Amaral, Margarida, 1958-Pissarra, Luísa Maria da Silva, 1977-2010-07-272010-07-272008http://hdl.handle.net/10451/1532Tese de doutoramento em Bioquímica (Genética Molecular), apresentada à Universidade de Lisboa através da Faculdade de Ciências, 2008Cystic Fibrosis (CF) is the most common genetic disease among Caucasians. This lifethreateningdisease is caused by mutations in the CFTR (cystic fibrosis transmembrane conductance regulator) protein, a chloride (Cl-) channel expressed in the apical cell membrane of epithelial cells, essential for the normal functioning of epithelia lining the airways, the intestinal tract, including pancreatic and bile ducts, the sweat glands and the male reproductive tract. Hallmarks of CF disease include exocrine pancreatic insufficiency, increased sweat NaCl concentration and male infertility, but the major cause of morbidity and mortality is pulmonary disease with chronic bacterial colonization. Lung infection and inflammation generate a vicious cycle leading to the progressive loss of lung function and premature death. CFTR, a member of the ATP-Binding Cassette (ABC) transporter superfamily, comprises two membrane spanning domains (MSD1 and 2), forming the pore, two cytosolic nucleotide binding domains (NBD1 and 2), which dimerize to bind and hydrolyze ATP, and a central regulatory domain (RD), unique to CFTR in the ABC family, which regulates channel activity upon phosphorylation. Despite the large number (over 1500) of mutations found in the CFTR gene of CF patients, a single one, F508del, in NBD1, accounts for approximately 70 % of CF chromosomes. The second and third most common mutations, occurring worldwide at frequencies slightly over 1 %, are G542X and G551D. Given the high incidence of F508del, most research and drug-development efforts have focused on this mutation: understanding its basic defect and how it initiates the CF pathological cascade, so as to design stategies to correct it. The most obvious cellular defect caused by the F508del-CFTR mutants is protein intracellular mislocalization due to its retention in the endoplasmic reticulum (ER), where it is synthesized as an immature core-glycosylated form, and subsequent degradation in the proteasome. F508del-CFTR is thus prevented from progressing along the secretory pathway and from insertion in the plasma membrane. The ER quality control, mostly composed by molecular chaperones, discriminates between native (folded) and non-native (misfolded) conformations, ensuring that only correctly folded proteins progress beyond the ER to theGolgi. It is plausible to envisage that the trafficking defect of F508del-CFTR may result from an impaired folding efficiency of the mutant NBD1 domain, in spite of the fact that folded F508del-NBD1 is thermodynamically nearly as stable as the wild-type domain. Nevertheless, in vivo, besides exhibiting defective processing, full-length F508del-CFTR is also more unstable than wt-CFTR and, when manipulated to be inserted into the cell membrane, it still displays a channel gating defect, suggesting it still maintains a distorted conformation. As the mutant still retains some residual activity, the search for pharmacological agents that promote F508del-CFTR folding and/or exit from the ER is currently a major venture in both academia and industry. These efforts have already come up with candidate lead compounds and drugs, some of which are already in phases 1 and 2 of clinical trials. Knowledge of a detailed 3D structure of a protein is important not only to understand its folding, function and specificity and the effects of disease-causing mutations at a molecular level, but is also a pre-requisite for rational drug design, allowing the identification of the best candidate docking sites for the binding of therapeutic compounds. To this end, several CFTR-NBD1 high-resolution structures have been published, including the crystal structure of human F508del-NBD1, which was shown to display only subtle differences relative to wt-NBD1. The first published structure of F508del-NBD1 was nevertheless that of a genetically engineered domain, where seven additional mutations, including known revertants of the F508del trafficking defect, had been added to improve domain solubility and hence ease crystallization. Subsequently, the high-resolution structures of new constructs, including solely either two (F494N-Q637R) or three (F429S-F494N-Q637R) solubilizing mutations, were published. The first objective of this doctoral work was to evaluate whether these double (F494NQ637R) or triple (F429S-F494N-Q637R) sets of mutations, included in the F508del-NBD1 crystal in order to increase its solubility, might also change the overall folding