The Toxicologist - Society of Toxicology

this class of drug would have few side effects because angiogenesis does not occur in
adult, but an unanticipated toxicity has been hypertension (HTN) and proteinuria
in 10 – 80% of patients receiving these drugs. It is now appreciated that these toxicities
reflect inhibition of homeostatic VEGF signaling. In vasculature, VSP inhibition
reduces VEGF-dependent nitric oxide levels, causing systemic vasoconstriction
and HTN. In the kidney, podocyte-derived VEGF maintains glomerular endothelial
cell health, and VSP inhibition induces glomerular endothelial damage, thrombotic
microangiopathy and proteinuria. Intriguingly, patients developing grade 3 or
4 HTN experience superior anti-tumor efficacy, suggesting that development of
HTN might be a useful pharmacodynamic marker of effective VSP inhibition. Two
lessons can be drawn: First, in this era of targeted therapies, predicting toxicity from
a specific and highly potent drug requires a thorough understanding of the drug
target’s normal role in homeostasis. Rodent models allowing conditional gene inactivation
in the adult will be increasingly required to predict these kinds of toxicities
when knowledge of a signaling pathway’s role in normal physiology is incomplete,
and several examples will be discussed. Second, because newer agents inhibit fewer
targets, toxicities must be classified as either mechanism-dependent, or mechanismindependent
(“off-target”). Mechanism-dependent toxicities may in fact represent
novel surrogate biomarkers of signal inhibition efficacy, such as HTN or proteinuria,
to which drugs can be titrated in individual patients. Such a personalized dosing
strategy may define the optimal therapeutic index for individuals, and perhaps
help predict which patients are more likely to respond to a given agent.
1444 LEAD OPTIMIZATION STRATEGIES AND NOVEL
BIOMARKER TECHNOLOGIES TO ENSURE DRUG
SAFETY.
J. S. Ozer. Pharmacokinetics, Dynamics and Metabolism, Pfizer, Chesterfield, MO.
Sponsor: S. Ramaiah.
Because of the increased awareness and regulatory emphasis on drug safety issues
during clinical trials and product post-marketing use, there is a need for newer biomarker
technologies and approaches to better quantify and validate the most sensitive
and specific safety biomarkers. Integration of multiple platform technologies
including LC/MS/MS, LTQ-Orbitrap, ELISA, and Luminex is used to analytically
and biologically validate safety biomarkers throughout drug development. Limiting
antibody reagents across preclinical species for a particular assay can be overcome by
employing MS technologies strategically. TFF3 is one safety biomarker for renal
(urine) and GI (serum) injury where multiple platform technologies have been employed
to analytically validate biomarker values. These approaches and technologies
can be applied during clinical trials and are feasible for translational use, whereas
antibody technology shows limitations in cross-validations across species.
Lead optimization approaches are employed to utilize safety biomarkers early in the
pipeline using samples from discovery PK studies. Lead optimization allows chemistry
support to be maximized in the pipeline, while screening for known toxicity liabilities.
Prodromal renal functional markers of early organ toxicity are presented in
a lead optimization format that is earlier than traditional histopathology endpoints.
The renal model utilizes bile acids, leukotriene and proximal tuble enzymatic activity.
It is critical that the biomarkers developed for use lead optimization are highly
applicable to several preclinical models.
1445 TOXICITY TESTING IN THE 21ST CENTURY FOR
ECOTOXICOLOGY.
S. Edwards 1 and G. T. Ankley 2 . 1 ORD/NHEERL, U.S. EPA, Research Triangle
Park, NC and 2 ORD/NHEERL, U.S. EPA, Duluth, MN.
The National Research Council (NRC) report, Toxicity Testing in the Twenty-first
Century: A Vision and a Strategy has relevance for ecological as well as human
health risk assessment. In April 2009, the Society of Environmental Toxicology and
Chemistry (SETAC) held a workshop that considered key elements of the scientific
foundation that would be needed to implement the vision of toxicity pathwaybased
testing in support of ecological risk assessment The term adverse outcome
pathway (AOP) was used to describe the linkage of molecular events modeled in a
toxicity pathway assay to downstream biologic effects considered adverse from an
ecotoxicological perspective (i.e., effects on survival, reproduction). Five challenges
related to the elucidation and description of AOPs were considered. First, consistent
with the NRC strategy concerning toxicity pathway elucidation and linkage to
adversity, the challenge of describing AOPs from the extant literature and quantitatively
modeling key components was addressed. Second, approaches for reverse engineering
AOPs from combinations of transcriptomic, proteomic, metabolomic,
and/or phenotypic data were examined. Because adversity in an ecological risk context
is typically considered at the population level, approaches for translating toxicity
pathway outputs into appropriate parameters for population modeling was a
third challenge discussed. The fourth related to the challenge of discriminating
adaptive (i.e., homeostatic, allostatic) responses from adverse ones and incorporating
that knowledge into AOP models. Finally, because species extrapolation is a
central challenge in ecological risk assessment, the workshop examined how to determine
conservation of AOPs among species and use this information in predicting
species sensitivity to support ecological risk assessments. This session will summarize
the results of the SETAC effort and invite discussion with SOT members
regarding development of an integrated toxicity testing paradigm that supports
both human health and ecological risk assessment.
