Marchantia polymorpha, a New Model Plant for Autophagy Studies

Autophagy is a catabolic process for bulk and selective degradation of cytoplasmic components in the vacuole/lysosome. In Saccharomyces cerevisiae, ATG genes were identified as essential genes for autophagy, and most ATG genes are highly conserved among eukaryotes, including plants. Although reverse genetic analyses have revealed that autophagy is involved in responses to abiotic and biotic stresses in land plants, our knowledge of its molecular mechanism remains limited. This limitation is partly because of the multiplication of some ATG genes, including ATG8, in widely used model plants such as Arabidopsis thaliana, which adds complexity to functional studies. Furthermore, due to limited information on the composition and functions of the ATG genes in basal land plants and charophytes, it remains unclear whether multiplication of ATG genes is associated with neofunctionalization of these genes. To gain insight into the diversification of ATG genes during plant evolution, we compared the composition of ATG genes in plants with a special focus on a liverwort and two charophytes, which have not previously been analyzed. Our results showed that the liverwort Marchantia polymorpha and the charophytes Klebsormidium nitens and Chara braunii harbor fundamental sets of ATG genes with low redundancy compared with those of A. thaliana and the moss Physcomitrella patens, suggesting that multiplication of ATG genes occurred during land plant evolution. We also attempted to establish an experimental system for analyzing autophagy in M. polymorpha. We generated transgenic plants expressing fluorescently tagged MpATG8 to observe its dynamics in M. polymorpha and produced autophagy-defective mutants by genome editing using the CRISPR/Cas9 system. These tools allowed us to demonstrate that MpATG8 is transported into the vacuole in an MpATG2-, MpATG5-, and MpATG7-dependent manner, suggesting that fluorescently tagged MpATG8 can be used as an autophagosome marker in M. polymorpha. M. polymorpha can provide a powerful system for studying the mechanisms and evolution of autophagy in plants.


INTRODUCTION
Autophagy is a highly conserved catabolic process in eukaryotes for degrading and recycling cytoplasmic components. Among several types of autophagy reported thus far, the molecular mechanisms and physiological significance of macroautophagy have been the most intensively studied (Mizushima et al., 2011). During macroautophagy (hereafter simply referred to as "autophagy"), a cup-shaped membrane sac called the isolation membrane or phagophore elongates and sequesters cytoplasmic components, and its edge is closed to form a double-membranebounded structure called the autophagosome (Baba et al., 1994). The outer membrane of the autophagosome fuses with the vacuolar/lysosomal membrane, and a single-membraned structure called the autophagic body is released into the luminal space of the vacuole/lysosome to be degraded by lytic enzymes (Takeshige et al., 1992;Baba et al., 1994Baba et al., , 1995. In addition to bulk degradation, certain substrates, including organelles, are selectively recognized and degraded by autophagy . In Saccharomyces cerevisiae, autophagy-related processes are also involved in biosynthetic delivery; the newly synthesized precursor form of aminopeptidase I is transported into the vacuole by small double-membrane-bounded vesicles (Baba et al., 1997), which are formed through a mechanism similar to autophagy (Harding et al., 1996;Scott et al., 1996).
Most ATG genes are also conserved in plants, and deletion of many ATG genes in Arabidopsis thaliana results in defective autophagy, suggesting that ATG proteins in A. thaliana have the same functions as those in S. cerevisiae and mammals (Doelling et al., 2002;Hanaoka et al., 2002;Yoshimoto et al., 2004;Thompson et al., 2005;Xiong et al., 2005;Inoue et al., 2006;Phillips et al., 2008;Chung et al., 2010;Suttangkakul et al., 2011;Li et al., 2014;Young et al., 2019). Genetic analyses have revealed that autophagy is involved in responses to abiotic and biotic stresses (Doelling et al., 2002;Liu et al., 2005Liu et al., , 2009Xiong et al., 2007;Zhou et al., 2013;Chen et al., 2015). Whereas almost all atg mutants of A. thaliana are fertile under favorable growth conditions, autophagy is required for male gamete differentiation in Oryza sativa and Physcomitrella patens (Kurusu et al., 2014;Sanchez-Vera et al., 2017). Furthermore, selective degradation of various organelles, including peroxisomes, by autophagy has been reported (Kim et al., 2013;Shibata et al., 2013;Yoshimoto et al., 2014); the molecular mechanisms of this phenomenon, however, remain largely unknown (Yoshimoto and Ohsumi, 2018). The difficulty in studying autophagy in plants is partly due to the genetic redundancy of key ATG genes in model plants, including A. thaliana. For example, A. thaliana, O. sativa, and P. patens harbor nine, four, and six ATG8 homologs, respectively, which makes it difficult to unravel the functions of ATG8 proteins in these plants (Kellner et al., 2017). Although functional differentiation among mammalian ATG8 homologs has been reported (Weidberg et al., 2010), it remains completely unknown whether each ATG8 homolog in plants acts at a different step in autophagy. The significance of the duplication of ATG genes during plant evolution also remains obscure, since information from basal land plants and algal species remains sparse. To obtain more insights into the diversification and evolution of autophagy in the plant lineage, information from charophytes, the closest living relatives of land plants, and from additional bryophytes would be needed.
In this study, we identified homologs of ATG genes in the basal land plants Marchantia polymorpha and P. patens and the charophytes Klebsormidium nitens and Chara braunii. A comparison among these species, as well as Chlamydomonas reinhardtii and A. thaliana, indicated that M. polymorpha shares a common set of ATG genes with low redundancy compared with those of other land plants; this finding suggested that M. polymorpha would be a good system to investigate the molecular mechanisms and physiological significance of autophagy in land plants. Many molecular genetic techniques and cell biological tools have been established for M. polymorpha (Era et al., 2009;Ishizaki et al., 2016;Kanazawa et al., 2016;Minamino et al., 2018), which would also be a good reason to use this plant for autophagy studies. In this study, as the first step toward the study of autophagy in M. polymorpha, we generated transgenic plants expressing fluorescently tagged MpATG8 proteins to monitor autophagosomes and produced the autophagy-defective mutants Mpatg2, Mpatg5, and Mpatg7 to analyze the effect of defective autophagy on thallus development and their responses to nutrient starvation. Our results indicated that the number of ATG genes gradually increased during plant evolution. We also succeeded in monitoring the dynamics of MpATG8, which was transported into the vacuole in an MpATG2-, MpATG5-, and MpATG7-dependent manner. These tools would be useful for future studies to understand the basic mechanisms and physiological significance of autophagy in plants.

