The innate immune system is the human body’ first line of defense against invading pathogens. This defense mechanism recognizes PAMPs through PRRs. PRRs are divided into different families, including the TLR family. The TLR family comprises 10 human members that each recognize their own set of ligands. Upon ligand detection, adaptor proteins are recruited to the TLR and the intracellular signaling pathway becomes activated. Two possible signaling pathways are induced depending on the stimulated TLR. The first signaling pathway is triggered by the adaptor protein MyD88 and results in the activation of the NF-κB transcription factor family and the production of pro-inflammatory cytokines. The second signaling pathway is guided by the adaptor protein TRIF and induces IFN production. The interactions between receptor and adaptor proteins are crucial in both signaling pathways and are coordinated via a common domain, called the TIR domain. To date, multiple TIR domain structures are determined in both receptor and adaptor proteins. How TIR domains exactly interact, however, remains unclear. In addition, the activation of MyD88 is also intensively studied. On the one hand, MyD88 activation is controlled through phosphorylation of two serine residues in the MyD88 TIR domain via an unknown mechanism. Diverse mutations, on the other hand, cause uncontrolled MyD88 activation. These mutations were recently identified in different diseases, including lymphoma and arthritis. More insight in the function of MyD88 can be important for the development of therapeutic treatments. In this thesis, we studied the interactions of the adaptor proteins MyD88 and TRIF. First, we randomly mutated their TIR domains. Then, we analyzed the effect of each mutation on the binding of the mutated protein with its interaction partners via the MAPPIT technique. In this technique, the TIR domains of two interaction partners were coupled as bait and prey. The interaction between bait and prey induced a MAPPIT luciferase signal that correlates with the strength of the interaction. The mutational effect on the TIR domain structure was analyzed via the FoldX prediction server. Mutations that disturbed the protein structure were omitted for further analysis. Putative interaction sites in the TIR domains of MyD88 and TRIF were subsequently studied via Co-IP experiments and NF-κB and IRF3 activation assays. Our results confirmed known interaction sites and led to the discovery of novel interaction interfaces. This knowledge allowed us to generate computer models that display the interactions of MyD88 or TRIF with their interaction partners. We also illustrated how TIR domains of MyD88 form an oligomer complex, called the Myddosome. During this thesis, the cryo-EM structure of the Mal TIR domain was solved. This structure provided new insights in TIR-TIR interactions. We therefore decided to re-interpret our previous data about the Mal and TLR4 TIR domains and our current data about the MyD88 and TRIF TIR domains with the cryo-EM structure of the Mal TIR domain as a template. We were able to further clarify TIR-TIR interactions and to propose models for the TLR4/Mal/MyD88, TLR4/TRAM/TRIF, and TLR3/TRIF complexes. We also focused on the activation mechanism of MyD88. We mimicked MyD88 TIR domain phosphorylation via two serine-to-aspartate mutations (S242D and S244D). We also studied two specific mutations that induce MyD88 auto-activation. The first mutation, MyD88 L265P, is associated with different types of human B-cell lymphoma and in particular with WM. The second mutation, MyD88 S222R, is recently discovered in severe destructive arthritis. The effect of these four mutations on MyD88 interactions was extensively studied via MAPPIT experiments and NF- κB activation assays. In addition, the mutational effects on the structure of the MyD88 TIR domain were analyzed via MD simulations. These approaches resulted in unique insights in MyD88 TIR domain activation. Structural alterations caused by the mutations induced the active conformation of the MyD88 TIR domain and thus explained why MyD88 activation requires phosphorylation. The simulations also further clarified the uncontrolled MyD88 signaling in both WM and severe destructive arthritis. This knowledge might contribute to the development of new therapies.