Much is amiss with mitochondria during the pathology of amyotrophic lateral sclerosis. Two new papers report on how the cell’s powerhouse organelle is thrown off balance by mutant proteins linked to the disease. In the January 18 Journal of Neuroscience, researchers from Children’s Hospital of Philadelphia, Pennsylvania, describe how blocking AMP activated protein kinase—normally a beneficial modulator of cellular metabolism—protects motor neurons in cell and animal models of the disease. Going beyond the motor neuron and into the muscle, a team based at the University of Alabama School of Medicine in Birmingham report, in the January 19 Developmental Cell online, that a mutation associated with ALS disrupts normal mitochondrial architecture in worms and flies.
Together, the two works provide more fodder for the theory that ALS involves major energetic dysfunction, commented Jordi Magrané of the Weill Medical College of Cornell University in New York City. Magrané, who was not involved in the current studies, has described abnormal mitochondrial transport in motor neurons modeling ALS (see ARF related news story on Magrané et al., 2012).
Neuronal Metabolism Out of Control
Robert Kalb, senior author on the Journal of Neuroscience paper, was inspired to examine energy regulation in ALS by previous work linking defective metabolism to the disease in mice (Dupuis et al., 2004). He recruited graduate student and first author Ria Lim to examine the role of AMP activated protein kinase (AMPK), a master regulator of cellular metabolism, in the disease. AMPK is activated when the ratio of ATP:AMP drops too low. For example, during exercise, AMPK turns on in muscle tissue and helps the cell burn the extra fuel it needs. The kinase, Kalb and Lim suspected, would mediate the cell’s response to abnormal ATP levels in ALS as well.
The researchers cultured neurons from embryonic rat spinal cords and expressed the gene for human superoxide dismutase 1 (SOD1) in the cells. Mutant SOD1, which causes ALS, is known to bind the anti-apoptotic protein Bcl-2 in spinal cord mitochondria (see ARF related news story on Pasinelli et al., 2004) and block mitochondrial pores (see ARF related news story on Israelson et al., 2010). In Lim’s assay, the G37R mutant SOD1 killed the majority of cultured neurons within three days, while cells expressing the wild-type dismutase survived. The mSOD1 cells also exhibited depressed respiration. While it is not easy to directly measure the ATP:AMP ratio in cultures, Kalb said, the reduced oxygen use that Lim measured indicated reduced metabolism. As an energy sensor, AMPK should respond to this deficiency, with a conformational change that makes it more likely to be phosphorylated by a handful of upstream kinases. Sure enough, Lim observed 50 percent more active, phosphorylated AMPK in cells making mSOD1 compared to wild-type neurons.
In normal cells, AMPK keeps ATP levels steady by boosting uptake of fatty acids and glucose, while downregulating other activities like protein translation that use up fuel. Would this activity protect the sickened neuron cultures expressing mSOD1? To find out, Lim treated the cells with an AMPK antagonist, Compound C, or an activator, AICAR. Compound C kept the neurons alive, but AICAR did not, indicating AMPK activation is bad news for motor neurons with mSOD1, and that dampening its function is protective.
Lim and Kalb took their experiments in vivo with C. elegans that express human G85R SOD1 in their neurons. These nematodes are impaired in movement (Wang et al., 2009). Lim crossed these worms with a strain lacking aak-2, a subunit of AMPK. The double mutants didn’t move quite like wild-type worms, but they crawled and swam more than twice as quickly as mSOD1 single mutant worms. Similarly, aak-2 deletion alleviated some of the movement problems of worms expressing mutant TDP-43, another ALS-linked gene.
Although the complete pathway from mutant SOD1 expression to metabolic defects remains unknown, “we think that cells are sensing a bioenergetic defect, they are engaging [AMPK] to correct that defect, and the systems they engage are, in fact, detrimental,” Kalb concluded. Next, he intends to examine which body tissues are most strongly affected by the energy-managing dysfunction, and look for potential drug targets. Neurons, he speculated, could be affected because they expend loads of ATP to maintain their ionic gradients.
