Upregulation of PAIP1 promotes the gallbladder tumorigenesis through regulating PLK1 level.

BACKGROUND: Increasing evidence suggests that elevated expression of polyA-binding protein-interacting protein 1 (PAIP1) is associated with cancer development and progression. However, how PAIP1 promotes gallbladder cancer (GBC) is still unclear.
METHODS: Two GBC tissue-derived cell lines, NOZ and GBC-SD cells, were used in this study. Assays of cell proliferation, colony formation, apoptosis, and xenograft tumor model were performed to examine the tumorigenic effects of PAIP1. Immunohistochemical (IHC) staining was used to examine the expression level of PAIP1 in both patient GBC tissues and mouse tumors. Microarray and bioinformatics analysis were used to explore the targets of PAIP1. Quantitative polymerase chain reaction (qPCR) and western blot analysis were used to validate PAIP1-mediated targets.
RESULTS: We found that upregulated PAIP1 expression was correlated with GBC. Knockdown of PAIP1 in gallbladder cells alleviated cell proliferation, promoted apoptosis, and inhibited xenograft tumor growth. Gene microarray analysis showed that stable silencing of PAIP1 altered various gene expressions. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis suggested that PAIP1 regulates cell cycle progression. Finally, we found that the PLK1 kinase, a key regulator of cell cycle, was regulated by PAIP1 at the transcriptional and protein levels. PLK1 level was positively correlated with PAIP1 level in both mouse tumors and GBC tissues. PAIP1 interacted with PLK1, and rescue of PAIP1 could recover PLK1 protein level and inhibit apoptosis.
CONCLUSIONS: Our data suggest that PAIP1 contributes to GBC progression likely through regulating PLK1 level. Since upregulated PAIP1 expression is positively associated with GBC, PAIP1 may act as a clinical prognostic biomarker of GBC. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

Gallbladder cancer (GBC) is a rare biliary tract malignancy and ranks fifth in tumors of the digestive tract worldwide (1). The 5-year survival rate of GBC patients is less than 5%, likely because of its early metastasis, late diagnosis, as well as poor prognosis (2,3). Currently, radical surgery is the only powerful curative treatment option for GBC (3), whereas chemotherapeutic approaches can only extend overall survival by a few months (4). Furthermore, due to its late and nonspecific symptoms, GBC is commonly diagnosed at advanced stages, which makes treatment choices are limited and prognosis is poor. Therefore, it is of paramount importance to explore the molecular mechanisms and potential therapeutic targets for GBC.

PolyA-binding protein interacting protein 1 (PAIP1) is encoded by the PAIP1 gene, which contains 2 interacting motifs (PAM1 and PAM2) for binding the PolyA-binding protein (PABP) (5,6). While PABP forms a complex with the eukaryotic initiation factor 4G (eIF4G) to regulate messenger RNA (mRNA) circularization, the association of PAIP1 and PABP further promotes PABP’s activity on translation initiation (5,7). Specifically, PAIP1 simultaneously interacts with PABP along with the translation initiation factors, eIF4 and eIF3, to facilitate 5’ cap-dependent translation (5,7). Furthermore, ribosomal protein S6 kinase 1 and 2 (S6K1/2) are capable of phosphorylating eIF3 to enhance PAIP1-eIF3 interaction and translation initiation in a nutrient-dependent manner (8). Meanwhile, addition of the inhibitors, rapamycin and PP242, which target the mechanistic target of rapamycin complex 1 (mTORC1), can disrupt their association and reduce the translation efficiency (8). In mice, PAIP1 specifically binds to YBX2, a RNA-binding protein that participates in mRNA storage, which stimulates translation of spermatogenic mRNAs during spermiogenesis (9). The results indicate a critical role of PAIP1 in male germ cell development. Recently, it has been reported that the eukaryotic polypeptide chain release factor, eRF3, competes with PAIP1 in binding PABP at the same domain, which attenuates PAIP1’s activity towards PABP and promotes translation termination via the PABP-eRF3 complex (10). In addition, PAIP1 forms a multiprotein complex to impede RNA deadenylation and protect mRNA decay (11). These findings clearly suggest PAIP1 plays important roles in mRNA translation and stability.

