- Open Access
The oncogenic role of circPVT1 in head and neck squamous cell carcinoma is mediated through the mutant p53/YAP/TEAD transcription-competent complex
- Lorena Verduci†1,
- Maria Ferraiuolo†1, 2,
- Andrea Sacconi1,
- Federica Ganci1,
- Jlenia Vitale1,
- Teresa Colombo3,
- Paola Paci4,
- Sabrina Strano2,
- Giuseppe Macino5,
- Nikolaus Rajewsky6 and
- Giovanni Blandino1Email author
© The Author(s). 2017
- Received: 10 June 2017
- Accepted: 1 December 2017
- Published: 20 December 2017
Circular RNAs are a class of endogenous RNAs with various functions in eukaryotic cells. Worthy of note, circular RNAs play a critical role in cancer. Currently, nothing is known about their role in head and neck squamous cell carcinoma (HNSCC). The identification of circular RNAs in HNSCC might become useful for diagnostic and therapeutic strategies in HNSCC.
Using samples from 115 HNSCC patients, we find that circPVT1 is over-expressed in tumors compared to matched non-tumoral tissues, with particular enrichment in patients with TP53 mutations. circPVT1 up- and down-regulation determine, respectively, an increase and a reduction of the malignant phenotype in HNSCC cell lines. We show that circPVT1 expression is transcriptionally enhanced by the mut-p53/YAP/TEAD complex. circPVT1 acts as an oncogene modulating the expression of miR-497-5p and genes involved in the control of cell proliferation.
This study shows the oncogenic role of circPVT1 in HNSCC, extending current knowledge about the role of circular RNAs in cancer.
Circular RNAs (circRNAs) are a class of endogenous RNAs that have been known for more than a decade . Original hypotheses on circRNAs ascribed them the role of plant viroids  and Hepatitis delta virus molecules  or considered them the result of transcriptional noise . More recently, circRNAs have emerged for their potential functions in the regulation of gene expression .
CircRNAs are mostly generated from coding or noncoding exons, but they can also derive from intronic, antisense, 5′ or 3′ untranslated, and intergenic genomic regions . Very high and conserved circRNA expression is typically found in neural tissues , muscle, and some hematopoietic cells, for hundreds of circRNAs [8, 9].
CircRNAs comprised of exonic sequences are produced by a poorly characterized mechanism called “back-splicing”, where a downstream 5′ splice site is joined to an upstream 3′ splice site, involving single or multiple exons [5, 10–14]. Due to lack of accessible ends, circRNAs are resistant to exonucleases and, as a result, they are more stable than linear RNA isoforms. Despite advancements in the study of circRNAs, their function in eukaryotes is not clear. According to the recent literature, they may regulate alternative splicing [10, 14], bind and sequester RNA-binding proteins and ribonucleoprotein complexes [6, 15], be translated, bind in trans to other RNA sequences , or regulate miRNA expression [6, 15].
Of particular interest is the recently discovered role of circRNAs in cancer [16–18]. In the same line, our work investigates the role of a human circRNA, circPVT1, in head and neck squamous cell carcinoma (HNSCC). CircPVT1 was first identified as circ6 by Memczak et al.  and then named circPVT1 after its host gene PVT1 in subsequent work [19, 20]. The PVT1 gene is frequently up-regulated in many types of cancers, including HNSCC [21–25]. The circPVT1 locus is embedded in the long non-coding RNA PVT1 and it originates from exon 2 of the PVT1 gene (human genome GRch38/hg38).
HNSCC is the sixth leading cancer by incidence worldwide and the eighth most common cause of cancer death [26, 27]. Although in the past two decades new surgical and medical treatments have improved the quality of life of patients [28–30], the 5-year survival rate is achieved by only 40–50% of patients .
We started our study investigating the oncogenic role of circPVT1 in HNSCC using a robust collection of human tissue samples. circPVT1 was found significantly up-regulated in tumors compared with matched non-tumoral tissues. More importantly, we have discovered that circPVT1 expression was enriched in tumors carrying mutant p53 proteins (mut-p53). Genomic data have shown that p53 is the most frequent mutated gene in HNSCC; indeed it is mutated in up to 85% of HNSCC cases and these involve mainly exons 5–8 [31–34]. We recently reported that mut-p53 cooperates with the transcriptional co-factor YAP (Yes-Associated Protein) in breast cancer cell lines . YAP as an oncogene acts as an effector of the Hippo pathway, playing a critical role in the initiation and progression of several human cancers, including HNSCC [36–39]. YAP and mut-p53 proteins are able to physically interact and share a common set of transcriptional programs in cancer . In our study, we found that the circPVT1 was regulated through the mut-p53/YAP/TEAD complex via its regulatory region. Moreover, our data show that circPVT1 was able to regulate its own expression through binding YAP. To date, the role of circRNAs in HNSCC is unexplored. Collectively, these findings mirror a novel alteration in the circRNA network that might contribute to the fine deciphering of the tumorigenesis occurring in mut-p53 HNSSC patients.
circPVT1 is up-regulated in HNSCC patients with TP53 mutations
Previous studies have shown that PVT1 resides in the well-known cancer risk region 8q24 and is amplified in HNSCC [21–25]. To analyze in detail the PVT1 amplification, we used the HNSCC cancer data set provided by The Cancer Genome Atlas (TCGA) .
circPVT1 was shown to be up-regulated in tumor samples (Fig. 1a, c). Interestingly, circPVT1 was significantly up-regulated in tumors carrying TP53 mutations but not in those with intact TP53 (Fig. 1b, d). This might suggest that the presence of the mut-p53 protein accounts for circPVT1 up-regulation observed in HNSSC patients. To further investigate the specificity of the association between TP53 mutations and circPVT1 amplification we analyzed the PVT1 amplification in relation to the FAT1 and CDKN2A mutations, the second (22.46%) and the third (22.07%) most frequent gene mutated in HNSCC patients according to TCGA. Focusing on the PVT1 chromosome interval representing circPVT1, we did not observe any difference between tumor samples with either FAT1 or CDKN2A mutations and those with an intact FAT1 or CDKN2A gene, which were both up-regulated in comparison to non-tumoral samples (Additional file 1: Figure S1a, b). This might further support that the circPVT1 amplification observed in HNSCC with TP53 mutations is strictly connected to the mut-p53 status and is not generally related to any cancer mutations.
To further validate our bioinformatic approach we selected two other circRNAs deregulated in cancer, SMG7 and RPN2 . Using the HNSCC cancer data, we found both circRNAs were up-regulated in the tumors compared to non-tumoral samples (Fig. 1e, g). Unlike circPVT1, we did not observe any specific up-regulation of SMG7 and RPN2 when comparing TP53 mutated versus wild-type p53 tumors (Fig. 1f, h). This indicated that TP53 mutations did not affect the deregulation of both SMG7 and RPN2 in HNSCC.
In previous work , we reported the incidence of p53 mutations in HNSCC patients used for this study, determined by direct sequencing of p53 exons 2 through 11. Of the original 121 patients used in , we studied here 114, adding one patient not included in , whose clinical characteristics are shown in Additional file 2: Table S1. In this work, 67 out of the 115 patient samples used (58%) exhibited single or multiple p53 mutations in the tumor tissue .
In order to determine whether the up-regulation of circPVT1 correlated with the TP53 mutations, we compared the circPVT1 expression of mut-p53 and wild-type tumoral samples. As shown in Fig. 2b, there was a statistically significant up-regulation of circPVT1 in mut-p53 compared to wild-type samples.
Furthermore, we compared circPVT1 expression between tumoral and non-tumoral samples, considering separately the mut-p53 and wild-type patients. We found that circPVT1 was significantly up-regulated only in patients with TP53 mutations, confirming the correlation between mut-p53 and circPVT1 in HNSCC (Fig. 2c, d).
The identity of circPVT1 was also analyzed by Sanger-sequencing following RT-qPCR reactions in a subgroup of patients. We selected 20 tumoral samples, including 10 wild-type and 10 mut-p53 samples. For all patients selected, we confirmed the identity of circPVT1 as shown in Additional file 1: Supplementary results 1. The list of patients ordered on the basis of circPVT1 expression and p53 status is included in Additional file 2: Table S2.
