Long noncoding RNA FER1L4 suppresses cancer cell growth by acting as a competing endogenous RNA and regulating PTEN expression

Aberrantly expressed long noncoding RNAs (lncRNAs) are associated with various cancers. However, the roles of lncRNAs in the pathogenesis of most cancers are unclear. Here, we report that the lncRNA FER1L4 (fer-1-like family member 4, pseudogene) acts as a competing endogenous RNA (ceRNA) to regulate the expression of PTEN (a well-known tumor suppressor gene) by taking up miR-106a-5p in gastric cancer. We observed that FER1L4 was downregulated in gastric cancer and that its level corresponded with that of PTEN mRNA. Both FER1L4 and PTEN mRNA were targets of miR-106a-5p. Further experiments demonstrated that FER1L4 downregulation liberates miR-106a-5p and decreases the abundances of PTEN mRNA and protein. More importantly, FER1L4 downregulation accelerated cell proliferation by promoting the G0/G1 to S phase transition. We conclude that one mechanism by which lncRNAs function in in tumorigenesis is as ceRNAs for tumor suppressor mRNAs.

In this study, we investigated whether FER1L4 can act as a ceRNA for PTEN mRNA through miR-106a-5p. We found that FER1L4 acted as a ceRNA to regulate PTEN expression by acting as a sponge for miR-106a-5p in gastric cancer. FER1L4 downregulation has been found to be a characteristic molecular change in gastric cancer 6 and to lead to decreases in PTEN mRNA and protein levels. Because PTEN is a negative regulator of the cell cycle, we studied the effects of the downregulation of FER1L4 on the cell cycle; downregulation of FER1L4 by small interfering RNAs (siRNAs) increased cell proliferation by promoting the G 0 /G 1 to S phase transition.

