Deregulated expression and activity of Farnesyl Diphosphate Synthase (FDPS) in Glioblastoma

Abstract

Glioblastoma (GBM), the most aggressive brain cancer, is highly dependent on the mevalonate (MVA) pathway for the synthesis of lipid moieties critical for cell proliferation but the function and regulation of key intermediate enzymes like farnesyl-diphosphate synthase (FDPS), up to now, remained unknown. A deregulated expression and activity of FDPS was the central research idea of the present study. FDPS mRNA, protein and enzyme activity were analyzed in a cohort of stage III-IV glioma patients (N = 49) and primary derived cells. FDPS silencing helped to clarify its function in the maintenance of malignant phenotype. Interestingly, compared to tumor-free peripheral (TFB) brain and normal human astrocytes (NHA), FDPS protein expression and enzyme activity were detected at high degree in tumor mass where a correlation with canonical oncogenic signaling pathways such as STAT3, ERK and AKT was also documented. Further, FDPS knockdown in U87 and GBM primary cells but not in NHA, enhanced apoptosis. With the effort to develop a more refined map of the connectivity between signal transduction pathways and metabolic networks in cancer FDPS as a new candidate metabolic oncogene in glioblastoma, might suggest to further target MVA pathway as valid therapeutic tool.

Figure 1

Dysregulated FDPS expression in GBMs. (A) Real-time PCR analysis of FDPS in NHA, 7 tumor free brain tissues and 39 glioma tissues (G2-G42). Data, expressed as fold change units, were normalized with β-actin and referred to the NHA considered as calibrator. Columns represent mean ± SD of the results performed in triplicates. (B) Representative Western blot showing the basal protein levels of p-STAT3, STAT3, p-AKT, AKT, p-ERK, FDPS and PCNA in NHA and 27 tumor brains (G16 astrocytoma grade II, G2-23 astrocytoma grade III, G5-G31 glioma grade III, G7-G42 glioblastoma grade IV); α-Tubulin was used as loading control. Each set of samples (NHA-G21) and (G24-G42) has been run in parallel. The two panels show the representative western blots of three different experiments performed with similar results whose densitometric analysis is showed in Supplementary Fig. 2. The graphs below represent correlation analysis of western blot results for the protein expression of FDPS and the protein expression of p-STAT3, p-AKT, p-ERK and PCNA. Plotted values are normalized on corresponding ones in NHA (C) Western blot analysis for p-STAT3, STAT3, p-AKT, AKT, p-ERK, FDPS and PCNA in 7 tumor brains divided into the Central (C), Intermediate (I), and Peripheral (P) fraction as explained in Supplementary Fig. 1. β-actin was used as loading control. Each set of samples (G20c-G28p) and (G45c-G50p) has been run in parallel. The two panels show representative Western blots of three different experiments performed with similar results whose densitometric analysis is showed in Supplementary Fig. 3. The graphs below represent correlation analysis of western blot results for the protein expression of FDPS and the protein expression of p-STAT3, p-AKT, p-ERK and PCNA. Plotted values are normalized on corresponding ones in tumor infiltrated Central tissue (C). (D) FDPS quantitative analyses of the western blot results of Fig. 1C. FDPS values in I and P fractions are normalized on corresponding ones in tumor infiltrated Central tissue. Histograms represent mean ± SD in densitometry units of scanned immunoblots from the 3 different experiments. (E) FDPS gene (left panel) and protein (right panel) expression in tumor free brains, low grade gliomas and high grade gliomas groups as respectively detected by RT-PCR and western blotting. Quantitative analyses of the results are shown in histograms. Data are presented as the mean ± SD of the results performed in triplicates. (ANOVA, ***p < 0.001).

Figure 2

Functional analysis of the role of FDPS in GBMs. (A,B) Representative Western blot showing pSTAT3, STAT3, p-AKT, AKT, p-ERK, ERK FDPS and PCNA protein levels in 7 human primary glioma cell lines established from the indicated cancer patients (G18, G23, G24, G27, G37, G39, G50) (A), in NHA and in the four indicated glioma cell lines; β-actin or α-Tubulin were used as loading control. Panels show representative blots of three different experiments performed with similar results. The tables below report correlation analysis of western blot results for the protein expression of FDPS and the protein expression of p-STAT3, p-AKT, p-ERK and PCNA. (C) U87MG cells or U87MG cells transfected with siRNA FDPS were cultured for 48 h in presence or absence of EGF in the last 8 minutes before cells lysis; cell lysates were immunoblotted for p-STAT3, STAT3, p-AKT, AKT, p-ERK, ERK, FDPS, Mcl-1, BCL-XL and α-Tubulin as loading control. Data are representative of 3 independent experiments performed with similar results. (D) Distribution of U87MG cells or U87MG cells transfected with siRNA FDPS in the different cell cycle phases. All the results shown are representative of three independent experiments performed in duplicate, expressed as mean ± SD (ANOVA, **p < 0.01vs control). (E) Cytofluorimetric assessment of apoptosis in U87MG cells or U87MG cells transfected with siRNA FDPS or Scramble siRNA. Histograms indicate the total percentage of early (AV + /PI- cells) and late apoptotic events (AV + /PI + cells) as well as necrotic cells (AV-/PI + cells). All the results shown are representative of three independent experiments (ANOVA, *** p < 0.001, ** p < 0.01). (F) Patient-derived primary cell line (GBM39) or GBM39 transfected with siRNA FDPS were cultured for 48 h in presence or absence of EGF in the last 8 minutes before cells lysis; cell lysates were immunoblotted for p-STAT3, STAT3, p-AKT, AKT, p-ERK, ERK, FDPS, Mcl-1, BCL-XL and α-Tubulin as loading control. Data are representative of 3 independent experiments performed with similar results. (G) NHA cells or NHA cells transfected with siRNA FDPS were cultured for 48 h; cell lysates were immunoblotted for p-STAT3, STAT3, p-AKT, AKT, p-ERK, ERK, FDPS, Mcl-1 and α-Tubulin as loading control. Data are representative of 3 independent experiments performed with similar results. (H,I) Detection of apoptosis in patients-derived primary cell line (GBM39), both in basal condition and after FPDS silencing (H) and in NHA cells or NHA cells transfected with siRNA FDPS or Scramble siRNA (I). All the results shown are representative of three independent experiments (ANOVA, ***p < 0.001, **p < 0.01, *p < 0.05).

Figure 3

Densitometric analysis of FDPS protein expression and FDPS activity in tumor brain and in tumor free brain tissues. (A) Table of densitometric analysis values of FDPS protein expression normalized on own β-actin or α-Tubulin (arbitrary units) and FDPS activity values (expressed in nmol/min/mg prot) in G4-G48 glioblastoma tumor brain and in tumor free brain tissues; the same values were plotted in the next scatter plots (B,C) where bars indicate median expression. Results are representative of 3 different independent determination for each indicated patient brain tissues.

Figure 4

Schematic representation of the main molecular findings. Looking for novel brain cancer biomarkers, the protein, the enzymatic and the gene expression determinations of FDPS have been conducted in different intraoperative brain sample sites of a series of GBM patients. Compared with the normal human astrocyte (NHA) and Tumor Free Brains (TFB), human tumor infiltrated brains (TIB) displayed an aberrant expression of this enzyme (green). This led us to begin to shed light on the role of FDPS in cell survival mainly by modulating the oncogenic signalling pathways in GBM cancer cells.