PNO1 promotes human cholangiocarcinoma tumorigenesis and chemosensitivity


 Background

The partner of NOB1 homolog (PNO1) is important for ribosome biogenesis and serves as an oncogene in several cancers. However, the role of PNO1 in cholangiocarcinoma (CHOL) remains largely unknown.
Methods

The method of mRNA microarray analysis, high-throughput screening technologies, on-line databases analysis, PDXs models, biochemistry and molecular biology have been utilized to reveal the role of PNO1 in the progression of CHOL.
Results

PNO1 was significantly upregulated in CHOL tissues and predicted poor prognosis. PNO1 could induce cell proliferation and metastasis in vitro and in vivo. In CHOL cells expressing wild-type p53, PNO1 knockdown increased expression of p53 and its downstream gene p21 and decreased cell viability; these effects were blocked by p53 knockout and attenuated by the p53 inhibitor PFT-a. Furthermore, PNO1 knockdown reduced global protein synthesis and inhibiting MDM2-mediated ubiquitination and p53 degradation. Moreover, PNO1 overexpression enhanced the sensitivity to bortezomib treatment in CHOL. In addition, MYC overexpression promoted PNO1 promoter activity while MYC knockdown decreased PNO1 mRNA and protein, which led to decreased cell viability and clone formation. The expression of MYC was found positively correlated with PNO1 in CHOL patients.
Conclusions

Collectively, upregulation of PNO1 by MYC promotes cholangiocarcinoma tumorigenesis through suppressing p53 signaling pathway and enhanced the sensitivity to bortezomib treatment. PNO1 serves as a candidate of therapeutic target in CHOL treatment and clinical chemotherapy regimen.


Background
The partner of NOB1 homolog (PNO1) is important for ribosome biogenesis and serves as an oncogene in several cancers. However, the role of PNO1 in cholangiocarcinoma (CHOL) remains largely unknown.

Methods
The method of mRNA microarray analysis, high-throughput screening technologies, on-line databases analysis, PDXs models, biochemistry and molecular biology have been utilized to reveal the role of PNO1 in the progression of CHOL.

Results
PNO1 was signi cantly upregulated in CHOL tissues and predicted poor prognosis. PNO1 could induce cell proliferation and metastasis in vitro and in vivo. In CHOL cells expressing wild-type p53, PNO1 knockdown increased expression of p53 and its downstream gene p21 and decreased cell viability; these effects were blocked by p53 knockout and attenuated by the p53 inhibitor PFT-a. Furthermore, PNO1 knockdown reduced global protein synthesis and inhibiting MDM2-mediated ubiquitination and p53 degradation. Moreover, PNO1 overexpression enhanced the sensitivity to bortezomib treatment in CHOL.
In addition, MYC overexpression promoted PNO1 promoter activity while MYC knockdown decreased PNO1 mRNA and protein, which led to decreased cell viability and clone formation. The expression of MYC was found positively correlated with PNO1 in CHOL patients.

Conclusions
Collectively, upregulation of PNO1 by MYC promotes cholangiocarcinoma tumorigenesis through suppressing p53 signaling pathway and enhanced the sensitivity to bortezomib treatment. PNO1 serves as a candidate of therapeutic target in CHOL treatment and clinical chemotherapy regimen.

