Key Points Raised by Individual Speakers
• Because of their high energy requirements, neurons are especially vulnerable to injury and death from dysfunctional mitochondria.
• Pathological and physiological evidence reveals mitochondrial dysfunction in all major neurodegenerative diseases.
• Questions remain as to whether mitochondrial dysfunction is causal to neurodegenerative disease. Even if is not causal, mitochondrial dysfunction is still highly important and likely contributory to disease. Identifying therapies to improve mitochondrial function or to degrade dysfunctional mitochondria may make sense.
• Studying primary mitochondrial diseases can shed light on neurodegenerative diseases that show similar pathology. Because both types of diseases affect multiple pathways and organ systems, they require the approach of systems biology.
• Potential therapeutic approaches include medications that induce mitochondrial genesis, catalytic antioxidants to protect against reactive oxygen species, regulators of intracellular calcium, and regulators of redox potential across mitochondrial membrane. Maintenance of redox potential is crucial for mitochondrial integrity and control over oxidative phosphorylation.
Mitochondria are cellular organelles responsible for oxidative phosphorylation, the vital process of converting nutrients into adenosine triphosphate (ATP) molecules that provide the power for normal cell functions. Each neuron has at least hundreds of mitochondria. Because nerve cells are postmitotic, any mitochondrial damage that is sustained will accumulate with age and lead to dysfunction. Widespread damage to mitochondria causes cells to die because they can no longer produce enough energy. Indeed, mitochondria themselves unleash the enzymes responsible for cell death. The brain is especially vulnerable to mitochondrial dysfunction because its energy needs are higher than that of any other organ in the body. The brain accounts for only 2 percent of body weight yet consumes 20 percent of oxygen.
Mitochondrial functioning is determined by two separate genomes, one in the mitochondria, known as mitochondrial DNA (mtDNA), and the other in the nucleus. The mitochondrial genome encodes 13 proteins, all of which are vital to oxidative phosphorylation. The nuclear genome encodes approximately 1,500 genes involved in mitochondrial biology, including proteins necessary for replication of mtDNA, transcription, translation, and posttranslational modifications. There is only one copy of mtDNA, inherited from the mother, versus two copies of nuclear DNA, one from the mother and the other from the father. Mitochondria not only are responsible for oxidative phosphorylation, but they also play significant roles in metabolism and signaling, including fatty acid synthesis, ketone body metabolism, calcium homeostasis, and apoptosis. More specifically, mitochondria provide the majority of cellular energy in the form of ATP. They generate and regulate reactive oxygen species, they buffer calcium levels inside the cell, and they control apoptosis (Wallace, 2005, 2010).
Mitochondrial defects are found in pathological studies of all major neurodegenerative diseases, said Vamsi Mootha of Harvard Medical School. The range of mitochondrial defects includes fragmentation and other morphological changes, increased mutation rates in mtDNA, changes in permeability of mitochondrial membranes, changes in redox potential, accumulation of mutant proteins, and impaired oxidative phosphorylation (Reddy and Reddy, 2011). But whether these mitochondrial defects are causal in neurodegenerative disease is the fundamental question, Mootha said. The potential roles of mitochondria in neurodegenerative disease are, in his view, threefold: (1) they harbor primary lesions and thus serve as the primary source of disease pathology; (2) they function properly, but serve as mediators or amplifiers of disease; or (3) they are bystanders that do not contribute to pathology. Even if mitochondrial defects are not causal, they are likely contributory, noted Neil Kowall of Boston University, and thus any therapy that preserves, enhances, or corrects mitochondrial function is likely to be beneficial in forestalling cell death and disease progression.
This chapter summarizes workshop presentations that provide evidence of mitochondrial dysfunction in major neurodegenerative diseases. Because the evidence is unclear as to whether mitochondrial dysfunction is causal, it may be valuable to look at primary mitochondrial diseases and adopt a systems approach to research, several participants said.
