Using machine learning to combat rejection in medical transplants

Dr Andreas Tiffeau-Mayer (UCL Division of Infection and Immunity) used the UCL – University of Sydney Strategic Ignition Grant to further knowledge about why some medical transplants fail.

It is widely accepted that medical transplants in humans need to take place between a donor and a recipient who ‘match’. If there isn’t a match, the transplant can fail. Transplant rejection relates to donor tissue being recognised as foreign by the recipient’s immune system. This is a response that happens on a molecular level. Donor tissue has so-called Major Histocompatibility Complex (MHC) molecules present on its cells, which can differ from those of the transplant recipient. Donor MHC molecules form complexes with immunogenic self-peptides – fragments of proteins. It is these specific complexes that elicit the immune response towards the donor tissue. T cells – a type of white blood cell, are believed to be responsible for the immune response from the recipient’s side. However, more work is needed to understand the specifics of why the immune response and subsequent rejection happens.

Dr Andreas Tiffeau-Mayer and his lab at the UCL Division of Infection and Immunity have previously used computational methods to understand how T cell receptors engage their molecular targets in the context of infectious disease. Associate Professor Alexandra Sharland from the University of Sydney works on transplant immunology, and her lab has generated large amounts of sequencing data through the wet lab experiments they do. Believing there were synergies in their work that could further the molecular understanding of transplant rejection, they applied for the UCL – University of Sydney Strategic Ignition Grant. On being successfully awarded the funds, they embarked on a year-long project to develop the science in this field.

Understanding transplant rejection

“A key idea in transplantation is that there’s going to be a mismatch,” Andreas explained. “It’s a molecular interaction between receptors on one side, and what they recognise on the other side. We know we need to try to match people with a similar genetic background to do a transplant. But we also know it’s never going to be perfect. Therefore we’re trying to understand why the immune system sees a transplant as foreign.”

Alexandra’s lab has explored this in depth through experiments with mice in her lab. In particular, they have focused on identifying T cells that have a specificity to peptides that are foreign to the recipient. They believe that analysing the immune responses to particular donor MHC-peptide complexes in tissue will provide insights into the immune response against the transplanted organ. 

Andreas’s lab came to this subject matter from a different direction. They had previously developed statistical techniques to detect patterns in the sequence of receptors that drive the immune response to different infectious diseases. 

Andreas and his lab subsequently analysed some of the data generated by Alexandra’s lab, using the computational methods they had previously developed. Combining this computational and clinical expertise, the team discovered that immune responses – leading to transplant rejection – are primarily driven by the contact between T cell receptors and donor MHC molecules. The specific presented peptides play a lesser, secondary role in the immune response, in contrast to T cell responses to infectious diseases. This is a key finding that is helping to further work in the transplant field.

“This has been a thoroughly enjoyable experience, and a fantastic opportunity, not only for me but for the PhD students and postdocs in my group to realise first-hand the benefits of interdisciplinary collaboration,” Alexandra said. “As biologists, we tend to focus on interactions at the cellular and molecular level. Andreas and his group were able to zoom out to look at transplant immune responses on a much broader scale, and reveal patterns that we would not have discerned otherwise.”

Developing an important area of science

“Ultimately, if we can better understand the rules of how the immune system sees transplanted tissue as being different, there will be positive therapeutic implications,” Andreas said. “We’ve never had a chance to study T cell receptors in the context of transplant immunology before this. It’s been very beneficial to test the generality of our computational methods, and gain insights into the molecular interplay between MHC molecules.”

As a result of the findings from this project, Andreas and Alexandra have the evidence they need to apply for further funding to pursue related research. They are also working on writing up the results of this project for publication. 

As well as furthering an important aspect of science, this project has created opportunities for collaboration and networking for others active in this field. Both labs had early career researchers involved in the project – which they believe was a great career development opportunity – and the funding also enabled Alexandra and two members of her team to visit London. They organised a scientific symposium – the London Quantitative Immunology Day – during this visit. Attended by more than 80 people involved in the field, and with Alexandra as the keynote speaker, this was an opportunity for the collaborators to showcase their work to the scientific community.

Andreas says this funding has enabled a high-quality collaboration with experts he would otherwise have been unlikely to work with due to the geographic distance between the UK and Australia. “I’m a physicist by training, and I’ve worked in computational immunology for a while now,” Andreas said. “Alexandra comes from the world of transplant immunology and a more medical background. This was an international, interdisciplinary collaboration. It’s expanded our networks, and enabled our labs to learn more about each other’s expertise. We definitely hope to continue working together.”

Links

• Dr Andreas Tiffeau-Mayer's academic profile
• UCL Division of Infection and Immunity
• UCL Faculty of Medical Sciences
• The London Quantitative Immunology Day 

Featured image

Credit: iStock
Description: BioScience Antibody, T-Cell, Antibody cell, Molecule of immunoglobulin.