Starting from the demographic and clinical characteristics of the patients, we used a univariate linear regression analysis to test the association of circPVT1 with prognostically relevant risk factors. Considering tumor samples, we observed a significant correlation between circPVT1 and mut-p53 (Additional file 2: Table S3). This result correlates with the circPVT1 up-regulation seen in patients with TP53 mutations, as shown above. Among other factors analyzed, we also found a correlation between circPVT1 and alcohol use, with a clear tendency towards statistical significance (Additional file 2: Table S3).
We then evaluated the impact of circPVT1 on the outcomes of patients. Using the HNSCC cancer data set and our collection of HNSCC samples, we determined that patients with high circPVT1 levels had a poorer overall survival in comparison to those exhibiting low circPVT1 levels (Fig. 2e, f). Multivariable analysis confirmed that high circPVT1 levels were associated with reduced overall survival and that such association was dependent on TP53 mutations in both sample populations (Fig. 2e, f).
circPVT1 cellular localization in HNSCC cell lines
Modulation of mut-p53 expression affects circPVT1 but not PVT1 expression
We also investigated the existence of an inverse relationship between circPVT1 and mut-p53, i.e., whether modified circPVT1 levels influence p53 expression. The p53 protein level was not affected by circPVT1 down-regulation, therefore excluding the inverse regulation (Additional file 1: Figure S2d).
We also showed that the reduction of p53 affected circPVT1 level in Detroit 562 cells. Again, siRNAp53, siRNAp53 3′ UTR, and siRNAp53 smart pool significantly reduced circPVT1 expression with no effect on PVT1 expression (Additional file 1: Figure S2e–g). As seen before, also in the Detroit 562 cell line, modified circPVT1 levels had no influence on p53 expression (data not shown).
A specific DNA binding consensus for mut-p53 protein has not been characterized so far. For this reason, we focused first on possible post-transcriptional regulation of mut-p53 on circPVT1, investigating if mut-p53 was able to bind the mature form of circPVT1. Using RNA immunoprecipitation (RIP), we evaluated the circPVT1 level in CAL27 cells after p53 immunoprecipitation using input and IgG as controls. No circPVT1 signal was detected by RT-qPCR after p53 immunoprecipitation, showing that p53 did not influence circPVT1 expression through direct binding (Additional file 1: Figure S2h).
Finally, to confirm the absence of p53 binding to circPVT1, we co-expressed mut-p53 and circPVT1 in H1299 cells, a p53-devoid human non-small cell lung carcinoma cell line (p53 null). The two vectors expressing circPVT1 were generated according to the methods described in [15, 41] (Additional file 1: Figure S2i). After the co-expression of mut-p53 and circPVT1 in H1299 cells (Additional file 1: Figure S2j-l) we performed p53 immunoprecipitation (Additional file 1: Figure S2m) and evaluated the circPVT1 level by RT-qPCR. The circPVT1 signal obtained was similar to that of the housekeeping gene GAPDH, showing the absence of direct binding of p53 to circPVT1 (Additional file 1: Figure S2n, o). Moreover, we measured the expression of four different intronic regions upstream of circPVT1, retained in the RNA population, putatively involved in its circularization and indicated as circPVT1_UP1, circPVT1_UP2, circPVT1_UP3, and circPVT1_UP4. We did not observe any significant signal, showing that p53 did not bind these regions (Additional file 1: Figure S2p-s).
circPVT1 expression is regulated through the mut-p53/YAP/TEAD complex on its own promoter
YAP and its paralogue TAZ (Tafazzin) are the major downstream effectors of the Hippo pathway, and TEAD family proteins (TEA Domain Family Member 1) mainly mediate their biological functions [42–44]. In previous work, we showed that YAP physically interacts with mut-p53 proteins, enhancing their pro-proliferative transcriptional activity . In this role, YAP was shown to have distinct functions from its paralog TAZ in relation to the oncogenic pathway involved. In fact, YAP depletion reduced the expression of cell cycle genes regulated by mut-p53 , while TAZ depletion did not. Since YAP has been shown to be an important co-factor of mut-p53 in cancer, we investigated the impact of the down-regulation of YAP and its partners, TAZ and the TEAD family proteins, on the expression level of circPVT1 and PVT1 (Fig. 4d–f; Additional file 1: Figure S3a–c).
YAP affected circPVT1 expression but not that of PVT1, while TAZ and the TEAD family proteins affected the expression of both circPVT1 and PVT1, in the CAL27 cell line (Fig. 4g). These data indicate that mut-p53 and YAP specifically regulated circPVT1 while TAZ and TEAD regulated both circPVT1 and its host gene PVT1.
To further verify the impact of YAP down-regulation on circPVT1 expression, we used another siRNA against YAP, siRNA-YAP2 (Additional file 1: Figure S3d). We obtained a statistically significant down-regulation of circPVT1 expression at 24 and 48 h, whereas no PVT1 down-regulation was observed (Additional file 1: Figure S3e, f).
Reduction of YAP using both siRNA-YAP and siRNA-YAP2 (Additional file 1: Figure S3g) affected circPVT1 expression also in the Detroit 562 head and neck cell line with no effect on the PVT1 expression level (Additional file 1: Figure S3h, i).
To assess whether mut-p53 and YAP finely regulate the circPVT1 expression level, we performed a rescue experiment in the A253 cell line, a human submandibular gland cell line (p53 null). After mut-p53 overexpression (mutp53R175H) in A253 cells, we observed an increase in the circPVT1 level (Fig. 4h, j). On the contrary, circPVT1 expression was reduced as a consequence of YAP down-regulation (Fig. 4i, j). As expected, circPVT1 expression was restored to control levels when YAP down-regulation was performed concomitantly with mut-p53 over-expression (Fig. 4j). These data show that mut-p53 and YAP worked together to regulate circPVT1 expression levels. No modulation of the PVT1 expression level was observed under these experimental conditions (Fig. 4k).
In order to thoroughly determine how circPVT1 is transcriptionally regulated, we used the metabolic tagging of newly transcribed RNAs by 4-thiouridine (4sU) after 5,6-dichlorobenzimidazole 1-β-D-ribofuranoside (DRB) treatment to monitor the nascent circRNA following the depletion of mut-p53, YAP, and TEAD1 proteins (Fig. 5m). We found that reduced expression of YAP, mut-p53, and TEAD1 decreased nascent circPVT1 expression when compared to control cells (Fig. 5n). The concomitant analysis of PVT1 expression confirmed that TEAD1 impacted on expression of both circPVT1 and its host gene PVT1, while a slight effect was observed on PVT1 transcription upon depletion of YAP and mut-p53 protein expression (Fig. 5o).
In aggregate, these data show that the newly identified circPVT1 promoter is a transcriptionally active region, distinct from the promoter of its host gene PVT1, and regulated by the mut-p53/YAP/TEAD complex. The PVT1 promoter also exhibited transcriptional control of circPVT1 expression.
circPVT1 autoregulates its own expression via YAP binding
To dissect further the role of YAP in the regulation of circPVT1 at the post-transcriptional level, and the role of mut-p53 in this regulation, we performed a RIP assay after the transfection of siRNA-YAP and siRNAp53 separately, in CAL27 cells. The down-regulation of YAP and mut-p53 expression was confirmed as shown in Additional file 1: Figure S4e–g. The immunoprecipitation of YAP after the down-regulation of YAP and mut-p53 was confirmed as shown in Additional file 1: Figure S4h. Upon YAP down-regulation the binding to circPVT1 was lost (Fig. 6g, h). Moreover, YAP did not bind circPVT1 also as a consequence of mut-p53 down-regulation (Fig. 6g, h). These results highlight the role of mut-p53 protein in the stabilization of the YAP/circPVT1 complex. We also tested the binding of YAP to the four intronic regions, circPVT1_UP1, circPVT1_UP2, circPVT1_UP3, and circPVT1_UP4. We found that the two closest regions to the circPVT1 transcription start site, circPVT1_UP1 and circPVT1_UP2, were involved in YAP binding and were affected by YAP down-regulation (Fig. 6i, j). Although to a lesser extent than that of circPVT1, these regions were also affected by p53 down-regulation (Fig. 6i, j). The other two regions included in the analysis, circPVT1_UP3 and circPVT1_UP4, were not affected by YAP or p53 down-regulation (Fig. 6k, l).
