Structural flexibility and shape similarity contribute to exclusive functions of certain ATG8 isoforms in the autophagy process

doi:10.21203/rs.3.rs-4831294/v1

Download PDF

Research Article

Structural flexibility and shape similarity contribute to exclusive functions of certain ATG8 isoforms in the autophagy process

https://doi.org/10.21203/rs.3.rs-4831294/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

One of them relates to the curious interactions between certain ATG8 isoforms and its receptors/adaptors in plants during the autophagy process. These proteins ultimately regulate the initiation, expansion and maturation of the phagophore, which acts as a docking platform for numerous autophagy proteins. The bioinformatician analysis of the human, yeast and plant amino acid sequences allowed us to split them into groups and determine the head-liners taking the experimental data into account. We conducted a structure-based study aimed to identify some hidden reasons for the differences in selectivity of ATG8 isoforms. A series of molecular dynamics simulations were run in attempt to explain the ATG8s’ stage-dependent functionality. The canonic secondary structure organization and folding of all ATG8 proteins and therefore almost identical PPI interface areas make this investigation particularly important and interesting. Aware of the dual role of the ATG8 interacting motif (W/F/Y-X-X-X-L/I/V) in autophagosome biogenesis and the recruitment of an anchored selective autophagy receptor (SAR), we conducted a mobility domain analysis. To this end, sequence of amino acids associated with the LDS interface was localized and underwent an RMSD-based clustering analysis. The resulting ATG8-peptide complexes reflected significant binding preferences of different ATG8 isotypes.

Autophagy

ATG8

adaptors

molecular dynamics

protein-protein docking

Starting from the Atg8 focused protein-peptide docking, we tried to identify conformational changes that occur in the contact region of the corresponding adaptors and receptors involved in the overall biogenesis events in autophagy.

Eukaryotic cells are continuously exposed to external stressors from the environment throughout their lifetime. This demands a highly controlled adaptive cellular response, in which cells undergo rapid changes to cope with harsh conditions. By adapting their metabolism and protecting themselves against potential damage, cells can maintain homeostasis and function, also in the presence of environmental challenges. However, if not managed carefully and correctly, the consequences can be detrimental. Autophagy is one such adaptive cellular process, which functions at basal levels and is highly induced in response to various types of stress. A number of autophagy-related (ATG) proteins are directly involved in this process as core components of the cellular autophagic machinery. The function of the ATG proteins, in particular their ability to interact with a number of macroautophagic regulators, is modulated by posttranslational modifications (PTMs) such as phosphorylation, glycosylation, ubiquitination, acetylation, lipidation, and proteolysis (Y. Xie et al., 2015).

ATG proteins are grouped into 1) a Class III phosphatidylinositol-3-kinase (PI3K) complex functioning in vesicle nucleation, 2) a serine-threonine kinase complex involved in induction of autophagy, and 3) ubiquitin-like protein conjugating systems ATG12 and ATG8 that promote maturation of vesicles (Bu, Yang, Guo, Huang, & Chen, 2020; Cao, Li, Yang, & Cao, 2021). Microtubules are polymers of tubulin dimers whose dynamics are regulated by microtubule-associated proteins. They constantly polymerize and depolymerize to facilitate trafficking of organelles along microtubular tracks and chromosomal segregation in mitosis (R. Xie, Nguyen, McKeehan, & Liu, 2010; Galjart, 2010). After assembly, microtubules are constantly modified in different patterns to enhance their functions. One type of modification is acetylation that results in acetylated microtubules that recruit molecular motors enabling increased flux of vesicles along microtubular tracks (R. Xie et al., 2010; Dompierre et al., 2007). All these post-translational modifications are of a great interest for both animal and plant organisms (A. V. Rayevsky, Sharifi, Samofalova, Karpov, & Blume, 2019; Samofalova, Karpov, Raevsky, & Blume, 2019)

Central to autophagy is the ubiquitin-fold protein ATG8, which could be processed by the cysteine protease ATG4 (Kirisako et al., 2000), which leads to the exposure of the c-terminal glycine of ATG8 (Kabeya et al., 2004), allowing its recruitment to the phagophore through the attachment of the phospholipid phosphatidylethanolamine (PE). Atg8-family proteins have been extensively described for their implication in autophagosome formation and cargo selection in the cytoplasm. Although Atg8 family proteins also localize in the nucleus, their role in this compartment remains largely unexplored. Structurally ATG8s are ubiquitin-like proteins that share a similar structure consisting of two amino- terminal α-helices and a ubiquitin-like core. Atg8 proteins undergo a unique ubiquitin- like conjugation to phosphatidylethanolamine on the autophagic membrane, a process essential for autophagosome formation (Shpilka, Weidberg H Fau - Pietrokovski, Pietrokovski S Fau - Elazar, & Elazar, 2011). ATG8 is activated by the E1 enzyme ATG7, transferred to the E2 enzyme ATG3, and finally conjugated to PE using a ligase complex containing an ATG12-ATG5 conjugate and ATG16 (Ohsumi, 2001; Marshall & Vierstra, 2018). When inserted into autophagic membranes, lipidated ATG8 provides a docking platform for numerous autophagy adaptors and receptors that drive autophagic vesicle dynamics and cargo selection, respectively (Khaminets, Behl, & Dikic, 2016; Gatica, Lahiri, & Klionsky, 2018; Marshall & Vierstra, 2018). Subsequently, the lipidated ATG8 (ATG8–PE) proteins further recruit numbers of receptors and adaptors to regulate the initiation, expansion, and maturation of the phagophore, and, ultimately, fusion of the autophagosome with the vacuole (Martens & Fracchiolla, 2020; Sun et al., 2022). Atg8–PE thus provides a docking platform for autophagy adaptors to facilitate autophagosome formation, or for receptor recognition to selectively recruit the cargos to the nascent phagophore membrane, respectively.

