PTTG3P promotes gastric tumour cell proliferation and invasion and is an indicator of poor prognosis

Abstract Pseudogenes play a crucial role in cancer progression. However, the role of pituitary tumour‐transforming 3, pseudogene (PTTG3P) in gastric cancer (GC) remains unknown. Here, we showed that PTTG3P expression was abnormally up‐regulated in GC tissues compared with that in normal tissues both in our 198 cases of clinical samples and the cohort from The Cancer Genome Atlas (TCGA) database. High PTTG3P expression was correlated with increased tumour size and enhanced tumour invasiveness and served as an independent negative prognostic predictor. Moreover, up‐regulation of PTTG3P in GC cells stimulated cell proliferation, migration and invasion both in vitro in cell experiments and in vivo in nude mouse models, and the pseudogene functioned independently of its parent genes. Overall, these results reveal that PTTG3P is a novel prognostic biomarker with independent oncogenic functions in GC.


Introduction
Gastric cancer (GC) is one of the most common malignancies in the world, ranking as the second most prevalent carcinoma and the third leading cause of cancer-related death in China [1]. As it is difficult to detect at early stage, GC often leads to unresectable primary tumours and poor chemotherapy effects. Therefore, a better understanding of the general genetic profile associated with the pathogenesis and progression of GC is urgently needed.
Pseudogenes, genomic loci that resemble real genes, were once regarded as functionless entities, harbouring premature stop codons, deletions/insertions or frameshift mutations that abrogate the normal transcription and translation of 'real' genes [2]. In recent years, however, several studies have shown that pseudogenes also play critical roles in tumourigenesis/tumour suppression by competing with the expression of their true gene counterparts or through processing parent gene-targeted siRNAs [2,3]. Subsequently, various pseudogenes that are critically involved in carcinogenesis and cancer progression have been disclosed [4][5][6][7], but investigation into their functions in GC remains limited.
Pituitary tumour-transforming 3, pseudogene (PTTG3P), an intronless gene that is highly homologous to its family members pituitary tumour-transforming 1 (PTTG1) and pituitary tumour-transforming 2 (PTTG2), was first identified by Kakar and colleagues in 2000 [8]. Both PTTG1 and PTTG2 have been reported to serve oncogenic functions in human cancers [9][10][11], but the role of PTTG3P in GC remains unclear, and this pseudogene has previously been regarded as functionless.
In this study, we assessed PTTG3P expression using our previously described microarray analysis [12] and subsequently validated its expression in GC tissue specimens. We found that PTTG3P was significantly up-regulated in GC tissues and served as an independent risk factor for poor disease-free survival (DFS) and overall survival (OS). In addition, PTTG3P overexpression stimulated cell proliferation, potentially by inducing the G 1 -S transition, and promoted cell invasion both in vitro and in vivo. Moreover, the expression and function of PTTG3P were found to be independent of its true gene counterpart. Overall, these results reveal that PTTG3P is a novel prognostic predictor in GC that might enable the development of new therapeutic strategies for GC.

Materials and methods Patients
A retrospective cohort study was conducted among 198 patients who underwent surgical resection of primary gastric carcinoma from 2008 to 2010 at the Fudan University Shanghai Cancer Center. GC diagnoses were histopathologically confirmed. None of the patients received preoperative therapy. The resected tissue samples were immediately frozen in liquid nitrogen and stored at À80°C until RNA extraction. Clinicopathological features from all patients were obtained from medical records, pathological reports and personal interviews; these included age, gender, DFS, OS and tumour features (such as tumour location, size, differentiation, depth of invasion and the presence/absence of lymphatic metastasis). GC stage was defined according to the TNM classification system. The patients were followed up every 3 months during the first year after surgery and every 3-6 months thereafter until 31 May 2015. All patients had complete follow-up information. DFS was calculated from the date of surgery to the date of disease progression (local and/or distal tumour recurrence) or to the date of death. OS was defined as the length of time between surgery and death of the patient or the last follow-up date. The study was approved by the Clinical Research Ethics Committee of the Fudan University Shanghai Cancer Center. Written informed consent was obtained from all participants for the use of their tissues in this study.

