Pancreatic cancer tissues were collected from 68 patients during surgery at the First Affiliated Hospital of Dalian Medical University, China between 2002 and 2006. Ten adjunct non-cancerous tissues were also collected as the controls. All the tissues were fixed by formalin and embedded in paraffin wax for histological and immunohistochemical experiments. The pancreatic cancer tissues were pancreatic ductal adenocarcinoma. The cancers were staged according to the American Joint Committee on Cancer (AJCC) standards (29). All the procedures with regard to patient recruitment, informed consent, sample collection and processing were approved by the IRB of Dalian Medical University.

Earle’s balanced salt solution (EBSS), 3-methyl-adenine (3-MA) and monodansylcadaverine (MDC) were purchased from Sigma. LysoTracker Red and kinase inhibitors (PD98059 and LY294002) were purchased from the Beyotime Institute of Biotechnology. 2′,2′-Difluorodeoxycytidine gemcitabine (GEM) was obtained from Dalian Melone, Biotechnology Co., Ltd. USP22-specific shRNA, negative control shRNA (shNC), USP22 expression construct and the blank vector were designed and synthesized by GenePharma. USP22 antibody was purchased from Abcam, LC3 antibody from Sigma, antibodies against BECN1, SQSTM1, Bcl-2, caspase-3, AKT1 from Proteintech, and antibodies against phospho-AKT1 (Ser473), ERK1/2 and phospho-ERK1/2 from Bioworld.

The human pancreatic cell line (Panc-1) was purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Panc-1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; HyClone) supplemented with 10% fetal bovine serum, 100 IU/ml penicillin and 100 μg/ml streptomycin at 37°C with 5% CO₂. Transfection of Panc-1 cells was performed using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s instructions.

The general procedure of IHC was performed according to the protocol described (30). The USP22 antibody was used at 1:400 dilution and LC3 antibody at 1:200 dilution. Biotin-labeled secondary antibodies were used for visualization through HRP-streptavidin with 3,3′-diaminobenzidine (DAB) substrate. The counterstaining was carried out with haematoxylin. Staining with pre-immune IgG was used as the control.

Panc-1 cells were collected by trypsin digestion method and fixed with 2.5% glutaraldehyde. The cell pellets were then fixed with 1% osmic acid. After a series of dehydration, cell pellets were embedded in Embedding Medium (Sigma). Ultrathin sections (50–70 nm) were prepared using Leica EM UC6 ultramicrotome and stained with uranyl acetate and lead citrate, followed by observation on a JEM-2000EX transmission electron microscope.

Cells were stained with LysoTracker Red in phosphate-buffered saline (PBS) or 0.1 mM MDC in DMEM at 37°C for 30 min in the dark. After washing three times with PBS, the cells were examined using fluorescence microscopy.

Panc-1 cells were fixed with 4% paraformaldehyde, and then permeabilized with 0.1% Triton X. After blocking with goat serum, the cells were incubated with USP22 or LC3 antibody overnight at 4°C, followed by incubation with corresponding secondary antibody for 40 min at room temperature. The cells were counterstained with DAPI for observation by fluorescence microscopy.

Cell proliferation was measured by the CCK-8 Kit (Beyotime Institute of Biotechnology) according to the manufacturer’s instructions. Cells in 100 μl media were reacted with 10 μl CCK-8 reagent at 37°C for 2 h, followed by measuring the absorbance at 450 nm on a microplate reader. The data were analyzed using SPSS software.

Cells were harvested and fixed with 70% ice-cold ethanol for 24 h, and then stained with propidium iodide (PI) for cell cycle analysis by flow cytometry. The data were analyzed using SPSS.

Cells were harvested and stained with the Annexin V-FITC Apoptosis Detection kit (Beyotime Institute of Biotechnology). The apoptotic cells were determined by flow cytometry and the data were analyzed using SPSS.

Cells grown in 96-well plates were incubated with 10 μM DCF-DA for 20 min at 37°C. The DCF fluorescence (Ex _{485 nm} and Em _{535 nm}) was measured using a multimode plate reader. The data were analyzed by SPSS.

Cells were lysed using RIPA buffer [25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS] with protease and phosphatase inhibitors (EMD Biosciences). After sonication for 2 min at 4°C and centrifugation for 10 min at 4°C, the supernatant was taken as the total cell lysate. Protein concentration was quantified using the Bradford method. Equal amounts of total protein (50 μg) were analyzed by SDS-polyacrylamide gel (SDS-PAGE). Following transfer to the NC membrane, the blot was probed with primary antibodies as indicated and then incubated with HRP-labeled secondary antibodies for visualization using enhanced chemiluminesence reagents (Thermo). The images were obtained by the Bio-Rad Imaging System.

