Unveiling the Molecular Mechanisms of Plant Autophagy—From Autophagosomes to Vacuoles in Plants

Abstract

Autophagy is an evolutionarily conserved intracellular vacuolar process. Since Christian de Duve first coined the term ‘autophagy’ in 1963, it had not been well understood at the molecular level until much later, due to limitations in biochemical approaches and/or morphological approaches posed by electron microscopy. An important milestone was achieved with the isolation and identification of autophagy-related (ATG) genes by genetic screening using yeast Saccharomyces cerevisiae. ATG genes are well conserved in most eukaryotic organisms, which allowed the subsequent isolation of ATG gene-knockouts in plants. From the phenotypic analyses of the autophagy-defective plants, the physiological roles of autophagy have been predicted. However, in some cases, all the phenotypes cannot be simply explained by defects in autophagy. Therefore, in order to fully understand the physiological implications of plant autophagy, it is quite important to elucidate the molecular mechanisms involved in each process in macro-/micro-autophagy. Although, until recently, our understanding of the molecular mechanisms of plant autophagy was lagging compared to similar research in yeast and animals, current studies have made many great advances in the plant research field. In this review, we discuss current knowledge of the molecular mechanisms of plant autophagy, from autophagy-induction/autophagosome-formation to vacuolar degradation, comparing these to processes in yeast and mammals. We also review aspects of plant autophagy research that require further investigation in the future.

Introduction

Since plants are immobile organisms, they frequently encounter various types of adverse environmental stresses. To survive under such severe conditions, plants have to adapt and cope with external environments. Autophagy is one of the key processes for such adaptation and coping, which allows plants to undergo intracellular remodeling and physiological changes in response to those conditions. Autophagy degrades cytoplasmic components in the vacuole to provide raw materials and energy for the plant, and also to eliminate damaged/toxic components for the maintenance of essential cellular functions. Two types of autophagy, called macro- and micro-autophagy, have been well described in plants (Bassham et al. 2006). Macroautophagy is achieved via a unique double membrane vesicle, called an autophagosome, which encloses a portion of the cytoplasmic constituents. Microautophagy is achieved by the invagination of the vacuolar membrane to incorporate cytoplasmic constituents into vacuolar lumens. In both pathways, the vacuole is the final destination for the degradation of cytoplasmic constituents. In this mini review article, we will focus on macroautophagy, which is most extensively studied in many organisms.

In plants, the existence of autophagic processes have been reported since the late 1960s by morphological analyses using electron microscopy (EM). However, due to the limitations of static EM analysis, the physiological roles and molecular mechanisms of plant autophagy had not been well understood until around the year 2000. Similar to autophagy studies in other organisms, the isolation and identification of ATG genes by genetic screening using yeast Saccharomyces cerevisiae made a breakthrough in plant autophagy research. Genes for most core ATG proteins have been well conserved beyond kingdoms in eukaryotes and their universality also allowed for the isolation of autophagy mutants in plants. The phenotypes of the autophagy mutants suggested that autophagy affects many aspects of physiological functions in plants [e.g. resistance to abiotic stresses such as nutrition starvation (Yoshimoto et al. 2004), drought (Liu et al. 2009), salt (Liu et al. 2009), high temperature (Zhou et al. 2013), oxidative stresses (Xiong et al. 2007) and biotic stresses such as necrotrophic pathogens (Lenz et al. 2011)]. However, those phenotypes are in some cases caused by indirect effects. Therefore, in order to truly understand the physiological roles of plant autophagy, it is very important to elucidate how autophagy is induced, how autophagosomes are formed, what the cargoes in each stress condition are, and how cytoplasmic constituents are delivered to the vacuole (Fig. 1). Studies in the last decade in the plant autophagy field have elucidated these issues at the molecular level although research is still ongoing and in its infancy. Here, we describe autophagic processes at the molecular level and discuss aspects of plant autophagy research that require future investigation.

