Vacuolar accumulation and colocalization is not a proper criterion for cytoplasmic soluble proteins undergoing selective autophagy

ABSTRACT

Autophagy is an important cytoprotective process that mediates degradation of dysfunctional or unnecessary cellular components. In the process of autophagy, a double-membrane organelle termed the autophagosome is formed to sequestrate portions of cytoplasm and subsequently delivered into lysosome or vacuole for degradation. The accumulation of autophagic bodies in the vacuoles after treatment with concanamycin A (ConcA) is a widely used protocol for monitoring the occurrence of autophagy in plants. Here, it was found that the cytoplasmic soluble GFP was accumulated in vacuoles upon ConcA treatment. Importantly, the GFP signal showed good colocalization with the autophagic marker mCherry-ATG8f in vacuoles based on two commonly used methods, the Pearson-Spearman correlation colocalization analysis and the plot profile analysis. Further results showed that the free GFP did not interact with ATG8s. Thus, analysis of accumulation and colocalization only in vacuoles is not a trustworthy way to judge whether degradation of cytoplasmic protein is dependent on the selective autophagy pathway in plants. In this short perspective, we propose several primary steps to distinguish that the cytoplasmic proteins are degraded by selective or bulk autophagy, hoping they could contribute to identify and clarify the selective autophagic cargos and receptors in plants.

KEYWORDS:

• Autophagy
• concanamycin A
• cytoplasmic soluble protein
• colocalization analysis
• cargos
• receptors

Autophagy plays an important role in the maintenance of intracellular homeostasis by degradation and recycling of cytoplasmic components, including dysfunctional organelles, unnecessary proteins, and invading pathogens. Based on their different morphological features, three types of autophagic pathways (macroautophagy, chaperone-mediated autophagy and microautophagy) have been characterized in eukaryotic cells.¹ During macroautophagy (hereafter autophagy), a double-membrane vesicle, called autophagosome, is formed from the expanding membrane of phagophore. The autophagosomes sequester the enclosed components and deliver them into the lysosome/vacuole for degradation. Mounting evidence suggests that autophagy is a highly regulated process by selectively recognizing the specific protein cargos.^2,3 The molecular machinery for selective autophagy requires autophagic receptors that recruit the targeted cargos to the growing autophagosomal membranes. ATG8s, the lipid-conjugated ubiquitin-like proteins, are core components of the autophagy machinery. Most of the proteins that are specifically turned over by selective autophagy are recognized by the presence of short ATG8 interacting motifs (AIMs) that facilitate their association with the autophagy apparatus.^4,5 Over the last decade, ground-breaking studies in yeast and mammalian models uncovered molecular details of selective cargo recognition. However, there is still a large gap in our understanding of selective autophagy process in plants. Although the precise molecular mechanisms of cargo selection by autophagy are yet to be established, an increasing number of autophagic receptors or cargos, including AtNBR1, DSK2, RPN10, Exo70E2, AtSec62 and C53 have been identified and characterized for selective autophagy degradation in plants (Table 1).^6–11 In the process of their function being clarified, colocalization analysis with the autophagic marker ATG8s is an important step in selective cargos identification.

Table 1. Autophagic cargos and/or receptors have been identified and characterized for selective autophagy in plants