MITOCHONDRIAL DYSFUNCTION AND NEURODEGENERATIVE DISEASES
As noted above, mitochondrial dysfunction is found in the major neurodegenerative diseases. This section outlines workshop presentations about mitochondrial dysfunction in Parkinson’s disease, amyotrophic lateral sclerosis (ALS), Huntington’s disease, and Alzheimer’s disease.
Parkinson’s Disease
Parkinson’s disease is characterized by a loss of dopamine-containing neurons in the brain region known as the substantia nigra. Pathological and other studies have convincingly shown that mitochondrial deficiency accumulates in this brain region upon aging, said Richard Youle of the National Institute of Neurological Disorders and Stroke. Youle’s talk focused on the function of two proteins that are mutated in familial, early-onset Parkinson’s disease: Parkin and PINK1 (PTEN-induced putative kinase 1).
The normal functions of Parkin and PINK1 have not been well understood until recently, Youle said. Evidence from multiple species is accumulating that these proteins normally work together to trigger clearance of damaged mitochondria, a process known as mitophagy. It stands to reason that, if mutated, they can fail to induce mitophagy, leaving dysfunctional mitochondria to accumulate within the cell and cause death. In this way, the failure of mitophagy is implicated in the etiology of early-onset Parkinson’s disease (Narendra and Youle, 2011).
When mitochondria are under stress or damaged, remarked Youle, they accumulate PINK1. PINK1 is a mitochondrial protein ordinarily anchored to the mitochondrion’s outer membrane at low concentrations. When the mitochondrial membrane loses its electrical potential—whether by DNA mutations, reactive oxygen species (ROS),1 or other perturbations—PINK1 increases. Increasing concentrations of PINK1, in turn, serve to recruit Parkin, which is a ubiquitin ligase, from the cytosol. Parkin marks the damaged mitochondrion with ubiquitin, a process that triggers formation of an autophagosome (see also Chapter 3). The autophagosome engulfs
1 ROS are produced as a byproduct of oxidative phosphorylation.
the damaged mitochondrion, then merges with a lysosome, which degrades it. The findings on PINK1/Parkin, which have been replicated in multiple laboratories, have established a novel pathway for mitochondrion quality control, said Youle. He noted that much of the earlier work in this area was done on cultured cell lines, because the compounds used to induce this pathway in cultured cell lines were too toxic to neurons. More recently, however, two groups have shown the pathway in neurons (Cai et al., 2012; Wang et al., 2011).
Youle relayed that his laboratory has started a drug screening program to identify compounds that stimulate the PINK1/Parkin pathway. While he acknowledged that people with early-onset Parkinson’s may not benefit from stimulating the pathway because their PINK1 or PARKIN are mutated, people with sporadic Parkinson’s disease may benefit, as might others with neurodegenerative disease whose mitochondria are dysfunctional.
Amyotrophic Lateral Sclerosis
ALS predominantly affects motor neurons, leading to progressive muscle wasting and paralysis. In animal models of ALS, mitochondrial abnormalities precede symptoms of disease (Manfredi and Xu, 2005). Electron microscopy has revealed structural abnormalities in mitochondria in spinal motor neurons and in the motor cortex of ALS patients. Neil Kowall of Boston University focused his presentation on SOD1 (Cu, Zn superoxide dismutase), the first identified gene responsible for causing ALS. The corresponding protein is mutated in about 20 percent of familial cases of ALS.2 The most widely used animal model of ALS is a transgenic mouse carrying a mutant SOD1 gene. The mouse develops muscle wasting similar to that of ALS.
