NephroTube: kidney toxicity screening in vitro
We have funded the creation of NephroScreen, an organ-on-a-chip platform that models human renal tubular injury and has the potential to replace in vivo nephrotoxicity studies in drug development.
Challenge Contractor: Dr Martijn Wilmer
Organisation: Radboud University Medical Centre, The Netherlands
Start date: 2013
Duration: 3 years
Sponsors: GSK | Pfizer | Roche
Preclinically, drug candidates are assessed for their nephrotoxic potential using cell assays and subsequently as part of chronic toxicity tests carried out for regulatory purposes, often in two species (a rodent and a non-rodent). Nephrotoxicity accounts for 2% of drug failures in the preclinical stages of development and 19% of all failures in Phase III clinical studies.
The translational gap between the preclinical toxicology models and their predictive value to the clinic means that compounds are tested in animals rather than being screened out early in development, or that standalone in vivo studies are required to investigate the relevance of a nephrotoxic signal.
To minimise the use of animals and address the difficulties of predicting nephrotoxicity, GSK, Pfizer and Roche posed the NephroTube Challenge to develop a complex, human relevant, microfluidic platform capable of accurately identifying nephrotoxic effects.
The NephroTube Challenge was awarded to a team led by Dr Martijn Wilmer at Radboud University Medical Centre in the Netherlands, that included expertise in drug transporters, advanced microfluidics and in vitro toxicology.
The team developed NephroScreen, a microfluidic human kidney proximal tubule-on-a-chip platform that can be used as an early in vitro screen for drug-induced nephrotoxicity. Consisting of 40 kidney proximal tubules per plate, the platform does not require complex pumps, making it suitable for high throughput and routine drug screening.
3Rs and scientific benefits
The team demonstrated that polarised 3D proximal tubules can be formed from epithelial cell lines cultured in a microtiter plate-based microfluidic chip (termed OrganoPlate®) under fluid flow, with cilia present on the apical side of the tubules. The tubules can be accessed from the apical and basal sides, enabling transporter mediated cellular efflux and influx to be measured. Importantly, the tubules from one of the cell lines used, have barrier integrity with the formation of tight junctions – the disruption of these is a key measure of nephrotoxicity.
NephroScreen was independently characterised at three different laboratories using two proximal tubule human cell lines and validated with 12 nephrotoxic compounds provided by the Sponsors. This demonstrated that the platform can accurately detect toxicities when compared with preclinical and/or clinical data, and that multiple endpoints such as drug transporter function, barrier integrity, cell viability and genetic variations – essential endpoints for assessing drug-induced toxicity – can be measured. Using NephroScreen, the team identified novel indicators of kidney injury, such as upregulated miRNAs, which may be better markers of cell viability than the currently used proliferation assays.
The development and preliminary characterisation of NephroScreen has been described in five publications to date. Workshops have been held to facilitate transfer of the technology to the Sponsors, with supporting Standard Operating Procedures for 20 nephrotoxicity endpoints published online. The utility of NephroScreen is also being evaluated by six other pharmaceutical companies. It is early days to assess the 3Rs impacts but as confidence and experience builds in the use of microphysiological systems, such as NephroScreen, there will be an impact on animal use – in the short-term to screen out compounds prior to in vivo tests and ultimately in the long-term potentially as part of a suite of in vitro approaches to replace chronic toxicity studies.
All components of NephroScreen are now available commercially. Martijn has recently co-founded Cell4pharma, a spin-out company from Radboud University, to provide human renal cell lines. For end-users wanting to purchase an off-the-shelf product, MIMETAS, a Dutch SME, is investigating the potential to turn NephroScreen into a single commercial offering to further increase the uptake and reproducibility of the model.
The OrganoPlate® was further adapted for neurotoxicity studies in the Neuratect Challenge posed by Abbvie, GSK and Sanofi in 2014. A team led by MIMETAS was awarded £1M to generate a human stem cell-based model for the assessment of neurotoxicity and seizure liability in safety pharmacology studies. Behavioural observations in animals, usually rats, are used to predict whether a drug may cause seizures in the clinic. In vitro modelling of seizurogenic activity offers the potential to screen large numbers of compounds in early development and mitigate downstream seizure concerns without using animals. The team developed a microfluidic cell culture platform, termed NeuroScreen 3D, using the OrganoPlate® system coupled with a co-culture of human-induced pluripotent stem cell-derived neurons and astrocytes from Cellular Dynamics International. The cultures form neuronal networks and demonstrate burst activity which can be measured by calcium imaging. Measurements of burst activity showed that NeuroScreen 3D can discriminate between compounds with known seizurogenic activity selected by the Sponsors. The platform provides the ability to multiplex calcium imaging assays with neurite outgrowth and cell viability measurements and integrates single neuron activity detection for highly granular data analysis. The Sponsors are at various stages of evaluating the technology in-house. The components of NeuroScreen are available commercially.
Sponsor in-kind contributions
The Sponsors provided a blinded compound set for nephrotoxicity testing across multiple laboratory sites and related preclinical and clinical data for validation purposes.
This case study was published in our 2019 CRACK IT Review.