MULTIPLEX ASSAY OF IN VITRO CYTOKINE SECRETION AS A USEFUL PREDICTOR OF IN VIVO TOXICITY OF CHEMICALS: THE A-CUTE-TOX PROJECT AND THE EUROPEAN REACH POLICY

MARIA SOLEDAD ALVAREZ; SABRINA DEVOS; RON KOOIJMAN; ENRIQUE O´CONNOR

Budapest

Congreso; XXIV International Congress Cytometry in the Age of Systems Biology; 2008

International Society for Analytical Cytology

The different patterns of cytokine correspond with different functions as immune effectors. it has been proposed that in vivo cytokine production is a sensitive indicator of immunotoxicity chemicals. The European Commission is fostering research on in vivo methods to replace anial studies in the novel policy for registry, Evaluation and Classificationof chemicals (REACH policy). We have applied both ELISA and flow cytometry multiplex assay of 11 cytokines (Flow Cytomix,Bender Med systems) using a 96 well flow cytometer (Cytomics FC500 MPL Beckman-Coulter) to assay the immunotoxicity of 60chemical compound of reference in the context of the EuropeanProject "A-cute-Tox", aimed to design a testing strategy for the REACH policy. The effect of test chemicals and dexamethasome as a positive control has been examined on in vitro cytokine production by human PBCM stimulated with 10 ug/ml PHA and in the presence of every test compound. We obtained IC50 or EC50values in 58 chemicals for cytokine production by human PBMC.The multiplexed assay for 11 cytokine secretion is a promising tool for a testing strategy. INF-G, IL-5 and TNF-a exhibit in general the best separate correlations with other toxicity parameters. However, centralizing descriptors like the average toxicity or higher toxicity values for the whole set of cytokines are predictive parameters performing better than each single cytokine, especially if two poorly performing cytokine, IL-2 and IL-4 are withdrawn of the multiplexed assay. When the few chemicals identified as outliers in the correlation of IC50 of Il-5 measured by IC-50 by ELISA are excluded form the Flow Cytomics set of data, excellent coefficient of correlations with human in vivo data are obtained for sets of 45-52 test chemicals (R values of 0.7-0.73) for average and minimal IC50/Ec50 values in monocyte secreted cytokines or in the whole set excluding IL-2 and IL-4. These correlations are in the same range as rodent vs human toxicity and their calculation do not require the use of animals. In addition, xenobiotic-induced suppression of TNF-a secretion is strongly correlated to general in vitro cytotoxicity (Neutral redUpdate assay. Soponsored by Ec project LSHB-CT-2004-512051.A Cute Tox Project)