Identification of Orthologs of Core Autophagy Machinery Components in Plants
Extensive studies using S. cerevisiae have revealed that ATG1-10, 12-14, 16-18, 29, and 31, and VPS15 and VPS34 are required for autophagosome formation (Mizushima et al., 2011). Although some ATG genes are highly duplicated in angiosperms, the precise significance of the expansion of ATG genes remains obscure. To gain insight into this phenomenon, we searched the genome sequences of charophytes (K. nitens and C. braunii) and bryophytes (M. polymorpha and P. patens) for homologs of ATG genes, a number of which were then compared with those of A. thaliana and C. reinhardtii. As a query, we used the sequences of the ATG genes of A. thaliana (Shemi et al., 2015;Liu et al., 2018). We also examined ATG11 and ATG101. ATG11 is not required for starvation-induced bulk autophagy but is essential for selective autophagy in S. cerevisiae (Kim et al., 2001); Atg11 interacts with Atg1 and cargo receptors, which is crucial for selective autophagy (Yorimitsu and Klionsky, 2005;Farre et al., 2008;Okamoto et al., 2009). Atg101 forms a complex with ULK1 (homologous to Atg1) in mammalian cells, although S. cerevisiae does not harbor a homolog of ATG101 (Hosokawa et al., 2014;Mercer et al., 2014). In A. thaliana, ATG11 and ATG101 are thought to form a complex with ATG1, and the atg11 mutant exhibits similar phenotypes to those of other atg mutants (Li et al., 2014), suggesting that ATG11 and ATG101 are involved in general autophagy in plants. Therefore, ATG11 and ATG101 were also included among the "core autophagy machinery genes" in this study.
As shown in Table 1, almost all core autophagy machinery genes were highly conserved among the plants we investigated, although ATG16 homologs in C. reinhardtii (Shemi et al., 2015) and ATG2 and ATG10 homologs in C. braunii were not detected. Homologs of ATG17, ATG29, and ATG31 were not detected in this study, consistent with previous studies (Kawamata et al., 2005;Li et al., 2014). While a considerable number of ATG genes are present in A. thaliana and P. patens, C. reinhardtii, charophytes, and M. polymorpha possess only one gene for each core autophagy machinery component, with a few exceptions; multiple genes for ATG8 and ATG18 exist in the genomes of these plants with lower redundancy than A. thaliana and P. patens (Table 1). Thus, core autophagy machinery genes have seemingly expanded gradually during plant evolution.
Recently, Pang et al. (2019) reported that ATG10 has been lost in quite a few lineages of eukaryotes, including Pichiaceae and the SAR supergroup, which comprises stramenopiles, alveolates, and Rhizaria. As summarized in Table 1, we detected an ATG10 homolog in K. nitens but not in C. braunii. We also looked for ATG10 homologs in other charophytes whose genome and/or transcriptome information was available (Spirogyra pratensis, Nitella mirabilis, Coleochaete orbicularis, and Mesostigma viride) and found that these algae, except for N. mirabilis, possess ATG10 homologs. This distribution of ATG10 suggested that secondary loss of ATG10 occurred in Charophyceae ( Table 2). The ATG10 protein is an E2-like enzyme required for covalent linkage between the C-terminal glycine residues of ATG12 and ATG5 (Mizushima et al., 1998;Shintani et al., 1999;Suzuki et al., 2005;Phillips et al., 2008;Chung et al., 2010). In Toxoplasma gondii and Komagataella phaffii, which do not possess ATG10 homologs, ATG12 forms a noncovalent complex with ATG5, which does not require the C-terminal glycine of ATG12 (Pang et al., 2019). Intriguingly, the ATG12 of C. braunii does not harbor glycine at the C-terminus, although the glycine residue is conserved at the C-terminus of N. mirabilis ATG12 ( Table 2). These features could reflect a similar mechanism of complex formation between ATG12 and ATG5 in C. braunii; ATG12 might form a noncovalent complex with ATG5.
A. thaliana and Nicotiana tabacum are reported to possess two types of ATG18: conventional ATG18, which is similar to yeast and mammalian ATG18, and plant-unique ATG18, which harbors the BCAS3 domain at the C-terminus (Xiong et al., 2005;Zhou et al., 2015). All of the land plant species we analyzed in this study harbor both types of ATG18, whereas C. reinhardtii does not possess plant-unique ATG18 ( Table 3). We then investigated whether plant-unique ATG18 is found in other chlorophytes. Although we did not detect ATG18 with the BCAS3 domain in Dunaliella salina, Volvox carteri, Micromonas pusilla, Ostreococcus lucimarinus, and Ostreococcus tauri, Coccomyxa subellipsoidea harbored this type of ATG18. Canonical ATG18 TABLE 1 | Composition of core autophagy machinery genes in Viridiplantae.