Background
Cholangiocarcinoma is a highly malignant tumor that often arises from biliary tract epithelia. It commonly groups into intrahepatic cholangiocarcinoma (iCCA), perihilar cholangiocarcinoma (pCCA) and distal cholangiocarcinoma (dCCA) by anatomical location [1][2][3]. CHOL accounts for 15-20% of primary hepatobiliary malignancies, while the number of cases caused by CHOL continues to increase worldwide [4,5]. It is mostly in the advanced stage when CHOL diagnosis is made when a patient is given a poor prognosis. Patients who are inoperable, often fail to chemotherapy and have limited of options for targeted therapy [6][7][8]. Therefore, exploring the underlying molecular mechanisms of CHOL development, especially identifying novel prognostic biomarkers improve the prognosis and help to diagnosis CHOL in an early stage, is urgently needed.
Ribosome assembly factor PNO1, is also known as the RNA-binding protein "partner of NOB1", plays an essential role in ribosome biogenesis [9][10][11]. It is responsible for the cleavage of 18S mediated by binding to NOB1, while its deletion could inhibit the synthesis of 18S rRNA and 40S subunit [12,13]. In addition to the numerous studies of ribosome-related function of PNO1, several studies have reported that PNO1 acts as an oncogene in cancer cells, including colorectal cancer, lung adenocarcinoma and urinary bladder carcinoma [14][15][16]. Therefore, exploring the relationship between PNO1 and CHOL is essential to the future exploration of new targeted therapies.
In this study, the method of mRNA microarray analysis, high-throughput screening technologies, on-line databases analysis, PDXs models, biochemistry and molecular biology have been utilized, leading to the con rmation that PNO1 is overexpressed in CHOL tissues and MYC-mediated upregulation of PNO1 could contribute to CHOL progression by inhibiting the p53/p21 signaling pathway and enhanced the sensitivity to bortezomib treatment. Furthermore, PNO1 might act as a speci c prognostic biomarker of CHOL.

Materials And Methods
Patients and tissue specimens 10 pairs of CHOL tissues (T) and para-carcinoma tissues (P) were obtained from (Shandong Provincial Hospital A liated to Shandong First Medical University). The para-carcinoma tissues were used for control. Tissue chips were purchased from OUTDO BIOTECH (shanghai, China). Clinicopathologic characteristics of patients could be seen in Supplementary Table S1. Samples stained with hematoxylin and eosin were veri ed by experienced pathologists.

Microarray analysis
Gene expression pro le was detected in 10 pairs of CHOL tissues and para-carcinoma tissues, as well as in RBE cells infected with shCtrl or shPNO1 by microarray analysis. The para-carcinoma tissues and cells transfected with shCtrl were used as the control group. The RNeasy kit (Sigma) method was implied to extract the total RNA, and the concentration of which was determined by Nanodrop 2000 (Thermo Fisher Scienti c). RIN value was evaluated with Agilent 2100 and Agilent RNA 6000 Nano Kit (Agilent). RNA sequencing was performed with Affymetrix human GeneChip PrimeView according to the manufacturer's instruction, and the outcomes were scanned by Affymetrix Scanner 3000 (Affymetrix). Welch t-test with Benjamin-Hochberg FDR. was used to perform the raw data statistical signi cance assessment. In addition, signi cant difference analysis and functional analysis based on KEGG pathway enrichment analysis were executed.
Cell culture and transfection Human cholangiocarcinoma cell lines (HuCCT1, RBE, HuH28 and TFK-1) and HEK293T were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Human biliary epithelial cells (HIBEpiC) were obtained from Fenghbio (Hunan, China). Wild-type RBE cells (RBE/p53 +/+ ) and RBE cells lacking the p53 gene (RBE/p53 -/-) were produced from laboratory. HuCCT1, RBE, HuH28, TFK-1, RBE/p53 +/+ and RBE/p53 -/cells were grown in 90% RPMI-1640 with 10% FBS (Corning); HEK293T cells were grown in 90% DMEM with 10% FBS (Corning). Cells were grown at 37 ˚C in a humidi ed atmosphere of 5% CO 2 . All the plasmids involved in this paper were obtained from Genechem (Shanghai, China). The information of sequences could be found in Supplementary Table S2. Cell lines in logarithmic phase were infected with lentivirus vectors with polybrene at a multiplicity of infection of 20 for 12~16 h, and then maintained in the new fresh complete medium. After 72 hours of infection, when the infection e ciency of lentivirus on cells was observed to be over 85% using uorescence microscope (Olympus), cells were harvested for the follow-up experiments. Cells transfected with shCtrl or si-Ctrl were used as the control group.

