Search

The most common cancer therapies including chemo- and radiation therapy are designed to induce cell death by DNA damage. However, DNA damage response (DDR) repair pathways initiated by tumor cells can resist the effect of anticancer drugs thus providing the rationale for combining DNA repair pathway inhibitors with radiation and DNA damaging anticancer drugs. The reliance of different cancer to different DDR mechanism can guide for targeted use of specific DDR inhibitor to maximize tumor cell killing. Here, we show the differential activities of two classes of DDR pathway inhibitors, Ataxia Telangiectasia- and Rad3-related kinase (ATRi) and poly(ADP-ribose) polymerases inhibitors (PARPi) across 20 glioma stem cell (GSC) lines. Using a panel of 3 ATRi, we identified a subgroup of GSCs, showing significant synergism with TMZ combination but resistant to ATRi single agent monotherapy. Interestingly, majority of the synergistic cell lines are associated with MGMT promoter methylation, suggesting that only tumor cells with no MGMT repair will benefit from ATR inhibition. Moreover, we also observe that PARPi + TMZ showed synergism only in MGMT unmethylated GSC cells suggesting that this synergism is a result of the dual function of PARPi in blocking BER repair and blocking MGMT PARylation thus abolishing MGMT mediated repair. Altogether, we believe MGMT methylation status in primary GBM could function as a strong biomarker for patient selection for ATRi or PARPi combination with TMZ in the clinic. We are currently exploring the activity of the triple combination of TMZ + PARPi + ATRi for the small subgroup of GSCs that do not show response to either TMZ + ATRi or PARPi + TMZ combinations.

Glioblastoma (GBM) is an aggressive brain tumor and has an extremely poor prognosis despite the use of multiple treatment modalities. DNA damage response (DDR) signaling plays an important role in inducing radiation and temozolomide (TMZ) resistance and hence has emerged as a molecular target for therapeutic development. The Ataxia Telangiectasia and Rad3-related protein (ATR) kinase is a key regulator of the DDR machinery, activated by DNA damage. Here, we show that three clinical-grade ATR inhibitors (Bay1895344, AZD6738 and Berzosertib) had similar selective sensitivity pattern across 16 glioma-like stem cell (GSC) lines tested. ATR inhibitors inhibited the growth of GSCs at low nanomolar range concentrations. Interestingly, all three ATR inhibitors showed a significant synergism with TMZ in a selective group of GSCs (Combination index and Bliss model). Importantly, we demonstrate that MGMT promoter methylation status was associated with cellular response to combination treatment with preferential inhibition of cell growth in MGMT promotor methylated (MGMT deficient) cell lines. Further, we compared the RNA-seq data from GSCs in the synergism and non-synergism group and used multiple complementary approaches to identify the response marker that confer sensitivity to combination therapy. Our preliminary data analysis identified several genes that confer sensitivity to combination treatment and studies are underway to validate the data. We also investigated the efficacy of BBB penetrant ATR inhibitor BAY 1895344 in orthotropic xenografts and administration of BAY 1895344 and TMZ combination significantly reduced tumor size and extended survival in an intracranial animal model. Combining ATR inhibitor with TMZ was well tolerated and did not confer additional toxicity as the body weights of TMZ and combination groups were comparable. The findings from this study provides a rationale to use ATR inhibitors in combination with TMZ in MGMT methylated tumors to improve therapeutic efficacy of standard of care for GBM patients.

Temozolomide (TMZ) is the standard chemotherapy for malignant gliomas (MG), and resistance to this drug is mediated by the DNA repair protein O6-methylguanine-DNA methyltransferase (MGMT). Epigenetic silencing of the MGMT gene by promoter methylation in about 40% patients is associated with loss of MGMT expression that compromises DNA repair leading to favorable response to TMZ therapy. Understanding the mechanism of MGMT mediated repair and modulating MGMT activity will enhance TMZ activity in MGMT unmethylated MG. Here we report a novel mode of regulation of MGMT protein activity by Poly-ADP-ribose polymerase (PARP). We found that PAPR physically interacts with MGMT and PARylates MGMT in response to TMZ treatment. We further showed PARylation of MGMT by PAPR is required for MGMT binding to DNA and to remove O6-methylguanine adducts in damaged DNA induced by TMZ. All 4 PARP inhibitors (trapping and non-trapping) tested (Talazoparib, Pamiparib, Veliparib, Olaparib) can inhibit PARP-MGMT binding, PARylation of MGMT, binding to DNA and subsequent removal of O6 - lesions in damaged DNA. We showed combination of PARP inhibitor with TMZ potentiated TMZ cytotoxicity in both MGMT methylated and unmethylated Glioma stem cell lines, but more profoundly in unmethylated group in vitro and in vivo. PARP inhibition acted as a double-edged sword in MGMT unmethylated MG: blocking BER/SSBR pathway to repair TMZ induced N7-MetG and O3-MetA, and more importantly, suppressing PARP-mediated PARylation of MGMT and thus silencing MGMT activity to repair O6-MetG, resulting in augmented cytotoxicity. This is the first study to show that PARylation of MGMT by PARP is required for repairing TMZ-induced O6-methylguanine adducts, and inhibition of MGMT PARylation abolishes MGMT function and renders sensitization to TMZ treatment. This finding provides a rationale for combining TMZ/CCNU and PARP inhibitors in MGMT unmethylated MG patients to enhance the benefit of adjuvant chemotherapy.

