431

Malignant glioma represents the most common form of primary human brain cancer. Despite therapeutic advances, it remains amongst the deadliest of cancers. The current standard of care for malignant gliomas consists of surgical resection followed by Temozolomide (TMZ)/Radiation therapy. We reasoned that the development of agents that exhibit simultaneous tumoricidal activity, TMZ chemo-sensitization, and radiation sensitization should offer a favorable strategy in glioma therapy.
 To this end, we undertook a SiRNA screen-based strategy. We screened for gene silencings that inhibited the Fanconi Anemia (FA) DNA repair pathway - a critical determinant of TMZ resistance. Simultaneously, a screen was carried out to identify SiRNAs with tumoricidal activity against the U87 glioma line. A third screen was carried out to identify SiRNAs with radiation sensitizing activity using the same cell line. Since radiation and TMZ exert their effects by DNA damage, we screened a commercial library (QIAGEN) consisting of 712 SiRNAs against 356 genes implicated in DNA repair/damage response.
 We searched for SiRNAs that scored strongly positive in all three screens and identified PSMA1, a critical component of the proteasome complex. To exclude “off-target” effects, we recapitulated the findings with pharmacologic inhibitors of proteasome function, including Bortezomib, MG132, Lactacystin, and ALLN. All inhibitors produced tumor killing at micromolar/sub-micromolar concentrations. Further, at sub-lethal doses, all compounds significantly sensitized (>10 fold) the U87 line to TMZ and radiation. This sensitization was not affected by the over-expression of a dominant negative mutant of P53, Epidermal Growth Factor Receptor Variant III (EGFRvIII), or O6-Methyl-Guanine Methyl Transferase (MGMT) - three well-studied determinants of glioma resistance to TMZ/radiation. To exclude cell-line specific effects, the TMZ/radiation sensitizing effects of Bortezomib was recapitulated using the T98g, A172, U343, and U373 glioma lines as well as the glioma stem cell line BT74. In vivo intracranial U87 xenograft model also recapitulated the in vitro synergy between TMZ and Bortezomib.
 Given the distinct modes of DNA damage by TMZ and radiation, we hypothesized that the chemo-sensitizing effects of proteasome inhibition may extend to other DNA damaging agents. Indeed, proteasome inhibition enhanced the tumoricidal activity of Cisplatin in breast and ovarian cell lines as well as 5 FU in colon cancer cell lines. The mechanism of sensitization appeared multi-factorial, involving FA pathway inactivation as well as G2/M checkpoint modulation.
 In sum, we offer pre-clinical data supporting proteasome inhibition as a strategy for augmenting the current regimen of malignant glioma treatment. Moreover, the broad-chemo-sensitizing effects of proteasome inhibition suggest application in a diverse set of tumor therapies.

99th AACR Annual Meeting-- Apr 12-16, 2008; San Diego, CA