PTTG3P upregulation is associated with enhanced proliferation and cell-cycle progression of HCC
To validate the dysregulation of PTTG3P in HCC, we retrospectively reviewed its expression in 10 databases with PTTG3P expression data in HCCDB, a database of HCC expression atlas 24. PTTG3P upregulation is observed in 8/10 datasets (Fig. 1A). Notably, HCCDB15 refers to TCGA-LIHC. By RNA-FISH assay, we confirmed the cytoplasmic distribution of PTTG3P in Hep3B and HepG2 cells (Fig. 1B). To explore the functional role of PTTG3P, we designed PTTG3P shRNA and overexpression lentivirus (Fig. 1C-D). PTTG3P knockdown slowed HCC cell proliferation (Fig. 1E-F), increased G1 phase accumulation, and decreased the proportion of cells in the S phase (Fig. 1G-I). In comparison, PTTG3P overexpression significantly enhanced cell proliferation (Fig. 1E-F), decreased G1 phase accumulation but increased the ratio of cells in the S phase (Fig. 1G-I).
PTTG3P upregulation reduces apoptosis and enhances invasion and migration of HCC cells
By performing Annexin V/PI staining, we found that PTTG3P depletion led to increased cell apoptosis. In contrast, its overexpression resulted in reduced cell apoptosis (Fig. 2A-C). Matrigel invasion and scratch assays showed that PTTG3P knockdown impaired HCC cell invasion (Fig. 2D-F) and migration (Fig. 2G-I). However, its overexpression significantly enhanced cell invasion (Fig. 2D-F) and migration (Fig. 2G-I).
PTTG3P physically interacts with ILF2 and ILF3 in HCC
To investigate the functional role of PTTG3P in HCC, we tried to identify its interacting proteins. The schematic map showing the construction of recombinant pBluescript II SK + plasmid is presented in Fig. 3A. Then, RNA pull-down assay was conducted using Biotin-labelled sense PTTG3P and antisense PTTG3P followed by silver staining (Fig. 3B). The bands specific to PTTG3P were excised and subjected to mass spectrometry (Fig. 3B, Supplementary Table 2). Results showed that ILF3 (also known as NF90, 90 kD) has the highest Sequest HT score (Fig. 3C). This protein usually forms a regulatory complex with ILF2 (also known as NF45, 45 kD), which is also observed in the MS results (Fig. 3C). Sense PTTG3P specific 45 kD and 90 kD bands were also observed (Fig. 3D). Then, RNA-IP assay was conducted using anti-ILF2 and anti-ILF3, respectively. The presence of ILF2 and ILF3 in the anti-ILF2 and anti-ILF3 immunoprecipitated samples were confirmed by western blotting (Fig. 3E-F). These samples had significantly higher enrichment of PTTG3P compared to the IgG control group (Fig. 3G).
By checking RNA-seq and survival data in TCGA-LIHC and RNA-seq data from GTEx, we confirmed the upregulation of ILF2 and ILF3 in HCC tissues compared to normal liver and tumor-adjacent normal tissues (Fig. 3H). Survival analysis confirmed the association between upregulated ILF2/ILF3 and unfavorable prognosis (Fig. 3I-L).
PTTG3P depletion increases the binding of ILF2 to the promoter of TOP2A
Since we characterized that PTTG3P interacts with ILF2 and ILF3 in HCC, while the ILF2/ILF3 complex as a transcriptional enhancer/factor, we tried to explore their downstream targets in common. Firstly, we overexpressed PTTG3P in Hep3B cells. 48 h later, total RNA was extracted for RNA-seq. By setting adj. p < 0.05 and log 2(Fold Change, FC) ≥ 1 as the cutoff, we identified 215 dysregulated genes (Supplementary Table 3). One recent study explored dysregulated genes in HepG2 cells after ILF2 depletion 25. Using their data, we identified 125 genes downregulated after ILF2 inhibition (adj.p < 0.05 and log 2(FC) ≤-1) (Supplementary Table 4). These two sets shared six genes, including TOP2A, KIF14, BUB1, SERPINA3, STRA6, and KREMEN1 (Fig. 4A). For validation purposes, we further assessed the correlation of these genes with PTTG3P and ILF2 in TCGA-LIHC (Fig. 4B). By setting moderate correlations as the cutoff (| Pearson's r ≥ 0.4,| adj. p < 0.05), we found that TOP2A, KIF14, and BUB1 as potential candidates. Besides, we manually added LOXL2 and ENO2, two overexpressed genes after PTTG3P overexpression (Supplementary Table 3), as potential candidates.
