SOD1 Trimers are Toxic to Motor Neurons

Trimeric SOD1 is toxic to motor neurons, and mutations that stabilize the trimeric form, are more toxic than ones that destabilize it, according to a study published in the December 30 Proceedings of the National Academy of Sciences online. These findings provide support to the view that soluble oligomers, rather than insoluble aggregates, are the toxic species in many neurodegenerative diseases, and suggests that destabilization of trimeric species may be therapeutic in amyotrophic lateral sclerosis (ALS) caused by mutations in superoxide dismutase (SOD1).

High-speed atomic force microscopy reveals that under stabilizing conditions (pH 3.5), SOD1 oligomers feature a tripartite structure. Image courtesy of N. Dokholyan.

High-speed atomic force microscopy reveals that under stabilizing conditions (pH 3.5), SOD1 oligomers feature a tripartite structure. Image courtesy of N. Dokholyan.

The study combined high-speed atomic force microscopy (HS-AFM) to directly image SOD1 trimers, structural molecular simulations to establish the trimeric structure, and rational design of stabilized and destabilized mutants to test toxicity of trimers in cell culture.

“It has yet to be shown that amyloid aggregates are toxic,” noted senior author Nikolay Dokholyan of the University of North Carolina at Chapel Hill. “In contrast, evidence has mounted that oligomeric species are toxic in ALS, Alzheimer’s disease, and other neurodegenerative diseases” (see Sept 2014 news; Dec 2015 news). In ALS, SOD1 has also been shown to form soluble oligomers, but the precise nature of the oligomers has been unclear. Their small size and transient nature prevents the use of traditional tools for structural determination, Dokholyan said.

To overcome that limitation, first author Emily Proctor worked with co-author Yuri Lyubchenko of the University of Nebraska Medical Center in Omaha to capture images of the oligomers using HS-AFM. Metastable SOD1 oligomers were stabilized at low pH, and volumetric determinations of both the monomer and the oligomer were made from the images. The monomer’s volume was about 25 cubic nanometers, while the oligomer’s was about 72 cubic nanometers. The evidence for a 3:1 ratio was further supported by direct visual imaging of the oligomer, in which three regions of density could be seen clustering together. As the pH rose, these trimers dissociated back into monomers. “This was about five years of work,” Dokholyan noted wryly.

How the SOD1 monomers linked together in a trimer structure remained unclear. To explore this issue, the researchers conducted limited proteolysis, in order to cut only those sites that were accessible to solvent, and therefore were not involved in protein-protein interactions. The most common form of SOD1 is as a dimer, and so one possibility was that a third unit simply linked onto a preexisting dimer. If that were so, then proteolysis should not occur at the dimer interface. But that is exactly what occurred.

“We concluded from the cleavage pattern that the interface(s) between monomers in the SOD1 trimer must be located in a different region of the monomer from the native SOD1 dimer interface,” the authors wrote, indicating that the protein undergoes a significant change in quaternary structure in order to adopt the trimeric conformation. “SOD1 undergoes major, dramatic conformational changes in this rearrangement”, Dokholyan said.

First author Emily Proctor - Image courtesy of N. Dokholyan.

First author Emily Proctor – Image courtesy of N. Dokholyan.

The authors turned next to dynamic molecular modeling, allowing the monomers in their simulations to interact and reach a metastable state, with the restriction that the proteolytic cut sites remain exposed. In their derived model, residues of the dimer interface are separated and exposed to solvent, while new contacts contribute to the stability of the trimer. A newly discovered epitope of an antibody (C4F6) that binds disease-linked SOD1 species (Rotunno et. al., 2014) was exposed on the surface of the trimer in the model, suggesting a structural mechanism for its preference for binding oligomers rather than monomers or dimers and further supporting the model.

Senior author Nikolay Dokholyan Image courtesy of N. Dokholyan.

Senior author Nikolay Dokholyan Image courtesy of N. Dokholyan.

To verify the model, the authors next created a series of SOD1 monomers with mutations introduced into the predicted trimer interface regions. Mutations were chosen for their ability to stabilize, or destabilize, the trimer. Destabilizing mutants formed large aggregates, “supporting the hypothesis that large, insoluble aggregates, such as fibrils, lie on a different, competing pathway from the formation of small, soluble aggregates,” the authors wrote in the paper.

To test the effect of trimerization on cell viability, the team introduced their mutations into NSC-34 cells, a motor neuron-like cell line. After three days, cell death in stabilizing mutants matched or exceeded that from the known ALS-causing mutant A4V, while cells with destabilizing mutants survived. “It turned out that all the mutants that stabilized the trimers killed cells very rapidly,” Dokholyan said, “while those in which we destabilized the trimers lived as long as wild-type cells. That suggested to us that the stability of the trimer is critical for cell viability.”

But is trimerization a key step in SOD1-linked ALS pathogenesis? Evidence is, as yet, indirect. “Nearly all of the known ALS-linked SOD1 mutations affect the stability of monomer and/or the dimer,” Dokholyan said, “which excluded them from our study due to the additional complications of untangling effects. However, in reality, destabilization of the monomer or dimer in addition to stabilization of the trimer would have an even stronger effect in creating SOD1 trimer.”

Moreover, he added, 62% of residues currently known to feature disease-relevant point mutations are located in the proposed SOD1 trimer interfaces, and at least two-thirds of disease mutations located in the trimer interfaces help stabilize trimer formation, while the effect of the others is currently unknown.

While they were not the focus of this study, Dokholyan said, preliminary results indicate that both the common ALS mutations A4V and G93A “form this trimer quite aggressively.” Work is currently underway to design compounds that specifically destabilize trimers, to determine their therapeutic potential in SOD1-related disease. The lab is also beginning to explore the pathways involved in trimer-induced cell death.

Neil Cashman. Image courtesy of N. Cashman

Neil Cashman. Image courtesy of N. Cashman

Neil Cashman from the University of British Columbia in Vancouver, who did not participate in the study, commented, “I think that this study is very important from the standpoint of elucidating the toxicity of misfolded SOD1 in ALS, both familial forms of SOD1-ALS and in sporadic ALS. Stefan Marklund’s group has previously identified trimers as a molecular species in mouse models of familial ALS (Zetterstroem et. al., 2007), but Dr. Dokholyan and colleagues have identified that the trimer form is preferentially toxic in the NSC-34 motor neuron cell line model, and that mutations that enhance or reduce stability of the trimer form are more or less toxic respectively.

I think that the most important contributions of this paper are a plausible and rational series of models for the structure of the non-native SOD1 trimer, through innovative application of ‘knowledge-based’ and ‘physics-based’ approaches. Heuristically, there is remarkable concordance between the trimer structure(s) and intelligent mutagenesis to modify stability and toxicity of this transient species.”


Proctor EA, Fee L, Tao Y, Redler RL, Fay JM, Zhang Y, Lv Z, Mercer IP, Deshmukh M, Lyubchenko YL, Dokholyan NV. Nonnative SOD1 trimer is toxic to motor neurons in a model of amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A. 2016 Jan 19;113(3):614-9. [Pubmed].

disease-als SOD1 topic-preclinical
Share this: