Susceptibility of Mutant SOD1 to Form a Destabilized Monomer Predicts Cellular Aggregation and Toxicity but Not In vitro Aggregation Propensity

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by the rapid and progressive degeneration of upper and lower motor neurons in the spinal cord, brain stem and motor cortex. The first gene linked to ALS was the gene encoding the free radical scavenging enzyme superoxide dismutase-1 (SOD1) that currently has over 180, mostly missense, ALS-associated mutations identified. SOD1-associated fALS patients show remarkably broad mean survival times (<1 year to ~17 years death post-diagnosis) that are mutation dependent. A hallmark of SOD1-associated ALS is the deposition of SOD1 into large insoluble aggregates in motor neurons. This is thought to be a consequence of mutation induced structural destabilization and/or oxidative damage leading to the misfolding and aggregation of SOD1 into a neurotoxic species. Here we aim to understand the relationship between SOD1 variant toxicity, structural stability, and aggregation propensity using a combination of cell culture and purified protein assays. Cell based assays indicated that aggregation of SOD1 variants correlate closely to cellular toxicity. However, the relationship between cellular toxicity and disease severity was less clear. We next utilized mass spectrometry to interrogate the structural consequences of metal loss and disulfide reduction on fALS-associated SOD1 variant structure. All variants showed evidence of unfolded, intermediate, and compact conformations, with SOD1G37R, SOD1G93A and SOD1V148G having the greatest abundance of intermediate and unfolded SOD1. SOD1G37R was an informative outlier as it had a high propensity to unfold and form oligomeric aggregates, but it did not aggregate to the same extent as SOD1G93A and SOD1V148G in in vitro aggregation assays. Furthermore, seeding the aggregation of DTT/EDTA-treated SOD1G37R with preformed SOD1G93A fibrils elicited minimal aggregation response, suggesting that the arginine substitution at position-37 blocks the templating of SOD1 onto preformed fibrils. We propose that this difference may be explained by multiple strains of SOD1 aggregate and this may also help explain the slow disease progression observed in patients with SOD1G37R.