Autophagic cargos/receptors	Descriptions	Intracellular distribution	Colocalization pattern	Interaction analysis	AIMs site	References
AtNBR1	AtNBR1, a selective autophagic receptor, anchors the ubiquitinated targets to autophagosomes	cytoplasmic soluble and punctate protein	Forms both ring-like and punctate structures, and colocalizes with ATG8 in cytoplasm	Pull down; Y2H; FRET; in vivo recruitment	660-WDPI-665	^Citation6,Citation7
AtSec62	AtSec62, an ER-phagy receptor, is recruited to autophagosomes during ER stress	ER	Forms ring-like structures and colocalizes with ATG8 in cytoplasm	Co-IP	AIM1, 160-WQTL-165; AIM2, 215-WILL-220	^Citation8
C53	C53, an ER-phagy receptor, is recruited to ATG8 proteins upon ER stress	Cytoplasmic soluble protein	Forms ring-like structures and colocalized with GFP-ATG8a in cytoplasm	pull-down; SPR	Shuffled AIM	^Citation9
DSK2	DSK2, an autophagy receptor, recruits BES1 to the autophagy pathway.	Localizes in the cytosol and nucleus	Colocalizes with BES1 and ATG8 in cytoplasm	Pull down; Y2H; BIFC; Co-IP	AIM1, 42-FKEL-47; AIM2, 248-FNML-253	^Citation10
RPN10	RPN10, a selective autophagy receptor, mediates degradation of the proteasome	Localizes in the cytosol and nucleus	Colocalizes with ATG8 in cytoplasm	Y2H; BIFC; Pull down	UIM	^Citation11
ATI1&ATI2	Serve as cargo receptors for chlorophagy and ER-phagy	Associates with the ER or ER vesicles	Colocalizes with ATG8 in vacuole upon ConcA treatment	Y2H; 1D 1 H NMR spectra	13-WEVV-18	^Citation12–14
EXO70B2	Interacts with ATG8 and transport into the vacuole	Cytoplasmic soluble protein	Colocalizes with ATG8a and AtNBR1 in vacuole after BTH and ConcA treatment	Pull down; BIFC	AIM1, 509-YPKL-514; AIM2, 523-FDEV-528	^Citation15
EXO70D1/2/3	EXO70D proteins act as selective autophagy receptors to target type-A ARR cargos for autophagic degradation	Cytoplasmic soluble proteins	Colocalizes with ATG8 in vacuole upon ConcA treatment	Y2H	Not test	^Citation16
ORM1&ORM2	Function as selective autophagy receptors for FLS2 degradation	ER	Not test	Y2H; Co-IP	ORM1, 37-WLVV-42; ORM2, 34-WLVV-39	^Citation17
GSNOR1	GSNOR1, a regulator of NO signaling, is degraded by selective autophagy after S-nitrosylation at Cys-10	Cytoplasmic soluble proteins	Colocalizes with ATG8 in cytoplasm, and in vacuole upon ConcA treatment	Co-IP; LCI	151-YTVV-156	^Citation18
TSPO	TSPO, a heme binding membrane protein, is downregulated by heme binding, and autophagy pathway	Membrane-anchored protein	Partially colocalizes with ATG8 in cytoplasm	Not test	121-YLYL-126	^Citation19
EXO70E2	Delivery of EXO70E2 into the vacuole is mediated by AtNBR1 upon autophagic induction	Cytoplasmic punctate protein	Colocalizes with ATG8 in the presence of AtNBR1 in cytoplasm	Interacts with AtNBR1 by Y2H; FRET; Co-IP; in vivo recruitment	Not test	^Citation7
HSP90&ROF1	HSP90&ROF1, are degraded by AtNBR1-mediated autophagy during the heat shock stress.	Cytoplasmic soluble proteins	Colocalizes with AtNBR1 in vacuole upon ConcA treatment	Interacts with AtNBR1 by Co-IP; BIFC	Not test	^Citation20
AtSar1d	A small GTPase, regulates autophagosome progression through specific recognition of ATG8e	Coat protein complex II (COPII) vesicles	Colocalizes with ATG8e upon BTH and ConcA treatment	Co-IP; in vivo recruitment; FRET	a noncanonical motif	^Citation21

Y2H, yeast two-hybrid; FRET, Förster (Fluorescence) resonance energy transfer; Co-IP, co-immunoprecipitation; BIFC, bimolecular fluorescence complementation; NMR, nuclear magnetic resonance; SPR, surface plasmon resonance; Shuffled AIM, the sequence ‘IDWG’, representing a shuffled version of the canonical AIM sequence (W/F/Y-X-X-L/I/V); UIM, Ubiquitin interaction motif; LCI, luciferase complementation imaging.

Concanamycin A (ConcA) is one of the widely used inhibitors to study autophagy in plants. ConcA interferes with vacuolar-type H⁺-translocating ATPase by binding to the subunit c of the proton translocating Vo complex.²² Inhibition of the V-ATPases results in blocking vacuolar acidification and impeding hydrolases activity in the vacuoles.²³ Thus, the accumulation of autophagic bodies in vacuoles upon treatment with ConcA is applied for monitoring the occurrence of autophagy in plants.²⁴ Here, it was found that the cytoplasmic soluble GFP were accumulated in vacuoles, showing globular clusters (indicated by arrows) and punctate structures (indicated by arrowheads), after treatment with ConcA for 6 hours. These two patterns might be due to their different extents of degradation. However, the delivery of GFP protein into vacuoles was abolished in autophagic defective mutant atg7-2 (Figure 1(a)). Upon BTH and ConcA (BTH+ConcA) treatment, a common autophagy-inducible condition,²⁵ the GFP signal were colocalized with the autophagic marker mCherry-ATG8f in vacuoles, showing punctate (Inset 2) and globular structures (Inset 3), respectively (Figure 1(b,c)). These signals in vacuoles showed highly dynamic movement under time-lapse imaging (Supplemental movie 1). Remarkably, the GFP protein was not recruited to mCherry-ATG8f puncta in the cytoplasm under the control condition (Inset 1 of Figure 1(c,d)). To qualitative the level of co-occurrence of these fluorescing proteins, two common colocalization methods, the Pearson-Spearman correlation analysis and the plot profile analysis were used. Both analyses showed that the GFP signals had good colocalization with the autophagic marker mCherry-ATG8f (Figure 1(d,e)).

Figure 1. Cytoplasmic soluble GFP protein is nonselectively engulfed by autophagosomes and delivered into vacuole upon autophagy induction condition

(A) Accumulation of soluble GFP protein in vacuole is dependent on autophagy pathway. Five-day-old Arabidopsis seedlings expressing GFP were treated with ConcA (0.5 μM) for 6 hours and then were subjected to confocal observing. The GFP signal was observed in vacuole upon ConcA treatment, but was not observed in autophagic defective mutant atg7-2. Arrows and arrowheads indicated globular and punctate structures of GFP in vacuole, respectively. Bar = 50 μm. (B) Colocalization between the GFP protein and the autophagic marker mCherry-ATG8f in vacuole. Five-day-old Arabidopsis seedlings co-expressing GFP and mCherry-ATG8f were treated with BTH and ConcA (BTH+ConcA) for 6 hours to induce autophagy, and then were subjected to confocal imaging. Bar = 20 μm. BTH, Benzo-(1,2,3)-thiadiazole-7-carbothioic acid S-methyl ester. (C) The enlarged images of insets in (B). Inset 1, the GFP signal was not colocalized with mCherry-ATG8f puncta in the cytoplasm. Arrowheads indicated autophagosome-like structures. Inset 2, colocalization between the punctate structures of GFP and mCherry-ATG8f in vacuole. Inset 3, colocalization between the globular structures of GFP and mCherry-ATG8f in vacuole. Bar = 10 μm. (D) Double intensity plots of colocalizing GFP and mCherry-ATG8f from the inset 1 (Left) and inset 2 (Right). The green and red lines represent gray values of GFP and mCherry-ATG8f signals, respectively. The intensities were analyzed with Image J. (E) Colocalization analysis with the PSC plug-in of ImageJ. The linear Pearson correlation coefficient (r_p = .73) and the nonlinear Spearman correlation coefficient (r_s = 0.59).

GFP has a capacity to bind aggregated proteins like amyloid fibrils in mammal.²⁶ To test whether GFP could be degraded by the selective autophagy pathway, an in vivo recruitment assay was adopted. In this system, the transmembrane domain of calnexin (CNX) is fused to N-terminal of fluorescent tag RFP and then connected with ATG8e (CNX-RFP-ATG8e). When co-expression of the interest cargos, they would be recruited to endoplasmic reticulum (ER)-localized CNX-RFP-ATG8e, showing the linear ER patterns.⁷ Compared with the positive control AtNBR1-GFP, the free GFP cannot be recruited to CNX-RFP-ATG8e (Figure 2(b)), indicating that GFP was not recognized by ATG8e. In general, the macroautophagy and microautophagy confers two mainly pathways for vacuolar targeted degradation.^1,27–29 The soluble GFP protein distributed in the cytoplasm might be nonselectively engulfed by autophagosome or directly invaginated at the tonoplast and then delivered into vacuole (Figure 2(c,f)). Thus, the use of vacuole-targeted accumulation and colocalization analysis as a reporter in autophagy study should be carefully considered. However, looking at the autophagic cargos or receptors identified over the last decade in the field of plant research, we found that nearly half of the findings were based on co-localization analysis in vacuoles after treatment with ConcA (Table 1).

Figure 2. Proposed model for the nonselective cytoplasmic soluble protein and the selective cargos engulfed by autophagosomes

(A) in vivo recruitment model to identify cargos/receptors that degraded by selective autophagy. (B) The free GFP cannot be recruited to ER-localized CNX-RFP-ATG8e. The paired constructs were transiently co-expressed in Arabidopsis mesophyll protoplasts. AtNBR1-GFP was used as a positive control.⁷ Bar = 10 μm. (C-F) Different pathways that may contribute to cytosol proteins accumulated and colocalized with ATG8 in the vacuole. The autophagosomes are formed to sequestrate partial cellular components containing cytosolic soluble proteins, like GFP, and subsequently fused with the vacuole to deliver the inclusions into the vacuolar lumen (C). For the soluble cargos, they are specially recognized and recruited by ATG8 to the double membrane of autophagosome. In this case, ring-like structures of these cytosolic cargos, such as C53, sometimes can be observed in cytoplasm (D). For the protein aggregates, they are recognized and engulfed by autophagosomes. In cytoplasm, the aggregates are colocalized with the autophagic marker ATG8 in a punctate pattern. Both of nonselective and the selective cytosolic materials were showed colocalization after the autophagic bodies accumulated in the vacuole (E). The cytoplasmic components might be directly engulfed by microautophagy at the level of the tonoplast or of the membranes of the MVB (multivesicular body) and then enter into the vacuolar lumen by invagination and scission (F).

How to identify one cargo, especially the cytoplasmic soluble protein, was degraded by selective autophagy or bulk autophagy or microautophagy? To better distinguish these pathways, it is easier to get some hints from the autophagic process shown in Figure 2(c,f)). For the nonselective proteins, the autophagosomes are formed to sequestrate portions of the cytoplasm that contains cytosolic materials like GFP protein, and subsequently deliver the inclusions into vacuole for degradation. The nonselective cytosolic materials are shown by colocalization with the autophagic marker ATG8s after the autophagic bodies accumulated in vacuoles (Figure 2(c)). For the soluble cargos/receptors, they are specially recognized and recruited by ATG8s to the double membrane of autophagosomes. In this case, these proteins should be colocalized with ATG8s and sometimes show ring-like structures in cytoplasm (Figure 2(d)), which have been already verified in some receptors like C53 and AtNBR1.^7,9 For the insoluble cargos, like protein aggregates, they are recognized by ATG8s and engulfed by autophagosome, showing colocalization in cytoplasm (Figure 2(e)). Additionally, microautophagy might directly invaginate the cytoplasmic components at the level of the tonoplast or of the membranes of the MVB (multivesicular body) and then deliver into the vacuolar lumen (Figure 2(f)). Conspicuously, both the nonselective and the selective cytosolic materials are shown to colocalization after the autophagic bodies accumulated in vacuoles. Thus, colocalization analysis of interest protein with the autophagic marker ATG8s in cytoplasm, but not in vacuoles in the presence of ConcA, is an important and simple method to distinguish whether the degradation is dependent upon the selective autophagy pathway in plants.

Previously, a basic set of guidelines for the selection and interpretation of assays for studying autophagy in different organisms has been well documented.³⁰ Here, we propose several primary steps to distinguish that the cytoplasmic proteins are degraded by selective or nonselective autophagy, hoping these suggested points could contribute to standardize research of autophagy in plants.

Step 1. Screening ATG8s-interacted cargos/receptors

Identifying the interacted proteins of ATG8s is the first and most important step. To elucidate and identify the interacting cargo/receptor of ATG8s, some commonly methods including the yeast two-hybrid (Y2H) screen and immunoprecipitation followed by mass spectrometry (IP-MS) analysis have been widely used. Recently, new methods like proximity labeling have been developed for capturing protein–protein interactions in a vast array of biological contexts.³¹ The effectiveness of proximity labeling in mapping ATG8s-interacted networks will open up a new line of research in the autophagy field in plants.

Step 2. Performing colocalization analysis

Colocalization with the fluorescent tag fused ATG8s in images is an important way to verify special autophagic cargos or receptors. At this step, to avoid random separation of cytoplasmic components by nonselective autophagy, we strongly suggest to do colocalization analysis in cytoplasm but not in vacuoles (Figure 1). For the soluble proteins, the ring-like structures sometimes could be observed in the cytoplasm. Since autophagy is activated under various abiotic and biotic stresses, it is necessary to consider induction of autophagy with a proper condition. Starvation by deprivation of carbon or nitrogen source is one of the most used methods. To monitor distinct organelle-specific autophagy, specific treatment conditions need to be carried out, such as induction of endoplasmic reticulum (ER) stress by tunicamycin (TM) and dithiothreitol (DTT),⁸ as well as induction of mitophagy by some uncouplers like 2,4-dinitrophenol (DNP) and carbonyl cyanide ptrifluoro-methoxyphenyl hydrazone (FCCP).³² Additionally, data obtained from overexpression assays should always be considered with caution, because overexpressing proteins may affect its distribution causing aggregate formation and autophagy.

Step 3. Interaction analysis and AIM sites identification

Physical interaction between the cargos/receptors and ATG8-family proteins is required for subjecting them to autophagic degradation. It is important to verify protein–protein interactions both in vitro and in vivo through methods such as yeast two-hybrid assays (Y2H), bimolecular fluorescence complementation (BiFC), Förster (Fluorescence) resonance energy transfer (FRET), co-immunoprecipitation (Co-IP), and pull-down assay. Recently, we have developed an ER-targeted in vivo recruitment assay to screen the cargos recruited by autophagic receptor AtNBR1.⁷ The in vivo recruitment assay is a powerful tool to analyze the interact partners of ATG8s, especially for the cytoplasmic soluble cargos. By relocating the targeted proteins at the subcellular level, this system could effectively avoid background interference caused by the dispersive distribution of soluble proteins.

Most of ATG8-family interacting proteins contain a core consensus sequence W/F/Y-x-x-L/I/V (x = any amino acid) called LIR (LC3-interacting region) motif, also named ATG8 interacting motif (AIM).⁴ AIM binds to the hydrophobic pocket of ATG8s that comprise a docking surface known as LIR docking site (LDS). Mutational analysis of the AIM sites is important to validate the interest proteins as an autophagy substrate. Recently, different from the canonical AIM, new ATG8 binding interfaces like the ubiquitin-interacting motif (UIM) and the shuffled ATG8 interacting motifs (sAIM) have been found.^9,11,33 These findings extend the range of autophagic receptors and cargos, and reveal a higher complexity in autophagic cargos selection.

Step 4. Degradation pathway analysis

The proteasome and the autophagosome constitute two major cytoplasmic protein degradation pathways in eukaryotic cells.³⁴ Their activities are carefully orchestrated to control various protein abundance and maintain cellular homeostasis. The GFP tagged cargo protein is often used to monitor autophagic flux due to their different sensibility to degradation in vacuole (GFP is relatively resistant to hydrolysis). A concentration-curve and time-course analysis with or without autophagic flux inhibitors could help to evaluate the extent of vacuolar delivery and proteolysis of interest cargos. Moreover, genetic approaches to block autophagy with several core autophagy-related gene mutants are required for validating the autophagic protein degradation pathway. Importantly, studies should also compare the effects of proteasomal degradation inhibitors like MG132 to determine the contribution from the proteasome. Recent findings strongly suggest that there is crosstalk and cooperation between the proteasome and the autophagosome pathways. Actually, these two degradation pathways have been found coordinately regulating BES1 (BRI1-EMS SUPRESSOR 1) degradation.¹⁰ The proteasome is itself degraded by autophagy upon nitrogen starvation and proteasome inhibition.¹¹ Thus, comparative analyses of the proteasomal and the autophagy pathways are helpful to clarify their substrate selectivity and compensatory regulation of protein degradation.

Conclusions

The selective autophagy has emerged as a new and potent modulator of various abiotic and biotic stress responses in plants. Even though a growing number of ATG8s-interacting proteins have been identified, there remain challenges to our understanding of how specific cargos are targeted for degradation. Colocalization analysis with autophagic marker ATG8s has always been considered as an important step to determine whether the degradation of a protein is dependent on autophagy. Here, the cytoplasmic soluble GFP protein was found accumulating in the vacuoles and showing good colocalization with the autophagic marker mCherry-ATG8f after treatment with ConcA. Therefore, when reporting special cytosol proteins degraded by selective autophagy, we need to do colocalization analysis in cytoplasm rather than in vacuoles to avoid the cytoplasmic components that are nonselectively delivered into vacuoles by bulk autophagy or microautophagy. Additionally, colocalization analysis is just a beginning, and multiple parallel verifiable experiments, especially protein–protein interaction between ATG8 and the cargo/receptor, need to be carried out to reach a proper conclusion.

Disclosure Statement

No potential conflicts of interest are disclosed.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by National Science Foundation of China (31600288); Guangdong Provincial Science and Technology Project (2016A020210127); SCNU Youth Teacher Research and Development Fund Project (671075); Scientific Research Projects of Guangzhou (201805010002);National Natural Science Foundation of China [31600288].