Fig. 1

Schematic diagram of entire macro-autophagic processes at the molecular level. Left; list of core ATG proteins essential for autophagosome formation. ATG molecules marked by asterisks are still not identified in plants. Right; schematic diagram of macroautophagy (hereafter referred to as autophagy). Autophagy is induced by stresses such as nutrient starvation and so on. The TOR kinase complex negatively regulates autophagy through direct phosphorylation of ATG13, which inhibits formation of the ATG1/ATG13 protein kinase complex, leading to inactivation of ATG1 activity. ATG9 localized small vesicles derived from Golgi bodies serve a source for initiating autophagosome formation. PI3P produced by the PI3K complex is essential for autophagosome formation. Two ubiquitination-like reactions consisting of ATG8- and ATG12-conjugation systems are important for autophagosome expansion and enclosure. During autophagy, ATG8 confers selectivity for specific cargo degradation. The outer membrane of the autophagosome fuses with the vacuolar membrane for transportation of cytoplasmic constituents for vacuolar degradation. ATG molecules shown by ‘Atg’ with two small letters have not yet been identified or are not confirmed to form complexes in plants. In this review, the following four points are made and discussed (in some instances alongside cases of yeast and mammals) about the entire autophagic process: 1. How is plant autophagy induced? 2. What does IM originate from? 3. What are the specific cargoes in plant autophagy? 4. How do autophagosomes fuse with the vacuolar membrane?

Induction of Plant Autophagy

One of the major physiological roles of autophagy in plants is cellular adaptation to nutrient starvation, i.e. the transient gain of nutrition and energy by the degradation of a cell’s own proteins and organelles. Therefore, autophagy has to be induced properly depending on nutritional conditions. In yeast, it is well known that target of rapamycin (TOR), which is a central protein kinase of the nutrient-sensing pathway, negatively regulates autophagy (Blommaart et al. 1995, Noda and Ohsumi 1998). This negative regulation of autophagy is achieved through the direct hyperphosphorylation of Atg13 at multiple serine residues by TOR complex 1 (TORC1) (Kamada et al. 2010). During nutrient starvation, TOR kinase is inactivated, leading to the dephosphorylation of Atg13 by an unknown phosphatase. This facilitates the activatation of the Atg1 kinase via association with Atg13 and Atg17, which forms the Atg1-Atg13-Atg17-Atg29-Atg31 protein complex required for autophagy induction.

In plants, especially in Arabidopsis, one TOR homolog (Menand et al. 2002), four ATG1 homologs (Suttangkakul et al. 2011), and two ATG13 homologs (Suttangkakul et al. 2011) have all been identified, but those for ATG17, ATG29, or ATG31 have not. Instead, other ATG proteins have been identified in Arabidopsis: ATG101, which forms a complex with the ATG1 homolog, ULK1 (Unc-51 like autophagy activating kinase), and a homolog of the mammalian ATG13 protein (Li et al. 2014). It has been shown by yeast two hybrid system (Y2H) and bimolecular fluorescence complementation (BiFC) analyses that Arabidopsis ATG101 interacts with ATG1a, ATG13a, and itself (Li et al. 2014). Interestingly, it also interacts with ATG11 probably due to the existence of ATG17-like domain in Arabidopsis ATG11. In yeast, ATG11 functions in selective autophagy, but not in bulk autophagy. Since Arabidopsis ATG11 has an ATG17-like domain, it may function both in bulk and selective autophagy. Thus, ATG1/13 protein kinase complexes seemingly exist in plants as in mammals. In Arabidopsis, ATG13a is phosphorylated under non-starvation conditions and dephosphorylated under starvation conditions. Additionally, ATG1a is phosphorylated during carbon and nitrogen starvation, as in other eukaryotes (Suttangkakul et al. 2011). Furthermore, the atg13a atg13b double mutant has been indicated to be defective in autophagy (Suttangkakul et al. 2011). However, it is still unknown whether the TOR kinase directly phosphorylates ATG13 resulting in the inactivation of ATG1 activity in plants. Interestingly, unlike in mammals, Arabidopsis ATG1a and ATG13a proteins are dramatically degraded under nutrient-limiting conditions by an autophagic route, suggesting that ATG1/13 kinase complexes are targets of autophagy (Suttangkakul et al. 2011). This may guarantee a negative feedback loop of autophagic activity in a plant-specific manner.

