High USP22 expression indicates poor prognosis in hepatocellular carcinoma

Abstract

Ubiquitin-specific protease 22 (USP22) removes ubiquitin from histones, thus regulating gene transcription. The expression frequency and expression levels of USP22 were significantly higher in hepatocellular carcinoma (HCC) than in normal liver tissues. High USP22 expression in HCC was significantly correlated with clinical stage and tumor grade. Kaplan-Meier analysis showed that elevated USP22 expression predicted poorer overall survival and recurrence-free survival. High USP22 expression was also associated with shortened survival time in patients at advanced tumor stages and with high grade HCC. Multivariate analyses revealed that USP22 expression is an independent prognostic parameter in HCC. These findings provide evidence that high USP22 expression might be important in tumor progression and serves as an independent molecular marker for poor HCC prognosis. Thus, USP22 overexpression identifies patients at high risk and represents a novel therapeutic molecular target for this tumor.

Keywords: cancer biomarker; hepatocellular carcinoma; prognosis; ubiquitin-specific protease 22.

Figure 1. USP22 expression in HCC tissues and cell lines

(A) Semi-quantitative RT-PCR analysis of USP22 mRNA expression in seven HCC cell lines (Bel-7402, HepG2, SK-Hep-1, HuH-7, Hep3B, QGY-7701, and SMMC-7721) and a normal hepatic cell line (LO2); (B) Western blot analysis of USP22 protein expression in seven HCC cell lines (Bel-7402, HepG2, SK-Hep-1, HuH-7, Hep3B, QGY-7701, and SMMC-7721) and a normal hepatic cell line (LO2). β-actin was used as internal control for RT-PCR and Western blot.

Figure 2. USP22 expression in HCC and normal liver tissues

(A) Semi-quantitative RT-PCR analysis of USP22 mRNA expression in HCC specimens (T) and normal adjacent hepatic tissue (N); (B) Western blot analysis of USP22 protein expression in representative HCC (T) and normal adjacent tissue (N); (C) Quantitative RT-PCR analysis of USP22 expression in HCC specimens (T) and normal adjacent hepatic tissue (N). β-actin served as an internal control.

Figure 3. siRNA silencing of USP22

HepG2 cells were transfected with USP22 siRNA and negative control siRNA. At 48 h post-transfection, changes in the mRNA and protein levels of USP22 were determined using RT-PCR (A) and Western blot (B) analyses. β-actin was used as internal control for Western blot and RT-PCR. (C) Relative protein levels were quantified using Image J (NIH, Bethesda, MO, USA) and normalized versus internal control. Three independent experiments were performed and representative results are shown.

Figure 4. USP22 gene silencing inhibits HCC cell growth

HepG2 cells were transfected with USP22 siRNA, negative control siRNA and vehicle control, respectively. At 12, 24, 48, 72 and 96 h post-transfection, cell viability was determined using the MTT assay. Each independent experiment was performed 3 times.

Figure 5. USP22 gene silencing leads to HCC cell apoptosis

HepG2 cells were transfected with USP22 siRNA and negative control siRNA. The cells were analyzed at 24 h post-transfection. (A) Flow cytometry were performed to analyze cell apoptosis rate. Each independent experiment was performed 3 times. (B) USP22 expression was observed in HCC cells via immunohistochemical staining with an anti-USP22 antibody under a confocal laser scanning microscope.

Figure 6. USP22 gene silencing induces mitochondrial apoptosis in HepG2 cells

At 48 h post-transfection with USP22 siRNA and control siRNA in HepG2, the mitochondria and protein in the cytoplasm were isolated and subjected to Western blot analysis. β-actin was used as internal control. (A). Protein levels of caspase-3(Cas-3), cleaved caspase-3 (C-Cas-3), Bcl-2, Bax, Bcl-XL, and cytochrome c (Cyt C) in the cytoplasm; (B) Relative protein expression levels in the cytoplasm were quantified using Image J and normalized versus international control; (C). Protein levels of Bcl-2, Bax and CytC in the mitochondria.(D) Relative protein expression levels in the mitochondria were quantified using Image J and normalized versus international control density. Three independent experiments were performed and representative results are shown.

Figure 7. Immunohistochemical staining for USP22 in HCC and normal adjacent hepatic tissues

The level of USP22 protein was determined by immunohistochemical staining using a USP22 antibody, and the nuclei were counterstained with hematoxylin. (A) USP22 was overexpressed in the cytoplasm in HCC tissue; (B) USP22 was expressed at negative levels in matched normal hepatic tissues from the same patient (100×). (C) and (D) are images recorded at higher magnification (400×) from the areas shown in boxes in (A) and (B), respectively.

Figure 8. Receiver operating characteristic (ROC) curve analysis used to select a USP22 cutoff score based on the training set

(A) USP22 cutoff for overall survival in the training set. (B) USP22 cutoff for recurrence-free survival in the training set. At each immunohistochemical score, the sensitivity and specificity for the outcome studied were plotted, thus generating a ROC curve.

Figure 9. Kaplan-Meier survival analysis of USP22 expression in the test set and in all patients

Higher USP22 expression was closely correlated with (A) poor overall survival and (B) recurrence-free survival in the test set. (C) Higher USP22 expression also correlated with inferior overall survival (D) and recurrence-free survival in all patients. In the test set and in all patients, the median durations of overall survival for patients with low and high USP22 expression were 78.0 vs. 15.0 months (p < 0.001) and 82.0 vs. 22.0 months (p < 0.001), respectively.