Intriguingly, an abnormally high expression of PAIP1 has been demonstrated to be linked with poor overall survival in various human cancers, including breast, gastric, liver, pancreatic, and lung cancers (12-16). For example, using different tissue-derived cell lines, Lin group has shown that overexpression of PAIP1 promotes cell proliferation, metastasis, and angiogenesis, whereas knockdown of PAIP1 inhibits these phenotypes, in both breast and pancreatic cancers, suggesting its pathological roles in tumorigenesis (12). Although PAIP1 may participate in lung adenocarcinoma via the AKT/GSK3β pathway (13) or in pancreatic cancer via AKT pathway (12), whether PAIP1 takes part in GBC remains unknown, and thus exploring the physiological function of PAIP1 and its potential relevance in GBC may be a fruitful avenue of research.

In this study, we discovered that upregulated PAIP1 expression was positively correlated with GBC. We further observed that knockdown of PAIP1 in two gallbladder cell lines inhibited cell proliferation, reduced colony formation, and induced apoptosis. At the animal level, we showed that knockdown of PAIP1 remarkably reduced xenograft tumor growth, reinforcing a critical role of PAIP1 in gallbladder oncogenesis. Mechanistically, using gene expression profiling microarray analysis we found that stable silencing of PAIP1 altered various gene expressions, including many genes that regulate cell cycle progression. Finally, found that PLK1 kinase, a key regulator of cell cycle progression, is regulated at the transcription and protein levels by PAIP1, and that rescue of PAIP1 level can restore the PLK1 level and activate cell growth. Therefore, we propose that PAIP1 likely contributes to GBC progression through regulating PLK1 and thus may be a clinical prognostic biomarker of GBC. We present the following article in accordance with the MDAR reporting checklist (available at https://dx.doi.org/10.21037/atm-21-2417).

Methods

Cell lines and culture

The human GBC cell lines, GBC-SD and NOZ, used in this study were purchased from the Cell Bank of the Chinese Academy of Science (Shanghai, China) and the Fenghui Bio Company (Changsha, China), respectively. Cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 100 U/mL penicillin/streptomycin in a humidified incubator with 5% CO2 and at 37 °C.

RNA interference and vectors

Small interfering RNAs (siRNAs) that specifically target human PAIP1 were purchased from GenePharma (Shanghai, China). The siRNA sequences were listed in Table S1. The vectors, pCS2-3xFlag-PAIP1 and pCS2-3xHA-PAIP1, were cloned from complement DNA (cDNA) extracted from HEK 293T cells, and the DNA sequences were validated by sequencing (TsingKe Company, China). Cells were cultured on 6-well plates to ~80% confluence and transfected with siRNAs or negative control using Lipofectamine 2000 (Invitrogen, USA) according to the manufacturer’s instructions. The shPAIP1 and scrambled short hairpin control (shCtrl) lentivirus vectors were purchased from GeneChem (Shanghai, China). The shPAIP1 sequences in the vector were listed in Table S1. Stably shPAIP1-transfected cells were collected though treatment with puromycin (1 μg/mL, Solarbio, China).

Collection of GBC microarray data

Microarray data sets (GSE139682, GSE132223, and GSE139682) available online were analyzed using the Gene Expression Omnibus (GEO) of the National Center for Biotechnology Information (NCBI). The cholangiocarcinoma data set was obtained from The Cancer Genome Atlas (TCGA).

Cell proliferation and cell cycle analysis

For cell growth assays, 2×103 cells per well were seeded into 96-well plates, and 3 biological wells per group were analyzed. Cell numbers in each well were evaluated using a cell counting kit-8 (CCK-8; Engreen Biosystem, Ltd., Auckland, New Zealand) after 5 days. Briefly, 10 microliters of CCK-8 reagent was added to each well, and the plate was then incubated at 37 °C for 2 hours. Subsequently, the absorbance at 490 nm was measured in each well with a spectrophotometer. For cell cycle progression analysis, the indicated cells were fixed with precooled 70% ethanol overnight at 4 °C and incubated in staining solution (5 U/mL RNase A and 10 μg/mL propidium iodide) at 37 °C for 30 min. The cell cycle distribution was assayed by flow cytometry on a FACSCalibur (BD Biosciences, USA), and the data were analyzed with FlowJo software.