Results
FER1L4 and PTEN mRNA are downregulated in gastric cancer. Our previous work revealed that the FER1L4 level is significantly lower in gastric cancer tissues than in paracancerous tissues 6 . In this study, we further found that FER1L4 levels in the human gastric cancer cell lines AGS, MGC-803 and SGC-7901 were lower than those in the human normal gastric epithelial cell line GES-1 (Fig. 1a). These results suggest that decreases in the level of FER1L4 may be associated with the occurrence of gastric cancer.
Considering that FER1L4 and PTEN mRNA are targets of miR-106a-5p 12,24 and that miR-106a-5p is highly expressed in gastric cancer 25 , we speculated that FER1L4 and PTEN mRNA may act as a pair of ceRNAs that are linked by miR-106a-5p. An important corollary of the ceRNA hypothesis is that ceRNAs coordinately regulate each other's expression 8 . Based on the above findings, we investigated whether FER1L4 was coexpressed with PTEN in human samples. We measured FER1L4 and PTEN expression in 20 gastric cancer tissue samples by quantitative reverse transcription-polymerase chain reaction (qRT-PCR). The samples were sorted into two subsets (n = 10) according to the FER1L4 expression level (i.e., high and low FER1L4 groups, Fig. 1b). As shown in Fig. 1c, we observed significantly higher levels of PTEN mRNA in the high FER1L4 group than in the low FER1L4 group and vice versa. This coexpression is consistent with the ceRNA hypothesis.
FER1L4 and PTEN are targeted by miR-106a-5p in gastric cells. PTEN mRNA is one of the validated targets of miR-106a-5p 12 . In our previous study, the interaction between FER1L4 and miR-106a-5p was first predicted by miRcode and then confirmed by dual luciferase reporter assays 24 . However, we do not know whether miR-106a-5p effectively regulates FER1L4 and PTEN in gastric cells. To increase the miR-106a-5p level, we transfected miR-106a-5p mimics into the normal human gastric epithelial cell line GES-1 and the human gastric cancer cell lines, AGS, MGC-803 and SGC-7901. Next, we utilized qRT-PCR analysis to reveal that miR-106a-5p suppressed both FER1L4 and PTEN mRNA abundance in all of the tested gastric cell lines (Fig. 2a,b). Moreover, to further test whether FER1L4 and PTEN expression levels were linked by miR-106a-5p, we decreased the miR-106a-5p level by transfection of its inhibitors into GES-1 and AGS cells. qRT-PCR analyses indicated that the transfection of miR-106a-5p inhibitors not only increased FER1L4 levels but also increased PTEN levels in both GES-1 and AGS cells (Fig. 2c). Because FER1L4 and PTEN mRNA are targets of miR-106a-5p, the knockdown of miR-106a-5p led to increases in free FER1L4 and PTEN mRNA.
Our previous study showed that by decreasing the expression of cyclin-dependent kinase (CDK) 1 and CDK2, the miR-106a inhibitors arrested gastric cancer cells at the G 0 /G 1 and G 2 /M phases and then suppressed cell proliferation; and the results of animal experiments showed that the miR-106a inhibitors significantly suppressed tumor growth in a dose-dependent manner 26 . As a result, we assumed that the tumor suppression effect of miR-106a inhibitors might indirectly through the overexpression of FER1L4.
Effects of FER1L4 downregulation on miR-106a-5p and PTEN expression. The ceRNA hypothesis proposes that the downregulation of miRNA targets will result in freeing of the same miRNA molecules 8 . For example, GAS5-siRNA significantly reduces the endogenous GAS5 level while simultaneously increasing the miR-21 level 27 . Thus, we sought to determine whether the downregulation of FER1L4 would influence miR-106a-5p and the levels of its targets. In our previous study, we designed a siRNA against FER1L4 and effectively reduced FER1L4 levels in the normal human gastric epithelial cell line GES-1 and the human gastric cancer cell lines AGS, MGC-803 and SGC-7901 24 . Here, we transfected FER1L4-siRNA into GES-1, AGS, MGC-803 and SGC-7901 cells, and qRT-PCR analyses revealed that the miR-106a-5p levels in the gastric cells were increased by the knockdown of FER1L4 (Fig. 3).
If FER1L4 functions as a ceRNA, its downregulation might free additional miR-106a-5p. This miR-106a-5p would target PTEN mRNA and trigger the downregulation of PTEN. We knocked down FER1L4 in GES-1, AGS, MGC-803 and SGC-7901 cells and monitored PTEN expression by qRT-PCR and Western blot. FER1L4 knockdown resulted in decreased levels of PTEN mRNA and protein (Fig. 4).
FER1L4 regulates the cell cycle and cell proliferation. PTEN acts as a negative regulator of the cell cycle by suppressing the phosphoinositide 3-kinase (PI3K)-AKT pathway 28 . Because perturbations of the FER1L4 level markedly affected PTEN expression, we decided to investigate the effects of disrupting FER1L4 on the cell cycle and cell proliferation. Flow cytometry demonstrated that FER1L4 downregulation promoted the G 0 /G 1 to S phase transition in GES-1, AGS, MGC-803 and SGC-7901 cells (Fig. 5). Moreover, FER1L4 knockdown also accelerated cell proliferation in all of the gastric cell lines (Fig. 6).

