Hitchhiking on selective autophagy

A Publisher Correction to this article was published on 19 April 2018

This article has been updated

Selective autophagy is important for controlled degradation of cellular components. However, a selective autophagic degradation mechanism for ribosomes in mammals has remained unclear. A study now describes non-selective and selective ribosome degradation and a significant role for ‘bystander’ non-selective autophagy.

Autophagic processes entail the engulfment of cellular components into double-membrane vesicles, called autophagosomes1. These then fuse with lysosomes to degrade their contents. Autophagy can be divided into non-selective (bulk) autophagy and selective autophagy that specifically targets damaged or superfluous organelles or cellular structures, such as mitochondria (mitophagy), peroxisomes (pexophagy), lysosomes (lysophagy) or protein aggregates (aggrephagy)2,3. Bulk autophagy occurs on a basal, cell-type-dependent level and comprises non-specific engulfment of cellular content by autophagosomes as well as delivery to, and degradation by, lysosomes1. It is regulated by various physiological stimuli, such as starvation, which increases bulk autophagy to supplement the amino acid pool. As such, non-selective autophagy serves an important function in cellular homeostasis following different types of stresses. In contrast, selective autophagy relies on recognition of ‘eat-me’ signals on the cargo by specific autophagy receptors that then bind to lipidated ATG8 homologues, such as LC3, on autophagosomes (Fig. 1a). ATG8 lipidation involves ATG7 and ATG3, is regulated by the ATG12–ATG5–ATG16 complex, and ultimately leads to ATG8-coated autophagosomes. Strikingly, autophagy also occurs in cells deficient for ATG5 and ATG7, despite the loss of ATG8 lipidation4, increasing the importance of understanding the role of ATG5 and ATG7 in selective autophagy processes. Whereas signalling factors, autophagy receptors and modulating factors have been identified and characterized for several selective autophagy pathways3, the molecular basis of selectivity in ribophagy remains unclear5. In yeast, nitrogen starvation causes activation of a selective ribophagy pathway, dependent on ATG7 and the AAA ATPase, Cdc48. This process is controlled by ribosome ubiquitination — Ltn1-mediated ubiquitination prevents ribophagy, whereas deubiquitination by Ubp3–Bre5 drives ribophagy6,7,8. It is not known whether similar mechanisms could also govern a selective ribophagy pathway in mammalian cells.

Fig. 1: Selectivity of ribophagy pathways and selective autophagy in general.

a, Molecular mechanisms driving non-selective/bulk (top) and selective autophagy pathways (bottom). In selective autophagy, autophagy receptors selectively bind cargo and recruit it into autophagosomes by binding to lipidated ATG8. b, Summary of the effects on ribophagy elicited by different stresses and described by An and Harper9. Increase in ribophagy is ATG5/ATG7-dependent, as indicated. ER, endoplasmic reticulum. c, Depiction of bystander autophagy observed during selective autophagy (mitophagy) by An and Harper9. Damaged and marked (ubiquitinated) mitochondria are recognized by autophagy receptors and targeted to autophagosomes. However, autophagosomes also engulf bystander cargo, such as ribosomes, proteasomes and actin. Similar processes are also observed under ribophagy, pexophagy and lysophagy.

In this issue of Nature Cell Biology, An and Harper9 describe a systematic study to search for a putative selective ribophagy pathway in mammalian cells under various stress conditions. To monitor ribophagy flux, the authors used an innovative approach utilizing endogenously Keima-tagged 40S and 60S ribosomal subunits. Keima is a coral-derived fluorescent protein that exhibits a shift in excitation wavelength following pH change, which can be used to monitor protein flux to the lysosome and has been widely used in the study of mitophagy10,11. Employing these Keima-tagged ribosome reporters, the authors were able to show that bulk autophagy, induced by inhibition of mTOR signalling, increases ribophagy flux to a similar extent as that observed for non-related cytosolic cargo. They found that this increase in ribophagy flux is largely unaffected by ATG5 depletion, confirming that the observed mechanism represents non-selective autophagic ribosome degradation (Fig. 1b).

