Predicting HLA-specific humoral alloimmunity to improve histocompatibility and immunological risk assessment in transplantation.

Renal transplantation prolongs and enhances quality of life and is the optimal treatment for patients with end-stage renal failure. Despite efforts to increase the availability of donor organs, the size of the kidney transplant waiting list has remained largely unchanged reflecting the constant demand for more organs. A major challenge, therefore, in organ transplantation is to ensure that graft survival is maximised. Although early results after transplantation are excellent, long-term outcomes have remained static and over 30% of kidney transplants fail within 10 years. This is due to a process of chronic graft injury which occurs mainly due to development of antibodies (called alloantibodies) against special proteins expressed by the donor organ (called Human Leukocyte Antigens or HLA). Alloantibodies are also detrimental to the function and longevity of other solid-organ transplants such as heart, lung and pancreas transplants.

The risk of antibody-mediated damage can be offset when organ donors and recipients have a good tissue-match (donor and recipient share as many HLA proteins as possible) but this is often not possible because HLA proteins vary widely in individuals, both within and between different ethnic groups. Current methods for determining tissue compatibility and HLA-related immunological risk are inadequate. This is because they simply enumerate HLA differences between a donor and a potential recipient without considering the molecular and structural differences between their HLA proteins.

The primary focus of this research is to investigate better ways to assess tissue compatibility in transplantation and to accurately define the risk of alloantibody development. Building on preliminary studies, we will use advanced bioinformatics modelling tools and experimental data to develop a computational algorithm that determines the degree of donor and recipient HLA incompatibility and quantifies the associated immunological risk. We will do that by investigating two components of the immune response against donor HLA. First, we will assess how unique structural and physical properties of donor HLA are recognised by specialised cells in the recipient, called B-cells, driving the immune response. Second, we will also examine the potential of another cell population, called T-cells, to provide help to recipient B-cells which is an essential step for the development of alloantibodies. For this to happen, T-cells have to recognise small fragments of donor HLA that are presented to them by recipient HLA on the surface of B-cells. We will use specialised computational techniques to predict the capacity of recipient HLA to bind donor HLA-derived fragments and the accuracy of our predictions will be examined in patients who develop strong alloantibody responses after exposure to foreign HLA.

The computational algorithm for assessing donor and recipient HLA incompatibility will be validated in large clinical datasets in collaboration with NHS Blood and Transplant and with international transplant organisations. The datasets will include thousands of patients that underwent kidney transplantation. These studies will examine whether assessment of donor and recipient histocompatibility based on our algorithm is superior to conventional HLA matching as a strategy to reduce the detrimental effect of post-transplant alloantibody responses and improve long-term graft survival after transplantation.

An important strategy to offset the risk of alloimmunity and maximise transplant outcomes is to minimise the number of HLA mismatches between donor and recipient. However, current assessment of HLA incompatibility is inadequate, limited by the assumption that all mismatches within an HLA locus are equally significant. The principal aim of the proposed research is to explore the full potential of applying advanced bioinformatics modelling techniques, incorporating different mechanistic aspects of the immune system and capitalising on analysis of large clinical datasets, to develop new and improved methods for predicting HLA immunogenicity, determining HLA compatibility and defining HLA related immunological risk.

Building on previous work, we will develop a computational algorithm to analyse the structure and physicochemical properties of HLA molecules and identify immunodominant B-cell epitopes on donor HLA based on their unique surface electrostatic properties compared to recipient HLA. Similarly, immunodominant CD4+ T-cell epitopes on donor HLA will be detected by investigating the potential of recipient HLA class II molecules to present donor HLA-derived peptides, thus accounting for the indirect pathway of alloreactivity. This immunoinformatics approach to determining donor HLA immunogenicity will be tested and refined by analysis of HLA-specific alloantibody responses and by detection of CD4+ allospecific T-cells in HLA-sensitised patients.

The combined HLA immunogenicity algorithm, accounting for both B-cell and T-cell aspects of humoral alloreactivity, will then be validated in large, national, clinical datasets by examining its capacity to assess the risk of alloantibody development after kidney transplantation and whether it leads to improved assessment of HLA incompatibility, compared to conventional HLA mismatching, by enabling selection of HLA incompatible donor-recipient combinations that are favourable in terms of long-term kidney graft survival.

Societal impact: Renal transplantation is the optimal treatment for patients with end-stage renal failure (over 3,267 kidney transplants were performed last year in the UK) but long-term benefits are compromised by antibody mediated graft injury leading to graft loss. I anticipate that the proposed research will enable selection of donor-recipient HLA combinations that are clinically favourable in terms of alloantibody development and allograft survival. There is currently great interest in incorporating more accurate ways of assessing HLA incompatibility into the national deceased donor organ allocation scheme and this is a realistic prospect in the near future. Improved histocompatibility is likely to lessen the requirement for heavy immunosuppression post-transplantation impacting favourably on patient morbidity and mortality (death with a functioning graft is a major cause of allograft loss). Findings from this research have the potential to reduce the requirement for re-transplantation, thus reducing the pressure on the limited resource of donor organs. HLA antibodies severely limit the chance of finding an antibody-compatible donor kidney and the proposed approach will be particularly important for recipients (e.g. children and young adults) who are likely to require repeat transplantation in the future (approximately 23% of patients listed for renal transplantation in the UK are awaiting a repeat transplant and most have developed HLA alloantibodies). The computational matching algorithm will be more permissive and can improve access to transplantation by enabling allocation of kidneys to HLA mismatched patients that are currently considered poor tissue matches. Humoral alloimmunity is a significant burden in cell (e.g. pancreatic islet transplantation), tissue and solid organ transplantation (including heart, lung and pancreas transplantation). The proposed research will help quantify the HLA-related immunological risk associated with transplantation and will likely be applicable into the wider transplant field.

Economic impact: End-stage renal failure and the need for renal replacement therapy represent a significant economic burden for the NHS. Renal transplantation is associated with a cost benefit compared to renal replacement therapy but the full benefits to the health service are not realised because of long-term graft attrition rates that have remained largely unchanged despite modern patient management. My research has the potential to improve long-term transplant outcomes and increase access to transplantation, thus reducing the economic impact of renal diseases on the NHS.

Non-academic beneficiaries: The proposed research is likely to generate intellectual property (IP) rights. A novel computational scoring system will be developed that assesses donor and recipient HLA compatibility. This system may be implemented in national and international deceased-donor kidney transplant allocation schemes. It may also be used by individual transplant centres to make clinical management decisions on transplant patients. This computational approach will be implemented into a new software that will be copyrighted (open source) and distributed as an executable file.