RIP assays also revealed that the nuclear co-factor YAP bound to the mature circPVT1 and that mut-p53 had a role in the stabilization of the YAP/circPVT1 complex. These data show the presence of an operating circPVT1 in the nucleus, suggesting the capability to regulate its own expression.
circPVT1 modulation affects the oncogenic phenotype of HNSCC cell lines
We further demonstrated the specificity of the siRNAs against circPVT1 using four mismatched control siRNAs, generated according to the methods described in . Two mismatched control siRNAs were designed from the sequence of siRNAcircPVT1, designated siRNAct1 and siRNAct2; the other two mismatched control siRNAs were designed from the sequence of siRNAb-circPVT1, designated siRNAb-ct1 and siRNAb-ct2. We tested these siRNAs in CAL27 and Detroit 562 cells. siRNAct1 and siRNAct2 had no effect on the circPVT1 expression level in comparison to siRNAcircPVT1, which had an effect in both CAL27 and Detroit 562 cells (Additional file 1: Figure S5b, c). Similar results were obtained for siRNAb-ct1 and siRNAb-ct2 when these siRNAs were compared to siRNAb-circPVT1 (Additional file 1: Figure S5d, e). We conclude that the mismatched control siRNAs behaved similarly to siRNAct in CAL27 and Detroit 562 cell lines.
Down-regulation of circPVT1 significantly reduced the proliferation rate of CAL27 cells, determined by cell counting at different time points (Fig. 8b; Additional file 1: Figure S6a). To better understand the regulation of CAL27 cell proliferation by circPVT1, we examined the cell cycle profile. As shown in Fig. 8c and in Additional file 1: Figure S6b, down-regulation of circPVT1 expression led to a significant decrease in the cell population in S phase and a significant increase in the cell population in G2 phase. An inhibitory effect on proliferation after circPVT1 down-regulation was also observed in colony-forming assay (Fig. 8d; Additional file 1: Figure S6c). Interestingly, circPVT1 down-regulation did not affect the migration of CAL27 cells (Fig. 8e; Additional file 1: Figure S6d). To better investigate the specific effect of circPVT1 in determining the phenotype of CAL27 cells, we reduced the level of its host gene PVT1 using a siRNA specific for the long noncoding RNA. siRNA-PVT1 decreased PVT1 expression by about 65% (Additional file 1: Figure S6e) at 48 h after transfection, but it did not affect the circPVT1 expression level (Additional file 1: Figure S6f). The siRNA-PVT1 effect was compared to the siRNAct control used above. Down-regulation of PVT1 did not impact on the proliferation rate of CAL27 cells, determined by cell counting and cell cycle analysis (Additional file 1: Figure S6g, h). Moreover, we did not obtain any statistically significant difference between siRNA-PVT1 and siRNAct with regard to the ability of CAL27 cells to form colonies (Additional file 1: Figure S6i). Finally, we evaluated the CAL27 cell line behavior in a migration assay upon down-regulation of PVT1. No specific role of PVT1 was observed in the migration of CAL27 cells (Additional file 1: Figure S6j).
The reduction of circPVT1 expression using both siRNA-circPVT1 and siRNAb-circPVT1 was also performed in Detroit 562 cells (Fig. 8a). The two siRNAs decreased circPVT1 expression by about 70 and 90%, respectively (Fig. 8a), compared to siRNAct. As seen in CAL27, the down-regulation of circPVT1 in Detroit 562 cells caused a significant reduction in their ability to form colonies compared to the control (Fig. 8f).
The phenotype following the down-regulation of circPVT1 was also observed in a human pharynx squamous cell carcinoma cell line (FaDu). The FaDu cell line has one missense mutation in codon 248 of exon 7 (mutp53R248L) and an additional heterozygous splicing site mutation of intron 6 with a frameshift and a premature stop codon, resulting in a truncated variant of p53. siRNA-circPVT1 and siRNAb-circPVT1 decreased circPVT1 expression in FaDu cells by about 70 and 90%, respectively, compared to siRNAct (Additional file 1: Figure S6k). After the down-regulation of circPVT1 in FaDu cells, their ability to form colonies was significantly reduced in comparison to the control (Additional file 1: Figure S6l).
Since cisplatin (CDDP) is one of the standard treatments for HNSCC patients, we assessed the potential positive effect of the down-regulation of circPVT1 expression on cisplatin-induced killing effects, using Detroit 562 cells. We found that circPVT1 down-regulation rendered Detroit 562 cells more prone to CDDP-induced killing effects (Fig. 8g, h). Both the half maximum effective concentration (EC50) and the half lethal concentration (LC50) were significantly reduced (Fig. 8i).
Finally, we studied the effect of circPVT1 over-expression on the CAL27 cell phenotype using the two vectors pcDNA3-circPVT1-a and pcDNA3-circPVT1-b (Additional file 1: Figure S7). circPVT1 over-expression with both vectors enhanced the proliferation rate of the CAL27 cells, determined by cell counting at different time points (Additional file 1: Figure S7a). Particularly, the vector pcDNA3-circPVT1-a had a stronger effect on CAL27 proliferation as confirmed by cell cycle analysis (Additional file 1: Figure S7a, b). It is also worth mentioning that circPVT1 over-expression by both vectors enhanced the capacity of CAL27 cells to form colonies (Additional file 1: Figure S7c).
mut-p53 down-regulation reduces the proliferation of HNSCC cell lines
To investigate in more detail the relationship between mut-p53 and circPVT1, we evaluated whether the CAL27 phenotype after mut-p53 down-regulation followed the phenotype observed after circPVT1 down-regulation. Down-regulation of p53 significantly decreased the proliferation rate of CAL27 cells, determined by cell counting at different time points (Additional file 1: Figure S8a). To elucidate the details of the cell proliferation reduction, we performed cell cycle analysis, observing a significant increase in the cell population in G1 phase and a decrease in the cell population in S and G2 phases (Additional file 1: Figure S8b). These results indicate that mut-p53 down-regulation suppresses cell growth, promoting the block in G1 phase of the cell cycle. As for circPVT1, we observed a reduction in the colony number in a colony-forming assay (Additional file 1: Figure S8c). In contrast to circPVT1, mut-p53 protein was also involved in migration. In Additional file 1: Figure S8d we show a reduction of cell migration after p53 down-regulation of about 30% compared to the control. Finally, we performed an experiment to test the possible additive effect on CAL27 phenotype after concomitant down-regulation of p53 and circPVT1. We observed an additive effect on the proliferation rate as determined by cell counting (Additional file 1: Figure S8e). No other additive effect was evidenced from the cell cycle analysis or colony assay compared to the single down-regulation of p53 and circPVT1 (Additional file 1: Figure S8f, g).
circPVT1 up-regulation pairs with the down-regulation of mut-p53-associated miR-497-5p
Among other putative circRNA functions, the current literature demonstrates that circRNAs have the potential to regulate miRNA expression [6, 15, 19]. Since circPVT1 expression was up-regulated in HNSCC, we focused our attention on the miRNAs down-regulated in our previous work .
Secondly, we measured the correlation between circPVT1 and these miRNAs by Spearman and Pearson’s correlation (Additional file 2: Table S4) using the data on circPVT1 expression obtained from RT-qPCR as shown above, and the data on the five down-regulated miRNAs obtained from the miRNA expression profiling . In detail, we analyzed 68 tumor samples with p53 mutations against 37 non-tumoral matched counterparts. We observed a significant negative relationship between circPVT1 and two of the miRNA selected: miR-99-5p and miR-497-5p (Additional file 2: Table S4). We measured the miRNA level in CAL27 cells after circPVT1 down-regulation using siRNAcircPVT1. In our in vitro model miR-99-5p was down-regulated after circPVT1 down-regulation (data not shown); in contrast, miR-139-3p, miR-370, and miR-1224-5p were undetectable by RT-qPCR due to their low expression levels. miR-497-5p was detectable by RT-qPCR and it was up-regulated after circPVT1 down-regulation, as expected from the data from patient analysis (Fig. 9b).