Mitochondria from motor neurons in this animal model exhibit smaller size, fewer number, defective membrane potential, and impaired fusion. Fusion of mitochondria is designed to distribute mtDNA to the mitochondrial population and preserve the capacity for oxidative phosphorylation. These morphological and physiological changes in mutSOD1 motor neurons are not seen in wild type SOD1 motor neurons (Magrane et al., 2012). mutSOD1 also alters the levels of at least 50 different mitochondrial proteins, including proteins involved in the electron transport chain and in fusion, suggesting a possible widespread effect of mutSOD1 (Karbowski and Neutzner, 2012).
mutSOD1 also inflicts mitochondrial damage, as assessed by an increase of cytochrome c in the cytosol. Because cytochrome c is an essential component of the electron transport chain, which is situated in the inner
2 About 5 to 10 percent of ALS cases are familial.
membrane of the mitochondria, its release into the cytoplasm indicates disruption of mitochondria membranes. But this toxic effect only occurs in the presence of the protein Bcl-2, which can reverse its functional phenotype and become a toxic protein (Pedrini et al., 2010). The identification of Bcl-2 as a necessary contributor to SOD1 toxicity suggests that Bcl-2 could be used as a molecular target for drugs designed to inhibit its action (Pedrini et al., 2010). Bcl-2 may also be an important target not only in familial ALS, but possibly also sporadic ALS, said Kowall. That is because research has recently found that, in a subset of ALS patients with bulbar onset, wild-type SOD1 becomes hyperoxidized. In concert with Bcl-2, hyperoxidized wtSOD1 displays mitochondrial toxicity similar to that seen with mutSOD1. Thus, Bcl-2 represents a common link between familial and a subtype of sporadic ALS, and thus appears to be a good target for therapeutics that inhibit it.
Huntington’s Disease
Huntington’s disease is an autosomal dominant disease in which the mutated protein, mhuntingtin (mHTT), displays excess polyglutamine repeats. mHTT localizes to the outer mitochondrial membrane, where it exerts widespread and deleterious effects on mitochondria and selective loss of neurons in the striatum. Kowall said a great deal of evidence shows that mHTT reduces mitochondrial motility, alters mitochondrial morphology, causes calcium dysregulation, reduces oxidative phosphorylation, and depolarizes the mitochondrial membrane in lymphoblasts of Huntington’s disease patients. The depolarization is increased with greater numbers of polyglutamine repeats. mHTT also alters the balance between mitochondrial fusion and fission (Lin and Beal, 2006; Reddy and Reddy, 2011).
Several therapeutic strategies have recently emerged for Huntington’s disease, Kowall noted. One avenue is to target a mitochondrial fission3 protein to which mHTT binds, GTPase dynamin-related protein 1 (DRP1). The targeting of DRP1 is suggested by the finding that a dominant-negative DRP1K38A mutant, which reduces DRP1 activity, rescues mitochondria from the following adverse effects of mHTT: mitochondrial fragmentation, defects in anterograde and retrograde mitochondrial transport, and neuronal cell death. These findings were reported in cells from humans with Huntington’s disease and from mice (Song et al., 2011). In other words, compounds that inhibit DRP1 might be useful as potential therapies.
Kowall described two more novel therapies. The first aims to detox-
3 Mitochondrial fission is a quality control mechanism in which the mitochondrion divides into two, one healthy and the other containing the damaged portion of the mitochondria. The latter portion is degraded.
ify HTT. It involves intraventricular infusion of ganglioside GM1, which phosphorylates mutant HTT at specific serine amino acid residues. The approach not only curtailed the toxicity of HTT, but also restored normal motor function in symptomatic Huntington’s disease mice (Di Pardo et al., 2012). The second therapy is with the already approved drug meclizine. This drug suppresses mitochondrial respiration and activates cellular survival pathways. In several models of Huntington’s disease, meclizine was found to be neuroprotective (Gohil et al., 2011).