RNA isolation, reverse transcription and quantitative real-time PCR
Total RNA was extracted from tissue samples and cell lines using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Reverse transcription (RT) and quantitative real-time PCR (qRT-PCR) kits (Takara, Dalian, China) were utilized to evaluate PTTG3P expression. RT and qRT-PCR were performed as previously described [13]. Specific primer sets for PTTG1, PTTG2 and PTTG3P were designed based on regions with low homology among these genes. PCR of the relevant genes and sequencing of the PCR products were carried out to ensure specificity. b-actin was included as an endogenous control to normalize the data.

Cell proliferation assays
Cell proliferation was evaluated using Cell Counting Kit-8 (CCK-8) (Dojindo, Kumamoto, Japan), an EdU DNA imaging kit (Life Technology) and colony-formation assays. The former two assays were performed on 2 9 10 3 cells grown in 96-well plates according to the recommended protocols. An automatic microplate reader (BioTek, Winooski, VT, USA) was used to determine the absorbance at 450 nm. Images of the cells used in the EdU assays were taken at 1009 and counted at 2009 under an immunofluorescence microscope (Olympus, Tokyo, Japan).
For colony-formation assays, 800 cells were seeded onto 6-well plates and incubated for 2 weeks. Then, the cells were fixed with ethanol and stained with crystal violet. The number of colonies containing more than 30 cells was counted.  with Annexin V-FITC and/or propidium iodide (PI) (BD Bioscience, USA) for apoptosis analysis or fixed with ethanol overnight at À20°C followed by subsequent PI (Calbiochem) staining for DNA content (cell cycle) analysis.

Cell motility and invasion assays
A wound-healing assay was performed to assess cell motility. Transfected cells were plated at equal densities in 6-well plates and grown to 100% confluence. The cells were pre-treated with Mitomycin C (Sigma-Aldrich) for 1 hr at 37°C to separate the role of cell motility from that of cell proliferation. Wounds were scratched with sterile pipette tips, loose cells were removed by rinsing with PBS, and serum-free medium was added. The wounds were observed at 0 and 6 hrs after scratching under a microscope at 1009 (Olympus).
Transwell chambers (8 lm, 24-well format) (Corning) were employed for cell invasion assays. A total of 4 9 10 4 cells in 100 ll of serumfree medium was loaded into the upper inserts, and 500 ll of culture medium containing 10% FBS was loaded into the lower chambers as a chemo-attractant. After a 24-hr incubation at 37°C, the cells that had migrated through the filters were fixed with ethanol and stained with crystal violet. Photographs were taken under a microscope at 2009 (Olympus), and the number of invaded cells was counted at 4009.

Tumour formation and metastasis assays in a nude mouse model
Athymic female BALB/c nude mice were maintained under specificpathogen-free conditions. HGC-27 cells stably expressing PTTG3P or the vector control were harvested and resuspended with RMPI-1640. For tumour formation, a total of 5 9 10 6 cells were subcutaneously injected into the right flank of each 5-week-old mouse (five mice for each group). For tumour metastasis, 1 9 10 6 cells were injected into the tail vein of each mouse (three mice per group). After transplantation, the weight of the mice and the growth of the subcutaneous tumours were assessed every 2 days. The mice were killed after a period of 4-6 weeks. Subcutaneous tumours and lungs were excised and measured. Tumour size was measured with the following formula: (L 9 W 2 )/2, where L is the length and W is the width of each tumour.

Statistical analysis
All statistical analyses were performed using SPSS 20.0 (IBM, Chicago, IL, USA).

3363
Correlations between PTTG3P expression and clinicopathological parameters were analysed using the Chi-square test. PTTG3P expression was assessed using the Chi-square test or Fisher's exact probability test. Survival was calculated using the Kaplan-Meier method and compared with the log-rank test. The results of the functional assays were analysed using Student's t-test. Variables with a value of P < 0.05 in univariate analysis were used in multivariate analysis based on the Cox proportional hazards model. P values less than 0.05 were considered significant.