It has been demonstrated that plant autophagy is dynamically induced under nitrogen and/or carbon starvation (Merkulova et al. 2014). The TOR kinase is negatively involved in this process. Down-regulation of TOR in Arabidopsis leads to constitutive autophagy, meaning that autophagy is induced even in nutrient-sufficient conditions. Furthermore, autophagy is more enhanced in the TOR RNAi lines than in wild-type plants under nutrient starvation conditions (Liu and Bassham 2010, Merkulova et al. 2014). It is now widely observed that autophagy is upregulated under a wide range of abiotic stress conditions in plants (Pu et al. 2017). For example, salt-, osmotic-, oxidative-, ER (endoplasmic reticulum)-stresses can all induce autophagy in plants. In addition, treatment of salicylic acid (SA) related molecules such as BTH, an SA agonist, is also indicated to induce autophagy in plants in order to negatively regulate cell death (Yoshimoto et al. 2009). A recent, interesting study revealed that overexpressing TOR kinase only inhibits salt and osmotic stress-induced autophagy, but not oxidative or ER stress-induced and SA-induced autophagy, suggesting the existence of TOR-dependent and TOR-independent pathways to regulate autophagy in plants (Pu et al. 2017). In yeast, it has been reported that zinc depletion triggers autophagy and this induction is controlled by inactivation of TORC1 (Kawamata et al. 2017). Recently, we found that zinc deficiency induces autophagy also in plants (K. Yoshimoto, unpublished data; Eguchi et al. 2017). However, it is still unknown whether this process is controlled by TOR kinases or not.

Origin of the Isolation Membrane and Its Formation

One of the important and interesting, long-standing issues in the autophagy field is the elucidation of the origin of the isolation membrane (IM). In 2009, it was revealed by electron tomography that autophagosomes are formed in a subdomain of the ER, which formed a cradle encircling the IM, suggesting that the ER participates in autophagosome formation (Hayashi-Nishino et al. 2009). However, another report published at roughly the same time suggested that IM are derived from mitochondrial outer membranes during starvation (Hailey et al. 2010). Thereafter, imaging data revealed that the pre-autophagosome structure (PAS)/autophagosome marker, ATG14, and the autophagosome-formation marker, ATG5, localizes to the ER-mitochondria contact site after starvation. From these data, the now widely accepted model is that the ER-mitochondria contact site is important in autophagosome formation in mammalian cells (Hamasaki et al. 2013). However, this does not exclude the possibility that other membranes, such as the plasma membrane, participate in autophagosome formation. Sources of IMs may be different depending on the cellular situations.

In yeast, it is known that autophagosomes form in proximity to the vacuolar membrane. However, current imaging data indicated that the ER is closely connected to PAS/IM. Although the imaging is conducted in an artificial system in which aminopeptidase I (Ape1) is overexpressed and thus resulting in forming stable cup-shaped IMs, it was indicated that the ER exit site in which COPII vesicles form is adjacent to the opening edge of the IM. Indeed, it is known that yeast mutants defective in the formation of COPII vesicles fail to form autophagosomes (Suzuki et al. 2013). These results support an idea that ER participates in autophagosome formation through COPII vesicles. However, direct evidence is still lacking. In addition, during autophagosome membrane formation, the most important molecule is Atg9 because it is the sole Atg protein harboring membrane-spanning regions. High-sensitive CCD camera imaging has revealed that Atg9 localizes on 30–60 nm cytoplasmic small vesicles derived from the Golgi apparatus. Atg9 vesicles have long been considered a source of lipids in the formation of autophagosomes. However, recent quantitative analysis using fluorescent microscopy in yeast indicated that in response to starvation, only 3 to 5 Atg9 vesicles are recruited to the PAS through interaction with Atg13 inside Atg1 complexes. From these results, Atg9 is now proposed to function at an early stage as an initiator of IM formation rather than contributing as a lipid source for membrane expansion (Yamamoto et al. 2012, Suzuki et al. 2015).

In plants, information about membrane sources for the IM and the molecular machineries of autophagosome formation is in its infancy compared to in yeast and mammals. However, recent findings provide molecular evidence that plant autophagosomes are derived from the ER. In Arabidopsis, although autophagy is not completely impaired, ATG9 disrupted mutants accumulate less autophagic bodies, suggesting that the atg9 mutants have defects in normal autophagosome formation. Consistent with this fact, upon autophagy induction, atg9 mutants accumulate abnormal autophagosome-related tubules, which are related to the ER. Electron tomography revealed that in the atg9 mutant, the ER connects to developing autophagosomal membranes via multiple narrow membranes. Additionally, time-lapse imaging showed that ATG9 vesicles transiently interact with the autophagosomes and move away after a quick interaction (Zhuang et al. 2017). These results imply that ATG9 is required for efficient budding of IM from the ER membranes unlike in yeast and mammals. Another paper also supports the model that plant autophagosomes originate from the ER. By simultaneous imaging of the autophagosome-formation marker, ATG5, the autophagosome marker, ATG8, and an ER marker, the close association of autophagosomal structures and ER membranes has been observed in Arabidopsis. These in vivo imaging data suggest that the ER plays an important role in autophagosome formation also in plants (Le Bars et al. 2014).