Western blotting

Western blotting was performed as described previously (17). In brief, cells were collected and lysed by RIPA buffer with a protease inhibitor cocktail (Bimaker, China). Equal amounts of protein were electrophoresed on sodium dodecyl sulfate polyacrylamide gel electrophoresis gels (SDS-PAGE) and then transferred to polyvinylidene difluoride membranes (Millipore, MA, USA). After that, the membranes were blocked in buffer [5% free-fat milk in tris-buffered saline with Tween20 (TBST)] before being probed with the primary antibodies listed in the Table S2.

Apoptotic assay

shCtrl or shPAIP1 cells were trypsinized in the logarithmic growth phase for 1 hour. Both adherent and floating cells were collected and washed 3 times with ice-cold phosphate-buffered saline. Subsequently, flow cytometry was used to detect apoptotic cells double stained by Annexin V-APC staining (88-8007; eBioscience, San Diego, CA, USA) and propidium iodide within 15 minutes according to the manufacturer’s instructions. Each experiment was performed in triplicate.

Colony formation assay

Cells were seeded at a density of 1×103 cells in six-well plates and incubated at 37 °C for 10 d. The cell colonies were rinsed with PBS, then fixed with methanol for 10 min, and stained with methylthionine chloride. The number of colonies with more than 50 cells was counted with phase contrast imaging using a microscope (XDS-100, CAIKON, China).

Caspase 3/7 assay

Caspase 3/7 assay was performed according to the manufacturer’s protocol for cells cultured in a 96-well plate (G8091; Promega, Madison, WI, USA).

Transwell assays

NOZ cells were seeded into D60 plates and then cultured to 50% confluence with a concentration of 1×106 cells. Cells were transfected with 200 pmol siRNAs targeting PAIP1 using lipofectamine 2000 (Invitrogen, USA) according to the manufacturer’s instructions. Cells were transfected for 48 hours and seeded in 24-well plates (5,000 cells each) with 8-µm pore sizes (BD Bioscience, USA) for the migration assay. The invasion assay was analyzed using the same transwell inserts coated with Matrigel. Cells were seeded into the upper insert in serum-free media, and complete medium containing 10% FBS was added to the bottom wells incubated at 37 °C for 36 hours. Cells on the lower filter surface were fixed with 4% paraformaldehyde for 30 min, and stained with Giemsa for 15 min. The number of cells in three fields from three independent wells was counted.

Tumor xenograft experiments

Female BALB/c nude mice (4 weeks old) were obtained from LingChang Biotech (Shanghai, China) and raised following a regular procedure. NOZ cells (4×106/mL) transfected with the luciferase (Luc)-labeled shPAIP1 or shCtrl lentivirus construct were injected subcutaneously into the right flanks of BALB/C 10 nude mice. Tumor growth was measured following the indicated time points, and tumor sizes were calculated by the following formula: volume (cm3) = 0.5 × length × width2. After 4 weeks, 10 µL/g of D-Luciferin (15 mg/mL) was injected intraperitoneally into these mice before they were sacrificed, after which the tumors were collected and weighed. The fluorescence in each mouse was detected using the Lumina LT imaging system (PerkinElmer, Waltham, MA, USA). All procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory. Animal experiments were approved by the Ethics Committee of the Tongji Medical College, Huazhong University of Science and Technology (No. GSGC0162883).

IHC staining

A total of 22 GBC tissues were collected from patients in Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Tissue histosection slides were de-paraffinized and rehydrated following the protocols described previously (18). The slides were incubated with rabbit anti-human polyclonal antibodies against PAIP1 (1:200 dilution) and Ki-67 (1:100 dilution) at 4 °C overnight and then probed with biotin-conjugated goat anti–rabbit secondary antibody at room temperature for 30 min. Slides were incubated with streptavidin-peroxidase complex. The peroxidase reaction was developed with 3, 3’-diaminobenzidine and counterstained with Mayer’s hematoxylin. All procedures performed in this study involving human participants were in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Tongji Medical College, Huazhong University of Science and Technology (LLHBCH2019LW-003), and informed consent was provided by all participants.

The intensity score was determined by evaluating staining intensity of positive staining as described previously (19). Samples with histoscore of more than 4 were considered to be high, while those with a score less than 4 were considered to be low.