PTTG3P is up-regulated in GC tissues and correlates with poor prognosis
We previously identified systemic variations in lncRNA expression between GC and paired non-tumour samples performed with microarray analysis [12] and noted that the pseudogene PTTG3P was up-regulated (2.008-fold change; P = 0.022) in GC tissues. A similar result was also found in The Cancer Genome Atlas (TCGA) database (P = 3.87EÀ10, Fig. 1A). Therefore, we analysed the mRNA expression levels of PTTG3P in 63 pairs of GC tissues and adjacent nontumours (ANTs) and found that PTTG3P was significantly up-regulated in 68.3% (43 of 63) of the GC tissues compared with the ANTs (P = 0.021, Fig. 1B). We next analysed the correlation between PTTG3P expression and clinicopathological characteristics in another 136 patients with GC. As shown in Table 1, high PTTG3P expression levels divided by the median value [14] were tightly correlated with larger tumour sizes (P = 0.043) and higher recurrence rates (P = 0.022).
DFS and OS curves were plotted according to PTTG3P expression level by the Kaplan-Meier method and log-rank tests. As shown in Figure 1C, high expression of PTTG3P was correlated with both significantly shorter DFS (P < 0.001) and OS (P < 0.001; Fig. 1C). Univariate analysis of survival revealed that the relative level of PTTG3P expression (P < 0.001), tumour grade (P < 0.001), lymphatic metastasis (P = 0.002) and TNM stage (P < 0.001) was prognostic indicators of DFS (Table 2) and OS (Table 3). Moreover, multivariate Cox regression analysis showed that both advanced TNM stage and high expression of PTTG3P were independent risk factors of DFS (Table 2) and OS (Table 3; P < 0.05). These results suggest that PTTG3P is a potential independent prognostic factor for GC.

PTTG3P stimulates GC tumour cell proliferation
To investigate the biological effects of PTTG3P in GC progression, baseline levels of PTTG3P expression were examined in five GC cell lines and a normal human gastric epithelial cell line. PTTG3P expression was significantly elevated in the GC cell lines ( Fig. 2A). AGS and HGC-27 cells were selected for overexpression experiments, and the efficiency of overexpression was validated by qRT-PCR (Fig. 2B). Then, we utilized CCK8 and colony-formation assays to elucidate the potential effect of PTTG3P on GC tumour cell proliferation. The CCK8 assay showed that the absorbance of PTTG3P-overexpressing cells was significantly higher than that of vector control-transfected GC cells,  suggesting that overexpression of PTTG3P in AGS and HGC-27 cells leads to increased accumulation of living tumour cells (P < 0.05; Fig. 2C). Similarly, the colony-formation assays showed that overexpression of PTTG3P led to the formation of more colonies in both AGS and HGC-27 cells than in controls (P < 0.01; Fig. 2D). These data suggest that PTTG3P stimulates GC tumour cell proliferation.

PTTG3P promotes G 1 -S cell cycle transition in GC cells
Next, we performed EdU immunofluorescence and flow cytometry assays to further explore the potential influence of PTTG3P on cell cycle progression in GC cells. An EdU labelling assay showed a  greater number of EdU-positive cells in cultures overexpressing PTTG3P than in the vector group, suggesting that the up-regulation of PTTG3P caused more cells to enter S phase compared to the control (Fig. 3A). Furthermore, cell cycle distribution analysis also showed that the percentage of PTTG3P-overexpressing AGS and HGC-27 cells in S phase (34.1% and 35.3%, respectively) was significantly higher than that in the control group (30.3% and 30.3%, respectively; P < 0.05), coupled with a decrease in the number of cells in G 1 phase. These results suggest that PTTG3P might affect the G 1 -S transition (Fig. 3B). Finally, Western blotting results also demonstrated that overexpression of PTTG3P increased cyclin D1 and p27 protein levels in AGS and HGC-27 cells (Fig. 3C). Overall, these data suggest that PTTG3P prompts cell cycle progression by promoting the G 1 -S transition in GC cells.

