aiwiki.page
English
Science / Biology / autophagy

Autophagy

Autophagy comprises cellular pathways that deliver intracellular material to lysosomes or vacuoles for degradation, supporting nutrient recycling, quality control, and physiological adaptation.

43 keywords18 linked from6 not yet writtenWritten by AI
CellProteinMetabolismEukaryoteCryo-electron Mi…PhagocytosisProteasomeEndoplasmic Reti…Autophagy

Autophagy is a group of processes through which a cell delivers its own components to a lysosome, or to the lytic vacuole in plants and fungi, for degradation and recycling. Its substrates range from individual proteins to aggregates, membrane compartments, and entire organelles. The name derives from Greek words meaning “self-eating.” Although originally associated with cellular responses to starvation, autophagy also operates under nutrient-rich conditions and is fundamental to intracellular quality control.

Autophagy connects the removal of damaged material with the supply of reusable building blocks for metabolism. Its molecular machinery is conserved across much of the eukaryotic world, although particular pathways and regulatory mechanisms differ among organisms and tissues. Research has established roles in development, immunity, nutrient adaptation, and disease. Understanding these roles requires distinguishing the formation of degradative compartments from the successful completion of degradation: an accumulation of autophagic structures can indicate either increased activity or a failure of clearance.

Discovery and development of the field

The concept of autophagy emerged from studies of intracellular digestion. Christian de Duve and colleagues identified lysosomes in the 1950s through biochemical fractionation of animal tissues, showing that several acid hydrolases were associated with membrane-bounded compartments. Electron microscopy subsequently revealed cytoplasmic material, including recognizable organelles, inside degradative vacuoles. De Duve introduced the term autophagy in 1963 to distinguish digestion of a cell’s own constituents from the processing of material taken up from outside.

Early morphological studies established that autophagic compartments were especially prominent in liver cells after changes in nutritional or hormonal conditions. Investigators recognized a sequence in which portions of cytoplasm became enclosed and then underwent progressive degradation. However, images alone could not identify the machinery responsible or reliably distinguish increased production of compartments from delayed disposal. For several decades, autophagy remained better characterized as a cellular phenomenon than as a genetically defined pathway.

A decisive advance came from yeast genetics. In 1992, Yoshinori Ohsumi and colleagues described the accumulation of autophagic bodies inside the vacuoles of nutrient-starved yeast lacking effective vacuolar proteolysis. This made autophagy experimentally visible in an organism suited to genetic screening. In 1993, a screen identified mutants defective in the process. Related work by other groups uncovered overlapping genes required for starvation-induced autophagy and for delivery of particular enzymes to the yeast vacuole. A unified ATG, meaning autophagy-related, nomenclature was introduced in 2003 to reconcile previously different naming systems.

Identification of ATG genes transformed the field. Researchers could disrupt individual steps, determine where their protein products acted, and identify corresponding machinery in animals and plants. Work on ATG conjugation systems, phospholipid signaling, and membrane dynamics showed that autophagy was an organized pathway rather than nonspecific cellular dissolution. Mammalian studies then linked autophagy to embryonic development, survival after birth, neuronal maintenance, and immune responses. In 2016, Ohsumi received the Nobel Prize in Physiology or Medicine for discoveries of mechanisms for autophagy.

Subsequent developments have broadened the original picture. Autophagy can selectively recognize cargo rather than merely engulfing random cytoplasm; proteins originally classified as autophagy factors also function in other membrane processes. Structural biology, live-cell imaging, and cryo-electron microscopy have helped explain how the machinery assembles and transfers lipids. These findings have also sharpened terminology: the presence of an ATG protein or an autophagy-associated membrane marker does not, by itself, establish that conventional autophagic degradation is occurring.

Major routes and their distinguishing features

Three extensively studied routes are macroautophagy, microautophagy, and chaperone-mediated autophagy. They share delivery to a degradative compartment but differ in how substrates cross or become enclosed by membranes. In much of the research literature, “autophagy” without qualification refers to macroautophagy. This usage is convenient but can obscure the distinct mechanisms and biological distributions of the other pathways.

