Funktionelle Charakterisierung des DNA-Sensors DDX41 und des Mediators STING, die an der angeborenen Immunantwort auf zytosolische DNA und zyklische Dinukleotide involviert sind
Zusammenfassung der Projektergebnisse
DNA recognition in the cytosol involves a number of sensors and activates interferon production. STING is a transmembrane protein, which is central to this process. STING binds 2´3´-cGAMP, a native second messenger molecule produced in response to DNA detection, as well as cyclic dinucleotides (CDNs) of bacterial origin. A second protein that has been shown to be involved in STING-dependent activation of interferon signaling in response to cytosolic DNA and CDNs is DEAD-box protein DDX41. We aimed to use biochemical methods to characterize DDX41´s interaction with DNA, CDN and STING. We encountered difficulties in reproducibly expressing sufficient amounts of DDX41 and observed that the protein has a tendency to aggregate even in the presence of different solubility tags. We could determine that DDX41´s DEAD domain binds DNA and other domains weaken this interaction. However, despite our efforts, we were not able to detect any CDN binding by DDX41. The difficulties we encountered in handling DDX41 led us to concentrate our efforts on our second aim, the mechanism of STING activation upon CDN binding. STING consists of an N-terminal transmembrane connected via a flexible linker to its C-terminal cytosolic domain, which includes a dimerization motif. In vitro STING had always until now been found in a dimeric form. Crystallographic studies revealed that the STING dimer undergoes a conformational change upon cGAMP binding, which is correlated to its ability to activate downstream response. We could determine that STING dimer dissociates at low nM concentrations. We showed that C148 located in the flexible linker is important for dimer and thermal stability. We then used single cysteine mutants to attach fluorescent dyes and probe the conformational changes that STING’s cytosolic domain undergoes in solution upon cGAMP and CDN binding. We observed that STING dimer dissociated at single molecule (pM) concentrations. Our efforts to stabilize the dimer by N-terminal dimerization tags were unsuccessful. When cGAMP or bacterial CDN cyclic-di-GMP (cdG) were present, STING dimer was stabilized. We could observe that these two ligands stabilize two STING different conformations in solution, a closed one in the presence of c-GAMP and an open one in the presence of cdG, confirming findings by other groups using X-ray crystallography. Using the same approach, we could then compare the effect CDN binding has on different STING natural variants and observed that these vary both in the ability of ligands to induce dimer stability and the respective stabilized conformations. Our findings suggest that ligand induced dimerization of STING could be the first step in the activation of the protein. In addition to developing a smFRET assay to monitor conformational changes of STING’s CDN binding domain upon ligand binding, we also developed a toolbox with which we could monitor the position the C-terminal tail (CTT) of STING. However, due to time limitations only preliminary data was obtained.
