NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Madame Curie Bioscience Database [Internet]. Austin (TX): Landes Bioscience; 2000-2013.

Cover of Madame Curie Bioscience Database

Madame Curie Bioscience Database [Internet].

Show details

Origin and Evolution of Self-Consumption: Autophagy

and *.

Corresponding Author: Department of Biochemistry, The Norwegian Radium Hospital, Montebello, N-0310 Oslo, Norway. Email:


While misfolded and short-lived proteins are degraded in proteasomes located in the nucleus and cytoplasm, the degradation of organelles and long-lived proteins in the lysosome occurs by the process of autophagy. Central and necessary to the autophagic process are two conserved ubiquitin-like conjugation machineries. These conjugation machineries appear to be specific for autophagy and can together with genetic and morphological data be used to trace the natural history of autophagy. Here we discuss the origin and evolution of autophagy.


The term autophagy, meaning eating (phagy) oneself (auto), refers to the transport of cytoplasmic components to the degradative organelle called the vacuole in yeast and plants, and lysosome in other eukaryotes. For simplicity we will use the term lysosome in this chapter. The transport of cytoplasmic cargo can be membrane-driven or carried out by selective transport of certain proteins over the lysosomal membrane (Fig. 1). The latter process, called chaperone-mediated autophagy, is unrelated to other autophagy pathways and will not be further discussed. Autophagosomes en route to the lysosome can readily be detected and recognised by electron microscopy and have been observed mis-regulated in a number of pathological conditions like cancer, myopathies, neurodegeneration and bacterial infections.1 The relevance of autophagy in these disease conditions has been unclear until recently since no good markers or means of manipulating the activity of the process were known. The ability to study autophagy was greatly increased by the identification of the central autophagy machinery and the identification of specific subcellular markers for autophagy in yeast. Autophagy is now currently being intensively studied in a number of model organisms.2

Figure 1. Schematic model of proteasomal and autophagic degradation pathways in pro- and eukaryotes.

Figure 1

Schematic model of proteasomal and autophagic degradation pathways in pro- and eukaryotes. Proteins are degraded in prokaryotes by chambered proteases termed the 19S proteasome upon recognition of primary amino acid target sequences (red). The more complex 26S (more...)

To gain a better foundation for understanding the mechanism and relevance of autophagy to extant eukaryotes, it is useful to look at the evolution and presence of autophagy across different groups. In this chapter, we first review the different modes of autophagy and the specific molecular machinery involved. To assign the presence or absence of the process in different groups of organisms, we rely on positive identification of autophagy based on morphological and genetic data as well as presence of the autophagy-specific molecular machinery within taxa. We then infer the likely origin and subsequent loss of autophagic capacity within eukaryotic lineages based on the current understanding of eukaryotic phylogeny.3,4 Finally, we discuss the relationship of autophagy to that of other degradation pathways used for degradation of transmembrane proteins (endocytosis) and short-lived proteins (proteasomes).

Degradation of Proteins in Chambered Proteases in Pro- and Eukaryotes

Degradation of intracellular proteins and organelles is important for cell homeostasis, as it regulates enzymatic activity, removes toxic or misfolded proteins and produces free amino acids. Both bacterial and eukaryotic cells degrade proteins by the use of proteases.5 A set of threonine proteases work in concert in a superstructure called the proteasome, with a catalytically active lumen. By controlling entry of proteins into the proteasomal lumen, general cytoplasmic proteins are shielded from its activity. Whereas proteins in bacteria destined for degradation carry degradation signals embedded in their primary sequence, eukaryotes have evolved a protein conjugation process which tags proteins with polyubiquitin allowing specific recognition and degradation by the proteasome.

Different Modes of Autophagy Degrade Long-Lived Proteins and Organelles in Eukaryotes

Eukaryotic cells differ from prokaryotes by the presence of membrane bound organelles and a nucleus. Turnover and degradation of these structures cannot be performed by the proteasome and instead occurs by the process of autophagy. Autophagy occurs by the transport of cytoplasmic components into the lysosome for degradation. Two modes of autophagy exist, macroautophagy and microautophagy (Fig. 1, reviewed in refs. 2, 6). Microautophagy is the direct engulfment of cytoplasm and organelles by the lysosomal membrane. Macroautophagy, on the other hand, involves the isolation of a part of the cytoplasm by a cup-shaped double membrane structure and its subsequent degradation in the lysosome. Upon closure of the isolation membrane, the resulting autophagosome fuses with the lysosomal membrane releasing the inner vesicle with the cargo that is ultimately degraded. The capture of cytoplasmic components is unspecific and generally consists of endoplasmic reticulum, cytosol and organelles like mitochondria. Two variants of macroautophagy called cytoplasm-to-vacuole targeting (Cvt) and pexophagy differ from macroautophagy by being specific in cargo selection and by forming smaller autophagic vesicles. The Cvt pathway in the yeast Saccharomyces cerevisiae is a constitutive biosynthetic transport pathway for the vacuolar-resident enzyme aminopeptidase I. To date no evidence for this pathway has been found in other organisms than S. cerevisiae.2 Pexophagy is, as the name suggests, specific autophagy of peroxisomes and has been studied extensively in the yeasts Pichia pastoris, Hanensula polymorpha, and Pichia methanolica.7 Pexophagy in yeast, is induced by the rapid adaptation from a feeding substrate requiring large amounts of peroxisomes to one that does not. Based on biochemical data, selective pexophagy has also been suggested to occur in rat hepatocytes.8

The Autophagy-Specific Molecular Machinery Involves Two Ubiquitin-Like Conjugation Pathways

Genetic studies in yeast for genes necessary for macroautophagy, Cvt and pexophagy have revealed overlapping genetic requirements for all the pathways (reviewed in refs. 6, 9). Many of the genes involved in completion of autophagy are clearly involved in other cellular processes and thus serve as poor indicators for the presence of autophagic capacity. Others, here termed the central autophagy machinery, appear specific for autophagy. Yeast mutants for genes of the central machinery, and thus autophagy deficient, grow normally under nonstarvation conditions. Moreover, this apparent nonrequirement in other developmental processes extends to metazoa since mouse mutants not capable of performing autophagy develop virtually normally until birth, upon which they appear to die from starvation.10 The central autophagy machinery necessary for the membrane dynamics of autophagosome formation, consists of two ubiquitin-related conjugation pathways (Fig. 2, reviewed in refs. 11,12). The ubiquitin related molecules (Ubl) Atg12 and Atg8 both share limited sequence identity with ubiquitin but engage in protein modification systems strikingly similar to that of ubiquitin. The carboxy-terminal glycine of Atg12 is covalently attached to a lysine in the centre of Atg5 (Fig. 2).13 Like in the ubiquitin conjugation pathway the C-terminal glycine of Atg12 is first activated by an E1 enzyme, Atg7, creating a high energy thioester bond. Subsequently Atg12 is transferred to a E2-like enzyme called Atg10 forming a second thioester bond. The conjugation of Atg12 to the target Atg5 is necessary for the progression of the second conjugation reaction of Atg8 to its substrate phosphatidylethanolamine (PE). A cysteine protease, Atg4, is necessary for activation of Atg8 revealing the C-terminal glycine. As for Atg12, the E1-like enzyme transferring Atg8 to a E2 like enzyme is Atg7 making it the only known E1 like enzyme to be conjugating two separate Ubl's. The E2 enzyme of Atg8, Atg3 transfers the Atg8 to phosphatidylethanolamine. Finally, like the case for ubiquitin, Atg8 is deconjugated by a protease and can potentially be used for another round of conjugation. The protease responsible is again Atg4.

Figure 2. Comparison of conjugation pathways of selected ubiquitin superfamily members and the biosynthetic pathway of thiamine.

Figure 2

Comparison of conjugation pathways of selected ubiquitin superfamily members and the biosynthetic pathway of thiamine. The ubiquitin-like proteins, Atg12 and Atg8 necessary for autophagy undergo conjugation reactions closely resembling that of bacterial (more...)