In the model plant Arabidopsis thaliana, ATG8 proteins are encoded by 9 genes that have been grouped into 3 clusters based on amino acid sequence alignment (Wang, Nolan, Yin, & Bassham, 2020; Kellner, De la Concepcion, Maqbool, Kamoun, & Dagdas, 2017). The first cluster includes four members (ATG8A, ATG8C, ATG8D and ATG8F), the second one contains three members (ATG8B, ATG8E and ATG8G), and the third one – only two members (ATG8H and ATG8I). Analysis showed that the presence of ATG8h-i group members is not strictly associated with the total number of ATG8 homologs in each species (Wang et al., 2020; Kellner et al., 2017). Transcriptional regulation o plant ARG8 genes can affect the autophagy process. ATG8a, ATG8b and ATG8h demonstrated greater expression than ATG8d, ATG8e and ATGi, in turn the expression of other genes did not change significantly (Wang et al., 2020).

Previously we investigated structural rearrangements of microtubules which accompanies autophagy development, and touched on the preconditions for ATG8a interaction with α-tubulin molecules (A. Rayevsky et al., 2023; Rayevsky A, 2021). However, functional discrepancies between the individual mammalian ATG8 proteins are currently not fully understood. By combining all the data of the A. thaliana, we found that there also remains much uncertainty regarding the deep underlying structural changes responsible for some specific functions inherent to individual isotypes. From these, we used the content of the opensource databases to prepare and run a series of molecular dynamics simulations aimed to determine some difference between the stable conformations of the interface patterns or vice versa explain the observed isotype-specific features (functions, protein partners) by capturing the transition state in the core elements. Summarizing the results, we suppose the main role in selectivity of ATG8 belongs to its partners, which can recruit its structure in a different way depending on the flexibility degree. The successful reproduction of ATG8-adaptor binding also could become a starting point for implementation of competitive peptides and ability to estimate the probability to block certain ATG8 isotype, and arrest the related autophagy state, in silico.

Structure organization, functional patterns and protein partners of ATG8 proteins.

All necessary information concerning structure organization, functional patterns and protein partners of ATG8 proteins was gathered from the UniProt database and String network (Szklarczyk et al., 2019). Three-dimensional structures were taken from the AlphaFold depositary using UniProt codes as a search query (Varadi et al., 2022).

We used a BioEdit to perform and analyze a multiple alignment of sequence isotypes from human, yeast and plant species (Eargle, Wright, & Luthey-Schulten, 2006; Hall, 1999). A parameter set consisted of a Blosum62 comparison matrix together with a Zero (Zero End-gap Global Alignment) pairwise alignment method and absolute gap penalty of 2.4 (optimal for structural similarity recognition value). Then we analyzed all the plant sequences, divided them into groups by similarity, and calculated corresponding similarity degrees. Considering the information is critical for understanding the mechanism of ATG8 selectivity, we also determined a number of common functional areas based on the available data.

Molecular Dynamics

A molecular dynamics simulation was carried out in a Gromacs 2020 package (Páll, Abraham, Kutzner, Hess, & Lindahl, 2015). The CHARMM36 force field was chosen to calculate the dynamics of a protein molecule (Huang & MacKerell, 2013). A Nose-Hoover and Parrinello-Rahman coupling integration was applied to perform an accurate long-range simulation of 400 ns in an aqueous environment (the water molecules model TIP3P). The parameters of the algorithms (thermo- and barostat) for maintaining the temperature and pressure in the system were 310 K and 1 bar. We chose the radius of the Coulomb and van der Waals interactions equal to 1.2 nm. The Particle Mesh Ewald (PME) algorithm was used to calculate long electrostatic interactions between atoms and to maintain equilibrium in the system after each iteration (Essmann et al., 1995; Liu, Hu, & Zhao, 2011). The equilibration (settling) of the water molecules and counterions, compensating the charge of the protein molecule, was achieved by subsequent replacement of NVE with NVT ensemble.

We applied a RMSD- based (Root-mean squared deviation) clustering method with a standard set of parameters to process a conformational change of the binding site area. For thee we created a separate section in the index file to focus on the AIM surface. All clusters were ranged by the number of conformations per cluster to identify the largest one and then we estimated the distribution of these conformations across the MD simulation period. The largest cluster, covering the entire trajectory, corresponded to the most stable mode of the binding site and the centroid structure was utilized for further investigation. The RMSF (Root-mean squared fluctuation) plot was generated with a Gromacs built-in tools to analyze the motility of distinct secondary elements. The knowledge could be useful for interpretation of a sequence dependent difference in a binding site shape.

Docking

Protein-peptide docking for ATG8 and its adaptor/receptor partners was performed using Haddock 2.0 server in advanced mode (van Zundert et al., 2016). AIR constraints for docking were applied based on the extrapolated information from the crystal data of different orthologs (PDBID: 5LXH, 2L8J, 7AA7, 5D94).

