Expression of USP22 and Survivin is an indicator of malignant behavior in hepatocellular carcinoma

Subsequently to the publication of the above article, an interested reader drew to the authors' attention that a pair of data panels showing the results of immunohistochemical staining experiments in Fig. 1 on p. 2210 had already appeared in a previously published paper (Zhang Y, Li Y, Lin C, Ding J, Liao G and Tang B: Aberrant upregulation of 14‑3‑3σ and EZH2 expression serves as an inferior prognostic biomarker for hepatocellular carcinoma. PLos ONE 9: e107251, 2014).  Owing to the fact that the contentious data in the above article had already been published prior to its submission to International Journal of Oncology, the Editor has decided that this paper should be retracted from the Journal. The authors were asked for an explanation to account for these concerns, but the Editorial Office did not receive a satisfactory reply. The Editor apologizes to the readership for any inconvenience caused. [International Journal of Oncology 47: 2208‑2216, 2015; DOI: 10.3892/ijo.2015.3214].


Introduction
Hepatocellular carcinoma (HCC) is the most common primary malignant liver tumor type, ranking as the third leading cause of all cancer-related deaths in the world (1,2).The post-surgical 5-year survival rate is low, largely due to the high recurrence rate, as observed in Shanghai, China (3).Since tumorigenesis and tumor progression of hepatic cells are the result of multiple genetic alterations, a single molecule targeting therapy has yet to be discovered.Thus, the identification of target molecules that control the biological characteristics of HCC is of great importance.
Ubiquitin-specific protease 22 (USP22) is a novel deubiquitinating enzyme that cleaves ubiquitin (Ub) from Ub-conjugated protein substrates (4).In humans, the USP22 gene is located on chromosome 17 and is expressed in various adult tissues and at an early embryonic stage (4).USP22 is considered to be a cancer stem cell marker that induces therapy resistance, tumor aggressiveness and metastatic dissemination; its expression is associated with a poor outcome in various cancers including colon, breast and liver cancer (5)(6)(7)(8)(9).USP22 is also an enzymatic subunit of the hSAGA transcriptional cofactor complex, which is required for activator-driven transcription, cell cycle progression and tumorigenesis (5,6).USP22 is required for the transcription of target genes regulated by the MYC oncoprotein and other sequence-specific activators that require hSAGA activity (5).However, the detailed mechanism by which USP22 affects these processes remains unknown.USP22 plays a role in telomere maintenance by deubiquitinating the shelterin protein TRF1 to prevent its degradation (10).More recent studies have demonstrated that USP22 can inhibit the transcription of the p21 gene by deubiquitinating the transcriptional regulator FBP1, leading to cell proliferation and tumorigenesis (11).It remains unclear whether USP22 plays a role in deubiquitinating other proteins.
Survivin is a member of the inhibitor of apoptosis protein (IAP) family and inhibits apoptosis (12).Recently, it has also been shown to function as a subunit of the chromosomal passenger complex (CPC) to regulate cell division (13).Survivin expression is associated with a poor outcome in a variety of human cancers (14).It has been shown that protein level of survivin is regulated by the ubiquitin-proteasome pathway (15,16).Although the detailed mechanism of survivin degradation is still unclear, translocation of survivin to the nucleus accelerated its degradation in an APCCdh1-dependent manner (17).USP22 may play a role in regulating survivin through deubiquitination.
In the present study, we investigated the USP22 expression and its correlation with clinicopathological features and survivin expression in HCC.Moreover, we examined the biological role of USP22 in HCC cells.

