Some amyloid-β accumulates in mitochondria, where it has been blamed for all manner of malfunctions, from respiration problems to wonky organelle morphology. How could one small peptide be responsible for so much mayhem? Researchers think they may have uncovered a mechanism. Aβ throws a wrench into the assembly line that produces mature mitochondrial proteins, according to a paper in the August 28 Cell Metabolism. Researchers led by Chris Meisinger at the University of Freiburg, Germany, reported that Aβ retards this coming-of-age process, triggering the destabilization of several key mitochondrial proteins. 

“There is substantial evidence for mitochondrial deficits in Alzheimer’s disease; however, the reasons have been obscure,” commented Flint Beal of Weill Cornell Medical College in New York. “This paper presents a potential mechanism.” 

The vast majority of proteins in the mitochondria are cytosolic emigrants. Most flash an N-terminal passport sequence to hitch a ride on the organelle’s import machinery (see Vögtle et al., 2009). Once inside, the proteins surrender this presequence to mitochondrial processing peptidase (MPP), which releases mature, stable proteins that tackle various jobs. A metalloprotease called PreP then degrades the remnant presequence peptides. 

Protein Interrupted. Normally, immature mitochondrial proteins carrying N-terminal presequences are trimmed by MPP, then Cym1 disposes of the clippings (left panel). In AD (right), Aβ seems to derail this process by blocking Cym1, leading to a presequence build-up that blocks maturation of preproteins. [Image courtesy of Mossmann et al., Cell Metabolism, 2014.]

Meisinger and colleagues had previously noticed that immature preproteins accumulated in yeast cells lacking PreP activity. Interestingly, researchers have reported that PreP also degrades Aβ in mitochondria, and that the activity of the protease is weak in AD patients (see Jul 2006 news story and Alikhani et al., 2011). Meisinger and colleagues wondered whether degradation of presequence peptides was necessary for the protein maturation machinery to work, and if so, whether interference from Aβ could foul up the process.

To address this question, co-first authors Dirk Mossmann and F.-Nora Vögtle started by generating yeast mutants that lacked the PreP homolog, Cym1. Mass spectroscopic analysis revealed an increase in immature preproteins in the mutants’ mitochondria. The functions of those proteins ran the gamut from ATP synthesis to maintenance of mitochondrial DNA, and oxidative stress response. In a cell-free system using mitochondrial extracts spiked with a known presequence peptide from the mitochondrial protein cytochrome oxidase 4 (Cox4), the researchers determined that the peptides failed to be degraded in extracts from Cym1-deficient yeast. These peptides, in turn, inhibited MPP’s ability to cleave preproteins. Interestingly, the yeast lacking Cym1 displayed many of the mitochondrial dysfunctions observed in people with AD, including boosted levels of reactive oxygen species (ROS), reduced oxygen consumption, and decreased membrane potential.

Because Aβ accumulates in AD mitochondria and is a known substrate for PreP (Cym1), the researchers tested whether an abundance of Aβ could thwart protein maturation in yeast mitochondria. In a cell-free system, the researchers found that Aβ was indeed degraded by Cym1. However, the peptide blocked the efficient degradation of the Cox4 preseqence peptide, indicating that Aβ may derail the turnover of presequence peptides, which in turn slows preprotein maturation. To confirm this in whole cells, they generated a yeast strain that produces Aβ peptides driven by the galactose promoter, and found that after three days on galactose, several immature proteins had accumulated. After five days, signs of mitochondrial dysfunction started to appear, including a 30 percent bump in ROS production and a dip in both membrane potential and oxygen consumption to 80 percent of wild-type levels.

The researchers next searched for hints of the same mechanism in human tissue. Using a combination of western blotting and mass spectrometry, they compared mitochondrial protein maturation in postmortem brain samples from four AD patients to that of four controls. Mossmann and Vögtle identified an abundance of several immature preproteins in AD patients, including the mitochondrial ribosomal subunit MRPL23, respiratory chain protein NDUFA9, several enzymes in the citric acid cycle, and MPP itself.

The human data suggests that this mechanism could be relevant in for AD, commented Xiongwei Zhu of Case Western Reserve University in Cleveland. However, he added that the impairment appeared to be limited to just several mitochondrial proteins, and wondered whether this apparent selectivity would be sufficient to cause the broad mitochondrial defects observed in AD.

