BIITE: A Tool to Determine HLA Class II Epitopes from T Cell ELISpot Data

Guidelines for analysis of low-frequency antigen-specific T cell results: Dye-based proliferation assay vs 3H-thymidine incorporation

It is generally recognized that dysregulation of the immune system plays a critical role in many diseases, including autoimmune diseases and cancer. T cells play a crucial role in maintaining self-tolerance, while loss of immune tolerance and T cell activation can lead to severe inflammation and tissue damage. T cell responses have a key role in the effectiveness of vaccination strategies and immunomodulating therapies. Immunomonitoring methods have the ability to elucidate immunological processes, monitor the development of disease and assess therapeutic effects. In this respect, it is of particular interest to evaluate antigen (Ag)-specific T cells by determining their frequency, type and functionality in cellular assays. Nevertheless, Ag-specific T cells are detected infrequently in most diseases using current techniques. Many efforts have been made to develop more sensitive, reproducible, and reliable methods for Ag-specific T cell detection. It has been found that analysis of cellular proliferation can be a useful tool to determine the presence and frequency of Ag-specific T cell and to provides insight into modulation of the T cell response by a specific antigen or therapy. However, the selection of a cut-off value for a positive response and therefore a more accurate interpretation of the data, continues to be a major concern. Here, we provide guidelines to select a proper cut-off for monitoring of Ag-specific CD4⁺ T cell responses. In vitro Ag-stimulation has been assessed with two methods; a dye-based proliferation assay and ³H-thymidine-based assay. Two cut-off approaches are compared; mean and variance of control wells, and the stimulation index. By evaluating the proliferative response to the in vitro Ag-stimulation using these two methods, we demonstrate the importance of taking into consideration the variability of the control wells to distinguish a positive from a false positive response.

Predicting HLA class II antigen presentation through integrated deep learning

Accurate prediction of antigen presentation by human leukocyte antigen (HLA) class II molecules would be valuable for vaccine development and cancer immunotherapies. Current computational methods trained on in vitro binding data are limited by insufficient training data and algorithmic constraints. Here we describe MARIA (major histocompatibility complex analysis with recurrent integrated architecture; https://maria.stanford.edu/ ), a multimodal recurrent neural network for predicting the likelihood of antigen presentation from a gene of interest in the context of specific HLA class II alleles. In addition to in vitro binding measurements, MARIA is trained on peptide HLA ligand sequences identified by mass spectrometry, expression levels of antigen genes and protease cleavage signatures. Because it leverages these diverse training data and our improved machine learning framework, MARIA (area under the curve = 0.89-0.92) outperformed existing methods in validation datasets. Across independent cancer neoantigen studies, peptides with high MARIA scores are more likely to elicit strong CD4+ T cell responses. MARIA allows identification of immunogenic epitopes in diverse cancers and autoimmune disease.

Infection with Burkholderia pseudomallei - immune correlates of survival in acute melioidosis

Melioidosis, caused by Burkholderia pseudomallei, is a potentially lethal infection with no licensed vaccine. There is little understanding of why some exposed individuals have no symptoms, while others rapidly progress to sepsis and death, or why diabetes confers increased susceptibility. We prospectively recruited a cohort of 183 acute melioidosis patients and 21 control subjects from Northeast Thailand and studied immune parameters in the context of survival status and the presence or absence of diabetes. HLA-B*46 (one of the commonest HLA class I alleles in SE Asia) and HLA-C*01 were associated with an increased risk of death (odds ratio 2.8 and 3.1 respectively). Transcriptomic analysis during acute infection in diabetics indicated the importance of interplay between immune pathways including those involved in antigen presentation, chemotaxis, innate and adaptive immunity and their regulation. Survival was associated with enhanced T cell immunity to nine of fifteen immunodominant antigens analysed including AhpC (BPSL2096), BopE (BPSS1525), PilO (BPSS1599), ATP binding protein (BPSS1385) and an uncharacterised protein (BPSL2520). T cell immunity to GroEL (BPSL2697) was specifically impaired in diabetic individuals. This characterization of immunity associated with survival during acute infection offers insights into correlates of protection and a foundation for design of an effective multivalent vaccine.