circPVT1 binds to and regulates miR-497-5p expression
To investigate the effect of circPVT1 activity on miR-497-5p expression, we used the two vectors expressing circPVT1, pcDNA3-circPVT1-a and pcDNA3-circPVT1-b. Upon circPVT1 overexpression we evaluated the miR-497-5p level by RT-qPCR. miR-497-5p expression decreased by about 70% after the transfection of pcDNA3-circPVT1-a, and by about 50% after the transfection of pcDNA3-circPVT1-b (Fig. 9c). To verify if the activity of circPVT1 on miRNA-497-5p was specific, we evaluated the expression of the other two molecules possibly involved in the circPVT1 mechanism of action, i.e., mut-p53 and PVT1. After pcDNA3-circPVT1-a or -b transfection, we did not observe any modification in mut-p53 or PVT1 expression levels (Fig. 9d, e). To confirm that the circPVT1 was a direct regulator of miR-497-5p, we searched for the miR-497-5p binding site on the circPVT1 sequence (Fig. 9f). Then, we deleted the miR-497-5p binding site in the pcDNA3-circPVT1-a vector, generating the vector pcDNA3-circPVT1-a-del. The pcDNA3-circPVT1-a-del and pcDNA3-circPVT1-a vectors induced circPVT1 expression to similar levels (Fig. 9g), and after the co-transfection with miR-497-5p, the circPVT1 expression was still comparable between the two vectors (Fig. 9g). miR-497-5p expression was reduced after the transfection of pcDNA3-circPVT1-a alone or in combination with miR-497-5p (Fig. 9h). We did not observe any change in miR-497-5p expression after the transfection of pcDNA3-circPVT1-a-del with or without miR-497-5p (Fig. 9h). These data show that circPVT1 affected miR-497-5p expression through the selected binding site, which was specific and required for the miR-497-5p regulation. Finally, the regulation of miR-497-5p by circPVT1 might explain the presence of circPVT1 in the cytoplasm as well as the nucleus.
miR-497-5p up-regulation in CAL27 cells mimics the phenotype induced by circPVT1 down-regulation
Aurka, mki67, and bub1 are genes involved in circPVT1 downstream oncogenic effects
CircRNAs have recently re-emerged as a class of endogenous RNAs with different roles in eukaryotic cells . Of particular interest is the emerging oncogenic function of circRNAs, which might make them candidates for new biomarkers and therapeutic targets in cancer. Here, we investigated the role of circPVT1, a member of the circRNA family, in the pathogenesis of HNSCC.
We first investigated circPVT1 expression using the HNSCC cancer data set provided by TCGA. circPVT1 was found to be up-regulated in tumors hosting mut-p53 in comparison to unmatched non-tumoral samples. Similar findings were also found using a well-characterized collection of 115 HNSCC samples in which each tumor sample was compared to its matched non-tumoral tissue, minimizing inter-individual variation . Indeed, circPVT1 expression determined by RT-qPCR was significantly up-regulated in tumor samples and in particular in those carrying TP53 mutations. This was not evidenced when circPVT1 expression was associated with FAT1 and CDKNA2 gene mutations in HNSCC, thereby mirroring circPVT1 as a non-coding mediator of mutant p53 oncogenic activities. Indeed, linear regression analysis showed a significant correlation between circPVT1 and mut-p53 in tumoral samples.
It is known that TP53 is the most frequent mutated gene in human cancers . Moreover, p53 missense mutations not only determine the loss of its tumor-suppressive functions, but can also generate new proteins with oncogenic activities [47–50]. TP53 mutations are associated with decreased survival rate and increased risk of locoregional recurrence in HNSCC . Using both the HNSCC cancer data set and our collection of HNSCC samples, we showed that circPVT1 also impacts on the survival of these patients. In particular, high circPVT1 levels are associated with a poorer overall survival in association with TP53 mutations. Moreover, circPVT1 plays a role in the resistance to cisplatin of HNSCC cell lines, but only in those carrying mutant p53 proteins. This further supports the role of circPVT1 in mut-p53-dependent pathways.
Our results show that circPVT1 is regulated by mut-p53 independent of its host gene PVT1. This mechanism appears to be unidirectional; in fact the down-regulation of circPVT1 did not influence mut-p53 protein expression. The analysis of the phenotype of HNSCC cell lines after down-regulation of either mut-p53 or circPVT1 expression revealed similar effects. We obtained in both cases a reduction of the malignant phenotype, showing that the two molecules work within the same molecular pathway. The role of circPVT1 as an oncogene was assessed also by circPVT1 overexpression. Cells over-expressing circPVT1 showed an increased capacity to proliferate and form colonies, further supporting the importance of circPVT1 in determining the oncogenic phenotype.
In an attempt to explain the link between mut-p53 and circPVT1 in tumors, we have established that YAP is a main regulator of circPVT1 expression at the transcriptional and post-transcriptional levels. Our data show that mut-p53 and TEAD are recruited at the same site where YAP binds the circPVT1 promoter, confirming the capacity of YAP and mut-p53 to interact with each other at promoter sites, and the important role of TEAD family transcription factors in mediating YAP-dependent gene expression [35, 43]. In fact, performing a DRB-4sU assay, we observed a decrease of nascent circPVT1 as a consequence of mut-p53, YAP, and TEAD1 down-regulation. YAP regulates circPVT1 also at the post-transcriptional level by binding the mature form of circular RNA, and mut-p53 acts as the stabilizer of the YAP/circPVT1 complex. In support of this model, we observed that when the mut-p53 protein was depleted, circPVT1 expression levels were reduced and the YAP/circPVT1 complex was lost.
Of note, we showed that circPVT1 uses its own promoter and not the promoter of its host gene. By ChIP assay experiments for p-Pol II, we showed that the newly identified circPVT1 promoter is a region that is transcriptionally active. Even if we did not exclude that the mut-p53/YAP/TEAD complex is also recruited at the PVT1 promoter level, our data support a mechanism according to which the mut-p53/YAP/TEAD complex works preferentially at the circPVT1 promoter level in the HNSCC context and in strict relation with mut-p53. It is reasonable to conclude that this status is achieved through the capacity of circPVT1 to regulate its own expression and to recruit preferentially the mut-p53/YAP/TEAD complex at its promoter region. In fact, we showed that the presence of circPVT1 in the nucleus is characterized by circPVT1 binding to the nuclear co-factor YAP. Moreover, by DRB-4sU and ChIP assay in circPVT1 over-expression conditions, we found, respectively, an increase in nascent circPVT1 production and an enrichment of Pol II and YAP recruitment at the circPVT1 promoter level.
We found that circPVT1 acts as an oncogene repressing the function of miRNA-497-5p, a miRNA associated with mut-p53 and that has been reported to have a tumor suppressor role in several cancers [51–54]. Here, we showed that miR-497-5p up-regulation significantly decreased the proliferation rate of CAL27 cells, which is consistent with its reported tumor suppressor function. Moreover, we originally demonstrated that circPVT1 is a direct regulator of miR-497-5p, impairing its tumor suppressor activity.
The circPVT1 downstream oncogenic effect included the up-regulation of aurka, mki67, and bub1, which are all putative targets of miRNA-497-5p and are involved in cell proliferation (Fig. 11h). We do not exclude that circPVT1 is also able to regulate other miRNAs. Indeed, we observed a significant negative relationship between circPVT1 and miR-99-5p (Additional file 2: Table S4), even if this relationship was not confirmed in our in vitro model. The study of other miRNAs regulated by circPVT1 will be the subject of future work.
Collectively, these findings depict a network in which gain of function mutant p53 proteins trigger the activation of a downstream non-coding effector, demonstrated here by the binding of circPVT1 and miR-497-5p, which leads to unrestrained proliferation through the aberrant enhancement of the expression of cell cycle regulated genes. The involvement of YAP and TEAD as critical components of the Hippo tumor suppressor pathway in the context of mutant p53 activity might represent an additional proof of the aberrant crosstalk between two distinct tumor suppressor pathways. Indeed, TP53 mutations lead to the production of mutant p53 proteins that engage physically with YAP and TEAD and might subvert the tumor suppressor into oncogenic activities.
We found that circPVT1 behaves as an oncogene in HNSCC and that the mut-p53/YAP/TEAD complex transcriptionally regulates its expression. Although further studies are necessary to thoroughly elucidate the molecular pathway in which circPVT1 is involved, these findings significantly advance our understanding of the circRNAs’ mechanism of action. Further to providing new insights into the biological function of circRNAs in cancer, our data might contribute to identify new candidate non-coding biomarkers such as circRNAs specifically associated with TP53 mutations in HNSCC. This might be useful for diagnostic and therapeutic strategies in HNSCC.
