The aim of the proposal is the establishment of a yeast-three hybrid system for the identification of mycobacterial proteins that bind to selected small molecules. It is envisioned to use the system to (i) find targets of active molecules that have been identified in whole cell screenings; (ii) identify mycobacterial proteins that activate or deactivate drugs; (iii) establish simple assays for proteins previously identified as key targets. Since the identification of protein targets of a small molecule drug is challenging, improved methods are asked that facilitate their discovery. With the recent development of an optimised yeast three-hybrid system in the Johnsson group, the stage is set for its application to identify small molecule-protein interactions. This adapted version of the classical yeast two-hybrid system features an additional third synthetic hybrid ligand. It uses SNAP-tag (a self-labelling protein tag) as an anchor protein and O6-benzylguanine (BG) derivatives as synthetic ligands. Currently it allows the screening of human cDNA libraries for small molecule-protein interactions in yeast. To apply this system for drug discovery against tuberculosis, the construction of genomic libraries of M. smegmatis was initiated and libraries of M. tuberculosis will follow. We would start with conjugating known TB drugs (with known or unknown targets) to BG and screen them against genomic libraries. Suggested drugs to start are rifampicin and BTZ as positive controls and clofazimine and thiocarlide as drugs with unclear mechanism of action. Noteworthy that the three-hybrid screen can also be applied for the identification of proteins that activate a certain drug. Furthermore, once a protein-ligand is identified, the hybrid assay can be used to screen for new drugs against this target in HTS compatible competition assay.