The genetic requirements for microautophagy have mostly been studied in the context of micropexophagy in Pichia pastoris, and overlap with that of macroautophagy in requiring the Atg8 conjugation machinery.14,15 Indeed, Atg8 gets recruited to the site of microautophagy on the lysosomal membrane and its activation by Atg4 is needed for micropexophagy.14 It is less clear whether the Atg12 conjugation machinery is needed. To date, neither Atg5, Atg10, nor Atg12 has been shown to be required for micropexophagy or the Cvt pathway.9 This suggests that the Atg12 conjugation machinery is not strictly required for Atg8 conjugation and insertion into autophagosomal membranes in all types of autophagy.

In summary, it appears reasonable to assume that the presence of both the Atg12 and Atg8 machineries reflects the capacity for both macro and microautophagy and the presence of the Atg8 machinery alone reflects a capacity to conduct only microautophagy.

Is Autophagy a Pan-Eukaryotic Process?

Since prokaryotes neither have internal membranes nor lysosomes, autophagy necessarily has originated at a later point of evolution. The question arises whether autophagy is a pan-eukaryotic process and would be present in the last common eukaryotic ancestor (LCEA). Autophagy can be positively identified on morphological grounds based on electron microscopy, presence of autophagy-specific genes and genetic evidence for their necessity for autophagy in the respective organisms. The identification of the central autophagy machinery in yeast spurred the molecular genetic study of autophagy in other model systems.7 Homologues of many of the central autophagy encoding genes like Atg5, Atg7, Atg10, Atg12 and Atg8 are present in all eukaryotic model organisms investigated to date (reviewed in ref. 2). Genetic, morphological and sequence-based evidence for autophagy exists from the major eukaryotic groups of plants (Arabidopsis thaliana), amoebozoa (Dictyostelium discoideum), fungi (S. cerevisiae, H. polymorpha, P. pastoris and P. methanolica), metazoa (Caenorbhabditis elegans, Drosophila melanogaster and Mus musculus).2,16,17 The above findings are all obtained in model organisms used for molecular genetic cell biological studies. The situation is less clear when looking at model organisms within excavata and chromalveolata. Often, these organisms are chosen because they are important parasites relevant to human disease. Based on ultrastructural observations, macroautophagy has been suggested to occur in both Leishmania donovani (excavata) and Tetrahymena thermophila (chromalveolata) upon induction of cell death. These findings have, however, not been confirmed by molecular means.18,19 Since the genome of L. donovani is not being sequenced a meaningful search for presence of autophagic genes within this organism is not feasible yet.20 We searched for homologues of the core autophagic machinery in genomes within excavata and chromalveolata (Fig. 3A). We find that the genomes of T. thermophila and T. pseudonana, that are both free-swimming chromalveolates, encode homologs for the proteins of both the ATG8 and the ATG12 pathways. Interestingly, the situation is different in intra- and extracellular parasites within both taxa. The genomes of T. cruzi, L major and T. vaginalis, and the chromalveolate P. falciparum, encode only homologs for the proteins of the ATG8 pathway.21,22 Thus, there is a clear tendency for parasitic species whether excavata or chromalveolata to have lost the ATG12 pathway. G. lamblia appears to have also lost the proteins of the ATG8 pathway. This is perhaps not so surprising given that G. lamblia is known to lack organelles, such as mitochondria, peroxisomes and lysosomes, normally found in eukaryotes.23

Figure 3. Evolutionary distribution of autophagy.

Figure 3

Evolutionary distribution of autophagy. a) The presence or absence of the Atg12 pathway, Atg8 pathway, and Atg7 needed for both pathways in each taxon, are illustrated (full circle: strong evidence of at least one homolog, half circle: weak evidence of (more...)

The Atg12 machinery is strictly necessary for macroautophagy, but has to our knowledge not been shown to be necessary for microautophagy. By extension, it suggests that microautophagic, but not macroautophagic capacity is present within several parasites. Interestingly, one of the most studied functions of macroautophagy is to provide free amino acids upon starvation conditions by unspecific consumption of cytosol. The extracellular parasites G. lamblia and T. vaginalis and the intracellular parasites L. major, T. cruzi, P. faliciparum and C. Pavrum all spend parts or all their lifecycle intracellularly, or bathed in body fluids of the host(s). Presumably then, they are living in an environment where supply of amino acids is not limiting. It could be that macroautophagy for this reason is not strictly necessary for these highly specialized parasites. In line with this idea, neither members of the Atg12 nor the Atg8 group were found by BLAST analysis of the highly reduced genome of the obligate intracellular parasite Encephalitozoon cuniculi, which is considered to be a member of a fungi sister group.24

