Science Inventory

Application of Signaling Pathway-Based Adverse Outcome Pathways and High Throughput Toxicokinetic-PBPK for Developmental Cardiac Malformations

Citation:

Hunter, S., K. Grode, Mitch Rosen, D. Kapraun, N. Sipes, N. Baker, R. Judson, AND T. Knudsen. Application of Signaling Pathway-Based Adverse Outcome Pathways and High Throughput Toxicokinetic-PBPK for Developmental Cardiac Malformations. Society of Toxicology, Baltimire, MD, March 12 - 18, 2017.

Impact/Purpose:

Identification of putative signaling pathway molecular initiating events (MIE) for cardiac malformations was performed by analyzing the Mouse Genome Informatics Phenotype database. Genes associated with valvulo-septal phenotypes were linked to curated pathways using Panther (pantherdb.org). A dataset containing in vitro cellular and molecular effects for 1061 ToxCast chemicals and gene scores for 249 target genes was used for analysis. The ToxCast gene targets were linked to pathways using Panther. For each chemical a pathway score was created as the sum of gene scores for the constituent genes in each pathway. A human pregnancy PBPK model was used to calculate the peak fetal serum concentration for a subset of 225 ToxCast chemicals using both 1 and 100 mg/kg/day doses during the 12th week of gestation. For each chemical, bioactivity was defined as reaching a peak serum concentration greater than the gene score (AC50) for the targets. Using the gene targets of the signaling pathways associated with valvulo-septal phenotypes, 46 chemicals were identified as bioactive at a 1 mg/kg/day dose and 109 chemicals at 100 mg/kg/day. Acetochlor, Rotenone, Tebufenpyrad, Thiabendazole, Perfluorooctanoic acid were ranked as the most potent chemicals at 1mg/kg/day. Oxytetracycline dehydrate, Acetochlor, Rotenone, Pyridaben and Thiabendazole at 100mg/kg/day. For the 4 toxcast chemicals that are rat or rabbit in vivo cardiac teratogens and have PBPK and gene score data , at the 100mg/kg/day dose all were identified as bioactive in the TGFbeta, Wnt, GPCR or gonadotropin signaling pathways; none at 1mg/kg/day. This analysis reveals a broad range of diverse signaling pathway MIEs that correlated with apical endpoints of prenatal cardiac developmental toxicity. Analysis of the relationships between altered pathway activity and phenotype induction are needed to define the specificity between altered pathways and phenotypes; however, this analysis is a starting point for pathway-based mechanistic prediction in defining the biological applicability domain for prenatal cardiac developmental toxicity. This abstract does not represent EPA policy.

Description:

Associating putative molecular initiating events (MIE) with downstream cell signaling pathways and modeling fetal exposure kinetics is an important challenge for integration in developmental systems toxicology. Here, we describe an integrative systems toxicology model for developmental cardiac malformations in general and valvulo-septal defects in particular. The Mouse Genome Informatics (MGI) Phenotype database was mined to recall genes associated with valvulo-septal phenotypes. This gene list was linked to curated pathways using the Panther database (pantherdb.org). Next, we mined the ToxCast database for effects of 1061 chemicals on 249 target assay-associated gene scores linked to the Panther pathways. For each chemical, a target pathway score was created by summing gene scores for all constituent genes in each associated pathway. This yielded a matrix of predicted chemical-pathway interactions for those elements linked to valvulo-septal defects derived from the mouse phenotype ontology database. A human pregnancy PBPK model was used to estimate the peak fetal serum concentration during the 12th week of gestation for a subset of 416 ToxCast chemicals following 1 or 100 mg/kg/day doses. For each chemical, bioactivity was defined by peak serum concentration in relationship to the gene score (AC50) for each target element. Using the gene targets of the signaling pathways associated with valvulo-septal phenotypes, 88 chemicals were predicted as bioactive at a 1 mg/kg/day exposure and 184 chemicals at 100 mg/kg/day. Acetochlor, Rotenone, Tebufenpyrad, and Valproic acid were ranked as potent chemicals. PBPK and gene score data were available for 4 ToxCast chemicals that are rat or rabbit in vivo cardiac teratogens in the ToxCast Toxicity Reference Database. The integrated exposure-response model predicted all four to be bioactive in the TGFbeta, Wnt, GPCR or gonadotropin signaling pathways at 100 mg/kg/day, but not at 1 mg/kg/day. This analysis reveals a broad range of diverse signaling pathway MIEs that correlated with apical endpoints of prenatal cardiac developmental toxicity. The integrative model provides a starting point for pathway-based mechanistic prediction in defining the biological applicability domain for prenatal cardiac developmental toxicity. This abstract does not represent EPA policy.

Record Details:

Record Type:DOCUMENT( PRESENTATION/ POSTER)
Product Published Date:03/18/2017
Record Last Revised:06/15/2018
OMB Category:Other
Record ID: 341165