SGK1 protein expression is a prognostic factor of lung adenocarcinoma that regulates cell proliferation and survival

Abstract

The etiological or clinicopathological significance of serum glucocorticoid-induced protein kinase 1 (SGK1) in lung adenocarcinoma remains unclear. This study aimed to investigate the role of SGK1 in the development and progression of human lung adenocarcinoma and the effects of targeted inhibition of intrinsic SGK1 expression on the proliferation of lung adenocarcinoma cells. SGK1 protein expression in 150 human cases of lung adenocarcinoma was detected by immunohistochemical analysis, and the relationships between SGK1 expression and clinicopathological features were assessed. In addition, endogenous SGK1 profiles were determined in seven lung adenocarcinoma cell lines. Cell proliferation, cell cycle distribution, and apoptosis were characterized in the absence and presence of SGK1 inhibitors. Compared to the adjacent normal tissues, significantly higher SGK1 expression levels were detected in the cytoplasm in cancerous lung adenocarcinoma tissues. Besides, SGK1 expression correlated with lymph node metastasis, distant metastasis, and pathological staging. Univariate analysis suggested that overexpression of this protein correlated significantly with a poor prognosis. Cultured lung adenocarcinoma cells expressed relatively high SGK1 levels, and inhibition of this protein was associated with G2 cell cycle arrest and reduced cyclin B1 and cdc2 expression. Pharmacological SGK1 inhibition experiments corroborated the role of this protein in cell cycle progression. SGK1 expression correlated closely with lung adenocarcinoma progression and could be used as a prognostic marker. Endogenous SGK1 inhibition abrogated lung adenocarcinoma cell proliferation via G2/M-phase cell cycle arrest, which was likely mediated by the concerted actions of cell cycle regulators.

Introduction

Lung cancer, which contributes to nearly 1 million deaths per year, is the leading cause of cancer-related deaths worldwide [1]. Lung cancers are commonly classified by histopathologic type, of which adenocarcinoma is the most common type [2]. Although diagnostic advances have increased the rate of early detection and chemotherapeutic agents have yielded significant improvements in patient outcomes, much remains unknown about the mechanisms that drive lung adenocarcinoma progression.

Expression of the protein serum glucocorticoid-induced protein kinase 1 (SGK1) has been linked to nodular lymphocyte-predominant Hodgkin lymphoma [3] and has been identified as a prognostic factor for adrenocortical carcinoma [4]. In squamous cell carcinomas, SGK1 mRNA levels were found to correlate with tumor grade, size, and clinical stage, further supporting a role for this protein as an effective prognostic indicator [5]. In an animal model, SGK1 inhibition was shown to induce radiation-induced apoptosis in colon tumor cells [6] and reduce prostate cancer cell growth [7]. Similarly, SGK1 has been identified as a therapeutic target in neurofibromatosis type 2 (NF2)-associated meningiomas [8]. In addition to these therapeutic effects, SGK1 inhibition was found to behave synergistically when combined with other cancer therapies [9].

SGK1 plays a regulatory role in many cellular mechanisms, including transport, hormone release, neuroexcitability, inflammation, cell proliferation, and apoptosis. This kinase is activated by insulin and growth factors via phosphatidylinositide-3-kinase (PI3K) and 3-phosphoinositide dependent kinase (PDK1) signaling pathways. Together with protein kinase B (PKB)/Akt, this key mediator of the PI3K cell viability pathway can inhibit the mitogen-activated protein kinase (MAPK) pathway by phosphorylating serine (Ser) 364 on B-Raf [10,11].

Because PI3K signaling is a common event in the origin of cancer, the downstream effectors Akt and SGK1 have been investigated as potential therapeutic targets in multiple cancers. Akt and SGK1 have both overlapping and unique substrates and functions. Interestingly, a large subset of Akt inhibition-resistant cancers was found to constitutively upregulate SGK1 [12], and other researchers have suggested that this upregulation contributes to cancer cell resistance to PI3K alpha-inhibition by perpetuating mammalian target of rapamycin complex 1 (mTORC1) signaling [13].

The role of SGK1 in various cancers remains to be fully characterized, but a greater understanding of its pathophysiology may be rooted in the capacity of SGK1 to regulate cellular growth and proliferation. In hepatocytes, SGK1 signaling induces mitogen-activated protein kinase kinase (MEK)/extracellular signal-related kinase (ERK) signaling, which then regulates the expression of genes that encode important regulators of growth and survival such as c-fos and cyclin D1 [14]. Evidence from a non-small-cell lung cancer model suggests that SGK1 promotes tumor cell growth and migration by activating beta-catenin/T-cell factor (TCF) [15]. Another report proposed that SGK1 kinase activity may regulate the growth of various cancer cells through the insulin-dependent phosphorylation of Mdm2 (a p53 regulator) [16]. Other reported SGK1-sensitive tumor growth mechanisms include the activation of K⁺ and Ca²⁺ channels and amino acid and glucose transporters and the regulation of various transcription factors [7].

