Abstracts / Toxicology Letters 238S (2015) S56–S383 S233 P10-017 Volatile biomarkers of cisplatin-induced toxicity in vitro R. Fijten ∗ , A. Smolinska, A. Boots, J. Dallinga, F.-J. van Schooten Maastricht University, Pharmacology and Toxicology, Maastricht, Netherlands Cisplatin is an anti-cancer drug used in a wide variety of can- cers including bladder, ovarian, testicular and non-small cell lung cancer (NSCLC). A major issue occurring in cancer treatment is drug resistance. Cisplatin resistance can occur through mechanisms such as increased efflux or influx of the drug or by alterations in the expression of proteins involved in pathways essential for cisplatin- induced damage. Since cisplatin resistance is an increasing clinical problem, there is an arising need to identify patients who are or will become resistant to cisplatin chemotherapy. Exhaled air con- tains metabolites called Volatile Organic Compounds (VOCs) that can be used as a non-invasive diagnostic and monitoring tool. VOCs in exhaled air are indicators of metabolic changes and malfunctions occurring throughout the human body. Therefore, it can be antici- pated that cisplatin resistance will induce a different exhaled VOC profile than is present in the breath of non-resistant patients. As a first attempt, we studied VOCs released by the NSCLC cell line A549 after exposure to a cisplatin concentration of 50 M (GI50) and 400 M (cell viability < 10%). Random Forest (RF) was used to identify VOCs released by the cells in response to these cisplatin treatments. At 50 M, 9 VOCs could be identified as only excreted by the cells in response to the exposure. At 400 M, this was the case for 19 VOCs. Finally, only 4 VOCs were found to be statistically different between the two studied cisplatin concentrations. These results indicate that specific VOCs are secreted in the headspace of cells as a response to cisplatin. Currently, chemical identification of these VOCs is underway. After chemical identification, VOCs can be linked to cellular mechanisms specific to the damaging effects of cisplatin or the response mechanisms of the cell toward this expo- sure. This feasibility study is the first step in searching for VOCs specific for cisplatin treatment, and later on, possible resistance in lung cancer patients. http://dx.doi.org/10.1016/j.toxlet.2015.08.687 P10-018 Transcriptome profiling reveals allergy associated gene expression in human bronchial epithelial cells and dendritic cells exposed to diesel exhaust particulate chemicals S. Macchiarulo, T. Gant, M. Leonard ∗ Public Health England, Toxicology, Chilton, Didcot, United Kingdom Chemical pollutants are often associated with fuel combustion and particulate matter generated from diesel engines. The impacts these chemicals have as a consequence of inhalation exposure include effects on respiratory conditions such as asthma. Bronchial epithelial and dendritic cells in the airways represent some of the more proximal sentinels of inhaled material. As they have a promi- nent role in the development of inflammatory responses in allergic asthma it was the primary aim of this work to characterise whether diesel exhaust particulate chemical extracts (DEPE) alter allergy rel- evant gene expression patterns in these cells. We profiled in vitro cultures of human primary bronchial epithelial cells (HPBECs), from 5 individual donors, exposed to DEPE (0.5–20 g/ml resid- ual mass) for global mRNA expression changes using an RNA-Seq poly(A) library based sequencing method (Illumina HiSeq 2000). Selected immune related gene expression changes were further characterised for corresponding protein levels released into the cell culture media and revealed increased expression of Il-1alpha and TGF-alpha among others. The TLR4 ligand lipopolysaccharide (LPS) is a co-factor for allergic responses and was observed to further enhance IL-1alpha levels in DEPE treated HPBECs. Signals such as Il-1alpha originating from epithelial cells can direct den- dritic cell function. We therefore used conditioned medium from HPBECs exposed to DEPE and LPS to treat dendritic cell differen- tiated human MUTZ-3 cells and examined gene expression using RNA-Seq profiling. CCR7 expression induced by LPS conditioned media was further enhanced when cells were treated with media from HPBECs, exposed to both DEPE + LPS in combination. DEPE alone did not elicit any changes in CCR7 expression from MUTZ-3 cells. As CCR7 is an important chemokine receptor for dendritic cell trafficking to lymph nodes, increased levels of expression, altered by signals derived from DEPE treated bronchial epithelial cells, point towards specific mechanisms through which diesel exhaust material may alter allergic responses upon inhalation. http://dx.doi.org/10.1016/j.toxlet.2015.08.688 P10-019 Exploratory urinary metabolomic profiling of renal cell carcinoma using 1 H NMR spectroscopy and multivariate analysis M. Monteiro 1,∗ , A. Barros 2 , M. Carvalho 3,1 , A.S. Pires-Luís 4,5 , C. Jerónimo 4,5 , R. Henrique 4,5,6 , B. Maria de Lourdes 1 , A. Gil 7 , P. Guedes de Pinho 1 1 Faculty of Pharmacy, UCIBIO@REQUIMTE, Laboratory of Toxicology, Porto, Portugal 2 University of Aveiro, Department of Chemistry, Aveiro, Portugal 3 University Fernando Pessoa, Faculty of Health Sciences, Porto, Portugal 4 Portuguese Oncology Institute-Porto, Cancer Biology & Epigenetics Group, Porto, Portugal 5 Biomedical Sciences Institute (ICBAS), Department of Pathology and Molecular Immunology, Porto, Portugal 6 Portuguese Oncology Institute-Porto, Department of Pathology, Porto, Portugal 7 University of Aveiro, CICECO-Instituto de Materiais de Aveiro (CICECO/UA), Aveiro, Portugal A 1 H NMR-based metabolomics method was applied to first void urine samples from healthy controls (n = 48) and renal cell carcinoma (RCC) patients (n = 52) to assess its predictive ability for the detection of cancer induced metabolic changes. After raw data processing, principal component analysis (PCA) and partial least-squares discriminant analysis (PLS-DA) were applied. A vari- able selection procedure was used in tandem with PLS-DA in order to extract information from the NMR data mainly related to the discrimination between control and case. As a result, the number of variables (spectral signals) was reduced by ca. 30%, improving model robustness and interpretability. Monte Carlo Cross Valida- tion (MCCV) and receiver operating characteristic (ROC) analysis of the reduced and original data sets showed improved predic- tive power [(Q 2 (median), 0.59 vs 0.35] and higher classification rate, sensitivity and specificity (respectively, 85.3%, 84.2%, and 86.5%, compared to 81.4%, 78.9%, and 84.1%), thus expressing satisfactory robustness. In the particular case of renal cancer, urine composition was observed to carry important impacts of