Real-time quantitative PCR
Total RNA from the cell lines or tissue samples was isolated, respectively using TRIZOL (Sigma), reverse transcription of RNA (2.0 µg) to cDNA was performed using the PrimeScript RT reagent kit (Takara) following the manufacturer's instructions. RT-PCR was carried out using the SYBR Premix Ex Tag (Takara) on VII7 real time PCR instrument. 2 -ΔΔCt Method was used to analysis the relative quantitative of target genes with GAPDH as the internal reference. Primers for genes were synthesized by Sangon and shown in Supplementary Table S3. The para-carcinoma tissues and cells transfected with shCtrl or si-Ctrl were used as the control group.

IHC staining
Tissue sections were incubated with PNO1 antibody (1:100, Sigma) at 4°C overnight, then washed with phosphate-buffered saline (PBS) for 3 times, and incubated with horseradish peroxidase (HRP) conjugated goat anti-mouse IgG polyclonal antibody for 30 min at room temperature. DAB and hematoxylin were used to stain tissue slides. IHC scoring of patients were determined based on the combined evaluation of the staining intensity and staining extent scores. The para-carcinoma tissues were used as the control group.

Western Blotting
Cells and tissues were lysed in ice-cold RIPA buffer (Millipore) and protein concentration was detected by BCA Protein Assay Kit (HyClone-Pierce). The para-carcinoma tissues and cells transfected with shCtrl or si-Ctrl were used as the control group. The same amount of total protein from each group was separated by 12% SDS-PAGE and transferred onto a PVDF membranes (Bio-Rad Laboratory). The membranes were blocked with primary antibodies in TBST plus 5% non-fat milk. After being washed with TBST, membranes were incubated with the appropriate secondary antibody. The signals were visualized using ECL-PLUS Kit (Amersham). Band intensities were quanti ed using Image J software (NIH). The information of antibodies was shown in Supplementary Table S4. Celigo cell counting assay 72 h after the infection, cells were seeded onto a 96-well plate with the cell density of 2,000 cells/well/100 µL in RPMI-1640 medium containing 10% FBS and then further cultured for 5 days. Cells transfected with shCtrl were used as the control group. Cell counting was accomplished every day by Celigo image cytometer (Nexcelom Bioscience) and the cell proliferation curve was drawn.

CCK8 assay
Lentivirus infected CHOL cells were measured by CCK8 assay. Cells in the logarithmic phase were seeded onto 96-well plates (2,000 cells/well). CCK8 solution (10 µL, MCE) was incubated with cells for 2.5 h. The OD450 was measured by microplate reader (Tecan) and the cell viability ratio was calculated. Cells transfected with shCtrl were used as the control group.

Cell apoptosis
Infected cells were inoculated in a 6-well plate and cells were harvested when the cell density reached over 50 %. After being washed with 4°C ice-cold PBS, cells were centrifuged at 1000 rpm and resuspended with 200 µL binding buffer. Evaluation of apoptosis was performed by Annexin V-APC staining ow cytometry method (eBioscience) following the manufacturer's protocol. Cells transfected with shCtrl were used as the control group.

Cell-cycle analysis
Infected cells were inoculated in a 6-well plate and cells were harvested when the cell density reached 70 %. After being washed with 4°C ice-cold PBS, cells were centrifuged at 1000 rpm and xed with 70% ethanol at 4°C overnight. Then the cells were washed again extensively, and then incubated with FxCycle PI/RNase Staining Solution (Thermo Fisher Scienti c) for 20 minutes. DNA content was analyzed by FACS and the proportion of DNA in different phases was analyzed using FlowJo version 10.0. Cells transfected with shCtrl were used as the control group.

