Mismanaged mitochondria are a pivotal feature of Alzheimer’s (see ARF related news story and Hirai et al., 2001) and, indeed, a common theme in neurodegenerative disease. The Case Western group has shown that in fibroblasts from people with sporadic AD, mitochondria are extra long, likely because the fission protein DLP1 is underexpressed (Wang et al., 2008). Another group recently reported that DLP1 (also called Drp1) function is altered in cultured neurons exposed to Aβ (see ARF related news story on Cho et al., 2009). Two genes associated with Parkinson disease, pink1 and parkin, have also been linked to mitochondrial fission and fusion in Drosophila (see ARF related news story on Poole et al., 2008) and in human cells (Exner et al., 2007). And HeLa cells expressing polyglutamine-expanded huntingtin, which causes Huntington disease, are particularly sensitive to oxidative stress causing mitochondrial fragmentation (Wang et al., 2009).

Wang and colleagues went looking for the cause of mitochondrial derangement in tissue samples from AD brains. Comparing them to age-matched control samples, they found reduced levels of several mitochondrial managers: DLP1 (reduced by 74 percent) as well as the fusion proteins OPA1, Mfn1, and Mfn2 (down by 61, 28, and 34 percent, respectively). The amount of fission protein Fis1, in contrast, was elevated nearly fivefold in AD samples. The mitochondria in control samples filled both the cell body and its projections; in AD samples, the mitochondria and associated proteins were generally limited to the cell body.

Next, the researchers attempted to mimic AD’s changes to fission and fusion proteins in cell culture. They overexpressed Fis1 and used RNAi to diminish expression of the rest in M17 human neuroblastoma cultures. These changes caused predictable mitochondrial morphology: DLP1 RNAi resulted in elongated mitochondria, while excess Fis1 and reductions in OPA1, Mfn1 and Mfn2 resulted in fragmented organelles. Mirroring the changes in the AD brain, the altered cells showed mitochondria vacating the cell periphery and redistributing to the perinuclear area. The researchers repeated the experiments in differentiated neurons, primary rat hippocampal cultures, with similar results.

Having mitochondria in the wrong place likely impacts cellular function. “Due to the complex morphology of neurons and high energy demand at the synapses, reduced mitochondrial coverage in neurites will likely have significant deleterious effect on synaptic function,” Zhu wrote in an e-mail to ARF. To determine the downstream effects of mitochondrial morphology on synaptic plasticity, the researchers analyzed dendritic spines. Control rat hippocampal neurons averaged four or five spines per 10 microns of dendrite; those modified with a Fis1 overexpression vector or RNAi for the other proteins had only one or two over the same length. Dendritic spines in the treated cells also contained fewer mitochondria.

In Alzheimer disease, the mitochondrial proteins must be altered by some part of the disease, such as amyloid-β. Wang and colleagues had already shown that M17 cells overexpressing APP have fragmented mitochondria that stay in the cell body (Wang et al., 2008). In the current study, they analyzed the effect of oligomeric Aβ-derived diffusible ligands (ADDLs; Klein et al., 2002) on mitochondria and their fission and fusion proteins in primary cultures of rat hippocampal neurons. As a control, they treated some cells with a reverse Aβ peptide containing amino acids 42-1.

The cells treated with oligomers had shorter mitochondria that mostly clustered in the cell body. These cells had fewer dendritic spines as well. Using time-lapse microscopy, the researchers were able to monitor labeled mitochondria in rat hippocampal neurons for fission and fusion events. The organelles coalesced and fragmented constantly in control cells, but did so less often in oligomer-treated cells, suggesting both processes were impaired. ADDLs also induced mitochondria to produce more reactive oxygen species than control cells.

Cells treated with oligomeric Aβ also had lower levels of both DLP1 and OPA1 than control cells. Mfn1, Mfn2, and Fis1 levels were unchanged. “Changes in Mfn1/2 and Fis1 may require coordinated signals from other sources or contribution from other cell types,” Zhu wrote. DLP1 overexpression returned dendritic spine numbers to normal in ADDL-treated cells, and OPA1 overexpression was able to return ROS generation to control levels.

DLP1 was a bit of an outlier in the study—alone among the proteins examined, its reduced expression seen in AD brains caused mitochondria to elongate. “I don’t have a good explanation on why DLP1 level is reduced,” Zhu wrote. DLP1 is involved in apoptosis, he noted, so perhaps its diminished levels in tissue samples are part of the neuron’s response to apoptotic signals.

Based on these data, Zhu hypothesizes that Aβ causes mitochondrial fragmentation and abnormal distribution, which causes metabolic dysfunction, particularly in axons and dendrites. Notably, he saw no evidence of apoptosis in the altered cells, suggesting that changes to fission and fusion happen early in the disease process. However, other interpretations are possible, said Mark Cookson of the National Institute on Aging in Bethesda, Maryland. He suggested that instead, axonal damage might lead to downstream effects on the local mitochondria.

Zhu hopes next to find a similar mechanism for disruption of mitochondrial dynamics in mouse models of AD. Cookson also wondered if the same changes to fission and fusion proteins are a common theme in neurodegenerative disease. “Generally, when neurons are sick, do you see those changes?” he said. “It would be interesting to know if it happened in Huntington’s or in ALS.”

“Our study strongly suggests that it may be a valid therapeutic target to restore normal mitochondria distribution in neuritis so as to restore synaptic function,” Zhu wrote. But Cookson cautioned that messing with something as essential and ubiquitous as mitochondria might be a risky business, requiring finely targeted treatment. “My guess would be you’d have a very narrow window in which things would work properly. If Aβ is indeed upstream of the mitochondrial defect, it might be a better target, he suggested.—Amber Dance.

Reference:
Wang X, Su B, Lee H, Li X, Perry G, Smith MA, Zhu X. Impaired balance of mitochondrial fission and fusion in Alzheimer’s disease. J Neurosci. 2009 Jul 15;29(28):9090-9103. Abstract