The study was started from the investigation of all known and classified ATG8’s from Homo sapiens, Saccharomyces cerevisiae and Arabidopsis thaliana organisms. Based on known crystal structures for human and yeast species and different degree of homology to the plant proteins, a comparative alignment was generated. To date, three types of autophagy have been well described in plants, including macroautophagy, microautophagy, and mega-autophagy.

Ubiquitin interaction motif docking site sensing

The macroautophagy (hereafter referred to autophagy) is the most well-characterized one in plants. In plants, autophagy is generally activated upon stress and its regulation is executed by numbers of AuTophaGy-related genes (ATGs), of which the ATG8 plays a dual role in both biogenesis of autophagosomes and recruitment of ATG8-interacting motif (W/F/Y-X-X-L/I/V) with anchored selective autophagy receptors (SARs). In addition to the contribution to autophagosome formation mentioned above, ATG8 plays an additional key role in the selection of specific SARs to be sequestered prior to its degradation. In this process, ATG8 proteins bind to these SARs via their ATG8-interacting motifs. Such motif is either termed as AIM or ubiquitin-interacting motif (UIM), corresponding to the LC3-interacting region (LIR)/AIM docking site (LDS) or the UIM docking site (UDS) of ATG8, respectively. Through these motifs different ATG8 forms interact with various adaptor/receptor proteins to recruit specific cargos for degradation by selective autophagy (Fig. 1).

Figure 1. LIR-docking site - LDS is represented with two hydrophobic pockets responsible for the recognition of Trp and Leu in the AIM (WXXL) motif are labeled W-site and L-site, residues constituting these pockets are indicated with red and green arrows, respectively. Basic residues responsible for ionic interaction with acidic residues in AIM are enclosed with a black square. Potential acetylation sites (K49 and K51 in A. thaliana) are marked with asterisks, the UDS (ubiquitin-interacting motif-docking site) core sequence is countered with yellow braces, and the glycine site for PE conjugation is shown in the red box.

Autophagic adaptor/receptor proteins of choice

In humans, there are six Atg8 orthologs: LC3A, LC3B, LC3C (encoded by respective MAP1LC3, microtubule associated protein light chain 3, alpha, beta, and gamma genes); and GABARAP (γ-aminobutyric acid receptor-associated proteins), GABARAPL1 (GABARAP-like protein 1) and GABARAPL2 (also known as GATE-16) (Nguyen & Lazarou, 2022; Wesch, Kirkin, & Rogov, 2020; von Muhlinen et al., 2013; Wesch et al., 2020). Most of LC3/GABARAP functions are considered dependent on their post-translational modifications and addressing to membranes through a conjugation to a lipid, the phosphatidyl‐ ethanolamine (Leboutet et al., 2023). In yeast, there is single ATG8 protein (Varga, Keresztes, Sigmond, Vellai, & Kovács, 2022; Shpilka et al., 2011). In turn Arabidopsis’s ATG8 family consists of nine Atg8 genes, termed Atg8a to Atg8i. The proteins encoded by this family show between 41% and 99% identity with each other and from 47–73% identity with the yeast protein. The AtAtg8 proteins also show considerable identity (between 33% and 41% identity) with light-chain 3 (LC3) from the mammalian MAP1A and 1B; the Golgi membrane transport modulator protein Gate 16 (between 48% and 60%) and to GABARAP (42–59%) (Ketelaar, Voss, Dimmock, Thumm, & Hussey, 2004). Interestingly, that localization of ATG8 depends on nutrient conditions and could be localized in either endoplasmic reticulum (ER)-associated vesicles, or in the cytoplasm.

RMSD-based clusters

Canonical ATG8-Interacting Motif, (cAIM), also known as an LC3-Interacting Region (LIR), is a well-characterized short linear motif that interacts with ATG8 by forming a parallel β-sheet with β-sheet 2 in ATG8 (Birgisdottir Å, Lamark, & Johansen, 2013). The cAIM is represented by the WXXL consensus sequence, where W is an aromatic residue (W/F/Y), L is an aliphatic hydrophobic residue (L/I/V), and X can be any residue (Johansen & Lamark, 2020) (Fig. 2). We conducted a series of molecular dynamics simulations aimed to determine the most frequently observed geometry of the LDS. The cAIM contacts a hydrophobic patch on ATG8 known as the LIR/AIM docking site (LDS). The residues that compose HP1 and HP2 hydrophobic pockets (in ATG8a there are Q18, I22, R29, K47, K49, Y50, L51, V52, P53, L56, V64 and F78), and potential acetylation sites (K49 and K51) are individually colored (by residue name) and labeled in the Fig. 2. To analyze the mobility of the area we localized LDS interface-associated amino acids and performed an RMSD-based analysis. We applied a clustering analysis, focused on the area and found the most abundant separate clusters combining the major number of slices from the trajectory.