Cellular survival might also be a means by which SGK1 contributes to cancer etiology. In MCF-7 breast cancer cells, SGK1 inhibition enhanced the pro-apoptotic membrane androgen receptor-activated focal adhesion kinase (FAK)/PI3K/Rac1/Cdc42 signaling response [17]. Conversely, SGK1-mediated survival signaling was previously reported to inhibit apoptosis in epithelial tumor tissues [18]. Taken together, these reports support the findings of Towhid et al., who observed increased apoptosis and cell death in Caco-2 colon carcinoma cells in response to a combination of ionizing radiation and SGK1 inhibition in vitro [6].

Although SGK1 signaling and downstream biological pathways continue to be expansively explored and elucidated in various cancer models, information about the role of SGK1 as a prognostic factor in human cancers is much more limited. Upregulated SGK1 mRNA expression has been observed in some squamous cell carcinoma samples and was found to correlate with various clinical parameters, including tumor size and clinical stage [5]. However, neither the capacity of SGK1 as a prognostic factor nor its relationships with clinical disease parameters have been adequately investigated in the context of lung adenocarcinoma. Similarly, the biological pathways associated with SGK1 expression in lung adenocarcinoma are not well-known. In this study, we sought to answer these questions and elucidate the role of SGK1 in the origin of lung adenocarcinoma both in vivo with tissues from 150 patients and in vitro using lung cancer cell lines that could be manipulated with small interfering RNA (siRNA) and pharmacological inhibitors. We report for the first time that in lung adenocarcinomas, SGK1 expression is closely correlated with tumor progression and could be used as a prognostic marker. Furthermore, the inhibition of endogenous SGK1 reduced the proliferation of lung adenocarcinoma cells via cell cycle arrest at the G2/M phase. These findings collectively suggest a therapeutic role for SGK1 in lung adenocarcinoma.

Materials and methods

Lung adenocarcinoma tissue

This work was performed in accordance with the Declaration of Helsinki. All subjects provided informed consent before enrollment in the study. Ethical approval was obtained from the Ethics Committee of Zhejiang University. All experiments in this study were performed in accordance with the principles of Declaration of Helsinki.

Tumor tissues were collected from 150 patients with stage I to IV lung adenocarcinoma at the First Affiliated Hospital of the Medical College of Zhejiang University between January 2008 and December 2010, including 77 with stage I disease, 35 had stage II disease, and 38 had stage III or IV disease. The patients’ ages ranged from 20 to 84 years, with a median of 59 years. None of the patients had undergone preoperative radiotherapy or chemotherapy or had a history of other tumors, and all adhered to the clinical and pathological data integrity thresholds.

Tumor tissues were collected via surgical resection or puncture. Adjacent tissue samples exceeding 5 cm were collected from the edges of tumor tissues. All tissue specimens were fixed in a 10% neutral formalin solution and paraffin-embedded after routine dehydration. All pathological sections were confirmed by two pathologists. The patients were followed up until January 2016.

siRNA, inhibitors and cell lines

All siRNA oligos were purchased from Shanghai Ji Ma Pharmaceutical Technology Co., Ltd. and are listed in Table 1. The SGK1 inhibitor GSK650394 (C25H22N2O2, molecular weight: 382.45) was purchased from Selleckchem (S7209; Houston, TX). The 50-mM stock solution in DMSO was stored at -80°C and diluted to the target concentration in complete medium at the time of use.

Table 1.

SiRNA oligo structures

Gene	Company	Sense (5’-3’)	Antisense (5’-3’)
siSGK1#1	GenePharma, China	CCCUCACUUACUCCAGGAUTT	AUCCUGGAGUAAGUGAGGGTT
siSGK1#2	GenePharma, China	GCCAAUAACUCCUAUGCAUTT	AUGCAUAGGAGUUAUUGGCTT
siSGK1#3	GenePharma, China	GGAAUGUUCUGUUGAAGAATT	UUCUUCAACAGAACAUUCCTT
Control siRNA	GenePharma, China	GUAUGACAACAGCCUCAAGTT	CUUGAGGCUGUUGUCAUACTT

The human lung adenocarcinoma cell lines A549, PC9, HCC827, H292, H1299, H1650, and H1975 were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, P.R. China).

