(a) Autophagosome formation. Schematic of autophagosome formation and degradation: Within the cytoplasm, double-membraned, sac-like structures called phagophores are the first morphologically recognizable autophagic precursors and can be distinguished within cells by the proteins that associate with their membranes. A complex comprising ATG12–ATG5–ATG16L1 proteins enables the conjugation of LC3-II to the membranes. The edges of the phagophore elongate and eventually fuse while engulfing a portion of the cytoplasm. As the phagophore enlarges and approaches closure, the ATG5–ATG12–ATG16L1 complex dissociates from the outer membrane, whereas LC3-II remains associated. The resulting structure is a spherical double-membrane organelle, called the autophagosome. Following closure, autophagosomes are trafficked by dynein motors along microtubules to the perinuclear region where they fuse with the lysosomes and their contents are degraded. (b) Lipidation of LC3-II. During autophagosome formation, LC3 (and other ATG8 ubiquitin-like family proteins) are conjugated to the lipid PE in autophagosome membranes. This lipidation requires a protease and two ubiquitin-like conjugation systems (explained in detail in [1,2]). ATG4 is a cysteine protease which cleaves the C-terminus of LC3 exposing a glycine residue. This first cleaved form of LC3 is called LC3-I. A further reaction then occurs involving a complex of ATG12–5 and ATG16L1, which together act as an E3-like ligase. This determines the site of LC3 lipidation and assists the transfer of LC3-I to PE in membranes to form LC3-II. ATG8/LC3 proteins may assist in the expansion and closure of autophagosomal membranes, in autophagosome-lysosome fusion and inner autophagosomal membrane degradation.

Schematic diagram of conventional methods to measure rates of autophagy. (a,b) Western blots for LC3-II: Measuring LC3 lipidation by western blotting is one of the best-established methods for measuring autophagic flux. However, care must be taken in interpreting increases in LC3 levels as this may occur as a result of an increase in autophagosome formation (upregulation) or a blockage in clearance. To discriminate between these two scenarios, assays should be performed in basal conditions and in the presence of an agent that prevents lysosomal degradation such as bafilomycin A1 (Baf) or ammonium chloride (NH₄Cl). (a) When autophagy is induced, LC3-II levels increase as more autophagosomes are formed. In the presence of a lysosomal blocker, LC3-II levels increase further because increased autophagosome formation still occurs, but autophagosomes cannot be cleared and therefore build up within the cell. (b) In some conditions when autophagy is blocked (for example, if fusion with the lysosome is prevented), LC3-II levels can also increase because autophagosomes may form but are not degraded. In this scenario, when LC3-II levels are measured in the presence of Baf or NH₄Cl, LC3-II levels are unchanged. The difference in patterns between (a) and (b) can be used to discriminate between autophagy induction and blockage. (c) When LC3-labelled vesicles (puncta) are measured within cells with a single fluorophore (e.g. cells expressing GFP-tagged LC3 or immunofluorescence labelling of the endogenous protein), an increase in puncta can be observed both in autophagy inducing and autophagy blockage conditions. N.B. Commercially available antibodies with cross-reactivity to zebrafish LC3 are widely available from suppliers such as from Novus Biologicals (used in [7–9]) and Cell Signaling Technology (used in [10]).

