Plants, as sessile organisms, persistently encounter diverse biotic and abiotic stresses that can adversely impact on their growth and thus severely impair crop production. One of the main objectives in agriculture is therefore to obtain high-yielding plants, resilient not only to diseases but also to increasingly higher temperature (Kim et al., 2021). From basic research in the model plant
At a cellular-molecular level, plants have evolved a plethora of mechanisms to be able to respond quickly and accurately to their ever-changing, often stressful environment. The plasma membrane (PM), acting as an interface between the extracellular surroundings and cellular constituents, is densely packed with an array of proteins involved in the sensing and transmitting of stimuli essential for adaptive responses. The transport and perception of, for example, plant hormones, which are indispensable for plant development, are centered around PM-localized proteins. Thus, their localization and abundance need to be tightly regulated and understanding this regulation is of exceptional interest. It is therefore not surprising that several regulatory pathways participate in the control of PM proteins, underlining the key functions for the spatiotemporal control of PM protein turnover in plant development and various adaptive growth responses (Korbei and Luschnig, 2013; Luschnig and Vert, 2014).
In eukaryotes, protein homeostasis at membranes is modulated by the endomembrane network. This intricate system of internal membranes provides spatial organization for cell activities by compartmentalizing processes and functions in transporting proteins to their site of action. Vesicular transport and membrane trafficking are pivotal for higher plants, as they not only maintain the functions of organelles and cells, but are also required for fast and dynamic reactions to fluctuating inputs and stimuli (Inada and Ueda, 2014). The plant endosomal system is responsible for the transport of endocytic and biosynthetic cargo, and therefore contributes to the regulation of the protein composition of the PM, the trans-Golgi network (TGN), which in plants coincides with early endosomes (EEs) (Dettmer et al., 2006), lytic and protein storage vacuoles, as well as the cell wall. At the TGN/EE, two major trafficking pathways meet: the endocytic pathway, which represents a set of trafficking routes with cargo sorting, recycling, and degradative functions, and the secretory pathway, which exports proteins from the ER via the Golgi apparatus to the TGN/EE (Robinson et al., 2008; Viotti et al., 2010; Paez Valencia et al., 2016).
Through endocytosis, which can be either dependent on the scaffolding protein clathrin (clathrin-mediated endocytosis [CME]) or independent of it, cells internalize PM-localized proteins, lipids, and extracellular material (Robinson, 2015; Sandvig et al., 2018). CME was first demonstrated in plants only quite recently (Dhonukshe et al., 2007), and extensive descriptions of the events that encompass CME are often times still based on elaborate studies in animal and yeast systems (reviewed in McMahon and Boucrot, 2011). The clathrin coat self-assembles into units composed of three clathrin heavy chains and three clathrin light chains that oligomerize together to form three-legged structures called triskelia (Chen et al., 2011; Robinson, 2015).
PM proteins in plant cells are primarily internalized by CME, which can be divided into five successive steps: nucleation, cargo selection, clathrin coat assembly, membrane scission, and uncoating (McMahon and Boucrot, 2011). In plants, CME first manifests in the bending of the PM toward the cytoplasm into a clathrin-coated pit. As clathrin does not interact directly with the membranes or cargos, the formation of these clathrin-coated pits is initiated by the recruitment of adaptor and accessory proteins to the proximity of the membrane. These hetero-tetrameric adaptor protein complexes, like the adaptor protein 2 (AP-2) complex and the plant-specific TPLATE complex (Zhang et al., 2015; Reynolds et al., 2018), as well as monomeric adaptor proteins interact with the membrane lipids and sorting motifs in cargo proteins as well as clathrin and assist in guiding them from the cytoplasm to nucleation sites. Adaptors are thus key to the formation of transport vesicles and to selection of cargo for incorporation into vesicles (Cocucci et al., 2012; Arora and Van Damme, 2021).
The clathrin-coated pits eventually mature and bud off to form clathrin-coated vesicles (CCVs). After budding from the PM, the clathrin coat is shed, allowing for recycling of its monomeric constituents in further rounds of endocytosis. In mammals and yeast, this happens directly following scission from the membrane, whereas the coat of plant CCVs appears to be retained and the components are only gradually discarded on route to the TGN/EEs (Narasimhan et al., 2020). The removal of the clathrin coat frees the vesicles for fusion with TGN/EE, the site where the decision for further sorting, either recycling back to the PM or to the vacuole for degradation, takes place (Chen et al., 2011; Reynolds et al., 2018; Rodriguez-Furlan et al., 2019).
