Overexpression of geranylgeranyl diphosphate synthase contributes to tumour metastasis and correlates with poor prognosis of lung adenocarcinoma

Abstract This study aimed to evaluate the biological role of geranylgeranyl diphosphate synthase (GGPPS) in the progression of lung adenocarcinoma. GGPPS expression was detected in lung adenocarcinoma tissues by qRT‐PCR, tissue microarray (TMA) and western blotting. The relationships between GGPPS expression and the clinicopathological characteristics and prognosis of lung adenocarcinoma patients were assessed. GGPPS was down‐regulated in SPCA‐1, PC9 and A549 cells using siRNA and up‐regulated in A549 cells using an adenoviral vector. The biological roles of GGPPS in cell proliferation, apoptosis, migration and invasion were determined by MTT and colony formation assays, flow cytometry, and transwell and wound‐healing assays, respectively. In addition, the regulatory roles of GGPPS on the expression of several epithelial‐mesenchymal transition (EMT) markers were determined. Furthermore, the Rac1/Cdc42 prenylation was detected after knockdown of GGPPS in SPCA‐1 and PC9 cells. GGPPS expression was significantly increased in lung adenocarcinoma tissues compared to that in adjacent normal tissues. Overexpression of GGPPS was correlated with large tumours, high TNM stage, lymph node metastasis and poor prognosis in patients. Knockdown of GGPPS inhibited the migration and invasion of lung adenocarcinoma cells, but did not affect cell proliferation and apoptosis. Meanwhile, GGPPS inhibition significantly increased the expression of E‐cadherin and reduced the expression of N‐cadherin and vimentin in lung adenocarcinoma cells. In addition, the Rac1/Cdc42 geranylgeranylation was reduced by GGPPS knockdown. Overexpression of GGPPS correlates with poor prognosis of lung adenocarcinoma and contributes to metastasis through regulating EMT.


Introduction
Lung adenocarcinoma is a common and severe tumour that belongs to non-small cell lung cancers (NSCLC). Although great advances have been made in the diagnosis and treatment of NSCLC, the worldwide mortality rate remains high [1]. Patients with NSCLC often relapse and develop metastases after surgery, chemotherapy and/or radiation therapy, resulting in an overall five-year survival rate of less than 18% [2]. Researches into the molecular mechanisms underlying NSCLC progression and discovery of novel biomarkers and therapeutic targets have become hot topics in the clinic.
Geranylgeranyl diphosphate (GGPP) is an isoprenoid synthesized through the cellular mevalonate pathway [3]. Defects in isoprenoid biosynthesis can be accompanied by various disease, including cancers, and metabolic and cardiovascular diseases [4]. Geranylgeranyl diphosphate synthase (GGPPS) is an important branch-point enzyme in the mevalonate metabolic pathway, which can synthesize GGPP from farnesyl diphosphate (FPP) [5,6]. Abnormal GGPPS expression can affect the relative levels of FPP and GGPP, participating in the regulation of many diseases such as cigarette smoke-induced pulmonary disease [7], insulin resistance [8,9], male infertility and cardiac hypertrophy [10,11]. Furthermore, digeranyl bisphosphonate inhibits GGPPS, resulting in the inhibition of protein geranylgeranylation in MDA-MB-231 cells, and contributing to the suppression of breast cancer cell migration [12]. GGPPS is also important in the occurrence and progression of cirrhosis-induced hepatocellular carcinoma (HCC) and can be used to predict the biological characteristics of HCC [13]. However, related researches on the specific biological roles of GGPPS in lung adenocarcinoma are still limited.
In this study, GGPPS expression was detected in lung adenocarcinoma, and its relationships with clinicopathological characteristics and prognosis of patients were assessed. The regulatory roles of GGPPS on cell proliferation, apoptosis, migration and invasion of lung adenocarcinoma cells were analysed, and the regulatory mechanisms of GGPPS in EMT were further evaluated. Our findings may reveal the biological roles of GGPPS in the occurrence and progression of lung adenocarcinoma. GGPPS may thus be considered an independent biomarker for the prognosis of lung adenocarcinoma and a potential therapeutic target.

