If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Gliomas are the most common form of central nervous system tumor, and high-grade gliomas (WHO III-IV) are notoriously difficult to treat. Although standard treatment involves surgical resection and radio/chemotherapy, the incidence of recurrence is very high because of the diffusing nature of these tumors. Previous studies have shown that polyunsaturated fatty acids such as gamma-linolenic acid (GLA) and eicosapentaenoic acid are cytotoxic to tumor cells both in vitro and in vivo (1–4). GLA can alter tumor cell metabolism and proliferation through the production of lipid peroxides and reactive oxygen species, changes in mitochondrial membrane lipid composition, loss of mitochondrial membrane potential, induction of apoptosis, and cell cycle arrest (3–5). Several PUFAs including GLA have been reported to alter bcl2 expression thereby inducing apoptosis (6). GLA also has synergistic effects on both radiotherapy and chemotherapy (7). The difficulty in treating patients with gliomas has led to the proposed use of GLA as an adjuvant therapy (7, 8). However, the mechanisms by which GLA causes its antitumor effects remain largely unknown in gliomas (5, 7). At subcytotoxic concentrations, GLA reduces both rates of glioma cell proliferation and migratory capacity (9), without inducing the apoptosis typically seen with higher concentrations of GLA (2, 5, 8). This study aimed to identify the possible molecular targets of GLA treatment to provide a better understanding of the mechanisms behind these previous findings in gliomas. Because GLA may alter DNA cell cycle, control and repair mechanisms was analyzed as lipid peroxides can cause substantial DNA and protein damage (6). The expression of E2F1, cyclin D1, bax, bcl2, Ku70, and Ku80 and the effects of the natural antioxidant vitamin E on expression were analyzed. The Ku proteins were chosen as previous studies have shown that loss or reduction in their expression causes increased DNA damage and micronucleus formation in the presence of radiation (10). The fact that GLA exposure is known to enhance the efficacy of radiation treatment raised the question whether the Ku proteins could be involved in this effect as seen for other molecules such as roscovitine (10) and flavopiridol (11).
MATERIALS AND METHODS
C6 rat glioma cells were obtained from the ATCC and grown in DMEM containing 10% fetal calf serum and antibiotics (penicillin 50 U/mL, streptomycin 50 μg/mL). Cells in the exponential phase of growth were used, growing in 75 cm2 flasks in a humidified atmosphere of 5% CO2, 95% air at 37°C. Cells were exposed to 150 μM GLA complexed with 0.5% albumin (9), whereas the control cells were exposed to 0.5% albumin. Cell proliferation and viability assays followed standard methods (12). The concentration of 150 μM was chosen based on the previous studies allowing inhibition of cell proliferation without extensive cytotoxicity and apoptosis (9). This concentration is clinically relevant as it can be readily achieved by dietary supplementation with fatty acids. A recent study in pregnant women detected plasma EPA concentrations of 41.5 ± 17.8 mg/L after 17 weeks supplementation (13). Cell proliferation and gene expression studies were also performed in the presence or absence of 10 μM vitamin E as a natural antioxidant.
Control and treated cells were prepared and collected following standard procedures for FACS analysis after 24 hours ± 150 μM GLA (3).
The cells were grown on glass cover slips ±150 μM GLA for 24 hours before fixation with ice-cold 4% formaldehyde in 0.1 M phosphate buffer. Standard immunohistochemical methods were used following the sequence: primary antibody, biotinylated secondary antibody, and streptavidin-Alexa 488 (9). All cells were triple-labeled for the respective protein of interest, actin (rhodamine-phalloidin) and nucleus (Hoechst 33342), and mounted with Vectashield (Vector Labs, USA).