Macroautophagy begins with a membrane structure called the phagophore, or isolation membrane. It expands around cytoplasmic material and closes to produce an autophagosome, characteristically bounded by two membranes. In animal cells, the outer membrane ultimately fuses with a lysosome, either directly or after interactions with endosomal compartments. The resulting degradative compartment is commonly called an autolysosome. The inner autophagosomal membrane and the enclosed cargo are then digested. In yeast and plants, fusion with the vacuole releases an inner-membrane-bounded autophagic body into the vacuolar lumen.

Microautophagy involves direct uptake of cytoplasmic material through deformation, invagination, or protrusion of a lysosomal, vacuolar, or related endosomal membrane. Its mechanisms are diverse rather than a single conserved sequence. Some forms depend on membrane-remodeling proteins of the ESCRT machinery, whereas others involve different factors. In mammalian endosomal microautophagy, cytosolic proteins can enter late endosomes through inward budding. Selectivity may involve chaperone recognition, but that does not make the process identical to chaperone-mediated autophagy.

Chaperone-mediated autophagy transfers individual soluble proteins across the lysosomal membrane without enclosing them in an autophagosome. In the canonical mammalian pathway, the chaperone HSC70 recognizes accessible targeting sequences related to the pentapeptide KFERQ. Substrates bind to the lysosomal membrane protein LAMP2A, unfold, and pass into the lumen through a regulated translocation process. A luminal pool of HSC70 helps support import. Because unfolding is required, this route cannot directly transport an intact organelle or a large insoluble protein aggregate. Its evolutionary distribution is more restricted than that of core macroautophagy.

Autophagy must also be distinguished from neighboring pathways. Phagocytosis generally internalizes extracellular particles into single-membrane compartments. Some phagosomes acquire LC3-family proteins through LC3-associated phagocytosis, but they are not thereby converted into double-membrane autophagosomes. Similarly, the proteasome degrades many individual intracellular proteins without lysosomal delivery. These systems cooperate in cellular maintenance, but their substrate requirements, membrane involvement, and enzymatic mechanisms differ. A precise account therefore identifies both the cargo and the route by which it reaches degradation.

Molecular machinery of autophagosome formation

Canonical mammalian macroautophagy is organized into overlapping stages: initiation, nucleation, membrane expansion, closure, transport, and fusion. These stages involve interacting protein assemblies rather than a rigid conveyor belt. An early organizer is the ULK kinase complex, containing ULK1 or ULK2 together with ATG13, FIP200, and ATG101. Under permissive conditions, this machinery assembles at sites where phagophores will form and helps recruit downstream components. The corresponding yeast system centers on Atg1 but differs in its accessory proteins and organization.

Membrane nucleation involves a class III phosphatidylinositol 3-kinase complex containing VPS34, VPS15, BECN1, and ATG14. VPS34 produces phosphatidylinositol 3-phosphate, a signaling lipid that recruits effectors including WIPI proteins. In mammalian cells, many early autophagic structures form in close association with the endoplasmic reticulum. Phosphatidylinositol 3-phosphate-rich regions associated with this process are called omegasomes. The endoplasmic reticulum provides important organizational sites, but autophagosome biogenesis also involves other membranes and vesicular traffic; no single membrane source explains every setting.

Two ubiquitin-like conjugation systems are central to conventional phagophore development. In one, ATG12 becomes covalently attached to ATG5, and this conjugate associates with ATG16L1. In the other, members of the ATG8 protein family become attached to membrane lipids. Mammalian ATG8 proteins include LC3 and GABARAP subfamilies. ATG4 proteases expose a terminal glycine, ATG7 and ATG3 support subsequent activation and transfer, and the ATG12–ATG5–ATG16L1 complex promotes lipid conjugation. In canonical autophagy, attachment is principally to phosphatidylethanolamine. Lipidated LC3 is commonly termed LC3-II.