In summary, present available genomic information as well as ultrastructural and genetic evidence exists for both macro- and microautophagy within bikonts and unikonts (Fig. 3B). The core autophagy machinery was therefore present in the LCEA. Partial or complete loss of autophagic capacity, most likely have occurred secondarily in parasitic species within both bikonts and unikonts. We postulate that macro- and microautophagy will be present in most nonparasitic eukaryotes.


Sampling core molecular machineries used for distinct cellular processes across eukaryotic taxa allows deduction of the minimal set of these molecules present in the LCEA. Caution should be shown when interpreting the original function of these machineries, since cooption of molecules for new functions is one of the main driving forces of evolution. This caveat can be partially counteracted by ensuring that the molecules involved perform the same functions within both unikonts and bikonts by genetic studies, since it would be less likely that the same function would be acquired independently.

Using this approach, it has been proposed that the LCEA contained a complex endomembrane system involved in both exocytic and endocytic events.25 Autophagy is intimately linked with endocytosis. Not only do the two pathways end up in the same acidic degradative compartment, there is also evidence for fusion of autophagosomes with late endocytic compartments before the cargo is released in lysosomes.26 It is therefore important to understand the extent to which the endocytic machinery was present in the LCEA. The core components of endocytosis including specific components of early endosomal, as well as recycling and sorting endosomal compartments were all present in the LCEA (see the chapter by Field, Gabernet-Castello and Dacks in this book). Importantly, the HOPS SNAREcomplex of proteins necessary for the fusion of autophagosomes with lysosomes was also present.27 Thus, the wide pan-eukaryotic distribution of the core autophagic machinery and the essential endosomal machinery with which it interacts argues for the presence of autophagy in the LCEA.

An interesting and related question is: what was the original function of autophagy? Perhaps the most widely known role of autophagy is its involvement in survival under starvation conditions. Unspecific macroautophagy of cytoplasm can be induced in plants, amoeba, fungi, insects and mammals allowing intracellular components to be rapidly recycled and reused during starvation. 2,10,16 This allows near normal activity of cells even though steady supply of nutrients is not available. It is conceivable that keeping up cellular activity rather than forming spores during starvation conditions gave a selective advantage in early eukaryotic evolution by allowing cells to localise to new foraging locations while consuming intracellular pools of amino acids.

Other suggested functions of autophagy, that have recently gained functional evidence from genetic studies, are: immune response to bacterial infections, promotion of longevity, removal of intracellular protein aggregates, prevention of neurodegeneration and finally the need for autophagy to control cell death and cancer.2,28,29 While it is hard to speculate how the latter roles were relevant to early eukaryotic life it is easier to hypothesise an important role for autophagy in preventing infections. Some intracellular bacterial pathogens, like Shigella and Streptococcus escape the phagosome or endosome and multiply in the cytoplasm. Autophagy can counteract such infections by recapturing and degrading these escaped bacteria.30,31 Thus, autophagy may represent a part of the very first innate immune system.

From a pragmatic point of view, the necessity for autophagy seems to complement the two other intracellular degradation pathways, the proteasomal and endocytic pathway. Proteasomes degrade intracellular short-lived proteins tagged by polyubiquitin. Endocytic and autophagic trafficking shuttles membrane associated molecules or organelles to be degraded in the lysosome. Endocytic trafficking ensures that the cell is able to digest transmembrane proteins within the lysosomal lumen. Some transmembrane cargo, like the Egf receptor, need to be monoubiquitinated to follow the degradation route.32 Finally, autophagy is responsible for degradation of long-lived proteins, turnover of intracellular membrane-bound organelles and protein aggregates. The appearance of endocytosis and autophagy correlate with appearance of the LCEA. Their simultaneous acquisition could have arisen because of the need for degradation pathways to deal with complex intracellular membrane compartments not degraded by the proteasome.

