The folding of proteins into 3D structures is essential for their function and for life. Because malfunction of proteins causes disease, many therapeutics modify protein structure and function. There is, however, a large need for new and improved drugs. Understanding protein folding and its relationship with function will support the development of novel therapies. A large fraction of (potential) drug targets are transmembrane proteins. In this thesis we therefore aimed to improve understanding of multidomain multispanning transmembrane protein folding and folding-function relationships using CFTR as a model protein. CFTR consists of two multispanning transmembrane domains, two cytosolic nucleotide-binding domains, and a largely unstructured regulatory region. The protein is the best-studied member of the large ABC transporter family and forms a chloride channel across the plasma membrane. Misfolding of mutant CFTR causes the lethal disease cystic fibrosis (CF). As all transmembrane proteins, CFTR has parts that need to fold in the ER lumen, in the membrane, and in the cytosol. This compartmentalization increases the complexity of the folding process because each of these compartments provides a different folding environment and the protein domains need to assemble together to obtain the final 3D structure. In this thesis we investigated the hierarchy and coherence of domain folding and assembly, the structural requirements for export from the ER, folding-function relationships of CFTR and the modes of action of the current FDA-approved drugs.
Using radiolabeling and limited proteolysis folding assays we found that the domains of CFTR assemble largely in the order of synthesis. We found that all domains of CFTR and the N-terminus, which folds around the TM helices of TMD1 and TMD2, are essential for assembly of CFTR. Assembly barely changes the structure of the individual domains, which will have obtained near-native conformations at the time of assembly. We found that defects in many aspects of CFTR folding, including domain assembly steps, are detected by the quality control system, but not all mistakes are recognized. This leads to the expression of misfolded and nonfunctional proteins at the cell surface. The structure of CFTR is conserved throughout all type I ABC exporters, and thus our results may be translatable to the other type I ABC exporters.
Because many cystic fibrosis-causing mutants of CFTR have assembly defects, an effective therapy for cystic fibrosis could be developed with assembly correcting compounds. Because current correctors do not repair assembly of TMD1,
TMD2 and NBD2, especially compounds that can amend this would be beneficial. Our work suggests that for the design of new correctors, CFTR’s long N-terminus and ICLs should be considered as targets. For successfully treating patients with the most occurring mutant F508del-CFTR, the assembly-correcting compounds are best combined with NBD1 correctors. Our results additionally indicate that reducing the translation rate may be considered as a therapy for folding diseases.