Reagents

Fetal bovine serum (10099), RPMI 1640 Medium (11875119), and Trypsin-EDTA (0.05%; 25300054) were purchased from Gibco (Grand Island, NY), and the MTT assay was purchased from Amresco (0793; Solon, OH). EnVision Detection Systems Peroxidase/DAB (K5007) was purchased from DAKO (Glostrup, Denmark), the cell cycle staining kit was purchased from BD Biosciences (550825; San Jose, CA), and the Annexin V FITC Apoptosis Detection Kit was purchased from BD Biosciences (556547; San Jose, CA). Antibodies specific for SGK1 (ab59337), cyclin B1 (12231), and Cdc2 (28439) were purchased from Abcam (Cambridge, UK), and a horseradish peroxidase (HRP)-labeled goat anti-rabbit IgG antibody (7074) was purchased from Cell Signaling Technology, Inc. (Danvers, MA). RapidStep ECL Reagent was obtained from EMD Millipore (345818; Billerica, MA), and TRIzol (R). First Strand cDNA Synthesis Kit (6210A), and SYBR Premix Ex Taq (RR820Q) were obtained from the Takara Corporation (Shiga, Japan).

Instruments and equipment

The microtome and pathological imaging system used in this study were obtained from Leica Biosystems (Wetzlar, Germany). An iMark microplate reader, protein electrophoresis and transfer devices, and the gel imaging system were purchased from Bio-Rad (Hercules, CA), and flow cytometry equipment was purchased from BD Biosciences.

Immunohistochemistry

Immunohistochemical staining was used to detect SGK1 protein expression in lung adenocarcinoma and adjacent non-cancerous tissues. Paraffin sections were oven baked at 60°C for 2 h, dewaxed, hydrated, and subjected to antigen retrieval. The samples were then blocked in 3% sheep serum and incubated with a rabbit anti-human SGK1 polyclonal antibody (1:500 dilution) at room temperature for 2 h. After washing with PBS, the samples were incubated with a biotinylated secondary antibody and SABC enzyme complex at room temperature for 30 min. After another washing step, the samples were treated with DAB to initiate the colorimetric reaction, counterstained with hematoxylin, dehydrated, transparent encapsulated, and then subjected to microscopic observation and imaging to evaluate tissue staining.

SGK1 protein was mainly expressed in the cytoplasm and appeared as brown granules. Tumor tissues were immunostained alongside adjacent non-cancerous control tissues, and the integrated intensity of staining intensity and percentage of positive cells were used to quantify SGK1 expression, as described in the Methods. The intensity of staining was scored as 0 (no signal), 1 (weak), 2 (moderate), or 3 (strong signal), and the percentage of positive cells was scored as 0 (< 1%), 1 (1% to 25%), 2 (25% to 50%), or 3 (> 50%). High expression of SGK1 was defined as a total score of ≥3 points; all others were classified as having low expression.

Cell culture

The human lung adenocarcinoma cell lines A549, PC9, HCC827, H292, H1299, H1650, and H1975 were cultured in RPMI 1640 medium containing 10% fetal bovine serum and were maintained in a humid incubator at 37°C with 5% CO₂.

Targeted inhibition of endogenous SGK1 expression or kinase activity

RNA interference (RNAi) technology designed to inhibit endogenous SGK1 expression was used in this study. SGK1 siRNA and Lipofectamine RNAiMAX (Invitrogen) were each diluted to the working concentration in culture medium according to the manufacturer’s protocol. These working solutions were then uniformly mixed, incubated for 20 min at room temperature, and added to tissue culture plates containing A549 or H1975 cells for a 48-h culture. Three oligos (siRNA-1, siRNA-2, and siRNA-3) were used, and a scrambled siRNA was used as a control (NC).

The small molecule inhibitor GSK650394, which specifically inhibits SGK1 activity, was purchased from Selleckchem and stored at -80°C in DMSO. For subsequent experiments, this stock solution was diluted in culture medium to the target concentration and added to A549 or H1975 cells in the logarithmic growth phase.

MTT and colony cloning cell proliferation assays

A549 and H1975 cells in the logarithmic growth phase were plated in 96-well tissue culture plates and treated with 5 × 10⁴/ml siRNA or 0- to 10-µM GSK650394 for 24, 48, or 72 h. After adding MTT solution to each well, and cultures were further incubated for 4 h, at which time the culture medium was replaced with dimethyl sulfoxide. The absorbance (A value) of each well was measured at 490 nm, and these values were used to calculate cell viability.

Twenty-four hours after siRNA infection or 12 h after GSK650394 incubation, cells from each experimental group were collected, plated at a density of 5 × 10³/well in six-well tissue culture plates, and cultured for 2 additional weeks, after which the supernatants were removed. The cells were fixed in methanol for 15 min and stained with an appropriate amount of crystal violet dye. The cell colony formation in each experimental group was then analyzed via microscopy (seven colonies were randomly selected; those with ≥10 or more cells were defined single clones). The experiment was repeated three times.

Flow cytometry analysis of cell cycle distribution and apoptosis

Cells were collected for flow cytometry and apoptosis analyses at 48 h after siRNA transfection. A549 and H1975 cells were treated with GSK650394 (5- and 10-μM) for 24 h and fixed overnight in 70% ethanol at 4°C. Fixed cells were incubated with FITC-Annexin V and propidium iodide (PI) for 30 min at room temperature in the dark and subjected to an evaluation of cell cycle distribution. The experiment was repeated three times.

