SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Amplification of broad regions of 8q is one of the most frequent genetic alterations in hepatocellular carcinoma (HCC), suggesting the existence of oncogenes in addition to MYC at 8q24.21. In this report we examine the potential role of the candidate amplified oncogene serum and glucocorticoid kinase 3 (SGK3) at 8q13.1 in HCC pathogenesis. We found amplification and overexpression of SGK3 was frequently detected in clinical HCC specimens and that SGK3 genomic activation was significantly associated with poor outcome of patients (P = 0.028). Functionally, we found that overexpression of SGK3 in HCC cells increased cell cycle progression through G1, cell survival, clonogenicity, anchorage-independent growth, and tumor formation in nude mice. In contrast, RNA interference (RNAi) silencing of SGK3 inhibited its oncogenic effects. We provide evidence that SGK3 promotes HCC growth and survival through inactivating glycogen synthase kinase 3 beta and Bcl-2-associated death promoter, respectively. We also found that expression of SGK3, which like AKT is activated by PI3K/PDK1 signaling, has more significance than overexpression of AKT in predicting poor outcome in HCC patients. Taken together, our findings in the present study suggests that the SGK3 pathway may function in parallel with the AKT pathway and undergoes an AKT-independent signaling pathway in the pathogenesis of HCC. Further characterization of SGK3 may provide a prognostic biomarker for HCC outcome prediction and a novel therapeutic target in HCC treatment. (HEPATOLOGY 2012;55:1754–1765)

Hepatocellular carcinoma (HCC) is one of the most frequent human malignancies worldwide, with very poor prognosis. It is generally believed that the development of HCC is associated with chronic inflammation caused by hepatitis B virus (HBV) or hepatitis C virus (HCV) infection.1 Accumulation of irreversible alterations of critical genes during long-term inflammation finally leads to hepatocellular pathogenesis.2 Although under intensive investigation, detailed knowledge of molecular pathogenesis of HCC still remains to be elucidated.

Like other solid tumors, genomic instability often occurs during the development and progression of HCC.3 Chromosomal segmental gains (i.e., 1q and 8q) and losses (i.e. 1p, 4q, 8p, 16q and 17p) are common events in HCC patients.4 8q is one of the most frequently amplified regions detected by comparative genomic hybridization (CGH).5 Two known oncogenes, c-Myc and FAK at 8q24, have been characterized for their oncogenic effects on HCC development.6, 7 Based on the CGH data, the whole long arm of chromosome 8 is frequently amplified in HCC.5 In particular, multivariate analysis revealed that gain of 8q11 is an independent prognostic indicator in HCC patients.8 These findings suggest that chromosome 8q, especially the region proximal to the centromere, may harbor other important oncogenes related to the development and progression of HCC.

In order to identify candidate oncogenes in 8q, we used the publicly available Oncomine microarray database to screen the genes located in 8q with both genomic copy number gain and messenger RNA (mRNA) up-regulation. Serum and glucocorticoid kinase 3 (SGK3) at 8q13.1 is one of the significantly up-regulated genes. SGK3 belongs to the SGK family of serine/threonine kinase, which shares great similarity with AKT.9 The kinase domain of the SGKs share 55% identity with the AKT kinase domain, suggesting that SGKs may have a similar function as AKT in cancer development.10 SGKs have been reported to be involved in the regulation of cell growth, proliferation, survival, and migration.11, 12 As a member of the SGK family, SGK3 has been implicated as a regulator in the interleukin (IL)-3-dependent survival pathway.13 A recent study suggests that SGK3 can promote breast cancer through an AKT-independent mechanism.14 In the present study, we extensively investigated the role of SGK3 in HCC development. Our results found that SGK3 was frequently amplified and up-regulated in HCC, and possessed a strong oncogenic function by promoting cell proliferation and survival.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Database.

The Oncomine microarray database (http://www.oncomine.org) was used to screen the gene expression level in paired liver cancer samples compared with normal control tissues. “8q11, 8q12, 8q13” were used as keywords in the Oncomine query and “liver cancer” was used as the primary filter. Genes reported with both genomic copy number gain and mRNA up-regulation were selected for further study. All gene expression data were log-transformed, median-centered per array, and the standard deviation (SD) was normalized to one per array.15

HCC Samples and Cell Lines.

Primary HCC specimens were obtained from patients who underwent hepatectomy for HCC at the Cancer Center of Sun Yat-Sen University (Guangzhou, China). All HCC patients gave written informed consent on the use of clinical specimens for medical research. Normal liver tissues and cirrhotic liver tissues were obtained from patients who underwent liver transplantation at Queen Mary Hospital. Human liver cancer cell lines BEL-7402,16 QGY-7703,17 SMMC-7721, and QGY-770118 were obtained from the Institute of Virology of the Chinese Academy of Medical Sciences (Beijing, China). H4M was previously established in our laboratory.19

Plasmids and Reagents.