Materials and methods
Patients and tissue samples.HCC and normal tissues were obtained from 4 patients who underwent surgery at the Affiliated Hospital of Guilin Medical University.These tissue specimens were frozen and stored at -80˚C.Paraffin-embedded pathology specimens from 151 patients with HCC were obtained from the archives of the Department of Pathology, the Affiliated Hospital of Guilin Medical University, Guilin, China.All patients underwent complete surgical resection between 2001 and 2007.All samples were obtained after approval by the Ethics Committees of Guilin Medical University.These HCC cases included 133 men (88.1%) and 18 women (11.9%).The ages of the patients ranged from 24 to 77 years, with a mean age of 51 years.Histologically, 73 cases were classified as well differentiated, 48 cases were moderately differentiated and 30 cases were poorly differentiated HCC.The clinicopathological characteristics for these patients included the age, gender, serum α-fetoprotein (AFP) level, tumor size, tumor differentiation, vascular invasion and tumor stage.The tumors were classified and graded based on the pTNM classification advocated by the International Union Against Cancer.In addition, 30 samples of normal liver tissues were examined as a control.The tissues were fixed in 10% buffered formalin and embedded in paraffin.
Immunohistochemical staining.For the immunohistochemical examination, serial 4-µm sections were stained with hematoxylin and eosin and used for immunohistochemical analyses.The slides were incubated with primary USP22 antibody (monoclonal goat; 1:50; Abcam, ab71732) and polyclonal antisurvivin antibody (1:200, sc-10811; Santa Cruz Biotechnology) at 4˚C overnight after antigen retrieval by microwave treatment in citrate buffer (pH 6.0); detection was performed by the streptavidin-biotin peroxidase system (Universal LSAB™2 kit; Dako, Tokyo, Japan).The immunohistochemistry grades were defined as high and low according to the number of cells stained and the intensity of the reaction in individual cells.The expression of USP22 and survivin was graded as high (over 20% of tumor cells showed strong or diffuse immunopositivity) and low (<20% of the tumor cells showed weak or patchy immunopositivity or no staining).
Western blot analysis.The hepatic tissue and cell lines were incubated with RIPA buffer on ice before being subjected to western blot analysis.The protein concentration was detected by the Bradford method with BSA (Sigma-Aldrich) as the standard.Equal amounts of cell and tissue extract (40 µg) were subjected to SDS-PAGE and transferred to nitrocellulose membrane (Bio-Rad Laboratories) for antibody blotting.The membrane was then blocked and incubated with primary and secondary antibodies.The primary antibodies were USP22 (ab71732; Abcam), survivin (sc-10811; Santa Cruz Biotechnology), cyclin B and p21 (Transduction Laboratories, San Jose, CA, USA) and β-actin (TA-09; ZSGB-BIO, Beijing, China).
Silencing by siRNA and cell viability.HepG2 and SK-Hep-1 cells were transfected with 150 pmol of siRNA USP-22 in 1 ml of OPTI-MEM according to the manufacturer's instructions.For the RNAi downregulation of USP22, the USP22 sequencespecific siRNA and negative control siRNA (Guangzhou Ribobio, Co., Ltd., Guangzhou, China) were designed and synthesized as previously described.Following siRNA treatment (48 h), HepG2 and SK-Hep-1 cells were used for in vitro MTT assay as described above.Cells (2x10 4 cells/well) were cultured in 96-well culture plates for 1 day.Cells were collected after 24, 48 and 72 h to detect each indicator.MTT solution (5 mg/ml; 20 µl) was added to each well, and cells were cultured in a CO 2 incubator for 4 h.Then, the culture solution was removed, and 150 µl of DMSO was added to each well and agitated at room temperature for 10 min.The OD values of each well were measured using a microplate reader.The analysis for each experimental group was performed in 6-double wells.The average values were calculated and growth curves were constructed.
Statistical analysis.The SPSS v.17.0 (SPSS, Inc., Chicago, IL, USA) software package was used for analysis.All data are presented as the means ± standard deviation.The α2 test and Fisher's and t-test were employed to evaluate the relationship between USP22, survivin expression and clinicopathological variables.The survival analyses were conducted according to the Kaplan-Meier method, and survival characteristics were compared using log-rank tests.All P-values quoted were twosided, and P<0.05 was considered statistically significant.

