The Toxicologist - Society of Toxicology

1011 IN VITRO GENE EXPRESSION PROFILING TO
PREDICT SKIN SENSITIZING POTENTIAL.
J. van der Veen 1, 2 , J. Ezendam 1 , T. Pronk 1, 2 , R. Vandebriel 1 , J. Pennings 1 and
H. van Loveren 1, 2 . 1 Laboratory for Health Protection Research, National Institute for
Public Health and the Environment, Bilthoven, Netherlands and 2 Department of
Health Risk Analysis and Toxicology, Maastricht University, Maastricht, Netherlands.
The upcoming ban on animal testing in the Cosmetics Directive and costliness of
the high number of animals required under REACH, have prompted a high demand
for the development of alternative methods to replace the currently used animal
models for skin sensitization. One promising in vitro system is the use of gene
expression profiling in human keratinocytes (KC), which are key players in the initiation
of skin sensitization. In a previous study using the human keratinocyte cell
line HaCaT the in vitro effects of sensitizers on gene expression profiles were evaluated.
That study showed that a set of 13 genes was able to classify sensitizers with
70% accuracy. Pathway analysis showed that pathways related to oxidative stress
were significantly upregulated by sensitizers. To improve classification and to further
explore the relevance of this pathway the experiment was extended with additional
chemicals. HaCaT cells were exposed to 18 sensitizers and 8 irritants in a
concentration leading to 80% viability after 24 hours exposure. Gene profiling was
used to classify sensitizers and find regulated pathways. Classification using the
Random Forest algorithm in a leave-one-compound-out cross validation scheme
resulted in a gene list that was able to classify skin sensitizers with 96% accuracy.
The only incorrectly classified compound was the irritant sodium lauryl sulphate,
which is a false-positive compound in the murine LLNA. Interestingly, it proved
possible to correctly classify the false-positive compound xylene and the false-negative
compound nickel. Pathway analysis was performed using ToxProfiler. Pathways
related to the antioxidant response and the innate immune response were identified.
In conclusion, KC gene profiling can be applied to identify contact sensitizers
and may thus be a useful in vitro model to predict skin sensitization potential of
chemicals.
1012 COMPARATIVE INVESTIGATION OF PHOTOTOXICITY
OF CHEMICAL COMPOUNDS BY USING THE
STANDARD BALB 3T3 NRU PHOTOTOXICITY TEST
AND A THREE-DIMENSIONAL HUMAN SKIN MODEL
(EST 1000).
A. Heppenheimer and A. H. Poth. Genetic and Alternative Toxicology, Harlan
Cytotest Cell Research GmbH, Rossdorf, Germany.
Where substances are intended for use in personal care products applied to the skin
an assessment of potential phototoxic hazard is required. The initial test is the
measurement of a UV/visible absorption spectrum to identify absorption at relevant
wavelength, followed by in an in vitro assay for phototoxicity, the Balb 3T3
neutral red uptake phototoxicity test (OECD 432). However, this test has its limitations,
as non-hydrosoluble chemicals can be tested only at low concentrations due
to their lack of aqueous solubility and also many complex mixtures or formulations
cannot be tested. Consequently, it does not take into account the bioavailability of
test chemicals topically applied to skin and in many cases e.g. after having a phototoxic
effect in this test system, such information is required. To overcome these limitations,
the use of reconstructed skin models is an interesting alternative and a useful
follow-up test. In the present study the EST 1000 skin model (Cell Systems,
Germany) was used. Chloropromazine and sodium dodecyl sulphate were used as
reference compounds. 6-methylcoumarin (6-MC) and Rose Bengal were selected as
model compounds. 6-MC is known to possess a photoxic potential in vivo but it
does not show effects in vitro in the Balb 3T3 test, whereas Rose Bengal is known
to show no in vivo photoxicity, but effects in the Balb 3T3 test. Both compounds
were comparatively evaluated in the the Balb 3T3 neutral red uptake phototoxicity
test and the EST 1000 skin model. In contrast to the Balb 3T3 model, the human
skin equivalents were able to predict the phototoxic potential of both compounds.
The obtained data indicate that the EST 1000 model is a useful model for the prediction
of phototoxicity and also useful as a follow-up test.
1013 FUNCTIONAL IN VITRO ASSAYS OF DIFFERENT CELL
SOURCES ARE ABLE TO PREDICT SPECIFIC
IMMUNOSUPPRESSIVE PROPERTIES OF CHEMICAL
COMPOUNDS.
A. Fischer, L. Koeper and H. Vohr. Toxicology, Bayer Schering Pharmacology
Wuppertal, Germany. Sponsor: H. Ellinger-Ziegelbauer.