To actively regulate the abundance of proteins at the PM, they need to be recognized by components of the CME machinery, specifically by adaptors. These distinguish the proteins to be internalized by motifs found in the cytosolic domains of such prospective cargo. Sorting motifs can be present in the protein sequence (intrinsic motifs) or can occur through covalently linked specific post-translational modifications like phosphorylation or ubiquitination. Importantly, several distinct sorting signals in the same cargo protein can participate in cargo recognition, adding some flexibility to subsequent sorting events (Traub, 2009; Traub and Bonifacino, 2013; Arora and Van Damme, 2021).
Linear sorting motifs found in the cytosolic domain of the proteins, like the di-leucine- and the tyrosine (YXXΦ)-based sorting motif (where Y is tyrosine, X is any amino acid, and Φ is a bulky hydrophobic residue), have also been identified in plant PM proteins (Arora and Van Damme, 2021). Nevertheless, only the tyrosine motifs have so far been linked to endocytic sorting from the PM (Geldner and Robatzek, 2008), where they were found to be important for the internalization of several PM proteins like the pin-formed (PIN) auxin efflux facilitators (Kleine-Vehn et al., 2011; Sancho-Andres et al., 2016), the brassinosteroid hormone receptor brassinosteroid insensitive 1 (BRI1) (Liu et al., 2020), the tomato pathogen-related receptor-like protein
The best documented post-translational modification for endocytosis of PM proteins in plants is ubiquitination. This reversible protein modification serves as a signal triggering endocytosis and consequent sorting for degradation, and thus plays a key role in directing the entry of PM proteins into the endosomal system (Clague et al., 2012; Piper et al., 2014; Dubeaux and Vert, 2017). Protein ubiquitination is catalyzed by a series of highly conserved enzymes: ubiquitin activating enzymes (E1), ubiquitin conjugating enzymes (E2), and ubiquitin ligases (E3), resulting in the formation of an isopeptide bond between the C-terminus of ubiquitin and the free amine of a lysine residue on the target protein (Vierstra, 2012; Callis, 2014; Piper et al., 2014). Ubiquitin conjugation is extraordinarily complex in plants, exemplified by more than 1500 E3 ubiquitin ligases, found in the model plant
Target proteins can be monoubiquitinated or poly-monoubiquitinated, where individual ubiquitin molecules are conjugated to cargo proteins or additional ubiquitin molecules are linked to an already attached ubiquitin to form chains. Different types of such polyubiquitin chains can form, depending on which lysine residues of the ubiquitin molecule are conjugated (Husnjak and Dikic, 2012). Two common forms are K48- and K63-linked ubiquitin chains, where the E2/E3 pairing determines substrate specificity and the type of chain linkage, although the same E3 may form different linkage types together with different E2s (Tomanov et al., 2014; Stewart et al., 2016). Ubiquitin binding domain (UBD)-containing proteins, which noncovalently associate with ubiquitin, are responsible for the decoding and translation of the ubiquitination code (Husnjak and Dikic, 2012). These ubiquitin receptors typically contain short amino acid stretches, lacking a strict consensus sequence, that display a low binding affinity toward ubiquitin (Husnjak and Dikic, 2012). Importantly, protein ubiquitination is a reversible process with de-ubiquitinating enzymes remodeling the ubiquitin code, contributing to the highly dynamic and reversible characteristics of endocytic trafficking decisions (Isono and Nagel, 2014). Ubiquitination of PM proteins, especially the addition of K63-linked polyubiquitin chains, serves as a signal for endocytosis and further sorting to the vacuole (Husnjak and Dikic, 2012; Romero-Barrios and Vert, 2018). This was shown for several different PM proteins like the iron transporter IRT1 (Dubeaux et al., 2018), the abscisic acid (ABA) receptors PYL4 and PYR1 (Bueso et al., 2014), and the auxin efflux facilitator PIN2 (Leitner et al., 2012b) as well as the brassinosteroid receptor BRI1 (Di Rubbo et al., 2013; Martins et al., 2015). Endocytic vesicles co-localize with K63-linked polyubiquitinated cargo (Johnson and Vert, 2016), and the molecular mechanisms driving endocytosis of such cargos are essentially conserved in plants. Nevertheless, some key players involved in this process appear plant-specific (Dubeaux and Vert, 2017; Gao et al., 2017; Schwihla and Korbei, 2020).