Study population and clinical samples
The study population and clinical samples are described in . Briefly, the study population includes 115 prospectively enrolled patients with histologically confirmed primary HNSCC undergoing curative treatment at the Otolaryngology Head and Neck Surgery Department. Only patients who did not receive any anticancer therapy before surgery were included in the study. Only HNSCC patients who developed local recurrence 1 month after surgery and with a follow-up ≥ 12 months were considered for the prognostic study. Two biopsies, from tumor and histologically normal tissue, were collected at surgery and preserved in RNA later (Ambion, Austin, TX, USA) from each patient. The histologically normal tissue was taken in correspondence with surgical resection margins and are described in the text as non-tumoral tissue.
P53 mutational analysis
The p53 mutational analysis is described in .
CAL27, FaDu, A253, and H1299 cell lines were maintained in RPMI-1640 medium (Invitrogen); the Detroit 562 cell line was maintained in DMEM (Invitrogen). The medium was supplemented with 10% fetal bovine serum (FBS; Invitrogen), penicillin (100 U/ml; GIBCO), and streptomycin (200 mg/ml; GIBCO) at 37 °C in a 5% CO2 atmosphere.
Quantification of circPVT1 molecules per cell
Finally, we divided the above result by the number of cells (5.88 × 106 for Detroit cells and 4.35 × 106 for CAL27 cells).
RNAse R treatment
The RNAse treatment was performed using Ribonuclease R (RNase R) according to the manufacturer’s recommendations (Epicentre). In particular, 2 μg of RNA were treated with 10 U of RNase R at 37 °C for 30 minutes in 10× Reaction Buffer. The untreated samples were incubated at the same conditions in 10× Reaction Buffer without the RNase R.
The transfections of siRNAs and mimicRNAs were performed using Lipofectamine RNAiMax (Life Technologies). The transfections of vectors were performed using Lipofectamine 2000 (Life Technologies). All experiments were conducted according to the manufacturer’s recommendations. For mature miR-497-5p expression, we used the mirVana miR-495-5p mimic (Ambion) at a final concentration of 5 nM and as control we used the mirVana miRNA mimic, Negative Control #1 (Ambion), at the same concentration. The transfection of vectors and mimicRNAs in the same experiment were performed using Lipofectamine RNAiMax (Life Technologies) for mimicRNAs and, 24 h after the first transfection, a second transfection for the vectors was performed using Lipofectamine 2000 (Life Technologies). The mut-p53 overexpression in A253 cells was performed using 5 μg pcDNA3-p53-R175H , and the same amount of pcDNA3 was used as control. For the H1299 RIP assay we used 5 μg pcDNA3-p53-R175H + 8 μg pcDNA3- circPVT1-a and 13 μg pcDNA3 vector as control. For the CAL27 RIP assay of overexpression, we used 5 μg YAP-Flag  + 8 μg pcDNA3-circPVT1-a and 13 μg pcDNA3 vector as control. For the CAL27 RIP assay in silencing conditions we used 320 pmol siRNAct, 320 pmol siRNA-YAP, and 320 pmol siRNAp53. The circPVT1 overexpression in CAL27 cells was performed using 4 μg pcDNA3-circPVT1-a or -b and 4 μg pcDNA3 vector as control. siRNAp53 smart pool was used at 80 pmol final concentration. Custom siRNAs were used at the following final concentrations: siRNAct, 150 pmol; siRNAct1, 150 pmol; siRNAct2, 150 pmol; siRNAbct1, 150 pmol; siRNAbct2, 150 pmol; siRNAcircPVT1, 150 pmol (siRNA overlaps the splice junction); siRNAb-circPVT1, 150 pmol (siRNA overlaps the splice junction); siRNA-PVT1, 300 pmol; siRNAp53, 150 pmol; siRNAp53 3′ UTR, 80 pmol; siRNA-YAP, 160 pmol; siRNA-YAP2, 80 pmol; siRNA-TAZ, 80 pmol siRNA-TAZ-1 + 80 pmol siRNA-TAZ-2; siRNA-TEAD, 40 pmol siTEAD1/3/4 + 40 pmol siTEAD2 + 40 pmol siTEAD1 ex 5 + 40 pmol siTEAD1 ex 8; siRNA-BUB1_1, 160 pmol; siRNA-BUB1_2, 160 pmol; siRNA-BUB1_3, 160 pmol.
Construction of circPVT1 expression vectors
pcDNA3-circPVT1-a was constructed by PCR amplification of the circPVT1 locus, including 954 bp upstream and 53 bp downstream of the nonlinear splice sites. We used the primers F1 and R1 designed to incorporate HindIII and NotI restriction sites and 6 bp of extra random sequence to aid in restriction digestion. The amplified region was cloned into the HindIII/NotI sites of pcDNA3 (Invitrogen). For construction of pcDNA3-circPVT1-b, an 895 bp DNA stretch upstream of the splice acceptor was amplified and inserted downstream in the reverse orientation in an XhoI/ApaI digested pcDNA3-circPVT1-a, using the primers F2 and R2 designed to incorporate ApaI and XhoI restriction sites and 6 bp of extra random sequence to aid in restriction digestion.
The pcDNA3-circPVT1-a-del was generated utilizing the pcDNA3-circPVT1-a plasmid as template. We deleted the miR-497-5p binding site using the QuikChange II XL site-directed mutagenesis kit (Agilent Technologies). Primer sequences are listed in Additional file 3: Table S5 and designated circPVT1_delmiR-497-5p F and circPVT1_delmiR-497-5p R. The miR-497-5p binding site was identified by RNAhybrid .
Cell proliferation assay
Cell proliferation was determined by viable cell counting. We seeded 1.5 × 105 cells in six-well plates in duplicate and grown for 72 h. Cell counting was performed after 24, 48, and 72 h mixing an aliquot of cells 1:1 with Trypan Blue dye (Invitrogen).
Cell cycle analysis
We fixed 106 cells in cold ethanol and these were then washed, resuspended in phosphate-buffered saline containing 50 μg/ml propidium iodide (PI; Sigma-Aldrich) and 50 μg/ml RNAase A, and analyzed by flow cytometry on a Millipore Guava® easyCyte 8HT.
We plated 103 cells in 60 mm dishes and incubated them at 37 °C in a 5% CO2 atmosphere for colony formation. After 10 days, colonies were stained with crystal violet (Sigma-Aldrich) for 30 min and counted.
Cell viability was assessed using an ATPlite assay (Perkin Elmer, Massachusetts, USA) accordingly to the manufacturer’s instructions. We seeded 800 cells in 96-well plates in 200 μl of media. Silencing was performed in reverse transfection and, after 24 h, cells were treated as indicated with cisplatin (CDDP) for a further 72 h. Each plate was evaluated immediately on a microplate reader (Expire Technology, Perkin Elmer). Each sample was assayed in triplicate.
Transwell migration assays
Transfected cells were detached and counted. A migration assay was performed using a 24-well plate. We seeded 6 × 104 cells, in a volume of 500 μl RPMI-1640 without FBS, in the upper chamber with an 8-mm pore size filter (BD Falcon, Franklin Lakes, NJ, USA) while the bottom chamber of the transwell was filled with 700 μl of RPMI-1640 with 10% FBS. Cells were allowed to migrate for 36 h in a humidified incubator at 37 °C and 5% CO2. Migrated cells, which had attached to the outside of the filter, were visualized by staining with DAPI (Sigma-Aldrich) and counted under a Zeiss LSM 510 laser scanning fluorescence confocal microscope.
Western blot analysis
Cells were homogenized on ice for 30 min in a lysis buffer composed by 50 mM, Hepes pH 7.5, 5 mM EDTA pH 8.0, 10 mM MgCl2, 150 mM NaCl, 50 mM NaF, 20 mM β-glicerophosphate, 0.5% NP40, 0.1 mM sodium orthovanadate, 1 mM PMSF, 1 mM dithiothreitol (DTT), and protease inhibitor cocktail (Roche). Lysates were clarified by centrifugation for 10 min, max speed, at 4 °C. Proteins (30 μg/lane) were separated on 10% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Immunoblots were probed with the following primary antibodies: mouse monoclonal anti-p53 (DO1; Oncogene Science Uniondale, NY, USA), rabbit polyclonal anti-YAP (Santa Cruz Biotechnology), and mouse monoclonal anti-GAPDH (Calbiochem). Immunostained bands were detected by a chemiluminescent method.
