Reduction Of Immune Repertoire Diversity Through DNA Sequence Constraints At VDJ Junctions

B and T lymphocytes are effector cells of the adaptive immune system. These cells express surface receptors that bind a huge variety of antigens, and together they comprise a person’s immune repertoire. A diverse repertoire is essential for mounting robust immune responses against a wide range of pathogens, and repertoire diversity affects the probability that DNA sequencing can uniquely tag a clonally expanded population of cells for the detection of minimum residual disease (MRD) during cancer treatment.

Immune repertoire diversity arises partly through the combinatorial splicing of gene segments from the variable (V), diversity (D), and joining (J) regions of a B or T cell receptor locus. Much additional diversity is created through the stochastic insertion and deletion of nucleotides at the splice junctions, and by somatic hypermutation (SHM) in maturing lymphocytes. The generation of junctional diversity is an important part of this process, but it may be constrained by the underlying biological mechanisms.

To explore the landscape of junctional diversity among immune receptor loci, we developed a likelihood model that can annotate VDJ junctions in the presence of SHM and compute the probability that a given receptor sequence was generated only once in a person’s repertoire, which is essential for tracking MRD. Using high-throughput sequencing data from several individuals and a range of receptor loci, we identify mechanistic constraints that limit B and T cell receptor diversity. For example, we show that the usual variability in CDR3 length is reduced at the immunoglobulin kappa (IgK) locus, and we connect this finding to sequence motifs that constrain nucleotide deletion at the ends of IgK gene segments. Our findings will inform future genetic studies of the adaptive immune system, and they provide quantitative guidance for deciding which cancer clones can be tracked for reliable MRD detection.