Unlike the induction of bulk autophagy, selective autophagy pathways are typically driven by accumulation of aberrant or damaged structures (for example, protein aggregation or loss of mitochondrial membrane potential) that lead to exposure of specific ‘eat-me’ signals, which are then detected by autophagy receptors3. Most likely, such signals are missing from ribosomes under conditions in which they remain largely functional. For this reason, An and Harper9 induced ribosomal dysfunction by different types of stresses. Utilizing this approach, they were able to identify specific conditions under which a potential selective ribophagy mechanism is induced in mammalian cells.

The first such condition is induction of proteotoxic stress by arsenite. Treatment with sodium arsenite is known to cause translational attenuation and formation of stress granules (SGs) composed of cytosolic RNA–protein complexes. An and Harper9 show that ATG5-dependent autophagic degradation of both 40S and 60S ribosomal subunits is triggered under these conditions — suggesting the function of a selective ribophagy mechanism. One could speculate that the observed phenomenon represents granulophagy rather than ribophagy, as previous research shows that SGs undergo selective autophagy in an ATG7-dependent manner12. However, the 60S subunit was not found to initiate SG formation, and An and Harper9 clearly observed both subunits being degraded with levels of arsenite that do not induce SGs or ribosome aggregates, strongly pointing towards a ribophagy process.

The second condition found by An and Harper9 to induce selective ribophagy is the induction of chromosomal mis-segregation by reversine. Reversine causes unbalanced karyotypes known to trigger autophagy13. Similar to arsenite treatment, the authors identified an ATG5-dependent increase in 60S ribosomal subunit flux following reversine application.

In contrast, selective ribophagy could not be detected when translation was perturbed by treatment with a nascent chain terminator, puromycin; translation elongation inhibitor, cycloheximide; or by prevented extraction and clearance of aberrant ubiquitinated nascent chains through inhibition of AAA ATPase, p97. Thus, An and Harper found that ribosomal malfunction was not sufficient to expose signals for their selective degradation, consistent with previous findings in yeast8.

Following this demonstration of the existence of mammalian selective ribophagy, it is essential to identify specific receptors acting in ribophagy and to decipher the underlying molecular mechanisms. Previous studies in yeast showed an important role for ribosome ubiquitination in starvation-induced ribophagy, possibly through masking the protein surface detected by ribophagy receptors in a process driven by Ubp3–Bre5 (interacting with Cdc48) and Ltn1 (ref.8). In contrast, An and Harper9 show that neither inhibition of p97 (an analogue of yeast Cdc48) alone, nor p97 inhibition in the context of starvation, has an effect on ribophagy, unlike the situation in yeast7. However, the specific degradation of ribosomes following arsenite treatment was shown to be dependent on p97 activity, supporting a role for p97 in selective ribophagy. In addition, inhibition of USP10 (Ubp3 in yeast) reduces starvation-induced ribophagy, as observed in yeast6. However, it may be more likely that this effect is due to the role of USP10 in controlling the Beclin1–Vps34 complex14, rather than a direct effect on ribosome ubiquitination.