Results
High expression of USP22 and survivin in HCC cases.We examined the expression of USP22 and survivin in 30 normal adjacent liver samples and 151 HCC tissues by immunohistochemistry.In normal adjacent liver tissues, all cases showed low expression of USP22 (Fig. 1).Among 151 HCC cases, 80 (53.0%) cases showed high expression of USP22 in their cytoplasm (Fig. 1).Similarly to USP2 expression, 92 (60.9%) cases showed high expression of survivin in their cytoplasm.Then, we compared USP22 expression or survivin expression with clinicopathological features including age, gender, tumor stage, tumor size, serum α-fetoprotein (AFP) level, tumor differentiation and vascular invasion in 151 HCC cases (Table Ⅰ).Notably, high expression of USP22 was significantly correlated with tumor size, poor differentiation and tumor stage (Fig. 2A).Survivin expression was also significantly correlated with tumor size, poor differentiation and tumor stage (Fig. 2B and Table Ⅰ).In addition, survivin expression was significantly correlated with vascular invasion and serum level of AFP protein (Fig. 2B and Table Ⅰ).Thus, high expression of USP22 and survivin was frequently observed and well correlated with malignant behavior in HCC cases.

Correlation between USP22 and survivin expression in HCC.
Next, we examined the correlation between USP22 and survivin in HCC cases.Among 151 HCC cases, 80 cases showed high expression of USP22, and 71 cases showed low expression of USP22 (Table Ⅰ).Among 80 cases with high expression of USP22, 69 cases showed high expression of survivin.Moreover, among 71 cases with low expression of USP22, 48 cases showed low expression of survivin.Thus, we demonstrated that USP22 expression was well correlated with survivin expression in HCC (Table Ⅱ).
To confirm the correlation between USP22 and survivin, we examined the expression of USP22 and survivin in HCC cell lines and tissues by western blot analysis and RT-PCR.In this experiment, we used 7 HCC cell lines (Hep3B, Bel-7402, HepG2, SK-Hep-1, SMMC-7721, HuH-7 and QGY-7701), a normal hepatic cell line (LO2), 4 HCC tissues and 4 normal adjacent liver tissues.USP22 expression was observed in 7 HCC cell lines, but not in LO2 cells (Fig. 3B).Similarly to USP22, survivin expression was also observed in 7 HCC cell lines, but not in LO2 cells (Fig. 3B).Expression levels of USP22 mRNA and protein in HCC tissues were higher than those in normal adjacent liver tissues (Fig. 3C).Survivin showed similar expression pattern in HCC tissues.These findings support the immunohistochemical data showing that: i) higher expression of USP22 and survivin was observed in HCC, but not in normal liver tissue; and ii) USP22 expression was well correlated with survivin expression in HCC.

Poor prognosis in HCC patients with high expression of USP22 and survivin.
To evaluate the prognostic value of USP22 and survivin, we examined the survival rate of 151 patients with high or low expression of USP22 or survivin by Kaplan-Meier analysis.The 5-year survival rate of HCC cases with high expression of USP22 was significantly lower than that with low expression of USP22 (P<0.001;Fig. 4A).Similarity, the 5-year survival rate of HCC cases with high expression survivin was significantly lower than that with low expression of survivin (P<0.01;Fig. 4B).

Suppression of cell growth by USP22 knockdown in HCC cells.
To investigate the role of USP22 in malignant behavior of HCC, USP22 was silenced by using siRNA oligonucleotides in HCC cell lines, HepG2 and SK-Hep-1 with high expression of USP22.After transfection of USP22 siRNA, reduced expression of USP22 protein and mRNA was confirmed by western blot analysis and RT-PCR (Fig. 5A).Then, we examined the effect of USP22 knockdown on cell growth of HCC cells by MTT assay.Growth of USP22 siRNA-treated HCC cells was slower than that of control siRNA-treated cells (Fig. 5B).
As shown above, we found the possible correlation between USP22 and survivin (Fig. 3).In USP22 knockdown cells, expression of survivin mRNA and protein was examined.USP22 knockdown decreased protein level of survivin (Fig. 5A).However, decreased level of survivin mRNA was low (Fig. 5A).These finding suggest that increased survivin may be translationally as well as transcriptionally regulated by USP22.
Table Ⅰ.USP22 and survivin expression and its correlation with clinicopathological features in HCC cases.