The full-length SGK3 was cloned into expression vector pcDNA3.1(+) (Invitrogen, Carlsbad, CA). Cloned SGK3 was transfected into HCC cell lines BEL-7402 and QGY-7703 cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Rabbit-anti-SGK3, Mouse-anti-GSK3-β (glycogen synthase kinase 3 beta), Mouse-anti-ACTB antibody were obtained from Santa Cruz Biologicals (Santa Cruz, CA). Mouse-anti-CCND1, Rabbit-anti-phospho-GSK3-β (S9), Mouse-anti-phospho-BAD (Bcl-2-associated death promoter) (S112), Rabbit-anti-BAD, Rabbit-anti-phospho-AKT, Rabbit-anti-AKT, Mouse-anti-Bcl2, Rabbit-anti-Bcl-XL, and Rabbit-anti-Bax antibodies were obtained from Cell Signaling Technology (Danvers, MA). Small interfering RNA (siRNA) against SGK3 was obtained from Ambion (Carlsbad, CA). Short hairpin RNA (shRNA) against SGK3 was obtained from Origene (Rockville, MD).

RNA Extraction and Quantitative Real-Time Polymerase Chain Reaction (PCR).

Total RNA was extracted using TRIZOL Reagent (Invitrogen), and reverse transcription was performed using a reverse-transcription (RT)-PCR Kit (Roche, Basel, Switzerland) according to the manufacturer's instructions. For quantitative real-time PCR (qPCR), the complementary DNA (cDNA) was amplified using an SYBR Green PCR Kit (Applied Biosystems, Carlsbad, CA) and an ABI PRISM 7900 Sequence Detector. Sequences of primers used are listed in Supporting Information Table 1.

Fluorescence In Situ Hybridization (FISH).

A BAC clone at 8q13.1 containing the SGK3 gene (RP11-705C3) was labeled with Spectrum-red (Vysis, Downers Grove, IL). Chromosome 8 centromere probe was labeled with Spectrum-green (Vysis). FISH reaction was performed according to the described method.20

Functional Assays for Tumorigenicity.

Cell proliferation assay, foci formation assay, and colony formation in soft agar were carried out as described.21 For cell proliferation assay and foci formation assay, cells were seeded at a density of 1,000 per well. For soft agar assay, cells were seeded at a density of 5,000 per well. The results are expressed as mean ±SD of three independent experiments. For the xenograft tumor growth assay, the cell number used for transplantation was 5 × 106 (BEL-7402), 5 × 106 (QGY-7703), and 8 × 106 (QGY-7701). Tumor volume was measured weekly and calculated by the formula V = 0.5 × L × W2. All animal experiments were approved by and performed in accordance with the Committee on the Use of Live Animals in Teaching and Research at the University of Hong Kong.

Detecting DNA Content and Apoptosis by Flow Cytometry.

Cells were seeded to 6-well plates at 30% confluence. Serum-free medium with L-Mimosine (400 μM) was added for G1 synchronization. After 24 hours, medium containing 10% fetal bovine serum (FBS) was added for an additional 12 hours. Cells were fixed in 75% ethanol, stained with propidium iodide (PI), and analyzed by flow cytometry. The results were analyzed with ModFit LT2.0 software. The apoptosis assay was done with the Annexin V-FITC apoptosis detection kit (BD biosciences, San Jose, CA) according to the manufacturer's instructions.

Statistics.

Statistical analysis was carried out with SPSS v. 16 (Chicago, IL). Pearson's χ2 test was used for categorical variables and independent Student t test for continuous data. The mRNA level of SGK3 in HCCs and the matched nontumor tissue was compared using paired Student t test. The correlation between mRNA levels of SGK3 and its genomic content was analyzed using Pearson's χ2 test. Kaplan-Meier plots and log-rank tests were used for survival analysis. The association of SGK3 expression with clinical features was examined by Pearson's χ2 test. A Cox proportional hazard regression model was used to analyze the independent prognostic factors. Differences were considered significant when P < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Amplification and Overexpression of SGK3 in HCC.