Schematic diagram of the tandem mRFP-EGFP-LC3 reporter to monitor autophagic flux. (a) Representation of the reporter construct mRFP-EGFP-LC3 and the behaviour of the encoded protein under different pH conditions. Under neutral pH conditions, both EGFP and RFP fluorescence is observed. Under acidic pH conditions, EGFP fluorescence is quenched and only red fluorescence is observed. (b) mRFP-EGFP-LC3 labelling during autophagosome biogenesis, maturation and degradation. Unlipidated mRFP-EGFP-LC3 remains in the cytoplasm (light yellow) whereas lipidated mRFP-EGFP-LC3 is recruited to both inner and outer membranes of phagophores and double-membrane autophagosomes. During these steps of autophagosome formation, the fluorescent signal of both fluorophores, mRFP and EGFP, is visible and vesicles appear as yellow puncta. Autophagosomes eventually fuse with lysosomes to form autolysosomes. Under these acidic conditions, the contents within the inner membrane are eventually degraded. The green fluorescent signal from EGFP is quenched in the acidic lysosomal conditions whereas the mRFP signal remains, resulting in red autolysosomes. (c) Representative images of a cell expressing mRFP-EGFP-LC3 with different levels of autophagy. The combination of green and red fluorescent signals from unlipidated mRFP-EGFP-LC3 results in a yellow background in the cytoplasm of the cells. The intensity of this yellow may change dependent upon changes in the autophagy flux. Under low autophagy conditions, most of mRFP-EGFP-LC3 remains unlipidated resulting in a yellow background and only a few yellow or red vesicles (autophagosomes and autolysosomes) are seen. After autophagy induction, many new autophagosomes form and are labelled with lipidated LC3. These rapidly fuse with lysosomes. This can be observed as an increase in the number of total vesicles and the ratio of red:yellow vesicles as well as reduced yellow background. When autophagic flux is blocked, autophagosome formation may still occur. In this scenario, autophagosomes and autolysosomes accumulate but cannot be degraded and can be observed as yellow puncta. The continuous lipidation of mRFP-EGFP-LC3 as new autophagosomes form reduces the yellow background of the cytoplasm.

Schematic diagram of the GFP-LC3-RFP-LC3DG reporter to measure autophagic flux. (a) Schematic diagram of the GFP-LC3-RFP-LC3DG reporter construct. The GFP-LC3-RFP-LC3DG protein is cleaved by ATG4 resulting in the release of GFP-LC3 and RFP-LC3DG in equimolar amounts. (b) GFP-LC3 becomes lipidated and binds to autophagosomes and autophagosome precursors, and can be visualized as green vesicles (puncta), whereas unlipidated RFP-LC3DG remains in the cytoplasm. The GFP signal is quenched when autophagosomes fuse to lysosomes to form autolysosomes. The green signal can therefore be used as a marker for phagophores and autophagosomes, but autolysosomes are not labelled. (c) Representative images of a cell expressing GFP-LC3-RFP-LC3DG with different levels of autophagy. The unlipidated RFP-LC3DG is released as an internal control at the same rate and amount as GFP-LC3 and always remains cytosolic. Levels of red signal are independent of autophagy degradation and remain unchanged upon autophagy perturbation. GFP-LC3, however, can be found unlipidated free in the cytoplasm or lipidated, hence bound to autophagic membranes, and therefore susceptible to autophagy degradation. Under high levels of autophagic flux, GFP-LC3 becomes lipidated and degraded, and thereby the level of green signal is reduced. When autophagy is blocked, the accumulation of unlipidated GFP-LC3 and the lack of degradation of the lipidated form results in an increase in the GFP signal. The ratio of the GFP:RFP (i.e. the green signal from GFP-LC3 and the unchanged mRFP-LC3DG) is then used to measure the rate of autophagic flux.

Measuring autophagy substrate clearance in vivo. (a) Zebrafish were generated which express the fluorescent, photoconvertible protein Dendra tagged to human tau, a known autophagy substrate. The green fluorescent Dendra protein can be photoconverted to a red fluorescent protein by exposure to 405 nm light. (b) Mosaic expression of the transgene allows individual neurons in the spinal cord to be identified and selected for photoconversion. Images of the same neurons were taken before and immediately after photoconversion and then at 12 h intervals. The amount of red fluorescent signal was quantified at each time point and used to calculate the clearance of tau protein. (c) Schematic diagram of the clearance kinetics of tau in response to manipulation of autophagic flux. Treatment with autophagy inducers (green) accelerates the clearance of tau protein whereas treatment with autophagy blockers (red) slows the clearance kinetics.