A plant-specific factor that is involved in autophagosome formation and is colocalized with an ER marker has also been reported (Zhuang et al. 2013): SH3 DOMAIN-CONTAINING PROTEIN2 (SH3P2), which is a Bin-Amphiphysin-Rvs domain-containing protein, binds to phosphatidylinositol 3-phosphate (PI3P) and ATG8. SH3P2 colocalized to the autophagosomal membrane and Dex-inducible SH3P2 RNAi knockdown plants showed less autophagosomes and autophagic bodies during autophagy-induced conditions, suggesting that SH3P2 regulates autophagosome formation. In addition, immuno-EM analysis using Carnexin, an ER marker, and SH3P2 antibodies showed their colocalization on the same autophagosome structures, further suggesting that ER membranes are sources of autophagosomal membranes in plants. The fact that SH3P2 binds to PI3P and associates with ATG6 and VPS34, which are the components of PI3K, suggests that the function of the PI3K complex is important for autophagosome formation in plants in the same way as in yeast and mammals. Although it is still unknown whether plant autophagosome membranes contain PI3P, as in yeast (Obara et al. 2008), chemicals inhibiting PI3K activity (3-methyladenine and wortmannin) inhibit plant autophagy as well (Moriyasu and Ohsumi 1996, Merkulova et al. 2014), further supporting the above suggestion that PI3P produced by PI3K is essential for autophagosome formation in plants.

It is not yet known whether PAS exist in plants and, if so, whether IM formation occurs from the PAS in plant cells, as in yeast. Little is known about IM formation in plants as the equivalent molecules that localize on the initiation site of the IM in yeast, such as ATG14, ATG16, and ATG17, have not yet been identified or characterized due to a lack of readily identifiable homologues based on their primary structures. Future studies should shed further light on the detailed molecular mechanisms of IM formation in plants.

Selective Cargoes for Plant Autophagy

Since the phenotypes of plant autophagy mutants are rather pleiotropic, it is difficult to determine the true physiological role(s) of plant autophagy. In order to fully appreciate the importance of autophagy in many aspects of the plant life cycle, it is critical to know exactly which cargoes are degraded by autophagy under each condition. Selective autophagy is indeed now a hot topic in the autophagy field. Recently, cellular constituents including protein aggregates and organelles, as well as pathogens, have been shown to be cargoes for selective autophagy in yeast and mammals (Farré and Subramani 2016). It is well known in yeast and mammals that selective autophagy is executed by the direct interaction between ATG8, which localizes on the IM, and the receptor/adaptor protein(s) harboring ATG8 interacting motifs (AIM)/LC3 interaction region (LIR) (Noda et al. 2010).

Since autophagy-defective plants are sensitive to environmental stresses such as salt-, drought- (osmotic-), high temperature-, and hypoxia-stress (Liu et al. 2009, Chen et al. 2015), autophagy is believed to help plants protect themselves against such stresses. However, the molecular mechanisms for this have not been well elucidated. During high temperature-stress conditions, plant autophagy seems to be selective, which is considered to be mediated by a cargo receptor, Neighbor of BRCA1 gene 1 (NBR1) (Table 1). Plant NBR1 is a hybrid protein which harbors functions of both p62 and NBR1 proteins in mammals. It binds to ATG8 and ubiquitin for the selective recognition of cargoes. Arabidopsis nbr1 mutants highly accumulate ubiquitylated unsoluble proteins and show early cell death compared to wild-type plants, similar to atg mutants. However, nbr1 mutants do not show hypersensitivity to carbon starvation unlike atg mutants, suggesting that NBR1 is involved in the selective degradation of denatured proteins under high temperature conditions, but is not involved in bulk autophagy (Zhou et al. 2013). Another cargo receptor, tryptophan-rich sensory protein (TSPO), which plays an important role in adaptation to high temperature- and drought-stresses, has been identified in plants (Table 1). TSPO is transiently induced by abiotic stresses and acts as a haem scavenger via ATG8-mediated selective autophagy, managing excess and deleterious haem levels in the cell (Vanhee et al. 2011). Interestingly, TSPO binds to another target protein, plasma membrane intrinsic protein 2;7 (PIP2;7), which is a plasma membrane-localized aquaporin. This interaction facilitates the selective degradation of aquaporin, resulting in decreased levels of aquaporin in the plasma membrane, which suppresses water permeation from cells. Thus, TSPO is a selective autophagy receptor for PIP proteins that is thought to confer adaptation to high temperature- and drought-stresses (Hachez et al. 2014).