Reverse Transcriptase-Polymerase Chain Reaction
GLA-exposed and control C6 cells were used for total RNA extraction by Trizol (Life Technologies, USA). The first strand of complementary DNA (cDNA) was generated from 1 μg RNA as previously described (12). PCR amplification cycle: 1 min at 94°C, 1 min at primer-specific temperature (55°C for Ku80; 60°C for E2F1, Ku70, GAPDH; 65°C for p53, cyclin D1, bcl-2, c-myc; and 68°C for bax), and 1 min at 72°C. To ensure the exponential phase of amplification, the number of PCR cycles was determined and optimized for each of the proteins. Controls for nonspecific amplification showed no bands on gel (data not shown). Semiquantitative gene expression data was calculated by the ratio of respective gene/GAPDH density after ethidium bromide staining. GAPDH expression remained unchanged after GLA exposure. Primers used were as follows (sense/antisense): cyclin D1 (435 bp): 5′-TGTTCGTGGCC TCTAAGATGA-3′, 5′-GCTTGACTCCAGAAGGG CTT-3′; E2F-1 (143 bp): 5′-ACGCTATGAAACCTCACT AAA-3′, 5′-AGGACATTGGTGATG TCATA-3′; Bax (352 bp): 5′-GCAGGGAGGATGGCT GGGGAGA-3′, 5′-TCCAGACAAGC AGCCGCTCACG-3′; Bcl-2 (780 bp): 5′-GCGAAGTGCTATTG GTACCTG-3′, 5′-ATATT TGTTTGGGGCAGGTCT-3′; Ku70 (414 bp): 5′-AAGAGGATCATGCTATTCACCAA-3′, 5′-TCCTTGTCTTAGTTTTCACTGG-3′; Ku80 (433 bp): 5′-GATGAGGCAGCAGCTGTT-3′, 5′-CAAAATGTGCTGCTGAAT-3′; GAPDH (306 bp): 5′-GTCGGTGTGAACGGATTTG-3′, 5′-ACAAACATGGGGGCATCAG-3′.
Western Blotting Analysis
Twenty micrograms of sample proteins were electrophoresed through 7.5% SDS-PAGE and transferred to nitrocellulose membranes (Hybond ECL membrane, Amersham Pharmacia Biotech). Immunoblotting followed standard methods (9), and fluorescent bands (Alexa-488 label) were visualized in an image system (Molecular Dynamics Typhoon 8600 Variable Mode Imager). The antibodies used in this study were as follows: E2F-1 (H-137, rabbit), GAPDH—as internal control (V-18, goat) (Santa Cruz Biotechnology, USA). Bcl-2 (ab7973-1, rabbit), bax (ab16837-100, rabbit), and cyclin D1 (ab31450-250, rabbit) were purchased from Abcam, USA. Ku70/80 (ab1358, rabbit) was obtained from Chemicon International, USA. Secondary antibodies conjugated with Alexa-488 (anti-goat, anti-mouse, and anti-rabbit) used were produced in donkey (Invitrogen–Molecular Probes).
The semiquantitative analysis of reverse transcriptase-polymerase chain reaction (RT-PCR) products was performed using a Molecular Dynamics Typhoon 8600 Variable Mode Imager. All data are presented as the mean ± standard error (SE). The results were submitted to analysis of variance followed by the unpaired t-test to compare gene expression results between treated and untreated cells. Differences were considered significant at P < 0.05.
After 24 hours of growth in the presence of 150 μM GLA, C6 cells were counted, and then a 44% inhibition of cell proliferation was seen without inducing significant apoptosis. A 49% decrease in S-phase was found in the GLA-treated cells with no significant changes seen in sub-G1, G1, and polyploid phases (Fig. 1). Hoechst labeling of DNA highlighted a significant decrease (25%) in mitotic index in the GLA-treated cells without a significant change in apoptosis (Fig. 1). However, the most important finding in the morphological analysis of DNA was the 39% increase in micronucleus formation in the GLA-treated cells. Examples of micronucleus formation and cyclin D1 staining of these structures can be seen in Fig. 1.
The effects of 150 μM GLA on mRNA expression of E2F-1, cyclin D1, Ku70, Ku80, bax, and bcl2 were analyzed during a 24-hour period, at intervals of 0, 1, 2, 3, 6, 12, and 24 hours (Figs. 2 and 3). Protein expression was determined at the end of the 24-hour period (Figs. 2 and 3). The expression of cyclin D1 mRNA was seen to increase only after 1 hour of GLA exposure and remained higher than the control value until the end of the 24-hour period (Fig. 2). However, no significant alteration in protein levels of cyclin D1 was observed by Western blot analysis (Fig. 2). In contrast, E2F1 mRNA started to decrease at 1 hour of treatment and E2F1 protein expression was seen to decrease by 45% after GLA exposure (Fig. 2).
The mRNA expression of bcl2, Ku70, and Ku80 were unaltered by GLA (Figs. 2 and 3). In contrast, Ku80 protein expression was decreased by 71% after 24 hours GLA treatment (Fig. 2). The high concentration of Ku70/80 immunolabeling in mitotic cells can be seen in Fig. 2, and loss of Ku80 expression may affect cell proliferation as well as DNA repair. Both mRNA and protein expression of the bax monomer and dimer was increased after treatment (Fig. 3). The bax/bcl2 ratio increased in the GLA-treated cells for bax 21 kDa, bax 50 kDa, and total ratio (217%) in relation to the control group (Fig. 3). The presence of the natural antioxidant vitamin E did not significantly alter the mRNA expression patterns of the proteins studied although it did partially reverse the effects of 150 μM GLA on cell proliferation (Fig. 4). The effects of GLA are therefore only, in part, dependent on peroxidation, confirming previous findings in this cell line where 150 μM GLA causes only a small increase in reactive oxygen species generation and TBARS formation (9).