Colony formation assay
Lentivirus infected cells were inoculated in a 6-well plate with 800 cells/well/ 2 mL complete growth medium. Medium was exchanged every 2-3 days. Cell clones were photographed by an Olympus digital camera, then xed and stained by 4% paraformaldehyde and Giemsa (Solarbio), respectively. The numbers of colonies were counted using photoshop 2020. Cells transfected with shCtrl were used as the control group.
Transwell migration assay Transwell assay was performed using Corning chamber. Lentivirus infected cells were seeded in a 24-well plate with 5×10 4 cells/well the upper chambers lled with 100 µL serum-free medium. 600 µL DMEM medium plus 30% FBS were lled in the lower chamber and incubated at 37°C for 24 h (HuCCT1: 48 h). At the end of incubation, oating cells were removed and cells in lower chamber were xed and stained with crystal violet. Images of cells were taken and analyzed using NIH image J software. Cells transfected with shCtrl were used as the control group.
Promoter activity assay RBE cells were transduced with a plasmid expressing luciferase reporter under the control of a p53-driven promoter and lentivirus encoding anti-PNO1 or control shRNAs (the control group) for 4-6 hours. After 48 hours, cells were washed twice with PBS, and harvested with 1× PLB from the Dual-Luciferase Reporter Assay System (Promega). Then cells were lysed, centrifuged, and the supernatant was collected. Aliquots of supernatant (40 µL) were added to 96-well plates, followed by 20 µL luciferase assay reagent (Promega) at room temperature. Luciferase activity was measured immediately using a luminometer (Orion II Microplate Luminometer, Berthold Detection Systems). Data were normalized to the results obtained for the internal control Renilla luciferase.
To examine whether the 5 predicted transcription factors have indeed modulated PNO1 transcription, plasmids were constructed via encoding the full-length predicted factors. A separate plasmid was constructed encoding luciferase reporter downstream of a 2.0-kb fragment of the PNO1 promoter. Plasmid integrity was con rmed by DNA sequencing. Brie y, HEK293T or RBE cells were co-transfected with each of 5 different overexpression plasmids or MYC overexpression plasmid (500 ng), Renilla luciferase reporter (20 ng), and luciferase reporter (500 ng) using Lipofectamine 3000 Transfection reagent (Thermo Fisher Scienti c). Promoter activities were measured after 48-hour transfection.
Northern blot analysis of 18S rRNA Extracted total RNA (10 µg /sample) was fractionated on a 1.2% agarose-formaldehyde gel and transferred to Hybond NÞ membranes (Amersham). BIO-labeled probe was purchased from Thermo Fisher Scienti c. The membrane was hybridized overnight in hybridization buffer (Thermo Fisher Scienti c), washed twice (5 minutes each) under low-stringency conditions (2 × SSC, 0.1% SDS), blocked with blocking buffer at 50 ˚C for 30 minutes, then incubated at 50 ˚C for 20 minutes with stabilized streptavidin-horseradish peroxidase conjugate (1:300). Membranes were exposed to a phosphor storage screen and visualized using the Chemi Doc XRS + System (Bio-Rad). Cells transfected with shCtrl were used as the control group.

Protein synthesis assay
Global protein synthesis was assayed using a commercial kit (Cayman Chemical). Transfected RBE cells were incubated for 30 minutes with o-propargyl-puromycin in complete medium in 96-well plates (100 µL/well). Cells transfected with si-Ctrl were used as the control group. Cells were xed with cell-based assay xative (100 µL/well), stained with 5 FAM-azide (Cayman Chemical), and photographed under a uorescence microscope (Accu-Scope). Images were collected using MetaMorph image acquisition software (Molecular Devices). Fluorescence intensity in single cells was quantitated using Image Pro Plus software (Media Cybernetics).

Xenograft tumor studies
All animal procedures were carried out in accordance with the Helsinki Declaration and were approved by the Ethics Committee of Shandong Provincial Hospital A liated to Shandong First Medical University (NO.2019-156). For cell line tumor xenografts, 0.2 mL exponentially growing lentivirus infected HuCCT1 cell suspensions at a destiny of 2× 10 7 cell/ml were injected subcutaneously into BALB/c nude mice (4-6 weeks old). For metastasis mouse models, 0.1 ml lentivirus infected HuCCT1 cell suspensions at a destiny of 1× 10 7 cell/ml were injected intravenously into BALB/c nude mice (4-6 weeks old). For the PDX model, tumor tissues were obtained directly from the operating room, dissected carefully, trimmed into small pieces (approximately 2~3-mm 3 ) under sterile conditions and implanted subcutaneously into the anks of female NOD/SCID mice. After 2 months of tumor growth, the established PDX tumors were transplanted subcutaneously into the anks of female NOD/SCID mice. When the tumors reached about 100 mm 3 , the mice received an intraperitoneal dose of bortezomib (0.5 mg/kg, twice a week) for bortezomib treatment, and the normal group were treated with Saline for comparison at the same time. Tumor growth was assessed twice a week using a caliper and tumor volumes (V) were estimated as V=π/6×L×W×W (L and W was tumor length and width, respectively). Mice were sacri ced at humane endpoints and the tumor tissues were removed for immunostaining.

