You are here:
Integrating in vitro and in silico data in a model of neurovascular developmental toxicity
Citation:
Saili, K., T. Zurlinden, N. Baker, AND T. Knudsen. Integrating in vitro and in silico data in a model of neurovascular developmental toxicity. Presented at NC-SOT, RTP, NC, October 30, 2017. https://doi.org/10.23645/epacomptox.6822947
Impact/Purpose:
Poster presentation at the NC SOT meeting. Although perturbed blood-brain barrier (BBB) function has been implicated in some neurobehavioral disorders, it is unclear whether developmental toxicant exposure impacts BBB formation. The results of this study highlight potential molecular initiating events that may impact neurogenesis through BBB disruption.
Description:
Although perturbed blood-brain barrier (BBB) function has been implicated in some neurobehavioral disorders, it is unclear whether developmental toxicant exposure impacts BBB formation. We utilized a computational systems model of the neurovascular unit (cNVU) in concert with in vitro ToxCast chemical effects data on neurogenesis and angiogenesis assays to build a predictive model of BBB developmental disruption. A subset of 38 ToxCast chemicals was tested for effects on endothelial cell (HUVEC) and neuroprogenitor cell (hNPC, hNC, hNN) behaviors at a range of concentrations. ToxPi potency scores were derived for each chemical across a range of 8 neurogenesis and 9 angiogenesis cellular behaviors, including cell titer, migration, tubule formation, and neurite outgrowth. Chemical clustering by cellular effects resulted in three major groups characterized predominantly by neurogenic effects, neuro-/endothelial effects, or negligible effects. There were 311 ToxCast assays associated with chemicals in either of the positive effects groups, 32 of which could be mapped to one of 18 critical nodes in a nascent systems model of NVU development that was used to predict chemicals that may perturb BBB formation. Based on a second ToxPi analysis of the 311 ToxCast assays, PFOS, a perfluorinated compound of growing human health concern, had the highest NVU-relevant ToxPi score suggesting a potential effect on BBB development. PFOS affected network formation in the neurogenic, but not angiogenic systems. It also tested positive in several inflammatory ToxCast assays (e.g., TGFβ1, VCAM1, MCSF, IL8, CSF1R, TIE2) suggesting an integrative model would reveal an impact on cytokine signaling. Both PFOS and a known angiogenesis disruptor, 5HPP-33, also tested positive in several ToxCast assays representing angiogenesis signaling (e.g., VEGFR1, VEGFR2, VEGFR3). Both of these compounds also impacted neurogenic endpoints in vitro and suppressed TGFβ1, one of three ToxCast assays directly implicated in neuroprogenitor development in the cNVU model. The results for PFOS and other putative BBB disrupting compounds highlight potential molecular initiating events that may impact neurogenesis through BBB disruption. Taken together, the integration of new in vitro cell-based tests with existing ToxCast data facilitates translation of cell-cell interactions within the NVU into a predictive toxicology framework of BBB disruption. This abstract does not reflect US EPA policy.
URLs/Downloads:
DOI: Integrating in vitro and in silico data in a model of neurovascular developmental toxicity
SAILI_NCSOT_FALL2017_POSTER_20171024_FINAL_STICS_M.PDF (PDF, NA pp, 2004.439 KB, about PDF)