Alzheimer’s Disease
Mitochondrial dysfunction precedes the pathological changes that are the hallmarks of Alzheimer’s disease (Yao et al., 2009). Douglas Wallace of the Children’s Hospital of Philadelphia proposed that the cause of Alzheimer’s disease—and dementia more broadly—is from underlying dysfunction of the mitochondria. Beginning in 1993, Wallace’s team found a mutation in one of the mitochondrial tRNA genes. The mutation correlated with 3 percent of late-onset Alzheimer’s cases, 5 percent of Parkinson’s, and 7 percent of the combined population. The finding was later corroborated by others. He asserted that subtle defects in tRNA will generate more global mitochondrial protein synthesis defects. Subsequently, his team began to study a mutation at the nucleotide position 414, which is adjacent to the control region promoter of mtDNA. The mutation previously had been shown to be increased with age in human fibroblasts (Michikawa et al., 1999). Wallace’s team found the mutation in 65 percent of Alzheimer’s brains and 57 percent of Down syndrome–dementia brains versus 0 percent of age-matched controls (Coskun et al., 2004).
Most recently, Wallace’s team examined more globally the control region of mtDNA in tissue taken from the frontal cortex of the brain. The control region is responsible for regulating transcription of mitochondrial genes and helps copy mtDNA. They found the highest rate of somatic mutations in the Alzheimer’s brain (Coskun et al., 2010). The mutation frequency was also elevated in Down syndrome–dementia cases relative to controls, but it was lower than that in Alzheimer’s. Control tissue did show an age-related increase in mutation frequency, although the level was lower than that found in the other groups. The heightened rate of mutations was also found in serum and other tissues of Alzheimer’s and Down syndrome cases, suggesting that the phenomenon is systemic. But, said Wallace, the brain is the most deeply affected tissue because of its disproportionately high energy demands. The study also found reduction in transcription of mtDNA and a reduction in the mtDNA copy number, implying a reduction in oxidative phosphorylation.
Turning to causation of neurodegenerative disease, Wallace expressed
the view that formation of Aβ plaques is not causal; rather, he hypothesized, Aβ protein is initially produced by cells as a compensatory means of protecting mitochondria. But as the protein continues to be produced, it begins to aggregate to form oligomers and larger aggregates that inhibit mitochondria, leading to cell injury and death. According to this model, the protein aggregates are contributory to the death of neurons in neurodegenerative disease, but not causal. Primary causation, according to his hypothesis, rests with dysfunctional mitochondria. He expressed the opinion that “bioenergetics is the common pathophysiological mechanism for all of these neurodegenerative diseases.” He was then questioned in the discussion by several skeptical participants who did not agree with his causal attribution. In reply, Wallace described how his team had developed a way to introduce a cytochrome oxidase point mutation in mtDNA and found that the animal developed cardiomyopathy, myopathy, and pathological changes in hippocampal neurons, in retinal ganglion cells, and in the optic nerve. “This one particular point mutation—it has nothing to do with the nucleus—shows that energetics can affect all of those different functions,” he asserted.
PROTEIN DEPOSITS AND TOXICITY TO MITOCHONDRIA
Multiple lines of evidence suggest that toxic proteins such as Aβ, apolipoprotein E (ApoE) fragments, and α-Synuclein can impair mitochondria, said Lennart Mucke of the Gladstone Institutes and the University of California at San Francisco. In this case, the damaged mitochondria would not be the primary cause of the disease, but rather would be secondary to the actions of aggregated proteins, which would be the primary cause. A significant amount of research shows that Aβ peptide accumulates in mitochondria, where they cause dysfunction and apoptosis (Manczak et al., 2006; Yao et al., 2009).
One possible mechanism by which protein deposits are toxic to neurons is by impairment of axonal transport of mitochondria. Mitochondria are generated largely in the cell body and need to be actively transported to the synapse, where energy need is high. Devoid of mitochondria, synaptic function can be impaired. Mucke and his team assessed the effects of Aβ and tau proteins on axonal transport of mitochondria (Vossel et al., 2010). They found that adding Aβ oligomers in culture quickly inhibited axonal transport of mitochondria in healthy neurons, a finding supported by earlier research. They also were interested in determining whether tau played a role. Reducing tau levels prevented Aβ oligomer-induced disruption of axon transport without affecting baseline axonal transport. The complete elimination of tau by gene knockdown also had the same effect. They concluded, “Aβ requires tau to impair axonal transport, and that tau
reduction protects against defects in Aβ-induced axonal transport” (Vossel et al., 2010, p. 198-a).