Because 8q amplification is one of the most frequent genetic alterations in HCC, a publicly available Oncomine microarray database was used to screen the expression of the genes located in the region proximal to the centromere of chromosome 8q with both genomic copy number gain and mRNA up-regulation to identify candidate oncogenes. A total of 35 genes were among the results of the Oncomine query22, 23 (Supporting Information Table 2). Genes involved in cellular signal transduction such as transcriptional factors and protein kinases were selected and their expression levels were examined by qPCR in five randomly selected HCC specimens (Supporting Information Fig. 1). Due to the similarity with AKT and the significant up-regulation in HCC specimens, SGK3 was selected for further study. To confirm the up-regulation of SGK3 in HCC, qPCR was used to compare the SGK3 expression level between tumor and their paired nontumor specimens in 91 primary HCCs. The results showed that the average fold change of SGK3 expression was significantly higher in tumor tissues compared with that in nontumor tissues (3.85 versus 1.23; P < 0.001, paired Student t test, Fig. 1A). The expression level of SGK3 in HCC patients was also significantly higher than in normal liver donors and cirrhotic livers (P < 0.001, independent Student t test, Fig. 1A). The expression of SGK3 in five HCC cell lines was also detected by qPCR (Fig. 1B).

thumbnail image

Figure 1. Amplification and overexpression of SGK3 in HCC. (A) Relative expression level of SGK3 was detected by qPCR in seven healthy liver donors, nine cirrhotic livers, and 91 HCCs (NL, normal liver; CL, cirrhotic liver; NT, non-tumor tissue; T, tumor tissue). The boxes represent the lower and upper fold change; lines with boxes and whiskers denote mean and SD, respectively. **P < 0.001, paired Student t test, independent Student t test. (B) Expression of SGK3 in five HCC cell lines was detected by qPCR. The fold change of SGK3 was derived by comparing SGK3 expressed in HCC cell lines with that in pooled nontumor liver tissues (five specimens). (C) Expression of SGK3 in mRNA level was significantly correlated with its genomic content in 40 paired HCC samples (P < 0.05, Pearson's χ2 test). β-Actin and 18s rRNA were used as internal controls for fold changes of genomic content of SGK3 locus and mRNA level of SGK3 in HCC samples, respectively. (D) The copy number of SGK3 in normal metaphase lymphocyte from a healthy donor is shown in the upper-left panel. The magnified image of the selected region is shown in the upper-right panel. Representative images of SGK3 copy number in HCC nontumor tissue is shown in the lower-left panel. Representative images of SGK3 copy number in HCC tumor tissue is shown in the lower-right panel. Original magnification, 1,000×. Red signals represent BAC probe containing SGK3 gene and green signals represent centromere of chromosome 8. (E) Representatives of immunostaining of SGK3 expression detected in primary HCC sample and its corresponding nontumor tissue. Original magnification, 100× (upper panel), 400× (lower panel).

Download figure to PowerPoint

To test whether SGK3 overexpression was associated with gene amplification, the relative genomic DNA content of SGK3 in 40 paired HCC tissue samples was investigated by qPCR. The results show that the expression of SGK3 at the mRNA level was significantly correlated with its genomic content in HCCs (P = 0.009, Pearson χ2 test, Fig. 1C). To further confirm the amplification of SGK3 in HCC tissues, FISH was applied to detect the DNA copy number of SGK3 in eight randomly selected HCCs using a BAC probe containing SGK3 and a chromosome 8 centromere probe as control. The DNA copy number gain at the SGK3 locus was detected in three of eight of HCCs (Fig. 1D). Interestingly, SGK3 overexpression was detected in three of three and one of five HCCs with or without DNA copy number gain, respectively (Fig. 1E).

SGK3 Has Strong Oncogenic Ability.

To determine the tumorigenic potential of SGK3 in HCC pathogenesis, SGK3 was cloned into expressing vector pcDNA3.1 and stably transfected into BEL-7402 and QGY-7703 cells. Empty vector was used as a control. To avoid clone deviation, two clones of the transfected cell line were picked for functional studies. The enhanced expression of SGK3 in stably transfected cells was confirmed by RT-PCR and western blotting (Fig. 2A). In vitro functional assays demonstrated that SGK3 had strong tumorigenic ability. Cell growth assay showed that the cell growth rates in SGK3-transfectants were significantly faster than that in empty vector-transfectants at the end of day 6 (P < 0.001, Student t test, Fig. 2B). SGK3 could significantly enhance the colony formation ability in foci formation assay (P < 0.05, Student t test, Fig. 2C) and anchorage-independent growth in soft agar (P < 0.05, Student t test, Fig. 2D).

