2. General Aspects of Autophagy and Molecular Mechanisms
The term autophagy (Greek, “self-eating”) refers to a tightly evolutionary self-degradation system that is essential for the maintenance of the physiological homeostasis that occurs in response to several cellular and environmental stresses
[2]. Over the past ten years, substantial research has been carried out in regard to understanding the regulation and molecular mechanisms of autophagy, and considerable interest has emerged in this area of research, enough to confer the Nobel Prize in Medicine or Physiology to Yoshinori Oshumi for his groundbreaking work on the autophagic mechanism
[3].
In mammalian cells, at least three distinct forms of autophagy—macroautophagy, microautophagy, and chaperone-mediated autophagy (CMA)—can be distinguished based on the mode of cargo delivery to lysosomes. During microautophagy, the lysosomal membrane can directly invaginate to trap cytosolic cargo for degradation; this process can include intact organelles and can be defined in different terms, such as micromitophagy (for mitochondria), microlipophagy (for lipid droplets), and micropexophagy (for peroxisomes)
[4]. CMA uses chaperones to identify cargo proteins containing a specific pentapeptide-targeting motif (KFERQ), which is recognized by heat shock cognate 71 kDa protein (HSC70) in the cytosol, that are targeted and bound to the lysosomal membrane protein LAMP2A
[5]. In contrast, macroautophagy (hereafter referred to as autophagy) involves the sequestration of cytoplasmic cargo by de novo double-membrane vesicles (named autophagosomes) that ultimately fuse to lysosomes for degradation
[6].
Autophagy is a highly catabolic process, conserved from yeast to humans, used by eukaryotic cells to maintain cellular homeostasis and cellular and organellar quality control in response to multiple forms of stress, including energy or nutrients deprivation, hypoxia, oxidative stress, endoplasmic reticulum (ER) stress, and infection
[7][8][9].
From a molecular point of view (
Figure 1), autophagy can be divided into four critical steps: initiation, nucleation, maturation, and degradation, strictly orchestrated by at least 37-autophagy-related proteins (ATGs), that deliver cytoplasmic cargo to the lysosome for degradation. Two nutrient-responsive kinases, 5′ AMP-activated protein kinase (AMPK) and mechanistic target of rapamycin (MTOR), are the main regulators (positive and negative, respectively) of autophagy and quickly respond to external stimuli, nutrient fluctuations, and phosphorylation/dephosphorylation events. AMPK is a heterotrimeric protein that acts as a monitor of intracellular energy levels by controlling the AMP/ATP ratio. AMPK also controls the cellular processes that increase the ATP amounts (such as glycolysis and β-oxidation), controls the fatty acids and cholesterol synthesis, and regulates the ATP-consuming mechanisms
[10][11]. MTOR is a serine/threonine protein kinase belonging to the phosphatidylinositol kinase-related kinase (PIKK) family. Inside the cell, MTOR is found in two distinct complexes, MTORC1 and MTORC2, but only MTORC1 regulates the cellular metabolism and responds to external and internal signals, such as amino acids, glucose, and growth factors
[12]. Apart from MTOR itself, MTORC1 is composed of various regulatory subunits (including Raptor, MLST8, PRAS40, and DEPTOR)
[13], and, under physiological conditions, MTORC1 results in activity and represses the autophagy machinery by phosphorylating the unc-51-like kinase 1 (ULK1) serine threonine kinase complex (consisting of ULK1, ATG13, RIB-inducible coiled-coil protein 1 (FIP200), and ATG101) at Ser-757, Ser-638, and Ser-758 residues
[14][15]. The inactivation of MTORC1 causes the dephosphorylation of the target ribosomal protein S6 kinase (p70S6K) and the translation initiation factor 4E binding protein-1 (4E-BP1), provoking a downregulation of the cellular protein synthesis translation and the activation of the autophagy-dependent self-digestive process. In this latter event, the AMPK inhibits MTORC1 directly by phosphorylating Raptor
[16] and indirectly via the activation of the tuberous sclerosis 2 (TSC2) complex
[17] to determine the dephosphorylation of ULK1 at Ser-757, Ser-638, and Ser-758 residues. Then, the AMPK phosphorylates ULK1 and ATG13 in specific amino acid residues (Ser-555, Ser-777, Ser-317, Ser-467) to activate the ULK1 complex
[14].
