Human pluripotent stem cell cardiomyocytes and hepatocytes with engineered genotypes for drug safety evaluation

Lead Research Organisation: University of Nottingham
Department Name: School of Clinical Sciences


An essential part of drug development by pharmaceutical companies is safety assessment before the drugs are given to the general patient population. Current regulation necessitates that much of this testing is done in cells or organs derived from animals, or in living animals, which primarily comprise mice, rats, dogs and monkeys. New EU regulations on toxicity testing (known as Registration, Evaluation, Authorisation and restriction of Chemicals or 'REACH') mean that up to 54 million animals will be needed over the next 10 years to evaluate 30,000 compounds, which equates to 1800 animals per drug. Even with this level of animal use, published information suggests that preclinical testing is only 71% predictive of whether a drug will be toxic in humans. Poor predictability is due to species differences, where the physiology of humans is very different to the various animals used for the tests. In addition, current systems cannot mirror the genetic diversity of human populations, whereby one person can tolerate a very high dose of drug, while someone else will have an adverse reaction and end up in hospital. Indeed, adverse drug reactions account for 100,000 deaths per year in the US alone and this is mainly due to problems that develop with the person's heart or liver.

To try to reduce or replace animal use and to increase predictability to human toxicity, this proposal will form a new collaboration between scientists with skills in genetic engineering (Skarnes, Rosen; Wellcome Trust Sanger Centre), human pluripotent stem cell biology and robotic automation (hPSC; Denning, Young; University of Nottingham) and toxicity testing (Goldring, Park; University of Liverpool). The aim is to genetically engineer human stem cells that are healthy except for the newly-added mutation, which has been selected because it is believed to cause susceptibility or resistance to certain drugs. In the lab, we will then coax the stem cells into becoming liver or heart cells. By investigating the function and survival of these cells after treatment with a range of drugs, we will be able to quantitate the amount of positive or negative change associated with a particular mutation. This has not previously been possible because human heart or liver cells with these mutations have not been available, or because the background genetics vary so much from one patient to another that accurately determining the impact of a single mutation on drug response is difficult. Keeping background genetics the same but varying the mutation overcomes these problems and will provide the pharmaceutical industry with a new tool to evaluate drug response in healthy cells and those carrying mutations that confer drug susceptibility or resistance. This will not only reduce animal use but also increase the safety of drugs taken by people, including those who are predisposed to adverse drug reactions. If this technology only led to a 0.1% reduction in animal-based testing, it would have the potential to save up to 54,000 animals in Europe alone.

Technical Summary

Safety assessment of each new drug developed requires up to 1800 animals, which typically comprise non-rodent species (monkeys, dogs) and rodents (rats and mice). New EU regulations on toxicity testing (termed 'REACH') will use up to 54 million animals over the next 10 years to evaluate 30,000 compounds. Even with this level of animal use, preclinical assays are cited as only 71% predictive of whether a drug will be toxic in humans. Poor predictability is due to species differences and the inability to develop test platforms that mirror diverse human genotypes, particularly those associated with the high drug susceptibility seen in cardio- and hepato-toxicity. Adverse drug reactions account for 100,000 deaths per year in the US alone. To facilitate reduction and replacement of animal use, and to increase predictability to human toxicity, this proposal brings together a new consortium with skills in genome engineering (Skarnes, Rosen; Sanger Centre), human pluripotent stem cell biology & robotic automation (hPSC; Denning, Young; Nottingham) and toxicity testing (Goldring, Park; Liverpool). The aim is to engineer different patient-relevant mutations associated with drug susceptibility or resistance into the protein coding regions of otherwise genetically healthy hPSCs. Differentiating these cells will produce cardiomyocytes and hepatocytes that carry specific mutations within a common genetic background, allowing unbiased evaluation of how genotype-drug interaction affects cell structure, function and viability. Co-culturing cardiomyocytes with hepatocytes of different genotypes will allow the impact of altered hepatocyte function on cardiomyocytes to be assessed. Comparing these results with data from the literature and industrial partners will allow the predictive value of this humanised in vitro model to be determined. A 0.1% reduction in animal-based in vitro, ex vivo and in vivo tests has the potential to save up to 54,000 animals in Europe alone.

Planned Impact

Engineering and differentiating hPSCs to cardiomyocytes and hepatocytes for use as novel in vitro genotyped human models in drug evaluation has the capacity to make substantial impact on the 3Rs, industry, student / staff training and public perception of science. Text below shows how the work will have direct impact and highlights our existing routes to impact, which will be continued into the NC3Rs project.
Impact on 3Rs
Drug safety assessment & toxicity testing requires 475,290 animal procedures each year in the UK, while predicted numbers to satisfy EU REACH regulations may be up to 54 million over the next 10 years. Although combinations of mice, rats, dogs and monkeys are used, species differences mean that prediction of human toxicity is cited as only 71% (May et al., Br J Biochem Sci. 66:160-5, 2009). Other animals, such as guinea-pigs, are also used to carry out ex vivo studies (e.g. heart ventricular wedge assays for cardiotox). Any new platforms that can improve on existing toxicity tests would have a substantial impact on replacing and / or reducing animal use. Development of the genotypically-relevant human-based in vitro resource described in this proposal has the capacity to complement existing tests. If this leads to a modest reduction of 0.1% in animal use, this would equate to sparing up to 54,000 animals.
Industrial Impact
The pharmaceutical industry has reached crisis point with 98% of sales based on products of more than 5 years old, 110,000 jobs lost in the US and $130 billion dollars lost due to patent expiry. Escalating costs of drug development, estimated as $1.5 billion per drug, are contributing to this crisis. Developing better predictive tools to complement existing safety assays would help to reduce costs by reducing numbers of poor drugs reaching pre-clinical (hence animals) or clinical testing phases. Members of Safety Screening within AstraZeneca estimate that incremental (i.e. less that 1%) increases in predictivity would save tens if not hundreds of hundreds of millions of dollars per drug.
In this project, we will closely involve our industrial partners (Syngenta / BRIC partners via Nottingham team; various industrial partners via Liverpool's MRC Centre for Drug Safety Science [CDSS;] - See also 'Communication Plan' of JeS form). This will ensure the assays developed are relevant to industry (existing partners and newly-scoped partners), which will promote uptake of the technology and hence benefit a reduction in animal use. The use of the robotic platforms in this proposal will be of particular interest to industry.
The Tech transfer Offices at each of the institutions will facilitate industrial interaction and exploitation. By way of example, in Nottingham ~80 new patent applications are filed each year, providing licensing revenues of £700K p/a and leveraging >£10m p/a from industry. The University has 24 spin-out companies in its portfolio, of which 12 have products in the market place and have collectively raised more than £60m in venture finance in recent years.
The current application will train a PDRA and 0.5FTE technical post, while Nottingham will provide 1 PhD student and 3 MSc students. This builds on the team's repertoire of training that includes a Masters student (in excess of 100 trained), an EPSRC-funded Doctoral Training Centre (50 PhD students over 5 years) and a track record of supplying skilled workers to the industrial sector. Through these strategies and others (e.g. Erasmus, lab exchanges, sabbaticals), we have trained personnel from 30 countries, which often leads to longer term collaboration. This has supplied a workforce to academia (e.g. PhDs, postdocs, academics), government (e.g. NHS, teaching), charities (Anthony Nolan trust), SMEs (e.g. Nutricia, Oxford Agricultural Trials) and major industry (e.g. GSK, AZ, Syngenta, Pfizer, GE Healthcare, MedImmune).
Public engagement of science will be as in the Communication Plan


10 25 50