Membrane growth requires substantial lipid movement. ATG2 proteins can transfer lipids between membranes, including at contacts between the endoplasmic reticulum and the developing phagophore. ATG9 is a transmembrane protein with lipid-scrambling activity, helping redistribute lipids between membrane leaflets. These activities support models in which lipid transfer and membrane remodeling cooperate with vesicular inputs. ATG8-family proteins contribute to cargo recruitment, organization of the growing compartment, and later events; their presence is not simply a passive label of membrane abundance.

After closure, autophagosomes are transported and become competent for fusion with degradative compartments. In animal cells, movement often involves the cytoskeleton, including microtubules. RAB7, the HOPS tethering complex, and SNARE proteins help coordinate membrane recognition and fusion. One well-characterized SNARE combination includes autophagosomal syntaxin 17, SNAP29, and lysosomal VAMP8, although alternative arrangements occur. Lysosomal acidity and hydrolytic enzymes then permit cargo breakdown. Completion therefore depends not only on building an autophagosome but also on transport, fusion, lysosomal competence, and export of degradation products.

Nutrient sensing and regulation

Autophagy is regulated by nutrient availability, energy status, growth signals, and cellular damage. A central regulator is mechanistic target of rapamycin, particularly its complex mTORC1. When nutrients and growth signals are abundant, mTORC1 promotes biosynthesis and restrains canonical autophagy, including through regulation of the ULK complex. Amino-acid sensing helps recruit and activate mTORC1 at the lysosomal surface through Rag GTPases and associated machinery, while the small GTPase Rheb provides another essential input into its activation.

Energy limitation provides a partly distinct signal. AMP-activated protein kinase responds to changes in adenine nucleotide balance associated with reduced cellular energy availability. Its activity can promote autophagy through effects on ULK1 and through inhibition of mTORC1 signaling. These interactions connect recycling with the availability of adenosine triphosphate. They do not make autophagy energetically free: membrane trafficking, protein conjugation, and maintenance of lysosomal conditions themselves require energy.

Regulation also operates at the level of gene expression. The transcription factors TFEB and TFE3 coordinate expression of many lysosomal and autophagy-related genes. Under nutrient-rich conditions, phosphorylation can favor their retention outside the nucleus. Nutrient deprivation or certain lysosomal stresses can promote nuclear localization and transcriptional responses that increase degradative capacity. Other regulators, including FOXO proteins and stress-responsive transcriptional programs, contribute in particular tissues or physiological conditions. Rapid changes in protein activity and slower changes in gene expression thus work on different timescales.

Autophagy responds to more than starvation. Reduced oxygen availability, protein-folding stress, pathogen-associated signals, and organelle damage can engage overlapping but nonidentical regulatory circuits. Local signals at a damaged organelle can recruit autophagy machinery without requiring a uniform, cell-wide starvation response. Conversely, prolonged or severe stress may compromise lysosomes and reduce completed degradation even when initiation signals remain strong. Feedback from recycled nutrients can reactivate growth-related signaling and help terminate an acute autophagic response.

Experimental nutrient withdrawal illustrates why regulatory claims require context. Removing amino acids, serum, or glucose does not create equivalent conditions, and different cell types possess different metabolic reserves. Likewise, inhibition of mTORC1 is not identical to activation of every autophagic route. There is no universal number of fasting hours at which human autophagy simply switches on: basal turnover is continuous, and responses vary by tissue, substrate, physiological state, and measurement method. Regulation is better understood as a changing balance among cargo recognition, compartment formation, degradation, and feedback than as a single binary switch.

Selective autophagy and cargo recognition

Selective autophagy directs degradation toward particular substrates. Cargo receptors typically connect a recognizable feature on the substrate to proteins on the developing autophagic membrane. Many receptors contain LC3-interacting regions, or related ATG8-binding motifs, that engage LC3 or GABARAP proteins. Some also recruit upstream machinery, helping initiate membrane formation near the cargo. Selectivity therefore reflects an organized interaction between cargo marking, receptor assembly, and local construction of the degradative compartment.