The lack of knowledge about the relationship between tumor genotypes and therapeutic responses remains one of the most important gaps in enabling the development of effective targeted cancer therapies. Increasing evidence supports the notion that DNA damage response (DDR) signaling plays an important role in inducing radiation and temozolomide (TMZ) resistance in glioblastoma (GBM) and hence, it has emerged as a molecular target for therapeutic development. The cellular response to TMZ therapy is critically mediated by MGMT and DDR signaling networks that are regulated by Ataxia Telangiectasia Mutated (ATM) and Ataxia Telangiectasia and Rad3-related protein (ATR). ATR is a primary sensor of DNA damage initiating subsequent repair leading to cell survival. Here, we show that three clinical grade ATR inhibitors (Bay1895344, AZD6738 and VE-822) had similar selective sensitivity pattern across eight glioma-like stem cells (GSC) lines tested with three (GSC28, GSC11, GSC274) being the most sensitive, and three (GSC300, GSC272, and GSC20) being the most resistant cell lines. Interestingly, all three ATR inhibitors showed a significant synergism with TMZ (Combination index and Bliss model) in GSC300, GSC272, and GSC20, and antagonism in GSC28, GSC11, and GSC274. Further analysis showed that the synergistic group was resistant to single agent ATR inhibitor, whereas antagonistic group was inherently sensitive to ATR inhibitor. The synergistic response was not associated with any major genetic alterations such as MGMT status, EGFR amplification, PTEN deletion/mutation, or p53 deletion/mutation. However, GSEA analysis of baseline RNA-seq data of the 6 cell lines showed an enhanced expression of six DNA damage/repair/response pathways in the antagonistic group. It suggests that a low level of DNA damage/repair/response pathways could be a marker of synergy for TMZ with ATR inhibitors in GBM. Further studies are underway to validate this observation in more GSC lines in vitro and in vivo.

Deregulated uptake of glucose and amino acids is one of the hallmarks of cancer metabolism. Clinical trials and numerous drugs have been designed to target cancer cell metabolism, however, the effect on cell growth inhibition and how tumor cell metabolism couples with cell growth remain unclear. In this study, we report the glioblastoma stem cells (GSCs) growth dependence on nutrients (glucose, amino acid, and fatty acid) and oxygen. We show that the basal metabolic activities are unable to reflect the extent of proliferation inhibition by metabolic inhibitors. 2-DG (20mM) downregulated extracellular acidification rate (ECAR) more effectively in GSCs with higher glycolytic activity (Pearson r=0.71, p=0.08), the basal glycolytic activity, however, was not significantly correlated with glycolysis inhibitor (2-Deoxy-D-glucose, 2-DG) sensitivity (Pearson r=0.35, p=0.44), indicating basal glycolytic activity cannot predict GSC response to glycolysis inhibitor. Similarly, basal oxidative phosphorylation (OXPHOS) activity also failed to predict GSC responses to OXPHOS inhibition. Although GSCs showed the consistent susceptibility against oligomycin (complex V inhibitor) and hypoxia (Pearson r=0.61, p=0.15), there is no significant correlation between basal oxygen consumption rate (OCR) and oligomycin/hypoxia-induced proliferation inhibition. Thus, both basal glycolytic and basal OXPHOS activities were unable to explain their roles in GSC growth. Studies are underway to investigate the effect of fatty acid and amino acid on tumor growth. As drugs targeting cancer metabolic processes always show dose-limiting toxicity, our study will focus on identifying new “druggable” targets according to GSC nutrient dependences and developing specific therapeutic strategy for GBM populations with certain nutrient dependence.

Hypoxia-inducible factor 1 (HIF-1), a master heterodimeric transcriptional regulator consisting of HIF-1alpha and HIF-1beta subunits for cellular response to hypoxia, plays an important role in carcinogenesis, while CCAAT/enhancer-binding protein alpha (C/EBPalpha) is proposed to act as a tumor suppressor in C/EBPalpha-expressing tissues. Previously, we reported that ectopically expressed HIF-1alpha protein interacts with and enhances transcriptional activity of C/EBPalpha, which favors leukemic cell differentiation. Here we further showed that such an interaction also occurred in their endogenously expressing state of leukemic U937 cells. Glutathione S-transferase pull-down assay proposed that the protein-protein interaction was direct, and transactivation domains of C/EBPalpha and the basic helix-loop-helix domain of HIF-1alpha were essential for such an interaction. More intriguingly, we provided the first demonstration that C/EBPalpha competed with HIF-1beta for direct binding to HIF-1alpha protein. Correspondingly, C/EBPalpha overexpression significantly inhibited the DNA-binding ability of HIF-1 and expressions of hypoxia-responsive element-driven luciferase and HIF-1-targeted genes vascular endothelial growth factor, glucose transporter-1 and phosphoglycerate kinase 1. In parallel, suppression of C/EBPalpha expression by specific small hairpin RNA increased DNA-binding ability of HIF-1 and expression of these HIF-1-targeted genes in leukemic U937 cells. These results would provide new insights for antitumor potential of C/EBPalpha protein.

