A hyperthermophilic protein G variant engineered via directed evolution prevents the formation of toxic SOD1 oligomers

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disorder characterized by selective death of motor neurons in the brainstem, motor cortex, and spinal cord, leading to muscle atrophy and eventually to death. It is currently held that various oligomerization-inducing mutations in superoxide dismutase 1 (SOD1), an amyloid-forming protein, may be implicated in the familial form of this fast-progressing highly lethal neurodegenerative disease. A possible therapeutic approach could therefore lie in developing inhibitors to SOD1 mutants. By screening a focused mutagenesis library, mutated randomly in specific “stability patch” positions of the B1 domain of protein G (HTB1), we previously identified low affinity inhibitors of aggregation of SOD1_G93A and SOD1_G85R mutants. Herein, with the aim to generate a more potent inhibitor with higher affinity to SOD1 mutants, we employed an unbiased, random mutagenesis approach covering the entire sequence space of HTB1 to optimize as yet undefined positions for improved interactions with SOD1. Using affinity maturation screens in yeast, we identified a variant, which we designated HTB1_M3, that bound strongly to SOD1 misfolded mutants but not to wild-type SOD1. In-vitro aggregation assays indicated that in the presence of HTB1_M3 misfolded SOD1 assembled into oligomeric species that were not toxic to NSC-34 neuronal cells. In addition, when NSC-34 cells were exposed to misfolded SOD1 mutants, either soluble or preaggregated, in the presence of HTB1_M3, this inhibitor prevented the prion-like propagation of SOD1 from one neuronal cell to another by blocking the penetration of SOD1 into the neuronal cells.