Meisinger suggested that the maturation impairment probably was broader than reported in this initial paper. Because the researchers had access to limited amounts of each postmortem sample, identifying preproteins was difficult, especially since such preproteins are known to be rapidly degraded. He believes that many more proteins may have been affected. If the level of each mitochondrial preprotein increased by just a few percent, it could impair total mitochondrial proteostasis, he wrote in an email to Alzforum.

Russell Swerdlow of the University of Kansas in Kansas City praised the study. Swerdlow is a proponent of a mitochondrial hypothesis of Alzheimer’s disease (Swerdlow et al., 2014). He added that the findings stimulate more questions. “The authors do not discuss data that indicate mitochondrial function in AD subjects is also altered outside the brain, for example in platelets and fibroblasts,” Swerdlow said. “Does Aβ also account for that biochemical phenomenon?”

Meisinger told Alzforum that answering this question is a major focus of his lab. “Our main interest at the moment is to test if we can find immature mitochondrial proteins in blood cells from AD patients,” he said. If so, he predicted that such preproteins could eventually become useful diagnostics.—Jessica Shugart


  1. I’m always encouraged to see investigators consider the possibility that mitochondria may play a role in AD. Over the past decade, data has accumulated from a number of labs that Aβ localizes to mitochondria not only in transgenic animals and in vitro systems, but also in brains from human AD subjects. Based on this, more and more people are wondering whether mitochondria may mediate the toxic effects of Aβ and thereby play an important role in AD neurodegeneration. I’m sure that to many investigators an additional attractive feature of this view is that it is consistent with the amyloid cascade hypothesis. Under this scenario Aβ is still the upstream problem in AD, but you need to explain why and how it causes neurodegeneration, and mitochondria provide that explanation.

    This paper is consistent with this view.  From a molecular biology perspective it is a very impressive study.  The demonstration that mitochondrial protein presequence processing and peptide turnover can be functionally coupled is simply beautiful.  Showing that Aβ, when added to the systems studied, perturbs mitochondrial protein processing and induces protean mitochondrial changes further establishes a potential link back to AD.

    While I have no concerns about the experimental data, the authors do not discuss data that indicate mitochondrial function in AD subjects is also altered outside the brain, for example in platelets and fibroblasts.  Does Aβ also account for that biochemical phenomenon?  They also do not touch on the cybrid literature, which suggests AD-specific mitochondrial functional changes can perpetuate in cell culture independent of the addition of exogenous Aβ. 

    I don’t think this study is necessarily inconsistent with a mitochondrial cascade hypothesis that presumes mitochondrial dysfunction precedes Aβ production in AD.  I would also point out that the physiologic role of intracellular Aβ (or extracellular Aβ, for that matter) is poorly understood.  In this spirit, data from this paper could be consistent with the existing hypothesis that considers the possibility that one physiologic role Aβ plays is to essentially turn off or destroy dysfunctional mitochondria.

    Interpretative issues aside, I think this study does a very good job of justifying mitochondria as an AD therapeutic target.  If for no other reason, based on this and the basic biology it reveals, this is a very valuable paper.

    View all comments by Russell Swerdlow
  2. This is a very interesting and important paper. There is substantial evidence for mitochondrial deficits in Alzheimer's disease, however, the mechanism has been obscure. We and others found evidence for an impairment of cytochrome oxidase in both postmortem brain tissue and in platelets. However, it has been difficult to find mitochondrial DNA mutations to account for this observation. In collaboration with Doug Wallace, we found increased mutations in non-coding control regions of the mitochondrial genome, but sequencing of the three mtDNA encoded COX subunits has not shown point mutations. The possibility of another mechanism therefore has been important to explore. This paper presents a potential mechanism.

    As the authors report, most nuclear encoded mitochondrial proteins are imported with presequences that are then cleaved off by proteases in the mitochondrial matrix. The authors show that the peptidasome Cym1/PreP plays a critical role in this processing. It has also been shown that there are abnormalities in mitochondrial respiration and increased ROS production in AD. It is also known that PreP can degrade Aβ. The authors showed that Aβ impairs the processing of the presequence peptides. They showed that processing of COX4 preprotein is impaired in mitochondria from PS2/APP mice. They then showed that the presequence peptide for MDH2 accumulates in mitochondria from AD brains. In yeast mitochondria deficient in cym1, ATP production was impaired, respiration decreases, and ROS increased.