of full-length wt- and F508del-CFTR. To this end, we studied the properties of novel cell lines stably expressing wt- or F508del-CFTR in cis with these solubilizing mutations. Data presented here reveal that these mutations have in fact a revertant effect, by partially rescuing the trafficking of F508del-CFTR and also by contributing to the partial correction of its gating. Additionally, and in contrast to previously described F508del revertants, these mutations do not significantly alter the processing or gating of wt-CFTR, thus appearing to rescue CFTR in a F508del-specific manner. Altogether, these data suggest that the published crystal structure of F508del-NBD1 is that of a partially corrected NBD1. Therefore it is still reasonable to expect the resolution of the structure of F508del-NBD1 without any added mutations will identify changes caused by this deletion. Moreover, these data bring into focus the motifs where the solubilizing mutations lie as potentially interesting structural sites for docking of small molecules to correct the folding defect of F508del-CFTR. Subsequent to the conclusion of this work, a new F508del-NBD1 high-resolution structure of a construct without any solubilizing mutations, but lacking a 32-aminoacid stretch (including one of the residues involved in the solubilizing mutations), was described. This new structure again shows that F508del produces changes only in its vicinity. As the previous ones, this structure suggests that the most relevant impact of this deletion would be primarily on the interaction of NBD1 with other domains, thus affecting the full-length CFTR, rather than the NBD1 structure itself. The second objective of this work was to further understand how some missense mutations in NBD1 affect the processing of CFTR. Towards this goal we produced and studied, biochemically and functionally, engineered NBD1-CFTR mutants which have been detected in CF patients but had not been molecularly characterized. Two NBD1 mutations were studied, namely S549N, in the signature sequence, and Y563N, a buried residue close to the F508 residue. Despite its location in a loop involved in the ATP binding interface, S549N was found to have only a mild negative influence in the processing of CFTR, allowing for the appearance of functional channels at the cellular surface. Y563N, on the other hand, severely disrupts the processing of CFTR, likely affecting NBD1 folding in a different manner than F508del, as it could not be rescued by either G550E or 4K, two known revertants of F508del- CFTR. The third objective of this doctoral work was to study the mechanisms by which second-site mutations located in NBD1 achieve correction of F508del-CFTR defects. Understanding such rescuing mechanisms is also important for the identification of structural motifs and to suggest chemical structures for candidate small molecule correctors. Analysis of the combined effect of previously described individual F508del-CFTR revertants, namely G550E, R553M, R553Q and R555K, was performed. Our observations, indicating a cumulative effect, suggest that these mutations promote correction by different mechanisms or at different points of the NBD1 structure, although further studies are required to understand and discriminate between those mechanisms. The fourth objective here was to assess the effects of the two best characterized F508del-CFTR revertants, namely G550E and 4RK, on G551D, a rather frequent and extensively studied CF mutation. In a previous study by our group these mutations were proposed to act on F508del-CFTR by different mechanisms: while G550E seems to act directly on NBD1 folding, 4RK appears to act indirectly, by promoting exit of F508del-CFTR from the ER through abolishment of RXR retention motifs. G551D-CFTR channels are known to be properly processed and inserted in the cellular membrane but they hardly respond to ATP stimulation. Located in the periphery of NBD1, the G551 residue is part of the signature sequence, in one of the ATP-binding sites, and G551D is believed to perturb ATP binding and hydrolysis, interfering with NBD dimerization and thereby deeply affecting the gating of the resulting CFTR channels. We produced novel cell lines expressing CFTR with G551D in cis with either G550E or 4RK. Analysis of these novel G551D mutants shows that the gating defect of G551D-CFTR is more efficiently corrected by 4RK than by G550E, suggesting that the 4RK set of mutations may also have a direct structural impact on the mutant domain in addition to their known impact on trafficking, thus indicating that the effect of revertants is mutation-specific.application/pdfengGenética molecularTeses de doutoramentodoctoral thesis