Coimmunoprecipitation (co-IP) experiment

Co-IP experiments were performed following the procedures as described before (20). The resulting immune-precipitates were subjected to immunoblotting with the indicated antibodies. To construct the pCS2-3xFlag-PLK1-TurboID plasmid, the pCS2-3xFlag-PLK1 reported previously (21) was digested with XbaI and XhoI restriction enzymes. The DNA fragment of TurboID was cloned into this digested vector by a HieffClone Plus multi One-Step Cloning Kit (10912ES10, Yesen, China). The TurboID fragment was amplified by PCR from a commercial gene (USA126049; Addgene, Watertown, MA, USA). The sequences of the primers were listed in Table S1. For biotin labeling of transiently transfected cells, the pCS2-3xFlag-PLK1-TurboID construct was transfected into HEK 293T cells, and biotin at a final concentration of 50 µM was added just 6 hours before cells were harvested (22). To isolate the biotinylated proteins that bind with PLK1, cells lysates were denatured and incubated with NeutrAvidin agarose (29204; Thermo Fisher Scientific, USA) at 4 °C for 4 hours with rotation. After being washed with IP buffer 3 times, the resins were subjected to run gel and immunoblot with the indicated antibodies.

Human gene expression array

The total RNA sample was analyzed by Agilent 2100 using GeneChip 3’ in vitro translation Express Kit (Affymetrix, USA). The Affymetrix gene expression profiling microarray chip was treated as described before (20), and scanned to obtain images and raw data. A differentially expressed gene list is available in Table S1.

Statistical analysis

Data are presented as mean ± standard deviation (SD) from three experimental replicates. The Student’s t-test was used to analyze the differences between groups using GraphPad Prism 5. Statistical differences between two groups were analyzed using an independent Student’s t-test (2-tailed). A P value less than 0.05 was considered to be statistically significant. For quantification of the western blot signals, ImageJ software was used to measure the relative intensity of each band.

Results

PAIP1 expression was upregulated in GBC tissues

To investigate whether PAIP1 plays a pathologic role in GBC, relative mRNA expression levels of PAIP1 in normal tissue vs. tumor tissues in two different data sets from the GEO were analyzed. We found that PAIP1 expression levels were significantly upregulated in GBC tissues compared to normal tissues in both data sets (). It has been illustrated that GBC together with cholangiocarcinoma (CCA) constitutes a group of biliary tract cancers (19,23). To clarify whether PAIP1 is only upregulated in GBC tissues, we further analyzed expression levels of PAIP1 in CCA tissues obtained from the TCGA database. We found that PAIP1 was markedly upregulated in CCA, suggesting PAIP1’s broader effect in biliary tract cancers (). To confirm whether higher mRNA levels of PAIP1 are correlated with higher protein levels of PAIP1 in GBC tissues, we assessed PAIP1 protein levels in 22 GBC tissues by IHC staining (). Based on the histoscores, PAIP1 protein was assessed as having a higher expression in most GBC tissues (15 out of 22 gallbladder tumor tissue samples), which confirmed that PAIP1 expression is upregulated in GBC tissues ().

Figure 1

Abnormal expression of PAIP1 was associated with gallbladder cancers. (A,B) The mRNA levels of PAIP1 in each pair of normal tissues and GBC tissues obtained from 2 different expression profiling data sets (GSE139682 and GSE132223) were analyzed respectively. (C) The mRNA levels of PAIP1 in normal cholangio tissues and cholangiocarcinoma tissues obtained from TCGA were compared. (D,E) The expression levels of PAIP1 from different GBC patients were examined by IHC staining, and the representative images with low or high PAIP1 expression are shown (D). Scores of PAIP1 staining in different types of GBC tissue samples were plotted (E). Data are presented as mean ± SD. *P<0.05; **P<0.01; ***P<0.001. GBC, gallbladder cancer; TCGA, The Cancer Genome Atlas.