An additional group of cells was collected 48 h after siRNA transfection or 24 h after GSK650394 (5- and 10-μM) treatment, labeled with FITC-Annexin V and PI, and incubated at room temperature for 15 min in the dark. Cells from each experimental group were analyzed by flow cytometry to evaluate apoptosis.

Western blotting

Total proteins were extracted from A549, PC9, HCC827, H292, H1299, H1650, and H1975 human lung adenocarcinoma cells using the DC method. The proteins were separated using SDS-PAGE and wet-transferred to PVDF membranes, which were subsequently blocked in 3% bovine serum albumin (BSA) for 1 h. The blocked membranes were incubated with primary antibody for 2 h at room temperature, washed in Tris-buffered saline + Tween-20% (TBST), incubated for 1 h with an HRP-labeled secondary antibody, and washed three additional times with TBST. SGK1 protein expression was analyzed via enhanced chemiluminescence (ECL) on a gel imaging system. GAPDH was used as the internal reference protein.

For cell cycle analysis, the cells were cultured for 48 h after siRNA transfection or 24 h after GSK650394 treatment (5- and 10-μM). Total proteins were extracted using RIPA lysis buffer, and protein concentrations were determined using the DC method. The proteins were separated by SDS-PAGE and transferred to membranes that were blocked for 1 h at room temperature with 3% BSA. The membranes were then incubated for 2 h at room temperature with primary antibodies, washed three times with TBST, and incubated with a HRP-labeled secondary antibody at room temperature for 1 h. Target protein expression was detected via ECL on a gel imaging system. Cyclin B1 and Cdc2 were detected in the G2/M phase of cell cycle. The apoptosis-related proteins caspase 3 and poly-ADP ribose polymerase (PARP) were also detected. GAPDH was used as the internal reference protein.

Quantitative PCR detection of gene expression

Total RNA was extracted from A549, PC9, HCC827, H292, H1299, H1650, and H1975 cells using TRIzol reagent. Twenty microliters of each sample were used as a template for cDNA synthesis. A Takara reverse transcription kit and Bio-Rad gradient PCR device were also used for this step (conditions: 37°C for 15 min, 85°C for 5 s). To analyze SGK1 mRNA expression, the SYBR Premix Ex Taq kit was used for quantitative PCR, and reactions were run on the Applied Biosystems 7900HT fluorescent quantitative PCR apparatus (Thermo Fisher, Waltham, MA) using the following conditions: 30 cycles of 95°C for 5 s and 60°C for 30 s. GAPDH mRNA was used as the internal reference to calculate SGK1 mRNA level.

Statistical analysis

SPSS 20.0 software (SPSS, Inc., Chicago, IL) was used for the statistical analysis. Quantitative data were analyzed using the T test and analysis of variance. Categorical data were analyzed using the chi-square test, Kaplan-Meier-based univariate survival analysis, and log-rank test. A Cox regression model was used for the multivariate survival analysis. A P value of less than 0.05 was considered to indicate a significant difference between groups.

Results

SGK1 expression and clinicopathological features of human lung adenocarcinomas

The relationships of SGK1 protein expression with clinicopathological parameters were evaluated by immunohistochemical analysis in surgical or puncture samples from 150 patients with stage I to IV lung adenocarcinoma (Figure 1A). A significantly higher number of patients in the low SGK1 category had non-cancerous adjacent tissues (P = 0.032). By contrast, no difference was detected between high and low SGK1 expression in adenocarcinoma tissues (Figure 1B).

Figure 1.

SGK1 expression in lung adenocarcinoma tissues. A. Representative photomicrographs of immunohistochemical staining for SGK1 in lung adenocarcinoma tissues and adjacent normal tissues. 1-Low expression (adjacent normal tissue), 2-Low expression (lung adenocarcinoma tissue), 3-High expression (lung adenocarcinoma tissue). B. Statistical analysis of 150 cases of lung adenocarcinoma and adjacent normal tissues immunostained using antibodies against SGK1.

Nodal stage, metastasis, and tumor stage were found to correlate with SGK1 expression. Specifically, SGK1 levels were higher in adenocarcinomas with a more severe nodal stage and in stage III to IV tumors. Low SGK1 expression correlated with a lack of metastasis (Table 2). SGK1 was not found to correlate with age, sex, smoking history, histopathologic grade, or T stage.

Table 2.