An increasing aim in safety assessment of chemicals and drugs is to reduce, refine
and replace animal testing. Great efforts were and still are made to evaluate potential
adverse effects of compounds on the immune system in vitro. Regarding im-
216 SOT 2011 ANNUAL MEETING
munosuppression, most methods are based on mitogen stimulation assays, although
it has been demonstrated that a functional assay is one of the most sensitive
indicators for primary immunotoxic effects in vivo.
The Mishell-Dutton culture (MD), a well known method to investigate immunosuppression,
represents a functional assay that requires the interaction of different
cell types to generate an immune response in vitro. To our knowledge the test has
never been considered as an in vitro alternative to the existing animal tests.
In a previous study we were able to demonstrate the ability of MD assays with
mouse spleen cells to correctly predict different immunosuppressant compounds 1 .
To expand its potential, we modified the Mishell-Dutton culture and are now able
to measure the in vitro antibody response of peripheral blood mononuclear cells
(PBMC) as well as spleen cells of different species.
With this in vitro assay we are able to demonstrate immunosuppressive effects of
test items and to clearly discriminate between specific immunosuppression and
non-specific cytotoxicity. Preliminary data show an excellent concordance between
species as well as different cell sources.
1 L.-M. Koeper and H.-W. Vohr, 2009. Functional assays are mandatory for a correct
prediction of immunotoxic properties of compounds in vitro. Food and
Chemical Toxicology 47, 110-118
1014 PRECISION-CUT LIVER SLICES AS AN EX VIVO
MODEL TO STUDY IDIOSYNCRATIC
HEPATOTOXICITY IN MOUSE AND HUMAN.
M. Hadi 1 , Y. Chen 1 , M. Stitzinger 2 , H. Emmen 2 and G. Groothuis 1 .
1 Pharmacokinetics, Toxicology & Targeting, University of Groningen, Groningen,
Netherlands and 2 Toxicology, NOTOX BV, s’-Hertogenbosch, Netherlands. Sponsor:
A. Vickers.
Idiosyncratic drug reactions (IDRs) are adverse drug reactions that are rare, sporadic,
unpredicted by clinical trials, unrelated to drugs’ pharmacology and occur
without relation to time or dose. IDRs may arise from drug interaction with inflammation
that render the liver more sensitive to injury resulting in increased toxicity.
With the aim to develop a translational model to unravel the mechanism behind
IDRs and to find biomarkers that can detect them, we used mouse (M) and
human (H) precision-cut liver slices (PCLS) to study the influence of inflammatory
reactions on the toxicity of drugs. PCLS technology is receiving increased attention
as a potential ex vivo toxicological model because PCLS retain the normal tissue architecture
of an intact liver with all its cell types in their natural environment.
PCLS from M and H were incubated with paracetamol (APAP), diclofenac (DF),
ketoconazole (KT) or clozapine (CZ) alone and in the presence of lipopolysaccharide
(LPS), an inflammation inducer. Cell viability [ATP] and cytokine production
[CBA] were assessed. APAP, DF, and CZ (but not KT) were more toxic in M than
H. LPS had no influence on APAP and DF toxicity in M or H. APAP and DF
slightly increased LPS-induced TNF-α production in M but strongly reduced it in
H PCLS and both decreased the LPS-induced IL-1β production in both species. In
contrast, toxicity of KT (only in M) and CZ (in both M and H) was aggravated in
the presence of a non-toxic dose of LPS and an increase in LPS-induced IL-1β production
was observed here. In conclusion, clear species differences were identified
in the effect of the drugs on toxicity and on inflammatory reactions due to LPS.
Toxicity of KT (only in M) and CZ (in both M and H) was aggravated by LPS, not
observed in the cases of APAP or DF. A correlation appeared between increased toxicity
and increased LPS-induced IL-1β production in the co-incubation of LPS and
drug. Based on these results, PCLS appear to be a promising ex vivo model to study
mechanisms behind IDRs.
1015 TRANSPORT OF CHLORPROMAZINE ACROSS CACO-2
CELLS AND MEASURING THE FREE
CONCENTRATION BY ND-SPME.
J. Broeders, B. Blaauboer and J. Hermens. Institute for Risk Assessment Sciences,
Utrecht University, Utrecht, Netherlands.
Animal experiments are used during drug development to determine the pharmacokinetic
parameters of a compound. The use of in vitro and in silico techniques
could replace some of these in vivo tests. However, to improve the predictability of
these in vitro tests, it is also necessary to fully understand the kinetic behavior of the
compound in the in vitro system. Therefore, the aim of this project, part of the
European project Predict-IV, is to improve our understanding of the kinetics of
compounds in in vitro test systems.
The kinetic behavior of chlorpromazine (CPZ) in the Caco-2 cell system was studied.
Bi-directional transport was measured and the influence of bovine serum albumin
(BSA) added basolaterally was determined. A negligible depletion-solid phase
microextraction (nd-SPME) method was developed to measure the free concentration
of CPZ in the medium. This concentration was used to correct the Caco-2
permeability results. All samples were analyzed by HPLV-UV.