In this study, 150 μM GLA treatment for 24 hours in vitro caused a downregulation of E2F1 mRNA and protein expression showing a rapid response of mRNA, starting only after 1 hour of treatment and lasting until the end of the 24-hour period (Fig. 2). Because of the importance of E2F1 in the control of S-phase entry, significant decreases in its expression may directly influence cell proliferation. Recent studies with Minerval (2-hydroxy-9-cis-octadecenoic acid), a derivative of oleic acid, showed that this drug induced cell cycle arrest before entry into S phase, causing human lung adenocarcinoma (A549) cells to arrest in the G0/G1 phase. This cell cycle arrest was associated with a marked decrease in the expression of E2F1 (14), in a similar manner to our present findings in GLA exposed C6 glioma cells.
GLA increased the expression of bax 21 kDa and 50 kDa, and the bax/bcl-2 ratio was altered demonstrating the higher levels of bax in relation to bcl-2 levels after treatment. However, at 150 μM, GLA did not increase apoptotic rates in C6 cells in vitro, Fig. 1 and (9). Although these results suggest that bax would stimulate the apoptotic response, other proteins also form heterodimers with bax and can thus modulate or inhibit the apoptotic response. In neuroblastoma cells, bax associates with cytoplasmic Ku70, blocking apoptosis in response to certain chemotherapeutic agents (15, 16). Ku70 can also function as a physiological inhibitor of bax-induced apoptosis (17).
A single report showed that Ku proteins can bind to the E2F motif thereby regulating the cell cycle response to DNA damage and other insults (18). High levels of the DNA repair proteins Ku70 and Ku80 are related to hyperproliferation and carcinogenesis, and the heterodimeric Ku70/Ku80 protein complex is important for DNA repair in resistant tumor cells (16). Ku70/Ku80 heterodimers represent the regulatory subunit of DNA-dependent protein kinase (DNA-PKc), which plays an important role during the repair of DNA double strand breaks. Ku proteins have been implicated in the maintenance of proliferating cell nuclear antigen (PCNA) on chromatin following ionizing radiation (19).
In this study, GLA caused a 71% decrease in Ku80 protein expression. However, this was not accompanied by decreased mRNA expression and may point to an increase in Ku80 proteolysis in GLA-treated cells. The decreased Ku80 levels in GLA-treated cells could subsequently lead to increased Ku70 protein availability for interaction with other proteins such as bax, which could thus block the apoptotic response to GLA (20, 21). The oxidative stress caused by GLA in tumor cells (3, 9) may be involved in reduced Ku80 protein levels, as oxidative stress was able to induce nuclear loss of Ku70 and Ku80 in pancreatic acinar AR42J cells (22).
Taken together, we may speculate that this mechanism is at least partly responsible for the nonapoptotic response caused by 150 μM GLA in these cells. A recent study has reported that reducing Ku80 to 10% of original levels by RNAi caused impaired DNA replication and activated a DNA replication checkpoint blocking cell division in G1 (23). The RNAi study did not, however, induce increased apoptosis but led to a decrease in S-phase and caused a decrease in PCNA expression. An important finding in this study was the significant increase in micronucleus formation in the GLA-treated cells, and this is indicative of altered DNA replication. The reduced Ku80 concentration could be directly related to impaired DNA replication resulting in cell cycle changes and micronucleus formation. A recent study reported that shRNA inhibition of Ku80 expression caused decreased cell proliferation and increased sensitivity to both radio and chemotherapy (24). In addition, Ku knockout mice are hypersensitive to ionizing radiation (16). In cells with Ku80 knockout, irradiation leads to increased micronucleus formation versus Ku80 positive cells (25). A recent study has shown that the cyclin-dependent kinase inhibitor roscovitine also induces reduced S-phase, inhibition of Ku80 expression, and to a lesser degree Ku70 expression, and then increases A549 nonsmall cell lung cancer sensitivity to radiation (10). To our knowledge, this is the first report of polyunsaturated fatty acids altering Ku protein expression, and this finding is of importance as it may explain one of the reasons why GLA is able to increase the sensitivity of tumor cells to both radiotherapy and chemotherapy (7). Further studies are underway to determine the mechanism by which GLA or one of its metabolites alters the metabolism and function of the Ku proteins in glioma cells.
This research was funded by the Brazilian research foundations FAPESP, CNPq, and CAPES. The authors thank the collaboration of Thais Martins de Lima and Rui Curi in the use of the FACS system.