To further determine the selectivity of mammalian ribophagy compared to other selective autophagy mechanisms, An and Harper9 generated a set of autophagy flux reporters for different cellular compartments or for autophagy cargoes by fusing Keima to LC3 (autophagosomes), RPS3 and RPL28 (ribosomes), TOMM20 (mitochondria), LDHB (cytosol), ACTB (actin cytoskeleton), or PSMD12 (19S proteasome) (Fig. 1c). They used these reporters in cells treated with mTOR inhibitor Torin1 or arsenite, thereby specifically triggering a non-selective form of autophagy versus a selective form. Strikingly, the authors found that all reporters show a similar flux increase in both treatments. Only in the case of ribosomal flux was an increase observed following arsenite treatment when compared to mTOR inhibition, suggesting a certain degree of specificity. This brings back several longstanding questions in the field: how selective is selective autophagy? Are there mechanisms preventing engulfment of non-targeted cargo, so called ‘bystanders’? Do autophagosomes initiated by a selective mechanism also engulf these bystanders, in a manner similar to bulk autophagy3? These observations led the authors to monitor the flux of the same diverse cargoes following selective induction of mitophagy, pexophagy and lysophagy by chemical agents typically used to promote degradation of these organelles. Thus also revealing that extensive bystander autophagy occurred under these conditions. These observations support the plausible model that, to a large extent, induction of selective autophagy leads to bulk degradation of a portion of the cytosol (that is, so called bystander cargo) with only some selective enrichment of specific cargo through receptor-mediated delivery to autophagosomes. However, it is also possible that some chemical agents used to promote selective autophagy may trigger more general proteotoxic stress, which then induces bulk autophagy in parallel with a selective form of autophagy. In this context, it will be interesting to apply autophagosomal content profiling technology, recently published by Le Guerroué et al.15, to further explore the quantitative cargo composition following stimuli for specific selective autophagy mechanisms.

Taken together, the results by An and Harper9 provide critical insight into specific stress conditions that lead to selective increase in ribophagic flux in mammalian cells. An additional take-home message can be extrapolated from the important role of bystander autophagy leading to significant degradation of non-specific cargo in selective autophagy — always test alternative cargo before drawing any conclusions about the selectivity of an autophagy process.

Change history

  • 19 April 2018

    In the version of this News & Views originally published, owing to a technical error some incorrect characters were introduced into the second sentence of the main text, starting “Autophagy can be divided into non-selective...”, where “caspringDE@2017n” should have read ‘can’. This has now been corrected in all versions of this News & Views.


  1. 1.

    Lamb, C. A., Yoshimori, T. & Tooze, S. A. Nat. Rev. Mol. Cell Biol. 14, 759–774 (2013).

    CAS  Article  PubMed  Google Scholar 

  2. 2.

    Dikic, I. Annu. Rev. Biochem. 86, 193–224 (2017).

    CAS  Article  PubMed  Google Scholar 

  3. 3.

    Anding, A. L. & Baehrecke, E. H. Dev. Cell 41, 10–22 (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  4. 4.

    Nishida, Y. et al. Nature 461, 654–658 (2009).

    CAS  Article  PubMed  Google Scholar 

  5. 5.

    Stolz, A., Ernst, A. & Dikic, I. Nat. Cell Biol. 16, 495–501 (2014).

    CAS  Article  PubMed  Google Scholar 

  6. 6.

    Kraft, C., Deplazes, A., Sohrmann, M. & Peter, M. Nat. Cell Biol. 10, 602–610 (2008).

    CAS  Article  PubMed  Google Scholar 

  7. 7.

    Ossareh-Nazari, B. et al. EMBO Rep. 11, 548–554 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Ossareh-Nazari, B. et al. J. Cell Biol. 204, 909–917 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  9. 9.

    An, H. & Harper, J. W. Nat. Cell Biol. (2017).

  10. 10.

    Katayama, H., Kogure, T., Mizushima, N., Yoshimori, T. & Miyawaki, A. Chem. Biol. 18, 1042–1052 (2011).

    CAS  Article  PubMed  Google Scholar 

  11. 11.

    Narendra, D. P., Wang, C., Youle, R. J. & Walker, J. E. Hum. Mol. Genet. 22, 2572–2589 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Buchan, J. R., Kolaitis, R.-M., Taylor, J. P. & Parker, R. Cell 153, 1461–1474 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Santaguida, S., Vasile, E., White, E. & Amon, A. Genes Dev. 29, 2010–2021 (2015).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Liu, J. et al. Cell 147, 223–234 (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. 15.

    Le Guerroué, F. et al. Mol. Cell 68, 786–796 (2017).

    Article  PubMed  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Christian Münch or Ivan Dikic.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Münch, C., Dikic, I. Hitchhiking on selective autophagy. Nat Cell Biol 20, 122–124 (2018).

Download citation

Further reading


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing