4
Abstracts / Toxicology 290 (2011) 1–46
P005 A genomic approach to overcoming temozolomide resistance in glioblastoma multiforme
a novel mechanism of MGMT regulation. This work demonstrates that transcription patterns can be used to establish drug–target relationships, and techniques such as connectivity mapping could provide powerful drug discovery tools.
Joshua L. Smalley 1,∗ , Sarah K. Smalley 2 , Shu-Dong Zhang 3 , Simon J. Morley 2 , Timothy W. Gant 1 1
MRC Toxicology Unit, University of Leicester, LE1 9HN, United Kingdom 2 School of Life Sciences, University of Sussex, BN1 9QG, United Kingdom 3 Centre for Cancer Research and Cell Biology, QUB, BT9 7BL, United Kingdom E-mail address:
[email protected] (J.L. Smalley). Chemotherapeutic treatment of cancer is often rendered inactive by resistance that is either intrinsic or acquired. Glioblastoma multiforme (GBM) is an aggressive primary brain cancer treated with the DNA methylating agent temozolomide (TMZ). Acquired resistance of GBM to TMZ is mediated by increased levels of the DNA repair enzyme O-6-methylguanine-DNA methyltransferase (MGMT) (Zhang et al., 2010). MGMT translation is controlled by the mTOR pathway, implicating mTOR inhibitors as possible TMZ sensitizers (Caporali et al., 2008). Global gene expression analysis has been used to predict the individual therapeutic response to GBM, and can differentiate between TMZ sensitive and resistant cells using transcriptional patterns (Mischel et al., 2004). This work takes that concept a step further, by using transcription patterns for TMZ resistant GBM cells to predict chemicals that enhance TMZ sensitivity. A TMZ resistant transcription pattern was generated from a publically available microarray dataset (ArrayExpress Acc No. EGEOD-7696) and used to query an in-house optimized connectivity map - a transcription pattern matching program that utilizes a database containing the microarray data of >1400 chemical treatment instances in cancer cell lines (Zhang and Gant, 2009; Smalley et al., 2010). Connectivity mapping identified 50 chemicals that produce inverse transcription patterns to that of TMZ resistance. Using a transcription pattern for the mTOR inhibitor KU0063794, we used connectivity mapping to mine our results for chemicals also displaying mTOR inhibition transcription patterns (Fig. 1). The connectivity map identified two chemicals already known to enhance TMZ sensitivity: the PI3Kinase inhibitor LY294002 and thalidomide. Resveratrol was identified by both the TMZ resistance query and the mTOR inhibitor query. In the presence of TMZ, resveratrol displayed mTOR inhibitor characteristics – reducing 4EBP1 phosphorylation and MGMT levels. Compounds A, B and C were identified by the TMZ resistance query alone. They reduce MGMT levels, but have little effect on 4EBP1, potentially indicating
Fig. 1. Using transcription patterns to describe TMZ resistance and mTOR inhibition, connectivity mapping identified chemicals that down-regulate MGMT protein, which is up-regulated in TMZ resistance. Eukaryotic translation initiation factor 4Ebinding protein 1 (4EBP1) is regulated by mTOR and controls MGMT translation. 4EBP1 protein was measured to assess mTOR activity.
Reference Caporali, S., Levati, L., Starace, G., Ragone, G., Bonmassar, E., Alvino, E., D’Atri, S., 2008. Mol. Pharmacol. 74, 173–183. Mischel, P.S., Cloughesy, T.F., Nelson, S.F., 2004. Nat. Rev. Neurosci. 5, 782–792. Smalley, J.L., Gant, T.W., Zhang, S.D., 2010. Toxicology 268, 143–146. Zhang, J., Stevens, M.F., Laughton, C.A., Madhusudan, S., Bradshaw, T.D., 2010. Oncology 78, 103–114. Zhang, S.D., Gant, T.W., 2009. BMC Bioinformatics 10, 236.
doi:10.1016/j.tox.2011.09.013 P006 Dio3 mRNA is translationally down-regulated with druginduced liver damage Kate M. Dudek 1,∗ , Emma L. Marczylo 1 , Laura Suter-Dick 2 , Anne E. Willis 1 , Timothy W. Gant 1 1
MRC Toxicology Unit, University of Leicester, LE1 9HN, United Kingdom 2 InnoMed PredTox Consortium, F. Hoffmann-La Roche, Ltd., Basel, Switzerland E-mail address:
[email protected] (K.M. Dudek). Regulation of gene expression at the level of mRNA translation is important in cellular responses to toxic stress, particularly in vivo. However, this genomic response to toxicity has been much less explored than the transcriptional response. Here we used genomic methods to determine if any mRNA transcripts were being altered translationally in response to drug induced liver injury. For analysis we used samples from the EU-FP6 PredTox study. In this study 14 compounds that had failed in development due to hepato- or nephro-toxicity were investigated to find gene expression profiles that were indicative of, but preceding, the onset of the toxicity. The aim was to use these profiles in other studies to enhance pathological assessment and provide earlier warning of toxicity. Each compound was delivered to male Wistar rats (n = 5) once per day orally, at a low or high dose with matched vehicle-treated controls, for up to 15 days. Sampling took place at 2, 4 and 15 days. Drugs producing the most severe hepato-toxic effects were selected for mRNA translational analysis. Sucrose density fractionation, combined with microarray analysis and qRT-PCR verification, was used to identify mRNAs showing altered translational regulation with or without transcriptional alteration. Changes in mRNA translational activity were confirmed using Western blotting. From these data we have identified Dio3 mRNA as translationally altered in a linear manner with increasing liver damage following treatment with several of the compounds, one of which is shown in Fig. 1. Dio3 is involved in the bioavailability of thyroid hormone in developing tissue (Hernandez, 2005). Originally thought to be solely expressed in the foetus, more recent work indicates reexpression in certain cases in adult tissue. This includes work by Kester et al. (2009) who showed an induction in expression of Dio3 in the regenerating liver of mice and rats who had undergone partial hepatectomy. We believe the translational change may be mediated by RNA binding proteins as these favour a CG rich environment, such as that observed in the 5 flanking region of the Dio3 gene (Hernandez, 2005). Work is currently underway on developing an in vitro model to enable further understanding of the mechanism behind the reduction in Dio3 protein expression.