Ki-67 staining assay
Tissue slides were blocked with 3% PBS-H 2 O 2 and were incubated with primary antibody Ki-67 at 4°C overnight. Then slides were incubated with HRP goat anti-rabbit IgG at room temperature for 2 h. Finally, all slides were stained by Hematoxylin (# BA4041, Baso) and Eosin (# BA4022, Baso). Tumors transfected with shCtrl were used as the control group.

Statistical Analysis
Student's T-Test and one-way ANOVA were used to analyze the statistical signi cance. All statistical analysis was performed using SPSS 22.0 (IBM, SPSS) and GraphPad Prism 8.01 (GraphPad Software). Each bar represents the mean values ± SD of three independent experiments. P < 0.05 was considered statistically signi cant. Kaplan-Meier survival analysis was performed using the log-rank test. The primary endpoint of the animal study was de ned as the tumor growth, tumor volumes and KI67 expression in tumor tissues.

Results
Identi cation of PNO1 as a novel target in CHOL At present study, we rst utilized 10 pairs of CHOL tissues and non-tumorous bile ducts tissues for microarray analysis, which revealed 2,394 upregulated DEGs and 1,562 downregulated DEGs ( Figure 1A and 1B). To identify potential oncogenes, we selected 20 up-regulated DEGs that have not been reported in CHOL-related study ( Figure 1C). Using high-content screening, we have detected cell growth and the knockdown expression of selected genes in RBE cells with lentivirus-transduced target-shRNA (Supplementary Figure S1A). The results suggested that cell growth was distinctly inhibited in shSYT13, shCDCA5, shTMED3, shPNO1 and shDEPDC1 groups ( Figure 1D-F). Considering the most obvious inhibitory effects of PNO1 knockdown, PNO1 was selected for subsequent studies.

PNO1 is upregulated in CHOL tissue samples and predicts poor prognosis
To further evaluate the expression pattern of PNO1 in CHOL, the protein expression of PNO1 in 83 CHOL tissues and 31 para-carcinoma tissues were measured using IHC. The results showed that PNO1 was signi cantly upregulated in tumor tissues ( Figure 1G and 1H). Based on detailed tumor characteristics information from the CHOL patients, we found that high expression of PNO1 was positively correlated with advanced pathological grading, TNM stage, recurrence patients and poor prognosis of CHOL patients ( Figure 1I-L). The PNO1 mRNA expression levels were also con rmed in CHOL by RT-PCR ( Figure  1M). The results suggested that PNO1 might serve as an oncogenic role in CHOL tumor tumorigenesis.

PNO1 promotes CHOL cell proliferation and migration in vitro.
PNO1 is highly expressed in cultured CHOL cell lines (HuCCT1, RBE, HuH28 and TFK-1), compared to normal bile duct epithelial cell line HIBEpic (Supplementary Figure S1B). The decreased of PNO1 mRNA expression in three shRNA or si-RNA (PNO1-target) suggested successful transfection, and shPNO1-1 was selected for subsequent research because of its most effective interference veri ed by CCK8 assay in RBE cells (Supplementary Figure S1C-E). In cultured HuCCT1 and RBE cells, the downregulation of both PNO1 protein and mRNA levels were veri ed (Figure 2A), and the inhibition effects of shPNO1 on cell proliferation and migration were con rmed, including decreased in cell viability ( Figure 2B), percentage of cells in S phase ( Figure 2C), colony formation ( Figure 2E) and chemotaxis cells ( Figure 2F), and also inducing the percentage of cells to undergo apoptosis ( Figure 2D). On the contrary, the upregulation of PNO1 in HuH28 and TFK-1 cells could increase cell viability, colony formation and chemotaxis cells ( Figure 2G-J).