    I think that these findings may tie together a number of observations in AD, and link mitochondrial dysfunction to effects of Aβ in mitochondria. There is evidence for this, particularly in synaptic mitochondria. I think it would have been nice if the authors had showed that the impaired processing of COX4 presequence led to an impairment of COX activity in the transgenic mice and postmortem AD tissue. Nevertheless, these are important observations that suggest a mechanism by which Aβ can directly contribute to mitochondrial dysfunction in AD.  

    View all comments by M. Flint Beal
  3. Mitochondrial dysfunction is one of the most prominent and earliest deficits in Alzheimer's and likely plays a critical role in the pathogenesis of this disease. This interesting study demonstrated a novel mechanism by which Aβ causes mitochondrial dysfunction. Most mitochondrial proteins are encoded by the nucleus and are imported into mitochondria with the guidance of presequence that is cleaved and degraded after successful import. First, the authors used a yeast model to elegantly demonstrate a functional coupling between mitochondrial presequence processing and the presequence peptide turnover. They went on to convincingly show in yeast that Aβ could specifically inhibit the presequence peptide degradation, which in turn impaired the cleavage of presequence. Mitochondrial deficits ensued.

    More importantly, the authors confirmed that the processing of the presequence of several nuclear-encoded mitochondrial proteins was indeed impaired in brain mitochondria from AD patients, where mitochondrial Aβ is abundant, suggesting that such a mechanism is probably of pathophysiological relevance to human disease.

    It was unexpectedly found almost a decade ago that Aβ is present in mitochondria. In fact, Aβ could be imported through the mitochondrial import machinery and likely impairs mitochondrial function, at least in part, through specific interactions with several mitochondrial proteins such as ABAD and cypD. Mitochondria even have their own enzyme (e.g., matrix prepeptidase) to degrade Aβ. This study adds a new twist to the mitochondrial Aβ story in that it suggests this exact Aβ-degrading enzyme in mitochondria is also a target of the toxic effect of mitochondrial Aβ. This not only leaves the mitochondrial level of Aβ unchecked, but also results in dysfunctional mitochondrial preprotein maturation and an imbalance in the organelle proteome, which could collectively contribute to the pleiotropic mitochondrial deficits in AD.  This study suggests that mitochondrial matrix prepeptidase could be a critical target for preventing the toxic effects of mitochondrial Aβ. 

    While there is no question that this is a very intriguing idea and the authors convincingly demonstrated it in the yeast model, we need to be cautious about its implication for human disease without a similar mechanistic study in mammalian systems. Moreover, this study raises additional questions. For example, impaired processing of the mitochondrial presequence in AD brain mitochondria appeared limited to only several, instead of the full spectrum, of the mitochondrial proteins as would be expected due to the universal effect of mitochondrial Aβ. What causes such selectivity? What is the functional importance of such selectivity? Is such selective vulnerability sufficient to cause overall imbalance in mitochondrial proteome and pleotropic mitochondrial deficits in AD? One should also bear in mind that Aβ may impact mitochondrial function through other cytosolic mechanisms such as calcium-induced abnormal changes in mitochondrial dynamics.

    View all comments by Xiongwei Zhu

Make a Comment

To make a comment you must login or register.


News Citations

  1. Novel Aβ Protease Found in Mitochondria

Paper Citations

  1. . Global analysis of the mitochondrial N-proteome identifies a processing peptidase critical for protein stability. Cell. 2009 Oct 16;139(2):428-39. PubMed.
  2. . Decreased proteolytic activity of the mitochondrial amyloid-β degrading enzyme, PreP peptidasome, in Alzheimer's disease brain mitochondria. J Alzheimers Dis. 2011 Jan 1;27(1):75-87. PubMed.
  3. . The Alzheimer's Disease Mitochondrial Cascade Hypothesis: Progress and Perspectives. Biochim Biophys Acta. 2013 Sep 23; PubMed.

Further Reading


  1. . Processing of mitochondrial presequences. Biochim Biophys Acta. 2012 Sep-Oct;1819(9-10):1098-106. Epub 2011 Dec 7 PubMed.
  2. . Is Alzheimer's disease a systemic disease?. Biochim Biophys Acta. 2014 Sep;1842(9):1340-9. Epub 2014 Apr 18 PubMed.

Primary Papers

  1. . Amyloid-β peptide induces mitochondrial dysfunction by inhibition of preprotein maturation. Cell Metab. 2014 Oct 7;20(4):662-9. Epub 2014 Aug 28 PubMed.