Figure 2. A surface-based representation of the LDS binding sites, derived from the series of MD simulations (isotypes Atg8a-d as A-D figures). Here we visualized the most frequent conformations of the LDS according to the conformational distribution across the correspondent trajectories and the number of frames in the top-ranked RMSD-based clusters (a bar-like diagram). UBA site is also shown from the side. The surface is mapped using a residual properties color scheme (grouped according to their side chains' pKa values and charges)

LDS ATG8a mapped with widest cluster plot

For example, the LDS of ATG8a is represented with the largest cluster of 1863 slices (the entire number of frames per trajectory was 2500), sharing similar disposition of the Ca atoms and side chains which form the c of the binding site. Another cluster is represented with only 263 instances and the LDS of both centroid structures from these two clusters are slightly different. In general, the total number of clusters per trajectory also serves as an evaluation of the binding site conformation stability. From these, the most stable binding sites shape were showed for ATG8a, ATG8c and ATG8f. In the case of ATG8d we also identified two close-enough clusters, which conformations are evenly distributed along the trajectory and diluted with excessive outlined conformations. The trajectory analysis revealed, that two cavities (HP1 and HP2) necessary for the interaction with known AIMs were well-outlined in ATG8a, ATG8b, ATG8d, ATG8h and ATG8i. In addition, these proteins are characterized with a common shape of the LDS on the side, forming something like a cleft. At the same time, ATG8c, ATG8e and ATG8g shared another common shape of the interface profile (Fig. 3).

In each case of the ATG8 family we chose only those clusters which contained more than 100 frames of the trajectory. We also used the information on the distribution of the conformations from the leader cluster within the frames from the molecular trajectory. In general, conformational clustering focused on the binding site area showed the existence of a singleton predominant cluster, covering all the trajectory. That means, that each centroid structure from the leader cluster reflects the overall conformation stability of the studied isotype area across the MD. From these, we used the output of each simulation to represent the amino acid composition and structure-depending surface of the binding site.

Clustering score

Form the figures it is evident, that despite the sequence similarity, the binding site can take different conformation and it, probably, can have impact on the features of protein-protein interactions. One of the most important findings is that the accessibility of specific amino acid residues (for example, HP1 and HP2, surface curvature) of the LDS site varies when comparing ATG8a-ATG8d isotypes (see Fig. 3). We analyzed secondary structure stability and the allocation of the loop elements during the simulation using a Root-mean square fluctuation (RMSF) plot to understand which part of ATG8 and from which isotype is less stable (Fig. 4) Upon reviewing the literature, it became apparent that proteins belonging to the homologous group ATG8 exhibit a preference for certain partners. As a result, we selected three model proteins to examine LDS/LIR interactions, categorized into three groups based on sequence alignment results. We chose ATG8a, ATG8e, and ATG8i due to the frequency of references in publications and the abundance of experimental data available for them (Yemets, Shadrina, Blume, Plokhovska, & Blume, 2024). On the contrary, the LIR motif may be provisionally categorized into four groups based on their functions and the time when they first appear at the beginning stages of autophagy. Based on the known formula of AIM sequence we searched for the fragments of the ATg8 partners. Therefore, knowing about the approved selectivity of diverse ATG8 isotypes towards certain protein partners, we attempted to evaluate the significance of a specific protein-protein interactions. From the partner functions we can suppose which ATG8 isotype can take part in the autophagy process and at which phase.


Figure 3. Another set of isotypes (Atg8e-g are at E, F and G figures; Atg8h and ATG8i are at H and I figures) are represented in a surface-based mode - the LDS binding sites derived from the series of MD simulations. The clustering result is reflected in a bar-like diagram, showing the less stable conformations for ATG8i isotype.

ATG8 pockets

Additionally, we utilized an established crystallographic data from human and protozoan homologs to accurately allocate the docked poses of AIM-containing peptides. These peptides have necessary AIM pattern (the motif binds canonically to the W- and L-sites of ATG8) and a few terminal residues for electrostatic interaction with the negatively charged region near the LDS site.

After identifying the most consistently present conformations of the binding site, some of which are illustrated in Figs. 2 and 3, we conducted a flexible molecular docking of the peptides to the rigid ATG8s site structures. As we had hoped, the docking results of different AIM variants corresponded to the experimental data. The table of values displays distinct preferences for interaction based on the docked poses formed by three model ATG8 isotypes and specific motifs (Fig. 4).

At the same time, dynamic processes within the ATG8 structure, which may modify selective binding, have been already highlighted in some publications, arguing it is superior to the docking. For this reason, we selected the best 12 complexes, based on their scoring function values and quality of the interaction with the hydrophobic pockets, and then relaxed them for an additional 150 ns. In several cases the simulations resulted in a thick inter-protein contact or in a strengthening of hydrogen bonds. However, in some cases we observed a complete dissociation of the molecules. We also noted some changes in the contour of the interaction interfaces between ATG8 and its partners (especially on the ATG8 side).

First group of the partners - SARs, that take part in the autophagy mechanisms of cargo linkage and can interact with ATG8 (Kirkin, 2020). Notably, NBR1 is responsible for the dispatch of such polyubiquitinated aggregates to nascent autophagosomes. Almost all ATG8 isotypes contain a set of residues (I22D, Y26E, R29E, K47E, V64D, R68E) that interact with a so-called ubiquitin-associated area (UBA) of NBR1 (660EWDPIL). One of the ATG8 pockets, namely W-site, is formed with a loop containing Tyr26 and Arg29 between helix-2 and strand-1. This loop can stick in a charge-dependent manner to the AIM site of NBR1. Interestingly, the NBR1 AIM motif (WDPI) is quite similar to the PexRD54 motif (WEIV) which drives the autophagosome formation by linking its vesicles to autophagic the compartment marked with the ATG8 protein (Pandey et al., 2021).

Figure 4. A cartoon-like representation of an average ATG8 protein with colored disordered regions and labeled secondary structure elements – coloring scheme is similar for 2D prediction and the cartoon model. Peptide-docking results for ATG8s and four representatives of SARs are represented in the table (A). RMSF-based visualization of different level of flexibility calculated for all known ATG8 isotypes from A. thaliana. Helices and b-sheets are labeled according to the structure assessment after the MD and delineated analyzing sequence alignments (B).