Table 1

Selective autophagy in plants

Types of selective autophagy	Selective cargo	Cargo receptor(s)/ adaptor(s)	Conditions	Refs
Quality control
Aggrephagy	Insoluble ubiquitinated protein aggregates	NBR1	Heat stress	Zhou et al. 2013
Proteaphagy	26S proteasome	RPN10	Proteasome inhibitor treatment	Marshall et al. 2015
Entire chlorophagy	Entire photodamaged chloroplasts	Unknown	UV-B irradiation, strong light	Izumi et al. 2017
Pexophagy	Damaged/abnormal leaf peroxisomes	Unknown	Normal growth conditions	Yoshimoto et al. 2014
Reticulophagy?	Damaged fragmented ER	Unknown	ER stress (tunicamycin/DTT)	Liu et al. 2012
Stress response
TSPO-mediated autophagy	Porphyrins (Haem)	TSPO	ABA treatment	Vanhee et al. 2011
TSPO-mediated autophagy	Aquaporin PIP2;7	TSPO	Drought? ABA treatment	Hachez et al. 2014
Nutrient recycle
Piecemeal chlorophagy	RCBs, ATI-PS bodies	Unknown for RCB, ATI1	Darkness (carbon starvation), senescence	Izumi et al. 2010*; MichaeLi et al. 2014*
Reticulophagy	ATI-ER	ATI1, ATI2	Carbon starvation	Honig et al. 2012*
Mitophagy?	Mitochondria	Unknown	Dark-induced senescence	Li et al. 2014
Anti-viral
Virophagy	CaMV	NBR1	Virus infection	Hafrén et al. 2017*

RCBs, rubisco-containing bodies; ATI, ATG8-interacting protein; ATI-PS, plastid-associated ATI; ATI-ER, ER-associated ATI; CaMV, cauliflower mosaic virus;

TSPO, tryptophan-rich sensory protein; ABA, abscisic acid; PIP, plasma membrane intrinsic protein.

References marked by asterisks are not mentioned in this review article.

One of the targets of selective autophagy is protein complexes (Table 1). Like in mammals, plant proteasome complexes are selectively recognized and degraded by autophagy. In Arabidopsis, the inactivated 26S proteasome is degraded by ATG8-mediated selective autophagy via a bridge of ubiquitin receptor, RPN10, between ATG8 localized autophagosomes and ubiquitylated proteasomes (Marshall et al. 2015).

Other types of selective autophagy are the degradation of organelles for quality control (Table 1). A typical function of plants that is distinct from other organisms is photosynthesis. During photosynthesis, when light energy is in excess, reactive oxygen species (ROS) is generated in various steps and thus chloroplasts can become damaged. Therefore the quality control of chloroplasts is important for the maintenance of photosynthetic activity and cellular homeostasis, and plant yield. Recently it has been reported that whole chloroplasts damaged by excessive light energy are transported to and degraded in vacuoles (Izumi et al. 2017). The atg mutants are hypersensitive to UV-B irradiation compared to wild-type plants and highly accumulates damaged chloroplasts in cytoplasm. How damaged chloroplasts are enwrapped by autophagosomes and are transported to vacuolar lumens for degradation remains to be elucidated.