PNO1 knockdown inhibits CHOL tumorigenesis in vivo
Mice xenograft models were used through subcutaneous injection of HuCCT1 cell models to verify the effect of PNO1 knockdown on tumor growth in vivo. We selected a small sample size (n=10), because initial intention was to gather basic evidence regarding the role of PNO1 in CHOL in more complex experimental designs. The measurement of tumor volume started at day 11 post tumor-inoculation and end up with day 32. The results of which illustrated a lower tumor weight and signi cantly smaller tumor size compared with controls ( Figure 3A-C). Moreover, Ki-67 staining was used to detect proliferative activity of tumors, which was shown to be signi cantly weaker in shPNO1 group than shCtrl group ( Figure 3D). The result also showed the expression of PNO1 at the mRNA level in tumors ( Figure 3E). In addition, metastasis mouse models by intravenous injection were established. After six weeks, the mice were sacri ced and the lung, liver, organ in gastrointestinal tract were harvested. We found that PNO1 downregulation HuCCT1 cells formed fewer lung and liver metastatic nodules by a Living Image System ( Figure 3F).
Oncogenic activities of PNO1 depend on p53/p21 signaling To further explore the mechanism of PNO1 on CHOL development, microarray analysis was performed to determine DEGs in RBE cells transduced with lentivirus encoding either shPNO1 or shCtrl and revealed 2,083 upregulated DEGs and 1,805 downregulated DEGs ( Figure 4A and 4B). Combined with KEGG pathway enrichment analysis, the results indicated that PNO1 may participate in several functional pathways involved in cell proliferation and migration ( Figure 4C). The expression of 15 selected DEGs from the 10 most enriched signaling pathways were consistent with the microarray analysis and p53 was the most strongly upregulated gene in response to PNO1 knockdown by RT-PCR ( Figure 4D). Furthermore, the analysis of STRING database indicated that p53 could bind to other 14 DEGs and may be the most potential effector of PNO1 in CHOL ( Figure 4E). Indeed, we demonstrated that PNO1 knockdown increased luciferase activity in the p53-driven luciferase reporter assay ( Figure 4F). Western blot analysis con rmed increased protein expression of both p53 and p21 ( Figure 4G). The inhibitory effects of PNO1 knockdown were prevented when p53 was knocked out ( Figure 4H and I), and the inhibitory effects were attenuated when cells were treated with the p53 inhibitor PFT-a ( Figure 4J). PNO1 knockdown activates p53 by inhibiting MDM2-mediated ubiquitination and degradation of p53 RNA pro ling and Northern blotting veri ed that PNO1 knockdown in RBE cells signi cantly decreased amounts of 18S rRNA ( Figure 5A), 40S and 60S subunits, and the 80S ribosome ( Figure 5B). Protein synthesis assay showed that PNO1 knockdown in RBE cells visibly inhibited of global protein synthesis ( Figure 5C). Western blot analysis and coimmunoprecipitation (co-IP) assay indicated that PNO1 knockdown led to a reduction in degradation of p53 ( Figure 5D) and ubiquitination of p53 ( Figure 5E). Moreover, co-IP analysis indicated that PNO1 knockdown increased the binding of RPL11 to MDM2( Figure 5F).

PNO1 facilitates the therapeutic effect of bortezomib in CHOL
Meanwhile, we found that PNO1 knockdown in RBE cells could increase the protein level of PTEN while overexpression of PNO1 could reduce the expression of PTEN ( Figure 6A). In addition, co-IP analysis indicated that PNO1 knockdown increased the binding of p53 to PTEN ( Figure 6B). It has been reported that PTEN attenuated bortezomib e cacy in CHOL treatment [17]. To explore the effect of manipulating PNO1 abundance on sensitivity to bortezomib, cell models in vitro and PDXs models in vivo were established. As shown in CCK8 assays, RBE cell viability was inhibited by 0.5 µM concentration of bortezomib when compared with saline, whereas an enhanced therapeutic effect was discovered upon PNO1 overexpression ( Figure 6C). Immunohistochemistry (IHC) staining revealed low expression of PNO1 and concurrent up-regulation of p53 and PTEN expression in tumor tissues from CC1 when compared with CC2 ( Figure 6D). Consistent with the in vitro ndings, bortezomib treatment was effective at limiting tumor growth in PNO1-high expression PDXs when compared with PNO1-low expression PDXs ( Figure 6E and 6F).