Figure 1. Molecular aspects of autophagy. Autophagy is a catabolic process used by eukaryotic cells to maintain cellular homeostasis and cellular and organellar quality control in response to multiple forms of stress. Different factors are involved during the main stages of autophagy. During the initiation step, the Unc-51-like kinase 1 (ULK1) serine threonine kinase complex receives input from the mechanistic target of rapamycin (MTOR) and from the kinase 5′ AMP-activated protein kinase (AMPK). Meanwhile, MTOR is the main autophagic repressor; AMPK is an autophagy activator. The ULK1 complex then activates nucleation of the phagophore by activating members of class III phosphatidylinositol 3 kinase (PI3KC3) complexes, such as AMBRA1, VSP34, and BECN1. BCL2 regulates the activity of BECN1. Ubiquitin-like conjugation systems ATG12-ATG5-ATG16 and LC3-ATG4 are necessary for the maturation of the vesicle. Other proteins, such as p62, complete the autophagosome formation. The last phase is characterized by fusion of the autophagosome with lysosome and the degradation of the autophagic cargo by acidic hydrolases.
The ULK1 complex activation represents the initiation step of autophagy
[18] and triggers the nucleation of the phagophore by phosphorylating and activating class III phosphatidylinositol 3 kinase (PI3KC3) complexes. The core of the PI3KC3 complex consists of class III PI3K, BECN1
[19], vacuolar protein sorting 34 (VSP34)
[20], and general vesicular transport factor (p115)
[21], then it can be associated with both ATG14
[22] and the activating molecule in BECN1-regulated autophagy protein 1 (AMBRA1) (PI3KC3-C1)
[23] or UV radiation resistance-associated gene protein (UVRAG) (PI3KC3-C2)
[19], which both, in turn, generate local phosphatidylinositol-3-phosphate (PIP3) production at a characteristic ER structure called the omegasomes, the membranous regions that elongate to generate the phagophores
[24]. Both the initiation and nucleation steps promote the formation of the autophagic vesicle membrane that can also include membranes from ER-mitochondria and ER-plasma membrane contact sites, mitochondria, recycling endosomes, and Golgi complex
[25][26], as well as ATG9-containing vesicles. PIP3 then recruits, through the PIP3-binding domain, WD repeat domain phosphoinositide-interacting proteins (WIPI) and zinc-finger FYVE domain-containing protein 1 (DFCP1) to promote membrane elongation
[27]. This process represents the transition from omegasomes to the phagophores.
Two ubiquitin (Ub)-like conjugation systems are necessary for autophagosome formation (maturation step). One system involves the covalent conjugation of the Ub-like mammalian ATG12 to ATG5, which, together with RAB37
[28], further establishes a complex with ATG16, which, in turn, links to WIPI2
[27] and associates with the phagophore membrane. Although ATG7 and ATG5 represent fundamental proteins for autophagy execution, recent investigations have unveiled that autophagy may also occur in
Atg5 or
Atg7-deficient cells. In this case, autophagy is termed “alternative autophagy” and the elongation process depends on the activity of RAB9 GTPase
[29]. The second pathway, the ATG8 family proteins, includes microtubule-associated protein 1A/B light-chain 3A (LC3) and the GABARAP (gamma-aminobutyric acid receptor-associated protein) subfamily to membrane-resident phosphatidylethanolamine (PE) thanks to the sequential action of the proteases ATG4 and ATG3
[30]. These complexes, both regulated by the E1-like enzyme ATG7, bring the conversion of the soluble form of LC3 (LC3-I) to the lipidated autophagic vesicles-associated form, known as LC3-II. LC3-II is one of the most commonly used markers of autophagy as it migrates faster than LC3-I by electrophoresis detection; additionally, green fluorescent protein-LC3 fusion protein can be used to detect autophagosomes shifting from diffuse to punctuate staining by microscopy
[2]. The lipidation process of LC3 is critical for the expansion and closure of the phagophore. However, during this process, it has been unveiled that the association of WIPI4 with ATG2, which together recruit ATG9, is essential. Indeed, the WIPI4–ATG2 complex moves to the ER and mediates the tethering between the ER and phagophore to regulate the lipid transfers necessary for the expansion of the phagophore
[31]. Furthermore, ATG9, whose recruitment to ER is determined by the WIPI4–ATG2 complex, is equally important for the lipid and membrane supply to the phagophore
[32][33].
The closure of the autophagic vesicle engages members of the endosomal sorting complex required for transport (ESCRT), in particular the charged multivesicular body protein 2A (CHMP2A) and the vacuolar protein sorting-associated-4 (VPS4)
[34]. Both proteins move on the outer leaf of the autophagic vesicle, but, meanwhile, CHMP2A is responsible for the closure of the autophagosome, and VSP4 determines the disassembly of ESCRT molecules. Consistent with this, the genetic inhibition of CHMP2A and VPS4 alters the closure of the vesicle and delays fusion with the lysosome
[34].
