Masterprüfung mit Defensio, Maximilian von der Linde

B-cells diversify their antibody repertoire to specifically react to invading antigens by two mechanisms of antibody maturation: Somatic hypermutation (SHM) and class- switch recombination (CSR). For both, activation-induced deaminase (AID) which deaminates cytosines to uracils acts on target immunoglobin (IG) loci and downstream repair mechanisms lead to manifested mutations in the regions of the antibody interacting with antigens or new effector functions, thereby supporting the adaptive immune system. AID also hits non-IG genes, so called AID off-targets, in a process called aberrant somatic hypermutation (aSHM) which occurs at low frequencies but might cause lymphomas if oncogenes are affected. In general, both replication and transcription mechanisms and regulation thereof has been implicated in AID targeting which needs accessible single stranded DNA (ssDNA), less is known about why some transcribed genes are targeted and others are not. In this project, we used a transgenic version of the human Burkitt’s lymphoma cell line Ramos that allowed us to constitutively activate overexpression and shuttling of AID into the nucleus where it can access DNA with greater frequency than wild-type AID to extract information about AID off-targets. We performed a whole-genome sequencing (WGS) of two samples, an untreated day 0 sample and a treated sample with AID overexpression over 16 days, so that the day 0 sample could be used for the subtraction of germline mutations. We aimed to identify somatic AID induced mutations and compare them to the transcriptional profile. First results from common somatic mutation callers showed that they were unable to reliably extract positions of AID-dependent mutations. I concluded that the used biological background made the WGS data unsuitable to find AID off-targets with classical somatic mutation calling because the effect of AID in the day 16 sample was too weak and was further diminished by WT AID in the day 0 sample to be detected by the coverage used in WGS. Consequently, I developed a custom pipeline to extract the mutation rate at individual positions and compare AID on target sites (within the IGH locus) with off-target sites. Using this custom approach, I found differences in AID targeting at a regional level by averaging mutation rates and could now find AID- related effects where traditional mutation callers could not. We hope to use this information on the whole genome and thus enable a comparison the transcriptome of Ramos in the future.