STRUCTURAL DETERMINANTS AFFECTING THE OLIGOMERIZATION TENDENCY OF SOME PANCREATIC RIBONUCLEASES

The three dimensional domain swapping (3D-DS) mechanism represents a useful strategy that proteins can follow to self-associate. This mechanism requires the presence of a flexible portion that generally links the N- and/or C-terminus of the protein to its core. This portion, called hinge loop, can adopt different conformations as a function of the environmental conditions and allows the mentioned N- and/or C-terminal domains to be detached from the protein core and to be swapped with an identical domain of another protomer. This induces the formation of dimer(s) or larger oligomer(s) that often display, or enforce, biological properties that are absent, or attenuated, in the native monomer. The 3D-DS mechanism is shared by some proteins involved in amyloidosis, such as the human prion protein, β2-microglobulin, or some cystatins. Although not being amyloidogenic, the “pancreatic-type” RNases, i.e. resembling the features of the well-known pancreatic bovine RNase A, have become milestones to comprehend the determinants ruling out the 3D-DS mechanism. This is true especially for RNase A and for its natively dimeric homolog bovine seminal (BS)-RNase, whose structural determinants characterizing their self-association through 3D-DS have been deeply studied in the recent past. Moreover, also other ribonucleases included in the same RNase A super-family can oligomerize through this mechanism, as for example onconase (ONC), that can be induced to dimerize through the 3D-DS of its N-terminal domain. In this thesis, my aim was to elucidate some structural determinants that settle the tendency of some pancreatic-type RNases to self-associate, keeping the known features of RNase A as a reference. Therefore, we firstly investigated if the different methods that can be used to obtain the oligomers of RNase A might influence their properties. We treated RNase A with two different protocols: in particular, i. lyophilization of acetic acid solutions of the protein, or, ii. thermal incubations performed at 60°C in 20 or 40% aqueous ethanol. We observed that the enzymatic activity of the monomers and of the N- and C-swapped dimers (ND and CD, respectively) are slightly affected by the particular method used to induce their self-association. Then, I focused my attention to onconase (ONC), an amphibian member of the “pancreatic-type” RNase super-family lead by RNase A. In particular my aim was to unlock its C-terminal domain, that in the wild-type is blocked by a disulfide bond involving its C-terminal Cys104 residue, in order to investigate if this RNase variant may dimerize also through the C-terminal-end swapping. Unfortunately, none of numerous ONC mutant that I produced could form either the C-dimer, or also larger oligomers. In addition, all these variants formed less N-dimer than the wild type, confirming the actual mutual influence existing between the N- and C-termini of the protein, as it occurs for other RNases. We also supposed that the impossibility for ONC to form a C-dimer could be ascribable to the elongation of its C-terminal domain, with respect to RNase A. Therefore, we analyzed also the aggregation propensity of human pancreatic RNase, a variant that displays a C-terminus elongation of four AA residues in comparison with RNase A, similarly to ONC. Both the wt and a mutant obtained by the deleting the four mentioned residues displayed SEC profiles qualitatively very similar to the one of RNase A. Therefore, the human variant actually displayed that it can extensively oligomerize, but we could also deduce that the C-terminal elongation influences only marginally this process. The last project I developed was focused on the analysis of the oligomerization tendency of human angiogenin (ANG), a 14 kDa RNase variant characterized by a low ribonucleolytic activity (10-5/10-6 fold less than of RNase A), however necessary for its crucial angiogenic effects. ANG is involved in tumorigenesis but it exerts also a survival-promoting effect on the central nervous system (CNS) neuronal progenitors. However, some ANG variants are involved in neurodegenerative diseases, such as Amyotrophic Lateral Sclerosis (ALS) and Parkinson Disease (PD). ALS is a multifactor disease, but one hypothesis to explain the pathogenic effect could be related to a possible oligomerization and precipitation of the mutants in the CNS. Among the numerous pathogenic ANG variants existing, one candidate retained by some scientists prone to undergo self-association is the S28N mutant. In our hands, this variant showed to dimerize at a slightly higher extent than the wild type, that in turn dimerized as well. Then, other ANG pathogenic mutants, in particular H13A and Q117G, also showed to dimerize, but with less reproducible results. Anyway, we detected for the first time that also ANG can dimerize through the 3D-DS mechanism, as many other RNases do. In conclusion, all the results of this thesis can be considered a step forward to comprehend the determinants settling the 3D-DS dimerization, or oligomerization tendency of many pancreatic-type RNases, and to compare the determinants that induce the differences emerging within them.