Species
Class ATG10 C-terminal Gly in ATG12 Zygnematophyceae 1 + "ND," not detected in our analysis; +, conserved; −, not conserved. homologs were found in all of these species (Table 3). Therefore, ATG18 containing the BCAS3 domain seems to have been acquired before the emergence of Streptophyta, which was also supported by the result of phylogenetic analysis (Figure 1).

Dynamics of MpATG8 in M. polymorpha
Because of the low genetic redundancy of ATG components ( Table 1) and molecular genetic tools available (Ishizaki et al., 2016), M. polymorpha is expected to be a good model for understanding the fundamental molecular mechanisms of autophagy in land plants. As the first step of the autophagy study using M. polymorpha, we generated transgenic plants expressing fluorescently tagged MpATG8 proteins under the regulation of their own promoters. ATG8 localizes to the isolation membrane during autophagosome formation and stays on the inner membranes of autophagosomes until their degradation in the vacuole/lysosome in various organisms, including A. thaliana (Kirisako et al., 1999;Kabeya et al., 2000;Yoshimoto et al., 2004;Izumi et al., 2015). Therefore, fluorescently tagged MpATG8 is expected to be a good tool to visualize autophagosome dynamics in M. polymorpha. As shown in Figure 2A, monomeric Citrine (mCitrine)-tagged MpATG8a and MpATG8b were localized to punctate structures distributed throughout the cytosol. Fluorescence was also detected in the nucleus (asterisks in Figure 2A), which might reflect a function of the nucleus as a reservoir of MpATG8 as reported in mammalian cells . As autophagic bodies bearing ATG8 transported to the vacuolar lumen are generally degraded immediately by vacuolar enzymes (Kirisako et al., 1999;Huang et al., 2000), we treated transgenic M. polymorpha with concanamycin A (concA), which inhibits acidification of the vacuole and inactivates vacuolar lytic enzymes, to visualize autophagic bodies in the vacuole (Yoshimoto et al., 2004). After treatment with concA, both mCitrine-MpATG8a and mCitrine-MpATG8b were visible as punctate structures in the vacuole ( Figure 2C). Vacuolar localization was not observed in mock-treated cells, suggesting that these punctate structures are autophagic bodies. We then investigated whether MpATG8a and MpATG8b are localized to the same structure. We expressed monomeric RFP (mRFP)tagged MpATG8a and mCitrine-MpATG8b in the same plant and observed strong colocalization at the same punctate structures in the cytosol; colocalization was also observed in the vacuole in concA-treated cells, suggesting that MpATG8a and MpATG8b are localized to the same autophagosomes/autophagic bodies (Figures 2B,D). These observations indicated that MpATG8a and MpATG8b behave in a similar manner to ATG8 in other organisms, and these molecules with fluorescent tags would be useful as autophagosome markers.