Another disease-related protein that impairs mitochondria is ApoE. The ApoE gene is the main susceptibility gene identified for late-onset Alzheimer’s disease, and it is found on chromosome 19. Neurons produce ApoE when they are stressed by a host of factors, including aging, oxidative stress, trauma, and protein deposition. ApoE synthesis is thought to protect neurons from damage and to repair and remodel them. However, research has shown that the cleavage products of ApoE impair mitochondria (Brecht et al., 2004). ApoE e4, the allele associated with Alzheimer’s disease, is most sensitive to being cleaved, whereas the other ApoE alleles are less so, said Mucke.
MITOCHONDRIAL DISEASES AND THEIR UTILITY FOR NEURODEGENERATIVE DISEASE
Given the uncertainty as to what roles mitochondrial dysfunction plays in neurodegerative disease, Mootha suggested the value of studying primary mitochondrial diseases, which refer to nearly 150 genetic diseases in which the lesion lies in a gene encoding a protein that is directly involved in mitochondrial biology. The diseases are heterogeneous, with dozens being the focus of study over many decades. Caused by genetic single-gene mutations or deletions, they follow Mendelian or a maternal pattern of inheritance.
Mitochondrial disease can shed light on neurodegenerative disease, said Mootha, in part because disease phenotypes are similar. For example, some mitochondrial disease phenotypes include ataxia, neuropathy, myopathy, deafness, and blindness. Indeed, several subsequent presentations focused on mitochondrial pathology in neurodegenerative disease, such as Parkinson’s and ALS. Another reason why mitochondrial diseases carry import for neurodegenerative disease is that multiple organ systems are involved, just like neurodegenerative diseases, and their genetics are better characterized through an ambitious project known as the Mitocarta, which is an inventory of more than 1,000 mouse genes encoding proteins that localize to the mitochondria (Pagliarini et al., 2008). Finally, mitochondrial diseases are valuable, in his view, in providing “genetic extremes” that can help to determine whether or not a particular neurodegenerative disease may have mitochondrial defects as the root cause. Mootha advised looking for connections between mitochondrial and neurodegenerative diseases when there is at least some common ground, such as in pathogenesis, pathology, or biomarkers.
Even though mitochondria look similar upon microscopy, looks are deceiving. Mootha remarked on the enormous heterogeneity of mitochondria across different tissues. He reported that, after studying 14 different
tissues, research has found that mitochondria from 2 different tissues share only 75 percent of their proteins, whereas the remaining mitochondrial proteins are tissue specific. There is even physiological heterogeneity within an individual cell—mitochondria, for example, can possess different patterns of fuel usage. Given the diversity of phenotypes and genotypes, Mootha advocated for a systems approach to the study of mitochondrial function. Such an approach combines genomics, proteomics, metabolomics, biochemistry, and computer modeling to capture the dynamic range of complex interactions within cells and across tissues. Applying systems biology to neurodegenerative diseases would require identifying component parts, building wiring diagrams to connect these parts, identifying circuitry causal for the disease, and using the knowledge to develop therapies, he observed.
MITOCHONDRIA AND CELL DEATH
One commonality across neurodegenerative diseases is that they all feature a high degree of cell death. Here the focus is on mitochondria; mitochondria play a key role in regulating cell death, which occurs in specific brain regions across all neurodegenerative diseases. Cell death is of three types: (1) necrosis, which is the most chaotic form of death that involves cytoplasmic swelling, nuclear dissolution, and lysis; (2) apoptosis, an orderly form of death, reliant on ATP, that produces cell fragments that phagocytic cells are able to engulf and remove before the cell’s contents disgorge onto surrounding cells and cause damage; and (3) autophagy, in which the cell degrades its cytoplasm and organelles via lysosomes (Martin et al., 2010). Mitochondria are the sites where antiapoptotic and proapoptotic proteins interact, and they regulate signals for cell death.