Stable silencing of PAIP1 suppressed cell proliferation, migration, and invasion

Next, two different gall bladder-derived cancer cell lines, NOZ and GBC-SD cells, were used to investigate the causal role of PAIP1 in GBC progression. Lentivirus vector-based shRNA construct bearing a specific sequence targeting PAIP1 gene (shPAIP1) was generated and transfected into NOZ and GBC-SD cells with shCtrl serving as a control. After virus-infected cells treated with puromycin were harvested, the PAIP1 knockdown efficiency was validated by qPCR analysis and western blotting. We observed that mRNA levels of PAIP1 were significantly decreased in both cell lines upon shRNA treatment (, Figure S1A), and PAIP1 protein levels were also decreased accordingly (, supplemental Figure S1B). As expected, knockdown of PAIP1 significantly retarded cell growth, as assessed by the counting of cell numbers over time (, Figure S1C). Cell proliferation assays revealed a suppressed capacity of cell proliferation in PAIP1-silenced cells as compared with controls (, Figure S1D). Furthermore, colony formation assays also showed decreased cell colony numbers in cells expressing shPAIP1 than in cells expressing shCtrl, regardless of which gallbladder cell lines were used (, Figure S1E,F). Meanwhile, transwell assays were performed to examine the effect of PAIP1 in GBC cell migration and invasion. We found that knockdown of PAIP1 markedly inhibited cell migration and invasion abilities (). Altogether, these data support that PAIP1 plays a role in GBC progression, migration and invasion.

Figure 2

Reduction of PAIP1 expression suppressed gallbladder NOZ tumor cell growth, colony formation, cell migration and invasion in vitro. (A,B) PAIP1 mRNA or protein levels in NOZ cells infected with shCtrl or shPAIP1 lentivirus construct were examined by qPCR (A) and western blotting (B), respectively. (C,D) Knockdown of PAIP1 retarded cell proliferation, as revealed by cell counting over time (C) and measures of the optical density value with the CCK-8 kit (D). (E,F) Colony formation assays were performed upon knockdown of PAIP. The representative images are displayed by staining with methylthionine chloride shown in (E) and the quantifications of colony numbers from 3 independent plates are plotted (F). (G,H) Transwell assays were performed to examine the effects of PAIP1 on the migration and invasion of NOZ cells. The representative images stained with Giemsa are shown in (G) and the quantification of cell numbers from 3 biological replicates are plotted (H). Data are presented as mean ± SD. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001. CCK-8, cell counting kit-8.

Reduction of PAIP1 level promoted gallbladder cell apoptosis and caused cell cycle arrest

Since knockdown of PAIP1 suppresses cell proliferation, we wanted to know whether this results from induced apoptosis of living cells or the arrest of cell cycle progression. Toward this end, annexin-V marked apoptotic cells were counted by flow cytometry, in which annexin-V specifically bound to phosphatidylserine exposed on the outer cellular membrane in dead cells (24). From experiments performed in triplicate, we observed that knockdown of PAIP1 increased the cell population staining with annexin-V (, Figure S2A) and facilitated cell apoptosis (, Figure S2B) in NOZ and GBC-SD cells, respectively. Caspases, including caspase-3 and caspase-7, are crucial for final apoptotic execution, and caspases become cleaved upon activation (25). Thus, a sensitive, single-step method based on bioluminescence protease reaction was carried out using peptide-conjugated aminoluciferin as the protease substrate and a firefly luciferase that has been molecularly evolved for increased stability (26). Consistently, an increased portion of apoptotic cells were observed upon knockdown of PAIP1 (, Figure S2C). To rule out the possibility that induced apoptosis in cells with stable silencing of PAIP1 might have been an off-target effect, 2 siRNA oligos were transiently transfected into NOZ cells, and the protein levels of caspases were examined. We found that both siRNAs targeting PAIP1 genes induced caspase-3 and caspase-7 activation as confirmed by the increase in protein levels and accumulation of cleaved caspase-3 forms, which indicated apoptotic cell progress had occurred (). Additionally, flow cytometry analysis demonstrated that knockdown of PAIP1 caused cell cycle arrest, but somehow, with an increased G1 cell population in NOZ cells () but an increased G2/M population in GBC-SD cells (Figure S2D). Thus, our data suggest that the suppression of cell proliferation upon knockdown of PAIP1 in GBC cells is likely due to induced apoptosis and cell cycle arrest.

