ENDOCRINE DISRUPTORS: MODELING THE INTRACELLULAR RESPONSE
Michael Breen, Rory Conolly
National Center for Computational Toxicology, U.S. EPA, Research Triangle Park, NC, USA
R16-14
ABSTRACT
Scientists have identified alterations in the concentration dynamics of specific hormones as risk factors for common cancers such as breast cancer (estrogen, progesterone), endometrial cancer (estrogen), and prostate cancer (estrogen, testosterone). These adverse hormonal changes in the tightly regulated endocrine pathways can be induced from exposure to exogenous endocrine disruptors. Chemicals capable of acting as endocrine disruptors are ubiquitous with environmental sources that include household detergents, pesticides, plastics, pharmaceutical estrogens, industrial chemicals, and byproducts of incineration, paper production, and fuel combustion. Ecological exposures to endocrine disruptors are primarily from industrial and waste water treatment effluents, while human exposures are mainly through the food chain. The adverse effects induced by exposure to endocrine disruptors can be mediated through alterations in the enzymes involved in steroid synthesis. We are developing a mechanistic mathematical model of the intratesticular and intraovarian metabolic network that mediates steroid synthesis to describe the dose-response for endocrine disruptors, and to identify and link new robust molecular biomarkers of exposure that are indicative of the ultimate adverse effects. The deterministic model describes the biosynthetic pathways for the conversion of cholesterol to the sex steroid hormones (estradiol, testosterone, and 11-ketotestosterone) secreted by the testes in fish. The model includes the intermediate metabolites and enzymatic reactions for the multiple pathways involved in the biosynthesis of the sex steroids. Changes in the concentration dynamics of the secreted hormones are used as an index of the endocrine disruption. The initial concentrations and enzyme kinetic reaction rates were taken from the literature or set to biologically reasonable values. This mechanistic model allows for an improved understanding of the source-to-outcome linkages and dynamic dose-response behavior at the molecular level for effective use of biomarkers for risk assessments with endocrine disruptors, including their possible effects on endocrine-induced cancers. Since the biosynthetic pathways for the sex steroids are evolutionarily conserved to a significant extent, this model is likely to also be relevant for mammalian species.
LINKING BIOMARKERS OF EXPOSURE TO EFFECTS
Molecular Biological Effects
Receptor-ligand interaction, DNA binding, enzyme activity
COMPUTATIONAL MODEL
Intratesticular Steroidogenic Pathway
v1
CHOL
EDC EXPOSURES
• • • • •
Exposure of male and female fathead minnows to EE2 (synthetic estrogen): high ecological/regulatory relevance Dose levels: 0 (control), 10, 100 mg/L Dosing phase: 8 days Recovery phase: 8 days Tissue sampling: day 1, 4, 8, and 16
Cellular
Altered signaling, gene activation, protein synthesis
Organ
Altered physiology and tissue morphology
Individual
Impaired development and reproduction, cancer, death
Population
Structure, Extinction
v2 P450c17 (hydroxylase) 3ßHSD v5
PROG PREG
P450scc
17 PREG
v3 P450c17 (lyase) v6
DHA
v4 17ßHSD v7
DIOL
v15 v8 v14 v11
11-T
v16
KT
11ßHSD
Biomarkers
mRNA, protein, enzyme levels
Metabolite profiles
Functional and structural change (pathology)
Altered reproduction or development
Decreased number of animals
v9
17 PROG
v10
DIONE
P45011ß
v17
T
v18 P450arom v12 v13 v19
19-T E2
v20
ESTRO NE
Computational model Small fish model
Systems biology models
Fathead Minnow Partially characterized genome High ecological/regulatory relevance Molecular markers, metabolomics
Deterministic Model
d CHOL = −v1 dt
d PROG = v9 − v5 dt
d 17 PROG = v 6 + v9 − v10 dt d DIONE = v10 + v7 − v11 − v12 dt
d ESTRONE = v12 − v13 dt
Small fish exposure system
Fathead minnows
d E 2 = v13 + v19 − v 20 dt
PARAMETER ESTIMATION
Objective function:
where:
d PREG = v1 − v 2 − v5 dt
HYPOTHALAMIC-PITUITARY-GONADAL (HPG) AXIS
HPG Axis
Arterial blood
Conceptual Systems Model
Dopamine GABA taurine
5HTR Y2 R A B D1 R
Brain, GnRH neuronal system
NPY GnRH D2 R
Venous blood
d 17 PREG = v 2 − v3 − v 6 dt d DHA = v3 − v 4 − v7 dt
d 19T = v18 − v19 dt d 11T = v14 − v15 dt
f = ∑∑ [Si ,n − Si (t n ,θ )]
I N i =1 n =1
2
I
N
= number of species (metabolites) = number of time samples = concentration of species (metabolite) = adjustable model parameters
d KT = v15 − v16 dt
Hypothalamus
E2 T KT
5HT
d DIOL = v 4 − v8 dt
d T = v11 + v8 − v14 − v17 − v18 dt
θ
• •
S
Apply an iterative optimization algorithm Simultaneously estimate parameters for all dose concentrations
GnRH Negative Feedback
Inhibin Activin Pituitary, gonadotrophs Activin
GnRH Follistatin PACAP NPY
Enzyme Kinetics
GnRH R
MODEL SIMPLIFICATION
LH FSH
Anterior Pituitary
Activin R
PACAP R LHβ GPα
Y1 R FSHβ
D2 R
GnRH R
LH, FSH Gonads (Ovaries, Testes)
VTG
Outer mitochondrial membrane
Preg
3β HSD Prog
P450c17 (lyase) Dione P450 arom estradiol E2
E+S + I
Ki
Km
ES
E+P
v (reaction rate)
Vmax
Vmax 2
Increasing inhibition
Rate-limiting reaction
Michaelis-Menten model
P450scc
Inner mitochondrial membrane
Competitive inhibition
P450c17 (hydroxylase) 17-prog
17βHSD
EI
• •
Km
Motivation
StAR
E2, T, KT
Cholesterol 21-hydroxylase 11-deoxycortisol 20βHSD 20βS
testosterone
T
• • • •
More intuitive understanding of dynamic functional behavior Reduces number of parameters Identify rate limiting step(s): quasi-steady state approximations Identify preferred pathways
P45011β
Method
Steroid hormone responsive tissues (e.g. liver, gonads)
Feedback control system of the HPG axis that regulates synthesis and secretion of primary steroid hormones (estradiol (E2), testosterone (T), and 11-ketotestoterone (KT, only in male fish)) by the release of gonadotropin releasing hormone (GnRH) from the hypothalamus, and luteinizing hormone (LH) and follicle stimulating hormone (FSH) from the pituitary.
Inhibin Activin
LDL HDL LH FSH
LDL R HDL R LH R FSH R Generalized gonad
11-testosterone 20βHSD 17α,20β-P (MIS) 11βHSD ketotestosterone
S (substrate conc.)
ACKNOWLEDGMENTS
KT
Mathematical Model
NHEERL, U.S. EPA, Duluth, MN Gerald Ankley, PhD Dan Villeneuve, PhD
E2
ERα,β1,β2
VTG ZRP
Liver
VTG ZRP
v=
Vmax S S + α Km
I α = 1+ Ki
DISCLAIMER
This work was reviewed by the U.S. EPA and approved for publication but does not necessarily reflect Agency policy.
Conceptual systems model shows key regulatory components of HPG axis. Green and red arrows indicate activation and negative feedback (inhibition), respectively. White boxes indicate proteins and peptides. Small molecules (e.g. steroids and neurotransmitters) are shown in italics.
3 parameters: V max , K m , K i