Lee Martin of Johns Hopkins University cautioned that cell death in humans versus animal models of neurodegenerative diseases may not be by similar mechanisms. He reported mouse–human species differences in the factors controlling the mitochondrial permeability transition (MPT), that is, an increase in permeability of mitochondrial membranes to small molecular weight molecules. MPT results from opening the mitochondrial permeability transition pore, a protein pore formed in mitochondrial membranes under certain pathological conditions. Induction of the permeability transition pore can lead to swelling of mitochondria and necrosis, and it also plays a major role in some types of apoptosis. Martin also noted species differences in signaling mechanisms of caspases, which are enzymes under the control of the mitochondria that are crucial to apoptosis, differences in caspace substrates, differences in mitochondrial fusion machinery, and in signaling mechanisms for DNA repair and metabolism, among others. Species differences in cell death confound the translation of findings from animal models into human clinical trials, he observed. He suggested modi-
fying the design of preclinical studies to rely less on mouse as models and more on human neural stem cell–derived neurons.
POTENTIAL BIOMARKERS AND THERAPIES
There are no established biomarkers or therapies for treating mitochondrial dysfunction in neurodegenerative disease. Wallace said his laboratory is working to develop them. One biomarker under development is near-infrared spectroscopy across the skin, using different infrared diodes that interrogate the redox potential of the respiratory chains. Wallace said his laboratory is also developing a biomarker using micro-organic breath analysis. They are hoping to get some surrogate variables that change in real time, and then go into a Phase I clinical trial and have at least a safety/efficacy indication.
Regarding therapies, this chapter has already mentioned a few in relation to specific neurodegenerative diseases. Focusing instead on generic therapies for mitochondrial dysfunction, Wallace said his first priority for therapy would be to stimulate formation of more mitochondria. Drugs to generate mitochondria are being tested in various animal models and cell culture systems. In particular, he noted that the drug bezafibrate has been found to increase mitochondrial biogenesis in cancer cells and ameliorate mitochondrial dysfunction (Wang and Moraes, 2011). It has not yet been tested in brain cells.
Other therapeutic options, Wallace explained, include (1) catalytic antioxidants to protect against ROS; (2) regulators of intracellular calcium; and (3) regulators of redox potential across mitochondrial membrane. Maintenance of redox potential is crucial for mitochondrial integrity and control over oxidative phosphorylation. One participant pointed out that antioxidant therapies have been uniformly ineffective in clinical trials, but Wallace responded that the doses may have not been high enough. Another participant advised targeting mitochondrial therapies in cases of threshold effects, that is, the point at which there is significant compromise of mitochondrial function. The participant also noted the possibility of mitochondrial therapies having secondary downsides.
RESEARCH NEEDS AND NEXT STEPS SUGGESTED BY INDIVIDUAL PARTICIPANTS
The workshop speakers identified many questions for future research and other opportunities for future action. The suggestions related to mitochondrial dysfunction are compiled here to provide a sense of the range of suggestions made. The suggestions are identified with the speaker who
made them and should not be construed as reflecting consensus from the workshop or endorsement by the Institute of Medicine.
• Develop deeper understanding of energy biology and interactions between bioenergetics and environmental influences. (Wallace)
• Identify biomarkers to follow mitochondrial functioning. (Lee, Mootha)
• Find biomarkers of mitochondrial decline. (Mootha)
• Identify new therapies that increase mitochondrial biogenesis. (Wallace)
• Identify therapies that interfere with mitochondrial contribution to pathogenesis of neurodegenerative disease, including therapies that increase mitophagy or stimulate activity of PINK1 and Parkin. (Mootha, Youle)
• Find therapies that detoxify mutant protein aggregates that interact with mitochondria. (Kowall)