The wide distribution of autophagy within unikonts and bikonts argues for a fundamental need for autophagy in nonparasitic species. It will be interesting to see to what extent the molecular control and use of autophagy has been adapted to meet different needs in different taxa.


The authors would like to thank Per O. Seglen, Harald Stenmark and Karine Lindmo for critically reading the manuscript. T. E. Rusten is supported by the Norwegian Research Council.


Cuervo AM. Autophagy: In sickness and in health. Trends Cell Biol. 2004;14(2):70–77. [PubMed: 15102438]
Levine B, Klionsky DJ. Development by self-digestion: Molecular mechanisms and biological functions of autophagy. Dev Cell. 2004;6(4):463–477. [PubMed: 15068787]
Richards TA, Cavalier-Smith T. Myosin domain evolution and the primary divergence of eukaryotes. Nature. 2005;436(7054):1113–1118. [PubMed: 16121172]
Delsuc F, Brinkmann H, Philippe H. Phylogenomics and the reconstruction of the tree of life. Nat Rev Genet. 2005;6(5):361–375. [PubMed: 15861208]
Pickart CM, Cohen RE. Proteasomes and their kin: Proteases in the machine age. Nat Rev Mol Cell Biol. 2004;5(3):177–187. [PubMed: 14990998]
Ohsumi Y. Molecular dissection of autophagy: Two ubiquitin-like systems. Nat Rev Mol Cell Biol. 2001;2(3):211–216. [PubMed: 11265251]
Klionsky DJ. Autophagy. Georgetown, Texas, USA: Landes Bioscience. 2004
Luiken JJ, van den Berg M, Heikoop JC. et al. Autophagic degradation of peroxisomes in isolated rat hepatocytes. FEBS Lett. 1992;304(1):93–97. [PubMed: 1618306]
Klionsky DJ, Cregg JM, Dunn Jr WA. et al. A unified nomenclature for yeast autophagy-related genes. Dev Cell. 2003;5(4):539–545. [PubMed: 14536056]
Kuma A, Hatano M, Matsui M. et al. The role of autophagy during the early neonatal starvation period. Nature. 2004;432(7020):1032–1036. [PubMed: 15525940]
Noda T, Suzuki K, Ohsumi Y. Yeast autophagosomes: De novo formation of a membrane structure. Trends Cell Biol. 2002;12(5):231–235. [PubMed: 12062171]
Klionsky DJ, Emr SD. Autophagy as a regulated pathway of cellular degradation. Science. 2000;290(5497):1717–1721. [PMC free article: PMC2732363] [PubMed: 11099404]
Mizushima N, Noda T, Yoshimori T. et al. A protein conjugation system essential for autophagy. Nature. 1998;395(6700):395–398. [PubMed: 9759731]
Mukaiyama H, Baba M, Osumi M. et al. Modification of a ubiquitin-like protein Paz2 conducted micropexophagy through formation of a novel membrane structure. Mol Biol Cell. 2004;15(1):58–70. [PMC free article: PMC307527] [PubMed: 13679515]
Strømhaug PE, Bevan A, Dunn Jr WA. GSA11 encodes a unique 208-kDa protein required for pexophagy and autophagy in Pichia pastoris. J Biol Chem. 2001;276(45):42422–42435. [PubMed: 11533052]
Scott RC, Schuldiner O, Neufeld TP. Starvation-induced autophagy is suppressed by TOR signaling independently of S6K and ameliorates the effects of loss of TOR function. Dev Cell. 2004;7(2):167–178. [PubMed: 15296714]
Rusten TE, Lindmo K, Juhâasz G. et al. Programmed autophagy in the Drosophila fat body is induced by ecdysone through regulation of the PI3K pathway. Dev Cell. 2004;7(2):179–192. [PubMed: 15296715]
Bera A, Singh S, Nagaraj R. et al. Induction of autophagic cell death in Leishmania donovani by antimicrobial peptides. Mol Biochem Parasitol. 2003;127(1):23–35. [PubMed: 12615333]
Christensen ST, Chemnitz J, Straarup EM. et al. Staurosporine-induced cell death in Tetrahymena thermophila has mixed characteristics of both apoptotic and autophagic degeneration. Cell Biol Int. 1998;22(7-8):591–598. [PubMed: 10452827]
