Automated Hepatocyte Differentiation From Pluripotent Stem Cells and Their Use in Drug Screening
David Hay, Principal Investigator, MRC Centre for Regenerative Medicine, University of Edinburgh
In both preclinical and clinical testing, drug induced liver injury
(DILI) is the major cause of drug attrition, costing pharmaceutical
companies millions of dollars every year (Ware & Kehtani 2017,
Trends in Biotechnology). It is therefore essential to rapidly identify
and remove drug candidates that pose a risk, decreasing development
costs and improving post market success. Currently primary human
hepatocytes (PHH) are the most widely used cell type for the prediction
of DILI. However when cultured in vitro PHH do not proliferate, rapidly
lose their phenotype, in particular drug metabolism. Furthermore,
sources are limited and usually isolated from diseased tissue limiting
their utility in large scale drug screening. The cost effective delivery
of human tissue from a renewable resource, such as pluripotent stem
cells (PSCs), offers a possible solution to the issues associated with
human somatic cells for in vitro testing. Of note, PSC derived
hepatocytes have previously been shown to be as sensitive and
predictable as the pharmaceutical ‘gold standard’ PHH (Szkolnicka et al
2014, Stem Cells Translational Medicine and Szkolnicka et al 2016, Stem
Cells Translational Medicine). More recently, improvements in culture
methods and protocols have led to further improvement in hepatocyte
function and stability. The focus of our recent research has been to
combine defined differentiation with automation to deliver reliable and
stable liver models at scale. In order to multiplex our cell based
screening tool we have used a cytological assay that ‘paints the cell’,
specifically marking intracellular components of the cell, allowing high
through put and multi-parametric analysis (Gustafsdottir et al 2013,
PLoS One; Bray et al 2016, Nature Protocols). PSCs were differentiated
into hepatocyte like cells (HLCs) using a previously published protocol
(Cameron et al 2015, Stem Cell Reports). Once mature HLCs were incubated
with a pharmaceutical grade compound library provided by AstraZeneca.
Following 48 hours exposure, we scored for drug toxicity, using multiple
endpoints and correctly identified drug toxicity in compounds tested.
In conclusion, we have developed an automated, reproducible, and
scalable platform to generate functional hepatocytes for human drug
screening. Importantly, the production method is GLP compliant and cost
effective (Cameron et al 2015, Stem Cell Reports). Going forward our
automated system will be further miniaturised and fine-tuned to study
genetic variation in the human population.
|