Materials and methods
Patients and tissue samples A total of 32 adenocarcinoma tissues and adjacent normal tissues were collected from patients with lung adenocarcinoma. These patients had undergone surgical resection at the Department of Thoracic Surgery, Jinling Hospital, Nanjing, China, between July 2015 and February 2016. Patients who had received preoperative radiotherapy or chemotherapy were excluded from this study. All tissues were confirmed histopathologically as lung adenocarcinoma or adjacent normal lung tissue and stored in liquid nitrogen. This study was approved by the Institutional Review Board of Southern Medical University. Informed consents were obtained from all enrolled patients.

Cell transfection with GGPPS adenovirus vector
GGPPS (Ad-GGPPS) and control (Ad-GFP) adenovirus vectors were kind gifts from Professor Chao-jun Li [7]. A549 cells were transfected when the cell density reached 50%. multiplicity of infection was 20. Green fluorescence was found in cells at 60 hr post-transfection, and cells were harvested for qRT-PCR and western blot analysis at 72 hr post-transfection.

MTT and colony formation assay
Cell viability of transfected cells was assessed every 24 hr by MTT, using a Cell Proliferation Reagent Kit I (Roche Applied Science Indianapolis, USA). A total of 750 transfected cells were plated onto six-well plates with fresh medium containing 10% FBS. Medium was refreshed every 4 days. After 10-14 days of culturing, colonies were fixed with methanol, stained with 0.1% crystal violet (Sigma-Aldrich, St. Louis, MO, USA) and manually counted.

Flow cytometry
Apoptosis in transfected cells was analysed by flow cytometry (FACScan, BD Biosciences, San Jose, CA, USA) using CellQuest software (BD Biosciences). By staining with FITC-annexin V and propidium iodide (PI), late apoptotic cells, early apoptotic cells, viable cells and dead cells were classified, and the ratio of apoptotic cells was calculated.

Transwell migration and invasion assay
A total of 5 9 10 4 cells and 1 9 10 5 cells, grown in medium containing 1% FBS, were isolated for migration and invasion assays, respectively. These cells were seeded into the upper chamber of an insert (8-lm pore size, Millipore). For invasion assays, the upper chamber of insert was pre-coated with Matrigel (Corning, 356234). Medium containing 10% FBS was added to the lower chamber. After incubation at 37°C with 5% CO 2 for 24 hr, cells in the upper layer were removed with cotton wool, and cells in lower chamber were fixed with methanol and stained with 0.1% crystal violet (Sigma-Aldrich). Stained cells were photographed and counted in five random fields.

Wound-healing assay
A total of 5 9 10 4 transfected cells were isolated for wound-healing assay. After seeded onto dishes, a wound track was scored into each dish with a plastic scraper. Debris was removed by washing with PBS. After 0, 24 and 48 hr of culturing, photographs were captured using a microscope and the extent of wound healing was measured.

Prenylation and membrane association measurements
Protein prenylation was measured as described previously [8]. In brief, SPCA-1 and PC9 cells with the indicated treatment lysed in 500 ll lysis buffer. The total protein concentration was diluted to 1 mg/ml and partitioned with same volume of 4% Triton X-114 for 5 min. at 37°C to solubilize and fractionate the lipid-rich cell membrane. The aqueous upper phase contains the water-soluble small GTPases, and the organic lower phase contains the lipid-soluble small GTPases. The membrane proteins of transfected cells were extracted using Mem-PER Plus Membrane Protein Extraction Kit (Thermo Scientific, 89842). Simply, 5 9 10 6 cells were harvested by 5 min. of centrifugation at 300 9 g and washed with 3 ml cell wash solution. About 0.75 ml permeabilization buffer was added into cell depositions and incubated for 10 min. at 4°C with constant mixing. After 15 min. of centrifugation at 16,000 9 g, the supernatant containing cytosolic proteins was collected. Then, 0.5 ml solubilization buffer was added into cell depositions and incubated for 30 min. at 4°C with constant mixing. After 15 min. of centrifugation at 16,000 9 g, supernatant containing solubilized membrane and membrane-associated proteins was collected. The subcellular fractions were immunoprecipitated with Rac1/ Cdc42 antibody (Cell Signaling Technology, Danvers, MA, USA), and subsequently subjected to western blot analysis.

Immunoprecipitation
Immunoprecipitation was performed according to a standard protocol [14]. Rac1/Cdc42 antibody was used to form immune complex with Rac1/Cdc42 proteins in lysates and was immunoprecipitated down with protein A/G plus agarose. Finally, the equivalent protein samples were subjected to Western blot analysis against Rac1/Cdc42.

Western blot assay
Total proteins were isolated from tissues and cells, using lysis buffer containing mammalian protein extraction reagent RIPA (Beyotime, China), PMSF (Roche) and protease inhibitor cocktail (Roche, Basel, Switzerland). Protein concentration was determined using a Bio-Rad Protein Assay kit

Statistical analysis
All experiments were performed in triplicate, and data were expressed as mean AE standard deviation (S.D.). All statistical analyses were performed using SPSS 20.0 software (IBM, Chicago, IL, USA). An independent sample t-test was performed to compare the mean values of 2181 different groups. A chi-square test was performed to evaluate the correlation between GGPPS and the clinical characteristics of patients. Kaplan-Meier survival analysis was performed to assess the correlation between GGPPS and overall survival of patients. A Cox regression model was used to analyse independent prognostic risk factors. A Pvalue less than 0.05 was considered significantly different.

GGPPS is up-regulated in lung adenocarcinoma
The expression of GGPPS has been reported to be significantly higher in lung adenocarcinoma tissues than in normal lung tissues [15][16][17][18][19] (Figs 1A and S1). In consistent with these findings, we found that Ggps1 mRNA expression was significantly higher in lung adenocarcinoma tissues than in adjacent normal tissues (P < 0.001, Fig. 1B). At the protein level, TMA showed a significantly higher GGPPS content in lung adenocarcinoma tissues than in adjacent normal tissues (P < 0.001, Fig. 1C). Furthermore, the increased GGPPS protein expression in lung adenocarcinoma tissues was also confirmed by western blotting (P < 0.01, Fig. 1D).

Relationship between GGPPS expression and the clinicopathological characteristics and prognosis of patients with lung adenocarcinoma
The relationship between GGPPS expression (staining score) and the clinicopathological parameters of patients was assessed. The results showed that the increased GGPPS expression in lung adenocarcinoma was significantly associated with large tumour size (>3 cm, P < 0.05), lymph node metastasis (P < 0.001) and high TNM stage (P < 0.001, Fig. 2A) ( Table 1). The 85 patients could be divided into high-expression and low-expression groups, according to GGPPS expression levels, using 6.4 as a cut-off value (<6.4) (Fig. 2B).
The relationship between GGPPS expression and patient prognosis was further analysed by Kaplan-Meier survival analysis. As shown in Fig. 2Be, patients with high expression levels of GGPPS exhibited a significantly shorter OS than those with low expression levels (P < 0.001, Fig. 2Be). Furthermore, Cox regression analysis confirmed that high GGPPS expression contributed to the poor prognosis of lung adenocarcinoma (HR = 3.539, 95% CI: 1.652-7.581) ( Table 2).

