Cells dress up the enzyme SOD1—the cause of some cases of inherited amyotrophic lateral sclerosis—in a variety of ways. They accessorize it with copper and zinc ions. They tie it together with an intramolecular disulfide bond. They link it with another well-dressed SOD1 to form a dimer. All those modifications mean there are many kinds of SOD1 that could lead to ALS—but lately studies have focused on the most immature, naked Apo-SOD1 as the variety most likely to form aggregates and make motor neurons sick. A study posted in the January 21 Proceedings of the National Academy of Sciences USA online questions that focus. Researchers from the University of Waterloo in Ontario, led by Elizabeth Meiering, report that Apo-SOD1 mutants do not aggregate much at all.
More than 100 different SOD1 mutations have been linked to ALS. Some result in rapidly progressing, severe disease; others allow people to survive upwards of a decade following diagnosis. If unstable, aggregated Apo-SOD1 causes ALS, scientists reason; then the more unstable a SOD1 mutant is, the faster it should fell its victims. Meiering and joint first authors Helen Stubbs and Kenrick Vassall (now at the University of Guelph in Ontario) analyzed Apo-SOD1 mutants in vitro and attempted to align their physical properties with the severity of disease they cause in people.
Given Apo-SOD1’s reputation as an unstable aggregator (see ARF related news story on Banci et al., 2009 and ARF related news story on Nordlund et al., 2009), the authors did not expect they could squeeze much data out of the protein using biophysics techniques to analyze protein stability and the formation of small, soluble oligomers. They were pleasantly surprised. “The protein in this most immature form is marginally stable,” Meiering said.
SOD1 contains four cysteines. Two link up in an intramolecular disulfide bridge; the other two, at positions 6 and 111, are usually reduced. But as aggregates form, these free cysteines may hook up with the cysteines on other SOD1 molecules, promoting formation of multiple-SOD1 species. To avoid this confounding factor, Meiering and colleagues used a pseudo-wild-type (pseudo-WT) construct in which those free cysteines are swapped for a serine and alanine. This SOD1 is more stable than normal; while the wild-type Apo-SOD1 is approximately 75 percent folded at equilibrium, the pseudo-WT construct is 95 percent folded.
Some scientists contacted by ARF criticized the use of pseudo-WT SOD1. Nikolay Dokholyan and Rachel Redler, of the University of North Carolina at Chapel Hill, suggested in an e-mail that since the free cysteines are likely involved in aggregation, eliminating them is inappropriate for studies of inclusion formation (see full comment below). Meiering explained that she wanted to examine SOD1 aggregation in a state mimicking the earliest stages of disease, before the protein is likely to encounter oxidative stress and form unnatural disulfide bridges via the cysteines in question (see ARF related news story on Karch et al., 2009).
To analyze the effects of different mutations on stability and aggregation, the researchers piggybacked several kinds of disease-linked SOD1 mutations onto the pseudo-WT construct: the substituted amino acids at the dimer interface (A4V, A4T, A4S, and V148I); in a β-barrel motif (G37R and H43R); in metal-binding regions (H46R and G85R); in other structural elements (G93R, G93S, G93A, G93D); and a salt bridge participant (E100G).
At physiological temperatures, most of the mutants were unstable—more likely to be unfolded than the pseudo-WT. However, H46R and V148I were more likely to be folded than the pseudo-WT, suggesting they are unusually stable. Therefore, not all SOD1 mutations destabilize Apo-SOD1, even though they cause disease. The finding suggests that unstable Apo-SOD1 must not be the cause—at least, not the only cause—for ALS.
Could a greater inclination to aggregate in the Apo form be the common factor that pathogenic SOD1 mutants share? To pursue this hunch, the scientists used ultracentrifugation and sedimentation to examine how large groups of SOD1 molecules grew. The pseudo-WT, in Apo form, tended to hang out by itself with a molecular weight of 15 kDa, monomer size. But some mutations formed larger structures: A4V and E100G weighed in at somewhere between 1 and 2 monomers, and H43R was closer to the size of a dimer. Using light scattering, the scientists determined that, over a course of days, both A4V and H43R grew into aggregates as large as 100 nm across.
But when the scientists attempted to correlate aggregation with disease duration in people with the mutations, they found no pattern. Having the least stable, most aggregation-prone Apo-SOD1 did not lead to a faster-progressing disease course.
The current analysis of Apo-SOD1 is the most complete to date, wrote Jeffrey Agar of Brandeis University in Waltham, Massachusetts, when asked by ARF (see full comment below). However, the results clearly contradict those of scientists who found that Apo-SOD1 is unstable and aggregates readily, he added.
Meiering and colleagues suggested that the different results come down to whether the protein was shaken or still. “When you shake protein, you can make pretty much any protein aggregate,” she said. Thus, scientists can find aggregation in vitro that would be unlikely to occur in vivo. In these experiments, the Waterloo researchers preferred to mimic natural conditions. “We were trying to be as gentle as possible,” Meiering said—and that gentleness meant less aggregation.
The study suggests Apo-SOD1 is not necessarily the place to look for the cause of ALS. Instead, Meiering advised, researchers should consider all of the protein’s myriad stages of maturation as potential unstable aggregators. Redler and Dokholyan also noted that Apo-SOD1 is not completely off the hook—for example, cellular factors could team up with the protein to promote aggregation in a manner that would not occur in pure solution.—Amber Dance