When it comes to cleaning house, neurons are no slouches. Considering the number of neurodegenerative diseases that are triggered by the accumulation of protein garbage, it is clear that these cells cannot tolerate messes. For quick pickups, the ubiquitin-proteasome system tags and destroys some proteins. But, as two papers in this week’s Nature online demonstrate, keeping neurons spic, span, and healthy, requires sweeping with a broader broom.

That broom is autophagy, the process of bulk protein degradation, where bits of cytosol, including some organelles, are incorporated into lysosomes. Work from two independent Japanese groups shows that disrupting autophagy, by knocking out either of two genes critical to the process, causes widespread neurodegeneration in mice. In neurons, loss of this constitutive clearance system leads to an age-dependent accumulation of ubiquitinated proteins in the cytosol and the formation of inclusion bodies. Together, the two papers, one from the lab of Noboru Mizushima and the other from Keiji Tanaka, both at the Tokyo Metropolitan Institute of Medical Science, reveal that autophagy is essential to the normal function and health of neurons. Their work highlights the role of this housekeeping pathway in neurodegenerative disease, where it offers a potent counterforce to the accumulation of toxic proteins and formation of inclusion bodies (see ARF related news story).

Cells use two major protein degradation systems for quality control of new protein synthesis and disposal of old or damaged proteins. The ubiquitin-proteasome system removes individual proteins that have been marked for degradation. The other system, autophagy, relies on vesicles to engulf portions of cytoplasm and deliver them to lysosomes. Macroautophagy, one type of “self-eating,” is the main mechanism cells have for turning over long-lived proteins and organelles. Autophagy is activated in Huntington, Parkinson, and Alzheimer diseases, presumably as a protective response to the accumulation of toxic proteins or aggregates (see ARF related news story and Yu et al., 2005).

Autophagy is an important cell homeostatic process, and knockout of either of two key genes in the autophagy pathway, Atg5 or Atg7, causes perinatal death in mice. To look at the function of autophagy specifically in neurons, the researchers, working with collaborators across Japan, used the nestin promoter and a Cre/lox system to produce live mice with either Atg5 (in the case of the Mizushima group) or Atg7 (by Tanaka and colleagues) deleted specifically in neurons. They showed that most neurons lacked Atg proteins and autophagic activity. Both sets of mice revealed strikingly similar phenotypes of neurodegeneration and inclusion body formation starting early in life. By three weeks, the Atg5-lacking mice displayed problems with balance, coordination, and strength, including abnormal walking, limb clasping defects, and failure to balance on a rotating rod for even a short time. Atg7 knockouts had very similar progressive deficits, but were even sicker, showing lowered survival by 4 weeks of age, and most died by 28 weeks old.

For the Atg5 knockouts, the gross anatomy of their brain was normal, but on closer inspection the researchers, led by first author Taichi Hara, found evidence of neuron loss, most prominently in the cerebellar Purkinje cells, and apoptosis (in neighboring granular layer cells). Hippocampal pyramidal cells were also affected, and throughout the brain there was evidence of axonal swelling. In the case of the Atg7 knockout, joint first authors Masaaki Komatsu and Satoshi Waguri documented a more pronounced neuronal loss in both the cerebellum and cerebral cortex, and saw apoptosis of neurons in the cortex.

Staining of brain tissues with anti-ubiquitin antibodies revealed a dramatic phenotype of progressive accumulation of large, ubiquitin-positive inclusion bodies throughout the brain. The aggregates of misfolded proteins were seen in neurons but not glial cells, consistent with the conditional knockout of the two genes and loss of autophagy only in neurons. In the Atg5 knockout, inclusion formation was preceded by an increase of diffuse ubiquitin staining, starting in embryonic cells. From this data, Hara et al. conclude that the primary cellular phenotype of Atg5-deficient neurons is the accumulation of diffuse abnormal cytosolic proteins, with inclusion body formation occurring later. Their results suggest that the main function of autophagic vesicles is to take up and destroy diffuse proteins, not to gobble up inclusion bodies, as has been previously suggested. Formation of inclusion bodies follows on the inadequate clearance of the soluble proteins. Experiments in mice with an inducible liver-specific loss of Atg5 supported this idea, as those mice also demonstrated a rapid build-up of diffuse cytosolic ubiquitinated proteins, followed by formation of inclusion bodies after induction. Komatsu et al. showed that Atg7 deficiency did not affect proteasome function, so that the accumulation of ubiquitinated proteins was a direct result of ineffective autophagy.