The relationship between SGK1 expression and clinicopathologic parameters of lung adenocarcinoma patients

	n	SGK1		χ2	P-value
		High	Low
Age (years)	150	57.85±9.67	60.04±12.25	t = 1.223	0.223
< 59	66	32	34
≥ 59	84	33	51	1.274	0.259
Gender
Male	79	35	44
Female	71	30	41	0.064	0.800
Smoking
No	97	45	52
Yes	53	20	33	1.046	0.306
Pathological Grade
G1	14	3	11
G2-G3	136	62	74	3.017	0.082
T Stage
T1	42	12	30
T2	81	39	42
T3-4	27	14	13	5.290	0.071
Nodal Stage
Nx	26	10	16
N0	79	28	51
N1-3	45	27	18	7.345	0.025
Metastasis
Yes	4	4	0
No	146	61	85	/	0.033
Tumor Stage
I	77	24	53
II	35	15	20
III-IV	38	26	12	14.383	0.001

A Kaplan-Meier analysis of overall survival (OS) was conducted to evaluate the potential use of SGK1 as a marker of lung adenocarcinoma (Figure 2). High SGK1 expression correlated significantly with worse OS (log rank P = 0.026) in the study cohort. Patients in the elevated SGK1 group had worse survival outcomes than did those in the low SGK1 group, with respective median OS durations of 40 and 50 months. Taken together, SGK1 expression appears to be an adequate predictor of OS, although it was not verified as an independent prognosticator.

Figure 2.

Overexpression of SGK1 in lung adenocarcinoma confers poor prognosis. Kaplan-Meier Overall Survival (OS) analysis showed that the OS of SGK1 high-expression group patients were significantly lower than the SGK1 low-expression group patients.

In the study cohort, the OS prognostic values of other clinical parameters were also evaluated using univariate and multivariate analyses. The tumor stage (P = 0.003), lymph node type (P < 0.001), and SGK1 expression (P = 0.029) were all deemed adequate prognosticators for survival outcomes in the univariate analysis. Of these, only the lymph node type remained an independent prognosticator of OS in a detailed multivariate analysis (P = 0.017). Age, sex, smoking history, histological grade, and metastasis were not found to be adequate predictors of OS (Table 3).

Table 3.

Univariate and multivariate analysis of different prognostic factors for overall survival in 150 patients with lung adenocarcinoma

Variable	Univariate analysis			Multivariate analysis
	Risk ratio	95% CI	P-value	Risk ratio	95% CI	P-value
Age (years)
≥ 59 vs < 59	0.647	0.374-1.119	0.119
Gender
Male vs Female	1.311	0.772-2.227	0.316
Smoking
Yes vs No	0.957	0.554-1.654	0.876
Pathological grade
G2-3 vs G1	0.807	0.322-2.027	0.649
Tumor stage
T2-4 vs T1	0.306	0.139-0.677	0.003	0.469	0.199-1.106	0.084
Lymph nodes
N1-3 vs No	0.454	0.295-0.699	< 0.001	0.574	0.364-0.905	0.017
Metastasis
Yes vs No	0.957	0.233-3.928	0.951
SGK1
High vs low	0.557	0.329-0.943	0.029	0.677	0.396-1.156	0.153

Regulatory role of SGK1 in cell proliferation and survival

To investigate the mechanistic role of SGK1 in lung adenocarcinoma proliferation and progression, SGK1 expression was evaluated in the A549, PC9, HCC827, NCI-H292, NCI-H1299, NCI-H1650, and NCI-H1975 human lung adenocarcinoma cell lines. A549 and NCI-H1975 exhibited the highest relative expression levels of SGK1 mRNA (0.91 and 0.81, respectively) and protein (Figure 3A, 3B) and were selected for subsequent SGK1 loss-of-function experiments.

Figure 3.

SGK1 was highly expressed in lung adenocarcinoma cells and its expression was suppressed by SGK1 siRNA transfection. A. Real-time PCR analysis of SGK1 mRNA expression in 7 lung cancer cell lines (A549, PC9, HCC827, NCI-H292, NCI-H1299, NCI-H1650 and NCI-H1975). B. SGK1 protein expression was determined by Western blot. Quantification (right panel) based on at least 3 independent experiments was performed by normalizing against GAPDH levels. C. The efficiency of SGK1 knockdown in A549 and NCI-H1975 cells was determined by Western blot analysis where GAPDH was used as an internal control. WT: wild-type cells. NC: control siRNA-transfected cells. siRNA1, 2 and 3: SGK1 siRNA1, 2 or 3-transfected cells. ***P < 0.05 versus NC.

SGK1 knockdown was performed using three SGK1-targeting siRNAs, and a scrambled siRNA was used as the negative control (NC). Western blotting of the transfected cells indicated that all three siRNAs efficiently knocked down SGK1 protein expression (relative to NC and wild-type cells; Figure 3C, P < 0.001). Of these, siRNA target 3 yielded the best performance and was used for the remaining experiments.

An MTT cell viability assay was used for a further evaluation of the effects of SGK1 knockdown in A549 and NCI-H1975 cells. Cell viability was evaluated 24, 48, and 72 h after siRNA knockdown. Compared with NC-siRNA, siRNA 3 clearly and significantly reduced cell viability at 48 h (P = 0.029 for A549, P = 0.018 for H1975) and 72 h in both cell lines (P = 0.008 for A549, P = 0.013 for H1975), although not during the immediate 24 h after knockdown. In addition to reduced cell viability, the siRNA3-transfected cells exhibited greatly reduced colony-forming capacities (P < 0.001 vs NC siRNA, Figure 4C, 4D).

