CONTRIBUTION OF EXPERIMENTAL AND INTER AND INTRASPECIES VARIABILITY IN A COMPUTATIONAL MODEL FOR ETH

1
CONTRIBUTION OF EXPERIMENTAL AND INTER AND
INTRASPECIES VARIABILITY IN A COMPUTATIONAL MODEL
FOR ETHANOL-INDUCED PERTURBATIONS OF NEOCORTICAL
DEVELOPMENT

• J.M. Gohlke, W.C. Griffith, E.M. Faustman
• Institute for Risk Analysis and Risk
  Communication and Department of Environmental
  Health, University of Washington, Seattle,
  Washington, USA

OBJECTIVE To quantitate inter and intraspecies
toxicodynamic variability in the neocortex during
pre and postnatal development in the rat, mouse
and primate brain we analyzed
Interexperimental variation (See Section
I.) Growth Fraction (sensitivity analysis
based on 7 key studies) Clearance time
(sensitivity analysis based on 5 key
studies) Intraspecies variation (See Section
II.) Interspecies variation (See Section II.
and III.)
MODEL FOR NEOCORTICAL DEVELOPMENT
A.
Figure 1. Mechanism based framework for
evaluation of neocortical development. The
anatomy of neocortical neurogenesis, migration,
apoptosis, and differentiation and how it relates
to model building. A. Basic model framework
from Leroux (1996), which is color coded to
illustrate model for neocortical development
where Type X cells represent neuronal progenitor
cells in the ventricular zone and Type Y cells,
represent postmitotic neurons. B. During
neurogenesis (G13-G19 in rat) progenitor cells
are generated in the ventricular zone. During G1
newly generated cells either stay in the
proliferative population or become postmitotic
and begin migration through the intermediate zone
(IZ) to the cortical plate (CP). Differentiation
along with ubiquitous apoptosis of the new
neuronal population takes place postnatally
(P0-P14) in the rat neocortex.

• EXAMPLE ANALYSES OF INTER-EXPERIMENTAL
  VARIABILITY IN KEY PARAMETERS
• A. Clearance Time

B.
Figure 5. Comparison of estimated maximum
biological variability from our model to
variability reported in studies used for
parameter estimation

• Intraspecies Variability Key Points
• The experimental design may falsely increase
  variability estimates because the animals have to
  be sacrificed and variability in time of
  conception, therefore intraindividual and
  intralitter correlations in parameters through
  time are unknown.
• Results suggest experimental error can account
  for the majority of the observed variability in
  parameter estimation data, highlighting the
  importance of reduced experimental variability to
  discover subtle toxicant-induced effects.

• II. ANALYSIS OF INTRASPECIES AND INTERSPECIES
  VARIABILITY
• Intraspecies observed variation experimental
  variation biological variation
• We used the maximum variability in
  stereological studies in our model to provide an
  estimate of true biological variability in model
  parameters, assuming biological variability in
  parameters in independent of variability in X and
  Y cell number.
• Figure 5 shows comparison of model output using
  maximum variability reported in stereology
  studies to observed variability reported in
  studies used for parameter estimation.

Figure 2. Analysis of the impact of experimental
variation in clearance time estimates for TUNEL
and pyknotic nuclei in neocortical synaptogenesis
model. We varied estimations of death rates
during the postnatal differentiation period based
on differing estimations of clearance times in
the rat model using the Thomaidou et al. (1997)
dataset for the rat model (A.) and using the
Verney et al. (2001) dataset for the mouse
model(B.). Clearance times shown are 1 hr.
1,2, 2.5 hrs. 3, and 4 hrs. 4,5.
Stereological estimates ( SEM) of total neuronal
number are plotted for comparison (?).
B. Growth Fraction
III. ETHANOL-INDUCED INHIBITION OF NEOCORTICAL
DEVELOPMENT INTERSPECIES COMPARISONS
Figure 7. Interspecies comparisons of
dose-response relationships in final neuronal
output based on lengthening of cell cycle in the
rat neocortex after exposure to ethanol during
neurogenesis. A Weibull d-r function was used
(rate untreated baseline e (drpdose2) where
drp is the time-dependent dose-response
parameter) and 95 population intervals
estimating intraspeices variability based on
stereological studies are shown as dotted lines.

• Key Point
• Interspecies variability in application of rat
  toxicity data to the primate neurogenesis model
  indicates primates may be more sensitive to
  ethanol-induced neocortical neuronal loss during
  neurogenesis at BECs above 100 mg/dl based on the
  steeper slope of neocortical neurogenesis during
  primate development.

Figure 3. Analysis of experimental variation for
the growth fraction (GF) parameter in our
neocortical neurogenesis model. Analyses was
performed to examine various growth fraction (GF)
estimates 80 in mouse 6 and rat 7, 93 in
mouse 8and rat 9, 97 in mouse10 or 100 in
mouse11 of cells actively cycling or a
time-dependent GF with 100 at beginning of
neurogenesis and falling to 80 at end of
neurogenesis in mouse 12. A. Analysis of the
growth fraction (GF) parameter in rat model using
the Thomaidou et al. (1997) dataset for
synaptogenesis model for rat B. Analysis of
the growth fraction (GF) parameter in mouse model
using the Verney et al. (2000) dataset for
synaptogenesis model. Stereological estimates
(SEM) of total neuronal number are plotted for
comparison (?).
IV. CONCLUSIONS

• We have developed stochastic models for mouse,
  rat and primate neocortical neurogenesis allowing
  for direct analyses of contributions of
  variability in toxicodynamic processes.
• Experiments used for parameter estimation
  report large amounts of variability compared with
  stereology studies used for validation of model
  construct, suggesting experimental error may mask
  toxicant induced effects at this level.
• Interspecies variability between rodent and
  monkey is large compared with intraspecies and
  interexperimental variability, while the opposite
  is true when comparing rat with mouse.
• Our model predictions indicate dynamic
  differences in normal neocortical neurogenesis
  which may enhance the susceptibility of the
  primate brain to neuron loss after exposure to
  ethanol during neocortical neurogenesis.

• Key Points
• Interexperimental variability can be evaluated
  within our model construct to inform hypotheses
  regarding the relative contributions of key
  parameters driving neocortical neurogenesis.
• Our sensitivity analysis demonstrated that
  higher clearance times of dead cells (4 hrs.)
  and a high growth fraction (100) agree well with
  independent stereological data in the rat, while
  the opposite is true in our mouse model.

• Interspecies Variability Key Point
• The monkey neocortex is estimated to contain
  400 times as many neurons as the rat and the
  mouse neocortex is estimated to contain 1.23
  times more neurons than the rat neocortex.

Acknowledgements This study was supported by the
Center for Child Environmental Health Risks
research through EPA grant R826886 and NIEHS
1PO1ES09601, NIEHS P30ES07033 (CEEH Center),
T32ES07032 (EPT Training Grant), and the ARCS
Foundation.