Discussion
Studies have revealed that ceRNAs play important roles in post-transcriptional regulation and are involved in oncogenesis and cancer progression 12,13,17,19,20,23,29-33 . Our previous study found that several lncRNAs may be associated with gastric cancer via their actions as ceRNAs 24 . In the present study, we   focused on FER1L4 and PTEN mRNA because both of these RNAs are targets of miR-106a-5p, a typical onco-miRNA 12,24,25 . FER1L4 was downregulated in gastric cancer tissues 6 and gastric cancer cells (Fig. 1a). FER1L4 was coexpressed with PTEN such that the upregulation of FER1L4 led to greater expression of PTEN and vice versa (Fig. 1b,c). To test whether FER1L4 downregulation resulted in a reduction in the level of PTEN expression by freeing miR-106a-5p, we knocked down FER1L4. This knockdown resulted in an increase in miR-106a-5p (Fig. 3) and a decrease in PTEN mRNA and protein levels (Fig. 4). Furthermore, FER1L4 downregulation accelerated cell proliferation (Fig. 6) by promoting the G 0 /G 1 to S phase transition (Fig. 5). These results indicate that FER1L4 and PTEN mRNA are a pair of ceRNAs that are linked by miR-106a-5p.
Indeed, the FER1L4-PTEN interaction may be associated with additional miRNAs. FER1L4 and PTEN share 33 types of MREs as predicted by miRcode 34 . It has been found that several transcripts, such as PTENP1 29,35 , CNOT6L 12 , VAPA 12 , VCAN 30 , ZEB2 13 , and lncRNA-BGL3 23 , act as ceRNAs of PTEN mRNA that are mediated by pools of miRNAs (Fig. 7). These transcripts share different types of MREs with PTEN mRNA and are connected in a complex ceRNA network. Furthermore, these transcripts may also be ceRNAs of other RNAs with which they share different MREs. For example, VCAN acts as a ceRNA in the regulation of not only PTEN but also RB1 and CD34 30,36 . ceRNAs may contain various MREs, and each miRNA may target hundreds of transcripts. Different pathways may interact with each other via ceRNA cross-talk 14 . In addition, whether ceRNAs can alter miRNA function in vivo requires further investigation 37,38 . The study of ceRNA interactions represents a new approach to examining complex post-transcriptional regulatory networks. Cancer is caused by alterations in various pathways. We could design specific ceR-NAs to alter these pathways for therapeutic purposes.
Recently, circular RNAs (circRNAs) were included as new members of ceRNAs [39][40][41][42][43] . A circRNA named CDR1as that contains approximately 70 MREs functions as a ceRNA that sequesters miR-7 away from its targets 44,45 . The ceRNA world is becoming increasingly more complex and fascinating. To better understand ceRNAs systematically, several groups have developed in silico mathematical ceRNA models to describe the characteristics of ceRNAs and ceRNA networks [46][47][48][49][50] . Moreover, several algorithms and databases, such as Linc2GO 51 , starBase 52 , lnCeDB 53 , and Cupid 54 , have been developed to predict ceRNA interactions. These tools facilitate ceRNA studies.  ceRNAs can not only be used to explain biological phenomena, such as autophagy, apoptosis, and morphogenesis [55][56][57] , but can also be developed as miRNA inhibitors, such as short tandem target mimic (STTM) 58 and circular miRNA sponges 59 . For example, circular miRNA sponges have been found to exhibit excellent anticancer effects 59 . These findings indicate that ceRNAs might represent new therapeutic approaches to cancer and other diseases 60 . Further investigations may help us to understand how ceRNAs contribute to oncogenesis and tumor metastasis. ceRNA disturbances may be associated with many diseases [61][62][63][64] .
In conclusion, we found that FER1L4 was minimally expressed in gastric cancer. Via its functions as a ceRNA, FER1L4 liberated miR-106a-5p, downregulated PTEN expression, and affected cell growth.

Methods
The methods were carried out in accordance with the approved guidelines.
All experimental protocols were approved by The Human Research Ethics Committee from Ningbo University.
All experiments were performed in triplicate. Relative quantification of gene expression was performed by the 2 −ΔΔCt method 65,66 . Western blot. The cells were collected and lysed with cell lysis buffer for Western blotting (Beyotime, Haimen, China). The proteins (30 μ g per lane) were separated on 12% SDS-polyacrylamide gels and transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Immunoblotting of the membranes was performed using the following primary antibodies: anti-PTEN (CST, Danvers, MA, USA) and anti-β -actin (4A Biotech, Beijing, China). The signals were revealed after incubation with the recommended secondary antibodies using an Odyssey Infrared Imaging System (LI-COR, Lincoln, NE, USA). β -actin was used as the control.