thumbnail image

Figure 2. Overexpression of SGK3 increases tumorigenicity in HCC. (A) Ectopic expression of SGK3 was detected in SGK3-transfected cells by RT-PCR and western blot. (B) Cell growth rate between SGK3- and empty vector-transfected cells was compared by XTT assay. The results are expressed as mean ±SD of three independent experiments. *P < 0.05, **P < 0.001, independent Student t test. (C) Representative of increased foci formation in monolayer culture induced by SGK3. Quantitative analyses of foci numbers are shown in the right panel. Values are the mean ±SD of at least three independent experiments. *P < 0.05, independent Student t test. Original magnification, 40×. (D) Soft agar assay showed that SGK3 could increase the colony formation frequency. *P < 0.05, independent Student t test. Original magnification, 40×. (E) Representative images of the tumors (arrows) formed in nude mice induced by empty vector-transfected cells (left dorsal flank) and SGK3-transfected cells (right dorsal flank). Tumor volumes of mice transplanted with BEL-7402 were monitored for 5 weeks. Tumor volumes of mice transplanted with QGY-7703 were monitored for 6 weeks. The average tumor volume is expressed as mean ±SD in five mice. *P < 0.05, independent Student t test.

Download figure to PowerPoint

To assess the capacity of SGK3 in promoting tumor growth in vivo, BEL-7402 cells stably expressing SGK3 or empty vector were subcutaneously injected into nude mice and tumors were monitored for 5 weeks. No tumor formed in the mice injected with BEL-7402 cells transfected with empty vector (Vec-7402) at the end of week 5. Tumor formation was observed in four of five mice injected with SGK3-transfected BEL-7402 cells (SGK3-7402). For QGY-7703 cells, tumor formation was observed in one of five and three of five nude mice injected with Vec-7703 and SGK3-7703 cells, respectively (Fig. 2E). The average volume of tumors induced by SGK3-7703 cells was significantly larger than that induced by Vec-7703 cells on the sixth week after injection (P = 0.02, Student t test, Fig. 2E).

Silencing SGK3 Expression Inhibits Its Tumorigenicity.

To confirm the oncogenic function of SGK3, two siRNAs (siSGK3-1 and siSGK3-2) specifically targeting SGK3 were transfected into QGY-7701 cells that express a relatively high level of SGK3. The silencing effect was detected by qPCR. Both siRNAs could specifically knockdown the expression of SGK3 (Fig. 3A). The knockdown of SGK3 was further confirmed by western blotting (Fig. 3A). Functional assays demonstrated that knockdown of SGK3 could significantly inhibit cell growth rate (P = 0.001, Student t test, Fig. 3B), foci formation (P = 0.02, Student t test, Fig. 3C), and anchorage-independent growth in soft agar (P = 0.001, Student t test, Fig. 3D). To investigate the tumor-inhibitory effect of SGK3 in vivo, QGY-7701 cells stably expressing shSGK3 or Scramble shRNA were subcutaneously injected into nude mice and tumor was monitored for 5 weeks. Tumor formation was observed in four of five and one of five nude mice injected with Scramble shRNA and shSGK3 cells, respectively (Fig. 3E). Silencing SGK3 significantly inhibited the tumor formation of QGY-7701 cells in vivo (P = 0.02, Student t test, Fig. 3E).

thumbnail image

Figure 3. Silencing SGK3 by RNAi inhibits its tumorigenicity. (A) Two siRNAs against SGK3 could effectively decrease SGK3 expression in mRNA and protein levels detected by qPCR (left) and western blot (right). Scramble siRNA and β-actin were used as negative and loading controls, respectively. Silencing SGK3 expression could inhibit cell growth (B), foci formation (C), and colony formation in soft agar (D). Original magnification, 40×. The results are expressed as mean ±SD of three independent experiments. *P < 0.05, independent Student t test. (E) The specific inhibition of SGK3 by shRNA was detected by qPCR (left panel). Representative images of the tumors (arrows) formed in nude mice induced by scramble shRNA-transfected cells (left dorsal flank) and shSGK3-transfected cells (right dorsal flank). Tumor volumes of mice were monitored for 5 weeks. The average tumor volume is expressed as mean ±SD in five mice. *P < 0.05, independent Student t test.

Download figure to PowerPoint

SGK3 Promotes G1/S Transition in HCC Cell Lines.

To characterize the oncogenic mechanism of SGK3 in HCC tumorigenesis, the role of SGK3 in cell cycle progression was investigated. Vec-7402 or SGK3-7402 cells were treated with L-Mimosine in serum-free medium for G1 phase synchronization. After 24 hours, serum was added to the culture medium and the DNA content of the cells was analyzed by flow cytometry at the indicated timepoints. After synchronization, most of the cells were blocked in G1 phase. The percentages of cells in G1 and S phases were similar between Vec-7402 and SGK3-7402 cells. However, SGK3-7402 cells entered into S phase much faster than Vec-7402 cells after serum stimulation. The percentage of cells in S phase was significantly higher in SGK3-7402 cells than in Vec-7402 cells 12 hours after serum stimulation (P = 0.03, Student t test, Fig. 4A). Similar results could be observed in QGY-7703 cells (P = 0.01, Student t test, Fig. 4B). These results indicated that SGK3 could facilitate DNA synthesis and promote G1/S transition.