Generation of atg Mutants of M. polymorpha by Genome Editing
To investigate the physiological significance of autophagy in M. polymorpha, we generated Mpatg5, Mpatg7, and Mpatg2 mutants (hereafter referred to as Mpatg5-1 ge , Mpatg7-1 ge , and Mpatg2-1 ge , respectively) by genome editing using the clustered regularly interspaced short palindromic repeats (CRISPRs)-associated endonuclease Cas9 (CRISPR/Cas9) system (Sugano et al., 2014(Sugano et al., , 2018. ATG5, ATG7, and ATG2 are required for autophagosome formation, and deletion of these genes causes defects in autophagy in various organisms, including S. cerevisiae, mammals, and A. thaliana (Tsukada and Ohsumi, 1993;Mizushima et al., 2001;Doelling et al., 2002;Kuma et al., 2004;Komatsu et al., 2005;Thompson et al., 2005;Inoue et al., 2006;Velikkakath et al., 2012). The mutations detected in Mpatg5-1 ge , Mpatg7-1 ge , and Mpatg2-1 ge result in frame shifts, and functional full-length proteins cannot be produced in these mutants (Figure 3). To investigate whether autophagy occurs in these mutants, we observed the dynamics of mCitrine-MpATG8a in these mutants. In wild-type (WT) plants, mCitrine-MpATG8a was localized to the punctate structures in the cytoplasm and vacuolar lumen (Figures 2A,C, 4A). In contrast, vacuolar localization of mCitrine-MpATG8a was not observed in any of the Mpatg mutants, while mCitrine-MpATG8a was observed as puncta in the cytoplasm (Figures 4A,B). We then performed a cleavage assay of mCitrine-MpATG8a in the Mpatg mutants. As ATG8 is rapidly degraded in the vacuole, whereas mCitrine/YFP is more resistant to lytic enzymes, translocation of mCitrine-MpATG8 to the vacuole can be monitored by examining the accumulation of free mCitrine by immunoblotting (Shintani and Klionsky, 2004;Chung et al., 2010). In WT, two bands were observed via immunoblotting using an anti-green fluorescent protein (GFP) antibody, at approximately 43 and 27 kDa ( Figure 4C). The 43 kDa band represented full-length mCitrine-MpATG8a, and the 27 kDa product represented free mCitrine. In contrast, only the 43 kDa product was detected in Mpatg mutants, confirming that mCitrine-MpATG8a was not transported into the vacuole in these mutants ( Figure 4C). This result indicated that Mpatg5-1 ge , Mpatg7-1 ge , and Mpatg2-1 ge are defective in autophagy and that MpATG8 is transported into the vacuole in an MpATG5-, MpATG7-, and MpATG2-dependent manner in M. polymorpha. We then observed the macroscopic phenotypes of Mpatg mutants. Under normal laboratory conditions, the proximal regions of the thalli of all Mpatg mutants exhibited a yellowish chlorotic phenotype, which is similar to the early senescence phenotype of atg mutants of A. thaliana (Doelling et al., 2002;Hanaoka et al., 2002; Figure 5A). Thus, autophagy appears to also be required for preventing early senescence in M. polymorpha. We then investigated whether Mpatg mutants are hypersensitive to nutrient starvation, as reported for atg mutants of A. thaliana (Doelling et al., 2002;Hanaoka et al., 2002). We cultured thalli of M. polymorpha in liquid 1/2× Gamborg's B5 medium under continuous light (control) or in 10 mM 2-(N-morpholino)ethanesulfonic acid (MES) under the dark condition, which induces both bulk autophagy and piecemeal autophagy of chloroplasts in A. thaliana (Izumi et al., 2010), and measured chlorophyll contents. Consistent with the early senescence phenotype shown in Figure 5A, Mpatg mutants exhibited lower chlorophyll contents after incubation in 1/2× Gamborg's B5 medium for 3 days under continuous light compared with WT, confirming that autophagy plays critical roles in preventing early senescence in M. polymorpha (Figures 5B,C). We also found that Mpatg mutants cultured in 10 mM MES for 3 days under dark condition exhibited significantly lower chlorophyll contents than that in WT cultured in 10 mM MES ( Figure 5D). These data indicated that autophagy is required for normal response to nutrient starvation in M. polymorpha.

DISCUSSION
ATG genes were first identified in S. cerevisiae, and most ATG genes have now been shown to be conserved in various lineages of eukaryotes. In this study, we showed that charophyte species and M. polymorpha also possess a set of genes for the core autophagy machinery with lower redundancy than those of P. patens and A. thaliana. This result suggests that the core autophagy machinery has been expanded during land plant evolution. Some of the ATG genes, such as ATG8 and ATG18, are reported to have multiplied in various organisms, including land plants. Intriguingly, M. polymorpha also harbors multiple ATG8 and ATG18 genes, and the plant-unique ATG18 group comprising the BCAS3 domain is conserved in M. polymorpha, some charophytes, and the chlorophyte C. subellipsoidea. The distribution of the plant-unique ATG18 suggests that this type of ATG18 was acquired before the emergence of Streptophyta, and secondary losses of this gene occurred independently during plant evolution, although its molecular function remains unknown. Moreover, it remains unclear whether the members of ATG8 and ATG18 are functionally differentiated in plants.
In A. thaliana, knockdown of ATG18a, one of eight ATG18 homologs, results in similar but weaker phenotypes to those of mutants of other ATG genes, suggesting that ATG18a plays a major role in autophagy, although the other seven ATG18 homologs could have a partly redundant function (Xiong et al., 2005;Kang et al., 2018). In addition to its localization to the PAS, Atg18 of S. cerevisiae is also localized to the vacuolar membrane via phosphatidylinositol 3,5-bisphosphate binding and acts in retrograde transport from the vacuole, which is independent of autophagy (Dove et al., 2004;Efe et al., 2007). ATG18 homologs in A. thaliana might also have autophagyindependent functions. In mammals, two groups of ATG8related proteins, the LC3 and GABARAP/GATE-16 subfamilies, are both localized to the autophagosomes but play distinct roles in autophagosome formation (Kabeya et al., 2004;Weidberg et al., 2010). The LC3 group is involved in the elongation of the phagophore membrane, whereas GABARAP/GATE-16 plays an essential role in a later stage of autophagosome maturation (Weidberg et al., 2010). Caenorhabditis elegans possesses two ATG8 homologs: LGG-1 and LGG-2.
LGG-1 is involved in the formation of autophagosomes. In contrast, LGG-2 mediates the maturation of autophagosomes and facilitates their tethering with lysosomes through interaction with VPS39 (Manil-Segalen et al., 2014). In plants, distinct binding affinities of potato ATG8 members with PexRD54, an effector protein in the pathogenic fungus Phytophthora infestans, have been reported (Dagdas et al., 2016). Thus, the plant ATG8 and ATG18 groups could comprise functionally differentiated members; this possibility should be verified in future studies.
Although most of the core machinery of autophagy is conserved in plants, ATG10 homologs were not detected in two charophyte species. Loss of ATG10 has been observed in various lineages of eukaryotes, some of which utilize a noncovalent complex of ATG5 and ATG12 without the C-terminal glycine residue (Pang et al., 2019). Given that ATG10 mediates covalent linkage between ATG12 and ATG5 at the C-terminal glycine residue of ATG12, secondary loss of ATG10 could be associated with conversion from covalent to noncovalent ATG12-ATG5 complexes during evolution, which could be followed by deletion of the glycine residue of the C-terminus of ATG12. Consistent with this notion, we did not detect ATG10 homologs in two charophyte species, one of which (C. braunii) possesses ATG12 without C-terminal glycine (Tables 1, 2). Further characterization of ATG12 and ATG5 in C. braunii would be needed to verify the possible convergence of the ubiquitin-like conjugation system involving ATG12 and ATG5 across a wide range of eukaryotic lineages.
M. polymorpha has been recognized as a new model for analyzing the developmental processes of land plants (Bowman, 2016;Bowman et al., 2017;Ishizaki, 2017). This plant would also be useful for analyses of molecular mechanisms of autophagy due to its low genetic redundancy and widely available molecular genetic tools (Ishizaki et al., 2016;Bowman et al., 2017). In the interest of understanding the molecular mechanisms of autophagy in plants, it would be beneficial to reveal the molecular functions of ATG8 members, given that nonplant ATG8 acts at various steps of autophagy, such as expansion and closure of the isolation membrane, cargo recognition, and transport of autophagosomes to the vacuole/lysosome (Abeliovich et al., 2000;Nakatogawa et al., 2007;Fujita et al., 2008a;Kimura et al., 2008;Xie et al., 2008;Noda et al., 2010;Manil-Segalen et al., 2014;Nguyen et al., 2016;Tsuboyama et al., 2016). Although several ATG8-interacting proteins, such as ATI1/2, ATI3, DSK2, NBR1, ORM1/2, PUX, RPN10, and TSPO, which are involved in autophagy induced by certain stresses, have been identified (Svenning et al., 2011;Vanhee et al., 2011;Honig et al., 2012;Zhou et al., 2013Zhou et al., , 2014Zhou et al., , 2018Hachez et al., 2014;Michaeli et al., 2014;Marshall et al., 2015Marshall et al., , 2019Hafren et al., 2017;Nolan et al., 2017;Yang et al., 2019), it remains unclear how various substrates are selectively targeted by autophagy in plants (Yoshimoto and Ohsumi, 2018). Given that M. polymorpha possesses fewer copies of ATG8 than other model plants, this plant would be useful for revealing the molecular functions of ATG8 members and functional diversification of ATG8 in land plants.
In this study, we succeeded in visualizing autophagosomes using fluorescently tagged ATG8 proteins in M. polymorpha. Both MpATG8a and MpATG8b were localized to punctate structures in the cytosol and vacuole. Deletion of MpATG2, MpATG5, or MpATG7 resulted in defective transport of MpATG8a into the vacuole, indicating that vacuolar transport of ATG8 is autophagy-dependent in M. polymorpha, as reported in other organisms. Punctate localization of MpATG8a was also detected even in the Mpatg mutants, and similar localization is also observed in Arabidopsis atg mutants (Yoshimoto et al., 2004;Kang et al., 2018). ATG8/LC3 is reported to be incorporated into protein aggregates in an autophagy-independent manner in mammalian cells (Kuma et al., 2007;Tanida et al., 2008). Therefore, it is highly likely that MpATG8a also aggregates in the cytosol independently of autophagic activities in the Mpatg mutants. It would be also possible that some population of fluorescently tagged MpATG8a/b-positive puncta observed in WT plants represents unfunctional protein aggregates, which should be verified in future studies.
The autophagy-defective mutants of M. polymorpha exhibited an early senescence-like phenotype and hypersensitivity to nutrient starvation (Figure 5), which resembles the phenotypes observed in atg mutants of other plants (Doelling et al., 2002;Hanaoka et al., 2002;Mukae et al., 2015;Wada et al., 2015). In A. thaliana, salicylic acid (SA) signaling is involved in early senescence in atg mutants (Yoshimoto et al., 2009). Although the relevance of SA signaling to the senescence of Mpatg mutants remains unknown, autophagy might play a common role in preventing senescence among land plants.
A forward genetic approach using M. polymorpha would also be effective to reveal the molecular mechanisms of autophagy in land plants, as in other systems (Tsukada and Ohsumi, 1993;Thumm et al., 1994;Harding, 1995;Tian et al., 2010;Morita et al., 2018). Its haploid-dominant life cycle and low genetic redundancy could make this plant even more amenable to forward genetic analyses than other model plants, including A. thaliana (Ishizaki et al., 2016). Screening of mutants defective in autophagy and functional analyses of obtained factors would facilitate understanding of the molecular mechanisms of autophagy in land plants. In conclusion, M. polymorpha is a suitable system for analyzing autophagy in land plants. Further studies in this plant will contribute to revealing the molecular mechanisms of autophagy in plants, which would also be useful to gain insights into how the autophagy machinery has been functionally diversified and how autophagy has been recruited to support plant physiology.