Analysis of trajectories

The stable trajectories of the NBR1 complexes with two of chosen ATG8 isotypes revealed, that major part of isotypes can bind to NBR1, which is active at different stages of autophagy. Amino acids W661/L665 of NBR1 occupy both hydrophobic pockets on the surface of ATG8a and ATG8e. Stable hydrogen bonds were found at the interfaces of both ATG8a (K47, Y50 to L665; Y50 to W661; E18, K25 to E660; R29 to P663; K47 to D662) and ATG8e (K48 to L665, I664; K69 to L665; Y61 to W661; E18, K26 to E660; V53 to P663; K50 to D662). The ATG8i complex also forms a hydrogen mesh, but only single anchor, a W661 amino acid, makes contacts with its hydrophobic pocket (Fig. 5).

Second group of proteins that interact with ATG8 are represented with a Bin/Amphiphysin/Rvs family, which members contain both BAR- and SSH3like domains (localized on the autophagosomal membrane). These proteins are involved in the promotion of the phagophore formation and its closure during the maturation of autophagosomes. The adaptor protein, known as SH3P2, plays a role in embedding of the phagophore in the membrane of plants, but it is not necessary for its movement through the endocytosis process.

From the analyzed trajectories it is evident that ATG8a and ATG8i almost completely lost its interactions with the SH3P2 fragment (321STGEYVVVVVRKV). While ATG8i and SH3P2 totally dissociate after the first half of the simulation period. As we determined, the interactions between V326/Y325 and the hydrophobic pockets diminished in the beginning of MD, but the hydrogen contacts (L51, V52 to Y325; K49, K68 to E324; K49 to S321) with ATG8a persisted across all the simulation. The strongest interaction with SH3P2 was determined for ATG8e isotype. Both the tight anchoring of V320/Y325 in the HP1 and HP2 pockets and a stable network of hydrogen contacts (L52, V53 - V326, V328; E19 - Y325; K50 - V326, S321) were observed. It is well known that ATG8e and ATG8f interacts with SH3P2 during the exposure to carbon simulation in in vitro tests (Zhuang et al., 2013). As these results were previously published, now we can suggest that our model is a reasonably accurate reflection of the available data. Additionally, it is these two, ATG8e and ATG8f, interact with ATI1 and ATI2 to contribute to endoplasmic reticulum degradation and form ATI bodies (Honig, Avin-Wittenberg, Ufaz, & Galili, 2012; Avin-Wittenberg, Michaeli, Honig, & Galili, 2012).

Molecular dynamic residue interactions

Of particular note, ATG8e is the only model that readily dissociates from the next peptide, the LIR motif of ATG7 (679FNLDWE). ATG7 interacts solely with the C-terminal domain, which is essential for ATP-mediated activation of the C-terminus of Gly116 Atg8 (serving as an E1 enzyme). To the extent that ATG7 is a component of the third autophagic group that mediates its lipidation with PE, this complex comprising of ATG1, ATG3, ATG4, and ATG7, ATG8 may directly regulate ATG8 in its lipidation with PE. These interactions could potentially induce PAS assembly and aid in the autophagic turnover, whereby ATG1a is transported into the vacuole alongside the ATG8-decorated autophagic cells. This transfers ATG8 from a catalytic cysteine in one protomer of Atg7 to Atg3, which is linked to the N-terminal domain in the N-terminal domain of the opposite protomer of the Atg7 complexed as a subunit of a homodimer.


Figure 5. The results of molecular dynamics simulations performed to predict which of ATG8s can interact with certain SARs are shown in a table-like view. The analysis of H-bond network and HP1/HP2 pockets occupation revealed the most stable complexes and those which were tend to dissociate after 100 ns of MD.

In the dynamics of the ATG7 complexes, we found that ATG8a and ATG8i maintain hydrophobic contacts between HP1/HP2 and W683/L681. Additionally, the ATG7 fragment forms tight hydrogen bonds with both ATG8a (L51 to L681; Y26 to E684; R29, K49 to E684) and ATG8i (L49 to L681; Y24 to E684; R27, K47 to E684).

To recap, the crucial interaction involves the connection between ATG8 and PI3P on autophagosomes, and the animal protein that binds to both objects is FYVE and coiled coil domain containing adaptor proteins 1 (FYCO1). FYCO1 belongs to the fourth group of adaptor proteins, and up to nine FYCO1 homologs have been identified in Arabidopsis at the moment (Wywial & Singh, 2010). Our study focuses on FREE1, which selectively interacts with ATG8 isotypes (Zeng et al., 2023). Being lipidated in the autophagosome membrane ATG8 proteins trigger the interaction of the autophagosome with phosphorylated FREE1, particularly through the action of KIN10, launching the process of autophagosome closure.

ATG8a and ATG8i keep both amino acids (I441/W438) of the FREE1 fragment (436GDWMNIIK) together in the HP1/HP2 pockets consistently. The hydrogen network between the amino acids of the fragment is found to be more stable in ATG8i (R27 - I441, N440; L49, R66 - N440; K45, K47, E16 - G436) compared to ATG8a (L51 - M439, I441; K49 - M439, G436; E18 - W438; K47 - D437).