Muscle Mitochondria Mislocalized
While motor neurons are the major focus of ALS research, muscle garners more attention lately (see ARF related news story on Palma et al., 2011). In the Developmental Cell paper, first author Sung Min Han and senior author Michael Miller, both at the University of Alabama, reported how motor neurons affect mitochondria in muscles.
Han and Miller focused on VAPB, another ALS gene (Nishimura et al., 2004). Miller noticed that VAPB contains a motif called a major sperm protein (MSP) domain, named for the factor that worms' sperm secrete to cause oocyte maturation (Miller et al., 2001). He suspected that the MSP domain of VAPB could also be secreted, a fact which he and Han confirmed in collaboration with Hiroshi Tsuda, then a postdoctoral researcher in the laboratory of Hugo Bellen at Baylor College of Medicine in Houston, and now at McGill University in Montréal, Canada. They also found that the ALS-associated P58S mutation in the MSP domain prevents this secretion (see ARF related news story on Tsuda et al., 2008).
In the current work, Han and Miller, Tsuda and Bellen again collaborated to elucidate the importance of VAPB MSP secretion. First, they examined Drosophila lacking the VAPB homolog dvap. Studying flight muscles, they found that mitochondria neatly line up with actin in between the myosin-containing striations in wild-type flies. But in the dvap-null insects, the mitochondria were deformed; most were puny, while some were swollen. Similarly, in worms missing the VAPB homolog vpr-1, muscle mitochondria failed to form their normal organized lines and instead sprawled across the cytoplasm in highly interconnected networks. The worms’ mitochondria also took up only small amounts of the marker MitoTracker, which depends on a healthy membrane potential for its uptake, indicating poor mitochondrial function. The worms themselves were “lethargic,” Miller said, further confirming their metabolic syndrome.
If lack of VAPB, or its homologs, causes problems, then replacing it might fix the phenotype. When the researchers provided dvap to the muscle of dvap-null flies, it did not rescue the muscle defects; however, putting dvap specifically in the neurons did. The same was true in the nematodes. “These VAP proteins are essentially like secreted motor neuron factors that regulate energy production in muscle,” Miller concluded. Given how the ALS-linked mutation disables VAPB MSP secretion, “that provides some compelling evidence that the secreted form is crucial for the pathogenesis,” he added. VAPB was also recently reported to have a key cell-autonomous role in maintaining proper calcium uptake by mitochondria (De Vos et al., 2011).
In further experiments, the team identified SAX-3 Robo, and CLR-1 Lar receptors as the key muscle receptors for the VAPB MSP domain. Signaling pathways downstream of these receptors regulate the actin cytoskeleton, which controls mitochondrial localization (see figure below). The same receptors are expressed in neurons, so the VAPB MSP domain might influence neural mitochondrial arrangements as well, the authors note.
Motor Neuron VAPB Fragment Energizes Muscle
Motor neurons cleave and secrete the VAPB MSP subunit. ROBO and LAR receptors on muscle cells pick up the MSP, which alters downstream signaling to promote integration of mitochondria into I-bands, the actin-rich striations where they belong. ALS-linked VAPB mutants fail to release MSP, linking these mutations to muscle dysfunction in disease. Image credit: Cell Press
“One major unresolved question is how the VAP-MSP fragment reaches the extracellular space. Another question is whether a similar mechanism operates in vertebrates and, in particular, in mammals,” noted Dick Jaarsma of the University Medical Center Rotterdam, The Netherlands, in an e-mail to ARF (see full comment below). “The model of Han et al. predicts that the functional deficits resulting from VAPB deficiency in motor neurons occur in the muscle fibers. However, electromyography and muscle biopsy point to a neurogenic basis of the disease…implying functional loss of motor neurons or their axons (rather than muscle fibers) as the prime event in the disease,” Jaarsma continued. Miller noted that in ALS, “You see a lot wrong with the muscle, too,” and suggested there is a renewed appreciation for the role of muscle in the ALS disease process.
Magrané was excited to see that the VAPB, like SOD1, causes mitochondrial mislocalizations. So do mutations in the ALS gene TDP-43, as the Kalb team and others have observed (see ARF related news story on Xu et al., 2010). “It looks like [ALS pathology] is all about positioning; it is all about having mitochondria where they are supposed to be,” Magrané said.—Amber Dance