Figure 4.

Knockdown of SGK1 reduced the proliferation of lung adenocarcinoma cells. A, B. MTT assay results of A549 and NCI-H1975 cells at 24 h, 48 h and 72 h after SGK1 expression was inhibited by siRNA3 transfection. C, D. Knockdown of SGK1 reduced colony formation capacities of A549 and NCI-H1975 cells. **P < 0.01, ***P < 0.001 compared to control group (siC).

SGK1 inhibition was further evaluated using the small molecule inhibitor GSK650394. A similarly timed cell viability evaluation was performed using cells treated with 0-, 5-, and 10-µM GSK650394. At 24, 48, and 72 h, both 5- and 10-µM GSK650394 significantly diminished the viability of both lung cancer cell lines, compared to cells not treated with the agent (analysis of variance-A549: P = 0.013 for 24 h, < 0.001 for 48 h and 72 h; H1975: P = 0.007 for 24 h, < 0.001 for 48 h and 72 h, Figure 5A). Inhibitor-treated A549 cells exhibited a six-fold reduction in colony-forming capacity relative to NC (P < 0.001, Figure 5B), whereas H1975 cells exhibited a 9-fold reduction (P < 0.001, Figure 5C).

Figure 5.

Inhibition of SGK1 reduced the proliferation of lung cancer cells. A, B. MTT assay results of A549 and NCI-H1975 cells at 24 h, 48 h and 72 h after SGK1 was inhibited by GSK650394. C, D. Imbibition of SGK1 reduced colony formation capacities of A549 and NCI-H1975 cells. **P < 0.01, ***P < 0.001 compared to control group.

Regulatory role of SGK1 in cell cycle progression

Cell cycle distribution in the SGK1 knockdown cells was evaluated using flow cytometry. In both cell lines, SGK1 knockdown cells exhibited a significantly reduced proportion in G1 (P < 0.001 vs NC-siRNA treated cells) and increased proportion in G2 (P < 0.001 vs NC si-RNA treated cells; Figure 6A, 6B). Furthermore, siRNA 3-treated cells exhibited a trend toward a reduced proportion of cells in S phase relative to NCs, although this difference was not significant. Further analysis of the cell cycle-regulating proteins cyclin B1 and cdc2 showed that SGK1 knockdown reduced the expression of both proteins (P < 0.001). An analysis of GSK650394-treated cells confirmed that SGK1 inhibition induced G2/M phase cell cycle arrest, as in both cell lines, the proportion of cells in G1 phase (but not S phase, Figure 7A, 7B) decreased in response to GSK650394 concentrations as low as 5 µM (P < 0.001 vs 0 µM). The finding that the proportion of cells in G2 phase had nearly doubled in response to GSK650394 treatment corroborated this finding (P < 0.001). Once again, SGK1 inhibition (by a small molecule inhibitor) led to consequent decreases in the expression of the cell cycle proteins cyclin B1 and cdc2 (P < 0.001, Figure 7C).

Figure 6.

Knockdown of SGK1 induces G2/M arrest. A: Representative cell cycle distribution of SGK1 knockdown assessed by flow cytometry in A549 and NCI-H1975 cells. Non-targeting siRNA (siC) was used as negative control for siRNA treatment. B: Flow cytometry analysis was used to determine the percentage of cells in each cell-cycle phase. Each bar represents the mean ± SD. ***P < 0.001. C: Knockdown of SGK1 resulted in the decrease of Cyclin B1 and Cdc2 protein expressions in A549 and NCI-H1975 cells. GAPDH was used as a loading control. Band intensity was normalized by loading amount as indicated below each panel.

Figure 7.

Inhibition of SGK1 affects lung adenocarcinoma cell cycle distribution. A: Representative cell cycle distribution of SGK1 inhibited by GSK650394 assessed by flow cytometry in A549 and NCI-H1975 cells. B: Flow cytometry analysis was used to determine the percentage of cells in each cell-cycle phase as shown in the diagram. Each bar represents the mean ± SD. ***P < 0.001. C: Inhibition of SGK1 by GSK650394 resulted in the decreases of Cyclin B1 and Cdc2 protein expression in A549 and NCI-H1975 cells. GAPDH was used as a loading control. Band intensity was normalized by loading amount as indicated below each panel.

Regulatory role of SGK1 in apoptosis signaling

Finally, an apoptosis assay was performed to address whether SGK1 silencing would affect the apoptosis of lung cancer cells. In this FITC-Annexin V and PI cell labeling and flow cytometry experiment, neither the proportion of early (Ann⁺/PI^-) nor late (Ann⁺/PI⁺) apoptotic cells was significantly affected by the siRNA knockdown of SGK1 in either A549 or NCI-H1975 cells, as indicated in Figure 8A, 8B. A proteomic analysis of the apoptotic proteins Caspase 3 and PARP further demonstrated the relative insensitivity of the apoptotic pathway to SGK1 inhibition (Figure 8C). Similar results were observed in GSK650394-treated cells (Figure 9), where neither treatment with 5- nor 10-µM inhibitor affected the apoptosis rates or expression of canonical apoptosis proteins.

