Predictive value of human stem cells for developmental neurotoxicity testing


John C. Livesey: 0000-0001-9010-5970

Document Type


Conference Title/Conference Publication

British Neuroscience Association 2013: Festival of Neuroscience


Barbican Centre, London, UK

Conference Dates

April 7-10, 2013

Date of Presentation



The developing human nervous system is vulnerable to injury because cell growth, proliferation and maturation occur over a prolonged time period in utero and post-partum. Animal models of developmental neurotoxicity (DNT) are costly and have relatively low predictive value for humans. In this study, we have utilized human pluripotent stem cells (hPSCs) to develop an in vitro model of neurogenesis and have tested its sensitivity to 4 widely used antiepileptic drugs (AEDs) associated with risks of neurological impairments in humans with perinatal exposure.

The hPSC line was differentiated toward neural phenotypes with retinoic acid. The effects of phenobarbital (PHB), valproic acid (VPA), lamotrigine (LTG) and carbamazepine (CBZ) were examined on stem cell viability by the MTT assay, proliferation and apoptosis by flow cytometry and pluripotency and neurogenesis were investigated by immunocytochemistry. VPA dose-dependently reduced viable cell number and induced apoptosis whereas only high concentrations of PHB and CBZ induced apoptosis. Of the 4 drugs tested, VPA caused the greatest reduction in cell proliferation. All concentrations of PHB significantly reduced neurogenesis.

In contrast, lower concentrations of VPA (10μM and 100μM) increased neurogenesis and only reduced it at the highest concentration tested (1000μM). LTG dose-dependently reduced viable cell number and moderately impaired neurogenesis. Only VPA reduced Oct-4 expression suggesting a reduction in stem cell pluripotency.

In conclusion our model shows differential sensitivity to the 4 antiepileptic drugs investigated, with PHB and VPA having the highest potential risk for DNT, correlating with an increased risk of neurological malformations observed with these agents in epidemiological studies. Additionally, our data corresponds with in vivo data showing that VPA inhibits cell proliferation in the rodent brain and PHB impairs neurogenesis. We therefore report for the first time that the use of hPSCs as an in vitro model of neurogenesis might be a powerful alternative to animal-based methods of DNT testing and enable rapid evaluation of drug safety in the future.


Volume 22, P2-G-028

Click here to view the festival website

This document is currently not available here.