RNA extraction, reverse transcription, and RT-qPCR
The total RNA was extracted with TRizol (Thermo Fisher Scientific) following the manufacturer’s instructions and quantified using a Nanodrop (Thermo Scientific). The concentration, purity, and quality of total RNA were assessed using a Nanodrop TM 1000 spectrophotometer (Nanodrop Technologies). Reverse Transcription of miRNAs and RT-qPCR quantification of miRNA expression were performed by TaqMan MicroRNA RT assay and TaqMan MiRNA® Assay (Thermo Fisher Scientific), respectively, according to the manufacturer’s protocol. RNU44 and RNU48 were used as endogenous controls to standardize miRNA expression. For circPVT1, PVT1, and gene expression analysis, reverse transcription and RT-qPCR were performed using MMLV RT (Invitrogen) and SYBR Green® Assays (Applied Biosystems), respectively, according to the manufacturers’ instructions. For the detection of the circPVT1 expression level we used divergent primers that produce a PCR product only when the template is circular. circPVT1 expression in patients was normalized to the geometric mean of H3, GAPDH, and RPL19. In the other RT-qPCRs the expression of the genes was normalized to the GAPDH level.
We performed 1% formaldehyde cross-linking and ChIP as described in . The chromatin solution was immunoprecipitated with sheep polyclonal anti-p53 Ab7 antibody (Merck Millipore), rabbit polyclonal anti-YAP (H-125; Santa Cruz), rabbit polyclonal anti-Pol II (N-20; Santa Cruz), mouse monoclonal anti-TEF-1 (anti-TEAD-1, BD Transduction Laboratories), and ChIP-grade mouse monoclonal [4H8] anti-RNA polymerase II CTD repeat YSPTSPS (phospho S5; Abcam). The immunoprecipitations were performed using Pierce ChIP-grade Protein A/G magnetic beads (Thermo Fisher Scientific). The immunoprecipitated and purified chromatin was subjected to RT-qPCR. The promoter occupancy was analyzed by RT-qPCR using the SYBR Green assay (Applied Biosystems). Normalization was performed to the amount of input chromatin.
Untreated CAL27 cells were detached and resuspended in freshly prepared nuclear isolation buffer, kept on ice for 5 min (with frequent mixing) and stored at −80 °C. The cell lysate was twice frozen and thawed before use. Lysate was passed through a 27.5 gauge needle four times to promote nuclear lysis and centrifuged at 14,000 × g for 10 min at 4 °C. An aliquot of lysate (10 μl) was used as input for western blot analysis. For immunoprecipitation, the lysates were incubated with 50 μl of Dynabeads protein G (Thermo Fisher Scientific) and 7 μl of anti-p53 Ab7 antibody (Merck Millipore), rabbit polyclonal anti-YAP (H-125; Santa Cruz Biotechnology) or sheep serum, used as control, under constant shaking at room temperature for 1 h. Dynabeads protein G were prepared according to the manufacturer’s instructions. Following the immunoprecipitation, 10% of the beads were used for weatern blot analysis and the other part was used to isolate co-precipitated RNAs by resuspending beads in TRIzol (Thermo Fisher Scientific) according to the manufacturer’s instructions.
Transfection of CAL27 cells was performed as indicated above. Cells were incubated with 100 μM of 5,6-dichlorobenzimidazole 1-β-D-ribofuranoside (DRB; Sigma, D1916) for 3 h to block Pol II transcription. Transcription was recovered after DRB release and newly transcribed RNA was labeled with 350 μM of 4-thiouridine (4sU; Sigma, T4509) for 2 h. Cells were trypsinized and pelleted. TRIzol (Invitrogen) was added to stop transcription and total RNA was extracted. Total RNA (15 μg) was used for biotinylation and purification of 4sU-labeled nascent RNA, for each point. 4sU-labeled RNA was incubated in biotinylation buffer (10 mM HEPES pH 7.5, 1 mM EDTA) with 5 μg of MTSEA Biotin-XX (90066/9006-1 Biotium) for 1.5 h with rotation. MTSEA Biotin-XX was dissolved in dimethylformamide (DMF; Sigma, D4551) at a concentration of 10 mg/ml. Next, unbound Biotin-XX was removed by an equal volume of phenol:chloroform:isoamyl alcohol (15593-031, Invitrogen) and RNA was precipitated at 13,000 rpm for 20 min at 4 °C with 1:10 volume of 5 M NaCl and an equal volume of isopropanol. The RNA pellet was washed with 75% ethanol and resuspended in 50 μl nuclease-free H2O. 4sU-labeled and unlabeled RNA was separated by using 200 μl Dynabeads MyOne Streptavidin C1 (65001, Invitrogen) at room temperature for 20 min according to the manufacturer’s instructions. Beads were washed five times with 0.8 ml of washing buffer (100 mM Tris pH 7.4, 10 mM EDTA, 1 M NaCl, 0.1% Tween 20). Nascent RNA was eluted with 100 μl 0.1 M DTT twice and was precipitated with 40 μl of 4 M LiCl, 2 μl glycogen, and 600 μl ice-cold ethanol. The RNA pellet was resuspended in 20 μl nuclease free H2O and incubated at 65 °C for 10 min at 400 rpm. Reverse transcription and RT-qPCR quantification were performed as described above.
circPVT1 and clinical features
A logistic regression model was fitted to evaluate association with clinical variables. Statistical significance was set at 5%.
Kaplan-Meier analysis was used for overall survival and the log-rank test was used to assess differences between curves with high and low circPVT1 intensity. We defined the group of patients with low circPVT1 expression as samples with expression lower than median(X)-σ/2, where X is the circPVT1 expression distribution and σ the standard deviation.
The EC50 and the LD50 were determined using the Compusyn Software .
Analysis of TCGA HNSCC exon expression data
Normalized TCGA HNSCC exon expression analysis was performed using 31 non-tumoral and 263 tumor samples (we obtained information for 129 mut-p53 and 57 wt-p53 samples). The chromosomal intervals are related to the human genome GRch37/hg19. Deregulation of specific exons in different subgroups of samples was assessed by a two-sided Wilcoxon rank sum test.
Analysis of PVT1 and circPVT1 promoters
We used the Promoter 2.0 prediction and the Lasagna 2.0 web-tool to analyze the PVT1 and circPVT1 promoters. The promoter sequences are related to the human genome GRch38/hg38.
In silico prediction targets
In silico prediction of miRNA-497-5p targets was performed using miRwalk2 . We selected targets predicted from three different software and chose according to the negative correlation between miRNA-497-5p expression  and mRNA expression analysis of 14 patients with TP53 mutations and nine normal matched patients.
Quantification and statistical analysis
GraphPad Prism was used to determine the statistical significance of the in vitro experiments. MATLAB was used to determine the statistical significance of the data involving patient samples. For each figure, relevant information for assessing the accuracy and precision of the analysis is included in the accompanying legend.
We thank AIRC for supporting Maria Ferraiuolo with the FIRC-AIRC Fellowship (ID 19371) and Francesca Torti for kindly providing PVT1 and circPVT1 primers and for the sequences of the following siRNAs: siRNAcircPVT1, siRNAb-circPVT1, and siRNA-PVT1.
Contribution of AIRC (AIRC 20613), EPIGEN Flagship Project (13/05/R/42) to G.B. and of ABOCA S.P.A. to S.S. was greatly appreciated.
Availability of data and materials
Normalized TCGA HNSC exon expression data analyzed during the current study are available in the Broad Institute repository (http://gdac.broadinstitute.org/; doi:10.7908/C11G0KM9) . The mRNA expression analysis used during the current study has been deposited in NCBI’s Gene Expression Omnibus and is accessible through GEO series accession number GSE107591 (https://0-www-ncbi-nlm-nih-gov.brum.beds.ac.uk/geo/query/acc.cgi?acc=GSE107591) .