One interesting suggestion is that tight and advantageous interaction with the cognate peptide can cause a change in the binding surface, resulting in clearer hydrophobic pockets and site topography. However, a complete or partial dissociation of peptide molecules from the surface of each of the three ATG8 isotypes causes the surface to return to its initial apo-form conformation. This, based on the computational modelling, could be one of the mechanisms of selectivity against ATG8’s targets.

We performed multiple alignments and analyzed the comparison of all the sequences from different ATG8s in plants and human. All the ATG8s from A. thaliana can be divided in three groups with high degree of similarity in each group and low between groups.

Transcriptional regulation o plant ARG8 genes can affect the autophagy process. ATG8a, ATG8b and ATG8h demonstrated greater expression than ATG8d, ATG8e and ATGi, in turn the expression of other genes did not change significantly (Wang et al., 2020). It is also shown, that ATG8b co-localization with LDs in the cotyledon cells isotype is essentially important for the A. thaliana seed germination (Miklaszewska et al., 2023). Importantly, ATG8's surface is relatively flat and lacks distinct features, except the trench where the complementary amino acid chain fits.

In general, in addition to the different level of ATG8 transcription activity on certain stages of cell cycle, there some structural determinants of isotype-specific functions. Our molecular dynamics simulations demonstrated distinct conformational stability of the LDS binding site for ATG8s from separate groups. Similarly, to the interaction of ATG8 with microtubules, the arginine/lysine-rich patch on the surface initiates the recognition of the partner protein. Our molecular docking procedure resulted in different values of scoring functions for these isotypes. After comparing the probabilities of formation of various ATG8/adapter combinations, we decided to determine the degree of binding strength of the complex elements.

In summary, the peptide binding causes an ATG8 interface changes trying to wrap the peptide with the highest affinity. Additionally, a crucial aspect is the interactions between hydrophobic elements and a set of hydrogen bonds between protein and adaptor/receptor fragments. Recognition appears to occur because of the quick interaction of electronegative amino acids near the LIR site with a cluster of positively charged residues. Then, the peptide is spatially recognized by the immersion of hydrophobic amino acids into HP1/HP2 pockets accompanied by the formation of hydrogen bonds. With even a slight mismatch in the position of amino acids, some of the peptide can dissociate from the ATG8 surface and is held in place by only one of the anchored hydrophobic residues and adjacent hydrogen bonds. Finally, we can suggest that the protein partner’s linker can recruit a particular isotype of ATG8. Probably, tight interaction with the binding site depends on the entire structure flexibility despite the almost similar LDS binding site. In such situation the LDS site, figuratively speaking, envelops the linker and that increases the binding interaction between cognate pairs of ATG8 and protein partner.

Funding

The work was supported by the budget project of the National Academy of Sciences of Ukraine (state registration number 0120U100937, 2020-24) and the project of National Research Foundation of Ukraine (state registration number 0120U104883, 2020-23) using the capacities of the Grid Cluster of the Institute of Food Biotechnology and Genomics of the National Academy of Sciences of Ukraine and the virtual organization CSLabGrid. Authors are grateful to Dr. Sergio Pantano from the Institut Pasteur de Montevideo for his kind cooperation and Mr. Volodymyr Rayevsky for the valuable assistance with graph visualizations of the manuscript.

Competing Interests

The authors declare no conflicts of interest.

Author contributions EB and AR contributed equally. Conceived funding and designed the experiments: AR and MS. Performed the experiments: SS, MS and AR under guidance of YB. Analyzed the data: EB, AR, MS, PK and SO. Wrote the manuscript: EB and AR. All authors contributed to the final version of the manuscript.

Data Availability

The datasets generated during and/or analyzed during the current study are not publicly available due to unstable power supply, military action, etc. but are available from the corresponding author on reasonable request.

