Generation and Shaping of The Human B cell Receptor Repertoire

This thesis describes novel features in the generation and shaping of the primary B cell receptor repertoire. By high-throughput sequencing, we have obtained 9969 unique DJH-rearrangements and 5919 unique VHDJH-rearrangements obtained from CD19+ cells from 110 healthy adult blood donors. We found that DJH-rearrangements and non-productive VHDJH-rearrangements had many features in common but differed significantly in their use of D-genes. During D- to JH-gene rearrangement, the D-genes proximal to the JH-locus were rearranged more frequently than JH-locus distal D-genes while VH-locus proximal D-genes were observed more frequently in non-productive VHDJH-rearrangements than VH-locus distal D-genes. We further demonstrated that the distance between VH-, D- and JH-gene segments influenced their ability to rearrange within the human immunoglobulin locus.

In the attempt to identify structural differences in N regions we studied the nucleotide composition of N regions within 668 unique and unmutated non-productive human VHDJHrearrangements and 9832 unique DJH-rearrangements containing one D-gene. We demonstrated that the composition of N regions was highly compatible with a model in which N regions were generated by two concatenating 3’-overhangs and that the net result of N addition was an overweight of C in the 3’-overhangs of gene segments. During joint formation this surplus generated a positive G/C ratio gradient across the N regions. Surprisingly, this G/C gradient differed in N regions between VH-D and D-JH gene segments and was highly dependent on gene segment trimming.

Mouse studies have suggested that B cells with a non-productive VHDJH-rearrangement or autoreactive antigen receptors may be rescued by performing a secondary rearrangement called VHgene replacement. The presence of a cryptic RSS (5’-TACTGTG-3’) located at the 3’-end of ~80% of all VH-genes enables an upstream positioned VH-gene to replace the VH-gene within an already rearranged IgH locus. After VH-gene replacement, an identifiable five to ten nucleotide footprint may remain as a remnant of the first VH-gene. Demonstrating the contribution of VH-gene replacement in the generation of the human B cell receptor repertoire has however been controversial. We have previously failed to demonstrate a significant contribution of VH-gene replacement in humans while others have found them to occur at a frequency of 5%. Previous studies have been limited by the size of the datasets studied. In order to find evidence for rare occurring events such as VH-gene replacement it is vital to have large datasets. We have in this thesis investigated 29802 unique human VHDJH-rearrangements downloaded from NCBI for the presence of footprints after VH-gene replacement. We did not find any evidence of footprints after VH-gene replacement in N-additions ranging from 5-15 nucleotides in length in either mutated or unmutated VHDJH-rearrangements.

The recombination-activating genes (RAG) 1 and 2 are only expressed in lymphocytes and they are the main facilitators of V(D)J-rearrangement. Their mechanisms are similar to many other DNA binding and cleaving enzymes such as transposons that cut semi-randomly within the genome. Transposons have been domesticated and their functions have successfully been modified for the purpose of gene therapy. We have in this thesis explored the opportunity of using RAG1 and RAG2 to introduce a transgene into the human immunoglobulin heavy chain loci. These enzymes recognize specific recombination signaling sequences (RSS) that could act as potential integration sites for transgenes. We have designed and generated a prototype for a non-viral two-plasmid targeted gene intergrational system consisting of one vector containing both RAG1 and RAG2 (RAGVec1.0) and one vector containing the gene of interest, in this case RFP, (GeneVec1.0) flanked by 23-spacer RSS. We have used the model cell line HEK293T for testing the expression potential of both plasmids and achieved high mRNA expression of both RAG1 and RAG2 and demonstrated translation into protein for RAG1. We further demonstrated that RFP was expressed within 72 hours of transfection. Taking the so-far accumulated data together, it has not been possible to demonstrate targeted gene integration yet using the current vector design.