Identification of Orthologs for Core Autophagy Machinery
Amino acid sequences of core autophagy machinery in K. nitens, P. patens, and M. polymorpha were obtained in MarpolBase 1 using ATG genes of A. thaliana (Li et al., 2014;Shemi et al., 2015;Liu et al., 2018) as queries. ATG genes in C. braunii were searched in the C. braunii portal site 2 . For core autophagy machinery orthologs in C. reinhardtii except for ATG11, we referred to the previous study (Jiang et al., 2012;Shemi et al., 2015). The ATG11 homolog in C. reinhardtii was searched in MarpolBase. For ATG17, ATG29, and ATG31, whose homologs have not been identified in A. thaliana thus far, ATG genes in S. cerevisiae were used as queries. ATG10 and ATG12 homologs in S. pratensis, N. mirabilis, C. orbicularis, and M. viride were searched in the transcriptome database in MarpolBase. ATG18 homologs in D. salina, V. carteri, C. subellipsoidea C-169, M. pusilla CCMP1545, and O. lucimarinus were searched in Phytozome v12.1.6 3 . ATG18 homologs in O. tauri were searched in MarpolBase. A domain search was performed using SMART 4 (Letunic et al., 2014;Letunic and Bork, 2017). The accession numbers and amino acid sequences analyzed in this study are included in the Supplementary Material. We followed the nomenclature proposed in Bowman et al. (2016) for nomenclature of genes, proteins, and mutants of M. polymorpha.

