We have funded the creation of an in silico model of the host response to chronic infection by Leishmania parasites, to reduce the number of animals used in the development of drug treatments and vaccines.
Challenge Contractor: Professor Paul Kaye
Organisation: University of York
Start date: 2013
Duration: 3 years
Large numbers of animals, typically rodents, are used in efficacy studies for new antibiotics or vaccines each year. The animals are infected with the pathogen of interest after vaccination or treated with the experimental drug. Untreated controls are always used. The resulting disease in control animals and those in which the vaccine or drug are ineffective can cause significant suffering. The use of in silico approaches to study disease biology and predict efficacy has the potential to replace some animal use.
Addressing the animal welfare concerns associated with predicting drug efficacy, the NC3Rs posed the Virtual Infectious Disease Research (VIDR) Challenge to develop a model of infection and the host response to pathogen insult for use in basic research and target development.
The VIDR Challenge was awarded to a team led by Professor Paul Kaye, an expert in the immunology of the tropical disease leishmaniasis, and his collaborators which included Simomics, a software SME that develops tools for pharmaceutical and life sciences industries.
The team developed the LeishSim Virtual Laboratory, a computational model of a mouse spleen chronically infected with the parasites that cause visceral leishmaniasis. The model is parametrised to allow the interrogation of various genetic, immune and pharmacokinetic factors over the time-course of the infection and the impact of potential treatment interventions.
3Rs and scientific benefits
Around 40,000 animals are used globally each year for leishmaniasis drug and vaccine research and development. The LeishSim Virtual Laboratory provides the opportunity to minimise animal use by testing hypotheses in silico and prioritising those experiments where an in vivo approach is essential. For example, Paul and colleagues have used LeishSim to investigate monotherapy treatments. 1,248 simulations were run over four days to analyse potential knockout phenotypes across approximately 95 independent variables (genes) with each simulation providing data for as many timepoints as required. The comparable study in mice would 36 | 37 CRACK IT Review 2019 have taken at least three years to run the assays and required more than 3,500 animals. Using LeishSim, the team were able to identify the best genes to focus on, using 62 mice for the in vivo validation of the best “hits” from the model output.
LeishSim has been validated for use with both the major parasite species that cause visceral leishmaniasis (L. donovani and L. infantum). Generating the necessary parameterisation data has delivered new understanding of leishmaniasis immunology, including tissue-specific alterations in the transcriptome; new understanding of immune changes during and after treatments; and novel data on drug pharmacokinetics that will inform the formulation of topically applied treatments. The model also has the potential to be applied to other forms of leishmaniasis such as post kala-azar dermal leishmaniasis where there is no animal model available.
The tissue transcriptomic data and histopathology on the early and late response of mice to L. donovani infection in the presence or absence of drug treatment, and associated metadata, generated through the VIDR Challenge has been published in Wellcome Open Research and whole slide images are available on the global pathology network, LeishPathNet. The samples used to generate the data are banked at the University of York, enabling other researchers to request additional tests on the existing samples rather than conduct new animal studies.
The development of LeishSim has resulted in the growth of Simomics from two to eight staff. The underpinning software can be adapted and customised for other areas relevant to the 3Rs including toxicological risk assessments. For example, Simomics has secured a further £0.2M through CRACK IT for Challenges which focus on in silico modelling of reproductive and developmental toxicity (the DARTpaths Challenge) and acute toxicity for the classification and labelling of chemicals (the Maximise Challenge). They also led a consortium awarded £0.45M from Innovate UK (in a competition co-funded by the NC3Rs) to provide a Virtual Fish Ecotoxicology Laboratory for environmental toxicity testing of new and legacy drugs that were tested prior to current environmental regulations coming into force. The Virtual Fish software, developed in partnership with AstraZeneca and scientists at the University of York, integrates mathematical models for toxicity, exposure, uptake and metabolism. It forms the basis of a €10M Innovative Medicines Initiative partnership of nine pharmaceutical companies, academics and European regulators on the intelligent prediction of environmental risks posed by human medicinal products. Over the next decade the partnership is expected to save one million fish, and industry over £500M in unnecessary testing, without compromising environmental protection.
Paul and Simomics are also part of a consortium that has received $148k from the Foundation for Sarcoidosis Research to adapt the model for sarcoidosis, an inflammatory disease which, like leishmaniasis, is characterised by the formation of granulomas.
Sponsor in-kind contributions
This Challenge was sponsored by the NC3Rs. There were no in-kind contributions.
This case study was published in our 2019 CRACK IT Review.