Table of Contents:
SUMMARY Background: The primary and secondary immunoglobulin repertoires represent an important basis of the pathogen defence in the vertebrate. The mucosa-associated lymphoid tissue plays a crucial role in the establishment and maintenance of mucosal homeostasis against invasive infections. Inside mucosal tissues, Immunoglobulin A (Ig A) is the predominating immunoglobulin that is secreted into the blood plasma, as well as onto the mucous membrane. The differentiation of antigen-activated B-cells into plasma and memory cells allows a quicker and more effective immune response in case of a recurrent confrontation with the same pathogen. Basic mechanisms concerning the development of B-cell and antibody repertoires are still unclear. Objective: It remains uncertain which selective pressure acts upon the B-cell repertoire during the class switch recombination from IgM to IgA in germinal centres. Furthermore it is unknown which criteria drive the recruitment of lymphocytes into the pool of IgA producing memory and plasma cells. In this work, we tested the hypothesis that subpopulations of IgA producing memory and plasma cells are not recruited by chance, but on the basis of their antigen binding properties. Methods: IgM producing B-cells as well as IgA producing memory and plasma cells were separated from Peyer’s patches of four adult BALB/C wildtype mice using flow cytometry. Using reverse transcriptase-PCR, we selectively amplified and cloned IgM and IgA transcripts of the immunoglobulin heavy chains. The transcripts were sequenced and nucleotide sequences were compared with the IMGT databases of germline sequences (Alignment via IMGT VQuest). Results: We collected a total number of 413 immunoglobulin sequences in four independent experiments of which 323 (78%) were unique. The IgA sequences of memory and plasma cells showed a significantly lower clonal diversity than IgM sequences (74% and 72% vs. 94%; p <0.01). IgA sequences of memory and plasma cells contained more somatic mutations than IgM sequences (37‰ and 35‰ vs. 14‰, p <0.01). IgA transcripts from memory cells contained a lower number of random, so-called n-nucleotides within the CDR3-regions than IgA transcripts from plasma cells (p <0.01). In comparison to the IgM sequences, the IgA sequences from memory cells showed significantly shorter CDR3 regions (p = 0.02) and significantly shorter NDN regions (p=0.0002). A similar trend was observed with IgA plasma sequences; however this was not statistically significant. Moreover, we found a preference of single JH segments depending on the rearranged VH family. Discussion: IgA showed the typical signs of classical antigen selection. In comparison to IgM transcripts from the same tissue sample, the limited clonal diversity and the increased mutation rate of IgA memory and plasma cell transcripts document an antigen dependent focused selection during the class switch from IgM to IgA. Interestingly, CDR3-regions from IgA of memory cells and IgM were differently composed. As a consequence, IgA of memory cells showed a lower structural variety and, on average, a deeper antigen binding grove compared to IgM facilitating a closer interaction with the antigen. In contrast to earlier assumptions, the use of the VH family influences the JH utilization and the composition of the CDR3 region. Conclusion: IgA producing memory and plasma cells represent separate populations with differing repertoires of antigen binding sites inside their antibodies. They both show characteristic features of an individual antigen dependent selection. Our data demonstrate that the CDR3 region is an important selection factor during class switch recombination to IgA in germinal centers.