Biological function of GGPPS in lung adenocarcinoma cells
To evaluate the biological function of GGPPS in lung adenocarcinoma cells, GGPPS expression levels were measured in five NSCLC cell lines and HBE cells. Results showed that the expression of GGPPS was relatively higher in SPCA-1, H1975 and PC9 cells, slightly higher in A549, but relatively lower in H1299 than in HBE cells (Fig. 3A).
GGPPS was down-regulated in SPCA-1, PC9 and A549 cells using siRNA and up-regulated in A549 cells using adenoviral vector Ad-GGPPS (P < 0.001, Figs 3B and S2A). The effects of GGPPS expression on cell proliferation were determined by MTT and colony formation assays. However, neither down-regulation nor up-regulation of GGPPS influenced the growth of SPCA1, PC9 and A549 cells (Figs 4A and B;  Fig. S2B,C). In addition, flow cytometry showed that the changed expression of GGPPS also did not affect the apoptosis of those cells (Figs 4C and S2D). However, transwell assay revealed that the downregulation of GGPPS significantly inhibited the migration and invasion of lung adenocarcinoma cells (SPCA-1, PC9 and A549 cells), and up-regulation of GGPPS significantly increased the migration and invasion of A549 cells (P < 0.01, Figs 5 and S3D). Importantly, GGPP administration recovered the inhibited migration and invasion of SPCA-1 and PC9 cells transfected with siRNA-GGPPS (P < 0.01, Fig. 5A1, B1, A2, B2). Moreover, the inhibited migration induced by GGPPS down-regulation was also identified in SPCA-1, PC9 and A549 cells by wound-healing assay (P < 0.01, Fig. S3A, B, C).  Mechanism underlying the regulatory role of GGPPS in the migration and invasion of lung adenocarcinoma cells EMT is a key mechanism underlying the migration and invasion of tumour cells [20,21]. Thus, we assayed the expression of several important EMT markers in GGPPS downexpressed SPCA-1, PC9, A549 cells and in GGPPS overexpressed A549 cells. As shown in Figs 6 and S4, the expression of E-cadherin was significantly increased in GGPPS downexpressed SPCA-1, PC9, A549 cells, while the expression of N-cadherin and vimentin was reduced (P < 0.01, Figs 6A1, B1, A2, B2 and S4). On the contrary, the expression of Ecadherin in GGPPS overexpressed A549 cells was significantly decreased, and the expression of N-cadherin and vimentin was increased (P < 0.01, Fig. 6C1, C2). Besides, GGPP administration recovered the expression of E-cadherin, N-cadherin and vimentin in SPCA-1 and PC9 cells transfected with siRNA-GGPPS (P < 0.01, Fig. 6A1, B1, A2, B2).

Down-regulation of GGPPS in SPCA-1 and PC9 cells reduced Rac1/Cdc42 geranylgeranylation
Previous studies have indicated that GGPPS deletion can reduce the content of GGPP and increase the content of FPP [10,11]. To determine whether protein prenylation was disrupted by downexpressed GGPPS in lung adenocarcinoma cells, the level of GGPP was detected. The HPLC-MS/MS showed that the GGPP level was significantly decreased in SPCA-1 and PC9 cells transfected with siRNA-GGPPS (P < 0.01, Fig. 7A and B). The Rho GTPases Rac1 and Cdc42, representative geranylgeranylated proteins, have been reported to be able to regulate cell migration [22][23][24]. We examined the hydrophobicity and membrane association of Rac1/Cdc42 in SPCA-1 and PC9 cells transfected with siRNA-GGPPS. The results showed the Rac1/Cdc42 geranylgeranylation was significantly reduced in GGPPS downexpressed SPCA-1 and PC9 cells (P < 0.01, Fig. 7C and D). The Rac1/ Cdc42 membrane association and its expression in organic phase were significantly decreased (P < 0.01, Fig. 7E-H). Further analysis revealed that the decreased Rac1/Cdc42 prenylation following knockdown of GGPPS in SPCA-1 and PC9 cells was recovered with GGPP administration (P < 0.01, Fig. 7I and J).