If failure of autophagy leads to the accumulation of ubiquitinated proteins, the formation of inclusion bodies, and neurodegeneration in otherwise normal neurons, what about its role in human neurodegenerative disease? From a number of studies, it’s clear that this constitutive housecleaning pathway has an important role to play in conditions like Parkinson and Huntington diseases, which feature polyubiquitinated inclusions. Autophagy also contributes to the metabolism of amyloid-β peptides in Alzheimer disease. Attempts to boost autophagy using the drug rapamycin show some beneficial effects in enhancing clearance of toxic proteins in models of neurodegeneration (see ARF related news story and Berger et al., 2006). Going forward, it will be essential to sort out the contributions of autophagy versus proteasomal degradation pathways (see ARF related news story) in diseases where cells are fighting off the assault of mutated or misfolded proteins.—Pat McCaffrey


  1. The extreme scarcity of autophagic vacuoles in normal brain and their appearance in states of disease have previously led many to assume that autophagy in neurons is mainly an inducible process. Autophagy is solely responsible for organelle turnover, however, and the large cytoplasmic mass of neurons would suggest, therefore, that autophagy might have a significant constitutive component. The two papers by Komatsu et al. and Hara et al. have now provided elegant and definitive evidence in neurons for constitutive autophagy and have demonstrated that it is required for neuron survival. In fact, the results imply that the brain may actually be one of the tissues most vulnerable to a possible impairment of autophagy. These findings, therefore, offer insight into why neurons are preferentially victimized in diseases that disrupt the lysosomal system, even when the disease is a systemic one.

    This new evidence for actively ongoing autophagy in neurons, which normally proceeds in the absence of readily detectable morphological intermediates (i.e., autophagic vacuoles), indicates that this process in healthy neurons is exceptionally efficient. Another implication from these observations is that autophagic vacuole accumulation in neurodegenerative disease states may signify a failing autophagy system, rather than simply an activation of autophagy as is frequently proposed. The findings are highly relevant to Alzheimer disease (AD) where autophagic function is impaired as evidenced by a massive build-up of autophagy intermediates especially within dystrophic dendrites of affected neurons. This indicates that the usually efficient progression of autophagosomes to lysosomes is impeded (Nixon et al. 2005). Autophagosome-lysosome fusion is already known to be slowed by normal cell aging (Martinez-Vincente et al. 2005) and additional risk factors for AD are likely to be found to impair autophagy. Autophagic vacuoles are highly enriched in γ-secretase and actively generate Aβ during autophagy (Yu et al., 2005). Although normally most of the generated Aβ would be degraded within lysosomes, in AD and transgenic AD models, the marked build-up of autophagic intermediates within an impaired autophagy pathway is a significant source and intracellular reservoir of Aβ (Yu et al., 2005). The two new papers, in the context of these other recent observations, therefore, support potential links between autophagic failure and neurodegeneration, amyloidogenesis, and possibly the intracellular accumulation of other disease-related proteins in AD. Therapeutic strategies based on facilitating efficient autophagy show glimpses of promise in neurodegenerative disease models (e.g., Ravikumar et al., 2004).


    . Extensive involvement of autophagy in Alzheimer disease: an immuno-electron microscopy study. J Neuropathol Exp Neurol. 2005 Feb;64(2):113-22. PubMed.

    . Protein degradation and aging. Exp Gerontol. 2005 Aug-Sep;40(8-9):622-33. PubMed.

    . Macroautophagy--a novel Beta-amyloid peptide-generating pathway activated in Alzheimer's disease. J Cell Biol. 2005 Oct 10;171(1):87-98. PubMed.

    . Inhibition of mTOR induces autophagy and reduces toxicity of polyglutamine expansions in fly and mouse models of Huntington disease. Nat Genet. 2004 Jun;36(6):585-95. PubMed.

