Periodic Reporting for period 2 - COSMIC (COmbatting disorders of adaptive immunity with Systems MedICine)
Reporting period: 2020-01-01 to 2022-09-30
EU countries face large health challenges to combat chronic diseases. COSMIC pursued a Systems medicine approach comprising the integration of laboratory and computational approaches to combat B-cell neoplasia (BCN) and rheumatoid arthritis (RA) both prototypical diseases originating from abnormal functioning of immune cells. These immune-related diseases pose a variety of challenges including (1) an incomplete understanding of the underlying molecular/cellular mechanisms, (2) uncertain diagnosis/prognosis and (3) variable treatment efficacies of existing drugs. This results in suboptimal patient management and huge costs for healthcare systems. In particular, COSMIC focussed on the incorrect functioning of the germinal center (GC) which contributes to the emergence of B-cell clones expressing autoreactive antibodies in RA, or showing a malignant behaviour in BCN. GCs are specialised anatomical sites for example found in lymph nodes and make part of the adaptive immune system. To determine the role of the GC in RA and BCN is challenging as it is not easily accessible for experimental investigation. Therefore, we developed new experimental approaches and computational models to gain more insight in the (patho)physiological cellular and molecular mechanisms underlying the (dis)function of the GC. This will contribute to the elucidation of the the GC reaction and is essential to understand the ontogeny and evolution of BCN and RA.
(1) We developed a methodology to isolate single GCs and determine their clonal B-cell composition. (2) To characterize GCs at the single cell level, it is crucial to analyze RNA from highly enriched viable cells. We developed an approach to isolate a fraction of murine GC B cells from lymph nodes of immunized mice to obtain a time-resolved assessment of the developmental trajectories imposed on B cells while transiting through the GC. (3) We have set up Oxford Nanopore Technologies RNA-sequencing, to analyze specific transcript isoforms. This allows the interrogation of the structure, sequence and abundance of RNA with its modifications, including splicing and polyadenylation features, for the in-depth characterization of GC B cells.
Advances in bioinformatics approaches
We (contributed to) developing and implementation of the following tools/approaches:
(1) A new tool for the detection of RNA isoforms using the Nanopore data. (2) The most up-to-date 3D prediction structure algorithms to construct 3D models: Studying antibodies in B cell lymphomas is important from both a biological and a clinical perspective. (3) A bioinformatics pipeline to analyze and visualize gene segment usage, clonotype diversity, phylogenetic clonal trees and SHM for human BCR repertoires. (4) A graph network-based approach (IgIDivA) for the BCR repertoire analysis of the intraclonal diversification to gain insight into the ontogeny and evolution of B cell clones in health and disease. (5) A cross-(multi) platform normalization of gene-expression microarray data (CuBlock) that allows metanalysis across different experiments. (6) A tool for HLA class II-binding peptides prediction to screen for potential human autoepitopes linked to different infectious agents. (7) AbSolution: to annotate BCR sequences with physiochemical properties, level of SHM, N-glycosylation sites, etc to facilitate their analysis.
Understanding the normal GC
The B-cell repertoires from single GCs gave important insight into the nature and variation of GCs in human, allowing analysis of the source and the fate of (non-)functional B-cell receptors (BCR). We could determine the effects of affinity-based B-cell selection in relation to specific mutations and gain insight into the functional convergence of dominant clones across GCs.
Understanding B-cell neoplasia
(1) Somatic hypermutation (SHM) of BCR genes is critical for the production of high-affinity antibodies but may lead to auto-antibodies (in RA) or BCN. We identified a mechanism that explains why SHMs are not repaired in the cell, which has important implications for our understanding of GC B-cell transformation in malignancies. (2) Using single-cell RNA sequencing of GC B cells, we identified transcriptional signatures associated with the two main functional subsets of GC B cells, which led to the identification of GC-derived diffuse large B-cell lymphomas originating from unique subsets of GC cells. This will help to determine the putative cell-of-origin of these tumors. (3) Our algorithms to construct 3D antibody protein structures were applied to antibodies from 925 chronic lymphocytic leukemia (CLL) patients providing insight in their function and helping development of a prognostic stratification scheme in CLL as part of diagnosis.
Understanding RA
(1) We established the receptor repertoire of autoreactive GC B cells in a mouse RA model leading to identification of B suppressor cells with predetermined collagen type II specificity. Different B-cell clones were characterized functionally regarding induction of arthritis, pain and bone erosions. Validation of a series of antibodies protecting against arthritis, may open a new avenue towards RA therapy. (2) Characterizing and phenotyping dominant B-cell clones in blood of individuals at-risk for developing RA showed they correspond most to clones found in the plasma cell compartment and less to those in the memory B-cells compartment. Moreover, dominant clones change over time and don’t have common BCR characteristics. (3) Nanopore sequencing gave insight in the post-transcriptional regulation in positively selected GC B-cells in human and mice. These data will help to understand the (defects in) molecular mechanisms of B cell selection and affinity maturation in autoimmune disease and B-cell malignancies.
Computational modelling
We developed (spatio)temporal dynamic computational models of the GC to increase our understanding of the normal GC reaction and in case of BCN and RA: (1) An agent-based model for the simulation of multiple GCs to allow the simulation of asynchronous initiation of GCs and study how different mechanisms can impact the GC shutdown, including mechanisms for the suppression of self-reactivity. (2) A multiscale GC model to study plasma cell differentiation, integrating an agent-based model representing cellular dynamics with a core gene regulatory network (CGRN) describing plasma cell differentiation. We applied this model to simulate different genetic aberrations underlying BCN, and facilitate the interpretation of BCR repertoires. (3) A stochastic model of the GC to understand the processes that underlie the emergence of clonal diversity and B-cell affinity maturation. (4) A tool for analysis and interpretation of single cell transcriptomic data (a stochastic model for the CGRN that is central to plasma cell differentiation).