect regulation of their translation.…”

“…For instance, the binding of hnRNP E1 to p27 kip1 3' UTR via its KH1 domain stabilizes p27 kip1 mRNA, fueling p27 kip1 protein expression by enhancing its translation prior to degradation. The upregulated p27 kip1 protein consequently inhibits cell proliferation, cell cycle progression, and tumorigenesis, and can concurrently promote cell apoptosis under paclitaxel treatment [39] . Interestingly, other cyclin-dependent kinase inhibitors such as p21 Waf1 are also regulated through hnRNP E1 mediated mRNA stability.…”

Explaining decisions of Graph Convolutional Neural Networks: patient-specific molecular subnetworks responsible for metastasis prediction in breast cancer

“…Next to the common genes found in all four networks, each network was characterized by several special, cancer-associated genes which are of high interest because they might represent patient-specific central signaling nodes and therapeutic vulnerabilities. Some examples are PTPN11 that is known to activate a transcriptional program associated with cancer stem cells or the EMT-related genes SOX4 or VIM that might be responsible for the high invasive capacity of the tumors and their early metastasis formation (Bentires-Alj et al, 2004;Aceto et al, 2012;Sharma et al, 2019;Zhang et al, 2012) Interestingly, the network of the metastatic patient GSM615233 harbored the genes FABP4 and LPL which both have been shown to interact with CD36, another highly expressed node in the network, to support cell proliferation and counteract apoptosis (Guaita-Esteruelas et al, 2016;Liang et al, 2018;Kuemmerle et al, 2011) In contrast, in the non-metastatic patient GSM150990 especially the interleukin receptor IL6ST and the Ras GTPase-activating protein 1 (RASA1) seem to be interesting because for both high expression levels have been linked with a favorable prognosis (Liu et al, 2014;Mathe et al, 2015) In the other non-metastatic patient GSM615695 high levels of HMGN2 and PCBP1 were identified which both have been shown to be able to inhibit cell proliferation (Shi et al, 2018;Fan et al, 2018) Although the experimental validation for the networks is still missing, it is tempting to speculate that these genes might contribute to the benign phenotype of the tumor in these patients.…”

“…The comparison of the subnetworks of the non-metastatic and the metastatic patients furthermore revealed some patient-specific genes which might give valuable information about specific mechanisms of tumorigenesis and therapeutic vulnerabilities in the respective patient. In general, it seemed that the subnetworks of the non-metastatic patients contained more genes that have been linked to better prognostic outcomes such as JUP, PCBP1 and HMGN2 in GSM615695 (Bailey et al, 2012;Shi et al, 2018;Fan et al, 2018) or RASA1, IL6ST, KRT19 and RPS14 in GSM150990 (Liu et al, 2014;Mathe et al, 2015;Saha et al, 2018;Zhou et al, 2013) while the networks of both metastatic patients harbored genes that are known to be involved in aggressive tumor growth or therapy resistance which might explain the early metastatic spread in these patients. Some examples are CDK1, SFN and XPO1 in GSM519217 (Alexandrou et al, 2019;Neve et al, 2006;Taylor et al, 2019) or CAV1, PTPN11 and FTL in GSM615233 (Qian et al, 2019;Aceto et al, 2012;Chekhun et al, 2013) However, not only the presence of specific genes might be important, but also their overall expression level.…”

Novel Role of Heterogeneous Nuclear Ribonucleoprotein E1 in Regulation of Apoptosis and Autophagy by a Triazole Derivative in Vascular Endothelial Cells

“…Over the last two decades researches have demonstrated the multifunction of hnRNP E1 plays important roles in maintaining cellular homeostasis during different physiological processes. HnRNP E1 has been proven to promote cell apoptosis in several cancer cell lines [20][21][22]. Recently, a research group reported that hnRNP E1 repressed autophagy through downregulation of LC3B to promote tumor cell apoptosis [23].…”