thumbnail image

Figure 4. SGK3 promotes G1/S transition in HCC cells. Flow cytometry was used to compare DNA content between empty vector- and SGK3-transfected BEL-7402 cells (A) and QGY-7703 cells (B). Cells were treated with L-Mimosine (400 μM) in serum-free medium for G1 phase synchronization, and then stimulated with medium containing 10% FBS 24 hours later. Summarized results are expressed as mean ±SD of three independent experiments (lower panel). *P < 0.05, independent Student t test. (C) Silencing SGK3 by RNAi could significantly inhibit the G1/S transition in QGY-7701 cells detected by flow cytometry. Summarized results are expressed as mean ±SD of three independent experiments (left panel). *P < 0.05, independent Student t test.

Download figure to PowerPoint

To further clarify whether the knockdown of SGK3 could arrest the cell cycle at the G1/S checkpoint, siRNAs against SGK3 (siSGK3-1 and siSGK3-2) were transfected into QGY-7701 cells. DNA content was analyzed by flow cytometry 48 hours after the treatment and the results showed that both siRNAs were able to arrest the cell cycle at the G1/S checkpoint. The percentage of cells in S phase was significantly reduced in siSGK3-transfected cells compared with the control-siRNA-transfected cells (P = 0.04, Student t test, Fig. 4C).

SGK3 Inactivates GSK3-β and Stabilizes CCND1.

To understand the molecular mechanism of SGK3 on promoting G1/S transition, the expression level of CCND1, the major cyclin controlling the G1/S checkpoint, was tested. As shown in Fig. 5A, the expression of CCND1 was slightly up-regulated in SGK3-7402 cells compared with Vec-7402 cells. Expression of CCND1 was then compared between Vec-7402 and SGK3-7402 cells when cells were stimulated with serum after L-Mimosine/free serum-mediated G1 phase synchronization. Western blot analysis showed that the expression of CCND1 was significantly higher in SGK3-7402 cells compared with Vec-7402 cells 12 hours after serum stimulation (Fig. 5B). Furthermore, decreased CCND1 expression was detected when the expression of SGK3 was silenced in another cell line, QGY-7701 (Fig. 5C).

thumbnail image

Figure 5. SGK3 inactivates GSK3-β and stabilizes CCND1. (A) Expressions of SGK3 and CCND1 were compared by western blot analysis between empty vector- and SGK3-transfected BEL-7402 cells (clones 1 and 2). β-Actin was used as a loading control. (B) Vec-7402 or SGK3-7402 cells were treated with L-Mimosine (400 μM) in serum-free medium for 24 hours, and then stimulated with medium containing 10% FBS for 12 hours. Expression of CCND1 was determined by western blotting at the indicated timepoints. (C) After silencing SGK3 expression in QGY-7701 cells, the expression of CCND1 was decreased. Scramble siRNA was used as negative control (left panel). Treatment of MG132 (20 μM) for 6 hours prevented the degradation of CCND1 induced by silencing SGK3 expression (right panel). (D) Expression of phospho-GSK3-β (S9), phospho-AKT (S112), total AKT, and total GSK3-β was detected by western blot. β-Actin was used as a loading control. (E) Treatment of LiCl (10 mM) for 48 hours inhibited the degradation of CCND1 induced by silencing SGK3 expression in QGY-7701 cells. NaCl was used as a control. (F) The effect of LiCl on cell proliferation was measured by XTT assay. QGY-7701 cells transfected with siSGK3 or scramble siRNA were treated with different concentrations of LiCl for 48 hours. NaCl was used as a control. Data are differently expressed in the histogram on the right panel. The viabilities of cells treated with siSGK3 are shown as percentage of the viabilities of control cells.