Discussion
USP22 belongs to a large family of proteins with ubiquitin hydrolase activity, and recombinant USP22 is able to cleave a synthetic ubiquitin molecule in vitro.USP22 has been identified as a death-from-cancer signature gene, that is, a marker for predicting the likelihood of treatment failure in cancer patients (21,22).Indeed, it has been reported that overexpression of USP22 was associated with a poor outcome in various cancers including breast, colorectal, non-small cell lung cancer, salivary duct carcinoma and oral squamous cell carcinoma (6)(7)(8)(23)(24)(25).In the present study, we found that 53.0% of the 151 HCC cases showed high expression of USP22 by immunohistochemistry. Notably, high expression of USP22 was significantly associated with malignant behavior including tumor size, tumor differentiation and tumor stage and prognosis in HCC cases (Table Ⅰ and Fig. 2A).Thus, high expression of USP22 is widely observed in various cancers, suggesting that elevated expression of USP22 may be a common event in cancer.However, the mechanism of upregulation of USP22 in cancer is still unclear.Importantly, USP22 can be a strong prognostic marker for various cancers including HCC.Therefore, USP22 may be a useful cancer therapeutic target.
USP22 is a key subunit of the SAGA (Spt-Ada-Gcn5 acetyltransferase) complex, a chromatin-modifying transcription coactivator complex, which regulates the expression of genes related to tumorigenicity and proliferation (6).Within SAGA, USP22 deubiquitylates histone H2B, which is intimately linked to transcription activation of certain genes (5).Furthermore, USP22 is recruited to specific genes by activators such as the MYC oncoprotein, where it is required for transcription (5).Thus, USP22 is a positive regulator of the growth of tumors.Indeed, USP22 depletion results in cell cycle arrest at G1 phase and compromises MYC functions including transformation (5).In the present study, we also observed the suppression of cell growth by USP22 knockdown in HCC cells (Fig. 5B).The critical role of USP22 in cell cycle progression is supported by previous and present findings; i) USP22 regulates cell proliferation and affects p21 expression by deubiquitinating the transcriptional regulator FBP1 (11); ii) USP22 knockdown reduced the expression of cyclin B and increased the expression of p21 (Fig. 5A); iii) USP22 positively regulates cell cycle via both BMI-1-mediated INK4a/ARF and Akt signaling pathway which further emphasizes its role as an oncogene (26); and iv) USP22 stabilizes Sirt1 by removing polyubiquitin chains conjugated onto Sirt1 leading to decreased levels of p53 acetylation and suppression of p53 functions (27).Interestingly, USP22 levels are significantly upregulated in colorectal cancer with associated increase in the expression of several cell cyclerelated genes such as BMI-1, c-Myc and both, pAkt (Ser473) and pAkt (Thr308) (7,26).These findings suggest that USP22 may be involved in progression of HCC by promoting cell growth via activated cell cycle as well as evading apoptosis and stem cell renewal (Fig. 5C).
Here, we found the possible correlation between USP22 and survivin in HCC cell lines, tissues and cases.Survivin is a bifunctional protein that suppresses apoptosis and regulates cell division (28,29) and is highly expressed in various cancers (13).Importantly, we and other groups have shown that high expression of survivin was significantly associated with a poor prognosis in HCC (Fig. 4) (30,31).In our previous study, cytoplasmic survivin overexpression was found to be associated with a poor prognosis in colon and head and neck cancer (32)(33)(34)(35).Recent report shows that BMI-1, a key component of the Polycomb Repressive Complex 1, induces repressive epigenetic regulation of the promoter of survivin (36).Moreover, overexpression of BMI-1 correlates with drug resistance in B-cell lymphoma cells via the stabilization of survivin expression (37).As USP22 upregulates BMI-1, USP22 may upregulate survivin mediated by BMI-1 overexpression (Fig. 5C).Indeed, we found that USP22 knockdown decreased survivin expression in HCC cells (Fig. 4).However, decreased level of survivin mRNA was not parallel to decreased level of survivin protein (Fig. 5A), suggesting that increased survivin may be translationally as well as transcriptionally regulated by USP22.To demonstrate this hypothesis, further experiments will be required.
USP22 is part of an 11-gene polycomb/cancer stem cell signature that uniformly exhibits a marked propensity toward metastatic dissemination as well as a therapy-resistance phenotype (5,6).In this signature gene, cyclin B1 (CCNB1), which is critical regulator of mitosis, is included.Interestingly, we found that USP22 knockdown decreased cyclin B expression in HCC cells.This finding suggests that cyclin B may be regulated by USP22 via deubiquitination (Fig. 5C).Further investigation of the USP22-mediated deubiquitination of cell cycle-related proteins could be infirmative.
In conclusion, our present findings suggest that USP22 may be involved in HCC progression in cooperation with survivin.We suggest that USP22 can be useful as a new diagnostic marker and therapeutic target in HCC patients.However, the detailed mechanism of USP22 overexpression, involved in the progression of human cancer, remains unexplained.

