E2F1-mediated KDM4A-AS1 up-regulation promotes EMT of hepatocellular carcinoma cells by recruiting ILF3 to stabilize AURKA mRNA

Abstract

Hepatocellular carcinoma (HCC) is a gastrointestinal tumor with high clinical incidence. Long non-coding RNAs (lncRNAs) play vital roles in modulating the growth and epithelial-mesenchymal transition (EMT) of HCC. However, the underlying mechanism of lncRNA KDM4A antisense RNA 1 (KDM4A-AS1) in HCC remains elusive. In our study, the role of KDM4A-AS1 in HCC was systematically investigated. The levels of KDM4A-AS1, interleukin enhancer-binding factor 3 (ILF3), Aurora kinase A (AURKA), and E2F transcription factor 1 (E2F1) were determined by RT-qPCR or western blot. ChIP and dual luciferase reporter experiments were performed to detect the binding relationship between E2F1 and KDM4A-AS1 promoter sequence. RIP and RNA-pull down confirmed the interaction of ILF3 with KDM4A-AS1/AURKA. Cellular functions were analyzed by MTT, flow cytometry, wound healing and transwell assays. IHC was performed to detect Ki67 in vivo. We found that KDM4A-AS1 was increased in HCC tissues and cells. Elevated KDM4A-AS1 level was correlated to poor prognosis of HCC. Knockdown of KDM4A-AS1 inhibited the proliferation, migration, invasion and EMT of HCC cells. ILF3 bound to KDM4A-AS1 and AURKA. KDM4A-AS1 maintained the stability of AURKA mRNA by recruiting ILF3. E2F1 transcriptionally activated KDM4A-AS1. Overexpressed KDM4A-AS1 reversed the contribution of E2F1 depletion to AURKA expression and EMT in HCC cells. KDM4A-AS1 promoted tumor formation in vivo through the PI3K/AKT pathway. These results revealed that E2F1 transcriptionally activated KDM4A-AS1 to regulate HCC progression via the PI3K/AKT pathway. E2F1 and KDM4A-AS1 may serve as good prognostic targets for HCC treatment.

Fig. 1. KDM4A-AS1 was up-regulated in HCC.

A Starbase database predicted KDM4A-AS1 level in HCC tissues and normal adjacent tissues. B Kaplan–Meier analysis was used to detect the survival rate of HCC patients. C RT-qPCR was performed to examine KDM4A-AS1 level in HCC tissues (n = 48) and adjacent normal tissues (n = 48). D RT-PCR detected the level of KDM4A-AS1 in clinical phase I + II and III + IV tissues. E RT-qPCR analysis of KDM4A-AS1 level in human normal liver cells (THLE-3) and HCC cell lines. F FISH assay was utilized to test KDM4A-AS1 location in HCC cells. Results were expressed as means ± SD for at least triplicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 2. Knockdown of KDM4A-AS1 inhibited EMT of HCC cells.

A Transfection efficiency was verified by RT-qPCR. B, C Flow cytometry was applied for cell apoptosis analysis. D Western blot was used to examine Bax and Bcl-2 expression. E MTT assay was conducted to determine cell viability. F, G Wound healing experiment examined cell migration. H, I Transwell analysis of cell invasion ability. J Western blot was performed to detect EMT-related proteins. Results were expressed as means ± SD for at least triplicate experiments. **P < 0.01, ***P < 0.001.

Fig. 3. KDM4A-AS1 recruited ILF3 to maintain the stability of AURKA mRNA in HCC cells.

A RNA pull-down verified KDM4A-AS1 could directly bind to ILF3. B RNA pull-down verified ILF3 specifically bonded to AURKA. C RIP confirmed the targeting relationship between KDM4A-AS1 and ILF3. D RIP assay confirmed AURKA as a target of ILF3. E The expression of ILF3 was tested by RT-qPCR. F Cells were exposed to 5 μg/ml actinomycin D for 0, 3, 6, 9, 12 h, respectively. RT-qPCR analysis of the mRNA stability of AURKA. G, H AURKA level was tested by RT-qPCR and Western blot after knocking down ILF3. I Detection of AURKA mRNA stability. J, K RT-qPCR and Western blot were used to detect AURKA level after knocking down KDM4A-AS1. Data were shown as means ± SD for three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4. ILF3 knockdown inhibited EMT of HCC cells through inhibiting AURKA.

A, B RT-qPCR and western blot were utilized to assess ILF3 level in THLE-3, Hep 3B, and Huh-7 cells. C, D The transfection efficiency of pcDNA 3.1 AURKA was verified by RT-qPCR and western blot. E, F RT-qPCR and western blot were used to detect AURKA expression in HCC cells with sh-ILF3 and pcDNA 3.1 AURKA. G, H Cell apoptosis was tested by flow cytometry. I Western blot was conducted to examine Bax and Bcl-2 expression. J MTT assay was performed to determine cell viability. K, L Wound healing experiments checked cell migration. M, N Transwell analysis of cell invasion ability. O Western blot detection of EMT-related proteins. Results are expressed as mean ± SD for at least triplicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 5. E2F1 facilitated KDM4A-AS1 transcription in HCC cells.

A, B RT-qPCR and western blot were used to assess E2F1 levels in THLE-3, Hep 3B, and Huh-7 cells. C, D The transfection efficiency of sh-E2F1 was verified by RT-qPCR and western blot. E RT-qPCR was performed to assess KDM4A-AS1 level in HCC cells with sh-E2F1. F The binding site of E2F1 in the KDM4A-AS1 promoter was predicted by JASPER. G, H ChIP and luciferase reporter assays detected the binding relationship between E2F1 and KDM4A-AS1 promoter sequence. Data were shown as means ± SD for three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 6. E2F1 knockdown inhibited HCC EMT through KDM4A-AS1.

A The transfection efficiency of OE-KDM4A-AS1 was verified using RT-qPCR. B RT-qPCR was used to assess KDM4A-AS1 level in HCC cells with sh-E2F1 and OE-KDM4A-AS1. C RT-qPCR was utilized to detect AURKA expression. D Western blot was performed to detect AURKA and PI3K/AKT pathway protein levels. E, F Flow cytometry was employed to measure cell apoptosis. G Western blot analysis of Bax and Bcl-2 expression. H MTT assay was used to determine cell viability. I, J Wound healing experiment tested cell migration. K, L Transwell analysis of cell invasion ability. M Western blot was conducted to detect EMT-related proteins. Results were expressed as means ± SD for at least triplicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 7. KDM4A-AS1 medicated by E2F1 promoted tumor formation in vivo via activating PI3K/AKT signaling.

BALB/C male nude mice were injected subcutaneously with HCC cells transfected with stable sh-E2F1 and OE-KDM4A-AS1. A The images of the xenograft tumors. B The tumor growth curve. C Changes in tumor weight. D IHC was used to examine Ki67 expression (Scale bar: 100 μm). E, F RT-qPCR analysis of KDM4A-AS1 and AURKA expression. G Western blot analysis of AURKA, PI3K/AKT pathway proteins, and EMT-related proteins. Results expressed as mean ± SD for at least triplicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 8. The mechanism diagram.

E2F1 transcriptionally activated KDM4A-AS1. The up-regulated KDM4A-AS1 stabilized the AURKA mRNA by recruiting ILF3, and then promoted HCC cell proliferation, migration, invasion, and EMT via PI3K/AKT pathway.