Download figure to PowerPoint

Interestingly, our results indicated that silencing SGK3 could not alter the mRNA level of CCND1 in QGY-7701 cells (data not shown), suggesting that SGK3 may regulate CCND1 expression at the protein level. To further study how SGK3 regulates CCND1 expression, siSGK3-transfected and control-siRNA-transfected cells were treated with MG132, a specific proteasome inhibitor that inhibits the proteolysis of the ubiquitinated protein. Western blot analysis showed that MG132 could prevent the degradation of CCND1 induced by the down-regulation of SGK3 (Fig. 5C), suggesting that SGK3 could stabilize CCND1 through inhibiting its proteasomal degradation. It is well established that CCND1 undergoes ubiquitin-dependent degradation upon phosphorylation by GSK3-β.24 Because SGK3 belongs to the AGC kinase family, it may share some of the same substrates of AKT. Interestingly, SGK3 was reported to interact with GSK3-β, and phosphorylate GSK3-β on Ser9.25 We therefore investigated whether the phosphorylation of GSK3-β was involved in the SGK3-induced stabilization of CCND1 in HCC. The results show that overexpression of SGK3 increased the phosphorylation level of GSK3-β on Ser9 in BEL-7402 cells. Conversely, depletion of SGK3 reduced the phosphorylation level of GSK3-β on Ser9 in QGY-7701 cells. The phosphorylation level of AKT was not affected, which suggested that SGK3 functions independently of AKT (Fig. 5D). Furthermore, lithium chloride (LiCl), a specific inhibitor of GSK3-β, could significantly inhibit the degradation of CCND1 induced by the downregulation of SGK3 (Fig. 5E). We further evaluated the effect of LiCl on the inhibition of cell proliferation induced by the down-regulation of SGK3. Although LiCl inhibited cell proliferation due to its own cytotoxicity, it prevented the inhibition of cell proliferation induced by the down-regulation of SGK3 in a dose-dependent manner (Fig. 5F).

SGK3 Inhibits Serum Starvation-Induced Apoptosis of HCC Cells.

Given the important role of PI3K/AKT in cell survival and apoptosis, we further checked the effect of SGK3 on HCC cell survival. Vec-7402 or SGK3-7402 cells were cultured in serum-free medium for 6 days. Apoptosis was measured by fluorescence-activated cell sorting (FACS)-based Annexin-V/PI double staining at the indicated timepoints. As shown in Fig. 6A, the apoptotic rate in SGK3-7402 cells was significantly lower than in Vec-7402 cells after serum was withdrawn for 6 days. This indicated that SGK3 can protect HCC cells from serum starvation-induced apoptosis. To confirm the observed findings, SGK3 was knocked down in QGY-7701 cells by siRNAs and subjected to the same assay. Consistently, inhibition of SGK3 sensitized HCC cells to serum starvation-induced apoptosis (Fig. 6B).

thumbnail image

Figure 6. SGK3 inhibits serum-starvation-induced apoptosis of HCC cells. (A) Vec-7402 or SGK3-7402 cells were seeded at 20% confluent and cultured in serum-free medium for 6 days. (B) control-siRNA-transfected or siSGK3-transfected cells were seeded at 20% confluence and cultured in serum-free medium for 4 days. The apoptosis rate was measured by FACS-based Annexin-V/PI double staining at the indicated timepoints. Annexin-V-negative and PI-negative cells were counted as viable cells. Annexin-V-positive cells were counted as apoptotic cells. Summarized results are expressed as mean ±SD of three independent experiments. *P < 0.05, independent Student t test. (C) Vec-7402 or SGK3-7402 cells were cultured in serum-free medium and harvested at the indicated timepoints. Expression of phospho-BAD (S112), total BAD, phospho-AKT (S473), total AKT, Bcl2, Bcl-XL, and Bax were detected by western blot. β-Actin was used as a loading control.

Download figure to PowerPoint

To understand the molecular mechanism involved in SGK3-induced HCC cell survival, we further checked the expression of some critical apoptotic components related to the PI3K/AKT signaling pathway. Consistent with previous findings,14 the phosphorylation level of BAD was significantly higher in SGK3-7402 cells compared with Vec-7402 cells. (Fig. 6C) It is well identified that phosphorylation of BAD will prevent its binding to Bcl-XL, and further inhibit its proapoptotic effect in the mitochondria-related apoptotic pathway.26 Thus, SGK3 may also enhance HCC cell survival through phosphorylating BAD. The phosphorylation level of AKT was not affected, further supporting that the antiapoptotic effect of SGK3 was independent of AKT (Fig. 6C).

Clinical Significance of SGK3 in HCC.

To investigate the clinical significance of SGK3 in HCC patients, we evaluated the correlation of SGK3 overexpression with HCC clinicopathological features. An association study found that overexpression of SGK3 was significantly associated with poor overall survival of HCC patients (P = 0.028, Fig. 7A). Univariate analysis revealed that the tumor stage, recurrence status, and vascular invasion status were associated with overall survival. Expression of SGK3 was also significantly correlated with overall survival of the patients (P = 0.039, Table 1). Furthermore, multivariate Cox proportional hazard regression analysis showed that overexpression of SGK3 was an independent prognostic factor for overall survival of HCC patients (P = 0.047, Table 1). However, no significant correlation was found between expression of SGK3 and other clinicopathological features of the HCC patients examined (Supporting Information Table 3).