Conceptualization, LV, GB; methodology, LV, MF, AS, GB; validation, LV, MF, AS, FG, JV; formal analysis, LV, AS; investigation, LV, MF, FG; resources, PP, TC, SS, GM, NR, GB; writing of original draft, LV; writing, review and editing, LV, NR, GB; supervisor, GB; project administrator, GB; funding acquisition, GB and SS. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The study was approved by the scientific ethics committee of the Italian National Cancer Institute ‘Regina Elena’ (Rome) (protocol CE/379/08). All patients have given written informed consent for the study. The experimental methods comply with the Helsinki Declaration.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Hsu MT, Coca-Prados M. Electron microscopic evidence for the circular form of RNA in the cytoplasm of eukaryotic cells. Nature. 1979;280:339–40.View ArticlePubMedGoogle Scholar
- Sanger HL, Klotz G, Riesner D, Gross HJ, Kleinschmidt AK. Viroids are single-stranded covalently closed circular RNA molecules existing as highly base-paired rod-like structures. Proc Natl Acad Sci U S A. 1976;73:3852–856.View ArticlePubMedPubMed CentralGoogle Scholar
- Kos A, Dijkema R, Arnberg AC, van der Meide PH, Schellekens H. The hepatitis delta (delta) virus possesses a circular RNA. Nature. 1986;323:558–60.View ArticlePubMedGoogle Scholar
- Cocquerelle C, Mascrez B, Hétuin D, Bailleul B. Mis-splicing yields circular RNA molecules. FASEB J. 1993;7(1):155–60.PubMedGoogle Scholar
- Jeck WR, Sharpless NE. Detecting and characterizing circular RNAs. Nat Biotechnol. 2014;32:453–61.View ArticlePubMedPubMed CentralGoogle Scholar
- Memczak S, Jens M, Elefsinioti A, Torti F, Krueger J, Rybak A, et al. Circular RNAs are a large class of animal RNAs with regulatory potency. Nature. 2013;495:333–38.View ArticlePubMedGoogle Scholar
- Rybak-Wolf A, Stottmeister C, Glažar P, Jens M, Pino N, Giusti S, et al. Circular RNAs in the mammalian brain are highly abundant, conserved, and dynamically expressed. Mol Cell. 2015;58(5):870–75.View ArticlePubMedGoogle Scholar
- Alhasan A, Izuogu OG, Al-Balool HH, Steyn JS, Evans A, Colzani M, et al. Circular RNA enrichment in platelets is a signature of transcriptome degradation. Blood. 2016;127(9):e1–e11. doi:10.1182/blood-2015-06-649434.View ArticlePubMedPubMed CentralGoogle Scholar
- Memczak S, Papavasileiou P, Peters O, Rajewsky N. Identification and characterization of circular RNAs as a new class of putative biomarkers in human blood. PLoS One. 2015;10(10):e0141214. doi:10.1371/journal.pone.0141214.View ArticlePubMedPubMed CentralGoogle Scholar
- Ashwal-Fluss R, Meyer M, Pamudurti NR, Ivanov A, Bartok O, Hanan M, et al. circRNA biogenesis competes with pre-mRNA splicing. Mol Cell. 2014;56:55–6.View ArticlePubMedGoogle Scholar
- Ivanov A, Memczak S, Wyler E, Torti F, Porath HT, Orejuela MR, et al. Analysis of intron sequences reveals hallmarks of circular RNA biogenesis in animals. Cell Rep. 2015;10(2):170–7. doi:10.1016/j.celrep.2014.12.019.View ArticlePubMedGoogle Scholar
- Liang D, Wilusz JE. Short intronic repeat sequences facilitate circular RNA production. Genes Dev. 2014;28(20):2233–247.View ArticlePubMedPubMed CentralGoogle Scholar
- Starke S, Jost I, Rossbach O, Schneider T, Schreiner S, Hung LH, et al. Exon circularization requires canonical splice signals. Cell Rep. 2015;10(1):103–11.View ArticlePubMedGoogle Scholar
- Zhang XO, Wang HB, Zhang Y, Lu X, Chen LL, Yang L. Complementary sequence-mediated exon circularization. Cell. 2014;159:134–47.View ArticlePubMedGoogle Scholar
- Hansen TB, Jensen TI, Clausen BH, Bramsen JB, Finsen B, Damgaard CK, et al. Natural RNA circles function as efficient microRNA sponges. Nature. 2013;495:384–88.View ArticlePubMedGoogle Scholar
- Guarnerio J, Bezzi M, Jeong JC, Paffenholz SV, Berry K, Naldini MM, et al. Oncogenic role of fusion-circRNAs derived from cancer-associated chromosomal translocations. Cell. 2016;165:289–02.View ArticlePubMedGoogle Scholar
- Li P, Chen S, Chen H, Mo X, Li T, Shao Y, et al. Using circular RNA as a novel type of biomarker in the screening of gastric cancer. Clin Chim Acta. 2015;444:132–36.View ArticlePubMedGoogle Scholar
- Li Y, Zheng Q, Bao C, Li S, Guo W, Zhao J, et al. Circular RNA is enriched and stable in exosomes: a promising biomarker for cancer diagnosis. Cell Res. 2015;25:981–84.View ArticlePubMedPubMed CentralGoogle Scholar
- Chen J, Li Y, Zheng Q, Bao C, He J, Chen B, et al. Circular RNA profile identifies circPVT1 as a proliferative factor and prognostic marker in gastric cancer. Cancer Lett. 2017;388:208–19.View ArticlePubMedGoogle Scholar
- Panda AC, Grammatikakis I, Kim KM, De S, Martindale JL, Munk R, et al. Identification of senescence-associated circular RNAs (SAC-RNAs) reveals senescence suppressor CircPVT1. Nucleic Acids Res. 2017;45(7):4021–035.View ArticlePubMedGoogle Scholar
- Chapman MH, Tidswell R, Dooley JS, Sandanayake NS, Cerec V, Deheragoda M, et al. Whole genome RNA expression profiling of endoscopic biliary brushings provides data suitable for biomarker discovery in cholangiocarcinoma. J Hepatol. 2012;56:877–85.View ArticlePubMedGoogle Scholar
- Colombo T, Farina L, Macino G, Paci P. PVT1: a rising star among oncogenic long noncoding RNAs. Biomed Res Int. 2015;2015:304208. doi:10.1155/2015/304208.View ArticlePubMedPubMed CentralGoogle Scholar
- Guan Y, Kuo WL, Stilwell JL, Takano H, Lapuk AV, Fridlyand J, et al. Amplification of PVT1 contributes to the pathophysiology of ovarian and breast cancer. Clin Cancer Res. 2007;13:5745–755.View ArticlePubMedGoogle Scholar
- Haverty PM, Hon LS, Kaminker JS, Chant J, Zhang Z. High-resolution analysis of copy number alterations and associated expression changes in ovarian tumors. BMC Med Genomics. 2009;2:21. doi:10.1186/1755-8794-2-21.View ArticlePubMedPubMed CentralGoogle Scholar
- Wang F, Yuang JH, Wang SB, Yang F, Yuan SX, Ye C, et al. Oncofetal long noncoding RNA PVT1 promotes proliferation and stem cell-like property of hepatocellular carcinoma cells by stabilizing NOP2. Hepatology. 2014;60:1278–290.View ArticlePubMedGoogle Scholar
- Leemans CR, Braakhuis BJ, Brakenhoff RH. The molecular biology of head and neck cancer. Nat Rev Cancer. 2011;11:9–22.View ArticlePubMedGoogle Scholar
- O’Rorke MA, Ellison MV, Murray LJ, Moran M, James J, Anderson LA. Human papillomavirus related head and neck cancer survival: a systematic review and meta-analysis. Oral Oncol. 2012;48:1191–201.View ArticlePubMedGoogle Scholar
- Boscolo-Rizzo P, Maronato F, Marchiori C, Gava A, Da Mosto MC. Long-term quality of life after total laryngectomy and postoperative radiotherapy versus concurrent chemoradiotherapy for laryngeal preservation. Laryngoscope. 2008;118:300–06.View ArticlePubMedGoogle Scholar
- Smith RB, Sniezek JC, Weed DT, Wax MK. Utilization of free tissue transfer in head and neck surgery. Otolaryngol Head Neck Surg. 2007;137:182–91.View ArticlePubMedGoogle Scholar
- Vergeer MR, Doornaert PA, Rietveld DH, Leemans CR, Slotman BJ, Langendijk JA. Intensity-modulated radiotherapy reduces radiation-induced morbidity and improves health-related quality of life: results of a nonrandomized prospective study using a standardized follow-up program. Int J Radiat Oncol Biol Phys. 2009;74:1–8.View ArticlePubMedGoogle Scholar