The overall survival curves were generated using the Kaplan-Meier method. The relationship between USP22 and LC3 in pancreatic cancer tissues was analyzed by the Spearman rank correlation analysis. The relationship between the level of LC3 or USP22 expression and clinicopathological characteristics was examined using the χ² test. The differences among groups were analyzed by one-way ANOVA and the Student-Newman-Keuls (SNK)-q test using SPSS 17.0 software. Differences were considered to indicate a statistically significant result with a P-value <0.05.

To examine whether USP22 is overexpressed in pancreatic cancer and determine its relationship with autophagy and pancreatic cancer prognosis, 68 pancreatic cancer tissue samples and 10 adjacent normal pancreatic tissue samples were collected and analyzed in the present study. IHC staining was used to detect USP22 and LC3 in situ. By analyzing several slides each for all 78 samples, it was found that both USP22 and LC3 were expressed in these patient samples to different extents. The representative images are shown in Fig. 1. Notably, USP22 was not expressed in normal adjacent tissues, but was overexpressed in advanced pancreatic cancer tissues, suggesting that USP22 may play an important role in pancreatic cancer progression. By contrast, LC3 was expressed at basal level in normal pancreatic tissue and also elevated to a high level in advanced pancreatic cancer tissue, indicating that it is required for physiological and pathological autophagy. In addition, it was found that USP22 was localized to the cytoplasm and nucleus, whereas LC3 was localized to the cytoplasm only, both of which are consistent with their physiological functions.

Quantitative analysis of all IHC results revealed that there were 66.2% of pancreatic cancer samples expressing USP22 and 52.9% for LC3. However, in adjacent normal tissues the expression of USP22 and LC3 was 0 and 10%, respectivley (Table I). The USP22 and LC3 proteins showed a significant difference between pancreatic cancer and the adjacent normal tissue (P=0.000, 0.011). Further statistical analysis demonstrated that 44.1% of pancreatic cancer patients were USP22- and LC3-positive, whereas the percentage for double-negative in USP22 and LC3 was 25%. Tumors (22.1%) were USP22-positive but LC3-negative, while 8.8% of tumors were LC3-positive but USP22-negative (Table II). The association analysis using SPSS indicated that the correlation between USP22 and LC3 was strong in pancreatic cancer (ρ=0.385, P=0.001). These results suggested that the overexpression of USP22 is highly related with autophagy in pancreatic cancer, indicating the significance of investigating the pathological role of USP22 in pancreatic cancer.

To clarify the relativity of USP22 and autophagy to pancreatic cancer prognosis, we systematically analyzed the correlation of USP22 expression and autophagy with all clinicopathological characterics of pancreatic cancer from the 68 patients. The cancer stages were determined according to the AJCC system. Based on the IHC analyses of all 68 samples as above, we identified the following relationships (Table III): i) tumor differentiation, lymphatic vessel infiltration and cancer stage were associated with the expression of USP22 and LC3 (P<0.05); ii) pancreatic external invasion was associated with USP22 (P<0.05) but not with LC3 (P>0.05); iii) age, gender, tumor location and tumor size were not associated with the expression of USP22 and LC3 (P>0.05). These results proved the positive relationship between the expression of USP22 and LC3 and the progression of pancreatic cancer, suggesting that USP22 overexpression may be a causal factor for pancreatic cancer.

To confirm the relationship between USP22 and LC3 and the outcome of pancreatic cancer, the survival curve of patients was analyzed using the Kaplan-Meier method. The results demonstrated that the overexpression of USP22 and LC3 was significantly correlated with short survival time (Fig. 2), indicating a close relationship between USP22 overexpression and/or enhanced autophagy and the poor prognosis of pancreatic cancer patients.

Results from the clinical samples above suggested that USP22 was able to promote LC3 expression, therefore leading to autophagy. To clarify this possibility, the Panc-1 pancreatic cancer cell line was utilized for the experiments. As shown in Fig. 3A, when USP22 was downregulated by shRNA or overexpressed by USP22 cDNA construct, USP22 levels were altered accordingly. USP22 shRNA was found to markedly knock down USP22, concomitant with a notable decrease in LC3-II, indicating that USP22 was a factor for promoting LC3 processing to generate a more active form LC3-II. SQSTM1, a selective substrate of autophagy, was also decreased following USP22 overexpression, demonstrating that USP22 promoted authophagy. However, the other autophagy component beclin-1 was not affected by USP22, suggesting the early stage of autophagy may not be the target of USP22. Immunofluorescent staining experiments further supported the high efficiency of USP22 knockdown or overexpression (Fig. 3B) and the corresponding change in LC3 (Fig. 3C).