Figure 1 .
Figure 1.Immunohistochemical staining for USP22 and survivin in normal adjacent tissues and HCC.A representative case, respectively, of USP22 expression in normal hepatic tissue and in HCC tissue is shown (original magnification, x200).

Figure 2 .
Figure 2. Involvement of USP22 and survivin in malignant behavior of HCC.(A) In 151 HCC cases, USP22 expression was examined by immunohistochemical analysis.Correlation between USP22 and clinicopathological features was examined as shown in Table I.Among clinicopathological features, tumor size, tumor stage and tumor differentiation were significantly correlated with USP22 expression.Graph shows the correlation of USP22 expression with tumor size, tumor stage and tumor differentiation in 151 HCC cases.(B) In 151 HCC cases, survivin expression was examined by immunohistochemical analysis.Correlation between survivin and clinicopathological features was examined as shown in Table I.Among clinicopathological features, tumor size, tumor stage, tumor differentiation, vascular invasion and serum level of AFP protein were significantly correlated with survivin expression.The correlation of survivin expression with tumor size, tumor stage, tumor differentiation, vascular invasion and serum level of AFP protein in 151 HCC cases is shown.

Figure 3 .
Figure 3. Correlation between USP22 and survivin in HCC.(A) In 151 HCC cases, USP22 expression and survivin expression were examined by immunohistochemical analysis.Correlation between USP22 and survivin was examined as shown in Table II.Graph shows the correlation of USP22 expression with expression in 151 HCC cases.USP22 expression in HCC cell lines and tissues.(B) The expression levels of USP22 and survivin in a normal hepatic cell line (LO2) and seven HCC cell lines (Hep3B, Bel-7402, HepG2, SK-Hep-1, SMMC-7721, HuH-7 and QGY-7701) was examined by western blot analysis.β-actin expression was used as a loading control.(C) Expression of USP22 and survivin in 4 HCC tissues and 4 normal adjacent liver tissues were examined by RT-PCR and western blot analysis (WB).β-actin expression was used as a control.

Figure 4 .
Figure 4. Prognostic value of USP22 and survivin in HCC cases.(A) The relation of USP22 expression to survival is shown.Kaplan-Meier plots for disease-free survival show the association of survival and expression of USP22 in patients with HCC.The prognoses for patients who were high for expression of USP22 (n=80) was poorer than the prognoses for patients with low expression of USP22 (n=71).The statistical significance of these data was measured by the using log-rank tests.(B) The relation of survivin expression to survival is shown.Kaplan-Meier plots for disease-free survival show the association of survival and expression of survivin in patients with HCC.The prognoses for patients who were high for expression of survivin (n=92) was poorer than the prognoses for patients with low expression of survivin (n=59).The statistical significance of these data was measured by the using log-rank tests.

Figure 5 .
Figure 5.Effect of USP22 knockdown on cell growth in HepG2 and SK-Hep-1 cells.(A) After 48 h of USP22 siRNA transfection, cells were collected and the expression of survivin, cyclin B and p21 was examined by western blot analysis.β-actin expression was used as a loading control.Expression of USP22 and survivin mRNA was examined by RT-PCR.β-actin mRNA expression was used as a control.(B) After 48 h of USP22 siRNA transfection, cells were used for MTT assay.Briefly, cells (2x10 4 cells/well) were cultured in 96-well culture plates for 24 h.The cells were collected after 24, 48 and 72 h for analysis.The graph shows cell growth of USP22 siRNA and control siRNA-treated HepG2 and SK-Hep-1 cells by MTT assay.(C) Schematic model of the role of USP22 in HCC progression.