  2. The recent papers by the Mizushima and Tanaka labs provide compelling support of a role for autophagy in the constitutive turnover of cellular material and the importance of this process in maintenance of neuronal health. However, the conclusion that autophagy has no role in the clearance of inclusion bodies is premature. There is now strong evidence from conditional models of polyglutamine disease (e.g., Yamamoto et al., 2000 and Zu et al., 2004) to indicate that neurons can eliminate inclusion bodies and can recover from the toxic effects of aggregated protein—once expression is turned off. While there is no direct evidence yet that autophagy is required for this process, the Mizushima and Tanaka groups are now in an excellent position to test this hypothesis.


    . Reversal of neuropathology and motor dysfunction in a conditional model of Huntington's disease. Cell. 2000 Mar 31;101(1):57-66. PubMed.

    . Recovery from polyglutamine-induced neurodegeneration in conditional SCA1 transgenic mice. J Neurosci. 2004 Oct 6;24(40):8853-61. PubMed.

  3. This pair of papers shows that disruption of the autophagy pathway
    through deletion of the genes that encode critical components of the
    pathway (i.e., either Atg5 or Atg7) within neurons leads to
    behavioral abnormalities, neurodegeneration, and inclusion formation.
    The papers are interesting for several reasons.

    First, although features of autophagy are known to be involved in
    normal protein turnover and may be part of a coping response to
    nutrient deficiency, it also is believed to be a programmed cell death
    pathway. Thus, it was unclear whether disruption of this pathway
    would lead to greater cell death because of impaired protein turnover
    or greater cell survival as seen when another cell death pathway,
    apoptosis, is disrupted. The fact that abnormal protein accumulation
    and greater cell death is seen indicates that autophagy plays a
    critical role in normal protein turnover in mammalian systems.

    Second, ubiquitin immunoreactive inclusions were found in both Atg5-
    and Atg7-deficient mice, despite the fact that these mice were not
    known to otherwise harbor a genetic mutation producing
    aggregation-prone proteins. Although not shown in these papers,
    inclusion formation can be observed following inhibition of the
    proteasome, the other major protein degradation pathway. One of the
    two groups examined proteasome function in the Atg7-deficient mice
    and found no proteasome impairment (although it would have been interesting to examine proteasome function in vivo). Taken at face value, these results add
    further support to the notion that inclusion formation is a
    downstream cellular response to the accumulation of proteins that are
    otherwise destined for degradation. Notably, the paper by Tanaka and
    colleagues found that the accumulation of "diffuse" cytosolic
    ubiquitin immunoreactivity occurred first, before inclusion
    formation, and was a more consistent phenotype of autophagy
    disruption than inclusion formation.

    Although it is generally assumed that the proteins that are
    ubiquitinated and accumulate in inclusion bodies are misfolded and
    possibly non-functional, the finding here raises the provocative
    possibility that inclusions may form, in part, from normal proteins
    that accumulate when degradation is impaired. In such a scenario,
    pathogenesis might arise from having too much of a good thing.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Eat 'Em Up Early—Autophagy Might Delay Huntington's Disease
  2. Lysosomes and Proteasomes Compete for PD Researchers' Attention

Paper Citations

  1. . Macroautophagy--a novel Beta-amyloid peptide-generating pathway activated in Alzheimer's disease. J Cell Biol. 2005 Oct 10;171(1):87-98. PubMed.
  2. . Rapamycin alleviates toxicity of different aggregate-prone proteins. Hum Mol Genet. 2006 Feb 1;15(3):433-42. PubMed.

Further Reading


  1. . Macroautophagy--a novel Beta-amyloid peptide-generating pathway activated in Alzheimer's disease. J Cell Biol. 2005 Oct 10;171(1):87-98. PubMed.
  2. . Extensive involvement of autophagy in Alzheimer disease: an immuno-electron microscopy study. J Neuropathol Exp Neurol. 2005 Feb;64(2):113-22. PubMed.
  3. . Autophagic vacuoles are enriched in amyloid precursor protein-secretase activities: implications for beta-amyloid peptide over-production and localization in Alzheimer's disease. Int J Biochem Cell Biol. 2004 Dec;36(12):2531-40. PubMed.

Primary Papers

  1. . Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature. 2006 Jun 15;441(7095):885-9. PubMed.
  2. . Loss of autophagy in the central nervous system causes neurodegeneration in mice. Nature. 2006 Jun 15;441(7095):880-4. PubMed.