Plants contain specific type of peroxisomes called leaf peroxisome in photosynthetic tissues. Leaf peroxisomes are involved in photorespiration together with chloroplasts and mitochondria, and produce hydrogen peroxide (H₂O₂) in this metabolism, as a result, peroxisomal proteins invariably damaged by H₂O₂ in it. Therefore, quality control of leaf peroxisomes is important for maintaining cellular homeostasis in plants. The atg mutants highly accumulate abnormal peroxisomes only in photosynthetic tissues (Yoshimoto et al. 2014). The abnormality can be detected in electron microscopic pictures as electron-dense regions, which probably correspond to protein aggregates inside peroxisomes. Interestingly, IMs were always associated with these electron-dense regions, suggesting that autophagy specifically targets and selectively degrades damaged peroxisomes for quality control. The molecular mechanism of how damaged peroxisomes are selectively recognized by IMs is still unknown. However, preliminary data suggest that plant pexophagy is NBR1-independent, unlike in mammals (Yoshimoto et al. 2014). Additionally, Atg30 and Atg36, which are involved in yeast pexophagy, are only found in yeast species (Farré et al. 2008, Motley et al. 2012). Therefore, plant-specific mechanisms ought to exist for the selective degradation of peroxisomes.

Artificially induced ER-stress by tunicamycin treatment induces autophagy, whereby the ER is transported to the vacuolar lumens for degradation (Liu et al. 2012). This process requires inositol-requiring enzyme-1b (IRE1b), which is a sensor for ER stress, but not bZIP60, which is a target of IRE1b in the ER stress signal transduction pathway. Autophagy is considered to be involved in the quality control of the ER by degrading abnormal ER that contains unfolded and denatured proteins. However, it is unclear whether such reticulophagy occurs under natural growth conditions or not.

Work on the molecular analysis of Arabidopsis ATG11 suggested that mitophagy, i.e. selective mitochondria degradation by autophagy, is induced in darkened leaves (Li et al. 2014). Although Atg11 is involved in selective autophagy in yeast, further confirmation of whether Arabidopsis ATG11 has the same function as in yeast is needed. Furthermore, the identification and characterization of a receptor/adaptor for plant mitophagy will further expand our molecular knowledge of selective autophagy in plants.

Autophagosome Fusion to Vacuoles

One of the important processes in the autophagy pathway is autophagosome fusion to vacuolar membranes for the transport of cytoplasmic constituents into vacuolar lumens for degradation. Information about this process is accumulating, especially in mammals and yeast, although elucidation of the molecular details is still in progress. Previous studies have revealed that autophagosome fusion to vacuoles requires soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs), Rab GTPases, and the homotypic vacuole fusion and protein sorting (HOPS) tethering complex, similar to other fusion events for intracellular vesicle trafficking (Reggiori and Ungermann 2017). In yeast, three Q-SNAREs, Vam3, Vti1, and Vam7, and the R-SNARE Ykt6 are involved in the fusion. However, yeast autophagosomal SNARE has not yet been identified. In mammals, syntaxin17 (Stx17), which interacts with SNAP-29 and the endosomal/lysosomal SNARE VAMP8, has been identified as an autophagosomal SNARE required for fusion with the lysosomes (Itakura et al. 2012). Stx17 localizes to the outer membrane of fully formed autophagosomes but not to the IM intermediates, thus only fully formed autophagosomes can fuse with lysosomes. Endosomal Rab7/Ypt7, which controls transport to late endosomes and lysosomes, is known to associate with autophagosomes and is involved in their fusion with lysosomes/vacuoles in mammals and yeast (Reggiori and Ungermann 2017). Recent work further indicated that the Mon1-Ccz1 complex, a guanine nucleotide exchange factor (GEF) for Rab GTPase Rab7, directly binds to LC3, a mammalian homolog of ATG8, via a LC3 interacting region (LIR) for its activation and is involved in autophagosome fusion to lysosomes (Gao et al. 2018).

In plants, the involvement of SNAREs for autophagosome fusion to vacuoles has also been reported. In Arabidopsis, the VTI1 SNARE family consists of three closely related members, VTI11, VTI12, and VTI13. Of these members, VTI12 seems to be involved in autophagy, probably at the fusion step of autophagosomes with vacuoles (Surpin et al. 2003). The VTI12-disrupted mutant, vti12, showed an early senescence phenotype under nutrient starvation conditions, which is reminiscent of autophagy-defective mutants. In addition, the mutant accumulates many small vesicles that are bubble-like in appearance in the cytoplasm, which are not observed in wild-type plants. These unique multivesicular structures seem to be autophagosome-related structures. As in yeast, there is no apparent homolog of autophagosomal SNARE Stx17 in plants, although plant autophagosome-specific SNARE proteins are predicted to exist for the fusion to vacuoles.

Unlike Rab7/Ypt7 in mammal and yeast, plant RAB7 may not be involved in the fusion step. Arabidopsis possesses eight RAB7 homologs in the genome, which are further divided into three subgroups (RABG1, G2, and G3a-f) (Pereira-Leal and Seabra 2001). Of these, the RABG3 subgroup is shown to be involved in the autophagic process (Kwon et al. 2010). Autophagy is induced during tracheary element differentiation; however, this process was significantly stimulated by the overexpression of RABG3bCA, a constitutively active mutant, and inhibited by overexpression of RABG3bDN, a dominant negative mutant and RABG3b RNAi. This suggests that RABG3b functions in autophagy induction or autophagosome formation in plants, especially in the vascular system. Recent studies indicated that plant homologs of the Mon1-Ccz1 complex activate RAB7, leading to vacuolar trafficking (Cui et al. 2014, Ebine et al. 2014). Therefore, it is possible that RAB7 homologs and its effector (GEF) function in the fusion of autophagosomes with vacuoles in plants. In mammals, another effector of RAB7, the FYVE and coiled-coil (CC) domain-containing protein (FYCO1), has been shown to interact with ATG8 and PI3P on autophagosomes and microtubules (MTs), to promote MT plus end-directed transport of autophagosomes (Pankiv et al. 2010). FYCO1 homologs have been identified in Arabidopsis, but it is still unknown whether the homologs are involved in autophagosome transport in plants. Given that organelle movements in plant cells depend on an actin-myosin system, in contrast to a microtubule system in mammals, the plant FYCO1 homolog may not be involved in autophagosome movement.

In addition to the above proteins, core ESCRT machinery and ESCRT-associated proteins also seem to have important roles in autophagosome fusion to vacuoles in plants. Defects in ESCRT-I-interacting protein FYVE1/FREE1 and the ESCRT-III-associated de-ubiquitylating enzyme, AMSH3, resulted in the accumulation of autophagosomes in the cytoplasm, suggesting that these ESCRT related proteins are involved in the fusion step (Isono et al. 2010, Gao et al. 2015). However, the accumulation of autophagosomes may be due to a lack of central vacuoles in these mutants. Mutants of an ESCRT-III-associated de-ubiquitylating enzyme, amsh1, and dominant-negative mutants of ESCRT-III-core protein, vps2.1, accumulated fewer autophagic bodies inside their vacuolar lumens, suggesting that these proteins partially participate in the fusion process or in autophagosome formation (Katsiarimpa et al. 2013). Further investigation will be needed to elucidate the molecular interactions of ESCRT related proteins in this process.

Concluding Remarks

At present, through the use of reverse genetic analyses of atg mutants, the physiological role of plant autophagy is becoming ever clearer. Many processes are thought to be unique to plants because of their autotrophic nature, and thus attract attention. Elucidation of the detailed molecular mechanisms of plant autophagy will prove whether autophagy functions directly in response to each type of environmental stress and/or differentiation/developmental condition. In particular, the characterization of specific cargoes depending on conditions and the elucidation of those selective mechanism(s) are important aspects for future investigation. Additionally, precise knowledge of when and in which tissues, and under which conditions autophagy is induced is important to understand the physiological functions of autophagy in plants. However, it is still hard to quantify autophagic activity in plants unlike in yeast. Therefore the establishment of a novel quantification method such as alkaline phosphatase (ALP) assays in yeast would help expand our knowledge of plant autophagy in future studies.

Acknowledgments

This work was supported in part by Grants-in-aid for Scientific Research 16H07255 (to K.Y.), 23000015 (to Y.O.), and 16H06375 (to Y.O.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Disclosures

The authors have no conflicts of interest to declare.

Footnotes

Subject areas: (2) environmental and stress responses, (7) membrane and transport

Abbreviations

Abbreviations
 
• AIM

  ATG8-interacting motif

 
• ATI

  ATG8-interacting protein

 
• ATG

  autophagy-related

 
• IM

  isolation membrane

 
• LIR

  LC3-interacting region

 
• NBR1

  neighbor of Brca1 gene

 
• PE

  phosphatidylethanolamine

 
• PI3K

  phosphatidylinositol 3-kinase

 
• PI3P

  phosphatidylinositol 3-phosphate

 
• SA

  salicylic acid

 
• SNAREs

  soluble N-ethylmaleimide-sensitive factor attachment protein receptors

 
• TOR

  target of rapamycin

 
• TSPO

  a tryptophan-rich sensory protein