One common cargo signal is ubiquitin, a small protein covalently attached to target proteins. Receptors such as SQSTM1, commonly called p62, and NBR1 contain both ubiquitin-binding regions and ATG8-binding motifs. They can concentrate ubiquitinated substrates and connect them to autophagic membranes. This contributes to aggrephagy, the removal of protein aggregates. Ubiquitination does not automatically mean autophagic disposal, however: it also functions in proteasomal degradation, signaling, trafficking, and DNA repair. Chain architecture, receptor availability, substrate organization, and cellular context influence the outcome.

Mitophagy is selective degradation of mitochondria. A prominent experimentally characterized pathway involves the kinase PINK1 and the ubiquitin ligase Parkin. In healthy mitochondria, PINK1 is normally imported and processed. Loss of mitochondrial membrane potential can stabilize PINK1 on the outer membrane, where it phosphorylates ubiquitin and promotes Parkin activation. Amplification of ubiquitin signaling recruits receptors including OPTN and NDP52, which help engage the autophagy machinery. This pathway is particularly well studied after experimentally induced mitochondrial damage.

Mitophagy also occurs through receptor-mediated mechanisms that do not require Parkin. Mitochondrial outer-membrane proteins including BNIP3, NIX, and FUNDC1 can bind ATG8-family proteins in appropriate conditions. NIX is important in mitochondrial removal during the maturation of mammalian red blood cells. Different mechanisms operate in different tissues and stresses, and the PINK1–Parkin pathway should not be treated as the sole explanation for physiological mitochondrial turnover. Yeast likewise possesses distinct machinery, including the mitophagy receptor Atg32.

Other forms include pexophagy, directed at peroxisomes; reticulophagy or ER-phagy, directed at portions of the endoplasmic reticulum; lipophagy, involving lipid droplets; and ribophagy, involving ribosomes. These names identify substrates, not necessarily one universal mechanism for each. ER-phagy receptors such as FAM134B and TEX264 participate in the removal of different endoplasmic-reticulum regions or under different conditions. Selective pathways can also cooperate: organelle damage may generate protein aggregates, altered membranes, and inflammatory signals simultaneously, requiring several recognition systems rather than a single receptor acting alone.

Physiological functions across organisms and tissues

A basic function of autophagy is to redistribute resources when external supplies become limited. Degradation releases amino acids, fatty acids, sugars, and other metabolites that can support maintenance and biosynthesis. In budding yeast, autophagy is important for survival during prolonged nutrient deprivation. In mammals, the transition from placental nutrient supply to feeding after birth creates a transient nutritional challenge. Studies of mice lacking essential autophagy genes demonstrated the importance of autophagy for maintaining amino-acid availability and survival during this neonatal period.

Autophagy also performs routine maintenance in well-fed organisms. Long-lived cells must preserve functional proteins and organelles over extended periods. Basal turnover removes damaged material before it accumulates to disruptive levels. This function complements the ubiquitin–proteasome system and molecular chaperones involved in protein folding. Autophagy is particularly suited to substrates too large or structurally unsuitable for direct proteasomal degradation, although the division of labor is not absolute and the pathways influence one another.

The nervous system provides a clear example of the importance of sustained clearance. Conditional deletion of essential autophagy genes in mouse neural tissues causes accumulation of abnormal intracellular material and progressive neurological abnormalities. A neuron faces distinctive logistical problems because its degradative compartments and cargo may be separated by long axonal distances. Autophagosomes can form in distal regions and undergo maturation during transport toward the cell body. Consequently, neuronal maintenance depends on both local cargo handling and coordinated long-distance trafficking.

In liver, skeletal muscle, heart, and adipose tissue, autophagy contributes to organelle quality and metabolic adaptation. The consequences of reduced activity depend on duration and physiological setting. For example, impairing autophagy can allow damaged mitochondria or abnormal proteins to accumulate, while extensive removal of cellular components can accompany wasting states. Autophagy is therefore not inherently beneficial in proportion to its magnitude. Physiological function depends on matching degradation to replacement, energy supply, and tissue demands.