After internalization from the cell surface, endocytosed vesicles deliver their content via fusion to the TGN/EE and the cargo proteins can either be recycled back to the PM or be further sorted to multi-vesicular bodies (MVBs) and on to the vacuole for degradation (Paez Valencia et al., 2016). The recycling machinery includes small GTPases and their regulators as well as the retromer complex, while the Endosomal Sorting Complex Required for Transport (ESCRT) machinery is in charge of sorting PM proteins to their degradation (Buono et al., 2017; Isono and Kalinowska, 2017; Rodriguez-Furlan et al., 2019). PM proteins can be recycled back to the PM by being actively diverted from the default vacuolar degradation pathway at the TGN/EE. Plants do not seem to have endosomes dedicated specifically to a recycling pathway, but this rather involves components of the TGN/EE and potentially early stages of the MVBs (Robinson and Neuhaus, 2016).
Proteins destined for degradation are not recycled back to the PM, but internalized into the intraluminal vesicles (ILVs) of the MVBs (Otegui, 2018). The signal that marks PM proteins for degradation is ubiquitination, and the ESCRT machinery is thought to exclusively capture ubiquitinated membrane cargoes and sort them into ILVs of MVBs (Shields and Piper, 2011; Dubeaux and Vert, 2017). MVBs originate from the TGN/EE in a process that requires the ESCRT machinery for ILV formation and annexins for releasing MVBs to fuse with the vacuole, whereupon ILVs and their cargo is degraded (Paez Valencia et al., 2016; Isono and Kalinowska, 2017; Cui et al., 2018; Figure 1).
The ESCRT machinery is a conserved, multi-subunit membrane remodeling complex. It functions in the formation of membrane invaginations, which bud away from the cytoplasm, followed by scission. This stepwise process, performed by protein complexes termed ESCRT-0 to ESCRT-III and accessory components, acts in the recognition, concentration, and sequestering of cargo into the ILVs of MVBs and membrane-deforming events (Henne et al., 2011; Paez Valencia et al., 2016; Gao et al., 2017; Isono and Kalinowska, 2017).
The ESCRT-0 complex is required for the initial targeting and concentration of ubiquitinated cargo and further recruits ESCRT-I, to which it passes on the cargo (Hurley, 2010). The ESCRT-I then recruits the ESCRT-II to the endosomes and the presence of both complexes induces invagination of the limiting membrane toward the endosomal lumen (Hurley and Hanson, 2010). ESCRT-I and -II complexes might also act in parallel to cluster ubiquitinated cargo for internalization (Hurley, 2008; Hurley and Ren, 2009). ESCRT-II then recruits and possibly activates the ESCRT-III complex, which consists of small soluble subunits that assemble into higher-ordered multimers on endosomal membranes (Teo et al., 2004). Once the ESCRTs are assembled, the ubiquitin molecule is removed from the cargo proteins by deubiquitinating enzymes, while an AAA ATPase recycles ESCRT-III back into its monomeric form (Babst et al., 1998). The recruitment and activity of the AAA ATPase is mediated by ESCRT-III–related proteins (Gatta and Carlton, 2019).
In yeast/mammals, the ESCRT-0 complex is made up of two subunits and found at the EEs and the PM (Raiborg and Stenmark, 2009; Henne et al., 2011), where it appears to preassemble with cargoes to enhance the sorting efficiency without affecting vesicle formation
The
The ESCRT-I complex, which in yeast/mammals forms a hetero-tetrameric complex containing one copy of each of the subunits (Schuh and Audhya, 2014), is responsible for binding ubiquitinated cargo and ESCRT-0 subunits (Bache et al., 2003).