Figure 3

Reduction of PAIP1 expression promoted NOZ cell apoptosis and caused cell cycle arrest. (A,B) Flow cytometry analysis was performed determine the percentage of apoptotic cells upon knockdown of PAIP1. The numbers in the top right corner represent apoptotic cells. Three independent images in each group are shown (A), and quantitative data are plotted (B). (C) Caspase-Glo 3/7 assays were performed in cells expressing shCtrl or shPAIP1. The caspase 3/7 activity was measured by luminescence. (D,E) Two different siRNA oligos targeting PAIP1 were transfected into NOZ cells, and protein levels of caspase-3 and caspase-7 were examined by immunoblotting (D). The indicated protein levels from 3 biological replicates were quantified (E) using ImageJ. (F) Flow cytometry was performed to examine the cell populations at different cell cycle stages. Data are presented as mean ± SD. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.

Knockdown of PAIP1 attenuated tumorigenesis in mice

To evaluate the pathological role of PAIP1 in vivo, we developed a xenograft tumor model by inoculating equal amounts of Luc-labelled shCtrl or shPAIP1 cells into the right flanks of nude mice (, left panel). We followed the growth of tumors over time by measuring the tumor size and weight. Compared with the control mice that exhibited a gradual increase of tumor size, the shPAIP1 mice exhibited much slower growth of tumors (). At 22 days post injection, the mice were sacrificed and the tumors were weighed. The tumors were consistently found to be much smaller upon knockdown of PAIP1 as compared to those of control mice (). In line with our results, upon addition of D-Luciferin into mice before they were killed, we observed stronger luminescent signals in the control mice than in the shPAIP1 mice (, right panel). IHC staining of mice tumor tissues revealed that knockdown of PAIP1 led to reduced expression of Ki-67, a cell proliferation marker, confirming PAIP1 level to be directly correlated with cell proliferation and tumorigenesis ().

Figure 4

Silencing of PAIP1 expression attenuated the malignant phenotype in mice. (A,B,C,D) Xenograft assays were performed using NOZ cells stably expressing shCtrl or shPAIP1. The representative images of nude mice (from n=10) with tumors are shown in the left panel, and luminescent imaging is shown in right panel (A). Tumor sizes were measured at the indicated time points (B). The tumor images (C) and tumor weight after 22-day injection are presented (D). (E) Representative IHC staining images of tumors from mice injected with shCtrl or shPAIP1 cells were stained with the indicated antibodies. HE, hematoxylin and eosin. Data are presented as mean ± SD. ***P<0.001; ****P<0.0001. IHC, immunohistochemical.

PAIP1 regulated expression of multiple genes involved in cell cycle progression

To explore the molecular mechanism of PAIP1 in gallbladder cell proliferation and tumorigenesis, an Affymetrix microarray chip was used to identify potential downstream targets of PAIP1. Subsequently, NOZ cells with shCtrl or shPAIP1 construct were harvested, and total mRNAs were purified and subjected to hybridization of the microarray chip, with the expression profiles being analyzed. We repeatedly observed that knockdown of PAIP1 markedly altered multiple gene expression levels as compared to the shCtrl cells in 3 independent samples ( and in total online: https://cdn.amegroups.cn/static/public/atm-21-2417-1.xlsx). When we set a selected threshold with 2-fold changes and a P value of less than 0.05, 983 upregulated genes and 1212 downregulated genes were identified upon knockdown of PAIP1 relative to the shCtrl group (). The hierarchical clustering analysis revealed that the differentially expressed genes were mainly enriched in DNA replication and cell cycle (). The expression changes of several selected genes involved in these 2 pathways were validated by qPCR analysis. We noticed that knockdown of PAIP1 significantly inhibited expression of MKI67 (marker of proliferation Ki-67), PLK1 (Polo-like kinase 1), and SKP2 (S-phase kinase-associated protein 2), and promoted expression of CDKN1B (cyclin-dependent kinase inhibitor 1B), ITGA2 (alpha 2 subunit of VLA-2 receptor), and TXNIP (thioredoxin interacting protein), which is consistent with the microarray data; meanwhile, there were no changes of expression of JUN (jun proto-oncogene), which is inconsistent with the microarray data (). To identify which genes were indeed affected by PAIP1, the protein levels encoded by some of these genes were examined by immunoblotting. Surprisingly, we found that PLK1 level, but not other protein levels, was dramatically decreased upon knockdown of PAIP1, which might indicate the relevance of PLK1 and PAIP1 ().

Figure 5

Microarray analysis indicated that PAIP1 may affect cell cycle progression to regulate gallbladder cell proliferation. (A) Heatmap shows the differential expression of genes upon stable silencing of PAIP1. Three independent replicates of each group were compared. (B) Volcano plots indicating the differentially expressed genes from the overlapping of 3 replicates in shPAIP1 cells relative to shCtrl cells. The vertical dashed lines in the plot represent logarithm values normalized fold changes of ±1.0. The horizontal dashed line represents an adjusted P value of 0.05. [log2(FC) >1, P<0.05]. Red dots represent upregulated genes [983], and blue dots represent downregulated genes [1,212]. (C) KEGG pathway shows the top enriched significantly changed gene clusters. The corresponding gene numbers in each cluster are present. (D,E) mRNA or protein levels of several identified factors upon knockdown of PAIP1 in NOZ cells were examined by qPCR (D) and western blotting (E), respectively. Data are presented as mean ± SD. **P<0.01; ****P<0.0001; n.s, not significant. KEGG, Kyoto Encyclopedia of Genes and Genomes.

PAIP1 promoted gallbladder tumorigenesis possibly through regulation of PLK1 level

It has been well demonstrated that PLK1 is a crucial regulator of cell cycle progression (27,28). Elevated PLK1 levels are present in many type of cancers, including GBC, and are closely linked with poor prognosis (29,30). Consistently, we also found that PLK1 levels were higher in GBC tissues obtained from the aforementioned datasets (). Interestingly, PLK1 expression levels were positively correlated with PAIP1 expression levels in these tissue samples (R=0.37, P=2.9e-05; ). We conducted immunostaining of the 22 collected human GBC tissues with the PLK1 antibody, and we found that 9 samples exhibited low expression levels, while 13 samples exhibited high expression levels of PLK1 (Figure S3A,B). Consistently, PLK1 protein levels were positively correlated with PAIP1 protein levels in the 22 GBC tissue samples (R=0.7656, P<0.001; Figure S3C). To further confirm the relationship of PLK1 and PAIP1, we examined protein levels of PLK1 in xenograft tumor tissues bearing either the shCtrl or shPAIP1 lentivirus construct. The PLK1 levels significantly were reduced in shPAIP1 samples compared with shCtrl samples (), which underscores our point that PLK1 levels are positively correlated with PAIP1 levels in both the mouse model and patient GBC tissues.

Figure 6

PAIP1 regulates gallbladder tumors likely through regulating PLK1. (A) The mRNA levels of PLK1 in each pair of normal tissues and GBC tissues obtained from expression profiling of the GSE139682 data sets are shown. **P<0.01. (B) The correlation profiling patterns of normalized expression between PAIP1 and PLK1 in 118 cholangiocarcinoma patient tissues from GSE89749 are shown. The correlation coefficient R value and statistical P value are shown. (C,D) The protein levels of PLK1 from mouse xenograft tumor tissues bearing shCtrl and shPAIP1 constructs were examined by immunohistochemical staining, and the representative images with low or high PLK1 expression are shown (C). Scores of PLK1 staining of mouse tumor samples are plotted (D). Data are presented as mean ± SD. *P<0.05. (E) Western blot analysis showed the interaction of PAIP1 and PLK1 in cells expressing Flag-PLK1-TurboID in the presence or absence of biotin. The interaction of PLK1 and Rad51 served as a positive control. (F,G) Reciprocal co-IP assays were performed in HEK 293T cells to show the interaction of PAIP1 with PLK1. (H) Western blot analysis and examination of caspase protein levels showed that reintroduction of PAIP1 rescued apoptotic phenotypes in NOZ cells. GBC, gallbladder cancer.