To investigate whether autophagy flux is promoted by USP22, LysoTracker Red staining (specific staining for lysosome) and MDC staining (specific staining for autophagosome) were performed. As shown in Fig. 3D, the number of lysosome and autophagosome were increased by the overexpression of USP22, suggesting that the formation of autophagosome fused with lysosome was promoted by USP22. Furthermore, TCM analysis also showed the increased number of autophagosome following the overexpression of USP22 (Fig. 3E). Taken together, these results demonstrated that USP22 is a factor that promotes autophagy in pancreatic cancer cells.

To investigate the molecular mechanism underlying the augmented autophagy by USP22, MAPK and PI3K/AKT pathways were specifically analyzed in the present study. Using RNAi and gene overexpression combined with the application of kinase inhibitors, it was found that USP22 overexpression slightly increased ERK1/2 activity as reflected by the phosphorylated form of ERK1/2. The level of LC3-II was concomitantly increased (Fig. 4A). In addition, inactivation of ERK1/2 by a specific inhibitor (PS98059) abolished LC3-II formation, suggesting that ERK1/2 is one of the kinases involved in the promotion of LC3-II formation and subsequent autophagy, which may occur through the phosphorylation of LC3 as observed in the regulation of LC3 by PKA and PKC (31,32). However, the inhibition of AKT1 by a PI3K inhibitor (LY294002) did not alter the level of LC3-II (Fig. 4B). Thus, these results demonstrated that ERK activation is involved in USP22-induced autophagy.

To elucidate the mechanism of USP22-induced autophagy on the prognosis of pancreatic cancer patients, cell death and survival pathways were further examined. When an autophagy inhibitor 3-MA was used to inhibit autophagy in Panc-1 cells, it was found to be effective as indicated by the decrease of LC3-II (Fig. 5A). The effect of USP22-induced autophagy on apoptosis was then analyzed. Western blotting of anti-apoptotic Bcl-2 and pro-apoptotic caspase-3 showed almost no difference following the overexpression or knockdown of USP22 or treatment with 3-MA (Fig. 5B). Further analysis of apoptosis with flow cytometry did not show any obvious changes (Fig. 5C, upper panel). These results demonstrated that USP22-induced autophagy has no significant effect on apoptosis in Panc-1 cells. We also examined Panc-1 cell proliferation under various levels of USP22. It was found that the proliferation of USP22 shRNA-tranfected cells was reduced by ~30% as compared to USP22-overexpressed cells. The 3-MA treatment reduced the proliferation of USP22 cells by ~20%, but only by 10% for USP22 shRNA-transfected cells (Fig. 5C, bottom panel). Flow cytometric analysis revealed that the percentage of S phase in USP22-overexpressed cells was increased by 2-fold over the USP22 shRNA-transfected cells and by 1.4-fold over the USP22-overexpressed and 3-MA-treated cells (Fig. 5D). These data confirm that the function of USP22-induced autophagy promoted Panc-1 cell proliferation.

Nutritional deficiency is a characteristic of the pancreatic cancer microenvironment (33), which is closely related to the development of pancreatic cancer (34,35). To determine whether USP22-induced autophagy enhances resistance to starvation, Panc-1 cells were starved in EBSS for various timepoints. USP22-overexpressed cells showed enhanced resistance to starvation as compared to USP22 shRNA-transfected cells. However, the capacity of resistance to starvation was decreased after treatment with 3-MA (Fig. 5E), suggesting that USP22 conferred resistance to nutritional starvation is mediated by autophagy.

When Panc-1 cells were treated with a chemotherapy drug gemcitabine (36–38), it was found that the apoptotic rate of USP22-overexpressed cells was less than that of the USP22 shRNA-transfected cells. However, the apoptotic rate was increased by combinational treatment with gemcitabine and 3-MA (Fig. 5F, upper panel). The cell proliferation assay demonstrated that the inhibition rate of proliferation by USP22 overexpression and gemcitabine treatment was less than that by the knockdown of USP22 and gemcitabine treatment. Combined treatment with gemcitabine and 3-MA further inhibited Panc-1 cell proliferation (Fig. 5F, lower panel). These results demonstrated that USP22 also promoted cell survival under chemotherapeutic condition through autophagy. Since gemcitabine functions as a nucleoside metabolic inhibitor, it generates cellular stresses, such as ROS. As shown in Fig. 5G, knockdown of USP22 plus gemcitabine treatment increased the ROS level as compared to USP22 overexpression and gemcitabine treatment. Additional treatment with 3-MA increased the level of ROS, suggesting that the prevention of ROS production by USP22 may be another mechanism for USP22-mediated cell survival.

The aforementioned results indicated that USP22-induced autophagy increases cell proliferation and the resistance to stresses, such as starvation and chemotherapy, thereby synergistically promoting the cell survival of pancreatic cancer.