The ESCRT-II, which has a pivotal role in linking the upstream ubiquitin-binding ESCRT complexes to the downstream ESCRT-III complex, is a Y-shaped heteromer of the subunits Vps22, Vps25, and Vps36 (Hurley, 2010). In
ESCRT-III does not contain any known UBDs, and thus is unlikely to recognize ubiquitinated cargoes. Instead, ESCRT-III is essential for membrane scission during sorting of the endocytic cargo into ILVs (Henne et al., 2012). In contrast to the models proposed for animal systems, plant ILVs were shown to form networks of concatenated vesicles instead of individual vesicles in the lumen of the MVB. These networks remain connected by narrow bridges, thus trapping the cargo. The potential role of the ESCRT-III is to act as a diffusion barrier to prevent escape of the cargos destined for degradation (Buono et al., 2017). Consistent with this key function, overexpression of dominant negative alleles of ESCRT-III core subunits results in defects in MVB biogenesis and vacuolar degradation of PM proteins (Cai et al., 2014). Ultimately VPS4/SKD1, an ortholog of yeast/mammalian AAA ATPase, hydrolyzes ATP, leading to the disassembly of the ESCRT-III complex, allowing for recycling of the subunits. This is an essential step for the completion of the membrane scission and the formation of ILVs and, furthermore, for the MVBs to fuse with the vacuole/lysosomes (Hurley, 2010; Schuh and Audhya, 2014). All core subunits of the conserved ESCRT-III machinery have two homologs in
A further protein interacting with the ESCRT-III is ALIX, which also interacts with ESCRT-I subunits, therefore having the potential to link ESCRT-I and ESCRT-III (Bissig and Gruenberg, 2014). ALIX binds ubiquitin, specifically K63-linked ubiquitin chains (Dowlatshahi et al., 2012), suggesting it may function as a ubiquitin receptor for protein sorting into MVBs (Pashkova et al., 2013). Similar functions have been attributed to
Plant genomes encode unique ESCRT components with limited sequence similarities to mammalian/yeast subunits, which interact with known ESCRT subunits to regulate endocytic sorting processes. The SH3 DOMAIN-CONTAINING PROTEIN 2 (SH3P2; Figure 1) has been implicated in vacuolar trafficking of ubiquitinated cargoes as it localizes to CCVs and binds K63-linked ubiquitin chains, VPS23A, and deubiquitinating enzymes (Kolb et al., 2015; Nagel et al., 2017). FYVE 1 (Fab 1, YOTB, Vac 1, and EEA1)/FREE1 (FYVE DOMAIN PROTEIN REQUIRED FOR ENDOSOMAL SORTING 1) localizes to MVBs and binds ubiquitin, interacts with SH3P2, both VPS23 subunits, and is incorporated into the ESCRT-III complex (Barberon et al., 2014; Gao et al., 2014; Kolb et al., 2015; Belda-Palazon et al., 2016; Figure 1). It is essential for ILV formation, vacuolar biogenesis, autophagic degradation, seedling development, and localization and degradation of PM proteins and cytosolic ABA receptors (Barberon et al., 2014; Gao et al., 2014; Kolb et al., 2015; Belda-Palazon et al., 2016).
The high number of plant-specific factors, next to the conserved canonical trafficking machinery responsible for transporting PM proteins destined for degradation, reflects the need of plants to precisely adjust the abundance of their PM proteins in order to be able to respond quickly and accurately to their surroundings. Furthermore, some ESCRT subunits, as, for example, the TOL proteins, have undergone drastic gene expansions (Gao et al., 2017; Cui et al., 2018; Mosesso et al., 2019; Vietri et al., 2020). Thus, to fully understand plant adaptation mechanisms, deciphering the molecular mechanisms underlying the regulation of ESCRT-dependent degradation of ubiquitinated PM proteins is crucial.
While it is evident that the ESCRT machinery participates in virtually every aspect of plant growth and development, its contribution to mediating adaptive responses to environmental stimuli appears remarkable. Specifically, a variety of loci associated with ESCRT-mediated sorting have been linked to often times highly defined aspects of higher plants’ responses to biotic as well as abiotic stress conditions. Prominent examples are the ESCRT-I subunits, where a double mutant line (
Among the abiotic and biotic stressors, drought conditions receive particular attention as one of the most important factors that limit crop production on a global scale (Gupta et al., 2020). Drought stress restricts many aspects of plant growth and development from plant height to root morphogenesis (Gray and Brady, 2016). Thus, water represents a limiting factor for crop production in agriculture, aggravated further by current signs of climate change, which represents a serious threat for a sustainable supply with staple food (Nuccio et al., 2018; Osmolovskaya et al., 2018; Bertolino et al., 2019). Plants in their natural environment have evolved strategies to cope with limited water supply, which is reflected in a spectrum of mechanisms resulting in an increased tolerance to drought. A plant's approach to cope with such limitations involves either drought escape, which allows plants to complete their life cycle before the onset of drought, drought avoidance, antagonizing enhanced water loss, or drought tolerance characterized by (oftentimes long-term) osmotic adjustments (Zhang, 2007; Osmolovskaya et al., 2018; VanWallendael et al., 2019). However, water shortage beyond a critical threshold level will inevitably result in irreversible tissue damage together with major effects on photosynthesis, respiration, and nutrient uptake (Zhu, 2002; Osmolovskaya et al., 2018). These effects are linked to drought-related aberrations in stomata function, interfering with gas exchange, thus leading to dramatic consequences for crop plants (Zhu, 2002; Bertolino et al., 2019).
