In silico design and optimisation of novel host-directed therapies

Project background

Host directed therapies augment the natural immune response to infection by supplementing host defence mechanisms or reducing excessive inflammation and can shorten the treatment period of anti-infective agents. These therapies can also provide an alternative where infections are unresponsive to anti-infective agents. Aspergillus fumigatus causes the most common lung fungal infection, Aspergillosis, and experimental data from murine models suggests pro-inflammatory cytokines play an important role in controlling this infection. A host directed therapy treating with the pro-inflammatory cytokine, IFNγ, has become an accepted treatment for patients unresponsive to anti-fungal drugs. For these therapies to be successful, a mechanistic understanding of pathogen activity and host immunity is required.

Why we funded it

This PhD Studentship aims to develop an in silico model of underlying mechanisms of IFNγ therapy for A. fumigatus infection as a replacement for mice used in these studies.

Dr Tanaka and Dr Bignell have previously developed a simulated dose-finding experiment, used to validate predictions and inform experiments. This replaced 2000 animals that would have been required reducing the overall animal requirements for the study by 98.5%. It is estimated a similar level of reduction can be achieved through the proposed computational approach regarding the studies of IFNγ therapy. This will replace the use of up to 400 animals per annum in the Bignell laboratory.

Research methods

Routine use of simulations in infectious disease research has the potential to replace initial experiments by quantitatively assessing the impact of infection and identifying parameters which most potently influence the infection outcome. Additionally, it can reduce numbers of animals in further experiments by calculating minimal sample sizes required for the validation of simulation data. The proposed in silico model will be used to determine the optimal treatment strategy for IFNγ therapy of Aspergillosis in patient-specific cases. To demonstrate the scientific utility of the computational approach, the predictions of the optimal treatment strategies will be validated in mice. The simulation programmes developed in this proposal will then be incorporated into a prototype graphical-user-interface simulator suitable for use in the fungal research community.

Adjunctive host-directed therapies which augment the natural immune response to infection have the potential to shorten duration of anti-infective treatments, prevent drug resistance and reduce tissue injury by promoting autophagy, antimicrobial peptide production and other macrophage effector mechanisms. Maximal exploitation of this approach demands an integrated, mechanistic understanding of host immunity and pathogen activities, usually obtained from whole animal experiments, which most researchers use as the gold standard model for human infectious disease. However, a practical and ethical assessment of mechanism can never be credibly achieved via murine infection models, because prohibitively large numbers of animals are required to explore all variables governing therapeutic regimens. We therefore contest that a paradigm shift is required to convert the discovery process to one which is more computer-, rather than animal-driven.

We have research track records in a simple mathematical modelling approach, amenable to learning by nonmathematicians, which might accelerate this shift. This studentship builds upon the approach we have recently validated and published (Tanaka et al. 2015). While our previous efforts revealed the effects of low-dose fungal spore inhalation, this new programme of research will convert the theoretical framework to design the optimal treatment strategies. The student will use it to model, in silico, the underlying mechanisms of a clinically accepted IFN-gamma therapy for invasive fungal infection. By allowing the exhaustive and systematic interrogation of therapeutic options, including combinatorial effectsand optimal regimens, this approach provides a mathematical framework to perform in silico-guided design of novel optimal therapies.