1446 ADVERSE OUTCOME PATHWAY (AOP) MODELING OF
KNOWN PATHWAYS.
M. E. Andersen 1 , K. Watanabe 2 and I. R. Schultz 3 . 1 The Hamner Institutes,
Research Triangle Park, NC, 2 Oregon Health and Science University School of
Medicine, Portland, OR and 3 Battelle Pacific Northwest, Sequim, WA.
A nine-member workgroup representing disciplines of neurotoxicology, wildlife biology,
ecotoxicology, and engineering developed a case study for domoic acid, an
algal toxin with adverse effects on both wildlife and humans. The case study
demonstrated strategies to develop computational models of known AOPs in an effort
to move beyond the current toxicity testing paradigm focused on chemical-specific
toxicity to a focus on toxicity pathway perturbations applicable for ecological
risk assessment. Domoic acid is a potent agonist for kainate receptors - ionotropic
glutamate receptors whose activation leads to the influx of sodium and calcium.
Increasing intracellular calcium concentrations lead to toxicity and cell death. The
development of a conceptual framework for the AOP required an iterative process
with two important outcomes: (1) a critically reviewed pathway from exposure to
adverse outcome that is stressor specific and (2) identification of a key cellular
process (or processes) suitable for evaluation in a mechanistic assay in vitro. The
toxicity pathway indicated that in vitro cellular assays of altered neuronal calcium
should serve as a measure of the key response and that the results of these assays
would be amenable to mechanistic modeling for identifying perturbations (and domoic
acid treatments) that are within normal, those that are adaptive, and those
that are clearly toxic. In vitro assays with outputs that are also amenable to measurement
in exposed populations would link in vitro to in vivo conditions.
Toxicokinetic information with domoic acid also aids in linking in vitro results to
the individual. Another required step is taking projected responses of individuals
and creating population models. All these linkages were considered in group considerations
for the specific domoic acid example, leading to a series of more generic
recommendations about strategies for literature mining and model development for
known pathways.
1447 REVERSE ENGINEERING ADVERSE OUTCOME
PATHWAYS FROM ‘’OMICS DATA.
E. J. Perkins 1 , K. Chipman 2 , F. Falciani 2 , S. Edwards 3 , T. Habib 4 , R. Taylor 5 ,
G. Van Aggelen 6 , C. Vulpe 7 and N. V. Garcia-Reyero 8 . 1 Environmental
Laboratory, U.S. Army ERDC, Vicksburg, MS, 2
College of Life and Environmental
Sciences, University of Birmingham, Birmingham, United Kingdom, 3
NHRL, U.S.
EPA, Research Triangle Park, NC, 4
Biological Sciences, University of Southern
Mississippi, Hattiesburg, MS, 5
Systems Biology, Pacific Northwest National
Laboratories, Richland, WA, 6
Pacific Environmental Science Center, Environment
Canada, Vancouver, BC, Canada, 7
Nutritional Science and Toxicology, University of
California at Berkeley, Berkeley, CA and 8 Chemistry, Jackson State University,
Jackson, MS.
While many toxicologically-relevant pathways are known, many responses are mediated
via unknown, or poorly characterized, mechanisms and modes of action.
The advent of global analysis tools provides new capabilities for probing entire biological
systems. Complex interaction networks can be reverse engineered or inferred
from gene, protein, metabolic, signaling data enabling exploration of potential
modes or mechanisms of toxic action. We describe reverse engineering of interaction
networks from genes, proteins, and metabolites altered by toxicants to identify
adverse outcome pathways in ecotoxicology. We then examined the utility of this
for deducing toxicologically-relevant networks that regulate response to a defined
stressor and dictate outcomes. A large data set of 868 arrays was used that focused
on expression changes in fathead minnow ovary tissue representing exposure to 7
different chemicals, over different times, and in vivo versus in vitro conditions. By
applying different approaches, we demonstrate how this data set can be used to
infer gene regulatory networks. The network path from stressor to adverse outcome
can be considered a candidate adverse outcome pathway. Identification of candidate
adverse outcome pathways allows for the formation of testable hypotheses about
the key biologic processes, biomarkers or alternative endpoints, which could be
used to monitor an adverse outcome pathway. Finally, we identify the unique challenges
facing the application of this approach in this field, and attempt to provide a
road map for the utilization of these tools
306 SOT 2010 ANNUAL MEETING