thumbnail image

Figure 7. Clinical significance of SGK3 and AKT1 in HCCs. (A) Kaplan-Meier analysis shows that overexpression of SGK3 (defined as greater than a 1.5-fold increase) was significantly associated with poorer overall survival in HCC patients. (B) No statistical difference in AKT1 expression was detected in 91 paired HCC samples (paired Student t test). (C) The relative fold changes of SGK3 and AKT1 in 91 HCC samples was detected by qPCR. The histogram shows the percentage of patients with different fold changes of SGK3 or AKT1 expression in tumor samples compared with their paired nontumor samples. (D) Kaplan-Meier overall survival curve of HCC patients divided into three subgroups according to their expression levels of SGK3 and AKT1. Patients in Group 1 have neither overexpression of SGK3 nor overexpression of AKT1. Patients in Group 2 have overexpression of SGK3 but not overexpression of AKT1. Patients in Group 3 have overexpression of both SGK3 and AKT1. Patients in Group 3 have the poorest overall survival rate. Overexpression was defined as greater than a 1.5-fold increase. (E) A model illustrating the oncogenic effect of SGK3 in HCC pathogenesis.

Download figure to PowerPoint

Table 1. Univariate and Multivariate Analyses
Clinicopathological FeaturesUnivariate AnalysisMultivariate Analysis
HR95% CIPHR95% CIP
  1. Abbreviations: CI, confidence interval; HR, hazard ratio.

Age      
< 60 years versus > 60 years0.8740.339-2.2520.7801.1100.305-4.0370.874
Gender      
Male versus female0.6120.265-1.4130.2501.0000.324-3.0871.000
Tumor stage      
I & II versus III & IV2.4781.152-5.3290.0202.6450.837-8.3610.098
Recurrence      
No versus yes3.5121.643-7.5060.0012.3250.880-6.1460.089
Vascular invasion      
No versus yes2.6101.081-6.3040.0330.8650.234-3.1940.827
SGK3 overexpression      
No versus yes3.5251.063-11.6890.0394.7721.019-22.3450.047

SGK3 Pathway Is More Common in HCC Compared with AKT1 Pathway.

Because SGK3 shares great similarity with AKT, it may function in parallel with AKT in cancer development. First, we compared the expression level of AKT1 between tumor and nontumor tissues in 91 primary HCCs and no significant difference was detected (Fig. 7B). The relative expression level of SGK3 and AKT1 in the same HCC cohort was compared by qPCR. Interestingly, the result showed that the frequency of SGK3 overexpression was obviously higher than that with AKT1 overexpression (Fig. 7C), suggesting that SGK3 overexpression is a more common event in HCC development.

To investigate the significance of the combined expression of SGK3 and AKT1 in HCC prognostic prediction, we further divided the patients into three subgroups according to their relative SGK3 and AKT1 expression level. Patients in Group 1 have neither overexpression of SGK3 nor overexpression of AKT1. Patients in Group 2 have overexpression of SGK3 but do not have overexpression of AKT1. Patients in Group 3 have both overexpression of SGK3 and AKT1. We found that the mean survival time of patients in Group 1 (88.64 weeks) was significantly higher than that in Group 2 (52.37 weeks) and Group 3 (44.09 weeks). Patients in Group 3 have the lowest mean survival time, suggesting that combined overexpression of SGK3 and AKT1 predicted poorer overall survival of HCC patients (Fig. 7D).

Discussion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Many oncogenes including c-MYC, CCND1, EGFR, and RAS are frequently amplified in different kinds of cancers, suggesting that gene amplification is one of the important mechanisms for the activation of an oncogene.27, 28 Like other solid tumors, genomic DNA amplification is also frequently detected in HCC, including amplification of 8q.5 Previous studies mainly focused on the terminal region of 8q, which harbors well-established oncogenes including c-MYC and FAK. However, the regions proximal to the centromere of 8q were usually neglected. In this study we aimed to identify novel oncogenes in the region proximal to the centromere of 8q. Due to the similarity with AKT and the significant up-regulation in HCC specimens, SGK3 was selected for further study. SGK3 belongs to the serum- and glucocorticoid-induced kinase family of serine/threonine kinase.9 The SGK family exhibits structural similarity to AKT, arousing interest in their functions, because AKT plays important roles in cell survival, proliferation, and metabolism. In this study, overexpression of SGK3 was frequently detected in HCCs, which was significantly associated with its genomic copy number gain (P = 0.009).

