For glioblastomas, COX-2 expression is linked to poor survival. COX-2 effects are mediated by the receptors EP2 and EP4, whose regulation is poorly understood. The expression of EP4, and activation or inhibition of EP4 activity in human glioblastoma T98G cells, was found to correlate with growth on soft agar. Chemoprevention drugs, troglitazone (TGZ) and some COX inhibitors, significantly suppressed EP4 expression in T98G cells in a dose dependant manner. Specificity protein 1 (Sp-1) binding sites, located within region -197 to -160 of the human EP4 promoter, are important for the transcription initiation of the human EP4 gene and are responsible for the EP4 suppression by TGZ. Mutation in the Sp-1 sites altered the promoter activity of luciferase constructs and TGZ effects on the promoter. The inhibitory effect of TGZ on EP4 expression was reversed by PD98059, a MEK-1/Erk inhibitor. Immunoprecipitation-Western blot analysis detected Sp-1 phosphorylation that was dependent on TGZ-induced Erks activation. ChIP assay confirmed that Sp-1 phosphorylation decreases its binding to DNA and as a result, leads to the suppression of EP4 expression. Thus, we propose that the expression of EP4 is regulated by Sp-1, but phosphorylation of Sp-1 induced by TGZ suppresses this expression. This represents a new and unique mechanism for the regulation of the EP4 receptor expression.

The expression of nonsteroidal anti-inflammatory drug-activated gene-1 (NAG-1) is regulated by the p53 and Egr-1 tumor suppressor pathways. Many anti-cancer drugs and chemicals induce NAG-1 expression, but the mechanisms are not fully understood. Transgenic mice expressing human NAG-1 are resistant to intestinal and prostate cancer, suggesting that NAG-1 is a tumor suppressor. Proteasome inhibitors exhibit anti-glioblastoma activities in preclinical studies. Here, we show that the proteasome inhibitors MG132 and bortezomib induced NAG-1 expression and secretion in glioblastoma cells. MG132 increased NAG-1 expression through transcriptional and post-transcriptional mechanisms. At the transcriptional level, the induction of NAG-1 required the −133 to +41 bp region of the promoter. At post-transcriptional levels, MG132 stabilized NAG-1 mRNA by increasing the half-life from 1.5 h to > 8 h. Because of the dramatic increase in mRNA stability, this is likely the major contributor to MG132-mediated NAG-1 induction. Further probing into the mechanism revealed that MG132 increased phosphorylation of the p38 MAPK pathway. Consequently, inhibiting p38 phosphorylation blocked activation of the NAG-1 promoter and decreased mRNA stability, indicating that p38 MAPK activation mediates both MG132-dependent promoter activation and mRNA stabilization of NAG-1. We propose that the induction of NAG-1 by p38 MAPK is a potential contributor to the anti-glioblastoma activity of proteasome inhibitors.

Background The Delphi consensus statements on the management of germ cell tumors (GCTs) failed to reach agreements on the statement that the cases with 1) pineal and neurohypophyseal bifocal lesion, 2) with diabetes insipidus, and 3) with negative tumor markers can be diagnosed as germinoma without histological verification. To answer this, multicenter retrospective analysis was performed. Methods A questionnaire on clinical findings, histological diagnosis, and details of surgical procedures was sent to 86 neurosurgical and 35 pediatrics departments in Japan. Results Fifty-one institutes reported 132 cases that fulfilled the three criteria. Tissue sampling was performed in 91 cases from pineal (n = 44), neurohypophyseal (n = 32), both (n = 6) and distant (n = 9) lesions. Histological diagnosis was established in 89 cases: pure germinoma or germinoma with syncytiotrophoblastic giant cells in 82 (92.1%) cases, germinoma and mature teratoma in two cases, and granulomatous inflammation in two cases. Histological diagnosis was not established in two cases. Although no tumors other than GCTs were identified, three (3.4%) patients had non-germinomatous GCTs (NGGCTs). None of the patients developed permanent complications after endoscopic or stereotactic biopsy. Thirty-nine patients underwent simultaneous procedure for acute hydrocephalus without permanent complications, and hydrocephalus was controlled in 94.9% of them. Conclusion All patients who fulfilled the three criteria had GCTs or granulomatous inflammation, but not other types of tumors. However, no less than 3.4% of the patients had NGGCTs. Considering the safety and the effects of simultaneous procedures for acute hydrocephalus, biopsy was recommended in such patients.

EP4 expression in human glioblastoma cells correlates with growth on soft agar. The cyclooxygenase inhibitor sulindac sulfide first altered specificity protein-1 (Sp-1) and early growth response gene-1 expression, then increased the expression of nonsteroidal antiinflammatory drug-activated gene 1 and activating transcription factor 3, and then decreased EP4 expression. EP4 suppression was dependent on blocking the Sp-1 binding sites in the human EP4 promoter. Mutation in the Sp-1 sites in EP4 altered the promoter activity and abolished sulindac sulfide effects. The inhibitory effect of sulindac sulfide on EP4 expression was reversed by PD98059, a mitogen-activated protein/extracellular signal-regulated kinase kinase-1/extracellular signal-regulated kinase inhibitor. Sp-1 phosphorylation was dependent on sulindac sulfide-induced Erk activation. Chromatin immunoprecipitation assay confirmed that Sp-1 phosphorylation decreases Sp-1 binding to DNA and leads to the suppression of EP4. Inhibition of cell growth on soft agar assay was found to be a highly complex process and seems to require not only the inhibition of cyclooxygenase activity but also increased expression of nonsteroidal anti-inflammatory drug-activated gene 1 and activating transcription factor 3 and suppression of EP4 expression. Our data suggest that the suppression of EP4 expression by sulindac sulfide represents a new mechanism for understanding the tumor suppressor activity.