Avin-Wittenberg T, Michaeli S, Honig A, Galili G (2012) ATI1, a newly identified atg8-interacting protein, binds two different Atg8 homologs. Plant Signal Behav 7(6):685-687. https://doi.org/10.4161/psb.20030
Birgisdottir ÅB, Lamark T, Johansen T (2013) The LIR motif - crucial for selective autophagy. J Cell Sci 126(Pt 15):3237-3247. https://doi.org/10.1242/jcs.126128
Bu F, Yang M, Guo X, Huang W, Chen L (2020) Multiple Functions of ATG8 Family Proteins in Plant Autophagy. Front Cell Dev Biol 8:466. https://doi.org/10.3389/fcell.2020.00466
Cao W, Li J, Yang K, Cao D (2021) An overview of autophagy: Mechanism, regulation and research progress. Bull Cancer, 108(3):304-322. https://doi.org/10.1016/j.bulcan.2020.11.004
Dompierre JP, Godin JD, Charrin BC, Cordelières FP, King SJ, Humbert S, Saudou F (2007) Histone Deacetylase 6 Inhibition Compensates for the Transport Deficit in Huntington's Disease by Increasing Tubulin Acetylation. The Journal of Neuroscience 27(13):3571-3583. https://doi.org/10.1523/jneurosci.0037-07.2007
Eargle J, Wright D, Luthey-Schulten Z (2006) Multiple Alignment of protein structures and sequences for VMD. Bioinformatics 22(4):504-506. https://doi.org/10.1093/bioinformatics/bti825
Essmann U, Perera L, Berkowitz M, Darden T, Lee H, Pedersen L (1995) A Smooth Particle Mesh Ewald Method. J. Chem. Phys 103:8577. https://doi.org/10.1063/1.470117
Galjart N. (2010) Plus-end-tracking proteins and their interactions at microtubule ends. Curr Biol 20(12):R528-537. https://doi.org/10.1016/j.cub.2010.05.022
Gatica D, Lahiri V, Klionsky DJ (2018) Cargo recognition and degradation by selective autophagy. Nat Cell Biol 20(3):233-242. https://doi.org/10.1038/s41556-018-0037-z
Hall TA. (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. NUCLEIC ACIDS SYMPOSIUM SERIES 41:95-98.
Honig A, Avin-Wittenberg T, Ufaz S, Galili G (2012) A new type of compartment, defined by plant-specific Atg8-interacting proteins, is induced upon exposure of Arabidopsis plants to carbon starvation. Plant Cell 24(1):288-303. https://doi.org/10.1105/tpc.111.093112
Huang J, MacKerell AD Jr (2013) CHARMM36 all-atom additive protein force field: validation based on comparison to NMR data. J Comput Chem 34(25):2135-2145. https://doi.org/10.1002/jcc.23354
Johansen T, Lamark T (2020) Selective Autophagy: ATG8 Family Proteins, LIR Motifs and Cargo Receptors. J Mol Biol 432(1):80-103. https://doi.org/10.1016/j.jmb.2019.07.016
Kabeya Y, Mizushima N, Yamamoto A, Oshitani-Okamoto S, Ohsumi Y, Yoshimori T (2004) LC3, GABARAP and GATE16 localize to autophagosomal membrane depending on form-II formation. Journal of Cell Science 117(13):2805-2812. https://doi.org/10.1242/jcs.01131
Kellner R, De la Concepcion JC, Maqbool A, Kamoun S, Dagdas YF (2017) ATG8 Expansion: A Driver of Selective Autophagy Diversification? Trends Plant Sci 22(3):204-214. https://doi.org/10.1016/j.tplants.2016.11.015
Ketelaar T, Voss C, Dimmock SA, Thumm M, Hussey PJ (2004) Arabidopsis homologues of the autophagy protein Atg8 are a novel family of microtubule binding proteins. FEBS Lett 567(2-3):302-306. https://doi.org/10.1016/j.febslet.2004.04.088
Khaminets A, Behl C, Dikic I (2016) Ubiquitin-Dependent And Independent Signals In Selective Autophagy. Trends Cell Biol 26(1):6-16. https://doi.org/10.1016/j.tcb.2015.08.010
Kirisako T, Ichimura Y, Okada H, Kabeya Y, Mizushima N, Yoshimori T, Ohsumi Y (2000) The reversible modification regulates the membrane-binding state of Apg8/Aut7 essential for autophagy and the cytoplasm to vacuole targeting pathway. J Cell Biol 151(2):263-276. https://doi.org/10.1083/jcb.151.2.263
Kirkin V (2020) History of the Selective Autophagy Research: How Did It Begin and Where Does It Stand Today? J Mol Biol 432(1):3-27. https://doi.org/10.1016/j.jmb.2019.05.010
Leboutet R, Largeau C, Müller L, Prigent M, Quinet G, Rodriguez MS, Legouis R (2023) LGG-1/GABARAP lipidation is not required for autophagy and development in Caenorhabditis elegans. eLife 12:e85748. https://doi.org/10.7554/eLife.85748
Liu Y, Hu C, Zhao C (2011) Efficient parallel implementation of Ewald summation in molecular dynamics simulations on multi-core platforms. Computer Physics Communications 182:1111-1119. https://doi.org/10.1016/j.cpc.2011.01.007
Marshall RS, Vierstra RD (2018) Autophagy: The Master of Bulk and Selective Recycling. Annu Rev Plant Biol 69:173-208. https://doi.org/10.1146/annurev-arplant-042817-040606
Martens S, Fracchiolla D (2020) Activation and targeting of ATG8 protein lipidation. Cell Discov 6:23. https://doi.org/10.1038/s41421-020-0155-1
Miklaszewska M, Zienkiewicz K, Klugier-Borowska E, Rygielski M, Feussner I, Zienkiewicz A (2023) CALEOSIN 1 interaction with AUTOPHAGY-RELATED PROTEIN 8 facilitates lipid droplet microautophagy in seedlings. Plant Physiol 193(4): 2361-2380. https://doi.org/10.1093/plphys/kiad471
Nguyen TN, Lazarou M (2022) A unifying model for the role of the ATG8 system in autophagy. J Cell Sci 135:(11) https://doi.org/10.1242/jcs.258997
Ohsumi Y (2001) Molecular dissection of autophagy: two ubiquitin-like systems. Nat Rev Mol Cell Biol 2(3):211-6 https://doi.org/10.1038/35056522