The oncogenic effect of SGK3 on HCC development was demonstrated by both in vitro and in vivo functional assays. Our data demonstrate that SGK3 was able to promote cell growth and increase abilities of foci formation and colony formation in soft agar. The tumorigenicity of SGK3 was further confirmed by tumor formation in a nude mice model. Promotion of cell proliferation is a major mechanism of oncogenesis in transforming a normal cell to a cancer cell. In this study we found that SGK3 could enhance the proliferative ability of HCC cells by promoting G1/S transition. Moreover, down-regulation of SGK3 inhibited cell proliferation by inducing G1/S checkpoint arrest. CCND1 is the major cyclin that regulates the G1/S checkpoint.29 Increased expression of CCND1 is often observed in cancer caused by defective posttranscriptional regulation.30, 31 The PI3K/AKT pathway plays an important role in regulating cell cycle progression.32 One major mechanism by which AKT promotes cell cycle progression is through the inactivation of GSK3-β and subsequently inhibiting the degradation of CCND1.33 In the present study we found that SGK3 was also able to modulate the stability of CCND1 with a mechanism similar to AKT. Overexpression of SGK3 enhanced the phosphorylation of GSK3-β on Thr9 and inhibited the proteasomal degradation of CCND1. Lithium chloride, a specific inhibitor for GSK3-β, could inhibit the degradation of CCND1 induced by silencing SGK3 expression. These findings suggested the existence of an SGK3/GSK3-β/CCND1 pathway in regulating HCC development. In addition to enhanced proliferation, resistance to apoptosis is also a hallmark of cancer cells. In this study we also found that SGK3 can protect HCC cells from serum starvation-induced apoptosis. SGK3 has been reported to regulate the IL-3-dependent survival of hematopoietic cells by phosphorylating BAD. Consistent with a previous report,13 we also observed that the phosphorylation of BAD was up-regulated in SGK3-7402 cells compared with Vec-7402 cells. Thus, enhanced phosphorylation of BAD may account for the antiapoptotic property of SGK3 in HCC cells. Given the complexity of the PI3K/AKT pathway, the activation of SGK3 in HCC may affect multiple downstream pathways involved in cell proliferation and survival. Other signaling pathways downstream of SGK3 still remain to be further elucidated.

One interesting finding in this study is that amplification and overexpression of SGK3 is a more common event in HCC compared with AKT1 overexpression, suggesting that the SGK3 pathway may be a more dominant event in HCC development. Abnormal activation of the PI3K/AKT pathway and its downstream targets has been observed in a wide spectrum of human malignancies including breast cancer,34 ovarian cancer,35 and lung cancer.36 A recent study finds that SGK3 functions independently of AKT in breast cancer progression,14 suggesting the existence of an AKT-independent oncogenic signaling pathway downstream of PI3K in cancer development. In the present study we also demonstrated that SGK3 might play an important role in HCC development independently of AKT1. Because overexpression of SGK3 was detected in over 50% of HCC patients, it is likely that abnormal activation of the SGK3 signaling pathway is more commonly involved in the development of HCC. In the clinical association analysis, we did not find a significant association between SGK3 expression and common clinicopathological features. We noticed that the expression level of SGK3 was already slightly up-regulated in the nontumor tissues compared with the normal liver tissues (Fig. 1A). This suggested that amplification and overexpression of SGK3 may be an early event in the development of HCC. In addition, we found that SGK3 mainly affected the proliferation and survival of HCC cells. However, the effect of SGK3 on cell motility and metastasis is weak (data not shown). Because many clinicopathological features (e.g., Vascular invasion, tumor differentiation) are often associated with late-stage tumors, it is possible that the association of SGK3 expression and these clinicopathological features is weak. Although SGK3 overexpression was not associated with most clinicopathological features in HCC, our study found that overexpression of SGK3 was significantly correlated with poor overall survival of HCC patients (P = 0.028). Furthermore, the expression level of SGK3 was significantly lower in healthy liver donors and in the preneoplastic stages of hepatocarcinogenesis (fibrosis, cirrhosis) compared with the tumor tissues. Therefore, abnormal expression of SGK3 could be used as a novel biomarker in prediction of outcome of HCC patients.

In conclusion, our results indicate that amplification and overexpression of SGK3 can promote cell proliferation and survival of HCC cells and plays an important role in HCC pathogenesis. Further characterization of the oncogenic function of SGK3 in HCC development may provide a novel therapeutic target in HCC treatment.

References

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_25584_sm_SuppFig1.tif11933KSupporting Information Figure
HEP_25584_sm_SuppTab1.doc64KSupporting information Table 1. Gene list of candidate oncogenes in 8q11–13.
HEP_25584_sm_SuppTab2.doc49KSupporting information Table 2. The DNA sequence of PCR primers
HEP_25584_sm_SuppTab3.doc67KSupporting Information Table 3.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.