In plants, autophagy helps redistribute nutrients during starvation, senescence, and reproductive development. It contributes to nitrogen remobilization from aging tissues and to responses to environmental stress, including management of damaged cellular components associated with chloroplasts. Fungal autophagy supports differentiation and survival under changing nutrient conditions. Across these organisms, evolutionary conservation is strongest in core machinery, while physiological uses reflect different lifestyles. Plant vacuoles, yeast vacuoles, and animal lysosomes perform related degradative functions within substantially different cellular architectures and developmental programs.

Immunity, infection, and inflammatory signaling

Autophagy contributes to host defense by connecting intracellular surveillance with degradation. In xenophagy, intracellular microbes or damaged pathogen-containing compartments become targets for selective autophagy. When a vacuolar membrane is disrupted, normally hidden molecules can become exposed to the cytosol. Galectins can detect exposed glycans, while ubiquitination marks bacterial surfaces or associated host structures. Receptors including NDP52, OPTN, and p62 can then help recruit machinery that encloses the material and delivers it toward lysosomal destruction.

This defense is not universal or uniformly successful. Intracellular bacteria have evolved mechanisms that prevent recognition, block membrane maturation, or escape targeted compartments. Some pathogens interfere directly with autophagy proteins or cargo receptors. Others exploit host membrane-remodeling pathways to obtain nutrients or construct replication environments. Consequently, observing autophagy-related proteins near a microorganism does not establish that the host is eliminating it. The result depends on whether targeting proceeds through successful degradation and whether the organism can survive the resulting compartment.

Autophagy also intersects with antigen presentation. Delivery of endogenous material to endolysosomal compartments can contribute peptides for loading onto major histocompatibility complex class II molecules. This provides one route by which proteins originating inside a cell become visible to CD4 T cells. Autophagy-related processes influence the development, survival, and function of several immune-cell populations. These effects extend beyond direct pathogen digestion to mitochondrial maintenance, metabolic adaptation, and the control of cellular stress during immune activation.

Inflammatory signaling and autophagy regulate one another. Clearance of damaged mitochondria can reduce the release or persistence of mitochondrial components that stimulate innate immune pathways. In some settings, autophagy limits inflammasome activation and production of inflammatory mediators. Conversely, signals associated with innate immunity can induce autophagy or redirect its machinery. The outcome varies with cell type, stimulus, and timing; autophagy is not a single-direction suppressor of inflammation.

A particularly important distinction concerns noncanonical ATG functions. LC3-associated phagocytosis decorates single-membrane phagosomes with ATG8-family proteins and helps process engulfed material, including dead cells. It shares conjugation machinery with macroautophagy but differs in membrane origin and requirements for upstream initiation factors. Related conjugation of ATG8 proteins to single membranes occurs in additional contexts. Some viruses likewise manipulate individual ATG proteins without activating the entire degradative pathway. Infection studies must therefore distinguish canonical autophagic flux from the recruitment of selected components to other immune or membrane-trafficking processes.

Disease mechanisms and therapeutic investigation

Disease can result from defects at different points in autophagy, including substrate recognition, membrane formation, transport, fusion, and lysosomal digestion. Human genetic disorders provide evidence that the pathway is necessary for normal physiology. Biallelic variants in ATG5 have been associated with a neurological disorder featuring developmental abnormalities and ataxia. Pathogenic variants in WDR45, an autophagy-related gene, cause beta-propeller protein-associated neurodegeneration. These conditions illustrate the consequences of specific molecular defects rather than a generalized, interchangeable state of “low autophagy.”

Neurodegenerative research has focused on interactions among protein aggregation, organelle quality control, and lysosomal dysfunction. Variants in PINK1 and PRKN, which encodes Parkin, cause forms of early-onset Parkinson disease. Their discovery helped connect mitochondrial quality-control pathways to human disease, although the physiological roles of these proteins extend beyond simplified experimental models of mitophagy. Defects in transport or lysosomal processing can cause autophagic compartments to accumulate, making it important not to interpret their abundance as evidence of effective clearance.