Figure 8.

Apoptosis was not affected by the silencing of SGK1 in lung adenocarcinoma cells. A: Neither the proportion of early (Ann⁺/PI^-) or late (Ann⁺/PI⁺) stage apoptotic cells was significantly affected by siRNA knockdown of SGK1 in A549 cells. B: Neither the proportion of early (Ann⁺/PI^-) or late (Ann⁺/PI⁺) stage apoptotic cells was significantly affected by siRNA knockdown of SGK1 in H1975 cells. C: Apoptotic proteins Caspase-3 and PARP were insensitive to the inhibition of SGK1 in both cells. GAPDH was used as a loading control. Band intensity was normalized by loading amount as indicated below each panel.

Figure 9.

Apoptosis was not affected by GSK650394 in lung adenocarcinoma cells. A: Neither the proportion of early (Ann⁺/PI^-) or late (Ann⁺/PI⁺) stage apoptotic cells was significantly inhibition of SGK1 by GSK650394 in A549 cells. B: Neither the proportion of early (Ann⁺/PI^-) or late (Ann⁺/PI⁺) stage apoptotic cells was significantly inhibition of SGK1 by GSK650394 in H1975 cells. C: Apoptotic proteins Caspase-3 and PARP were insensitive to the inhibition of SGK1 in both cells. GAPDH was used as a loading control. Band intensity was normalized by loading amount as indicated below each panel.

Discussion

Although the Ser/Thr protein kinase SGK1 has been implicated in various cancers, much remains unknown about the importance of the SGK kinase family in the field of cancer research. The high sequence homology of SGK and Akt led to a previous belief that SGKs were functionally redundant and a consequent focus of therapeutic strategies toward Akt inhibition. The earliest indication of a PDK1-activated, Akt-independent SGK1 growth signal was reported in a breast cancer screen of targeted short hairpin RNAs [19]. The unique PI3K-independent activity of SGK1 in the presence of PI3Kα inhibition (i.e., complete Akt inactivation), has since been revealed [13]. Additional studies have provided further evidence of a unique role for SGK1 in the origin of cancer by revealing the overexpression of this protein in tumor cells with little to no Akt activation [20].

In recent years, SGK1 inhibition strategies have been used to inhibit the growth of androgen-dependent prostate cancer cells [21], to induce apoptosis in colon cancer cells (alone and synergistically with other compounds) [16], and to prevent chemically-induced colon carcinogenesis in experimental animal models [6]. Despite these advances, little is known about the etiological role of SGK1 in lung adenocarcinoma. Given the emerging role of this protein as a regulator of cancer cell progression and survival and its potential involvement in cell cycle control, we sought to investigate the patterns of SGK1 expression in human lung adenocarcinoma tissues and its contributions to lung cancer cell proliferation, cell cycle regulation, and apoptosis.

We observed significantly elevated SGK1 protein expression in the cytoplasm of tumor cells, relative to cells in adjacent noncancerous sections. We also found significant correlations of SGK1 expression in lung adenocarcinoma with lymph node metastasis, distant metastasis, and pathological staging (P < 0.01). These results echo the findings of Abbruzzeze et al., who reported that increased SGK1 mRNA levels in 66 NSCLC samples were concordant with an increased tumor size, elevated tumor grade, and worse clinical stage [5]. Interestingly, that earlier study found no correlation between SGK1 protein expression and various clinical parameters, which could be attributed to the use of different tumor pathologic types and the sample size. This study focused on lung adenocarcinoma and included a relatively larger sample size (150 cases).

The prognostic use of SGK1 in adrenocortical tumor samples has been reported [4]. Subsequent multivariate and univariate analyses of clinical parameters, including SGK1 expression and patient survival, revealed that SGK1 expression was an effective predictor of OS (P = 0.029), although not an independent predictor of lung adenocarcinoma. Other prognostic clinical parameters in our patient cohort included the tumor stage and lymph node type and represent the first examples of SGK1 as a prognosticator in human lung adenocarcinoma samples.

A previous study of lung cancer tissues identified TRIM11 as a prognostic marker, but that study featured a relatively larger histopathological characteristic distribution bias, and most cases involved grade 2 to 3 disease. Furthermore, the inclusion of only four stage IV cases might have limited the ability of the authors to systematically assess the prognostic value of TRIM11. In vitro, TRIM11 knockdown led only to reduced cell growth and invasiveness, possibly by suppressing ERK and PI3K/Akt signaling [22]. Incidentally, the study reported the reduced levels of PCNA and cyclin D1 protein expression, thus corroborating a possible role for PI3K/Akt signaling in the proliferation of lung cancer cells via the regulation of cell cycle proteins.