We conducted ChIP-qPCR assay to check whether PTTG3P depletion alters the binding of ILF2 to the promoter of these genes. Results showed that after PTTG3P knockdown, ILF2 presented enhanced binding to the promoter of TOP2A, but showed weakened binding to the promoter of ENO2 (Fig. 4C-D). By checking protein level expression, we observed negative ENO2 expression in both normal liver and HCC tissues (Fig. 4E). In consistence with previous reports, we confirmed negative TOP2A expression in hepatocytes in normal liver (all negative by three primary antibodies in the HPA, HPA006458, HPA026773, and CAB002448) (Fig. 4F, green frame). In comparison, medium to high TOP2A expression was observed in HCC tissues (Fig. 4F, brown frame).
PTTG3P inhibition results in nuclear translocation of ILF2
After PTTG3P depletion, we observed nuclear accumulation of ILF2 by IF staining (Fig. 5A). However, no alteration was observed in the cellular distribution of ILF3 (Fig. 5A). Colocalization of ILF2 and ILF3 was observed in both cytoplasma and nuclear (Fig. 5A). By performing fractional western blotting, we confirmed this distributional alteration (Fig. 5B).
Via scanning the promoter region of TOP2A, we observed two potential ILF2/ILF3 binding sites (Fig. 5C). Then, we generated luciferase reporter plasmids carrying wild-type (pGL3-TOP2A-WT) and mutant TOP2A promoter sequences (pGL3-TOP2A-MT1, pGL3-TOP2A-MT2, and pGL3-TOP2A-MT3) based on a pGL3-basic backbone. ILF2 inhibition significantly reduced the luciferase activity of the reporters carrying at least one wild-type ILF2/ILF3 binding site (Fig. 5D-F), but had no influence on the reporter with two mutant sites (Fig. 5G). In contrast, PTTG3P inhibition significantly increased the luciferase activity of the reporters carrying at least one wild-type ILF2/ILF3 binding site (Fig. 5D-F), but did not influence the luciferase activity of the reporter with two mutant sites (Fig. 5G). These findings suggest that PTTG3P depletion might enhance TOP2A expression via promoting nuclear translocation of ILF2. To test this hypothesis, we checked the mRNA expression of TOP2A expression after PTTG3P knockdown. Interestingly, PTTG3P depletion did not increase but decreased TOP2A mRNA levels (Fig. 5H). This discrepancy implies that some other mechanisms are involved in regulating TOP2A expression/stability.
ILF2/ILF3 complex stabilizes TOP2A mRNA in the cytoplasm
Since the ILF2/ILF3 complex may act as an mRNA stabilizer, we tested whether ILF2/ILF3 binds to the 3'UTR of TOP2A mRNA and regulates its stability. RIP-qPCR confirmed significantly higher enrichment of TOP2A 3'UTR in the samples precipitated by anti-ILF2 than the IgG control (Fig. 6A-B). PTTG3P depletion significantly reduced the enrichment level (Fig. 6A-B). These findings suggest that ILF2 binds to the 3'UTR of TOP2A mRNA, the effect of which was enhanced by PTTG3P. ILF2 or PTTG3P depletion significantly reduced the stability of TOP2A mRNA (Fig. 6C-D), the effect of which was enhanced by simultaneous ILF2 and PTTG3P inhibition (Fig. 6C-D).
PTTG3P or TOP2A inhibition impairs HCC growth in-vivo
By generating Hep3B derived in vivo tumor model in mice, we found that PTTG3P overexpression significantly enhanced tumor growth (Fig. 7A-B). PTTG3P or TOP2A inhibition significantly slowed the growth rate (Fig. 7A-B). HE and TUNEL staining showed that PTTG3P and TOP2A inhibition are associated with enhanced tumor necrosis and apoptosis (Fig. 7C, top and bottom panels). In agree with these data, Ki-67 staining indicated a decreased level of cell proliferation in both PTTG3P KD and TOP2A KD groups (Fig. 7C, middle panel).
Based on these findings, we infer that PTTG3P exerts dual-regulation in TOP2A expression at both transcriptional and post-transcriptional levels. It is usually overexpressed in HCC tumor cells and accumulated in cytoplasm. Under this circumstance, it forms a complex with ILF2/ILF3, which binds to the 3'UTR of TOP2A mRNA and increases its stability. However, PTTG3P depletion can trigger nuclear translocation of ILF2, which binds to the promoter of the TOP2A gene and enhances its transcription (Fig. 8).