Most cases of ALS are sporadic, but some familial cases are caused by mutations in superoxide dismutase-1, or SOD1, a protein that normally protects cells by clearing potentially damaging free radicals. Mutated, SOD1 forms aggregates and damages motor neurons. The most common animal model for ALS is based on rodents that overexpress human, mutant SOD1.

Bcl-2 is another protective protein. It forms pores in the mitochondrial membrane to allow ions through and maintain a healthy membrane potential. It also inhibits pro-death proteins such as Bax and BaK. However, Bcl-2 has a dark side: tucked away in a pocket of the protein is a toxic BH3 domain, and when that domain is unleashed, Bcl-2 promotes cell death instead of survival.

Researchers knew that mSOD1 entered mitochondria, and that mitochondria looked and behaved abnormally in the spinal cords of people with ALS and animal models. But until now, there was nothing to link the two, said study senior author Piera Pasinelli. With first author Steve Pedrini and colleagues, Pasinelli determined that mSOD1 pulls Bcl-2 out of its protective pocket, and the two wreak havoc together. This is one possible explanation, but unlikely the sole mechanism, by which mitochondria lose their form and function in ALS.

Several years ago, Pasinelli and colleagues demonstrated that mSOD1 binds Bcl-2 (see ARF related news story on Pasinelli et al., 2004). Originally, Pasinelli said, the researchers assumed mSOD1 “stole Bcl-2 away from the mitochondria.” But they were surprised. “It turns out that, in fact, here we have, through the action of mutant SOD1, a new gain of toxic function,” she said.

To examine the mSOD1-Bcl-2 interaction, the researchers transfected HEK293T human embryonic kidney cells, which normally do not express Bcl-2, with wild-type or mutant SOD1 in the presence or absence of a Bcl-2 construct. Only in the case of mSOD1 plus Bcl-2 did they note any change in the cells. Under the microscope, cytochrome c antibody staining showed this small protein, part of the electron transport chain, leaked out of mitochondria to fill the cell. Biochemical assays confirmed that cytochrome c levels were diminished in the mitochondria of cells expressing mSOD1 and Bcl-2. And according to a commercial assay of cellular metabolism, cell viability dropped by 25 percent in cells containing this combination of constructs.

Further, the researchers performed transmission electron microscopy on the double-transfected cells. Unlike the healthy mitochondrial networks in cells transfected with WT-SOD1, Bcl-2, or a combination, the Bcl-2/mSOD1 cells evinced swollen, rounded mitochondria with disorganized cristae and vacuolization. “It looks really similar to the strange-looking mitochondria that you see in the ALS mouse model, in the spinal cord,” said Thomas Gould, a postdoctoral researcher at the Salk Institute in La Jolla, California, who was not involved with the study.

The scientists reasoned that if mSOD1 and Bcl-2 were acting via Bcl-2’s toxic BH3 domain, then mutating that domain should limit the toxicity. When they used a Bcl-2 mutant that can still bind SOD1, but with an altered BH3 domain, all the toxic effects of the pair reversed. Cytochrome c stayed in the mitochondria where it belonged, and cell viability returned to normal.

Then, the researchers used a set of Bcl-2 antibodies to analyze the protein’s conformation in vivo. They had one antibody that selectively bound the BH3 domain when it was exposed, and another that only bound Bcl-2’s protective pocket when the BH3 domain was safely secreted within. In mice overexpressing human mSOD1, the BH3 antibody signal appeared in spinal cord homogenates at 30 days of age, before symptom onset, and rose to peak at 85 days, when the disease becomes noticeable. BH3 exposure continued until the end-stage of disease. Conversely, signaling from the pocket antibody decreased with the disease. The researchers saw high levels of BH3 signaling in one human sample as well, from a person who had ALS because of a SOD1 mutation.

Overall, the results imply that Bcl-2, upon mSOD1 binding, flips conformation to expose the toxic BH3 domain. Among its effects are allowing the mitochondrial membrane to leak, releasing mitochondrial components such as cytochrome c. Lacking fully functional mitochondria, the neuron starves and sustains damage. “It is indeed one of the triggers of mitochondrial dysfunction,” Pasinelli said. The authors call the pair “partners in crime,” since neither can cause this damage alone.

Although mSOD1 makes use of the cell’s apoptotic machinery, it is not causing apoptosis, Pasinelli and Gould agreed. Mitochondrial abnormalities show up very early in the disease process in mSOD1 animals, well before motor neurons actually die. Instead, Pasinelli suggested mSOD1 simply damages mitochondria, which could itself be a trigger for motor neuron disease.

Other research papers support the hypothesis that mSOD1 uses the apoptotic system to cause problems, but not apoptosis. Messenger RNA for another member of the Bcl-2 family, Bim, is increased in the spinal cords of symptomatic mSOD1 mice. Bim activates Bax, a pro-death protein. Knocking down Bim protects cultured motor neurons from mSOD1, suggesting Bim, too, is part of the mSOD1 damage pathway (Hetz et al., 2007). In another paper, Gould and colleagues at Wake Forest University in Winston-Salem, North Carolina, where he worked in the laboratory of Robert Oppenheimer, suggest that motor neuron death and motor neuron dysfunction are distinct in an ALS mouse model. They crossed Bax-deficient mice with mSOD1 animals, so that the resulting double mutants would have motor neurons incapable of normal apoptosis. The motor neurons did not die, but they still suffered denervation of muscle tissue and mitochondrial vacuolization, suggesting motor neurons need not go through apoptosis to degenerate in ALS (Gould et al., 2006).

Gould said he is not fully convinced that Bcl-2 is required for mSOD1 mitochondrial toxicity. For one, he would prefer to see the pair work together to kill cells more akin to motor neurons than the HEK293T cells used in the study. In addition, he said an in vivo mouse experiment would buoy the theory. Gould suggested crossing Bcl-2 knockouts—which survive for six weeks—with mSOD1 mice. If Bcl-2 is required for mSOD1 mitochondrial damage, then the mitochondria of these mice should remain unchanged compared to Bcl-2 single mutants.

For her part, Pasinelli is forging ahead with her group to look for ways to turn their finding into a treatment. A small molecule that blocks the SOD1-Bcl-2 interaction, she suggested, might be protective. The researchers are working to delineate the exact protein sequences involved in mSOD1-Bcl-2 binding, and to find peptides that will get in the way. However, Pasinelli concedes, the effectiveness of such a treatment would probably be limited to the minority of people with ALS who carry SOD1 mutations.—Amber Dance.

Reference:
Pedrini S, Sau D, Guareschi S, Bogush M, Brown RH Jr, Naniche N, Kia A, Trotti D, Pasinelli P. ALS-linked mutant SOD1 damages mitochondria by promoting conformational changes in Bcl-2. Hum Mol Genet 2010 May 22. Abstract