It has been reported that PAIP1 regulates mRNA translation efficiency and development through interaction with several proteins (5,9,10). To test whether PAIP1 regulates GBC cell proliferation by binding PLK1, co-IP experiments were performed. Using a TurboID technique that could efficiently identify proximal and interacting proteins in their natural cellular environment (22), we found that PLK1 was able to bind with PAIP1 in HEK 293T cells (). Reciprocal co-IP results further confirmed that PLK1 interacts with PAIP1 (). Finally, we wondered if reintroduction of PAIP1 in PAIP1-silenced cells could recover the reduced PLK1 level. Thus, a Flag-tagged PAIP1 construct was transfected into NOZ cells with two siPAIP1 oligos, and we found that PLK1 levels were increased upon PAIP1 level recovery (). Moreover, caspases activities were also rescued, indicating the importance of PAIP1 in apoptosis (). Collectively, the results support the supposition that PAIP1 regulates gallbladder tumors through the regulation of PLK 1.

Discussion

GBC remains an understudied disease. Although its incidence is less than 2 per 100,000 individuals worldwide, patients with GBC are predominantly diagnosed at the advanced stages with 5-year survival rates being <5% (3). In contrast to more common human tumors, evidence about the molecular changes involved in the development of GBC are less known thus far. Therefore, it is crucial to identify specific regulators that are required for gallbladder carcinogenesis. The tumor suppressor genes, including TP53, ARID1A, and SMAD4, and the oncogenes, including CDKN2A/B, KRAS, and EGFR, have been characterized with mutations or genomic alterations in GBC, and this may provide direction for clinical treatment (31,32). However, only a subset of patients is directly linked with these mutations. It is thus critical to discover other key factors functioning in GBC development and progression.

In the present study, we identified poly A-binding protein-interacting protein PAIP1 as a potential regulators involved in GBC development and progression. Elevated PAIP1 level is associated with gallbladder carcinogenesis, including GBC and CCA. Using two different GBC cell lines, we further demonstrated that reduction of PAIP1 levels in tumor cells retarded cell proliferation, colony formation, promoted apoptosis, and caused cell cycle arrest. Given that we failed to obtain cholangiocarcinoma-derived cell lines, we did not examine whether the similar phenotypes would occur upon silencing of PAIP1. However, the fact that upregulated PAIP1 levels were correlated with cholangiocarcinoma tissues may lead us to investigate whether PAIP1 is also an important regulator in CCA.

To explore the possible molecular mechanism underlying how PAIP1 affects GBC, we performed high-throughput microarray chip analysis. This revealed that PAIP1 may participate in regulating cell cycle progression. After a comprehensive survey, we found that PAIP1 significantly affect a key cell cycle regulator, PLK1, thereby activating functions in GBC development and progression. Overexpression of PLK1 has been shown to occur in a wide range of tumors, and this phenomenon has attracted considerable attention from researchers in academic and commercial pharmaceutical settings seeking to develop PLK1 inhibitors as a means of cancer treatment (29). The present study is the first of its kind to link PAIP1 with PLK1 genetically and physically, which might open a window for clinical treatment of GBC.

There are a few reasons to assume PAIP1 is a regulator in gallbladder tumors acting via PLK1. First, it has been reported that PAIP1 exerts its tumor-promoting role by facilitating cell cycle progression through regulating the expression of cyclin D1 in gastric cancer (14). Researchers have shown that PAIP1 knockdown results in decreased expression of cyclin D1, whereas overexpression of PAIP1 enhances its expression. Unfortunately, we did not find that decreased PAIP1 altered cyclin D1 expression in our microarray dataset in total online: https://cdn.amegroups.cn/static/public/atm-21-2417-1.xlsx. As PLK1 is a critical cell cycle regulator, and knockdown of PAIP1 causes cell cycle arrest, it is reasonable to assume that they are linked. Second, we found that PAIP1 significantly affected he PLK1 protein level (), and an elevated PAIP1 level is positively correlated with higher PLK1 level in gallbladder tumors. Since PLK1 inhibition has been shown to cause cell cycle block and apoptosis, PLK1 inhibitors may be a potential cancer therapy for treatment of GBC.

However, how PAIP1 regulates PLK1 is still unclear, and several possible interactions underlying this relationship may be explored in future research. Given that PAIP1 interacts with the polyA-interacting protein PABP to promote translational efficiency, it is likely that PAIP1 affects the mRNA translation or mRNA stability of PLK1. Alternatively, PAIP1 may form a complex with PLK1 to affect cell cycle progression. Testing these hypotheses is a worthwhile endeavor and may provide insights into the clinical treatment of gallbladder carcinogenesis.

The article’s supplementary files as