- Agrawal N, Frederick MJ, Pickering CR, Bettegowda C, Chang K, Li RJ, et al. Exome sequencing of head and neck squamous cell carcinoma reveals inactivating mutations in NOTCH1. Science. 2011;333:1154–157.View ArticlePubMedPubMed CentralGoogle Scholar
- Balz V, Scheckenbach K, Götte K, Bockmühl U, Petersen I, Bier H. Is the p53 inactivation frequency in squamous cell carcinomas of the head and neck underestimated? Analysis of p53 exons 2-11 and human papillomavirus 16/18 E6 transcripts in 123 unselected tumor specimens. Cancer Res. 2003;63(6):1188–191.PubMedGoogle Scholar
- Network CGA. Comprehensive genomic characterization of head and neck squamous cell carcinomas. Nature. 2015;517:576–82.View ArticleGoogle Scholar
- Stransky N, Egloff AM, Tward AD, Kostic AD, Cibulskis K, Sivachenko A, et al. The mutational landscape of head and neck squamous cell carcinoma. Science. 2011;333:1157–160.View ArticlePubMedPubMed CentralGoogle Scholar
- Di Agostino S, Sorrentino G, Ingallina E, Valenti F, Ferraiuolo M, Bicciato S, et al. YAP enhances the pro-proliferative transcriptional activity of mutant p53 proteins. EMBO Rep. 2016;17(2):188–01.View ArticlePubMedGoogle Scholar
- He C, Mao D, Hua G, Lv X, Chen X, Angeletti PC, et al. The Hippo/YAP pathway interacts with EGFR signaling and HPV oncoproteins to regulate cervical cancer progression. EMBO Mol Med. 2015;7(11):1426–449.View ArticlePubMedPubMed CentralGoogle Scholar
- Hua G, Lv X, He C, Remmenga SW, Rodabough KJ, Dong J, et al. YAP induces high-grade serous carcinoma in fallopian tube secretory epithelial cells. Oncogene. 2016;35(17):2247–265.View ArticlePubMedGoogle Scholar
- Snijders AM, Schmidt BL, Fridlyand J, Dekker N, Pinkel D, Jordan RC, et al. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene. 2005;24(26):4232–242.View ArticlePubMedGoogle Scholar
- Zhang L, Yang S, Chen X, Stauffer S, Yu F, Lele SM, et al. The hippo pathway effector YAP regulates motility, invasion, and castration-resistant growth of prostate cancer cells. Mol Cell Biol. 2015;35(8):1350–362.View ArticlePubMedPubMed CentralGoogle Scholar
- Ganci F, Sacconi A, Bossel Ben-Moshe N, Manciocco V, Sperduti I, Strigari L, et al. Expression of TP53 mutation-associated microRNAs predicts clinical outcome in head and neck squamous cell carcinoma patients. Ann Oncol. 2013;24:3082–088.View ArticlePubMedPubMed CentralGoogle Scholar
- Dubin RA, Kazmi MA, Ostrer H. Inverted repeats are necessary for circularization of the mouse testis Sry transcript. Gene. 1995;167:245–48.View ArticlePubMedGoogle Scholar
- Zanconato F, Forcato M, Battilana G, Azzolin L, Quaranta E, Bodega B, et al. Genome-wide association between YAP/TAZ/TEAD and AP-1 at enhancers drives oncogenic growth. Nat Cell Biol. 2015;17(9):1218–227.View ArticlePubMedGoogle Scholar
- Zhang H, Liu CY, Zha ZY, Zhao B, Yao J, Zhao S, et al. TEAD transcription factors mediate the function of TAZ in cell growth and epithelial-mesenchymal transition. J Biol Chem. 2009;284:13355–3362.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhao B, Ye X, Yu J, Li L, Li W, Li S, et al. TEAD mediates YAP-dependent gene induction and growth control. Genes Dev. 2008;22:1962–971.View ArticlePubMedPubMed CentralGoogle Scholar
- Li Z, Huang C, Bao C, Chen L, Lin M, Wang X, et al. Exon-intron circular RNAs regulate transcription in the nucleus. Nat Struct Mol Biol. 2015;22(3):256–64.View ArticlePubMedGoogle Scholar
- Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C, et al. Mutational landscape and significance across 12 major cancer types. Nature. 2013;502:333–39.View ArticlePubMedPubMed CentralGoogle Scholar
- Morton JP, Timpson P, Karim SA, Ridgway R, Athineos D, Doyle B, et al. Mutant p53 drives metastasis and overcomes growth arrest/senescence in pancreatic cancer. Proc Natl Acad Sci U S A. 2010;107(Karim SA):246–51.View ArticlePubMedGoogle Scholar
- Muller PA, Vousden KH. Mutant p53 in cancer: new functions and therapeutic opportunities. Cancer Cell. 2014;25:304–17.View ArticlePubMedPubMed CentralGoogle Scholar
- Wang W, Cheng B, Miao L, Mei Y, Wu M. Mutant p53-R273H gains new function in sustained activation of EGFR signaling via suppressing miR-27a expression. Cell Death Dis. 2013;4:e574. doi:10.1038/cddis.2013.97.View ArticlePubMedPubMed CentralGoogle Scholar
- Mantovani F, Walerych D, Sal GD. Targeting mutant p53 in cancer: a long road to precision therapy. FEBS J. 2017;284(6):837–50. doi:10.1111/febs.13948.View ArticlePubMedGoogle Scholar
- Guo ST, Jiang CC, Wang GP, Li YP, Wang CY, Guo XY, et al. MicroRNA-497 targets insulin-like growth factor 1 receptor and has a tumour suppressive role in human colorectal cancer. Oncogene. 2013;32(15):1910–920.View ArticlePubMedGoogle Scholar
- Huang C, Ma R, Yue J, Li N, Li Z, Qi D. MiR-497 suppresses YAP1 and inhibits tumor growth in non-small cell lung cancer. Cell Physiol Biochem. 2015;37(1):342–52.View ArticlePubMedGoogle Scholar
- Li D, Zhao Y, Liu C, Chen X, Qi Y, Jiang Y, et al. Analysis of miR-195 and miR-497 expression, regulation and role in breast cancer. Clin Cancer Res. 2011;17(7):1722–730.View ArticlePubMedGoogle Scholar
- Wang S, Mo Y, Midorikawa K, Zhang Z, Huang G, Ma N, et al. The potent tumor suppressor miR-497 inhibits cancer phenotypes in nasopharyngeal carcinoma by targeting ANLN and HSPA4L. Oncotarget. 2015;6(34):35893–5907.PubMedPubMed CentralGoogle Scholar
- Valenti F, Ganci F, Fontemaggi G, Sacconi A, Strano S, Blandino G, et al. Gain of function mutant p53 proteins cooperate with E2F4 to transcriptionally downregulate RAD17 and BRCA1 gene expression. Oncotarget. 2015;6(8):5547–566.View ArticlePubMedPubMed CentralGoogle Scholar
- Gordon M, El-Kalla M, Zhao Y, Fiteih Y, Law J, Volodko N, et al. The tumor suppressor gene, RASSF1A, is essential for protection against inflammation-induced injury. PLoS One. 2015. doi: 10.1371/journal.pone.0075483Google Scholar
- Rehmsmeier M, Steffen P, Hochsmann M, Giegerich R. Fast and effective prediction of microRNA/target duplexes. RNA. 2004;10(10):1507–517.View ArticlePubMedPubMed CentralGoogle Scholar
- Chou TC, Martin N. Compusyn for drug combinations: PC software and user’s guide: a computer program for quantitation of synergism and antagonism in drug combinations, and the determination of IC50 and ED50 and LD50 values. Paramus: ComboSyn; 2005.Google Scholar
- Dweep H, Gretz N. miRWalk2.0: a comprensive atlas of microRNA-target interactions. Nat Methods. 2015;12(8):697.View ArticlePubMedGoogle Scholar
- Broad Institute TCGA Genome Data Analysis Center. Analysis-ready standardized TCGA data from Broad GDAC Firehose 2016_01_28 run. Broad Institute of MIT and Harvard. 2016. https://doi.org/10.7908/C11G0KM9.
- Verduci, L, Ferraiuolo M, Sacconi, A, Ganci, F, Vitale J, Colombo T, Paci, P, Strano, S, Macino, G, Rajewsky, N, Blandino, G. The oncogenic role of circPVT1 in head and neck squamous cell carcinoma is mediated through the mutant p53/YAP/TEAD complex. Gene Expression Omnibus. 2017. https://0-www-ncbi-nlm-nih-gov.brum.beds.ac.uk/geo/query/acc.cgi?ac=GSE107591