Plants carefully protect their cells from dehydration, coordinated by the phytohormone ABA, which rapidly accumulates in plants in response to drought/dehydration stress. This is accomplished by an array of ABA-mediated responses, such as stomatal closure to limit water loss and the production of protective metabolites (Yamaguchi-Shinozaki and Shinozaki, 2006; Gomez-Cadenas et al., 2015). Apart from that, ABA influences adaptive growth responses, including the control of seed germination as well as overall plant organ growth (Cutler et al., 2010; Finkelstein, 2013; Nakashima and Yamaguchi-Shinozaki, 2013). These latter ABA responses impact on plant morphogenesis as a result of an intimate interplay with responses triggered by additional growth regulators. Auxin, for example, is indispensable for root morphogenesis and functions as a key regulator of root system architecture (Lavenus et al., 2013). Ongoing research, unraveling the mechanisms of ABA–auxin crosstalk in the regulation of root growth, thus produces essential insights into the interplay between stress responses and morphogenetic signaling events and how they jointly control adaptive responses (Xie et al., 2021).
Members of the TOL protein family function in the first steps in the ESCRT-mediated degradation of ubiquitinated PM-associated proteins (Korbei et al., 2013; Moulinier-Anzola et al., 2020). Nevertheless, the precise function of the individual TOLs in this enlarged protein family is not yet described. There are apparent functional redundancies within members of the TOL family, as
PIN2, an intrinsic PM protein required for directional cellular efflux of the phytohormone auxin, represents an excellent example for such regulation (Gallei et al., 2020; Konstantinova et al., 2021). This protein is modified by K63-linked ubiquitin chains in dependence of RING DOMAIN LIGASE (RGLG)-type E3 ubiquitin ligases, as such ubiquitination acts as a principal signal for PIN2 endocytosis (Leitner et al., 2012a, 2012b). This is underlined by the expression analyses of mutant
As higher-order
Next to the regulation of auxin homeostasis (Korbei et al., 2013), TOLs could therefore also be relevant for the perception and/or downstream interpretation of signals triggered by ABA, potentially via affecting the half-life of proteins involved in the control of ABA signaling and/or homeostasis (Korbei et al., 2013; Moulinier-Anzola et al., 2020). Thus, it will be essential to decipher if TOL proteins function in modulating the ABA pathway in plants to elucidate, in addition to their general role in the endosomal degradation pathway, if and how TOLs play a more differentiated role in ABA signaling.
Overall plant performance can be considered as a product of signaling pathways defining plant morphogenesis and stress tolerance under favorable and comparably adverse growth conditions (Nuccio et al., 2018; VanWallendael et al., 2019). TOL substrate specificity and activity in the control of protein turnover positions members of this protein family at the intersection of these antagonistically acting growth determinants. Understanding TOL function in further detail is thus likely to result in important insights into the mechanisms by which plants define the trade-off between stress tolerance and plant performance under highly volatile environmental conditions (Maggio et al., 2018).
In summary, these observations led to a working hypothesis in which elevated tolerance to abiotic stressors such as drought may be due to increased ABA receptor activity caused by mis-functioning members of the ESCRT machinery like potentially also the TOLs. On the other hand, root development, a process coordinated by auxin, also affects plant tolerance to abiotic stresses, including drought (Uga et al., 2013). Thus, in a
Understanding the genes responsible for modulating the responses to phytohormones, as key mediators of plant responses to drought stress, could be a great approach to maintain as well as improve the productivity of crop plants. Enhanced drought tolerance without affecting further phenotypic traits represents a highly desirable agronomic trait (Waltz, 2014). Nevertheless, such plant lines have to be subjected to extensive phenotypic analysis. Basic growth parameters, including plant and organ shape/size, root architecture, and flowering time, have to be assessed carefully, as well as the plant responses to further environmental parameters, including light quantity/quality, variable temperatures, and additional abiotic stresses as well as responses to selected pathogens. These experiments should reveal potential trade-offs in overall plant development as a consequence of the enhanced drought tolerance (Nuccio et al., 2018; Vaidya et al., 2019).
In the light of recent climate extremes, it appears imperative to generate novel crop varieties that are well adapted to adverse environmental conditions and devoid of undesirable side effects affecting overall plant fitness and/or crop yield. Green biotechnology tried to address this issue, aiming at the generation of less drought-responsive crop varieties, by exploiting a range of different approaches. Different from labor-intense and time-consuming standard breeding strategies, solutions offered by biotechnology most often involve the generation of “conventional” genetically modified organisms (GMOs), utilizing the overexpression or silencing of determinants of drought responses in plants. An alternative to these rather blunt approaches is offered by a combination of knowledge obtained in the model plant