PTTG3P inhibits GC tumour cell apoptosis
We also performed an Annexin V apoptosis assay to determine the influence of PTTG3P on GC cell apoptosis. The results showed that after overexpression of PTTG3P, both AGS and HGC-27 cells showed significantly decreased proportions of early apoptotic cells (Annexin V + PI À ) (1.8% and 2.3%, respectively) and late apoptotic cells (Annexin V + PI + ) (1.3% and 3.0%, respectively) compared with controls (early apoptotic cells, 5.5% and 4.5%; late apoptotic cells, 2.9% and 4.7%, respectively; P < 0.05; Fig. 4A). Next, a 5-FU-induced apoptosis assay was performed to confirm the influence of PTTG3P on cell apoptosis, and similar results were obtained. Western blotting showed that overexpression of PTTG3P attenuated PARP1 and caspase-3 cleavage in both AGS and HGC-27 cells (Fig. 4B). Thus, PTTG3P reduces GC tumour cell apoptosis.

PTTG3P overexpression promotes GC cell migration and invasion
To determine whether PTTG3P promotes GC cell invasion and migration, we performed wound healing and transwell assays. Overexpression of PTTG3P resulted in accelerated wound-healing rates (P < 0.05; Fig. 5A) and an increased number of invaded cells and MMP9 and down-regulated the invasion-inhibiting protein E-cadherin (Fig. 5C). Collectively, these results suggest that PTTG3P enhances GC cell migration and invasion.

PTTG3P promotes GC tumour growth and metastasis in vivo
Finally, we used nude mouse xenograft and metastasis models to investigate the functions of PTTG3P in vivo. Consistent with the ex vivo data, overexpression of PTTG3P remarkably accelerated tumour growth compared with controls in the xenograft models (Fig. 6A). In addition, the xenografts derived from PTTG3P-overexpressing HGC-27 cells exceeded those derived from control cells in both size and weight (Fig. 6B). Metastasis assays showed that potential pulmonary metastases could be detected by CT scanning in the PTTG3P overexpression group (Fig. 6C bottom, red arrows indicate the suspicious tumour masses), whereas tumour masses were not detected in the empty-vector control group (Fig. 6C top), which was further confirmed by H&E staining (Fig. 6D). These results suggest that PTTG3P facilitates GC tumour growth and metastasis in vivo.

PTTG3P expression is independent of its parent genes PTTG1 and PTTG2
Because the parent genes PTTG1 and PTTG2 were previously confirmed to be oncogenes in various cancer types [11,[15][16][17] and PTTG3P shares high similarity with these genes (Fig. S1), we suggested that PTTG3P plays a role in modulating PTTG1 or PTTG2 that is similar to previously reported mechanisms [18,19]. To test this hypothesis, we designed specific primer sets for PTTG3P, PTTG1 and PTTG2. PCR products were examined by agarose gel electrophoresis followed by sequencing to confirm primer specificity (Fig. S2). Similar to PTTG3P, both PTTG1 and PTTG2 were up-regulated in GC tissues compared with ANTs (Fig. 7A), but correlation analysis showed that the expression levels of these two parent genes were not related to PTTG3P expression (P > 0.05, Fig. 7B). In vitro overexpression of PTTG3P in GC cell lines also failed to significantly alter the mRNA or protein expression level of either PTTG1 or PTTG2 (Fig. 7C and D). Thus, we concluded that PTTG3P expression and function are independent of its true gene counterparts.