Phylogenetic Analysis of ATG18
Amino acid sequences of ATG18 in various plant species were aligned with ClustalX 2.1 (Larkin et al., 2007), and alignment gaps were removed using Gblocks 5 . Phylogenetic analysis was performed using PhyML 3.0 6 (Guindon et al., 2010) under the LG+G+I+F model, which was selected by Smart Model Selection in PhyML (Lefort et al., 2017). Bootstrap analysis was performed by resampling 1,000 sets. The sequences used in the phylogenetic analysis and the alignment from which gaps were removed were included in the Supplementary Material.

Vector Construction
Genomic sequences of MpATG8a (Mapoly0001s0494.1) and MpATG8b (Mapoly0027s0034.1) were amplified by PCR from genomic DNA prepared from gemmae of M. polymorpha accession Takaragaike-1 (Tak-1, male) (Ishizaki et al., 2008), and the amplified products were subcloned into pENTR/D-TOPO (Invitrogen) according to the manufacturer's instructions. To construct mCitrine-and mRFP-MpATG8, genomic sequences comprising the protein-coding regions and 3 -flanking sequences (2 kb) were amplified with the SmaI site followed by a flexible linker sequence (Gly-Gly-Ser-Gly) attached at the 5 -end and subcloned into the pENTR vector. Then, cDNA for mRFP or mCitrine containing the SmaI site at the 5end was inserted into the SmaI site of pENTR vectors containing the MpATG8 genes using the In-Fusion HD Cloning System (Clontech) according to the manufacturer's instructions. The 5 kb 5 -sequence [promoter + 5 -untranslated region (UTR)] of each MpATG8 was then amplified and inserted into the SmaI site of the mRFP/mCitrine-MpATG8 vectors. The resultant chimeric genes were then introduced into pMpGWB301 (mCitrine-tagged MpATG8a and MpATG8b) or pMpGWB101 (mRFP-MpATG8a) (Ishizaki et al., 2015) using the Gateway LR Clonase TM II Enzyme Mix (Invitrogen) according to the manufacturer's instructions. To construct CRISPR/Cas9 vectors, two complementary oligonucleotides in the sequences of MpATG2, MpATG5, and MpATG7 were synthesized and annealed, and the resulting double-stranded fragments were subcloned at the BsaI site of the pMpGE_En03 vector (Sugano et al., 2018). The resultant gRNA cassette flanked by the attL1 and attL2 sequences in pMpGE_En03 were then introduced into the pMpGE010 vector (Sugano et al., 2018) using the Gateway LR Clonase II Enzyme Mix. The list of primer sequences used in this study is included in the Supplementary Material.

