Structural versatility of BS-RNase: different oligomeric isomers formed through 3D domain swapping of N- and C-termini

Bovine seminal RNase (BS RNase) is the only naturally dimeric member of the pancreatic-type RNase superfamily. The enzyme is a mixture of two isoforms: 1) M=M, dimeric through two antiparallel disulfides occurring between the two subunits; 2) MxM (70% of the total), characterized by the swapping of the N-termini besides the mentioned disulfides1. When dissolved in 40% acetic acid and subjected to lyophilisation, BS RNase forms a mixture of oligomeric aggregates2, as does RNase A, the proto-type of the super-family3. However, while the oligomers of the pancreatic variant have been extensively characterized3-5, the multimers of BS RNase6 are presently less known. To deepen their characterization we induced BS RNase self-association by using the same conditions used with RNase A, i.e. lyophilisation of enzyme solutions at low pH, or thermally-induced aggregation of concentrated protein solutions in various media. The multimers obtained were analyzed by SEC, cation-exchange chromatography, cross-linking, electrophoresis, proteolysis, and enzymatic assays. The main results obtained are: - BS-RNase forms several 3D domain-swapped conformers, in particular at least two tetramers; one is probably a totally N-terminal-swapped isoform4, while the other contains a C-terminal swapping. - The C-swapped tetramer is less stable than the N-swapped one. - BS RNase multimers larger than tetramers appear to be a mixture of various isoforms, similarly to what occurs with RNase A3. These multimers seem to contain, again, a C-terminus swapping. - The validity of the results reported have been strengthened by the studies performed with a K113N BS RNase mutant.