BACKGROUND: Glioblastoma represents a compelling disease for signal transduction targeted therapy. Despite the evidence defining the PI3K network as a high quality target, regulatory feedback loops as well as cross talk with other networks make targeting the PI3K network a challenging task. Complete control of GBM is dependent on the elucidation of the escape mechanisms and the rational design of combination therapies. A major objective in our laboratory is to develop robust predictive molecular and genetic markers/signature that will lead to selection precision therapy. These predictive markers will allow us to determine rational combinatorial approaches to target aberrations in the PI3K network and at the same time prevent deleterious effects of feedback loops and cross talk. METHODS: We have utilized a large panel of glioma stem cell (GSC) lines to perform drug sensitivity/resistance screens. Reponses profiles of GSCs using multiple PI3K inhibitors in cytotoxicity assay were generated. Molecular targets/pathways contributing to alternate signaling pathway activation by inhibitors were identified using gene expression profiling and protein expression/activity profiling. Top gene targets were subjected to loss of function studies using RNAi. Synergistic effect of knocking down the target along with PI3K inhibitors on cell proliferation, and tumorigenicity were assessed. RESULTS: We have utilized several system approaches to identify targets in compensatory or parallel pathways after PI3K inhibition. Our results suggest that different molecular profiles in GSC dictate the differences in viability observed after drug treatment, despite a common blockade of PI3K pathway. Resistant and sensitive cell lines were profiled using the Affymetrix and RPPA platforms to identify genes that are differentially expressed in sensitive and resistant GSCs responsible for intrinsic resistance. Using gene expression profiling, we also identified adaptive response gene targets in pre- and post-treatment resistant GSC. This approach has uncovered alternate pathways and shutting down these pathways would synergize the effect on cell growth when combined with PI3K pathway inhibitors. We have selected the top 10 candidate genes and two targets are being studied for rational combination with PI3K inhibitors CONCLUSIONS: Failure of PI3K pathway inhibitors is due to escape through overlapping or parallel signaling pathways, and that complete control of GBM depend on the elucidation of these escape mechanisms and we show that therapies to target essential pathways overcome or circumvent treatment resistance and can provide the next generation of improved therapeutic options. SECONDARY CATEGORY: Tumor Biology.

Glioblastoma, with aberrant activation of growth factor signaling pathways, represents a compelling disease for kinase inhibitor therapy. Despite the evidence defining the PI3K network as a high quality target, regulatory feedback loops as well as cross-talk with other networks make understanding and targeting the PI3K network a challenging task. As a result of selective pressure by PI3 kinase inhibitors, glioblastoma: 1) up-regulates compensatory survival pathways and 2) utilizes complimentary pathways as a predominant mechanism of escape from anti-PI3K drugs, which can be eliminated by concomitant use of complimentary inhibitors. Here, we tested these hypotheses, using a large panel of glioma cells and glioma initiating stem cell (GIC) lines treated with PI3K inhibitors. A differential response to PI3K pathway inhibitors divided GICs into responder and non-responder groups, identifying molecular profiles that distinguish the two groups. Gene expression profiling before and after PI3K inhibitor treatment was analyzed by Affymetrix microarrays to identify molecular targets/signatures contributing resistance to PI3K inhibitors. We identified mitogen- and stress-activated protein kinase 1 (MSK1) was markedly induced after PI3K/mTOR inhibitor treatment. MSK1 depletion attenuated GIC resistance to PI3K/mTOR inhibitors, and overexpression of MSK1 potentiated the resistance to PI3K/mTOR inhibitors. Further investigation identified β-catenin as a direct interacting protein of MSK1. Interestingly, MSK1 phosphorylated β-catenin at ser552 and regulated β-catenin nuclear translocation and transcriptional activity. More interestingly, we showed that overexpression of S552D-β-catenin reversed MSK1-shRNA-attenuated resistance to PI3K/mTOR inhibition. Disruption of MSK1 synergized with PI3K/mTOR inhibitors to extend survival in an intracranial animal model, accompanied by decreased phosphorylation of β-catenin at Ser552. Taken together, our results suggested that PI3K/mTOR inhibitors activated MSK1/β-catenin signaling, which may serve as an escape survival signaling mechanism upon PI3K/mTOR inhibition in glioblastoma. Therefore targeting PI3K/mTOR pathway together with MSK-1 might show increased therapeutic benefit when used as combination regimen.