COMPUTATIONAL STEROIDOGENESIS MODEL TO PREDICT BIOCHEMICAL RESPONSE TO ENDOCRINE ACTIVE CHEMICALS: MODEL DEVELOPMENT AND CROSS VALIDATION Miyuki Breen1,2, Michael S. Breen3, Natsuko Terasaki4, Makoto Yamazaki4, Rory B. Conolly1 1National Center for Computational Toxicology, U.S. EPA, Research Triangle Park, NC, USA, 2Biomathematics Program, Department of Statistics, North Carolina State University, Raleigh, NC, USA, 3National Exposure Research Laboratory, U.S. EPA, Research Triangle Park, NC, USA, 4Molecular Toxicology Group, Safety Research Laboratory, Mitsubishi Tanabe Pharma Corporation, Chiba, Japan

ABSTRACT CELL PROLIFERATION EXPERIMENTS Dynamic Mass Balances: Metabolic Pathway MODEL PARAMETERS Steroids, which have an important role in a wide range of physiological processes, are d synthesized primarily in the gonads and adrenal glands through a series of Incubate cells in medium for 72 hr Estimated parameters with data from metyrapone study • VCcellxxxx ,cell V med C ,med P ,cell U ,cell • enzyme-mediated reactions. The activity of steroidogenic enzymes can be altered by a • Incubate cells in new medium + stimuli + carrier for 72 hr dt variety of endocrine active chemicals (EAC), some of which are therapeutics and others that are environmental contaminants. We are developing a dynamic mathematical model • Collect samples at 0, 72, 96, 120, 144 hr Vcell  volume of viable cells of the metabolic network of adrenal steroidogenesis to predict the synthesis and • Measure number of viable cells using cell viability analyzer V  volume of medium secretion of adrenocortical steroids (e.g. mineralocorticoids, glucocorticoids, androgens med and estrogens), and their biochemical response to EAC. We previously developed a Pxx,cell  production rate of steroid in cells deterministic model which describes the biosynthetic pathways for the conversion of CELL PROLIFERATION MODEL dN kt Uxx,cell  utilization rate of steroid in cells cholesterol to adrenocortical steroids, and the kinetics for enzyme inhibition by the EAC, kN N Nep metyrapone. In this study, we extended our model for a multiple enzyme inhibiter, dt p 0 aminoglutethimide. Experiments were performed using the H295R human adrenocortical Competitive Enzyme Inhibition Equation cells, and concentrations of 12 steroids were simultaneously measured with a newly N  number of viable cells developed LC/MS/MS method. We performed cross validation of our model for the baseline data across multiple experimental studies. Results show that the model k p  growth rate ki simulation closely corresponds to the time-course baseline data. Our study vi  demonstrates the feasibility of using the in silico mechanistic computational model to N0  initial number of viable cells CEAC,cell predict the in vitro adrenocortical steroid concentrations using H296R cells. This 1 capability could be useful to help define mechanisms of actions for poorly characterized Estimated:k p  0.00878 kEAC,i chemicals and mixtures in support of the H295R steroidogenesis screening system, and • Direct least-squares solution by to screen drug candidates based on steroidogenic effects in the early phase of drug vii  overall metabolic reaction rate for reaction development. linearization kii  first-order metabolic reaction rate for reaction C  concentration of EAC in cells EFFECTS OF ENDOCRINE ACTIVE EAC,cell STEROIDOGENESIS EXPERIMENTS ki enzyme inhibition constant of EAC for reaction CHEMICALS ON HPA AXIS • Incubate cells in medium for 72 hr EAC,i Start clock: Incubate cells in new medium + stimuli + carrier for 72 hr • Metyrapone Model CROSS VALIDATION Metyrapone Study • Validated model with baseline data from aminoglutethimide • Baseline and two concentrations (1, 10 μM) • 17 first-order study • Collect samples at 0, 8, 24, 48, 72 hr metabolic • Measure cell and medium concentrations of E1 and E2 using reaction rates ELISA and remaining 12 steroids using LC/MS/MS • 14 steroid Aminoglutethimide Study equilibrium constants • Baseline and one concentration (100 μM) • Collect samples at 24, 48, 72 hr • First-order • Measure cell and medium concentrations of 12 steroids except E1 CHOL uptake and E2 using LC/MS/MS rate • MET partition STEROIDOGENESIS MODEL coefficient • 2 enzyme Mathematical model based on in vitro experimental design • inhibition • Two compartments: medium and H295R cells constants • Two pathways: transport and metabolism • First-order metabolic reaction rates • Assumed quasi-equilibrium for steroid transport Aminoglutethimide Model - rapid & reversible steroid transport between medium and cells • First-order cholesterol transport rate • 17 first-order Feedback control system of hypothalamus-pituitary-adrenal (HPA) axis regulates synthesis and secretion of adrenocortical steroids (aldosterone (Aldo), cortisol (Cort), • Endocrine active chemicals: partition coefficient for transport, metabolic testosterone (T), estradiol (E2)) by release of corticotropin releasing hormone (CRH) competitive enzyme inhibition reaction rates from hypothalamus, and adrenocorticotropic hormone (ACTH) from pituitary • 14 steroid Equilibrium Equations: Transport Pathway equilibrium H295R CELL LINE constants CqCxxx,med ,cell DISCUSSION • Established from human female • First-order • Dynamic behavior of model simulation and baseline data in medium qxx  equilibrium constant for steroid adrenocortical carcinoma CHOL uptake correspond across both studies Cxx,cell  concentration of steroid in cells rate • Maintains ability to secrete all major • Mechanistic model can improve understanding of dose-response Cx concentration of steroid in medium AGT partition adrenocortical steroids x,med • behavior for chemicals that alter activity of steroidogenic enzymes coefficient • Proposed as EPA screening assay for environmental chemicals • Yields algebraic equations • This capability could help define mechanisms of action for poorly • 6 enzyme capable of disrupting or modulating steroidogenesis Decouples equations for steroids in medium from equations for characterized chemicals and mixtures in support of the H295R • inhibition • Being evaluated to screen drug candidates based on steroids in cells steroidogenesis screening system constants Miyuki Breen was supported by the NCSU/EPA Cooperative Training Program in Environmental Sciences steroidogenic effects in early phase of drug development - simplify large inverse problem Research, Training Agreement CT833235-01-0 with North Carolina State University. This work was reviewed by the US EPA and approved for publication but does not necessarily reflect Agency policy.