Hybrid structural modelling of alloantibody binding to human leukocyte antigen with rapid and reproducible cross-linking mass spectrometry (Set 2 of 3)

Alloantibody recognition of donor human leukocyte antigen (HLA) is associated with poor clinical transplantation outcomes. However, the molecular and structural basis for the alloantibody-HLA interaction is not well understood. Here, we used a hybrid structural modelling approach on a previously studied alloantibody-HLA interacting pair with inputs from ab initio, in silico and in vitro data. Highly reproducible cross-linking mass spectrometry data was obtained with both discovery and targeted-based approaches. The cross-link information was then used together with predicted antibody F_v structure, predicted antibody paratope and in silico predicted interacting surface to model the antibody-HLA interaction. This hybrid structural modelling approach closely recapitulates the key interacting residues from a previously solved crystal structure of an alloantibody-HLA-A*11:01 pair. These results suggest that a predictive-based hybrid structural modelling approach supplemented with cross-linking mass spectrometry data can provide functionally relevant structural models to understand the structural basis of antibody-HLA mismatch in transplantation.

Set 2 of 3 uploaded dataset and files. Discovery cross-linking mass spectrometry data and search files.