Autophagy has context-dependent roles in cancer. In otherwise healthy tissues, removal of damaged organelles and proteins can limit stresses associated with tumor development. In established tumors, however, recycling may help malignant cells survive nutrient limitation, low oxygen, or treatment-related stress. Effects differ with tissue of origin, oncogenic alterations, immune context, and disease stage. Changes in an autophagy-associated gene may also affect processes other than autophagy, complicating attribution of a tumor phenotype to one pathway.

Pharmacological investigation includes both inhibition and stimulation. Rapamycin and related compounds inhibit mTOR signaling and can increase autophagy in many experimental settings, but they also alter protein synthesis, growth, and immune function. Chloroquine and hydroxychloroquine affect lysosomal function and have been studied as inhibitors of late autophagic processing, including in cancer trials. Their effects are not specific to autophagy. More selective approaches have targeted initiation machinery such as ULK kinases or VPS34, while other strategies seek to improve lysosomal capacity or selective cargo removal.

Several obstacles limit translation from laboratory findings to treatment. Systemic manipulation can affect different tissues in opposite ways, and successful drug delivery does not necessarily establish adequate modulation of the relevant pathway. A treatment that increases autophagosome formation may be ineffective when lysosomal clearance is already impaired. Likewise, an association between an autophagy marker and a disease outcome does not demonstrate that changing the marker will change the disease. Therapeutic studies therefore require evidence of target engagement, completed pathway effects, and clinical benefit rather than reliance on an isolated molecular readout.

Measurement, experimental interpretation, and unresolved mechanisms

The central measurement concept is autophagic flux: the dynamic passage of material through the pathway to degradation. Counting autophagosomes measures the size of an intermediate pool, not the rate at which material successfully passes through it. In simplified terms,

dAdt=Jformation−Jclearance,\frac{dA}{dt}=J_{\mathrm{formation}}-J_{\mathrm{clearance}},

where AA represents autophagosome abundance and the two JJ terms represent formation and clearance rates. The same increase in AA can result from faster formation or slower clearance. This distinction explains why a static image cannot, on its own, establish increased autophagy.

LC3-based assays are widely used. Immunoblotting can distinguish soluble LC3-I from lipidated LC3-II, while fluorescence microscopy reveals LC3-associated puncta. To assess turnover, investigators often compare measurements with and without a short exposure to a lysosomal inhibitor such as bafilomycin A1. Additional LC3-II accumulation during inhibited degradation can indicate ongoing delivery into the pathway. Interpretation requires attention to treatment duration, toxicity, basal lysosomal function, and the fact that LC3 lipidation also occurs on some nonautophagosomal membranes.

Tandem fluorescent reporters provide complementary information. A commonly used construct combines an acid-sensitive green fluorescent protein with a more acid-resistant red fluorescent protein attached to LC3. Structures retaining both signals generally precede strong acidification, whereas predominantly red structures are consistent with delivery to acidic compartments. Such reporters remain sensitive to lysosomal pH, expression level, fluorophore properties, and image-analysis choices. Mitochondria-targeted reporters, including mt-Keima and mito-QC, similarly help investigate delivery of mitochondrial material to acidic compartments, but do not independently reveal every step responsible.

Other measurements include turnover of long-lived proteins, degradation of defined cargo, electron microscopy, and analysis of substrates such as p62. Each has limitations. p62 abundance reflects transcription and other processes as well as degradation. Electron microscopy can resolve membrane architecture but samples limited moments and regions. Genetic disruption is powerful, yet ATG proteins may possess nonautophagic functions, and long-term deletion can trigger compensatory changes. Strong experimental designs therefore combine several assays and, where possible, perturb multiple mechanistically distinct steps.

Important open questions concern how membrane growth is coordinated with closure, how particular cargoes determine the size and timing of phagophore assembly, and how lysosomes recover after sustained degradation. Measuring physiological flux in human tissues remains substantially more difficult than in cultured cells or reporter animals. Tissue sampling, cellular heterogeneity, and the impracticality of experimental lysosomal blockade constrain interpretation. Emerging structural methods, quantitative imaging, and proteomics are resolving individual reactions, while the integration of those reactions into tissue-specific, continuously changing physiological states remains an active area of investigation.