The newly formed double-membrane vesicle, just termed autophagosome, traps the engulfed cytosolic material, recruited by cargo receptors, such as SQSTM1/P62 and NBR1, as autophagic cargo destinated for degradation
[35]. Then, the autophagosome is transported along microtubules to the perinuclear region to fuse with lysosomal membranes and form an autophagolysosome. This event requires the changing in lysosomal pH, tethering factors, such as the multiprotein homotypic fusion and vacuole protein sorting (HOPS) complex, RAB7, and small GTPase, along with SNAREs proteins (syntaxin 17 (STX17) and synaptosomal-associated protein 29 (SNAP29) in the autophagosome and vesicle associated membrane protein 8 on the lysosome (VAMP8)
[36][37]. Emerging evidence suggests that other proteins, such as the adaptor protein EPG5 and members of the LC3/GABARAP family, are involved in the autophagosome–lysosome fusion event
[38][39].
The last phase encompasses the degradation of the autophagic cargo by acidic hydrolases in the lysosome, and the resulting breakdown products are released back to the cytoplasm to provide new sources of energy in response to the nutritional needs of the cell (Figure 1).
Autophagy also exists in multiple variant forms that are characterized to sequester and degrade specific intracellular elements, such as proteins (proteinphagy), lipid droplet (lipophagy), xenobiotics (xenophagy), and organelles. Probably, the most studied selective form of autophagy toward an intracellular organelle is mitophagy, the autophagic removal of autophagy, which is an essential mitochondrial quality control mechanism
[40]. Mitophagy may occur under physiological conditions (development, aging, and differentiation), it may be induced by diverse harmful conditions, such as loss of mitochondrial functioning, excessive reactive oxygen species (ROS) production, hypoxia, and accumulation of unfolded proteins, and it is involved during the pathogenesis of a number of human diseases, including neurodegeneration
[41], cancer
[42], and cardiovascular disease
[43]. The possibility that, inside a cell, mitochondria could be removed was suggested in late 1915
[44] and confirmed in 1962 when, by using electron microscopy, mitochondrial fragments were identified in liver lysosomes
[45]. However, the molecular mechanisms at the basis of mitophagy started to be unveiled in the early 2000s when it was demonstrated that, during the differentiative process, red blood cells
[46][47][48] eliminate their mitochondria by a mechanism orchestrated by the OMM resident NIP3-like protein X (NIX/BNIP3L), which, thanks to a WXXL-like motif facing the cytosol, can bind LC3 and GABARAP proteins to recruit isolation membranes to mitochondria
[49][50]. FUN14 domain-containing protein 1 (FUNDC1) represents another OMM protein that acts as a mitophagy receptor in response to determinate stimuli
[51]. Under normal physiological conditions, SRC kinase phosphorylates FUNDC1. Upon hypoxia, SRC becomes inactivated and FUNDC1 is subsequently dephosphorylated. This results in an increase in association with LC3 and the incorporation of the damaged mitochondria into the autophagic vesicle
[51]. However, probably, the best characterized molecular axis regulating mitophagy in mammals is the PTEN-induced kinase 1 (PINK1) and Parkinson protein 2 (parkin). In functional mitochondria, PINK1 is imported into the mitochondria in the IMM, processed by several protease, and then degraded by the proteasome. Following mitochondria uncoupling, PINK1 accumulates and stabilizes on the OMM where it phosphorylates Parkin to change its conformation, recruits Parkin on the mitochondria, and activates Parkin into an active phospho-Ub-dependent enzyme
[52][53][54][55]. In this state, Parkin determines the ubiquitinization of diverse OMM proteins and the consequent recruitment of the Ub-binding autophagic receptors LC3, NBR1, MDP52, TAX1BP1 (TBK1), and p62/Sequestome
[56][57]. Interestingly, it has been proven that, during Parkin-mediated mitophagy, the OMM results ruptured
[55][58][59], and the IMM is potentially exposed to interaction with the autophagic vesicle. A recent work identifies the IMM protein prohibitin 2 (PHB2) as a crucial mitophagy receptor and demonstrates that PHB2 binds the LC3-interaction region (LIR) domain following the mitochondrial depolarization and proteasome-dependent rupture of the OMM
[60]. Another IMM protein responsible for the regulation of mitophagy is cardiolipin (CL). Following mitophagy stimuli, CL translocates to the OMM and acts as a signal for the identification and elimination of harmed mitochondria
[61]. Furthermore, it has been demonstrated that this process is facilitated by the fact that an LC3 protein possesses a CL-binding site
[61]. Recent investigations also account for mitochondrial matrix proteins having an important role in mitophagy regulation. An example may be found in the nod-like receptor (NLR) family member NLRX1. NLRX1 contains a LIR domain, and, upon infection with the pathogen
Listeria, oligomerizes to induce the binding of the LIR motif to LC3 to activate mitophagy
[62]. Recently, two other mitochondrial matrix resident proteins have been associated with mitophagy regulation and activation in yeast. Consistently, a knockout of both mitochondrial kinases Pkp1 and Pkp2 abolishes the mitophagy program
[63].