Transcription factor MYC positively regulates PNO1 expression in CHOL
To explore the upstream mechanism of PNO1 on CHOL, we identi ed MYC as an important transcription factor of PNO1 using TFBIND screening of transcription factor binding sites combined with DEGs from mRNA microarray analysis of RBE cells ( Figure 7A-C). We found that MYC knockdown distinctly decreased expression levels of PNO1 mRNA and protein in RBE cells ( Figure 7D and 7E), while also inducing the mRNA expression of p53 and p21. Moreover, MYC knockdown in RBE cells visibly decreased cell viability and colony formation ( Figure 7F and 7G). In addition, MYC knockdown also induced cell cycle arrest and cell apoptosis ( Figure 7H and 7I). The result demonstrated that MYC was highly expressed in CHOL tissues and was positively correlated with PNO1 based on data from on-line GEPIA2 ( Figure 7J and 7H). All of the above results led to the suggestion that MYC overexpression may help drive CHOL by triggering PNO1 overexpression.

Discussion
Several past studies have con rmed PNO1 as an oncogene in tumorigenic processes. For example, Peng revealed that mammalian PNO1 worked as a ribosome assembly factor in colorectal cancer cells, and its oncogenic effects could serve as a novel therapeutic strategy in colorectal cancer [14]. Guo offered the evidence that PNO1, a direct functional target of miR-340-5p, was related with Notch pathway and EMT, which acts as promising target for the treatment of lung adenocarcinoma in the future [16]. In the present study, we found that PNO1 was upregulated in CHOL tissues and knockdown the expression of PNO1 could inhibit RBE cell growth in vitro, which might be predictive of the potential of PNO1 as a prognostic biomarker and its oncogenic effects in CHOL progression.
To further con rm the expression pattern of PNO1 in CHOL, we detected the PNO1 protein levels of 83 CHOL clinical samples by IHC staining. Indeed, these results were consistent with the previous mRNA microarray analysis: PNO1 was upregulated in CHOL tissues. Moreover, based on tissue microarray analysis, high PNO1 expression was positively correlated with tumor grade and poor survival of patients with CHOL. These results highlighted the potential of PNO1 as a prognostic biomarker and its oncogenic effects in CHOL progression.
In our study, we demonstrated that PNO1 was signi cantly upregulated in CHOL cell lines. The knockdown of PNO1 repressed cell growth, colony formation, cell migration, induced cell cycle arrest, cell apoptosis, and also inhibited tumor growth and metastasis in vivo and vitro. On the contrary, overexpression of PNO1 promoted cell proliferation and migration. The tumor suppressor effects of PNO1 knockdown had been previously reported in cell cultures, and this is the rst work in which the suppressor activity of shPNO1 is reported for in xenograft models in vivo. Our ndings may result in extrapolating our results to humans and showing a promising eld of research emerges in the area of CHOL. These results suggested that PNO1 might serve as a potential target molecule in CHOL patients. Subsequently, we explored the possible mechanism of oncogenic function of PNO1 using mRNA microarray analysis. Based on the DEGs between RBE cells treated with shPNO1 or shCtrl, KEGG pathway enrichment analysis was used to identify the possible functional pathways. The results showed that PNO1 could possibly participate in several functional pathways involved in cell cycle progression, cell proliferation and metastasis. The mRNA levels of selected DEGs involved in the 10 most enrichment signaling pathway were consistent with the mRNA microarray analysis by RT-PCR. Combined with the result from the STRING database, we focused on the function of p53 at the follow-up study. It is known that tumor suppressor p53 involved in a variety of cellular events, for instance, p53/p21 signaling could induce cell cycle arrest and apoptosis as well as repress cell proliferation; oncogenes could promote MDM2-mediated degradation and ubiquitination of p53 [18][19][20][21][22][23][24][25][26][27][28][29]. In our study, we con rmed that the knockdown of PNO1 could increase luciferase activity by a p53-driven luciferase reporter assay and could enhance the protein levels of p53, CDKN1A (p21) and MDM2. Moreover, the knockout of p53 or the addition of p53 inhibitor PFT-a could enhance cell viability, which could be reduced by the knockdown of PNO1.