Páll S, Abraham MJ, Kutzner C, Hess B, Lindahl E (2015) Tackling Exascale Software Challenges in Molecular Dynamics Simulations with GROMACS. Lecture Notes in Computer Science 8759:3-27 https://doi.org/10.1007/978-3-319-15976-8_1.
Pandey P, Leary AY, Tumtas Y, Savage Z, Dagvadorj B, Duggan C, Bozkurt TO (2021) An oomycete effector subverts host vesicle trafficking to channel starvation-induced autophagy to the pathogen interface. Elife 10:e65285 https://doi.org/10.7554/eLife.65285
Rayevsky A, Bulgakov E, Sharifi M, Samofalova D, Ozheredov D, Karpov P, Blume Y (2023) In silico induced effect of N-ε-lysine acetylation on microtubule stability and subsequent interaction of microtubule-associated proteins. Cell Biol Int 47(9):1547-1557. https://doi.org/10.1002/cbin.12052
Rayevsky A, Ozheredov DS, Samofalova DA, Ozheredov SP, Karpov PA, Blume YB (2021) The Role of Posttranslational Acetylation in the Association of Autophagy Protein ATG8 with Microtubules in Plant Cells. Cytology and Genetics 55:510-518.
Rayevsky AV, Sharifi M, Samofalova DA, Karpov PA, Blume YB (2019) Structural and functional features of lysine acetylation of plant and animal tubulins. Cell Biol Int 43(9):1040-1048. https://doi.org/10.1002/cbin.10887
Samofalova DO, Karpov PA, Raevsky AV, Blume YB (2019) Protein phosphatases potentially associated with regulation of microtubules, their spatial structure reconstruction and analysis. Cell Biol Int 43(9):1081-1090. https://doi.org/10.1002/cbin.10810
Shpilka T, Weidberg H, Pietrokovski S, Elazar Z (2011) Atg8: an autophagy-related ubiquitin-like protein family. Genome Biol 12(7):226. doi:10.1186/gb-2011-12-7-226
Sun S, Feng L, Chung KP, Lee KM, Cheung HH, Luo M, Zhuang X (2022) Mechanistic insights into an atypical interaction between ATG8 and SH3P2 in Arabidopsis thaliana. Autophagy 18(6):1350-1366. https://doi.org/10.1080/15548627.2021.1976965
Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Mering CV (2019) STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47(D1):D607-D613. doi:10.1093/nar/gky1131
van Zundert GCP, Rodrigues J, Trellet M, Schmitz C, Kastritis PL, Karaca E, Bonvin A (2016) The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes. J Mol Biol 428(4):720-725. https://doi.org/10.1016/j.jmb.2015.09.014
Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Velankar S (2022) AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res 50(D1):D439-D444. https://doi.org/10.1093/nar/gkab1061
Varga VB, Keresztes F, Sigmond T, Vellai T, Kovács T (2022) The evolutionary and functional divergence of the Atg8 autophagy protein superfamily. Biol Futur 73(4):375-384. https://doi.org/10.1007/s42977-022-00123-6
von Muhlinen N, Akutsu M, Ravenhill BJ, Foeglein Á, Bloor S, Rutherford TJ, Randow F (2013) An essential role for the ATG8 ortholog LC3C in antibacterial autophagy. Autophagy 9(5):784-786. https://doi.org/10.4161/auto.23698
Wang P, Nolan TM, Yin Y, Bassham DC (2020) Identification of transcription factors that regulate ATG8 expression and autophagy in Arabidopsis. Autophagy 16(1):123-139. https://doi.org/10.1080/15548627.2019.1598753
Wesch N, Kirkin V, Rogov VV (2020) Atg8-Family Proteins-Structural Features and Molecular Interactions in Autophagy and Beyond Cells 9:(9). https://doi.org/10.3390/cells9092008
Wywial E, Singh SM (2010) Identification and structural characterization of FYVE domain-containing proteins of Arabidopsis thaliana. BMC Plant Biol 10:157. https://doi.org/10.1186/1471-2229-10-157
Xie R, Nguyen S, McKeehan WL, Liu L (2010) Acetylated microtubules are required for fusion of autophagosomes with lysosomes. BMC Cell Biol 11:89. https://doi.org/10.1186/1471-2121-11-89
Xie Y, Kang R, Sun X, Zhong M, Huang J, Klionsky DJ, Tang D (2015) Posttranslational modification of autophagy-related proteins in macroautophagy. Autophagy 11(1):28-45. https://doi.org/10.4161/15548627.2014.984267
Yemets A, Shadrina R, Blume R, Plokhovska S, Blume Y (2024) Autophagy formation, microtubule disorientation, and alteration of ATG8 and tubulin gene expression under simulated microgravity in Arabidopsis thaliana. NPJ Microgravity 10(1):31. https://doi.org/10.1038/s41526-024-00381-9
Zeng Y, Li B, Huang S, Li H, Cao W, Chen Y, Jiang L (2023) The plant unique ESCRT component FREE1 regulates autophagosome closure. Nat Commun 14(1):1768. https://doi.org/10.1038/s41467-023-37185-6
Zhuang X, Wang H, Lam SK, Gao C, Wang X, Cai Y, Jiang L (2013) A BAR-domain protein SH3P2, which binds to phosphatidylinositol 3-phosphate and ATG8, regulates autophagosome formation in Arabidopsis. Plant Cell 25(11):4596-4615. https://doi.org/10.1105/tpc.113.118307

Download PDF

Reviewers invited by journal
01 Aug, 2024
Editor assigned by journal
31 Jul, 2024
First submitted to journal
30 Jul, 2024

You are reading this latest preprint version

Structural flexibility and shape similarity contribute to exclusive functions of certain ATG8 isoforms in the autophagy process

Status:

Version 1

Abstract

Figures

Key message

Introduction

Materials and Methods

Molecular Dynamics

Docking

Results and discussion

Ubiquitin interaction motif docking site sensing

Autophagic adaptor/receptor proteins of choice

RMSD-based clusters

Clustering score

ATG8 pockets

Analysis of trajectories

Molecular dynamic residue interactions

Conclusions

Declarations

References

Status:

Version 1