The Mediator of transcriptional regulation is the central coactivator that enables a response of RNA polymerase II to activators and repressors. It is conserved from yeast to human and consists of 25 subunits in yeast that are organized in four modules called head, middle, tail, and CDK8/Cyclin C module. Despite its central role in transcription the functional mechanism remains enigmatic. To overcome the lack of detailed structural data on the Mediator a recombinant expression system was established that allows large-scale purifications of Mediator head module subcomplexes. It has been shown that via limited proteolysis assays and multicistronic expression the problems of insolubility and low expression rates of Mediator subunits can be overcome, paving the way for structural studies on subcomplexes of the Mediator head module. First data indicated that a reconstitution of the complete head module is within close reach. Large-scale copurification data led to a detailed interaction map of subunits and subcomplexes from within the head module and towards the middle module. The second part of this work describes the structure solution of a subunit in the CDK8/Cyclin C module – Cyclin C. Cyclin C binds the cyclin-dependent kinases CDK8 and CDK3, which regulate mRNA transcription and the cell cycle, respectively. The crystal structure of Cyclin C reveals two canonical five-helix repeats and a specific N-terminal helix. In contrast to other cyclins, the N-terminal helix is short, mobile, and in an exposed position that allows for interactions with proteins other than the CDKs. A model of the CDK8/Cyclin C pair reveals two regions in the interface with apparently distinct roles. A conserved region explains promiscuous binding of cyclin C to CDK8 and CDK3, and a non-conserved region may be responsible for discrimination of CDK8 against other CDKs involved in transcription. A conserved and Cyclin C-specific surface groove may recruit substrates near the CDK8 active site. Activation of CDKs generally involves phosphorylation of a loop at a threonine residue. In CDK8, this loop is longer and the threonine is absent suggesting an alternative mechanism of activation is discussed based on a CDK8-Cyclin C model.