Discussion
Protein prenylation is an important post-translational protein modification that is critical for the membrane association of a number of signalling proteins involved in cell differentiation, proliferation and migration [25]. A series of signalling pathways will be altered once this modification is disrupted, which accounts for the pathological progression of many diseases including neurodegenerative disorders and cancers [4]. The mevalonate pathway is considered as a target in the treatment of cancers [26][27][28]. It has been reported that geranylgeranyl transferase inhibitors (GGTIs) can act as anti-cancer agents by inhibiting the proliferation and invasion of cancer cells [29][30][31]. Statins (HMG-CoA reductase inhibitor) and bisphosphonates (farnesyl diphosphate synthase inhibitor) have been shown to promote apoptosis and inhibit cancer cell migration [32,33], while the anticancer effects of statins and bisphosphonates are inhibited by GGPP [34][35][36]. Geranylgeranyl diphosphate synthase (GGPPS) is an important branch-point enzyme in the mevalonate metabolic pathway and participates in regulating the protein prenylation. Therefore, it is worthy to study the biological role of GGPPS in cancer, and whether GGPPS can serve as a target in the treatment of cancer.
In the present study, significantly higher GGPPS expression was found in lung adenocarcinoma tissues than in adjacent normal tissues. This result is consistent with that of previous research [13], and further identifies an important role for GGPPS in lung adenocarcinoma. GGPPS up-regulation was associated with increased tumour size, lymph node metastasis, advanced TNM stage and poor prognosis of lung adenocarcinoma. These results indicate that GGPPS is closely related to tumour development and may thus be used as a biomarker and therapeutic target in lung adenocarcinoma.
GGPPS inhibition can reduce RhoA geranylgeranylation and membrane localization through GGPP depletion, and thus inhibit breast cancer cell migration [12]. Previous studies have indicated that GGPPS knockout altered protein prenylation in Sertoli cells and Neonatal rat ventricular myocytes (NRVMs) [10,11]. In the present study, we found that the Rac1/Cdc42 geranylgeranylation was decreased by knockdown of GGPPS, and the GGPP administration could recover the changes. Furthermore, GGPPS down-regulation inhibited the migration and invasion of lung adenocarcinoma cells, and GGPPS up-regulation significantly increased the migration and invasion of A549 cells. However, the proliferation and apoptosis of lung adenocarcinoma cells were not affected. These results may suggest that GGPPS down-regulation reduces prenylation and membrane association of small GTPase protein which regulates cell migration, thereby inhibiting the migration and invasion of lung adenocarcinoma cells.
The molecular mechanisms of GGPPS in the migration and invasion of lung adenocarcinoma cells were further evaluated in the present study. E-cadherin expression was significantly increased following knockdown of GGPPS, while the expression of N-cadherin and vimentin was reduced. In addition, the expression of E-cadherin in GGPPS overexpressed A549 cells was significantly decreased, while the expression of N-cadherin and vimentin was increased. Ecadherin, N-cadherin and vimentin are well-known markers of EMT [37][38][39], which is a vital process in the invasion and metastasis of tumour cells [40,41]. Our study demonstrated for the first time that GGPPS contributes to the migration and invasion of lung adenocarcinoma cells, at least partially through the regulation of EMT. Additionally, the Rho GTPases, Rac1 and Cdc42 are known to regulate cell migration, which are related to lung cancer metastasis [22][23][24]42]. These Rho GTPases are preferentially geranylgeranylated by GGPP through GGTase I [43]. We found that Rac1/Cdc42 geranylgeranylation was reduced by GGPPS knockdown. This results suggest that GGPPS down-regulation may reduce Rho GTPase geranylgeranylation and inhibit downstream signalling pathways related to EMT [44], thereby inhibiting tumour cell metastasis. In addition, the recovery effects of exogenous GGPP indicate that GGPPS knockdown is contribute to enzyme deactivation in GGPP production.

Conclusion
In conclusion, GGPPS is overexpressed in lung adenocarcinoma, and thus may function as a prognostic factor. Knockdown of GGPPS may inhibit the migration and invasion of lung adenocarcinoma through EMT regulation. However, the precise signalling pathways responsible for the biological functions of GGPPS in EMT still need to be studied. Further researches on the action mechanisms of GGPPS in lung adenocarcinoma are needed.
Fig. S1 Ggps1 mRNA expression was significantly increased in lung adenocarcinoma tissues compared to that in normal lung tissues in three studies. Data were obtained from the publicly available database Oncomine.