Immuno-wars: attack of the clones

In the autoimmune disease rheumatoid arthritis (RA) the immune system is attacking our own body. Bioinformatician Barbera van Schaik (Amsterdam UMC) developed data analysis methods to characterize specific immune cell receptors. This makes it possible to identify disease-specific T cells that may be targeted to temper the immune response in RA.  

Human blood with red blood cells, T cells (orange) and platelets (green) - foto Zeiss Microscopy

The immune system is a complex system that protects us from a wide range of invading antigens - molecules that stimulate an immune response, e.g. pathogens such as bacteria and viruses. Many different immune cell types are involved in fighting the invaders. The cells of the so-called innate immune system can act quickly but without a targeted attack. In contrast, the adaptive immune system reacts slower but can eliminate intruders using a very specific and, therefore, much more effective strategy. T cells and B cells are white blood cells that play an important role in the adaptive immune response. These cells express T-cell receptors (TCR) and B-cell receptors (BCR) on their surface to recognize specific pathogens. When a T/B-cell recognizes a specific antigen via its receptor the cell starts to proliferate to generate many copies of itself. This army of T/B-cell clones ensures a robust and targeted clearance of the antigen. 

To recognize the large variety of foreign antigens each individual T/B-cell carries a unique receptor with unique specificities. This unique specificity is created by random recombination of genes that encode the receptor. Moreover, extra variation is introduced by random removal and addition of nucleotides (the building blocks of DNA) and, for B cells, also by the introduction of mutations during the proliferation. Therefore, each single receptor sequence defines a unique T/B-cell able to recognize a specific antigen.

Attack on our own body

In autoimmune diseases such as RA the adaptive immune systems turns against our own body and starts to attack specific non-foreign body cells (autoantigens). In RA this eventually leads to the destruction of joints. Since 2007 the group of prof. dr. Niek de Vries (Amsterdam UMC) uses next generation sequencing to study the role of T/B cells in RA. Laboratory protocols were developed to sequence T/B-cell receptor of the cells present in blood or synovial tissue. The identification of unique TCR or BCR sequences and their frequency provides insight in the changes or development of T/B-cell repertoires in different tissues. This methodology, largely used in the immunology field, is called Repertoire Sequencing or Adaptive Immune Receptor Repertoire Sequencing (AIRR).

Bioinformatician Barbera van Schaik who works in the group of prof. dr. Antoine van Kampen (Amsterdam UMC) developed specific data analysis methods to characterize and quantify BCR and TCR repertoires. “In the past 10 years around 100 sequencing runs have been performed with 100 to 200 samples on each single run. Although the resulting repertoire data is limited in size, it takes about 5 days to analyse all sequences from a single sequencing run. Datasets are frequently re-analysed once improvements are made or additional analysis methods are developed.”

Analysing and sharing data

Barbera van Schaik

In this work Van Schaik uses multiple SURFsara services to facilitate the data analysis and the sharing of raw and analysed data. “We use the Research Drive to share data between project participants and with collaborators in other institutes. The HPC Cloud is used to support our data analysis. For every run, descriptions of all samples (metadata) are shared via the Research Drive. In this metadata document we describe for each sample the type of receptor, the organism from which the sample was taken (e.g. human or mouse), and to which project the sample belongs.”

In the first step of the repertoire analysis the metadata is converted into JSON format, divided into pieces, and uploaded to the server. Subsequently, multiple virtual machines are started on the HPC Cloud. Each machine first updates the scripts for the data analysis from a repository and then requests a job from the server. The data is automatically downloaded from the Research Drive, and after successful analysis the results are uploaded again together with a description about the analysis, such as software versions and parameter values. This process is repeated until there are no more jobs on the server. As soon as the analysis of an entire run is complete the data is further analysed by the researchers.

Results and further experiments

“There is clear evidence that T cells play a role in the origin and development of RA. Identification of specific pathogenic T cells may allow selective and highly effective targeting without having severe side effects. However, such selective targeting may only be feasible if the same T cells dominate the immune response at different sites of inflammation. Using T-cell repertoire sequencing we showed that in individual RA patients a limited number of specific T cells dominate the immune response in the inflamed joints. The limited breadth of the T-cell response in synovial tissue of the individual RA patient indicates that development of immunotherapies that selectively modulate dominant T-cell responses might be feasible.”

“In 2018 our groups initiated COSMIC (Combatting Disorders of Adaptive Immunity with Systems Medicine;, which is a Marie Curie European Training Network funded from the European Union's Horizon 2020 research and innovation programme. SURFsara is partner organisation in COSMIC. In this project we will develop computational models and experimental approaches to gain further knowledge about the adaptive immune system.”


Anne Musters, Paul L. Klarenbeek, Marieke E. Doorenspleet, Giulia Balzaretti, Rebecca E. E. Esveldt, Barbera D. C. van Schaik, Aldo Jongejan, Sander W. Tas, Antoine H. C. van Kampen, Frank Baas and Niek de Vries (2018) In Rheumatoid Arthritis, Synovitis at Different Inflammatory Sites Is Dominated by Shared but Patient-Specific T Cell Clones. J. Immunol . 201(2), 417-422