To more clearly understand the role of SGK1 in lung carcinogenesis, we manipulated SGK1 expression in two lung cancer cell lines. Most notably, both siRNA-mediated SGK1 knockdown and SGK1 inhibitor treatment reduced the viability and colony-forming capacities of the cells, thus implicating SGK1 in the survival and proliferation of lung cancer cells. An additional cell cycle distribution analysis of these cells revealed that SGK1 inhibition caused G2 phase arrest and down-regulated the protein expression of cyclin B1 and cdc2. Cyclin B1 is a G2/M phase-specific regulatory protein that forms complexes with cdc2 and thus activates the complexes’ protein kinase activity. The mechanism by which SGK1 reduces the expression of these complex subunits will require further investigation.

Previously, SGK1 was found to promote G1 cell cycle progression through the cytoplasmic phosphorylation of p27, which facilitates cyclin D-cdk complex formation during the early G1 phase [23,24]. Constitutive SGK1 overexpression, however, led to the misexpression of phosphorylated p27 and subsequent dysregulation of cell cycle progression. Other evidence to support SGK1-mediated cell cycle regulation includes MDM2-directed ubiquitination and subsequent degradation of p53, a regulator of G1/S exit and G2/M phase cyclin B1-cdk inhibition that is frequently mutated in various cancers [18]. In our study, we demonstrated that SGK1 inhibition is an effective strategy for inducing cell cycle arrest in lung cancer cells that may be experiencing SGK1-mediated cell cycle dysregulation.

SGK1 might also contribute to disease through survival signaling. In our study, we observed that SGK1 knockdown decreased lung cell viability, thus implicating SGK1 expression as a promoter of lung cancer cell survival. This suggestion agrees with the findings of a previous report, whereby SGK1 was shown to play a role in the IL-2-mediated survival of kidney cancer cells and to inhibit drug-induced apoptosis through a possible drug resistance mechanism [25]. Nasir et al. alternatively proposed evidence to suggest that SGK1 might promote tumor cell survival by negatively regulating the pro-apoptotic FOXO3a transcription factor, which then promotes the pro-apoptotic Bcl-2 family member BIM [26]. Further evidence of a role for SGK1 in survival signaling came from a report that described SGK1-related beta-catenin overexpression in intestinal tumors, which was attributed to the phosphorylation of SGK1 and consequent inactivation of the beta-catenin degradation enzyme, GSK3 beta [27]. Finally, a colon cancer cell study combined the propositions of both Nasir et al. and Wang et al. by suggesting that SGK1 is a direct downstream target of beta-catenin. Following the Wnt signal-induced enhancement of beta catenin, increased levels of SGK1 act on FOXO3a mislocalization and inhibit the expression of its downstream pro-apoptotic targets [28]. We note, however, that although we observed diminished cell viability and proliferation with SGK1 inhibition, neither the apoptosis rate nor the expression of apoptosis proteins (e.g., caspase 3, PARP) were affected by either siRNA or small molecule inhibition. Possibly, the detailed mechanisms vary among tumor types. In lung adenocarcinoma, SGK1 inhibition did not induce apoptosis. This finding suggests the involvement of different mechanisms in different cells and warrants further studies to validate this issue.

In summary, SGK1 protein expression was reliably upregulated in lung adenocarcinoma samples and was found to correlate with OS in a univariate analysis, although this correlation did not persist in a multivariate analysis. In other words, SGK1 histologic detection could help to determine a patient’s prognosis but cannot be used independently to predict OS. We also present the first evidence of differential SGK1 expression in human lung adenocarcinomas, and our data support findings previously revealed only in human lung squamous carcinomas. We elucidated a nascent profile of SGK1 expression in the context of lung cancer cell proliferation; namely, we demonstrated the regulatory control of SGK1 over cell cycle progression from the G1 to G2 phase and determined that this is likely mediated by controlling cyclinB1-cdc2 complex formation and function. Although SGK1 inhibition clearly affected cell survival in this study, the underlying signaling mechanisms remain less clear. Furthermore, SGK1 inhibition did not obviously affect apoptosis rates or canonical apoptosis protein expression, despite previous evidence of a role for SGK1-mediated apoptosis and survival regulation. Additional studies are clearly required to form a more complete image of the signaling pathways by which SGK1 governs lung cancer cell proliferation, survival, and growth, particularly in a preclinical disease model. Until then, the information reported in this study supports SGK1 inhibition as a therapeutic strategy for lung adenocarcinoma and offers a prognostic role for SGK1 expression in the pathological evaluation of human patients with lung adenocarcinoma. Finally, our findings and those of previous studies suggest an important carcinogenic role for SGK family members and therefore merit further investigation.

Disclosure of conflict of interest

None.