Discussion
To the best of our knowledge, our study is the first to clarify the biological functions of PTTG3P in somatic malignancies, and to demonstrate that its oncogenic functions occur independently of its parent genes. We found that PTTG3P expression was elevated in GC tissues at the mRNA level and that PTTG3P served as an independent prognostic factor for GC. Up-regulation of PTTG3P promoted in vitro and in vivo GC cell proliferation and invasion/metastasis. Based on these results, we identified PTTG3P as a functional oncogenic pseudogene and a possible negative prognostic predictor of GC.
It was previously reported that PTTG3P is expressed at extremely low levels or is absent in normal tissues but is detectable in ovarian tumour tissues and cell lines [8]. In the present study, we showed that PTTG3P was significantly overexpressed in GC tissues compared with ANTs, suggesting that this pseudogene has a cancer-specific expression pattern in GC tissues. Clinically, the up-regulation of PTTG3P was highly correlated with increased tumour burden and a higher after treatment recurrence rate, suggesting that PTTG3P facilitates tumour progression by enhancing tumour cell proliferation and invasion. This hypothesis was verified in subsequent in vitro cell experiments and nude mouse models. The enhanced tumour cell proliferation and invasiveness caused by abnormally high expression of PTTG3P thus resulted in more aggressive biological behaviour, leading to shortened DFS and OS in patients with GC. Future studies might verify the clinical significance of PTTG3P expression in a larger number of GC samples to identify its prognostic value for patients with GC.
As a member of the PTTG family, PTTG3P is recognized as an intronless pseudogene that shares high homology with its parent genes PTTG1 and PTTG2, which led us to presume that PTTG3P might exert its biological effect via PTTG1/2 modulation, similar to other reported pseudogenes [18,19]. However, correlation analysis of PTTG3P with PTTG1 and PTTG2 expression in GC tissue did not reveal a significant relationship. Additionally, in vitro overexpression of PTTG3P in GC cell lines failed to alter the expression of the other two genes at the mRNA and protein levels. These results suggest that PTTG3P exerts its oncogenic role independently of its parent genes.
Because pseudogenes might not possess transcriptional control regions, they may be subject to transcriptional elements that are different from paralogous functional genes, thus exerting different functions than the parent genes [20,21]. Poliseno and colleagues revealed that the PTENP1 3 0 UTR significantly suppressed cell proliferation in PTEN-null PC3 cells, supporting the notion that PTENP1 could exert a tumour-suppressive role independent of the parent PTEN gene [2]. Likewise, wPPM1K, a partial retrotranscript pseudogene containing inverted repeats capable of being processed into two endo-siRNAs, regulates cell growth-related target genes and exerts tumour-suppressive activity independent of its cognate gene PPM1K [3]. The pseudogene wCx43 can be translated into a 43 kD protein that bears an amino acid change from arginine R202 to cysteine, resulting in a deficiency in the ability to mediate intercellular communication [22]. Collectively, these results suggest that pseudogenes, especially those subject to different transcriptional elements, may exert their biological roles independently of their parent genes. Furthermore, PTTG3P is located at 8q13.1, whereas the ancestral genes are located on chromosomes 5 and 4 [8]; therefore, it is possible that PTTG3P might be subject to different transcriptional regulation and might possess a different regulatory function compared to those of its cognate counterpart, thus allowing it to exert its oncogenic role independently of its parent genes.
In summary, the current study suggests that PTTG3P exhibits a novel oncogenic role in the regulation of pathways related to cell cycle progression that is independent of its parent gene and that this pseudogene predicts poor prognosis in patients with GC. Our findings provide new insights into the molecular details of GC and the potential therapeutic targets that can be used to combat this disease.

Figure S1
The homologous sequences of PTTG3P, PTTG1, and PTTG2. Black represents the matched nucleotides among PTTG3P, PTTG1 and PTTG2, while white represents the differences Figure S2 Verifying the qRT-PCR primers. (A) After designing specific primer sets for PTTG3P, PTTG1, and PTTG2, PCR was performed to verify the efficacy of the primers. (B) Sequencing the PCR product to verify the specificity of PTTG3P primers. (C) Sequencing the PCR product to verify the specificity of PTTG1 primers. (D) Sequencing the PCR product to verify the specificity of PTTG2 primers