It was well known the role of ribosome biogenesis in the control of a cell to synthesize proteins and hence acts an essential role in cell proliferation and migration, which was associated with many diseases including cancer [30][31][32]. In the present study, we found that PNO1 knockdown in RBE cells decreased the amounts of 18S rRNA, 40S subunits, 60S subunits, and the 80S ribosome, which led to signi cant inhibition of global protein synthesis by RNA pro ling and Northern blotting. Meanwhile, a p53-driven luciferase reporter assay and quantitative PCR analysis indicated that PNO1 knockdown could increase luciferase activity and p53 mRNA expression, while also leading to a decrease in the degradation and ubiquitination of p53. The results suggested that PNO1 knockdown could activate p53 signaling. It has been revealed that ribosome protein RPL11 binds to MDM2 and inhibits its ubiquitin ligase activity toward p53, resulting in p53 accumulation [18,20,30,[33][34][35][36][37][38]. We con rmed that PNO1 knockdown distinctly increased the binding of RPL11 to MDM2, suggesting that ribosomal stress increased the binding of RPL11 to MDM2 and decreased ubiquitination and degradation of p53, which may contribute to the accumulation of p53.
In addition, results from cell models and patient-derived xenografts (PDX) models indicated that PNO1 overexpression enhanced the therapeutic effect for BZ in CHOL. Meanwhile, one rejoices in surprise realizing that PNO1 de cient caused a signi cant increase in protein levels of PTEN and p53 in vivo. A previous study reported that PTEN inhibited proteasome activity and attenuated bortezomib e cacy in CHOL [39]. PTEN might modulate the function of proteasomes [40]. In our study, PNO1 knockdown suppressed ubiquitin degradation of p53. As well know that The PTEN and p53 tumor suppressors are among the most commonly inactivated or mutated genes in human cancer [41]. Of note, the p53 protein cooperates with PTEN and might be an essential blockage in development of many tumors [42].
Collectively, we guessed that knockdown of PNO1 elevated PTEN/p53 expression and then depressed the therapeutic effect for BZ in CHOL.
Indeed, MYC could act as a regulator of ribosome biogenesis and an important target for cooperative actions of p53 and PTEN in the regulation of tumor development [43][44][45]. In our study, MYC was identi ed as a potential transcription factor of PNO1. We suggested that MYC knockdown reduced expression levels of PNO1 mRNA and protein in RBE cells, while also increasing the levels of p53and p21. Furthermore, MYC knockdown in RBE cells visibly reduced cell viability and induced cell cycle arrest as well as cell apoptosis. We also identi ed that MYC expression was positively correlated with PNO1 based on on-line CHOL databases analysis. These results suggested that MYC overexpression may contribute to driving CHOL via triggering PNO1 overexpression, which was consistent with previous studies stating that MYC was an oncogene in cancers[46-48]. However, detailed mechanism of how MYC regulates PNO1 expression needs to be explored through further studies.

Conclusions
In conclusion, our study identi ed PNO1 as a tumor promotor in CHOL, which could be possibly used as a prognostic indicator and novel therapeutic target in the treatment of CHOL. We found that PNO1 was upregulated in CHOL tissues and its upregulation or downregulation could promote or suppress the development and progression of CHOL respectively. Furthermore, our results provided evidence that PNO1 could be regulated by transcription factor MYC, and its oncogenic effects were regulated by p53 signaling pathway. Our ndings also revealed that the relationship between PNO1/PTEN/p53 axis and bortezomib sensitivity in CHOL treatment. These ndings highlighted the potential of PNO1 as a novel targeted therapy in CHOL.

Consent for publication
Not applicable.

Data availability
The datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.