Bernal A, Ear U, Kyrpides N. Genomes OnLine Database (GOLD): A monitor of genome projects world-wide. Nucleic Acids Res. 2001;29(1):126–127. [PMC free article: PMC29859] [PubMed: 11125068]
El-Sayed NM, Myler PJ, Bartholomeu DC. et al. The genome sequence of Trypanosoma cruzi, etiologic agent of Chagas disease. Science. 2005;309(5733):409–415. [PubMed: 16020725]
Gardner MJ, Hall N, Fung E. et al. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002;419(6906):498–511. [PMC free article: PMC3836256] [PubMed: 12368864]
Lanfredi-Rangel A, Attias M, de Carvalho TM. et al. The peripheral vesicles of trophozoites of the primitive protozoan Giardia lamblia may correspond to early and late endosomes and to lysosomes. J Struct Biol. 1998;123(3):225–235. [PubMed: 9878577]
Katinka MD, Duprat S, Cornillot E. et al. Genome sequence and gene compaction of the eukaryote parasite Encephalitozoon cuniculi. Nature. 2001;414(6862):450–453. [PubMed: 11719806]
Jekely G. Small GTPases and the evolution of the eukaryotic cell. Bioessays. 2003;25(11):1129–1138. [PubMed: 14579253]
Gordon PB, Seglen PO. Prelysosomal convergence of autophagic and endocytic pathways. Biochem Biophys Res Commun. 1988;151(1):40–47. [PubMed: 3126737]
Pulipparacharuvil S, Akbar MA, Ray S. et al. Drosophila Vps16A is required for trafficking to lysosomes and biogenesis of pigment granules. J Cell Sci. 2005;118(Pt):3663–3673. [PubMed: 16046475]
Pattingre S, Tassa A, Qu X. et al. Bcl-2 antiapoptotic proteins inhibit Beclin 1-dependent autophagy. Cell. 2005;122(6):927–939. [PubMed: 16179260]
Ravikumar B, Vacher C, Berger Z. et al. Inhibition of mTOR induces autophagy and reduces toxicity of polyglutamine expansions in fly and mouse models of Huntington disease. Nat Genet. 2004;36(6):585–595. [PubMed: 15146184]
Nakagawa I, Amano A, Mizushima N. et al. Autophagy defends cells against invading group A Streptococcus. Science. 2004;306(5698):1037–1040. [PubMed: 15528445]
Ogawa M, Yoshimori T, Suzuki T. et al. Escape of intracellular Shigella from autophagy. Science. 2005;307(5710):727–731. [PubMed: 15576571]
Raiborg C, Rusten TE, Stenmark H. Protein sorting into multivesicular endosomes. Curr Opin Cell Biol. 2003;15(4):446–455. [PubMed: 12892785]
Hochstrasser M. Evolution and function of ubiquitin-like protein-conjugation systems. Nat Cell Biol. 2000;2(8):E153–157. [PubMed: 10934491]
Hertz-Fowler C, Peacock CS, Wood V. et al. GeneDB: A resource for prokaryotic and eukaryotic organisms Nucleic Acids Res 200432D339–343. (Database issue) [PMC free article: PMC308742] [PubMed: 14681429]
Altschul SF, Madden TL, Schè;affer AA. et al. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 1997;25(17):3389–3402. [PMC free article: PMC146917] [PubMed: 9254694]
Ivens AC, Peacock CS, Worthey EA. et al. The genome of the kinetoplastid parasite, Leishmania major. Science. 2005;309(5733):436–442. [PMC free article: PMC1470643] [PubMed: 16020728]
Abrahamsen MS, Templeton TJ, Enomoto S. et al. Complete genome sequence of the apicomplexan, Cryptosporidium parvum. Science. 2004;304(5669):441–445. [PubMed: 15044751]
Armbrust EV, Berges JA, Bowler C. et al. The genome of the diatom Thalassiosira pseudonana: Ecology, evolution, and metabolism. Science. 2004;306(5693):79–86. [PubMed: 15459382]
Copyright © 2000-2013, Landes Bioscience.
Bookshelf ID: NBK6274


  • PubReader
  • Print View
  • Cite this Page

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...