Plant Material and Transformation
The M. polymorpha accession Tak-1 was grown asexually and maintained on 1/2× Gamborg's B5 medium containing 1.4% agar at 22 • C under continuous white light. Transformation was performed as previously described (Kubota et al., 2013). Transformants were selected on plates containing 10 mg/l hygromycin B and 250 mg/l cefotaxime for the pMpGWB101 and pMpGE010 vectors and 0.5 µM chlorsulfuron and 250 mg/l cefotaxime for the pMpGWB301 vector.

Confocal Laser Scanning Microscopy
Five-day-old thalli grown on 1/2× Gamborg's B5 medium containing 1.4% agar at 22 • C under continuous white light were used for observation. Dorsal thallus tissues were observed using an LSM 780 confocal microscope (Carl Zeiss) as previously described (Kanazawa et al., 2016). For concA treatment, 4day-old thalli were incubated in liquid 1/2× Gamborg's B5 medium plus 1 µM concA (Santa Cruz Biotechnology, sc-202111) for 14 h at 22 • C under continuous white light. concA was dissolved in dimethyl sulfoxide (DMSO) at 1 mM as a stock solution. For the mock treatment, samples were treated with DMSO at a concentration equal to that used for the inhibitortreated samples.

Identification of Mutation Points
For genotyping of mutants generated by CRISPR/Cas9, total RNA was extracted from 5-day-old thalli of Mpatg5-1 ge , Mpatg7-1 ge , and Mpatg2-1 ge using the RNeasy Plant Mini Kit (Qiagen) and used as a template for reverse transcription using SuperScript III Reverse Transcriptase (Invitrogen) and the oligo (dT) (18-mer) primer according to the manufacturer's instructions. Mutations in the obtained cDNA fragments were analyzed by direct sequencing.

Immunoblot Analysis
Five-day-old thalli were used for the immunoblot analysis. One hundred milligrams of plants were homogenized in 200 µl of grinding buffer [50 mM HEPES-KOH, pH 7.5, 340 mM sorbitol, 5 mM MgCl 2 , and 1× Complete TM Protease Inhibitor Cocktail (Roche)] for each genotype and centrifuged at 1,000 × g for 10 min. The supernatants were centrifuged at 3,000 × g for 10 min, and the resulting supernatants were used for immunoblotting. The polyclonal anti-GFP antibody (Kanazawa et al., 2016) was purified by affinity column chromatography using the GST-mCitrine protein bound to the HiTrap TM NHS-activated HP Column (GE Healthcare) and used at 500× dilution. The peroxidaseconjugated donkey anti-rabbit immunoglobulin antibody (GE Healthcare) was used as the secondary antibody. Signals were detected using Immobilon Western Chemiluminescent HRP Substrate (Merck).

Measurement of Chlorophyll Content
Five-day-old thalli were incubated in 1 ml of liquid 1/2× Gamborg's B5 medium for 3 days under continuous light or in 10 mM MES (pH 5.5) for 3 days under dark condition, and chlorophyll was extracted by soaking in 500 µl N,N-dimethylformamide overnight. Calculation of chlorophyll concentrations was done according to Porra et al. (1989). Three thalli were treated as one sample for correct measurement of fresh weights.

DATA AVAILABILITY
The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.

AUTHOR CONTRIBUTIONS
TN and TU designed the research and wrote the manuscript. TN performed a major part of the experiment and analyzed the data. TK and NM prepared the anti-GFP antibody. TK assisted with vector construction. HT and TU supervised the study.

FUNDING
This study was financially supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan [Grant Nos. 24114003, 17K19412, and 18H02470 (to TU) and 17H07333 and 18K14738 (to TK)], a Grant-in-Aid for the Japan Society for the Promotion of Science (JSPS) fellows (to TN, Grant No. 19J13751), the BIO-NEXT project from the Okazaki Institute for Integrative Bioscience (to HT), the Mitsubishi Foundation, and the Yamada Science Foundation.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2019.00935/ full#supplementary-material DATA SET S1 | This data set contains all amino acid sequences analyzed in this study, the alignment generated for phylogenetic analysis, and lists of primers and accession numbers.